Method and device for intracellular delivery

Abstract

The invention generally relates to the field of intracellular delivery In an aspect of the invention, there is provided a method for delivering a cargo into a cells, the method comprising providing a mixture comprising of the cell and the cargo suspended in a viscoelastic fluid, and deforming the cells in the mixture, the deformation creates transient pores in the cell's membrane allowing delivery of the cargo into the cells.

Description

DESCRIPTIONTITLE OF INVENTION: METHOD AND DEVICE FOR INTRACELLULAR DELIVERYFIELD OF THE INVENTION

[0001] The invention generally relates to the field of intracellular delivery.BACKGROUND

[0002] Intracellular delivery of functional substances — such as drugs, genetic material, or other molecules — has been widely adopted across various biological and medical fields, including drug delivery, vaccines production, cell imaging, cell reprogramming and gene editing.

[0003] Intracellular delivery strategies can be broadly categorized into two approaches: carrier-mediated transportation and membrane disruption - or as viral vector methods, chemical methods and physical methods according to the different delivery principles. Carrier- mediated transportation uses carriers such as viruses, lipid layers, virus-like particles, and polymers to encapsulate biomolecules. These carriers facilitate the transfer of their contents into cells via viral infection, endocytosis, or fusion. For instance, viruses have been proven effective in transfection for gene editing. Lipofection has been utilized for COVID-19 vaccine manufacturing. However, these methods have at least one of the following drawbacks: toxicity, limited DNA packaging capacity, high cost, safety concerns, and low delivery efficiency in suspension cells.

[0004] Generally, viral vector methods and chemical delivery methods package therapeutics or genes by modified virus, liposome, and polymer, followed by injection into organism. The packaged cargos deliver therapeutic agents through mechanisms such as infection, endocytosis, and fusion with the cell membrane. Some of these methods have been successfully applied in vaccine production and CAR T cell therapy.

[0005] Nevertheless, concerns regarding cytotoxicity, risk of insertional mutagenesis, and high manufacturing costs present significant obstacles for viral vectors. Furthermore, chemical delivery methods are often limited by their low delivery efficiency in suspension cells and associated cytotoxicity.

[0006] Membrane disruption involves creating temporary pores in the cell membrane to facilitate biomolecule entry. Microinjection is one of the earliest techniques in this category. Inthis process, a cell is held in position by a micromanipulator, and a microinjector then delivers biomolecules into the cell. This method has been proven effective for gene editing in a small number of cells. Nevertheless, its low throughput restricts its applicability in medical and biological fields. In addition, various external energy sources, such as electrical, thermal, optical, ultrasonic, and mechanical stimuli, have been explored to create transient pores in cell membranes. Among these, electroporation is a widely used laboratory technique, where an electric field induces temporary pores in the membrane. However, excessive field strength or prolonged pulse duration can cause irreversible membrane damage and cell death, making it challenging to balance delivery efficiency and cell viability. Consequently, this balance remains difficult to regulate.

[0007] In recent years, mechanical deformation methods, particularly cell squeezing and shearing, have emerged as effective intracellular delivery techniques with minimal impact on cell function and viability. These approaches employ microfluidic chips featuring single or parallel constrictions to enable the intracellular transport of fluorescent molecules, RNA, DNA, and plasmids. Notably, the CRISPR-Cas9 gene editing system has been successfully implemented using this methodology, underscoring its potential for advancing biological research. Furthermore, the throughput can be enhanced by replacing microfluidic constrictions with filter membranes containing thousands of micropores. However, since effective cell squeezing requires micropore diameters smaller than those of the target cells, challenges such as pore clogging and cell loss persist. To address this, Majid et al. developed microfabricated silicon nitride (Si N) microsieves, which enhance cell viability while maintaining delivery efficiency. Nevertheless, the specialized microfabrication processes involved in producing such devices hinder their accessibility for biological laboratories lacking advanced manufacturing infrastructure.

[0008] Physical methods for intracellular delivery involve generating transient membrane disruptions. Microinjection uses a fine glass needle to directly inject cargo into the cytoplasm or nucleus of a cell. This technique is commonly applied in genetic manipulation, cloning, in vitro fertilization, and single-cell analysis.

[0009] However, gene therapies often require millions of cells, making it challenging to scale up the microinjection technique for large scale applications. To address these challenges, various methods such as electroporation, magnetic microrobot, photoporation, acoustofluidic sonoporation, and mechanoporation, have been proposed to enhance the throughput of transient membrane disruption techniques. Among these methods, electroporation is notable for its high efficiency and throughput. Additionally, commercial electroporation machines have been developed for research purposes.

[0010] However, while electroporation is effective for introducing small molecules, ions, and nucleic acids, it may be less efficient for larger molecules or particles. Increasing the voltage can create larger pores in the cell membranes, thereby enhancing delivery efficiency.

[0011] However, the strong electric field can also compromise cell viability. Over the past two decades, mechanoporation methods have proven effective in the field of intracellular delivery. Specifically, microchannels with narrow constrictions apply compressive and shear forces to generate membrane disruptions at stress concentration points. Sub-microscale needle arrays can puncture the cell membrane to create large pores for accommodating large molecules.

[0012] Nevertheless, direct physical contact between cells and microstructures can introduce additional problems. The diameters of cells within the same cell line can vary. Squeezing conditions that are suitable for smaller cells may lead to cell death in larger cells.

[0013] For cell types with different size, the microstructure size should also be readjusted to maintain a consistent delivery efficiency. Compared to direct physical contact-induced membrane disruption, hydrodynamic methods, such as vortex shedding, utilize micropost arrays with spacings larger than a cell’s diameter, which significantly reduces the risk of clogging.

[0014] However, current hydrodynamic methods typically exhibit low uniformity in terms of membrane permeabilization, and thus result in relatively low delivery efficiency. In recent years, viscoelastic fluid was found excellent to apply stress on cells for deformability characterization with a low flow velocity. This phenomenon has also been applied for intracellular delivery.

[0015] Chung et al. performed successful RNA transfection using a micro constriction with viscoelastic fluid. However, the constriction width was designed to be 75% of the average diameter of cells to ensure high efficiency. Therefore, the small constriction width may result in low viability for high transfection efficiency. Mehmet et al. showed a continuous, contact- free intracellular delivery method using a constriction wider than cell diameter and viscoelastic fluid. Although the cell viability was improved to 74%, it is still difficult to strike a balance between high viability and transfection efficiency.

[0016] There is thus a need for a more efficient intracellular delivery method that overcomes the drawbacks of the prior art. Furthermore, other desirable features and characteristics will become apparent from the subsequent detailed description and the appended claims, taken in conjunction with the accompanying drawings and this background of the disclosure.SUMMARY

[0017] In an aspect of the invention, there is provided a method for delivering a cargo into a cells, the method comprising providing a mixture comprising of the cell and the cargo suspended in a viscoelastic fluid, and deforming the cells in the mixture, the deformation creates transient pores in the cell’s membrane allowing delivery of the cargo into the cells.

[0018] By “cargo”, it is meant to include any molecule or moiety intended for delivery into a cell or cells.

[0019] In various embodiments, the cargo is delivered into a plurality of cells suspended in the viscoelastic fluid.

[0020] By “viscoelastic fluid”, it is meant to include any fluid that exhibits both viscosiy and elasticity — in other words, it deforms under shear, but it also has a tendency to return to its original form when the stress is removed, much like a solid. The elastic of the fluid depends on the molecular weight and concentration of polymer.

[0021] Advantageously, as will be described below, the viscoelastic fluid allows for the deformation of the cells allowing entry of the cargo across the cell membrane. By “deformation”, it is meant to include any manipulation or structures put in place to allow the cells to deform, i.e. change its shape.

[0022] In various embodiments, the cell is deformed by passing the mixture through a micro-post array, the micro-post array comprises a first set of a plurality of microposts, wherein a pair of adjacent microposts forming a constriction point and the width of each constriction point is equal to or greater than the diameter of the cell, reducing direct contact between the cell and the microposts. The width should be 0% to 20% larger than the average cell diameter. The cell being the cell which the cargo is intended to be delivered into. Advantageously, this provided excellent results for cargo delivery into the cell and prevents clogging

[0023] In various embodiments, a micropost array consists of a grid of microscopic, vertical pillars, typically arranged on a flat substrate. Here, micro-constriction means the narrow space between the microposts. As can be seen in Figure 1 F, two microposts are included. The microconstriction or constriction point is the place where the cell was passed through (Figure 1F iii).

[0024] In various embodiments, the width of each constriction point in the first set of microposts is equal to or greater than the diameter of the cell, preferably between 4pm to 28pm, preferably 16pm to 28pm, more preferably 20pm. These dimensions are used in the experiments described here, e.g. for MDA-MB-231 cells. The width required should be larger than the average cell diameter to reduce clogging problem. As such, if the cells are smaller (8pm), the constriction width can be smaller (8 pm). If the cells are larger (30 pm), the constriction width can be larger (40 pm).

