Peptide concentration gradients and aligned microfiber topography synergize to speed and direct Schwann cell migration

Abstract

Conjugating precise concentrations of bioactive peptides on aligned topographies holds a promising application in directionally guiding Schwann cell migration, a significant step in peripheral nerve regeneration. To harness this behavior, we have developed aligned fiber scaffolds functionalized with variable concentration gradients of YIGSR, a laminin-derived peptide known to promote Schwann cell motility. Using thiol-ene click chemistries, we generated uniform and gradient patterns of YIGSR on the aligned fibers with spatial control over tethered peptide concentration during fabrication, yielding two uniform concentration scaffolds of 100 pmol/cm² and 420 pmol/cm² YIGSR, and three gradient profiles of slopes 7 pmol·(cm²·mm)⁻¹, 15 pmol·(cm²·mm)⁻¹, and 60 pmol·(cm²·mm)⁻¹. Schwann cell migration on scaffolds revealed that uniform YIGSR functionalization enhanced migration in a sex-specific and concentration-dependent manner. Female Schwann cells responded with greater migration on 100 pmol/cm² uniform YIGSR-functionalized fibers while male Schwann cell migration was enhanced on fibers with both 100 and 420 pmol/cm² compared to non-functionalized fibers. However, guidance of cell migration can not be achieved with increasing cell speed alone. Therefore, gradients were fabricated directly on the fiber scaffolds and quantified. While shallow YIGSR gradients (7 and 15 pmol·(cm²·mm)⁻¹) did not consistently bias Schwann cell directionality in the direction of the gradient, 60 pmol·(cm²·mm)⁻¹ gradient profiles induced a haptotactic response, measured by directional velocity and haptotactic index, with both sexes migrating toward regions of higher peptide concentration. Thus, along with contact guidance effects provided by aligned fibers, precisely-defined peptide-functionalized gradients can be used to further bias Schwann cell migration for nerve regenerative applications.

Statement of Significance:

Peripheral nerve injuries often result in incomplete recovery, partly because cells crucial for repair cannot efficiently move into injury sites. While researchers have developed aligned fibers that act as a pathway for the cells into the injury site, cells are free to move in any direction along the path, reducing their ability to support repair. This study demonstrates that by combining aligned fibers with bound chemical gradients to act as guard rails, cells move preferentially in one direction along the pathway. By precisely controlling both the fibers’ physical alignment and chemical gradients, we achieved unidirectional cell migration. This dualcue approach represents a significant advancement in biomaterial design for nerve repair, offering a promising strategy to enhance regeneration across nerve defects.

1. Introduction

The peripheral nervous system is known to regenerate after injury, although clinical outcomes are often inconsistent or delayed [1,2]. In cases of segmental injuries, surgical intervention is required to maximize recovery of sensory and motor function [3,4]. Typically, for critical size gaps (> 3 cm in humans), autografting is required. This procedure involves the harvest of a “less-critical” nerve from the patient and suturing it into the injury site to serve as an immunologically inert, bioactive scaffold [3,5]. While autografting is the gold standard for repair, it necessitates a secondary procedure, introduces significant morbidity at the donor site, and is limited by other issues including fascicle incompatibility and availability of tissue [6]. Synthetic nerve guidance conduits and scaffolds offer a compelling alternative to grafting; however, their functional performance has so far remained suboptimal in comparison [7]. As such, peripheral nerve reconstruction remains a complex and challenging surgical task [8].

A critical component of this regenerative process is the behavior of Schwann cells; axonal regeneration can only proceed to the extent of which Schwann cells are able to migrate to and within the injury site [9]. Synthetic matrices can be designed to introduce longitudinal features that mimic blood vessels within a native nerve [10]. With features such as alignment and curvature, fibers have been used to take advantage of contact guidance, a long-studied phenomena in both cell migration [11] and axonal guidance [12]. In the development of materials for nerve regeneration, aligned polymer fibers with diameters from 0.2 to 1.2 μm have resulted in various levels of Schwann cell axial morphological alignment [13], maturation [13], actin alignment [14–16], and cell movement along the fiber axis [14,16,17]. While contact guidance clearly aligns Schwann cells uniaxially, aligned fibers alone have not been sufficient to speed infiltration in one direction over another. To enhance nerve regeneration, a design must recruit or direct cells into the gap, unidirectionally.

In the context of cell migration, molecular and mechanical gradients have long been known to direct cells. While initially studied as chemotaxis or directed cell migration in response to a soluble gradient, other forms of directed taxis have been uncovered, such as haptotaxis [18–21], response to bound cues, or mechanotaxis [22], response to mechanical gradients. Biomaterial design has long sought to mimic these cues, from soluble gradients due to controlled release to bound gradients of peptides and proteins on scaffolds [23,24]. Apart from enhancing the bioactivity of a material, bound gradients can spatially recruit cells with wide ranging examples from the immune response [25,26] to metastasis [27,28] to nerve development and regeneration [29]. Laminin was originally noted to increase the haptotaxis of Schwann cells [21], and therefore, laminin peptides, such as YIGSR (Try-Ile-Gly-Ser-Arg) [30] and IKVAV [31], bound to solid substrates have been further used to enhance Schwann cell migration [32,33]. More specifically, Motta et al. showed significant haptotactic migration of Schwann cells on YIGSR-gradient functionalized glass substrates [33]. However, to translate these types of materials, gradients that bias cell migration need to be developed on materials that can be implanted into a wound.

Aligned topographies and bioactive gradients have independently demonstrated promising results on Schwann cells. However, synergistic investigations of peptide functionalized topographical substrates on directing Schwann cell migration have been limited [32]. Polymers have been formed into topographical structures, including aligned fibers or patterned ridges, that support contact guidance, or uniaxial migration. Thiol-ene click strategies, highly selective reactions that are initiated by relatively mild reaction conditions, including photochemistry [34], have become increasingly popular as crosslinking methods in the fabrication of biomaterials and for bioconjugation [35,36]. Due to its photo-reactivity, thiol-ene reactions afford excellent control of the placement of bioactive species and consequently allow for precise substrate patterning, making it a scalable method to advance material functionalization in gradient concentration profiles and patterns [37,38]. The development of aligned fiber scaffolds with precise gradients, therefore, can lead to synergistic methods to enhance and direct cell migration – through both contact guidance and haptotaxis.

This work aims to maximize unidirectional migration of Schwann cells to improve peripheral nerve regeneration. We utilized a thiol-ene click modification of allyl functionalized polycaprolactone (PCL), along with touch spinning [39] to fabricate aligned polymer fibers. These fibers were functionalized with varying degrees of concentration gradients of YIGSR peptides, to ultimately bias Schwann cell migration. Schwann cells of both biological sexes were measured, as previous work has demonstrated clear differences in biological responses, such as migration, myelination, and protein production of glial cells based on sex [16,40–43]. Quantification of cell migration provided a comprehensive view of the gradients on fibers necessary that direct cells when contact guidance cues are also present. The scope and extent of this impact is detailed below.

