Ablation of Lrp4 in Schwann Cells Promotes Peripheral Nerve Regeneration in Mice

Simple Summary Schwann cells (SCs) are the primary glial cells in the peripheral nervous system and play an indispensable role in peripheral nerve regeneration. A recent study indicated the potential involvement of low-density lipoprotein receptor-related protein 4 (Lrp4) in axonal regeneration in larval zebrafish. However, whether Lrp4 participates in nerve regeneration in mammals remains unknown. Herein, we constructed conditional knockout (cKO) mice in which Lrp4 was specifically deleted in SCs under the control of the Dhh-cre promoter. We found that the peripheral nerve regeneration in cKO mice was more robust than that in control mice. Assessment of SC proliferation via BrdU staining revealed significantly increased cell numbers in the cKO mice. Furthermore, as a master transcription factor participating in the onset of myelination, Krox-20 was dramatically downregulated in the injured nerves of the cKO group compared with the control group nerves. The enhanced demyelination speed in cKO mice was confirmed by electron microscopy. Altogether, these results suggest that Krox-20 downregulation in mutant mice leads to SC proliferation upon injury and that SCs can remove myelin debris, thereby promoting nerve regeneration. Abstract Low-density lipoprotein receptor-related protein 4 (Lrp4) is a critical protein involved in the Agrin-Lrp4-MuSK signaling pathway that drives the clustering of acetylcholine receptors (AChRs) at the neuromuscular junction (NMJ). Many studies have shown that Lrp4 also functions in kidney development, bone formation, nervous system development, etc. However, whether Lrp4 participates in nerve regeneration in mammals remains unknown. Herein, we show that Lrp4 is expressed in SCs and that conditional knockout (cKO) of Lrp4 in SCs promotes peripheral nerve regeneration. In Lrp4 cKO mice, the demyelination of SCs was accelerated, and the proliferation of SCs was increased in the injured nerve. Furthermore, we identified that two myelination-related genes, Krox-20 and Mpz, were downregulated more dramatically in the cKO group than in the control group. Our results elucidate a novel role of Lrp4 in peripheral nerve regeneration and thereby provide a potential therapeutic target for peripheral nerve recovery.


Introduction
Peripheral nerve injuries are common (incidence of 1 in 1000 individuals per year) and result from systemic diseases or disabling diseases causing localized damager [1]. However, unlike injuries in the central nervous system (CNS) that result in loss of axon regeneration ability, the extra environment in the peripheral nervous system (PNS) permits

Immunofluorescence
Muscles were fixed in 4% paraformaldehyde (PFA) at room temperature for 30 min and incubated with 0.1 M glycine for 30 min. Next, the tissue sections were incubated with blocking buffer (2% BSA, 7% goat serum, and 0.5% Triton X-100 in PBS) for 2-4 h at room temperature or overnight at 4 • C. After washing three times each for 15 min, the samples were incubated with the primary antibodies in blocking buffer overnight at 4 • C. The tissues were washed three times and incubated with fluorescently labeled secondary antibodies for 2-4 h at 25 • C. After washing three times for 30 min in PBS, the tissues were flat-mounted using Hydromount (National Diagnostics, Kolkata, India). For sciatic nerve staining, the sciatic nerves were embedded in OCT. The frozen samples were cut into 20-µm-thick sections, fixed in 4% PFA, and washed in PBS at room temperature for 30 min. Next, the sections were incubated in blocking buffer (2% BSA, 7% goat serum, and 0.5% Triton X-100 in PBS) for 2-4 h at room temperature, and then incubated overnight with primary antibodies at 4 • C. After washing three times with PBS, the samples were incubated at room temperature for 2-4 h with the Alexa-488 goat anti-rabbit or Alexa-594 goat antimouse secondary antibody. The samples were then washed three times and mounted in Hydromount (National Diagnostics, Kolkata, India). For all NMJ staining, α-BTX was added to the secondary antibody solution. For Dhh-tdTomato and Dhh-tdTomato-Lrp4 −/− mouse NMJ staining, CF568-conjugated α-BTX was replaced with CF488-conjugated α-BTX. Images were collected using florescence microscope (FSX100, Olympus, Tokyo, Japan), and image analysis and NMJ counting were conducted using ImageJ.

Western Blot Analysis
Western blotting was performed as previously described [11]. Briefly, the sciatic nerves were homogenized and lysed in RIPA buffer. After centrifugation the supernatant was collected, and the lysate protein concentrations were measured. Protein samples (50 µg of protein) were separated by SDS-PAGE and transferred onto nitrocellulose membranes. The membrane was incubated in 5% milk and then incubated with the primary antibodies, anti-LRP4 (1:1000, clone N207/27, UC Davis/NIH NeuroMab Facility) and anti-GAPDH (1:2000, HRP-60004, Abcam, Cambridge, United Kingdom). After washing, the membranes were incubated with the anti-rabbit or anti-mouse horseradish peroxidase (HRP)-conjugated secondary antibody (1:2000, Invitrogen, Carlsbad, CA, USA), and the signal was detected by Western HRP chemiluminescence reagents (Thermo, Waltham, MA, USA). The grayscale values of the bands were quantified using ImageJ.

