Application of In Vivo Cryotechnique for Morphological Visualization of Nerve and Vascular Structures in a Sciatic Nerve Crush Model

Abstract

After a crush injury in sciatic nerve fibers, dynamic changes in blood circulation and immune-cell mobilization occur during axonal regeneration. High-resolution visualization under near-physiological conditions is crucial for understanding these mechanisms. Conventional histological techniques introduce perfusion- and dehydration-induced artifacts that obscure circulation. We employed the in vivo cryotechnique (IVCT) to visualize blood flow within sciatic nerve fibers and assess temporal changes during regeneration. In uninjured mice, IVCT preserved native tissue architecture with minimal shrinkage compared to perfusion fixation, with superiority quantitatively shown by fractal analysis. In the crush model, hematoxylin-eosin, Luxol fast blue, and immunohistochemical staining of IVCT-prepared, freeze-substituted sections revealed axonal degeneration and regrowth. The close association between regenerating fibers and vascular structures, along with erythrocyte distribution, indicates a morphological link between nerves and blood vessels. Electrophysiological assessment using compound muscle action potentials and functional recovery measured by the sciatic functional index demonstrated restored nerve function at 28 days, consistent with histology. These findings suggest that IVCT is a useful method for analyzing peripheral nerve regeneration and vascular dynamics, thereby highlighting its potential as a novel approach in peripheral nerve research.

I. Introduction

Peripheral nerve injury induces sensorimotor dysfunction that significantly impairs activities of daily living [1–3]. Peripheral nerve regeneration involves sequential processes, including injury-induced Wallerian degeneration, axonal elongation, remyelination, and target-muscle reinnervation [4–6]. Microenvironmental changes, including in blood-flow dynamics and immune-cell mobilization, crucially modulate nerve regeneration [7]. To understand the nerve-regenerative mechanisms, high-resolution visualization of these events under near-physiological conditions is essential, and studies using histological evaluations have been conducted [8, 9] in paraffin sections prepared through perfusion fixation (PF) of peripheral nerves. PF is prone to perfusion-pressure and tissue-shrinkage-induced artifacts during alcohol dehydration, with impaired preservation of fine structures to that in vivo [9, 10]; these artifacts hinder the accurate evaluation of the regenerative microenvironment, including vascular morphology, immune-cell localization, and soluble-factor distribution. The in vivo cryotechnique (IVCT), which involves direct cryofixation of living tissues, is a promising method that overcomes these limitations and avoids PF-induced artifacts while enabling both morphological and immunohistochemical (IHC) analyses [11–13]. IVCT facilitates instantaneous preservation of all biological components in situ, which enables precise detection of rapidly changing signaling molecules and soluble factors. Although angiogenesis and blood-circulatory changes are closely associated with peripheral nerve regeneration, detailed observation using PF is difficult owing to the structural collapse of blood vessels associated with perfusion and dehydration. Given the ability of IVCT to preserve the vascular microenvironment under near-physiological conditions, it constitutes a rational approach for the detailed morphological evaluation of peripheral nerve regeneration. A previous study applied IVCT to analyze the structural responses of sciatic nerves under mechanical stretching conditions [10]. However, IVCT has not yet been used to evaluate the peripheral nerve regeneration process in a traumatic injury model. The sciatic nerve crush model was selected because it is a well-established and reproducible model that effectively recapitulates key aspects of peripheral nerve regeneration, including Wallerian degeneration, axonal regrowth, and functional recovery [14]. This model is widely used in preclinical research due to its suitability for evaluating both biological and therapeutic aspects of nerve regeneration. This study aimed to demonstrate the utility of IVCT in visualizing nerve regeneration and blood circulation in a sciatic nerve crush model.

II. Materials and Methods

Animal experiments

A total of 64 male C57BL/6J mice (12–24 weeks old, 23–28 g) were used in this study. Of these, 41 mice were used for histological analyses, 7 mice for nerve conduction studies (NCS), and 16 mice for functional behavioral assessments. Animals were age-matched and evenly allocated across experimental groups to minimize variability.

The mice were housed individually in a ventilated incubator within the animal facility under a 12-h light/dark cycle and a constant temperature of 23 ± 1°C. Food (LabDiet 5040, St. Louis, MO) and water were provided ad libitum. For surgical procedures, anesthesia was administered via intraperitoneal injection of 0.1 ml (per 10 g bodyweight) of a mixed anesthetic cocktail (0.75 mg/kg medetomidine hydrochloride, Nippon Zenyaku Kogyo Co., Koriyama, Japan; 4 mg/kg midazolam, Sandoz K.K., Tokyo, Japan; and 5 mg/kg butorphanol tartrate, Meiji Seika Pharma Co., Tokyo, Japan). For euthanasia, the same cocktail was administered at an overdose level to ensure euthanasia. All experimental protocols were conducted in accordance with institutional guidelines and were approved by the Animal Ethics Committee of Health Science University (protocol code: 2023-001). The study was conducted and reported in accordance with the ARRIVE guidelines [15].

