Local Release of Paclitaxel from Aligned, Electrospun Microfibers Promotes Axonal Extension

Abstract

Traumatic spinal cord injuries ultimately result in an inhibitory environment that prevents axonal regeneration from occurring. A low concentration administration of paclitaxel has been previously shown to promote axonal extension and attenuate the upregulation of inhibitory molecules after a spinal cord injury. In this study we incorporated paclitaxel into electrospun poly(l-lactic acid) (PLA) microfibers and established that a local release of paclitaxel from aligned, electrospun microfibers promotes neurite extension in a growth-conducive and inhibitory environment. Isolated dorsal root ganglion cells were cultured for five days directly on tissue culture polystyrene surface, PLA film, random, or aligned electrospun PLA microfibers (1.44 ± 0.03 µm) with paclitaxel incorporated at various concentrations (0 – 5.0% w/w in reference to fiber weight). To determine the effect of a local release of paclitaxel, paclitaxel-loaded microfibers were placed in CellCrown™ inserts above cultured neurons. Average neurite extension rate was quantified for each sample. A local release of paclitaxel maintained neuronal survival and neurite extension in a concentration-dependent manner when coupled with aligned microfibers when cultured on laminin or an inhibitory surface of aggrecan. Our findings provide a targeted approach to improve axonal extension across the inhibitory environment present after a traumatic injury in the spinal cord.

1. Introduction

Spinal cord injury (SCI) is a detrimental condition that currently affects approximately 250,000 patients with an additional 12,000 new cases each year in the US.^[1] SCIs typically result in functional deficiencies below the injury site such as a loss of bladder and locomotor control. An SCI is characterized by an initial, physical insult followed by a secondary, deleterious biological cascade that culminates in the formation of a cystic cavity lined by a primarily astrocytic scar (reactive gliosis).^{[2, 3]} This reactive gliosis not only physically obstructs axonal regrowth and regeneration, but also upregulates and secretes inhibitory compounds such as chondroitin sulfate proteoglycans (CSPGs) that further inhibit axonal extension.^[4] After an SCI occurs, this inhibitory upregulation and deposition is sustained for up to eight weeks after the initial injury.^[5] Although this complex formation is localized to the central nervous system (CNS), adult CNS axons have the inherent capability to regenerate, but are limited by this inhibitory environment.^{[4, 6]} Therefore, the administration of a growth-promoting scaffold could promote axonal regeneration through this injury site.

Systemic drug delivery approaches have had restricted access to the spinal cord due to the limited diffusion of the blood spinal cord barrier (BSCB).^[7] To improve spinal cord tissue regrowth, various localized drug delivery methodologies using hydrogels,^{[8, 9]} nanoparticles,^{[10, 11]} and foam scaffolds^{[12, 13]} as the delivery carriers have been proposed to overcome this challenge by incorporating pro-regeneration agents, such as neurotrophic growth factors, directly into the injury site. In order to advance the capability of a local administration to promote spinal cord tissue repair, additional treatments, such as methods to incorporate hydrophobic treatments, that can maintain delivery for at least two weeks must also be established. Furthermore, recent SCI therapies have been leveraging the advantages of multiple delivery platforms to target the multi-faceted biological response that is induced after an SCI.^[14–17]

Paclitaxel (PTX) is a hydrophobic molecule and mitotic inhibitor that mechanistically stabilizes microtubule formation. After an axotomy, the DLK pathway is activated, which prevents the rapid depolymerization of microtubules and retraction bulb formation, which in turn supports and maintains microtubule stabilization and axonal extension.^{[18, 19]} Clinically, paclitaxel is administered as an anticancer agent (taxol) at high concentration (µM) dosages. Additionally, because of paclitaxel’s poor solubility, it is administered in a toxic solvent, Kolliphor® EL, that can induce peripheral neuropathy. Studies incorporating a low concentration administration of paclitaxel have shown that paclitaxel can promote axonal extension in vitro^[20] as well as mitigate the production of CSPGs and other inhibitory molecules at the glial scar in vivo after an SCI.^[21] However, paclitaxel delivery systems have been limited due to paclitaxel’s poor solubility, non-specific binding, and large molecular size that prevents it from traversing the BSCB.^[22] One delivery approach has circumvented these issues by delivering paclitaxel directly into the injury site via an osmotic mini-pump.^[21] However, these surgically implanted pumps can be dislodged, induce infections, and have a limited loading capacity, which limits their translational applications. Therefore, biomaterial-based drug delivery systems such as microspheres and microfibers present an additional platform to locally deliver PTX and promote spinal cord tissue repair.