BrdU Staining
Bromodeoxyuridine (BrdU) staining was performed as previously described with modifications [29]. The mice were injected with BrdU (B5002 Sigma-Aldrich, St. Louis, Biology 2021, 10, 452 4 of 17 MO, USA) at 5 days post crush (dpc, 100 mg/kg, three times at 6 h intervals). Two hours after the final injection, the mouse hearts were perfused with 4% PFA. Sciatic nerves were embedded in OCT compound, and longitudinal sections were cut into 20-µm-thick sections. The samples were rinsed with 0.5% Triton X-100 for 30 min and incubated with 2 N HCl for 1 h at 37 • C. The samples were then rinsed with borate buffer (0.1 M, pH 8.5) for 8-12 min and washed with PBS three times. Next, the samples were blocked in blocking buffer (2% BSA, 7% goat serum, and 0.5% Triton X-100 in PBS) for 2-4 h at room temperature. The samples were then immunostained with an anti-BrdU antibody (1:1000 in blocking buffer) overnight at 4 • C. After washing 3 times for 15 min each with 0.5% Triton X-100 in PBS, the tissues were incubated with an Alexa Fluor 488-conjugated antibody and DAPI overnight at 4 • C. After washing 3 times for 30 min each with 0.5% Triton X-100 in PBS, the tissues were flat-mounted in Hydromount (National Diagnostics, Atlanta, GA, USA).

TUNEL Staining
Terminal deoxynucleotidyl transferase dUTP nick end (TUNEL) staining was performed using a TUNEL apoptosis detection kit (Sigma 11684795910) according to the manufacturer's protocol with some modifications. At 5 dpc, the tissues were sectioned and fixed with 4% PFA for 1 h at room temperature, after which the sciatic nerves were embedded in OCT compound and longitudinal sections were cut into 20-µm-thick sections. The samples were washed three times with PBS for 15 min each time and then permeabilized with 0.5% Triton X-100 in PBS for 2 h at room temperature. The samples were then incubated with a TUNEL reaction mixture for 60 min at 37 • C and rinsed with DAPI for 30 min. After washing with PBS three times for 15 min each time, the tissues were mounted in Hydromount (National Diagnostics).

Denervation Experiment
For the nerve crush experiments, mice (male, 2 months old) were anesthetized, and their right thigh fur was shaved and sterilized with 75% alcohol. A 1 cm incision was made in the skin over the lateral femur. The sciatic nerves at the mid-thigh were exposed and clamped with micro forceps for 10 s. The incisions were closed with steel clips, and the mice were placed on a thermostatic rubber pad and kept warm until their revival. In the nerve transplant experiment, sciatic nerve segments (approximately 1 cm long) were removed from the control and cKO mice and exchanged between the mice. The nerve segments were sutured to the existing nerve at both ends in the correct orientation. After 8 weeks, the soleus was isolated and stained for NMJs.

Biotin Tracing
At 7 dpc, the mice were anesthetized, the sciatic nerve was removed from the injury site, and the distal stump was incubated with biotin. Two hours after incubation, the mice were perfused with 4% PFA. The sciatic nerve cryosections were labeled with streptavidin Alexa 568 for 1 h at room temperature. After washing with PBS 3 times, the sections were mounted with Hydromount (National Diagnostics).

Electron Microscopy
At 7 dpc, the mice were anesthetized and cardiac perfused with 4% PFA and 2% glutaraldehyde in 0.1 M sodium cacodylate buffer (NaCac, pH 7.4). The distal segments of the sciatic nerves were separated and fixed overnight at 4 • C in the same perfusion fixative for 24 h. Electron microscopy (EM) analysis of myelin was performed as described previously with modifications [29]. Ultrathin sections were photographed with an electron microscope.

Statistical Analysis
Before being analyzed, all data in our study were checked by the D'Agostino-Pearson omnibus. Data were analyzed by the unpaired t-test, one-way ANOVA, and two-way ANOVA using GraphPad Prism. Data are shown in dot plots as the mean ± SEM. The significance criterion was set at p < 0.05. *, **, ***, and **** represent p < 0.05, p < 0.01, p < 0.001, and p < 0.0001, respectively.

