Conjugating Micropatches to Living Cells Through Membrane Intercalation

Abstract

Existing clinical cell therapies, which rely on the use of biological functionalities of living cells, can be further enhanced by conjugating functional particles to the cells to form cell-particle complexes. Disk-shaped microparticles produced by the top-down microfabrication approach possess unique advantages for this application. However, none of current mechanisms for conjugating the microfabricated microparticles to the cells is principally applicable to all types of cells with therapeutic potentials. On the other hand, membrane intercalation is a well-established mechanism for attaching fluorescence molecules to living cells or immobilizing cells to a solid surface. This paper reports a study on conjugating the disk-shaped microparticles referred to as micropatches to living cells through membrane intercalation for the first time. The procedure for producing the cell-micropatch complexes features an unprecedented integration of microcontact printing of micropatches, end-grafting of linear molecules of octadecyl chain and poly(ethylene glycol) to the printed micropatches, and use of gelatin as a temperature-sensitive sacrificial layer to allow the formation and subsequent release of the cell-micropatch complexes. Complexes composed of mouse neuroblastoma cells were found to be stable in vitro and the micropatch-bound cells were viable, proliferative, and differentiable. Moreover, complexes composed of four other types of cells were produced. The membrane-intercalation mechanism and the corresponding fabrication technique developed in this study are potentially applicable to a wide range of therapeutic cells and thus promise to be useful for developing new cell therapies enhanced by the disk-shaped microparticles.

Graphical Abstract

1. Introduction

Recent years have seen tremendous advances in research, development and clinical application of cell therapies.^1,2 Common to the mainstream cell therapies is their reliance on the biological functionalities of living cells such as the abilities of stem cells to differentiate into specialized cells and cytotoxic T lymphocytes to kill cancer cells.^3,4 On the other hand, artificial micro/nano-particles can be engineered to possess a wide range of functionalities that are radically distinct from those of the cells. For examples, micro/nano-particles can be used to deliver drugs, enable advanced medical imaging, kill tumor cells by converting magnetic energy to mechanical force, or induce and monitor stem-cell differentiation.^5,6,7,8 Physical integration of a therapeutic cell with micro/nano-particles therefore promises to create a hybrid structure with a combined or even synergistic therapeutic capability unmatchable by its individual components. The hybrid structure is herein termed cell-particle complex.

Attempts have been made to create functional cell-particle complexes. Irvine’s group chemically attached drug-laden nanoparticles to thiol groups on the surface of T lymphocytes or hematopoietic stem cells (HSCs) to activate the T lymphocytes or promote proliferation of the HSCs in vivo.⁹ However, the nanoparticles were produced by the bottom-up approach, which generally suffers from limited ability to control shape, size and structure of the particles. On the contrary, the top-down approach can produce particles with precisely controlled shapes, sizes and compositions. Rubner’s, Mitragotri’s and Batrakova’s groups combined photolithography and layer-by-layer (LbL) assembly to fabricate disk-shaped microparticles termed cellular backpacks and conjugated the backpacks to cells through the ligand-receptor binding mechanism.^10,11,12,13 Recently, Mitragotri’s group fabricated cellular backpacks loaded with interferon-γ for macrophage immunotherapy with a microcontact printing (μCP)-based method.¹⁴ The ligand-receptor binding was used again to conjugate the backpacks to the macrophages. While various ligand-receptor pairs exist, probably none is applicable to all types of therapeutic cells. We originally developed the μCP-based method for fabricating disk-shaped microparticles and conjugated the microparticles to various types of cells by coating the microparticles with a polycation.^15,16,17,18 Electrostatic attraction between the polycation and sialic acids on cell surface was hypothesized to be the conjugation mechanism. However, the hypothesis has not been strictly confirmed and it is unknown whether all existing or potential therapeutic cells have a negative surface charge. Taken together, a method that, in principle, allows for conjugating disk-shaped microparticles to all therapeutic cells, has not been developed. Such a universally applicable method would be desirable for simplifying the development of clinically useful products based on the cell-microparticle complexes.

On the other hand, membrane intercalation as a mechanism of conjugation has been widely used to label living cells with fluorescence dyes such as octadecyl rhodamine B (R18) and DiI series dyes.¹⁹ Moreover, Froelich et al. found that DiI labeling of human adipose-derived stem cells was not toxic to the cells and did not impair the proliferation, migration or differentiation potential of the cells.²⁰ Each R18 molecule has an octadecyl chain (-C₁₈H₃₇) and each DiI-series molecule has two octadecyl chains. The octadecyl chains can intercalate into the lipid bilayer of the cell membrane. This mechanism has also been used to immobilize animal and human cells to a solid surface by Nagamune’s group.^21,22,23 In their method, a linear molecular chain composed of an oleyl segment at one end, a linear poly(ethylene glycol) (PEG) segment in the middle, and an N-hydroxy-succinimidyl ester (NHS) group at the other end was grafted to a solid surface through the NHS-amine reaction. Intercalation of the dangling oleyl segments into the cell membrane allowed immobilization of the cells. The PEG chain functioned as a flexible linker to prevent steric hindrance between the cell and the surface. Using this technique, the group immobilized a wide variety of animal and human cells. Since every human cell has a lipid bilayer as the major component of its membrane, this membrane-intercalation mechanism is, in principle, applicable to all human cells.

We hypothesize the membrane intercalation can be utilized to conjugate the disk-shaped microparticles, which are termed micropatches here, to the therapeutic cells as illustrated in Figure 1. This paper reports the production of cell-micropatch complexes based on this mechanism through a novel procedure. Mouse neuro-2a (N2a) neuroblastoma cells, which are widely used to study neuronal differentiation, were used as the major model cell in this study.²⁴ Stability of the N2a-micropatch complexes, and the effects of micropatch attachment on viability, proliferation and differentiation of the N2a cells were characterized. Four other types of cells, including tumor cells, endothelial cells, macrophages and mesenchymal stem cells, were used to demonstrate the broad applicability of this technique.

Figure 1.

Schematic of conjugating a micropatch to a living cell via the membrane-intercalation mechanism. One face of the disk-shaped micropatch bears dangling octadecyl chains with PEG segments as linkers. The octadecyl chains intercalate into the lipid bilayer of the cell membrane, conjugating the micropatch to the cell.