The outcomes of biological approaches to improve traumatic SCI have been limited due to several challenges related to delivery of biologics, such as removal of active agents through cerebrospinal fluid flow, poor survival of the transplanted cells, and limited control over the delivery duration of therapeutic agents. Engineered technologies have increased in frequency as a preferred treatment mechanism for SCI tissue repair due to their ease of manufacturing and reproducibility.^[23] Biomaterials-based approaches for spinal cord tissue repair have been shown to improve neuronal regeneration^{[8, 9, 24]}, but have resulted in limited functional recovery primarily due to their inability to attenuate or target the formation of the glial scar.^{[24, 25]} Effective approaches to attenuate the inhibitory components and simultaneously stimulate axonal extension are needed to significantly improve regeneration outcomes in SCI repair.

Polylactic acid (PLA) aligned, electrospun microfibers have been shown to provide a directional guidance that enhances axonal extension in vitro^{[17, 26, 27]} and decrease the cystic cavity volume in vivo after an SCI.^{[25, 28]} While this system does not directly overcome the inhibitory environment that prevents axonal regeneration and extension after an SCI, paclitaxel has been previously incorporated in poly(lactic-co-glycolic acid) (PLGA) electrospun microfibers as a targeted anti-cancer therapy,^[29] which could address this issue. In this study, we encapsulate PTX into PLA microfibers and test the hypothesis that a local release of PTX from aligned fibers promotes axonal extension compared to PLA fibers alone or soluble PTX dosing, and characterize the ability of axons to overcome an inhibitory environment similar to that seen in a traumatic SCI, when guided by aligned, PTX-releasing fibers.

2. Results

2.1 Paclitaxel incorporation and release from aligned microfibers

After a spinal cord injury occurs, the secondary, inhibitory biological cascade commences and is sustained for weeks to months after the initial injury.^{[5, 30]} We, therefore, wanted to first develop a platform that could sustain a treatment release throughout this duration and fabricated microfiber substrates by electrospinning approximately 1 milligram of an 8% PLA polymer solution onto glass coverslips. Paclitaxel was incorporated into the polymer solution at an expected loading concentration, assuming a 100% loading efficiency, from 0 – 5% (mass paclitaxel/mass microfiber). Electrospun fibers were deposited on an aluminum disk that was either rotating for aligned fibers or stationary for random-oriented fibers. Fiber diameter and alignment was determined with scanning electron micrographs (Figure 1, Table 1). Approximately 100 individual fiber diameters were measured using ImageJ software and averaged per paclitaxel incorporation condition. The fiber diameter remained between 1.37–1.51 µm for all paclitaxel incorporation groups and median fiber alignment was quantified between 2–8 degrees for aligned microfibers (Table 1). Paclitaxel incorporation into aligned microfibers does not alter the diameter or alignment of these fibers.

Figure 1.

Incorporating paclitaxel in electrospun PLA microfibers does not alter fiber morphology, diameter, or alignment. Aligned PLA microfibers were produced by electrospinning a PLA solution with increasing expected concentrations of paclitaxel 0% (A), 0.05% (B), 0.1% (C), 0.5% (D), 1.0% (E), 5.0% (F) (w/w 1 milligram of PLA) and visualized by scanning electron microscopy. Scale bar: 5 µm.

Table 1.

Characterization of Aligned Fiber Diameter with Increasing Concentrations of Paclitaxel.

Expected Paclitaxel Loading (%)	Paclitaxel Loading Efficiency	Actual Paclitaxel Loading (%)	Fiber Diameter (µm ± STD)	Fiber Alignment (° ± STD)
0 (PLA)	N/A	0	1.41 ± 0.29	4.79 ± 4.36
0.05	0.394	0.02	1.56 ± 0.45	3.21 ± 2.47
0.10	0.424	0.04	1.39 ± 0.37	7.56 ± 6.84
0.50	0.507	0.25	1.50 ± 0.51	6.32 ± 3.83
1.00	0.582	0.58	1.36 ± 0.43	2.32 ± 1.99
5.00	0.652	3.26	1.44 ± 0.41	4.45 ± 4.16

Paclitaxel incorporation efficiency was determined by dissolving the electrospun microfibers into an organic solvent solution and processing them through an ultraviolet detector and maximum absorbance was quantified for each sample (Table 1). Paclitaxel incorporation was determined to be 39–65% for groups with an expected loading level of paclitaxel from 0.05–5.0%, respectively.

