Surface-Modified Electrospun Poly(ε-Caprolactone) Scaffold With Improved Optical Transparency and Bioactivity for Damaged Ocular Surface Reconstruction

Abstract

Purpose.: The purpose of this study was to modify and functionalize the surface of synthetic poly-ε-caprolactone (PCL) nanofibrous scaffolds to improve their biocompatibility in order to provide better “cell-substrate” interaction.

Methods.: Poly-ε-caprolactone solution was electrospun and its surface functionality was modified by helium–oxygen (He/O₂) plasma discharge. Scaffolds were characterized for their morphology, wetting ability, mechanical strength, and optical properties by using scanning electron microscopy (SEM), water contact angle measurement, tensile strength, and ultraviolet-visible (UV-Vis) spectrophotometer, respectively. The biocompatibility of nanofibers was explored by culturing human corneal epithelial (HCE-T) cell line. Subsequently, human limbal epithelial cells (LECs) were cultured to evaluate the bioactivity. Cell proliferation was checked by MTT (3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay. Immunofluorescent staining and reverse transcription–polymerase chain reaction were done to check the gene expression; SEM was used to study the morphology.

Results.: Plasma-treated and untreated scaffolds showed almost similar morphology and tensile strength. Water contact angle measurement and optical transparency data showed that the plasma-treated PCL (pPCL) exhibited significantly improved wettability and transparency as compared to the untreated PCL scaffolds. Biocompatibility results indicated that both scaffolds are biocompatible in terms of cell survival and proliferation. However, pPCL showed better cell adhesion and proliferation. Results supported that LEC cultured on pPCL scaffolds had enhanced cell adhesion and proliferation, in comparison to untreated PCL. Gene expression study showed cultures were able to retain their normal phenotype on both scaffolds.

Conclusions.: The hydrophilicity of the surface achieved by plasma treatment effectively enhanced the transparency and promoted the biocompatibility of scaffolds. These nanofibers may act as biological cues for endorsing ocular surface engineering.

Introduction

Dysfunction of limbal epithelial cells (LECs) can lead to limbal stem cell deficiency (LSCD), which is associated with conjunctival ingrowth, inflammation, ocular discomfort, and vision impairment. ¹ Transplantation of ex vivo expanded LECs or oral mucosal epithelial cells has been found to be a promising procedure to treat corneas manifesting LSCD. ^2–5  

One of the obstacles in ocular surface engineering is unavailability of a suitable carrier for the transplantation of LECs. So far, various natural and synthetic materials have been explored. ^6–9 In our previous report, we have proposed electrospun poly-ε-caprolactone (PCL) as an alternative for human amniotic membrane (HAM). ⁹ Poly-ε-caprolactone is being extensively investigated as a scaffold and drug delivery agent in tissue engineering because of its biocompatibility and efficacy. ^10–16 One of the challenges that we have faced in our previous work is the opacity of PCL membranes. Hence, in the present study we addressed this issue by enhancing the transparency of scaffolds by using plasma discharge treatment. Also, the plasma treatment further enhanced the cell adhesion properties of these scaffolds. Surface modification by plasma treatment is a well-established method for modification of surface chemistry without changing its morphology in an ecofriendly way. ¹⁷ Plasma is a partially ionized gas that contains a mixture of ions, electrons, neutral molecules, and free radicals that are able to create active species on a plasma-treated surface. When exposed to air these active species react with O₂ to create hydrophilic functional groups. In this study, electrospun PCL was subjected to helium–oxygen (He/O₂) plasma treatment to impart hydrophilicity, as hydrophilic surfaces can promote cell attachment and growth. 

We also compared the efficacy of plasma-treated PCL (pPCL) with untreated PCL nanofibers by using LECs. To the best of our knowledge, so far there has been no report in the literature that uses pPCL nanofibrous scaffolds for ocular surface engineering. 

