Synthesis and Characterization of a Novel Peptide Targeting Human Tenon Fibroblast Cells To Modulate Fibrosis: An Integrated Empirical Approach

Abstract

Fibrosis is the primary factor influencing the prognosis of glaucoma post-trabeculectomy surgery, an eye condition characterized by increased intraocular pressure (IOP). Despite advancements in surgical procedures and aftercare, it continues to be a serious impediment. During the clinical intervention of scarring, fibrosis is managed by using topical application of combined antifibrotic drugs (mitomycin C). But still, scarring remains a key problem due to minimal drug penetration and nonbioavailability. In this study, we synthesized a cell-specific peptide for modulating scarring in human tenon fibroblasts undergoing epithelial–mesenchymal transition (EMT). The peptide was also conjugated with mitomycin C in order to investigate the effect of the drug conjugation on human tenon fibroblasts from the nanofiber composite system and to evaluate the fibrosis process. Peptide VRF2019 was identified using a subtractive proteomics approach, including solubility, cell penetration, and amphipathic properties. The peptide structure was determined using circular dichroism spectroscopy. The peptide and drug was conjugated using N-ethyl-N′-(3-(dimethylamino)propyl) carbodiimide/N-hydroxysuccinimide (EDC-NHS) chemistry, and the conjugation efficiency was evaluated using high-pressure liquid chromatography. The conjugated product and polycaprolactone (PCL) were electrospun to form a composite nanofiber, which was tested for cytotoxicity and drug release on human tenon fibroblast cells. The modeled VRF2019 peptide structure formed an α-helical structure with all residues spanning the allowed regions of the Ramachandran plot. Subsequent molecular dynamics simulations also demonstrated its membrane penetration potential. The peptide uptake was also studied in human tenon fibroblast cells. High-pressure liquid chromatography (HPLC) and mass spectrometry measurements confirmed peptide–drug conjugation and stability. Furthermore, scanning electron microscopy (SEM) investigation revealed the structure and size of the PCL composite nanofiber. We infer from early research that the PCL composite nanofiber matrix can greatly increase drug delivery and bioavailability.

Glaucoma is an eye condition that leads to permanent visual impairment. It is a category of optic neuropathies characterized by retinal ganglion cell degeneration. The significant loss of ganglion cells occurs as a result of the increased intraocular pressure caused inside the eye. Excess aqueous fluid is drained from the eye via the trabecular meshwork and uveoscleral outflow to maintain intraocular pressure. If any of these fail to function properly, it causes an increase in intraocular pressure.¹ The only possible solution is targeted at reducing intraocular pressure.² Early diagnosis paves the way to reduce the pressure with medications like prostaglandin analogues, carbonic anhydrases, and β-adrenergic blockers.³ At the severe stage, medical treatment does not help to reduce adequate pressure. Hence, laser or incisional surgeries are performed.⁴ The most frequent incisional surgery is trabeculectomy, which includes the removal of a small part of the trabecular meshwork to drain the excess aqueous fluid trapped inside the eye. In some cases, if trabeculectomy fails to control pressure, glaucoma-draining implants or devices are implanted. Despite advances in surgical techniques and postoperative care, fibrosis remains a major impediment. Management of fibrosis is done through the application of topical or a combination of antifibrotic drugs at the time of post-surgery to modulate scarring.

The most commonly used agents are cytotoxic antimetabolites such as 5-fluorouracil and mitomycin C (MMC).⁵ Mitomycin C is an alkylating agent that inhibits DNA synthesis in human cells. Because of this property, MMC has become the gold standard drug for glaucoma infiltration surgery (GFS) procedures.^1,6,7 In some cases, GFS fails and patients need to restart the medication or have surgery. Due to the poor penetration and bioavailability of drugs, scarring of tissue happens. Excessive scarring leads to failure of the filtering bleb after glaucoma filtration surgery.⁸ Therefore, there is a need to manage fibrosis through different medications and bioavailability. To influence the efficacy of MMC in the target area of the tissue, the drug must be delivered at the time of wound closure. To do so, cell-penetrating peptides (CPPs) as carrier molecules for targeting the human tenon fibroblast are very crucial. CPPs have gained popularity because they act as delivery vehicles.⁹ CPPs have been used to deliver small molecules, DNA, and interfering RNA.¹⁰ CPPs can be conjugated with various drugs and nucleic acids using linkers to form peptide–drug complexes. Recent advancement has been drawn to noncovalent drug delivery due to its easy preparation protocols involving creating noncovalent bonds, like electrostatic and hydrophobic bonds, than covalent bonds, like amide bonds, which also produce drug-delivery systems.¹¹

