Experimental Neurology 358 (2022) 114227 Contents lists available at ScienceDirect Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr Research paper Preferential regeneration and collateral dynamics of motor and sensory neurons after nerve injury in mice Sara Bolívar, Esther Udina * Institute of Neurosciences, Department Cell Biology, Physiology and Immunology, Universitat Autònoma de Barcelona, and Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), 08193 Bellaterra, Spain A R T I C L E I N F O A B S T R A C T Keywords: Specificity in regeneration after peripheral nerve injuries is a major determinant of functional recovery. Un­ Preferential motor reinnervation fortunately, regenerating motor and sensory axons rarely find their original pathways to reinnervate appropriate Specific regeneration target organs. Although a preference of motor axons to regenerate towards the muscle has been described, little is Axon collaterals known about the specificity of the heterogeneous sensory populations. Here, we propose the comparative study Motoneuron of regeneration in different neuron subtypes. Using female and male reporter mice, we assessed the regenerative Sensory neuron Nerve injury preference of motoneurons (ChAT-Cre/Ai9), proprioceptors (PV-Cre/Ai9), and cutaneous mechanoreceptors Regeneration (Npy2r-Cre/Ai9). The femoral nerve of these animals was transected above the bifurcation and repaired with fibrin glue. Regeneration was assessed by applying retrograde tracers in the distal branches of the nerve 1 or 8 weeks after the lesion and counting the retrotraced somas and the axons in the branches. We found that cuta­ neous mechanoreceptors regenerated faster than other populations, followed by motoneurons and, lastly, pro­ prioceptors. All neuron types had an early preference to regenerate into the cutaneous branch whereas, at long term, all neurons regenerated more through their original branch. Finally, we found that myelinated neurons extend more regenerative sprouts in the cutaneous than in the muscle branch of the femoral nerve and, particularly, that motoneurons have more collaterals than proprioceptors. Our findings reveal novel differences in regeneration dynamics and specificity, which indicate distinct regenerative mechanisms between neuron subtypes that can be potentially modulated to improve functional recovery after nerve injury. Significance statement 1. Introduction The lack of regeneration specificity after nerve transections is one of Peripheral neurons can regenerate after nerve injury, but functional the largest challenges in nerve regeneration, which directly affects recovery might fail if reinnervation is not specific. After a transection, functional target regeneration. Most studies have focused on the axons elongate following environmental cues from Schwann cells and regenerative environment, but only a few have examined the potential target organs (Bolívar et al., 2020). Classical studies in mammals have differences in the intrinsic growth response of different peripheral described a “preferential motor reinnervation” (PMR), a preference of neurons. Using reporter mice, we demonstrate that three paradigmatic the motoneurons to regenerate towards the muscle branch rather than to populations of peripheral neurons (motoneurons, proprioceptors, and the cutaneous branch of the nerve (Brushart, 1988, 1993; Madison et al., cutaneous mechanoreceptors) share a common mechanism guiding their 2007; Madison et al., 1996). However, this phenomenon is controversial axons to the cutaneous branch shortly after injury, but at long term this and depends on many factors such as the species (Martini et al., 1994; mechanism differs, and each neuron type prefers to regenerate through Robinson and Madison, 2006), the age of the animal (Brushart, 1988; the “correct” branch. Our findings evidence distinct regenerative Robinson and Madison, 2006), or the repair method (Robinson and mechanisms that could be potentially modulated to guide axons, thus Madison, 2005; Robinson and Madison, 2003). improving functional recovery after nerve transection. Most of these studies have focused on PMR, but only a few investi­ gated preferential regeneration in sensory neurons. Peripheral sensory neurons are a highly heterogeneous population that includes * Corresponding author. E-mail address: esther.udina@uab.cat (E. Udina). https://doi.org/10.1016/j.expneurol.2022.114227 Received 16 June 2022; Received in revised form 26 August 2022; Accepted 8 September 2022 Available online 13 September 2022 0014-4886/© 2022 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 proprioceptors, several touch receptors, and nociceptors and, therefore, from The Jackson Laboratory (Bar Harbor, Maine, USA): ChAT-IRES-Cre these can be found in both the cutaneous and the muscle branch of the (#006410), B6 PV-Cre (#017320) and Npy2r-ires-Cre (#029285). We nerve in uninjured nerves. Thus, there is not a general “correct path” for obtained three mice lines that expressed the red fluorescent protein all the dorsal root ganglia (DRG) neurons and addressing the question of TdTomato under the control of a specific neuronal promoter: ChAT-Cre/ specific regeneration is more complex. There are, however, few studies Ai9, PV-Cre/Ai9, and Npy2r-Cre/Ai9, respectively. Mice were housed in suggesting that sensory axons might preferentially regenerate through a controlled environment (12-h light-dark cycle, 22 ± 2 ◦ C), in open the cutaneous branch (Maki et al., 2005; Maki et al., 1996; Mears et al., cages with water and food ad libitum. 2003), although the diversity of DRG neurons was not taken into ac­ count. Furthermore, muscle afferents may have a preference for the 2.2. Surgical procedures muscle branch (Brushart et al., 2005; Madison et al., 1996). However, the evidence is scarce and further studies are needed to confirm that Fifty-three adult female (20) and male (37) mice (7- to 12-weeks old) preferential sensory regeneration exists, especially in mice. were used to evaluating the preferential regeneration of some neuron Reporter mice that express a fluorescent protein in specific neuron subtypes (Fig. 1). Mice were anesthetized with intraperitoneal ketamine subpopulations can facilitate the study of regeneration in the different (90 mg/kg) and xylazine (10 mg/kg) and the left femoral nerve was types of neurons. Genetic labeling of the neuron soma can help us exposed using an inguinal approach. The subcutaneous adipose tissue differentiate the preference in regeneration in different types of neurons was carefully removed until the nerve and its ramifications were in the DRG. Moreover, fluorescence in the axons of these neurons allows accessible. Then, the femoral nerve was cut above the bifurcation, 6 mm us to study collateral branching during regeneration, a process that proximal to the entry of the quadriceps branch into the quadriceps might be relevant in specificity (Brushart et al., 1998; Redett et al., 2005; muscle. Immediately afterward, the injured nerve was repaired with Robinson and Madison, 2013). 