Methods

Electrospinning of Scaffolds and Plasma Treatment

Poly-ε-caprolactone solution (10% w/v) was made by dissolving PCL pellets in trifluoroethanol. The solution was electrospun by using an electrospinning setup consisting of dual-polarity high-voltage DC power supply unit (Gamma High Voltage Research, Ormond Beach, FL), a syringe pump (KDS 100; KD Scientific, Holliston, MA), syringe (Dispovan; New Delhi, India), and a needle (24-G) with blunted tip. The positive terminal of the high-voltage supply was connected to the needle tip, while the negative terminal was connected to a metallic collector plate, and a voltage of 15 kV was maintained between them. Electrospun fibers were collected on coverslips kept over the metallic collector plate. Flow rate was maintained at 0.5 mL/h and needle tip to collector distance was maintained at 13 cm. All chemicals were purchased from Sigma-Aldrich (St. Louis, MO). 

Plasma treatment was conducted in an indigenously designed atmospheric pressure glow plasma reactor on plain PCL nanofibrous scaffolds deposited on glass coverslips. These samples were placed between two rectangular aluminum electrodes covered with glass dielectric sheets. Helium–oxygen gas mixture (3:1 ratio) was passed between the electrodes and glow plasma was created at a discharge voltage of 3.5 kV and frequency of 15 kHz. The treatment time was optimized to 2 minutes. The samples were removed by using tweezers and were used for further experiments as formed. 

Characterization of Scaffolds

Surface Morphology of Nanofibers.

Surface morphology of the PCL and pPCL nanofibers was examined by using SEM (Quanta, 200F; FEI, Eindhoven, the Netherlands), at an accelerating voltage of 20 kV, and fiber diameter was analyzed by using image analysis software (ImageJ; National Institutes of Health, Bethesda, MD). 

Contact Angle Measurement.

Hydrophobicity of PCL scaffolds before and after plasma treatment was measured by water contact angle measurement using sessile drop method. A water droplet of 30 μL was placed on the membranes with a microsyringe and an image was taken with a digital camera (EOS 40D SLR; Canon, New Delhi, India). Finally, contact angle was calculated by using image analysis software. The test was performed for six samples and an average value is quoted. 

Tensile Strength Measurement.

Tensile properties of the two samples were measured on thick free-standing membranes of nanofibers electrospun using the same parameters. Both membranes were tested by using Instron 5848 Microtester (High Wycombe, UK), at a crosshead speed of 5 mm/min and load cell of 10-N capacity. Rectangular specimens (membranes) of width of 10 mm, length of 30 mm, and thickness of nearly 30 μm were used for the studies. Six specimens were tested for each sample and their average value is reported. 

Optical Transparency.

Dry PCL and pPCL membranes were kept in PBS for wetting and checked subsequently for optical transparency. While pPCL membranes showed wetting immediately, PCL membranes got completely wet only after 48 hours of dipping in PBS. These wet-treated and untreated samples were compared with wet HAM, dry PCL membrane, and glass coverslip, which was taken as control. Percentage transmittance was measured by UV-Vis spectrophotometer (Model lambda 35; PerkinElmer, Singapore) in the visible range from 350- to 700-nm wavelength. 

Visual assessment of optical transparency of HAM, wet PCL, and wet pPCL was also done by keeping these membranes on a printed text and taking photographs with a digital camera. 

Assessment of Cell Attachment and Proliferation

To compare the effect of scaffold on cell attachment and proliferation, HCE-T cells (Riken cell bank, Ibaraki, Japan) were seeded on plasma-treated and untreated surfaces as a step to study “cell-substrate” interaction. Finally, cells on these scaffolds were analyzed by using SEM at the predetermined time points. Briefly, a total of 2.6 × 10 ⁴ cells/cm² were seeded and cultures were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemented with 5% fetal bovine serum (FBS) and 100 U/mL penicillin/streptomycin, 5 μg/mL insulin, and 10 ng/mL human epidermal growth factor (hEGF) for 6 hours; 1, 3, and 7 days at 37°C. Finally, cells were fixed in 2.5% glutaraldehyde and were further dehydrated by sequential immersion in alcohol series, followed by dehydration with 1,1,1,3,3,3-hexamethyldisilazane (Merck, New Delhi, India). The samples were sputter coated with gold and observed under SEM. 

Isolation and Culture of LECs

Donor Tissue.