The conjugate CPPs generally enter the cells via direct translocation or through endocytosis. Some of the peptides like AG30, AH90, CW49, Esculentin-1a, etc. have therapeutic potential.¹²⁻¹⁵ One of the major concerns is CPP’s stability, which can be achieved by making a nanocomposite through electrospinning.¹⁶ Nanofibers are well-known for their good absorption and release properties.¹⁷ They have achieved the unique features of a localized drug-delivery system, including microscale or nanoscale diameters with a similar structure to the extracellular matrix, controllable surface morphology, very high surface area, porosity, drug loading capacity, etc.^18,19 Poly(ε-caprolactone) (PCL) is an FDA-approved polymer and well-established delivery molecule, particularly in the ocular system because of its nontoxicity, biocompatibility, and biodegradability. It can be easily fabricated into nanofibers and nanoparticles.²⁰ However, there is very limited study on therapeutic peptide-based drug delivery to modulate the wound-healing process, especially in post glaucoma surgery. As a result, with guidance from human eye proteomics, as reported by Semba et al.,²¹ we developed human tenon fibroblast cell-specific CPPs (VRF2019 and penetratin) that can penetrate cells undergoing EMT to modulate the scarring process. We were able to identify two CPPs (one of them was found to be a reported peptide) using a subtractive proteomic approach combined with in silico approaches. Biophysical characteristics and cellular uptake were tested in experimental validations. Peptides were then conjugated with mitomycin C and electrospun to create a nanofiber composite matrix that may be used as a drug carrier to improve the drug conjugate availability. We have designed peptides that penetrate the cells that undergo EMT transition.

Materials and Methods

Design of Cell-Penetrating Peptides

The human epithelial–mesenchymal transition (EMT) genes (377) pertaining to cancer conditions were retrieved from dbEMT (http://dbemt.bioinfo-minzhao.org/index.html).²² All of the genes retrieved were mapped to their corresponding Uniprot IDs based on their HGNC (HUGO Gene Nomenclature Committee) identifiers. Only 357 of the 377 genes were successfully mapped to UniProtKB IDs, with the remainder being microRNA coding genes that were not considered. All 357 protein-coding sequences were obtained in FASTA format from the Uniprot database for further study.

To design a human EMT-based peptide with cell-penetrating properties, the EMT proteins were subjected to BlastP analysis against the 731 nonredundant cell-penetrating peptides (CPPs) data set, among 1845 from the CPPsite database (https://webs.iiitd.edu.in/raghava/cppsite/information.php/).²³ The BlastP analysis was performed with default parameters using CLC Genomics Workbench 0.1 to identify the conserved sequence between the EMT and the CPP. The resulting conserved matches of EMT sequences of >60% sequence identity matched with CPPs, and lengths of 6–16 were considered for further analysis. The resultant EMT-based peptides were subjected to physicochemical property analysis such as solubility, charge, hydrophobicity, amphipathicity, hydrophilicity, charge, and molecular weight, followed by cell-penetrating property analysis using online servers, Innovagen peptide solubility calculator (https://pepcalc.com/peptide-solubility-calculator.php), and CellPPD²⁴ (https://webs.iiitd.edu.in/raghava/cellppd/algo.php/), respectively. Moreover, the secondary structure–property and its amphipathic property were predicted using AMPHIPASEEK²⁵ (https://npsa-prabi.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_amphipaseek.html), and the toxicity of peptides was also predicted by the ToxinPredtool²⁶ (https://webs.iiitd.edu.in/raghava/toxinpred//). Among these peptides, the ones with an SVM score >1.0, stable secondary structure elements, and polyarginine and polylysine at the N-terminal were prioritized for further screening. Later, the potential CPPs with desirable properties were scrambled randomly with the Scrambled-Peptide generation server. For the randomly generated peptides, amphipathic properties and cell-penetrating properties were validated (Figure 1).

Figure 1.

Schematic representation of peptide design strategy based on in silico methods.

Molecular Modeling and Molecular Dynamics Studies of Designed Peptides

For the peptides that surpass all validation criteria, the three-dimensional structure was obtained from the PDB database; for nonhomologous peptides the 3D structures were predicted using I-TASSER,²⁷ and the model with a high C-score value (signifies a model with higher confidence) was chosen as the best model. Subsequently, these peptides were further subjected to molecular dynamics (MD) simulation studies to determine the structural stability of the modeled peptides. Prior to simulation, the peptides were preprocessed and optimized using the Schrö̈dinger protein preparation wizard. The preprocessed peptides were subjected to structural stability analysis using molecular dynamics simulation using the GROMACS 5.1.4 package²⁸ with CHARMM36m set as a force field. The peptides were solvated using a SPC (single point charge) water model in a cubic box with a dimension extending to 0.1 nm as a periodic boundary condition in all directions from the protein molecules, and the charge of both systems was neutralized by adding appropriate counterions. Further, the system was minimized using the steepest descent algorithm followed by the conjugate gradient method through 50 000 steps and was equilibrated using both NVT and NPT methodology for 100 ps. The LINCS algorithm was used to limit all bound interactions, while the SETTLE algorithm was employed to constrain the shape of the water molecules. The NVT equilibration was carried out at a constant temperature of 300 K using the V-rescale weak coupling method and the Parrinello–Rahman method for the NPT equilibration that was set at 1 atm, respectively. Finally, the well-equilibrated system was subjected to MD simulation for 300 ns, and the structural conformations were sampled every 2 ps. Furthermore, from the MD trajectory, the root mean square deviation (RMSD) and root mean square fluctuation (RMSF) of the peptides were plotted using XMGRACE software.

In the case of membrane simulation, the peptides were positioned and solvated in a 1-palmitoyl-2-oleoylphosphatidylcholine (POPC) membrane bilayer using OPM and CHARMM-GUI²⁹ software packages. Further solvation, neutralization, and equilibration of the system at NVT and NPT conditions were carried out similarly in the GROMACS 5.1.4 package, as mentioned earlier. The equilibrated systems were further subjected to MD simulation for 300 ns and their respective trajectories were analyzed as mentioned earlier.

