Biomimetic device

Abstract

A method of manufacturing a biomimetic device (100) suitable for myelination assays, the method comprising: forming a plurality of pillars (110) extending from a base (112), the plurality of pillars (110) and the base (112) comprising a hydrogel material (114), wherein forming the plurality of pillars (110) and the base (112) comprises: applying an uncured hydrogel precursor material (114a) to a mould; curing the hydrogel precursor material (114a) to form a hydrogel material (114) in the mould; and removing the hydrogel material (114) from the mould.

Description

[0001] BIOMIMETIC DEVICE

[0002] FIELD OF THE INVENTION

[0003] The present invention relates to biomimetic devices suitable for myelination assays, and methods of manufacturing such biomimetic devices.

[0004] BACKGROUND

[0005] Oligodendrocytes (OLs) play a crucial role in the functioning of the central nervous system (CNS) by myelinating neuronal axons. Breakdown of the myelin can impact neuronal communication, which can result in neurodegeneration and neurological disorders (such as multiple sclerosis - MS).

[0006] Currently, there are no approved therapies for MS, or other demyelinating diseases, that stimulate myelin repair. Analysis of candidate drugs is difficult to conduct in vivo. The intricate and complicated structure of the brain and spinal cord mean that it is challenging to observe and study myelination in mammals. Even if such studies can be performed, animal models may not provide accurate predictions for humans. Furthermore, it is not possible to manipulate mechanical or physical aspects of tissues in vivo, meaning that many of the factors that impact OL-mediated myelination in the CNS are negated during studies and are not well understood.

[0007] Thus, there is widespread interest in developing in vitro models to study myelination. Some progress has been made in the field; 3D structures can be produced to mimic axonlike fibres, around which OLs can grow and myelinate. However, current in vitro models have a number of drawbacks that limit their suitability for and effectiveness in myelination assays.

[0008] Generally, fabrication techniques that are currently used to produce axon-like fibres of a suitable size / shape require precise control and are time consuming. Techniques such as electrospinning and micro-stereolithography mean that current models have long fabrication times and are relatively expensive, which may limit their use in commercialscale research. Furthermore, these fabrication techniques inherently restrict the materials that may be used; certain materials may not be suitable for use with certain techniques. Many models therefore use glass, plastic or extremely stiff polymer-based materials for construction of the fibres. This may result in models that do not accurately reflect the properties of real brain / CNS tissue, which may result in myelination assays providing incorrect conclusions.

[0009] Therefore, it is desirable to develop in vitro models that address at least some of these issues, so that myelination assays can be performed more effectively and / or more accurately.

[0010] SUMMARY

[0011] According to a first aspect of the invention, there is provided a biomimetic device suitable for myelination assays, the device comprising: a plurality of pillars extending from a base, wherein the plurality of pillars and the base comprise a hydrogel material.

[0012] The hydrogel material may comprise a non-photocured polymer. The device may be moulded (i.e. replicated from a mould).

[0013] The hydrogel material may not comprise a photoinitiator (e.g., Irgacure). The hydrogel may comprise a polymer that is cured using a polymerisation initiator. The device may be formed using a mould, the hydrogel material being cured within the mould. Other means of curing may be used to produce the hydrogel material, rather than using photocuring (such as micro-stereolithography). Using a non-photocured polymer may mean that a wider range of polymer materials may be used for the hydrogel. The hydrogel materials may be formed using cheaper or more readily available polymers.

[0014] The hydrogel material may comprise a polymeric constituent and water, wherein the polymeric constituent comprises at least 70% w / w polyacrylamide (PAM) or at least 70% w / w polyethylene glycol (PEG). For example, the polymeric constituent may comprise a poly (acrylamide-co-acrylic acid) in which the copolymer comprises at least 70% acrylamide monomers.

[0015] The polymeric constituent may comprise at least 80% w / w PAM, or 80% w / w PEG, or 90% w / w PAM or 90% PEG. The use of a polymeric constituent comprising at least 70% w / w the relevant polymer may mean that the hydrogel is more biocompatible and / or more straightforward to produce. The polymeric constituent may be regarded as all of the hydrogel material that is not water. The majority of the polymeric constituent may be a hydrogel polymer (e.g., PAM or PEG). The remaining portion of the polymeric constituent may comprise an initiator and / or a catalyst, for example. Each of these additional components may comprise less than 10% or 5% w / w of the polymeric constituent.

[0016] The PAM or PEG-based polymeric constituent may be used in a hydrogel that is not photocured, i.e., the PAM or PEG may be polymerised not via photocuring (e.g., microstereolithography), but via a polymerisation initiator, for example. This may mean that the polymeric constituent does not have to comprise a photoinitiator.

[0017] The plurality of pillars may be arranged in an array. Each of the pillars may be substantially cylindrical. The pillars are coupled at one end to the base and may be free standing (i.e., each pillar is connected at one end to the base, with no further connections along the length of the pillar or at the opposing end). The pillars may be vertical or substantially vertical with respect to the base. The pillars may mimic axons. The geometry, surface and mechanical and chemical properties may be selected to enable investigation of candidate treatments (e.g. drugs and / or biopharmaceuticals) that may influence myelination.

[0018] By looking down on the top of the device (by using established techniques such as confocal microscopy or electron microscopy), the user may obtain a cross-sectional view of the pillars (mimicking axons) and their myelination. By altering the depth of field and / or the focus, different slices may be taken along the height of the pillar. The use of vertical, freestanding pillars may improve the ease with which myelination can be observed compared to devices which use horizontal or scaffold structures (for example, devices produced using electrospinning or printing), where alignment and entanglements can present challenges. The biomimetic device may enable better analysis of the number of layers of myelination, the percentage of each pillar circumference that is myelinated and / or the number of connections between pillars.