A total of 40 human limbal tissue specimens were obtained from donor sclerocorneal rims, with a mean donor age of 32.63 ± 11.20 (range, 6–62) years. The specimens were stored at 4°C in MK medium before using them for cell culture. The average time interval from tissue retrieval to culture was 1.25 ± 2.0 (range, 1–4) days. The study was conducted on approval from the Institutional Ethical Research Committee (All India Institute of Medical Sciences, New Delhi, India), in accordance with the tenets of the Declaration of Helsinki and with appropriate research consent obtained from donors' families. 

Isolation and Culture of Human LECs.

The explant culture was performed according to our previously reported method. ¹⁸ Briefly, the limbal tissue pieces were placed on the surface of preconditioned nanofibers and left for 10 minutes, then medium consisting of DMEM/F12 nutrient mixture (3:1) supplemented with FBS (10%), insulin (5 μg/mL), hydrocortisone (0.5 μg/mL), glutamine (2 mM), hEGF (20 ng/mL), and penicillin/streptomycin (100 U/mL) was added and incubated at 37°C with 5% CO₂. The cultures were maintained for 14 days and the medium was changed twice a week. All cell culture reagents were purchased from Sigma-Aldrich. 

Viability Staining for Cytotoxicity Analysis

Cultures were subjected to viability staining by using Live-Dead Cell Staining Kit (Biovision Research, Milpitas, CA) according to the manufacture's protocol. Cultures were stained with 1 mM Live-Dye (Biovision Research) and 2.5 mg/mL propidium iodide (PI) and viewed under a fluorescence microscope (Nikon, Tokyo, Japan). The dyes had an excitation at 488 nm with an emission at 518 nm and 615 nm for Live-Dye and PI, respectively. A total of 500 cells were counted in five fields and the percentage of viable cells was calculated. 

MTT (3-(4, 5-Dimethylthiazol-2-yl)-2,5-Diphenyltetrazolium Bromide) Staining

Growth kinetics of LECs was measured colorimetrically by MTT assay using standard protocol. Briefly, 2.6 × 10⁴ cells/cm² cells were seeded on scaffolds and glass coverslip. Cultures were stained at 1, 3, 7, 10, and 14 days. All experiments were performed in triplicate and repeated three times. 

Morphologic Analysis

Morphology of LECs cultured for 2 weeks was analyzed by SEM according to the above mentioned protocol. 

Immunofluorescence Staining

Cultures were stained with mouse monoclonal antibody cytokeratin (K) 3/12 (clone AE5; Chemicon, Temecula, CA), ATP-binding cassette subfamily G member 2 (ABCG2, clone 5D3; BD Pharmingen, San Jose, CA) and integrin β1 (clone LM534; Chemicon) at a dilution of 1:100. Cultures were fixed with cold methanol and acetone (3:1 for K3/12) and 2% paraformaldehyde (for integrin β1). Nonspecific sites were blocked by using 2% BSA (Jackson ImmunoResearch Lab, West Grove, PA), followed by incubation with primary antibodies overnight at 4°C. The cells were then incubated with secondary fluorescein isothiocyanate rat anti-mouse IgG1 antibodies (1:500) (BD Pharmingen) for 1 hour and counterstained with 4′,6-diamidino-2-phenylindole (1:1000; Invitrogen, Carlsbad, CA). Finally, cells were mounted with antifade reagent (FluoroGuard; Bio-Rad Laboratories, Hercules, CA) and observed under a fluorescence microscope. 

Reverse Transcription–PCR

Ribonucleic acid was extracted by using a Trizol reagent (Invitrogen) according to the manufacturer's protocol. Complementary deoxyribonucleic acid was synthesized by using a Moloney murine leukemia virus reverse transcriptase enzyme (Promega, Madison, WI). β2 Microglobulin was used as an internal control. The expression of stem cell K19 and ABCG2 and differentiation markers (K3 and K12) was checked by using routine PCR. ⁹  

Results

Scanning Electron Microscopy

The SEM images (Figs. 1A, 1B) of PCL nanofibers before and after plasma treatment showed an average fiber diameter of 140 ± 31 nm and 131 ± 23 nm, respectively. There was no significant change in fiber diameter of PCL after plasma treatment and the surface morphology remained unchanged. Both membranes showed similar porous structure and no flattening or melting of fibers was seen after plasma treatment. 