Building the Peptide–Mitomycin C Complex

Initially, the mitomycin C (CID-5746) retrieved from the PubChem database³⁰ was prepared through the LigPrep module. Subsequently, the minimum potential structures of peptides (penetratin and Random_pep2 (VRF2019)) were linked to prepare the mitomycin C at its C-terminal using the build structure module of Chimera,³¹ wherein the free carboxyl of the last amino acid of each peptide was linked to the free amino group of the mitomycin C moiety. Further, the linked peptide–mitomycin C complex was minimized by 200 steps of steepest descent followed by 100 steps of conjugate gradient using Chimera. Furthermore, the peptide–mitomycin C complexes were preprocessed using the Protein Preparation Wizard of Schrödinger suite³² to optimize the stereochemistry, and the energy was minimized using the optimized potentials for liquid simulations (OPLS3) force field. The resultant minimized complexes were considered for further membrane simulations, as mentioned earlier.

Peptide Synthesis

Peptide VRF2019 (KKQIRFWRKQWNIMR) and penetratin were synthesized using solid-state synthesis and procured from M/s. GenScript (https://www.genscript.com/) with an HPLC purity of >95%. The N terminal of the peptide was linked with the fluorescein isothiocyanate (FITC) label.

Peptide Secondary Structure and Stability Analysis

Far-ultraviolet (UV) circular dichroism (CD) spectra of peptides, VRF2019, and penetratin (500 nM) were recorded on a spectropolarimeter (J810; Jasco International Co., Ltd., Tokyo, Japan) using a 0.1 cm path length quartz cuvette at 37 °C. Spectra of VRF2019 and penetratin were recorded from 260 to 190 nm at 37 °C at a scan rate of 50 nm/min. The CD measurements were recorded in triplicate. Data was analyzed using the online tool BESTSEL (https://bestsel.elte.hu/index.php).

The stability of the peptide was analyzed through HPLC. Peptides stored at 4 and 37 °C for a month were checked for their stability using HPLC. Further, the peptide of 1 μM concentration was run on 15% native PAGE to validate the stability. The aggregation of peptide VRF2019 at pH 2, 4, 7, and 10.5 was further analyzed using Zetasizer Nano ZS (Malvern Instrument, Malvern, U.K.) equipment with a 4 mW He–Ne laser operating at 633 nm. A dip cell cuvette with a 1 cm path light was used. The intensity of scattering light was measured at a 173° angle to assess surface potential.

Ethical Approval

The primary human tenon fibroblast (HTF) cells used in the study were collected from patients who underwent glaucoma infiltration surgery after obtaining informed signed consent from the patient. The study was approved and reviewed by the local ethics committee at the Vision Research Foundation (Ethics no.: 635-2017-P), Chennai, India, and the committee deemed that it conformed to the principles of research in accordance with the Declaration of Helsinki.

Establishing Primary Human Tenon Fibroblast and Myofibroblast Cells

The explant samples were collected in sterile tubes from the patient post infiltration surgery. The tissues were washed with phosphate-buffered saline (1× PBS) thrice to remove red blood cells (RBCs). Further, the tissues were carefully explanted on the Petri dish coated with collagen type 1 matrix. DMEM/F12 (Gibco) medium with 20% fetal bovine serum (FBS) (Gibco, heat-inactivated) and 1% antibiotics (Antibiotics-Antimycotic, Gibco) were added to the explant and incubated for 48 h. The established human tenon fibroblast was passaged and maintained in the same medium. All of the experiments were conducted with passages 3 and 4. For myofibroblast differentiation, passage 2 was subcultured in DMEM/F-12 medium with 20% FBS (Gibco). Once the cells were 80% confluent, they were serum-starved with DMEM/F-12 with 5% FBS overnight. The next day, the flask was treated with human recombinant TGF-ß1 (Sigma, cat. no. SRP3170), 5 ng/mL, and allowed to incubate for 48 h.

Characterization of Fibroblasts and Myofibroblasts

HTF cells were allowed to grow on the coverslip for 24 h. After being fixed with 4% paraformaldehyde and washed with phosphate-buffered saline, the cells were permeabilized with 0.5% Triton X-100 and washed with PBS. The cells were then blocked with 1% bovine serum albumin (BSA) prior to overnight incubation with primary antibodies vimentin, pan-keratin, and alpha-smooth muscle actin (alpha-SMA). The detection was done using Cy3.5 and FITC secondary antibodies and counterstained with Hoechst (Thermo Fisher, cat. no. 33342). The cells were incubated for 45 min in dark and imaged using a fluorescent microscope (Axio Observer, Zeiss GmbH, Germany).

Cells were extracted and lysed with RIPA buffer and then separated on sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) system at 100 V in electrophoresis buffer (25 mM Tris, 190 mM glycine, and 0.1% SDS). Using a semidry transblot apparatus (Hoefer, Holliston, MA, U.S.A.), the separated proteins were transferred to a nitrocellulose membrane (Amersham Protran Premium, cat. no. GE10600002, 0.45 M pore size) at 1.50 mA/cm². Post-transfer the membrane was blocked with 5% BSA in 1× TBST (20 mM Tris HCl pH 7.5, 150 mM NaCl, and 0.1% Tween 20). Further, the membrane was washed thrice with 1× TBST and incubated with vimentin (CST, cat. no. D21H3) and alpha SMA (Sigma, cat. no. A2547) antibodies at 4 °C overnight. Each antibody was 500-fold diluted in 5% (w/v) BSA (Hi-Media, Mumbai, India) solution.