[0019] Each of the pillars may have a diameter of less than 10 microns. Each of the pillars may have a diameter 0.1 to 10 microns, or 0.1 to 5 microns, or 1 to 10 microns. For example, the pillars may have a diameter of 3 microns, or 5 microns or 10 microns. The pillars may have a diameter of greater than 10 microns (for example, up to 20 microns). However, diameters of less than 10 microns may better represent neuronal axons found in human brain. The diameter of the pillars may be made arbitrarily large (within the physical limits of the manufacturing apparatus). The pillars may have a diameter corresponding to axon diameters in human cortical white matter (which may range from 0.1 to 10 microns, with the majority of diameters being less than 1 micron). Pillars of less than 10 microns or less than 1 micron may provide a more accurate assay for real neural tissue. The pillars may have a diameter configured to promote myelination. Axons of less than 0.2 microns in diameter may not get myelinated in vivo; while the threshold may be 0.4 microns in vitro. Axons with diameters ranging from 0.2 to 0.8 microns may be myelinated or unmyelinated. Thick axons with diameters larger than 1 micron may be preferably myelinated in vivo and more often fully myelinated in vitro. Each of the pillars may have a diameter of at least 0.2 microns, or at least 0.4 microns, or at least 1 micron, which may assist in forming and removing the pillars from a mould.

[0020] The diameter of each pillar may be the diameter immediately following a manufacturing process (e.g., at the point that the device is removed from a mould). The diameter of each of the pillars may increase once a sample is applied to the device or the device is immersed in fluid (e.g., phosphate buffered saline). The average pillar diameter may increase by 10 to 30%, for example. The plurality of pillars may be arranged on the base so as to account for this increase in diameter.

[0021] Each of the pillars may have a height of 10 to 30 microns, or 10 to 50 microns, 10 to 200 microns, or 20 to 200 microns, or 20 to 50 microns. A height of at least 20 microns may provide a better assay for myelination of axons than existing devices, particularly those that use flat surfaces for cell propagation. A height of less than 50 microns may assist in the manufacturing of the device, particularly via moulding, as it may be easier to remove from the moulds. A height of less than 50 microns may reduce the likelihood that the pillars would collapse.

[0022] Each of the pillars may have a height to diameter aspect ratio of at least 2. Each of the pillars may have a height to diameter aspect ratio of between 2 and 10 or between 2 and 5. For example, the pillars may have a diameter of approximately 3, or 5 or 10 microns, and a height of between 20 and 25 microns. The pillars may have a diameter of between 15 and 20 microns, and a height of between 50 and 150 microns (for example, approximately 100 microns). High aspect ratio pillars (e.g., those with an aspect ratio of at least 5), may provide a better assay for axons. The aspect ratio may be less than 10, particularly for soft / ultrasoft hydrogel materials, in order to reduce the likelihood that pillars will collapse during demoulding.

[0023] Each of the pillars may have an elastic modulus (Young’s modulus) of 0.1 to 100 kPa, or 0.5 to 100 kPa or 0.5 to 50 kPa. The elastic modulus of the pillars may correspond to a characteristic elastic modulus of a biological tissue / material, which may mean that the device provides a better assay for myelination of axons. The elastic modulus may be between 0.1 and 1 kPa, or 0.4 and 0.6 kPa (‘ultrasoft’, which may correspond to brain tissue elasticity); between 3 and 7 kPa, or 4 and 6 kPa (‘soft’, which may correspond to the stiffness of individual neuronal axons); between 15 and 25 kPa (‘intermediate’); or between 40 and 60 kPa (‘stiff’, which may correspond to the stiffness of other body tissues and may be used as a control). Human brain matter may have a typical stiffness of 0.1 to 1 kPa, while individual axons may have a stiffness of approximately 5 kPa. Each of the pillars may have an elastic modulus corresponding to said tissue stiffness - e.g., 0.1 to 10 kPa or 0.1 to 5 kPa. Being able to vary and control the stiffness of the pillars may enable a better or more adaptable myelination assay. Myelinating cells (e.g., oligodendrocytes) may be mechanosensitive, and therefore it may be important to be able to replicate the biomechanical properties of various tissues.

[0024] The distance between nearest neighbour pillars (i.e., an interpillar distance) may be less than 50 microns, or less than 35 microns or less than 20 microns. The interpillar distance may be 1 to 50 microns, 25 to 35 microns, 1 to 20 microns, 3 to 20 microns, 5 to 20 microns, or 5 to 15 microns. The plurality of pillars may be arranged in a vertical / horizontal square grid formation, meaning that the interpillar distance in the x and y directions are equal. An interpillar distance of less than 20 microns or less than 35 microns may more closely match the bundled nature of brain / CNS tissue and or may allow for more myelination / interpillar connections. The interpillar distance may be greater than 20 or 35 microns (and may be arbitrarily large, within the physical limits of the device). However, interpillar distances of less than 50 microns reduce the size of the device and may increase the possibility of one cell contacting multiple pillars. An interpillar distance of at least 1 micron, or 3 microns, or 5 microns or 10 microns may provide sufficient space to promote oligodendrocytes to propagate.

[0025] The hydrogel material may comprise polyacrylamide (PAM). The plurality of pillars and / or the base may be made essentially of PAM. The use of hydrogel materials for the biomimetic device may provide a better assay for myelination of axons than other materials (for example, glass or plastics / polymers). The mechanical properties of PAM (e.g., elastic modulus) may better correspond to in vivo tissues / axons. The porosity of PAM may provide an adequate substrate for cell cultures and may provide a sufficient substitute for myelination to occur on. The use of hydrogels, such as PAM, may provide for faster, cheaper or easier manufacture of biomimetic assays, as the need for more sophisticated microfabrication techniques / tools may be reduced or omitted. Alternatively, the hydrogel material may comprise polyethylene glycol (PEG), which may also provide all or some of the advantages discussed above with regard to PAM. Other biocompatible hydrogel materials may be used.

[0026] The biometric device may further comprise a well, the well surrounding the plurality of pillars and defining a receptacle for holding a sample. The well may be formed as a recess within a container that is placed around the pillars. The container / well may comprise a biocompatible polymer. The container / well may be a siloxane polymer, for example polydimethylsiloxane (PDMS). The well may reduce the volume of sample and / or the number of cells used during a myelination study compared to using a multiwell plate or a Petri dish.

[0027] At least some of the pillars may comprise neuronal molecules encapsulated within the hydrogel material. The molecules may be mixed into the hydrogel material such that it is contained within the hydrogel’s porous structure. The molecules may be an enzyme, a neurotransmitter, a growth agent or a guidance molecule that is a released by neuronal axons. Molecules released by neuronal axons (such as neuregulin-145), small molecule drugs, extracellular vesicles or short interfering RNAs (siRNAs) may be encapsulated within / adsorbed into the pillars and / or the base. The molecules may be slowly released into a sample medium or captured via cell-targeted uptake into myelin-forming cells of the sample, in order to study the effects of such molecules on myelination.