2. Materials and methods

2.1. Materials

All materials were used as received unless otherwise specified. Hexanes (mixture of isomers, ACS reagent, ≥ 98.5%), chloroform (contains amylene as stabilizer, ≥ 99.5%), dichloromethane (ACS reagent, contains amylene as stabilizer, ≥ 99.5%), 3-chloroperbenzoic acid (≤ 77.0%), sodium bicarbonate (ACS reagent, ≥ 99.5%), magnesium sulfate (anhydrous, ≥ 99.5%), and sodium chloride (ACS reagent, ≥ 99.0%) were purchased from Sigma-Aldrich (St. Louis, MO). Isopropanol (extra dry over molecular sieves, AcroSeal, 99.5%) and 2-allyl-cyclohexanone (≥ 97.0%) were purchased from Thermo Fisher Scientific (Waltham, MA). Toluene (HPLC grade, 99.9%, Sigma-Aldrich) was dried on an Inert Pure Solv system (model PS-MD-3) and degassed via three cycles of freeze-pump-thaw. ε-Caprolactone (97%, Sigma-Aldrich) was dried over calcium hydride under nitrogen flow and distilled using active vacuum. Methanol (ACS Reagent, ≥ 99.8%) was purchased from Sigma-Aldrich (St. Louis, MO). Lithium phenyl(2,4,6-trimethylbenzoyl) phosphinate (LAP, > 98.0%) was purchased from TCI America (Mountain View, CA). CYIGSRK-FITC (≥ 90%) was purchased from Proteogenix (Schiltigheim, France). C-Peg11-YIGSR (≥ 90%) was purchased from Genscript (Piscataway, New Jersey).

Dulbecco’s Modified Eagle Medium (DMEM) (catalog #: 90-013-PB), N2 supplement (catalog #: 17502048), bovine pituitary extract (BPE) (catalog #: 13028014), goat Alexa Fluor^™ 568 anti-rabbit IgG (catalog #: A-11011), Alexa Fluor^™ 488 phalloidin (catalog #: A12379), paraformaldehyde (PFA) (catalog #: O4042–500), bovine serum albumin (catalog #: BP9706), Hoechst nuclei stain 33342 (catalog #: H1399), rectangular dishes (4-well, catalog #: 267061), borosilicate 25 × 25 mm coverslips (catalog #: 12-541-039) and CellTracker^™ Deep Red Dye (catalog #: C34565) were purchased from Thermo Fisher Scientific (Waltham, MA). Laminin (catalog #: L2020), fetal bovine serum (FBS) (catalog #: F2442), forskolin (catalog #: F6886), L-glutamine (catalog #: 25030081), Trypsin-EDTA (catalog #: T4174), poly-L-lysine hydrobromide (catalog #: P9155), phosphate buffered saline (PBS) (catalog #: BP2944100), rabbit anti S-100 IgG (catalog #: S2644), Hank’s Balanced Salt Solution with and without Ca²⁺ and Mg²⁺ (catalog #: H1387 and H4891) and sodium borohydride (catalog #: 102894) was purchased from MilliporeSigma (Burlington, MA). Triton X-100 (catalog #: 8698.5–16) was purchased from Ricca Chemical Company (Arlington, TX). Breath-Easy film (catalog #: BEM-1) was purchased from Diversified Biotech (Dedham, MA). Loctite silicone sealant (catalog #: 908570) was purchased from Henkel (Germany).

2.2. Polymer synthesis and fiber formation

Polymers were synthesized and fibers were fabricated via touch spinning. Briefly, Mg(BHT)₂(THF)₂ was synthesized as previously described [44]. The synthesis of 6-allyl-ε-caprolactone is further detailed in the SI and in Scheme S1. To synthesize 25% allyl-functionalized polycaprolactone (allyl-PCL), in a glovebox (H₂O < 0.1 ppm, O₂ < 0.1 ppm), a solution of Mg(BHT)₂(THF)₂ (catalyst, 2.96 g, 4.86 mmol, 0.01 eq.) and isopropanol (initiator, 0.97 mmol, 0.002 eq.) was dissolved in dry toluene (8 mL). This mixture was subsequently added to a solution of ε-caprolactone (41.63 g, 36.48 mmol) and 6-allyl-ε-caprolactone (18.75 g, 12.16 mmol; NMR: Figures S1 and S2) in dry toluene (350 mL) to yield 25 mol% allyl-PCL (Fig. 1A). The vessel was sealed, removed from the glovebox, and left to stir for 5 h at 60°C. The reaction was then quenched with trifluoroacetic acid (TFA) and the polymer was thrice precipitated into cold n-hexanes. The resulting polymer, allyl-PCL, was dried in vacuo overnight and collected as a yellow/white solid. Both proton (¹H) and carbon (¹³C) nuclear magnetic resonance (NMR) spectroscopy were performed in CDCl₃ on a Bruker Advance NEO at 500 MHz (23°C) with a 5 mm iProbe. Further characterization with SEC, TGA, and DSC are detailed in the SI.

Fig. 1.

(A) Allyl-functionalized poly(ε-caprolactone) (allyl-PCL) was synthesized and touch-spun into aligned scaffolds. The polymer was synthesized via ring-opening copolymerization using a 3:1 molar ratio of ε-caprolactone to 6-allyl-ε-caprolactone. Mg(BHT)₂(THF)₂ and isopropanol were used as the catalyst and initiator, respectively. (B) The resulting fiber scaffold collected on a coverslip is shown. Scale bar = 5 mm. (C) An SEM image of touch-spun, aligned fibers. Scale bar = 20 μm. D) The fibers were spun at a constant rate of 2000 RPM, yielding a diameter of 1.2 ± 0.1 μm. N = 77 fibers were measured. (E) Fiber alignment was measured, showing a narrow angle distribution. Three samples (N = 3) were imaged and quantified for angular distribution. (F) To visualize the gradients, fluorescently labeled peptide (CYIGSRK-FITC) was functionalized on to the scaffolds. The full scaffold is shown at 25 mm. (i) 7 pmol·(cm²·mm)⁻¹; (ii) 15 pmol·(cm²·mm)⁻¹; (iii) 60 pmol·(cm²·mm)⁻¹. (G) Within each Region, concentration of peptide was quantified along the gradients, resulting in 7, 15, or 60 pmol·(cm²·mm)⁻¹. Closed symbols indicate the measured concentrations; open symbols are the points along the regression line that match Positions a-e. (H) This quantified concentration was then further correlated to the applied light intensity to demonstrate the fidelity of the method. Resulting linear fit was [Concentration = 40.1*(Light Intensity) +70.58]. Linear correlation (r²) was 69.4% for all of the scaffolds quantified, regardless of the gradient applied.

The allyl-PCL (M_w = 88 kDa) was dissolved in chloroform (25 w/w %), drawn into a glass syringe (1 mL) fixed to a blunt-end needle (18G), and loaded onto a syringe pump. The syringe pump was positioned so that the tip of the needle was in tangential contact with rotating, perpendicular rods connected to the rotating disk [39]. Polymer was dispensed out of the syringe at a constant flow rate (5 μL/min) and a rotation speed of 2000RPM. Borosilicate coverslips (25 × 25 mm) were used as the collection plates for the fibers.

The resulting fibers were sputter coated with gold and imaged using scanning electron microscopy (SEM). Sputter coating was done using a Denton Desk V Sputter-Coater with gold at a coating rate of 1.39 nm/min for 300 s forming approximately a 6.95 nm uniform layer. SEM was performed on a Thermo Fisher Scientific Apreo S using the secondary electron detector. Fiber diameter was measured using the measurement function of NIH FIJI and directionality was analyzed using the directionality plugin of the same software. The average diameter over N = 77 fibers, measured across 10 images, was calculated and reported. Directionality was calculated using NIH FIJI Directionality plugin [45], images were uploaded and rotated to achieve a direction of 0° ± 0.2; the center of the gaussian distribution was defined as 0°. The angular dispersion (°) is reported and represents the average standard deviation of the gaussian distribution across 3 (N=3) fiber scaffolds.