[0025] In various embodiments, the method further comprises passing the mixture through a second set of a plurality of microposts prior to the first set of micropost, wherein the width of each constriction point in the second set of microposts is between 32pm to 24pm. In various embodiments, constrictions with a width of 20 pm can also realize the delivery.

[0026] In various embodiments, the viscoelastic fluid comprises a biocompatible polymer dissolved in a Newtonian fluid. In various embodiments, the polymer may be any one of polyethylene oxide (PEO), guar gum, xanthan gum, polyvinyl alcohol, hyaluronic acid, alginate or methylcellulose. In various embodiments, the Newtonian fluid may be PBS.

[0027] In various embodiments, the concentration of the biocompatible polymer is 0.5 to 20mg / mL in the Newtonian fluid.

[0028] In various embodiments, the flow rate of the mixture the micro-post array is less than 100pL / min, preferably 50pL / min. In various embodiments where there are a plurality of micro-constrictions, the flow rate can be much higher (over 10 mL / s).

[0029] In various embodiments, the cargo concentration in the mixture is between 10pg / mL to 1000pg / mL, preferably 300pg / mL. The concentration of the cargo depends on the dosage of the cargo delivered into cells. If more cargo needs to be delivered into the cells, then the cargo concentration can be increased. For instance, the concentration of FITC-dextran ranges from 10 to 1000 pg / mL. But for RNA, 0.5 pg / mL is enough for observation.

[0030] In another example of the invention, the cells are deformed by passing the mixture through a stainless steel filter.

[0031] In various embodiments, the viscoelastic fluid is a biocompatible polymer in a Newtonian fluid, wherein the polymer is polyethylene oxide (PEO), guar gum, xanthan gum, polyvinyl alcohol, hyaluronic acid, alginate or methylcellulose, and the concentration of the polymer in the Newtonian fluid is between 0.5 to 20 mg / mL. In this embodiment, the polymer used is PEO and the high molecular weight of the PEO used here is 2000 kDa. In the example above of a micropost array, the molecular weight of PEO used in micropost array is 600 kDa. Therefore, the concentrations in these examples are naturally different. If the PEO molecular weight is high, the concentration should be lower to adjust the elasticity of the fluid.

[0032] In various embodiments, the flow rate of the mixture through the filter is between 2mL / min to 5ml_ / min. If the diameter of the stainless-steel filter is increased, the flow rate should also be increased to ensure enough deformation of cells.

[0033] In various embodiments, the aperture size of the filter equal to or greater than the diameter of the cell. In various embodiments, the aperture size should be 0% to 20% larger than the average cell diameter.

[0034] In various embodiments, the aperture size of the filter is between 12pm to 25pm.

[0035] In various embodiments, the filter comprises a plain weave mesh or a Dutch weave mesh.

[0036] In another aspect of the invention, there is provided a microfluidic device for delivering a cargo into a cell, the device comprising a micro-post array or a filter, wherein the width of the micro-constrictions in the micro-post array or aperture size of the filter is equal to or greater than the diameter of the cell. The width of the micro-constrictions in the micro-post array or aperture size of the filter should be 0% to 20% larger than the average cell diameter. Advantageously, this provided excellent results for cargo delivery into the cell and prevents clogging.

[0037] In various embodiments, the micro-post array comprises a first set of a plurality of microposts, wherein a pair of adjacent microposts forming a constriction point and the width of each constriction point is equal to or greater than the diameter of the cell, reducing direct contact between the cell and the microposts.

[0038] In various embodiments, the width of each constriction point in the first set of microposts is equal to or greater than the diameter of the cell, preferably between 4pm to 28pm, preferably 16pm to 28pm, more preferably 20pm.

[0039] In various embodiments, the device further comprises a second set of a plurality of microposts prior to the first set of micropost, wherein the width of each constriction point in the second set of microposts is between 32pm to 24pm.This second set of microposts improves delivery results.

[0040] In various embodiments, the distance between the first and second sets of microposts is about 100pm. The distance between the first and second sets of microposts depends on the time for cells to recover to their original shape. As such, this value may range between 100 to 1000 pm.

[0041] In various embodiments, the device further comprises a plurality of sets of microposts arranged in a series, wherein the width of the constriction points in each array is the same, and wherein the width of the constriction point in each array in the series progressively decreases in size.

[0042] In various embodiments, each micropost has a cylinder diameter of 40pm and a height of 20-400pm.

[0043] In the example where the embodiment is a filter, the filter is a stainless steel filter and comprises a plain weave mesh or a Dutch weave mesh.

[0044] In various embodiments, the filter is a stainless steel filter with an aperture size 0% to 20% larger than the average cell diameter. In various embodiments, the aperture size ranges from 3 pm to 100 pm.

[0045] In various embodiments, the filter is integrated into the dispensing tip or a filter holder connected to a syringe.

[0046]

[0047] Taking advantage of the ability of viscoelastic fluids to provide strong elastic and shear force, here we introduce a microfluidic system based on viscoelastic fluids and a microconstriction array for intracellular delivery. Unlike micropost array platforms with small spacing, the width of the micro-constriction in this system is equal to or larger than the cell’s diameter, which reduces clogging issue. To compensate for the deformation loss from the wider constriction, viscoelastic fluids are introduced to provide strong deformation to cells. Differ from intracellular delivery through a single constriction, a carefully designed micropost array platform enables repetitive and progressive cell deformation, facilitating multiple exchanges of substances between the cell interior and exterior in a single delivery process. Therefore, a uniform and high efficiency intracellular delivery without the risk of low viability can be achieved. Optimized device parameters showed successful delivery of dextran with different sizes to MDA-MB-231 and Jurkat cell. The transfection of mRNA to Jurkat cell showed a high enhanced green fluorescent protein (eGFP)-expressing efficiency (>99%) and high viability (>90%), making it a promising for biological and medical applications.BRIEF DESCRIPTION OF THE DRAWINGS

[0048] The invention will be better understood with reference to the detailed description when considered in conjunction with the non-limiting examples and the accompanying drawings, in which:

[0049] Figure 1. Delivery mechanism, structure design of viscoelastic fluid induced intracellular delivery process. (A) Illustration of the viscoelastic fluid induced cell deformation and intracellular delivery when cells pass through the constriction array. (B) Schematic illustration of the mechanism whereby membrane disruption occurs when viscoelastic fluids squeeze cells at constriction area. (C) Optical image of a fabricated microfluidic device. (D) Magnification of the whole constriction array. (E) Scanning electron microscopy (SEM) of theconstriction array. (F) High speed images of a cell passing through a micro-constriction in the PEO solution. (G) Area and velocity of a cell passing through a micro-constriction.

[0050] Figure 2. Cell deformation depends on the PEO concentration and constriction design. (A) Image of MDA-MB-231 cell at a range of PEO concentrations when passing through constrictions. (B) Deformability vs. x position for ten cells at different PEO concentrations. (C) Maximum deformability vs. PEO concentrations for MDA-MB-231 cells. (D) Deformability by different constriction width. (E) Comparison of cell deformability in a constriction array with width decreasing from 32 pm to 20 pm to one constriction with a width of 20 pm. *** indicates a P-value of less than 0.001 , calculated using t-test.

[0051] Figure 3. Delivery performance depends on PEO concentration, constriction width and flow rate. (A) Bright field and fluorescence microscopy image after dextran delivery of MDA-MB-231 cells by endocytosis, constriction array in EDTA solution and constriction in 2% PEO. (B) Fluorescence intensity histograms of a control / endocytosis MDA-MB-231 cell population and one treated by 32 pm-20 pm device in 2% PEO solution. (C) Delivery efficiency and mean fluorescence intensity (MFI) fold change as a function of PEO concentration range from 0% to 2%. (D) Viability and recovery as a function of PEO concentration. (E) Delivery efficiency and MFI fold change as function of minimum constriction width from 8 to 20 pm. (F) Viability and recovery as functions of minimum constriction width. (G) Delivery efficiency and MFI fold change of with flow rate varies from 10 to 200 pL / min. (H) Viability and recovery as a function of flow rate. ** indicates a P-value less than 0.01 , and *** indicates a P-value of less than 0.001 , calculated using Dunnett's 1-way ANOVA method.

[0052] Figure 4. Delivery performance depends on dextran concentration and cargo size. (A) Fluorescence intensity histograms of MDA-MB-231 cells treated with 70 kDa FITC-dextran at different concentrations. (B) Delivery efficiency and MFI fold change of MDA-MB-231 cells as a function of dextran concentration. (C) Viability and recovery for different dextran concentrations. (D) Bright field image, fluorescence microscopy image and (E) fluorescence intensity histograms of MDA-MB-231 cells treated with 4 kDa, 70 kDa and 2,000 kDa FITC- dextran in 2% PEO solution. (F) Fluorescence intensity histograms of MDA-MB-231 cells treated with 4 kDa, 70 kDa and 2,000 kDa FITC-dextran in 0.5% PEO solution. (G) Fluorescence intensity histograms of Jurkat cells treated with 4 kDa, 70 kDa and 2,000 kDa FITC-dextran in 2% PEO solution.