[0028] At least some of the pillars may be coated with a ligand solution. Chemical compounds within the coating may provide biochemical cues to cells in a sample. The coating may promote or inhibit oligodendrocyte (OL) maturation or myelin formation, for example. The ligand solution may comprise a synthetic and non-specific adhesion promoter, which may be used as a control (for example, Poly-D-lysine (PDL)). The ligand solution may comprise a binding partner of integrins which is configured to promote OL differentiation (for example laminin, for use with laminin-binding integrins). The ligand solution may comprise an agent configured to promote OL proliferation and survival (for example, fibronectin).

[0029] The plurality of pillars may comprise two or more subregions, each of the subregions of pillars having at least one of a different:

[0030] (i) pillar diameter;

[0031] (ii) distance between nearest neighbour pillars; and / or

[0032] (iii) pillar height.

[0033] Providing multiple subregions of pillars with different properties within the same biomimetic device may allow for direct comparison of the effects of said properties on a sample. For example, the device may comprise four quadrants, each of the four quadrants comprising pillars with different interpillar distances. A sample of OL cells may be loaded into to the device such that it covers the pillars in all of the subregions. Myelination may then be observed / studied for each of the subregions simultaneously. One of the subregions may have no pillars and may function as a control.

[0034] According to a second aspect of the invention, there is provided a method of manufacturing a biomimetic device suitable for myelination assays, the method comprising: forming a plurality of pillars extending from a base, the plurality of pillars and the base comprising a hydrogel material, wherein forming the plurality of pillars and the base comprises: applying an uncured hydrogel precursor material to a mould; curing the hydrogel precursor material to form a hydrogel material in the mould; and removing the hydrogel material from the mould.

[0035] Using a mould to produce the biomimetic device may enable faster, easier and / or cheaper manufacture of such devices. The mould may act as a ‘master’, enabling many devices to be produced using the same mould. This may mean that the more precise aspects of manufacture (namely, accurately placing and shaping the pillars) are only performed once to create the master, rather than for each device. The use of a mould may reduce the need for precision tools and equipment (for example, microfabrication tools).

[0036] The use of a moulding process may enable hydrogel pillars to be produced, or may at least enable hydrogel pillars with desirable physical characteristics (for example sufficient height) to be produced. Other manufacturing techniques may not be suitable for use with hydrogel. 3D micro-printing techniques may not be suitable for use with hydrogels - e.g., micro-stereolithography requires the printed material to be reactive to light (UV); if a hydrogel material such as PAM does not polymerise in response to UV, then it cannot be used with said technique. Therefore, desirable yet previously unsuitable materials may be used with the moulding process of the present invention.

[0037] The method may further comprise producing the mould. Producing the mould may comprise etching a substrate so as to produce a patterned region, the patterned region comprising a reverse of the plurality of pillars of the biomimetic device.

[0038] The substrate may comprise a silicon wafer. Use of a silicon wafer for the mould may enable the mould to be used many times, due the silicon’s hardness and stiffness. Silicon wafers may also allow for larger aspect ratio features to be etched into the mould compared to other substrates. This may allow for taller and / or higher aspect ratio pillars to be produced. Silicon wafers are commonly used in micromanufacturing, and therefore the use of a silicon wafer may mean that the mould can be readily produced using existing equipment and well-established techniques.

[0039] Patterning the substrate using etching may comprise using photolithography. Ultraviolet (UV) light may be used for the photolithography. The use of photolithography with a silicon wafer may enable sub-micron precision when generating the mould. This may allow pillar diameters and interpillar distances of < 1 micron to be produced.

[0040] The photolithography process may comprise applying a photoresist layer to a surface of the silicon wafer. A photomask may be provided through which the light (e.g., UV light) may be directed, the mask corresponding to a desired pattern of pillars for the biomimetic device. The method may further comprise applying a spacer material to the substrate around the patterned region, such that a well is formed around the patterned region. Applying uncured hydrogel precursor to the mould may comprise filling the well. The spacer material may be preformed into a spacer block, the spacer block having a hole through it that functions as the well. Use of a spacer / well may reduce the amount of hydrogel material used. Use of a spacer / well may allow multiple biomimetic devices to be manufactured concurrently on the same substrate (with each device having its own spacer / well). The spacer may be reused to make multiple devices. The spacer may be peeled from the mould substrate along with the hydrogel material, the hydrogel removed from the spacer, and the spacer washed (using water / ethanol, for example) so that the spacer can be reused.

[0041] The method may further comprise bonding a coverslip (or support) to the top of the hydrogel material. Removing the hydrogel material from the mould may comprise peeling the coverslip and hydrogel from the mould. The coverslip may be a Uonza GelBond® support film, for example. The coverslip may assist in removing the hydrogel from the mould and may support the hydrogel. The hydrogel may remain coupled to the coverslip during use, the coverslip acting as a substrate / support for the biomimetic device.

[0042] Curing the hydrogel precursor material may comprise applying a polymerisation initiator to the hydrogel precursor material while it is in the mould. The polymerisation initiator initiates polymerisation of the hydrogel precursor. The polymerisation initiator may be tetramethylethylenediamine (TEMED). The polymerisation initiator may be added to the hydrogel precursor and mixed while the precursor is in the mould. Once the polymerisation reaction is initiated, it may spread to the entire hydrogel volume.

[0043] The method may further comprise, prior to applying the uncured hydrogel precursor material to the mould, applying a hydrophobic coating to a surface of the mould. The coating may comprise a hydrophobic polymer or a superhydrophobic polymer. The coating may be a silicon-based polymer (e.g., a polysiloxane or a polysilane), a fluoropolymer or a silica-based material (e.g., a silica nano-coating), for example. The hydrophobic coating may be Rain-X. A layer of the hydrophobic coating may be sprayed or otherwise deposited on the surface of the mould. The hydrophobic coating may assist in removing the hydrogel material from the mould. The method may further comprise degassing the uncured hydrogel precursor material. After the precursor material has been applied to the mould, the mould may be placed in a desiccator and / or a vacuum chamber, thereby resulting in bubbles being removed from the hydrogel precursor. Degassing the uncured hydrogel precursor material may comprise vibrating the mould and material. For example, the uncured hydrogel precursor may be placed in a sonic or ultrasonic cleaner. Alternatively and / or additionally, the hydrogel may be degassed during and / or after curing in order to prevent bubbles becoming trapped in the hydrogel.

