Bioreactor for cell culture

Abstract

The invention relates to a bioreactor for cell culture, comprising: a multi-layer microfluidic device comprising a header layer and a gas permeable flow layer beneath the header layer, wherein: (a) the flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells; and (b) the header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises: (i) a first inlet port; (ii) a first outlet port; (iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and (iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port; wherein the one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein; and a cell capture composition physically adhered to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

Description

BIOREACTOR FOR CELL CULTURETechnical Field

[0001] The present invention relates to a bioreactor for cell culture and a method of preparing such a bioreactor for cell culture. The bioreactor comprises a multi-layer microfluidic device and a cell capture composition physically adhered to a grooved hydrophobic surface in a flow channel of the device. The method further relates to methods of culturing cells, transfecting / transducing cells, or selecting cells, wherein for each method the target cells are immobilised on a cell capture composition physically adhered to a grooved hydrophobic surface in a flow channel of a microfluidic device. The method further relates to a multi-layer microfluidic device which includes a flow layer comprising polymethylpentene, and a method of fabricating such a multi-layer microfluidic device.Background of Invention[2] Cell and gene therapy involves modification of cells and can offer a cure for some cancers and genetic diseases in a single treatment. As cell and gene therapies progress from clinical proof-of-concept through to regulatory approval and product licensing, a major hurdle will be the scale-out of patient-specific treatments. It is estimated that the market for cell and gene therapy in the US and EU will grow beyond 200,000 patient-specific treatments per year. While targeted in vivo gene therapy remains the ultimate goal, ex vivo gene therapy is already being applied clinically for treatment of some diseases and disorders. In particular, for a range of hematologic or immunologic disorders, ex vivo gene therapy offers great promise. By way of example, current ex vivo methods for modifying human cells with therapeutic genes may include a) purification of blood stem or immune cells from human blood b) introduction of therapeutic DNA or RNA into target cells using viral or non-viral delivery systems c) mitotic activation of transduced cells by tissue culture for several days to facilitate viral entry and stable genomic integration. Non-viral systems include gene editing platforms such as mega-nuclease, ZFN, TALEN, and CRISPR-Cas nucleases. Viral delivery systems include viral vectors such as adeno-associated viral (AAV), adenoviral, lentiviral (LV), and retroviral vectors. Adeno-associated viral vectors and lentiviral vectors are two of the more commonly used vectors for clinical trials because of theirrelative safety, minimal toxicity to cells and high rates of gene transfer. Traditional processes involve significant investment in large scale, centralised manufacturing facilities that utilise manual, flask-based culture processes. Not only are these processes complex and labour-intensive, they are prone to operator error and carry a contamination risk due to use of open flasks. Further, these processes are not easily scaled for clinical therapies.[3] Current ex vivo methods of culturing cells can be inefficient due to the challenge of scaling out the manufacturing of patient-specific cell products for use in gene therapy or autologous therapies, or the limitations on the number of cells that can be grown due to unfavourable oxygen gradients, particularly for cell types which have a high oxygen demand. These scale-limited production processes are cumbersome and cost-prohibitive. In addition, known tissue culture apparatus can be inefficient through the loss of non-adherent cells in bulk media flow through fluidic shear stress, particularly for cell types which have a low endurance to shear stress such as human embryonic cells. This can limit the density at which cells can be grown through the limitation on maximum perfusion rates.[4] Microfluidic bioreactors provide a promising approach to cell selection, culturing and / or modification, as an alternative to existing large-footprint apparatus that is often flask or bag based and requires skilled staff and complicated multi-step procedures, resulting in prohibitively expensive treatment costs and limiting accessibility and take up of life changing and disease-curing therapies. In principle, microfluidic bioreactors allow precise and typically automated control of critical parameters such as gas and fluid flows, temperature, humidity, timing and sample handling, in a closed system thus eliminating risk of contamination and human error which accompanies flask-based cell culture systems.[5] However, it remains desirable to develop microfluidic bioreactor apparatus and methods to effectively capture and retain target cells while avoiding or minimising non-specific binding of cells in the flow channels, to distribute target cells efficiently through the microfluidic device to maximise use of its working area, to effectively culture or improve the transfection / transduction rates of target cells captured in the flow channels, and / or to facilitate manufacturability and structural integrity of the microfluidicdevices, including microfluidic devices intended for single-use and / or patient-specific applications.[6] There is therefore an ongoing need for microfluidic bioreactors, and methods of cell culture, modification or selection using such bioreactors, which at least partially address one or more of the above-mentioned short-comings, or provide a useful alternative.[7] The discussion of the background to the invention included herein including reference to documents, acts, materials, devices, articles and the like is included to explain the context of the present invention. This is not to be taken as an admission or a suggestion that any of the material referred to was published, known or part of the common general knowledge as at the priority date of any of the claims.Summary of Invention[8] According to a first aspect, the invention provides a bioreactor for cell culture, comprising: a multi-layer microfluidic device comprising a header layer and a gas permeable flow layer beneath the header layer, wherein: (a) the flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells; and (b) the header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises: (i) a first inlet port; (ii) a first outlet port; (iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and (iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port; wherein the one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein; and a cell capture composition physically adhered to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.[9] In some embodiments, the hydrophobe is a hydrophobic polymeric segment, optionally a hydrophobic polyalkylene glycol segment, for example a polypropylene glycol segment.

[0010] In some embodiments, the one or more hydrophilic polymeric segments are polyethylene glycol segments.

[0011] In some embodiments, the amphiphilic polymer is a triblock copolymer comprising a hydrophobic polypropylene glycol segment located between two hydrophilic polyethylene glycol segments.

[0012] In some embodiments, the amphiphilic polymer has a molecular weight in the range of 10,000 to 20,000 g / mol.

[0013] In some embodiments, the antigen binding moiety is conjugated to at least one of, and optionally each of, the one or more hydrophilic polymeric segments such that the cell capture composition comprises a polymer-antigen binding moiety conjugate. The antigen binding moiety may be conjugated to a chain terminus of the at least one hydrophilic polymeric segment. Optionally, the antigen binding moiety is conjugated to the chain terminus by an amide linking group, wherein the amide linkage is produced by a conjugation reaction between a terminal amine-reactive functional group on the at least one hydrophilic polymeric segment with an amine functional group of the antigen binding moiety, preferably wherein the amine-reactive functional group is an activated ester functional group. The antigen binding moiety may be conjugated to the chain terminus by a click reaction between a terminal click-reactive functional group on the at least one hydrophilic polymeric segment and a complementary click- reactive functional group which is labelled on the antigen binding moiety.

[0014] In some embodiments where the cell capture composition comprises a polymer-antigen binding moiety conjugate, the cell capture composition further comprises an unconjugated amphiphilic polymer, such as a triblock copolymer comprising a hydrophobic polypropylene glycol segment located between two hydrophilic polyethylene glycol segments.

[0015] In some embodiments, the antigen binding moiety is not conjugated to the amphiphilic polymer. In such embodiments, the amphiphilic polymer may be physicallyadhered to the hydrophobic upper surface after the antigen binding moiety is physically adhered to the hydrophobic upper surface.

[0016] In some embodiments, the multi-layer microfluidic device further comprises a base layer beneath the gas-permeable flow layer, wherein the base layer provides a gas flow path for gas exchange across the flow layer.

[0017] In some embodiments, the header layer comprises: a second outlet port; and one or more second fluid outlets providing fluid communication between the flow channel and the second outlet port, wherein the one or more second fluid outlets are positioned in the header such that fluid exiting the flow channel through the second outlet port flows in a second flow direction toward the one or more second fluid outlets, the second flow direction being along the one or more grooves.

[0018] In some embodiments, the header layer further comprises: a second inlet port; and one or more second fluid inlets providing fluid communication between the second inlet port and the flow channel, wherein the one or more second fluid inlets are positioned along a dimension of the header which is substantially orthogonal to a longitudinal dimension of the one or more grooves so that fluid entering the flow channel from the one or more second fluid inlets flows toward the one or more second outlets in the second flow direction.

[0019] In some such embodiments, the one or more second fluid inlets comprise a diverging flow guide so that flows of fluid entering the flow channel from each second fluid inlet are progressively spread across an increasing flow width when passing though the diverging flow guide. The diverging flow guides of neighbouring second fluid inlets may optionally meet in the flow channel so that the flows of fluid entering the flow channel from the neighbouring second fluid inlets are substantially adjacent when merged. The nearest end of each groove may optionally be set back from the one or more second fluid inlets by an ungrooved portion of the hydrophobic upper surface in the flow channel. The ungrooved portion may have a length of at least 1 .5 mm in the second flow direction.

[0020] In some embodiments, the multi-layer microfluidic device comprises a flow resistor downstream of the second outlet port to regulate the flow rate of fluid flowingin the second flow direction toward the one or more second fluid outlets in the flow channel.

[0021] In some embodiments, the first flow direction and the second flow direction are substantially orthogonal.

[0022] In some embodiments, the gas permeable flow layer comprises polymethylpentene and the polymethylpentene is present at the hydrophobic upper surface of the flow layer. Optionally, the header layer and a base layer beneath the gas permeable flow layer also comprise polymethylpentene.

[0023] In some embodiments, the flow layer has a thickness, between bottom surfaces of the one or more grooves and the lower surface of the flow layer, of less than 100 pm, preferably less than 65 pm.

[0024] In some embodiments, the header layer comprises a plurality of structural posts configured to rest in abutment with the hydrophobic upper surface of the flow layer. The structural posts may have an elongated cross-section, such as an elliptical cross-section, having an axis of elongation substantially aligned with the grooves.

[0025] In some embodiments, the multi-layer microfluidic device comprises a flow resistor to regulate the flow rate of fluid flowing in the flow channel in the first flow direction toward the one or more first fluid outlets. Preferably the flow resistor is located downstream of the first outlet port.

[0026] In some embodiments, the one or more grooves have a depth of between 1 10 pm and 180 pm. In some embodiments, the one or more grooves have an aspect ratio (width:depth) of between 1 .5:1 and 1 :1 .5, such as between 1 .2:1 and 1 :1 .2.

[0027] In some embodiments, the multi-layer microfluidic device comprises one or more through bores extending colinearly through each layer of the multi-layer microfluidic device, and wherein the multi-layer microfluidic device is configured for arrangement in a stack including a plurality of the multi-layer microfluidic devices such that the through bores in each multi-layer microfluidic device in the stack provide fluid communication with corresponding through bores of at least one adjacent multi-layer microfluidic device in the stack.

[0028] In some embodiments, the antigen binding moiety is an antibody or antigen binding fragment thereof. In some embodiments, the antibody or antigen binding fragment specifically binds CD34 or CD3.

[0029] In some embodiments, the target cells are hematopoietic stem cells.

[0030] According to a second aspect, the invention provides a bioreactor cell culture system comprising: (a) a plurality of bioreactors according to any embodiment of the first aspect arranged in at least one stack; (b) a controller configured to individually control operation of one or more fluid pumps according to a cell culture protocol stored in a memory of the controller, the one or more fluid pumps delivering fluid to one or more through bores provided colinearly through the at least one stack, the delivered fluid entering the flow chamber of each bioreactor in the at least one stack according to the stored cell culture protocol.

[0031] According to a third aspect, the invention provides a method of preparing a bioreactor for cell culture, comprising: providing a multi-layer microfluidic device comprising a header layer and a gas permeable flow layer beneath the header layer, wherein: (a) the flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells; and (b) the header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises: (i) a first inlet port; (ii) a first outlet port; (iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and (iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port; wherein the one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein; and physically adhering a cell capture composition to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

[0032] In some embodiments, the antigen binding moiety is conjugated to at least one of, and optionally each of, the one or more hydrophilic polymeric segments such that the cell capture composition comprises a polymer-antigen binding moiety conjugate.

[0033] In some embodiments, physically adhering the cell capture composition to the hydrophobic upper surface comprises flowing a first aqueous composition comprising the polymer-antigen binding moiety conjugate into the flow channel, and allowing the polymer-antigen binding moiety conjugate to deposit on the hydrophobic upper surface. The first aqueous composition is optionally flowed into the flow channel from the one or more first fluid inlets toward the one or more first fluid outlets in the first flow direction.

[0034] In some embodiments, the method further comprises depositing unconjugated amphiphilic polymer on the hydrophobic upper surface.

[0035] According to a fourth aspect, the invention provides a method of culturing cells, the method comprising: a) providing a microfluidic device comprising: a flow channel located over a substrate comprising a hydrophobic upper surface, wherein the hydrophobic upper surface comprises an array of grooves that are substantially parallel; and a cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell; b) introducing cells into the flow channel under conditions to immobilise the target cells on the cell capture composition; c) flowing a culture medium through the flow channel over the groove array in a first flow direction that is across the grooves; and d) culturing the target cells under conditions to support viability, growth and / or differentiation.

[0036] In some embodiments, the method further comprises releasing the target cells from the capture composition before or while flowing the culture medium through the flow channel, wherein the released target cells are retained in the grooves during the culturing.

[0037] In some embodiments, introducing the target cells comprises flowing a sample comprising the target cells through the flow channel, wherein the target cells are selectively immobilised on the cell capture composition.

[0038] In some embodiments, the target cells are introduced by flow through the flow channel in a second flow direction that is along the grooves. Preferably, the second flow direction is substantially orthogonal to the first flow direction.

[0039] In some embodiments, the culture medium is flowed through the flow channel at a flow rate equivalent to 0.2-5 mL per million cells per day.

[0040] In some embodiments, the target cells are cultured for 1 -14 days.

[0041] In some embodiments, the method further comprises harvesting the cultured cells by flowing an extraction medium through the flow channel. Preferably, the extraction medium is flowed through the flow channel in a second flow direction that is along the grooves. The flow of extraction medium may be provided at a shear stress of at least 0.1 dyne / cm2, or at least 0.5 dyne / cm2, such as at least 1 dyne / cm2.

[0042] According to a fifth aspect, the invention provides a method of transfecting / transducing cells, the method comprising: a) providing a microfluidic device comprising a flow channel located over a substrate comprising a hydrophobic upper surface, wherein the hydrophobic upper surface comprises an array of grooves that are substantially parallel; and a cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell; b) introducing cells into the flow channel under conditions to immobilise target cells on the cell capture composition; c) introducing at least one exogenous molecule into the flow channel; and d) incubating the target cells and the at least one exogenous molecule under conditions to transfect / transduce the target cells with the exogenous molecule.

[0043] In some embodiments, introducing the at least one exogenous molecule into the flow channel comprises flowing the at least one exogenous molecule through the flow channel over the groove array in a first flow direction that is across the grooves.

[0044] In some embodiments, the method further comprises releasing the target cells from the capture composition before or while introducing the at least one exogenous molecule into the flow channel, wherein the released target cells are retained in the grooves during the incubating.

[0045] In some embodiments, introducing the target cells comprises flowing a sample comprising the target cells through the flow channel, wherein the target cells are selectively immobilised on the cell capture composition.

[0046] In some embodiments, wherein the target cells are introduced by flow through the flow channel in a second flow direction that is along the grooves.

[0047] In some embodiments, the method further comprises flowing a culture medium through the flow channel to culture the target cells in the culture medium. Preferably the culture medium is flowed through the flow channel in a first flow direction that is across the grooves. The target cells may be cultured in the culture medium prior to introducing the exogenous molecule into the flow channel. Alternatively or in addition, the target cells may be cultured in the culture medium after transfecting / transducing the target cells with the exogenous molecule.

[0048] In some embodiments, the method further comprises harvesting the transfected / transduced cells by flowing an extraction medium through the flow channel. Preferably the extraction medium is flowed through the flow channel in a second flow direction that is along the grooves. The flow of extraction medium may be provided at a shear stress of at least 0.1 dyne / cm2, or at least 0.5 dyne / cm2, such as at least 1 dyne / cm2.

[0049] According to a sixth aspect, the invention provides a method of selecting cells, the method comprising: a) providing a microfluidic device comprising: a flow channel located over a substrate comprising a hydrophobic upper surface, wherein the hydrophobic upper surface comprises an array of grooves that are substantially parallel; and a cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell; and b) flowing a sample comprising the target cells through theflow channel over the groove array in a second flow direction that is along the grooves under conditions to selectively immobilise the target cells on the cell capture composition.

[0050] In some embodiments, the method further comprises releasing the target cells from the capture composition prior to or during at least one step selected from (i) culturing the target cells in the flow chamber, (ii) transfecting / transducing the target cells in the flow chamber and (iii) harvesting the target cells from the flow chamber.

[0051] In some embodiments, the method further comprises flowing a culture medium through the flow channel over the groove array in a first flow direction that is across the grooves; and culturing the target cells in the presence of the culture medium. Preferably the second flow direction is substantially orthogonal to the first flow direction. The culture medium may be flowed through the flow channel at a flow rate equivalent to 0.2-5 mL per million cells per day. The target cells may be cultured for 1 -14 days.

[0052] In some embodiments, the method further comprises introducing at least one exogenous molecule into the flow channel; and incubating the target cells and the at least one exogenous molecule under conditions to transfect / transduce the target cells with the exogenous molecule. Introducing the at least one exogenous molecule into the flow channel may comprise flowing the at least one exogenous molecule through the flow channel over the groove array in a first flow direction that is across the grooves.

[0053] In some embodiments, the method further comprises culturing the transfected / transduced cells, optionally by flowing a culture medium through the flow channel over the groove array in a first flow direction that is across the grooves and culturing the transfected / transduced cells in the presence of the culture medium.

[0054] In some embodiments, the method further comprises harvesting the target cells by flowing an extraction medium through the flow channel. Preferably the extraction medium is flowed through the flow channel in the second flow direction. The flow of extraction medium may be provided at an average flow velocity of at least 0.1 dyne / cm2, or at least 0.5 dyne / cm2, such as at least 1 dyne / cm2.

[0055] In some embodiments of the fourth, fifth or sixth aspects, the substrate is a gas permeable layer and the microfluidic device comprises one or more gas flow paths for exchange of gas across the gas permeable layer. Preferably the one or more gas flow paths are beneath the gas-permeable layer.

[0056] In some embodiments of the fourth, fifth or sixth aspects, the substrate comprises polymethylpentene and the polymethylpentene is present at the hydrophobic upper surface of the substrate.

[0057] In some embodiments of the fourth, fifth or sixth aspects, the hydrophobe is a hydrophobic polymeric segment, such as a hydrophobic polyalkylene glycol segment, for example a polypropylene glycol segment.

[0058] In some embodiments of the fourth, fifth or sixth aspects, the one or more hydrophilic polymeric segments are polyethylene glycol segments.

[0059] In some embodiments of the fourth, fifth or sixth aspects, the amphiphilic polymer is a triblock copolymer comprising a hydrophobic polypropylene glycol segment located between two hydrophilic polyethylene glycol segments.

[0060] In some embodiments of the fourth, fifth or sixth aspects, the amphiphilic polymer has a molecular weight in the range of 10,000 to 20,000 g / mol.

[0061] In some embodiments of the fourth, fifth or sixth aspects, the antigen binding moiety is conjugated to at least one of, and optionally each of, the one or more hydrophilic polymeric segments such that the cell capture composition comprises a polymer-antigen binding moiety conjugate. The antigen binding moiety may be conjugated to a chain terminus of the at least one hydrophilic polymeric segment.

[0062] In some such embodiments, the antigen binding moiety is conjugated to the chain terminus by an amide linking group, wherein the amide linkage is produced by a conjugation reaction between a terminal amine-reactive functional group on the at least one hydrophilic polymeric segment with an amine functional group of the antigen binding moiety. The amine-reactive functional group is optionally an activated ester functional group.

[0063] In other such embodiment, the antigen binding moiety is conjugated to a chain terminus of the at least one hydrophilic polymeric segment by a click reaction between a terminal click-reactive functional group on the at least one hydrophilic polymeric segment and a complementary click-reactive functional group which is labelled on the antigen binding moiety.

[0064] In some embodiments of the fourth, fifth or sixth aspects where the cell capture composition comprises a polymer-antigen binding moiety conjugate, the cell capture composition further comprises an unconjugated amphiphilic polymer, such as a triblock copolymer comprising a hydrophobic polypropylene glycol segment located between two hydrophilic polyethylene glycol segments.

[0065] In some embodiments of the fourth, fifth or sixth aspects where the cell capture composition comprises a polymer-antigen binding moiety conjugate, providing the microfluidic device comprises flowing the polymer-antigen binding moiety conjugate into the flow channel, and incubating the hydrophobic upper surface with the polymer- antigen binding moiety conjugate for a sufficient period of time to allow the polymer- antigen binding moiety conjugate to physically adhere to the hydrophobic upper surface in the grooves. Providing the microfluidic device may further comprise flowing an unconjugated amphiphilic polymer into the flow channel, and incubating the hydrophobic upper surface with the unconjugated amphiphilic polymer for a sufficient period of time to allow the unconjugated amphiphilic polymer to physically adhere to the hydrophobic upper surface in the grooves, wherein the cell capture composition comprises the polymer-antigen binding moiety conjugate and the unconjugated amphiphilic polymer.

[0066] In some embodiments of the third, fourth, fifth or sixth aspects, the antigen binding moiety is an antibody or antigen binding fragment thereof. Optionally, the antibody or antigen binding fragment specifically binds CD34 or CD3.

[0067] In some embodiments of the third, fourth, fifth or sixth aspects, the target cells are hematopoietic stem cells.

[0068] According to a seventh aspect, the invention provides a multi-layer microfluidic device comprising a header layer and a gas permeable flow layer beneath the header layer, wherein: (a) the flow layer comprises a hydrophobic upper surfaceand a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells; and (b) the header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises: (i) a first inlet port; (ii) a first outlet port; (iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and (iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port; wherein the one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein, and wherein the gas permeable flow layer comprises polymethylpentene and the polymethylpentene is present at the hydrophobic upper surface of the flow layer.

[0069] In some embodiments, the header layer and a base layer beneath the gas permeable flow layer comprise polymethylpentene. Preferably, the header layer, the flow layer and the base layer are thermally fused above the glass transition temperature of the polymethylpentene to form the multi-layer microfluidic device.

[0070] In some embodiments, fluid flowing from the first inlet port towards the flow channel is distributed to a plurality of first fluid inlets via one or more stages of flow bifurcation, such as two or more stages of flow bifurcation, or three or more stages of flow bifurcation, such as four or more stages of flow bifurcation.

[0071] In some embodiments, the multi-layer microfluidic device further comprises a base layer beneath the gas-permeable flow layer, wherein the base layer provides a gas flow path for gas exchange across the flow layer.

[0072] In some embodiments, the header layer comprises: a second outlet port; and one or more second fluid outlets providing fluid communication between the flow channel and the second outlet port, wherein the one or more second fluid outlets are positioned in the header such that fluid exiting the flow channel through the second outlet port flows in a second flow direction toward the one or more second fluid outlets, the second flow direction being along the one or more grooves.

[0073] In some embodiments, the header layer further comprises: a second inlet port; and one or more second fluid inlets providing fluid communication between the second inlet port and the flow channel, wherein the one or more second fluid inlets are positioned along a dimension of the header which is substantially orthogonal to a longitudinal dimension of the one or more grooves so that fluid entering the flow channel from the one or more second fluid inlets flows toward the one or more second outlets in the second flow direction.

[0074] In some embodiments, fluid flowing from the second inlet port towards the flow channel is distributed to a plurality of second fluid inlets via one or more stages of flow bifurcation, or two or more stages of flow bifurcation, such as three or more stages of flow bifurcation, for example four or more stages of flow bifurcation.

[0075] In some embodiments, the one or more second fluid inlets comprise a diverging flow guide so that flows of fluid entering the flow channel from each second fluid inlet are progressively spread across an increasing flow width when passing though the diverging flow guide. The diverging flow guides of neighbouring second fluid inlets may meet in the flow channel so that the flows of fluid entering the flow channel from the neighbouring second fluid inlets are substantially adjacent when merged. The nearest end of each groove may be set back from the one or more second fluid inlets by an ungrooved portion of the hydrophobic upper surface in the flow channel. The ungrooved portion optionally has a length of at least 1 .5 mm in the second flow direction.

[0076] In some embodiments, the header layer is configured to define a filter for removing particles from fluid flowing towards the one or more second fluid inlets. The filter may comprise a filter channel with spaced apart posts arranged in at least one row extending across the filter channel. The posts may be arranged in a plurality of rows extending across the filter channel and the spacing between posts in each row decreases in the direction of flow through the filter.

[0077] In some embodiments, the first flow direction and the second flow direction are substantially orthogonal.

[0078] In some embodiments, the multi-layer microfluidic device comprises a flow resistor downstream of the second outlet port to regulate the flow rate of fluid flowingin the second flow direction toward the one or more second fluid outlets in the flow channel.

[0079] In some embodiments, the header comprises a plurality of structural posts configured to rest in abutment with the hydrophobic upper surface of the flow layer. The structural posts may have an elongated cross-section, such as an elliptical crosssection, having an axis of elongation substantially aligned with the grooves.

[0080] In some embodiments, the flow layer has a maximum thickness, beneath the flow channel, of no more than 400 pm, or between 100 and 300 pm, such as between 150 and 250 pm.

[0081] In some embodiments, the flow layer has a thickness, between bottom surfaces of the one or more grooves and the lower surface of the flow layer, of less than 100 pm, such as less than 65 pm.

[0082] In some embodiments, the one or more grooves have a depth of between 1 10 pm and 180 pm. In some embodiments, the one or more grooves have an aspect ratio (width:depth) of between 1 .5:1 and 1 :1 .5, such as between 1 .2:1 and 1 :1 .2.

[0083] In some embodiments, the header layer, the flow layer and a base layer beneath the gas-permeable flow layer each have a maximum thickness, beneath or above the flow channel, of no more than 400 pm, such as no more than 300 pm.

[0084] According to an eighth aspect, the invention provides a method of fabricating a multi-layer microfluidic device according to any embodiment of the seventh aspect, the method comprising: (i) producing the header layer, the flow layer and optionally a base layer as separate components by (a) moulding polymethylpentene under compression in a mould, at a temperature above the polymethylpentene melting point, to produce a structured layer, (b) cooling the structured layer, (c) demoulding the structured layer before the polymethylpentene cools to ambient temperature; and (d) annealing the structured layer under compression against a flat substrate at a temperature between the polymethylpentene glass transition temperature and the polymethylpentene melting point, thereby reducing the surface roughness of the structured layer; and (ii) aligning the header layer, the flow layer and optionally the base layer and compressing the aligned layers at a temperature above the glass transitiontemperature of the polymethylpentene to thermally fuse the layers, thereby forming the multi-layer microfluidic device.

[0085] In some embodiments, the polymethylpentene is moulded under compression of between 1 and 5 tons at a temperature of between 230 and 270°C, such as between 240 and 260°C, for example about 250°C.

[0086] In some embodiments, the polymethylpentene is above the polymethylpentene melting point for less than 60 seconds during moulding.

[0087] In some embodiments, the structured layer is annealed under compression of between 0.1 and 0.5 tons, such as between 0.2 and 0.4 tons, against an optically flat substrate at a temperature of between 180 and 200°C, such as between 175 and 190°C, for example about 180°C.

[0088] In some embodiments, the header layer, the flow layer and optionally the base layer are compressed at between 0.3 and 1 .5 tons, such as between 0.6 and 1 .0 tons, at a temperature of between 180 and 200°C, such as between 175 and 190°C, for example about 180°C, to thermally fuse the layers.

[0089] According to a ninth aspect, the invention provides a bioreactor for cell culture, comprising: the multi-layer microfluidic device according to any embodiment of the seventh aspect; and a cell capture composition physically adhered to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

[0090] According to a tenth aspect, the invention provides a method of culturing cells, the method comprising: i) providing a multi-layer microfluidic device according to any embodiment of the seventh aspect; ii) introducing cells into the flow channel under conditions to deposit the cells in the one or more grooves; ii) flowing a culture medium through the flow channel in a first flow direction that is across the one or more grooves; and iv) culturing the cells under conditions to support viability, growth and / or differentiation.

[0091] In some embodiments, the cells are introduced into the flow channel by flow in the first flow direction, wherein the cells deposit in the one or more grooves by sedimentation.

[0092] In some embodiments, the multi-layer microfluidic device comprises a cell capture composition physically adhered to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

[0093] According to an eleventh aspect, the invention provides a multi-layer microfluidic device comprising a header layer and a gas permeable flow layer beneath the header layer, wherein: (a) the flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells; and (b) the header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises: (i) a first inlet port; (ii) a first outlet port; (iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and (iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port; wherein the one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein; wherein the header layer further comprises: a second outlet port; and one or more second fluid outlets providing fluid communication between the flow channel and the second outlet port; wherein the one or more second fluid outlets are positioned in the header such that fluid exiting the flow channel through the second outlet port flows in a second flow direction toward the one or more second fluid outlets, the second flow direction being along the one or more grooves; wherein the header layer further comprises: a second inlet port; and one or more second fluid inlets providing fluid communication between the second inlet port and the flow channel; wherein the one or more second fluid inlets are positioned along a dimension of the header which is substantially orthogonal to alongitudinal dimension of the one or more grooves so that fluid entering the flow channel from the one or more second fluid inlets flows toward the one or more second outlets in the second flow direction; and wherein the multi-layer microfluidic device is characterized by one or more of the following features:• the one or more second fluid inlets comprise a diverging flow guide so that flows of fluid entering the flow channel from each second fluid inlet are progressively spread across an increasing flow width when passing though the diverging flow guide, preferably wherein the diverging flow guides of neighbouring second fluid inlets meet in the flow channel so that the flows of fluid entering the flow channel from the neighbouring second fluid inlets are substantially adjacent when merged, further preferably wherein the nearest end of each groove is set back from the one or more second fluid inlets by an ungrooved portion of the hydrophobic upper surface in the flow channel, optionally wherein the ungrooved portion has a length of at least 1 .5 mm in the second flow direction;• the multi-layer microfluidic device comprises a flow resistor downstream of the second outlet port to regulate the flow rate of fluid flowing in the second flow direction toward the one or more second fluid outlets in the flow channel;• the multi-layer microfluidic device comprises a flow resistor to regulate the flow rate of fluid flowing in the flow channel in the first flow direction toward the one or more first fluid outlets, preferably wherein the flow resistor is located downstream of the first outlet port;• the header layer comprises a plurality of structural posts configured to rest in abutment with the hydrophobic upper surface of the flow layer, wherein the structural posts have an elongated cross-section, preferably an elliptical crosssection, having an axis of elongation substantially aligned with the grooves;• the header layer is configured to define a filter for removing particles from fluid flowing towards the one or more second fluid inlets, preferably wherein the filter comprises a filter channel with spaced apart posts arranged in at least one row extending across the filter channel, more preferably wherein the posts arearranged in a plurality of rows extending across the filter channel and the spacing between posts in each row decreases in the direction of flow through the filter;• the one or more grooves have a depth of between 1 10 pm and 180 pm, and preferably an aspect ratio (width:depth) of between 1.5:1 and 1 :1.5, such as between 1.2:1 and 1 :1 .2;• the flow layer has a thickness, between bottom surfaces of the one or more grooves and the lower surface of the flow layer, of less than 100 pm, such as less than 65 pm;• the flow layer has a maximum thickness, beneath the flow channel, of no more than 400 pm, such as between 100 and 300 pm, for example between 150 and 250 pm; and• the gas permeable flow layer comprises, and preferably consists of, polymethylpentene and the polymethylpentene is present at the hydrophobic upper surface of the flow layer, preferably wherein the header layer comprises, and preferably consists of, polymethylpentene.

[0094] In some embodiments of any of the above aspects, the antigen binding moiety that specifically binds to a target cell is an antibody or an antigen binding fragment thereof.

[0095] In some such embodiments, the antibody or an antigen binding fragment thereof specifically binds to CD34 and comprises: a. a heavy chain variable region (VH) comprising a complementary determining region (CDR) H1 comprising a sequence as set forth in SEQ ID NO: 1 ; CDR H2 comprising a sequence as set forth in SEQ ID NO: 2; and CDR H3 comprising a sequence as set forth in SEQ ID NO: 3; and a light chain variable region (VL) comprising a CDR L1 comprising a sequence as set forth in SEQ ID NO: 4; CDR L2 comprising a sequence as set forth in SEQ ID NO: 5; and CDR L3 comprising a sequence as set forth in SEQ ID NO: 6; orb. a heavy chain variable region (VH) comprising complementary determining regions (CDRs) H1 , H2, and H3 of the sequence as set forth in SEQ ID NO: 7; and a light chain variable region (VL) comprising CDRs L1 , L2, and L3 of the sequence as set forth in SEQ ID NO: 8.

[0096] In another or further embodiments, the antibody or an antigen binding fragment thereof specifically binds to CD34 and comprises: a. a heavy chain variable region (VH) comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 7; and a light chain variable region (VL) comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 8; or b. a single-chain antibody fragment comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 9.

