Polyethylene glycol fusion repair of severed sciatic nerves accelerates recovery of nociceptive sensory perceptions in male and female rats of different strains

Abstract

Successful polyethylene glycol fusion (PEG-fusion) of severed axons following peripheral nerve injuries for PEG-fused axons has been reported to: (1) rapidly restore electrophysiological continuity; (2) prevent distal Wallerian Degeneration and maintain their myelin sheaths; (3) promote primarily motor, voluntary behavioral recoveries as assessed by the Sciatic Functional Index; and, (4) rapidly produce correct and incorrect connections in many possible combinations that produce rapid and extensive recovery of functional peripheral nervous system/central nervous system connections and reflex (e.g., toe twitch) or voluntary behaviors. The preceding companion paper describes sensory terminal field reorganization following PEG-fusion repair of sciatic nerve transections or ablations; however, sensory behavioral recovery has not been explicitly explored following PEG-fusion repair. In the current study, we confirmed the success of PEG-fusion surgeries according to criteria (1–3) above and more extensively investigated whether PEG-fusion enhanced mechanical nociceptive recovery following sciatic transection in male and female outbred Sprague–Dawley and inbred Lewis rats. Mechanical nociceptive responses were assessed by measuring withdrawal thresholds using von Frey filaments on the dorsal and midplantar regions of the hindpaws. Dorsal von Frey filament tests were a more reliable method than plantar von Frey filament tests to assess mechanical nociceptive sensitivity following sciatic nerve transections. Baseline withdrawal thresholds of the sciatic-mediated lateral dorsal region differed significantly across strain but not sex. Withdrawal thresholds did not change significantly from baseline in chronic Unoperated and Sham-operated rats. Following sciatic transection, all rats exhibited severe hyposensitivity to stimuli at the lateral dorsal region of the hindpaw ipsilateral to the injury. However, PEG-fused rats exhibited significantly earlier return to baseline withdrawal thresholds than Negative Control rats. Furthermore, PEG-fused rats with significantly improved Sciatic Functional Index scores at or after 4 weeks postoperatively exhibited yet-earlier von Frey filament recovery compared with those without Sciatic Functional Index recovery, suggesting a correlation between successful PEG-fusion and both motor-dominant and sensory-dominant behavioral recoveries. This correlation was independent of the sex or strain of the rat. Furthermore, our data showed that the acceleration of von Frey filament sensory recovery to baseline was solely due to the PEG-fused sciatic nerve and not saphenous nerve collateral outgrowths. No chronic hypersensitivity developed in any rat up to 12 weeks. All these data suggest that PEG-fusion repair of transection peripheral nerve injuries could have important clinical benefits.

Introduction

Severe peripheral nerve injuries (PNIs) in mammals (e.g., complete sciatic transections) result in immediate loss of motor functions and sensory perceptions (Gaudet et al., 2011; Scheib and Hoke, 2013; Bittner et al., 2016b; Vargas and Bittner, 2019; Mencel and Bittner, 2023) accompanied by possible chronic neuropathic pain (Jensen and Finnerup, 2014; Shamoun et al., 2022; Yang et al., 2022). Severe PNIs produce (1) immediate, complete axolemmal/axoplasmic/electrophysiological discontinuities and denervation of distal muscle fibers and sensory structures; (2) Wallerian Degeneration (WD) of severed distal sensory and motor axons within 1–3 days; (3) dissociation of all myelin sheaths within days to weeks; and, (4) atrophy of denervated muscle fibers that is often irreversible within weeks to months (Bittner et al., 2016b; Vargas and Bittner, 2019; Mencel and Bittner, 2023). Successful polyethylene glycol fusion (PEG-fusion) repair of severed axons has been reported to largely prevent (1–4) above and induce extensive, motor-dominant voluntary behavioral recoveries within weeks as assessed by the Sciatic Functional Index (SFI) in rats (Riley et al., 2015; Bittner et al., 2016b; Ghergherehchi et al., 2016, 2019b, 2021; Mikesh et al., 2018a, b; Vargas and Bittner, 2019; Mencel and Bittner, 2023).

PEG-fusion rapidly (seconds to minutes) produces correct and incorrect connections between proximal and distal motor axons following sciatic transection or ablation PNIs (Ghergherehchi et al., 2019a). PEG-fusion repair of such PNIs correlated with recovery of primarily motor behaviors as assessed by the SFI 4–6 weeks post-operatively (PO) could be partly due to re-organization of peripheral nervous system and central nervous system connections. In the preceding companion paper (Hibbard et al., 2025), we reported that PEG-fusion also rapidly produces correct and incorrect connections between proximal and distal sensory axons following sciatic transection or ablation PNIs. Morphological analyses showed extensive central reorganization of sensory terminal fields both contralateral and ipsilateral to the injury within weeks following neurorrhaphy with PEG-fusion repair (PEG-fused) and neurorrhaphy only (Negative Control [NC]) in both male and female Sprague–Dawley (SD) rats.

Except for the companion sensory axon labeling paper (Hibbard et al., 2025), most of the previous PEG-fusion studies concentrated on locomotor recovery mostly in female SD rats (Britt et al., 2010; Riley et al., 2015; Mikesh et al., 2018a, b; Ghergherehchi et al., 2021; Frost et al., 2023). In the current study, our goal was to determine the potential therapeutic effects of PEG-fusion repair on the loss of mechanical nociception following sciatic nerve transection injury. We systematically compared two strains of rats (outbred SD and inbred Lewis) of different sexes following PEG-fusion repairs because of previously reported strain and sex differences in animal models of neuropathic pain (Hestehave et al., 2020; Boullon et al., 2021). Two strains of rats were investigated also in part because SD rats, but not Lewis rats, often exhibit autophagia of toes that require euthanization of some animals prior to the study endpoint (Carr et al., 1992). Although data from non-noxious mechanosensory recovery assays (two-point discrimination, Semmes-Weinstein monofilament testing) following PEG-fusion repair have been reported in human cases after traumatic digital nerve transection injuries (Bamba et al., 2016; Lopez et al., 2022), the injuries were non-uniform, the nerves were sensory-only, and the sample size was extremely small (n = 4). In contrast, our animal injury and repair model was uniform to the sensorimotor mixed sciatic nerve in the mid thigh and assessed mechanical nociception.

We investigated sensory behavioral recovery by using von Frey filaments (VF) to estimate mechanical nociceptive withdrawal thresholds at different dorsal and midplantar regions of the hindpaws innervated by different sciatic nerve branches and the saphenous nerve. Several types of sciatic PNIs (crush, transection, and ablation) have been reported to rapidly produce hyposensitivity of the plantar and dorsal regions of the hindpaw (Devor et al., 1979; Kingery and Vallin, 1989; Kingery et al., 1994; Casals-Diaz et al., 2009; Cobianchi et al., 2014). Different regions of the hindpaws ipsilateral to the injury examined in the next 4–9 weeks PO were reported to recover to baseline values or became chronically hypersensitive possibly due to collateral sprouting of the neighboring saphenous nerve. Given these data, we hypothesized that (1) PEG-fusion repair of sciatic transections would produce better sensory-dominant behavioral recovery in both male and female SD and Lewis rats by returning to baseline withdrawal thresholds sooner than NC similar to our previously published PEG-fusion papers cited above, (2) This better sensory recovery in PEG-fused rats is mostly due to PEG-fused sciatic axons and possibly some saphenous nerve collaterals; and (3) PEG-fusion repair would not produce chronic hypersensitivity because it was not reported in previous experimental or clinical studies (Bamba et al., 2016; Mikesh et al., 2018a; Ghergherehchi et al., 2021; Lopez et al., 2022) and was not observed in our historical studies cited above nor in our initial pilot studies. Our results described herein are consistent with these hypotheses.