[0044] The hydrogel material may comprise polyacrylamide (PAM). Other biocompatible hydrogel materials may be used. The hydrogel material may comprise polyethylene glycol (PEG), for example.

[0045] The uncured hydrogel precursor material may comprise a solution with acrylamide (or, where the hydrogel material is not PAM, an equivalent monomer) in a percentage weight per volume of 3 to 12. The composition of the solution may be chosen based upon a desired pillar stiffness. The elastic modulus of the pillars may be controlled by altering the composition of the hydrogel material. For example, a higher concentration of solution used to produce the hydrogel (e.g., concentration of a hydrogel monomers such as acrylamide) may result in stiff pillars, while a lower concentration may result in ultrasoft / soft pillars. Thus, variation in the hydrogel material may allow for devices with a wide range of mechanical properties / stiffnesses to be produced using the same mould and / or methods.

[0046] The method may further comprise providing a well around the plurality of pillars, the well defining a receptacle for holding a sample. The well may be formed as a recess within a container that is placed around the pillars. The container / well may comprise a biocompatible polymer. The container / well may be a siloxane polymer, for example polydimethylsiloxane (PDMS). Where the hydrogel pillars are on / coupled to a coverslip or support, the container may be coupled to the coverslip around the pillars.

[0047] The method may further comprise applying a ligand solution to the plurality of pillars. The ligand solution may comprise Poly-D-lysine (PDL), laminin or fibronectin, for example. The hydrogel pillars may be incubated with the ligand solution. The mould may be configured such that the plurality of pillars comprises two or more sub-regions, each of the subregions of pillars having at least one of a different:

[0048] (i) pillar diameter;

[0049] (ii) distance between nearest neighbour pillars; and / or

[0050] (iii) pillar height.

[0051] For example, the mould may be configured to have four quadrants, each of the four quadrants configured to produce pillars with different interpillar distances. One of the subregions of the mould may be configured to produce a flat subregion (i.e., no pillars).

[0052] The method may comprise manufacturing a plurality of biomimetic devices, each biomimetic device being formed using a mould of a plurality of moulds, each of the moulds comprising a different patterned region of the same etched substrate. The substate may be etched so as to have multiple patterned regions arranged upon it, each of the patterned regions corresponding to a different mould that may be used to produce a biomimetic device. The different patterned regions may have a different pattern, so that each of the moulds produce biomimetic devices with different pillar diameters, interpillar distances and / or pillar heights, for example. The patterned regions may be sufficiently separated such that each can be used as a separate mould simultaneously. Thus, the same substrate may be used to produce multiple biomimetic devices concurrently. Each of the plurality of moulds may be the same or different. A different precursor material / formulation may be applied to each of the plurality of moulds, so that each of the resulting hydrogel materials / devices have different mechanical / chemical properties (e.g., elastic modulus).

[0053] The manufacturing method of the second aspect may be used to produce a biomimetic device according to the first aspect, including any optional features thereof. The biomimetic device may of the first aspect may include any of the features described with reference to the second aspect.

[0054] According to a third aspect, there is provided a method of performing a myelination assay, comprising culturing oligodendrocytes in contact with a biomimetic device according to the first aspect. The method of the third aspect may include any of the method steps and / or features described herein in connection with the second aspect. Culturing the oligodendrocytes in contact with the biomimetic device may comprise a first proliferation stage in a proliferation medium, and a second differentiation stage in a formulation comprising a compound or formulation under test.

[0055] The effect of the compound on myelination of the device may be determined by immunostaining myelin basic protein, and inspection via microscopy. The method may comprise determining a score. Determining a score may comprise assessing the number of pillars contacted with myelin, and / or a wrapping score for any myelin coated pillars.

[0056] DETAILED DESCRIPTION

[0057] Embodiments of the invention will be described, purely by way of example, with reference to the accompanying drawings, in which:

[0058] Figure 1 shows a schematic diagram of a biomimetic device;

[0059] Figure 2 shows a schematic diagram of a biomimetic device;

[0060] Figure 3 shows a scanning electron microscopy image of an array of hydrogel pillars;

[0061] Figure 4 shows a schematic diagram of a biomimetic device being manufactured in a mould;

[0062] Figure 5 shows a schematic diagram of a plurality of biomimetic devices being manufactured on a mould substrate;

[0063] Figure 6 shows a microscopy image of a mould comprising multiple subregions configured to produce different pillar parameters;

[0064] Figure 7 shows a diagram of a photolithographic method for producing a master mould;

[0065] Figure 8 shows z-stacked fluorescent images of a myelinated pillar;

[0066] Figure 9 shows a diagram of the number of contacted pillars during oligodendrocyte propagation; Figure 10 shows a diagram of a myelination wrapping score;

[0067] Figure 1 1 shows scanning electron microscopy images of pillars and oligodendrocytes;

[0068] Figure 12 shows transmission electron microscopy images of myelination around hydrogel pillars;

[0069] Figure 13 shows data illustrating the effects of pillar diameter and interpillar distance on cell attachment;

[0070] Figure 14 shows data the effect of pillar diameter and interpillar distance on wrapping score;

[0071] Figure 15 shows data illustrating the effects of a ligand coating of the pillars;

[0072] Figure 16 shows data illustrating the effects of pillar stiffness; and

[0073] Figure 17 shows data illustrating the effects of benztropine and clemastine

[0074] Biomimetic device

[0075] Referring to Figures 1 and 2, the biomimetic device 100 comprises a plurality of hydrogel pillars 1 10 extending vertically from a hydrogel base 1 12. As discussed further below, the pillars 1 10 and base 1 12 are formed as a single hydrogel component.

[0076] The pillars 1 10 are surrounded by a container 120, the container 120 having a hole through its centre that defines a well 122. A sample 102 comprising oligodendrocyte cells is deposited into the well, such that the sample 102 covers the pillars 1 10. The hydrogel base 1 12 and the container 120 are coupled to a substrate / support 130.

[0077] Referring to Figure 3, a scanning electron microscope (SEM) image of an array of pillars is shown (with a 20 micron scale bar). The array of pillars comprises a plurality of hydrogel pillars 1 10 extending from a hydrogel base 1 12. The pillars 1 10 have a diameter of 18 microns, an interpillar distance of 15 microns and a height of 100 microns (when immersed in liquid). The hydrogel material may be ‘stable’ such that the pillars remain substantially consistent while a myelination assay is being conducted - e.g., the pillars may not shrink during use of the device. However, the hydrogel material may be dehydrated to allow for certain imaging / analysis techniques to be conducted (such as electron microscopy). Shrunk pillars are shown later in the TEM images of Figure 12, for example.