2. Experimental section

2.1. Materials

Brij® S 100 polyoxyethylene (100) stearyl ether [PEG octadecyl ether, C₁₈H₃₇(OCH₂CH₂)_nOH, n ≈ 100, average Mn ≈ 4,670 Da], pyridine, gelatin from porcine skin, poly(D,L-lactide-co-glycolide) acid terminated (PLGA, lactide : glycolide = 75 : 25, Mw ≈ 4 – 15 kDa), thiazolyl blue tetrazolium bromide (MTT), bovine serum albumin (BSA), (3-aminopropyl)trimethoxysilane (APTES), succinic anhydride, N-hydroxysuccinimide (NHS), N,N’-dicyclohexylcarbodiimide (DCC), and CDCl₃ were purchased from Sigma-Aldrich. Poly(allylamine hydrochloride) (PAH) with Mw of 120 – 200 kDa, poly(styrene sulfonic acid) sodium salt (PSS) with Mw of 70 kDa, retinoic acid, trypsin (0.25%) EDTA, fetal bovine serum (FBS), and minimum essential medium eagle-alpha modification (α-MEM) were purchased from VWR. Poly(dimethyl siloxane) (PDMS) kit (Sylgard 184) was purchased from Dow Corning. Poly(n-propyl methacrylate) (PPMA) with Mw of 150 kDa was purchased from Scientific Polymer. Di-sodium tetraborate/sodium hydroxide buffer solution (pH 10) was purchased from Honeywell. Dulbecco’s modified eagle’s medium (DMEM, with 4.5 g/L glucose and 4 mM L-glutamine) was purchased from Lonza Bioscience. Octadecyl rhodamine B chloride (R18) and fluorescein isothiocyanate (FITC) were purchased from Biotium. Poloxamer 407 was purchased from Spectrum. NHS-functionalized PEG (PEG-NHS, Mw = 5 kDa) was purchased from Nanocs Inc.

2.2. Cell culture

Mouse neuroblastoma cells (N2a), mouse endothelial cells (bEnd.3), mouse macrophages (RAW 264.7) and human cervical cancer cells (HeLa) were cultured in complete growth medium (DMEM supplemented with 10% FBS, 100 units/mL of penicillin and 100 μg/mL streptomycin) at 37 °C and 5% CO₂. Human mesenchymal stem cells (hMSCs) were obtained from the Tulane Center for Stem Cell Research and Regenerative Medicine and cultured in α-MEM supplemented with 10% FBS, 100 units/mL penicillin and 100 μg/mL streptomycin under 37 °C and 5% CO₂.

2.3. Preparation of PDMS stamps

Prepolymer and curing agent of the PDMS kit were mixed at 10:1 weight ratio for 10 min. The mixture was poured on a silicon master with micron scale cavities prepared by photolithography. Then, the mixture was degassed using a vacuum desiccator and cured at 37 °C for 24 h. The cured PDMS slab was peeled off from the master and cut into 1 cm × 1 cm square stamps. Two types of stamps were used in this study. One carried 7 μm-diameter pillars in a square lattice with a center-to-center distance of 20 μm. The other carried 5 μm-diameter pillars in a square lattice with a center-to-center distance of 10 μm. The 7 μm-diameter-pillar stamps were used throughout this study unless otherwise noted.

2.4. Synthesis of NHS-PEG-C₁₈H₃₇

PEG octadecyl ether (4 g) was reacted with 2 equivalents of succinic anhydride (171.3 mg) to convert the terminal group from alcohol to carboxylic acid in the presence of 0.1 equivalent 4-dimethylaminopyridine (DMAP) (10.5 mg) as shown in Scheme 1A. All reagents were dissolved in degassed tetrahydrofuran (THF) and the reaction mixture was stirred at 70 °C for 24 h. Followed by stirring, the solution was concentrated by rotary evaporation. The synthesized HOOC-PEG-C₁₈H₃₇ was purified by column chromatography with a mobile phase of 7% MeOH/chloroform. The white crystal product, HOOC-PEG-C₁₈H₃₇, was characterized by ¹H NMR in CDCl₃ to determine chemical structure includes carboxylic acid groups using a Bruker Avance III 400 MHz. The carboxylic acid end of HOOC-PEG-C₁₈H₃₇ (1 g) was further reacted with NHS (2.2 eq., 53.8 mg) to covalently introduce NHS group to HOOC-PEG-C₁₈H₃₇ via esterification. The reaction was performed in the presence of DCC (96.2 mg) and pyridine (1–2 drops) in a mixture of dichloromethane (DCM) and dimethylformamide (DMF) (volume ratio: 1:1, volume: 5 mL) as shown in Scheme 1B. The reaction mixture was stirred at 0 °C for 2 h and then room temperature for 20 h. After stirring, white residue, dicyclohexylurea, was observed and removed by vacuum filtering. The filtrate was dried by rotary evaporator to remove solvent. Orange color, oil-like crude product was obtained. The crude product was dissolved in large excess of acetone (20 mL). The mixture was cooled to −78 °C in dry ice-acetone bath and then white solid was precipitated. The white solid was collected by vacuum filtering. After the cold solid melted to liquid at room temperature, the solution was concentrated again with a rotary evaporator. The white product was dissolved in acetone and precipitated in cold ether twice to obtain the final product. The white color product was characterized by ¹H NMR in CDCl₃ as described above.

Scheme 1.

Synthesis of (A) HOOC-PEG-C₁₈H₃₇ and (B) NHS-PEG-C₁₈H₃₇. RT: room temperature.

2.5. Preparation of surfaces for measuring water contact angles

Four types of surfaces were prepared by coating glass slides with APTES, PPMA, PAH/PSS/PAH, and PAH/PSS/PAH/BSA, respectively. The APTES-coated glass slides were prepared by vapor deposition. Briefly, a glass slide was treated with oxygen plasma using a Harrick plasma cleaner (duration: 30 s, power level: high, pressure: 600 mtorr) and then placed in a vacuum desiccator together with 200 μL APTES in a centrifuge tube. After creating vacuum inside, the desiccator was placed an oven at 37 °C and kept overnight. The slide was then washed with water and dried under a stream of nitrogen. The PPMA-coated glass slide was prepared as follows. PPMA in acetone (5 wt%) was spin-coated (3,000 rpm, 45 s) on a flat PDMS stamp. The stamp was placed on a slide, which was pre-treated with oxygen plasma as above, sitting on a hotplate at 100 °C. After 10 s, the stamp was peeled off from the slide. The PAH/PSS/PAH-coated glass slides were prepared as follows. An aqueous solution of PAH in water (1 wt%, 20 μL, pH 4) was dropped on a flat PDMS stamp. A glass coverslip was placed on the drop to spread it over the entire top surface of the stamp, so that the stamp surface was soaked in the PAH solution. After 15 min, the coverslip was removed, the stamp was rinsed with water and dried with a stream of nitrogen. This step was repeated twice with PSS in water (1 wt%, 20 μL) and the same PAH solution as above. The stamp was then placed on a glass that was pre-treated with oxygen plasma as above and kept for 10 sec. Finally, the stamp was peeled off from the slide. The PAH/PSS/PAH/BSA-coated glass slides were prepared by coating a flat stamp with BSA first as above except using BSA in PBS (1 wt%, 20 μL) and a soaking time of 30 min. The PAH/PSS/PAH trilayer was subsequently deposited on the stamp as above. The glass slides coated with APTES, PPMA, PAH/PSS/PAH, and PAH/PSS/PAH/BSA receptively were all soaked in di-sodium tetraborate/sodium hydroxide buffer. A PDMS pad (thickness: 7 mm) bearing a circular through hole (diameter: 9 mm) was placed on each slide to form a well, into which the pre-cooled buffer (400 μL, pH 10, 8 °C) was added. After 30 min, the buffer was aspirated and the slide was rinsed with pre-cooled PBS (8 °C, 400 μL) three times. Counterparts of the four types of native surfaces were prepared by treating them with NHS-PEG-C₁₈H₃₇. The procedure was the same as above except that NHS-PEG-C₁₈H₃₇ was added in the di-sodium tetraborate/sodium hydroxide buffer (5 mg/mL) in the soaking step.