5–10 μl of fibrin glue. The glue was prepared by mixing thrombin (25 U/ In the present study, we aimed to take advantage of reporter trans­ ml, ICN Biomedicals) in calcium chloride (45 mmol/l, Sigma), human genic mice to investigate PMR and preferential sensory regeneration. We fibrinogen (100 mg/ml, Sigma), and bovine fibronectin (8 mg/ml, have examined specificity in motoneurons and in two sensory pop­ Sigma) in a 2:1:1 ratio (Akhter et al., 2019; Guest et al., 1997). When the ulations that have different target organs: proprioceptors, which inner­ fibrin glue polymerized, the skin was closed with a 6–0 nylon suture vate muscle spindles, and cutaneous mechanoreceptors, whose endings (Aragó). Animals were monitored periodically until the end of the are found mainly in the skin. Our results show that there is a common experiments. mechanism in regeneration that guides neurons towards the cutaneous After 1 or 8 weeks, mice were anesthetized to re-expose the left branch shortly after injury. At long term, we find that motoneurons and femoral nerve (1 week: n = 7 Pv-Cre/Ai9, n = 5 Npy2r-Cre/Ai9; 8 weeks: proprioceptors have preference for the muscle branch, whereas cuta­ n = 15 Pv-Cre/Ai9, n = 14 Npy2r-Cre/Ai9, N = 7 ChAT-Cre/Ai9). neous mechanoreceptors prefer the cutaneous branch. These results Additionally, some uninjured mice were included in the process as extend the classical concept of PMR to sensory neurons and show that controls (n = 3 Pv-Cre/Ai9 and 3 Npy2r-Cre/Ai9). The quadriceps and there might be an intrinsic neuronal mechanism guiding them towards the saphenous branches were cut at the same distance from the injury specific target organs after an injury. (approximately at 6 mm), just before the entry of the quadriceps branch into the muscle. The end of the muscle and cutaneous branch was soaked 2. Methods in Fluorogold (4%, Fluorochrome) and TrueBlue (2.5%, Setareh Biotech), respectively, and protected from the light for 45 min. To avoid 2.1. Animals spilling, we set up a small well using a piece of parafilm covered by a Vaseline circumference, we introduced the nerve stump into the well All experimental procedures were approved by the Universitat and we filled it with the retrotracers. Control experiments were per­ Autònoma de Barcelona Animal Experimentation Ethical Committee formed to confirm both retrotracers labeled neurons to the same extent. and followed the European Communities Council Directive 2010/63/ We applied TrueBlue in the muscle branch and Fluorogold in the cuta­ EU. neous branch and counted the traced motoneurons and sensory neurons, Double transgenic mice were generated by breeding homozygous Ai9 finding no significant differences (Supplementary Fig. 1). (RCL-tdT) mice (#007909) with different homozygous Cre-driver lines Fig. 1. Experimental design of the study. Transgenic mice were obtained by breeding RCL-TdTomato mice and different Cre-driver lines. The femoral nerve was transected and glued proximally to the bifurcation. After 1 or 8 weeks, the nerve was re-exposed and the retrograde tracers Fluorogold and TrueBlue were applied to the muscle and cutaneous branch, respectively. Histological evaluation was performed 1 week later. We counted the number of retrotraced TdTomato+ neurons in each animal and the axons found in the terminal branches of the nerve. 2 S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 2.3. Histological evaluation Table 1 Mean diameter of motoneurons and proprioceptors. Characterization of transgenic mice. Three naïve mice of each type Animal Neuronal type Mean diameter Correction factor were euthanized with intraperitoneal pentobarbital (30 mg/kg) and ChAT-Cre/Ai9 Motoneurons 27,540 ± 0,54 1,057 perfused with 4% paraformaldehyde in PBS. Lumbar DRGs, spinal cords, PV-Cre/Ai9 Proprioceptors 33,454 ± 0,72 0,928 tibial anterior and soleus muscle, and skin from the hind paw were harvested and stored in PBS with 30% sucrose at 4 ◦ C. DRGs, spinal cords (15 μm), and muscles (20 μm) were serially cut in a cryostat 2.4. Data analysis (Leica) and collected in glass slides. The skin was also cut in a cryostat (60 μm thick) and stored free-floating in PBS. Samples were per­ GraphPad Prism 8 (version 8.0.2) was used for the statistical anal­ meabilized with PBS and 0.3% Triton X-100 (PBST) and blocked with ysis. The normal distribution of the samples was confirmed with 1.5% normal donkey serum (Vector Laboratories). DRGs were incubated Shapiro-Wilk test (p > 0.05). Sex differences among regenerated moto­ overnight at 4 ◦ C with mouse anti-neurofilament H antibody (1/000, neurons were assessed by a t-test. Percentage of regeneration, prefer­ Biolegend), rabbit anti-calbindin D-28 K (1/200, Millipore), rabbit anti- ential regeneration, and axon collaterals were analyzed using two-way Parv (1/800, Swant), rabbit anti-CGRP (1/200, Millipore) or Griffonia ANOVA, followed by a post hoc Tukey or Sidak test (indicated in the Simplicifolia Lectin I (IB4, 10 μg/m, Vector). Muscles were incubated figures). Differences were considered statistically significant if p < 0.05. with either mouse anti-sdMyHC (1/100, DSHB) or α-bungarotoxin All data are expressed as group mean ± standard error of the mean AlexaFluor 488 conjugate (1:200, Invitrogen). The skin was incubated (SEM). with rabbit anti-PGP9.5 (1:500, Cedarlane). Slices were washed and incubated with appropriate secondary anti-mouse or anti-rabbit anti­ 3. Results bodies (AlexaFluor 488, Invitrogen) for 2 h at room temperature. For IB4, a previous incubation with rabbit anti-IB4 (1:500, Vector) was 3.1. Reporter mice label different neuronal subtypes needed. After washing, slides were mounted with Fluoromount-G me­ dium (Southern Biotech) and imaged with a confocal microscope (Leica Choline acetyltransferase (ChAT) and parvalbumin (PV) are well- SP5, 20×, z-step size 1 μm). known markers of motoneurons and proprioceptors, respectively, and Preferential regeneration. One week after applying the retrotracers, reporter