[0078] Device manufacturing methods

[0079] Referring to Figures 4 and 5, a master mould 200 is used to produce a hydrogel-based biomimetic device. The mould 200 comprises a substrate (such as a silicon wafer), onto which is etched one or more patterned regions 202. The patterned regions on the substrate may be formed using deep reactive ion etching (DRIE). The regions etched by DRIE may be defined by a photolithographically defined mask (e.g., formed from resist). A hard mask may be defined by transferring a pattern from a photolithographically patterned resist layer into a hard inorganic material (such as silicon oxide). Such a hard mask may be eroded more slowly by DRIE and may therefore be more suitable for deep and high aspect ratio patterning.

[0080] The patterned regions 202 comprise a plurality of holes 204 arranged in an array. The holes 204 are configured such that, when uncured hydrogel precursor material 1 14a is cured within the mould 200, hydrogel pillars 1 10 are formed within the holes 204. The substrate may be coated with a release layer to improve the ease with which a hydrogel may be removed from the mould. The release layer may comprise a hydrophobic layer - for example a hydroxy termination PDMS material, a fluoropolymer or a fluorinated silane. The release layer may be thin - for example less than 1 micron or 500nm in thickness.

[0081] A plurality of patterned regions 202 are provided on the surface of the substrate 200. A spacer 210 comprising a hole in its centre is placed around each of the patterned regions 202, thereby defining a well 212. As such, each patterned region 202 / well 212 functions as an individual mould for a different biomimetic device. Uncured hydrogel precursor material 1 14a may be poured into each of the wells 212 and allowed to cure. The cured hydrogel 1 14 thus has an imprint of the patterned region 202 / well 212, thereby enabling the moulding of the hydrogel pillars 1 10 and base 1 12.

[0082] In one example, the hydrogel material 1 14 comprises polyacrylamide (PAM). The uncured hydrogel precursor material 1 14a comprises acrylamide (Aam), bis-acrylamide (bis-Aam) and acrylic acid (AA). In one example, these components were mixed together in phosphate buffered saline (PBS) at different ratios (measured as a percentage weight per volume) to produce hydrogel material with different stiffnesses (see table 1 below). In one example, ultrasoft (0.52 ± 0.02 kPa), soft (5.07 ± 0.13 kPa), intermediate (20.22 ± 1.85 kPa) and stiff (54.85 ± 2.78 kPa) hydrogel pillars were fabricated.

[0083] Table 1: Hydrogel material compositions with associated elastic modulus

[0084] The hydrogel material pH was raised to 8 using 10N NaOH, fluorescently labelled using fluorescein isothiocyanate (FITC) Dextran (5% v / v from 1 mg / ml stock solution) and kept in the fridge as a stock solution. Applying FITC labelling to the hydrogel may allow for easier observation of the pillar structures using fluorescent upright confocal microscopy, as discussed further below. A fresh 10% stock solution of ammonium persulfate (APS) was added to the uncured hydrogel precursor material 114a (a comonomer solution) at a final concentration of 0.1% and purged with nitrogen for 1 to 2 minutes to remove oxygen.

[0085] In some examples, the hydrogel may swell when immersed in aqueous solutions. A 26.03 ± 4.33% increase in the pillar diameter was observed between the mould and the actual hydrogel pillars when immersed in PBS. This may be taken into consideration when designing / manufacturing the mould and the biomimetic devices (e.g., by ensuring that the interpillar distances are sufficiently large to account for this swelling).

[0086] PDMS spacers 210 were placed on the mould substrate 200 around the patterned regions 202. 120 microlitres of the uncured hydrogel precursor material 114a was dispensed inside the wells 212 of the spacers 210. The mould substrate 200 was put in a desiccator under vacuum for 5 to 40 minutes (depending on the power of the pump) to remove all the air inside the wells 212. The mould substrate 200 was then placed in an ultrasonic cleaner for 1 minute to remove any trapped bubbles. Tetramethylethylenediamine (TEMED) was then added to the uncured hydrogel precursor material 114a at a working concentration of 0.1%, gently mixed and covered immediately with cut Lonza GelBond® support film sheets 130. This specific material for the support / backing is not essential, and other materials can be used. The comonomer solutions were kept covered for 10 to 45 minutes until complete polymerisation. The mould substrate 200 was then incubated in PBS for 5 minutes to facilitate the peeling of the hydrogel 114 from the mould. After removal of the supports 130 from the mould substrate 200, the hydrogel 114 was washed thoroughly in PBS and kept in the fridge until use.

[0087] Autoclaved PDMS containers 120 were glued (using uncured PDMS) to the support sheets 130 overnight in an incubator (with PBS to avoid the hydrogel drying), the containers 120 forming wells 122 around the hydrogel pillars 110. The biomimetic devices 100 were then sterilized under UV for 40 minutes inside a tissue culture hood. The hydrogel 114 was washed in 0.1 M of 4-morpholineethansulfonic acid (MES) hydrate with 0.5 M NaCl in de-ionised water (MES buffer, pH 6) for 10 minutes. The devices 100 were then incubated for 30 minutes in MES buffer (pH 6) supplemented with 0.1M N-hydroxysuccinimide (NHS) and 0.2M N-(3-dimethylaminopropyl)-N’- ethyl-carbodiimide hydrochloride (EDC).

[0088] Referring to Figure 6, a microscopy image of a mould with a plurality of patterned regions is shown. The mould (a silicon wafer substrate) comprises 16 patterned regions (which may be regarded as smaller sub-arrays). The 16 patterned regions comprise four replicates of four different pillar geometries: flat, and interpillar distances of d = 5, 10 or 15 microns. The pillar height and diameter are kept constant (except for the flat region). A spacer may be placed around any subset of the patterned regions. For example, a spacer could be placed around one of the four replicates, such that a biomimetic device is formed comprising four quadrants, each quadrant having one of the different pillar geometries. Alternatively, a spacer may be placed around the whole of the pattered region / array, such that a single device comprises sub-regions with a variety of different characteristics (e.g., pillar diameter and interpillar distance), such as the 16 sub-arrays of Figure 6. A device or multiple devices may be produced having a wide range of characteristics based on the patterned regions of the substrate. Referring to Figure 7, a master mould is produced using photolithography. UV light 220 is applied to a mould substate 200. The UV light 220 is passed through a mask 222. The mask 222 comprises an array of holes with a diameter D and spacing distance d. The configuration of the holes on the mask 222 corresponds to the desired patterned region 202 that is to be formed on the substrate 200. A photoresist layer 224 is applied to the surface of the substrate 200. The photoresist layer 224 and UV light 220 results in the pattern of the mask 222 being transferred to the surface of the substrate 200. Following the photolithography, the photoresist layer 224 is developed and the substrate is etched, thereby forming the patterned region 202. The remaining photoresist layer 224 is then removed, leaving the master mould (comprising the patterned region 202 on the substrate 200).