2.6. Measurement of water contact angles

Advancing contact angles of water were measured in air by the sessile-drop method using a KSV CAM 200 instrument. The volume of each water droplet was 5 μL. The samples were prepared and measured in triplicate.

2.7. Cell adhesion to different surfaces

Two pairs of surfaces were prepared. One pair, including the PAH/PSS/PAH/BSA surface and its counterpart treated with NHS-PEG-C₁₈H₃₇, were prepared as above. The other pair consisted of PPMA/PAH/PSS/PAH surface and its counterpart treated with NHS-PEG. The PPMA/PAH/PSS/PAH surface was prepared by depositing a PAH/PSS/PAH trilayer on a flat PDMS stamp by LbL as above and then a PPMA thin film on the trilayer by spin-coating as above. The PPMA/PAH/PSS/PAH multilayer was transfer-printed on a glass slide as above. The PPMA/PAH/PSS/PAH surface was treated with NHS-PEG as above except replacing NHS-PEG-C₁₈H₃₇ with NHS-PEG. A PDMS film (thickness: 7 mm) bearing a circular through hole (diameter: 9 mm) was placed on one of the surfaces to create a well. To inspect whether N2a cells bound to the surface after a 30 min incubation at 8 °C, a suspension of cells (cell number: 1.5 × 10⁵, 400 μL) containing calcein AM (1 μg/mL) pre-cooled at 8 °C was added in the well and kept at 8 °C for 30 min. After recording the images of the bottom surface of the well, the medium in the well was then gently pipetted and the bottom surface of the well was imaged again. To inspect whether the cells bound to the surface after an additional 2 h incubation at 37 °C, the cells were added to the well as above and was further incubated for 30 min at 8 °C and then 2 h at 37 °C and 5% CO₂. The samples were imaged as above. Numbers of cells were counted from the images using the Nikon NIS-Elements software and used to calculate percentages of the immobilized cells for the surfaces.

2.8. Preparation of gelatin sacrificial layers

A gelatin sacrificial layer was prepared by spreading a warm aqueous solution of gelatin (0.5 wt% unless otherwise noted, 20 mg, 50 °C) on a glass slide with a pipette. The diameter of the final circular gelatin solution film was 1.5 cm. The solution film was allowed to dry in air at the room temperature for overnight unless otherwise noted.

2.9. Fabrication of cell-micropatch complexes

Cell-micropatch complexes composed of PPMA/PAH/PASS/PAH-PEG-C₁₈H₃₇ micropatches were produced as follows (Figure 2). An aqueous solution of PAH in water (1 wt%, 20 μL, pH 10) was dropped on a micropillar stamp. A glass coverslip was placed on the drop to spread it over the entire top surface of the stamp. After 15 min, the coverslip was removed, the stamp was rinsed with water and dried with a stream of nitrogen. This step was repeated with PSS in water (1 wt%, 20 μL) and PAH in water (1 wt%, 20 μL, pH 4) as above. An acetone solution of PPMA (5 wt%) and R18 (25 μg/mL) was spin-coated (3,000 rpm, 45 s) on the stamp. The stamp was then placed on a gelatin-coated glass slide on a hotplate (100 °C) and kept for 10 s before being peeled off. A PDMS film (thickness: 7 mm) with a circular through hole (diameter: 9 mm) was placed on the gelatin-coated slide to form a well enclosing the printing area. The slide was then cooled to 8 °C. A pre-cooled NHS-PEG-C₁₈H₃₇ in di-sodium tetraborate/sodium hydroxide buffer (5 mg/mL, 400 μL, pH 10, 8 °C) was added into the well and kept at 8 °C for 30 min. After removing the NHS-PEG-C₁₈H₃₇ solution and washing the slide with cold water (8 °C) for three times, a pre-cooled (8 °C) suspension of cells (cell number: 1.5 × 10⁵, 400 μL) in the serum-free medium (DMEM for N2a, bEnd.3 RAW 264.7 and HeLa cells, and α-MEM for hMSCs) containing calcein AM (1 μg/mL) was added onto the slide and kept for 30 min. The slide was then transferred to a CO₂ incubator at 37 °C and kept for 2 h. The released complexes were purified by centrifuging (60 × g unless otherwise noted, 10 min) the suspension, removing the supernatant, and dispersing the pellet in the FBS-supplemented medium. Percentage of micropatch-bound cells was determined by using a homemade hemocytometer.¹⁵ The dispersed micropatch-bound cells and micropatch-free cells called bare cells in this study were re-seeded in a well of a 12 or 24 -well plate and cultured at 37 °C and 5% CO₂ for 24 h for N2a cells, hMSCs, bEnd.3 cells, and HeLa cells or 72 h for RAW 264.7 cells. The N2a cells and complexes were then treated with trypsin (0.25%) EDTA (100 μL, 5 min). The re-released cells were imaged and percentage of micropatch-bound N2a cells was determined as above.

Figure 2.

Procedure for producing the cell-micropatch complexes via the membrane-intercalation mechanism. The composition of the micropatches is denoted as PMMA(R18)/PAH/PSS/PAH-PEG-C₁₈H₃₇.