[0053] Figure 5. EGFP-mRNA transfection performance characterization. (A) Bright field image, fluorescence microscopy image and (B) fluorescence intensity histograms of Jurkat cells treated with EGFP-mRNA by endocytosis and viscoelastic fluid induced intracellulardelivery. (C) MFI fold change and delivery efficiency for EGFP-mRNA delivery into Jurkat cells. (D) Viability and recovery for EGFP-mRNA delivery into Jurkat cells.

[0054] Figure 6. Cell culture after delivery process. Cells were mixed with FITC-dextran and flow through microconstriction array in EDTA solution and PEO solution. In control group, cells were mixed with FITC-dextran for the same time. After delivery, cells were washed and resuspended in DMEM supplemented with 10% FBS and 1% penicillin and streptomycin for 96 hours. Similar cell densities after cell culture of the 3 groups were observed.

[0055] Figure 7. Microchannels with constriction widths of (A) 20-32 pm, (B) 16-28 pm, (C) 12-24 pm, (D) 8-20 pm. (E) Enlarged image of sections with different constriction width range from 32 to 8 pm. Scale bar: 50 pm.

[0056] Figure 8. Bright field image and fluorescence microscopy image of MDA-MB-231 cells treated with 4 kDa, 70 kDa and 2,000 kDa FITC-dextran in 0.5% PEO solution.

[0057] Figure 9. Bright field image, fluorescence microscopy image of Jurkat cells treated with 4 kDa, 70 kDa and 2,000 kDa FITC-dextran in 2% PEO solution.

[0058] Figure 10. Bright field image of a Jurkat cell when passing through a microconstriction with a width of 16 pm.

[0059] Figure 11. Schematic representation for transfection with stainless-steel filter and viscoelastic fluid, (a) Schematic of the structure and mechanism of intracellular delivery device with filter, (b) Image of stainless-steel filter with an aperture size of 20 pm. (c) Microscope image of twilled Dutch weave mesh structure, (d) Sectional view and (e) Schematic illustration of twilled Dutch weave mesh structure.

[0060] Figure 12. Optimization of viscoelastic fluid and filter induced intracellular delivery, (a) Delivery efficiency, MFI fold, (b) viability and recovery of FITC-dextran into MDA-MB-231 cells with various flow rates. PEO concentration, filter rating, and filter number were set as 20 mg / mL, 20 pm and 1 , respectively, (c) Delivery efficiency, MFI fold, (d) viability and recovery of FITC-dextran with various PEO concentrations. Flow rate, filter rating, and filter number were set as 1 mL / min, 20 pm and 1. (e) Delivery efficiency, MFI fold, (f) viability and recovery of FITC-dextran with various filter ratings. Flow rate, PEO concentration, and filter number were set as 1 mL / min, 1 mg / mL and 1. (g) Delivery efficiency, MFI fold, (h) viability and recovery of FITC-dextran with different filter numbers. Flow rate, PEO concentration, and filter rating were set as 1 mL / min, 1 mg / mL and 16 pm. (i) Fluorescence images of FITC-dextran delivery using PBS and PEO solution, (j) Comparison of fluorescence intensity using PBS and PEO solution.

[0061] Figure 13. Delivery performance to different concentrations, cell types, and cargo sizes, (a) Fluorescence intensity distribution of FITC-dextran into MDA-MB-231 cells withvarious concentrations, (b) Delivery efficiency and MFI fold change of MDA-MB-231 cells as a function of FITC-dextran concentration, (c) Fluorescence intensity distribution of FITC- dextran into Jurkat cells with various concentrations, (d) Delivery efficiency and MFI fold change of Jurkat cells as a function of FITC-dextran concentration, (e) Fluorescence images of FITC-dextran delivery into MDA-MB-231 cells with a molecule weight of 2,000 kDa, 70 kDa, and 4 kDa. To get a clear comparison between different molecule weights, the acquisition time was reduced from 300 ms to 100 ms. (f) Fluorescence intensity distribution of MDA-MB-231 cells delivered using FITC-dextran with different molecule weight.

[0062] Figure 14. RNA delivery into MDA-MB-231 cells, (a) Bright field and fluorescence images of control MDA-MB-231 cells, cells after lipofection, cells after filter delivery, and cells after lipofection and filter delivery. The acquisition time of fluorescence image of cells after filter delivery is 1500 ms. The acquisition time of the other fluorescence images is 300 ms. (b) Flow cytometric analysis of cells using different delivery methods, (c) Efficiency, (d) MFI fold change, and (e) viability of transfection using different methods.DETAILED DESCRIPTION

[0063] The invention illustratively described herein may suitably be practiced in the absence of any element or elements, limitation or limitations, not specifically disclosed herein. Thus, for example, the terms "comprising", "including", "containing", etc. shall be read expansively and without limitation. Additionally, the terms and expressions employed herein have been used as terms of description and not of limitation, and there is no intention in the use of such terms and expressions of excluding any equivalents of the features shown and described or portions thereof, but it is recognized that various modifications are possible within the scope of the invention claimed. Thus, it should be understood that although the present invention has been specifically disclosed by preferred embodiments and optional features, modification and variation of the inventions embodied therein herein disclosed may be resorted to by those skilled in the art, and that such modifications and variations are considered to be within the scope of this invention.

[0064] The invention has been described broadly and generically herein. Each of the narrower species and subgeneric groupings falling within the generic disclosure also form part of the invention. This includes the generic description of the invention with a proviso or negative limitation removing any subject matter from the genus, regardless of whether or not the excised material is specifically recited herein.

[0065] Other embodiments are within the following claims and non- limiting examples. In addition, where features or aspects of the invention are described in terms of Markush groups, those skilled in the art will recognize that the invention is also thereby described in terms of any individual member or subgroup of members of the Markush group.

[0066] EXAMPLES

[0067] Non-limiting examples of the invention and comparative examples will be further described in greater detail by reference to specific Examples, which should not be construed as in any way limiting the scope of the invention.

[0068] EXAMPLE 1

[0069] In this non-limiting example, we show a microfluidic device with viscoelastic fluids to implement highly efficient intracellular delivery and maintain high cell viability after delivery.

[0070] This method utilizes a micro-constriction array that generates progressive shear forces on the cell membrane via the viscoelastic fluids, facilitating the delivery of exogenous biomolecules into cells. Additionally, the constriction width is greater than the cell diameter to avoid physical squeezing by the device structures, which allows to maintain high cell viability and eliminate channel clogging issue.

[0071] Without altering the constriction width, we have successfully delivered a wide range of molecule sizes to MDA-MB-231 and Jurkat cells. We have also demonstrated the intracellular delivery of mRNA to Jurkat cells with high efficiency (>99%), high viability (90%), and high throughput (500,000 cells / min). This method holds significant potential for applications in biological sciences and medicine.

[0072] Experimental Section

[0073] Chip design and fabrication

[0074] The microfluidic channel pattern was designed with AutoCAD (Autodesk Inc., USA). Each chip contains 4 cell deformation zones. And each zone consists of 5 arrays of constrictions. The width of the constriction varies from 8 pm to 32 pm. The depth of the microchannel is 20 pm. The microchip layout was patterned using traditional photolithography technique with negative photoresist SU-8 on a 4-inch silicon wafer. PDMS (Sylgard 184, Dow Corning, USA) microfluidic channels were then fabricated using standard soft lithography. The inlet and outlet holes were created by biopsy punches (1.5 mm, Integra Miltex). The PDMS microfluidic channels were treated by a plasma system and then bonded with slide glass. The bonded microfluidic devices were baked at 60 °C for 24 hours.

[0075] Preparation of PEQ solution

[0076] Dulbecco's phosphate-buffered saline (DPBS) buffer was heated to 60 °C. Then, PEO (600 kDa, Sigma, No. 182028) was added slowly to the DPBS solution at a certainconcentration. The mixture was heated (60 °C) and stirred for 6 hours. A filter (0.22 pm) was used to remove the residual sediment in the mixture. The PEO solution was stored at 4 °C. Before the delivery process, the PEO solution was placed at room temperature for 1 h.

[0077] Cell culture

[0078] MDA-MB-231 and Jurkat cells were purchased from American Type Culture Collection (ATCC at No. HTB-26 and CRL-2899). MDA-MB-231 cells were cultured in high- glucose Dulbecco’s Modification of Eagle’s Medium (DMEM) (Thermo Fisher Scientific, USA) supplemented with 10% fetal bovine serum (FBS, Thermo Fisher Scientific, USA) to provide growth factors and 1% penicillin and streptomycin (Gibco, Thermo Fisher Scientific, USA) to prevent the growth of bacteria. Jurkat cells were cultured in RPMI-1640 medium (Corning, USA) supplemented with 10% FBS and 1% penicillin-streptomycin. Both the cell lines were maintained at 37°C, 5% (v / v) CO2 in a humidified incubator.