After overnight incubation, the membrane was washed thrice for 5 min with TBST, further incubated in the corresponding HRP-conjugated antirabbit secondary antibody (diluted to 5 000-fold in TBST), washed, and detected for HRP activity using HRP substrate (cat. no. 1705061; Bio-Rad Laboratories, Inc., Hercules, CA, U.S.A.) in Bio-Rad Gel Documentation system (Protein Simple, Biorad Hercules, CA, U.S.A.).

Peptide-uptake experiments were performed after the characterization of the HTF and myofibroblast cells. Peptides were dissolved at a stock concentration of 1 mM in distilled water, and 10 μM VRF2019 was incubated with the cells to study the intracellular localization. After 1 and 24 h of incubation, the cells were washed with PBS and visualized through fluorescence microscopy using an FITC filter. NCC-RbC-51-Riken (retinoblastoma tumor cells) were used as a control for the experiments. The cytoskeletal reorganization was captured by phalloidin staining (Cytoskeleton).

Cytotoxicity of the Peptides against Tenon Fibroblast Cells

To determine the vitality of the cells, the MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay was carried out in accordance with the manufacturer’s instructions (HiMedia Laboratories). Human tenon fibroblast cells were seeded in a 96-well plate with 2 000 cells per well. Peptides of various concentrations (10–100 μM) were added to each well and incubated for 24 h. Post-treatment MTT reagent was added to the cells and incubated for 2 h at 37 °C. The resulting formazan precipitate was solubilized with DMSO (dimethyl sulfoxide), and the absorbance was measured at 570 nm. Each sample was analyzed in triplicate, and an average value was taken for plotting the graph. Cell viability was determined for each peptide by taking the absorbance reading of treated cells corrected with the readings of cells treated with the medium controls. The percentage of cell survival in each well was calculated by subtracting untreated cells as 100%. The dose–response curves and the percentage of cell survival versus peptide concentration were determined.

Preparation of Cell-Penetrating Peptide–Drug Conjugate

The conjugates were made using standard EDC/NHS chemistry for covalent amide bond conjugation between peptides and the mitomycin C drug. The carboxyl group of peptides was activated through EDC/NHS and an amine-containing drug molecule reacted with the carboxyl-activated peptide. An equal concentration of the drug was allowed to react with the peptides in a 1:1 ratio of EDC/NHS. The reaction was incubated for 2 h, and the conjugates were used as such for further analysis.

Conjugate Characterization

The conjugates were identified using an amide bond between the peptide and the drug using Fourier transform infrared spectroscopy (FTIR) using the potassium bromide (KBr) pellet method. High-pressure liquid chromatography (HPLC) was employed to characterize the conjugates, and the C18 column was used to identify the conjugates with gradient elution of 80% acetonitrile and 0.1% trifluoroacetic acid. The column was kept at 26 °C, and the mobile phase was injected at a flow rate of 1.0 mL min^–1. The detection of eluent was carried out at a photodiode array detector (220 nm) with a total run time of 30 min. Purified component retention times were used to identify the peaks. Peptide standard graphs were obtained using the standard plot of peptide concentration versus area under the curve. The calibration curve ranged from 25 to 100 nanomoles for all of the analytes. Inter- and intraday precision, stability studies, and sample reanalysis were investigated for all of the analytes. The mass of the conjugate was also determined using electrospray ionization mass spectrometry (ESI-MS).

Preparation of PCL Nanofiber Composite

The PCL polymer (Sigma-Aldrich, molecular weight (MW) of 80 000 g/mol) was dissolved in nine parts of hexafluoroisopropanol (HFIP) (Sigma, cat. no. 440471) and one part of water containing 10 μM peptide or conjugate at room temperature for 24 h. The electrospinning process was carried out using the M/S ESPIN Nanotech instrument. About 5 mL of the solution was loaded into a plastic syringe, and the syringe was placed on a syringe pump. The electric potential was applied to the metallic needle (21 gauge) by the high voltage power. The fibers were gathered on a flat copper plate coated with aluminum foil. All of the experiments were conducted at room temperature and relative humidity within a closed chamber. The applied voltage was 10 kV, the solution flow rate was 10 μL/min, the tip of the needle to collector distance was 12 cm, and the duration of each experiment was 2 h. All of the experiments were conducted under the same set of conditions.

Fiber Morphology Characterization

Scanning electron microscopy was used for the characterization of the surface morphology of the nanofibers. All of the samples (PCL-VRF2019, PCL-MMC, PCL-conjugate, and 5% PCL) were sputter-coated with gold prior to their observation under the microscope (SNE-3200M) by SEC Korea. The samples were measured by the nano eye software. At least 200 different fiber diameters were analyzed for their dimension distribution per sample in order to ensure the accuracy of the measurements. The histogram of the fiber diameter was calculated for all of the samples, and graphs were created using Origin software.