Lrp4 Expression in SCs
Previous studies have demonstrated the function of Lrp4 in the glial cells of the CNS [22,23,27]. To initially investigate the function of Lrp4 in SCs, we generated SC Lrp4 cKO mice (Dhh-Lrp4 −/− ) by crossing Lrp4 f/f mice with Dhh::Cre mice ( Figure 1A). The sciatic nerves of the control (Lrp4 f/f ) and Lrp4 cKO mice were isolated for qRT-PCR and Western blotting. As shown in Figure 1B-D, both the mRNA and protein levels of Lrp4 were dramatically decreased in the Lrp4 cKO mice compared with the control mice. These results suggest that Lrp4 was successfully deleted from the SCs of Lrp4 cKO mice. Next, we examined the expression of Lrp4 in SCs from Lrp4-LacZ mice [27]. Exon 2 through 31 within the Lrp4 gene were replaced with LacZ, which encodesβ-gal. We found that β-gal was costained with the SC marker S100β, suggesting that Lrp4 was expressed in SCs ( Figure 1E).

Lrp4 Expression in SCs
Previous studies have demonstrated the function of Lrp4 in the glial cells of the CNS [22,23,27]. To initially investigate the function of Lrp4 in SCs, we generated SC Lrp4 cKO mice (Dhh-Lrp4 -/-) by crossing Lrp4 f/f mice with Dhh::Cre mice ( Figure 1A). The sciatic nerves of the control (Lrp4 f/f ) and Lrp4 cKO mice were isolated for qRT-PCR and Western blotting. As shown in Figure 1B-D, both the mRNA and protein levels of Lrp4 were dramatically decreased in the Lrp4 cKO mice compared with the control mice. These results suggest that Lrp4 was successfully deleted from the SCs of Lrp4 cKO mice. Next, we examined the expression of Lrp4 in SCs from Lrp4-LacZ mice [27]. Exon 2 through 31 within the Lrp4 gene were replaced with LacZ, which encodesβ-gal. We found that β-gal was costained with the SC marker S100β, suggesting that Lrp4 was expressed in SCs ( Figure 1E).

Nerve Regeneration Promotion in the Lrp4 cKO Mice
We previously showed that the motor axons can bypass AChR clusters to "overshoot" towards the periphery of muscle fibers during development in muscle-specific Lrp4 KO mice, implying that Lrp4 deficiency may promote axon growth [9]. Herein, to evaluate whether the ablation of Lrp4 in SCs promotes axon regeneration, we performed anterograde labeling with biocytin to trace the nerve axon regrowth trajectory at 7 dpc. Nerve crush was performed, as shown in Figure 2A. Significantly, the biotin-labeled NF were longer in Lrp4 cKO mice than in the control group ( Figure 2B, the dotted line indicates the crush site). Moreover, the regenerated nerve numbers at different distances from the crush site were increased ( Figure 2C), suggesting that Lrp4 expressed on mSCs may play a negative role in PNS regeneration. NMJ reinnervation is an excellent model for investigating whether regenerated axons completely retarget muscle fibers. At different days after nerve crush, the muscles were isolated and stained with CF568-conjugated α-BTX to visualize the AChR clusters and stained with antibodies against NF and SYN to visualize the axons and nerve terminals. We compared the ratios of regenerated NMJs in control and cKO mice. Depending on the overlapping extents of the nerve terminal and

Nerve Regeneration Promotion in the Lrp4 cKO Mice
We previously showed that the motor axons can bypass AChR clusters to "overshoot" towards the periphery of muscle fibers during development in muscle-specific Lrp4 KO mice, implying that Lrp4 deficiency may promote axon growth [9]. Herein, to evaluate whether the ablation of Lrp4 in SCs promotes axon regeneration, we performed anterograde labeling with biocytin to trace the nerve axon regrowth trajectory at 7 dpc. Nerve crush was performed, as shown in Figure 2A. Significantly, the biotin-labeled NF were longer in Lrp4 cKO mice than in the control group ( Figure 2B, the dotted line indicates the crush site). Moreover, the regenerated nerve numbers at different distances from the crush site were increased ( Figure 2C), suggesting that Lrp4 expressed on mSCs may play a negative role in PNS regeneration. NMJ reinnervation is an excellent model for investigating whether regenerated axons completely retarget muscle fibers. At different days after nerve crush, the muscles were isolated and stained with CF568-conjugated α-BTX to visualize the AChR clusters and stained with antibodies against NF and SYN to visualize the axons and nerve terminals. We compared the ratios of regenerated NMJs in control and cKO mice. Depending on the overlapping extents of the nerve terminal and AChR cluster, the endplates were categorized as either completely reinnervated (the AChR clusters fully overlapped with the regenerated nerve terminals) or fully denervated (AChR clusters did not overlap with the regenerated nerve terminals). Accordingly, the completely reinnervated NMJ ratios in cKO mice were noticeably increased at 10 and 14 dpc compared with those of control mice ( Figure 2D,E). The completely denervated NMJs in the Lrp4 cKO mice were significantly decreased at 10 and 14 dpc ( Figure 2F). These results suggest that the deletion of Lrp4 in SCs promotes axon regeneration in the PNS.