transgenic mice driven by their promoters have been shown to PV-Cre/Ai9 and Npy2r-Cre/Ai9 mice were euthanized with intraperi­ label these neuron subpopulations (Blum et al., 2021; Woo et al., 2015). toneal pentobarbital (30 mg/kg) and perfused with 4% para­ These neurons innervate the muscles and, thus, in the femoral nerve, formaldehyde in PBS. The ipsilateral L3 DRGs were extracted and stored their axons are found in the muscle branch. We observed that ChAT-Cre/ in PBS with 30% sucrose at 4 ◦ C, whereas lumbar spinal cords were Ai9 animals expressed fluorescence in the motoneurons of the ventral harvested and post-fixed for 3 h before storage. Spinal cords and DRGs horn of the spinal cord (Fig. 2A), but not in the DRGs. The area distri­ were serially cut in a cryostat (Leica, 15 μm thick) in a longitudinal bution of these neurons revealed two peaks, indicating that both γ- and orientation and picked up in glass slides. We photographed DRGs with α-motoneurons were labeled (Fig. 2B). Contrarily, PV-Cre/Ai9 showed an epifluorescence microscope (Olympus BX51, Olympus, Hamburg, fluorescence in large-diameter neurons in the DRG (Fig. 2B). These Germany) equipped with a digital camera (Olympus DP50, Olympus, neurons co-labeled mostly with the proprioceptive marker parvalbumin Hamburg, Germany) and CellSens Digital Imaging software (version 1.9, (84.62 ± 2.44%) and the marker for myelinated neurons NFH (82.37 ± Olympus, Hamburg, Germany). All neurons were counted in one out of 0.62%) (Fig. 2C). Furthermore, 9.82 ± 1.05% of these neurons were three slices (every 45 μm). We calculated the percentage of sensory positive for the Aβ-LTMR marker, Calb1. In the muscle, we saw that neurons that regenerated using the muscle or the cutaneous branch by ChAT-Cre/Ai9 had fluorescence in fibers innervating neuromuscular counting the TdTomato+/TrueBlue+ or TdTomato+/Fluorogold+ junctions, whereas PV-Cre/Ai9 fluorescence co-labeled with slow neurons in the DRG and dividing them by the total counted number of developmental myosin heavy chain, a muscle spindle marker (Fig. 2A). retrotraced neurons. The spinal cords of the same animals were used to Applying retrotracers to uninjured femoral nerves, we corroborated that count the motoneurons that regenerated through each branch. all motoneuron and essentially all proprioceptive axons (97.69 ± Axon collaterals. A subset of the long-term animals (8 weeks post- 0,82%) were found in the muscle branch of the nerve in our animal injury) was used to calculate the number of axon collaterals per models (Fig. 2D). neuron. After perfusion, the ipsilateral L2 and L3 DRGs and the femoral Axons of touch mechanoreceptors that innervate skin receptors are nerves of PV-Cre/Ai9 animals were harvested and stored in PBS with found in the cutaneous branch of the femoral nerve. Npy2r reporter mice 30% sucrose at 4 ◦ C. ChAT-Cre/Ai9 femoral nerves and lumbar spinal have been described to label low-threshold mechanoreceptors (LTMR) cords were extracted, the latter being post-fixed before storage. DRGs and nociceptors (Arcourt et al., 2017; Li et al., 2011). We characterized and spinal cords were processed as described above. The distal branches our Npy2r-Cre/Ai9 reporter mice and found that these animals label of the femoral nerve were put on glass slides and mounted with Fluo­ large- and small-diameter DRG neurons (Fig. 2B). Terminal axons could romount. We applied pressure on the coverslips to flatten the nerves and be found in the skin either as free endings or associated with hair fol­ image the distal part in the confocal microscope (Leica SP5, 20×, z-step licles (Fig. 2A). In the DRG, 40.7 ± 1.39% of Npy2r+ neurons were size 0.38 μm). The number of collaterals in the cutaneous branch was myelinated, whereas 61 ± 1.66% of the neurons co-labeled with the calculated by dividing the number of TdTomato+ axons in the saphenous peptidergic marker CGRP (Fig. 2C). Only 7.23 ± 0.69% of the Npy2r nerve by the number of TdTomato+/TrueBlue+ neurons. Likewise, the neurons were positive for the non-peptidergic marker IB4 and 1.54 ± collaterals in the muscle branch were obtained by dividing the number 0.4% for Calb1. We observed that most axons of Npy2r+ neurons were in of axons in the quadriceps branch by the number of TdTomato+/Fluo­ the cutaneous branch (82.68 ± 2,17%), but 17.32% of these neurons did rogold+ neurons. All neuron counts were corrected using the Aber­ not have their axons in this branch (Fig. 2D). Altogether, we used this crombie correction (Abercrombie, 1946), being the correction factor t/ transgenic animal as a reporter of cutaneous mechanoreceptors ((t + H)ssf), where t is the section thickness, H is the mean high of the (including touch and nociceptive neurons). object and ssf is the sampling factor. H was estimated by measuring the area of the soma of 100 neurons of each type and determining the mean 3.2. Regeneration does not differ between females and males but varies diameter of each neuronal population (Table 1). between neuron subtypes Regeneration was assessed in both females and males in all the 3 S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 Fig. 2. Characterization of ChAT-Cre/ Ai9, PV-Cre/Ai9, and Npy2r-Cre/Ai9 transgenic mice. A) Microscope images of the neuron somas in the spinal cord or the DRG and the terminal axons in the target organs. ChAT-Cre/Ai9 mice ex­ press fluorescence in the motoneurons of the ventral horn of the spinal cord and distally in the muscles, where their axons co-label with neuromuscular junctions (green). PV-Cre/Ai9 fluores­ cence is found in large neurons of the DRG and their axons co-label with sdMyHC, a marker of the muscle spindle (green). Sensory neurons of Npy2r-Cre/ Ai9 mice are found in the DRG and their axons in the skin, either as free endings or associated with hair follicles (indicated by the arrow, PGP9.5 in green). B) Size distribution of neurons according to their somatic area. C) Mi­ croscope images and percentage of co- labeling of TdTomato+ neurons in the DRG with the markers parvalbumin (Parv), NFH, CGRP, IB4, and Calb1 (n = 3/group). D) Percentage of neurons that have their axons in the muscle (orange) or cutaneous branch (purple) in unin­ jured nerves (n = 3/group). Scale bar: 50 μm. (For interpretation of the refer­ ences to colour in this figure legend, the reader is referred to the web version of this article.) 