Methods

Animals

All experimental procedures were approved and performed according to standards of the National Institutes of Health Guide for the Care and Use of Laboratory Animals (8^th ed., National Research Council, 2011), the Institutional Animal Care and Use Committee at the University of Texas at Austin (AUP 2022-00278, approved July 2, 2023) and Indiana University (20-038-5, approved August 12, 2020). Male (250–500 g) and female (225–300 g) outbred SD and inbred Lewis rats at the age of 3–12 months were housed 2–3/cage and maintained on a 12-hour light/dark cycle with food and water given ad libitum. A total of 148 animals were used in this study (See Table 1 for animal numbers in behavioral tests). Surgical and behavioral procedures were performed in the active cycle. The reduction in the number of SD rats with PO time shown in various graphs was due to autophagia of their digits that required IACUC-determined premature euthanasia prior to 6 weeks PO endpoint. Our historical SFI data was usually collected for 6–8 weeks PO because this is the time by which most individual SFI scores reached and maintained plateau values. Therefore, we initially confirmed in pilot studies in the current paper that 6 weeks PO would be the endpoint for most protocols. Based on these SFI data, we hypothesized that we would not expect to see hypersensitivity develop after 6 weeks PO. We subsequently examined whether animals would develop hypersensitivity at 6 weeks PO and did not see any significant changes in hypersensitivity between 6 weeks and 12 weeks PO in our initial cohort of animals. Therefore, to prevent IACUC undesired discomfort of animals, we did not retain as many animals from 6 weeks to 12 weeks PO. Another subset of rats was subjected to a different surgical protocol (sciatic and saphenous nerve ablation) and euthanized at 9 weeks PO.

Table 1.

Number of rats used in behavioral experiments

Experiments	Figure	Animal strain	Treatment group	Time point
				BL	3 d	1 w	2 w	3 w	4 w	5 w	6 w	7 w	8 w	9 w	10 w	11 w	12 w
Axon Morphology	2B,C	SD	Unop	2*
			PEG								2*
			NC								1
		Lewis	Unop	2*
			PEG								3
			NC								2
SFI	3A	SD	PEG			54	53	49	42	40	35	19	19	18	19	18	18
			NC			33	33	31	23	22	21	7	7	7	7	7	7
			Sham	18		18	18	18	18	18	18
	3B	Lewis	PEG			20	20	20	20	20	20
			NC			13	13	13	13	13	13
	3C	SD	PEG			31	31	30	26	25	21	10	10	10	10	9	9
			NC			18	18	17	14	13	12	6	6	6	6	6	6
			Sham	9		9	9	9	9	9	9
	3D	Lewis	PEG			10	10	10	10	10	10
			NC			2	2	2	2	2	2
	3E	SD	PEG			23	22	19	16	15	14	9	9	8	9	9	9
			NC			15	15	14	9	9	9	1	1	1	1	1	1
			Sham	9		9	9	9	9	9	9
	3F	Lewis	PEG			10	10	10	10	10	10
			NC			11	11	11	11	11	11
	3G	SD	Successful PEG			22	22	22	22	22	20	9	9	8	9	9	9
			Poor PEG			15	15	15	15	15	15	10	10	10	10	9	9
			NC			33	33	31	23	22	21	7	7	7	7	7	7
	3H	Lewis	Successful PEG			14	14	14	14	14	14
			Poor PEG			6	6	6	6	6	6
			NC			13	13	13	13	13	13
Plantar VF	4	SD	PEG	5	5	5	5	4	3	3	3
			NC	5	5	5	5	5	5	5	5
Dorsal VF Controls	5A	SD	Transections		3
	5E	SD	Female	9
			Male	9
	5F	SD	Female	10	10	10	10	10	10	10	10
			Male	10	10	10	10	10	10	10	10
	5G	SD	Female	9	9	9	9	9	9	9	9
			Male	9	9	9	9	9	9	9	9
Dorsal VF	6A	SD	PEG	45	45	45	42	38	30	28	25	20	20	20	20	20	20
			NC	24	24	24	24	23	15	14	13	10	10	10	10	10	10
	6B	SD	Successful PEG	13	13	13	13	13	12	12	12	9	9	9	9	9	9
			Poor PEG	14	14	14	14	14	14	13	13	11	11	11	11	11	11
			NC	24	24	24	24	23	15	14	13	10	10	10	10	10	10
	6C	SD	PEG	22	22	22	22	21	16	15	13	10	10	10	10	10	10
			NC	13	13	13	13	13	10	9	8	6	6	6	6	6	6
	6D	SD	Successful PEG	6	6	6	6	6	6	6	6	4	4	4	4	4	4
			Poor PEG	8	8	8	8	8	8	7	7	6	6	6	6	6	6
			NC	13	13	13	13	13	10	9	8	6	6	6	6	6	6
	6E	SD	PEG	23	23	23	20	17	14	13	12	10	10	10	10	10	10
			NC	11	11	11	11	10	5	5	5	4	4	4	4	4	4
	6F	SD	Successful PEG	7	7	7	7	7	6	6	6	5	5	5	5	5	5
			Poor PEG	6	6	6	6	6	6	6	6	5	5	5	5	5	5
			NC	11	11	11	11	10	5	5	5	4	4	4	4	4	4
	6G	Lewis	PEG	20	20	20	20	20	20	20	20
			NC	12	12	12	12	12	12	12	12
	6H	Lewis	Successful PEG	13	13	13	13	13	13	13	13
			Poor PEG	7	7	7	7	7	7	7	7
			NC	12	12	12	12	12	12	12	12
Dorsal VF multiple injuries	7A,B	SD	PEG	2	2	2	2	2	1	1	1	1	1	1
			NC	3	3	3	3	3	3	1	1	1	1	1
	7C,D	SD	PEG	1	1	1	1	1	1	1	1	1	1	1
			NC	1	1	1	1	1	1	1	1	1	1	1
Dorsal VF (2nd 3rd metatarsals)	8A	SD	PEG	5	5	5	5	5	5	5	4
			NC	8	8	8	8	6	5	5	5
	8B	SD	PEG														8
			NC														4

Note that all rats were tested for SFI but not necessarily for VF at all different regions. Decreased numbers are either due to autophagia or reaching endpoint at 6 weeks post-operatively. In Figure 2, * indicates groups with one male and one female. BL: Baseline; NC: negative control; PEG: polyethylene glycol; SD: Sprague–Dawley; VF: von Frey filaments; w: weeks.

Surgical procedure

The surgical procedure has been previously described (Mikesh et al., 2018a; Ghergherehchi et al., 2019b, 2021). Briefly, rats were anesthetized with 4% isoflurane and maintained at 2% isoflurane with a small isoflurane-inhaled anesthetic device (Handlebar Anesthesia, Pflugerville, TX, USA). Surgeries were performed on the left hindlimb following appropriate sterilization. The right hindlimb serves as the unoperated control for that rat. A 2–3.5 cm incision was made along the lateral aspect of the skin between the knee and the hip, and the bicep femoris muscle was divided to access the sciatic nerve. The connective tissue around the sciatic nerve was carefully peeled away with microforceps. The nerve was sharply transected by fine dissection scissors in Ca²⁺-containing isotonic extracellular fluid (e.g., 0.9% NaCl with 2 mM CaCl₂). In both PEG-fused and NC groups, severed sciatic nerve stumps were irrigated with 0.5% methylene blue (MB) followed by hypotonic diluted Normosol (250–255 mOsm). Axonal ends of both sciatic nerve stumps were carefully trimmed flush to enable their close apposition with at least four 10-0 microsutures through the epineurium and/or perineurium sheathes. For PEG-fused rats, lesion sites were then submerged in a sterile solution of 50% w/w 3.35 kDa PEG (Sigma-Aldrich, St. Louis, MO, USA) in distilled water for 1–2 minutes to non-specifically fuse closely apposed cut axonal ends. NC rats were sutured but not treated with PEG. Following neurorrhaphy, lesion sites in both PEG-fused and NC rats were flushed several times with Ca²⁺-containing isotonic extracellular fluid to accelerate Ca²⁺-induced vesicle accumulation to repair any remaining open axons (Mencel and Bittner, 2023). The muscle incision was closed with 5-0 sutures, and the skin was closed with wound clips. Sham-operated rats were subjected to the same initial procedure prior to nerve transection and then the same muscle and skin closure at the end. Rats recovered from surgery on heat pads and were returned to standard housing. Rats received 5 mg/kg subcutaneous injections of carprofen during surgery and daily carprofen injections for the next three days.

In this study, some rats after the initial sciatic transection and repair received a saphenous ablation injury at 3 or 6 weeks PO followed by a subsequent sciatic ablation injury at 9 weeks PO. For saphenous nerve ablations, a 1–1.5 cm skin incision was made along the medial aspect of the skin below the left knee to expose the saphenous nerve. The saphenous nerve was carefully separated from the adjacent saphenous vein with micro-forceps, and a 1 cm segment of the nerve was removed before closing the wound. For subsequent sciatic nerve ablations, the previously operated nerve was re-exposed, and a 1 cm segment of the sciatic nerve centered on the previous repair site was removed before closing the wound as described above.