The mSCs Rather Than tSCs Are Involved in Accelerating Nerve Regeneration in Lrp4 cKO Mice
Both mSCs and tSCs are involved in peripheral nerve regeneration, as mSCs promote axonal regeneration, while tSCs guide the nerve terminal to NMJs. To verify the role of mSCs, we performed a nerve transfer experiment, which is the gold standard for assessing mSC function [29] ( Figure 3A). We isolated the sciatic nerve from the control and cKO mice and exchanged the nerve segments. In this experiment, nerve allografts provided SCs from donor mice to host mice. Eight weeks after nerve transfer, the tibialis anterior (TA) muscles in the control mice appeared to display less relative atrophy than those from cKO mice ( Figure 3B), which was confirmed by quantifying the number of muscle fibers ( Figure 3C). We next quantified the number of reinnervated endplates in soleus muscles. As expected, the control mice displayed more reinnervated NMJs than the cKO mice ( Figure 3D,E). These results demonstrate that the ablation of Lrp4 in mSCs provides a favorable environment for axon regrowth. average tSC numbers, NMJ tSCs distribution, and tSC soma areas did not differ between the Dhh-tdTomato and Dhh-tdTomato-Lrp4 -/-mice ( Figure 4B-D). Furthermore, we investigated whether Lrp4 ablation influences tSCs under normal conditions ( Figure 4E). Consistently, the average tSC number and distribution of tSCs in each NMJ and soma area were comparable between the Dhh-tdTomato and Dhh-tdTomato-Lrp4 -/-mice ( Figure 4F-H). In summary, these results suggest that the deletion of Lrp4 in mSCs but not in tSCs contributes to accelerating NMJ regeneration. (E) Reduced fully reinnervated and increased denervated endplates in cKO after nerve transplantation. **** p < 0.0001, control, n = 75 slices; cKO, n = 93 slices, 6 mice/group. The fully innervated NMJ percentage was 20.3 ± 2.3% in the control mice, compared with 4.2 ± 0.7% in the cKO mice; the partially innervated NMJ percentage was 28.5 ± 2.1% in the control mice, compared with 13.7 ± 1.2% in the cKO mice; and the denervated NMJ percentage was 51.2 ± 3.6% in the control mice, compared with 82.0 ± 1.6% in the cKO mice. All the data are presented as the mean ± SEM. (E) Reduced fully reinnervated and increased denervated endplates in cKO after nerve transplantation. **** p < 0.0001, control, n = 75 slices; cKO, n = 93 slices, 6 mice/group. The fully innervated NMJ percentage was 20.3 ± 2.3% in the control mice, compared with 4.2 ± 0.7% in the cKO mice; the partially innervated NMJ percentage was 28.5 ± 2.1% in the control mice, compared with 13.7 ± 1.2% in the cKO mice; and the denervated NMJ percentage was 51.2 ± 3.6% in the control mice, compared with 82.0 ± 1.6% in the cKO mice. All the data are presented as the mean ± SEM.
To determine whether Lrp4 ablation in tSCs is involved in promoting NMJ regeneration, we characterized the morphology of tSCs by generating Dhh-tdTomato and Dhh-tdTomato-Lrp4 −/− mice whose SCs were labeled by red fluorescence. At 7 dpc, muscles were isolated and stained for NMJs. We found that each NMJ was covered by 2~4 tSCs, consistent with the tSC number under normal conditions ( Figure 4A) [31,32]. In addition, the average tSC numbers, NMJ tSCs distribution, and tSC soma areas did not differ between the Dhh-tdTomato and Dhh-tdTomato-Lrp4 −/− mice ( Figure 4B-D). Furthermore, we investigated whether Lrp4 ablation influences tSCs under normal conditions ( Figure 4E). Consistently, the average tSC number and distribution of tSCs in each NMJ and soma area were comparable between the Dhh-tdTomato and Dhh-tdTomato-Lrp4 −/− mice ( Figure 4F-H). In summary, these results suggest that the deletion of Lrp4 in mSCs but not in tSCs contributes to accelerating NMJ regeneration.