Release rate and duration of paclitaxel were quantified by submerging 1.5 mg of electrospun microfibers into 500 mL of phosphate buffer solution at 37°C for 84 days. Biweekly aliquots of the supernatant were removed and processed by HPLC for each sample (n = 3). The cumulative release of paclitaxel from electrospun microfibers was well maintained over 84 days in the higher expected loading level groups (0.5–5%; Figure 2A). Paclitaxel release from the microfibers is inversely correlated to the percent release of paclitaxel depending on the loading-level of paclitaxel in the microfibers. As such, 7–97% of the paclitaxel loaded into the fibers was released from the microfibers with an expected paclitaxel loading between 5–0.05% paclitaxel, respectively (Figure 2B).

Figure 2.

Paclitaxel release from PLA microfibers can be controlled and tuned based on the original loading level of paclitaxel. PLA microfibers were placed in PBS and aliquots were analyzed by HPLC over 84 days and the cumulative (A) and fraction (B) release were calculated. As the initial loading of paclitaxel increased, the release rate increased, and the fraction released increased at a slower rate than fibers with a lower incorporation of paclitaxel.

2.2 Effect of paclitaxel release from microfibers on axonal extension

To demonstrate that paclitaxel retains its activity after incorporation into microfibers and the effect of this local release mechanism on neuronal extension, dorsal root ganglion (DRG) neurons were isolated from P5 rat pups and cultured for five days on aligned electrospun microfibers with increasing concentrations of incorporated paclitaxel. Importantly, paclitaxel released from the aligned microfibers modulated neurite extension (Figure 3). Furthermore, this extension occurred in a paclitaxel concentration-dependent manner. As the concentration of paclitaxel increased, neurite extension decreased due to an over-stabilization of microtubules (Figure 3D). However, lower concentrations of paclitaxel (0.05% PTX/PLA fiber, w/w) promoted a greater and faster extension of DRGs than PLA fibers alone (p < 0.01) as shown in Figure 3E.

Figure 3.

Paclitaxel release from aligned microfibers modulates neurite extension as visualized by immunostaining for neurofilament. Neurite extension was quantified from isolated DRGs cultured on PLA fibers only (A), 0.02% PTX (B), 0.58% (C), and 3.26% (D). A small incorporation of paclitaxel (0.02%) promoted a significantly greater increase in neurite extension than fibers without paclitaxel and fibers with higher concentrations of paclitaxel (E). Scale bar: A–D, 100 µm. (p < 0.001, n = 6 – 10).

In order to determine the importance of a local release of paclitaxel from electrospun microfibers directly in contact with extending axons, sheets of microfibers loaded with various concentrations of paclitaxel were placed in CellCrown™ inserts in cell culture media above isolated DRG neurons (Figure 4A). DRGs were simultaneously cultured directly on either TCPS, aligned PLA microfibers, or paclitaxel-loaded aligned microfibers. Released paclitaxel from microfibers placed in the inserts does not significantly promote neurite extension when neurons are cultured on TCPS or aligned PLA microfibers (Figure 4D). Additionally, cells cultured directly on paclitaxel-releasing aligned microfibers induced maximum neurite extension (Figure 4D).

Figure 4.

A local delivery of paclitaxel is needed to promote significant DRG neurite extension. Paclitaxel-loaded PLA microfibers were placed in CellCrown™ inserts (A) and suspended above cells cultured on either TCPS (B) or aligned PLA microfibers (C). A modest increase in neurite extension is observed when paclitaxel is released from paclitaxel-loaded microfibers suspended in CellCrowns™ above DRG neurons cultured on TCPS and PLA microfibers alone, but the maximum neurite extension only occurred when the cells were directly in contact with the paclitaxel-loaded microfibers (D). Scale bar: B–C, 40 µm. (p < 0.001, n = 7 – 10).