2.3. Gradient functionalization of aligned fibers

The gradients were developed on the scaffolds using UV light (Fig. 2A) and photomasks (Fig. 2B). Gradient profiles of transparency were designed on Adobe Illustrator. The photomask that resulted in 7 pmol·(cm²·mm)⁻¹ gradient was designed as 50% to 100% transparency over 25 mm, 15 pmol·(cm²·mm)⁻¹ was 0% to 100% transparency over 25 mm, and 60 pmol·(cm²·mm)⁻¹ was 0% to 100% over 6.25 mm in the center (Fig. 2B). The resulting Illustrator file was then used for the fabrication of a chromium neutral density filter on quartz purchased from Advance Reproductions Corporation (North Andover, MA).

Fig. 2.

Workflow for materials and cellular characterization. (A) Aligned fibrous scaffolds were functionalized post-fabrication using thiol-ene click chemistry; graphical scheme demonstrating the functionalization process is depicted. (B) Schematic of gradient profiles applied to the fibers. The transparency percent of 0 indicates no light being transmitted, while 100 is full light being transmitted; three gradients were produced. Masks were 25 mm in length to develop three Regions that were characterized, each 6 mm in length. Small sections (3.5mm) at each end of the scaffold were removed for materials characterization. Region averaged concentrations were determined at the midsection of each region, denoted by black X. In addition, for the 6.5 mm gradient, concentrations were determined at the boundaries between Regions 1 and 2, and between Regions 2 and 3, denoted by red X. (C) For the cell studies, fiber scaffolds were analyzed using the schematic shown, with cells tracked in Regions 1, 2, and 3. Each box relates to a capture area (0.75mm × 1.1 mm) along the positional lines a (x=7.5 mm), b (x=10.5 mm), c (x=12.5 mm), d (x=14.5 mm), and e (x=17.5 mm). Cell migration was captured at least along 3 positional lines for each gradient. For consistency, the lengths of the drawings are consistent and meant to represent 25 mm, or the length of the slide.

Prior to addition of functional groups, the fibers were examine after exposure to UV light to confirm no changes due to UV light alone; methods and results are detailed in SI. Coverslips with touch-spun aligned fibers were submerged in a solution of the appropriate thiol (0.8 mM) and LAP (0.3 mM) in H₂O/MeOH (4:5 v/v). The photomask was placed over the fibers, with the gradient profile aligned with the direction of fibers. UV light (λ = 365 nm, I = 9 mW/cm² at scaffold surface) was applied through the photomask for 15 min. For uniform scaffolds, no photomask was used; uni-420 pmol/cm² was achieved using UV light (λ = 365 nm, I = 9 mW/cm²) for 15 mins, while uni-100 pmol/cm² achieved using UV light (λ = 365 nm, I = 5 mW/cm²) for 5 mins. The UV light intensity at the fiber surface through the photomask was measured directly and correlated to the photomask transparency (Figure S6).

For imaging, confirmation of gradient fidelity, and quantitative determination of the concentration, CYIGSRK-FITC was used as the thiol. The uniform scaffolds were made using no photomask, and functionalization concentration was controlled precisely by varying UV intensity and irradiation time using the photomask [46]. Fibers were imaged using a Keyence BZ-X710 fluorescence microscope. Fluorescence intensity along the fiber direction was examined using the ImageJ [47]. Using the “plot profile” tool, the normalized pixel intensity of the scaffold was measured and plotted along the gradient direction and distance. The pixel intensity measurement was also performed on the photomask to determine the fidelity of the mask transfer. The thiol was subsequently replaced with Cys-Peg11-YIGSR for cell migration studies.

2.4. Quantification of YIGSR on fibers

Scaffolds functionalized with CYIGSRK-FITC were used to quantify the functionalization density of immobilized YIGSR. For each gradient profile, 3.5 mm of fibers were removed from both the left and the right side of the scaffold because of aggregation and physical absorbance along the edges of the scaffold. Assuming position x = 0 mm is defined as the leftmost edge and x = 25 mm is the rightmost edge, the remaining fibers were divided into 3 equal sections along the fiber direction (6 mm wide sections) and the average concentration between 3.5 and 9.5 mm (Region 1), 9.5 and 15.5 mm (Region 2), and 15.5 and 21.5 mm (Region 3) was determined per pattern (Fig. 2C). Samples were trimmed to keep a consistent mass of 0.22 mg across all samples.

Sections were then dissolved in 300 μL of HFIP and the fluorescence was read using a Shimadzu UV-3600i UV-vis-NIR Spectrophotometer (λ_ex = 449 nm, emission λ_em = 511 nm). Resulting concentration was calculated using a CYIGSRK-FITC calibration curve. Linear regression lines were then calculated for each concentration profile using the measured concentrations per region. Using the regression curves, concentrations were calculated at x-distances x = 7.5, 12.5, and 17.5 mm for all gradients to align with cell studies. For the highest gradient, additional positions at x = 10.5 and 14.5 mm were calculated from the regression. Scaffolds were oriented so that distance x = 0 mm represented the lowest end of the concentration gradient. At least three independent scaffolds were tested and measured for each pattern.

2.5. Stability of peptide-functionalized scaffolds verification

Fiber scaffolds were functionalized with CYIGSRK-FITC using the medium gradient. The fiber scaffold was imaged using a Keyence BZ-X710 fluorescence microscope to verify the covalent tethering of CYIGSRK-FITC over the time of the migration studies. The scaffold was then submerged in proliferation media, incubated (37°C, 5% CO₂ and atmospheric O₂) for 48 h, and reimaged to verify the persistent presence and stability of functionalization.

2.6. Cell migration analysis

2.6.1. Substrate preparation and sterilization

Loctite silicone sealant was manually extruded atop fiber scaffolds to build a well for cell media. Resulting samples were sterilized using ethylene oxide (Andersen Sterilizer, Haw River, NC) and a 12 h sterilization cycle, followed by a 2 h aeration period. Sterilized substrates were then left for a minimum 72 h to allow for the release of any residual ethylene oxide before beginning migration experiments. Sterilized substrates were kept under vacuum until use.

2.7. Cell culture media formulations

Isolation media was specifically designed to act as a fibroblast inhibitor and contained DMEM D-valine, L-glutamine (1x), FBS (10% vol/vol), forskolin (5 μM), N2 supplement (1% vol/vol), and BPE (14.2 μg/mL) [48]. Proliferation media was identical to isolation media, except for the use of standard DMEM in place of DMEM D-valine. Seeding media contained only DMEM and L-glutamine. Migration media contained DMEM, FBS (10 vol/vol%), and L-glutamine (1x).

2.8. Isolation of Schwann cells

Primary Schwann cells were isolated from female and male adult (3–5 months) Sprague Dawley rat sciatic nerves that were recently sacrificed for other IACUC approved protocols. For each sex, a group of 2–3 rats (4–6 sciatic nerves) was designated as one biological replicate (N), resulting in a total of three separate biological replicates (N=3) used. For each isolation, nerves were removed, cells were dissociated, and then cells were cultured in isolation media on flasks coated with poly-L-lysine (1.5 μg/cm²) and laminin (0.1 μg/cm²) [33]. After ~14 days, cells were collected and counted. For cryopreservation, 0.5 × 10⁶ cells were resuspended in 900 μL proliferation media and 100 μL DMSO in a cryovial. The cryovial was stored at −80°C in a CoolCell^™ freezer container overnight, then transferred to a liquid nitrogen cell dewar the following day for long term storage.