Enhanced SC Proliferation in Lrp4 cKO Mice after Nerve Damage
To explore the potential mechanisms, we detected SC proliferation and survival using the BrdU labeling and TUNEL assays, respectively. In Dhh-tdTomato or Dhh-tdTomato-Lrp4 -/-mice, the right sciatic nerve was crushed while the left sciatic nerve was shammed to serve as the control. The mice were injected with BrdU 3 times at 5 dpc. Two hours after the final injection, the injured nerve segments and sham controls, including those at the proximal, distal, and injury sites, were isolated intact and then stained with the anti-BrdU antibody. As shown in Figure 5A, the number of BrdU+/Dhh+ cells, indicating newly formed SCs, was increased in the cKO mice. The dotted line indicates the crush site. Moreover, the SCs in cKO mice mainly proliferated in the distal nerve segments ( Figure 5B,C), which is consistent with previous reports [33]. Furthermore, the numbers of SCs under normal conditions were comparable between the cKO and control mice, suggesting that the accelerated proliferation of SCs in the cKO mice was nerve injury dependent ( Figure  5D,E). To determine whether the ablation of Lrp4 affects SC survival after injury, we used the TUNEL assay to detect apoptotic cells. As shown in Figures 6A,B, the numbers of TUNEL-positive cells did not differ between the control and cKO mice. These results suggest that Lrp4 deficiency in SCs leads to enhanced SC proliferation without disturbing SC survival after nerve injury.
During injury-induced SC reprogramming, macrophages are gradually recruited by SCs and cooperate with SCs to degrade myelin debris and thereby facilitate axonal regrowth [34,35]. To determine whether the increased SCs recruited more macrophages in cKO mice, we labeled the macrophages with an anti-CD68 antibody ( Figure 6C), and the CD68+ cell numbers were comparable between control and cKO mice ( Figure 6D). Taken together, these results suggest that the advantageous proliferation of SCs plays a potential role in promoting NMJ regeneration in Lrp4 cKO mice.

Enhanced SC Proliferation in Lrp4 cKO Mice after Nerve Damage
To explore the potential mechanisms, we detected SC proliferation and survival using the BrdU labeling and TUNEL assays, respectively. In Dhh-tdTomato or Dhh-tdTomato-Lrp4 −/− mice, the right sciatic nerve was crushed while the left sciatic nerve was shammed to serve as the control. The mice were injected with BrdU 3 times at 5 dpc. Two hours after the final injection, the injured nerve segments and sham controls, including those at the proximal, distal, and injury sites, were isolated intact and then stained with the anti-BrdU antibody. As shown in Figure 5A, the number of BrdU+/Dhh+ cells, indicating newly formed SCs, was increased in the cKO mice. The dotted line indicates the crush site. Moreover, the SCs in cKO mice mainly proliferated in the distal nerve segments ( Figure 5B,C), which is consistent with previous reports [33]. Furthermore, the numbers of SCs under normal conditions were comparable between the cKO and control mice, suggesting that the accelerated proliferation of SCs in the cKO mice was nerve injury dependent ( Figure 5D,E). To determine whether the ablation of Lrp4 affects SC survival after injury, we used the TUNEL assay to detect apoptotic cells. As shown in Figure 6A,B, the numbers of TUNEL-positive cells did not differ between the control and cKO mice. These results suggest that Lrp4 deficiency in SCs leads to enhanced SC proliferation without disturbing SC survival after nerve injury.
During injury-induced SC reprogramming, macrophages are gradually recruited by SCs and cooperate with SCs to degrade myelin debris and thereby facilitate axonal regrowth [34,35]. To determine whether the increased SCs recruited more macrophages in cKO mice, we labeled the macrophages with an anti-CD68 antibody ( Figure 6C), and the CD68+ cell numbers were comparable between control and cKO mice ( Figure 6D). Taken together, these results suggest that the advantageous proliferation of SCs plays a potential role in promoting NMJ regeneration in Lrp4 cKO mice.

Lrp4 Deficiency Promotes Demyelination and Downregulates Krox-20 after Nerve Injury
To investigate the mechanisms underlying LRP4 deficiency-induced SC proliferation, we focused on the Wnt and MAPK signaling pathways, which are involved in promoting the proliferation of SCs. As shown in Figure 7A, before and after nerve injury, the mRNA levels of β-catenin and MAPK were comparable between the control and Lrp4 cKO mice. Next, we focused on the factors regulating SC demyelination processes. After nerve injury, SCs undergo demyelination rapidly, and demyelinated SCs proliferate to promote nerve regeneration. Substantial progress has been identifying the intrinsic factors in SCs that participate in this process [36,37]. c-Jun, Notch, and Id2 are positive regulators of SC demyelination [38]. However, our qRT-PCR results showed no differences in their levels in control and cKO mice under both normal and nerve damage conditions ( Figure 7B). Next, we detected a negative factor related to demyelination, Krox-20, which is essential for SC demyelination after nerve injury [36,39]. Surprisingly, Krox-20 expression was significantly decreased in the mutant SCs compared with the control SCs. Accordingly, the expression of Mpz, which is required for proper myelin formation and maintenance, was reduced in the injury-induced cKO SCs compared with the control ( Figure 7C). This result suggests that the decreased Krox-20 and Mpz levels in the mutant SCs may have enhanced SC demyelination after nerve injury. Furthermore, the nerve segments were subjected to EM analysis. Although myelin breakdown or collapse was apparent in both cKO mice and control mice seven days after injury, the cKO mice displayed more demyelination and a lower amount of myelin debris ( Figure 7D). Statistically, the size of the myelin debris area was dramatically decreased (Figure 7E), and the regenerated axon numbers were increased in the cKO mice ( Figure 7F). These results suggest rapid demyelination and myelinophagy upon Lrp4 deletion in SCs.
Biology 2021, 10, x 11 of 16 mice. The white arrows indicate the macrophages. Scale bar = 25 μm (D) Quantification of the macrophage numbers revealed no significant difference between the control and cKO mice. Control and cKO, n = 10 slices, 3 mice/group.