[0089] In one example, the mould substrate 200 was a 3-to-4-inch (7.5 to 10.5 cm) diameter silicon wafer. A thin layer of positive photoresist (MICROPOSIT SI 805) was coated on the surface of the silicon wafer; the photoresist 224 was then exposed to UV. In some examples, the photoresist 224 was exposed to UV without using a mask 222 (using a Heidelberg DWU66+ direct write lithography system). In other examples, a photomask was used with a mask aligner (a Quintel mask aligner). The silicon wafer was then developed and mounted on a ceramic carrier using Santovac5 oil in preparation for deep reactive ion etching (DRIE) (using an STS inductively coupled plasma (ICP) advanced silicon etching (ASE) system). The number of etching cycles was chosen dependent upon the desired pillar height. The photoresist 224 was removed using dimethylsulfoxide (DMSO) and cleaned for 5 minutes using O2 plasma (Diener plasma asher). To ease the demoulding process of the hydrogel material, Rain-X solution was applied on the surface of the mould substrate 200.

[0090] Oligodendrocyte sample preparation

[0091] The biomimetic device of the present invention may be used to study myelination. A sample comprising oligodendrocytes (OLs) may be deposited into the device.

[0092] In one example, immunomagnetic isolation of O4+oligodendrocytes was performed using 50 nm superparamagnetic anti-04 microbeads. Cortical brain cells were extracted from two P6-P7 Sprague Dawley rat pups. A fresh solution of magnetic cell sorting buffer (MCS) was prepared by adding 1 mg / ml of D-glucose and 5 mg / ml (0.5%) of bovine serum albumin (BSA) into 0.5 mM ethylenediaminetetraacetic acid (EDTA)- PBS. Cells were resuspended in 85 microlitres of MCS buffer and 15 microlitres of anti- 04 bead solution per 107 cells, incubated for 15 minutes in the fridge, washed in MCS buffer and centrifuged at 200xg for 10 minutes. Cells were then resuspended in 500 microlitres of MCS buffer per 107 cells, passed through a magnetic separation (MS) column using a MiniMACS separator bound to a MACS multistand (separation apparatus available from Miltenyi Biotec) and washed 3 times with 500 microlitres of MCS buffer inside the column to remove unlabelled cells. Finally, the column was removed from the separator and flushed out with cell culture medium using a plunger.

[0093] Myelination studies

[0094] In order to test the function of the biomimetic device, tests were carried out to observe and analyse myelination caused by oligodendrocytes deposited into the device.

[0095] In an example study, the oligodendrocyte culture was separated into three distinct phases for analysis: (1) proliferation at day 2, (2) differentiation at day 7, and (3) myelination at day 14.

[0096] At day 0, primary rat OLs were counted and seeded at a density of 100,000 cells / cm2. The cells were then cultured in proliferation medium for 2 days. In one example, the proliferation medium consisted of Dulbecco’s modified Eagle’s medium (DMEM) mixed with neurobasal medium at a 1 : 1 ratio supplemented with IX N-2 supplement, IX B27 supplement, 5 pg / ml N-acetyl cysteine (NAC), 2 pg / ml (5 uM) forskolin, 10 ng / ml biotin, 10 ng / ml ciliary neurotrophic factor (CNTF), 1 ng / ml NT-3, 10 ng / ml platelet-derived growth factor (PDGF-AA) and 1% penicillin streptomycin. After 2 days, the proliferation medium was changed to a myelination medium by removing the growth factor PDGF-AA and adding 30 ng / ml T3.

[0097] In order to analyse the propagation of the OLs and the myelination, the cells may be fixed and immunocytochemistry (ICC) performed. In one example, cells were washed in PBS, fixed for 15 minutes in 4% paraformaldehyde (PFA) at room temperature and washed in PBS 3 times. The cells were then permeabilized using 0.1% saponin for 10 minutes and blocked using 5% normal goat serum (NGS) in PBS (blocking buffer) for 1 hour at room temperature. The cells were then incubated with the primary antibody rat anti-MBP at a concentration of 1 : 100 in blocking buffer for 1 hour at room temperature. Samples were washed 3 times in PBS and incubated with the secondary antibody goat anti-rat IgG 1 :500 in blocking buffer for 1 hour at room temperature. Cells were finally washed 3 times in PBS and stained with 1 pg / ml for 10 minutes at room temperature and stored in the fridge until imaged.

[0098] Referring to Figure 8, fluorescent images were acquired with an upright confocal microscope (Leica SP8 or Zeiss LSM 980) using a 25x water-dipping objective (NA > 0.95). Z-stacks were acquired with a step size of 2 microns and the images were processed using ImageJ. To visualise the fine details of myelin around the pillars, the z-step size can be reduced to 0.5 microns to ensure accurate 3D reconstruction of myelin (however, this will increase imaging acquisition time, which may not be suitable for a large number of pillars). Confocal z-stacks spanning the full height of the pillars (but excluding the base sections where majority of the cell bodies lie) were captured and a maximum intensity projection image was generated.

[0099] Referring to Figures 9 and 10, two main parameters for evaluating OL wrapping around pillars were quantified. The number of contacted pillars is shown in Figure 9, and the ‘wrapping score’ is shown in Figure 10. In contrast to Schwann cells, OLs can myelinate multiple axons simultaneously. Therefore, the average number of pillars that one OL is capable of contacting was quantified using the observed number of contacted pillars. The wrapping score (0-3) corresponds to the percentage of myelin coverage around the pillars with a score of 0 indicating no wrapping while a score of 3 implies a fully wrapped pillar (showing a complete ring). The number of fully contacted pillars was calculated by dividing the number of pillars having a score of 3 by the total number of cells.