Cell-micropatch complexes composed of PPMA/PAH/PASS/PAH micropatches and PPMA/PAH/PASS/PAH-PEG micropatches respectively were produced as follows. PPMA/PAH/PASS/PAH micropatches were prepared as above for producing the PPMA/PAH/PASS/PAH-PEG-C₁₈H₃₇ micropatches except in the following two aspects. First, the gelatin sacrificial layer was prepared from a 1 wt% solution and dried on a 50 °C hot plate for 30 min. Second, the step of treating the PPMA/PAH/PASS/PAH micropatches with NHS-PEG-C₁₈H₃₇ was removed. PPMA/PAH/PASS/PAH-PEG micropatches were prepared as above for preparing the PPMA/PAH/PASS/PAH-PEG-C₁₈H₃₇ micropatches except using the gelatin sacrificial layer prepared from a 1 wt% solution and dried on a 50 °C hot plate for 30 min and replacing NHS-PEG-C₁₈H₃₇ with NHS-PEG. Cell-micropatch complexes composed of these two types of micropatches were prepared as above except that a relative centrifugal force of 124 × g was used in the centrifugation step.

Cell-micropatch complexes composed of dot-on-pad micropatches were produced as follows. PAH labeled with FITC (PAH-FITC) was prepared first by adding FITC to an aqueous solution of PAH (1 wt%) at concentration of 20 μM. After 12 h at the room temperature, the solution was dialyzed against 1,000 mL water using a membrane with a molecular weight cutoff of 12,400 Da for 24 h. To enhance transfer of the micropatches to the gelatin sacrificial layer, PLGA nanoparticles were prepared as follows. PLGA in acetone solution (10 mg/mL, 1 mL) was slowly added into an aqueous solution of poloxamer 407 (0.25% w/v, 1 mL) under magnetic stirring (1,600 rpm) at the room temperature and kept for 15 min. The mixture was then placed in a desiccator, in which a low vacuum was created, and kept overnight. The mixture was centrifuged (20,000 × g, 10 min). The supernatant was removed and the pellet was re-dispersed in 1 mL water. To produce the dot-on-pad micropatches, an acetone solution of PPMA (5 wt%) and R18 (25 μg/mL) was firstly spin-coated onto a 7 μm-diameter pillar stamp (4,000 rpm, 45 s). A multilayer of PAH/(PSS/PAH-FITC)₅/PSS/PAH was then deposited onto the stamp as above. The PLGA nanoparticle suspension (30 μL) was added on the stamp and soaked for 30 min. After being rinsed with water and dried with a nitrogen stream, the stamp was placed on gelatin-coated glass slide on a hotplate at 75 °C. After 10 s, the stamp was peeled off from the slide. The rest of the process was the same as the production of cell-micropatch complexes composed of the other type of micropatches.

2.10. Microscopy

Optical micrographs were obtained using an inverted Nikon Ti epifluorescence microscope equipped with an Andor iXonEM + 885 EMCCD camera.

2.11. MTT assay

Samples were prepared in triplicate in 24-well plates. Each well contained about 8 × 10⁴ total cells. The complete growth medium in each well was replaced with MTT reagent in the complete growth medium (1 mg/mL, 200 μL). After being incubated at 37 °C and 5% CO₂ for 4 h, the MTT-containing medium was removed. Dimethyl sulfoxide (150 μL) was added to each well. The plate was covered with foil and mounted on a rocking shaker for 10 min. The mixture was centrifuged, photoabsorbance of supernatant was read at 570 nm using a microplate reader.

2.12. Proliferation assay

A square PDMS film (5 cm wide, 0.8 mm thick) bearing a 10 × 10 square array of 100 circular holes (3 mm diameter) was placed on the bottom of a cell culture dish (7 cm diameter) to form an array of wells. The dish was then filled with the complete growth medium. Approximately 200 bare cells and cell-micropatch complexes were added into the dish. The dish was then placed in an incubator. After 24 h, the wells were examined under a microscope. Wells that contained single micropatch-bound cells and single bare cells were identified. Three samples were prepared. The numbers of single micropatch-bound cells in a well and number of single bare cells in a well were 9 and 8, 11 and 9, and 7 and 12 respectively for the individual samples. The single micropatch-bound cells and single bare cells were examined at 48 h, 72 h and 120 h, respectively.

2.13. Differentiation assay

Each sample contained a mixture of about 5 × 10⁴ micropatch-bound and bare cells in a well of a 12 well plate. At each time point (0 h, 24 h, 48 h, 72 h), cells treated with the differentiation-promoting medium (DMEM supplemented with 2% FBS, 100 units/mL of penicillin and 100 μg/mL streptomycin, and 20 μM retinoic acid, four samples) or the regular medium (complete growth medium, three samples) were imaged at randomly selected areas (four for each differentiation-promoted sample and three for each control sample).²⁵ Moreover, the cells were stained with calcein AM (1 μg/mL) and Hoechst 33342 (1 μg/mL) at 72 h. Cells that developed one or more neurites with a length greater than twice the diameter of the cell body were assigned as differentiated neurons.²⁶ Numbers of differentiated micropatch-bound cells, total micropatch-bound cells, differentiated bare cells and total bare cells were counted. The numbers were used to calculate percentages of differentiation among micropatch-bound cells and bare cells respectively.

2.14. Statistical analysis

Experiment data are expressed as mean (± standard deviation). Student’s t test was performed to compare the differences of the data. The differences were considered as statistically significant at p < 0.05 (denoted as *), very significant at p < 0.01 (denoted as **), and highly significant at p < 0.001 (denoted as ***). The difference was regarded as not significant (n.s.) when p > 0.05.

3. Results and Discussion

3.1. Synthesis of NHS-PEG-C₁₈H₃₇

NHS-PEG-C₁₈H₃₇ was essential to the fabrication of the PMMA(R18)/PAH/PSS/PAH/NHS-PEG-C₁₈H₃₇ micropatches. In order to synthesize NHS-PEG-C₁₈H₃₇, HO-PEG-C₁₈H₃₇ was converted to HOOC-PEG-C₁₈H₃₇, which was an intermediate precursor to prepare NHS-PEG-C₁₈H₃₇ (Scheme 1A). As shown in the ¹H NMR (Figure 3A), the appearance of protons between ester and carboxylic group (i at 2.63 ppm and h at 2.61 ppm) confirmed successful synthesis of HOOC-PEG-C₁₈H₃₇. The prepared HOOC-PEG-C₁₈H₃₇ underwent carbodiimide-mediated esterification with N-hydroxysuccinimide to yield NHS-PEG-C₁₈H₃₇ (Scheme 1B). Figure 3B is ¹H NMR spectrum of the synthesized NHS-PEG-C₁₈H₃₇. The appearance of protons that belong to NHS group (-CH₂-CH₂-, j at 2.81 ppm) and the shift of -CH₂-CH₂- protons’ signals (h and i in the chemical structure of Figures 3A and 3B). Specifically, newly formed ester linkage led chemical shift to change from 2.63 to 2.94 ppm (-CH₂-, i) and from 2.61 to 2.76 ppm (-CH₂-, h), confirming successful synthesis of NHS-PEG-C₁₈H₃₇.