Electrophysiological recordings

Compound action potentials (CAPs) were recorded before sciatic transection and immediately following surgical procedures in rats to confirm electrophysiological continuity in PEG-fused rats and lack of continuity in NC rats using a PowerLab 4/35 (AD instruments, Sydney, Australia). Stimulating and recording electrodes were placed proximal and distal to the suture site at > 0.5 cm apart. Sciatic nerves were stimulated with incremental increases in voltages up to 4 V using 0.01 ms square wave depolarizations at 1 Hz with a 0.1 ms delay from the trigger signal.

Morphological analyses

Sciatic nerve segments were harvested for resin embedding as previously described (Mikesh et al., 2018a, b). Briefly, samples were fixed in 2% paraformaldehyde/3% glutaraldehyde in 0.1 M sodium cacodylate buffer, trimmed into proximal and distal segments that were 1–2 mm away from the suture site, and post-fixed in 1% osmium tetroxide/1% potassium ferrocyanide. Samples were further fixed in 1% aqueous uranyl acetate prior to dehydration and embedding in Hard Plus Resin 812 (Electron Microscopy Sciences, Hatfield, PA, USA). Samples were incubated at 60°C for 48–72 hours prior to sectioning. Glass knife thick sections (0.5 µm) were obtained using a Leica ultramicrotome, stained in toluidine blue, and imaged on a Zeiss Axiovert 200M fluorescent light microscope with HR3 camera (Hebron, KY, USA) for axonal morphological analyses using ImageJ (version 1.53o, National Institutes of Health, Bethesda, MD, USA). At least 200 axons from at least three regions of interest were analyzed in each sample by investigators blind to the treatment group for axon diameter and g-ratio.

Spinal cord labeling

To confirm sciatic innervation of the cutaneous site used for VF testing in this study, wheat germ agglutinin conjugated horseradish peroxidase (WGA-HRP) was used to anterogradely label sensory terminal fields. Animals received either bilateral, intradermal injections of WGA-HRP (0.3 μL, 2%; Invitrogen) into the dorsal aspect of the hindpaw between the 4th and 5th metatarsals (Unoperated, n = 2; NC; n = 2) or bilateral, intraneural injections of WGA-HRP into the sural nerve (Unoperated, n = 4). Forty-eight hours after injection, animals were anesthetized and perfused intracardially with saline followed by cold 1% paraformaldehyde/1.25% glutaraldehyde. The lumbar portion of the spinal cord of each animal was removed, postfixed in the same fixative for 5 hours, then transferred to sucrose phosphate buffer (10% w/v, pH 7.4) overnight for cryoprotection. Spinal cords were embedded in gelatin, frozen, and sectioned transversely at 40μm. For visualization of WGA-HRP, sections were immediately reacted using a modified tetramethyl benzidine protocol (Mesulam, 1982), mounted on gelatin-coated slides, and counterstained with thionin.

For each animal, the entire rostrocaudal range of spinal segments contributing to the sciatic nerve was examined under darkfield illumination. Spinal segment boundaries were determined using the midpoints between dorsal root entrances (Molander et al., 1984). We noted the presence or absence of WGA-HRP labeling in each spinal segment and its location in the gray matter. Beginning with the first section in which WGA-HRP terminal labeling was present, the entire rostrocaudal extent of labeling was assessed in sections at 320 µm intervals. Labeling was recorded in three dimensions using a computer-based morphometry system (Neurolucida, MBF Bioscience, Williston, VT, USA) at a final magnification of 500×. The extent of rostrocaudal labeling was identified by the spinal segment and expressed as a distance relative to the L3/L4 boundary. Digital light micrographs were obtained with an MDS 290 digital camera system (Eastman Kodak Company, Rochester, NY, USA). The brightness and contrast of these images were adjusted in Adobe Photoshop (Adobe Systems, San Jose, CA, USA).

Sciatic Functional Index

The SFI was used to assess motor behavioral recovery mediated by the sciatic nerve in rats as previously described (Mikesh et al., 2018a; Ghergherehchi et al., 2019b, 2021). Rats were handled and trained for at least three sessions prior to surgery to become acclimated to the testing apparatus and procedure. SFI tests were performed and scored by testers blinded to the experimental groups. Rat hind paws were marked with red ink (right, unoperated side) and blue ink (left, operated side). Rats were placed on one end of a slightly inclined board (1.52 m long, 10.2 cm wide) lined with paper strips to run back to the home cage on the other end. Inked paw prints were analyzed based on paw length, total toe spread, and intermediate toe spread to compute SFI scores. A successful SFI trial required three consecutive hindlimb steps on each side without hesitating or stopping. Two trials were obtained and averaged to calculate the final score at each given time point for each rat. Unoperated rats exhibit SFI scores of 0 ± 30, indicating symmetry of gait. Impaired movement of the injured hindpaw results in lower, negative SFI scores. Rats were tested weekly after surgery for up to 12 weeks PO.

Simplified dorsal/plantar von Frey filaments assay

The midplantar and various dorsal regions of hindpaws were tested using simplified up-down VF methods that were modified from previous publications (Chaplan et al., 1994; Kingery et al., 1994; Detloff et al., 2012; Bonin et al., 2014) and piloted in SD rats as described in subsequent paragraphs. Rats were handled and trained for at least four sessions prior to surgery to become acclimated to the testing apparatus and procedure. VF tests were performed and scored blind to the surgical procedures by different testers in different sessions. VF filaments (Stoelting Co., Wood Dale, IL, USA) were checked monthly to accurately represent intended gram forces and routinely replaced when necessary.

In all testing sessions, the starting hindpaw was randomly determined. The beginning filament (#5.07) was presented vertically to the targeted region until bent for two seconds and was repeated on the alternated paw. A positive or negative withdrawal response was recorded for each trial on each side. A positive withdrawal response was followed by a lower force filament whereas a negative response was followed by a higher force filament used for the next trial. A total of 5 trials were tested on each side. At least 30 seconds occurred between trials of both hindpaws. The final VF thresholds were calculated by first adding or subtracting an adjustment factor of 0.5 depending on the last response being negative or positive, respectively, then converted to gram force using the equation: Force = 10^{(0.904×filament #-3.54)}. Any response was voided and retried if the filament slipped or moved the hindpaw, or if the rat struggled in the towel during dorsal VF application. Note that the largest filament used in these tests was #6.10.

Midplantar VF testing was performed by a single tester for any given session. Rats were temporarily isolated in upside-down cages on top of a mesh-wire grid. Treats were given to lower the rat’s alertness during testing. Filaments were applied to the midplantar surface of the foot, approximately 1 cm posterior to the footpad of the middle phalange (Kloos et al., 2005). Following sciatic transections, many rats placed minimal support on the hindpaw ipsilateral to the injury. Consequently, larger-size filaments (starting at #5.46) would often not bend or elicit a withdrawal response even if they lifted the hindpaw (ankle flexion) or limb (hip flexion) (Figure 1A and Additional Video 1). This movement likely led to inaccurate estimates of VF gram force values and provided unwanted proprioceptive feedback.

Figure 1.

VF testing procedures.

(A) For midplantar VF, larger filaments (#5.46 and higher) often propped up the hindpaw ipsilateral to the injury without bending and failed to induce consistent withdrawal responses. (B) Illustration of dorsal VF testing. The rat was restrained in a towel by one tester while the other tester applied filaments to the dorsal regions of both hindpaws. (C) Tested dorsal regions between the metatarsals (red oval). Numbers indicate the positions of the knuckles. VF: von Frey filaments.

Dorsal VF assays were performed by two testers for any given session (Figure 1B). Rats were gently restrained in towels, which masked their head and upper torso, and elevated to an upright position by one tester to expose their hindpaws. Filaments were applied to the selected area between two metatarsals (Figure 1C) by the second tester when the rat was calm. Following sciatic transections, some rats had severe muscle atrophy and developed a clenched foot, leading to the filament easily slipping off the foot during its application. To avoid this problem, testers would gently extend the foot flat and hold the digits down until the trial was completed. No noticeable difference in withdrawal thresholds was observed with or without this foot extension. Note that larger filaments (starting at #6.45) did not elicit withdrawal responses and caused long-lasting wounds (data not shown). All filaments including and above #6.45 were excluded in subsequent experiments to avoid mechanical damage to cutaneous sensory structures. Overall, our dorsal VF method improved from previous protocols (Kingery et al., 1994; Detloff et al., 2012; Bonin et al., 2014) by reducing the number of repetitive testing (5 trials instead of more than 15 trials) and determining the maximum force that represent hyposensitivity without causing damage to the tested regions.