[0097] In some embodiments, the antibody or an antigen binding fragment thereof specifically binds to CD34 and comprises: a. a heavy chain variable region (VH) comprising a complementary determining region (CDR) H1 comprising a sequence as set forth in SEQ ID NO: 10; CDR H2 comprising a sequence as set forth in SEQ ID NO: 1 1 ; and CDR H3 comprising a sequence as set forth in SEQ ID NO: 12; and a light chain variable region (VL) comprising a CDR L1 comprising a sequence as set forth in SEQ ID NO: 13; CDR L2 comprising a sequence as set forth in SEQ ID NO: 14; and CDR L3 comprising a sequence as set forth in SEQ ID NO: 15; or b. a heavy chain variable region (VH) comprising complementary determining regions (CDRs) H1 , H2, and H3 of the sequence as set forth in SEQ ID NO: 16; and a light chain variable region (VL) comprising CDRs L1 , L2, and L3 of the sequence as set forth in SEQ ID NO: 17.

[0098] In some embodiments, the antibody or an antigen binding fragment thereof specifically binds to CD3 and comprises a heavy chain variable region (VH) comprising a sequence that is at least 90% identical to the sequence as set forth inSEQ ID NO: 16; and a light chain variable region (VL) comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 17.

[0099] In some embodiments, the antibody or an antigen binding fragment thereof specifically binds to CD34 and comprises: a. a heavy chain variable region (VH) comprising a complementary determining region (CDR) H1 comprising a sequence as set forth in SEQ ID NO: 18; CDR H2 comprising a sequence as set forth in SEQ ID NO: 19; and CDR H3 comprising a sequence as set forth in SEQ ID NO: 20; and a light chain variable region (VL) comprising a CDR L1 comprising a sequence as set forth in SEQ ID NO: 21 ; CDR L2 comprising a sequence as set forth in SEQ ID NO: 22; and CDR L3 comprising a sequence as set forth in SEQ ID NO: 23; or b. a heavy chain variable region (VH) comprising complementary determining regions (CDRs) H1 , H2, and H3 of the sequence as set forth in SEQ ID NO: 24; and a light chain variable region (VL) comprising CDRs L1 , L2, and L3 of the sequence as set forth in SEQ ID NO: 25.

[0100] In some embodiments, the antibody or an antigen binding fragment thereof specifically binds to CD34 and comprises: a. a heavy chain variable region (VH) comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 24; and a light chain variable region (VL) comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 25; or b. a heavy chain comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 26; and a light chain comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 27.

[0101] In some embodiments, the antibody or an antigen binding fragment thereof specifically binds to CD3 and comprises:a. a heavy chain variable region (VH) comprising a complementary determining region (CDR) H1 comprising a sequence as set forth in SEQ ID NO: 28; CDR H2 comprising a sequence as set forth in SEQ ID NO: 29; and CDR H3 comprising a sequence as set forth in SEQ ID NO: 30; and a light chain variable region (VL) comprising a CDR L1 comprising a sequence as set forth in SEQ ID NO: 31 ; CDR L2 comprising a sequence as set forth in SEQ ID NO: 32; and CDR L3 comprising a sequence as set forth in SEQ ID NO: 33; or b. a heavy chain variable region (VH) comprising complementary determining regions (CDRs) H1 , H2, and H3 of the sequence as set forth in SEQ ID NO: 34; and a light chain variable region (VL) comprising CDRs L1 , L2, and L3 of the sequence as set forth in SEQ ID NO: 35.

[0102] In some embodiments, the antibody or an antigen binding fragment thereof specifically binds to CD3 and comprises: a. a heavy chain variable region (VH) comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 34; and a light chain variable region (VL) comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 35; or b. a heavy chain comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 36; and a light chain comprising a sequence that is at least 90% identical to the sequence as set forth in SEQ ID NO: 37.

[0103] Unless the context requires otherwise, where the terms “comprise”, “comprises”, “comprised” or “comprising” are used in this specification (including the claims) they are to be interpreted as specifying the presence of the stated features, integers, steps or components, but not precluding the presence of one or more other features, integers, steps or components, or group thereof.

[0104] As used herein, the terms “first”, “second”, “third” etc in relation to various features of the disclosed devices, methods, systems etc are arbitrarily assigned andare merely intended to differentiate between two or more such features that the device, methods, systems etc may incorporate in various embodiments. The terms do not of themselves indicate any particular orientation or sequence. Moreover, it is to be understood that the presence of a “first” feature does not imply that a “second” feature is present, the presence of a “second” feature does not imply that a “first” feature is present, etc.Brief Description of Drawings

[0105] The present invention will now be described in greater detail with reference to the accompanying drawings. It is to be understood that the embodiments shown are examples only and are not to be taken as limiting the scope of the invention as defined in the claims appended hereto.

[0106] Figure 1 is an exploded view of a multi-layer microfluidic device according to an embodiment of the disclosure.

[0107] Figure 2 depicts a side cross-sectional view of the multi-layer microfluidic device of Figure 1 .

[0108] Figure 3 is an exploded view of a multi-layer microfluidic device according to another embodiment of the disclosure, which provides for bidirectional flow across and along the grooves.

[0109] Figure 4 depicts in plan view a header layer for a multi-layer microfluidic device according to another embodiment of the disclosure.

[0110] Figure 5 is an exploded view of a multi-layer microfluidic device according to another embodiment of the disclosure, which includes the header layer design of Figure 4.

[0111] Figure 6 is an exploded view of a multi-layer microfluidic device according to another embodiment of the disclosure.

[0112] Figure 7 depicts in plan view a header layer design for the multi-layer microfluidic device of Figure 6.

[0113] Figure 8 depicts in plan view the header layer of Figure 7 overlaid over the flow layer in the multi-layer microfluidic device of Figure 6.

[0114] Figure 9 depicts a magnified view of the region marked C in Figure 8.

[0115] Figure 10 depicts in plan view another header layer design for the multi-layer microfluidic device of Figure 6, having a downstream flow resistor.

[0116] Figure 11 is an exploded view of a multi-layer microfluidic device according to another embodiment of the disclosure, including a selectively permeable membrane.

[0117] Figure 12 depicts a side cross-sectional view of the multi-layer microfluidic device of Figure 1 1 .

[0118] Figure 13 is an exploded view of a stack of multi-layer microfluidic devices according to the present disclosure.

[0119] Figure 14 is an assembled view of the stack of Figure 13.

[0120] Figure 15 depicts structured sheets comprising four individual header layers.

[0121] Figure 16 is a graph showing the number of target CD34+ cells captured on a cell capture composition comprising poloxamer mixed with unconjugated MY10 antibody (Example 6).

[0122] Figure 17 is a graph showing the number of target CD34+ cells captured on a cell capture composition comprising poloxamer mixed with poloxamer-MY10 antibody conjugate (Example 6).

[0123] Figure 18 is a graph showing the number of target CD3+ cells captured on a cell capture composition comprising poloxamer mixed with unconjugated OKT3 antibody (Example 6).

[0124] Figure 19 is a graph showing the number of target CD3+ cells captured on a cell capture composition comprising poloxamer mixed with poloxamer-OKT3 antibody (Example 6).

[0125] Figure 20 is a graph showing the effect of competitive absorption of free poloxamer with unconjugated OKT3 antibody vs with poloxamer-OKT3 antibody on cell adhesion (Example 6).

[0126] Figure 21 is a graph showing the number of target CD34+ cells captured on different cell capture compositions comprising poloxamer and MY10 antibody, on TPX substrate immediately after cell selection (Example 7).

[0127] Figure 22 is a graph showing the number of target CD34+ cells captured on different cell capture compositions comprising poloxamer and MY10 antibody, on TPX substrate and after 85 hours of culturing (Example 7).

[0128] Figure 23 is a graph showing the number of target CD34+ cells captured on cell capture compositions comprising MY10 antibody, on polystyrene and TPX substrates (Example 8).

[0129] Figure 24 schematically depicts the flowcell apparatus described in Example 9.

[0130] Figure 25 is a graph showing the percentage of target CD34+ cells captured on different cell capture compositions comprising poloxamer and MY10 antibody, on TPX flowcell substrate (Example 10).

[0131] Figure 26 is a graph showing the percentage of target CD34+ cells captured on different cell capture compositions comprising poloxamer and MY10 antibody, on polystyrene and TPX flowcell substrates (Example 1 1 ).

[0132] Figure 27 is a graph showing the percentage of negative CD3+ cells captured on different cell capture compositions comprising poloxamer and MY10 antibody, on polystyrene and TPX flowcell substrates (Example 1 1 ).

[0133] Figure 28 is a graph showing the percentage of target antiCD34+ and negative CD3+ cells captured on different cell capture compositions comprising poloxamer and MY10 antibody, on TPX flowcell substrates (Example 12).

[0134] Figure 29 is a graph showing the percentage of target CD34+ cells captured on cell capture compositions comprising poloxamer-MY10 antibody conjugates prepared by NHS or click chemistry, on flowcell substrates (Example 13).

[0135] Figure 30 is a graph showing the percentage of target antiCD34+ and negative CD3+ cells captured on different cell capture compositions comprising poloxamer and QBEND10 antibody, on polystyrene flowcell substrates and with different flow stress during flowcell washing (Example 14).

[0136] Figure 31 is a graph showing the percentage of target antiCD34+ and negative CD3+ cells captured on different cell capture compositions comprising poloxamer and QBEND10 antibody, on TPX flowcell substrates and with different flow stress during flowcell washing (Example 14).

[0137] Figure 32 is a graph showing the percentage of target antiCD34+ and negative CD3+ cells captured on different cell capture compositions comprising different CD34 antibodies, on polystyrene flowcell substrates (Example 15).

[0138] Figure 33 is a graph showing the percentage of positive CD3+ cells captured on different cell capture compositions comprising poloxamer and OKT3 antibody, on polystyrene and TPX flowcell substrates (Example 16).

[0139] Figure 34 is a graph showing the percentage of target antiCD34+ and negative CD3+ cells captured on different cell capture compositions comprising poloxamer and different CD34 antibodies, on polymethylpentene flowcell substrates (Example 17).

[0140] Figure 35 is a graph showing the percentage of target antiCD34+ and negative CD3+ cells captured on different cell capture compositions comprising poloxamer and different CD34 antibodies, on polystyrene flowcell substrates (Example 17).

[0141] Figure 36 is a graph showing the % purity of CD34+ cells captured on different cell capture compositions comprising poloxamer and different CD34 antibodies, on polystyrene and polymethylpentene (TPX) flowcell substrates (Example 17).

[0142] Figure 37 depicts the header layer design of a TPX multilayer microfluidic device as produced in Example 23.

[0143] Figure 38 is a photograph of a clamped and fluidically connected TPX multilayer microfluidic device as produced in Example 24.

[0144] Figure 39 is a graph of the flow rate response to input pressure for a TPX multilayer microfluidic device as produced in Example 23 and tested in Example 25.

[0145] Figure 40 depicts a design change to the fluid inlets to the flow channel of a TPX multilayer microfluidic device as developed in Example 27.

[0146] Figure 41 depicts a design change to the fluid inlets, and groove array setback of a TPX multilayer microfluidic device as developed in Example 27.

[0147] Figure 42 depicts a design change to the structural posts separating flow and header layers of a TPX multilayer microfluidic device as developed in Example 27.

[0148] Figure 43 depicts design changes to the fluid inlets, groove array and structural posts of a TPX multilayer microfluidic device as developed in Example 27.

[0149] Figure 44 depicts a cell filter of a TPX multilayer microfluidic device as developed in Example 27.Sequence Listing

[0150] CDRs were identified by aligning sequences to mouse germ line sequences using IMGT Domain Gap Align tool.Detailed Description

[0151] The invention relates in various aspects to a multi-layer microfluidic device. The invention further relates in various aspects to a bioreactor for cell culture comprising a multi-layer microfluidic device and a cell capture composition physically adhered to a grooved, hydrophobic surface of a flow channel defined between layers of the microfluidic device.Multi-layer micro fluidic device

[0152] Referring firstly to Figures 1 and 2, there is shown an exploded schematic view of a multi-layer microfluidic device 100 according to an embodiment of the disclosure. Device 100 comprises a header layer 200, a gas permeable flow layer 300 located beneath the header layer, and optionally a base layer 400 located beneath the flow layer, with the flow layer thus located between the header layer and the base layer.

[0153] Flow layer 300 comprises a hydrophobic upper surface 310 and lower surface 320, and the upper surface comprises at least one and typically several grooves 312 forming a groove array in the flow layer upper surface for retention of cells 50 (as seen in Figure 2). Preferably, the hydrophobic upper surface comprises an array of grooves that are substantially parallel.

[0154] Header layer 200 has a first inlet port 210 which feeds one or more first fluid inlets 212 to flow channel 60 which is formed between the lower surface 221 of header 200 and the hydrophobic upper surface 310 of flow layer 300. Thus, grooves 312 form the floor of flow channel 60 in which cells 50 deposit and grow. First inlet port 210 may facilitate delivery of cells and culture medium through first inlets 212. One or more first fluid outlets 222 located at the opposite side of flow channel 60 provide fluidcommunication with first outlet port 220 which is provided to facilitate removal of waste fluid. In some embodiments, cell harvesting may also occur via outlet port 220.

[0155] First fluid inlets 212 and first fluid outlets 222 are positioned in the header layer such that fluid entering flow channel 60 moves in a first flow direction A from the fluid inlets toward the fluid outlets such that flow is directed across the grooves 312. This cross-flow direction can be used to load cells into one or more of the grooves 312. In some uses, cells introduced into flow channel 60 by cross-flow deposit into one or more grooves by sedimentation due to the difference in density of the cell and surrounding fluid even under flowing conditions. Cells deposit in regions of low shear stress and are retained there by depositional and hydrodynamic forces even under flow conditions. In some such uses, no cell capture composition may be required in the flow channel. In other uses, the cells introduced into flow channel 60 by cross-flow are immobilised by a cell capture composition physically adhered to the hydrophobic upper surface 310 of the flow layer, at least in the grooves thereof, as will be described further in subsequent sections. Optionally, the immobilised cells may then be released from the cell capture composition. In some embodiments cell release facilitates immediate harvesting of the separated cells. In other embodiments, the cells are released prior to or during a subsequent culturing step.

[0156] The cross-flow direction (flow in direction A) can also be used to provide cell culture medium across the top of the grooves 312 to provide nutrients and growth factors for cell metabolism and remove the waste products generated by the cells retained therein without displacing the cells. The cells are retained in the grooves and cultured in regions of low shear stress even when (i) a high flow rate of culture medium is flowed across the grooves to sustain the cells, and / or (ii) the cells are not immobilised in the grooves.

[0157] The flow rate may modulate the transport and deposition of cells into the grooves 312 and retention of cells in the grooves. In some uses, the cross flow rate is such that cells introduced by cross flow deposit into one or more grooves 312 near the fluid entrance to flow channel 60. The skilled person will appreciate that cell deposition will also be a function of the cell size and cell density and properties of the fluid in which they are carried. Similarly, cell retention within the grooves, e.g. during culturing, may be controlled by cell concentration and flow rate. The Reynolds number and devicegeometry (e.g., groove depth and aspect ratio) will also impact on cell retention. The Reynolds number is directly proportional to the cross flow velocity and inversely related to the plate separation. During cross flow, the retention of cells in the device will depend on the volume of flow recirculation in the base of grooves, which can be determined by fluid dynamic simulation. As the cells expand to fill the groove, they will be displaced from the groove, flow across the flow channel 60 and deposit by sedimentation into an adjacent groove until the groove array is filled.

[0158] In the arrangement shown in Figures 1 and 2, flow is in direction A, or “cross flow” which is across the longitudinal dimension of grooves 312 in flow layer 300. Typically, this cross flow is orthogonal (90°) to the longitudinal dimension of grooves 312 although it is also contemplated that cross flow need not be precisely orthogonal to achieve the beneficial effects of delivering cells and culture medium using cross flow. In an embodiment, the angle of cross flow is in the range of approximately 75° to 90° relative to the longitudinal dimension of the grooves, preferably approximately 80° to 90°, more preferably approximately 85° to 90°. In an embodiment, the cross flow is approximately 90° relative to the longitudinal dimension of the grooves.

[0159] Fluid flowing from first inlet port 220 towards flow channel 60 may be distributed to a plurality of first fluid inlets 212, which are preferably spread substantially uniformly across the full width of flow channel 60. The flow may be split to these first fluid inlets via one or more, typically four or more, such as five, stages of flow bifurcation, thus ensuring that a similar flow rate of fluid is emitted through each first fluid inlet 212 into the flow channel. Similarly, the flow at the opposite end of flow channel 60 is received at a plurality of first fluid outlets 222, again preferably spread substantially uniformly across the full width of the flow channel, and recombined via a flow path comprising one or more, typically four or more, such as five, stages of flow convergence (i.e. reverse to bifurcation) before reaching first outlet port 220. In this manner, the flow of fluid through the flow channel is uniformly directed in direction A and the flow stress and Reynolds number experienced at each point across the length of the grooves 312 will be substantially consistent.

[0160] Optional base layer 400 provides a gas flow path 410 for gas exchange across gas permeable flow layer 300. Respiratory gas feed (e.g. air + 10% CO2) can enter gas flow path 410 through gas inlet port 430 and expired gases escape throughgas outlet port 440. As used herein, a respiratory gas feed contains gases for sustaining cells, including O2 and typically also CO2. In some embodiments, the respiratory gas feed comprises air supplemented with CO2, for example up to 15%. Since flow layer 300 is gas permeable, gases in gas flow path 410 (including O2) perfuse through the grooved flow layer where they are exchanged with cells 50 retained in flow channel 60. Additionally, gases contained within the media circulated through flow channel 60 are also taken up by cells 50. Base layer 400 is sealed to header layer 200 either directly e.g. by bonding to a base layer flange surrounding flow layer 300, or indirectly by bonding together each of the three layers of the device.

[0161] Base layer 400 thus provides a useful means of controlling the gas composition for perfusion across the gas permeable flow layer, particularly when a plurality of the multi-layer microfluidic devices are arranged in a stack in a bioreactor cell culture system. However, it will be appreciated that the base layer is not essential in all embodiments, for example, when a single multi-layer microfluidic device is to be used in an incubator the base layer may not be essential for controlling the gas composition beneath the gas permeable flow layer.

[0162] Figure 3 is an exploded schematic view of a multi-layer microfluidic device 100 according to another embodiment of the disclosure which provides for flow in first flow direction A (cross flow), and flow in a second flow direction B which corresponds to the longitudinal dimension of grooves 312 (longitudinal flow). Here, header layer 200 comprises, additionally to the features previously described in the context of Figure 1 , a second inlet port 330 which feeds one or more second fluid inlets 332 to flow channel 60. One or more second fluid outlets 342 are located at the opposite side of flow channel 60 to second fluid inlets 332 and provide fluid communication with second outlet port 340, which is provided for removal of fluids cells from the flow channel.

[0163] In the embodiment illustrated in Figure 3, second inlet port 330 may be used to deliver fluid containing cells in direction B along the grooves 312, with waste removed via second outlet port 340. Target cells may be captured on a cell capture composition physically adhered to the hydrophobic upper surface 310 of the flow layer, at least in the grooves thereof, as will be described further in subsequent sections. After cell deposition, and optionally also after release of the captured cells from the cell capture composition, the cells may be cultured with a cross flow of cell culture medium indirection A delivered via first inlet port 210, with waste removed via first outlet port 220. The cells are retained in the grooves and cultured in regions of low shear stress even when high flow rates of culture medium are flowed across the grooves to sustain the cells. After culturing the cells, cell extraction fluid (also referred to herein as “extraction medium”) is delivered via second inlet port 330 to wash cultured cells along grooves 312 for removal through second fluid outlets 342 and second outlet port 340.

[0164] It is to be understood however, that second inlet 330 port and second fluid inlets 332 need not be provided. First inlet port 210 and first fluid inlets 212 may be used to deliver fluid containing cells for loading in grooves 312 (either with or without a cell capture composition present in the grooves), and for delivering a cell extraction fluid for washing cultured cells along grooves 312 for removal through second fluid outlets 342 and second outlet port 340. In both arrangements, removal of cells via second outlet port 340 is along grooves 312 in a longitudinal flow direction B, enabling the cells to be displaced from the grooves efficiently, for removal through second fluid outlets 342.

[0165] A further embodiment showing both cross flow and longitudinal flow is presented in Figure 4. Figure 4 illustrates the lower surface 221 , or underside, of a header layer 200. First inlet port 210 is in fluid communication with 32 evenly spacedapart first fluid inlets 212 to flow channel 60, with fluid flowing from first inlet port 210 being split via five stages of bifurcation to feed the first fluid inlets 212. On the opposite side of flow channel 60, 32 evenly spaced-apart first fluid outlets 222 are in fluid communication with first outlet port 220, with fluid flowing through the first fluid outlets 222 being re-combined via five stages of convergence before reaching first outlet port 220.

[0166] Similarly, second inlet port 330 is in fluid communication with 32 evenly spaced-apart second fluid inlets 332 to flow channel 60, with fluid flowing from second inlet port 330 being split via five stages of bifurcation to feed the second fluid inlets 332. On the opposite side of flow channel 60, 32 evenly spaced-apart second fluid outlets 342 are in fluid communication with second outlet port 340, with fluid flowing through the second fluid outlets 342 being re-combined via five stages of convergence before reaching second outlet port 340.

[0167] In this arrangement, it is possible to achieve removal of cells 50 by flow along grooves 312 in a second flow direction B (longitudinal flow). In a manner similar to the embodiment shown in Figure 3, fluid containing cells may enter via second inlet port 330, with the cells propagating along the grooves 312 for immobilisation of target cells on a cell capture composition physically adhered to hydrophobic upper surface 310 in the grooves. Alternatively, fluid containing cells may enter via first inlet port 210, with the cells depositing by sedimentation into grooves 312 (e.g. in the absence of any cell capture composition present in the grooves). In either case, the captured cells may then be cultured in the presence of culture medium delivered via first inlet port 210 and flowed in direction A across grooves 312 in the flow channel, optionally after first releasing the cells from a cell capture composition that was used to immobilise the target cells. At the conclusion of the cell culture process, cells may be removed through second outlet port 340 such that cultured cells are displaced along the grooves 312 to second outlet port 340. Laminar flow along the grooves 312 displaces the retained cells 50 toward second outlet port 340 so they can be collected in a sealed receptacle for removal from the bioreactor.

[0168] As would be understood by one of skill in the art, the term “port”, as it is used herein in the context of “inlet ports” and “outlet ports”, is to be broadly construed as part of a flow path that provides for the flowing in or flowing out of fluid relative to the flow channel 60. For example, first fluid inlet port 210 may be provided as an aperture (Figure 1 ) that provides for flowing in of fluid by direct fluid coupling with a first fluid source by a tube or the like (not shown). Alternatively, first fluid inlet port 210 may be provided as a junction (Figure 4) in a fluid flow path that provides for flowing in of fluid by a more indirect fluid pathway in fluid communication with first through bores 381 , 382.

[0169] As shown in Figure 3 and Figure 4, second inlet port 330 is in fluid communication with one or more second fluid inlets 332 providing fluid communication with flow channel 60. The one or more second fluid inlets 332 are positioned along a dimension of the header 200 which is orthogonal to the longitudinal dimension of the one or more grooves 312 so that fluid entering flow channel 60 from the one or more second fluid inlets 332 flows toward the one or more second fluid outlets 342 in the second fluid direction B. In the depicted embodiments, fluid flowing from second inletport 330 towards flow channel 60 is distributed to a plurality of second fluid inlets 332, preferably spread uniformly across the full length of flow channel 60 (i.e. along the dimension of the header which is orthogonal to the longitudinal dimension of the one or more grooves 312). The flow is split to these second fluid inlets via multiple stages of flow bifurcation, thus ensuring that a similar flow rate of fluid flows through each second fluid inlet 332 into the flow channel. Similarly, the fluid flowing from the flow channel through each second fluid outlet 342 is recombined via a flow path comprising multiple stages of flow convergence before reaching second outlet port 340.

[0170] Also shown in Figure 4 are structural posts 250 which, in some embodiments, are provided in the underside of header layer 200. Structural posts 250 are configured to rest in abutment with the hydrophobic upper surface 310 of flow layer 300, on the substrate surrounding and spanning between grooves 312, to impart structural support, and maintain patency and a substantially consistent channel height of flow channel 60.

[0171] Figure 5 is an exploded view of a multi-layer microfluidic device 100 according to another embodiment, incorporating header layer 200, flow layer 300 and base layer 400. Although not visible, the features of header 200 correspond to the features described in relation to Figure 4. Block 312 in flow layer 300 represents the elongate parallel grooves extending in direction B. Base layer 400 includes gas inlet 430 and gas outlet 440 which feed gas flow path 410.

[0172] Additionally, a number of through bores are shown. In the embodiment shown, through bores 381 , 382 provide a vertical channel for delivery of culture media and reagents, respectively, to first inlet port 210. Through bores 383 and 384 provide a vertical channel for delivery of cells and extraction fluid, respectively, to second inlet port 330. Through bores 391 and 392 provide a vertical channel for extraction of cells from second outlet port 340 and removal of expired media from first outlet port 220, respectively. Through bores 483 and 481 provide a vertical channel for flowing gas to gas inlet 430 and removing gas from gas outlet 440, respectively. The use of through bores extending through all three layers of the device for delivery and extraction of fluids and gases is particularly useful in embodiments where a plurality of multi-layer microfluidic devices 100 are arranged in a stack in a bioreactor cell culture system, as will be described hereafter.

[0173] Also seen in Figure 5 are structural posts 450 in base layer 400 which are configured to rest in abutment with the lower surface 321 of flow layer 300 to impart structural support and ensure patency of the gas flow path 410 during use. Additionally, a serpentine wall 460 directs gas flow in a serpentine path from gas inlet port 430 to gas outlet port 440. However, it is to be understood that wall 460 may be provided in wall portions or sections, or need not be provided at all with the gas simply flowing from gas inlet port 430 to gas outlet port 440. In some embodiments the structural posts 450 may alternatively be omitted where the bioreactor dimensions and / or materials are such that the fluid flow pathways will not collapse in use.

[0174] A multi-layer microfluidic device 100 according to another embodiment is depicted in Figures 6, 7, 8 and 9. Figure 6 is an exploded view of the multi-layer microfluidic device incorporating header layer 200, gas permeable flow layer 300 and base layer 400. Figure 7 displays the features present on lower surface 221 (underside) of header layer 200, when viewed from above through the header layer.

[0175] Flow layer 300 comprises a hydrophobic upper surface 310 and lower surface 320, and the upper surface comprises an array of parallel grooves (represented by block 312) extending in direction B, for retention of cells. In an exemplary embodiment, the groove array has the following characteristics: groove depth of 150 pm, groove width of 150 pm, groove pitch of 250 pm, groove spacing thus 100 pm, groove length of 54 mm, number of grooves is 238. In another exemplary embodiment, the groove array has the following characteristics: groove depth of 150 pm, groove width of 150 pm, groove pitch of 250 pm, groove spacing thus 100 pm, groove length of 54 mm, number of grooves is 302. Flow layer 300 is preferably formed of a gas permeable hydrophobic polymer, which in some preferred embodiments is polymethylpentene (TPX) as will be further described in a subsequent section. Upper surface 310, including in the grooves, is thus hydrophobic due to the non-polar properties of the gas permeable hydrophobic polymer.

[0176] Base layer 400 includes gas inlet 430 and gas outlet 440 which feed gas flow path 410. Respiratory gas feed entering via gas inlet 430 is thus directed to flow along gas flow path 410 under flow channel 60 by wall 460.

[0177] As seen in Figure 7, header layer 200 includes a first inlet port 210 which feeds via a plurality of first fluid inlets 212 to flow channel 60, which is formed between the lower surface 221 of header 200 and the hydrophobic upper surface 310 of flow layer 300. Thus, grooves 312 form the floor of the flow channel 60 in which cells can deposit and grow. Header layer 200 also includes a first outlet port 220 which receives flow from flow channel 60 via a plurality of first fluid outlets 222.

[0178] First fluid inlets 212 and first fluid outlets 222 are positioned in the header layer such that fluid entering the flow channel 60 moves in a first flow direction A from the first fluid inlets toward the first fluid outlets such that flow is directed across the grooves 312. The flow in direction A, or “cross flow”, is typically orthogonal (90°) to the longitudinal dimension of grooves 312 although it is also contemplated that cross flow need not be precisely orthogonal. In an embodiment, the angle of cross flow is in the range of approximately 75° to 90° relative to the longitudinal dimension of the grooves, preferably approximately 80° to 90°, more preferably approximately 85° to 90°. In an embodiment, the cross flow is approximately 90° relative to the longitudinal dimension of the grooves.

[0179] First fluid inlets 212 are preferably distributed substantially uniformly across the full width (in direction B) of flow channel 60. Fluid flowing from first inlet port 220 towards flow channel 60 is thus distributed substantially uniformly across the full width of flow channel 60. The flow is split to the first fluid inlets via one or more, typically four or more, stages of flow bifurcation, thus ensuring that a similar flow rate of fluid is emitted through each first fluid inlet 212 into the flow channel. In the depicted embodiment, flow from inlet port 210 is distributed via five stages of bifurcation to 32 first fluid inlets 212.

[0180] Similarly, the flow at the opposite end of flow channel 60 is received at a plurality of first fluid outlets 222 preferably distributed substantially uniformly across the full width of the flow channel and recombined via a flow path comprising one or more, typically four or more, stages of flow convergence (i.e. reverse to bifurcation) before reaching first outlet port 220. In the depicted embodiment, fluid flowing from flow channel 60 flows through 32 first fluid outlets and is recombined via five stages of flow convergence before reaching first fluid outlet 220.

[0181] In this manner, the cross flow of fluid through the flow channel is uniformly directed in direction A and the flow stress and Reynolds number experienced in the grooves, at each point along the full length of the grooves, will be substantially consistent.

[0182] Header layer 200 also includes a second inlet port 330 which feeds a plurality of second fluid inlets 332 to flow channel 60, and a second outlet port 340 which receives flow from flow channel 60 via a plurality of second fluid outlets 342.

[0183] Second fluid inlets 332 and second fluid outlets 342 are positioned in the header layer such that fluid entering the flow channel 60 moves in a second flow direction B from the second fluid inlets toward the second fluid outlets such that flow is directed along grooves 312. The flow in direction B, or “longitudinal flow”, is typically closely aligned with the longitudinal dimension of the grooves, thus allowing fluid to flow with minimal resistance along the grooves in the flow channel. However, it will be appreciated that some variation from perfect alignment may be accommodated while still allowing the flow in direction B to carry cells along the grooves during cell capture and cell harvesting.

[0184] The flow in direction B is typically orthogonal (90°) to flow in direction A, although it is also contemplated that the two flow directions need not be precisely orthogonal. In an embodiment, the angle between the two flows is in the range of approximately 75° to 90°, preferably approximately 80° to 90°, more preferably approximately 85° to 90°. In an embodiment, the angle between the two flows is approximately 90°.

[0185] Second fluid inlets 332 are preferably distributed substantially uniformly across the full width (in direction A) of flow channel 60. Fluid flowing from second inlet port 330 towards flow channel 60 is thus distributed substantially uniformly across the full width of flow channel 60. The flow is split to the second fluid inlets via one or more, typically four or more, stages of flow bifurcation, thus ensuring that a similar flow rate of fluid is emitted through each second fluid inlet 332 into the flow channel. In the depicted embodiment, flow from inlet port 210 is distributed via five stages of bifurcation to 32 second fluid inlets 332.

[0186] Similarly, the flow at the opposite end of flow channel 60 is received at a plurality of second fluid outlets 342, preferably distributed substantially uniformly across the full width of the flow channel (in direction A) and recombined via a flow path comprising one or more, typically four or more, stages of flow convergence (i.e. reverse to bifurcation) before reaching second outlet port 340. In the depicted embodiment, fluid flowing from flow channel 60 flows through 32 second fluid outlets and is recombined via five stages of flow convergence before reaching second fluid outlet 340.

[0187] In this manner, the longitudinal flow of fluid through the flow channel is uniformly directed in direction B and the flow rate and flow stress experienced along each groove 312 will be substantially consistent.

[0188] As seen in Figure 6, a number of through bores are provided to fluidically couple the microfluidic device 100 to external sources of fluid inputs and to receptacles for fluid outputs. Through bores 381 , 382 provide a vertical channel for delivery of culture media and reagents (e.g. exogenous molecules for transfection / transduction of the captured cells), respectively, to first inlet port 210. Through bores 383 and 384 provide a vertical channel for delivery of cells and extraction fluid, respectively, to second inlet port 330. Through bores 391 and 392 provide a vertical channel for extraction of cells from second outlet port 340 and removal of expired media from first outlet port 220, respectively. Through bores 483 and 481 provide a vertical channel for flowing gas to gas inlet 430 and removing gas from gas outlet 440, respectively.

[0189] The header also includes filters 351 for removing undesirably large particles from fluid flowing towards the first and second fluid inlets. Figure 44 depicts a magnified view of the rectangular region marked D in Figure 7, showing a filter 351 for removing particles from fluid flowing from through bore 383 towards second fluid inlets 332. The filter comprises a filter channel 740 with spaced apart posts 742 arranged in three rows extending across the filter channel. The spacing between posts in each row decreases in the direction of flow through the filter. The filter is thus configured to remove large particles, such as clumps of cells or other debris, from a fluid sample containing target cells to be captured, before these particles enter the flow bifurcation pathway where they may block or occlude the flow path.