4 S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 studies (Fig. 3A). To ensure that the counted number of regenerated = 0.002, cutaneous vs muscle branch). At long term, neurons showed a neurons does not differ between sexes, we analyzed the total number of preferential regeneration towards the branch they originally innervated. regenerated motoneurons 8 weeks after the injury (Fig. 3B). We chose Motoneurons and proprioceptors regenerated significantly more in the motoneurons because the three transgenic mice (ChAT-Cre/Ai9, PV- muscle branch (motoneurons: 40.50 ± 1.68 vs 46.76 ± 1.26, p = 0.023; Cre/Ai9, and Npy2r-Cre/Ai9) were used for the retrotracer counts in proprioceptors: 12.99 ± 1.20 vs 20.04 ± 1.58, p < 0.0001, cutaneous vs the spinal cord. We found that regeneration did not differ between fe­ muscle branch). Contrarily, more cutaneous mechanoreceptors regen­ males and males (p = 0.7884) and included both sexes for the statistical erated in the cutaneous than in the muscle branch of the nerve (193 ± analysis of the following experiments. 18.96 vs 118.36 ± 8.88, p = 0.0003, cutaneous vs muscle branch). There The percentage of regenerated motoneurons, proprioceptors, and was some variability between animals, but most of them showed the cutaneous mechanoreceptors was assessed by counting the total number same tendency to project to the “preferred” branch (Table 2). of retrotraced neurons that regenerated after the injury, including those in the cutaneous branch, the muscle branch, and those double-labeled. 3.4. Motoneurons extend more collaterals than proprioceptors After 1 week, 51.43 ± 11.7% of the cutaneous mechanoreceptors re­ generated up to the evaluation point, significantly more than moto­ We counted the number of fluorescent axons in the cutaneous and neurons (24.96 ± 6.27%, p = 0.017) and proprioceptors (13.64 ± muscle branch of the ChAT-Cre/Ai9 and PV-Cre/Ai9 animals 8 weeks 4.45%, p < 0.001). Eight weeks after the injury, the population that had after the femoral nerve injury to compare two myelinated populations regenerated more was the Npy2r+ (89.29 ± 6.83%), followed by mo­ (Fig. 5). For the estimation of the number of collaterals, we divided the toneurons (72.92 ± 1.64%, p = 0.012 vs Npy2r) and, lastly, pro­ number of axons by the number of traced neurons in each animal prioceptors (51.03 ± 3.57%, p < 0.001 vs motoneurons and Npy2r). (Fig. 4). Since Npy2r-Cre/Ai9 animals have an extensive number of axons, most of them unmyelinated, these animals were not included in 3.3. Motor and sensory neurons have preferential regeneration this analysis. We found that motoneurons had a mean of 2.25 (± 0.12) sprouts in the muscle branch and 2.46 (± 0.22) in the cutaneous branch. The proportion of neurons that regenerated through the muscle Proprioceptors extended significantly fewer collaterals in the muscle branch, the cutaneous branch, or both were assessed 1 and 8 weeks after (1.57 ± 0.14, p = 0.023) compared to motoneurons and, although non- the injury (Fig. 4). One week after the injury, all the studied neuron significant, these showed the same tendency in the cutaneous branch populations regenerated up to the evaluation point more through the (1.90 ± 0.18, p = 0.064). The two-way ANOVA revealed a significant cutaneous branch than the muscle branch (motoneurons: 21.8 ± 7.73 vs effect of the branch (p = 0.0498), but multiple comparisons failed to find 9.13 ± 2.03, p = 0.006; proprioceptors: 7.06 ± 2.64 vs 2.14 ± 0.77, p = differences between the collaterals in the different branches. 0,046; cutaneous mechanoreceptors: 146.8 ± 44.68 vs 39.6 ± 11.03, p Fig. 3. Total regeneration in both femoral branches. A) Regeneration in females and males 8 weeks after the injury. The total counted number of neurons that regenerated in both branches of the femoral nerve 8 weeks after the injury of each neuronal type is similar between females (green bars) and males (blue bars). B) No differences were found in the number of motoneurons that had regenerated 8 weeks after the injury between males and females (t-test, p = 0.788). C) Percentage of neurons that re­ generated 1 or 8 weeks after injury. Com­ plete regeneration (100%) is considered when reaching control values. Individual values are represented as orange circles (fe­ males) and blue squares (males). Two-way ANOVA and Tukey post hoc test indicated differences between neuron types. *p < 0.05 motoneurons vs Npy2r; $$$p < 0.001, $$$$p < 0.0001 PV vs Npy2r; ###p < 0.001 mo­ toneurons vs PV. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 5 S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 Fig. 4. Evaluation of the preferential regeneration of motoneurons, proprioceptors, and cutaneous mechanoreceptors. A) Counted number of neurons of each type in the muscle branch (black bars), the cutaneous branch (white bars), or both (striped bars). Individual values are represented as orange circles (females) and blue squares (males). B) Microscope images of the somas of motoneurons, proprioceptors, and cutaneous mechanoreceptors 8 weeks after the injury. Neurons were counted when the TdTomato (red) co-localized with TrueBlue (blue) or Fluorogold (yellow). C) Percentage of retrotraced neurons in the muscle or the cutaneous branch calculated over the total number of regenerated neurons. The preference for the muscle branch is represented in the left axis, whereas the cutaneous preference is shown in the right axis. Differences between branches were assessed by a two-way ANOVA followed by a Tukey post hoc test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 muscle vs cutaneous branch. Scale bar: 100 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) 4. Discussion Table 2 Number of animals showing preference towards each branch. In this study, we used reporter mice to examine peripheral regen­ Preferred Motoneurons Proprioceptors Cutaneous eration of motor and different sensory neurons. Our findings indicate branch mechanoreceptors that each type of neuron has an intrinsic mechanism that leads to a 1 8 1 8 1 8 regenerative preference towards its original target. week weeks week weeks weeks weeks Cutaneous 5/7 4/21 5/7 2/15 4/5 13/14 4.1. Reporter mice for the study of regeneration Muscle 2/7 16/21 2/7 12/15 1/5 1/14 Equal 0/7 1/21 0/7 1/15 0/5 0/14 Preference in regeneration has been controversial over the decades. 