Statistical analysis

No statistical methods were used to predetermine sample sizes; however, our sample sizes are similar to those reported in previous publications (Mikesh et al., 2018a; Ghergherehchi et al., 2019b, 2021). All statistical analyses were performed using GraphPad Prism (version 8, GraphPad Software, Boston, MA, USA, www.graphpad.com). Morphological data were analyzed by one-way analysis of variance followed by post hoc Tukey’s multiple comparisons. SFI and VF data were analyzed by a mixed-effects model (SD rats; used due to a decrease in n numbers caused by autophagia) or two-way analysis of variance (Lewis rats) with repeated measures followed by post hoc Tukey’s multiple comparisons. When comparing SD and Lewis rats for potential strain differences, 7–12 weeks PO data from SD rats were excluded because there were no Lewis data at those time points. A 95% confidence interval was used for all data. All data are presented as mean ± SEM. Animal and/or axon numbers are indicated in each figure panel for morphological and labeling data. All animal numbers are listed in Table 1. All data points and animal subjects used in this study were included in the composite SFI and VF analysis regardless of early euthanasia. Some analyses correlating SFI and VF data only included animals having at least six weeks of both SFI and VF data.

Results

Confirmation of polyethylene glycol-fusion repair of some sciatic axons

The status of axonal fusion/repair following a complete sciatic transection was confirmed using several well-published criteria (Bittner et al., 2016a; Ghergherehchi et al., 2016, 2019a, b, 2021; Mikesh et al., 2018a, b) as documented in Figure 2.

Figure 2.

PEG-fusion produces immediate restoration of electrophysiological continuity and preservation of many distal axons.

(A) Example CAP recorded immediately after calcium-containing saline irrigation in PEG-fused (neurorrhaphy and PEG application) and NC (neurorrhaphy only) rats. Unoperated nerve was recorded immediately after isolating the sciatic nerve. (B) Representative images of transverse thick sections taken from the sciatic nerve in Unoperated rats as well as in PEG-fused and NC rats distal to the injury at 6 weeks PO. Scale bars: 10 μm. (C, D) Violin plots of axon diameter and g-ratio obtained from Unoperated rats and operated rats at 6 weeks PO obtained from at least 200 axons from at least 3 regions of interest. n numbers of each group are indicated below in brackets (top: n of rat subjects; bottom: n of axons) in C and are the same for D (not labeled). One-way analysis of variance. For C, F_(5,2702) = 352, P < 0.0001. For D, F_(5,2702) = 14, P < 0.0001. Tukey’s test, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. CAP: Compound action potential; NC: negative control; PEG: polyethylene glycol; PO: post-operatively; SA: stimulus artifact; SD: Sprague–Dawley.

Compound action potential recordings

Figure 2A shows compound action potential (CAP) recordings of Unoperated sciatic nerve as well as PEG-fused and NC sciatic nerves immediately following calcium-containing saline application. A prominent CAP after the stimulus artifact was observed in PEG-fused rats but not in NC rats. This restoration of electrophysiological continuity strongly implies the restoration of axolemmal and axoplasmic continuity in some axons in PEG-fused sciatic nerves.

Axon morphology

Figure 2B illustrates axon morphologies in distal segments of sciatic nerves in Unoperated rats as well as PEG-fused and NC rats at 6 weeks PO. Unoperated rats exhibit similar (P > 0.05) axon diameter (SD: 4.14 ± 0.07 µm, Lewis: 4.34 ± 0.08 µm) and g-ratio (SD: 0.599 ± 0.003 µm, Lewis: 0.585 ± 0.003 µm) across strains (Figure 2C and D). Some statistical significance in g-ratio, but not axon diameter, was noted among the Unoperated subjects (data not shown), though such differences often even exist within the same animal at different locations (see Mikesh et al., 2018b for variation of axon diameter and g-ratio across proximal, graft, and distal portion of Unoperated rats). Most of the sampled operated rats were males (7 out of 8 cases; Table 1) to enable comparisons with historical data characterizing females (Mikesh et al., 2018a, b). NC distal nerves contained much debris and mostly small diameter (1–4 µm), thinly myelinated axons identified as regenerating axonal outgrowth in previous studies (Mikesh et al., 2018a, b). In contrast, PEG-fused distal nerves contained relatively little debris, some regenerative axons, but also some large diameter axons (4–7 µm) with thick myelination. These axons have been identified as successfully PEG-fused axons that do not undergo WD (Mikesh et al., 2018a, b). Overall, axons in PEG-fused nerves (SD: 2.34 ± 0.06 µm; Lewis: 2.44 ± 0.04 µm) had significantly larger diameters (P < 0.0001; Figure 2C) than those in NC nerves (SD: 1.89 ± 0.05 µm; Lewis: 1.80 ± 0.03 µm) in both strains. Axons in PEG-fused nerves (SD: 0.580 ± 0.004; Lewis: 0.572 ± 0.003) had higher g-ratio than those in NC nerves (SD: 0.558 ± 0.007; Lewis: 0.561 ± 0.004) in both strains (Figure 2D), though there was significance in SD rats (P < 0.05) but not in Lewis rats (P = 0.13). No significant difference (P > 0.05) was found in axon diameter or g-ratio in the same treatment group across strains. Furthermore, because axon diameters and g-ratio of all groups in this study using mostly males were within the range previously reported for female SD rats (Mikesh et al., 2018a, b), our results suggested that sex difference in axon morphologies is also unlikely.

Locomotor recovery

Figure 3 shows SFI behavioral recovery of PEG-fused and NC rats. Sham-operated Control SD rats had SFI scores between –1.5 ± 3.1 and 2.8 ± 2.5 (Figure 3A), which did not change over time compared with baseline scores of –1.3 ± 2.8 (data not shown). PEG-fused and NC SFI scores in SD rats were not significantly different (P > 0.05) from 7–28 days PO (Figure 3A). However, PEG-fused rats had significantly better (P < 0.05 or better) SFI scores than NC rats starting at 5 weeks PO. At 6 weeks PO, PEG-fused and NC SD rats had mean SFI scores of –67 ± 4.9 and –97 ± 4.6, respectively. Similarly, PEG-fused Lewis rats had significantly better (P < 0.05) SFI scores than NC rats at 6w PO with means of –64 ± 6.7 and 95 ± 6.3, respectively (Figure 3B). Both PEG-fused females (Figure 3C [SD] and 3D [Lewis]) and males (Figure 3E [SD] and 3F [Lewis]) demonstrated significantly (P < 0.05) better SFI scores at 6 weeks PO compared with same-sex NC rats. No significant difference (P > 0.05) in SFI recovery was observed across sexes or strains within each group. However, not all PEG-fused rats exhibited successful SFI recovery (individual SFI score ≥ –69 by 6 weeks PO (Mikesh et al., 2018a)). Figure 3G and H showed that successfully PEG-fused rats that began to recover SFI scores at 4 weeks PO exhibited significantly (P < 0.01 or better) better SFI scores than both poorly PEG-fused (individual SFI score < –69) and NC rats starting at 5 weeks PO (for 6 weeks PO, SD: –45 ± 3.2; Lewis: –48 ± 4.2). Following 6 weeks PO, successfully PEG-fused SD rats continued to improve SFI scores, reaching –26 ± 6.9 at 12 weeks PO. In contrast, no significant difference was found between poorly PEG-fused rats and NC rats at any time point (at 12 weeks PO, Poor PEG: –106 ± 5.1; NC: –110 ± 4.3) in both strains despite regenerating sciatic sprouts shown in Figure 2B (Sakuma et al., 2016; Mikesh et al., 2018a).

Figure 3.

Successful PEG-fusion restores voluntary motor behavioral recovery.