Determination of Surface Topography

The surface topography of the nanofibers was determined using the nitrogen adsorption isotherm method. The surface analysis method developed by Brunauer, Emmett, and Teller (BET) was used to examine the nanofiber composites. Analysis of the nitrogen adsorption–desorption isotherm at 77.35 K was done using Quantactrome ASiQwin. The pore size distribution of nanofibers was investigated using a surface analyzer (Quantactrome Instruments version 5.0). Samples of PCL, PCL-conjugate, PCL-peptide, and PCL–drug composite nanofiber mat were placed on the machine and degassed at 200 °C with helium for 3 h, and then nitrogen was purged in order to investigate the porosity and surface area of PCL and PCL-conjugate nanofibers. We computed isotherms using the de Bout T-method. The graphs for the isotherm and pore size distribution were plotted using Origin 8.5.

Conjugate Characterization, Release, and Stability

The electrospun fiber mat was cut into a 2 × 2 cm area and placed in 1 mL of 1× PBS to obtain the release of conjugate at various time points: 30 min, 1, 2, 3, 4, 8, 16, and 24 h. The released solution was collected and injected into the HPLC column to estimate the relative concentration of conjugate release. The obtained area under the curve was correlated with the known concentration of conjugate present in the total surface area of the fiber deposited on the foil. The concentration of conjugate released was calculated using the standard plot of the peptide at 220 nm through HPLC and plotted in Prism 8.0 software. The released solution was subjected to CD spectroscopy and FTIR to check conjugate stability. Further, the released solution was syringe filtered and used for treating the HTF and myofibroblast to check the cytotoxicity by MTT assay.

RNA Extraction and Quantitative Polymerase Chain Reaction

The primary HTF cells were treated with 10 μM VRF2019, 10 μM conjugate, and 250 μM MMC for 1 h, and total RNA was extracted using the RNeasy Micro (QIAGEN) kit according to the manufacturer’s protocol. The total RNA concentration was determined using a NanoDrop Microvolume UV–vis spectrophotometer (Thermo Fisher Scientific). cDNA was synthesized using an Applied Bioscience High Capacity cDNA conversion kit (cat. no. 4368814). Real-time quantitative polymerase chain reaction (qPCR) was performed using an Applied Biosystems 7300 with SYBR green chemistry (Applied Biosystems). qPCR reactions were prepared using TB green Mastermix (TaKaRa). Two-way ANOVA was applied for statistical analysis. The primer details are mentioned in Table S6.

Conjugate Cellular Localization and Immunofluorescence

Primary HTF cells were treated with 10 μM conjugate for 1 h. The effect of conjugate treatment and cellular localization of TGF-β1 markers was studied by fixing the cells with 4% paraformaldehyde and permeabilizing with 0.5% Triton X-100, followed by washing with 1× PBS. Blocking was performed using 1% bovine serum albumin (BSA) for 45 min followed by overnight incubation with primary TGF-β1 antibody (Santa Cruz, cat. no. sc-146). The detection was done using Cy3.5 secondary antibody and counterstained with Hoechst (Thermo Fisher, cat. no. 33342). The imaging was carried out with Carl Zeiss Axio Observer Fluorescent Microscopy (Carl Zeiss, ZmBH, Germany).

Statistical Analysis

All of the data were presented as mean and SD in at least three independent experiments. GraphPad Prism 8 software was used for statistical analysis. The significance was defined by a P-value of <0.05.

Results and Discussion

Design of Cell-Penetrating Peptides

Among 357 EMT protein sequences, only 264 of lengths 6–16 had >60% sequence identity with CPP peptides. Furthermore, the physicochemical property analysis resulted in only 98 peptides with good solubility, the cell-penetrating property that surpassed the filtration criteria. Finally, based on the SVM (support vector machine) score (>1.0), a high prevalence of helical conformation with N-terminal arginine and lysine showcasing the cell-penetrating property as well as amphipathic properties was considered, and thereby only 2 peptides (Pep1 and Pep2) were prioritized. Among these prioritized peptides, Pep2 was highly homologous to the well-proven CPP peptide penetratin. Further, among the random peptides generated based on Pep1 and Pep2, the random peptides with significant CPP properties and good solubility were considered for further analysis (Table S1).

Molecular Modeling and Molecular Dynamics Simulation Studies of Designed Peptides

The three-dimensional structure of Pep2 (penetratin) was obtained from the crystal structure (PDB ID: 2ND6), while the structures of other peptides surpassing all of the validation criteria were predicted by the I-TASSER server; the models with the highest C-score values were chosen as the best models for each peptide. Moreover, the predicted three-dimensional structures were validated for stereochemical properties using a PROCHECK generated Ramachandran plot, which showed all residues in the allowed region (Figure 2). Finally, their structural stability was studied further by molecular dynamics simulation studies. On the basis of the molecular dynamics simulation trajectory analysis (Figure 3), it could be inferred that Pep1, Pep2, and Random_pep1 show higher deviations around 0.7 nm and evolve to converge after 200 ns, as contrasted with Random_pep2, which has deviations around 0.6 nm. Moreover, among these peptides, only Pep2 and Random_pep2 were able to maintain their secondary structure stability, while Random_pep1 showed transition and loss of secondary structures during the last 300 ns of the production run. Therefore, on the basis of all of the analyses, it could be inferred that Pep2 (penetratin) and Random_pep2 (VRF2019) peptides have higher structural stability than other peptides.

Figure 2.

Three-dimensional structures of peptides and their corresponding Ramachandran plots: (A) Pep1, (B) Pep2 (penetratin), (C) Random_pep1, and (D) Random_pep2.

Figure 3.

Molecular dynamics simulation analysis of peptides Pep1, Pep2, Random _pep1, and Random_pep2: (A) RMSD, (B) RMSF, and (C) secondary structures of Pep1, Pep2, Random _pep1, and Random_pep2.