[0100] In addition to fluorescence microscopy, the biomimetic device is compatible with electron microscopy techniques. The periodic nanostructure of myelin, consisting of concentric multilayers surrounding axons, is critical for the facilitation of rapid signal conduction in axons. Electron microscopy may effectively visualise and characterise this multilayered structure.

[0101] In one example, the cells were washed in PBS and fixed in 2% PFA / 1 .5% glutaraldehyde (both electron microscopy grade) in 0.1M phosphate buffer for 30 minutes at room temperature. After washing in PBS, cells were stained in 1% osmium tetroxide / 1 .5% potassium ferricyanide for 1 hour at 4°C and washed 3 times. For scanning electron microscopy (SEM), samples were stained with 1% tannic acid in 0.05M cacodylate buffer for 40 minutes, washed and dehydrated in a series of ethanol dehydration treatments (70%, 90% and 2x 100% ethanol) for 10 minutes each. Finally, critical point drying was applied and the samples were mounted coated with gold (10-15 nm thick) before imaging. For transmission electron microscopy (TEM), samples were stained with thiocarbohydrazide, 2% osmium tetraoxide, 1% uranylacetate, lead aspartate at 60°C and dehydrated in a series of ethanol dehydration treatments (70%, 90% and 2x 100% ethanol) for 10 min each. Dehydradation was completed in propyleneoxide (PO) and samples were embedded in 1 : 1 mix of PO and eponresin followed by 100% eponresin overnight. Samples were placed in 100% fresh eponresin and cured at 60°C overnight. Ultrathin sections (70 nm) were obtained and visualised with TEM.

[0102] Referring to Figure 11, SEM images of OLs around 5 micron diameter pillars after a week of culture are shown. SEM may allow a 3D overview of OL-mediated wrapping of pillars to be observed.

[0103] Referring to Figure 12, TEM images of OLs around 5 micron (top row) and 3 micron (bottom row) diameter pillars after two weeks of culture are shown. As shown in the images, the freestanding, vertical pillar structure of the biomimetic device may enable horizontal cross-sections of myelin wraps to be obtained using TEM. The pillars may shrink due to hydrogel dehydration during sample preparation. However, multilayered myelin around the pillars with a nanostructure similar to that found in vivo can still be observed.

[0104] The biomimetic device may be used to investigate the effect of pillar geometry on myelination. In one example, the biomimetic device was manufactured to have a plurality of subregions with different pillar features (e.g., pillar diameter and interpillar distance). In one example, the number of cell bodies per field (250 x 250 pm2) were quantified by counting Hoechst-positive nuclei co-localised with the cell proliferation marker Ki67 on day two of the culture.

[0105] Referring to Figures 13 to 14, the results of investigation into the effect of pillar geometry on myelination. Biomimetic devices were manufactured with various properties (e.g., pillar diameter and interpillar spacing). Some devices comprised subregions of pillars with different properties, using a mould like that shown in Figure 6.

[0106] Referring to Figure 13, significant differences between the number of cells in regions with pillars compared to the flat regions were found, independent of the micropillar diameter D. No significant differences in cell number were found between the different interpillar distances when the pillar diameter was kept constant. These results indicate that higher cell densities can be achieved when using pillar arrays independent of their geometry (at least within the range of values tested here). This suggests that spatial constraints not only induce OL differentiation but may also influence OL survival.

[0107] Referring to Figure 14, the effect of pillar geometry on OL differentiation was investigated using the myelin basic protein (MBP) as a marker of mature OLs at day seven of the culture. Each pillar was assigned a wrapping score between 0 and 3 and the percentage of pillars that display 0, 1, 2, or 3 for each interpillar distance and each pillar diameter combination is depicted. For each of the plurality of pillars with a specific micropillar diameter D, the percentage of pillars that scored 1 or 2 mostly did not change when the interpillar distance d was modified. However, the proportion of pillars that scored 3 increased as the interpillar distance increased for 3 and 5 micron diameter pillars. However, this trend was not observed for 10 micron pillars, with the vast majority exhibiting a score of 3 (70.87 ± 13.36%, 76.16 ± 10.31% and 82.5 ± 8.43% when d = 5, 10 and 15 microns respectively).

[0108] The biomimetic device may be used to study the effect of biochemical cues on OL maturation and myelin formation. In one example, the devices 100 were washed thoroughly in PBS and incubated for at least 1 hour at 37°C with a ligand solution. A desired ECM proteins - poly-D-lysine (PDL) or fibronectin - were diluted in 50 mM 4- (2-hydroxyethyl)-l -piperazineethanesulfonic acid (HEPES) buffer (pH 8) at a concentration of 1, and 0.1 mg / ml respectively. Laminin was coated on top of PDL at a concentration of 0.02 mg / ml (in PBS) for at least 2 hours at 37°C. The hydrogel materials 114 were finally rinsed 3 times with PBS and stored immersed in PBS in the fridge for up to a week.

[0109] Referring to Figure 15, the results of an investigation into the effect of biochemical cues are shown. The investigation was performed using devices comprising an array of pillars with a diameter of 5 microns, an interpillar distance of 10 microns and a height of ~ 21 microns, the pillars comprising various coatings. While the different coatings did not have an effect on cell attachment (no statistical differences between cell numbers), laminin promoted myelination, measured as an increase in the percentage of fully covered pillars when compared to PDL coating. No statistical differences in the percentage of fully covered pillars were observed between PDL and fibronectin coatings. However, OLs cultured on fibronectin-coated pillars were able to contact more pillars than in the PDL-coated pillar control group. No differences were observed in the wrapping score 0, 1 and 2 of the different tested coatings except in the score 2 between PDL and laminin. These data suggest that fibronectin does not exert a significant effect on wrapping micropillars while laminin appears to promote wrapping. Since nonproliferative 04+ cells were used (which had already started the differentiation process), fibronectin did not prevent the differentiation / myelination process and provided results similar to the control group (PDL).

[0110] The biomimetic device may be used to study the effect of mechanical signals, such as substrate stiffness, on myelination. The stiffness of the hydrogel pillars may be characterised using atomic force microscopy (AFM). In one example, a JPK Nanowizard Cellhesion 200 module mounted on an inverted microscope (Olympus FV1000) was used. A 25 or 50 micron polystyrene bead was glued on a cantilever (MLCT, Brucker AFM probes) having an appropriate nominal spring constant k (ranging from 0.01-0.6 N / m). Hydrogel samples were glued into a Petri dish and indentation measurements (in a 3x3 grid) were performed in PBS; force-indentation curves were then processed. The stiffness (Young's modulus) of the hydrogel material was calculated as the average of 9 measurements per gel.

