Primary antibodies used in this study included cystathionine $\gamma$-lyase (CSE; 1:1000, 12217-1-AP, Proteintech Group Inc., Rosemont, IL, USA), $\beta$-catenin (1:1000, #9562, Cell Signaling Technology, Boston, MA, USA), phosphorylated Akt (p-Akt; 1:1000, #9271, Cell Signaling Technology, Boston, MA, USA), phosphorylated ERK1/2 (p-ERK1/2; 1:1000, #9101, Cell Signaling Technology, Boston, MA, USA), cyclin D1 (CCND1; 1:1000, #PA5-119058, Invitrogen, Carlsbad, CA, USA), Krox20 (1:1000, #ab43020, Abcam, Cambridge, MA, USA), c-Jun (1:1000, #558036, BD Biosciences, San Jose, CA, USA), neurofilament (NF; 1:1000, #841001, Biolegend, San Diego, CA, USA), myelin basic protein (MBP; 1:1000. #ab980, Millipore, Burlington, MA, USA) and Sox10 (1:1000, Sc-365692, Santa Cruz Biotechnology, Santa Cruz, CA, USA). Secondary antibodies included Alexa fluor-488 donkey anti-mouse (1:20,000, #A32766, Invitrogen, Carlsbad, CA, USA), alexa fluor-488 donkey anti-rabbit (1:20,000, #A32790, Invitrogen, Carlsbad, CA, USA), alexa fluor-594 donkey anti-mouse (1:20,000, #A10037, Invitrogen, Carlsbad, CA, USA), and alexa fluor-594 donkey anti-rabbit (1:20,000, #A10042, Invitrogen, Carlsbad, CA, USA).

Five-week-old healthy male C57BL/6 mice were procured from Samtako Bio Korea (Osan, Korea) and maintained under standard laboratory conditions (23 $\pm$ 1 °C, 12-hour light/dark cycle) with ad libitum access to food and water. All procedures in the mouse experiments were conducted in accordance with the Kyung Hee University Committee on Animal Research guidelines, with approval, ensuring efforts to minimize both animal suffering and the number of animals used [Protocol number #KHSASP-21-463].

For the in vivo peripheral nerve injury model, sciatic nerve axotomy was conducted under sterile conditions, as previously described [8]. Anesthesia was induced via intraperitoneal administration of pentobarbital sodium (35 mg/kg body weight, #P3761, Sigma-Aldrich, St. Louis, MO, USA). Pentobarbital sodium was dissolved in sterile saline to prepare a 50 mg/mL stock solution, filtered through a 0.22 µm syringe filter, and stored at 4 °C protected from light. The dosage was adjusted according to body weight to achieve surgical anesthesia. Following transection, the distal segment of each sciatic nerve was enclosed in a 10-mm blind-ended polyvinyl chloride (PVC) tube filled with gel foam pre-soaked in Flavo and maintained in situ for 1 to 4 days. Based on this gross observation, we conducted a pharmacoactivity assessment using an in vivo system with various concentrations ranging from 1 mM to 100 mM. On each designated post-operative day, animals were euthanized using CO₂ asphyxiation. Mice were euthanized by gradual exposure to CO₂ at a displacement rate of 20–30% of the chamber volume per minute. The distal nerve segments were collected for further analysis. Intact sciatic nerves from uninjured mice served as controls.

For ex vivo experimentation, sciatic nerve explant cultures were established based on a previously reported protocol [9]. In brief, sciatic nerves were aseptically isolated from 5-week-old male C57BL/6 mice using fine iris scissors (Fine Science Tools, Foster City, CA, USA) under sterile surgical conditions. The connective tissue surrounding each nerve was carefully removed in phosphate-buffered saline (PBS, #10010-023, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) under a stereomicroscope to ensure clean isolation. The nerves were then cut into 3–4 segments, each measuring approximately 2–3 mm in length, using the same fine scissors. The dissected nerve pieces were transferred to Dulbecco’s Modified Eagle Medium (DMEM; #SH30243.01, HyClone, Logan, UT, USA) supplemented with 10% fetal bovine serum (FBS, #16000-044, Gibco, Thermo Fisher Scientific, Waltham, MA, USA) and 1% penicillin-streptomycin (#15140-063, Gibco, Thermo Fisher Scientific, Waltham, MA, USA). Cultures were maintained with or without Flavo for three days in a humidified incubator at 37 °C and 5% CO₂.