Lrp4 Deficiency Promotes Demyelination and Downregulates Krox-20 after Nerve Injury
To investigate the mechanisms underlying LRP4 deficiency-induced SC proliferation, we focused on the Wnt and MAPK signaling pathways, which are involved in promoting the proliferation of SCs. As shown in Figure 7A, before and after nerve injury, the mRNA levels of β-catenin and MAPK were comparable between the control and Lrp4 cKO mice. Next, we focused on the factors regulating SC demyelination processes. After nerve injury, SCs undergo demyelination rapidly, and demyelinated SCs proliferate to promote nerve regeneration. Substantial progress has been identifying the intrinsic factors in SCs that participate in this process [36,37]. c-Jun, Notch, and Id2 are positive regulators of SC demyelination [38]. However, our qRT-PCR results showed no differences in their levels in control and cKO mice under both normal and nerve damage conditions ( Figure 7B). Next, we detected a negative factor related to demyelination, Krox-20, which is essential for SC demyelination after nerve injury [36,39]. Surprisingly, Krox-20 expression was significantly decreased in the mutant SCs compared with the control SCs. Accordingly, the expression of Mpz, which is required for proper myelin formation and maintenance, was reduced in the injury-induced cKO SCs compared with the control ( Figure 7C). This result suggests that the decreased Krox-20 and Mpz levels in the mutant SCs may have enhanced SC demyelination after nerve injury. Furthermore, the nerve segments were subjected to EM analysis. Although myelin breakdown or collapse was apparent in both cKO mice and control mice seven days after injury, the cKO mice displayed more demyelination and a lower amount of myelin debris ( Figure 7D). Statistically, the size of the myelin debris area was dramatically decreased (Figure 7E), and the regenerated axon numbers were increased in the cKO mice ( Figure 7F). These results suggest rapid demyelination and myelinophagy upon Lrp4 deletion in SCs.  The mRNA levels of c-Jun, Notch, and Id2 were comparable between the control and cKO mice before injury and at 7 dpc. **** p < 0.0001 (3 mice/group). (C) The mRNA levels of Mpz and Krox-20 were decreased in the cKO mouse sciatic nerves compared with the control mice at 7 dpc. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 (3 mice/group). (D) Representative electron micrograph of control and cKO sciatic nerves. The cKO mice showed lower amounts of myelin debris and more regenerated axons. BB indicates the Büngner band, Sc indicates Schwann cells. Scale bar = 5 µm. (E) The size of the myelin debris area was decreased in the cKO mice (13.4 ± 1.6) compared with the control mice (25.5 ± 1.1). **** p < 0.0001, control, n = 10 slices; cKO, n = 9 slices (3 mice/group). (F) The number of newly formed axons was increased in cKO mice (61.5 ± 7.3) compared with control mice. (30.4 ± 3.9). ** p = 0.0021, control, n = 15 slices; cKO, n = 22 slices (3 mice/group). All the data are presented as the mean ± SEM.

Normal NMJ Formation and Transmission in SC Lrp4 cKO Mice
Having demonstrated that the deletion of Lrp4 in mSCs promotes NMJ regeneration, we next examined whether SC-derived Lrp4 is required for NMJ development. We examined NMJs in the diaphragm muscles at E14.5 (initial NMJ formation) and P0 (during NMJ development) as previously reported [30]. The diaphragm and TA muscles were stained in whole-mount preparations with antibodies against NF and SYN to label axons and nerve terminal synaptic vesicles and with CF568-conjugated α-BTX to visualize the postsynaptic AChR clusters. The NMJ morphology in the cKO mice was similar to that of the control littermates at E14.5 and P0 ( Figure 8A). The endplate bandwidths, primary branch localizations, secondary branch numbers/lengths, AChR cluster numbers, and areas of AChR cluster number distribution did not differ between the two groups at either age ( Figure 8B-G). In addition, we characterized single NMJs of P0 and adult mice. As expected, the NMJ morphologies did not differ between the control and cKO mice ( Figure 8H), and nerve coverage was comparable between the control and cKO mice ( Figure 8I). These results indicate that Lrp4 ablation does not influence SC development. regenerated axons. BB indicates the Büngner band, Sc indicates Schwann cells. Scale bar = 5 μm. (E) The size of the myelin debris area was decreased in the cKO mice (13.4 ± 1.6) compared with the control mice (25.5 ± 1.1). **** p < 0.0001, control, n = 10 slices; cKO, n = 9 slices (3 mice/group). (F) The number of newly formed axons was increased in cKO mice (61.5 ± 7.3) compared with control mice. (30.4 ± 3.9). ** p = 0.0021, control, n = 15 slices; cKO, n = 22 slices (3 mice/group). All the data are presented as the mean ± SEM.