2.9. Schwann cell purity quantification

The purity of Schwann cells was assessed by seeding and culturing Schwann cells at 5,000 cell/cm² in a 24-well tissue culture plate for 48h. Schwann cells were fixed and labeled using previously published immunocytochemistry protocols [14,33]. In short, samples were fixed with 4% (v/v) paraformaldehyde solution for 10 min and permeabilized with 0.1% Triton X-100 for 10 min. Samples were subsequently incubated in 1 mg/mL sodium borohydride in PBS for 10 min (2x), washed with PBS, and blocked with 7.5% (w/v) BSA for 1 h at ambient temperature. Samples were then washed once with 1% (w/v) BSA and incubated in rabbit anti S-100 primary antibody (1:250 in 1% (w/v) BSA) at 4°C overnight. Cells were then incubated in Alexa Fluor^™ 568 (AF568) goat anti-rabbit secondary antibody (1:1000 in 1% (w/v) BSA) for 1 h at ambient temperature, washed with PBS, and labeled using Hoechst 33342 (1:500 in 1% w/v BSA). Three biological replicates were used (N = 3) and three technical replicates were conducted for each biological replicate. Cells were observed and imaged using fluorescence microscopy with 7 images of each substrate for a total of 21 images per biological replicate. Total S100⁺ cells were counted relative to all labeled nuclei to quantify purity. Only female and male Schwann cell cultures at high purities (> ~98%) were used for this study.

2.10. Culture and maintenance of primary Schwann cells

Flasks were coated with poly-L-lysine (PLL) (1.5 μg/cm²) for at least 2 h at room temperature or overnight 4°C. On the day of cell culture, the PLL solution was discarded and the flasks were washed once with PBS. The flasks were left to dry for at least 30 min in the biosafety cabinet before seeding cells. Frozen cells were taken from the liquid nitrogen dewar, quickly thawed in a warming bath, washed and then transferred for incubation (37°C, 5% CO₂ and atmospheric O₂). When cells reached ~80% confluence, cells were passaged to new PLL-coated flasks at 2,500 cells/cm². All cells used in the reported experiments were less than passage 5. Cell doubling time was calculated using a formula described previously [33].

2.11. Live cell labeling

Media was removed from flasks and cells were washed with HBSS with Ca²⁺ and Mg²⁺. For each T25 flask, 2 mL of 250 nM CellTracker^™ Deep Red Dye in HBSS with Ca²⁺ and Mg²⁺ was added. Each flask was then incubated for 30 min, followed by a wash with HBSS.

2.12. Morphological analysis

After 24 h migration studies, samples were fixed using 4% paraformaldehyde in PBS, permeabilized with 0.1 % Triton X-100 for 10 min, and quenched with 1 mg/mL sodium borohydride in PBS. Samples were blocked with 1 mg/mL BSA in PBS for 1 h at room temperature. To visualize actin filaments and nuclei, samples were thrice washed with PBS for 5 min and incubated in a cocktail of AF488 phalloidin (1:400) and Hoechst 33342 (1:500) for 1 h at ambient temperature. Three biological replicates (N = 3) and three technical replicates (n = 3) were utilized on each YIGSR concentration at each uniform and gradient profile. Cells were observed and imaged using fluorescence microscopy and 5 images were taken on each technical replicate for a total of 15 images per biological replicate at each YIGSR profiles. To quantify the influence of topography and tethered YIGSR to actin alignment, images were analyzed using a previously described MATLAB method. This method utilized edge detection and statistical analysis to measure percent alignment ± 10° of each cell relative to fiber orientation (100% indicates perfect alignment) [14,49].

2.13. Seeding cells for time-lapse experiments

YIGSR-tethered fiber substrates were soaked in DMEM under reduced pressure for 30 min to enhance wetting for cell adhesion. Labeled cells were removed from flasks using trypsin (1x), counted, and resuspended in seeding media. All cells were seeded at a concentration of 10,000 cells/cm², and incubated (37°C, 5% CO₂ and atmospheric O₂) for 4 h to allow for cell adhesion. The seeding media was removed and the cells were washed once with migration media. Each substrate then received an additional 1 mL of migration media and the plate was covered with Breath-Easy film to ensure minimal media evaporation over the duration of the experiment.

2.14. Live cell migration tracking

Cell migration was recorded using Zeiss Zen software every 15 min for 24 h on a Zeiss microscope equipped with a humidified and incubation module (37°C, 5% CO₂ and atmospheric O₂). To evaluate the impacts of YIGSR concentration and gradients on cell migration, cells at specified positions were tracked. The vertical center line of each scaffold was identified and a series of single capture regions along that line were defined (Fig. 2C). Similarly, 5 mm to the left and the right of the center line was identified, and capture regions were marked to explore the high and low ends of the concentration gradients, respectively. For the steepest gradient, additional capture areas 2 mm to the left and right of the center line were defined to guarantee imaging within the gradient portion of the mask.

2.15. Manual cell tracking on YIGSR-functionalized fibers

Schwann cells from three biological replicates were used for time lapse migration experiments. At least 30 individual areas were captured for constant YIGSR and 10 individual areas per position (Fig. 2C) were captured for YIGSR gradients. Each condition was repeated at least three times as technical replicates. For each capture area, focus was optimized for visualization of cell morphology and kept constant across the capture. Cell positions were captured by manual selection using a tracking plugin in NIH FIJI [50]. A minimum of 122 individual migrating cells were tracked resulting in path positions (x,y,t) per condition per sex. To mitigate position drifts that were sometimes observed during 24 h time lapse period, a fixed point was identified and tracked (x,y,t) in each single capture region. All cell positions captured within that region used this fixed point as an internal reference.

2.16. Criteria for Schwann cell migration

Criteria was embedded in the analysis of cell migration to remove cells that exhibited contact inhibition of locomotion from cell-cell contact the 24 h tracking period [51]. Cells that experienced interactions of Types 1–3 [51] were manually removed from analysis to focus on cell-substrate correlations. Each tracked cell must have displaced at least two times the cell diameter (~100 μm) from the initial position at any point during the 24 h experiment to be considered a motile cell. This criterion was imposed to ensure cells were migrating versus exhibiting membrane extension. It also excludes cells movement that may be due to thermal perturbations during incubation and the microscope’s motorized stage [52]. For comparison, a displacement criterion of 50 μm was also performed to ensure observed results were not artifacts (included in supplemental information). Cells that exited the imaging boundaries during the tracking period were excluded.

2.17. Mean squared displacement of migrating cells

Using the tracked cell positions (x,y), mean-squared displacement (MSD) was calculated with overlapping intervals by Eq. (1):

$$\mathit{\text{MSD}}={\left[x\left(t-\tau \right)-x\left(\tau \right)\right]}^2+{\left[y\left(t-\tau \right)-y\left(\tau \right)\right]}^2$$ (1)

Where t is time, τ = nΔt and n= 1,2, …96, and Δt is the time step size (15 min). The ensemble average MSD and individual cell MSD were calculated using publicly available MATLAB code (msdanalyzer) [53], and further details for the calculations of these parameters can be found in the reference.

2.18. Velocity and instantaneous speed

Using the tracked cell positions (x,y), velocity and speed were calculated. On the fiber substrates, the direction parallel to fiber alignment was defined as the x-axis (0°). Biased velocity ($〈\overrightarrow{V_{\Vert }}〉$(μm/min)) is the average parallel velocity of an individual cell at each 15 min interval over 24 h along the x-axis (parallel to fiber alignment). In contrast, $〈\overrightarrow{V_{\perp }}〉$(μm/min) is the average perpendicular velocity of an individual cell at each 15 min interval over 24 h. Speed in the x-axis (〈S_‖〉(μm/min)) is defined as the average absolute value of the velocity $\left(〈|\overrightarrow{V_{\Vert }}|〉\right)$ of an individual cell at each 15 min interval. The average perpendicular speed is denoted as 〈S_⊥〉 (μm/min) and was defined similarly. Overall speed was simply calculated as path distance traveled divided by total time.