[0190] Figure 8 displays the features present on both lower surface 221 (underside) of header layer 200 seen in Figure 7 and the features present on hydrophobic upper surface 310 of flow layer 300, when viewed from above the microfluidic device 100. The relative position of groove array 312 (present on flow layer 300, with grooves extending in direction B) and the flow features of header layer 200 can thus be seen. Figure 9 displays a magnified view of the rectangular region marked C in Figure 8.

[0191] As best seen in Figure 9, each second fluid inlet 332 includes a diverging flow guide 309 so that flows of fluid entering flow channel 60 are progressively spread across an increasing flow width (orthogonal to the groove orientation) when passing though the diverging flow guide. The diverging flow guides 309 of neighbouring second fluid inlets 332 meet in the flow channel at junctions 331 . Flows of fluid entering the flow channel from the neighbouring second fluid inlets are thus spread until they are substantially adjacent (in a direction orthogonal to the groove orientation) when merged after flowing past junctions 331. This configuration advantageously reduces or eliminates the “dead zones”, or regions of low flow stress, which can form between neighbouring second fluid inlets 332 which lack diverging flow guides (for example, as seen in Figure 4).

[0192] The configuration of the groove array, comprising parallel grooves 312 extending in direction B, is also clearly seen in Figure 9. The nearest end of each groove 312 is set back from the second fluid inlets 332, with an ungrooved, planar portion 313 of the flow layer hydrophobic upper surface 310 thus lying between the closest part of the second fluid inlets 332 (junctions 331 ) and the nearest end of grooves 312. The separation across the ungrooved portion (length in direction B, as indicated by arrow E) may be at least 1.5 mm, or at least 2 mm. This arrangement advantageously ensures that the flow conditions experienced in each groove 312 is not affected by any localised turbulence or dead zones caused by the merging of flows from the second fluid inlets.

[0193] The inventors have found that the diverging flow guides and the set back of grooves away from second fluid inlets ensures that a favourable, uniform distribution of captured target cells is obtained in flow chamber 60 when cells are loaded along the grooves. The undesirable accumulation of cells in “dead zones” between the secondfluid inlets is mitigated, and acceptably similar numbers of cells are captured in each groove due to the uniform flow stress experienced in the grooves.

[0194] Also shown in Figures 7, 8 and 9 are structural posts 250 which are provided in the underside of header layer 200 and configured to rest in abutment with, and preferably to bond with, the hydrophobic upper surface 310 of flow layer 300 in the assembled device. The structural posts are preferably located on the header layer for maximum contact with the hydrophobic upper surface, and may thus be centred on the elevated regions between the grooves 312. The structural posts thus maintain a desired separation between the header layer 200 and flow layer so that the flow channel remains open and of substantially consistent height above the groove array.

[0195] As best seen in Figure 9, the structural posts 250 have an elongated crosssection, with an axis of elongation substantially aligned with grooves 250. This configuration reduces the profile of the structural posts presented to fluid flowing through flow channel 60 in in flow direction B along the grooves, while ensuring that the structural posts still retain a sufficient cross-sectional footprint to space apart the layers and preferably to resiliently bond to the flow layer. The elongated cross-section of the structural posts is preferably configured to streamline fluid flow in flow direction B, thus reducing or avoiding the formation of “dead zones” (regions of low flow stress) behind the structural posts. The elongated cross-sectional shape may thus be rounded, for example substantially elliptical as depicted.

[0196] The inventors have found that microfluidic devices with structural posts configured in this manner are less susceptible to undesired accumulation and retention of cells in dead zones behind the posts when flowing cells for capture or harvesting in flow direction B.

[0197] Microfluidic devices require very precise control of flow, particularly for cell deposition and cell separation processes. Pneumatically controlled flow is preferable to syringe pumps or roller pumps since flow rate depends on the hydraulic resistance of flow resistors rather than the compressibility of the fluid. Thus, in some embodiments flow resistors are incorporated into the fluid pathways to govern the flow rates across the flow channel in flow direction A and / or flow direction B.

[0198] As seen in Figure 4, for example, flow resistors 371 and 372 regulate the rate of flow of fluids (e.g. culture medium and reagent respectively) from through bores 381 and 382 into first inlet port 210. Flow resistor 373, which provides approximately equivalent resistance to flow resistor 371 owing to its length, regulates the rate of flow (e.g. of sample containing target cells, and extraction medium respectively) from through bores 383 and 384 into second inlet port 330. Flow resistors 371 , 372 and 373 are thus located up-stream of flow channel 60.

[0199] In some preferred embodiments, however, one or more flow resistors is located downstream of the flow channel, and in particular downstream of the second fluid outlets so that the downstream flow resistor controls the flow rate of fluid through flow channel 60 in direction B, along the grooves. Figure 10 depicts a header layer according to such an embodiment, which is identical to the header depicted in Figure 7 except that a serpentine flow resistor 374 is present in the flow path downstream of second outlet port 340.

[0200] Surprising, the inventors found that a downstream location of the flow resistor provided an improved result when flowing a fluid containing cells to be captured from second inlet port 330 through flow channel 60 to second outlet port 340. When an upstream flow resistor was used, the distribution of captured cells to the grooves in flow channel 60 was uneven (even though dye tests confirmed that liquid-only flow was evenly distributed to the grooves in the same apparatus). Without limitation by theory, it is proposed that the poor cell distribution was caused by a hydrodynamic force called inertial migration that pushes cells away from the walls of microfluidic channels in regions of high fluid velocity, such as in the flow resistor. By placing the flow resistor downstream, the undesired effects of inertial migration on the distribution of cells sent to second fluid inlets 332 were mitigated.

[0201] One consequence of placing the flow resistor downstream is that flow channel 60 is exposed to the pneumatic pressure applied to the fluid entering second inlet port 330 to control the flow rate thereof. Typical operating pressures may be between 0.01 and 1 bar gauge. By using a rigid thermoplastic, such as polymethylpentene, as the material of construction, and adopting design features as disclosed herein with reference to Figures 6-10, the inventors were able to fabricate amicrofluidic device capable of withstanding pneumatic pressures of at least 1 bar gauge without failing or bulging.

[0202] In the various microfluidic devices disclosed herein, and particularly those described herein with reference to Figures 4-10, the flow layer 300 may be configured, according to application-specific requirements, to provide a plurality of grooves 312 with dimensions suitable to culture cells therein and to allow sufficient perfusion of gases through the flow layer to exchange with the cells. In some embodiments, the flow layer has a maximum thickness, beneath the flow channel, of no more than 400 pm, or no more than 350 pm, 325 pm, 300 pm, 275 pm or 250 pm. As used herein, the maximum thickness refers to the thickness of the flow layer in the regions between the grooves. The maximum thickness may be at least 90 pm, or at least 100 pm, 125 pm, or 150 pm Preferably the maximum thickness is between 100 and 300 pm, such as between 125 pm and 275 pm, or between 150 and 250 pm.

[0203] In some embodiments, the flow layer has a thickness within the grooves, i.e. between the bottom surfaces of the grooves and the lower surface of the flow layer, of less than 100 pm, or less than 80 pm, preferably less than 65 pm, such as between 25 and 65 pm, or between 30 and 60 pm. This may facilitate sufficient gas permeation through the flow layer to support cell viability and growth even when substantial concentrations of cells are captured in the grooves.

[0204] In some embodiments, the depth of the grooves in the flow layer is between 50 and 300 pm, such as between 100 and 200 pm, or between 110 and 180 pm, or between 130 pm and 165 pm, for example about 150 pm. In some embodiments, the grooves have an aspect ratio (i.e. width:depth ratio) of between 1.5:1 and 1 :1.5, or between 1.2:1 and 1 :1.2, such as about 1 :1 . Groove dimensions in this range have been found to avoid or mitigate loss of cells from the grooves during expansion under a cross flow of cell growth medium sufficient to sustain the cells.

[0205] Another embodiment of the disclosure is presented in the schematic illustration of Figures 1 1 and 12 which further comprises (in comparison to the embodiment of Figure 1 ) a selectively permeable membrane 500 which functionally separates flow channel 60 into two parallel flow channels across which low molecular weight metabolites can exchange. In some embodiments, selectively permeablemembrane 500 may be manufactured from a dialysis membrane or similar membrane material. Suitable membrane materials include but are not limited to regenerated cellulose, polysulfone and polymer blends including polyvinylpyrrolidone (PVP). In one embodiment, selectively permeable membrane 500 comprises a dialysis membrane having pore size suitable for exchanging molecules having a molecular weight of less than 20 kD, more preferably less than 15 kD and more preferably still, less than 10 kD. In this arrangement, the selectively permeable membrane 500 separates secondary header 700 and header 200, forming a second flow channel 70. Cells enter flow channel 60 via first inlet port 210 and deposit into grooves 312. Fluid, typically base medium containing nutrients with molecular weight less than 10kD, enters second flow channel 70 via secondary inlet port 730 and exits via secondary outlet port 740, where molecules less than 10 kD diffuse across membrane 500 entering flow channel 60 for nutrient delivery to cells in grooves 312. Metabolic waste molecules generated by cells in grooves 312 diffuse across membrane 500, enter flow channel 70 and exit via secondary outlet port 740. Meanwhile, growth medium enters flow channel 60 via first inlet port 210. Waste product is removed from flow channel 60 via first outlet port 220. As in the embodiment of Figure 1 , Figure 6 provides for cross flow of cells and media in direction A. Through bore 381 provides a vertical channel for delivery of fluids to first inlet port 210; through bore 391 provides a vertical channel for removal of fluids from first inlet port 220. Through bores 481 , 483 provide vertical channels for delivery and removal of gas. Figure 12 provides a sectional view of the device in Figure 1 1 .

[0206] A benefit associated with provision of selectively permeable membrane 500 is that delivery of fluid into second flow channel 70 can be controlled separately from delivery of fluid into flow channel 60 allowing for differential flow rates in channel 70 and channel 60 e.g. higher flow rates for base medium than growth medium. This can reduce quantities of growth media required, resulting in cost savings.

[0207] Notably, Figures 1 , 3, 5, 6 and 1 1 show gas input channel 481 and gas output channel 483 as through bores forming a vertical channel through header layer 200 and flow layer 300 for delivery of respiratory gas feed to gas flow path 410 and removal of expired gases. In some embodiments, through bores are provided forming vertical channels for delivery of fluid to each of the inlet ports, and removal of fluids from each of the outlet ports in the microfluidic device, each of the through boresextending colinearly through all three layers of the microfluidic device 100, and each fluid outlet is in fluid communication with a corresponding channel.

[0208] In some embodiments, the multi-layer microfluidic device further comprises one or more alignment features such as a tongue or protrusion (not shown) on the top layer (header 200 or secondary header 700) which cooperates with a corresponding groove or notch in the bottom surface of the base layer 400 (or vice versa).

[0209] Figures 13 and 14 show a plurality of multi-layer microfluidic devices 100 arranged in a stack 600 in exploded and assembled views respectively. The stack includes a substantially rigid base plate 610 and a substantially rigid cover plate 620 to impart structural support to the stack. In preferred embodiments, guide features such as posts (not shown) extend through colinear post channels (not shown) provided through the plurality of microfluidic devices to ensure alignment and flow patency in through bores forming fluid input channels 381 , 382, 383, 384, fluid output channels 391 , 392, gas input channel 481 and gas output channel 483 of the individual devices. Figures 13 and 14 illustrate a stack of twenty multi-layer microfluidic devices 100 with the 3-layered design shown in Figure 5. Layers 200, 300 and 400 from a single multilayer microfluidic bioreactor are shown on the exploded view (Figure 13). Advantageously, all twenty microfluidic devices are supplied by a common fluid source for each fluid required for the cell culture process. It is to be understood, however, that a stack of microfluidic devices could comprise a single microfluidic device, or e.g. 5, 10, 20, 30, 50, 60, 70, 80, 90, 100 or 100’s of microfluidic devices.

[0210] The multi-layer microfluidic devices 100, comprising header layer 200, gas permeable flow layer 300 and optional base layer 400, are depicted in Figures 1 -12 as individually fabricated devices which are not physically connected to other multi-layer microfluidic devices 100 (at least until post-fabrication assembly in a stack). However, it will be appreciated that other arrangements and methods of fabrication are possible. For example, a stack of multi-layer microfluidic devices 100 may be fabricated by bonding together multiple layers 200, 300 and 400, arranged in sequence, in one or more bonding steps.

[0211] In another example, each functional layer 200, 300 and 400 is initially manufactured on a structured sheet comprising multiple similar functional layers. Forexample, structured sheets may be formed comprising four individual header layers 200, as depicted in Figure 15. Similarly, structured sheets may be formed comprising four individual flow layers 300, and four individual base layers 400. When assembled by aligning and bonding the three structured sheets together, four physically connected multi-layer microfluidic devices 100 are arranged in parallel in a single composite sheet structure. Multiple multi-layered microfluidic devices may be arranged on sheets according to layouts and sheet sizes as may be determined by manufacturers or for optimised design or consumption of materials. The sheets may be cut to form individual multi-layer bioreactors, or the multi-layer sheets themselves may be stacked to form a sheet stack comprising a plurality of multi-layer microfluidic devices arranged both side- to-side and vertically stacked.Bioreactor for cell culture

[0212] In an aspect, the invention relates to a bioreactor for cell culture. The bioreactor comprises a multi-layer microfluidic device and a cell capture composition.

[0213] The multi-layer microfluidic device comprises a header layer and a gas permeable flow layer beneath the header layer. The flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells. The header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises: (i) a first inlet port; (ii) a first outlet port; (iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and (iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port. The one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein.

[0214] The multi-layer microfluidic device may generally be according to any of the embodiments disclosed in the preceding section.

[0215] The cell capture composition is physically adhered to the hydrophobic upper surface in the one or more grooves, and is thus located inside the flow channel of themicrofluidic device. The cell capture composition comprises at least (i) an antigen binding moiety that specifically binds to a target cell and optionally (ii) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments.

[0216] The cell capture composition is physically adhered to the hydrophobic upper surface, meaning that it is not chemically bonded to the surface by covalent bonds. Rather, the cell capture composition is retained on the surface by physical forces of attraction, i.e. van der Waals forces. Without limitation by theory, it is proposed that the hydrophobe of the amphiphilic polymer physically adsorbs to the hydrophobic upper surface due to the common non-polar character, while the one or more hydrophilic polymeric segments are repelled by and extend away from the hydrophobic upper surface due to the dissimilar polarity. The amphiphilic polymer likely forms a monolayer on the hydrophobic upper surface. In this way, the surface-active amphiphilic polymer modifies the physical and chemical properties of the hydrophobic upper surface, blocking non-specific attachment of cells to the surface and retaining the antigen binding moiety on the surface for specific attachment of target cells. Advantageously, this approach does not rely on complex surface chemistry to modify the surface.

[0217] The cell capture composition is present at least in the grooves of the flow layer but is typically present across the entire flow channel footprint on the hydrophobic upper surface. Preferably, the cell capture composition is physically adhered in each of the grooves, and may be present as a continuous coating over the hydrophobic upper surface including in the grooves and regions intermediate to the grooves. The physically adhered cell capture composition may be produced by a process comprising a step of flowing a fluid, typically an aqueous fluid, comprising the amphiphilic polymer into the flow channel and incubating the fluid therein to allow the amphiphilic polymer to deposit on the flow chamber walls. As used herein, the flow chamber walls refer to all interior surfaces of the flow chamber, including the grooved hydrophobic upper surface of the flow layer and the lower surface of the header layer. The amphiphilic polymer is thus preferably water-dispersible to allow an aqueous fluid composition comprising the polymer to be prepared. The fluid comprising amphiphilic polymer is preferably flowed across the grooves. In embodiments where the header layer comprises a second inlet port and a second outlet port to provide a longitudinal flow path (along the grooves), cross-flow treatment of the grooves avoids depositing the cellcapture composition in the longitudinal flow path before and after the grooves, where capture of cells is undesired.

[0218] A wide range of amphiphilic polymers comprising a hydrophobe and one or more hydrophilic polymeric segments are suitable provided that they are surface-active and include a hydrophobe capable of adsorbing to a hydrophobic surface, preferably a hydrophobic polymer surface, sufficiently strongly to physically adhere the polymer to the surface. Non-limiting examples of suitable amphiphilic polymers include non-ionic surfactant polymers (e.g. poloxamers), amphiphilic copolymers comprising one or more hydrophobic segments (e.g. hydrophobic polymerised blocks) and one or more hydrophilic zwitterionic polymerised blocks, and lipid bilayer polymers (e.g. comprising a phospholipid head and fatty aliphatic tails). In some embodiments, the amphiphilic polymer is a non-ionic surfactant polymer, such as a poloxamer.

[0219] The amphiphilic polymer may be an amphiphilic block copolymer. Thus, in some embodiments, the hydrophobe of the amphiphilic polymer is a hydrophobic polymeric segment, for example a hydrophobic polyalkylene glycol segment, i.e. a polyalkylene glycol segment which comprises or consists of C3+ alkylene groups such as C3-C6 alkylene groups. In some preferred embodiments, the hydrophobic polymeric segment is a polypropylene glycol (PPG) segment, i.e. it has the formula -(CH2-CHCH3- O-)n where n is the number of propylene glycol repeat units in the segment. The value of n may be greater than 30, or greater than 50. For example, n may be in the range of 30 to 100, or 50 to 80. Amphiphilic polymers comprising hydrophobic PPG chain segments in such ranges, together with one or more suitable hydrophilic chain segments, may be dispersible in aqueous solutions yet able to physically adhere to and thus surface-modify hydrophobic polymer surfaces.

[0220] One role of the hydrophilic polymeric segment is to block non-specific attachment of cells to hydrophobic surfaces modified by the amphiphilic polymer. The hydrophilic polymeric segment may be able to bind water. In some embodiments, the one or more hydrophilic polymeric segments are polyethylene glycol (PEG) segment(s), i.e. they have the formula -(CH2-CH2-O-)m where m is the number of ethylene glycol repeat units in the segment. The value of m for each PEG segment may independently be greater than 50, or greater than 70. For example, each m may independently be in the range of 50 to 150, or 70 to 130. Amphiphilic polymers comprising hydrophilic PEGchain segments in such ranges are expected to be dispersible in aqueous solutions, despite the presence of the hydrophobe capable to physically adhere to and thus surface-modify hydrophobic polymer surfaces. Moreover, the hydrophilic modification of the surface achieved with such PEG chain segments is considered suitable to block non-specific cell binding.

[0221] In some embodiments, the amphiphilic polymer is a triblock copolymer comprising a hydrophobic PPG segment located between two hydrophilic PEG segments. Such amphiphilic polymers, also known as poloxamers, are considered particularly suitable as they adopt a “polymer brush” configuration when physically adhered to hydrophobic polymer surfaces, with the intermediate PPG segment adsorbing to the polymer surface and the two terminal PEG segments extending away from the surface. Non-limiting examples of a suitable, commercially available poloxamer are Pluronic F-127 or Pluronic F-108 from BASF.

[0222] In some embodiments, the amphiphilic polymer, for example a poloxamer, has a molecular weight in the range of 10,000 to 20,000 g / mol, for example 10,000 to 15,000 g / mol. In some embodiments, the amphiphilic polymer, for example a poloxamer, has a molecular weight of approximately 12,500 g / mol or approximately 14,600 g / mol.Polymer-antiqen binding moiety conjugates

[0223] In some embodiments, the cell capture composition comprises an antigen binding moiety conjugated to at least one hydrophilic polymeric segment of an amphiphilic polymer. Specifically, the cell capture composition comprises a polymer- antigen binding moiety conjugate, in which the antigen binding moiety is chemically bonded to the amphiphilic polymer by at least one covalent bond. The antigen binding moiety specifically binds to a target cell, and the amphiphilic polymer comprises a hydrophobe and one or more hydrophilic polymeric segments, as previously discussed herein.

[0224] The inventors have found that physical adsorption of polymer-antibody conjugates to hydrophobic surfaces allows target cells to be immobilised on the surface with high efficiency, as measured by the percentage of target cells captured while flowing over the surface, and with high selectivity, due to the inhibition of negative cellnon-specific binding to the surface. In some embodiments, improved capture of target cells has been observed in comparison to physical mixtures of the unconjugated antibody and free amphiphilic polymer.

[0225] Where the amphiphilic polymer comprises two or more hydrophilic polymeric segments, antigen binding moieties may be conjugated to one, or more than one, or each, of the hydrophilic polymeric segments. Thus, for example, an antigen binding moiety may be conjugated to one or both PEG segments of a poloxamer. Preferably, the antigen binding moiety is conjugated to the chain terminus of the hydrophilic polymeric segment(s).

[0226] The antigen binding moiety may be conjugated to the hydrophilic polymeric segment by reaction with a conjugating functional group on the hydrophilic polymeric segment of the precursor amphiphilic polymer, preferably located at the chain terminus. It will be appreciated that the antigen binding moiety and hydrophilic polymeric segment are thus linked in the resultant polymer-antibody conjugate by a linking functionality which is the residue of the conjugating reaction.

[0227] Preferably, the conjugating functional groups in the precursor polymer are configured for facile and substantially irreversible conjugation reactions at ambient reaction conditions and without the need for chemical or photochemical activation.

[0228] In some embodiments, the conjugating functional groups are configured for conjugation to a native protein. This means that the conjugating functional groups are capable of reacting with moieties naturally present on proteins (such as the antibody), as distinguished from artificially introduced, conjugation-receptive functional groups which have been pre-labelled onto a protein. For example, the conjugating functional groups may be selected to target amine or thiol moieties naturally present on the antigen binding moiety. A wide range of conjugating functional groups are reportedly suitable for this purpose, and the hydrophilic chain segment prior to conjugation may incorporate any such groups according to the principles disclosed herein. In some embodiments, the conjugating groups are amine-reactive functional groups. In some embodiments, the conjugating functional groups configured for conjugation to a native protein, such as an antigen binding protein, are selected from the group consisting of activated esters, epoxides, isocyanates, isothiocyanates, sulfonyl chlorides,maleimides, anhydrides, chloroformates, aldehydes, ketones, carbodiimides and imidoesters.

[0229] In some embodiments, the conjugating functional groups in the precursor polymer are selected from activated esters (also known as active esters) and epoxides. As used herein, an activated ester is an ester functional group, including a carbonate ester functional group, which is highly susceptible to nucleophilic attack as a result of electronegative substituents. In particular, activated esters provide enhanced rates of reaction with amines to form amide linkages. Non-limiting examples of suitable activated esters may include N-hydroxy succinimide esters (NHS-esters), including sulfo-NHS esters (produced from N-hydroxysulfosuccinimide), fluorophenyl esters, and stabilised triazole esters (e.g. produced from HATU = 1 -[bis(dimethylamino)methylene]-1 H-1 ,2,3-triazolo[4,5-b]pyridinium 3-oxide hexafluorophosphate) and HBTU = 3-[bis(dimethylamino)methyliumyl]-3 / - / - benzotriazol-1 -oxide hexafluorophosphate). Epoxide groups are susceptible to ring opening conjugation reactions with amines or thiols to form beta-hydroxy secondary amine or thioether linkages respectively. In some embodiments, the conjugating functional groups are selected from N-hydroxy succinimide esters (NHS-esters) and terminal epoxides such as glycidyl groups.

[0230] When the conjugating functional groups of the precursor polymer are configured for conjugation to an antigen binding protein, the antigen binding moiety in the resultant polymer-antigen binding moiety conjugate may be conjugated to the hydrophobic polymeric segment by a linking functionality which is the residue of reaction between the conjugating functional group and an amine or thiol on the antigen binding moiety. For example, the antigen binding moiety may be conjugated to the hydrophilic polymeric segment by an amide linking group when the conjugation reaction involves reaction of an activated ester conjugating group on the amphiphilic polymer with an amine of the antigen binding moiety. In another approach, a Fc carbohydrate side chain of the antigen binding moiety may be conjugated to the hydrophilic polymeric segment via sialic acid residues, e.g. by reaction with hydrazide conjugates.

[0231] In other embodiments, the conjugating functional groups of the precursor amphiphilic polymer are configured for conjugation to a modified (synthetically labelled) protein, such as a modified antigen binding protein, via a click reaction. Click reactionsrefer generally to coupling reactions occurring between complementary pairs of functional groups which are thermodynamically predisposed to react with high selectivity, high conversion, a lack of complex by-products and under biological conditions. The resultant covalently bonded linking groups are chemically stable and biologically inert. Click chemistry is widely used for bioconjugation to biological macromolecules, for example to attach fluorophores or other reporter molecules. A particular advantage of click chemistry is that it allows selective bioconjugation of probe molecules to pre-functionalised target biological macromolecules in complex biological systems, also known as biorthogonal reactions. A wide range of click chemistries have been reported for bioconjugation reactions. These include [3+2] cycloaddition reactions, such as the azide-alkyne Huisgen cycloaddition between an azide functional group and an alkyne group to form a 1 ,2,3-triazole linker, inverse-demand Diels-Alder reactions between a tetrazine and a strained cycloalkene functional group, such as tetrazine and trans-cyclooctene (the tetrazene-TCO reaction), and the Staudinger ligation reaction between a tertiary phosphine and an azide functional group to form a phosphazo linker.

[0232] The conjugating functional groups on the precursor amphiphilic polymer, prior to the conjugation reaction, may thus comprise any of the functional groups known to participate in click reactions with a complementary functional group via these or other click reactions, according to the principles disclosed herein. In some embodiments, the conjugating functional groups are configured for conjugation to a modified protein (such as a modified antigen binding protein) via an alkyne-azide or a tetrazine-TCO click reaction.

[0233] Amphiphilic polymers having click conjugating groups on the hydrophilic polymeric segment(s) can be produced by a variety of synthetic methodologies known to those of skill in the art for labelling macromolecules with these groups. Hydroxy groups (-OH) present on the hydrophilic polymeric segments, for example at the terminus of a PEG segment, may thus be converted to click-reactive functional groups. For example, terminal hydroxy groups may be converted to azide groups (-N3) by tosylation (e.g. with tosyl chloride) and subsequent reaction with an azide salt (e.g. sodium azide). When the click-reactive functional groups are azide functional groups (-N3), the amphiphilic polymer is thus configured for copper-free alkyne-azide clickcoupling conjugation to an antigen binding moiety which has been labelled with the alkyne-based click group dibenzo-bicyclo-octyne (DBCO).

[0234] The antigen binding moiety may be labelled with a corresponding click functional group by methods known to those skilled in the art. Typically, the antigen binding moiety will be reacted with a click labelling reagent comprising an aminereactive conjugating group. For example, antigen binding moieties may be labelled with DBCO groups via commercially available reagents such as dibenzocyclooctyne- sulfo-A / -hydroxysuccinimidyl ester reagents.

[0235] In one set of exemplary embodiments, the polymer-antigen binding moiety conjugate is prepared according to Scheme 1. Thus, a poloxamer terminally functionalised with activated ester groups (NHS-carbonate ester, as depicted) conjugates with an antigen binding moiety (e.g. monoclonal antibody) by reaction with native amine groups thereon. The linking group between the polyethylene glycol chain segments and the antigen binding moiety is thus an amide group.Scheme 1.

[0236] In another set of exemplary embodiments, the polymer-antigen binding moiety conjugate is prepared according to Scheme 2. Thus, a poloxamer terminally functionalised with a click functional group (e.g. azide as depicted) conjugates with an antigen binding moiety (e.g. monoclonal antibody) by reaction with a complementary click functional group (e.g. DBCO as depicted) labelled thereon.Scheme 2.

[0237] In some embodiments, the cell capture composition comprises, in addition to the polymer-antigen binding moiety conjugate, an amphiphilic polymer which is not conjugated to an antigen binding moiety (an “unconjugated amphiphilic polymer” or “free amphiphilic polymer”). The unconjugated amphiphilic polymer may be according to any of the embodiments of amphiphilic polymer already disclosed herein, for example a triblock copolymer comprising a hydrophobic polypropylene glycol segment located between two hydrophilic polyethylene glycol segments (a poloxamer). The unconjugated amphiphilic polymer may be the same or a different type of amphiphilic polymer to that of the polymer-antibody conjugate. Typically, the unconjugated amphiphilic polymer will lack conjugating functional groups on the hydrophilic polymeric segment(s). For example, the unconjugated amphiphilic polymer may be a poloxamer with terminal hydroxy groups on the two polyethylene glycol segments. In some embodiments, polymer-antibody conjugates have been found effective in combination with free amphiphilic polymer across a wide range of ratios, providing the opportunity to minimise the amount of antibody reagent applied to the surface while utilising free amphiphilic polymer to block non-specific binding of non-target cells.

[0238] The unconjugated amphiphilic polymer may be deposited on the hydrophobic upper surface together with the polymer-antigen binding moiety conjugate,i.e. from the same fluid composition, or in a separate surface treatment before or after the polymer-antigen binding moiety conjugate.Unconjugated antigen binding moieties

[0239] In other embodiments, the cell capture composition comprises the amphiphilic polymer and the antigen binding moiety as chemically distinct components, i.e. the antigen binding moiety is not conjugated to the amphiphilic polymer. Typically, the unconjugated amphiphilic polymer will thus lack conjugating functional groups on the hydrophilic polymeric segment(s). For example, the unconjugated amphiphilic polymer may be a poloxamer with terminal hydroxy groups on the two PEG segments.

[0240] In some such embodiments, the antigen binding moiety is deposited on the grooved hydrophobic upper surface of the flow channel grooves in a first deposition step, and the unconjugated amphiphilic polymer is deposited over or around the antigen binding moiety in a second deposition step. Alternatively, the antigen binding moiety and amphiphilic composition are deposited on the grooved hydrophobic upper surface simultaneously, i.e. from the same fluid composition. In either case, the resultant cell capture composition comprising both unconjugated amphiphilic polymer and antigen binding moiety is physically adhered to the hydrophobic upper surface.

[0241] Non-conjugated combinations of antibody and poloxamer have been found effective in cell selection, particularly when (i) inherently high selectivity antibodies are used, (ii) the antibody is adsorbed to the hydrophobic upper surface before depositing the poloxamer thereon, and / or (iii) the ratio of antibody to poloxamer is controlled to avoid competitive displacement of antibody by the poloxamer.Antigen binding moiety

[0242] The cell capture composition comprises an antigen binding moiety that specifically binds to a target cell. The antigen binding moiety thus acts as the capture reagent in the cell capture composition by specifically binding to the target cells in use, thus immobilising the target cells in the grooves of the flow channel. As already described, the antigen binding moiety may be present in a polymer-antigen binding moiety conjugate, or it may be present as a chemically distinct component to the amphiphilic polymer.

[0243] The antigen binding moiety can be used to positively or negatively select cells as they flow over the cell capture composition based on the expression of certain surface antigens on the target cells, or the lack of expression of lineage-specific antigens on the target cells, respectively.

[0244] As used herein, the term “antigen binding moiety” refers to a polypeptide that specifically binds to an antigenic determinant. The antigen binding moiety may be recombinant or isolated.

[0245] The term "polypeptide" or "polypeptide chain" will be understood to mean a series of contiguous amino acids linked by peptide bonds.

[0246] As used herein, the term “antigenic determinant” is synonymous with “antigen” and “epitope,” and refers to a site (e.g. a contiguous stretch of amino acids or a conformational configuration made up of different regions of non-contiguous amino acids) on a polypeptide to which an antigen binding moiety binds, forming an antigen binding moiety-antigen complex.

[0247] As used herein, the term “specifically binds” refers to a specific antigen binding moiety’s preferential interaction with a given ligand on a cell over other ligands on the cell or components present in the sample. For example, a specific antigen binding moiety that specifically binds a given ligand, binds the given ligand, under suitable conditions, in an amount or a degree that is observable over that of any nonspecific interaction with other components in the sample. Suitable conditions are those that allow interaction between a given specific antigen binding agent and a given ligand. These conditions include pH, temperature, concentration, solvent, time of incubation, and the like, and may differ among given specific antigen binding moieties and ligand pairs, but may be readily determined by those skilled in the art.

[0248] The term "recombinant" shall be understood to mean the product of artificial genetic recombination.

[0249] The term “isolated” refers to a polypeptide that is not in its natural milieu. No particular level of purification is required. For example, an isolated polypeptide can be removed from its native or natural environment. Recombinantly produced polypeptides expressed in host cells are considered isolated for the purpose of the disclosure, as arenative or recombinant polypeptides which have been separated, fractionated, or partially or substantially purified by any suitable technique.

[0250] The antigen binding moiety may be an antibody. In some examples, the methods can be used to select or enrich CD34+ hematopoietic stem / progenitors by positive selection using an anti-CD34 antibody. In another example, the methods can be used to select or enrich T cells using anti-T-cell antibodies (e.g., CD2, CD3, CD4 / CD8, T-cell receptor ct / |3).

[0251] The captured cells can be released, when desired, from the antigen binding moiety (e.g. antibody) using a cell release agent. For example, the captured cells can be released from the antibody by chemical or enzymatic cleavage, such as the proteolytic enzyme chymopapain, or neuraminidase. Alternatively, the captured cells may be released from the antibody via competitive displacement, for example, the cell release agent may be a peptide which competitively binds to the antibody to displace the captured cells. Examples of such peptides, as cell release agents for the exemplified antibody 9C5, are disclosed in WO1995 / 034817 and W02001 / 090153. The cells may be subsequently expanded by culture methods disclosed herein and / or transduced / transfected by methods disclosed herein.

[0252] Particularly useful antibodies remain reactive after being adsorbed to a solid surface, and further, retain structural integrity when completely dried, and are reactive when rehydrated.