Tissue sections and teased nerve fibers were fixed with 4% paraformaldehyde (PFA, #158127, Sigma-Aldrich, St. Louis, MO, USA), rapidly frozen using dry ice, and cryosectioned into 6 µm slices. For immunostaining, the slides were incubated with 10% bovine serum albumin (BSA, #37525, Thermo Fisher Scientific, Waltham, MA, USA) for 1 hour at room temperature to block nonspecific binding sites. Subsequently, primary antibodies were applied and incubated overnight at 4 °C in phosphate-buffered saline (PBS) supplemented with 2% BSA. After thorough PBS washes, sections were exposed to Alexa Fluor 488- or 594-conjugated secondary antibodies for 2 hours at room temperature. Fluorescence imaging was performed using a laser scanning confocal microscope (LSM700, Carl Zeiss, Oberkochen, Germany). For quantitative evaluation, we acquired six randomly selected fields per slide using the confocal microscope (LSM 700, Carl Zeiss, Jena, Germany). Each field was captured as a single focal plane (not z-stack images). The total number of images analyzed per experimental group was 36 images from 6 slides, and all images were processed and quantified using the same parameters to ensure consistency.

Proteins were extracted and lysed in a buffer containing 2% (#L3771, Sigma-Aldrich, St. Louis, MO, USA) and 1% protease inhibitor cocktail (#11697498001, Roche Molecular Biochemicals, Nutley, NJ, USA). To assess protein integrity, samples were resolved on 10% SDS-polyacrylamide gels and stained with Coomassie Brilliant Blue. Following electrophoretic separation, proteins were transferred onto polyvinylidene difluoride (PVDF) membranes (#10600023, Amersham Biosciences, Buckinghamshire, UK). Membranes were then blocked overnight at 4 °C using 5% non-fat dry milk in Tris-buffered saline containing 0.05% Tween-20 (TBST, #7949, Sigma-Aldrich, St. Louis, MO, USA). Subsequently, membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies using goat anti-mouse IgG-HRP (#sc-2005), goat anti-rabbit IgG-HRP (#2004) and mouse anti-goat IgG-HRP (#sc-2354) from Santa Cruz Biotechnology (Santa Cruz, CA, USA) for signal detection.

To assess the morphological extent of PNP, four indices were utilized: strip, ovoid, myelin, and neurofilament (NF) indices. The strip index was defined as the count of black or white transverse bands observed within a 500 µm segment of sciatic nerve explants under a stereoscopic microscope. The ovoid index referred to the number of myelin ovoids identified within a 200 µm region of a single teased nerve fiber using differential interference contrast (DIC) microscopy (ZEN 2.3, Carl Zeiss Microscopy GmbH, Jena, Germany). The myelin index represented the proportion of teased nerve fibers among a total of 100 that retained continuous myelin staining exceeding 50 µm in length. The NF index denoted the number of nerve fibers exhibiting uninterrupted neurofilament-positive axonal segments of at least 100 µm in length, out of 100 fibers analyzed per sample [8].

All data are presented as mean $\pm$ standard error of the mean (SEM). Statistical analyses were performed using SPSS version 21.0 (IBM Corp., Armonk, NY, USA). Comparisons between two groups were evaluated using a two-tailed Student’s t-test, while differences among multiple groups were assessed by one-way analysis of variance (ANOVA) followed by Bonferroni post hoc testing for multiple comparisons. Each experiment was independently repeated at least four times. A p-value less than 0.05 was considered statistically significant.

During PNP, the transverse stripes (bands of Fontana) in sciatic nerves disappeared on day 3 after nerve injury [8]. With this gross property, we performed a pharmacoactivity using an in vivo system and various concentrations (1 mM~100 mM) [8]. Among the concentrations, Flavo associated with the CDK signaling pathway significantly inhibited the disappearance of the transverse stripes in sciatic nerves at 3 and 4 days after axotomy (Fig. 1). A toxic effect of 100 mM of Flavo caused the death of Schwann cells; therefore, the stripe index decreased (Fig. 1, left panel). Thus, we used 10 mM of Flavo for further in vivo evaluation.

Fig. 1.