SFI scores were obtained from weekly testing in SD rats for up to 12 weeks PO and in Lewis rats for up to 6 weeks PO. Both male and female Sham-operated Control (grey dashed) SD rats were included. Left panels, mixed-effects model; right panels, two-way analysis of variance. (A, B) SFI recovery of PEG-fused (red) and NC (black) SD and Lewis rats. For A, interaction F_(11,480) = 5.6, P < 0.0001; time F_{(2.3, 101)} = 2.9, ns; treatment F_(1,85) = 34, P < 0.0001. For B, interaction F_(5,155) = 6.1, P < 0.0001; time F_{(2.3, 72)} = 7.3, P = 0.0008; treatment F_(1,31) = 1.8, ns. (C, D) SFI recovery of female PEG-fused and NC SD and Lewis rats. For C, interaction F_(11,279) = 3.2, P = 0.0005; time F_{(1.9, 48)} = 1.3, ns; treatment F_(1,47) = 13, P = 0.0007. For D, interaction F_(5,50) = 2.5, P = 0.04; time F_{(2.3, 23)} = 1.1, ns; treatment F_(1,10) = 1.9, ns. (E, F) SFI recovery of male PEG-fused and NC SD and Lewis rats. For E, interaction F_(11,179) = 2.4, P = 0.01; time F_{(2.6, 42)} = 3.3, P = 0.035; treatment F_(1,36) = 19, P < 0.0001. For F, interaction F_(5,95) = 4.4, P = 0.0013; time F_{(2.1, 40)} = 5.2, P = 0.0091; treatment F_(1,19) = 2.4, ns. (G–H) PEG-fused rats were plotted after further separation based on successfully recovered SFI scores into successful PEG (red) and poor PEG (blue) in SD and Lewis rats. For G, interaction F_(22,433) = 17, P < 0.0001; time F_{(4.8, 189)} = 8.9, P < 0.0001; treatment F_(2,67) = 121, P < 0.0001. For H, interaction F_(10,150) = 10, P < 0.0001; time F_{(3, 90)} = 6.8, P = 0.0004; treatment F_(2,30) = 13, P < 0.0001. * at any time point indicates P values between PEG-fused and poorly PEG-fused rats (blue) and between PEG-fused and NC rats (black). For this and all subsequent figures, values are means ± SEM. Tukey’s test, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. NC: Negative control; ns: not significant; PEG: polyethylene glycol; PO: post-operatively; SD: Sprague–Dawley; SFI: Sciatic Functional Index; w: week.

Midplantar withdrawal thresholds

Baseline withdrawal thresholds of the midplantar region of both hindpaws in PEG-fused and NC rats (SD females) were between 34 ± 4.0 g and 46 ± 7.3 g (Figure 4). Subsequent to a sciatic nerve transection and repair, almost all rats at 3 days PO (9 out of 10 cases) experienced hyposensitivity at the midplantar region of the hindpaw ipsilateral to the injury, reflected by the lack of withdrawal responses to the largest filament used (#6.10), corresponding to a corrected gram force of 134 g. The one exception was an NC rat that became similarly hyposensitive after 4 more days at 1 week PO. Withdrawal thresholds of the unoperated hindpaw contralateral to the injury ranged from 39 ± 0 g to 63 ± 12 g between 3 days to 6 weeks PO, which were not significantly different (P > 0.05) from the baseline at any time point (P > 0.05). The earliest individual recovery of hindpaw withdrawal response ipsilateral to the injury in PEG-fused and NC animals occurred at 6 and 2 weeks PO, respectively. Additionally, PEG-fused and NC rats exhibited significant differences (P < 0.05 or better) between the ipsilateral side withdrawal thresholds and their respective baseline values up to 5 and 3 weeks PO, perhaps due to a delayed recovery in PEG-fused rats. However, the mean withdrawal thresholds between PEG-fused and NC rats were not significantly different (P > 0.05) at all time points. At 6 weeks PO, PEG-fused and NC rats exhibited a withdrawal response of the hindpaw ipsilateral to the injury to 102 ± 31.8 g and 97 ± 25 g, respectively. Only one PEG-fused and two NC rats showed consistent recovery up to 6 weeks PO. The remaining two PEG-fused rats did not exhibit any recovery, while the remaining three NC rats had partial recovery between 2 weeks and 5 weeks PO but became hyposensitive again at 6 weeks PO.

Figure 4.

Continual hyposensitivity at the midplantar region of the hindpaws ipsilateral to the injury in SD rats.

Following sciatic transection and repair, dorsal sensory behavioral recovery was assessed between 3 days and 6 weeks PO in SD rats. Mixed-effects model, interaction F_(21,98) = 2.8, P = 0.0003; time F_{(3.8, 53)} = 11, P < 0.0001; treatment F_(3,16) = 47, P < 0.0001. No significant difference (P > 0.05) between the ipsilateral sides or between the contralateral sides was found at any time point. In this and all subsequent figures, *above any time point indicate significant P values obtained from comparing the ipsilateral injured side thresholds with respective baselines in PEG-fused (red) and NC (black) rats. Brackets underneath * indicate all included time points were significant at the indicated level. Tukey’s test, *P < 0.05, **P < 0.01, ***P < 0.001. BL: Baseline; Cont: contralateral; Ipsi: ipsilateral; NC: negative control; PEG: polyethylene glycol; PO: post-operatively; SD: Sprague–Dawley; w: week.

Dorsal withdrawal thresholds –Unoperated and sham-operated controls

We first determined withdrawal thresholds at multiple dorsal hindpaw regions between different metatarsals in unoperated SD rats (Figure 1C). All regions had a withdrawal threshold of 38 g (corrected value of positive response to filament #5.88 or negative response to #5.46) prior to both sciatic and saphenous transections, after which rats experienced hyposensitivity at all regions (134 g; negative response to #6.10) at 3 days PO (Figure 5A). This result confirmed previous reports that the dorsal region of hindpaws of unoperated rats is innervated by both the saphenous and the sciatic nerves (Swett and Woolf, 1985; Kambiz et al., 2014; Hibbard et al., 2025).

Figure 5.

Dorsal von Frey testing controls in SD rats.

(A) Withdrawal threshold of the hindpaws ipsilateral and contralateral to the injury between different Mt at 3 days PO following both sciatic and saphenous transection. (B) Darkfield digital micrograph of a transverse hemisection showing WGA-HRP-labeled sensory terminal afferents from an Unoperated Control animal following intradermal injection into the area between the 4^th and 5^th Mt. (C) Representative reconstructions from Unoperated Control animals of the location of terminal labeling following intradermal injection of WGA-HRP into the area between the 4^th and 5^th Mt (shown in red) superimposed over the location of terminal labeling following intraneural injection of WGA-HRP into the sural branch of the sciatic nerve (shown in black). Sections were drawn at 320 µm intervals from matching areas of the L4 spinal segment. (D) Labeling in the rostrocaudal plane following intradermal injection into the area between the 4^th and 5^th Mt. Vertical bars indicate the extent of labeling in individual animals. For comparability, the extent of labeling is plotted across matching spinal segments, centered on the L3/L4 boundary. In Unoperated animals, WGA-HRP-labeled sensory afferents were confined to the L4 spinal segment; following transection of the sciatic nerve, all labeling was eliminated (n = 2). This is evidence of sensory projections to this cutaneous area being purely sciatic. Scale bars: 250 µm in B and C. (E–G) Withdrawal thresholds established between the 4^th and 5^th Mt. (E), Baseline withdrawal threshold obtained from three consecutive days of testing. (F) Withdrawal threshold obtained from 6 weeks of testing in Unoperated Control rats. (G) Withdrawal threshold obtained from six weeks of testing in Sham-operated Control rats. Note that Ipsi and Cont in E and F refer to their intact left and right hindpaws, respectively. No statistical significance (P > 0.05) was found between any curves on any graph. BL: Baseline; F: females; M: males; Mt: metatarsals; PO: post-operatively; SD: Sprague–Dawley; Unop: unoperated; w: week; WGA-HRP: wheat-germ-agglutinin-horseradish-peroxidase.

Following transection of the sciatic nerve, the saphenous nerve reinnervates the denervated sensory structures in the central and lateral dorsal region of hindpaws (Devor et al., 1979; Markus et al., 1984; Hibbard et al., 2025). Therefore, we focused on testing the most lateral region of the hindpaw between the 4^th and 5^th metatarsals to minimize/delay the influence of saphenous nerve reinnervation and isolate VF testing to an inherently, sciatic-only mediated cutaneous area. The sciatic only area was confirmed anatomically through intradermal injection of wheat germ agglutinin conjugated horseradish peroxidase (WGA-HRP) at this location in Unoperated Control SD animals (Figure 5B and D). The afferent terminal label in the spinal cord overlapped with the terminal field that was labeled following intraneural WGA-HRP injection into the sural branch of the sciatic nerve (Figure 5C). This afferent terminal label was completely absent in animals that also received sciatic nerve transections prior to intradermal injection between the 4^th and 5^th metatarsals (Figure 5D).