To examine the importance of the surface topography and alignment of a PLA substrate, DRG neurons were cultured on paclitaxel-incorporated thin films and random-oriented microfibers. A random microfiber orientation was constructed by electrospinning a PLA solution onto a stationary, grounded surface. As an additional control, a PLA thin film was casted onto glass coverslips with an approximate thickness of 50 microns. DRG neurons were cultured on paclitaxel-loaded random-oriented microfibers and films (Figure 5A–D). This local release of paclitaxel enhanced neurite extension, but not significantly more than films and random-oriented fibers without paclitaxel (Figure 5E). These results confirm that a local release of paclitaxel from aligned PLA microfibers was most effective to promote neurite extension (Figure 5E). Thus a local release of paclitaxel coupled with and from an aligned microfiber matrix is necessary to promote neurite extension in a synergistic and concentration-dependent manner.

Figure 5.

The mechanism and surface topography of the local delivery of paclitaxel determines DRG neurite extension. DRG neurons were cultured for five days on PLA films (A) and randomly oriented fibers (C) that were loaded with and without paclitaxel. Although a slight increase in neurite extension was observed among these groups (B, D, E), significant neurite extension only occurred when the microfibers were aligned and coupled with the local release of paclitaxel (E). Scale bars: A, C, 20 µm; B, D, 100 µm. (p < 0.001, n = 9 – 10).

2.3 Neuronal viability after paclitaxel administration

Previous studies have shown that paclitaxel administration can induce neuropathy and neuronal toxicity. Therefore, in order to assess the survival of DRG neurons after five days in culture, an alamarBlue Cell Viability Assay was conducted. Isolated DRG neurons were cultured on either TCPS or PLA microfibers with increasing concentrations of paclitaxel (0, 0.02, 0.58%). Measuring the fluorescence of the samples periodically over five days, yielded no significant difference between the normalized activities between the paclitaxel-loaded groups. (Figure 6). Therefore, incorporating paclitaxel into PLA microfibers does not inhibit neuronal survival after five days in vitro.

Figure 6.

Local paclitaxel administration does not affect neuronal viability. Isolated DRG neurons were cultured on tissue-culture polystyrene or PLA microfibers with various concentration of paclitaxel 0, 0.02%, or 0.58% for five days. Neuronal viability was calculated from an alamarBlue assay (C), normalized against the tissue-culture polystyrene group, and shown to be insignificant between all groups. (n = 5)

2.4 Confirmation of axonal extension mechanism from a local release of paclitaxel

Low concentrations of paclitaxel have been shown to induce microtubule stabilization in order to maintain axonal extension. In order to ensure that a local release of paclitaxel does not induce a novel neurite extension mechanism, 250 nM of nocodazole, a molecule that destabilizes microtubule formation, was added to DRG neurons cultured on PLA microfibers with and without paclitaxel. After five days of culture the neurite extension that was previously seen due to a local release of paclitaxel is abolished and there is no significant difference in neurite extension among the groups (Figure 7). Because this effect is ablated after the introduction of nocodazole, a local release of paclitaxel still activates the DLK pathway, which is necessary for microtubule stabilization and axonal extension.

Figure 7.

A blocking of microtubule stabilization ablates the neurite extension from paclitaxel. To ensure that this paclitaxel delivery mechanism maintains neurite extension via microtubule stabilization nocodazole, a microtubule-destabilizing agent, was added to DRG neurons cultured on PLA microfibers loaded with or without paclitaxel. Once this microtubule stabilization is blocked, the previous axonal extension due to the local release of paclitaxel no longer occurs (n = 3 – 6).

2.5 Neurite extension in an inhibitory environment

After an SCI occurs, inhibitory factors such as chondroitin sulfate proteoglycans are upregulated at the site of injury and inhibit axonal extension across the site of injury. Aggrecan is a CSPG that has been shown to inhibit axonal extension in vitro. Five hundred micrograms of aggrecan was coated on surfaces containing aligned PLA microfibers alone and PLA microfibers incorporated with 0.02% PTX. As expected, aggrecan diminished neurite extension among all DRG culture conditions tested (Figure 8). However, a local release of paclitaxel promoted a significant increase in neurite extension than PLA microfibers alone under these inhibitory conditions (p < 0.001, Figure 8). Consequently, paclitaxel release from aligned microfibers promotes neurite extension in an inhibitory environment similar to that seen after an SCI (p < 0.005).