To compare sciatic nerve morphology following IVCT and PF under uninjured conditions, 12 intact C57BL/6J mice were randomly assigned to either the IVCT or PF group (Fig. 1A; n = 6 per group). Subsequently, sciatic nerve morphology following IVCT and PF under injured conditions was compared using a sciatic nerve crush injury model. Specifically, 8 mice were randomly assigned to either the IVCT or PF group (n = 4 per group), and histological analyses were performed at 1, 3, 7, and 28 days after injury (one mouse per group at each timepoint). Morphological differences were evaluated by histological analyses (Fig. 1A).

Fig. 1.

Conceptual diagram of the experimental approach used to evaluate IVCT in peripheral nerve regeneration. (A) Sciatic nerve morphology was compared between in vivo cryotechnique (IVCT) and perfusion fixation (PF) under both uninjured and injured conditions. In intact C57BL/6J mice, animals were randomly assigned to IVCT or PF, and sciatic nerves were processed and subjected to histological sectioning and morphological evaluation. Under injured conditions, sciatic nerve morphology was analyzed using a sciatic nerve crush model, with injured nerves processed by IVCT or PF and evaluated using the same histological and morphological workflow. (B) To further assess the applicability of IVCT in a sciatic nerve crush model, mice were assigned to crush-injury or sham-operated groups and cryofixed using IVCT. Sciatic nerves were collected at 1, 3, 7, and 28 days after injury and subjected to histological sectioning and morphological analysis. Sham-operated sciatic nerves were collected at the corresponding 1-day timepoint. Figure created with BioRender.com (Publication license).

To further evaluate the applicability of IVCT in a sciatic nerve crush model, a total of 21 mice were used. In the newly assigned animals, the crush-injury group included 4 mice at each postoperative timepoint (days 1, 3, 7, and 28), resulting in 16 mice in total. The sham group consisted of 5 mice (n = 5) and was evaluated only at postoperative day 1. To equalize the number of samples across timepoints, four previously obtained IVCT-processed injured sciatic nerve specimens that had been used for comparison with perfusion fixation (one specimen at each timepoint: days 1, 3, 7, and 28) were incorporated into the present analysis. Consequently, the crush-injury group comprised 5 mice per timepoint, yielding a total of 20 IVCT-treated injured sciatic nerve samples for histological analysis (Fig. 1B).

For NCS, 7 mice were used, with the right and left sciatic nerves serving as crush and sham sides, respectively. Compound muscle action potential (CMAP) were recorded at baseline and 7, 28, and 42 days post-injury. To confirm reproducibility, additional baseline measurements were taken one week before injury. Eight mice per group were used for functional evaluation, conducted before injury and on postoperative days (PODs) 3, 5, 7, 10, 14, 21, and 28. Sample sizes were determined with reference to previous studies using analogous models and were considered sufficient to detect biologically relevant differences [16, 17].

Surgical procedures

The sciatic nerve crush procedure was performed as described in a previous study [18]. The mice were anesthetized and placed in the prone position. To expose the right sciatic nerve, a longitudinal skin incision was made from the gluteal region to the mid-thigh region. The sciatic nerve was then crushed at a 2-mm segment just distal to the sciatic notch using smooth forceps (F053-0802F; FRIGZ, Tokyo, Japan) for 180 sec. After compression, the nerve was repositioned, and the muscle and skin layers were sutured with a 3-0 nylon thread. In the sham-operated group, only exposure of the sciatic nerve was performed without inducing nerve injury, followed by suturing of the muscle and skin. In both groups, to prevent autotomy, a saturated aqueous solution of picric acid (2,4,6-trinitrophenol; Sigma-Aldrich, St. Louis, MO) was applied to the toes of the operated limb [19].

IVCT and freeze-substitution fixation for mouse sciatic nerves

IVCT for the mouse sciatic nerve was performed as previously described [10]. After exposing the right sciatic nerve, a cryogen composed of a −193°C isopentane–propane mixture pre-cooled in liquid nitrogen was directly applied to the nerve for cryofixation. Subsequently, additional liquid nitrogen was poured over the area to ensure complete freezing of the sciatic nerve, which was carefully excised from the body. The freeze-substitution solution consisted of an anhydrous acetone containing 2% paraformaldehyde (PFA). The excised sciatic nerves were immersed in this solution and stored at −80°C for 48 hr. Thereafter, the samples underwent a temperature-controlled substitution protocol: 2 hr at −30°C, 2 hr at −10°C, 2 hr at 4°C, and finally 2 hr at room temperature. The nerves were then rinsed with pure acetone, immersed in xylene, and embedded in paraffin. All sciatic nerve samples were sectioned using a sliding microtome (Tissue-Tek^® Sliding Microtome IVS-410; Sakura Finetek Japan Co., Ltd., Tokyo, Japan) and mounted onto Matsunami Adhesive Slide-coated glass slides (Matsunami Glass Ind., Ltd., Osaka, Japan).