When tested for VF sensitivity between the 4^th and 5^th metatarsals, Unoperated Control SD rats had withdrawal thresholds between 34 ± 6.1 g and 42 ± 4.0 g when tested for 3 days daily (Figure 5E) or between 28 ± 3.3 g and 57 ± 6.1 g when tested for 6 weeks (Figure 5F). Similarly, Sham-operated Control SD rats had withdrawal thresholds between 31 ± 6.3 g and 52 ± 7.4 g, which were not statistically different (P > 0.05) in both hindpaws before and after surgeries when tested for 6 weeks PO (Figure 5G). No significant difference (P > 0.05) was found between sexes at any time in any group.

Polyethylene glycol-fusion induces more rapid sensory recovery in a sciatic dorsal region

von Frey filaments sensitivities for Negative Control and polyethylene glycol-fused rats following sciatic transections

Prior to a sciatic nerve transection, baseline dorsal VF withdrawal thresholds between the 4^th and 5^th metatarsals of both hindpaws in PEG-fused and NC SD rats were between 37 ± 0.8 g and 42 ± 2.7 g (Figure 6A). Subsequent to a sciatic nerve transection and repair, almost all NC and PEG-fused rats at 3 days PO (37/38 cases) experienced hyposensitivity on the hindpaw ipsilateral to the injury to the 6.10 filament (134 g). The one exception was a PEG-fused rat that became similarly hyposensitive at 1 week PO. Between 1 week and 2 weeks PO, a few PEG-fused and NC rats started to exhibit lower withdrawal thresholds on the hindpaw ipsilateral to the injury, but the mean thresholds were not significantly different (P > 0.05). At 3 weeks PO, 58% (22/38 cases) of PEG-fused and 26% (6/23 cases) of NC rats showed some recovery, and the withdrawal threshold in PEG-fused rats (91 ± 6.4 g) was significantly (P < 0.05) lower than in NC rats (115 ± 6.9 g). At 5 weeks PO, 82% (23/28 cases) of PEG-fused and 50% (7/14 cases) of NC rats had some recovery, and the withdrawal threshold in PEG-fused rats (64 ± 7.0 g) was again significantly (P < 0.05) lower than in NC rats (94 ± 12 g). At 6 weeks PO, 92% (22/24) PEG-fused and 85% (11/13 cases) NC rats had some recovery, and the thresholds were 57 ± 5.9 g and 61 ± 9.9 g, which were no longer significantly different (P > 0.05). The VF withdrawal threshold of the hindpaw ipsilateral to the injury in PEG-fused rats became no longer significant (P > 0.05) from the baseline after 4 weeks PO, which was one week earlier than that exhibited by NC rats. Between 7 weeks and 12 weeks PO, VF withdrawal thresholds of the hindpaws ipsilateral to the injury in both groups did not decline below baseline between 38 g and 56 ± 7.1 g, suggesting neither group developed hypersensitivity. The unoperated hindpaws contralateral to the injury of both groups maintained VF sensitivities of 36 ± 2.0 g to 43 ± 4.0 g and did not differ significantly (P > 0.05) from baseline VF sensitivities at any time point.

Figure 6.

PEG-fusion promotes dorsal hindpaw sensory behavioral recovery in both sexes and strains.

Following sciatic transection and repair, dorsal sensory behavioral recovery was assessed between 3 days and 12 weeks PO in all SD rats (A, B), female SD rats only (C, D), male SD rats only (E, F), and 3 d–6 w PO in all Lewis rats (G, H) between the 4^th and 5^th metatarsals. Left panels compared all PEG-fused rats, regardless of their SFI scores, against negative control (NC) rats, and right panels compared PEG-fused rats that obtained significantly recovered SFI scores (PEG), PEG-fused rats that did not obtain recovered SFI scores (Poor PEG), and NC rats. For A–F, mixed-effects model; for G–H, two-way analysis of variance. For A, interaction F_(39,1088) = 21, P < 0.0001; time F_(13,1088) = 76, P < 0.0001; treatment F_(3,134) = 75, P < 0.0001. For B, interaction F_(26,462) = 2.2, P = 0.0006; time F_(6.9,246) = 74, P < 0.0001; treatment F_(2,48) = 6, P = 0.0047. For C, interaction F_(39,560) = 18, P < 0.0001; time F_(13,560) = 52, P < 0.0001; treatment F_(3,66) = 51, P < 0.0001. For D, interaction F_(26,232) = 3, P < 0.0001; time F_(4.8,85) = 62, P < 0.0001; treatment F_(2,24) = 10, P=0.0006. For E, interaction F_(39,476) = 7.5, P < 0.0001; time F_(6,228) = 33, P < 0.0001; treatment F_(3,64) = 50, P < 0.0001. For F, interaction F_(26,191) = 1.0, ns; time F_(5,72) = 26, P < 0.0001; treatment F_(2,21) = 7, P = 0.0047. For G, interaction F_(21,420) = 18, P < 0.0001; time F_(5.2,314) = 46, P < 0.0001; treatment F_(3,60) = 63, P < 0.0001. For H, interaction F_(14,203) = 0.82, ns; time F_(4.7,136) = 51, P < 0.0001; treatment F_(2,29) = 2.9, ns (P = 0.069). * above any time point indicate significant P values obtained from comparing the hindpaws ipsilateral to the injury to their baselines in PEG-fused (red), Poor PEG (blue), and NC (black) rats. Brackets underneath * indicate all included time points were significant at the indicated level. † above any time point indicate significant P values obtained from comparing the hindpaws ipsilateral to the injury between PEG-fused and NC rats (black) and between PEG-fused and poorly PEG-fused rats (blue). Tukey’s test, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. Same probability levels for † symbols. ns: Not significant; NC: negative control; PEG: polyethylene glycol; SD: Sprague–Dawley; PO: post-operatively; SFI: Sciatic Functional Index; w: weeks.

von Frey filaments sensitivities for successfully polyethylene glycol-fused rats following sciatic transections

As discussed above in Figure 3G–H, successful PEG-fusion of individual rats had significantly recovered motor-dominant voluntary behavioral recoveries by 6 weeks PO. Figure 6B shows dorsal VF data in PEG-fused animals based on significant SFI recovery (successfully PEG-fused) or not (poorly PEG-fused). Successfully PEG-fused rats exhibited earlier sensory recovery, which was significantly better (P < 0.05 or better) than poorly PEG-fused rats 2–3 weeks PO and better than NC rats 3–5 weeks PO (P < 0.05 or better). These data show a strong correlation between successful PEG-fusion and both motor-dominant (SFI) and sensory-dominant (dorsal VF) behavioral recovery.

Sex differences in von Frey filaments sensitivities

We explored potential sex differences in SD rats (Figure 6C–F). PEG-fused females exhibited significantly (P < 0.0001) lower withdrawal thresholds than NC females at both 3 weeks and 5 weeks PO (Figure 6C). The withdrawal threshold of the hindpaw ipsilateral to the injury in PEG-fused females became no longer significant (P > 0.05) from the baseline after 3 weeks PO, which was 2 weeks earlier than that return in NC females. An analysis of PEG-fused females based on successful SFI scores showed that successfully PEG-fused females exhibited significantly (P < 0.05 or better) better sensory recovery than both poorly PEG-fused and NC females as soon as 3 weeks PO (Figure 6D). Similarly, PEG-fused males exhibited significantly (P < 0.05) lower withdrawal thresholds than NC males at 4 weeks PO (Figure 6E). Both PEG-fused and NC males recovered to baseline after 4 weeks PO. An analysis of PEG-fused males based on successful SFI scores showed that successfully PEG-fused males exhibited significantly better (P < 0.05 or better) sensory recovery than both poorly PEG-fused and NC males at 4 weeks PO (Figure 6F). A possible sex difference in VF recovery occurred at 4 weeks PO when females exhibited significantly lower (P = 0.046) response on the hindpaw ipsilateral to the injury than males in the PEG-fused group. However, they did not exhibit this difference (P = 0.077) in the NC group.