2.19. Haptotactic index

Haptotactic index (HI) along the x-axis was defined as the ratio of the displacement in the positive x-direction between start and end positions normalized to the total path length traveled. Cells that moved in the positive x-direction (oriented from low to high concentration on gradients) yielded positive HI values. The average HI of Schwann cells on flat and fiber substrates was calculated to determine if cell migration exhibited bias in the direction of increasing concentration [54].

2.20. Statistical analysis

All experiments were conducted at least three times unless otherwise noted. At least 122 individual cells sourced from three biological replicates (N = 3) of each sex were analyzed for each YIGSR concentration region. Error for cellular results was reported as standard deviation. To compare differences between uniform YIGSR concentrations and non-functionalized fiber scaffolds, one-way ANOVA test with Tukey as a post hoc test was used. To compare differences between YIGSR gradients and uniform YIGSR profiles within each sex with regards to motility (speed and velocity), and percentage of actin alignment, two-way ANOVA test with Tukey as a post hoc test was used. To compare the difference between perpendicular and parallel speed within each sex at each position per YIGSR gradient and on uniform YIGSR concentrations, unpaired-t test was used. To compare MSDs, 2-sample Kolmogorov-Smirnov test was used. To compare the differences between utilizing 50 μm and 100 μm migratory displacement requirements to calculate $〈\overrightarrow{V_{\Vert }}〉$ and HI, a two-tailed unpaired t-test was used. Groups that were statistically different were marked with a solid line. Groups that were not statistically different were left unmarked. Representative images of nuclei morphology and actin alignment of female and male Schwann cells were within the third quartile of quantification.

3. Results

3.1. Fabrication and characterization of touch-spun aligned fibers

Allyl-PCL was successfully synthesized through the ring-opening copolymerization of ε–caprolactone and 6-allyl-ε-caprolactone (Figure S1–S3). Using touch-spinning aligned fibers were successfully fabricated around glass coverslips (Fig. 1B). The resulting fibers were quantified dimensionally using scanning electron microscopy (Fig. 1C), confirming fabrication of 1.20 ± 0.10 μm diameter fibers (Fig. 1D). Angular distribution was additionally measured to quantify fiber alignment and results showed a narrow distribution range of ± 6.30° relative to the average fiber axis (0°) (Fig. 1E).

3.2. Gradient functionalization of aligned fibers

Fiber scaffolds were first exposed to UV light and analyzed using DSC and NMR to confirm no post-processing chemical or crystallinity changes to the material. Both analyses showed negligible differences relative to the fibers before UV exposure (Figures S4 and S5). Concentration gradients of Cys-PEG₁₁-YIGSR were applied using thiol-ene click mechanisms between the allyl functional handles in the polymer and the thiol side chain of the cysteine amino acid, in the presence of the photocrosslinker LAP (Fig. 2A). Three linear gradients and two uniform profiles of varying concentrations were fabricated on fiber scaffolds by applying defined photomasks and irradiating UV light. The respective fluorescence intensity of FITC-functionalized scaffold images (Figure 1Fi–iii) was visualized for each gradient profile, with (i) from 50–100% over 25 mm, (ii) from 0–100% over 25 mm, and (iii) from 0–100% over 6.25 mm as shown in Fig. 2B. Fidelity of the mask to the resulting profile was further visualized by fluorescence (Figure S7).

Bound peptide concentration was measured using calibration curves and UV-vis spectroscopy. To quantify the peptide, regions were set up on the scaffolds (Fig. 2B). The 50 to 100% gradient yielded YIGSR concentrations of 271.67 ± 99.2, 295.50 ± 77.0, and 354.7 ± 91.8 pmol/cm² at in Region 1, 2, and 3 respectively (Fig. 1G), providing a slope of 6.9 pmol·(cm²·mm)⁻¹. The 0 to 100% gradient yielded concentrations of 175.00 ± 23.1, 292.67 ± 37.23, and 350.33 ± 28.3 pmol/cm² (Fig. 1G) representing a slope of 14.6 pmol·(cm²·mm)⁻¹. For the remaining gradient, which was localized to the center of the scaffold, 161.43 ± 115.1, 287.25 ± 128.4, and 512.67 ± 55.2 pmol/cm² (Fig. 1G) were measured for the same three regions, however, since both Region 1 and Region 3 should have been constant due to the mask, the gradient was calculated based on the edges of Region 1 and Region 3 (red Xs in Fig. 2B), giving a gradient of 58.5 pmol·(cm²·mm)⁻¹ (Fig. 1G). With this quantification, the fiber scaffolds and gradient profiles were named with regards to their resulting slopes and will be referred to as gradients 7 pmol·(cm²·mm)⁻¹, 15 pmol·(cm²·mm)⁻¹, and 60 pmol·(cm²·mm)⁻¹. Peptide concentrations showed correlation to the UV light intensity when measured using UV-vis spectroscopy (Fig. 1H). For 60 pmol·(cm²·mm)⁻¹, two additional x-positions at 10.5 and 14.5 mm were added for cell tracking to ensure multiple points within the gradient; these positions are denoted as positions b and d and concentrations of 200.1 and 434.2 pmol/cm² were calculated from the best fit profile. The uniform profiles gave concentrations of 103.7 ± 38.3 and 423.6 ± 80.0 pmol/cm² and thus will be referred to as uni-100 and uni-420. Stability of tethered peptides was verified by imaging functionalized fibers pre- and post- incubation in cell media for 48 h, confirming peptide functionalization patterns remained present through and after incubation (Figure S8).

3.3. Cell analysis on uniform YIGSR substrates

Cell purity and doubling time: Female and male Schwann cells had high purity of 98.8 ± 0.5% and 98.8 ± 0.3% at isolation. To assure that the cells being tracked remained highly purified, Schwann cells purity was again examined after the experiments were completed. The purity of the female and male Schwann cells after the experiments was 99.1 ± 0.3% and 98.9 ± 0.6%, respectively, demonstrating that the cells were unlikely to be contaminated by other cells such as fibroblasts (Table S1). The doubling time of female (42 ± 17 h) and male Schwann cell groups (40 ± 12 h) were found to have no statistical differences.

Actin alignment: After the 24 h migration period, actin alignment was measured with high levels of actin aligned to the fiber axis on both constant YIGSR substrates (79.4 ± 3.9% (female) and 78.8 ± 3.5% (male) on uni-100; 80.5 ± 2.5% (female) and 78.9 ± 6.0% (male) on uni-420). Previously published data for non-functionalized fibers was used as comparison and no differences were found between constant YIGSR and non-functionalized fibers (data reused from ref [16] with permission).

Cell trajectory: Cell paths were tracked and plotted to visualize migration. Similar to actin assessment, previously published data for non-functionalized fibers were used as comparisons. Based on visualization of the cell paths (Fig. 3A), female cells showed increased movement on uni-100 scaffolds compared to uni-420 and non-functionalized fibers (data reused from ref [16] with permission). Unlike female cells, cell paths of male cells were similar at both uni-100 and uni-420 YIGSR (Fig. 3B), which were increased compared to non-functionalized fibers (data reused from ref [16] with permission).

Fig. 3.