Characterization of the Peptide Using CD Spectroscopy and Zetasizer

Peptide uptake depends on the secondary structure and stability, which are influenced by temperature and pH.³³ The amount of net charge and aggregation are very crucial for peptide stability, solubility, and membrane penetration. Hence, we studied the secondary structure of peptide VRF2019 through CD spectroscopy at 37 °C. Both peptides showed a high intensity of the negative band at 202 and 208 nm and penetratin at 202 nm (Figure 4A). Penetratin peptide CD spectra agreed with a previous report.³⁴ Further, the peptide VRF2019 was characterized by native PAGE and dynamic light scattering for the aggregation property. VRF2019 showed a single band in native PAGE (Figure 4B). The predicted molecular weight of the peptide was also validated using 16% SDS-PAGE and found to be 2.3 kDa. The zeta potential of the peptide was −16.9 at pH 7.4, whereas at pH 4.5 and pH 10.2 the zeta potentials were found to be −8.79 and −8.38, respectively. The average size of the VRF2019 aggregation was not maintained throughout the pH range. Interestingly, the size of the aggregation increased at pH 10.2. The average sizes of the aggregation at different pH values (7.4, 4.5, and 10.2) were found to be 257, 285, and 1124 nm, respectively (Table S2). Additionally, membrane modeling was used to estimate membrane penetration, which was then supported by cellular uptake in HTF cells.

Figure 4.

Biophysical characterization of peptide: (A) CD spectra of penetratin and VRF2019 and (B) native PAGE of VRF2019 in triplicate.

Membrane Molecular Dynamics Simulation of the Validated Peptides

The average thicknesses of the lipid bilayer in the presence of Pep2 (penetratin) and Random_pep2 (VRF2019) are about 39.362 and 39.261 Å, respectively (Figure 5A and B and Table S3). On the basis of these values, it is noted that mild variability is shown for the lipid bilayer thickness in the presence of the Random_pep2 peptide. The area per lipid in the system is one of the parameters to quantify the disruption of membrane structures in the presence of peptides. In our study, the areas per lipid in the presence of peptides Pep2 (penetratin) and Random_pep2 (VRF2019) along the bilayer are 63.919 and 64.091 Ų, respectively (Figure 5C and D and Table S3). This indicates that the presence of both peptides in the lipid bilayer has given more rigidity to the membrane system when compared to the area per lipid of the POPC bilayer relative to the previously obtained data,³⁵⁻³⁷ which implies that these peptides maximally perturb the rigidity and order of the membrane system. On the basis of all of these analyses, it is clearly shown that both peptides, Pep2 (penetratin) and Random_pep2 (VRF2019), have the capability of entering the membrane system (Figure 6). Furthermore, the peptide VRF2019 uptake was also validated in HTF cells.

Figure 5.

Average thickness of lipid bilayer measured with respect to time: (A) Pep_2 (penetratin) and (C) Random_pep2 (VRF2019); area per lipid: (B) Pep_2 (penetratin) and (D) Random_pep2 (VRF2019).

Figure 6.

Structural snapshots of MD simulations of Pep2 (penetratin) and Random_pep2 (VRF2019) at 0, 100, and 300 ns in the presence of POPC lipid bilayer. Peptides Pep2 (penetratin) (magenta color) and Random_pep2 (VRF2019) (green) are shown in cartoon representation, while the POPC membrane is depicted in lines.

Evaluation of Peptide Toxicity and Its Uptake in Human Tenon Fibroblast Cells

The cultured cells from tissue explants were checked for the characteristic markers for fibroblast cells. Expression of fibroblast cell markers like vimentin and the absence of pan-keratin was observed in the cells isolated from the tissue (Figure 7A). HTF cells were then treated with TGF-ß1 (5 ng/mL) to induce differentiation into myofibroblast cells to simulate the fibrosis process. The differentiated cells expressed α-smooth muscle actin (α-SMA) (Figure 7A) (Figure S1A and C). To test the toxicity of peptides (penetratin and VRF2019), HTF cells were treated with different concentrations (10, 20, 30, 40, 50, 75, and 100 μM) of peptides for 24 h, and cell viability was determined using the MTT assay. Our findings indicated that peptides were nontoxic, with >70% of cells being viable for 24 h at 100 μM concentration (Figure 7B). Our findings are consistent with previous reports of CPP’s nontoxicity.³⁸ The proliferative activity was evaluated using BrdU assay. Our data revealed that peptides did not affect the proliferation of HTF cells (Figure 7C). CPPs generally penetrate cell membranes through two mechanisms, including direct translocation, which is temperature-independent, and endocytosis, which is temperature-dependent and energy-independent.³⁹⁻⁴² The mechanism of penetratin peptide cellular uptake has been reported to undergo direct translocation.^43,44 To understand the cellular uptake mechanism of the VRF2019 peptide, HTF cells were treated with VRF2019 and incubated at 4 and 37 °C. The uptake study showed that VRF2019 was able to penetrate HTF and myofibroblast cells at both 4 and 37 °C (Figure 7D). This confirms that VRF2019 also undergoes temperature-independent direct translocation. The time kinetics of the peptide uptake assay also revealed that the peptide can penetrate both the cytosol and the nucleus in 1 h (Figure 7E). Our data on actin staining also revealed there was a cytoskeletal reorganization between fibroblast and myofibroblast cells (Figure S1B). On peptide treatment, the cells did not show a significant change (Figure 7F).