Figure 3.

¹H NMR spectra of (A) HOOC-PEG-C₁₈H₃₇ and (B) NHS-PEG-C₁₈H₃₇ in CDCl₃.

3.2. Water contact angles of NHS-PEG-C₁₈H₃₇-grafted surfaces

To confirm that the synthesized NHS-PEG-C₁₈H₃₇ could be grafted to a surface bearing primary amine, an APTES-coated glass slide surface was used as a model surface and advancing water contact angles on the surface were measured before and after being treated with NHS-PEG-C₁₈H₃₇. As shown in Figure 4, the contact angle was 91.3° ± 4.9° before the treatment and 64.2° ± 0.93° after the treatment. The ~30% decrease indicates that NHS-PEG-C₁₈H₃₇ was successfully deposited the ATPES surface. The hydrophilic PEG segments were probably responsible for the decrease of the contact angle. Besides covalent bonding, it was possible that NHS-PEG-C₁₈H₃₇ was physically adsorbed to the ATPES surface due to the hydrophobic interaction between the octadecyl chains and the APTES surface. To investigate this possibility, PPMA, which is also hydrophobic and does not contain primary amine groups, was used as a control surface. As shown in Figure 4, the contact angle on the PPMA surface was 84.3° ± 0.94° before the NHS-PEG-C₁₈H₃₇ treatment and 87.1° ± 2.5° after the treatment. The ~3% increase in the contact angle suggests that physical adsorption of NHS-PEG-C₁₈H₃₇ to a hydrophobic surface was responsible for only a small change in the contact angle. Therefore, NHS-PEG-C₁₈H₃₇ was probably covalently grafted to the APTES surface. Contact angles on PAH/PSS/PAH/BSA-coated and PAH/PSS/PAH-coated glass slides were also measured before and after the NHS-PEG-C₁₈H₃₇ treatment because the two types of surfaces were used for probing cell adhesion and assembling cell-micropatch complexes in this study. The result (Figure 4) shows that NHS-PEG-C₁₈H₃₇ treatment caused a small, i.e., ~2%, decrease in the water contact angle to the PAH/PSS/PAH/BSA surface, and a large, i.e., ~16%, decrease to the PAH/PSS/PAH surface. Given that both BSA and PAH carried primary amine groups, the result indicates that NHS-PEG-C₁₈H₃₇ was probably grafted to the PAH/PSS/PAH surface, but it does not reveal whether NHS-PEG-C₁₈H₃₇ was grafted to the PAH/PSS/PAH/BSA surface.

Figure 4.

Advancing water contact angles on native APTES, PPMA, PAH/PSS/PAH/BSA, and PAH/PSS/PAH surfaces and the same surfaces treated with NHS-PEG-C₁₈H₃₇. Bars represent averages and whiskers represent standard deviations. n = 3.

3.3. Cell adhesion to different surfaces

To study whether an NHS-PEG-C₁₈H₃₇-grafted surface could immobilize living cells through membrane intercalation, we prepared a pair of surfaces. One was a PAH/PSS/PAH/BSA multilayer transfer-printed on a glass slide with the BSA layer as the top layer and this surface was called PAH/PSS/PAH/BSA surface. The other was the PAH/PSS/PAH/BSA surface treated with NHS-PEG-C₁₈H₃₇, denoted as PAH/PSS/PAH/BSA-PEG-C₁₈H₃₇ surface. N2a cells suspended in the cell-culture medium were added to both surfaces and incubated for 30 min at 8 °C. For the PAH/PSS/PAH/BSA surface, the cells were found to settle on the surface (Figure S1A1). After gently pipetting the medium, no cells remained on the surface (Figure S1A2), indicating that the cells did not adhere to the PAH/PSS/PAH/BSA surface. For the PAH/PSS/PAH/BSA-PEG-C₁₈H₃₇ surface, cells settled on the surface after the 30 min-8 °C incubation (Figure S1B1), and almost all of them remained on the surface after being gently pipetted (Figure S1B2). Result of quantitative analysis of three independent experiments is shown in Figure 5, revealing that approximately 81% ± 7% of cells adhered to the PAH/PSS/PAH/BSA-PEG-C₁₈H₃₇ surface, while 0% adhered to the PAH/PSS/PAH/BSA surface. These results indicate that NHS-PEG-C₁₈H₃₇ was grafted to the PAH/PSS/PAH/BSA surface and immobilized the cells.

Figure 5.

Adhesion of N2a cells to different surfaces. Bars represent averages and whiskers represent standard deviations. n = 3.

To confirm that PEG was non-adhesive to N2a cells, the above cell-adhesion experiment was performed on another pair of surfaces. One was a PMMA/PAH/PSS/PAH multilayer transfer-printed on a glass slide and was called PMMA/PAH/PSS/PAH surface. The other was the PMMA/PAH/PSS/PAH surface treated with NHS-PEG, denoted as PMMA/PAH/PSS/PAH-PEG surface. The result is shown in Figures 5 and S2, revealing that 78% ± 17% of the cells adhered to the PPMA/PAH/PSS/PAH surface, but only 14% ± 8% of the cells adhered to the PAH/PSS/PAH-PEG surface. The above results indicate that the surface-grafted PEG was inhibitive to the adhesion of N2a cells and the octadecyl chain of the PAH/PSS/PAH/BSA-PEG-C₁₈H₃₇ surface was therefore responsible for immobilizing the cells probably through membrane intercalation.

In addition to the 30 min-8 °C incubation, a 2 h-37 °C incubation was used to release the complexes from a glass slide by dissolving the gelatin sacrificial layer in the production of the cell-micropatch complexes. To determine whether this particular incubation step affected binding of the cells to the above surfaces, the above experiments were conducted again except including the 2 h-37 °C incubation after the 30 min-8 °C incubation. Representative images are shown in Figures S3 and S4. Quantitative analysis of three independent experiments revealed that approximately 94% ± 2% of cells were immobilized on the PAH/PSS/PAH/BSA-PEG-C₁₈H₃₇ surface, 0% was immobilized on the PAH/PSS/PAH/BSA surface, 97% ± 8% of cells were immobilized on the PPMA/PAH/PSS/PAH surface, and 38% ± 1% was immobilized on the PPMA/PAH/PSS/PAH-PEG surface following the 30 min-8 °C plus 2 h-37 °C incubation (Figure 5). This result indicates that the 2 h-37 °C incubation did not reduce adhesion of the cells to the PAH/PSS/PAH/BSA-PEG-C₁₈H₃₇ surface and confirms again that the octadecyl chains were responsible for immobilizing the cells.