Constant concentrations of YIGSR effect on cell motility. Pink represents female cells while black represents male cells. The cell paths were recentered to start at (0,0) and the trajectories were plotted over time for both (A) female cells and (B) male cells on YIGSR functionalized fibers on uniform 100 and 420 pmol/cm². The non-functionalized fiber (NF) data was reused from [16] with permission. (C) The speed parallel to the axis of the fibers was statistically the highest on 100 pmol/cm² fibers for female cells and 420 pmol/cm² for male cells. (D) The velocity, however, which has direction both in the + and – direction, were not different between the substrates. (E, F) The MSD was similar for the uniform and non-functionalized fibers over time. The dashed lines around each data set represents standard error for the lines; some error bars are small on the log plot and can not be distinguished from the line.

MSD: The MSD of Schwann cells was calculated from the cell positions to quantify the observations from the cell paths (Fig. 3E and 3F). The MSD of female cells on uni-100 scaffolds was statistically higher than on uni-420 substrates (p < 0.0001), which had no increase compared to non-functionalized fibers (data reused from ref [16] with permission). On the other hand, MSD of male Schwann cells on both uni-100 and uni-420 scaffolds were comparable, and both significantly increased compared to non-functionalized groups (p_uni-100—NF < 0.0001, p_uni-420—NF < 0.0001) (data reused from ref [16] with permission).

Velocity and Speed: Velocity $(〈\overrightarrow{V_{\Vert }}〉)$ in the axis of the fiber alignment was utilized to describe the possible positive or negative directional movement over 24 h. A positive $\overrightarrow{V_{\Vert }}$ would indicate a cell’s movement in the positive x-direction. Near-zero $\overrightarrow{V_{\Vert }}$ indicates bi-directional cell migration, or equal movement in both positive and negative direction. As expected, both female and male Schwann cells had near-zero $\overrightarrow{V_{\Vert }}$ on low and high uniform concentrations with no statistical differences between concentrations or with the non-functionalized fibers, with female cell velocities: ${〈\overrightarrow{V_{\Vert }}〉}_{\text{uni-100}}=0.01\pm 0.15\mu m/\text{min}$, ${〈\overrightarrow{V_{\Vert }}〉}_{\text{uni-420}}=0.01\pm 0.16\mu m/\text{min}$ and male cell velocities: ${〈\overrightarrow{V_{\Vert }}〉}_{\text{uni-100}}=0.02\pm 0.16\mu m/\text{min}$, ${〈\overrightarrow{V_{\Vert }}〉}_{\text{uni-420}}=0.01\pm 0.16\mu m/\text{min}$ (data reused from ref [16] with permission). Overall, the individual velocities of the cells ranged from ~−0.4 to 0.4 μm/min (Fig. 3D). Perpendicular velocities $〈\overrightarrow{V_{\perp }}〉$ for both female and male cells were also near-zero (Figure S9A and S10A), but the range was much smaller at ~−0.06 to 0.06 μm/min due to topographical contact guidance. Overall cell speed for uniform substrates was also calculated to allow for comparison to the literature. For female cells, speed 〈S〉 was found to be 0.43 ± 0.12 μm/min for uni-100 and 0.29 ± 0.08 μm/min for uni-420; for male cells, speed was 0.31 ± 0.08 μm/min for uni-100 and 0.34 ± 0.09 μm/min for uni-420. All velocities and speeds are detailed in Table S2 and S3. In summary, cells on constant YIGSR moved in both the +x and −x directions along the fiber axis, with minimal movement in the perpendicular axis.

Cell speed in the direction parallel to fibers (〈S_‖〉), was calculated from the respective velocity. Female Schwann cells had the highest speed on uni-100 samples (0.37 ± 0.13 μm/min), and no differences were observed between non-functionalized (0.26 ± 0.11 μm/min) and uni-420 fibers (0.23 ± 0.08 μm/min) (Fig. 3C). In contrast, male cells exhibited the highest speeds on uni-420 scaffolds, and no differences observed between non-functionalized and uni-100 fibers (Fig. 3C). The perpendicular cell speed (〈S_⊥〉) was similarly evaluated and, as expected, was significantly lower than parallel speed for all fiber substrates (data not shown).

3.4. Cell analysis on gradient YIGSR-functionalized aligned fibers

Actin alignment: Actin alignment was assessed on the three gradient profiles. No differences in actin alignment were detected between cells at positions a, c and e on 7 or 15 pmol·(cm²·mm)⁻¹ gradient profiles for both sexes (Fig. 4A–B (female) and Fig S11 (male)). However, female cells within the 60 pmol·(cm²·mm)⁻¹ gradient (b: 83.6 ± 4.9%, c: 83.4 ± 3.5%, d: 85.8 ± 4.3%) exhibited greater actin alignment compared to cells at positions a and e (a: 74.8 ± 4.8%; e: 77.2 ± 3.8%), which were just outside of the gradient (Fig. 4F). No differences were found between male cells at positions b-d compared to male cells at positions a and e on the 60 pmol·(cm²·mm)⁻¹ gradient (Figure S11).

Fig. 4.

Schwann cell actin filaments were highly aligned on all fiber conditions, with little or no difference between concentrations. (A, B, C) Images of female Schwann cell actin on the three different gradients, A = 7 pmol·(cm²·mm)⁻¹, B=15 pmol·(cm²·mm)⁻¹, C=60 pmol·(cm²·mm)⁻¹. The images displayed are from the third quartile of the images quantified. Scale bar (lower right corner of the image) is 50 μm. (D, E) The percentage of actin alignment of female cells on 7 (D) and 15 (E) showed no statistical differences between points along the gradient or when compared to uniform scaffolds. (F) On 60 pmol·(cm²·mm)⁻¹, female cells were more highly aligned along the 6.25 mm of the gradient (positions b-d). Percentage of actin alignment was calculated using a MATLAB edge detection program and 5 images (n = 5) per biological replicate (N = 3). (G, H, I) The cell paths were tracked at each position and displayed over the 24 h capture, with G for the 7 pmol·(cm²·mm)⁻¹ gradient, H for the 15 pmol·(cm²·mm)⁻¹ gradient, and I for 60 pmol·(cm²·mm)⁻¹ gradient. Statistical differences for actin alignment were calculated using two-way ANOVA test with Tukey as a post hoc test. Error bar = standard deviation. Similar male cell results are shown in SI.

Cell trajectory: As with the uniform substrates, cell trajectories were utilized to provide an overview of the migration. Both female (Fig. 4G–I) and male (Figure S11) Schwann cell paths were similar in magnitude but started to show preferential direction for the 60 pmol·(cm²·mm)⁻¹ gradient positions b, c, d.

MSD: The overall MSD for female and male cells across gradient profiles were calculated (Figure S12). Female Schwann cells in the middle of the gradients, for all gradients, showed higher MSD than those on constant substrates or in other positions of the gradient. A similar trend was not seen for the male cells, however, and therefore we moved from MSD to velocity and speed calculations.