Flavo inhibits the disappearance of transverse stripes during in vivo PNP. (a) Dose-dependent treatment for showing transverse stripe patterns in sciatic nerves (left panel). Scale bar = 600 µm. Quantitative data (right panel) shows the disappearance patterns of transverse stripe under a stereoscope (n = 3 mice). (b) Confocal images show immunolabeling for myelin basic protein as a myelin marker (green; upper panel) and neurofilament as an axon marker (NF, red; lower panel). Scale bar = 100 µm. (c) Ovoid (left panel), myelin (middle panel) and NF (right panel) indices show quantitative degree of in vivo PNP (n = 4 mice). ^{*, # or $}p 0.05 and ^{**, ## or $$}p 0.001. Comparisons between each experimental group (1 mM Flavo, 10 mM Flavo, and 100 mM Flavo) and the control group (Flavo untreated) are indicated as #, $, and *, respectively. ns, not significant. Flavo, flavopiridol; PNP, peripheral neurodegenerative process.

PNP is characterized by two primary morphological features: demyelination and axonal degradation. To evaluate whether Flavo can suppress these pathological changes, we conducted immunofluorescent staining using antibodies against myelin basic protein (MBP), a marker of the myelin sheath, and neurofilament (NF), a structural axonal marker, in an established in vivo PNP model. In uninjured control sciatic nerves, MBP staining appeared as continuous linear signals along the nerve fibers. By contrast, sciatic nerves examined 3 days post-injury exhibited fragmented MBP signals with irregular boundaries, indicative of Schwann cell demyelination during PNP (Fig. 1b, upper panel). Remarkably, treatment with Flavo preserved the structural integrity of MBP staining at day 3, resembling the intact and linear appearance observed in controls. Following peripheral nerve injury, axonal degradation can result either from intrinsic self-destructive mechanisms or secondary to Schwann cell disintegration [1]. In control samples, NF staining was detected as uninterrupted red fluorescence within the longitudinal axis of nerve fibers (Fig. 1b, lower panel). However, nerves examined 3 days after injury showed a notable increase in interrupted or fragmented NF signals, suggesting significant axonal breakdown. Importantly, Flavo prevented such disruption, maintaining the continuity of NF staining comparable to uninjured nerves.

To quantify these morphological effects, we employed validated morphometric parameters, namely the ovoid (left panel) and myelin (middle panel) indices (reflecting demyelination) and the NF index (reflecting axonal integrity, right panel) [8]. Analysis revealed that Flavo treatment led to a time-dependent reduction in both demyelination and axonal degeneration, as demonstrated by improvements in all three indices (Fig. 1c). Collectively, these findings support the conclusion that Flavo exerts robust protective effects on myelin and axonal structure during in vivo PNP.

Various signaling pathways affect Schwann cell dynamics during PNP. To determine which signaling pathways are involved downstream of H₂S in Schwann cells and whether Flavo regulates these signaling pathways during PNP, we checked the PI3K-Akt [10, 11], MAPK [12], and Wnt/$\beta$-catenin [5] signaling pathways using western blot analysis. The control and untreated explants exhibited no change of pAkt expression at 3DIV, and Flavo did not inhibit pAkt expression during PNP (Fig. 2a,b). Untreated explants upregulated pERK1/2 at 3 days in vitro (3DIV) compared with control explants, which did not express pERK1/2. Flavo did not inhibit the expression of pERK1/2 at 3DIV (Fig. 2a,b). Cyclin D1 is a cell proliferation marker and an effector that represents the pharmacoactivity of Flavo [7, 13, 14]. Cyclin D1 as a positive control was upregulated at 3DIV in untreated explants and it was suppressed by Flavo at 3DIV (Fig. 2a,b). Thus, we determined that PI3K-Akt and the MAPK/ERK signaling pathways are not regulated by Flavo in Schwann cells during PNP.

Fig. 2.