PF and alcohol-dehydration for mouse sciatic nerves

Under deep anesthesia, the mice were transcardially perfused with 4% PFA in 0.1 M phosphate-buffered saline (PBS). The excised sciatic nerves were post-fixed by overnight immersion in the same fixative. Subsequently, the tissues were dehydrated through a graded ethanol series, cleared with xylene, and embedded in paraffin.

Hematoxylin/eosin and Luxol fast blue staining

Hematoxylin and eosin (H&E) staining was performed on 10-μm paraffin sections. After deparaffinization in xylene, the sections were rehydrated using a graded ethanol series. Some sections were processed using the standard H&E staining protocol. Luxol fast blue (LFB) staining was performed according to the following procedure. The sections were immersed in xylene thrice for 10 min each, followed by 100% ethanol twice for 1 min, 90% ethanol for 1 min, and 80% ethanol for 1 min for stepwise rehydration. The slides were then transferred to a staining jar containing 0.1% LFB solution and incubated overnight at 50°C. After staining, the sections were immersed in 95% ethanol for 1 min and rinsed under running water for 10 min. Differentiation was performed by immersing the sections in 0.05% lithium carbonate solution for 40 sec, followed by immersion in 70% ethanol three times for 90 sec each. The sections were again rinsed under running water for 10 min and then dehydrated sequentially in 70%, 95%, and 100% ethanol for 1 min each. The final dehydration step included three 1-min immersions in 100% ethanol. The slides were then cleared in xylene for 1, 5, and 10 min and mounted using MOUNT-QUICK (Cosmo Bio Co., Ltd., Tokyo, Japan) and No. 1 cover glasses (Matsunami Glass Ind., Ltd.).

Immunohistochemistry

All analyses were performed on 10-μm paraffin sections. Initially, IHC staining was performed to visualize soluble proteins. After incubation in 0.3% hydrogen peroxide in PBS and 2% gelatin in PBS for 1 hr each, the sections were immunostained overnight at 4°C in 2% gelatin (from cold water fish skin, Sigma-Aldrich) containing a mouse polyclonal antibody against mouse immunoglobulin G1 (IgG1), heavy chain (1:1000, A90-105A, BETHYL Inc., Montgomery, TX), as previously described [20, 21]. Subsequently, the sections were incubated for 1 hr at room temperature with horseradish peroxidase-conjugated anti-goat IgG secondary antibody (whole molecule, 1:400, ab6667, Abcam, Cambridge, UK) for IgG1 detection. Immunoreactivity was visualized using 3-min development with 3,3'-diaminobenzidine (DAB) in a buffer containing hydrogen peroxide. Other sections were processed for immunofluorescence staining. After deparaffinization in xylene and rehydration through a graded ethanol series, the sections were stained in a humidified chamber. A PAP pen was used to prevent the staining solution from spreading. All sections were washed with 0.3% PBS containing Triton X-100 (PBS-Tx) and incubated in Blocking One Histo (Nacalai Tesque Inc., Kyoto, Japan) for 1 hr at room temperature. The antibody diluent was prepared using PBS containing 5% Blocking One Histo. Sections were incubated overnight at 4°C with the following primary antibodies diluted in the prepared buffer: mouse monoclonal anti-neurofilament 200 kDa (NF200), clone RT97, IgG1 (1:1000, MAB5262, Merck Millipore, Burlington, MA), and rabbit polyclonal anti-growth associated protein-43 (GAP-43) antibody (1:100, AB5220, Merck Millipore). For negative controls, the sections were treated with an antibody diluent without primary antibodies. After incubation with primary antibodies, sections were incubated for 1 hr at room temperature with the following secondary antibodies: Alexa Fluor^® 488 goat anti-mouse IgG (H+L) antibody (1:500, A-11001, Thermo Fisher Scientific, Waltham, MA) and Alexa Fluor^® 594 goat anti-rabbit IgG (H+L), cross-adsorbed secondary antibody (1:500, A-11012, Thermo Fisher Scientific), also diluted in antibody diluent. Following 2–4 washes with PBS, the sections were mounted using VECTASHIELD^® Antifade Mounting Medium with DAPI (H-1200, Vector Laboratories, Newark, CA) and sealed with No. 1 cover glasses.