[0079] Intracellular delivery process

[0080] Cells were filtered using a 30 pm filter (Miltenyi Biotec, No. 130-041-407) and washed twice by DPBS solution. FITC-dextran or EGFP mRNA were then added to the PEO solution at an appropriate concentration. The cells were resuspended in PEO solutions and loaded into a 1mL syringe. A syringe pump (Legato 111 , KD Scientific, USA) was used to inject the cell suspension into the PDMS microchannel. Cells were collected from the outlet and maintained at room temperature for 30 minutes to ensure the recovery of cell membrane disruption. For mRNA delivery group, cells were washed three times and suspended in DMEM medium, followed by incubation for 24 hours at 37°C, 5% (v / v) CO2 in a humidified incubator. The 0.4% trypan blue solution (Lonza, Switzerland) was used for cell viability analysis. During the delivery process, the images were recorded by a high-speed camera (Photron Inc., San Diego, CA). The frame rate was 20000 fps. The exposure time was set to 5 ps. The deformability of cells at each frame was analyzed and calculated by DMV software.

[0081] Flow field characterization

[0082] Firstly, to investigate the flow condition in microfluidic chip, the Reynolds number was calculated as Re = pvDh / r], where p is the density of fluid, v is the mean velocity, q is the viscosity of fluid, 1 Dh is the hydraulic diameter and can be calculated as Dh=2HW / (H + W). The H and W represent the height of the microchannel and constriction width, respectively. The parameter we use for Re calculation is shown in Table 1.

[0083] Table 1. Parameter for Reynolds number calculationName Value p 1 ,000 kg / m3

[0084] Deborah number calculation

[0085] The Deborah number (De) is defined as A / td, where / is the relaxation time, and td is the time of deformation. The relaxation time can be calculated by / \=18 / z(c / c')065, where Azis relaxation time in the Zimm theory, c is the PEO concentration, c is the minimum concentration which the polymer coils can overlap with each other.Azcan be calculated as4 = 0A63[q]Mwqs / NAkBT (1)

[0086] where [ry] is the intrinsic viscosity [q]=0.072MwOS5,2Mwis the molecular weight of PEO, qsis the solvent viscosity, NA is the Avogadro’s number, kB is the Boltzmann’s constant, T is the Kelvin temperature, c can be calculated as c*=0.77 / [ / 7].3The calculated results are summarized in Table 2.

[0087] Table 2. Parameter for Deborah number calculation

[0088] Flow cytometry

[0089] Before flow cytometry characterization, cells were washed twice by DPBS solution and suspended in DPBS solution. Next, the cells were examined by running through flow cytometry (MACSQuant Analyzer). The data was analyzed by software FlowJo 10.8.1 .

[0090] Results and Discussion

[0091] Working principle and system overview

[0092] The principle of cargo delivery was based on the transient membrane disruptions or holes formed when cells undergo rapid deformation. To achieve this, the microfluidic device was designed with a series of constrictions featuring progressively smaller widths. During the delivery process, cells and cargos were suspended in a viscoelastic fluid and pass through these constrictions (Figure 1A). Although the constriction width is larger than the cell diameter, the viscoelastic fluid can generate strong shear and compressive forces on the cells. These forces induce rapid cell deformation, creating transient disruptions or holes in the cell membrane. Besides, the fast compression on cells can cause a portion of the cytoplasm to be squeezed out of the cell. After compressing, cells absorb external fluids during the recovery process. Cargos, such as fluorescein isothiocyanate (FITC) labeled dextran and enhanced green fluorescent protein (EGFP) mRNA, are delivered into the cells with the external fluids (Figure 1 B).

[0093] The viscoelastic microfluidic devices were fabricated using traditional PDMS molding techniques, where PDMS with microstructures was bonded onto glass to form microchannels with multiple constrictions (Figure 1C). The microfluidic chip comprised four deformation zones, each containing four arrays of microposts with a cylinder diameter of 40 pm and a height of 20 pm (Figure 1 D, E). The distance between each array is 100 pm, and the constriction width decreases from 20 pm to 8 pm. There were 30 parallel constrictions in a cross section. Compared to a single microfluidic channel, the reticular microfluidic channel design mitigates the risk of potential channel clogging and enhances the throughput of the delivery process.

[0094] PEO solution was selected as the viscoelastic fluid due to its high biocompatibility, high lubricity and high solubility in water.44-46 To evaluate the deformation of cells in viscoelastic fluids, human breast cancer MDA-MB-231 cells were suspended in a 2% PEO solution for further experimentation.

[0095] Figure 1 F shows an MDA-MB-231 cell passing through a micro-constriction (20 pm wide) at a flow rate of 50 pL / min. The average flow rate and flow velocity at the constriction can be calculated to be 1.67 pL / min and 70 mm / s. It is noteworthy that, although the average diameter of MDA-MB-231 cells (16.51 pm) is smaller than the constriction width, the cells can deform without any physical contact with microposts, thereby reducing friction between the cell membrane and microposts. To further analyze the forces acting on cells, detailed quantification about the flow in the microchannel was conducted. It was calculated that the fluid motion in the microchannel was a laminar flow.

[0096] Therefore, the possibility of turbulence in the flow channel can be ruled out. The Deborah number (De) is used to characterize the viscoelasticity of the fluid, which is defined as A / td, where A is the relaxation time, and td is the time of deformation (0.6 ms, from the inlet of constriction to the center of constriction in Figure 1F).

[0097] The relaxation time can be calculated by A=18Az(c / c*)0'65, where Azis relaxation time in the Zimm theory, c is the PEO concentration, c* is the minimum concentration at which the polymer coils can overlap with each other.47 For 2% PEO solution, the A is calculated to be 3.8 ms.

[0098] Since the Deborah number around constriction was larger than 1 , the fluid showed elastic properties and demonstrating solid like behavior. As cells flowed to the center of the constriction, the shrinking of the cross section resulted in the compression on cells. The compression could also be validated by the decrease of cell area before the cell passed through the constriction (Figure 1G).

[0099] Similar to viscoelastic fluid, cells under compression exhibits both viscosity and elasticity. The relation of viscous and elastic forces on cells can be characterized by Ericksen number, defined as

[0100] where |jcis the viscosity of cell (0.72 Pa s), vc=Ad / td is the compression velocity, de is the relax diameter of cell (13.6 pm in Figure 1 F), Ec is the Young's Modulus of cell (182 Pa), A is the area of cell, Ad is the compression size (2.5 pm, decreased from 13.6 pm to 11.1 pm in Figure 1 F).

[0101] A high Er value suggests that cell’s elastic deformation is unable to catch up with the deformation of cells, indicating a portion of the cytoplasm were squeezed out of the cell. The Er value of the cell in Figure 1 F was calculated to be 0.39, which proves small cytoplasm losses of the cell. After the cell passed through the constriction, the elastic energy storage in the cell and viscoelastic fluid were released, resulting in a high flow rate and recovery of the cell area, as shown in Figure 1G.

[0102] Characterization of deformation

[0103] The role of PEO concentration and constriction width in the deformation process was further investigated while maintaining a constant flow rate of 50 pL / min. We evaluated cell shape changes around the constriction by analyzing high-speed video footage of individual cells passing through the micro-constriction.

[0104] Figure 2A shows cells within a 20 pm-wide constriction in PEO solutions at concentrations of 0%, 0.5%, 1%, 1.5%, and 2%. Each cell was approximated as an ellipsoid for the quantitative characterization of deformation. Deformability is defined as the ratio a / b, where a and b are the lengths of the major and minor axes of the cells, respectively. During the delivery process, the deformability increased with the PEO concentration, confirming that viscoelastic fluids enhance the cell deformation. We analyzed the deformability variation of 10 cells in PEO solutions of different concentrations (Figure 2B).

[0105] The deformability gradually increased, reaching a maximum at the constriction point, and then gradually returned to unity. When MDA-MB-231 cells were dispersed in DPBS, slight deformability changes occurred before the cells reached the constriction point, likely due to direct contact between cells and microposts at high flow rates. The maximum cell deformability was approximately 1.46 (Figure 2C). Compared with cells in DPBS solution, the PEO solution significantly increased deformability. In the 2% PEO solution, the average deformability at the constriction point reached 2.07. The increased deformability is attributed to the elasticity change of PEO solution. As the relaxation time is positively correlated with PEO concentration, a high concentration of PEO solution can generate strong compression on cells, leading to a large deformation of cells.

[0106] Additionally, 500 MDA-MB-231 cells in 2 wt % PEO solution were tested with constriction widths ranging from 20 pm to 8 pm (Figure 2D). Since 12 pm constriction is smaller than most of the MDA-MB-231 cells, direct physical contact between the cells and micropost is inevitable.

[0107] However, MDA-MB-231 cells exhibited similar deformability in PEO solutions across constrictions ranging from 20 pm to 12 pm. This suggests that a 20 pm-wide constriction provides sufficient deformation for cargo delivery.

[0108] Repeated cell squeezing can induce multiple deformation and recover process in the intracellular delivery indicating more cargo delivered into cells. Moreover, the deformation of cells can also be further improved by repeated cell squeezing. However, simply increasing the number of squeezing events may reduce cell viability.