Flavo inhibits the nuclear translocation of $\beta$-catenin in Schwann cells during ex vivo PNP. (a) Protein lysates (30 µg) from sciatic explants cultured for 3DIV were analyzed by Western blotting (n = 4 mice). DIV, days in vitro; pERK1/2, phosphorylated extracellularregulated kinase 1/2; pAkt, phosphorylated Akt; CCND1, cyclin D1. (b) Quantitative data shows the relative intensities of pAkt, pERK1/2, cyclin D1 and $\beta$-catenin in the absence or presence of Flavo (n = 3 mice). ^***p 0.0001. (c) Confocal images show immunolabeling for $\beta$-catenin ($\beta$-caten, red). DAPI (blue) and Sox10 (green) are used for nucleus markers. Arrows indicate nuclei double-positive for β-catenin and Sox10. Scale bar = 100 µm. (d) The number of cells was counted with $\beta$-catenin immunolabeling out of 200 DAPI-(+) nuclei in teased sciatic nerve fibers as compared with that of the control (n = 4 mice). ^*p 0.05. (e) The number of cells was counted with $\beta$-catenin immunolabeling out of 200 Sox10-(+) nuclei in teased sciatic nerve fibers as compared with that of the control (n = 4 mice). ^*p 0.05. ns, not significant. Flavo, flavopiridol; PNP, peripheral neurodegenerative process.

The Wnt/$\beta$-catenin signaling pathway affects the expression of myelin genes and the maintenance of the myelin structure in Schwann cells [5]. $\beta$-catenin was highly expressed in control sciatic nerves, whereas $\beta$-catenin expression decreased in untreated 3DIV explants (Fig. 2a,b). The Flavol treatment maintained $\beta$-catenin expression similar to that of the control at 3DIV (Fig. 2a,b). Nerve fibers at 3DIV also showed positive $\beta$-catenin signals, which overlapped with DAPI staining as a nuclear marker, while Flavo reduced the overlap between the $\beta$-catenin and DAPI signals at 3DIV (Fig. 2c). Quantitative data also indicated that Favo inhibited overlapping $\beta$-catenin with positive DAPI or SRY-related HMG-box 10 (Sox10) signals as a marker for nuclei in ex vivo culture (Fig. 2d,e). Sox10 is expressed in both myelinating and non-myelinating Schwann cells. Thus, our data suggest that Flavo effectively inhibits the nuclear translocation of $\beta$-catenin and affects the Wnt/$\beta$-catenin signaling pathway in Schwann cells during PNP.

Additionally, because Flavo is an inhibitor of CDKs, which act downstream of H₂S and induce cell cycle arrest and reduces cyclin D1 [7, 15, 16], Flavo should not inhibit upstream of CDKs, particularly CSE and 3-mercaptopyruvate sulfurtransferase (MST) as a specific enzyme that produces H₂S in Schwann cells [8]. To determine whether Flavo specifically modulates downstream targets of the H₂S pathway during PNP, we performed western blotting using anti-CSE and anti-MST antibodies in an ex vivo model. N-ethylmaleimide (NEM), a specific H₂S inhibitor [8], significantly reduced the expression levels of CSE and MST at 3DIV compared to untreated nerve fibers. In contrast, Flavo did not prevent the reduction in CSE and MST expression in degenerating Schwann cells (Fig. 3a,b), suggesting that CDK signaling acts downstream of the H₂S pathway. Taken together, these findings demonstrate that Flavo effectively suppresses PNP through modulation of the H₂S/$\beta$-catenin/CDK signaling pathways.

Fig. 3.

Flavo fails to suppress the H₂S signaling pathway during ex vivo PNP. (a) Protein lysates (30 µg) from sciatic explants cultured for 3DIV were analyzed by Western blotting (n = 4 mice). DIV, days in vitro; pERK1/2, phosphorylated extracellularregulated kinase 1/2; CCND1, cyclin D1; cystathionine $\gamma$-lyase (CSE); 3-mercaptopyruvate sulfurtransferase (MST). (b) Quantitative data shows the relative intensities of CSE, MST, pERK1/2 and CCND1 in the absence or presence of Flavo (n = 3 mice). ^**p 0.001 and ^***p 0.0001. ns, non-significant; PNP, peripheral neurodegenerative process.

Transcriptional regulation in Schwann cells is an essential control tower to exhibit the phenotypes under both normal and degenerating conditions [3]. To assess whether Flavo affects transcriptional regulation during PNP, we checked two transcription factors, such as c-Jun and Krox20, as negative and positive regulators of myelin maintenance, respectively, and performed immunostaining with specific anti- c-Jun and Krox20 antibodies. Krox20 was expressed on the nucleus in Schwann cells from control nerve fibers but did not appear in Schwann cells at 3DIV. Flavo effectively protected the disappearance of Krox20 signaling in Schwann cells at 3DIV (Fig. 4a).

Fig. 4.