[0253] The term "antibody", as used herein, refers to intact immunoglobulin molecules as well as fragments of immunoglobulin molecules that are capable of binding to an epitope of an antigen.

[0254] Antibodies can be of any class (e.g., IgG, IgE, IgM, IgD, IgA, and IgY), or subclass (e.g., IgG 1 , lgG2, lgG3, lgG4, Ig A1 and lgA2). The classes are distinguished by the type of heavy chain they contain. IgG molecules comprise heavy chains known as y-chains; IgMs have p-chains; IgAs have a-chains; IgEs have s-chains; and IgDs have 5-chains.

[0255] Antibodies can be formed by the association of one heavy chain type (e.g., p, 5, y, a, or s) with one light chain type (e.g., A or K). The basic (monomer) unit of aconventional antibody is a four polypeptide unit consisting of two identical heavy chains and two identical light chains connected together by disulfide bonds, and light chains (HC and LC) also contain intramolecular disulfide bonds for stabilization. The light chains are shorter, with lower molecular weights than the heavy chains. The general shape of an antibody is a Y, with a flexible hinge (interdomain) region at the centre of the Y. Each polypeptide chain has a constant region and a variable region.

[0256] The common notation for the light chain variable region is VL and for the light chain constant region is CL. The notation is similar for the heavy chain variable (VH) and constant regions (CH) with CH1 , CH2, and CH3 denoting the different constant region domains of the heavy chain. The fragment crystallizable (Fc) region contains only constant regions from the heavy chains (CH), but the fragment antigenbinding region (Fab) includes both a constant domain and the variable domains of both the heavy and light chains (VH and CL). The fragment variable region FV region contains only the two variable domains.

[0257] As used herein, “variable region” refers to the portions of the light and / or heavy chains of an antibody that specifically binds to an antigen and, for example, includes amino acid sequences of CDRs; i.e., CDR1 , CDR2, and CDR3, and framework regions (FRs). For example, the variable region comprises three or four FRs (e.g., FR1 , FR2, FR3 and optionally FR4) together with three CDRs. VH refers to the variable region of the heavy chain. VL refers to the variable region of the light chain.

[0258] As used herein, the term "complementarity determining regions” (syn. CDRs; i.e., CDR1 , CDR2, and CDR3) refers to the amino acid residues of an antibody variable region the presence of which contribute to specific antigen binding. Each variable region typically has three CDR regions identified as CDR1 , CDR2 and CDR3. In the case of a protein derived from an IgNAR, the protein may lack a CDR2. In one example, the amino acid positions assigned to CDRs and FRs are defined according to Kabat Sequences of Proteins of Immunological Interest, National Institutes of Health, Bethesda, Md., 1987 and 1991. For example, according to the numbering system of Kabat, VH FRs and CDRs are positioned as follows: residues 1 -30 (FR1 ), 31 -35 (CDR1 ), 36-49 (FR2), 50-65 (CDR2), 66-94 (FR3), 95-102 (CDR3) and 103- 1 13 (FR4). According to the numbering system of Kabat, VL FRs and CDRs are positioned as follows: residues 1 -23 (FR1 ), 24-34 (CDR1 ), 35-49 (FR2), 50-56 (CDR2), 57-88 (FR3),89-97 (CDR3) and 98-107 (FR4). The present disclosure is not limited to FRs and CDRs as defined by the Kabat numbering system, but includes all numbering systems, including those discussed above. In one embodiment, reference herein to a CDR (or a FR) is in respect of those regions according to the Kabat numbering system.

[0259] "Framework regions" (FRs) are those variable region residues other than the CDR residues.

[0260] The term "intact antibody” as used herein refers to whole or full-length antibodies that comprise heavy and light chains including a Fc region. The constant domains may be wild-type sequence constant domains (e.g., human wild-type sequence constant domains) or amino acid sequence variants thereof.

[0261] Antibodies can differ in valency. Some immunoglobulins can form multimers through linkage of their Fc domains via a J chain.

[0262] Antibodies can be polyclonal, monoclonal, recombinant.

[0263] The term “polyclonal antibody” as used herein refers to a mixed population of antibodies directed against different determinants (epitopes). The modifier "polyclonal" indicates the character of the antibody population and is not to be construed as requiring production of the antibody by any particular method. For example, polyclonal antibodies to be used in accordance with the present invention may be made introducing the whole pathogen or an isolated antigen by inoculation or infection into a host (e.g., goat) to induce the host to make antibodies against the pathogen or antigen.

[0264] The term "monoclonal antibody" (mAb) as used herein refers to an antibody obtained from a population of substantially homogeneous antibodies, or to said population of antibodies. The individual antibodies comprising the population are essentially identical, except for possible naturally occurring mutations that may be present in minor amounts. Monoclonal antibodies are highly specific, being directed against a single antigenic site. Furthermore, in contrast to conventional (polyclonal) antibody preparations that typically include different antibodies directed against different determinants (epitopes), each monoclonal antibody is directed against a single determinant on the antigen. The modifier "monoclonal" indicates the character of theantibody as being obtained from a substantially homogeneous population of antibodies and is not to be construed as requiring production of the antibody by any particular method. For example, the monoclonal antibodies to be used in accordance with the present invention may be made by the hybridoma method first described by Kohler et aL, Nature (1975) 256:495, or may be made by recombinant DNA methods. The "monoclonal antibodies" may also be isolated from phage antibody libraries using the techniques described in Clackson et aL, Nature (1991 ) 352:624-628 and Marks et aL, J. MoL BioL (1991 ) 222:581 -597, for example, or by other methods known in the art.

[0265] An “antigen binding fragment” of an antibody comprises one or more variable regions of an intact antibody. Fragments of immunoglobulin molecules such as F(ab')2, Fab, Fab' and Fv can be selectively cleaved from an immunoglobulin molecule using reducing agents and proteases to digest or cleave certain portions of the immunoglobulin protein structure. Fragments of immunoglobulin molecules can also be made recombinantly.

[0266] These antibody fragments, which retain some ability to selectively bind to an antigen (e.g., a polypeptide antigen) of the antibody from which they are derived, can be made using well known methods in the art (see, e.g., Harlow and Lane, supra), and are described further, as follows.

[0267] A Fab fragment consists of a monovalent antigen-binding fragment of an antibody molecule and can be produced by digestion of a whole antibody molecule with the enzyme papain, to yield a fragment consisting of an intact light chain and a portion of a heavy chain.

[0268] A Fab' fragment of an antibody molecule can be obtained by treating a whole antibody molecule with pepsin, followed by reduction, to yield a molecule consisting of an intact light chain and a portion of a heavy chain. Two Fab' fragments are obtained per antibody molecule treated in this manner.

[0269] A (Fab')2 fragment of an antibody can be obtained by treating a whole antibody molecule with the enzyme pepsin, without subsequent reduction. A (Fab')2 fragment is a dimer of two Fab' fragments, held together by two disulfide bonds.

[0270] A Fv fragment is defined as a genetically engineered fragment containing the variable region of a light chain and the variable region of a heavy chain expressed as two chains.

[0271] As used herein “recombinant antibody” refers to antibodies that are prepared, expressed, created or isolated by recombinant means, such as (a) antibodies isolated from an animal (e.g., mouse) that is transgenic for immunoglobulin genes (e.g., human immunoglobulin genes) or hybridoma prepared therefrom; (b) antibodies isolated from a host cell transformed to express the antibody; (c) antibodies isolated from a recombinant antibody library; and (d) antibodies prepared, expressed, created or isolated by any other means that involve splicing of immunoglobulin gene sequences to other DNA sequences.

[0272] Recombinant antibodies can be produced in any format. The smallest antigen binding fragment that retains its complete antigen binding site is the Fv fragment, which consists entirely of variable (V) regions. A soluble and flexible amino acid peptide linker is used to connect the V regions to a scFv (single chain fragment variable) fragment for stabilization of the molecule, or the constant (C) domains are added to the V regions to generate a Fab fragment (fragment, antigen-binding). scFv and Fab are widely used fragments that can be easily produced in prokaryotic hosts. Other antibody formats include disulfide-bond stabilized scFv (ds-scFv), single chain Fab (scFab), as well as di- and multimeric antibody formats like dia-, tria- and tetrabodies, or minibodies (miniAbs) that comprise different formats consisting of scFvs linked to oligomerization domains. The smallest fragments are VHH / VH of camelid heavy chain antibodies and single domain antibodies (sdAb).

[0273] The term “antibody” as used herein also encompasses humanized antibodies, primatized antibodies, human antibodies, synhumanized antibodies and chimeric antibodies.

[0274] The term “chimeric” antibody as used herein refers to an antibody having variable sequences derived from a non-human immunoglobulin, such as a rat or a mouse antibody, and human immunoglobulin constant regions, typically chosen from a human immunoglobulin template. Methods for producing chimeric antibodies are known in the art (see, e.g., Morrison, Science (1985) 229(4719):1202-7; Oi et al.,BioTechniques (1986) 4:214-221 ; Gillies et aL, J. Immunol. Methods (1985) 125:191 - 202; US 5,807,715; US 4,816,567; and US 4,816397).

[0275] “Humanized” forms of non-human (e.g., murine) antibodies are chimeric immunoglobulins that contain minimal sequences derived from non-human immunoglobulin. In general, a humanized antibody will comprise substantially all of at least one, and typically two, variable domains, in which all or substantially all of the CDR regions correspond to those of a non-human immunoglobulin and all or substantially all of the FR regions are those of a human immunoglobulin sequence. The humanized antibody can also comprise at least a portion of an immunoglobulin constant region (Fc), typically that of a human immunoglobulin consensus sequence. Methods of antibody humanization are known in the art (see, e.g., Riechmann et aL, Nature (1988) 332:323-7; US 5,530,101 ; US 5,585,089; US 5,693,761 ; US 5,693,762; and US 6,180,370; EP239400; WO 91 / 09967; US 5,225,539; EP592106; EP519596; Padlan, Mol. Immunol. (1991 ) 28:489-498; Studnicka et aL, Prot. Eng. (1994) 7:805- 814; Roguska et aL, Proc. NatL Acad. Sci. (1994) 91 :969-973; and US 5,565,332).

[0276] “Human antibodies” include antibodies having the amino acid sequence of a human immunoglobulin and include antibodies isolated from human immunoglobulin libraries or from animals transgenic for one or more human immunoglobulin and that do not express endogenous immunoglobulins. Human antibodies can be made by a variety of methods known in the art including phage display methods using antibody libraries derived from human immunoglobulin sequences (see, e.g., US 4,444,887 and US 4,716,1 11 ; WO 98 / 46645; WO 98 / 50433; WO 98 / 24893; WO 98 / 16654; WO 96 / 34096; WO 96 / 33735; and WO 91 / 10741 ). Human antibodies can also be produced using transgenic mice which are incapable of expressing functional endogenous immunoglobulins but which can express human immunoglobulin genes (see, e.g., WO 98 / 24893; WO 92 / 01047; WO 96 / 34096; WO 96 / 33735; US 5,413,923; US 5,625,126; US 5,633,425; US 5,569,825; US 5,661 ,016; US 5,545,806; US 5,814,318; US 5,885,793; US 5,916,771 ; and US 5,939,598, which are incorporated by reference herein in their entireties. Fully human antibodies that recognize a selected epitope can be generated using a technique referred to as “guided selection.” In this approach, a selected non-human monoclonal antibody, for example, a mouse antibody, is used toguide the selection of a completely human antibody recognizing the same epitope (see, e.g., Jespers et al., Biotechnology (1988) 12:899-903).

[0277] The term "synhumanized antibody" refers to an antibody prepared by a method described in W02007019620. A synhumanized antibody includes a variable region of an antibody, wherein the variable region comprises FRs from a New World primate antibody variable region and CDRs from a non-New World primate antibody variable region.

[0278] “Primatized antibodies” comprise monkey variable regions and human constant regions. Methods for producing primatized antibodies are known in the art. See, e.g., U.S. Pat. Nos. 5,658,570; 5,681 ,722; and 5,693,780, which are incorporated herein by reference in their entireties.

[0279] In some embodiments, the antibody is a CD34 or CD3 antibody. In some embodiments of the disclosure, the antibody is MY10 (CD34); QBEND10 (CD34); 9C5 (CD34), or OTK3 (CD3). In some embodiments, the antibody is QBEND10 or 9C5.Target cells

[0280] The cell capture composition comprises an antigen binding moiety that specifically binds to ligand expressed on a target cell.

[0281] In some embodiments of the disclosure the antigen binding moiety specifically binds to a CD2, CD4, CD5, CD7, CD8, CD1 1 a, CDw12, CD13, CD15, CD18, CD19, CD21 , CD22, CD25, CD28, CD29, CD30, CD33, CD34, CD36, CD38, CD40, CD41 , CD42a, CD42b, CD42c, CD42d, CD43, CD45, CD45RA, CD45RB, CD45RC, CD45RO, CD48, CD49b, CD49d (VLA-4), CD49e, CD49f (VLA-6), CD50, CD51 , CD53, CD55, CD58, CD64a, CD68, CD71 , CD72, CD73, CD81 , CD82, CD84, CD85A, CD85K, CD90, CD97, CD99, CD104, CD105, CD109, CD1 10, CD1 1 1 , CD1 12, CD1 14, CD115, CD1 17, CD123, CD124, CD126, CD127, CD130, CD131 , CD133, CD135, CD137, CD138, CD151 , CD157, CD162, CD164, CD166, CD168, CD172a, CD173, CD174, CD175, CD175s, CD176, CD183, CD184 (CXCR4), CD191 , CD200, CD201 , CD205, CD217, CD220, CD221 , CD222, CD223, CD224, CD225, CD226, CD227, CD228, CD229, CD230, CD235a, CD235b, CD236, CD236R, CD238, CD240, CD242, CD243, CD277, CD292, CDw293, CD295, CD298, CD309, CD318, CD324,CD325, CD338, CD344, CD349, CD350, CD361 , CD62L, CRTh2, CCR4, CXCR3, or CCR7 ligand.

[0282] In some embodiments of the disclosure the antigen binding moiety specifically binds to a CD3, CD4, CD8, CD1 1 a, CD18, CD25, CD28, CD34, CD37, CD45, CD47, CD52, CD58, CD62L, CD69, CD74, CD97, CD103, CD117, CD132, CD156a, CD179a, CD179b, CD184, CD232, CD244, CD252, CD302, CD305, CD317, or CD361 ligand.

[0283] In some embodiments of the disclosure the antigen binding moiety specifically binds to a CD34, CD3, or Protein G ligand.Bioreactor cell culture system

[0284] In an aspect, the invention relates to a bioreactor cell culture system. The system comprises a plurality of bioreactors for cell culture, as disclosed in the preceding section, arranged in at least one stack. One example of such a stack is depicted in Figures 13 and 14, where a plurality of multi-layer microfluidic devices 100, as disclosed herein, are arranged in a stack. In each multi-layer microfluidic device 100, a cell capture composition, as disclosed herein, is physically adhered to the hydrophobic upper surface 310 in the one or more grooves 312. The stack may comprise any number of microfluidic devices, for example between 5 and 40, or between 10 and 30.

[0285] The bioreactor cell culture system further comprises a controller configured to individually control operation of one or more fluid pumps according to a cell culture protocol stored in a memory of the controller, the one or more fluid pumps delivering fluid to one or more through bores provided colinearly through the at least one stack, the delivered fluid entering the flow chamber of each bioreactor in the at least one stack according to the stored cell culture protocol.

[0286] With reference again to Figures 5 and 6, for example, the one or more through bores may include (i) through bores 381 , 382 which provide a vertical channel for delivery of culture media and reagents, respectively, to each first inlet port 210 in the stack, and (ii) through bores 383 and 384 which provide a vertical channel for delivery of cells and extraction fluid, respectively, to each second inlet port 330 in the stack.

[0287] Culture medium may be delivered from a single bag, container, flask or other receptacle to all bioreactors in the stack by connecting a tube to through bore 381. Reagents may be delivered from a single bag, container flask or other receptacle to all bioreactors in the stack by connecting a tube to through bore 382.

[0288] In a protocol requiring cell selection, a sample comprising the target cells may be delivered to all bioreactors in the stack by connecting a tube to through bore 383. An extraction fluid may be delivered to all bioreactors in the stack from a supply (bag, container, flask or the like) by connecting a tube to through bore 384.

[0289] Respiratory gas feed (e.g. air + 10% CO2) may be supplied, e.g. by a gas pump, to through bore 483 for supply to the gas flow path 410 in each of the bioreactors.

[0290] In a preferred embodiment, one or more stacks of the type illustrated in Figures 13 and 14 are incorporated into a bioreactor cell culture system. Ideally, the cell culture system has a housing containing the stack(s) and fluid sources, in the form of plurality of reagent bags or containers containing the reagents necessary to perform cell culture protocols and which are fluidly coupled to the stacks through a manifold such as cover plate 620.

[0291] The system includes a controller which comprises a processor communicatively coupled with the controller memory and a user interface which is ideally incorporated into the housing and may comprise a touch screen, and / or display and keypad accessible to an operator of the system. The memory stores instructions for performing one or more cell culture protocols, and the instructions, when executed by the processor, cause the controller to operate the bioreactor cell culture system to perform the one or more cell culture protocols.

[0292] Inside the housing is one or more fluid pumps in communication with the one or more fluid sources. The pumps are operated by the controller to deliver fluid to individual cover plate openings in accordance with cell culture protocols stored the controller memory by controlling opening and closing of valves in the fluid lines. In some embodiments, the processor comprises a microcontroller configured to independently regulate pressure sources coupled to the bioreactor inputs.

[0293] The user interface receives one or more inputs from an operator to select from the memory, or to input to the memory, one or more cell culture protocols. Each culture protocol has a corresponding set of instructions stored in memory which, when executed by the processor, cause the controller to operate the pumps and other components of the system according to the selected / entered protocol.

[0294] For example, the controller is operable to cause extraction of cells from the grooves of individual bioreactors through a cell extraction channel extending colinearly through the stack and in fluid communication with the second outlet port in the individual bioreactors. Ideally, the housing includes a platform or area for receiving a receptacle such as a bag, flask or vial which is coupled to the extraction channel for uncontaminated collection of extracted cells.

[0295] Referring again to Figures 5 and 6, the cell extraction channel is through bore 391 which provides a vertical channel for extraction of cells from each second outlet port 340 in the stack. Through bore 392 provides a vertical channel for removing expired media, e.g. cell culture medium after passing across the grooves 312 of each microfluidic device, from each first outlet port 220 in the stack.

[0296] In some embodiments, the controller has internet connectivity via Wi-Fi or Ethernet connection enabling remote access to the processor for access to data, modification and control of protocols, and remote software updates. Additionally, the controller may be operatively coupled with a scanner or other device to automatically read or receive from a sample an identifier such as a barcode or electronic tag for secure tracking of patient material and customisation of protocols for each patient.Method of preparing a bioreactor for cell culture

[0297] In an aspect, the invention relates to a method of preparing a bioreactor for cell culture. The method comprises providing a multi-layer microfluidic device, as previously described herein, and physically adhering a cell capture composition to the hydrophobic upper surface of the flow layer in the one or more grooves. The cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell. The cell capture composition may generally be according to any of the embodiments previously disclosed herein. The components ofthe cell capture composition may be physically adhered to the hydrophobic upper surface in one or more than one deposition steps.

[0298] The cell capture composition must be physically adhered at least in the grooves of the flow layer, but will typically be adhered across the entire flow channel footprint on the hydrophobic upper surface. The cell capture composition may be produced by a process comprising a step of flowing a fluid, typically an aqueous fluid, comprising the amphiphilic polymer (with or without the antigen binding moiety) into the flow channel and incubating the fluid therein to allow the amphiphilic polymer to deposit on the flow chamber walls including the grooved hydrophobic upper surface. The fluid comprising amphiphilic polymer is preferably flowed across the grooves. In embodiments where the header layer comprises a second inlet port and a second outlet port to provide a longitudinal flow path (along the grooves), cross-flow treatment of the grooves avoids depositing the cell capture composition in the longitudinal flow path before and after the grooves (specifically, in the optionally bifurcating flow paths between the second inlet port and the one or more second fluid inlets, and between the one or more second fluid outlets and the second outlet port), where capture of cells is undesired.

[0299] In some embodiments, as previously discussed, the antigen binding moiety is conjugated to at least one hydrophilic polymeric segment of the amphiphilic polymer. The cell capture composition thus comprises a polymer-antigen binding moiety conjugate, in which the antigen binding moiety is chemically bonded to the amphiphilic polymer by at least one covalent bond.

[0300] In some such embodiments, the polymer-antigen binding moiety conjugate is prepared before introducing the cell capture composition into the microfluidic device. The pre-prepared polymer-antigen binding moiety conjugate may thus be dispersed in a fluid, preferably an aqueous fluid composition, flowed into the flow channel (preferably via cross-flow across the grooves) and incubated therein under conditions to allow the polymer-antigen binding moiety conjugate to deposit on the flow channel walls including in the grooves of the hydrophobic upper surface. The incubation may occur under static or flow conditions. Optionally, incubation may be at room temperature and approximately neutral pH. The inventors have found that incubation of a 50pg / mL aqueous solution (in DPBS buffer) of poloxamer-antibody conjugate for one hour (staticconditions after filling the flow channel) at room temperature is sufficient to produce a functionally acceptable deposition of cell capture composition in the grooves of a hydrophobic polymer flow layer.

[0301] It is also contemplated that a cell capture composition comprising polymer- antigen binding moiety conjugate may be provided on the hydrophobic upper surface of the flow layer by a process where the conjugation reaction between the precursor amphiphilic polymer and the antigen binding moiety is conducted inside the flow channel. Thus, for example, a poloxamer terminally functionalised with activated ester groups or with click functional groups may initially be deposited on the flow channel walls. The antigen binding moiety may then be flowed into the flow channel and allowed to react with the poloxamer in situ, for example according to the reactions depicted in Scheme 1 or 2.

[0302] In other embodiments, the cell capture composition comprises the amphiphilic polymer and the antigen binding moiety as chemically distinct components, i.e. the antigen binding moiety is not conjugated to the amphiphilic polymer.

[0303] In some such embodiments, the antigen binding moiety may be deposited on the grooved hydrophobic upper surface of the flow channel in a first deposition step, and the unconjugated amphiphilic polymer is deposited on the grooved hydrophobic upper surface of the flow channel in a second deposition step, thus producing a cell capture composition, comprising both the unconjugated amphiphilic polymer and the antigen binding moiety, physically adhered to the hydrophobic upper surface in at least the grooves thereof. In the first deposition step, the antigen binding moiety may be provided to the grooves by cross flow and / or longitudinal flow and left to incubate to adsorb thereto under static or flow conditions. For example, the antigen binding moiety dispersed in an aqueous fluid may be flowed into the flow channel and incubated therein under conditions (e.g. room temperature, approximately neutral pH) to allow the antigen binding moiety to deposit on the flow chamber walls, including in the grooves of the hydrophobic upper surface.

[0304] In the second deposition step, the unconjugated amphiphilic polymer dispersed in a fluid, preferably an aqueous fluid composition, may be flowed by cross flow and / or longitudinal flow into the flow channel and incubated therein underconditions to allow the unconjugated amphiphilic polymer to deposit on the flow channel walls. The incubation may occur under static or flow conditions. The inventors have found that incubation of a 50pg / mL aqueous solution of antibody for one hour at room temperature, followed by incubation of 1 % w / v aqueous solution of poloxamer for one hour at room temperature, is sufficient to produce a functional deposit of cell capture composition, comprising both components, in the grooves of the hydrophobic upper surface of the flow layer.

[0305] Alternatively, the antigen binding moiety and the unconjugated amphiphilic polymer may be deposited simultaneously from a fluid comprising a mixture both components. For example, the antigen binding moiety and the amphiphilic polymer, dispersed in an aqueous fluid, may be flowed into the flow channel and incubated therein under conditions (e.g. room temperature, approximately neutral pH) to allow both components to deposit on the flow chamber walls, including in the grooves of the hydrophobic upper surface.

[0306] Referring to Figures 1 -10, for example, the multi-layered microfluidic device 100 to be prepared for cell culture may be configured for uni-directional or bi-directional flow, as previously described herein. A fluid comprising a suitable polymer-antigen binding moiety conjugate, as disclosed herein, may be flowed into flow channel 60, either via first inlet port 210 or second inlet port 330 (when available). For bi-directional flow devices (e.g. as shown in Figure 3), it may be preferred to flow the fluid via first inlet port 210 (cross flow) to avoid depositing cell capture composition in the bifurcating flow path between second fluid port 330 and second fluid inlets 332. Therefore, target cells introduced to flow channel 60 via second fluid port 330 (longitudinal flow) may be less susceptible to premature immobilisation in the bifurcating flow path. The fluid present in flow chamber 60 is then incubated (under static or flow conditions) to allow the polymer-antigen binding moiety conjugate to deposit on the flow chamber walls, including in grooves 312 of hydrophobic upper surface 310.

[0307] For embodiments where the antigen binding moiety is not conjugated to the amphiphilic polymer, a fluid comprising unconjugated antigen binding moiety may be flowed into flow channel 60 and incubated therein to allow the antigen binding moiety to adsorb to the flow channel walls. A second fluid comprising unconjugated amphiphilic polymer may then be flowed into flow channel 60 and incubated therein toallow the unconjugated amphiphilic polymer to deposit on the flow channel walls together with the antigen binding moiety.Multi-layer micro fluidic device fabricated from polymethylpentene

[0308] In another aspect, the invention relates to a multi-layer microfluidic device which includes a gas permeable flow layer comprising, or consisting of, polymethylpentene. Preferably, the header layer and base layer (when present) of the multi-layer microfluidic device also comprises, or consists of, or is made from, polymethylpentene. Further preferably, each layer of the multi-layer microfluidic device comprises, or consists of, or is made from, polymethylpentene.

[0309] With regard to its structural configuration, the multi-layer microfluidic device comprises a header layer, a gas permeable flow layer beneath the header layer and optionally a base layer beneath the flow layer. The flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells. The base layer provides a gas flow path for gas exchange across the flow layer. The header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises: (i) a first inlet port; (ii) a first outlet port; (iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and (iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port. The one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein.

[0310] The multi-layer microfluidic device may generally be configured according to any of the embodiments disclosed in the preceding sections.

[0311] However, at least the gas permeable flow layer of multi-layer microfluidic devices according to this aspect comprises polymethylpentene. The polymethylpentene is present at the hydrophobic upper surface of the flow layer, thus providing the requisite hydrophobic character due to the non-polar hydrocarbonchemical structure of polymethylpentene. Typically, the gas permeable flow layer is fabricated from, and thus consists of, polymethylpentene.

[0312] As used herein, the term polymethylpentene refers to both homopolymers and copolymers of 4-methyl-1 -pentene, also referred to as PMP or TPX. Herein, these terms may be used interchangeably.

[0313] Fabrication of multi-layer microfluidic devices from thermoplastic polymers is desirable due to the short processing times involved in moulding the individual structured layers, in comparison to thermosetting polymers which may require several hours for curing. Furthermore, the use of thermoplastic also provides the opportunity to improve device performance, due to increased rigidity of the individual layers and improved bonding between the layers which can be created by thermal fusion.

[0314] In some embodiments, the header layer and the base layer thus also comprise, or consist, of polymethylpentene. This provides the opportunity to bond the header layer, the flow layer and the base layer together, by thermal fusion of similar thermoplastic materials, to form the multi-layer microfluidic device.

[0315] Polymethylpentene is a thermoplastic polyolefin with a melting point typically in the range of about 220°C-240°C. Its high softening and melting points, and low water absorption properties, make polymethylpentene a desirable material of construction for microfluidic devices which may require autoclave or steam-sterilisation. It is also desirably transparent, so that cell growth can be directly observed.

[0316] The density of polymethylpentene (about 0.83 g / cm3) is lower than many other polyolefins due to the branched structure of the isobutyl groups pendant to the main carbon chain. For the same reason, the gas permeability of polymethylpentene is relatively high. For example, polymethylpentene has an O2 permeability of about 9.4 x 10-15mol / m, in comparison to about 5.2 x 10’18mol / m for polypropylene. The O2 permeability is significantly lower than thermosetting PDMS (c.a. 25 times less). Nevertheless, the inventors recognised that the gas permeability properties of polymethylpentene may provide the opportunity to use thermoplastic polymethylpentene for the gas permeable flow layer in a bioreactor for cell culture, where exchange of respiratory gases (as needed for sustaining cells, including O2)across the flow layer is required to sustain cells retained in the flow channel, provided that thin enough layers (with sufficient rigidity) can be produced.

[0317] Polymethylpentene is also a highly hydrophobic polymer, with a surface tension (about 24 mN / m) closer to that of fluorine polymers such as PTFE than many other thermoplastic polymers.

[0318] However, as set out in Example 22, the inventors encountered significant difficulties in fabricating multi-layer microfluidic devices of the type disclosed herein from polymethylpentene. A particular challenge was to produce accurately structured grooved polymethylpentene flow layers which are sufficiently thin to allow permeation of respiratory gases (as needed for sustaining cells, including O2) in use. Despite its high gas permeability relative to other polyolefins, the gas permeability of polymethylpentene remains lower than that of some thermosetting polymers such as thermosetting PDMS previously used to fabricate microfluidic devices for cell culturing. Therefore, fabrication of a very thin flow layer was required to allow sufficient permeation of gases across the flow layer in use.

[0319] Only after intensive research were the inventors able to develop a method of fabricating suitably functional multi-layer microfluidic devices, as described in a subsequent section.

[0320] In some embodiments, the flow layer comprising polymethylpentene has a maximum thickness, beneath the flow channel, of no more than 400 pm, or no more than 350 pm, 325 pm, 300 pm, 275 pm or 250 pm. As used herein, the maximum thickness refers to the thickness of the flow layer in the regions between the grooves. The maximum thickness may be at least 90 pm, or at least 100 pm, 125 pm, or 150 pm Preferably the maximum thickness is between 100 and 300 pm, such as between 125 pm and 275 pm, or between 150 and 250 pm. Given that an array of grooves of sufficient depth to retain cells must be formed in this thickness, it will be appreciated that the resultant flow layer design is delicate and non-trivial to produce with a high melting thermoplastic such as polymethylpentene.

[0321] In some embodiments, the flow layer has a thickness within the grooves, i.e. between the bottom surfaces of the grooves and the lower surface of the flow layer, of less than 100 pm, or less than 80 pm, preferably less than 65 pm, such as between 25and 65 pm, or between 30 and 60 pm. This facilitates sufficient gas permeation, including O2 permeation, through the flow layer to support cell viability and growth even when substantial concentrations of cells are captured in the grooves.

[0322] In some embodiments, the depth of the grooves in the flow layer is between 50 and 300 pm, such as between 100 and 200 pm, or between 110 and 180 pm, or between 130 pm and 165 pm, for example about 150 pm. In some embodiments, the grooves have an aspect ratio (i.e. width:depth ratio) of between 1.5:1 and 1 :1.5, or between 1.2:1 and 1 :1.2, such as about 1 :1. Groove dimensions in this range have been found to avoid or mitigate loss of cells from the grooves during expansion under a cross flow of cell growth medium sufficient to sustain the cells.

[0323] The flow features defined in the header and base layers are also typically finely structured, so that fabrication of these layers from polymethylpentene was also challenging.

[0324] In some embodiments, the header layer, the flow layer and the base layer each comprise, or consist of, polymethylpentene and have a maximum thickness, beneath or above the flow channel in the assembled device, of no more than 400 pm, or no more than 350 pm, preferably no more than 300 pm. Despite the thinness of the structured layers, the inventors have demonstrated that multi-layer microfluidic devices with excellent structural fidelity, rigidity and pressure resistance can be produced with thermoplastic polymethylpentene.

[0325] Referring again to Figures 1 -12, each header layer 200, flow layer 300 and base layer 400 may comprise a structured layer consisting of moulded polymethylpentene. Each structured layer has a maximum thickness, beneath or above flow channel 60 in the assembled device, of no more than 400 pm, or no more than 350 pm, preferably no more than 300 pm. Grooves 312 formed in flow layer 300 preferably have a depth between 1 10 and 180 pm, or between 130 pm and 165 pm, for example about 150 pm.

[0326] As already described herein with reference to Figures 4, 5, 6, 7 and 9, the structured layers may include structural posts to rest in abutment with the adjacent layer, thus maintaining patency of flow channels formed between the layers. For example, structural posts 250 are formed on header layer 200, to maintain patency offlow channel 60. Structural posts 450 are formed on base layer 400, to maintain patency of gas flow path 410. A particular advantage of using thermoplastic polymethylpentene in the fabrication of the structured layers is that the structural posts may be thermally fused to the adjacent layer (i.e. flow layer 300, for both structural posts 250 and 450) in the assembled device. This may be achieved by a heat pressing step, conducted above the glass transition temperature of polymethylpentene. The resultant thermal bonding provides the opportunity to reduce the number of support posts and / or the dimensions of the support posts (particularly the profile of the support posts in the direction of cell flow, as already described herein), while still retaining good dimensional stability of flow channels in the device (including under elevated pressures).