3.4. Production of N2a-micropatch complexes

Figure 6A shows N2a cells on an array of PMMA(R18)/PAH/PSS/PAH/NHS-PEG-C₁₈H₃₇ micropatches on a gelatin sacrificial layer after a 30 min-8 °C incubation. After an additional 37 °C-2 h incubation, the cells and micropatches were released from the substrate with the micropatches being colocalized with the released cells (Figure 6B). The colocalization was observed after removing free micropatches with centrifugation (Figure 6C), indicating that the micropatches were stably bound to the cells. The centrifugation-purified N2a-micropatch complexes were then re-seeded on a cell-culture surface. The as-prepared micropatch-bound cells were able to adhere to and spread on the surface after 24 h (Figure 6D), qualitatively indicating that micropatch attachment did not interfere with this intrinsic ability of the cells. To characterize the stability of the cell-micropatch complexes, the re-seeded complexes were trypsinized. The micropatch-bound cells were released from the surface and the micropatches remained bound to the released cells as shown in Figure 6E. A quantitative measurement (Figure 6F) shows approximately 65% ± 5% of the cells were bound by micropatches after centrifugation (Figure 6C) and 45% ± 3% of the cells were bound by micropatches after trypsinization (Figure 6E). The high survival percentages of the complexes through the centrifugation, re-seeding, and trypsinization indicates that the membrane intercalation-based cell-micropatch conjugation was stable.

Figure 6.

Fabrication and characterization of stability of N2a-micropatch complexes. Overlaid micrographs of (A) N2a cells seeded on an array of PPMA/PAH/PSS/PAH-PEG-C₁₈H₃₇ micropatches following an 8 °C-30 min incubation, (B) suspended N2a cells and micropatches just released from the substrate by dissolving the gelatin sacrificial layer, (C) suspended micropatch-bound cells and bare cells after removing free micropatches through centrifugation, (D) N2a-micropatch complexes re-seeded on a cell-culture surface, and (E) suspended micropatch-bound and bare cells re-released from the substrate by trypsinizing the re-seeded cells. (F) Percentages of micropatch-bound N2a cells in samples prepared using PPMA/PAH/PSS/PAH-PEG-C₁₈H₃₇, PPMA/PAH/PSS/PAH, and PPMA/PAH/PSS/PAH-PEG micropatches respectively. The samples were as-prepared except that one was re-released as marked. (G) Schematic of a dot-on-pad micropatch bound to a cell. (H) Overlaid phase-contrast and fluorescence image of suspended N2a cells bound by the dot-on-pad micropatches via membrane intercalation. The inset in (H) shows a magnified cell bound by two dot-on-pad micropatches. Micropatches in (A-E) were labeled with R18 (red). The dots of the dot-on-pad micropatches were labeled with R18 (red) and pads of the micropatches were labeled with FITC (green). Cells in (A-C) were labeled with calcein AM (Green).

The technique developed in this study for producing the cell-micropatch complexes features the use of a gelatin sacrificial layer. Porcine gelatin forms hydrogel in water at a relatively low temperature (< 20 °C) but is dissolved in water at the human body temperature (37 °C).²⁷ This property of gelatin was utilized to allow firstly grafting NHS-PEG-C₁₈H₃₇ to the micropatches at 8 °C, then seeding the cells on the micropatches at 8 °C, and finally releasing the cell-micropatch complexes at 37 °C. This technique is different from a method that we previously developed for producing cell-micropatch complexes based on the use of a temperature-sensitive sacrificial layer made of poly(N-isopropylacrylamide) (PNIPAM).¹⁷ In the PNIPAM-based method, cells were seeded on the surface-bound micropatches at 37 °C and the complexes were released at the room temperature.

PAH-terminated rather than BSA-terminated micropatches were used to produce the cell-micropatch complexes because PAH, which contains a higher density of primary amine than BSA, could allow grafting a higher density of NHS-PEG-C₁₈H₃₇ to the micropatches. This hypothesis was supported by the fact that NHS-PEG-C₁₈H₃₇ treatment caused a significant decrease in water contact angle on the flat PAH/PSS/PAH surface, but not on the flat PAH/PSS/PAH/BSA surface (Figure 4). It is also of note that both the micropatches and the uncovered area of the gelatin sacrificial layer were exposed to the NHS-PEG-C₁₈H₃₇ in the step of grafting the NHS-PEG-C₁₈H₃₇ to the micropatches. Therefore, NHS-PEG-C₁₈H₃₇ should have been grafted to not only the micropatches, but also the gelatin layer because gelatin contains primary amine.²⁸ Given the fact that porcine gelatin contains a low percentage of residues carrying primary amine, we believe that the density of grafted NHS-PEG-C₁₈H₃₇ on the gelatin was much lower than that on the PAH-terminated micropatches. It is also worthwhile to note that the NHS-PEG-C₁₈H₃₇ grafted to the gelatin probably intercalated into the membrane of cells on the gelatin layer. Since each NHS-PEG-C₁₈H₃₇ molecule had only one reactive group, i.e., NHS, the gelatin grafted with the NHS-PEG-C₁₈H₃₇ could still be dissolved at 37 °C. While some NHS-PEG-C₁₈H₃₇-grafted gelatin molecules probably remained bound to the released cells through membrane intercalation, they apparently did not exert any significant impact on the morphology and behavior of the cells.

To demonstrate the versatility of production procedure in Figure 2, micropatches composed of a thermoplastic dot and a polyelectrolyte-multilayer pad, named dot-on-pad micropatches, were fabricated based on a method that we previously developed.²⁹ Briefly, the dot-on-pad micropatches were printed on the gelatin sacrificial layer and the top layer of the pads were grafted with NHS-PEG-C₁₈H₃₇. N2a cells were seeded on the micropatches as above and expected to bind to the micropatches through membrane intercalation as schematically illustrated in Figure 6G. Figure 6H shows released N2a-micropatch complexes containing dot-on-pad micropatches. This type of complexes potentially allows enhanced drug release from the dot to the cell bound by the micropatch.²⁹

This work was focused on proving the concept of conjugating the micropatches to living cells with the membrane-intercalation mechanism, so off-the-shelf materials were used to build the micropatches. In principle, the materials can be replaced with clinically approved materials such as PLGA as we demonstrated before.²⁹ Moreover, functional agents can be incorporated into the micropatches. We and Klyachko et al. previously incorporated catalase and Shields et al. recently loaded interferon-γ into the disk-shaped microparticles.^13,14,16 It is conceivable that other therapeutic proteins, polypeptides, or nucleic acids can be loaded into the micropatches by the same approach. It is worth noting that these drug molecules, which will be buried in a multilayer formed by themselves and polyelectrolytes, would be probably inaccessible by the NHS-PEG-C₁₈H₃₇ during the step of grafting NHS-PEG-C₁₈H₃₇ to the micropatches because of its large molecular size. So, the drug molecules will not be damaged by NHS-PEG-C₁₈H₃₇. In addition to the macromolecular drugs, small-molecule drugs can be used with this technique. The dot-on-pad micropatches may be particularly suitable to modulate the behavior of the cells bound by the micropatches because of its potential to enhance drug delivery to the cells bound by the micropatches. For example, retinoic acid may be loaded into the dots to promote neural differentiation of neural precursor cells that are bound by the micropatches.