Flavo controls transcriptional regulation in Schwann cells during ex vivo PNP. (a) Confocal images show immunolabeling for Krox20 (red, upper panel). DAPI are used for a nucleus marker (blue). Arrows indicate nuclei double-positive for Krox20 and DAPI. Scale bar = 100 µm. The number of cells was counted with krox20 immunolabeling out of 200 DAPI-(+) nuclei in teased sciatic nerve fibers as compared with that of the control (lower panel, n = 4 mice). ^*p 0.05. (b) Confocal images show immunolabeling for c-Jun (green, upper panel). DAPI are used for a nucleus marker (blue). Arrows indicate nuclei double-positive for c-Jun and DAPI. Scale bar = 100 µm. The number of cells was counted with c-jun immunolabeling out of 200 DAPI-(+) nuclei in sciatic nerve sections as compared with that of the control (lower panel, n = 4 mice). ns, not significant; DIV, days in vitro; Flavo, flavopiridol; PNP, peripheral neurodegenerative process.

In contrast, c-Jun was not expressed under normal conditions but was highly expressed on the nucleus in untreated Schwann cells at 3DIV (Fig. 4b). Flavo-treated nerve fibers showed no change in c-Jun nuclear expression in Schwann cells at 3DIV compared with that of untreated fibers at 3DIV (Fig. 4b). Taken together, these data indicate that Flavo effectively controls Krox20 expression as a positive transcriptional regulator to maintain myelin structure in Schwann cells in a c-Jun-independent manner.

H₂S functions as an antioxidant to reduce oxidative stress in living cells [17]. Thus, most of its downstream molecules are involved in overcoming oxidative stress and a return to normal conditions [17]. In previous studies, H₂S regulated cyclin D1 expression, and cyclin D1 protein levels were regulated by $\beta$-catenin [5, 15, 16, 18]. In other words, cyclin D1 expression is induced by the Wnt/$\beta$-catenin signaling pathway under increased H₂S, which provides a protective effect against oxidative stress in Schwann cells and activates the boost repair capacity of Schwann cells for effective PNP. In this study, we demonstrated that Flavo did not regulate H₂S production, while Flavo effectively inhibited cyclin D1 expression and nuclear translocation of $\beta$-catenin in degenerating Schwann cells (Figs. 2,3). Inhibiting Schwann cell proliferation may prevent the band of Büngner, which is a tunnel that guides the regenerating axon, leading to a delayed or terminated peripheral neurodegenerative process.

Peripheral nerves exhibit morphological and biochemical phenotypes, such as demyelination, axonal degradation, transcriptional regulation, Schwann cell trans-dedifferentiation, and Schwann cell proliferation during PNP. In our study, Flavo significantly inhibited morphologically these phenotypes and arrested PNP (Fig. 1). Schwann cells revert to a dedifferentiated form during PNP, which is also observed in immature Schwann cells during their development [3]. During PNP, axon degradation is regulated both physically and biochemically by Schwann cells [1]. Once PNP begins to progress, dedifferentiating Schwann cells physically dismantle and dissolve axons, thereby inducing axon degradation, while the myelin within Schwann cells undergoes fragmentation mediated by both Schwann cells by itself and macrophages infiltrating from outside of the nerves. Therefore, the effect of Flavo is thought to suppress axon degradation indirectly, rather than directly, by inhibiting CDKs in Schwann cells. In particular, Flavo possesses a potent ability to effectively suppress in vivo PNP, suggesting its potential for clinical application (Fig. 1). This suggests that Flavo can effectively penetrate the blood-nerve barrier (BNB), which protects peripheral nerves [19], thereby enabling control of in vivo PNP.

Oxidative stress induces c-Jun N-terminal kinase (JNK) activation in several cell types [20] and its inhibitory effect protects c-jun phosphorylation, leading to suppressed c-jun transcriptional activity. In our study, we determined that Flavo did not affect p-c-jun expression in Schwann cells during PNP while Krox20 (Erg2), as a positive regulator of myelin maintenance, was expressed in Flavo-treated Schwann cells at 3DIV (Fig. 4). A previous study reported that P-TEFb p-Tefb affects Krox20 gene expression [21]. Thus, Flavo could canonically affect the transcriptional function of P-TEFb during PNP and subsequently inhibit the decrease in krox20 expression by inactivated P-TEFb, leading to suppressed PNP. In other words, no change in krox20 expression in Flavo-treated Schwann cells could maintain the structure of the myelin and evoke saltatory conduction.