[0327] The inventors have also discovered that microfluidic devices comprising a grooved polymethylpentene surface in the flow channel provide excellent performance in cell capture and culture applications.Method of fabricating a multi-layer microfluidic device

[0328] In a further aspect, the invention relates to a method of fabricating a multilayer microfluidic device of the type described in the preceding sections, where the header layer, the optional base layer and the flow layer each comprise, or consist of, or are made from, polymethylpentene.

[0329] The method comprises producing the header layer, the flow layer and optionally the base layer as separate components by the following steps:(a) moulding polymethylpentene under compression in a mould, at a temperature above the polymethylpentene melting point, to produce a structured layer,(b) cooling the structured layer,(c) demoulding the structured layer before the polymethylpentene cools to ambient temperature; and(d) annealing the structured layer under compression against a flat substrate at a temperature between the polymethylpentene glass transition temperature and the polymethylpentene melting point, thereby reducing the surface roughness of the structured layer.

[0330] The method then comprises aligning the header layer, the flow layer and optionally the base layer and compressing the aligned layers at a temperature above the glass transition temperature of the polymethylpentene to thermally fuse the layers, thereby forming the multi-layer microfluidic device.Moulding

[0331] The polymethylpentene may be provided initially in the form of pellets. A sufficient amount of the polymethylpentene may be moulded in a heated press, such as a Carver hydraulic press, against a structured mould such as an aluminium micromachined mould or etched silicon mould. Moulding is conducted above the melting point of polymethylpentene, allowing the molten polymethylpentene to flow, and at a sufficiently high pressure to avoid bubble retention in the polymer. In some embodiments, the polymethylpentene is moulded under compression of between 1 and 5 tons at a temperature of between 230 and 270°C, such as between 240 and 260°C, for example about 250°C. Excessively high temperatures and moulding times should be avoided to limit oxidation or other thermal degradation of the polymer. In some embodiments, the polymethylpentene is above the polymethylpentene melting point for less than 60 seconds during moulding.Cooling and demoulding

[0332] After moulding the hot polymethylpentene, the resultant structured layer is cooled and demoulded before the polymethylpentene cools to ambient temperature. As used herein, ambient temperature refers to the atmospheric temperature in the production location (typically below 30 °C). In some embodiments, the structured layer is demoulded before the polymethylpentene cools to 30°C, or 40°C. In some embodiments, the structured layer is demoulded at between 40°C and 100°C, such as between about 60°C and about 80°C. Inadequate cooling and / or excessive removal force can result in deformation of the structured layer upon demolding, with the flow layer typically being most susceptible to defects due to its depth and fine features. On the other hand, excessive cooling to the point that the polymethylpentene becomes fully hardened risks damaging the mould and / or the structured layer during demoulding. In some embodiments, a hydrophobic coating layer (e.g. silane) applied to the mould before moulding is used to facilitate demoulding.Annealing

[0333] Annealing of the structured layers was found necessary to reduce the surface roughness, thereby improving successful bonding in the subsequent fabrication of the multilayer device. Pressure and time need to be precisely controlled at a temperature between the polymethylpentene glass transition temperature (about 170°C) and its melting point. Excessive pressure and / or time during annealing may damage moulded features formed in the structured layers, while inadequate pressure and / or time may leave an insufficiently smooth surface for subsequent thermal bonding. Annealing against a flat substrate, preferably an optically flat substrate, helps to produce the necessary surface smoothness.

[0334] In some embodiments, the structured layer is annealed under compression of between 0.1 and 0.5 tons, such as between 0.2 and 0.4 tons, against an optically flat substrate at a temperature of between 170 and 200°C, such as between 175 and 190°C, for example at about 180°C.Assembling

[0335] The multi-layer microfluidic device is produced by aligning the required structured layers, including the header layer, the flow layer and optionally the base layer, and compressing the aligned layers at a temperature above the glass transition temperature of the polymethylpentene to thermally fuse the layers. The compression and heating may be provided by use of a heated press, for example a Carver hydraulic press.

[0336] Pressure, temperature and time should be carefully controlled for bonding between layers to be successful. Extreme conditions can result in damage to moulded features and oxidation of the polymethylpentene. Conversely, insufficient heating and pressure leads to incomplete bonding. If pressure is not uniformly distributed across the layers, conformal contact may not be achieved resulting in non-uniform thermal distribution and a microfluidic device that bulges in the centre. The layers must be precisely aligned prior to thermal bonding as the process is irreversible and any misalignment will result in substandard device with an unpredictable flow pattern.

[0337] In some embodiments, the aligned layers are compressed at between 0.3 and 1 .5 tons, such as between 0.6 and 1 .0 tons, at a temperature of between 180 and 200°C, such as between 175 and 190°C, for example at about 180°C, to thermally fuse the layers.Method of culturing cells

[0338] In an aspect, the invention relates to a method of culturing cells. The method comprises providing a microfluidic device comprising a flow channel located over a substrate comprising a hydrophobic upper surface. In some embodiments, the microfluidic device is a multi-layer microfluidic device according to any of the embodiments previously described.

[0339] The hydrophobic upper surface comprises an array of grooves that are substantially parallel. Preferably, a cell capture composition is physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves. In some embodiments, the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell. In some embodiments, the microfluidic device with cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves is a bioreactor for cell culture according to any of the embodiments previously described.

[0340] The cell capture composition may be according to any of the embodiments previously described herein, and may also be physically adhered to the hydrophobic upper surface in the grooves by any of the methods previously described. Preferably, the cell capture composition is physically adhered in each of the grooves in the groove array, and may be present as a continuous coating over the hydrophobic upper surface including in the grooves and regions intermediate to the grooves.

[0341] The method of culturing cells comprises introducing cells into the flow channel of the microfluidic device under conditions to immobilise the target cells on the cell capture composition, at least in the grooves of the groove array. Culture medium is flowed through the flow channel over the groove array in a first flow direction that is across the grooves, and the target cells are cultured under conditions to support viability, growth and / or differentiation.Introducing cells

[0342] Cells, including at least the target cells to be cultured, may be introduced into the flow channel by flow across the grooves (cross flow orthogonal to the longitudinal dimension of grooves) or flow along the grooves (longitudinal flow). In some embodiments, the cells are introduced by flow along the grooves in a second flow direction that is substantially orthogonal to the first flow direction subsequently used to flow culture medium over the grooves. Flow of cells along the grooves may advantageously allow a more consistent number of target cells to be immobilised in each groove.

[0343] Before introducing the cells into the groove array, the microfluidic device may first be primed by flowing culture medium (cross flow and / or longitudinal flow) into the device until it is filled.

[0344] The microfluidic device can be configured with cell inlet and outlet ports. If depositing the cells into one or more grooves under static conditions the outlet port will be closed following inoculation of the cells by flow from the inlet port.

[0345] Referring to Figures 3-10, for example, the microfluidic device used in a method of cell culture may be a microfluidic device 100 configured for bi-directional flow, as previously described herein, having a cell capture composition physically adhered to hydrophobic upper surface 310 including in grooves 312. Cells may be introduced to flow channel 60 from second inlet port 330, via second fluid inlets 332. The resultant flow of cells is directed along grooves 312 in flow direction B. The target cells introduced in this way are thus immobilised on the cell capture composition, including (or primarily) in grooves 312. Advantageously, cells are delivered uniformly to each groove in the array with this approach, resulting in an efficient usage of the device’s working area. Waste product, optionally including non-target cells which are non-adherent to the cell capture composition, may be flowed out of flow channel via second fluid outlets 342 and second outlet port 340 during this cell capture process.

[0346] Referring to Figures 1 -2, as another example, cells may be introduced to flow channel 60 from first inlet port 210, via first fluid inlets 212. The resultant flow of cells is directed across grooves 312 in flow direction A. The target cells introduced inthis way are thus immobilised on the cell capture composition, including (or primarily) in grooves 312.

[0347] The cells are introduced into the flow channel under conditions to immobilise the target cells on the cell capture composition. The target cells are immobilised when the antigen binding moiety of the cell capture composition specifically binds to the binding site of the target cell, and the immobilisation conditions are selected to allow such specific binding to proceed.

[0348] The method may include flowing a sample comprising the target cells through the flow channel, wherein the target cells are selectively immobilised on the cell capture composition. Here, selective immobilisation means that the target cells are preferentially immobilised in comparison to other, non-target cells in the sample. Nonimmobilised cells, typically the non-target cells, are flowed out of the opposite end of the flow channel. The method thus allows target cells to be selected, where required, from a complex biological sample prior to the culturing step. The selectivity of capture may be enhanced by a washing step conducted at a flow stress sufficient to displace non-specifically bound cells, suitably at least 0.1 dyne / cm2, such as between 0.1 and 10 dyne / cm2, for example or at least 0.5 dyne / cm2, or at least 1 .0 dyne / cm2, such as about 1 .5 dyne / cm2.

[0349] The cells may be present in bodily fluids (e.g. blood), processed or partially processed bodily fluids (e.g., blood processed to remove platelets and / or serum), or resuspended in an appropriate fluid medium (e.g., culture medium) before being introduced into the device.

[0350] The cells may be prokaryotic cells or eukaryotic cells. The cells may be from a single-celled organism or a multi-celled organism. In some cases, the cells are genetically engineered, e.g., the cells may be chimeric cells. The cells may be bacterial, fungi, plant, or animal cells, etc. The cells may be from a human or a nonhuman animal or mammal. For instance, if the cell is from an animal, the cell may be a cardiac cell, a fibroblast, a keratinocyte, a hepatocyte, a chondrocyte, a neural cell, an osteocyte, an osteoblast, a muscle cell, a blood cell, an endothelial cell, a stem cell, an immune cell (e.g., a T-cell, a B-cell, a macrophage, a neutrophil, a basophil, a mast cell, an eosinophil), etc. In some cases, the cell is a cancer cell. The cells may be non-adherent cell lines (e.g. L1 .2, mouse immune B; Jurkat, KG1 a, human CD4+T ; Ramos, human immune B) or adherent cell lines (GPE86, mouse embryonic fibroblast), as well as primary cells. In a preferred embodiment, the cell is a primary cell such as a primary immune cell or stem cell. In one embodiment, the cell is a primary immune cell, for example, a T cell. In another embodiment, the cell is a stem cell, preferably a haemopoietic stem cell.

[0351] As used herein, the term “immune cells” refer to cells of the immune system, which defend the body against disease and foreign materials. Non-limiting examples of immune cells include dendritic cells, such as bone marrow-derived dendritic cells; lymphocytes, such as B cells, T cells, and natural killer cells; and macrophages. The immune cells may, in some embodiments, be derived from bone marrow, spleen, or blood from a suitable subject. For example, the immune cells may arise from a human or a non-human mammal, such as a monkey, ape, cow, sheep, goat, horse, donkey, llama, rabbit, pig, mouse, rat, guinea pig, hamster, dog, cat, etc. In a preferred embodiment, the immune cell is a T-cell, preferably a human T-cell.

[0352] As used herein, the term “stem cells” refers to clonogenic cells capable of both self-renewal and multilineage differentiation. Based on their origin, stem cells are categorised either as embryonic stem cells (ESCs) or as postnatal stem cells / somatic stem cells / adult stem cells (ASCs).

[0353] Embryonic stem cells (ESCs) can be derived from embryos that are 2-1 1 days old called blastocysts. They are totipotent - capable of differentiating into any type of cell including germ cells. ESCs are considered immortal as they can be propagated and maintained in an undifferentiated state indefinitely.

[0354] Adult stem cells (ASCs) are found in most adult tissues. They are multipotent - capable of differentiating into more than one cell type but not all cell types. Depending on their origin, ASCs can be further classified as hematopoietic stem cells (HSCs) and mesenchymal stem cells (MSCs). HSCs can be obtained either from cord blood or peripheral blood. MSCs are those that originate from the mesoderm layer of the fetus and in the adult reside in a variety of tissues such as the bone marrow stem cells (BMSCs), limbal stem cells, hepatic stem cells, dermal stem cells, etc.

[0355] HSCs can be divided into a long-term repopulating subset, capable of indefinite self-renewal, and a short-term subset that self-renew for a defined interval. HSCs give rise to nonself-renewing oligolineage progenitors, which in turn give rise to progeny that are more restricted in their differentiation potential, and finally to functionally mature cells including myeloid (monocytes and macrophages, neutrophils, basophils, eosinophils, dendritic cells), erythroid (erythrocytes), megakaryocytic (platelets) and lymphoid lineages (T-cells, B-cells, NK-cells). In some embodiments, the stem cell is a long-term repopulating haemopoietic stem cell.Cell culture

[0356] After the target cells are immobilised on the cell capture composition, culture medium is flowed through the flow channel over the groove array in a first flow direction that is across the grooves, and the target cells are cultured under conditions to support viability, growth and / or differentiation.

[0357] Optionally, the captured cells are released from the antigen binding moiety using a cell release agent, for example, the proteolytic enzyme chymopapain, or neuraminidase, before or during the culturing. Alternatively, the captured cells may be released from the antibody via competitive displacement, for example, the cell release agent may be a peptide which competitively binds to the antibody to displace the captured cells. The released target cells are retained in the grooves during the culturing because the flow of culture medium is directed across the grooves, thus avoiding or minimising disruption of the cells received therein.

[0358] The cells may be cultured, for example for 1 -14 days, under conditions that maintain viability and / or cell growth. For example, a time of not more than 48 hours may be preferred for HSC transduction, whereas up to 14 days may be preferred for CAR T cells. The growth can be characterized by cell division (mitosis) or by other processes, such as differentiation, during which the cells can change into specific types that are capable of functions analogous to tissues or organs in the whole organism.

[0359] During culture, cell culture medium is provided across the top of the groove array to provide nutrients and growth factors to cells retained in one or more grooves for cell metabolism and to remove the waste products generated by the cells.

[0360] The culture medium is flowed through the flow channel at a flow rate sufficient to support viability, growth and / or differentiation. In some embodiments, the flow rate is equivalent to 0.2 mL per million cells per day, for example at least 0.3 mL, at least 0.4 mL, at least 0.5mL, at least 0.6 mL, at least 0.7 mL, at least 0.8 mL, at least 0.9 mL, at least 1 mL, at least 1.1 mL, at least 1 .2 mL, at least 1 .3 mL, at least 1 .4 mL, at least 1 .5 mL, at least 1 .6 mL, at least 1 .7 mL, at least 1 .8 mL, at least 1 .9 mL, at least 2 mL, at least 2.5 mL, at least 3 mL, at least 3.5 mL, at least 4 mL, at least 4.5 mL, at least 5 mL per million cells per day. In some embodiments the flow rate is equivalent to between 0.2-5 mL per million cells per day, or 0.5-1 .5 mL per million cells per day, for example, 1 mL per million cells per day.

[0361] In some embodiments, the depth of the grooves in the flow layer is between 50 and 300 pm, such as between 100 and 200 pm, or between 110 and 180 pm, or between 130 pm and 165 pm, for example about 150 pm. In some embodiments, the grooves have an aspect ratio (i.e. width:depth ratio) of between 1.5:1 and 1 :1.5, or between 1 .2:1 and 1 :1 .2, such as about 1 :1 .

[0362] In one example, the culture medium continually flows across the groove array. This type of flow is referred to herein as “continuous flow”. Continuous flow is just that - continuous. The medium is continuously pumped with no breaks, resulting in a continuous stream of nutrients and growth factors. Alternatively, the culture medium may be provided by oscillatory flow or pulsatile flow. “Oscillatory flow” refers to an oscillating flow with zero mean velocity. “Pulsatile flow” is an oscillating flow superimposed on a steady flow with a non-zero mean velocity.

[0363] When a fluid is flowing through the flow channel of the microfluidic device disclosed herein, the flow may further be characterised depending on the velocity and viscosity of the fluid as laminar flow or turbulent flow. In one example, the cell culture medium is perfused across the groove array in a laminar flow. Laminar flow tends to occur at lower velocities, below a threshold at which it becomes turbulent. Turbulent flow is an unsteady flow regime that is characterized by eddies or small packets of fluid particles, which result in lateral mixing. In non-scientific terms, laminar flow is smooth, while turbulent flow is rough. Typically, the flow in devices useful in methods of the invention is laminar because of the low Reynolds number (typically less than 10).

[0364] The device can be configured with cell culture medium inlet and outlet ports. Media can be perfused across the grooves from the inlet port(s) by syringe pumps, peristaltic pumps, hydraulic pressure, or on-chip micropumps. Media perfusion provides controlled supply of growth factors and proteins as well as regulated exposure of fluid-induced forces to cells.

[0365] In some embodiments, either of two perfusion modes may be used: singlepass perfusion and recirculating perfusion. In single-pass perfusion, cell culture medium is perfused over the groove array (across or along the grooves) to flow waste container(s) connected to an outlet port(s), during which the perfused growth factors, cell metabolites, and secreted factors are exposed to cells for short periods of time. While in recirculating perfusion, medium, cell metabolites and secreted factors are recirculated across the groove array and exposed to cells for longer periods of time.

[0366] As used herein the terms “culture medium” or “culture media” refer to a medium / media designed to support the growth of cells. Cell culture media generally comprise an appropriate source of energy and compounds which regulate the cell cycle (often referred to as base or basal medium). A typical culture medium is composed of a complement of amino acids, vitamins, inorganic salts, glucose, and optionally, growth factors, hormones, and / or attachment factors. Cultures can be supplemented one or more times with critical nutrients (e.g., glucose). In some embodiments, the culture medium is supplemented with, for example, growth factors, transduction enhancers to, for example, modify the cells. In some embodiments, there are two flows of culture medium, one which typically contains base media components and the other which typically contains growth media components comprising high value nutrients / supplements. In addition to nutrients, the medium also helps maintain pH and osmolality. In some embodiments, the skilled person will appreciate that the cells to be cultured will determine the culture medium and conditions for cell culture. Selection of suitable cell culture medium and culture conditions (aerobic, anaerobic, temperature, pH) is within the routine skill of the skilled person.

[0367] O2 can be supplied to the culture by any suitable method, including sparging, diffusion into the flow chamber through a gas-permeable portion of the walls that define the flow chamber periphery, and via the culture medium. In the latter method, the culture medium may be perfused through an oxygenation chamber before it enters theflow channel, ensuring constant supplementation of O2. Typically, a continuous replenishment of oxygen is required to avoid hypoxic conditions adjacent to the culture. This is of particular concern in high cell-density culture.

[0368] In some embodiments, the substrate comprising the array of grooves is a gas-permeable substrate, thus allowing gas exchange across the gas-permeable substrate. At least a portion of the required O2 is then supplied to the culture by diffusion through the gas-permeable substrate. Waste gases from the cell culture may also diffuse out of the flow channel via the gas-permeable substrate.

[0369] In some embodiments, the gas-permeable substrate comprising the array of grooves comprises, or consists of, polymethylpentene. Preferably, such a substrate has a maximum thickness, beneath the flow channel, of no more than 400 pm, or no more than 350 pm, 325 pm, 300 pm, 275 pm or 250 pm. As used herein, the maximum thickness refers to the thickness of the substrate in the regions between the grooves. The maximum thickness may be at least 90 pm, or at least 100 pm, 125 pm, or 150 pm Preferably the maximum thickness is between 100 and 300 pm, such as between 125 pm and 275 pm, or between 150 and 250 pm. In some embodiments, the depth of the grooves in the substrate is between 50 and 300 pm, such as between 100 and 200 pm, or between 1 10 and 180 pm, or between 130 pm and 165 pm, for example about 150 pm. In some embodiments, the grooves have an aspect ratio (i.e. width:depth ratio) of between 1.5:1 and 1 :1.5, or between 1.2:1 and 1 :1.2, such as about 1 :1 . In some embodiments, the substrate has a thickness within the grooves, i.e. between the bottom surfaces of the grooves and the lower surface of the flow layer, of less than 100 pm, or less than 80 pm, preferably less than 65 pm, such as between 25 and 65 pm, or between 30 and 60 pm.

[0370] In some embodiments, the microfluidic device comprising the gas- permeable substrate is placed in a purpose-built incubator, preferably a CO2 incubator, to control gas exchange across the gas-permeable substrate and / or the culture temperature. Usually for mammalian cell culture, the level of CO2 in the incubator defines the level of oxygen. Controlling temperature and levels of CO2 and humidity in the incubator is important to the health and growth of cultured cells. For the majority of mammalian cell lines and primary cells, the optimal growth temperature is 37°C. A humidified atmosphere of approximately 95 % avoids desiccation of the cultures. CO2is typically needed as part of the medium buffer system to regulate the pH. The most commonly used CO2 - bicarbonate buffering system depends on a chamber atmosphere of 5-10 % CO2, providing a pH of 7.2 to 7.4.

[0371] In some embodiments, the microfluidic device comprises one or more gas flow paths beneath the gas-permeable substrate. At least a portion of the required O2 is then supplied to the culture by flowing a respiratory gas feed (e.g. air + 10% CO2) through the gas flow path(s), and allowing O2 to diffuse through the gas-permeable substrate. The inclusion of one or more gas flow paths, for example, immediately beneath the gas-permeable substrate that the cells are already cultured on (in one or more grooves) may provide better gas exchange between the cells and the desired atmosphere. It also eliminates the need for a gas-liquid interface.

[0372] Optionally, a pressure difference is applied across the gas-permeable substrate to modulate gas permeation across the gas-permeable substrate, for example between the flow channel and the one or more gas flow paths. For example, a higher pressure in the flow channel may mitigate or avoid the formation of gas bubbles in the flow channel during system priming, cell suspension loading and medium perfusion processes. Bubbles may be detrimental for microscale cell culture leading to blockage of microchannels or shear damage to cells.

[0373] Once culturing of the cells is complete, the cultured cells may be harvested from the device. This may be accomplished by flowing an extraction medium through the flow channel and out of an outlet port for collection. In some embodiments, the extraction medium is flowed through the flow channel in a second flow direction that is along the grooves. The flow of extraction medium may be provided at an average flow velocity of at least 10 mm / min and / or at a shear stress of at least 0.1 dyne / cm2, or at least 0.5 dyne / cm2, preferably at least 1 dyne / cm2, such as about 1 .5 dyne / cm2.

[0374] Referring to Figures 1 -10, for example, the microfluidic device used in a method of culturing cells may be a multi-layer microfluidic device 100 configured for uni-directional or bi-directional flow, as previously described herein, having a cell capture composition physically adhered to hydrophobic upper surface 310 including in grooves 312. After immobilisation of target cells on the cell capture composition, the immobilised cells may optionally be released by introducing a cell release agent to theflow channel, either by cross flow or longitudinal flow, preferably by cross flow to minimise disruption of the released cells from the grooves.

[0375] Culture medium is then introduced to flow channel 60 from first inlet port 210, via first fluid inlets 212. The resultant flow of culture medium is directed across grooves 312 in flow direction A, which is substantially orthogonal to the longitudinal dimension of the grooves. Waste product is flowed out of flow channel 60 via first fluid outlets 222 and first outlet port 392. During the culturing, a respiratory gas feed (e.g. air + 10% CO2) is flowed though gas flow path 410. Gas exchange across gas permeable flow layer 300 provides sufficient O2 in flow channel 60 to sustain the cells, and removes waste gases from the cell chamber.

[0376] After culturing is complete, the cells are harvested from the device by flowing an extraction medium through the flow channel. Preferably, the extraction medium is flowed along the grooves 312, as may be accomplished with the microfluidic devices described with reference to Figures 3-10. Thus, for example, the extraction medium is introduced to flow channel 60 from second inlet port 330, via second fluid inlets 332. The resultant flow of extraction fluid is directed along grooves 312 in flow direction B, with the harvested cells flowed out of flow channel 60 via second fluid outlets 342 and second outlet port 340 for collection.

[0377] The method disclosed herein advantageously allows target cells to be separated from a sample comprising both target and non-target cells, and subsequently cultured in and harvested from a single microfluidic device.Method of transfecting / transducing cells

[0378] In another aspect, the invention relates to a method of transfecting / transducing cells. The method comprises providing a microfluidic device comprising a flow channel located over a substrate comprising a hydrophobic upper surface. The hydrophobic upper surface comprises an array of grooves that are substantially parallel. Preferably, a cell capture composition is physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves. In some embodiments, the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

[0379] The cell capture composition may be according to any of the embodiments previously described, and may also be physically adhered to the hydrophobic upper surface in the grooves by any of the methods previously described. Preferably, the cell capture composition is physically adhered in each of the grooves in the groove array, and may be present as a continuous coating over the hydrophobic upper surface including in the grooves and regions intermediate to the grooves.

[0380] In some embodiments, the microfluidic device is a multi-layer microfluidic device according to any of the embodiments previously described. In some embodiments, the microfluidic device with cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves is a bioreactor for cell culture according to any of the embodiments previously described.Introducing cells

[0381] The method of transfecting / transducing cells comprises introducing cells into the flow channel under conditions to immobilise the target cells on the cell capture composition. The cells, and the method of introducing the cells to the flow channel, are generally as described in the previous section relating to methods of culturing cells. It is also envisaged that the target cells may be selected and isolated prior to introduction to the microfluidic device for transduction / transfection.Transfecting / transducing cells

[0382] At least one exogenous molecule is introduced into the flow channel, and the target cells and the at least one exogenous molecule are incubated under conditions to transfect / transduce the target cells with the exogenous molecule.

[0383] Optionally, the captured cells are released from the antigen binding moiety using a cell release agent, for example, the proteolytic enzyme chymopapain, or neuraminidase, before or while introducing the at least one exogenous molecule into the flow channel. Alternatively, the captured cells are released from the antibody via competitive displacement, for example, the cell release agent may be a peptide which competitively binds to the antibody to displace the captured cells before or while introducing the at least one exogenous molecule into the flow channel.

[0384] In some embodiments, the at least one exogenous molecule is introduced by flow through the flow channel over the groove array in a first flow direction that is across the grooves. This may be particularly advantageous if the cells are released prior to or during transfection / transduction, since the released target cells may be retained in the grooves while avoiding or minimising disruption of the cells received therein.

[0385] As used herein, an “exogenous” molecule refers to a molecule that is not normally present in the cell, but may be introduced into the cell, or an endogenous molecule that is over-expressed. Being “normally present in the cell” is determined by a certain developmental stage of the cell and environmental conditions. For instance, a molecule that is present only during embryonic development of a tissue / organ is an exogenous molecule with respect to an adult cell of said tissue / organ. Similarly, a molecule induced by heat shock is an exogenous molecule with respect to a non-heat- shocked cell. An exogenous molecule can comprise, for example, a functioning version of a malfunctioning endogenous molecule or a malfunctioning version of a normally- functioning endogenous molecule, or an endogenous molecule expressed in a different cellular compartment or at levels higher than endogenous levels.

[0386] An exogenous molecule may be a small molecule generated by chemical synthesis, or a macromolecule such as a protein, nucleic acid, carbohydrates, lipids, glycoproteins, lipoproteins, polysaccharides, any modified derivatives thereof, or any complex comprising one or more of the above molecules.

[0387] The deliberate delivery or introduction of nucleic acids into eukaryotic cells is typically referred to as “transfection”. Transfection may also refer to other methods and cell types, although other terms are often preferred. Genetic material (such as supercoiled plasmid DNA, mRNA or siRNA constructs), or even proteins such as antibodies, may be “transfected”. As used herein, “transfection” relates generally to non-viral delivery of a molecule into a cell including bacterial and non-animal eukaryotic cells.

[0388] Methods for the introduction of exogenous molecules into cells are known to those of skill in the art and include, but are not limited to, lipid-mediated transfer (i.e., liposomes or lipid-nanoparticle delivery, including neutral and cationic lipids),electroporation, direct injection, cell fusion, particle bombardment, calcium phosphate co-precipitation, DEAE-dextran-mediated transfer and viral vector-mediated transfer.

[0389] The present invention includes use of vectors for manipulation or transfer of genetic constructs. A vector is a nucleic acid molecule, preferably a DNA molecule, that can be used to artificially carry foreign genetic material; into another cell, where it can be replicated or expressed. A vector containing foreign DNA is referred to as a “recombinant vector”. Examples of vectors include, but are not limited to, plasmids, viral vectors, cosmids, extrachromosomal elements, minichromosomes.

[0390] The vector is generally a nucleotide sequence comprising an insert (transgene) and a larger sequence that serves as the "backbone" of the vector. The purpose of a vector which transfers genetic information to another cell is typically to isolate, multiply, or express the insert in the target cell. Vectors designed specifically for the expression of a transgene in a target cell are called “expression vectors”, and generally have a promoter sequence that drives expression of the transgene. Selection of appropriate vectors is within the knowledge of those having skill in the art.

[0391] As used herein, the term "promoter" is to be taken in its broadest context and includes the transcriptional regulatory sequences of a genomic gene, including the TATA box or initiator element, which is required for accurate transcription initiation, with or without additional regulatory elements (e.g., upstream activating sequences, transcription factor binding sites, enhancers and silencers) that alter expression of a nucleic acid, for example, in response to a developmental and / or external stimulus, or in a tissue-specific manner. In the present context, the term "promoter" is also used to describe a recombinant, synthetic or fusion nucleic acid, or derivative which confers, activates or enhances the expression of a nucleic acid to which it is operably linked. Exemplary promoters can contain additional copies of one or more specific regulatory elements to further enhance expression and / or alter the spatial expression and / or temporal expression of said nucleic acid.

[0392] As used herein, the term "operably linked to" means positioning a promoter relative to a nucleic acid such that expression of the nucleic acid is controlled by the promoter. A promoter can be operably linked to numerous nucleic acids, for example, through an internal ribosome entry site.

[0393] As used herein, “virus,” refers to any of a large group of infectious entities that cannot grow or replicate without a host cell. Viruses typically contain a protein coat surrounding an RNA or DNA core of genetic material, but no semipermeable membrane, and are capable of growth and multiplication only in living cells. Viruses include those that are formed when, such as when a vector containing all or a part of a viral genome, is transduced into an appropriate cell or cell line for the generation of such particles. The resulting viral particles have a variety of uses, including, but not limited to, transferring nucleic acids into cells. Thus, a virus is a packaged viral genome. A virus can refer to a single particle, a stock of particles or a viral genome.

[0394] As used herein, viral vector refers to a nucleic acid vector construct that includes at least one element of viral origin and can be packaged into a viral vector particle or virus. Reference to viral vector herein is used interchangeably with virus when it is packaged inside a protein coat. The viral vector particles or virus can be used for the purpose of transferring DNA, RNA or other nucleic acids into cells. Viral vectors include, but are not limited to, retroviral vectors, vaccinia vectors, lentiviral vectors, herpes virus vectors (e.g., HSV), baculoviral vectors, cytomegalovirus (CMV) vectors, papillomavirus vectors, simian virus (SV40) vectors, Sindbis vectors, Semliki Forest virus vectors, phage vectors, adenoviral vectors, and adeno-associated viral (AAV) vectors. Viral vectors typically include engineered viruses that are operatively linked to exogenous transgenes to transfer (as vehicles or shuttles) the exogenous genes into cells.

[0395] When a viral vector or viral particle is used to transfer genetic material of interest into a cell, the technique is referred to as "transduction". Thus generally, to "transduce" a cell is to use a viral vector or viral particle to transfer genetic material into a cell.

[0396] As used herein, nucleic acid molecule includes deoxyribonucleic acid (DNA) and ribonucleic acid (RNA), can be single- or double-stranded; can be linear, branched or circular; and can be of any length. Nucleic acids include those capable of forming duplexes, as well as triplex-forming nucleic acids. Nucleic acids also include analogs or derivatives of DNA or RNA. Analogs may comprise one or more of an altered backbone (e.g., phosphate-sugar backbone analogs), pentose sugar(s) or nucleobase(s) (e.g., universal bases, xeno nucleic acids). Examples include peptidenucleic acid (PNA), morpholino oligomers, locked nucleic acid (LNA), as well as glycol nucleic acid (GNA) and threose nucleic acid (TNA). Nucleic acids can encode gene products, such as, for example, polypeptides, regulatory RNAs (e.g., microRNAs, small interfering RNAs (siRNAs), messenger RNA (mRNA), guide RNAs (gRNA), and antisense), ribozymes, and small hairpins.

[0397] The exogenous molecule (e.g., vector) may be introduced into the device, simultaneously with or after introduction of the cells by cross flow and / or longitudinal flow. The exogenous molecule (e.g., vector) may be introduced into the device subsequent to cell expansion by the culture methods described herein. The exogenous molecule can be supplied in a medium that promotes transfection / transduction, for example, assist, promote, or facilitate entry into the cell, including liposome formulations, lipofectin or a culture medium that supports viral transduction. In some methods, the cells are activated prior to addition of a viral vector, for example, T cells may be activated by stimulation of the TCR activation pathway via CD3- and CD28- specific antibodies in combination with cytokines such as IL-7 and IL-15.

[0398] The methods may include static (e.g., single loading) or flow transfection / transduction, or a combination thereof. Flow transduction methods may include but are not limited to continuous perfusion where the exogenous molecule (e.g., vector) is continuously perfused through the device for the entire transfection / transduction duration, or recirculating perfusion where the viral vectors may be recirculated through the device, for example, using a peristaltic pump. In another example, consecutive perfusion is used where the exogenous molecule is loaded and allowed to incubate under static conditions. After a shortened transfection / transduction period, a further feed containing exogenous molecule together with e.g. culture medium can be loaded into the device.