Figure 8.

A local release of paclitaxel from aligned PLA microfibers significantly promotes neurite extension under inhibitory conditions similar to those seen after a spinal cord injury. Isolated DRG neurons were cultured on an uncoated surface, laminin only, or laminin/aggrecan-coated tissue culture polystyrene, aligned PLA microfibers only, or aligned paclitaxel-loaded PLA microfibers. Paclitaxel-loaded microfibers significantly promoted neurite extension than PLA microfibers under both a growth-conducive environment and inhibitory environment (*p < 0.001, **p < 0.005, n = 5 – 10).

3. Discussion

Following SCI, axonal extension is limited due to the upregulation and migration of inhibitory factors and cells to the site of injury.^{[4, 6]} In this study, we demonstrate that paclitaxel can be incorporated into electrospun PLA microfibers without modifying their alignment and fiber shape. A local release of paclitaxel from electrospun fibers, in a concentration-dependent manner, remains active to promote neurite extension after release by stabilizing microtubule formation. Additionally, this release of paclitaxel from microfibers can be controlled and tuned. Furthermore, incorporated paclitaxel in PLA microfibers can maintain neuronal survival and promote neurite extension under growth-conducive conditions as well as inhibitory conditions similar to that after an SCI.

We first wanted to determine if the release of paclitaxel from electrospun microfibers could be modified and controlled. To do this, we varied the amount of paclitaxel that we loaded into the PLA polymer solution and measured paclitaxel release from these fibers over 12 weeks and found that due to the slow diffusion of paclitaxel from these microfibers, larger incorporations of paclitaxel promoted a faster and larger release. In addition, modifying the fiber density can modify the amount of released paclitaxel as well as provide a more uniform growth conduit for axons to grow on as well. These findings are necessary to modify this platform for future animal models of SCI and translational applications.

Furthermore, paclitaxel incorporated into electrospun microfibers remained active in promoting axonal extension. To establish this, we used a well-characterized model of neurite extension with isolated dorsal root ganglion neurons cultured on approximately 1-milligram of PLA microfibers loaded with various concentrations of paclitaxel. We found that this administration technique retained paclitaxel’s axon extension modulation in a concentration-dependent and synergistic effect. By removing this direct release to cells, axons lost their ability to maintain their extensions. Even with the same PLA polymer solution, if the DRG neurons were cultured on this solution (film) or with a similar topography (random fiber), the local release of paclitaxel only slightly increased axonal extension, but not significantly. We hypothesize that this synergistic effect is due to the aligned fibers promoting axonal elongation coupled with the local release of paclitaxel that further maintains these extensions to maximize axonal growth. Only the local release coupled with the directional guidance benefits of aligned microfibers promoted neurite extension under these conditions showing that this synergistic effect is necessary for maximum neurite extension.

Although neurite extension under a growth-conducive environment is essential for understanding this mechanism, it is not the environment present after an SCI. Once an SCI occurs, various inhibitory factors are present or upregulated that prevent axonal extension from occurring such as chondroitin-sulfate proteoglycans.^[2–5] To properly mimic the inhibitory effects present after a spinal cord injury, dorsal root ganglion neurons were cultured on an inhibitory substrate of aggrecan. We established that this platform of a local release of paclitaxel can overcome this inhibitory environment and promote neurite extension.

Microtubule stabilization is integral for axonal polarization, extension, and formation during neurite development as well as after an axotomy.^{[18–20, 31]} Throughout axonal elongation, various intracellular components such as microtubules regulate the direction and formation of growth cones.^{[32, 33]} Paclitaxel prevents the formation of retraction bulbs after a lesion by polymerizing and stabilizing disorganized microtubules into an aligned morphology.^{[18, 34]} However, paclitaxel administration must be controlled to prevent neurite growth inhibition by an over-stabilization of microtubules^[35,36] as well as prevent neuronal toxicity.^[37,38] This study mitigates these concerns by administering paclitaxel at lower concentrations than those used in previous studies.^[21]