Figure 1

 

Architecture of electrospun nanofiber scaffold as seen under SEM at ×25,000 magnification. (A) Poly-ε-caprolactone nanofibers with average fiber diameter of 140 ± 32 nm. (B) Plasma-treated PCL nanofibers with average fiber diameter of 131 ± 23 nm. Scale bars: 1 μm. (C) Contact angle measurements on untreated PCL membranes showing water contact angle of 125° ± 4°.

Water Contact Angle

Assessment of surface modification by plasma treatment was done by water contact angle measurement. Poly-ε-caprolactone nanofibers showed water contact angle of 125° ± 4° and the droplet stayed on the surface till it dried (Fig. 1C). After plasma treatment the surface became highly hydrophilic with ∼0° contact angle. The water droplet quickly spread and disappeared within seconds and could not be photographed. 

Mechanical Strength

The average tensile strength of pPCL membrane was 1.99 ± 0.2 MPa and strain was 30.2% ± 1.2%. This was nearly similar to that of untreated PCL membranes, which showed average tensile strength of 1.93 ± 0.18 MPa and strain of 32.06% ± 2.66% (Figs. 2A, 2B). 

Figure 2

 

Stress-strain curve of (A) untreated PCL showing tensile strength of 1.93 ± 0.18 MPa and strain of 32.06% ± 2.66%; and (B) plasma-treated PCL showing tensile strength of 1.99 ± 0.2 MPa and strain of 30.2% ± 1.2%.

Optical Transparency

The transparency of pPCL membrane was compared quantitatively with wet and dry PCL membranes by using UV-Vis spectroscopy. From the UV-Vis spectra (Fig. 3D) it can be observed that glass coverslips (taken as control) showed maximum transmittance of approximately 99%, whereas the dry PCL membrane showed less than 3% transmittance at 700 nm. Wet PCL membrane showed a slightly higher transparency than dry PCL with approximately 13% transmittance throughout the spectral range. However, pPCL showed significantly higher transmittance, of more than 47% at wavelength 700 nm and 36% at wavelength 350 nm, than that of wet untreated PCL. This was very close to that of HAM, which showed transmittance of 77% at a wavelength of 700 nm and 41% at a wavelength of 350 nm. Figures 3A through 3C show the digital photographs of wet HAM, wet pPCL, and wet PCL for visual assessment of transparency. The text under the wet pPCL membrane was more clearly visible than that under untreated wet PCL membrane. The result for wet pPCL was comparable to that for HAM, which showed the best optical transparency. Hence, it may be inferred that pPCL membranes not only show better wetting behavior but also have enhanced optical transparency in the wet state, compared to the wet and dry PCL membranes. 

Figure 3

 

Digital photograph showing optical transparency of scaffolds. (A) Wet amniotic membrane through which printed text is clearly visible. (B) Wet PCL membrane showing translucency through which the printed text is slightly visible. (C) Wet pPCL showing transparency through which the printed text is clearly visible. (D) Ultraviolet-visible spectra for optical transmittance at a wavelength of 700 nm for (A) glass coverslip showing maximum transmittance of 99%; (B) wet amniotic membrane showing transmittance of 77%; (C) wet pPCL showing transmittance of 47%; (D) wet PCL showing transmittance of 13%; and (E) dry PCL showing least transmittance of 3%.

Biocompatibility of Scaffolds

Scanning electron microscopy was used to analyze the effect of surface chemistry on initial cell adhesion and proliferation. The SEM images showed that 6 hours (Fig. 4A) after cell seeding, some of the HCE-T cells adhered firmly to the pPCL surface, while no cells could be located on the PCL surface (Fig. 4E). After day 1 of culture, cells were flat and showed attachment toward both pPCL (Fig. 4B) and PCL (Fig. 4F). However, the cell number was higher on pPCL than untreated PCL. Day 3 SEM images for pPCL (Fig. 4C) and PCL (Fig. 4G) showed that cells were able to establish a firm cell-substrate attachment on both scaffolds by forming a monolayer. On day 7 of culture (Figs. 4D, 4H), cell populations on both pPCL (Fig. 4D) and PCL (Fig. 4H) demonstrated greater proliferation ability with multilayer formation. However, cells on pPCL nanofibers (Fig. 4D) showed more healthy and compact cellular morphology. 