Strain differences in von Frey filaments sensitivities

We also explored potential strain differences between SD and Lewis rats (Figure 6G and H). In Lewis rats, baseline withdrawal thresholds in PEG-fused and NC rats were between 31.4 ± 2.2 g and 38.1 ± 4.0 g, which was significantly (P < 0.01) lower than those of SD rats (Figure 6G). Subsequent to a sciatic nerve transection and repair, almost all rats at 3 days PO (17/20 cases) experienced hyposensitivity on the hindpaw ipsilateral to the injury (134 g). Of the three exceptions (all PEG-fused rats), one became similarly hyposensitive at 1 week PO, while the other two continued to recover. PEG-fused Lewis rats exhibited significantly (P < 0.05) lower withdrawal threshold of the hindpaws ipsilateral to the injury (105 ± 8.6 g) than NC rats (134 g) at 1 week PO. PEG-fused Lewis rats continued to exhibit a lower withdrawal threshold of the hindpaw ipsilateral to the injury than NC Lewis rats up to 4 weeks PO, at which time both groups exhibited no significant difference (P > 0.05) from baseline threshold of the hindpaw ipsilateral to the injury. A further analysis of successfully PEG-fused Lewis rats with SFI scores ≥ –69 showed that successfully PEG-fused rats exhibited marginally better sensory recovery than both poorly PEG-fused NC rats at 1 week PO (P < 0.053) and significantly at 4 weeks PO (P < 0.05; Figure 6H). PEG-fused Lewis rats on average recovered 2 weeks earlier than PEG-fused SD rats, resulting in a significantly lower (P < 0.01) threshold of the hindpaw ipsilateral to the injury at 2 weeks and 4 weeks PO. Similarly, NC Lewis rats exhibited a significantly lower (P < 0.01) threshold of the hindpaw ipsilateral to the injury than NC SD rats at 5 weeks PO.

Considered together, these data showed some strain differences and possible sex differences in VF recovery. Nevertheless, successful PEG-fusion strongly correlated with better motor and sensory behavioral recoveries across strains and sexes.

Contribution of saphenous collaterals to sensory recovery

Two populations of axons have been reported to produce mechanical nociception recovery following sciatic crush (Devor et al., 1979) or sciatic ligation and transection (Markus et al., 1984): (#1) invading intact saphenous nerve collaterals (saphenous collaterals) responsible for an initial recovery phase, and (#2) sciatic regenerating outgrowths (regenerating sciatic) responsible for a subsequent recovery phase. After complete sciatic transections and PEG-fusion repairs, rats have a third population (#3) of axons (PEG-fused sciatic) (Figure 2; Mikesh et al., 2018a) that could affect mechanical nociception recovery. Hence, we hypothesized that three axon populations (#1–3) in PEG-fused rats and two populations (#1–2) in NC rats could contribute to VF recovery shown in Figure 6. To determine which axon populations contributed to VF recovery after sciatic nerve transection, we examined a few PEG-fused and NC SD rats that received an initial sciatic transection and repair and then received another saphenous ablation at either (1) 3 weeks PO when regenerating sciatic (#2) should not be present or (2) 6 weeks PO when regenerating sciatic (#2) should be present (Mikesh et al., 2018a). All rats were tested for SFI weekly until the saphenous nerve was ablated and for VF 3 days after each injury and weekly thereafter. In both protocols, another sciatic ablation was performed at 9 weeks PO to assess whether sciatic axons were indeed responsible for any subsequent VF recovery after saphenous ablations. The combined saphenous ablation and initial sciatic transection injuries usually led to their premature euthanasia, thereby limiting the number of rats sampled in these protocols.

Saphenous ablations 3 weeks post-operatively

For PEG-fused (n = 2) and NC (n = 3) rats that exhibited VF recovery at 3 weeks PO (Figure 7A and B), a subsequent saphenous ablation (blue arrow) produced severe hyposensitivity in all rats tested 3 days and 1 week later, suggesting that saphenous collateral population (#1) was responsible for the VF recovery at 3 weeks PO. Three rats (PEG#1, NC#1, and NC#2) were euthanized at 4 weeks PO due to autophagia. The surviving rat PEG#2 started to recover VF sensitivity at 5 weeks PO, possibly due to both of the sciatic populations (#2–3). Rat NC #3 started to recover VF sensitivity at 8 weeks PO, almost certainly due to regenerating sciatic population (#2). This was confirmed by the sciatic ablation at 9 weeks PO that eliminated VF sensitivity in both rats. These data are consistent with the hypothesis that the PEG-fused sciatic population (#3) was primarily responsible for the more rapid behavioral recovery of PEG-fused SD rats compared with NC rats in Figure 6A and B.

Figure 7.

The promotion of VF recovery by PEG-fusion is independent of early saphenous collateral sprouting.

A few animals with sciatic nerve transections and repair at 0 day PO were subjected to a saphenous ablation injury at 3 weeks PO ([A] ipsilateral to the injury, [B] contralateral to the injury) or 6 weeks PO ([C] ipsilateral to the injury, [D] contralateral to the injury). At 9 weeks PO following the initial repair, all animals were subjected to another sciatic ablation. Arrows indicate sciatic transection and repair (black), saphenous ablation (blue), and sciatic ablation (green). X-axis labels indicate PO time points with respect to the initial sciatic transection and repair. The Y-axis was plotted using log10 scale according to discrete values of VF filaments (ticks at 19, 38, 75, and 134 g). All symbols are centered on these VF force values. The gray region indicates hyposensitivity (134 g). In brief, saphenous nerve collaterals are probably not responsible for more rapid sensory recovery after complete sciatic transections in PEG-fused rats. BL: Baseline; Cont: contralateral; Ipsi: ipsilateral; PEG: polyethylene glycol; PO: post-operatively; VF: von Frey filaments; w: weeks.

Saphenous ablations 6 weeks post-operatively

For rat PEG#3 that recovered both SFI and VF prior to 6 weeks PO and rat NC#4 recovered only VF at 6 weeks PO (Figure 7C and D), that a subsequent saphenous ablation at 6 weeks PO (blue arrow) produced no significant change in VF sensitivity in either rat. (Rat NC#4 exhibited slight fluctuations in VF scores commonly observed in baseline controls; see Figure 5). These data are consistent with the hypothesis that sciatic populations (#2–3), but not saphenous collateral population (#1), were responsible for recovery of mechanical nociception at 3 weeks PO or later. This was also confirmed by the sciatic ablation at 9 weeks PO that eliminated VF sensitivity in both rats.

In brief, data obtained from NC rats in Figure 7 are consistent with previous reports (Devor et al., 1979; Markus et al., 1984) that saphenous collaterals are responsible only for the initial phase of mechanical nociceptive recovery after various types of sciatic PNIs and retract upon sciatic reinnervation. For neurorrhaphy repair with PEG-fusion, saphenous collaterals and PEG-fused axons are each partly responsible for the early phase of VF recovery, and PEG-fused axons are responsible for more rapid VF recovery to baseline with (Figure 7A and B) or without (Figure 6A and B) saphenous nerve ablation.

Transient hypersensitivity of saphenous-sciatic overlapping innervation of dorsal hindpaw region

To more extensively investigate how PEG-fusion might affect innervation and sensitivity of hindpaw regions innervated also by a neighboring nerve, we conducted VF tests at another location between the 2^nd and 3^rd metatarsals on the dorsal region of rat hindpaws that are innervated by both the saphenous and the sciatic nerves (Swett and Woolf, 1985; Kambiz et al., 2014). Baseline VF withdrawal thresholds of both hindpaws in PEG-fused and NC SD rats were between 38.4 ± 0 g and 46 ± 7.3 g (Figure 8A). Subsequent to a sciatic transection and repair, almost all PEG-fused rats at 3 days PO (4/5 PEG-fused cases) and many NC rats, (4/8 NC cases) maintained baseline thresholds, but the others became hypersensitive. Both PEG-fused and NC rats exhibited significantly (P < 0.01) lower dorsal VF withdrawal thresholds of the hindpaws ipsilateral to the injury 2–6 weeks PO. However, at 12 weeks PO, PEG-fused and NC rats had mean withdrawal thresholds between 31 ± 3.6 g and 38.4 g that were no longer significantly different (P > 0.05) from baseline (Figure 8B). Individual threshold values were also within baseline values, therefore unlikely indicating hypersensitivity. No significant difference (P > 0.05) was observed between the hindpaws ipsilateral or contralateral to the injury of PEG-fused and NC rats at all tested time points.

Figure 8.

Transient hypersensitivity in a saphenous and sciatic nerve overlapped dorsal hindpaw region in SD rats.

Following sciatic transection and repair, dorsal sensory behavioral recovery was assessed between 3 days and 6 weeks PO in A and at 12 weeks PO in B between the 2^nd and 3^rd Mt. Mixed-effects model, interaction F_(21,130) = 1.3, ns; time F_{(2.8, 51)} = 6.2, P = 0.0015; treatment F_(3,22) = 9.2, P = 0.0004. Tukey’s test, **P < 0.01, ***P < 0.001. BL: Baseline; Mt: metatarsals; PO: post-operatively; SD: Sprague–Dawley; w: weeks.