Previous approaches of paclitaxel administration for anti-cancer treatments have involved osmotic mini-pumps, hydrogels, and nanoparticles.^{[21, 39, 40]} However, electrospun microfibers provide a unique platform for spinal cord injury applications due to their lack of incorporating paclitaxel in a neuropathic solvent, directed growth-permissive scaffold to promote axonal extension across the injury site, biodegradable platform, and long-term treatment release. A low concentration administration of paclitaxel has previously been shown to promote neurite extension.^{[20, 21]} By incorporating paclitaxel into aligned microfibers, neurons receive a controlled, low concentration dosage of paclitaxel that when coupled with a growth-permissive scaffold promoted axonal extension by stabilizing microtubule formation. Additionally, by releasing the paclitaxel locally to the cultured neurons, we hypothesize that the paclitaxel is directly uptaken by the neurons rather than dissolved into the media in other administration techniques. Future studies will elucidate this mechanism and leverage the multiple pathways that regulate microtubule stabilization. Furthermore, this platform can easily be incorporated and modified for in vivo spinal cord injury applications by directly placing a 2-dimensional sheet on top of an injured spinal cord or directly implanting a wrapped sheet of aligned microfibers into a conduit as shown in previous spinal cord injury applications.^[25] In addition, other therapeutic applications that require a localized treatment, such as anticancer or antibiotic delivery, could leverage the advantages of this platform to maximize potency and limit negative side-effects.^{[29, 41]}

4. Conclusions

Although preliminary studies have established that paclitaxel can promote neurite extension at low concentrations, we have shown the necessity of paclitaxel release from aligned, electrospun microfibers on neurite extension. Moreover, this study demonstrated this requirement of both components to promote a synergistic effect that maximally promotes neurite extension and maintains neuronal survival. In this study, we established that paclitaxel remains active after incorporation into PLA electrospun microfibers. The release of paclitaxel from these fibers is controllable and tunable for a prolonged period of time. The coupling of this release from aligned fibers enhances a greater neurite extension than either component alone under both a growth-conducive and inhibitory environment. Our findings show that aligned, electrospun microfibers incorporated with a low concentration of paclitaxel can provide a versatile, controllable, and readily tunable administration technique to provide a local, sustained, and tunable delivery of paclitaxel for therapeutic applications after a traumatic spinal cord injury.

5. Experimental Section

PLA Films

Glass coverslips (15 mm diameter, Fisher Scientific, Waltham, MA) were coated with an approximately 50-micron layer of an 8% w/w poly(l-lactic acid) (PLA; Grade 6201D, $\overline{M_w}=78\mathrm{k}\mathrm{D}\mathrm{a},\overline{M_n}=48\mathrm{k}\mathrm{D}\mathrm{a}$, NatureWorks LLC, Minnetonka, MN) solution.

Electrospinning PLA Fibers

PLA fibers were produced by electrospinning an 8% w/w polymer solution of PLA in chloroform and dimethylformamide (99:1 w/w) with increasing concentrations of paclitaxel (0, 0.025, 0.05, 0.1, 0.5, 1.0% w/w in reference to fiber weight; LC Laboratories, Woburn, MA). For aligned fibers, this solution was administered from a syringe for 2 h at a rate of 1.1 mL/h with a 10 kV electrical potential applied to the needle. This solution was applied onto coverslips grounded to a disc collector rotating at 1450 rpm from a separation distance of 6 cm (Figure 1). For random fibers, the solution was applied for 30 min at a rate of 0.65 mL/h with an 11 kV applied electrical potential with a 10 cm separation distance (Figure 5C). To evaporate any residual organic solvent, the coverslips were placed in a fume hood overnight. Sterilization occurred by exposing the samples to ultraviolet radiation in a biosafety cabinet for 45 min.

Paclitaxel Encapsulation Efficiency and Release

To determine the encapsulation efficiency of paclitaxel in the electrospun fibers, paclitaxel-loaded fibers (1 mg) were dissolved in chloroform (1 mL). Then, acetonitrile/water mixture (85:15, 9 mL) was added to the solution. A nitrogen stream was added to evaporate the chloroform at room temperature. The resulting solution was then processed through an absorbance microplate reader at 227 nm (Molecular Devices, Sunnyvale, CA), and maximum absorbance was quantified. The release of paclitaxel from the fibers was determined by placing coverslips of paclitaxel-loaded fibers (1.5 mg) in PBS (500 mL) and obtaining aliquots of the solution once every three days. These samples were analyzed by high-performance liquid chromatography (HPLC; Waters, Milford, MA) with a mixture of acetonitrile and water (85:15, v/v) as the mobile phase. Fifty microliters of the samples were passed in the mobile phase through a C-18 column (Agilent Technologies, Santa Clara, CA) at a rate of 1 mL/min. The column effluent was detected at 227 nm using an ultraviolet detector.