3.5. Effect of micropatch attachment on metabolic activity of N2a cells

The cell-micropatch complexes developed in this study are intended to be used for treating human diseases via enhancing the therapeutic capability of the cells with the micropatches. Therefore, the cells in the complexes must be alive and able to perform cellular activities required for the intended therapy. On the other hand, the plasma membrane not only is essential to cell survival, but also participates in numerous cellular activities. The micropatch developed in this study bound to the exterior of the plasma membrane via intercalating its dangling octadecyl chains into the membrane for an extended period of time. It is unknown whether this membrane-intercalation-based attachment of the micropatch would affect the cell viability and functionalities. In addition, the cells had been incubated at 8 °C for 30 min in our process of producing the complexes and it has been reported that a 48-h incubation of human embryonic stem cells at 4 °C caused a reduced metabolic activity of the cells measured by the MTT assay.³⁰ Although the exposure of cells to a low temperature was much shorter in our process and the cells in the complexes were apparently alive as indicated by their ability to adhere to and spread on a cell-culture surface (Figure 6D), a quantitative measurement of metabolic activity would provide a more accurate characterization of the technique.

To quantitatively measure metabolic activity of the N2a cells in the complexes, we performed MTT assay on a mixture of micropatch-bound (~60%) and bare (~40%) N2a cells after the cells were cultivated on cell-culture surfaces at 37 °C and 5% CO₂ for 2 d. A pure population of bare N2a cells cultured under normal condition was used a control. The mixture exhibited ~80% of the metabolic activity of the control (Figure 7). To explore whether the 30 min-8 °C incubation required for the production of the cell-micropatch complexes caused the decrease in metabolic activity, another pure population of bare N2a cells had been treated with a 30 min-8 °C incubation. The result (Figure 7) shows the 30 min-8 °C incubation was not responsible for the reduction of metabolic activity. While the exact cause of the reduction is unknown, these results reveal that N2a cells in the cell-micropatch complexes were largely metabolically active.

Figure 7.

Effect of micropatch attachment on metabolic activity of N2a cells evaluated by MTT assay.

3.6. Effect of micropatch attachment on proliferation of N2a cells

Certain cell therapy requires the therapeutic cells to be able to proliferate after administration. Cell division requires generation of new cell membrane and remodeling of the existing cell membrane. For a cell bound by a micropatch, these two processes may be affected because the micropatch-grafted octadecyl chains intercalated into the cell membrane may hinder its fluidity. Moreover, binding of a micropatch to a cell resembles culturing the cell on a solid substrate and it has been found that proliferation of cells can be profoundly affected by the mechanical and topological properties of the substrate.³¹ It is hence reasonable to speculate that micropatch attachment may affect proliferation of the cells bound by the micropatches. To unambiguously determine the effect of micropatch attachment on proliferation, we tracked single micropatch-bound and bare N2a cells. Figure 8A1 shows a single micropatch-bound N2a cell in a 3 mm-diameter well, which contained only one cell, 24 h after seeding. The cell divided into two cells at 48 h (Figure 8A2), four cells at 72 h (Figure 8A3) and nine cells at 120 h (Figure 8A4). The micropatch remained bound to one of the cells throughout this process. Figure 8B1 shows another single N2a cells bound by three micropatches. The cell divided into two cells at 72 h, with all three micropatches being bound to one of the two daughter cells (Figure 8B2). At 120 h, the micropatch-bound cell (Figure 8B2) divided into four cells with two micropatches attached to one cell and one micropatch attached to another cell (Figure 8B3). Corresponding quantitative analysis of 27 single micropatch-bound cells and 29 single bare cells is shown in Figure 8C. Proliferation ratio is defined as the ratio between the number of total cells (micropatch-bound or bare) divided from the initially identified single cells in a sample at a specified time point (48, 72, or 120 h) divided by the number of the single cells (micropatch-bound or bare) at the initial time (24 h). Averages and standard deviations of the proliferation ratios were calculated and no statistically significant differences between the micropatch-bound cells and bare cells was found at every time point of characterization. These results indicate that micropatch attachment did not affect proliferation of N2a cells. This result is not only consistent with the observation that DiI labeling did not impair the proliferation potential of a type of stem cells, but also further reveals that the membrane intercalation of the micropatch-grafted octadecyl chains did not affect proliferation of another type of mammalian cells.²⁰ From perspective of the application, the null effect of the micropatch attachment on cell proliferation is desirable for cell therapies requiring the administered cells to be proliferative.

Figure 8.

Effect of micropatch attachment on proliferation of N2a cells. Overlaid phase-contrast and fluorescence micrographs of (A1–A4) a single N2a cell bound by one micropatch (red) in a well, and (B1–B3) a single N2a cell bound by three micropatches (red) in another well at different time points. (C) Proliferation ratios of micropatch-bound cells and bare cells at different time points.