Discussion

Our electrophysiological, histological, and motor-dominant behavioral criteria in many previous publications (Riley et al., 2015; Ghergherehchi et al., 2016, 2019b, 2021; Mikesh et al., 2018a, b) were used to assess the success of PEG-fusion surgeries in the current study. To investigate sensory-dominant behavioral recovery following PEG-fusion, we developed fast and reliable VF testing protocols for different dorsal regions of the hindpaw before and after transection-type sciatic PNIs. Our VF protocols attempted to minimize mechanical damage to the cutaneous sensory structures by limiting the amount of repetitive testing and the maximum force applied by VF filaments (Chaplan et al., 1994; Detloff et al., 2012; Bonin et al., 2014). We showed that the restoration of mechanical nociception of a sciatic-mediated region (lateral dorsal region of the hindpaw) was enhanced by PEG-fusion repair of sciatic transection injuries in rats. Furthermore, PEG-fused rats that had successful SFI scores at or prior to 6 weeks PO exhibited an earlier return of withdrawal thresholds to baseline, suggesting a strong correlation between excellent motor-dominant and sensory-dominant behavioral recoveries and successful PEG-fusion repairs. This correlation was independent of the sex or the strain (SD, Lewis) of the rat. We demonstrated that the enhanced sensory behavioral recovery following 3 weeks PO was almost certainly due to PEG-fused sciatic nerves and independent of the early saphenous reinnervation. Finally, we showed that chronic hypersensitivity did not occur in a sciatic-mediated lateral dorsal region or a saphenous- and sciatic-mediated dorsal region.

Midplantar VF tests have been used to assess several types of sciatic PNIs with inconsistent reports of PO recovery. For example, rats having either crush injuries or complete sciatic transections and neurorrhaphy repair exhibited severe hyposensitivity 0–3 weeks PO followed by a return to baseline responses or hypersensitivity 4–9 weeks PO at the plantar region of the hindpaw ipsilateral to the injury (Devor et al., 1979; Casals-Diaz et al., 2009; Cobianchi et al., 2014). Other studies reported that rats that received sciatic transection or ablation PNIs immediately developed hyperalgesia and allodynia maintained up to 4 or 8 weeks PO (Dowdall et al., 2005; Chacur et al., 2010).

We observed many changes (e.g., weight shifts) in hyposensitive rats following sciatic transections that created trial-to-trial variability during midplantar VF testing (Additional Video 1). Such variability may explain the inconsistency in the literature. The original midplantar VF method was designed to specifically assess mechanical allodynia, not hyposensitivity, in a neuropathic pain PNI model (Chaplan et al., 1994). Therefore, the authors most likely never encountered problems (larger filaments lifting hindpaw or not bending) described in our study. In contrast, our dorsal VF method ensured proper bending of filaments of any size and showed consistent and reliable responses. Hence, we propose that PNI models that result in any hyposensitivity should employ dorsal VF and avoid midplantar VF testing unless a better-controlled, more-reliable midplantar method is developed.

Possible effects of polyethylene glycol-fusion on somatosensory and pain perceptions

In a previous publication (Ghergherehchi et al., 2019a) and the preceding companion paper (Hibbard et al., 2025), we have presented anatomical evidence that PEG-fusion produced correct and incorrect connections between proximal and distal motor and sensory axons. That is, PEG-fusion non-specifically fuses closely apposed, severed, open axons regardless of modality or original connectivity in various combinations. In the preceding paper, we observed both typical and aberrant afferent terminal labeling after intradermal WGA-HRP injection to the dorsal aspect of the foot in PEG-fused animals as early as 2 days PO (Hibbard et al., 2025), at which time very few PEG-fused animals exhibited any mechanical nociception. In the current study, we reported a beneficial effect of PEG-fusion to restore mechanical nociception that is likely contributed by unmyelinated C fibers of the smallest caliber (Purves et al., 2017). While the SFI is predominantly motor-driven, successful SFI recovery also depends on proprioceptive feedback such as muscle length, tendon tension, and hindpaw placement that are mediated by myelinated Aα fibers. PEG-fusion almost certainly rapidly (within minutes) joins the proximal and distal ends of sensory fibers with similar or different modalities as well as sensory-motor fusions (very rare) in various combinations (Bittner et al., 2016a; Mikesh et al., 2018a; Ghergherehchi et al., 2019a; Hibbard et al., 2025). Hence, PEG-fusion would be expected to produce various sensory perceptions and behavioral recoveries, as indeed reported in a recent case study (Lopez et al., 2022). Furthermore, PEG-fusion might maintain innervation of different distal sensory targets similar to previous work showing that PEG-fusion maintains distal motor targets (Mikesh et al., 2018a, b).

Polyethylene glycol-fusion does not prevent collateral sprouting of saphenous nerve axons

All PNIs usually trigger collateral sprouting starting at 1 week PO (Lemaitre and Court, 2021). However, no sciatic collateral sprouting was observed up to 7 weeks PO subsequent to capsaicin treatment that permanently removed some (~30%) unmyelinated nociceptive C fibers in the saphenous nerve (Sántha et al., 2022). These authors proposed that the remaining C fibers, intact myelinated fibers, and minimal WD may inhibit the invasion of sciatic collaterals via the release of signaling factors essential for collateral sprouting. Many PEG-fused sciatic axons (40%–60%) were rescued and persisted up to 6 weeks PO in our present and previous studies (Mikesh et al., 2018a, b). These surviving PEG-fused axons did not prevent the invasion of saphenous sensory collaterals. If the Sántha et al. (2022) proposal is correct, then increasing the number of successfully PEG-fused sciatic axons may provide adequate trophic signals and reduce WD to suppress saphenous collateral sprouting.

Rat strains and sensory perception

Mechanical allodynia using a partial nerve denervation model has been reported for eight strains of rats, including SD and Lewis rats (Shir et al., 2001). They reported that both SD and Lewis strains exhibited lower plantar withdrawal thresholds at 9 weeks PO than baseline thresholds, but the difference between the 9 weeks PO threshold and baseline threshold in Lewis rats was much smaller than that for SD rats, suggesting potentially earlier recovery in Lewis rats. This result is consistent with our observation that dorsal hindpaw withdrawal thresholds of Lewis rats on average recovered 2 weeks earlier than SD rats. One possible explanation for this earlier recovery in Lewis rats is an anatomical difference. Bobkiewicz et al. (2017) reported significantly shorter sciatic and sural nerve lengths in Lewis rats than SD rats (~10 mm in total) which may produce a shorter reinnervation time assuming the same rate of axonal regeneration. Another possibility is that saphenous nerve collaterals reinnervate the neighboring denervated region faster in Lewis rats than SD rats.

Limitations

Mechanical nociception is only one of the many sensory modalities that are mostly not addressed in this study. In addition to the mechanical nociception studied in this paper, other sensory modalities, such as heat and/or cold nociception, should be investigated. For other sensory modalities (e.g., mechano-sensations), the reinnervation of different types of distal sensory organs should be characterized to determine the effect of PEG-fusion.

Conclusion

Currently, there are five clinical trials using PEG-fusion technology to repair transection PNIs or ablation PNIs using autografts. The present study shows that mechanical nociceptive recovery is accelerated by PEG-fusion repair of sciatic nerves following transection PNIs in an experimental rat injury model. This result is independent of rat strain or sex. Furthermore, PEG-fusion did not lead to chronic hypersensitivity in hindpaws ipsilateral or contralateral to the injury in rats. Together, these data on mechanical nociception recovery in a sensorimotor mixed sciatic nerve are consistent with non-noxious sensory data from case studies of PEG-fusion repair of varying, non-uniform injuries to the human sensory-only digital nerves (Bamba et al., 2016; Lopez et al., 2022). Our results are consistent with suggestions that PEG-fusion repair would provide a better clinical alternative to conventional methods of repairing transection PNIs using only neurorrhaphy. Our results provide additional supporting evidence to current and future clinical trials on the efficacy and safety (do no harm) of PEG-fusion to enhance sensory perception.

Additional file:

Additional Video 1: Midplantar VF larger filaments do not elicit consistent withdrawal responses.

Funding Statement

Funding: This study was supported by DOD AFIRM III W81XWH-20-2-0029 subcontract; UT POC19-1774-13; Neuraptive Therapeutics Inc. 26-7724-56 and NIH R01-NS128086 grants; Lone Star Paralysis gift (to GDB).

Data availability statement:

All relevant data are within the paper and its Additional files.