Imaging and quantitative analysis

All sections were examined using an all-in-one fluorescence microscope (BZ-X800; KEYENCE, Osaka, Japan). After region selection under a 20× objective lens, images were captured using either a 20× or 40× objective. In the crush group, the region of interest was the injury site, whereas in the sham group, the corresponding anatomical region was selected. For each staining method, the exposure time and imaging parameters (gain, sharpness, and haze reduction) were standardized across all time points and groups. Quantitative image analysis was performed using MATLAB (MathWorks, Natick, MA) and ImageJ software (National Institutes of Health, Bethesda, MD). In the present study, blinding of the evaluators was not conducted [4, 22, 23]. Fractal analysis (FA) was employed to objectively assess morphological differences in the sciatic nerve between the fixation methods. FA is a validated quantitative approach used to evaluate collagen fiber alignment in healing ligaments and has been applied in various biological studies [23–26]. In this study, we hypothesized that artifacts such as perfusion pressure and alcohol dehydration induced by PF could affect the microstructural alignment of the sciatic nerve, and FA was used to compare these changes with those preserved by IVCT. One section was selected from each animal, and one image was acquired. The images were binarized using a threshold value automatically determined by MATLAB. The threshold was uniformly applied across all images, referencing the original images from both the IVCT and PF groups. Ten random regions of interest (ROIs) measuring 100 μm × 100 μm were selected from each image. The FA values were calculated based on the Minkowski–Bouligand dimension using a MATLAB routine. The average of the ten FA values obtained from each image was used as a representative value. Additionally, in H&E-stained sections, to compare the morphological features among the fixation methods, the number of nuclei within a 100 μm × 100 μm ROI was counted at 400× magnification. This analysis was based on the assumption that chemical fixation causes cellular shrinkage, leading to a higher nuclear density within a given area, and is a previously established method [27]. For IHC staining with the IgG1 antibody, positive areas were quantified using ImageJ to compare the preservation of soluble proteins. Three sections were selected from each animal, and three images per section (total of nine images) were acquired. Images were converted to 8-bit grayscale and binarized using the “Threshold” function in ImageJ. The threshold value was uniformly set across all images based on representative images from both groups. The positive area within each 100 μm × 100 μm ROI was measured using the “Rectangle” tool. The average of three measurements per section was used as a representative value. For LFB staining and the axonal marker NF200, positive areas were quantified using ImageJ. Three sections were selected from each animal, and three images per section (total of nine images) were acquired. The results are expressed as the percentage of the positive area in the crush group relative to that in the sham group. Images were converted to 8-bit grayscale and binarized using the “Threshold” function in ImageJ. The threshold value was uniformly set across all images based on the representative images from both groups. The positive area within each 100 μm × 100 μm ROI was measured using the “Rectangle” tool. The average of three measurements per section was used as a representative value.

Nerve conduction study

To longitudinally assess nerve regeneration after sciatic nerve crush injury, NCS was conducted using a transcutaneous method based on previously reported protocols [28–30]. Mice were sedated with 5% isoflurane in a plastic induction chamber and maintained under 3% isoflurane via a nose cone during the procedure. The body temperature was maintained at 34–36°C using a heating pad throughout the experiment. Stimulation was performed using sterile disposable needle electrodes (M-type SP No. 5, 0.25 × 30 mm; SEIRIN, Shizuoka, Japan) inserted near the sciatic nerve. A stimulator (SEN-7103, Nihon Kohden, Aichi, Japan) was used to deliver electrical pulses at an intensity of 1.0 mA and a frequency of 1 Hz (adjustable from 1–3 mA in 0.1-mA increments). Recording was conducted using sterile disposable needle electrodes connected to a BioAmp system (high-pass filter: 0.1 Hz; low-pass filter: 1 Hz; ADInstruments, Sydney, Australia) [31] and acquired using a PowerLab 4/35 data acquisition system (ADInstruments). The anode was placed 2 mm apart from the cathode. The recording electrode was inserted into the lateral belly of the gastrocnemius muscle, and the reference electrode was placed above the Achilles tendon of the right foot. The ground electrode was positioned at the base of the tail. The amplitude (mV) of the CMAP was analyzed.

Functional analysis

To assess functional recovery following sciatic nerve crush injury, the sciatic functional index (SFI) was used based on previously established protocols [32, 33]. Prior to injury, the mice were trained to walk along a walking track measuring 400 mm × 50 mm × 50 mm. Gait patterns were recorded using a video camera. During each gait cycle, the following parameters were measured: print length (PL), defined as the distance from the heel to the third toe, and toe spread (TS), defined as the distance between the first and the fifth toes. These parameters were measured in both the injured and uninjured hind limbs. The SFI was calculated using a regression formula reported in previous studies [32, 33].

Statistical analysis

Statistical analyses were performed using GraphPad Prism software (version 10; GraphPad Software, San Diego, CA). Data are presented as mean ± standard deviation (SD). Student’s t-test was used to compare data assessing differences between fixation methods and to evaluate CMAP amplitudes at baseline and one week later in the NCS. For all other comparisons, a repeated-measures analysis of variance (ANOVA) was performed. The normality of the data distribution was confirmed using the Shapiro–Wilk test. Statistical significance was set at p < 0.05.