Cell Culture

All animal procedures were conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee (IACUC) at The Johns Hopkins University. Dorsal root ganglion (DRG) neurons were isolated from P5 Sprague Dawley rat pups (Charles River, Wilmington, MA) and enzymatically digested in 0.1% collagenase type I (Sigma-Aldrich, St. Louis, MO) for 30 min and then in 0.25% trypsin (Invitrogen) for 45 min at 37°C. Enzymatic digestion was stopped by adding trypsin neutralizing solution (Invitrogen), and cells were recovered with centrifugation for 8 min at 1000 rpm at 4°C. Dissociated neurons were filtered through a 40-µm filter onto poly(l-lysine) (100 mg/ml, Sigma-Aldrich) and laminin-coated (10 µg/mL, Invitrogen) film, random-oriented, or aligned fiber samples for five days in neurobasal medium (Invitrogen, Eugene, OR) with B-27 supplement, L-glutamine, and penicillin-streptomycin. This medium was supplemented with decreasing concentrations of Cytarabine (Sigma-Aldrich) and NGF (Invitrogen). Cells were seeded at a density of 7,500 cells/cm². To assess the mechanism of neurite extension, some of the groups were supplemented with Nocodazole (250 nM, Sigma-Aldrich) at 6 – 24 h after cell seeding.

Transwell Culture with CellCrowns™

Sheets of random-oriented PLA microfibers (2 mg) with and without paclitaxel were incorporated into CellCrowns™ (Scaffdex Ltd., Tampere, Finland) according to the manufacture’s instructions. These CellCrowns™ were then placed over the cultured DRG neurons growing directly on tissue culture polystyrene (TCPS) or aligned PLA microfibers without paclitaxel.

Inhibitory Cell Culture System

A solution of aggrecan (200 µg/mL, Sigma-Aldrich) with laminin (10 µg/mL) was used to coat various surfaces. These concentrations were determined by observing neurite extension of DRG neurites after varying the concentration of aggrecan (50–700 µg/mL). As a negative control, DRG neurons were cultured on uncoated surfaces as well.

AlamarBlue Cell Viability Assay

Four hundred microliters of a ten percent alamarBlue (Invitrogen) in DRG media solution was added to isolated DRG neurons cultured at 20,000 cells/cm² on TCPS or aligned PLA microfibers with 0, 0.02, or 0.58% paclitaxel. After 4 hours of incubation, the plates were exposed to an excitation wavelength of 540 nm, and the emission of 590 nm was recorded in a plate reader (Biotek, Winooski, VT), repeated two times throughout the cell culture, and the results were normalized to TCPS growth.

Immunocytochemistry

Cells were fixed for 1 h with 4% paraformaldehyde (PFA) and blocked in PBS-5% normal goat serum (Sigma-Aldrich) and 0.1% Triton-X (Sigma-Aldrich). DRG cells were incubated overnight at 4°C with a chicken anti-neurofilament antibody (NF, 1:1000, Millipore, Temecula, CA). Specimens were subsequently incubated for 1 h with goat anti-chicken Alexa Fluor® 488-conjugated secondary antibody (1:300, Life Technologies, Carlsbad, California) and cover-slipped with Fluoroshield™ mounting medium (Sigma-Aldrich).

Quantitative Analysis

Using the Image J software, micrographs were acquired using bright field and confocal microscopes (Carl Zeiss, Inc., Oberkochen, Germany). Neurite extension was quantified by outlining the length from the edge of the soma to the end of the neurite for each neurite per neuron. These values were then averaged per well (N = 50 cells/group).

Statistical Analysis

Data are shown as mean ± standard error of the mean (SEM), and were analyzed using Matlab® (Mathworks, Natick, MA). The experiments involving a single determination of means between two independent groups were analyzed with the Student’s t-test or the one-way analysis of variance (ANOVA). Statistical significance was set at p < 0.05 unless otherwise noted.