Normal NMJ Formation and Transmission in SC Lrp4 cKO Mice
Having demonstrated that the deletion of Lrp4 in mSCs promotes NMJ regeneration, we next examined whether SC-derived Lrp4 is required for NMJ development. We examined NMJs in the diaphragm muscles at E14.5 (initial NMJ formation) and P0 (during NMJ development) as previously reported [30]. The diaphragm and TA muscles were stained in whole-mount preparations with antibodies against NF and SYN to label axons and nerve terminal synaptic vesicles and with CF568-conjugated α-BTX to visualize the postsynaptic AChR clusters. The NMJ morphology in the cKO mice was similar to that of the control littermates at E14.5 and P0 ( Figure 8A). The endplate bandwidths, primary branch localizations, secondary branch numbers/lengths, AChR cluster numbers, and areas of AChR cluster number distribution did not differ between the two groups at either age ( Figure 8B-G). In addition, we characterized single NMJs of P0 and adult mice. As expected, the NMJ morphologies did not differ between the control and cKO mice ( Figure  8H), and nerve coverage was comparable between the control and cKO mice ( Figure 8I). These results indicate that Lrp4 ablation does not influence SC development.
To investigate whether SC-derived Lrp4 is involved in the NMJ transmission, mEPPs, which replicate muscle membrane depolarizations due to spontaneous acetylcholine release, were measured by electrophysiology recordings as described previously [40][41][42]. The resting membrane potentials in the cKO mice were similar to those in the control mice ( Figure 8J). As shown in Figure 8K,L, both the frequency and amplitude of mEPPs in the cKO mice were comparable to those in the control mice. Moreover, the EPPs that indicate local electrical responses in response to nerve stimulation were comparable between the control and cKO mice ( Figure 8M,N). Taken together, these findings suggest normal NMJ development and maintenance in the SC Lrp4 cKO mice.  To investigate whether SC-derived Lrp4 is involved in the NMJ transmission, mEPPs, which replicate muscle membrane depolarizations due to spontaneous acetylcholine release, were measured by electrophysiology recordings as described previously [40][41][42]. The resting membrane potentials in the cKO mice were similar to those in the control mice ( Figure 8J). As shown in Figure 8K,L, both the frequency and amplitude of mEPPs in the cKO mice were comparable to those in the control mice. Moreover, the EPPs that indicate local electrical responses in response to nerve stimulation were comparable between the control and cKO mice ( Figure 8M,N). Taken together, these findings suggest normal NMJ development and maintenance in the SC Lrp4 cKO mice.