[0109] To address this issue, a series of constrictions with widths ranging from 32 pm to 24 pm were added before the 20 pm-wide constrictions to achieve progressive squeezing. Although the wider constriction can only provide smaller compression on cells, the shear force can also generate pores on cell membrane. Besides, the pre-deformation can influence the elasticity of cells, which reduces mechanical shock on cells when cells pass through narrow constrictions and enhance cell viability.

[0110] Figure 2E shows the deformability of 50 cells at each constriction point. The deformability slightly increased as the constriction width decreased from 32 pm to 20 pm. The average deformability at the final constriction was 2.28, higher than the deformability of cells at a single 20 pm-wide constriction. The higher deformation of cells can generate more and larger nanopores on cell membranes, resulting in better intracellular delivery performance than single constriction. Additionally, the similar viability of MDA-MB-231 cells before and after squeezing (92% and 90%) indicates that the squeezing process has minimal impact on cell viability.

[0111] Parameter optimization

[0112] To validate the efficacy of viscoelastic fluid in the micro-constriction array for improving the delivery process, we compared the delivery of FITC-dextran to MDA-MB-231 cells via endocytosis, squeezing in RoboSep™ Buffer (STEMCELL Technologies, Canada), and squeezing in a 2% PEO solution. In this process, the FITC-dextran concentration, cell density, constriction width, and flow rate were set to 100 pg / mL, 1 x107 / mL, 32 to 20 pm and 50 pL / min, respectively.

[0113] A high throughput of 500,000 cells / min was calculated based on the cell density and flow rate. Figure 3A shows that only cells treated with the PEO solution exhibited a significant fluorescence effect. For quantitative characterization of the delivery process, we define the threshold as the top 5% of fluorescence signals from cells in the control group (endocytosis). Delivery efficiency is then defined as the fraction of cells with fluorescence intensity exceeding this threshold (Figure 3B).

[0114] A delivery efficiency of 98.7% was observed for cells squeezed in the 2% PEO solution. Moreover, cells were cultured for 96 hours post-delivery, with no significant difference in cell density between the control group and cells squeezed in PEO solution, indicating minimal impact on cell viability (see Figure 6).

[0115] To further optimize the delivery performance, we considered PEO concentration, constriction dimensions, and flow rate as key parameters. The relationship between delivery efficiency and PEO concentration is shown in Figure 3C with constriction width and flow rate set as 20 to 32 pm and 50 pL / min, respectively. A positive correlation was observed between delivery efficiency and PEO concentration, with delivery efficiency approaching 100% at a PEO concentration of 1.5%. However, fluorescence signal intensity differed between cells treated with 1.5% and 2% PEO solutions.

[0116] To assess delivery performance, we calculated the mean fluorescence intensity (MFI) fold change, defined as the mean fluorescence intensity ratio between experimentalgroup and control group. Cells treated in the 2% PEO solution exhibited higher fluorescence intensity compared to those in the 1.5% PEO solution.

[0117] While trypan blue viability tests measure the fraction of live cells, some cells may remain in the microchip or be damaged during the squeezing process, which is not reflected in viability measurements. Therefore, we used recovery as a parameter to estimate the fraction of cells collected after the delivery process, defined as the ratio of cells collected to cells pumped into the microchip.

[0118] Cell number was estimated as the product of cell density and the volume of cell dispersion. The viability and recovery of MDA-MB-231 cells treated with PEO concentrations ranging from 0% to 2% are shown in Figure 3D. The average viability across all PEO concentrations remained relatively constant (89-93%). However, lower cell recovery was observed as PEO concentration decreased, with many cells accumulating at the inlet. Consequently, we selected a PEO concentration of 2% for high MFI fold change and recovery.

[0119] Additionally, the influence of constriction width on delivery performance was investigated using a 50 pUmin flow rate and 2% PEO solution. Microchannels with constriction widths of 8-20 pm, 12-24 pm, 16-28 pm, and 20-32 pm were fabricated for this study (see Figure 7). Narrowing the constriction slightly enhanced fluorescence intensity, and delivery efficiency remained high (above 97%) under all tested conditions (Figure 3E).

[0120] However, reducing constriction width also led to decreased viability. When the minimum constriction width was approximately half the average cell diameter (8 pm), viability dropped to 31% (Figure 3F). Similar to viability, recovery also decreased dramatically when the constriction width was less than 16 pm, with a large number of cells becoming blocked in the constriction. Only 16% of cells were collected from the microfluidic chip outlet when the minimum constriction width was 8 pm. Considering MFI fold change, viability, and recovery, we selected constriction widths of 16-28 pm for further investigation.

[0121] The effect of flow rate on delivery performance was then studied over a range of 10 pL / min to 200 pL / min. The relationship between delivery efficiency, MFI fold change, and flow rate is shown in Figure 3G. No significant difference in delivery efficiency was observed across different flow rates. However, MFI fold changes positively correlated with flow rate. The average viability exceeded 90% at flow rates below 100 pL / min (Figure 3H).

[0122] Recovery decreased as the flow rate increased, with an average recovery of 49% observed at a flow rate of 200 pL / min. The relatively low viability and recovery at higher flow rates can be attributed to the cell death due to the strong deformation at the constrictions. Therefore, the flow rate was set to 50 pL / min for subsequent experiments.

[0123] The key factors influencing the deformation of cells are the elasticity of fluid and translocation time when they pass through the constriction. For instance, the decrease in PEO concentration leads to smaller relaxation time, which can be compensated by increasing the flow rate to reduce the deformation time. On the other hand, a constriction with a width much larger than the diameter of cells can still generate strong stress on cells by using a polymer with high molecule weight and high flow rate to enhance the elasticity of fluid.

[0124] Influence of cargo concentration, size and cell type

[0125] Precise control of the delivery process is crucial, as excessively high or low cargo concentrations within cells may lead to undesirable outcomes, including reduced delivery efficiency and viability. In our proposed method, the quantity of cargo delivered into cells can be regulated by adjusting the cargo concentration in the PEO solutions.

[0126] An increase in fluorescence intensity was observed with dextran concentrations ranging from 10 to 1 ,000 pg / mL (Figure 4A). A concentration above 300 pg / mL ensured a delivery efficiency greater than 99% (Figure 4B). The MFI fold change exhibited a positive correlation with dextran concentration, indicating effective control over dextran levels within cells. Viability and recovery remained consistent across different concentrations, likely due to the uniform deformation process and the high biocompatibility of dextran (Figure 4C).

[0127] It is important to note that the impact of dextran size on delivery efficiency varies depending on the viscosity of the fluid. To investigate the intracellular delivery mechanism in viscoelastic fluids, we examined the delivery efficiency of cargos with different dextran sizes (3-5, 70, and 2,000 kDa) to MDA-MB-231 cells. The constriction width and flow rate were set as 16-28 pm and 50 pUmin, respectively.

[0128] Similar MFI fold changes were observed for different dextran sizes when cells were suspended in a 2% PEO solution (Figure 4D, E).

[0129] However, when suspended in a 0.5% PEO solution, lower fluorescence intensity was observed for 2,000 kDa FITC-dextran (Figure 4F and Figure 8).

[0130] A possible explanation is that delivery efficiency is influenced by a combination of pore size on the cell membrane and volume exchange during the delivery process. The deformability of cells in the 2% PEO solution is sufficient to create nanopores larger than the 2,000 kDa FITC-dextran (53.78 nm).

[0131] Therefore, the delivery dosage of dextran is controlled by substances exchange. At the same experimental conditions, the deformation of cells is the same, resulting in similar delivery performance for dextran with different molecular weights. However, in the low- viscosity fluid (0.5% PEO), the relaxation time of the fluid decreased from 3.8 ms to 1.54 ms, resulting in smaller deformation of cells. Therefore, the size of nanopores on the cellmembranes are too small for 70 and 2,000 kDa FITC-dextran to pass, leading to a lower dosage of large dextran delivered to cells compared to smaller dextran.

[0132] Even though directly compressing the cells with narrow constrictions can achieve a high delivery performance, the sensitivity to cell diameter limits the application. For instance, a constriction width of 4 pm can ensure a high delivery efficiency for primary human T-cells (6.7 pm). But for a cell line with large diameter (BxPc3 cells, 10.8 pm), the viability is less than 30% using the same constriction width.37 Therefore, the constriction width needs to be optimized for cell lines before intracellular delivery.

[0133] Compared to directly compressing cells, a significant advantage of the viscoelastic fluid-induced squeezing method is its low sensitivity to cell diameter. For viscoelastic fluid induced compress, even in the absence of cells passing through the channel, the fluid will still undergo elastic deformation, indicating low requirement to adjust the experimental conditions. To demonstrate the size insensitivity of our method, Jurkat cells with a small diameter (9.87±1.64 pm) were chosen. FITC-dextran at the same concentration (1 mg / mL) was delivered to Jurkat cells under the same conditions: flow rate (50 pL / min), constriction width (16-20 pm), and solution (2% PEO). The results showed that fluorescence intensity decreased as the FITC-dextran size increased (Figure 4G).