[0399] The method of transfecting / transducing cells may optionally comprise a step of flowing a culture medium through the flow channel to culture the target cells in the culture medium. Preferably, the culture medium is flowed through the flow channel in a first flow direction that is across the grooves. The target cells may be cultured in the culture medium prior to flowing the exogenous molecule through the flow channel.

[0400] Alternatively, the target cells may be cultured in the culture medium after transfecting / transducing the target cells with the exogenous molecule. Thus, following transfection / transduction, the cells may be expanded by culture methods previously described herein. Expansion of transfected / transduced cells may be accomplished by drug selection when an appropriate drug-resistance marker is included in the transfected / transduced DNA. In such embodiments, a selective medium containing the drug is used to expand the transfected / transduced cells.

[0401] The substrate comprising a hydrophobic upper surface, and grooves therein, may be configured (as disclosed in the previous section relating to methods of culturing cells) to retain the cells despite substantial cross-flows during culturing, and to allow sufficient gas exchange, including O2 permeation, across the substrate to sustain the cells.

[0402] Respiratory gases may be supplied to the cells during transfection / transduction by similar methods as disclosed in the previous section related to methods of cell culture. For example, the substrate comprising the array of grooves is a gas-permeable substrate and the device comprises one or more gas flow paths beneath the gas-permeable substrate. At least a portion of the required respiratory gases (including O2) is then supplied to the culture by flowing a respiratory gas feed (e.g. air + 10% CO2) through the gas flow path(s), and allowing the respiratory gases to diffuse through the gas-permeable substrate.

[0403] In some embodiments, the substrate comprising the array of grooves comprises, or consists of, polymethylpentene. The configuration of this substrate, including dimensions related to maximum thickness, groove depth, groove aspect ratio, and thickness within the grooves, may generally be according to any of the embodiments disclosed in the section relating to methods of culturing cells.

[0404] Once culturing of the cells is complete, the cultured cells may be harvested from the device. This may be accomplished by flowing an extraction medium through the flow channel and out of an outlet port for collection. In some embodiments, the extraction medium is flowed through the flow channel in a second flow direction that is along the grooves. This direction is preferred to efficiently wash the cultured cells along the grooves and out of the flow channel. The flow of extraction medium may beprovided at an average flow velocity of at least 10 mm / min and / or at a shear stress of at least 0.1 dyne / cm2, or at least 0.5 dyne / cm2, preferably at least 1 dyne / cm2, such as about 1 .5 dyne / cm2.

[0405] Referring to Figures 1 -10, for example, the microfluidic device used in a method of transfecting / transducing cells may be a multi-layer microfluidic device 100 configured for uni-directional or bi-directional flow, as previously described herein, having a cell capture composition physically adhered to hydrophobic upper surface 310 including in grooves 312. After immobilisation of target cells on the cell capture composition, the immobilised cells may optionally be released by introducing a cell release agent to the flow channel, either by cross flow or longitudinal flow, preferably by cross flow to minimise disruption of the released cells from the grooves.

[0406] A fluid containing an exogenous molecule is then introduced to flow channel 60, preferably from first inlet port 210, via first fluid inlets 212. The resultant flow of exogenous molecule is thus directed across grooves 312 in flow direction A, which is substantially orthogonal to the longitudinal dimension of the grooves. Waste product may be flowed out of flow channel 60 via first fluid outlets 222 and first outlet port 392. The target cells and the exogenous molecule are incubated in flow channel 60, including in grooves 312, under conditions to transfect / transduce the target cells with the exogenous molecule. During the immobilisation and transfection / transduction, a respiratory gas feed (e.g. air + 10% CO2) is flowed though gas flow path 410. Gas exchange across gas permeable flow layer 300 provides sufficient O2 in flow channel 60 to sustain the cells, and removes waste gases from the cell chamber.

[0407] Optionally, the target cells are cultured with a flow of culture medium, before and / or after transfecting / transducing the cells. In this case, culture medium is introduced to flow channel 60 from first inlet port 210, via first fluid inlets 212. The resultant flow of culture medium is directed across grooves 312 in flow direction A (cross flow). Waste product is flowed out of flow channel 60 via first fluid outlets 222 and first outlet port 392.

[0408] After transfection / transduction and optional subsequent culturing is complete, the cells are harvested from the device by flowing an extraction medium through the flow channel. Preferably, the extraction medium is flowed along thegrooves 312, as may be accomplished with the microfluidic devices described with reference to Figures 3-10. Thus, for example, the extraction medium is introduced to flow channel 60 from second inlet port 330, via second fluid inlets 332. The resultant flow of extraction fluid is directed along grooves 312 in flow direction B (longitudinal flow), with the harvested cells flowed out of flow channel 60 via second fluid outlets 342 and second outlet port 340 for collection.

[0409] The method disclosed herein advantageously allows target cells to be separated from a sample comprising both target and non-target cells, and subsequently transfected / transduced in and harvested from a single microfluidic device.Method of selecting cells

[0410] In another aspect, the invention relates to a method of selecting cells. The method comprises providing a microfluidic device comprising a flow channel located over a substrate comprising a hydrophobic upper surface. The hydrophobic upper surface comprises an array of grooves that are substantially parallel. Preferably, a cell capture composition is physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves. In some embodiments, the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

[0411] The cell capture composition may be according to any of the embodiments previously described, and may also be physically adhered to the hydrophobic upper surface in the grooves by any of the methods previously described. Preferably, the cell capture composition is physically adhered in each of the grooves, and may be present as a continuous coating over the hydrophobic upper surface including in the grooves and regions intermediate to the grooves.

[0412] In some embodiments, the microfluidic device is a multi-layer microfluidic device according to any embodiments previously described which allow flow along the grooves. In some embodiments, the microfluidic device with cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves is a bioreactor for cell culture according to any of the embodiments previously described.

[0413] The method of selecting cells comprises flowing a sample comprising target cells through the flow channel over the groove array in a second flow direction that is along the grooves under conditions to selectively immobilise the target cells on the cell capture composition.

[0414] Before introducing the sample into the groove array, the microfluidic device may first be primed by flowing culture medium into the device until it is filled.

[0415] The sample is flowed into the flow channel under conditions to selectively immobilise the target cells on the cell capture composition. The target cells may generally be cells as described in the previous section relating to methods of culturing cells. Typically, the sample will contain both target cells and other, non-target cells. Here, selective immobilisation means that the target cells are preferentially immobilised in comparison to the non-target cells. The target cells are immobilised when the antigen binding moiety of the cell capture composition specifically binds to the binding site of the target cell, and the immobilisation conditions are selected to allow such specific binding to proceed. The selective immobilisation of the target cells may be based on the expression of certain surface antigens on the target cells which are lacking in the non-target cells. Moreover, the amphiphilic polymer may block the adhesion of non- target cells to the hydrophobic surface, further enhancing the selectivity of immobilisation.

[0416] The selectivity of capture may be enhanced by a washing step conducted at a flow stress sufficient to displace non-specifically bound cells, suitably at least 0.1 dyne / cm2, such as between 0.1 and 10 dyne / cm2, for example or at least 0.5 dyne / cm2, or at least 1 .0 dyne / cm2, such as about 1 .5 dyne / cm2.

[0417] The method thus allows target cells to be selected from a complex biological sample. The cells may be present in bodily fluids (e.g. blood), processed or partially processed bodily fluids (e.g., blood processed to remove platelets and / or serum), or resuspended in an appropriate fluid medium (e.g., culture medium) before being introduced into the device.

[0418] The microfluidic device can be configured with fluid inlet and outlet ports. The sample may thus be flowed, continuously or non-continuously, through the fluid inlet port, then through the flow channel over the groove array in the second flowdirection that is along the grooves, and then out of the fluid outlet port. The target cells are selectively immobilised on the cell capture composition and the non-target cells are selectively removed from the flow channel via the fluid outlet port.

[0419] The method may be used to positively or negatively select cells. In positive cell selection methods, the target cells selectively immobilised on the cell capture composition are ultimately intended for harvesting and use, optionally after culturing and / or transfecting / transducing the target cells in the grooved flow channel according to methods disclosed herein. The selective immobilisation thus avoids or minimises the contamination of the desirable target cells with non-target cells, and allows productive use of the working area of the microfluidic device for capture and subsequent further processing of the target cells. Negative cell selection methods, by contrast, may be used to remove target cells from the sample, allowing the sample to be further processed or used without the target cells present, or with a reduced concentration of target cells relative to other useful cells in the sample.Positive cell selection methods

[0420] In some embodiments, the method of selecting cells further comprises releasing the target cells from the capture composition. The captured cells may be released from the antigen binding moiety using a cell release agent, for example, the proteolytic enzyme chymopapain, or neuraminidase. Alternatively, the captured cells are released from the antibody via competitive displacement, for example, the cell release agent may be a peptide which competitively binds to the antibody to displace the captured cells. The cells may be released prior to, during or after, preferably prior to or during, one or more subsequent processing steps, such as (i) culturing the target cells in the flow chamber, (ii) transfecting / transducing the target cells in the flow chamber and (iii) harvesting the target cells from the flow chamber.

[0421] In some embodiments, the target cells are cultured in the flow channel after being selectively immobilised therein. The method may thus include a step of flowing a culture medium through the flow channel over the groove array in a first flow direction that is across the grooves, and culturing the target cells in the presence of the culture medium. Preferably, the second flow direction is substantially orthogonal to the first flow direction. The step of cell culturing with a cross flow of culture medium maygenerally be conducted according to any of the embodiments previously disclosed in the section relating to methods of culturing cells.

[0422] In some embodiments, the target cells are transfected / transduced in the flow channel after being selectively immobilised therein. The method may thus include a step of introducing at least one exogenous molecule into the flow channel, and incubating the target cells and the at least one exogenous molecule under conditions to transfect / transduce the target cells with the exogenous molecule. Preferably, the at least one exogenous molecule is flowed through the flow channel over the groove array in a first flow direction that is across the grooves. The second flow direction is preferably substantially orthogonal to the first flow direction. The step of transfecting / transducing the cells with exogenous molecule may generally be conducted according to any of the embodiments previously disclosed in the section relating to methods of transfecting / transducing cells.

[0423] Respiratory gases may be supplied to the cells during capture and subsequent processing in the device by similar methods as disclosed in the previous section related to methods of cell culture. For example, the substrate comprising the array of grooves is a gas-permeable substrate and the device comprises one or more gas flow paths beneath the gas-permeable substrate. At least a portion of the required respiratory gases is then supplied to the culture by flowing a respiratory gas feed through the gas flow path(s), and allowing the respiratory gases (including O2) to diffuse through the gas-permeable substrate.

[0424] In some embodiments, the substrate comprising the array of grooves comprises, or consists of, polymethylpentene. The configuration of this substrate, including dimensions related to maximum thickness, groove depth, groove aspect ratio, and thickness within the grooves, may generally be according to any of the embodiments disclosed in the section relating to methods of culturing cells.

[0425] The method of selecting cells may include a step of harvesting the target cells, optionally after culturing and / or transfecting / transducing the target cells in the flow channel. The method may thus include a step of harvesting the target cells by flowing an extraction medium through the flow channel. Preferably, the extraction medium is flowed through the flow channel in the second flow direction, i.e. along the grooves(longitudinal flow). This direction is preferred to efficiently wash the cultured cells along the grooves and out of the flow channel. The flow of extraction medium may be provided at an average flow velocity of at least 10 mm / min and / or at a shear stress of at least 0.1 dyne / cm2, or at least 0.5 dyne / cm2, preferably at least 1 dyne / cm2, such as about 1 .5 dyne / cm2.

[0426] Referring to Figures 3-10, for example, the microfluidic device used in a method of selecting cells may be a microfluidic device 100 configured for bi-directional flow, as previously described herein, having a cell capture composition physically adhered to hydrophobic upper surface 310 including in grooves 312. A sample comprising target cells is introduced to flow channel 60 from second inlet port 330, via second fluid inlets 332. The resultant flow of cells is directed along grooves 312 in flow direction B (longitudinal flow). The target cells introduced in this way are thus selectively immobilised on the cell capture composition, including (or primarily) in grooves 312. Advantageously, the sample is delivered uniformly to each groove in the array with this approach, resulting in an efficient usage of the device’s working area for cell selection. Waste product, typically including non-target cells which are nonadherent to the cell capture composition, is flowed out of flow channel via second fluid outlets 342 and second outlet port 340 during the cell selection process.

[0427] During the cell capture process, a respiratory gas feed (e.g. air + 10% CO2) may be flowed though gas flow path 410. Gas exchange across gas permeable flow layer 300 provides sufficient O2 in flow channel 60 to sustain the cells and remove waste gases from the cell chamber.

[0428] After removal of the non-target cells, the selectively immobilised target cells may optionally be released by introducing a cell release agent to the flow channel, either by cross flow or longitudinal flow, preferably by cross flow to minimise disruption of the released cells from the grooves.

[0429] If the target cells are to be cultured in the device, culture medium is introduced to flow channel 60 from first inlet port 210, via first fluid inlets 212. The resultant flow of culture medium is directed across grooves 312 in flow direction A (cross flow), which is substantially orthogonal to the longitudinal dimension of the grooves. Waste product is flowed out of flow channel 60 via first fluid outlets 222 andfirst outlet port 392. During the culturing, the respiratory gas feed continues to flow though gas flow path 410, allowing gas permeation through gas permeable flow layer 300 to sustain the cells.

[0430] If the target cells are to be transfected / transduced in the device, a fluid containing an exogenous molecule is introduced to flow channel 60, preferably from first inlet port 210, via first fluid inlets 212. The resultant flow of exogenous molecule is directed across grooves 312 in flow direction A (cross flow), which is substantially orthogonal to the longitudinal dimension of the grooves. Waste product may be flowed out of flow channel 60 via first fluid outlets 222 and first outlet port 392. The target cells and the exogenous molecule are incubated in flow channel 60, including in grooves 312, under conditions to transfect / transduce the target cells with the exogenous molecule. During the transfection / transduction, the respiratory gas feed continues to flow though gas flow path 410, allowing gas permeation through gas permeable flow layer 300 to sustain the cells. Optionally, the target cells are also cultured with a flow of culture medium, as disclosed herein, either before or after transfecting / transducing the target cells.

[0431] After the selective immobilisation, and the optional culturing and transfection / transduction steps, the cells are harvested from the device by flowing an extraction medium into flow channel 60 from second inlet port 330, via second fluid inlets 332. The resultant flow of extraction fluid is directed along grooves 312 in flow direction B (longitudinal flow), with the harvested cells flowed out of flow channel 60 via second fluid outlets 342 and second outlet port 340 for collection.

[0432] The present disclosure is described further in the following non-limiting examples.ExamplesExample 1. Functionalisation of poloxamer with conjugating groups.

[0433] Pluronic F-127 (poloxamer; “Pluronic”) was terminally functionalised by two alternative methods described below in Scheme 1 and Scheme 2.

[0434] In a first method, Pluronic F-127 (poloxamer; “Pluronic” in Scheme 1 ) was terminally functionalised with N-hydroxy succinimide (NHS) carbonate groups, using N,N-disuccinimyl carbonate as reagent, by the methodology shown in Scheme 1 , where x = 100 and y = 65.Scheme 1 .

[0435] To a dry 50mL Schlenk tube equipped with a magnetic stirrer were added Pluronic F-127 (500mg) and dichloromethane (DCM, 6mL) under nitrogen at room temperature. Then, 40 equiv N,N-disuccinimidyl carbonate was added as a suspension in acetonitrile (3mL) , followed by addition of dry pyridine (40 equiv, 120pL). After 30min, a further portion of pyridine was added (80 equiv, 240pL). The reaction was maintained with stirring at room temperature for 22h. The mixture was transferred to a 50mL Falcon tube, and the solvent was evaporated to reach approximately 5mL and then precipitated with cold diethyl ether (40mL). The tube was centrifuged for 10min at 4000rpm. The ethereal phase was separated, and the solid was dissolved in dichloromethane (5mL) and reprecipitated. The process was repeated three additional times, and the solid was dried under vacuum to provide pluronic-NHS. The structure and acceptable purity of the Pluronic-NHS product was verified using NMR.

[0436] In a second method, Pluronic F-127 (poloxamer; “Pluronic” in Scheme 2) was terminally functionalised with azide groups by the methodology shown in Scheme 2, where x = 100 and y = 65.Scheme 2.

[0437] Pluronic F-127 and triethylamine were combined in DCM (15mL) under N2 and cooled to 0°C. Tosyl chloride (166mg, 5.5 eq.) was then added and the reaction stirred at 0°C for 30 minutes then at 50°C for 36h. The solution was added dropwise to ice-cold diethyl ether and the solid collected by centrifugation. This was repeated 3x with a 5-fold excess of diethyl ether to DCM. Solid was collected and dried under high vacuum. The crude solid was purified by dialysis (3500kDa dialysis tubing, 1 :1 MeCN:water), with the solvent being changed 4 times: three times every 1 ,5h and the fourth being left overnight. The acetonitrile was removed and the resulting viscous liquid freeze-dried to give the Pluronic-OTs intermediate as white solid.

[0438] Pluronic-OTs (300mg, 23pmol) and NaNa (30mg, 20 eq.) were combined in anhydrous dimethylformamide (DMF, 10mL) and heated at 90°C for 60h. The solution was concentrated under a stream of nitrogen to give a yellow oil. This was dissolved in DCM (40mL) and washed with brine (50mL x 2). The organic layer was separated and dried over magnesium sulfate. The solvent was removed under reduced pressure andunder high vacuum to give an off-white solid to provide pluronic-Na. The structure and acceptable purity of the Pluronic-Na product was verified using NMR.Example 2. Preparation of poloxamer-antibodv conjugates by NHS chemistry.

[0439] Poloxamer-antibody conjugates (Pluronic-mAb) were prepared by the methodology depicted in Scheme 3, where x = 100 and y = 65 and mAb is (i) an OKT3 clone, a CD3 monoclonal antibody, (ii) MY10, a CD34 monoclonal antibody, (iii) 9C5, a CD34 monoclonal antibody or (iv) QBEND10, a CD34 monoclonal antibody.Scheme 3.

[0440] OKT3 (1 mg / ml) and pluronic-NHS (10 mg / ml in DBPS) were combined in 1 .5 ml Eppendorf tubes at molar ratios of OKT3 to pluronic: 1 :5, 1 :10, 1 :20, 1 :40 and 1 :80. Samples were reacted at room temperature overnight, then purified by spin filtration. Sample volumes were made up to 2 mL with Dulbecco's Phosphate Buffered Saline (DPBS) and then spun down three times at 3000rcf for 15 minutes. The final concentration of the samples was measured using Nanodrop, the volume was adjusted to reach a concentration of 1 mg / ml for use in experiments. Samples were stored at 4°C until ready for characterisation via SDS-PAGE and cell separation assays.

[0441] To confirm the successful modification of an antibody with pluronic, SDS- PAGE was run on the conjugates using 4-12% Bis T ris Gels (ThermoFisher Scientific). Gels were run at 180V for 45 minutes and stained with Coomassie Blue and de-stainedovernight in MQ water. The appearance of new protein bands demonstrates the change in protein structure following pluronic conjugation to the antibody, consistent with an increase in molecular weight and quantitative (100%) conversion of antibody to antibody-pluronic conjugate (represented as Pluronic-mAb, Scheme 3). A gradual narrowing of the protein bands with increasing pluronic-NHS indicates that there is a more uniform labelling of pluronic-NHS to OKT3 antibodies with higher excess of pluronic-NHS molecules.

[0442] MY1 0 (1 mg / ml) and pluronic-NHS (10 mg / ml in DBPS) were conjugated and evaluated by the same methodology, at molar ratios of MY10 to pluronic: 1 :10 and 1 :20. The results demonstrated successful conjugation of the MY10 antibody, albeit at lower conversions: in the range of 45%-47% conversion of antibody to antibody-pluronic conjugate.

[0443] 9C5 (1 mg / ml) and pluronic-NHS (10 mg / ml in DBPS) were conjugated and evaluated by the same methodology, at molar ratios of 9C5 to pluronic: 1 :20. The results demonstrated successful conjugation of the 9C5 antibody at quantitative (100%) conversion of antibody to antibody-pluronic conjugate.

[0444] QBEND10 (2.73mg / ml) and pluronic-NHS (10 mg / ml in DBPS) were conjugated and evaluated by the same methodology, at molar ratios of QBEND10 to pluronic: 1 :10 and 1 :20. The results demonstrated successful conjugation of the QBEND10 antibody at quantitative (100%) conversion of antibody to antibody-pluronic conjugate.

[0445] Table 1 provides a summary of the antibody-polymer conjugates (Pluronic- mAb) prepared by NHS conjugation chemistry.Table 1.Example 3. Preparation of poloxamer-antibodv conjugates by click chemistry.

[0446] Click-reactive monoclonal antibodies (DBCO-labelled monoclonal antibody) were prepared by the methodology shown in Scheme 4, where mAb is (i) an OKT3 clone, a CD3 monoclonal antibody, or (ii) MY10, a CD34 monoclonal antibody. The click-reactive DBCO-labelled antibodies are named OKT3-DBCO mAb and MY10- DMCO mAb, and were produced with different degree of labelling, being the molar ratio of click labelling reagent (DBCO-PEG4-NHS) to mAb used in the synthesis.(DBCO-labeled monoclonal antibody)Scheme 4.

[0447] Poloxamer-antibody conjugates were then prepared by the methodology shown in Scheme 5, where x = 100 and y = 65.Scheme 5

[0448] OKT3-DBCO mAb with a degree of labelling of 2.0 (1 mg / ml) and pluronic- N3 (10 mg / ml in DBPS) were combined in 1 .5 ml Eppendorf tubes at the molar ratios of OKT3 to pluronic: 1 :5, 1 :10, 1 :20, and 1 :40. Samples were reacted at room temperature overnight, then purified by spin filtration. Sample volumes were made up to 2 mL with DPBS and then spun down three times at 3000 ref for 15 minutes. The final concentration of the samples was measured using Nanodrop to reach a concentration of 1 mg / ml for use in experiments. Samples were stored at 4°C until ready for characterisation via SDS-PAGE and cell adhesion assays.

[0449] To confirm successful conjugation of pluronic-OKT3 mAb via click chemistry, SDS-PAGE was run on the conjugates using a 4-12% Bis Tris Gels (ThermoFisher Scientific). Gels were run at 180V for 45 minutes and stained with Coomassie Blue and de-stained overnight in MQ water. The results demonstrated successful conjugation of the antibody by click chemistry, albeit at lower conversions: in the range of 65%-70% conversion of antibody to antibody-pluronic conjugate regardless of the reaction ratio.

[0450] MY10-DBCO mAb with a degree of labelling of 2.9 or 4.9 (1 mg / ml) and pluronic-Na (10 mg / ml in DBPS) were conjugated and evaluated by the same methodology, at molar ratio of MY10 to pluronic: 1 :20. The SDS-PAGE resultsdemonstrated successful conjugation of the MY10 antibody by click chemistry, providing 66% and 82% conversion of antibody to antibody-pluronic conjugate, respectively.

[0451] In another experiment to investigate the effect of time on conjugation, MY10- DBCO mAb with a degree of labelling of 1.1 (1 mg / ml) and pluronic-Na (10 mg / ml in DBPS) was combined in 1.5 ml Eppendorf tubes at a molar ratio of MY10 to pluronic: 1 :20. A time course reaction was conducted where the samples were reacted in room temperature for 1 , 2, 4, 8 and 24 hours. The SDS-PAGE results demonstrated successful conjugation of the MY10 antibody by click chemistry, with maximum conjugation achieved by 8 hours.

[0452] Table 2 provides a summary of the antibody-polymer conjugates prepared by click conjugation chemistry.Table 2.Example 4. Preparation of cell linesKG1 a cells (CD34+) and Jurkat mCherry cells (CD3+)

[0453] Human cell lines grown in tissue culture provide a source of pure phenotype for the modelling of cell separation. KG1 a is a promyeloblast macrophage cell line that expresses CD34+ antigen. Thus, KG1 a cells were utilized as a model for hematopoietic stem cells (HSC), which are CD34+ antigen presenting. Jurkat mCherry cells (T-cells) are T-cells which are CD3+ antigen presenting, and may for example be utilized for the purpose of researching optimum cell separation methods in the application of CAR-T therapies.

[0454] Mixtures of KG1 a cells and Jurkat mCherry cells were thus used in some cell separation experiments targeting (i) selective CD34+ cell capture (with KG1 a cells being the target cells and Jurkat mCherry cells being the negative cells); and (ii) selective CD3+ cell capture (with Jurkat mCherry cells being the target cells and KG1 a cells being the negative cells); and as a suitable cell line model to demonstrate separation of complex cell mixtures.

[0455] The cell lines were grown in suspension culture in media which were made up sterile and filtered, and only cell lines with passage number below 20 were used due to loss of antigen expression with increasing passage numbers.

[0456] KG1 a cell lines were grown in suspension culture in the following medium: Basal media: IMDMI; Supplement: 20% fetal calf serum; 1 % Penicillin-Streptomycin.

[0457] The KG1 a cells were harvested during mid log-growth phase. KG1 a cells were typically stained with H33342 Bisbenzimide 3HCL4H2O fluorochrome at a concentration of 5pg / mL at 37°C for 20 minutes before washing in DPBS with 0.5% BSA via centrifugation (300g for 5 minutes) for 2 cycles.

[0458] Jurkat mCherry cell lines were grown in suspension culture in the following medium: Basal media: RPMI-1640; Supplement: 10% fetal calf serum; 1% Penicillin- Streptomycin.

[0459] The Jurkat mCherry cells were harvested during mid log-growth phase and used unstained.Buffy coat (negative cells)

[0460] In further experiments, treated buffy coat samples was used to emulate negative cell populations in primary cell sources. Human buffy coat samples were provided by Red Cross Australia. The buffy coat samples were diluted to a volume of 180 mL in Buffer A, consisting of Dulbecco's phosphate-buffered saline (DPBS) supplemented with 0.5% bovine serum albumin (BSA). Individual 50 mL falcon tubes were prepared by adding 15 mL of lymphoprep™ media (Stem cell Technologies), followed by gentle pipetting of 30 mL of the diluted buffy coat sample into each tube. The tubes were then subjected to centrifugation at 1200 xg for 30 minutes without applying the brake. After centrifugation, the supernatant was meticulously discarded, and the remaining pellet underwent two washes by diluting it to a final volume of 50 mL in Buffer A, followed by centrifugation at 300 xg for 10 minutes. To eliminate platelets, the sample was washed twice in Buffer A, subjecting it to centrifugation at 130 xg for 15 minutes. The resulting cell pellet was resuspended in 1 mL of Buffer A, and subsequently incubated with 10 mL of ammonium chloride solution (Stem cell Technologies) on a gentle rocker at room temperature for 10 minutes. Following this, the cells were washed twice in Buffer B (DPBS + 0.5% BSA + 2mM EDTA), with centrifugation at 300 xg for 5 minutes. The final cell suspension was prepared in Buffer B.Mobilized Peripheral Blood (MPB) CD34+ cells

[0461] Cyropreserved Mobilized Peripheral Blood (MPB) CD34+ cells isolated using MACS beads (Miltenyi), were used as a positive cell as experimental protocols progressed from using cell lines to primary cells. MPB CD34+ cells were typically mixed in with buffy coat cells (simulating negative cell populations).

[0462] Fresh MPB CD34+ cells isolated using MACS beads isolated from apheresis product were also used in cell selection experiments. These cells were thawed 1 day prior to experiment, and cultured overnight in CellGenix® GMP SCGM + 2% human serum Albumin + 100ng / mL recombinant human stem cell factor (rhSCF, PeproTech), 100ng / ml rh thrombopoietin (TPO, PeproTech), 100ng / mL rh FLT3 receptor ligand (FLT3L, PeproTech).Example 5. General methodology for well plate cell separations methodology

[0463] The following general procedure was used to investigate cell separation on cell capture compositions present in well plates (as implemented in subsequent examples). 48 well plates are a suitable model system for initial optimisation of surface chemistry for microfluidic devices.

[0464] Well plate experiments were conducted using non-tissue culture treated 48- well plates, with a polystyrene surface or a polymethylpentene (TPX) surface unless otherwise specified. Each well plate was treated with 200pL of prepared samples of cell capture composition (concentration of protein varied for different experiments) and allowed to adsorb for 1 hour at room temperature. Simultaneously, cells were prepared following the Cell Culture Protocol (Example 4). After the surface chemistries had adsorbed for 1 hour, the well plates were washed twice with 500pL of DPBS.

[0465] Cells were mixed at a fixed ratio of negative: positive cells at a concentration of 5E6 cells / mL. Cell counts were determined using a Vi-Cell counter before seeding 150pL of the cell mixture onto each surface-treated well plate. The well plate was then left for 10 minutes on a gentle shaker. Subsequently, the well plates were washed with 500pL of Buffer A, consisting of DPBS supplemented with 2% fetal bovine serum (FBS). This washing step was repeated twice, and the washed, unbound cells were collected for analysis.

[0466] From the collected sample, 500pL was taken for Vi-Cell counts. The remaining volume of the sample was processed for flow cytometry by centrifuging at 300xg for 5 minutes. The supernatant was carefully discarded, and the remaining cell pellet was resuspended in 200pL of Buffer A. After this, 1 pL of the relevant primary monoclonal antibody (mAb) was used to stain the sample for 20 minutes on ice. Flow tubes were washed twice with 2mL of DPBS and centrifuged at 300xg for 5 minutes to remove any unbound antibodies.

[0467] The cells were resuspended in 200pL of Buffer A and a secondary goat antimouse antibody (APC ALEXA 647) was used to stain the sample for 20 minutes on ice. The sample was washed twice using 2mL of Buffer A, and finally suspended in 200pL of DPBS for flow cytometry.Example 6. Competitive adhesion experiments to determine plu-mAb affinity to hydrophobic substrates

[0468] This example sought to demonstrate that conjugating an antibody to a pluronic molecule can increase the affinity of the antibody to hydrophobic substrates, whilst supporting the natural conformation of the antibody.

[0469] A well plate assay was employed (per Example 5 general procedure), where different pluronic concentrations (range of 0 to 0.5% w / v in MQ water, DF = 10) were mixed at a 1 :1 volume ratio with either antibody or pluronic-antibody conjugate. The mixed samples were seeded onto a non-tissue culture treated 48-polystyrene well plate for 1 hour of physical adsorption in room temperature. The treated surfaces were washed with DPBS twice, and CD34+ cells were seeded onto the surfaces and allowed to bind for 10 minutes. The wells were then washed with DPBS twice, and after adding 200ul of DPBS + 0.5% BSA to the bound cells, the samples were imaged using a fluorescent microscope. A cell counting software was employed to count the number of positive and negative cells remaining on the surface, to determine the specific cell selectivity of the treated surfaces.

[0470] The results of an experiment comparing the effects of mixing pluronic with either i) unconjugated MY10 or ii) pluronic-MY10 mAb conjugate (prepared by NHS conjugation chemistry per Scheme 3 with mAb:plu ratio = 1 :40 mokmol; conversion 70% and used at 200ug / ml) were examined and are shown in Figure 16 and Figure 17 respectively. When pluronic was mixed with unconjugated MY10, there was a significant decrease in the capture of positive cells (target KG1 a cells) compared to mixing with pluronic-MY10 conjugate, indicating that pluronic inhibited the binding of antibodies to the surface. However, when pluronic-MY10 conjugate was used instead of MY10, the inhibitory effect of free pluronic was reduced. This suggests that pluronic- MY10 conjugate has an affinity for the surface that is greater than or similar to the affinity of free pluronic to a hydrophobic substrate. This was supported by the increased capture of positive cells when pluronic-MY10 conjugate was mixed with different concentrations of pluronic compared to unconjugated MY10 mixed with free pluronic. Even at high concentrations of pluronic, there was still a significant increase in positive cell capture, indicating that pluronic-MY10 conjugate may displace some pluronic molecules on the hydrophobic surface. Throughout the experiment, there was consistently low binding of negative cells (non-target Jurkat cells), as both pluronic and plu-MY10 conjugate effectively prevented non-specific cell binding.

[0471] The same experiment was conducted using a CD3 antibody (OKT3 clone) instead of a CD34 antibody, where Jurkat mCherry cells were used as positive cells (target CD3+ cells) and Hoechst33342 stained KG1 a cells were used as negative cells (non-target cells) in a 1 :1 mix. Results shown in Figure 18 (unconjugated OKT3) and Figure 19 (pluronic-OTK3 conjugates; prepared by click chemistry per Scheme 5 with mAb:plu ratio = 1 :20 mokmol; conversion 50% and used at 50ug / ml) indicate similar effects as observed using MY10 and pluronic-MY10 conjugates. In Figure 20, a clear shift is observed between the curve (of positive captured Jurkat cells) representing competition of pluronic with unconjugated OKT3 antibody (labelled “OKT3”), against the competition of pluronic with pluronic-OKT3 conjugates (labelled “OKT3-pluronic”). The results indicate a higher affinity of the OKT3-pluronic conjugate to the substrate than unconjugated OKT3.Example 7. Cell capture and expansion on polymethylpentene (TPX) substrates comprising cell capture compositions

[0472] A well plate assay (per Example 5 general procedure) was employed to assess the suitability of polymethylpentene (TPX) as a cell capture and expansion surface, as well as determine if pluronic-mAb has no cytotoxic properties. TPX well plates were fashioned using a PDMS cast well plate mould, which was then pressed onto a blank TPX sheet - the two materials are held together by the forces between them. No pressure or heat was applied when making these well plates.