6 S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 Fig. 5. Axon collaterals of motoneurons and proprioceptors in the lesioned femoral nerve. A) Microscope images of the distal branches of the femoral nerve 8 weeks after the axotomy. Motor (ChAT-Cre/Ai9) and proprioceptive (PV-Cre/Ai9) axons are labeled in red. B) Regenerative sprouts in each type of neuron were calculated by dividing the number of fluorescent axons by the number of traced somas in each branch. Individual values are represented as orange circles (females) and blue squares (males). Two-way ANOVA was followed by Sidak’s post hoc test. *p < 0.05, **p < 0.01 ChAT+ neurons vs PV+ neurons. Scale bar: 100 μm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.) Brushart first reported the existence of PMR in rats 35 years ago is expected. Specifically, we found 17.32% of these neurons are in the (Brushart, 1988), but results in mice were not as clear (Mears et al., muscle branch. Being the natural condition, this increases the 2003; Robinson and Madison, 2006; Robinson and Madison, 2003). complexity of the evaluation of sensory preferential regeneration. Additionally, little attention has been paid to sensory specificity. In the DRG, neurons are remarkably heterogeneous and have diverse target organs. Whereas muscle spindles are innervated by proprioceptors, 4.2. Regeneration speed of neurons LTMR innervate specialized structures in the skin (Li et al., 2011). Therefore, we need to find strategies to study each neuron subtype We found that the regeneration speed can widely vary depending on particularities. the type of fiber. One week after the transection, more cutaneous Three lines of reporter mice were used in our study, each labeling a mechanoreceptors arrived at the point where we applied the retrograde different neuron. Among sensory neurons, we chose two populations tracers compared to motoneurons and proprioceptors. Unmyelinated that should regenerate through opposite branches. Our results fibers have been described to recover their function earlier than confirmed that PV-Cre/Ai9 mice labeled proprioceptors as previously myelinated fibers (Navarro et al., 1994). As our cutaneous mechanore­ described (Woo et al., 2015). In contrast, Npy2r-Cre/Ai9 mice were used ceptors include myelinated and unmyelinated neurons, a faster speed in to study cutaneous mechanoreceptors. the regeneration of unmyelinated fibers could explain their advantage We found that the population labeled in these mice was heteroge­ over motoneurons and proprioceptors. Similarly, sensory neurons have neous, including small and large neurons. Previous reports described been reported to regenerate faster than motoneurons (Brushart et al., that the marker Npy2r was associated with large diameter Aβ-LTMR (Li 2020; Dolenc and Janko, 1976; Kawasaki et al., 2000). In contrast, using et al., 2011), but we found that only 1.54% of the labeled neurons were retrograde tracers, it has been described that motoneurons regenerate Calb1+. Our results are in line with recent reports that suggest that better than sensory neurons (da Silva et al., 1985) and that myelinated Npy2r is a marker for A-fiber nociceptors and A-high threshold mech­ sensory fibers can regenerate at the same speed as unmyelinated fibers anoreceptors (HTMR) (Arcourt et al., 2017; Chiu et al., 2014; Patil et al., (Lozeron et al., 2004). These contradictions can be explained by the 2018). These studies showed similar co-labeling with the markers NFH, heterogeneity of the sensory neurons. We found a greater advantage in CGRP, and IB4 and found terminal free endings and some lanceolate the regeneration of cutaneous mechanoreceptors compared to pro­ endings in the skin. The discrepancies between studies might be related prioceptors than to motoneurons. Therefore, regeneration should be to the different techniques used to design the transgenic mice (Li et al., analyzed in the different types of neurons considering sensory neurons 2011). Although our Npy2r-Cre/Ai9 mice do not specifically label this diversity. Aβ-LTMR population, the fluorescent neurons are mainly projecting to We also observed that the number of neurons that regenerated 8 the cutaneous branch and do not overlap with those from PV-Cre/Ai9 weeks after the injury differed between populations. Regenerating axons mice. Thus, Npy2r+ neurons represent a heterogeneous population of had sufficient time to reach the evaluation point after 8 weeks. However, mechanoreceptors, myelinated and unmyelinated, that generally we found that whereas cutaneous mechanoreceptors and motoneurons innervate the skin. Therefore, we used these mice to comparatively had substantially regenerated (89.29 and 72.92%, respectively), only study preferential regeneration in another sensory neuron 51.03% of proprioceptors were able to arrive at the distal point. An subpopulation. explanation for this phenomenon can be related to the worse regener­ Importantly, some LTMR and HTMR innervate muscles and articu­ ative capability of these neurons. If motoneurons regenerated and lations so a proportion of these neurons projecting to the muscle branch reinnervated target muscles faster than proprioceptors, the regenerative environment would change. Schwann cells might re-establish contact 7 S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 with axons and return to a mature non-reactive state. In this case, the capacity must be different. Our data reveal that not only motor axons, still regenerating proprioceptive axons would not have the proper pro- but also proprioceptors and cutaneous mechanoreceptors preferentially regenerative conditions to continue extending towards the muscle. regenerate towards their correct branch. However, although significant, Thus, those neurons with slower regeneration rates might fail to this preference is remarkably low (53.88% for motoneurons, 60,44% for continue extending their axons in a less permissive environment and proprioceptors, 60,97% for cutaneous mechanoreceptors) and, hence, their regeneration would remain incomplete. hardly relevant in terms of functional outcome. A better knowledge of the intrinsic growth programs of different neurons would help us design 4.3. Preferential regeneration in motor and sensory neurons strategies to enhance this preferential regeneration and, thus, the functional recovery after nerve lesions. It has been hypothesized that PMR in the femoral nerve occurs in two phases. During the first weeks, motor axons grow into both distal branches, forming collaterals (Brushart et al., 1998; Brushart, 1993; 4.4. Collateral branching in myelinated neurons Uschold et al., 2007). This phase results from pathway-axon interaction (Redett et al., 2005) and collaterals might prefer the branch with the After axotomy, regenerating axons have been described to sprout, most relative