[0134] However, since Jurkat cells in viscoelastic fluid still exhibited high deformability at the constriction, the delivery efficiency with 3-5, 70, and 2,000 kDa dextran remained above 99%, demonstrating the method's applicability to cells of varying diameters.

[0135] Applicability of mRNA transfection

[0136] The mRNA-based intracellular delivery has been proved to be efficient in CAR-T and coronavirus disease 2019 vaccines production. To evaluate the capability of our method for mRNA delivery, we investigated its performance in intracellular mRNA delivery for protein expression. Jurkat cells were transfected with EGFP-mRNA containing 996 nucleotides.

[0137] The parameters were set as follows: mRNA concentration of 10 pg / mL, PEO concentration of 2%, Jurkat cell density of 1x107ml_, flow rate of 50 pL / min, and constriction width of 16-28 pm. After the delivery process, the cells were incubated for 24 hours. As a negatively charged biological macromolecule, mRNA typically struggles to penetrate the negatively charged cell membrane.

[0138] Consequently, no green fluorescence of EGFP was observed in the control group (Figure 5A). Left is the control group with no PEO and no fluorescence. Right is the group using PEO solution, which showed strong fluorescence.

[0139] However, when Jurkat cells were treated using the viscoelastic fluid-induced delivery method, a strong fluorescence signal indicated successful mRNA delivery (Figure5B). The fluorescence intensity histogram revealed a significant change in Jurkat cells, achieving a high transfection efficiency of 99.9% and a substantial MFI fold change of 172 (Figure 5C). Additionally, the viability of Jurkat cells was approximately 95% (Figure 5D).

[0140] These results demonstrate the excellent transfection performance of the viscoelastic fluid-induced transfection method. The only limitation observed was a relatively low cell recovery rate of 71.5%, attributed to cell loss during the washing process, which requires further optimization.

[0141] Conclusions

[0142] Here, we present a novel microfluidic method capable of achieving cell deformation without physical contact, enabling high-efficiency and high-throughput cargo delivery to cells. Our results demonstrate that cells suspended in a viscoelastic fluid (PEO solution) experience significant stress when passing through micro-constrictions, even when the constriction width exceeds the cell diameter. The deformability of cells can be finely tuned by adjusting the PEO concentration.

[0143] Moreover, repetitive and progressively increasing viscoelastic squeezing enhances cell deformability without compromising cell viability. We achieved high-performance intracellular delivery of FITC-dextran to both MDA-MB-231 and Jurkat cell lines using the viscoelastic fluid-based squeezing method. Furthermore, EGFP-mRNA was successfully transfected into Jurkat cells with high transfection efficiency, high viability, and high throughput.

[0144] Compared with other microfluidic methods based on cell deformation, the viscoelastic fluid-based delivery method offers several advantages: 1. The use of a viscoelastic fluid allows for wider constrictions while maintaining strong cell deformability. This wider constriction significantly reduces the risk of direct contact between cells and microposts, thereby enhancing cell viability and preventing clogging during the delivery process. 2. Repeated squeezing of cells increases the uniformity of delivery efficiency. 3. The high delivery efficiency of FITC-dextran to MDA-MB-231 and Jurkat cells under the same delivery parameters demonstrates the method's low sensitivity to cell size.

[0145] In conclusion, our viscoelastic fluid-based repeated squeezing method enables uniform and efficient intracellular delivery, paving the way for future advancements in intracellular delivery technologies for biomedical applications.

[0146] EXAMPLE 2

[0147] In this non-limiting example, we present a novel intracellular delivery method utilizing a stainless-steel filter combined with polyethylene oxide (PEO) solution to transport biomolecules into cells. Optimal parameters were identified through systematic adjustmentsusing fluorescein isothiocyanate-dextran. The performance was further evaluated by mRNA encoding enhanced green fluorescent protein (eGFP), demonstrating a remarkable delivery efficiency of 94.7%, a cell viability rate of 94.3% and a throughput of 1x107 / mL. This new approach not only provides a cost-effective and user-friendly alternative but also enhances the potential for high-throughput applications in biological research and therapeutic interventions.

[0148] The findings suggest that the integration of stainless-steel filters and viscoelastic fluids enhances intracellular delivery capabilities, positioning this method as a promising tool for advancing cellular engineering technologies.

[0149] Advantageously, the stainless-steel filter provides controlled cross-sectional constrictions, with an aperture size comparable to or larger than the average cell diameter to minimize clogging. Since larger pores do not compress cells, a viscoelastic fluid is introduced to exert elastic and shear forces on the cell membrane as cells traverse the constriction. The efficacy of this method was evaluated using fluorescein isothiocyanate (FITC) labelled dextran and mRNA. After optimization, our approach demonstrated comparable transfection performance to lipofection. Additionally, combining this technique with lipofection achieved high delivery efficiency (94.7%) and cell viability (94.3%). The data here suggests that this easy-to-use, cost-effective and high-throughput approach with good delivery performance is an attractive intracellular delivery technology for engineering cell functions.

[0150] Materials and Methods

[0151] Preparation of PEO solution

[0152] A PEO solution (20 mg / mL) was prepared by dissolving 8 MDa PEO (Sigma, No. 372838) in PBS buffer. The mixture was heated to 60°C and stirred overnight. The resulting solution was then filtered using a bottle-top vacuum filter system (Sigma, CLS431097) and stored at 4°C. Before the delivery process, the PEO solution was diluted to the desired concentration using PBS and placed at room temperature for 1 hour.

[0153] Cell culture

[0154] MDA-MB-231 (HTB-26) and Jurkat (CRL-2899) cell lines were obtained from the American Type Culture Collection (ATCC) and cultured in Dulbecco's Modified Eagle Medium (DMEM) and Roswell Park Memorial Institute (RPMI) growth media (Thermo Fisher Scientific, USA), respectively. Both media were supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin (Gibco, Thermo Fisher Scientific, USA). The cells were maintained at 37°C in a humidified incubator with 5% CO2.

[0155] Intracellular delivery process

[0156] Before delivery, stainless-steel filters custom manufactured by Suzhou Linzhe Hardware Products Co., Ltd (China) and stainless-steel tips (Nordson, No. 7018107) were autoclaved at 120°C for 20 minutes following a standard sterilization cycle. The filters were then assembled into the stainless-steel tips. Cells were collected, washed, and filtered through a 30 pm filter (Miltenyi Biotec, No. 130-041-407). FITC-labelled dextran was prepared at the desired concentration in PEO solution. The cells were resuspended in PEO solution containing FITC-dextran. For each experimental condition, 1 mL of the mixture was loaded into a 5 mL syringe. A syringe pump (Legato 111 , KD Scientific, USA) was used to press the mixture of cells and cargo through the filters at a designed flow rate. The collected cells were incubated at room temperature for 10 minutes, then washed three times with PBS before characterization. Fluorescence intensity was analyzed using flow cytometry (MACSQuant Analyzer) and fluorescence microscopy. Cell viability and density were assessed using 0.4% trypan blue solution (Lonza, Switzerland) and a cell counter (Countess II FL, Thermo Fisher Scientific, USA). Flow cytometry data were analyzed using FlowJo 10.8.1.

[0157] Transfection of mRNA

[0158] For mRNA transfection, MDA-MB-231 cells were collected and seeded in 24-well culture plates 24 hours before transfection. In the lipofection group, 1.5 pL Lipofectamine 3000 (Thermo Fisher Scientific, USA) and 100 ng StemMACS™ Nuclear eGFP mRNA (Miltenyi Biotec, Germany) were each diluted in 25 pL DM EM. The diluted RNA was mixed with the diluted Lipofectamine 3000 and incubated at room temperature for 15 minutes before being added to cells in the culture plates.

[0159] For the filter-induced delivery group, cells were collected from culture plates and resuspended in 100 pL DMEM supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. The cells were mixed with 100 pL PEO solution (1 mg / mL) and 100 ng RNA before being passed through the filter.

[0160] For the combined filter and Lipofectamine 3000 group, liposomes were prepared as described for the lipofection group. The liposome solution (diluted in 50 pL DMEM) was mixed with 50 pL of the cell suspension and 100 pL of 1 mg / mL PEO solution, followed by flowing through a filter.

[0161] All groups were cultured in 24-well plates for 24 hours and analyzed using fluorescence microscopy and flow cytometry.

[0162] Results and discussion

[0163] Device design and mechanisms

[0164] The design principle of the filter and viscoelastic fluid induced intracellular delivery system is illustrated in Figure 11a. The system comprises a syringe, a filter, and a dispensingtip. The filter, with a diameter of 4 mm (Figure 11 b), is compact enough to be directly integrated into the dispensing tip via the syringe. Owing to the high pressure required to drive viscoelastic fluid through the filter, metal woven meshes were selected for their structural stability, precise aperture size, and minimal deformation under mechanical stress.

[0165] Among available metal mesh configurations, the Dutch weave mesh was prioritized due to its ability to achieve pore sizes comparable to cellular diameters. The detailed architecture of the Dutch weave mesh is depicted in Figure 11c and 11 d. This mesh comprises two distinct wire components: warp and weft.