Figure 7.

Peptide toxicity and uptake. (A) Immunofluorescence markers of HTF and myofibroblast cells showing vimentin and alpha-SMA positive. (B) MTT of peptide penetratin and VRF2019 at various concentration (5–100 μM). (C) Proliferation of HTF cells on VRF2019 treatment by BrdU assay. (D) Mechanism of cellular uptake of VRF2019 in HTF cells at 4 and 37 °C. (E) Time kinetics of VRF2019 cellular uptake in HTF cells (peptide uptake in green (FITC) and nucleus in blue (Hoechst)). (F) Actin reorganization by phalloidin staining in HTF cells post-treatment of VRF2019 and penetratin.

Peptide–Drug Conjugate and PCL-Composite Nanofiber Characterization

We conjugated the drug MMC with a peptide using standard EDC/NHS chemistry. The presence of conjugate in the reaction mixture was confirmed by HPLC and mass spectroscopy. HPLC data confirmed the presence of the conjugate by showing a difference in column retention time of 4.4 s when compared with either MMC or peptide alone. Our mass spectroscopy data also revealed the mass of the conjugate as 2467 kDa, which is equal to the theoretical mass of 2637 Da (Figure 8A and B and Figure S2A and B).

Figure 8.

Characterization of VRF2019-MMC conjugate. (A) Mass spectra of VRF2019, MMC, and conjugate with their respected masses. (B) HPLC peaks showing respected retention time of molecules (VRF2019, MMC, and conjugate). (C) SEM micrographs of PCL composites (PCL-VRF2019, PCL-MMC, PCL-conjugate, and PCL) and respective histograms showing the dimensions of nanofibers. (D) FTIR spectra of VRF2019, MMC, conjugate, PCL, PCL-VRF2019, PCL-MMC, and PCL-conjugate.

Further, to enhance the delivery of peptide–drug conjugates, we used 5% polycaprolactone to prepare nanofibers to release the VRF2019–MMC conjugate. SEM micrographs of electrospun matrices obtained from the peptide/PCL, drug/PCL, and conjugate/PCL blends are shown in Figure 8C. The diameters of the fibers are presented in the histogram. According to the data, the diameter of conjugate/PCL composite fibers is less than that of other PCL blend fibers. To analyze the interaction of PCL with VRF2019, MMC, and conjugate, the electrospun matrices obtained from various combinations were subjected to FTIR spectroscopy (Figure 8D). The IR spectra of PCL and PCL blends are shown with respect to conjugate, NH stretching vibrations, CH₂ vibrations, C=O vibrations of the amide linkage, and NH bending vibrations observed, respectively, at 3379, 2939, 2876, 1657, 1187, 1137, 1124, 1083, and 1039 cm^–1. The peaks observed for PCL alone at 2946, 2870, and 2821 cm^–1 are caused by CH₂ vibrations, whereas the intense sharp peak at 1653 cm^–1 is caused by C=O vibration. The slight shift of the characteristic peak is evidence of the interaction occurring between the conjugate and PCL.

To understand the surface topography of PCL composites, BET analysis was carried out. The collected nanofiber composites on the aluminum foil were analyzed for the BET analysis, and the adsorption/desorption isotherms obtained are shown in Table S4. The isotherm graphs of PCL, PCL-conjugate, PCL-peptide, and PCL-drug correspond to types I (a), V, I (a), and IV (a), respectively. Both PCL and PCL-peptide have been found to contain microspores, whereas PCL-conjugate and PCL-drug were found to have megapores according to IUPAC classification. The surface area and pore volume were found to be altered when compared to PCL alone (Figure S4a and b).

Conjugated Peptide Complex Stability Analysis by Molecular Dynamics

According to the RMSD plot (Figure 10A), among the peptides in complex with mitomycin C (Figure 9A–D), Pep2 (penetratin) has the fewest deviations compared to Random_pep2 (VRF2019), whereas Random_pep2 (VRF2019) has the greatest deviations without reaching any equilibrium until the end of the MD production run. In the case of both of the peptide–mitomycin C complexes, mitomycin C was found to be stable and intact with the peptides (Figure 10C and D) until 300 ns of MD simulation. The RMSF plot (Figure 10B) shows that, apart from the N- and C-terminal residues, the Pep2 (penetratin) residues were found to fluctuate less than Random_pep2 (VRF2019). It was also observed from the average area per lipid bilayer that both peptides in complex with mitomycin C increased the lipid bilayer disruption efficacy (Figure 10E and F and Table S5).

Figure 10.

MD analysis of peptide–mitomycin C complexes. (A) RMSD and (B) RMSF representing the highly fluctuating loop regions at N- and C-terminal. (C) RMSD of Pep2 (penetratin)–mitomycin C. (D) RMSD of Random_pep2 (VRF2019)–mitomycin C. Average thickness of lipid bilayer measured with respect to time: (E) Pep2 (penetratin)–mitomycin C and (F) Random_pep2 (VRF2019)–mitomycin C.

Figure 9.

Preprocessed and minimized peptide–mitomycin C complexes: (A) penetratin and (B) Random_pep2 (VRF2019). Structural snapshots of MD simulations of initial complexes: (C) Pep2 (penetratin)–mitomycin C and (D) Random_pep2 (VRF2019)–mitomycin C, in the presence of the POPC lipid bilayer.