amounts of basal lamina (Robinson and Madison, 2009). sending collaterals to up to 10 endoneurial tubes (Witzel et al., 2005). Then, PMR might be achieved by pruning the collaterals from the Sprouting is a pathfinding mechanism for the regenerating neurons since cutaneous branch while maintaining those in their correct path. increasing collaterals provide more opportunities for the axons to find Here, we analyzed PMR and sensory preference at earlier times than their correct targets. After reinnervation, sprouts are reduced in a pro­ previous studies and found that more neurons arrived at the cutaneous cess that can take several months (MacKinnon et al., 1991). In our branch evaluation point after 1 week than to the muscle branch, inde­ model, motoneurons and proprioceptors showed some collaterals in pendently of the neuron population. This favors the hypothesis that both branches of the nerve after axotomy that varied from 1.14 to 3.06 during the initial phases of regeneration, collaterals prefer the branch per neuron. Classical studies described an average of five daughter with more Schwann cells. Even though the femoral nerve has a similar axonal sprouts (Fu and Gordon, 1997). However, a significant degree of cross-sectional area between branches, there are proportionally more reinnervation is expected 8 weeks after the injury, and that could myelinated axons in the cutaneous branch than in the muscle branch explain the smaller number of collaterals in our samples. (Takahashi et al., 1999). More myelinated axons in a branch mean that We found that the cutaneous branch contained more collaterals than more Schwann cells are available for motoneurons to establish contacts the muscle branch. This phenomenon was previously described for and probably also more trophic support coming from these denervated motoneurons by Redett et al. (2005) in a rat femoral nerve injury model cells (Robinson and Madison, 2009; Takahashi et al., 1999). Further­ with DRG excision. Here, the use of reporter mice allowed us not only to more, several studies describe that Schwann cells from muscle and study collaterals without the need to manipulate sensory axons but also cutaneous branches secrete distinct trophic factors (Brushart et al., to study sprouting in a specific sensory population. 1998; Höke et al., 2006). Whereas denervated muscle nerves express Trophic factors such as NGF or BDNF have been described to increase higher levels of pleiotrophin (PTN), cutaneous nerves predominantly axonal sprouting (Boyd and Gordon, 2003; Diamond et al., 1992; Gallo upregulate hepatocyte growth factor (HGF), brain-derived neurotrophic and Letourneau, 1998; Streppel et al., 2002). Interestingly, these factors factor (BDNF), and nerve growth factor (NGF) (Brushart et al., 2013; are significantly upregulated in the cutaneous but not in the muscle Höke et al., 2006). BDNF is required for axonal growth and myelination branch (Höke et al., 2006). Thus, the differential expression of factors in (Zhang et al., 2000) and has been shown to improve motor axon cutaneous and muscle Schwann cells might influence the presence of regeneration (Boyd and Gordon, 2003), so it could be driving motor and regenerating collaterals. Additionally, 8 weeks after the lesion many sensory sprouts towards the cutaneous branch. axons can arrive to target organs. Both motoneurons and proprioceptors In contrast, at long term we found a preference of axons towards can reinnervate muscles, but they are not able to find a target in the skin. their correct branch: motor and proprioceptive neurons regenerated If these sprouts are a pathfinding mechanism, neurons reinnervating the more through the muscle branch, whereas cutaneous mechanoreceptors muscle would start reducing the sprouts when they reach their target were more prone to regenerate into the saphenous nerve. This prefer­ organ, which would explain the lower number of motor and proprio­ ence could depend on the type of trophic factors secreted by the distal ceptive collaterals in the muscle branch. target organs (Madison et al., 2007; Uschold et al., 2007) or the differ­ Our analysis indicated a significant difference in collaterals not only ences between Schwann cells of each branch (reviewed in Bolívar et al., between branches but also between neuronal populations. We found 2020). In contrast to classical studies, we did not find a high number of that motoneurons extended more sprouts than proprioceptors in the double-labeled neurons neither 1 week nor 8 weeks after the injury. This muscle branch. Interestingly, motoneurons increase the expression of might contradict the collateral pruning hypothesis since our results do BDNF receptor trkB after axotomy (Fernandes et al., 1998; Kobayashi not corroborate that neurons extend many collaterals to both branches et al., 1996), as well as the neurotrophin low-affinity receptor p75NTR that could account for the later specificity. Contrarily, our data suggest (Johnson et al., 1999). If NGF and BDNF are partly responsible for that signals from target organs predominate over those generated by increasing axonal sprouting, the increased expression of their receptors Schwann cells. These signals can arrive via axonal transport and in motoneurons may explain the increased number of collaterals found diffusion-driven movement, being the latter more significant (Madison in this population. and Robinson, 2014). If this is the case, different types of neurons would differentially respond to factors released by target organs. Denervated 5. Conclusion muscle has been shown to increase the expression of glial cell line- derived neurotrophic factor (GDNF), fibroblast growth factor (FGF), In this study we showed that, shortly after injury, there is a common and HGF (Lie and Weis, 1998; Yamaguchi et al., 2004; Zhao et al., 2004), mechanism in regeneration that guides axons towards the cutaneous whereas denervated skin upregulates NGF (Mearow et al., 1993). branch. At long term, axons prefer to regenerate through their “correct” However, evidence that these factors can specifically attract one type of branch. Moreover, regenerative capability and sprout dynamics differ axon is not consistent in the literature (reviewed in Allodi et al., 2012). between neuron subtypes. Our results evidence different regenerative Finally, most studies addressing specific regeneration have focused mechanisms of peripheral neurons, that can be potentially modulated to on the differences between the motor and sensory branches, but not on improve functional recovery after nerve injury. the intrinsic differences between neuronal populations. If axons have a Supplementary data to this article can be found online at https://doi. selective preference for a substrate or trophic factor, their regenerative org/10.1016/j.expneurol.2022.114227. 