Figure 4

 

Scanning electron microscopy images of HCE-T cells seeded over nanofibers to evaluate the attachment ability at different time intervals for pPCL (A) after 6 hours of cell seeding, (B) after 1 day, (C) after 3 days, and (D) after 7 days; and for PCL (E) after 6 hours of cell seeding, (F) after 1 day, (G) after 3 days, and (H) after 7 days.

Morphology and Activity of LECs on Scaffolds

The cellular morphology of LECs grown over pPCL surface was compared with that of cells cultured over untreated PCL. Figures 5A and 5B show that the epithelial cells grown on pPCL have more organized and uniform cellular morphology than on untreated PCL (Figs. 5C, 5D). Cells over pPCL surface were closely attached to each other with tightly opposed cell junctions and distinct cell borders, whereas cells cultured on PCL appeared to be randomly distributed, did not appear to be well attached to the cells beneath them, and had large intercellular spaces (Fig. 5C). Limbal epithelial cells cultured on pPCL showed presence of numerous short microvilli (Fig. 5B), a feature common to epithelial cells. However, cells grown over PCL surface were flattened, slightly larger, disorganized and had long, distended microvilli as compared to the former surface (Fig. 5D). Taken as a whole, SEM results showed that cells grown on plasma-treated surface had healthy confluent cell sheet firmly attached to the scaffold and formed tight cell to cell contact with abundant surface microvilli, which is most likely due to the higher cell affinity toward the plasma-treated surface. 

Figure 5

 

Scanning electron microscopy images of LECs cultivated on electrospun nanofibers for 2 weeks. (A) Human limbal epithelial cells grown on pPCL surface showing smooth cell sheet and close association with neighborhood cell by forming tightly opposed cell junctions. (B) Apical surface showing numerous short microvilli at higher magnification. (C) Human limbal epithelial cells on PCL surface showing random distribution with larger microvilli. (D) Apical surface of PCL showing distended microvilli.

Limbal epithelial cells cultured on scaffolds were stained with a mixture of Live-Dye, which is a cell permeable dye that gives green fluorescence, and PI, which is a nonpermeable dye that stains dead cells and gives red fluorescence, and were subjected to fluorescence microscopy. Figures 6A and 6B show fluorescence microscopy images of cells seeded on untreated PCL (Fig. 6A) and pPCL (Fig. 6B). Cells that showed red fluorescence (incorporated with PI) were counted as “dead” and those that showed green fluorescence (incorporated with Live-Dye) were counted as “live.” Plasma-treated PCL showed slightly higher cellular viability of 98.7 ± 1.2 as compared to PCL, which was 89.3 ± 1.4. 

Figure 6

 

Viability and cell proliferation staining of LECs cultivated over nanofibers for 2 weeks. (A) LECs cultivated on electrospun PCL nanofibers show confluent viable cell sheet with few dead cells (white arrowhead). (B) Limbal epithelial cells grown over pPCL surface depict high ratio of viable cells as demonstrated by positive green staining. The graph shows MTT analysis of LECs cultivated on electrospun nanofiber scaffolds at different time intervals (days 0, 7, 10, and 14) stained with MTT dye, and their proliferation potential compared with (A) glass coverslips (control), (B) pPCL, and (C) PCL. Data represent three independent experiments and all data points plotted as mean values ± SD (*P < 0.001, **P < 0.05).

MTT Analysis

The MTT assay was conducted to investigate the proliferation of LECs on PCL and pPCL scaffolds. It can be seen from the graph of OD versus days in Figure 6 that pPCL scaffolds were more suitable for LEC growth than untreated PCL scaffolds. A significant level of increase in cell proliferation (P < 0.001) was seen on pPCL surface, indicating that the hydrophilic nature of the pPCL scaffolds better supports cell adhesion and proliferation. 