[0166] The warp wires, characterized by a larger diameter, are aligned in parallel at predefined intervals, while the thinner weft wires interlace the warps, forming a tightly knit structure where adjacent weft wires maintain direct contact.

[0167] The filter contains three distinct cross-sectional reduction zones (Figure 11c and 11 e) : Area 1 , defined by two wefts and two warps; Area 2, formed by three wefts; and Area 3 formed by two wefts and one warps. As cells or particles traverse the filter, they must pass through Areas 1 and 2 twice and Area 3 once. Consequently, the filter’s aperture size is governed by the maximum diameter of spherical particles capable of traversing all three zones. During intracellular delivery, cells suspended in PEO solution are mixed with biomolecules and flow through the filter. As a viscoelastic fluid, PEO solution exhibits both elasticity and viscosity when undergoing deformation.

[0168] As the PEO-cell suspension enters the filter, the abrupt reduction in cross-sectional area induces a high flow rate and fluid deformation. This elevated flow rate generates substantial shear forces due to the relative velocity between cells and the filter structure, while rapid fluid deformation applies compressive stress to the cells. The combined effects of shear and compression deform the cells, creating transient pores in their membranes. These temporary pores enable passive diffusion of external cargo into the cells. Following filter traversal, the mechanical stress dissipates, allowing cellular membranes to reseal and return to their original state.

[0169] Parameter adjustment for optimization of delivery efficiency and cell viability

[0170] In the intracellular delivery process, the elasticity and viscosity of the PEO solution critically govern cell deformation. The concentration of PEO is the primary factor influencing the fluid’s viscosity and elasticity.

[0171] Additionally, elasticity of viscoelastic fluids is strongly dependent on the deformation time, quantified by the Deborah number De=A / t, where A is the relaxation time of the viscoelastic fluid solution, t denotes the deformation duration.

[0172] A high Deborah number corresponds to dominantly elastic fluid behavior under transient deformation.

[0173] Consequently, flow rate directly modulates the mechanical stress imposed on cell membranes. In addition to parameters that directly affect fluid properties, filter aperture size and the number of filters can also control the cell deformation. Optimizing these variables is essential to balance delivery efficiency and cell viability. To assess intracellular delivery performance, fluorescein isothiocyanate-dextran (FITC-dextran, 70 kDa) was used as a fluorescent tracer. MDA-MB-231 cells were suspended in PEO solution with 0.1 mg / mL FITC- dextran. Initial parameters included a cell density of 5 x 106 / ml_, PEO concentration of 20 mg / mL, flow rate of 15 mL / min, and a single filter with a 20 pm aperture.

[0174] Intracellular delivery efficiency was defined as the percentage of cells exhibiting fluorescence intensity exceeding a threshold set at the top 5% of the control group (cells suspended in PBS with FITC-dextran for identical durations).

[0175] As shown in Figure 12a, intracellular delivery efficiency approached nearly 100% at flow rates exceeding 2 mUmin. Besides the delivery efficiency, the mean fluorescence intensity (MFI) fold change, defined as the ratio of MFI of experimental group to the MFI of control group, was used to evaluate the intracellular dosage of FITC-dextran. The MFI fold change increased with flow rate, reaching values 80 fold higher than endocytosis-mediated uptake at 15 mL / min (Figure 12a).

[0176] However, cell viability decreased significantly at higher flow rates (Figure 12b), suggesting that transient nanopores on the membrane could transition into permanent damage at excessive flow rates. Moreover, cell recovery (the ratio of cell number collected after delivery to the cell number before delivery) was markedly reduced at flow rates above 5 mL / min, as many cells were fragmented during delivery.

[0177] Many cells are squeezed into fragments during delivery process. Given that delivery efficiency can be enhanced by adjusting other parameters, a flow rate of 2 mUmin was selected to ensure high cell viability and recovery. Under such a high flow rate, the throughput can be calculated to be 1 x 107cells / mL.

[0178] Subsequently, the influence of PEO concentration on delivery efficiency and cell viability was evaluated. As shown in Figure 12c, delivery efficiency remained consistently high (>95%) at PEO concentrations >1 mg / mL. However, at a lower concentration (0.5 mg / mL), the MFI fold change decreased to 10, reflecting diminished intracellular uptake of FITC-dextran. Cell viability exceeded 90% across the tested PEO concentration range (0.5-20 mg / mL), indicating minimal permanent membrane damage during delivery. Interestingly, cell recovery was lower than 80% at PEO concentrations above 2 mg / mL (Figure 12d).

[0179] Since high cell viability suggests that the delivery process does not induce significant permanent damage, the reduced recovery rate at higher PEO concentrations was attributed to cell loss during the post-delivery washing process. To mitigate cell loss while preserving delivery efficiency, a PEO concentration of 0.5 mg / mL was maintained in subsequent optimization steps.

[0180] In addition, the effect of filter aperture size on delivery performance was further investigated. A filter with an aperture size of 25 pm exhibited an efficiency ranging from 82% to 92%. Reducing the aperture size to 12 pm increased the average efficiency to 93%, with the mean fluorescence intensity (MFI) fold change rising from 7 to 16.8 (Figure 12e). No significant changes in delivery efficiency or cell viability were observed with varying aperture sizes (Figure 12f). But the average recovery decreased as the aperture size decreased from 16 pm to 12 pm, which can be attributed to the clogging occurred in the filter.

[0181] In contrast to aperture adjustments, augmenting the number of filters within stainless-steel dispensing tips failed to enhance delivery efficacy (Figure 12g).

[0182] In our system, the Dutch weave mesh structure inherently subjects cells to five deformation cycles per filter traversal, as cells pass through pores formed by intersecting stainless-steel wires. Consequently, a single filter suffices to induce multiple deformations. Beyond the minimal improvement in delivery efficiency, increasing the number of filters led to higher cell death rates and greater cell loss between filters (Figure 12h), failing to improve delivery performance and potentially degrading it. After optimization, the flow rate, PEO concentration, aperture size, and number of filters were set at 2 mL / min, 0.5 mg / mL, 16 pm, and 1 , respectively, for subsequent experiments.

[0183] A key challenge in using stainless-steel filters for intracellular delivery is manufacturing precision. Unlike silicon wafer-based micropores, which are fabricated via photolithography and wet etching to achieve uniform geometries, stainless-steel filters exhibit inherent variability in aperture size due to limitations in metal weaving processes. Furthermore, the cross-sectional geometry of stainless-steel micropores is triangular rather than circular, creating suboptimal contact between cells and the filter during transit. This geometric mismatch reduces the magnitude of cellular deformation compared to circular pore architectures.

[0184] However, the viscoelastic fluid medium mitigates this deformation deficit by augmenting stress transfer to cells. Figure 12i displays representative bright field / fluorescence overlay images comparing delivery efficacy with and without viscoelastic fluid. Under identical flow conditions, MDA-MB-231 cells processed in the absence of viscoelastic fluid exhibited negligible fluorescence signal. In contrast, cells treated with a 0.5 mg / mL (PEO) solutiondemonstrated a tenfold increase in fluorescence intensity (Figure 12j), confirming the critical role of viscoelasticity in enhancing membrane permeability.

[0185] Characterization of delivery performance based on cell and cargo properties.

[0186] As the dosage of delivered cargo into cells is crucial for achieving desirable biological and medical outcomes, we examined the dependency of delivery efficiency on cargo concentration, ranging from 10 pg / mL to 1 mg / mL. A positive correlation between cargo concentration and fluorescence intensity was observed (Figure 13a). The delivery efficiency increased by rising cargo concentration up to 100 pg / mL. Beyond this threshold, further increases in intracellular dosage could only be detected by measuring the MFI fold change, as the delivery efficiency approached nearly 100% (Figure 13b). Notably, at a cargo concentration of 10 pg / mL, the delivery efficiency remained above 70%.

[0187] To assess the method’s applicability across cell types, delivery performance was compared between MDA-MB-231 and Jurkat cells. A similar trend in fluorescence intensity was observed in both cell types. However, since Jurkat cells have a smaller average diameter than MDA-MB-231 cells, they experienced less mechanical stress when passing through the filters. As a result, the fluorescence intensity of FITC-dextran in Jurkat cells was significantly lower than in MDA-MB-231 cells (Figure 13c). At a FITC-dextran concentration of 10 pg / mL, the delivery efficiency in Jurkat cells was approximately 11 % (Figure 13d). Additionally, the MFI fold change in Jurkat cells was only one-third of that observed in MDA-MB-231 cells at the same concentration, indicating the need for further optimization in future studies.

[0188] Beyond cell type differences, we also investigated delivery performance based on cargo size. FITC-dextran molecules of varying molecular weights were delivered into MDA- MB-231 cells. As molecular weight decreased from 2,000 kDa to 4 kDa, fluorescence intensity increased, suggesting that intracellular dosage is size-dependent (Figure 13e). Specifically, the fluorescence intensity of 2,000 kDa FITC-dextran in cells was 89.4% lower than that of 4 kDa FITC-dextran (Figure 13f), indicating size-dependent cargo internalization. This phenomenon arises from two factors: (1) transient membrane nanopores generated during deformation impose steric limitations on large molecules, and (2) smaller cargos exhibit faster passive diffusion rates.