8 S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 Funding corticospinal tract axons to human Schwann cell grafts. J. Neurosci. Res. 50, 888–905. https://doi.org/10.1002/(SICI)1097-4547(19971201)50:5<888::AID- JNR24>3.0.CO;2-W. This work was supported by the project SAF2017–84464-R and the Höke, A., Redett, R., Hameed, H., Jari, R., Zhou, C., Li, Z.B., Griffin, J.W., Brushart, T.M., grant FPU17/03657 from Ministerio de Ciencia, Innovación y Uni­ 2006. Schwann cells express motor and sensory phenotypes that regulate axon versidades of Spain. regeneration. J. Neurosci. 26, 9646–9655. https://doi.org/10.1523/ JNEUROSCI.1620-06.2006. Johnson, H., Hökfelt, T., Ulfhake, B., 1999. Expression of p75NTR, trkB and trkC in nonmanipulated and axotomized motoneurons of aged rats. Mol. Brain Res. 69, Declaration of Competing Interest 21–34. https://doi.org/10.1016/S0169-328X(99)00068-6. Kawasaki, Y., Yoshimura, K., Harii, K., Park, S., 2000. Identification of myelinated motor The authors declare no competing financial interests. and sensory axons in a regenerating mixed nerve. J. Hand. Surg. [Am.] 25, 104–111. https://doi.org/10.1053/jhsu.2000.jhsu025a0104. Kobayashi, N.R., Bedard, A.M., Hincke, M.T., Tetzlaff, W., 1996. Increased expression of Data availability BDNF and trkB mRNA in rat facial Motoneurons after Axotomy. Eur. J. Neurosci. 8, 1018–1029. https://doi.org/10.1111/j.1460-9568.1996.tb01588.x. Li, L., Rutlin, M., Abraira, V.E., Cassidy, C., Kus, L., Gong, S., Jankowski, M.P., Luo, W., Data will be made available on request. Heintz, N., Koerber, H.R., Woodbury, C.J., Ginty, D.D., 2011. The functional organization of cutaneous low-threshold mechanosensory neurons. Cell 147, References 1615–1627. https://doi.org/10.1016/j.cell.2011.11.027. Lie, D.C., Weis, J., 1998. GDNF expression is increased in denervated human skeletal muscle. Neurosci. Lett. 250, 87–90. https://doi.org/10.1016/S0304-3940(98) Abercrombie, M., 1946. Estimation of nuclear population from microtome sections. Anat. 00434-0. Rec. 94, 239–247. https://doi.org/10.1002/ar.1090940210. Lozeron, P., Krarup, C., Schmalbruch, H., 2004. Regeneration of unmyelinated and Akhter, E., Rotterman, T., English, A., Alvarez, F., 2019. Sciatic nerve cut and repair myelinated sensory nerve fibres studied by a retrograde tracer method. J. Neurosci. using fibrin glue in adult mice. Bio-Protocol 9, 1–11. https://doi.org/10.21769/ Methods 138, 225–232. https://doi.org/10.1016/j.jneumeth.2004.04.012. bioprotoc.3363. MacKinnon, S.E., Dellon, A.L., O’Brien, J.P., 1991. Changes in nerve fiber numbers distal Allodi, I., Udina, E., Navarro, X., 2012. Specificity of peripheral nerve regeneration: to a nerve repair in the rat sciatic nerve model. Muscle Nerve 14, 1116–1122. interactions at the axon level. Prog. Neurobiol. 98, 16–37. https://doi.org/10.1016/ https://doi.org/10.1002/mus.880141113. j.pneurobio.2012.05.005. Madison, R.D., Robinson, G.A., 2014. Accuracy of regenerating motor neurons: influence Arcourt, A., Gorham, L., Dhandapani, R., Prato, V., Taberner, F.J., Wende, H., of diffusion in denervated nerve. Neuroscience 273, 128–140. https://doi.org/ Gangadharan, V., Birchmeier, C., Heppenstall, P.A., Lechner, S.G., 2017. Touch 10.1016/j.neuroscience.2014.05.016. receptor-derived sensory information alleviates acute pain signaling and fine-Tunes Madison, R.D., Archibald, S.J., Brushart, T.M., 1996. Reinnervation accuracy of the rat nociceptive reflex coordination. Neuron 93, 179–193. https://doi.org/10.1016/j. femoral nerve by motor and sensory neurons. J. Neurosci. 16, 5698–5703. https:// neuron.2016.11.027. doi.org/10.1523/jneurosci.16-18-05698.1996. Blum, J.A., Klemm, S., Shadrach, J.L., Guttenplan, K.A., Nakayama, L., Kathiria, A., Madison, R.D., Robinson, G.A., Chadaram, S.R., 2007. The specificity of motor neurone Hoang, P.T., Gautier, O., Kaltschmidt, J.A., Greenleaf, W.J., Gitler, A.D., 2021. regeneration (preferential reinnervation). Acta Physiol. 189, 201–206. https://doi. Single-cell transcriptomic analysis of the adult mouse spinal cord reveals molecular org/10.1111/j.1748-1716.2006.01657.x. diversity of autonomic and skeletal motor neurons. Nat. Neurosci. 24, 572–583. Maki, Y., Yoshizu, T., Tajima, T., Narisawa, H., 1996. The selectivity of motor axons https://doi.org/10.1038/s41593-020-00795-0. during regeneration with sensory axons. J. Reconstr. Microsurg. 12, 553–557. Bolívar, S., Navarro, X., Udina, E., 2020. Schwann cell role in selectivity of nerve https://doi.org/10.1055/s-2007-1006629. regeneration. Cells 9, 2131. https://doi.org/10.3390/cells9092131. Maki, Y., Yoshizu, T., Tsubokawa, N., 2005. Selective regeneration of motor and sensory Boyd, J.G., Gordon, T., 2003. Glial cell line-derived neurotrophic factor and brain- axons in an experimental peripheral nerve model without endorgans. Scand. J. Plast. derived neurotrophic factor sustain the axonal regeneration of chronically Reconstr. Surg. Hand Surg. 39, 257–260. https://doi.org/10.1080/ axotomized motoneurons in vivo. Exp. Neurol. 183, 610–619. https://doi.org/ 0284431051006510. 10.1016/S0014-4886(03)00183-3. Martini, R., Schachner, M., Brushart, T., 1994. The L2/HNK-1 carbohydrate is Brushart, T., 1988. Preferential reinnervation of motor nerves by regenerating motor preferentially expressed by previously motor axon-associated Schwann cells in axons. J. Neurosci. 8, 1026–1031. https://doi.org/10.1523/JNEUROSCI.08-03- reinnervated peripheral nerves. J. Neurosci. 14, 7180–7191. https://doi.org/ 01026.1988. 10.1523/JNEUROSCI.14-11-07180.1994. Brushart, T., Kebaish, F., Wolinsky, R., Skolasky, R., Li, Z., Barker, N., 2020. Sensory Mearow, K.M., Kril, Y., Diamond, J., 1993. Increased NGF mRNA expression in axons inhibit motor axon regeneration in vitro. Exp. Neurol. 323, 113073 https:// denervated rat skin. Neuroreport 4, 351–354. https://doi.org/10.1097/00001756- doi.org/10.1016/j.expneurol.2019.113073. 199304000-00002. Brushart, T.M., Gerber, J., Kessens, P., Chen, Y.G., Royall, R.M., 1998. Contributions of Mears, S., Schachner, M., Brushart, T.M., 2003. Antibodies to myelin-associated pathway and neuron to preferential motor reinnervation. J. Neurosci. 18, glycoprotein accelerate preferential motor reinnervation. J. Peripher. Nerv. Syst. 8, 8674–8681. https://doi.org/10.1523/jneurosci.18-21-08674.1998. 