3.7. Effect of micropatch attachment on differentiation of N2a cells

Some stem-cell-based therapies require the therapeutic cells to be differentiable at the site of transplantation.³² Since differentiation may involve extensive remodeling of the cell membrane and differentiation of stem cells are also affected by the physical properties of the substrate for culturing the cells,³¹ it is again reasonable to speculate that micropatch attachment may affect differentiation of the cells bound by the micropatches Effect of micropatch attachment on the differentiation was studied by culturing a mixture of the micropatch-bound N2a cells and bare N2a cells in a differentiation-promoting medium and the regular medium respectively. A cell having one or more neurites of a length more than twice the diameter of the cell body was identified as a differentiated cell.³³ A typical differentiated cell bound by a micropatch is shown at the center of Figure 9A, which also contained a few non-differentiated bare cells. Figure 9B shows a cluster of five differentiated cells, each of which was bound by one or two micropatches. Numbers of differentiated micropatch-bound cells, total micropatch-bound cells, differentiated bare cells, and total bare cells in each sample were counted and used to calculate parentages of differentiated cells among micropatch-bound and bare cells respectively. Statistical analysis of the result is shown in Figure 9C. Overall, the differentiation percentages increased with time for both micropatch-bound and bare cells in both media, but the rate of increase was much higher for the cells cultured in the differentiation-promoting medium than in the regular medium. More importantly, there was no significant difference in cell differentiation between the micropatch-bound cells and the bare cells in either the differentiation-promoting medium or the regular medium at every time points of characterization, indicating that micropatch attachment did not affect differentiation of N2a cells. Again, this result is consistent with the observation that DiI labeling did not impair the differentiation potential of a type of stem cells and further reveals that the membrane intercalation of the micropatch-grafted octadecyl chains did not affect differentiation of a different type of mammalian cells.²⁰ From perspective of the application, the null effect of the micropatch attachment on cell differentiation is desirable for cell therapies requiring the administered cells to be differentiable.

Figure 9.

Effect of micropatch attachment on differentiation of N2a cells. Overlaid fluorescence micrographs of (A) a differentiated cell bound by a micropatch and (B) a cluster of five differentiated cells bound by micropatches 72 h after being cultured in the differentiation-promoting medium. Micropatches are pointed by white arrows. Micropatches were stained with R18 (red). Cells were stained with calcein AM (green) and Hoechst 33342 (blue). (C) Differentiation percentages of micropatch-bound N2a cells and bare N2a cells in the differentiation-promoting medium and the regular medium respectively at different time points of cultivation.

3.8. Production of cell-micropatch complexes with other types of cells

To explore the applicability of the fabrication procedure illustrated in Figure 2 to other types of cells, hMSCs, bEnd.3 cells, RAW 264.7 cells, and HeLa cells were used to produce the cell-micropatch complexes. hMSC is a primary candidate for cell therapy and has also been used for cell-mediated drug delivery.^34,35,36 bEnd.3 is a mouse brain endothelial cell line that is commonly used to model blood vessels.³⁷ RAW 264.7 is a mouse macrophage-like cell line and widely used to model macrophages.³⁸ HeLa is a human cervical cancer line and one of the most extensively used cancer cell model. For hMSCs, 78% ± 3% of the cells were bound by the micropatches after centrifugal purification (Figure 10A1). The micropatch-bound cells could adhere to and spread on a cell-culture surface (Figure 10A2), indicating the cells were alive and healthy. Similar results were obtained for bEnd.3 (Figures 10B1 and 10B2), RAW 264.7 (Figures 10C1 and 10C2), and HeLa (Figures 10D1 and 10D2) cells. The percentages of the micropatch-bound cells were 67% ± 3%, 59% ± 6%, and 64% ± 1% for bEnd.3, RAW 264.7, and HeLa cells respectively. Overall, cell-micropatch complexes have been prepared with 59–78% yield for five types of cells tested in this study. It is worth noting that the membrane-intercalation mechanism has been used by Nagamune’s group to immobilize human embryonic kidney 293 cells, mouse fibroblast NIH/3T3 cells, human leukemia K562 cells, human T lymphocyte Jurkat cells, human B cell leukemia Daudi cells, mouse embryonic stem cells, mouse myeloid 32D cells, human umbilical vein endothelial cells, human aorta endothelial cells, human aorta smooth muscle cells, human epidermal keratinocytes, normal human dermal fibroblasts, and HeLa cells on a solid surface.^21,22,23 It is thus reasonable to believe that the membrane-intercalation mechanism and the corresponding fabrication technique developed in this study are potentially applicable to a wide range of cells, including the existing and future therapeutic cells.

Figure 10.

Micrographs of cell-micropatch complexes composed of (A1, A2) hMSCs, (B1, B2) bEnd.3 cells, (C1, C2) RAW 264.7 cells, and (D1, D2) HeLa cells. (A1, B1, C1, D1) As-prepared cell-micropatch complexes. (A2, B2, C2, D2) Complexes that were re-seeded on the cell-culture surfaces and had been cultured for 24 h (A2, B2, D2) and 72 h (C2). Micropatches in (A1, A2, B1, B2) were prepared using a stamp carrying 5 μm-diameter pillars. Micropatches were labeled with R18 (red). Cells in (A1, A2, B1, B2, C1, D1) were labeled with calcein AM (green). (A1), (A2), (B1), (B2), (C1), (D2) are overlaid fluorescence micrographs. (C2) and (D2) are overlaid phase-contrast and fluorescence micrographs.

4. Conclusions

Membrane intercalation has been successfully used to conjugate particulate micropatches to living cells. This mechanism does not require covalent modifications of the cells and has, in principle, a wide applicability to many cell types. The conjugation method is featured by the integration of microcontact printing of micropatches, grafting of NHS-PEG-C₁₈H₃₇ to the printed micropatches, and use of gelatin as a temperature-sensitive sacrificial layer for achieving the formation and release of the cell-micropatch complexes. The N2a-micropatch complexes produced by this technique were stable through centrifugation, re-seeding and trypsinization, and the micropatch-bound N2a cells were metabolically active, proliferative and differentiable. In addition to N2a cells, the technique has been successfully applied to hMSCs, bEnd.3 endothelial cells, RAW 264.7 macrophage-like cells, and HeLa cells. This membrane intercalation-based conjugation technique thus promises to be useful for developing new cell therapies enhanced by the micropatches.

ABBREVIATIONS

APTES

(3-aminopropyl)trimethoxysilane

BSA

bovine serum albumin

DCC

N,N’-dicyclohexylcarbodiimide

DCM

dichloromethane

DMAP

4-dimethylaminopyridine

DMEM

Dulbecco’s modified eagle’s medium

DMF

dimethylformamide

FBS

fetal bovine serum

FITC

fluorescein isothiocyanate

hMSCs

human mesenchymal stem cells

HSCs

hematopoietic stem cells

LbL

layer-by-layer

MTT

thiazolyl blue tetrazolium bromide

N2a

neuro-2a

NHS

N-hydroxysuccinimide

PAH

poly(allylamine hydrochloride)

PDMS

poly(dimethyl siloxane)

PEG

poly(ethylene glycol)

PLGA

poly(D,L-lactide-co-glycolide) acid

PNIPAM

poly(N-isopropylacrylamide)

PPMA

poly(n-propyl methacrylate)

PSS

poly(styrene sulfonic acid) sodium salt

R18

octadecyl rhodamine B chloride

THF

tetrahydrofuran

α-MEM

minimum essential medium eagle-alpha modification