III. Results

Morphological comparison between IVCT and PF in uninjured sciatic nerve

Representative IVCT-processed sciatic nerve tissues showed no visible ice crystals on either the surface or in the deeper tissue regions and widely open blood vessels containing flowing erythrocytes were well preserved (Fig. 2A). H&E staining revealed distinct morphological differences between the IVCT and PF groups. In the IVCT group, sciatic nerve fibers of various sizes were well preserved, and both the epineurium and perineurium were clearly distinguishable (Fig. 2A). In contrast, PF-treated nerves exhibited overall tissue and cellular condensation, presumably owing to perfusion pressure and alcohol dehydration, with compressed nerve fibers and blood vessels in close proximity (Fig. 2B). IgG1-antibody-based IHC staining revealed clear between-group differences in the localization of soluble proteins. In the IVCT group, IgG1 was well preserved in the interaxonal spaces and blood vessels (Fig. 2C) whereas, in the PF group, pronounced IgG1 leakage was observed (Fig. 2D), and this was consistent with the anatomical localization of open blood vessels containing erythrocytes observed in H&E-stained sections. Next, longitudinal sections were examined to compare the morphological characteristics of sciatic nerve fibers between the IVCT and PF groups. Similar to the transverse sections, longitudinal IVCT-preserved tissues showed well-aligned sciatic nerve fibers, whereas PF-treated nerves exhibited waviness and disorganization of fiber trajectories (Fig. 2E, F). Additionally, IHC staining revealed that IgG1 was well preserved within the blood vessels in the IVCT group (Fig. 2G, H). No signal was detected in the negative-control sections processed without primary antibody, supporting the specificity of IgG1 localization (Supplementary Fig. S1A–D). FA showed significantly lower values in the IVCT group than in the PF group (Fig. 2I). The number of nuclei within the same tissue area was significantly greater in the PF group than in the IVCT group (Fig. 2J). The IgG1-positive area was significantly larger in the IVCT group (Fig. 2K). Therefore, IVCT potentially effectively prevents structural shrinkage caused by perfusion pressure and dehydration and thereby preserves a morphology that more closely reflects the physiological state with retained soluble proteins.

Fig. 2.

Morphological preservation of sciatic nerves processed with IVCT or PF. (A) Low-magnification IVCT image. (i, ii) High-magnification views show no ice crystals and intact vessels with circulating erythrocytes. (A, B) H&E staining revealed preserved large/small fibers (i) and clear epineurium/perineurium (ii) in IVCT, versus tissue condensation with compressed fibers (iii) and vessels (iv) in PF. White arrowheads: perineurium; black arrowheads: vessels. (C, D) IgG1 immunostaining showed retention in interaxonal spaces (v) and vessels (vi) in IVCT, but leakage in PF (vii, viii). (E, F) Longitudinal H&E sections showed preserved fiber trajectories in IVCT, versus waviness/disorganization in PF. (G, H) IgG1 staining confirmed intravascular retention in IVCT and leakage in PF. Longitudinal sections underwent FA. (I) FA values were lower in IVCT; (J) nuclear counts higher in PF; (K) IgG1-positive area larger in IVCT. Bars = 100 μm (A–D), 50 μm (E–H, i–viii). *p < 0.05, **p < 0.01. IVCT, in vivo cryotechnique; PF, perfusion fixation; H&E, hematoxylin and eosin; FA, fractal analysis.

Morphological comparison of injured sciatic nerves processed with IVCT and PF

To directly compare morphological preservation between IVCT and PF in injured sciatic nerves, H&E-stained sections were analyzed at 1, 3, 7, and 28 days after sciatic nerve crush injury (Fig. 3). At 1 day after injury, IVCT-processed nerves showed discontinuity of nerve structures, with scattered erythrocytes distributed within the intervening gaps (Fig. 3A). In PF-processed nerves, the nerve trajectory appeared markedly tortuous, and some erythrocytes aggregated within blood vessels were also observed scattered within the nerve gaps (Fig. 3B). At 3 days after injury, IVCT preserved fine regenerating nerve fibers located adjacent to erythrocytes, and although blood vessels were dilated, erythrocytes were retained within the vascular lumen (Fig. 3C). In contrast, PF-processed nerves exhibited aggregation of erythrocytes at the injury site, making it difficult to observe local circulatory dynamics associated with dilated vessels (Fig. 3D). At 7 days after injury, IVCT-processed nerves demonstrated regenerating fibers aligned along erythrocytes and dilated blood vessels (Fig. 3E). In PF-processed nerves, erythrocyte aggregation was observed in some vessels, while erythrocyte leakage from dilated vessels was evident in others (Fig. 3F). By 28 days after injury, IVCT-processed nerves showed regression of vessel dilation and restoration of continuous nerve fibers traversing the injury site (Fig. 3G). PF-processed nerves also exhibited recovery of continuous nerve fibers across the injury site; however, dilated blood vessels were still observed in some regions (Fig. 3H).

Fig. 3.