91–99. https://doi.org/10.1046/j.1529-8027.2003.03012.x. Brushart, T.M., Jari, R., Verge, V., Rohde, C., Gordon, T., 2005. Electrical stimulation Navarro, X., Verdú, E., Butı́, M., 1994. Comparison of regenerative and Reinnervating restores the specificity of sensory axon regeneration. Exp. Neurol. 194, 221–229. capabilities of different functional types of nerve fibers. Exp. Neurol. 129, 217–224. https://doi.org/10.1016/j.expneurol.2005.02.007. https://doi.org/10.1006/exnr.1994.1163. Brushart, T.M., Aspalter, M., Griffin, J.W., Redett, R., Hameed, H., Zhou, C., Wright, M., Patil, M.J., Hovhannisyan, A.H., Akopian, A.N., 2018. Characteristics of sensory Vyas, A., Höke, A., 2013. Schwann cell phenotype is regulated by axon modality and neuronal groups in CGRP-cre-ER reporter mice: comparison to Nav1.8-cre, TRPV1- central–peripheral location, and persists in vitro. Exp. Neurol. 247, 272–281. cre and TRPV1-GFP mouse lines. PLoS One 13, e0198601. https://doi.org/10.1371/ https://doi.org/10.1016/j.expneurol.2013.05.007. journal.pone.0198601. Brushart, T.M.E., 1993. Motor axons preferentially reinnervate motor pathways. Redett, R., Jari, R., Crawford, T., Chen, Y.G., Rohde, C., Brushart, T.M., 2005. Peripheral J. Neurosci. 13, 2730–2738. https://doi.org/10.1523/jneurosci.13-06-02730.1993. pathways regulate motoneuron collateral dynamics. J. Neurosci. 25, 9406–9412. Chiu, I.M., Barrett, L.B., Williams, E.K., Strochlic, D.E., Lee, S., Weyer, A.D., Lou, S., https://doi.org/10.1523/JNEUROSCI.3105-05.2005. Bryman, G.S., Roberson, D.P., Ghasemlou, N., Piccoli, C., Ahat, E., Wang, V., Robinson, G.A., Madison, R.D., 2003. Preferential motor reinnervation in the mouse: Cobos, E.J., Stucky, C.L., Ma, Q., Liberles, S.D., Woolf, C.J., 2014. Transcriptional comparison of femoral nerve repair using a fibrin sealant or suture. Muscle Nerve 28, profiling at whole population and single cell levels reveals somatosensory neuron 227–231. https://doi.org/10.1002/mus.10422. molecular diversity. Elife 3, 1–32. https://doi.org/10.7554/eLife.04660. Robinson, G.A., Madison, R.D., 2005. Manipulations of the mouse femoral nerve Diamond, J., Holmes, M., Coughlin, M., 1992. Endogenous NGF and nerve impulses influence the accuracy of pathway reinnervation by motor neurons. Exp. Neurol. regulate the collateral sprouting of sensory axons in the skin of the adult rat. 192, 39–45. https://doi.org/10.1016/j.expneurol.2004.10.013. J. Neurosci. 12, 1454–1466. https://doi.org/10.1523/JNEUROSCI.12-04- Robinson, G.A., Madison, R.D., 2006. Developmentally regulated changes in femoral 01454.1992. nerve regeneration in the mouse and rat. Exp. Neurol. 197, 341–346. https://doi. Dolenc, V., Janko, M., 1976. Nerve regeneration following primary repair. Acta org/10.1016/j.expneurol.2005.10.007. Neurochir. 34, 223–234. https://doi.org/10.1007/BF01405877. Robinson, G.A., Madison, R.D., 2009. Influence of terminal nerve branch size on motor Fernandes, K.J.L., Kobayashi, N.R., Jasmin, B.J., Tetzlaff, W., 1998. Acetylcholinesterase neuron regeneration accuracy. Exp. Neurol. 215, 228–235. https://doi.org/10.1016/ gene expression in Axotomized rat facial Motoneurons is differentially regulated by j.expneurol.2008.10.002. Neurotrophins: correlation with trkB and trkC mRNA levels and isoforms. Robinson, G.A., Madison, R.D., 2013. Motor neuron target selectivity and survival after J. Neurosci. 18, 9936–9947. https://doi.org/10.1523/JNEUROSCI.18-23- prolonged axotomy. Restor. Neurol. Neurosci. 31, 451–460. https://doi.org/ 09936.1998. 10.3233/RNN-120301. Fu, S.Y., Gordon, T., 1997. The cellular and molecular basis of peripheral nerve da Silva, C.F., Madison, R., Dikkes, P., Chiu, T.-H., Sidman, R.L., 1985. An in vivo model regeneration. Mol. Neurobiol. 14, 67–116. https://doi.org/10.1007/BF02740621. to quantify motor and sensory peripheral nerve regeneration using bioresorbable Gallo, G., Letourneau, P.C., 1998. Localized sources of Neurotrophins initiate axon nerve guide tubes. Brain Res. 342, 307–315. https://doi.org/10.1016/0006-8993 collateral sprouting. J. Neurosci. 18, 5403–5414. https://doi.org/10.1523/ (85)91130-8. JNEUROSCI.18-14-05403.1998. Streppel, M., Azzolin, N., Dohm, S., Guntinas-Lichius, O., Haas, C., Grothe, C., Guest, J.D., Hesse, D., Schnell, L., Schwab, M.E., Bunge, M.B., Bunge, R.P., 1997. Wevers, A., Neiss, W.F., Angelov, D.N., 2002. Focal application of neutralizing Influence of IN-1 antibody and acidic FGF-fibrin glue on the response of injured 9 S. Bolívar and E. Udina Experimental Neurology 358 (2022) 114227 antibodies to soluble neurotrophic factors reduces collateral axonal branching after mechanotransduction channel for proprioception. Nat. Neurosci. 18, 1756–1762. peripheral nerve lesion. Eur. J. Neurosci. 15, 1327–1342. https://doi.org/10.1046/ https://doi.org/10.1038/nn.4162. j.1460-9568.2002.01971.x. Yamaguchi, A., Ishii, H., Morita, I., Oota, I., Takeda, H., 2004. mRNA expression of Takahashi, Y., Maki, Y., Yoshizu, T., Tajima, T., 1999. Both stump area and volume of fibroblast growth factors and hepatocyte growth factor in rat plantaris muscle distal sensory nerve segments influence the regeneration of sensory axons in rats. following denervation and compensatory overload. Pflugers Arch. - Eur. J. Physiol. Scand. J. Plast. Reconstr. Surg. Hand Surg. 33, 177–180. https://doi.org/10.1080/ 448, 539–546. https://doi.org/10.1007/s00424-004-1282-5. 02844319950159424. Zhang, J.-Y., Luo, X.-G., Xian, C.J., Liu, Z.-H., Zhou, X.-F., 2000. Endogenous BDNF is Uschold, T., Robinson, G.A., Madison, R.D., 2007. Motor neuron regeneration accuracy: required for myelination and regeneration of injured sciatic nerve in rodents. Eur. J. balancing trophic influences between pathways and end-organs. Exp. Neurol. 205, Neurosci. 12, 4171–4180. https://doi.org/10.1046/j.1460-9568.2000.01312.x. 250–256. https://doi.org/10.1016/j.expneurol.2007.02.005. Zhao, C., Veltri, K., Li, S., Bain, J.R., Fahnestock, M., 2004. NGF, BDNF, NT-3, and GDNF Witzel, C., Rohde, C., Brushart, T.M., 2005. Pathway sampling by regenerating mRNA expression in rat skeletal muscle following denervation and sensory peripheral axons. J. Comp. Neurol. 485, 183–190. https://doi.org/10.1002/ protection. J. Neurotrauma 21, 1468–1478. https://doi.org/10.1089/ cne.20436. neu.2004.21.1468. Woo, S.-H., Lukacs, V., de Nooij, J.C., Zaytseva, D., Criddle, C.R., Francisco, A., Jessell, T. M., Wilkinson, K.A., Patapoutian, A., 2015. Piezo2 is the principal 10