Velocity and Speed: Parallel velocity $(〈\overrightarrow{V_{\Vert }}〉)$ was investigated for both female and male cells on gradient fibers. $〈\overrightarrow{V_{\Vert }}〉$ of female and male cells on gradient 7 pmol·(cm²·mm)⁻¹ and gradient 15 pmol·(cm²·mm)⁻¹ were near-zero, similar to uniformly functionalized scaffolds and indicating no directional bias in migration (Figure S13). The highest gradient of 60 pmol·(cm²·mm)⁻¹ had increased directional bias up the gradient for both sexes (Figs. 5A and 5C) with a range of approximately −0.5 to 0.5 μm/min. $〈\overrightarrow{V_{\Vert }}〉$ for female cells at positions b, c, d within the gradient demonstrated significantly higher average velocities in the direction of greater YIGSR concentration, b: 0.07 ± 0.18 μm/min, c: 0.06 ± 0.15 μm/min, d: 0.06 ± 0.15 μm/min. Male cells behaved similarly within the 60 pmol·(cm²·mm)⁻¹ gradient, except at position b, where average biased velocity was non significantly different, b: 0.05 ± 0.15 μm/min, c: 0.07 ± 0.14 μm/min, d: 0.08 ± 0.13 μm/min. At positions a and e, which were outside of the applied gradient, both female and male cells demonstrated statistically lower $〈\overrightarrow{V_{\Vert }}〉$ compared to positions b-d. Similar to uniform YIGSR conditions, $〈\overrightarrow{V_{\perp }}〉$ of both female and male cells was near-zero (Figure S9 and S10) with a range of ~−0.06 to 0.06 μm/min.

Fig. 5.

60 pmol·(cm²·mm)⁻¹ gradient induced unidirectional migration. On 60 pmol·(cm²·mm)⁻¹ concentration gradients, the parallel velocity, vectorized in the axis of the fiber, is positive when moving in the direction of higher concentration, and negative when moving in the direction of lower velocity. This gradient profile induced statistical increases in parallel velocity for both female (A) and male (C) cells when cells were directly on the gradient positions of b, c, and d. This unidirectional migration was further quantified by haptotactic index, which also statistically increased at positions within the gradient (b, c, d) for both female (B) and male (D) cells.

All 〈S_‖〉 within gradients remained similar or lower than that of the best-performing uniform conditions per sex (Figures S16 and S17). On all gradients, female cell 〈S_‖〉 was reduced at all positions compared to uni-100. On the 7 pmol·(cm²·mm)⁻¹ gradient, male cell 〈S_‖〉 had no differences between the uniform YIGSR and any position on the gradient. On the 15 pmol·(cm²·mm)⁻¹ gradient, male cell 〈S_‖〉 was lower than the uni-420 YIGSR at any position on the gradient. On the 60 pmol·(cm²·mm)⁻¹ gradient, male cell 〈S_‖〉 at positions b and c was not statistically different than the uni-420; 〈S_‖〉 at c (0.26 ± 0.09 μm/min) was statistically higher than positions off the gradient (a and e). Similarly, female cells at position b (0.33 ± 0.13 μm/min) had statistically higher 〈S_‖〉 than at positions off the gradient (a and e). 〈S_⊥〉 for both sexes fluctuated across positions and gradient profiles, but all remained significantly lower than that of 〈S_‖〉. All velocities and speeds are detailed in Table S2 and S3. In summary, the steepest gradient profile impacted the speed and velocity parallel to the fibers, showing bias velocity in the direction of the gradient and higher speed.

Haptotactic Index: The haptotactic index (HI) was calculated to investigate Schwann cell directional persistence parallel to fiber alignment to the gradients. As expected, based on the minimal significance found with $〈\overrightarrow{V_{\Vert }}〉$, Schwann cells on 7 and 15 pmol·(cm²·mm)⁻¹ had a lack of directional bias at nearly all positions (Figure S18). Within the 60 pmol·(cm²·mm)⁻¹ gradient, female cells began to exhibit haptotaxis in the direction of increasing concentration along the gradient at positions b, c, and d (Fig. 5) (HI_b = 0.18 ± 0.45; HI_c = 0.18 ± 0.40; HI_d = 0.17 ± 0.40). Male cells demonstrated similar haptotaxis on 60 pmol·(cm²·mm)⁻¹, although only at positions c and d along the gradient (HI_b = 0.15 ± 0.42; HI_c = 0.20 ± 0.40; HI_d = 0.25 ± 0.38). Expectedly, the lack of a peptide gradient profile at positions a and e on the steepest gradient resulted in no haptotaxis for Schwann cells, providing a nice internal control for the migration.

4. Discussion

Polymeric fibers are explored widely for nerve regeneration applications because their topography mimics the natural matrix microenvironment, while simultaneously providing high surface area for cell attachment and proliferation [55–58]. In particular, oriented and aligned fibers have enhanced neurite outgrowth along the direction of fibers [59,60]. We and others have utilized polymer fiber-based constructs to influence cell migration [61,62], with both fiber diameter and alignment influencing the results. We have separately shown that contact guidance using aligned fiber topography supported Schwann cell axial migration [14,16] while peptide gradients of YIGSR on glass unidirectionally guided Schwann cells up the gradients [33]. Here we developed polymeric fibers that could be functionalized post-fabrication with constant or gradient YIGSR to characterize if Schwann cell response was synergistic to both cues.

The topographical cue in this report was aligned 1.2 μm PCL-based fibers. PCL is commonly used in implantable medical devices for various tissues and pathologies [63–66]. However, the utility of PCL has been limited by its high hydrophobicity and poor wettability, making it an unattractive surface for cell adhesion and growth [67]. Surface-modified materials with bioactive conjugates increase cell attachment and modulate hydrophilicity [68–71]. Previously, allyl-containing (14 mol%) functional PCL copolymer was synthesized using 6-allyl-ε-caprolactone to provide post-fabrication functionalization using thiol-ene click chemistry [38]. In this study, the mol% of allyl monomer was increased to 25% to increase the range of covalent concentration gradient profiles. The fiber diameter was selected based on several previous studies where micron-sized fibers significantly enhanced Schwann cell elongation and migration. Aligned PCL fibers of ~1 μm aligned human Schwann cells, with cells following individual fibers more closely than fibers that were randomly aligned [13]. Differentiation toward Schwann cells was increased on 0.6 and 1.6 μm fibers compared to 0.16 μm [72], and separately on ~1 μm fibers relative to 0.5 μm fibers [73]. Actin alignment, cell migration along the fiber axis, and Schwann cell velocities [14] were all enhanced on ~0.9 μm aligned fibers over 0.3 μm aligned fibers, and 1.2 μm aligned fibers were found to further improve Schwann cell actin alignment and migration parameters over 0.9 μm fibers [16], but differences were more modest between 1.2 μm and 1.8 μm fibers. In addition, sex-based differences in Schwann cell migration were also muted on 1.2 μm fibers [16]. In direct comparison, nano-sized features enhanced proliferation and adhesion while micron-sized features increased cell elongation, which is important for directing cell migration [74]. Others have used larger fiber diameters, including 5 μm and 8 μm fibers, and showed that primary rat Schwann cells possessed higher metabolic activity and viability on 5 μm polyhydroxyalkanoate (PHA) fibers relative to others studied [75]. Overall, these results indicated that fibers in the 1–5 μm size range have the potential to best support Schwann cell migratory responses, and 1.2 μm fibers were selected chosen for this study to further reduce sex-based differences.

To investigate synergy between contact guidance and haptotaxis, we quantified the peptide concentrations tethered to the fibers. Through characterization of the gradient profiles, we demonstrated thiol-ene reactions yielded a linear correlation between photomask transparency, UV light intensity, and concentration of immobilized peptide. For example, the photomask that resulted in 60 pmol·(cm²·mm)⁻¹ gradients was designed to target a slope four times the profile steepness of 15 pmol·(cm²·mm)⁻¹ and the resulting functionalization pattern exhibited a slope that was four times as steep. However, profile steepness was likely attenuated due to physical adsorption of the peptide in the low concentration end of the pattern. While a near-zero concentration was expected within Region 1 of the 60 pmol·(cm²·mm)⁻¹ gradient, a concentration of 161 ± 66 pmol/cm² was measured, indicating further optimization of post-functionalization washing processes may be needed to ensure complete removal of non-covalently bonded peptide. Other thiol-ene clicked fibers or functionalized peptide fibers reported higher amounts of peptide bound, up to 5 orders of magnitude higher, but the data detailing the relative concentrations (e.g., μmol/mass of scaffold) was not reported [76], peptides were not patterned as a gradient [77], or scaffolds did not contain fibers [32]. Specifically for YIGSR peptide, its controlled release was found to increase migratory rates of Schwann cells on ovalbumin ridge structures [78] and we previously reported that patterning the peptide in a gradient directionally biased migration [33] on flat, non-topographical, substrates. Therefore, by demonstrating that YIGSR density on fibers could be spatially controlled on aligned fiber scaffolds, we could further provide a detailed assessment of synergy between fiber topography and precise gradients.