Morphological preservation of injured sciatic nerves processed with IVCT or PF. H&E staining of injured sciatic nerves at 1, 3, 7, and 28 days after injury, processed with in vivo cryotechnique (IVCT) or perfusion fixation (PF). (A) IVCT at 1 day after injury, showing scattered erythrocytes at the injury site (i). (B) PF at 1 day after injury, showing disrupted axonal continuity with scattered erythrocytes among nerve fibers (ii). (C) IVCT at 3 days after injury, showing regenerating nerve fibers adjacent to erythrocytes and dilated blood vessels (iii). (D) PF at 3 days after injury, showing focal aggregation of erythrocytes at the injury site, precluding observation of continuously supplied circulatory dynamics (iv). (E) IVCT at 7 days after injury, showing regenerating nerve fibers adjacent to erythrocytes and dilated blood vessels (v). (F) PF at 7 days after injury, showing persistent erythrocyte leakage from dilated blood vessels (vi). (G) IVCT at 28 days after injury, showing regression of dilated blood vessels and restoration of continuous nerve fibers (vii). (H) PF at 28 days after injury, showing restoration of continuous nerve fibers; however, persistent erythrocyte leakage is observed in some dilated blood vessels (viii). Black arrowheads indicate blood vessels; white arrows indicate erythrocytes. Bars = 100 μm (A–H), 50 μm (i–viii). H&E, hematoxylin and eosin.

Collectively, these findings indicate that IVCT provides superior preservation of vascular integrity and axonal architecture during nerve regeneration following injury, whereas PF is associated with erythrocyte leakage and persistent distortion of regenerating nerve fibers.

Histological evaluation of nerve regeneration after injury using IVCT

To further evaluate the applicability of IVCT to the sciatic nerve crush model, detailed histological and IHC analyses focusing on nerve regeneration after injury were performed using IVCT-processed specimens.

LFB staining to assess myelin degeneration in the injured area showed that the LFB-positive area, expressed as the ratio of the crush group to the sham group, was significantly reduced at 7 days compared to 1 day (Fig. 4E–I). IHC demonstrated axonal regeneration within the crush region (Fig. 5A–D). On day 1, NF200 immunoreactivity markedly decreased, whereas GAP43-positive regenerating axons crossed the injury site from 3 days to 7 days. H&E staining confirmed the continuity of nerve fibers, whereas LFB staining revealed myelin sheaths surrounding the axons (Supplementary Fig. S2A, B). IHC showed minimal GAP43 expression, while continuous NF200-positive signals were observed (Supplementary Fig. S2C). Quantitative analysis revealed that the NF200-positive area within the crush site significantly increased at 28 days, compared to that at 1, 3, and 7 days (Fig. 6A–M). GAP43-positive axons were prominent at 3 and 7 days (Fig. 6I–L), and no signal was detected in the negative-control sections processed without primary antibodies (Supplementary Fig. S3A–J).

Fig. 4.

Time-course changes in myelin degeneration in injured sciatic nerves processed by IVCT. LFB staining evaluated myelin (A–D); LFB-positive area (crush/sham ratio) was reduced on 7 days vs 1 day (I). Bars = 100 μm (A–D). *p < 0.05, **p < 0.01. IVCT, in vivo cryotechnique; LFB, Luxol fast blue.

Fig. 5.

Time-course of axonal regeneration by immunohistochemistry. Immunofluorescence with NF200 (green) and GAP43 (red) antibodies. (A) 1 day: NF200 decreased in lesion. (B, C) 3–7 days: GAP43-positive regenerating axons traversed lesion. (D) 28 days: NF200 restored across crush site. Sham nerves showed continuous NF200 and minimal GAP43 throughout. Bar = 500 μm.

Fig. 6.

Quantitative analysis of regeneration. (A–H) 1 day: NF200 decreased; GAP43-positive axons present 3–7 days. (I–L) GAP43 prominent at 3 and 7 days. (M) NF200-positive area increased at 28 days vs. 1–7 days. Bars = 100 μm. *p < 0.05, **p < 0.01.

Functional recovery assessed by electrophysiological and behavioral analyses

Sciatic nerve regeneration was further confirmed by NCS. The reproducibility of the method was validated by comparing the CMAP amplitudes at baseline and 1 week later, without significant differences (Fig. 7A). Longitudinal evaluation revealed a significant decrease in CMAP amplitude at 7 days compared to pre-injury levels, followed by significant recovery at 28 and 42 days. By 42 days, CMAP amplitude returned to pre-injury levels (Fig. 7B). No significant changes were observed in the sham group at any timepoint (Fig. 7C). Functional recovery was assessed using the SFI. In the crush group, SFI was significantly reduced at 3 days compared to baseline, followed by a progressive recovery from 5 days onward. By 28 days, SFI values resembled the pre-injury levels (Fig. 7D). In the sham group, the SFI remained unchanged throughout the observation period (Fig. 7E).

Fig. 7.