[0473] The polymethylpentene wells (non-tissue culture treated) were surface treated with different cell capture compositions which were adsorbed onto the surface for 1 hour at room temperature. The cell capture compositions were: (1 ) blank (no surface treatment), (2) 1 % pluronic, (3) unconjugated MY10 (200 pg / ml), (4) unconjugated MY10 (200 pg / ml) followed by pluronic, (5) pluronic-MY10 conjugate (prepared by NHS conjugation chemistry per Scheme 3 with mAb:plu ratio = 1 :80 mokmol; conversion 76% and used at 200ug / ml), (6) pluronic-MY10 conjugate followed by free pluronic.

[0474] A 1 :1 mix of positive and negative cells (Hoechst33342 stained KG1 a cells and Jurkat mCherry cell lines) were then seeded into the treated wells. After washing away any unbound cells, the remaining cells on the surface were imaged using afluorescent microscope and counted using a cell counting software. The results shown in Figure 21 demonstrate that TPX is an appropriate hydrophobic substrate for selective capture of target cells using unconjugated antibody + pluronic mixtures or pluronic- antibody conjugates.

[0475] The cells captured on these cell capture compositions were also left for 85 hours of culture. Live cell images were obtained over this time, and final cell counts obtained at the end of the 85-hour cell culture period. The results shown in Figure 22 show successful expansion of cells, and no issues with cell viability when cells were left to expand following capture via all surface treatments.Example 8. Comparative cell capture on polymethylpentene vs polystyrene substrates comprising cell capture compositions

[0476] Further well plate assays (per Example 5 general procedure) were conducted to compare cell capture on different hydrophobic substrates, specifically polystyrene well plates vs polymethylpentene well plates (as described in Example 7). The cell capture compositions evaluated were: (1 ) unconjugated MY10, and (2) unconjugated MY10 followed by free pluronic.

[0477] A 1 :1 mix of positive target cells and negative non-target cells (Hoechst33342 stained KG1 a cells and Jurkat mCherry cell lines) were then seeded into the treated wells. After washing away any unbound cells, the remaining cells on the surface were imaged using a fluorescent microscope and counted using a cell counting software. The results are shown in Figure 23. A greater number of both positive and negative cells were captured on TPX substrate compared to polystyrene substrate, with the unmodified antibody providing unsatisfactory blocking of nonspecific binding. Greatly improved selectivity was obtained by using a cell capture composition comprising both unconjugated MY10 and pluronic. The results highlight the importance of effective strategies to block non-specific binding of negative nontarget cells on TPX surfaces.Example 9. General methodology for flowcell cell separations

[0478] The following general procedures were used to investigate cell separation on cell capture compositions present in a flowcell (as implemented in the subsequentexamples). Scaling up from well plate adhesion assays, flowcell cell separation assays were employed to determine the effect of fluid shear stress on the cell selection environment. Furthermore, it was critical to assess the system with this critical parameter, to understand the translatability of the surface chemistry to a large-scale microbioreactor system.

[0479] An experimental flowcell system was designed using an uncoated sticky slide (p-Slide VI 0.4; commercially available from ibidi; https: / / ibidi.com / channel- slides / 57--slide-vi-04.html). The sticky-slide family allows for experiments using custom-specific bottom materials such as plastic sheets, glass slides etc. The self adhesive (“sticky”) underside of the bottomless blank slide allows easy adaptation to specific substrates by sealing with pressure and heat. The six-channel format (defining 6 parallel flow microchannels with 30 pl channel volume, 0.44 mm channel height, 17 mm channel length 3.8 mm channel width, 2 Luer ports) when adhered to a substrate used in the experimental setup allowed for testing various experimental conditions in parallel, and the female leur-lock connectors allowed observation and analysis of the system under flow conditions.

[0480] After attaching the sticky slides to the substrate, the surface of the flow cell was primed with DPBS to create a suitable fluid environment. Following this, each channel was treated with the appropriate surface chemistry (aqueous composition comprising cell capture composition) with a 200ul syringe pull using a 1 ml plastic syringe with a male luer adapter to connect to the ibidi system. The cell capture composition was left to adsorb to the surface for 1 hour at room temperature. After this, the flow cell system was assembled by connecting the tygon tubing to a syringe pump and Hamilton syringe for the respective tasks of washing and loading cells onto the surface. Following 1 hour of physical adsorption of the cell capture composition, any unbound proteins were washed away by flowing DPBS through the flow cell using the syringe pump.

[0481] This process was repeated if the surface required more than one treatment to produce the cell capture composition (e.g. for unconjugated mAb followed by pluronic, the surface was adsorbed with mAb for 1 hour, washed, and then adsorbed with pluronic for 1 hour and washed again).

[0482] After surface treatment, cells with varying ratios of negative non-target cells to positive target cells were introduced into the flow cell by applying reverse pressure through a 200pL Hamilton syringe as schematically shown in Figure 24. Cells were left to settle for 4 minutes, corresponding to a cell sedimentation rate of 60 seconds per 100pm, to allow for maximum contact time between cells and the substrate to be treated with cell capture composition.

[0483] A controlled flow rate of DPBS was then applied to the system via the syringe pump to wash away unbound cells, leaving only the selectively captured cells on the surface. Throughout the washing process, fluorescent microscopy was used to capture images. A timelapse imaging protocol was established to document the cells before, during, and after the washing process. The collected images were processed using MATLAB, utilising custom cell counting software, designed specifically for this purpose.Example 10. Positive cell capture in a flow cell with different cell capture compositions.

[0484] Flowcell cell capture experiments were conducted (per Example 9 general procedure) using flowcells with hydrophobic polystyrene substrates treated with one of the sequence of treatments shown in Table 3. The MY10-pluronic conjugate was prepared by NHS conjugation chemistry per Scheme 3 with mAb:plu ratio = 1 :20 mokmol and used at 200ug / ml.Table 3.

[0485] Initially, only positive target cells (Hoechst33342 stained KG1 a cells) were used. The flowcell residence time during cell adhesion was 240 s. The removal of unbound cells was conducted at high flow stress, using a shear stress of c.a. 1.5 dynes / cm2. The results are shown in Figure 25. A significant increase in % capture of cells was observed for both MY10+pluronic (i.e. sequential treatments of unconjugated MY10, then free pluronic) and pluronic-MY10 conjugate treated flow cells, when compared against unconjugated antibody (MY10). The highest overall % capture of target CD34+ cells were shown in flow cell conditions treated with pluronic-MY10 conjugate alone and pluronic-MY10 conjugate followed by free pluronic. Conjugation of the antibody to the pluronic thus provides an improved result in comparison to separate addition of antibody followed by pluronic blocking agent.Example 11. Cell selection in a flow cell with polystyrene or polymethylpentene substrates and different cell capture compositions.

[0486] A flowcell cell selection experiment was conducted (per Example 9 general procedure) using flowcells with either a hydrophobic polystyrene substrate or hydrophobic polymethylpentene substrate. The substrates were treated with one of the sequence of treatments shown in Table 3 (but with unconjugated MY10 and pluronic- MY10 conjugate concentrations of 50 pg / mL). The MY10-pluronic conjugate was AL3 (Example 2, Table 1 ), prepared by NHS conjugation chemistry. After each treatment, the flow cells were washed using DPBS at a flow rate of 0.350 mL / min, which corresponds to a fluid shear stress of 0.2 dynes / cm2.

[0487] A 1 :1 mix of positive target KG1 a cells stained with Hoechst33342 at 5ug / ml and negative non-target Jurkat mCherry cells, at a total concentration of 1 .3E6 cells / ml, was used. The cells were introduced by pulling in 100 pL of the cell mix using a Hamilton Syringe. The cells were allowed to settle for a residence time of 240 seconds, and then washed with the same flow rate of 0.350 mL / min (fluid shear stress of 0.2 dynes / cm2) used for washing away the unbound protein in the capture composition deposition steps. During the washing phase, images were taken using a fluorescent microscope every 30 seconds at different locations along the flow cell. These images were laterused for counting the absolute number of positive target cells and negative non-target cells before and after cell separation using cell counting software.

[0488] The results for positive cell capture and negative cell capture are shown in Figure 26 and Figure 27 respectively. Flow cells treated with pluronic-MY10 conjugate (AL3) and the pluronic-MY10 conjugate followed by free pluronic captured a higher number of specifically bound positive cells compared to flow cells treated with unconjugated MY10 antibodies only. Surface treatment with unconjugated MY10 mAb followed by free pluronic produced a much smaller increase in positive cell selection. A low number of negative cells were observed to bind across all treatment conditions. This finding suggests that the combination of shear stress and the applied surface treatments effectively prevent the capture of negative cells within the system.

[0489] The results were similar for polystyrene and polymethylpentene (TPX) substrates, demonstrating the potential of polymethylpentene-based devices and confirming that polystyrene is a suitable proof of concept material for to model functionalisation of a polymethylpentene surface.Example 12. Cell selection in a flow cell treated with free pluronic followed by antibody compositions.

[0490] Further flowcell cell selection experiments were conducted (per Example 9 general procedure) where the order of pluronic addition was reversed compared to Example 11. Flowcells with hydrophobic polymethylpentene substrate were thus treated with one of the following treatment protocols: i. an unmodified flowcell without any pre-surface treatment (blank / control), ii. with only free pluronic treatment (pluronic; 1 % w / v), iii. with a first treatment of free pluronic at various different concentrations (ranging from 1 % w / v to 0.00001 % w / v), followed by a second treatment of My10 antibody (50 pg / mL) (Plu + MY10), iv. with a first treatment of free pluronic at various different pluronic solution concentrations (ranging from 1 % w / v to 0.00001 % w / v), followed by a second treatment of the pluronic-My10 conjugate AL3 (50 pg / mL) (Plu + plu-MY10); v. with only pluronic-My10 conjugate (50 pg / mL) (pluMY10); andvi. with a first treatment of pluronic-My10 conjugate followed by a second treatment of free pluronic (1 % w / v) (pluMy + plu).

[0491] Each reagent was exposed to the flowcell surface at room temperature for 1 hour, with subsequent washing of the flowcell with DPBS at a flow rate of 0.350 mL / min.

[0492] Figure 28 shows results from an experiment conducted using a 1 :1 mix of positive target KG1 a cells stained with Hoechst33342 at 5ug / ml and negative nontarget Jurkat mCherry cells, at a total concentration of 1.3E6 cells / ml, with the same cell capture procedure as Example 1 1 . The presence of pluronic as a pre-deposited agent, even at the lowest concentration (0.00001 % w / v) acted as a blocking agent to the adsorption of an unconjugated MY10 antibody. Contrastingly, when the pluronic- MY10 conjugate was used instead, a substantial increase in positive cell capture was observed. This suggests the existence of an interaction between the initially adsorbed pluronic layer and the subsequent layer of pluronic-MY10 conjugate introduced into the system which allows the antiCD34 to target the binding sites on the CD34+ cells. Notably, minimal binding of negative non-target cells was observed across all experimental treatments, except for the control group with a blank unmodified TPX substrate.Example 13. Cell selection in a flow cell treated with pluronic-antibodv conjugates prepared by -NHS or click chemistry

[0493] Further flowcell cell selection experiments were conducted (per Example 9 general procedure) to investigate whether pluronic-antibody conjugates produced by NHS chemistry (as described in Example 2) or click chemistry (as described in Example 3) have different effectiveness as a cell capture composition. Flowcells with hydrophobic polymethylpentene substrate were thus treated with one of the following pluronic-MY10 conjugates: AL2 or AL3 (NHS chemistry, Table 1 ), or FZ1 or FZ2 (click chemistry, Table 2). Each reagent was applied at room temperature for 1 hour (50 pg / mL), with subsequent washing of the flowcell with DPBS at a flow rate of 0.350 mL / min.

[0494] Figure 29 shows results from experiments conducted using a 1 :1 mix of positive target KG1 a cells stained with Hoechst33342 at 5ug / ml and negative nontarget Jurkat mCherry cells, at a total concentration of 1.3E6 cells / ml, with the sameprocedure as Example 1 1 . The results indicate that the conjugation method, whether - NHS or azide-DBCO click chemistry, has no significant impact on the performance of the conjugate when tested shortly after synthesis. Furthermore, no significant difference was evident between sample AL2 and AL3, which were synthesised via NHS chemistry using an antibody to pluronic ratio of 1 :10 and 1 :20, respectively. Moreover, no significant difference was evident between sample FZ1 and FZ2, which were synthesised via click chemistry with a degree of labelling (DOL) of 2.9 and 4.9, respectively (click-reactive labels on the antibody prior to click chemistry). Evidently, lower molar ratios for pluronic-NHS synthesis, and lower DOL of mAb for click chemistry, can be used without effecting the ability of the conjugates to capture positive target cells.Example 14. Effect of fluid shear stress on cell separation

[0495] Further flowcell cell selection experiments were conducted (per Example 9 general procedure) to investigate the effect of low and high shear stress on the enrichment of positive cells from a 1 :7 ratio of positive to negative cell population.

[0496] Flowcells with either hydrophobic polystyrene or hydrophobic polymethylpentene substrates were thus treated with the CD34 monoclonal antibody QBEND10 pluronic conjugate (plu-QBEND10; sample ALQ1 per Example 2, Table 1 ), optionally with a pre-treatment or post-treatment with free pluronic. Each reagent was exposed to the flowcell surface at room temperature for 1 hour, with subsequent washing of the flowcell with DPBS at a flow rate of either 0.350 mL / min (low shear stress; 0.2 dynes / cm2) or 2 mL / min (high shear stress; 1.5 dynes / cm2). Cell capture compositions were produced by one of the sequence of treatments shown in Table 4.Table 4.

[0497] In this study, MACS separated CD34+ cells, thawed, cultured overnight, and stained with Hoechst33342 (5ug / ml) were utilised as the positive target cell population, while processed buffy coat cells served as the negative non-target cell population (both as described in Example 4). A 1 :7 mix of positive cells at 5ug / ml and negative cells, at a total concentration of 1 .3E6 cells / ml, was used. The cells were introduced by pulling in 100 pL of the cell mix using a Hamilton Syringe. The cells were allowed to settle for a residence time of 240 seconds, and then washed with the same flow rate of either low-shear stress 0.350 mL / min (0.2 dynes / cm2) or high-shear stress 2.0 mL / min (1.5 dynes / cm2) as was used for washing away the unbound protein in the cell capture composition deposition steps. During the washing phase, images were taken using a fluorescent microscope every 30 seconds at different locations along the flow cell. These images were later used for counting the absolute number of positive and negative cells before and after cell separation using cell counting software.

[0498] Figure 30 and Figure 31 shows the results for polystyrene and polymethylpentene substrates respectively. On both substrates, the QBEND10- conjugate significantly enhanced the binding of positive target cells, whilst reducing the number of negative cells captured. On polystyrene substrate, shear stress had only a minor impact on the binding of negative cells to the surface. On polymethylpentene substrates, a higher % capture of negative non-target cells was observed at low shear stress, but this was completely negated at high shear stress with only minor impact on recovery of positive target cells.

[0499] The results demonstrate that sufficient fluid stress during the washing step may be important to ensure selective recovery of specifically bound positive cells, particularly on polymethylpentene substrates.Example 15. Cell selection in a flow cell treated with different antiCD34 antibodies

[0500] Further flowcell cell selection experiments were conducted (per Example 9 general procedure) to investigate cell selection on a polystyrene substrate, a suitable model system for other hydrophobic substrates such as TPX, using cell capture compositions containing different antiCD34 antibodies. Two different types of 9C5 were tested to evaluate potential variations between samples. Recombinant humanised 9C5.1 was purchased from a commercial manufacturer (Creative Biolabs, NY, USA), and mouse recombinant 9C5.C was prepared in-house. Capture of KG1 a cells from a mixture of KG1 a cells (with Hoechst33342 staining) and Jurkat mCherry cells (50:50 mixture; 1.3E6 cells / mL) was investigated. The flowcell residence time during cell adhesion was 240 s. The recovery of unbound cells was conducted at high shear stress, using a flow rate of 2 ml / min (c.a. 1 .5 dynes / cm2).

[0501] Experiments were conducted after treatment of the flowcell with My10, 9C5.1 and 9C5.5 unconjugated antibodies (50pg / mL), both with and without a subsequent treatment of free Pluronic F127 (1 % w / v) to block non-specific cell binding. The results are shown in Figure 32.

[0502] Both unconjugated antiCD34 9C5 clones, 9C5.1 and 9C5.C, exhibited a higher percentage of captured positive target cells compared to the MY10 clone. Furthermore, the addition of pluronic enhanced positive cell capture for all antibodies, but the unconjugated 9C5 + free pluronic treated flow cell displayed significantly higher percentages of positive cell capture compared to the unconjugated MY10 + free pluronic treated flow cell. Across all conditions, the capture of negative (Jurkat mCherry) non-target cells was relatively low. However, with the inclusion of pluronic, the 9C5 clones demonstrated nearly 0% capture of negative cells.

[0503] The results demonstrates that the addition of pluronic after the physical adsorption of an antibody consistently improves cell selection in a flow system, regardless of the antibody type. Secondly, the results highlight the importance ofantibody affinity to the antigen for efficient capture of target cells on a hydrophobic surface. The 9C5 clone was found to possess significantly higher affinity to the CD34+ antigen than MY10.Example 16. Cell selection in a flow cell treated with different antiCD34 antibodies

[0504] Further flowcell cell selection experiments were conducted (per Example 9 general procedure) to investigate cell selection on either polystyrene or polymethylpentene substrates using cell capture compositions containing antiCD3 antibodies and antiCD3 antibody-conjugates. Cell capture compositions were produced by the sequence of treatments shown in Table 5.Table 5.

[0505] Capture of negative non-target KG1 a cells from a mixture of KG 1 a cells (with Hoechst33342 staining) and positive target Jurkat mCherry cells (50:50 mixture; 1 .3E6 cells / mL) was then investigated. The flowcell residence time during cell adhesion was 240 s. The recovery of unbound cells was conducted at high flow stress, using a flow rate of 0.35 ml / min and a recovery shear stress of c.a. 0.2 dynes / cm2. The results for capture of positive target Jurkat cells are shown in Figure 33.

[0506] The behaviour of the CD3 antibody is similar to that of CD34 antibodies. The addition of free pluronic and the conjugation of pluronic to the antibody both enhance the capture of positive cells on a hydrophobic substrate when compared to unconjugated antibody alone.Example 17.

[0507] Further flowcell cell selection experiments were conducted (per Example 9 general procedure) to test different clones of monoclonal antibodies, and their conjugates.

[0508] Flowcells with either hydrophobic polystyrene or hydrophobic polymethylpentene substrates were thus treated with the following CD34 monoclonal antibodies, or conjugates thereof. The following conjugates were used: pluronic-MY10 conjugate (sample AL2), pluronic-QBENDI O conjugate (sample ALQ1 ), and pluronic- 9C5 conjugate (sample 9C5-AL), all per Example 2, Table 1 . Each reagent was applied at room temperature for 1 hour, with subsequent washing of the flowcell with DPBS at a flow rate of 2 mL / min (1 .5 dynes / cm2; high shear stress). Cell capture compositions were produced by the sequence of treatments shown in Table 6.Table 6.

[0509] Capture of target KG1 a cells from a mixture of KG1 a cells (with Hoechst33342 staining) and non-target purified buffy coat cells (1 :8 mixture; 1.2E6 cells / mL) was then investigated. The cells were introduced by pulling in 100 pL of the cell mix using a Hamilton Syringe. The cells were allowed to settle for a residence time of 240 seconds, and then washed with DPBS at a flow rate of 2.0 mL / min (fluid shear stress of 1.5 dynes / cm2). During the washing phase, images were taken using a fluorescent microscope every 30 seconds at different locations along the flow cell. These images were later used for counting the absolute number of positive and negative cells before and after cell separation using cell counting software. The results for capture of positive cells (CD34+) and negative cells (Buffy coat cells) are shown in Figure 34 for polymethylpentene substrate and Figure 35 for polystyrene substrate (95% confidence interval).

[0510] Utilisation of a conjugate led to the most significant improvement in positive cell selection for the MY10 clone. In contrast, higher affinity antibodies such as QBEND10 and 9C5 exhibited high percentages of positive cell capture even without conjugation. However, the key advantage of incorporating pluronic or using a mAb- conjugate became apparent in the remarkably low binding of negative cells at a substantially uneven ratio of cell population mix. This can be seen clearly in Figure 36 which shows the % purity of CD34+ cells, calculated using the formula = no. of target cells in output / (no of positive cells in output + no. of negative cells in output)Example 18. Methodology for cell separation on cell capture compositions in a bidirectional flow multi-layer micro fluidic device.

[0511] The following procedures were used to investigate cell separation on cell capture compositions present on the grooved surface of a flow channel in a bidirectional flow multi-layer microfluidic device, configured as described herein with reference to Figures 4 and 5.Reagent treatment: Adsorption of unconjuqated antibody followed by free pluronic• 1 .15mL of unconjugated antibody (50pg / mL) loaded from the reagent inlet bore (381 ) to the media outlet bore (392).• Allow to deposit on flow channel grooves of microfluidic device at room temp for 1 hour, under static conditions once filled.• Connect a 1.5mL Eppendorf tube containing 1 mL of 1% w / v Pluronic F-127 in water to the reagent inlet bore (381 )• Flow the pluronic F-127 from the reagent inlet bore (381 ) to the media outlet bore (392)• Allow to deposit on flow channel grooves of microfluidic device at room temp for 1 hour, under static conditions once filled.• Wash unbound reagents with 1 mL of DPBS loaded from the reagent inlet bore (381 ) to the media outlet bore (392)• Wash unbound reagents with 2mL of DPBS loaded from the cell inlet bore (384) to the cell outlet bore (391 )• Reservoirs and fluid connected prior to transferring out of the biosafety hood.Reagent treatment: Adsorption of pluronic-antibodv conjugates• 1 ,15mL of pluronic-antibody conjugate (50pg / mL) loaded from the reagent inlet bore (381 ) to the media outlet bore (392).• Allow to deposit on flow channel grooves of microfluidic device at room temp forI hour, under static conditions once filled.• Wash unbound reagents with 1 mL of DPBS loaded from the reagent inlet bore (381 ) to the media outlet bore (392).• Wash unbound reagents with 2mL of DPBS loaded from the cell inlet bore (384) to the cell outlet bore (391 ).• Reservoirs and fluid connected prior to transferring out of the biosafety hood.Cell separation• Load 1 .5mL of cell mix under loading pressure of +1 OOhPa to load the cells forI I mins along the grooves from the cell outlet bore (391 ) to the cell inlet bore (384). The cell mix contained negative and positive cell population suspended in buffer B (DPBS + 0.5% BSA + 2mM EDTA) at a concentration of 5 x 106cells / mL.• Decrease pressure to +50hPa and close cell inlet valve (connected to bore 384) then the cell outlet valve (connected to port 391 ).• Wash out upbound cells by flowing buffer A (DPBS + 0.5% BSA) under loading pressure of +999hPa for 4mins from the cell outlet bore (391 ) to the cell inlet bore (384).• Decrease pressure to +50hPa and close the cell inlet valve then the cell outlet valve• Image the cells captured in the device using an Olympus IX83 live cell imaging system.Example 19. Comparison of cell selection in PDMS vs polymethylpentene multilayer microfluidic devices

[0512] Bidirectional flow multi-layer microfluidic devices, configured as described herein with reference to Figures 4 and 5, were fabricated from either thermosetting polydimethylsiloxane (PDMS) or thermoplastic polymethylpentene (TPX). Basic methods for fabrication of PDMS-based devices is disclosed in WO2020 / 227772. Details of the methodology for fabricating the polymethylpentene (TPX) device is provided in subsequent examples.

[0513] Cell separation experiments were conducted using the general procedure of Example 18, with a surface treatment by pluronic-MY10 conjugate prepared by NHS chemistry (with mAb:plu ratio = 1 :20 mokmol; conversion 100% per Scheme 3 and Example 2) and a cell mixture comprising a 1 :9 ratio of positive target cells to negative non-target cells (Hoechst33342 stained KG1 a target cells and Jurkat mCherry cells respectively). Wide field fluorescence image scans indicated improved cell distribution and device utilisation with the polymethylpentene device. The local heat map of cell enrichment demonstrates improved selectivity of the antibody-treated polymethylpentene surface for capture of CD34+ cells, reducing the off-target capture rate (capture of non-target cells) and increasing the number of target cells captured. More efficient delivery of cells into the culture chamber allowed for better cell separation performance in polymethylpentene devices. The improved results were ascribed to the lack of channel deformation and improved control of fluid shear stress provided by the rigid thermally bonded polymethylpentene device in comparison to PDMS.Example 22. Development of multilayer microfluidic devices fabricated from polymethylpentene (TPX)

[0514] Multilayer microfluidic devices previously used for cell capture applications, as disclosed in WO2020 / 227772, included one or more gas-permeable layers fabricated from thermosetting polydimethylsiloxane (PDMS). PDMS is well suited for prototyping microfluidic devices and provides excellent permeability to O2 as required for the flow flayer.

[0515] However, the present inventors turned their attention to oxygen-permeable thermoplastics for fabrication of multilayer microfluidic devices for cell capture applications. This is due to the following advantages of thermoplastics, some or all of which may be obtainable when fabricating a thermoplastic multi-layer microfluidic device for cell culture:• short time required for moulding and hardening each structured thermoplastic layer for the device;• excellent rigidity of the structured thermoplastic layers, reducing the risk of variable layer separation distance or device failure at different applied pressures in use;• thermal bonding between adjacent layers when assembling the device does not adversely affect the hydrophobicity of the polymer (unlike e.g. plasma bonding used to bond thermoset layers).• bonding can be created between all contacting portions of adjacent layers when assembled to produce the device (including e.g. bonding between the flow layer and the structural posts of the header layer)

[0516] This led to an investigation into the use of polymethylpentene (TPX) in the fabrication of multilayer microfluidic devices. However, significant difficulties were encountered in producing devices with effective performance in cell capture applications.

[0517] The manufacturing process comprises of (1 ) layer production by moulding of TPX under compression in a mould at a temperature above the TPX melting point, 2) demoulding the resultant structured TPX layers, (2) annealing the TPX layers, (3)thermal bonding of the TPX layers to produce the multilayer microfluidic device, and (4) fluidic interfacing of the device for fluid and gas inputs and outputs.

[0518] Prolonged investigations revealed that precise control of temperature, time and pressure is critical to producing a functional device. TPX rapidly oxidises above its melting point (220°C) resulting in adhesion to moulds. Oxidation is also detrimental to cell separation selectivity as it promotes non-specific binding of antigen negative cells. Optical flatness of layers is required for effective bonding of TPX layers and must therefore be achieve during annealing. The following learnings about failure modes and mitigations were gained.

[0519] Moulding. Inadequate pressure during moulding or poor mould-cover alignment causes bubble formation in the layers. Bubbles can also be avoided by arranging TPX pellets as a monolayer with minimal distance between each grain. Excessive temperature or time during moulding results in oxidation or other degradation of the TPX, and this should be avoided. The use of N2 or other inert gas in TPX fabrication is also a potential solution for addressing this challenge.

[0520] Demoulding. Inadequate cooling time and excessive removal force can result in deformation (warping or stretching) of layers on demolding, with the flow layer typically being most susceptible to defects due to its depth and fine features. Excessive cooling time risks mould damage as the TPX hardens while conformed to the mould, requiring more force to remove causing breaking of delicate features. Demoulding of an oxidized layer was found to stretch the layer turning the stretched part a white colour. These issues were exacerbated by the need to produce TPX flow layers thin enough to provide adequate gas perfusion into the flow chamber in use, in view of the lower gas permeability of TPX compared to thermosetting PDMS.

[0521] Demoulding issues could be mitigated via the use of a hydrophobic coating layer (e.g. silane) on the mould, and by demoulding the TPX sheet before it reaches its full hardness with minimum force. Other release agents can also be used to improve demoulding performance.

[0522] Annealing. Annealing was found necessary to reduce surface roughness of the layers, thereby improving successful bonding in the subsequent fabrication of the multilayer device. Pressure (e.g. 50-500 kPa) and time (e.g. 2-5 minutes) need to beprecisely controlled at a temperature between TPX’s glass transition temperature (~170°C) and melting point (~220°C). Annealing against a flat substrate, preferably an optically flat substrate, is required to produce the necessary surface smoothness. As with demoulding, the flow layer is more susceptible to defects, however resistance channels and posts in the header layer are also delicate. Excessive pressure and / or time during annealing was found to narrow flow channel features in the layers, while inadequate pressure and / or time was found to leave an insufficiently smooth surface for subsequent thermal bonding. Annealing of a flow layer which was deformed during demoulding produced unacceptably “wavy” grooves in the annealed flow layer.

[0523] Thermal bonding. Pressure, temperature and time must be carefully controlled for bonding between layers to be successful. Extreme conditions (e.g. over 0.5- 1 MPa, 200°C or 30 mins in step 1 bonding and over 225°C or 10 minutes in step 2 bonding) can result in the collapse of moulded features and oxidation. Conversely, insufficient heating and pressure leads to incomplete bonding. If pressure is not uniformly distributed across the layers, conformal contact is not achieved resulting in non-uniform thermal distribution and a microfluidic device that bulges in the centre. The layers must be precisely aligned prior to thermal bonding as the process is irreversible and any misalignment will result in substandard device with an unpredictable flow pattern. Inadequately smooth surfaces, e.g. produced by ineffective annealing, can cause bonding to fail as the bonding surfaces are not in conformal contact.Example 23. Manufacturing of multilayer microfluidic devices fabricated from polymethylpentene (TPX)

[0524] After intensive research, as described in Example 22, the following fabrication methodology was developed for successful manufacture of TPX multilayer microfluidic device as depicted in Figure 5 and having a header layer design depicted in Figure 37. The header layer has a thickness of about 300 pm (other thicknesses in the range of 50 to 400 pm can also be produced), the flow layer has a total thickness (maximum thickness as defined herein) of about 200 pm (other thicknesses in the range of 300 to 400 pm can also be produced), the grooves in the flow layer have a depth of 150 pm (other depths in the range of 30 to 150 pm can also be produced), and the base layer has a thickness of about 200 pm (other thicknesses in the range of 200 to 400 pm can also be produced).

[0525] Moulding. Header, flow, and base layers are rapidly moulded on a rigid aluminium micromachined mould or etched silicon mould from TPX pellets at 250°C with 1 -5 tons of compression (1 -5 MPa, approximately) over 30 seconds, using a heated press (Carver Press). Full details of the optimised moulding step for each layer are provided in Table 7 below.

[0526] Cooling and demoulding. Each structured layer, still on the mould, is placed on a cooling plate and left to cool for 6-8 minutes before carefully removing the layer from the mould. It was estimated that the sheet temperature is around 60-80C when demoulded.

[0527] Annealing. Each layer is annealed on an optically flat silicon wafer with 0.2- 0.4 ton (c.a. 0.2-0.4 MPa) pressure for 4 minutes at 180°C, using a heated press. Full details of the optimised annealing step for each layer are provided in Table 8 below.

[0528] Thermal bonding. Header, flow, and base layers were then aligned and compressed at 0.6-1 ton (c.a. 0.6-1 MPa) at 180°C for 10-15 minutes to eliminate gas bubbles and make conformal contact, using a heated press (step 1 bonding). This was followed by a 5-7-minute bake on a hot plate at 220°C (step 2 bonding). Full details of the optimised thermal bonding step are provided in Table 9 below.Table 7. Moulding of three layersTable 8. Annealing of three layersTable 9. Thermal bonding three layers togetherExample 24. Fluidic coupling of multilayer microfluidic devices fabricated from polymethylpentene (TPX)

[0529] After successfully manufacturing the TPX multilayer microfluidic devices (per Example 23), a fluidic interface is necessary to transfer fluids and cells throughout the device under the pressures and flow rates required for device verification and cell separation, transduction, and expansion. Two methods of interfacing were developed to achieve this: 1 ) glued plastic ports and 2) metal clamps in combination with off-the- shelf fluidic fittings.

[0530] For the first method, block-shaped plastic ports were laser-cut and permanently glued in place on top of the header layer, in alignment with the inlet and outlet ports of the header layer. A biocompatible thermoplastic adhesive was used to ensure the device can withstand at least 1 bar inlet pressure. Tubing inserted into these ports required stenting and a Teflon tape wrapping to provide a leak-proof compression fit for cells and fluids to pass through. Although this method is irreversible and not sterilisable, it was found ideal for early-stage testing and prototyping.

[0531] The second method utilised metal clamps specifically designed with fluid coupling ports to align with the microfluidic device inlet and outlet port locations, a microscope viewing window, microscope holder and off-the-shelf fluidic connections. The microfluidic device, equipped with clamps and associated tubing, is shown in Figure 38. This method allows for the microfluidic device to be steam-sterilised. Moreover, the clamps can be readjusted to ensure alignment and can be reused after each cell experiment.