[0189] Since membrane pore size is primarily determined by cell deformation, molecular weight itself does not affect pore size. Consequently, larger molecules encounter greater steric hindrance during pore transit. Furthermore, slower diffusion kinetics of high-molecular-weight cargos reduce their effective intracellular accumulation within the transient pore lifetime.

[0190] RNA Delivery Analysis

[0191] Following the successful intracellular delivery of FITC-dextran, we extended our investigation to RNA molecules to evaluate the method’s potential for enabling gene expression - a critical capability for biomedical applications.

[0192] A mRNA encoding an enhanced green fluorescent protein (eGFP) linked to a nuclear localization signal was selected for intracellular delivery into MDA-MB-231 cells. Once inside the cells, the eGFP mRNA generates an eGFP signal, confirming successful mRNA expression. To compare our approach with a commercial transfection method, we also performed mRNA transfection using lipofection. In each group, 0.1 pg of mRNA was delivered to 5 x 104cells. Fluorescence characterization at 24 hours post-transfection revealed robust eGFP signals in lipofected cells (Figure 14a).

[0193] In contrast to lipofection, cells treated with the filter and PEG solution for mRNA delivery exhibited weak green fluorescence. Fluorescence was only detectable when the microscope acquisition time was extended fivefold compared to lipofection, suggesting limited mRNA expression. Given the high delivery efficiency observed with FITC-dextran, we hypothesized that PEG solution might degrade mRNA during the mixing process or inhibit its expression post-delivery. Since combining different intracellular delivery methods can enhance overall performance, 43, 44 we integrated our filter-based and viscoelastic fluid delivery method with lipofection. In this approach, liposomes were first prepared using conventional lipofection, after which cells were suspended in PEG solution and mixed with liposomes. The mixture was then passed through the filter and cultured for 24 hours. With the protective lipid layer, mRNA was successfully delivered and expressed in cells (Figure 14a).

[0194] Quantitative transfection analysis was performed using flow cytometry. Figure 14b displays the fluorescence intensity distribution across different delivery methods. Cells transfected via lipofection exhibited two peaks, one of which overlapped with the control group, indicating that some cells did not receive mRNA. In contrast, transfection using the filter and viscoelastic fluid showed greater uniformity, with only a minimal subset of cells exhibiting no fluorescence change. The average delivery efficiencies for lipofection, the filter-based method, and the combined approach were 74%, 97.8%, and 94.7%, respectively (Figure 14c). Although the MFI fold change for lipofection was higher than for the filter-based approach alone (Figure 14d), the combination of filtering and lipofection resulted in an MFI fold change nearly double that of lipofection alone (192%), demonstrating improved transfection uniformity. Compared to lipofection, using the filter alone slightly reduced cell viability from 97.3% to 94.3% (Figure 14e), an acceptable range for practical applications. Given these findings, we propose that this method not only serves as an alternative to commercial transfection techniques but also offers an improved strategy for enhancing delivery performance.

[0195] Conclusion

[0196] We developed a novel intracellular delivery method using a stainless-steel filter with a Dutch weave mesh structure in combination with a viscoelastic fluid. The fundamental mechanism of this approach relies on the generation of transient nanopores in the cell membrane through mechanical deformation, a principle previously demonstrated in studies utilizing filters and microfluidic chips.

[0197] However, we replaced silicon wafer-based filters fabricated with a stainless-steel filter. Unlike micropores in silicon wafers, the stainless-steel filter, with an aperture size larger than the average cell diameter, effectively reduces clogging and minimizes cell death. The integration of a viscoelastic fluid with the filter exerts adequate pressure and shear forces to induce significant cell deformation, enhancing delivery efficiency. Additionally, the use of stainless-steel filters significantly reduces the preparation time required for creating micropores or micro constrictions compared to microfluidic chips and silicon wafer filters. Following optimization of delivery parameters, we achieved efficient intracellular delivery while preserving high cell viability. Furthermore, we demonstrated that this method can successfully transfect mRNA, achieving delivery performance comparable to that of lipofection. Beyond serving as an independent transfection method, this approach can also be combined with lipofection to enhance delivery efficiency and increase the dosage delivered into cells. Given the potential of transient nanopore formation in gene editing, we believe our intracellular delivery approach — leveraging stainless-steel filters and viscoelastic fluid — holds significant promise for biological and medical applications.

[0198] While embodiments of the invention have been particularly shown and described with reference to specific embodiments, it should be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the scope of the invention as defined by the appended claims. The scope of the invention is thus indicated by the appended claims and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced.

Claims

CLAIMS1. A method for delivering a cargo into a cell, the method comprising providing a mixture comprising of the cell and the cargo suspended in a viscoelastic fluid, and deforming the cells in the mixture, the deformation creates transient pores in the cell’s membrane allowing delivery of the cargo into the cell.

2. The method according to claim 1 , wherein the cell is deformed by passing the mixture through a micro-post array, the micro-post array comprises a first set of a plurality of microposts, wherein a pair of adjacent microposts forms a constriction point and the width of each constriction point is equal to or greater than the diameter of the cell, reducing direct contact between the cell and the microposts.

3. The method according to claim 2, wherein the width of each constriction point in the first set of microposts is equal to or greater than the diameter of the cell, preferably between 4pm to 28pm, preferably 16pm to 28pm, more preferably 20pm.

4. The method according to claim 3, further comprising passing the mixture through a second set of a plurality of microposts prior to the first set of micropost, wherein the width of each constriction point in the second set of microposts is between 32pm to 24pm.

5. The method according to any one of the preceding claims, wherein the viscoelastic fluid a biocompatible polymer in a Newtonian fluid, wherein the biocompatible polymer is polyethylene oxide (PEO), guar gum, xanthan gum, polyvinyl alcohol, hyaluronic acid, alginate or methylcellulose.

6. The method according to claim 5, wherein the concentration of the biocompatible polymer is 0.5 to 20mg / mL or 0.1 to 2%.

7. The method according to any one of the preceding claims, wherein the flow rate of the mixture the micro-post array is less than 100pL / min, preferably 50pL / min, or over 10 mL / s.

8. The method according to any one of the preceding claims, wherein the cargo concentration in the mixture is between 10pg / mL to 1000pg / mL, preferably 300pg / mL.

9. The method according to claim 1, wherein the cells are deformed by passing the mixture through a stainless-steel filter.

10. The method according to claim 9, wherein the viscoelastic fluid is a biocompatible polymer in a Newtonian fluid, wherein the polymer is polyethylene oxide (PEO), guar gum, xanthan gum, polyvinyl alcohol, hyaluronic acid, alginate or methylcellulose, and the concentration of the polymer in the Newtonian fluid is between 0.5 to 20 mg / mL.11 . The method according to any one of claims 9 or 10, wherein the flow rate of the mixture through the filter is between 2mL / min to 5mL / min.

12. The method according to any one of claims 9 to 11, wherein the aperture size of the filter is equal to or greater than the diameter of the cell, preferably is between 12pm to 25pm.

13. The method according to any one of claims 9 to 12, wherein the filter comprises a plain weave mesh or a Dutch weave mesh.

14. A microfluidic device for delivering a cargo into a cell, the device comprising a micropost array or a filter, wherein the width of the micro-constriction in the micro-post array or aperture size of the filter is equal to or greater than the diameter of the cell.

15. The device according to claim 14, wherein the micro-post array comprises a first set of a plurality of microposts, wherein a pair of adjacent microposts forming a constriction point and the width of each constriction point is equal to or greater than the diameter of the cell, reducing direct contact between the cell and the microposts.

16. The device according to any one of claims 14 or 15 wherein the width of each constriction point in the first set of microposts is equal to or greater than the diameter of the cell, preferably between 4pm to 28pm, preferably 16pm to 28pm, more preferably 20pm.

17. The device according to any one of claims 14 to 16, further comprising a second set of a plurality of microposts prior to the first set of micropost, wherein the width of each constriction point in the second set of microposts is between 32pm to 24pm.

18. The device according to any one of claims 14 to 17, wherein the distance between the first and second sets of microposts is 100pm.

19. The device according to any one of claims 14 to 18, further comprising a plurality of sets of microposts arranged in a series, wherein the width of the constriction points in each array is the same, and wherein the width of the constriction point in each array in the series progressively decreases in size.

20. The device according to any one of claims 14 to 19, wherein each micropost has a cylinder diameter of 40pm and a height of 20-400pm.

21. The device according to claim 14, wherein the filter is a stainless steel filter and comprises a plain weave mesh or a Dutch weave mesh.

22. The device according to claim 21 , wherein the filter is a stainless steel filter with an aperture size ranges from 3 pm to 100 pm.

23. The device according to any one of claims 21 to 23, wherein the filter is integrated into the dispensing tip or a filter holder connected to a syringe.

Images