Electrophysiological and functional assessments. (A) NCS reproducibility confirmed by unchanged CMAP amplitudes at baseline and 1 week. (B) CMAP decreased at 7 days vs baseline, recovered at 28 and 42 days, reaching baseline by 42 days. Asterisks indicate the level of significance as follows: ** p < 0.01, *** p < 0.001, and **** p < 0.0001. (C) Sham CMAP unchanged. (D) SFI decreased at 3 days in crush group, improved from 5 days, no difference from baseline at 28 days. (E) Sham SFI unchanged. In D and E, different letters: p < 0.05; identical letters: no difference. NCS, nerve conduction study; CMAP, compound muscle action potential; SFI, sciatic functional index.

IV. Discussion

Methodological advantages of IVCT over chemical fixation

In this section, we discuss the methodological advantages of IVCT with freeze substitution over conventional chemical fixation for the structural evaluation of sciatic nerve morphology. Chemical fixation commonly causes artifacts, such as shrinkage of sciatic nerve fibers. These distortions, mainly induced by the fixative itself and subsequent alcohol dehydration, can be effectively avoided by employing IVCT with freeze substitution [10]. The sciatic nerve is a mixed nerve composed of motor, sensory, and autonomic fibers. H&E staining is widely used to assess sciatic nerve structure [34]. However, most previous histological evaluations relied on chemically fixed specimens, likely introducing artifacts that altered nerve-fiber morphology. In this study, we demonstrated preservation of morphology close to the physiological state by comparing H&E-stained images using FA and nuclear density within a standardized area. Therefore, IVCT with freeze substitution allows for a more accurate structural assessment of nerve morphology than conventional chemical fixation. FA and nuclear density evaluation are quantitative and reliable analytical methods that have been widely applied in various biological studies. Furthermore, IHC staining revealed that IgG1 was well preserved within sciatic nerve fibers in the IVCT group. IgG1 is a serum protein of intermediate molecular weight (approximately 150 kDa) and is typically localized within blood vessels and the interstitium [21, 35]. In the present study, the IgG1-immunopositive area was expanded in IVCT-treated samples, with particularly strong immunoreactivity within the blood vessels. However, IgG1 immunolocalization in PF-treated blood vessel samples was markedly altered, likely owing to the loss of blood circulation and the effects of alcohol dehydration-induced chemical fixation. These morphological findings indicate that IVCT with freeze substitution effectively avoids PF-induced artifacts. Furthermore, IVCT with freeze substitution allows sciatic nerve structures to be visualized in a state that more closely reflects physiological conditions.

Preservation of the vascular microenvironment and nerve–vessel interactions

This section focuses on the ability of IVCT to preserve the vascular microenvironment after nerve injury and to enable accurate visualization of anatomical and functional interactions between regenerating nerve fibers and blood vessels. A key advantage of IVCT over PF is its ability to prevent vascular collapse. Blood vessels accompanying sciatic nerve fibers supply essential trophic factors and play vital roles in neuronal homeostasis [36]. Angiogenesis is closely involved in peripheral nerve regeneration [6, 7]. With advances in microscopy and imaging techniques, more precise evaluations of the anatomical relationships between nerves and blood vessels are possible. However, artifacts from tissue excision and circulatory disruption in PF-based specimens may significantly distort the apparent volumes and densities of blood vessels. The problem is especially pronounced in crushed nerves. In such cases, increased vascular permeability makes the tissue more susceptible to perfusion pressure [37, 38].

In the present study, PF-treated injured sciatic nerve specimens exhibited pronounced erythrocyte leakage into the surrounding tissue at 3 and 7 days after injury, with erythrocytes no longer retained within the dilated blood vessels. This finding suggests that exsanguination and perfusion pressure may deform and collapse the structurally fragile vasculature after injury, resulting in vascular lumen instability. In contrast, IVCT-processed sciatic nerves at the same timepoints showed well-preserved dilated blood vessels containing circulating erythrocytes, and the anatomical relationships between the vascular lumen and adjacent nerve fibers were maintained in a manner closely reflecting physiological conditions.

Collectively, these observations indicate that PF-based specimens may underestimate or distort the post-injury vascular microenvironment, whereas IVCT enables more faithful visualization of vascular structures and blood flow during the regenerative process without the confounding effects of exsanguination and perfusion pressure.

IVCT-treated sciatic nerves showed well-preserved dilated blood vessels containing flowing erythrocytes at 3 and 7 days. Immature regenerating nerve fibers were observed in close proximity to the erythrocytes. Following nerve injury, oxygen demand increases dramatically, and hypoxia-inducible macrophages secrete vascular endothelial growth factor A (VEGF-A), which triggers angiogenesis from existing vessels [7]. Most macrophages are rapidly recruited from the bone marrow through the bloodstream [8]. This underscores the close anatomical and functional relationship between the regenerating nerve fibers and vasculature. In the present study, we simultaneously visualized regenerating nerve fibers and circulating erythrocytes using IVCT. This provides important morphological evidence that may support previous findings. Furthermore, it suggests a close anatomical and functional relationship between regenerating nerves and vasculature.