[0111] Referring to Figure 16, the results of an investigation into the effect of hydrogel stiffness are shown. The investigation was performed using devices comprising an array of laminin-coated pillars with a diameter of 5 microns, an interpillar distance of 10 microns and a height of ~ 21 microns. The hydrogel used in each of the biomimetic devices was formulated to have three different stiffnesses: ultrasoft (0.5 kPa), soft (5 kPa) and stiff (50 kPa). No differences in cell number between different pillar stiffnesses were observed. However, there was a lower percentage of fully covered pillars and a lower number of OL-contacted pillars on ultrasoft gels compared to soft and stiff gels. The ultrasoft gels also presented a higher percentage of pillars with a wrapping score of 0, 1 and 2 compared to soft and stiff gels. Moreover, no statistical differences were observed between soft and stiff gels, suggesting that only ultrasoft substrates impede the ability of OLs to wrap around the hydrogel pillars.

[0112] The biomimetic device may be used to study the effect of drugs on myelination was investigated. In one example, OL cells were cultured in proliferation medium for the first 2 days as usual. Then, the medium was switched to the differentiation formulation, but without the addition of T3. Instead, DMSO, benztropine at 1 pM or clemastine at 10 pM were used. All conditions included the same amount of DMSO (1 : 1000) and the medium was changed every other day.

[0113] Referring to Figure 17, the results of an investigation into the effect of benztropine and clemastine are shown. An increase in the percentage of pillars fully wrapped and in the number of OL-contacted pillars for both drugs when compared to the control (DMSO) was observed, without statistical differences in cell numbers. The percentage of pillars with a wrapping score of 0 or 1 was significantly higher when OLs were treated with DMSO compared to benztropine and clemastine, but no differences was found in pillars with a score of 2.

[0114] Although specific examples have been described, the skilled person will appreciate that variations are possible, within the scope of the invention, which should be determined with reference to the accompanying claims.

Claims

CLAIMS1 . A method of manufacturing a biomimetic device suitable for myelination assays, the method comprising: forming a plurality of pillars extending from a base, the plurality of pillars and the base comprising a hydrogel material, wherein forming the plurality of pillars and the base comprises: applying an uncured hydrogel precursor material to a mould; curing the hydrogel precursor material to form a hydrogel material in the mould; and removing the hydrogel material from the mould.

2. The method of claim 1, further comprising producing the mould, producing the mould comprising: patterning a substrate with an etch so as to produce a patterned region, the patterned region comprising a reverse of the plurality of pillars of the biomimetic device.

3. The method of claim 2, wherein the substrate is a silicon wafer.

4. The method of claim 2 or claim 3, wherein patterning the substrate comprises using photolithography.

5. The method of any of claims 2 to 5, further comprising: applying a spacer to the substrate around the patterned region, such that a well is formed around the patterned region, wherein applying uncured hydrogel to the mould comprises filling the well.

6. The method of claim 5, further comprising: bonding a coverslip to the top of the hydrogel material, wherein removing the hydrogel material from the mould comprises peeling the coverslip and hydrogel from the mould.

7. The method of preceding claim, wherein curing the hydrogel precursor material comprises applying a polymerisation initiator to the hydrogel precursor material while it is in the mould.

8. The method of any preceding claim, further comprising, prior to applying the uncured hydrogel precursor material to the mould, applying a hydrophobic coating to a surface of the mould.

9. The method of any preceding claim, further comprising degassing the uncured hydrogel precursor material.

10. The method of any preceding claim, further comprising providing a well around the plurality of pillars, the well defining a receptacle for holding a sample.1 1. The method of any preceding claim, wherein the mould is configured such that the plurality of pillars comprises two or more subregions, each of the subregions of pillars having at least one of a different:(i) pillar diameter;(ii) distance between nearest neighbour pillars; and(iii) pillar height.

12. The method of any preceding claim, the method comprising manufacturing a plurality of biomimetic devices, each biomimetic device being formed using a mould of a plurality of moulds, each of the moulds comprising a different patterned region of the same etched substrate.

13. A biomimetic device suitable for myelination assays, the device comprising: a plurality of pillars extending from a base, wherein the plurality of pillars and the base comprise a hydrogel material, the hydrogel material comprising a polymeric constituent and water, wherein the polymeric constituent comprises at least 70% w / w polyacrylamide PAM or at least 70% w / w polyethylene glycol PEG.

14. The biomimetic device of claim 13, wherein the device is produced according to the method of any preceding claim, for example moulded and / or not photocured.

15. The biomimetic device of claim 13 or claim 14 wherein each of the pillars has a diameter of less than 10 microns.

16. The biomimetic device of any of claims 13 to 15, wherein each of the pillars has a height of 10 to 30 microns.

17. The biomimetic device of any of claims 13 to 16, wherein each of the pillars has a height to diameter aspect ratio of at least 2.

18. The biomimetic device of any of claims 13 to 17, wherein each of the pillars has an elastic modulus of 0.1 to 100 kPa.

19. The biomimetic device of any of claims 13 to 18, wherein the distance between nearest neighbour pillars is 5 to 15 microns.

20. The biomimetic device of any of claims 13 to 19, further comprising a well, the well surrounding the plurality of pillars and defining a receptacle for holding a sample.

21. The biomimetic device of any of claims 13 to 20, wherein at least some of the pillars comprise neuronal molecules encapsulated within the hydrogel material.

22. The biomimetic device of any of claims 13 to 21, wherein at least some of the pillars are coated with a ligand solution configured to promote myelination.

23. The biomimetic device of any of claims 13 to 22, wherein the plurality of pillars comprises two or more subregions, each of the subregions of pillars having at least one of a different:(i) pillar diameter;(ii) distance between nearest neighbour pillars; and / or(iii) pillar height.

24. A method of performing a myelination assay, comprising culturing oligodendrocytes in contact with a biomimetic device according to any of claims 13 to25. The method of claim 24, wherein culturing the oligodendrocytes in contact with the biomimetic device comprises a first proliferation stage in a proliferation medium, and a second differentiation stage in a formulation comprising a compound or formulation under test.

Images