Discussion
In this study, we characterized the novel function of Lrp4 in mammalian SC by taking advantage of the specific conditional knockout of Lrp4 in the SCs of Dhh::cre;Lrp4 f/f mice and a reporter gene expressed in Lrp4 LacZ mice. We have shown for the first time that Lrp4 is expressed in SCs. As skeletal muscle fibers are not included in the sciatic nerve, our previous results indicate that motor neurons may contribute equal amounts of Lrp4 protein to SCs [19]. Here, the level of Lrp4 was decreased by approximately 50%, indicating that Lrp4 was successfully deleted from the SCs of Lrp4 cKO mice. Intriguingly, Lrp4 deficiency in SCs was advantageous in response to sciatic nerve injury. SC Lrp4 cKO mice exhibited accelerated NMJ regeneration compared with that of the control group. Moreover, the downregulation of Krox-20 may increase the demyelination and proliferation of mSCs upon Lrp4 deficiency in SCs, thereby promoting the regeneration of the PNS.
Initially, Lrp4 was largely known for its intricate role in NMJ formation by participating in the Agrin-Lrp4-MuSK signaling pathway. Ablation of the Lrp4 gene or addition of an anti-Lrp4-antibody will induce serious myasthenia gravis disorder across different organisms and in different model systems [43]. Loss of muscle-specific LRP4 in adult mice has demonstrated that Lrp4 is also important for NMJ maintenance [11]. To date, Lrp4 is known to be involved in a myriad of processes, such as limb anomalies [12][13][14], craniofacial organogenesis [15], kidney malformations [16], adult hippocampal neurogenesis [20], ALS [21], and Aβ clearance in Alzheimer's disease [22]. Recent evidence has indicated that Lrp4 serves as a necessary retrograde signal for presynaptic differentiation in the NMJ and CNS [9,10]; furthermore, it could modulate glutamatergic transmission and Aβ uptake in astrocytes [22]. In this study, we found a novel role of Lrp4 in the regeneration of PNS in mice, thereby providing a potential therapeutic target for peripheral nerve recovery.
Why Lrp4 promotes axon regeneration in low-vertebrate animals while playing an inhibitory role in mammals remains undetermined? First, this intriguing question may be explained by the evolutionary history of Lrp4. Although the sequence of Lrp4 is well conserved (80% identity), the timing of Lrp4 expression differs between zebrafish and mice during the period of AChR prepatterning. Zebrafish Lrp4 is not expressed in the muscle fiber prior to pioneer axon innervation; thus, AChR prepatterning is more dependent on the Wnt-MuSK signaling pathway [26]. Additionally, unlike the motor axons that fail to stop at the NMJ territory and instead crawl aimlessly across muscle fibers in mice, motor axons display only slight axon branching in the absence of Lrp4 in zebrafish [24]. The Lrp4 in mammals appears to have a stronger inhibitory effect on axon movement than that in zebrafish, which is a result of retrograde signal enhancement during evolution to regulate presynaptic differentiation [24]. Another homologous gene that induces differential axon regeneration was previously reported [44]. Second, the differential axon regeneration abilities of zebrafish and mice may also explain this phenomenon. Zebrafish have a robust regeneration ability in both the CNS [45,46] and PNS [47]. The axons in this laser ablation axon model regenerate faster than those in models of mechanistic injury [48]. The results from zebrafish showed that the pioneer axon across the injured site began to traverse approximately 12 h after injury, and the axons emerging later mostly traversed the injury site at approximately 20 h. Axon regeneration in the PNS generally begins approximately 3-7 days after injury, and broad regeneration occurs in the following several weeks [2]. According to the authors, zebrafish Lrp4 may play function via an extrinsic axon mechanism, through which the released Lrp4 ectodomain binds to the axons to act as a retrograde signal [25]. As axons in zebrafish have a sturdy intrinsic regenerating capacity, this resistance from the Lrp4 adhesion will be turned into the motivation of axons to move forward. However, in mammals, the resistance from the Lrp4 adhesion may be huge, thereby attenuating the speed of axon progression. The differences between zebrafish and mice need to be further researched in the future.
Regarding the mechanism by which Lrp4 promotes axon regeneration, we hypothesize that wild-type Lrp4 may have dual effects on both axon regeneration and SC proliferation, as in the development and maintenance of NMJs in mammals. The expression of wildtype Lrp4 on the SC membrane enhances its adherence to the extracellular matrix, which restricts its mobility and dedifferentiation and thereby reduces its debris clearance and proliferation abilities. In the regrowth cones, as the axon terminal highly expresses adhesion molecules (such as full-length transmembrane Agrin) [49,50], excessive adhesion to the Lrp4 ectodomain prevents regrowth. In Lrp4 ennui mutant zebrafish, in which Lrp4 encodes only the ectodomain, axon regeneration was unaffected [25]. Specifically, the overexpression of Lrp4 in motor neurons does not affect regeneration in zebrafish [25]. However, the downstream molecular mechanism of axon intrinsic factors has not been elucidated, and further research is needed.
A hallmark of SCs is their remarkable plasticity after peripheral nerve injury [51]. In this process, mSCs lose their characteristic myelin-related gene expression pattern. In turn, they start to activate a set of repair-related genes, and the balance between the opposing signaling systems determines the myelination of SCs. After nerve injury, the balance shifts towards the upregulation of myelination negative regulators such as c-Jun and Notch, which trigger SC dedifferentiation and downregulate Krox-20, Mpz, and other proteins positively regulating myelin [36,[52][53][54][55][56]. In our study, the expression of the positive myelin regulator Krox-20 was decreased dramatically in Lrp4 cKO mice after injury. Concomitantly, the myelin protein Mpz was also decreased, which is consistent with the conclusion that Krox-20 positively regulates Mpz. The decreased expression of myelin-related proteins may underlie the rapid SC demyelination in the mutant mice, thus making it more convenient for SCs to proliferate and clear myelin debris, thereby promoting nerve regeneration. Interestingly, the expression levels of the negative myelin regulators did not differ between the Lrp4 cKO and control mice, although they were all upregulated under injury conditions.

Conclusions
In summary, we first revealed that Lrp4 is expressed in the SCs of mice. Surprisingly, the ablation of Lrp4 in SCs was shown to promote peripheral nerve regeneration. The expression of Krox-20, a transcription factor essential for SC myelination during regeneration, was decreased in SC Lrp4 cKO mice, thus driving the demyelination and proliferation of mSCs. Our results reveal a novel role of Lrp4 in peripheral nerve regeneration, and may provide a new therapeutic target for peripheral nerve regeneration.

Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: All data generated or analyzed during this study are included in this published article.