Structural and functional correlates of nerve regeneration

Here, we discuss how the structural changes observed in IVCT-processed specimens correspond to the temporal progression of nerve regeneration and functional recovery following sciatic nerve crush injury. The ratio of LFB-positive area was significantly associated with nerve regeneration. LFB has a strong affinity for phospholipids, particularly myelin. After sciatic nerve crush, myelin distal to the lesion undergoes Wallerian degeneration, which is characterized by the degradation and phagocytosis of myelin by Schwann cells and macrophages. As degeneration progresses, the LFB staining intensity decreases [4]. We found a significant reduction in the LFB-positive area at 7 days compared to 1 day, which indicated that IVCT with freeze-substitution accurately captures myelin degeneration, similar to conventional fixation methods. To further characterize axonal regeneration, IHC was performed on the IVCT-prepared sections. The NF200-positive area was significantly associated with axonal regeneration. NF200 is a neurofilament marker that primarily localizes to the axonal cytoskeleton. During Wallerian degeneration, damaged axons are degraded by Schwann cells and macrophages, with decreased NF200 staining [6]. In contrast, regenerating axons extend from proximal stumps that remain connected to the cell bodies. In this study, we used both NF200 and GAP43 antibodies to distinguish between the degenerating and regenerating axons. GAP43 is a phosphorylated growth-associated protein that is selectively expressed in regenerating axons [39]. At 3 and 7 days, GAP43-positive axons were observed, indicating the elongation of regenerating axons within the injury site. By 28 days, GAP43-positive axons were no longer detected whereas the area of NF200-positive axons had significantly expanded. This shift suggests that regenerating axons had transitioned into a more mature state, with ongoing structural reconstruction. These structural changes were paralleled by the results of NCS. Specifically, CMAP amplitude was significantly reduced at 7 days, recovered markedly by 28 days, and returned to baseline by 42 days. Therefore, GAP43-positive regenerating axons had matured into NF200-positive axons, which led to functional reconnection with skeletal muscle. This temporal correspondence supports a consistent relationship between structural regeneration and functional recovery [40]. The SFI showed a transient decline at 3 days, followed by gradual improvement from 5 days onward, without significant difference from baseline by 28 days. This pattern corresponded with the expansion of NF200-positive axons and recovery of CMAP amplitude, which further supports the notion that structural maturation was accompanied by functional restoration. This study clearly demonstrates that the combination of IVCT and freeze-substitution enables high-fidelity preservation of both vascular flow and structural features in sciatic nerves following crush injury, while closely reflecting physiological conditions. The IHC findings and the results of NCS and SFI are consistent with previous reports. Moreover, they validate the present observations. The morphological data obtained via IVCT accurately correspond to the functional stages of nerve regeneration. Notably, the simultaneous visualization of regenerating axons and circulating erythrocytes provides morphological insight into the close relationship between nerves and vasculature. This approach may also capture phenomena that are difficult to detect using PF-based methods. This study demonstrated that the combination of IVCT and freeze-substitution enables detailed visualization of endogenous nerve-regeneration processes in crushed sciatic nerve fibers. This approach represents a valuable methodological platform for advancing basic research into mechanisms of nerve regeneration and evaluating the efficacy of conservative therapeutic strategies.

Limitations

This study has some limitations. First, our analysis did not capture subcellular structures. Angiogenesis and nerve regeneration are regulated by complex cellular processes–such as tip cell dynamics, Schwann cell-mediated Bungner band formation, and interactions with trophic and immune factors. These processes require higher-resolution techniques, including electron microscopy, for detailed evaluation. Second, image analysis was not blinded, potentially introducing detection bias, although we attempted to minimize this by following standardized and validated procedures. Finally, differences in fixative composition—2% PFA in acetone for IVCT versus 4% PFA in PBS for PF—could have confounded the observed morphological differences. These limitations should be considered when interpreting our findings.

Conclusion

Using IVCT with freeze-substitution, the morphological features of sciatic nerve fibers and the distribution of blood flow can be observed without artifacts that are typically introduced by conventional chemical fixation. In the injury region, sciatic nerve fibers were in close proximity to circulating erythrocytes. Axonal regeneration was precisely detected by IHC, and the observed findings coincided with axonal elongation. Functional recovery in the sciatic nerve crush model was further supported by NCS and functional analysis. Therefore, IVCT with freeze-substitution potentially is a valuable approach for evaluating the close interplay between nerve regeneration and blood circulation.

V. Grants Awarded

This research was funded by JSPS KAKENHI (Grant Number 23K19905) and a research grant from the Health Science University.

Supplementary Material

Supplementary Fig. S1

Supplementary Fig. S2

Supplementary Fig. S3