[0532] The microfluidic devices with ports fitted were then prepared as follows:Glued ports (Negative pressure degassing method, tube fitting)

[0533] The TPX microfluidic device is taped to a container that measures 12cm x 12cm x 4cm. The container is filled halfway with MilliQ water and placed in a vacuum chamber for 1 or 1 .5 hours. The container is then taken out and left to sit at room pressure with a cover for an hour. Five pieces of Tygon ND-100-80 tubing that are 60cm long and have a 0.02” inner diameter and a 0.06” outer diameter are obtained. A 22G stent is inserted into one end of each tubing and that end is wrapped with Teflon Tape. The other end of each tubing is connected to a fluid reservoir and the stented end is inserted into an acrylic port.Clamped ports (Positive pressure degassing method, tube fitting)

[0534] Prior to assembling the clamped device, five 60cm lengths of Tygon ND- 100-80 tubing with 23G stents inserted in one side are prepared. With the aid of the top plate of the clamp and a ferrule fitting, ferrules are pre-fitted onto the stented side of the tubing such that the end extends 2mm past the base of the top plate. The following details the assembly of the clamped device prior to steam sterilization: the smaller holes of the PDMS coned gaskets are aligned to the bottom of the holes in the top aluminium plate and the blank stamps are aligned to the sides of the TPX-microfluidic device that do not contain ports. The TPX-microfluidic device is then aligned to the larger holes of the coned stamps, adjusting stamps where necessary before firmly pressing the TPX- microfluidic device in place. The top plate arrangement is then placed onto the bottom plate, with the microfluidic device between the two plates. The plates are then temporarily held together using finger-tightened bolts, finally forming the clamp assembly. The pre-fitted tubing and clamp assembly are then autoclaved.

[0535] After a clamped device completes a sterilization cycle, the pre-fitted end of the tubing is inserted into the ports of the clamp assembly and secured in place with ferrule fittings. Each bolt is tightened to 17.5cNm of torque to secure the device. Free ends of the tubing are fitted with pinch clamps and secured into reservoirs containing either MilliQ water or DPBS. With unused ports closed with clamps, flow is encouraged between the media inlet and media outlet ports by applying 1 Bar of pressure to the media inlet reservoir. Once the media inlet resistive flow path is filled, the media inlet tubing is clamped, and the reagent inlet tubing is opened with 1 bar applied to the respective fluidic reservoir. When the edges of the cell chamber in the microfluidic device have been sufficiently coated, the reagent inlet and media outlet ports can be clamped. A flow path is established between the cell inlet and cell outlet ports, with 0.5 Bar of pressure applied to the inlet for 5 minutes. Finally, the microfluidic device is degassed by clamping the cell outlet and applying 0.8 bar of pressure to the cell inlet overnight to remove any remaining air pockets.Example 25. Characterisation of multilayer microfluidic devices fabricated from polymethylpentene (TPX)

[0536] The structural fidelity and suitable flow properties of the TPX multilayer microfluidic devices, as produced in Example 23 and fluidically coupled in Example 24, was confirmed by (1 ) flow calibration and 1 bar destructive test, (2) dye flow tests, (3) confocal microscopy and (4) in a cell capture experiment at 1 bar pressure.Flow calibration and 1 bar destructive test

[0537] A functional microfluidic device requires a consistent performance at up to 1 bar gauge working range (typical operating pressures between 0.01 to 1 bar), which is used when loading cells into the device. Flow calibration experiments were conducted to make sure there is a linear relationship between input pressure and flow rate, while also determining if any bubble remains in the system after degassing and if there is any leakage happening at 1 bar.

[0538] Using either fluidic interface device set up (per Example 24), the following steps were used to generate flow calibration curves and confirm the bond strength of the microfluidic device:1 ) Flow calibration with 50hPa pressure increment steps from 0 hPa up to 1000 hPa (1 bar). At the same time, the outlet flow rate was measured with an inline flow meter connected downstream.2) Clamp off all outlets and apply and input pressure of 1000hPa (1 bar) for a minimum of 2 minutes to make sure the device layers do not delaminate. If there is no visual bulging or audible splitting of the microfeatures then it can be concluded that the TPX-device can withstand 1 bar pressure in use.

[0539] As seen in Figure 39, devices produced according to Example 23 were found to provide linear responses of flow rate against input pressure from OhPa up to 10OOhPa (1 bar), when flowing fluids from media inlet to media outlet, from reagent inlet to media outlet and from cell inlet to cell outlet. The devices also passed the pressure test without failure.Dye flow tests

[0540] Dye flow tests were conducted to test whether channel heights (distance between floor and ceiling of layers) and resistance are consistent across the whole flow channel area of the microfluidic devices. The tests were conducted as follows (after flow calibration as described above):• Replace MilliQ water-filled with dye-filled reservoir, remove flow meter at outlet of the device.• Flow dye from Cell inlet to Cell outlet (with all other ports sealed) and record the flow profile.• Flow dye from Reagent inlet to Media outlet (with all other ports sealed) and record the flow profile.

[0541] The tests revealed that dye flowing from the Cell Inlet via the multiple fluid inlets into the flow channel moved at a consistent flow rate along the grooves, reaching the multiple outlets at the opposite side of the flow channel at the same time. The same result was obtained in flow across the grooves originating at the Reagent Inlet. The test revealed that the device is suitably configured to provide consistent flow conditions across the flow channel in both flow directions.Confocal microscopy

[0542] Confocal microscopy was used to investigate the internal structure of the microfluidic device. The protocol of confocal imaging was as follows:• Fill the TPX microfluidic device with FITC-dextran 20kDa at concentration of 1 mg / mL in water.• The device is imaged at different locations along the z-axis using confocal microscope Zeiss 880 inverted with 32x water immersion objective (NA= 0.85) or Confocal microscope Zeiss 900 upright with 20x air objective (NA=0.80).

[0543] The microscopy images confirmed the fidelity of the internal microfluidic device microfeatures, including flow resistors, flow bifurcations (between inlet ports and fluid inlets, or between fluid outlets and outlet ports) and the grooved flow channel.Example 26. Observations of cell distribution in multi-layer microfluidic devices

[0544] Further studies conducted with multi-layer microfluidic devices revealed a number of issues when using multi-layer microfluidic devices in cell capture experiments. As described in Examples 23-25, microfluidic devices configured for bidirection flow delivered by bifurcating header patterns were able to evenly distribute fluids over the bioreactor grooved bed as demonstrated by dye flow tests. When distributing cell media, however, the resultant cell distribution did not follow the same even distribution of dye molecules.

[0545] Cell distribution by flowing cell mixtures along the grooves was found to be non-uniform with increased cell capture toward the centre of the grooved bioreactor bed. The uneven distribution of captured cells results in a reduced utilization of bioreactor area for cell separation. Cells were also observed to deposit preferentially in ‘dead’ zones at the entry into the groove bioreactor bed between fluid inlets where fluid shear is reduced, and in low shear regions behind the bioreactor support posts.

[0546] Blockages of the upstream bifurcating flow distribution network were also observed. This was ascribed to cell clumps entering the narrow flow channels.

[0547] The separation of antigen negative and positive cells during the washing step depends on the washing fluid shear stress. If the washing shear stress is too low,there will be larger non-specific retention of antigen negative cells in the bioreactor. Also, during the stage of cell recovery from the bioreactor using cell releasing agents such as anti (mouse Fab), chymopapain or competing peptides, a low shear stress will reduce the recovery of antigen positive cells.

[0548] During subsequent culturing of the captured cells by cross-flow perfusion of cell culture medium, an undesirably high loss of cells was observed at high flow rates.

[0549] Cells were also retained in the flow channel behind the bioreactor support posts when harvesting the cells due to regions of low flow stress behind the support posts when flowing fluid along the grooves in the direction of the Cell Outlet port.Example 27. Further development of the multilayer microfluidic devices to improve cell distribution.

[0550] Through intensive research, the present inventors addressed the issues described in Example 26 by the following improvements to the design of the multi-layer microfluidic device.Location of the flow resistor

[0551] It was proposed that the poor distribution of cells in the flow channel was caused by a hydrodynamic force called inertial migration that pushes cells away from the walls of microfluidic channels, particularly in regions of high fluid velocity. Cells flowing through hydraulic flow resistors located upstream of the flow channel migrate across streamlines toward two equilibrium positions along the width of the hydraulic resistors causing uneven distribution of cells as they enter the culture chamber.

[0552] The issue was mitigated by an alternate design where the control of fluid flow rate through the flow chamber is provided by downstream hydraulic resistors, with the implication that the microfluidic groove bioreactor chamber is exposed to the full pressurisation of 1 Bar used to flow cells through the bioreactor. This operating pressure could be accommodated due to the improved design, strength and structural rigidity of the TPX microfluidic devices.

[0553] During the washing stage, where negative cells are flowed out of the bioreactor while retaining the positive cells retained on the cell capture composition, asignificant flow stress along the grooves is required. By-passing the flow resistors entirely in this stage assists to increase the flow stress, thus improving cell enrichment.Design of the fluid inlets to the flow channel

[0554] The preferential deposition of cells in ‘dead’ zones at the entry into the groove bioreactor bed, between the fluid inlets, was mitigated by two improvements to the microfluidic device. Firstly, each fluid inlet to the flow channel was redesigned to include a diverging flow guide (divergent channel) so that flows of fluid entering the flow channel are progressively spread across an increasing flow width (orthogonal to the flow direction) when passing though the fluid inlet into the flow channel. The diverging flow guides of neighbouring second fluid inlets meet in the flow channel so that the flows of fluid entering the flow channel from the neighbouring fluid inlets are substantially adjacent when merged, thus avoiding or reducing regions of low fluid stress between adjacent flows entering the grooved flow channel.

[0555] Figure 40 depicts the design change. In the initial design (on left), dead zones 600 of low fluid stress are formed between the separated fluid inlets 602 as fluid flows into the flow channel 604. In the improved configuration, each fluid inlet 702 comprises a diverging flow guide 708 which progressively spreads (diverges) the inflowing fluid across an increasing width (in the direction of arrows 710) as the fluid flows into flow channel 704. Diverging flow guides of neighbouring fluid inlets (e.g. 708a and 708b depicted) meet in the flow channel at junctions 712 so that the flows of fluid entering the flow channel from the neighbouring second fluid inlets are already diverged to be substantially adjacent when merged past the junctions. This design was found to reduce or eliminate dead zones of low flow stress, thus mitigating the preferential deposition of cells near the fluid inlets.

[0556] A second improvement in cell deposition was obtained by removing the groove array away from any remaining regions of low flow near the fluid inlets. This ensures that the cells entering the flow chamber experience an even fluid shear across all grooves in the culture chamber. As seen in Figure 41 , the nearest end 714 of each groove in groove array 716 is set back from the one or more second fluid inlets by an ungrooved, planar portion 718 of the flow channel’s floor. The set back distance may be, for example, 3 mm from the narrow ends of the diverging flow guides 708 asdepicted, or at least 1 .5 mm from the wide ends of the diverging flow guides 708, i.e. from the junctions.Groove design

[0557] Flow simulations may be used to optimise the groov...

Claims

Claims:1 . A bioreactor for cell culture, comprising: a multi-layer microfluidic device comprising a header layer and a gas permeable flow layer beneath the header layer, wherein:(a) the flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells; and(b) the header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises:(i) a first inlet port;(ii) a first outlet port;(iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and(iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port; wherein the one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein; and a cell capture composition physically adhered to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

2. The bioreactor according to claim 1 , wherein the hydrophobe is a hydrophobic polymeric segment.

3. The bioreactor according to claim 1 or claim 2, wherein the one or more hydrophilic polymeric segments are polyethylene glycol segments.

4. The bioreactor according to any one of claims 1 to 3, wherein the amphiphilic polymer is a triblock copolymer comprising a hydrophobic polypropylene glycol segment located between two hydrophilic polyethylene glycol segments.

5. The bioreactor according to any one of claims 1 to 4, wherein the amphiphilic polymer has a molecular weight in the range of 10,000 to 20,000 g / mol.

6. The bioreactor according to any one of claims 1 to 5, wherein the antigen binding moiety is conjugated to at least one of the one or more hydrophilic polymeric segments such that the cell capture composition comprises a polymer-antigen binding moiety conjugate.

7. The bioreactor according to claim 6, wherein the antigen binding moiety is conjugated to a chain terminus of the at least one hydrophilic polymeric segment.

8. The bioreactor according to claim 7, wherein the antigen binding moiety is conjugated to the chain terminus by an amide linking group, wherein the amide linkage is produced by a conjugation reaction between a terminal amine-reactive functional group on the at least one hydrophilic polymeric segment with an amine functional group of the antigen binding moiety.

9. The bioreactor according to claim 7, wherein the antigen binding moiety is conjugated to the chain terminus by a click reaction between a terminal click- reactive functional group on the at least one hydrophilic polymeric segment and a complementary click-reactive functional group which is labelled on the antigen binding moiety.

10. The bioreactor according to any one of claims 6 to 9, wherein the cell capture composition further comprises an unconjugated amphiphilic polymer.1 1 .The bioreactor according to any one of claims 1 to 5, wherein the antigen binding moiety is not conjugated to the amphiphilic polymer, and wherein the amphiphilicpolymer is physically adhered to the hydrophobic upper surface after the antigen binding moiety is physically adhered to the hydrophobic upper surface.

12. The bioreactor according to any one of claims 1 to 1 1 , wherein the multi-layer microfluidic device further comprises a base layer beneath the gas-permeable flow layer, wherein the base layer provides a gas flow path for gas exchange across the flow layer.

13. The bioreactor according to any one of claims 1 to 12, wherein the header layer comprises:- a second outlet port; and- one or more second fluid outlets providing fluid communication between the flow channel and the second outlet port; wherein the one or more second fluid outlets are positioned in the header such that fluid exiting the flow channel through the second outlet port flows in a second flow direction toward the one or more second fluid outlets, the second flow direction being along the one or more grooves.

14. The bioreactor according to claim 13, wherein the header layer further comprises:- a second inlet port; and- one or more second fluid inlets providing fluid communication between the second inlet port and the flow channel; wherein the one or more second fluid inlets are positioned along a dimension of the header which is substantially orthogonal to a longitudinal dimension of the one or more grooves so that fluid entering the flow channel from the one or more second fluid inlets flows toward the one or more second outlets in the second flow direction.

15. The bioreactor according to claim 14, wherein the one or more second fluid inlets comprise a diverging flow guide so that flows of fluid entering the flow channel from each second fluid inlet are progressively spread across an increasing flow width when passing though the diverging flow guide.

16. The bioreactor according to claim 15, wherein the diverging flow guides of neighbouring second fluid inlets meet in the flow channel so that the flows of fluid entering the flow channel from the neighbouring second fluid inlets are substantially adjacent when merged.

17. The bioreactor according to any one of claims 14 to 16, wherein the nearest end of each groove is set back from the one or more second fluid inlets by an ungrooved portion of the hydrophobic upper surface in the flow channel.

18. The bioreactor according to any one of claims 14 to 17, wherein the multi-layer microfluidic device comprises a flow resistor downstream of the second outlet port to regulate the flow rate of fluid flowing in the second flow direction toward the one or more second fluid outlets in the flow channel.

19. The bioreactor according to any one of claims 13 to 18, wherein the first flow direction and the second flow direction are substantially orthogonal.

20. The bioreactor according to any one of claims 1 to 19, wherein the gas permeable flow layer comprises polymethylpentene and the polymethylpentene is present at the hydrophobic upper surface of the flow layer.21 .The bioreactor according to any one of claims 1 to 20, wherein the flow layer has a thickness, between bottom surfaces of the one or more grooves and the lower surface of the flow layer, of less than 100 pm.

22. The bioreactor according to any one of claims 1 to 21 , wherein the header layer comprises a plurality of structural posts configured to rest in abutment with the hydrophobic upper surface of the flow layer.

23. The bioreactor according to claim 22 wherein the structural posts have an elongated cross-section having an axis of elongation substantially aligned with the grooves.

24. The bioreactor according to any one of claims 1 to 23, wherein the multi-layer microfluidic device comprises a flow resistor to regulate the flow rate of fluid flowing in the flow channel in the first flow direction toward the one or more first fluid outlets.

25. The bioreactor according to any one of claims 1 to 24, wherein the one or more grooves have a depth of between 110 pm and 180 pm.

26. The bioreactor according to any one of claims 1 to 25, wherein the multi-layer microfluidic device comprises one or more through bores extending colinearly through each layer of the multi-layer microfluidic device, and wherein the multilayer microfluidic device is configured for arrangement in a stack including a plurality of the multi-layer microfluidic devices such that the through bores in each multi-layer microfluidic device in the stack provide fluid communication with corresponding through bores of at least one adjacent multi-layer microfluidic device in the stack.

27. The bioreactor according to any one of claims 1 to 26, wherein the antigen binding moiety is an antibody or antigen binding fragment thereof, wherein the antibody or antigen binding fragment specifically binds CD34 or CD3.

28. The bioreactor according to any one of claims 1 to 27, wherein the target cells are hematopoietic stem cells.

29. A bioreactor cell culture system comprising:(a) a plurality of bioreactors according to any one of claims 1 to 28 arranged in at least one stack;(b) a controller configured to individually control operation of one or more fluid pumps according to a cell culture protocol stored in a memory of the controller, the one or more fluid pumps delivering fluid to one or more through bores provided colinearly through the at least one stack, the delivered fluid entering the flow chamber of each bioreactor in the at least one stack according to the stored cell culture protocol.

0. A method of preparing a bioreactor for cell culture, comprising: providing a multi-layer microfluidic device comprising a header layer and a gas permeable flow layer beneath the header layer, wherein:(a) the flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells; and(b) the header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises:(i) a first inlet port;(ii) a first outlet port;(iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and(iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port; wherein the one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein; and physically adhering a cell capture composition to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell. 1 .The method according to claim 30, wherein the antigen binding moiety is conjugated to at least one of the one or more hydrophilic polymeric segments such that the cell capture composition comprises a polymer-antigen binding moiety conjugate.

32. The method according to claim 31 , wherein physically adhering the cell capture composition to the hydrophobic upper surface comprises flowing a first aqueous composition comprising the polymer-antigen binding moiety conjugate into the flow channel, and allowing the polymer-antigen binding moiety conjugate to deposit on the hydrophobic upper surface.

33. The method according to claim 32, further comprising depositing unconjugated amphiphilic polymer on the hydrophobic upper surface.

34. A method of culturing cells, the method comprising: a) providing a microfluidic device comprising:- a flow channel located over a substrate comprising a hydrophobic upper surface, wherein the hydrophobic upper surface comprises an array of grooves that are substantially parallel; and- a cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell; b) introducing cells into the flow channel under conditions to immobilise the target cells on the cell capture composition; c) flowing a culture medium through the flow channel over the groove array in a first flow direction that is across the grooves; and d) culturing the target cells under conditions to support viability, growth and / or differentiation.

35. The method of claim 34, further comprising releasing the target cells from the capture composition before or while flowing the culture medium through the flow channel, wherein the released target cells are retained in the grooves during the culturing.

36. The method of claim 34 or claim 35, wherein introducing the target cells comprises flowing a sample comprising the target cells through the flow channel, wherein the target cells are selectively immobilised on the cell capture composition.

37. The method of any one of claims 34 to 36, wherein the target cells are introduced by flow through the flow channel in a second flow direction that is along the grooves.

38. The method of any one of claims 34 to 37, wherein the culture medium is flowed through the flow channel at a flow rate equivalent to 0.2-5 mL per million cells per day.

39. The method of any one of claims 34 to 38, wherein the target cells are cultured for 1 -14 days.

40. The method of any one of claims 34 to 39, further comprising harvesting the cultured cells by flowing an extraction medium through the flow channel.41 .The method of claim 40, wherein the flow of extraction medium is provided at a shear stress of at least 0.1 dyne / cm2.

42. A method of transfecting / transducing cells, the method comprising: a) providing a microfluidic device comprising:- a flow channel located over a substrate comprising a hydrophobic upper surface, wherein the hydrophobic upper surface comprises an array of grooves that are substantially parallel; and- a cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell; b) introducing cells into the flow channel under conditions to immobilise target cells on the cell capture composition;c) introducing at least one exogenous molecule into the flow channel; and d) incubating the target cells and the at least one exogenous molecule under conditions to transfect / transduce the target cells with the exogenous molecule.

43. The method of claim 42, wherein introducing the at least one exogenous molecule into the flow channel comprises flowing the at least one exogenous molecule through the flow channel over the groove array in a first flow direction that is across the grooves.

44. The method of claim 42 or claim 43, further comprising releasing the target cells from the capture composition before or while introducing the at least one exogenous molecule into the flow channel, wherein the released target cells are retained in the grooves during the incubating.

45. The method of any one of claims 42 to 44, wherein introducing the target cells comprises flowing a sample comprising the target cells through the flow channel, wherein the target cells are selectively immobilised on the cell capture composition.

46. The method of any one of claims 42 to 45, wherein the target cells are introduced by flow through the flow channel in a second flow direction that is along the grooves.

47. The method of any one of claims 42 to 46, further comprising flowing a culture medium through the flow channel to culture the target cells in the culture medium.

48. The method of claim 47, wherein the target cells are cultured in the culture medium prior to introducing the exogenous molecule into the flow channel.

49. The method of claim 47 or 48, wherein the target cells are cultured in the culture medium after transfecting / transducing the target cells with the exogenous molecule.

50. The method of any one of claims 42 to 49, further comprising harvesting the transfected / transduced cells by flowing an extraction medium through the flow channel.51 .The method of claim 50, wherein the flow of extraction medium is provided at a shear stress of at least 0.1 dyne / cm2.

52. A method of selecting cells, the method comprising: a) providing a microfluidic device comprising:- a flow channel located over a substrate comprising a hydrophobic upper surface, wherein the hydrophobic upper surface comprises an array of grooves that are substantially parallel; and- a cell capture composition physically adhered to the hydrophobic upper surface in one or more grooves of the array of grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell; and b) flowing a sample comprising the target cells through the flow channel over the groove array in a second flow direction that is along the grooves under conditions to selectively immobilise the target cells on the cell capture composition.

53. The method of claim 52, further comprising releasing the target cells from the capture composition prior to or during at least one step selected from (i) culturing the target cells in the flow chamber, (ii) transfecting / transducing the target cells in the flow chamber and (iii) harvesting the target cells from the flow chamber.

54. The method of claim 52 or claim 53, further comprising flowing a culture medium through the flow channel over the groove array in a first flow direction that is across the grooves; and culturing the target cells in the presence of the culture medium.

55. The method of claim 54, wherein the culture medium is flowed through the flow channel at a flow rate equivalent to 0.2-5 mL per million cells per day.

56. The method of claim 54 or claim 55, wherein the target cells are cultured for 1 -14 days.

57. The method of any one of claims 52 to 56, further comprising introducing at least one exogenous molecule into the flow channel; and incubating the target cells and the at least one exogenous molecule under conditions to transfect / transduce the target cells with the exogenous molecule.

58. The method of claim 57, further comprising culturing the transfected / transduced cells.

59. The method of any one of claims 52 to 58, further comprising harvesting the target cells by flowing an extraction medium through the flow channel.

60. The method of claim 59 wherein the flow of extraction medium is provided at a shear stress of at least 0.1 dyne / cm2.61 . The method of any one of claims 34 to 60, wherein the substrate is a gas permeable layer and the microfluidic device comprises one or more gas flow paths for exchange of gas across the gas permeable layer.

62. The method of any one of claims 34 to 61 , wherein the substrate comprises polymethylpentene and the polymethylpentene is present at the hydrophobic upper surface of the substrate.

63. The method of any one of claims 34 to 62, wherein the hydrophobe is a hydrophobic polymeric segment.

64. The method of any one of claims 34 to 63, wherein the one or more hydrophilic polymeric segments are polyethylene glycol segments.

65. The method of any one of claims 34 to 64, wherein the amphiphilic polymer is a triblock copolymer comprising a hydrophobic polypropylene glycol segment located between two hydrophilic polyethylene glycol segments.

66. The method of any one of claims 34 to 65, wherein the amphiphilic polymer has a molecular weight in the range of 10,000 to 20,000 g / mol.

67. The method of any one of claims 34 to 66, wherein the antigen binding moiety is conjugated to at least one of the one or more hydrophilic polymeric segments such that the cell capture composition comprises a polymer-antigen binding moiety conjugate.

68. The method of claim 67, wherein the antigen binding moiety is conjugated to a chain terminus of the at least one hydrophilic polymeric segment.

69. The method of claim 68, wherein the antigen binding moiety is conjugated to the chain terminus by an amide linking group, wherein the amide linkage is produced by a conjugation reaction between a terminal amine-reactive functional group on the at least one hydrophilic polymeric segment with an amine functional group of the antigen binding moiety.

70. The method according to 68, wherein the antigen binding moiety is conjugated to a chain terminus of the at least one hydrophilic polymeric segment by a click reaction between a terminal click-reactive functional group on the at least one hydrophilic polymeric segment and a complementary click-reactive functional group which is labelled on the antigen binding moiety.71 .The method of any one of claims 67 to 70, wherein the cell capture composition further comprises an unconjugated amphiphilic polymer.

72. The method of any one of claims 67 to 71 , wherein providing the microfluidic device comprises flowing the polymer-antigen binding moiety conjugate into theflow channel, and incubating the hydrophobic upper surface with the polymer- antigen binding moiety conjugate for a sufficient period of time to allow the polymer-antigen binding moiety conjugate to physically adhere to the hydrophobic upper surface in the grooves.

73. The method of claim 72, wherein providing the microfluidic device comprises flowing an unconjugated amphiphilic polymer into the flow channel, and incubating the hydrophobic upper surface with the unconjugated amphiphilic polymer for a sufficient period of time to allow the unconjugated amphiphilic polymer to physically adhere to the hydrophobic upper surface in the grooves, wherein the cell capture composition comprises the polymer-antigen binding moiety conjugate and the unconjugated amphiphilic polymer.

74. The method of any one of claims 31 to 73, wherein the antigen binding moiety is an antibody or antigen binding fragment thereof, wherein the antibody or antigen binding fragment specifically binds CD34 or CD3.

75. The method of any one of claims 31 to 73, wherein the target cells are hematopoietic stem cells.

76. A multi-layer microfluidic device comprising a header layer and a gas permeable flow layer beneath the header layer, wherein:(a) the flow layer comprises a hydrophobic upper surface and a lower surface, and the hydrophobic upper surface comprises one or more grooves for retention of cells; and(b) the header layer is configured to define a flow channel over the hydrophobic upper surface of the flow layer, and the header layer comprises:(i) a first inlet port;(ii) a first outlet port;(iii) one or more first fluid inlets providing fluid communication between the first inlet port and the flow channel; and(iv) one or more first fluid outlets providing fluid communication between the flow channel and the first outlet port;wherein the one or more fluid inlets and the one or more fluid outlets are positioned in the header layer such that fluid entering the flow channel from the one or more first fluid inlets moves in a first flow direction toward the one or more first fluid outlets, the first flow direction being across the one or more grooves to minimise disruption of cells received therein, and wherein the gas permeable flow layer comprises polymethylpentene and the polymethylpentene is present at the hydrophobic upper surface of the flow layer.

77. The multi-layer microfluidic device according to claim 76, wherein the header layer and a base layer beneath the gas permeable flow layer comprise polymethylpentene.

78. The multi-layer microfluidic device according to claim 76 or claim 77, wherein fluid flowing from the first inlet port towards the flow channel is distributed to a plurality of first fluid inlets via one or more stages of flow bifurcation.

79. The multi-layer microfluidic device according to any one of claims 76 to 78, further comprising a base layer beneath the gas-permeable flow layer, wherein the base layer provides a gas flow path for gas exchange across the flow layer.

80. The multi-layer microfluidic device according to any one of claims 76 to 79, wherein the header layer comprises:- a second outlet port; and- one or more second fluid outlets providing fluid communication between the flow channel and the second outlet port; wherein the one or more second fluid outlets are positioned in the header such that fluid exiting the flow channel through the second outlet port flows in a second flow direction toward the one or more second fluid outlets, the second flow direction being along the one or more grooves.81 .The multi-layer microfluidic device according to claim 80, wherein the header layer further comprises:- a second inlet port; and- one or more second fluid inlets providing fluid communication between the second inlet port and the flow channel; wherein the one or more second fluid inlets are positioned along a dimension of the header which is substantially orthogonal to a longitudinal dimension of the one or more grooves so that fluid entering the flow channel from the one or more second fluid inlets flows toward the one or more second outlets in the second flow direction.

82. The multi-layer microfluidic device according to claim 81 , wherein fluid flowing from the second inlet port towards the flow channel is distributed to a plurality of second fluid inlets via one or more stages of flow bifurcation.

83. The multi-layer microfluidic device according to claim 81 or claim 82, wherein the one or more second fluid inlets comprise a diverging flow guide so that flows of fluid entering the flow channel from each second fluid inlet are progressively spread across an increasing flow width when passing though the diverging flow guide.

84. The multi-layer microfluidic device according to claim 81 , wherein the diverging flow guides of neighbouring second fluid inlets meet in the flow channel so that the flows of fluid entering the flow channel from the neighbouring second fluid inlets are substantially adjacent when merged.

85. The multi-layer microfluidic device according to any one of claims 81 to 84, wherein the nearest end of each groove is set back from the one or more second fluid inlets by an ungrooved portion of the hydrophobic upper surface in the flow channel.

86. The multi-layer microfluidic device according to any one of claims 81 to 85, wherein the header layer is configured to define a filter for removing particles from fluid flowing towards the one or more second fluid inlets.

87. The multi-layer microfluidic device according to any one of claims 80 to 86, wherein the first flow direction and the second flow direction are substantially orthogonal.

88. The multi-layer microfluidic device according to any one of claims 80 to 87, comprising a flow resistor downstream of the second outlet port to regulate the flow rate of fluid flowing in the second flow direction toward the one or more second fluid outlets in the flow channel.

89. The multi-layer microfluidic device according to any one of claims 76 to 88, wherein the header comprises a plurality of structural posts configured to rest in abutment with the hydrophobic upper surface of the flow layer.

90. The multi-layer microfluidic device according to claim 89, wherein the structural posts have an elongated cross-section having an axis of elongation substantially aligned with the grooves.91 .The multi-layer microfluidic device according to any one of claims 76 to 90, wherein the flow layer has a maximum thickness, beneath the flow channel, of no more than 400 pm.

92. The multi-layer microfluidic device according to any one of claims 76 to 90, wherein the flow layer has a thickness, between bottom surfaces of the one or more grooves and the lower surface of the flow layer, of less than 100 pm.

93. The multi-layer microfluidic device according to any one of claims 76 to 92, wherein the one or more grooves have a depth of between 110 pm and 180 pm.

94. The multi-layer microfluidic device according to any one of claims 76 to 93, wherein the header layer, the flow layer and a base layer beneath the gas- permeable flow layer each have a maximum thickness, beneath or above the flow channel, of no more than 400 pm.

95. A method of fabricating a multi-layer microfluidic device according to any one of claims 76 to 94, the method comprising: i) producing the header layer, the flow layer and optionally a base layer as separate components by (a) moulding polymethylpentene under compression in a mould, at a temperature above the polymethylpentene melting point, to produce a structured layer, (b) cooling the structured layer, (c) demoulding the structured layer before the polymethylpentene cools to ambient temperature; and (d) annealing the structured layer under compression against a flat substrate at a temperature between the polymethylpentene glass transition temperature and the polymethylpentene melting point, thereby reducing the surface roughness of the structured layer; and ii) aligning the header layer, the flow layer and optionally the base layer and compressing the aligned layers at a temperature above the glass transition temperature of the polymethylpentene to thermally fuse the layers, thereby forming the multi-layer microfluidic device.

96. The method of claim 95, wherein the polymethylpentene is moulded under compression of between 1 and 5 tons at a temperature of between 230 and 270°C.

97. The method of claim 95 or claim 96, wherein the polymethylpentene is above the polymethylpentene melting point for less than 60 seconds during moulding.

98. The method of any one of claims 95 to 97, wherein the structured layer is annealed under compression of between 0.1 and 0.5 tons against an optically flat substrate at a temperature of between 180 and 200°C.

99. The method of any one of claims 92 to 96, wherein the header layer, the flow layer and optionally the base layer are compressed at between 0.3 and 1 .5 tons at a temperature of between 180 and 200°C to thermally fuse the layers.

100. A bioreactor for cell culture, comprising:the multi-layer microfluidic device according to any one of claims 76 to 94; and a cell capture composition physically adhered to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

101. A method of culturing cells, the method comprising: i) providing a multi-layer microfluidic device according to any one of claims 76 to 94; ii) introducing cells into the flow channel under conditions to deposit the cells in the one or more grooves; iii) flowing a culture medium through the flow channel in a first flow direction that is across the one or more grooves; and iv) culturing the cells under conditions to support viability, growth and / or differentiation.

102. The method of claim 101 , wherein the cells are introduced into the flow channel by flow in the first flow direction, wherein the cells deposit in the one or more grooves by sedimentation.

103. The method of claim 102, wherein the multi-layer microfluidic device comprises a cell capture composition physically adhered to the hydrophobic upper surface in the one or more grooves, wherein the cell capture composition comprises (i) an amphiphilic polymer comprising a hydrophobe and one or more hydrophilic polymeric segments; and (ii) an antigen binding moiety that specifically binds to a target cell.

104. The method of claim 103, wherein the cells are introduced into the flow channel by flow in a second flow direction that is along the grooves under conditions to immobilise the target cells on the cell capture composition.

Images