Conjugate Release from PCL Nanofibers and Its Effect on HTF Cells

The kinetics of peptide release from fibers containing two distinct doses of peptide–drug conjugate (100 and 200 μM) were studied. The conjugate release was monitored at various time intervals (30 min, 1, 2, 3, 4, 8, 16, and 24 h), and the cumulative release of conjugates released from PCL composite fibers up to 24 h is indicated in Figure 11A. Our results suggest that release is concentration-independent. The secondary structure of the released conjugates was studied and compared to that of the native peptide using CD spectroscopy. Our CD spectroscopy measurements demonstrate a structural difference between the released and native peptides (Figure 11B). When compared to native peptides with some helical structure, the released conjugates revealed a negative band. This demonstrates that the functionality of the retaining conjugates has been kept in the PCL nanofiber matrix. The MTT assay was used to assess the cytotoxicity of the conjugate and released conjugate on HTF. Our findings demonstrate that cell viability was significantly reduced after conjugate and released conjugate treatment (P < 0.001). The activity of the released conjugate is identical with that of the native conjugate. When compared to native MMC, the impact is ∼30 times stronger (Figure 11C and Figure S3). To ensure the conjugate’s cytotoxicity, we also treated the HTF cells by merely combining the drug and peptide. The findings showed that a simple combination of drug and peptide had no apparent cellular toxicity when compared to the conjugate alone (Figure 11D).

Figure 11.

Conjugate release and its effect on HTF cells. (A) Release kinetics of conjugate from PCL conjugate composite by HPLC with two different concentrations (100 and 200 μM). (B) Characterization of released conjugate by CD spectroscopy showing the spectra of VRF2019, VRF2019 release, conjugate, and conjugate release. (C) Validation of conjugates and released conjugates on HTF cells by MTT assay (P < 0.0001). (D) Effect of conjugate and VRF2019+MMC mix of HTF cells by MTT assay.

Validation of Conjugate in Fibrosis Modulation

To test the conjugate’s impact on HTF cells, primary HTF cells of 10 00 000 were treated for 1 h with 10 μM VRF2019–MMC conjugate. After treatment, the expression of collagen-secretion genes (COL1A1 and COL1A2) and TGF-β1 were measured by qPCR. The results (Figure 12C) demonstrated that conjugate-treated HTF cells had significant (P < 0.001) downregulation of COL1A2, although COL1A1 was not significantly downregulated when compared to drug-treated cells. TGF-β1 expression remains stable, with no substantial alterations after conjugate treatment. TGF-1 levels do not change in our immunofluorescence data (Figure 12A and B). TGF-β1 is a potent molecule that promotes extracellular matrix (ECM) production in wound healing, scarring, and fibrosis.^45,46 It also stimulates collagen type I secretion during fibrosis. According to our findings, VRF2019-MMC conjugate causes changes in COL1A1 and COL1A2 in HTF, demonstrating that peptide–drug conjugates influence collagen production by suppressing collagen gene expression.

Figure 12.

Conjugate treatment on HTF cells and its effect on biological functions. (A) Immunofluorescence of TGF-β1 expression in control HTF cells. (B) Immunofluorescence of TGF-β1 expression in conjugate-treated HTF cells. (C) Relative gene expression of COL1A1, COL1A2, and TGF-β1 mRNA with respect to GAPDH (n = 4) (P < 0.001).

Conclusion

In our present study, we have designed a cell-penetrating peptide, VRF2019, by using a subtractive proteomics approach from the EMT proteomic database. Furthermore, we characterized the peptide by both in silico and in vitro methods and established the mechanism of uptake in HTF cells. The prepared peptide was conjugated with MMC by standard EDC/NHS chemistry, and its delivery was enhanced through a PCL nanofiber matrix. The conjugate release was monitored through HPLC, and the maximum conjugate release was achieved at 2 h. Further, the effect of the conjugate and released conjugate was tested on HTF cells by MTT assay. Our data show that conjugate and released conjugate have enhanced the drug’s activity when compared with native MMC. Further, the conjugated peptide–mitomycin C was also observed to have maintained stability in the membrane. On the basis of the in vitro data, we conclude that the VRF2019 peptide–drug composite nanofiber matrix increased drug availability and had potential benefits in fibrosis. Our qPCR results also suggest that the peptide VRF2019–drug conjugate may influence fibrosis regulation. Future research could entail using animal experiments to verify and evaluate this premise.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsptsci.2c00148.

• List of primers used for qPCR analysis; peptides prioritized based on sequential in silico validations; zeta potential and zeta size of peptides at various pH values; MD simulation analysis of the peptides in POPC membrane; BET analysis of PCL and PCL composites; cellular uptake, F-actin staining, and Western blot; characterization of penetratin conjugate; IC₅₀ estimation of MMC by MTT assay in HTF cells; and BET analysis of PCL and PCL composites (PDF)

Author Present Address

^∇ Department of Virology and Biotechnology, ICMR–National Institute for Research in Tuberculosis/Bioinformatics Division, Chennai 600031, India

Author Contributions

^# D.G. and H.N. are co-first authors.

The authors declare no competing financial interest.

Special Issue

Published as part of the ACS Pharmacology & Translational Science virtual special issue “New Drug Modalities in Medicinal Chemistry, Pharmacology, and Translational Science”.