Molecular Characterization

The phenotype of LECs was examined by using immunofluorescence (K3/12, integrin β1, and ABCG2) and RT-PCR (K3, K12, K19, and ABCG2). Immunopositivity for integrin β1 and ABCG2 was detected over both scaffolds along with the glass coverslip control (Figs. 7A–F). Results showed that cells maintained their differentiated phenotype and expressed K3/12 (Figs. 7G, 7H) on both surfaces, similarly as control (Fig. 7I). Limbal epithelial cells cultivated on both synthetic nanofibers were able to express stemness-related (ABCG2, integrin β1) markers, showing that nanofiber surface chemistry is not inducing complete differentiation and supporting their stemness (Figs. 7D–I). Protein expression patterns of all these markers, in terms of fluorescence distribution and intensity in the cell, showed no difference when compared with control (glass coverslips). These findings were further validated by RT-PCR, and the results confirmed the expression of stem cell markers (ABCG2 and K19) as well as differentiation markers (K12 and K3) (Fig. 8). 

Figure 7

 

Expression of stem/progenitor cells (integrin β1, ABCG2) and differentiation-associated markers (K3/12) in ex vivo expanded human LECs. Immunofluorescence staining shows positive expression of cytokeratin on pPCL nanofibers: (A) integrin β1, (D) ABCG2, (G) K3/12; PCL nanofibers: (B) integrin β1, (E) ABCG2, (H) K3/12; and coverslip (control): (C) integrin β1, (F) ABCG2, (I) K3/12.

Figure 8

 

Reverse transcription–PCR analysis of LECs expanded over nanofibers and control surface (coverslip) for 2 weeks: ABCG2 (379 bp) and K19 (331 bp) represent stem cell–associated markers; K3 (145 bp) and K12 (150 bp) represent differentiation markers; B2M i.e. β microglobulin (210 bp) represents an internal control.

Discussion

A suitable ocular surface substitute must possess a biologically active and transparent surface, which should not induce any immunologic reaction upon transplantation. In recent years, many natural and synthetic matrices have been used to culture LECs ^{6–10,19–21} ; however, most previous studies have used HAM as a choice for carrier owing to its intrinsic properties. ^21,22 Nonetheless, owing to its biological origin, HAM carries inherent risks, such as disease transmission and infection, that cannot be totally avoided. ²⁰ The use of synthetic material can eliminate the risk factors associated with biological materials. Our previous study has shown that electrospun PCL nanofibers have potential as a scaffold for ocular tissue engineering, as they closely biomimic the extracellular matrix of ocular surface. ⁹ This valuable finding signifies a beginning point for establishing an alternative carrier to HAM for the future treatment of damaged ocular surface. The present study was conducted to improve the optical transparency and biocompatibility of PCL scaffolds by plasma treatment. 

Poly-ε-caprolactone nanofibers, fabricated using an electrospinning process, were inherently hydrophobic in nature and possessed poor wetting property. Though PCL is biocompatible, its hydrophobic nature poses difficulty in wettability and cell attachment. Also in our previous study we have found that PCL scaffolds remain opaque even after dipping in water for more than 48 hours. Hence, the surface property of PCL membranes was improved by atmospheric pressure plasma treatment, and a potential substrate with desired properties was created for ocular surface regeneration. Plasma treatment is widely used as a tool to efficiently modify the surface properties of polymeric materials by introducing desired functionalities for biomedical applications. ^23,24 It is a convenient and cost-effective method for improving the hydrophilic properties and permeability of polymer surfaces. Plasma-treated PCL nanofibers resulted in high hydrophilicity with zero water contact angle, leading to better adhesion and proliferation of epithelial cells as confirmed by SEM and MTT assay. Plasma is a mixture of ionized gas containing reactive species such as ions, free radicals, and molecules that react with the surface and create active sites, which in turn react with oxygen to form hydrophilic groups when exposed to atmosphere, thereby imparting hydrophilicity on the surfaces of the scaffolds. In this study both helium and oxygen were used simultaneously inside the plasma reactor to facilitate generation of high number of hydrophilic groups. Optical transparency is an important property that should be considered while developing a bioengineered ocular surface construct. A healthy ocular surface is required for clear vision and it contributes two-thirds of the total refractive power of the eye, which is the most remarkable property of the ocular surface. For this reason, an ocular surface equivalent fabricated by tissue engineering should be able to transmit most of the visible light, mimicking the natural behavior of the native ocular surface. Poly-ε-caprolactone nanofibrous membrane in its dry form is opaque, as almost all the light falling on its surface gets scattered as the light travels from air (refractive index [RI] ∼1.00) to PCL (RI ∼1.41). On dipping this hydrophobic membrane in water, the latter forms high contact angle with the fiber surface and air gets entrapped between the droplets, thereby forming a discontinuous layer of water over the PCL surface. When light falls on media with variable RI, it gets scattered again, with only a small part being able to transmit through the sample. On the other hand, He/O₂ pPCL surface is hydrophilic, which allows water to form a uniform layer around the fibers as well as to fill up the pores in between the fibers. Since the RI of PCL and water are fairly similar, that is, 1.41 and 1.33, respectively, a homogeneous medium is formed for light to travel through the sample, thereby increasing the transparency of the pPCL membranes (Figs. 9A–C). Our studies have shown that plasma-treated PCL can efficiently enhance the optical properties of the scaffolds. The results indicated that plasma-treated stromal equivalent can transmit 37% more light than the untreated plasma stromal equivalent. 