With the development of the gradient fiber substrates, we next measured the Schwann cell response. Uniaxial elongation of cells and actin alignment has been linked to cells’ ability to migrate [79,80]. Therefore, we first examined the morphology of Schwann cells after 24 h on 1.2 μm fibers, with actin showing >80% alignment with the fibers regardless of the peptide concentration or gradient profile. This level of alignment was higher than our previous work on flat substrate with YIGSR gradient profiles, where actin was ~60% aligned with the direction of the concentration change [33], but similar to the actin alignment on 0.9 – 1.8 μm fibers [16]. Others have quantified this elongated morphology through measurement of the cell aspect ratio, with cells that are highly elongated migrating faster with greater persistence [80]. Here we recorded some samples with significantly increased actin alignment, specifically the 60 pmol·(cm²·mm)⁻¹ profile positions for female cells. While female cell actin alignment on high gradients correlated with increased cell bias migration, male cells did not have significantly increased actin alignment even though they also had increased bias migration. These results indicated the possibility of a simple quantitative assessment, such as percent actin alignment, may be a strong indicator for biased migration when comparing substrates, but the correlation is imperfect, likely due to the high overall alignment due to the fibers alone. While we and others have proposed this metric, the quantification of the actin alignment remains limited in most reports. The ability for percent actin alignment to differentiate contact guidance and peptide concentration profiles for some cells showed the potential power of the simple assessment. Further work is required to determine why the male cells did not show a correlation.

Migration studies were defined by strict criteria to investigate cell-substrate correlations using individual cell-tracking data. We were interested in individual cells migrating rather than collective migration because endogenous nerve repair occurs via individual cells invading a gap. To reduce physiological variance within a cell population, more than 100 cells per sex were tracked and considered per condition. With individual cell migration, minimum displacements are typically set to differentiate motile cells [81,82], and a displacement of one to two times the cell diameter is a common minimum. In our previous study, we utilized a displacement requirement of 100 μm, which was approximately twice the observed average diameter of the Schwann cells [33]. While it was crucial to be consistent with our previous study for comparison, this criterion may exclude ‘slow’ moving cells. Therefore, we additionally imposed a displacement requirement of 50 μm to compare results to those calculated using the 100 μm criterion (analysis not shown). Across all substrates, 96% of female cells migrated over 50 μm and 65% exceeded 100 μm, while 98% and 72% of male cells moved over 50 μm and 100 μm, respectively. At all positions on YIGSR gradients profiles, parallel velocity $(〈\overrightarrow{V_{\Vert }}〉)$ and haptotactic index (HI) of female and male Schwann cells were not statistically different between 50 μm and 100 μm conditions, showing the directional bias was not an analytical artifact of displacement criteria.

Considering these migratory criteria, we first examined the effects of uniform YIGSR fibers at different concentrations. While 〈S_‖〉 on the uni-100 substrate increased for female cells over non-functionalized fibers, the addition of more YIGSR did not further improve 〈S_‖〉 for female cells. For male cells, 〈S_‖〉 on the uni-420 substrate was increased, while it was not different on uni-100 than non-functionalized fibers. These differences were not unexpected, as overall Schwann cell speed 〈S〉 on flat laminin substrates was higher for female cells than male [16]. Schwann cells utilize focal adhesions in binding to YIGSR [15] and therefore, the specific binding to YIGSR provides specificity to the cells in their migration over no specific adhesion proteins. Unfortunately, most migration studies do not quantify the amount of peptide or protein on the surface making it difficult to directly compare the responses relative to concentration. We previously found that the speed component parallel 〈S_‖〉 to fibers increased for Schwann cells, but the overall cell speed 〈S〉 was reduced for cells on fibers compared cells on flat laminin substrates [16]. With the addition of YIGSR, the overall speed 〈S〉 on fibers increased to the level of the overall speed reported on flat laminin substrates for both male cells and female cells [16]. This result further justified the testing of YIGSR gradients to determine if they could enhance directional migration.

Concentration gradient profiles of YIGSR were studied to induce haptotaxis in Schwann cell migration and synergize with contact guidance presented by the fiber topography. Slopes of YIGSR peptide concentration on glass previously studied were reduced compared to the slopes of YIGSR concentration, with ~1.75, 3.5, and 7 pmol·(cm²·mm)⁻¹ [33] on glass substrates compared to 7, 15, and 60 pmol·(cm²·mm)⁻¹ on fibers. While all of the reported gradients of YIGSR promoted some level of haptotaxis on glass [33], the gradients of 7 and 15 pmol·(cm²·mm)⁻¹ did not induce haptotaxis in Schwann cells on fibers. For dendritic cells, optimal chemokine gradient profiles for haptotaxis were exponential profiles, similar to those found in vivo [83]. Linear profiles were also found to be haptotactic if a threshold was exceeded [83], which agrees with the results presented here. The HI values were within a similar range to other studies across different cell types [27,83,84]. More specifically, HI values fell within the range of 0.2 – 0.3, consistent with biased Schwann cell migration on a 3.5 pmol·(cm²·mm)⁻¹ YIGSR gradient on flat glass substrates [33]. Gradients of IKVAV, another laminin peptide, were found to induce directed migration but the concentration profiles were significantly higher than those reported here [32]. Therefore, while haptotactic effects can be used to support unidirectional migration in addition to contact guidance, steeper gradients than what was necessary on flat surfaces were needed. Importantly, the highest gradients reported here enhanced the bias migration for both male and female cells on fibers, supporting the synergy between topography and bioactive gradients.

5. Conclusion

In this study, we utilized an allyl-containing polycaprolactone copolymer to create highly aligned fiber scaffolds. Post-fabrication, the fibers were functionalized with the bioactive peptide sequence YIGSR. Thiol-ene click chemistry was used for covalent tethering, enabling precise spatial control over functionalization densities. This discrete functionalization allowed for the quantification of Schwann cell migratory response to YIGSR profiles in a concentration and gradient-dependent manner. Uniform YIGSR on fibers enhanced Schwann cell motility over non-functionalized fibers; female Schwann cells exhibited peak migration speeds at lower YIGSR densities, while male cells responded more robustly to higher concentrations. Haptotaxis index confirmed the effects of gradients on fiber topographies by imparting bias migration. While shallow gradients of YIGSR did not reliably bias Schwann cell migration on aligned fibers, the steepest concentration gradient (60 pmol·(cm²·mm)⁻¹) successfully induced haptotaxis, guiding both female and male Schwann cells toward regions of higher peptide density. Together, these findings highlight the potential of engineering fiber-based scaffolds with precisely controlled biochemical and topographical cues to direct Schwann cell behavior. Such strategies may ultimately enhance axonal regeneration by guiding neuronal components across segmental nerve defects. Future in vivo studies are warranted to translate these principles into clinically effective nerve repair scaffolds.

Supplementary Material

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.actbio.2026.02.052.