Figure 9

 

Schematic representation for optical transparency: (A) dry PCL, (B) untreated wet PCL, and (C) plasma-treated wet PCL.

The in vitro biocompatibility of scaffold, assessed by HCE-T cell line, confirmed that the epithelial cells showed a greater affinity toward plasma-treated surface than untreated surface. This can be explained by the decreased hydrophobicity of pPCL membranes. The outer surface of cell membrane containing hydrophilic amino acids and the hydrophilic surface of He/O₂ pPCL show increased affinity toward each other, resulting in a high cell-substrate interaction, and demonstrate better proliferation than the untreated PCL surface. Raechelle et al. ²⁵ have modified the surface region of poly(methyl methacrylate) by plasma modification, with the aim to increase cellular response of human lens epithelial cells, while Notara et al. ²⁶ have demonstrated that plasma treatment can be used as a substrates for serum-free expansion of LECs. 

Our previous study has shown that our in-house fabricated scaffold has promising architecture to satisfy the requirements of cell growth and function. Our data demonstrated that LECs cultured on the membranes remained viable for up to 2 weeks without altering their phenotype and had the ability to penetrate inside the three-dimensional architecture of the scaffold. ¹⁰ The pore size and porosity are optimum for spreading and holding of cells. It conveniently accommodates a large number of cells by uniform distribution through the interconnected pores to facilitate optimum gas and nutrient diffusion. Scanning electron microscopy depicted healthy morphology of LECs on both scaffolds. However, cells cultivated on pPCL surface had typical corneal epithelial polygonal cell morphology with numerous short microvilli all over the apical surface. Marker expression studies also confirmed that cells were able to maintain their normal phenotype by expressing differentiation and stem cell markers. MTT results showed gradual increase in cell proliferation on both surfaces. It was speculated that LECs were able to interact and integrate well with the surrounding fibers in pPCL nanofibrous as compared to untreated PCL scaffolds. This finding has confirmed that surface topography plays a pivotal role with respect to cell attachment, survival, and proliferation. 

Conclusions

Surface modification of electrospun PCL scaffolds by atmospheric pressure plasma discharge is an efficient way to enhance biocompatibility and proliferation. In addition to this, the hydrophilicity of the PCL scaffolds developed owing to plasma treatment made them optically transparent in wet conditions. Highly porous nanofibrous scaffolds with excellent architecture, biocompatibility, and transparency that can mimic the natural extracellular matrix, as well as assist in maintaining a normal LEC phenotype, were successfully produced. These can be appropriately used as an alternative of HAM for repairing damaged ocular surface. 

Acknowledgments

Supported by a financial grant provided by the Department of Biotechnology (DBT), India. Senior Research Fellowship (SS) from Indian Council of Medical Research (ICMR), India, is gratefully acknowledged. 

Disclosure: S. Sharma, None; D. Gupta, None; S. Mohanty, None; M. Jassal, None; A.K. Agrawal, None; R. Tandon, None 

Footnotes

 RT and AKA contributed equally to the work presented here and should therefore be regarded as equivalent authors.