Response of terminal Schwann cells following volumetric muscle loss injury

Abstract

An often-overlooked component of traumatic skeletal muscle injuries is the impact on the nervous system and resultant innervation of the affected muscles. Recent work in a rodent model of volumetric muscle loss (VML) injury demonstrated a progressive, secondary loss of neuromuscular junction (NMJ) innervation, supporting a role of NMJ dysregulation in chronic functional deficits. Terminal Schwann cells (tSCs) are known to be vital for the maintenance of NMJ structure and function, in addition to guiding repair and regeneration after injury. However, the tSC response to a traumatic muscle injury such as VML is not known. Thus, a study was conducted to investigate the effect of VML on tSC morphological characteristics and neurotrophic signaling proteins in adult male Lewis rats that underwent VML injury to the tibialis anterior muscle using a temporal design with outcome assessments at 3, 7, 14, 21, and 48 days post-injury. The following salient observations were made; first, although there is a loss of innervation over time, the number of tSCs per NMJ increases, significantly so at 48 days post-injury compared to control. The degree of NMJ fragmentation was positively correlated with tSC number after injury. Moreover, neurotrophic factors such as NRG1 and BDNF are elevated after injury through at least 48 days. These results were unanticipated and in contrast to neurodegenerative disease models, in which there is a reduction in tSC number that precedes denervation. However, we found that while there are more tSCs per NMJ after injury, they cover a significantly smaller percent of the post-synaptic endplate area compared to control. These findings support a sustained increase in neurotrophic activity and tSC number after VML, which is a maladaptive response occurring in parallel to other aspects of the VML injury, such as over-accumulation of collagen and aberrant inflammatory signaling.

Keywords: Denervation; Musculoskeletal trauma; Neuromuscular junction; Neurotrophic factors.

Figure 1.

A) Percentage of NMJs considered fully innervated compared to partially or fully denervated (p<0.001). Roughly 255 individual NMJs were analyzed per timepoint; partial denervation was determined at the discretion of the blinded observer and counted as denervated. B) Example of a fully innervated NMJ with clear overlap of pre- and post-synaptic components, and a partially denervated NMJ. C) Counts of AchR objects per NMJ as determined by a blinded observer (p<0.001). Individual dots represent each NMJ analyzed, note object numbers are discrete and overlap is present. D) AchR objects per NMJ stratified by innervation status (p<0.001). E) Examples of AchR object counts; top row has few objects (range 1–3) while the bottom row shows NMJs fragmented into numerous objects (>4). All image evaluation was conducted by a blinded reviewer. Data analyzed by one- and two-way ANOVA; significantly different than * control (C), § 3, ¥ 7, † 14, or ‡ 21 days post-VML. Scale bar is 10μm.

Figure 2.

A) While no single NMJ can be representative of a whole group, examples from each group are shown. Cell bodies were identified by S100 (purple) and nuclei (DAPI, blue) overlap with consideration of their position over the post-synapse (α-BTX, red). B) The number of tSCs per NMJ was hand counted by a blinded observer. Individual dots (discrete numbers) represent each NMJ analyzed, ~255 across 8 tissues per timepoint, 1,527 in total (p<0.001). C) The number of tSCs per NMJ stratified by innervation status (p=0.016). D) Frequency of NMJs with at least one tSC sprout, as determined by a blinded observer (p<0.001). E) Frequency of tSC sprouts with an associated axon present (p<0.001). Data analyzed by one-way ANOVA with Tukey’s HSD post-hoc or mixed-effects model with Šidák’s correction for multiple comparisons; significantly different than * control (C), § 3, ¥ 7, † 14, or ‡ 21 days post-VML. Scale bar is 10μm.

Figure 3.

A) Plot of each NMJ analyzed by its tSC number and AchR object number. Best-fit lines for each group are overlaid, the slope and Pearson correlation coefficient of each as follows: control (m=0.009, r=0.015, p=0.819); day 3 (m=0.107, r=0.143, p=0.055); day 7 (m=0.037, r=0.169, p=0.006); day 14 (m=0.211, r=0.271, p<0.001); day 21 (m=0.095, r=0.202, p<0.001); day 48 (m=0.237, r=0.573, p<0.001). The same data stratified by B) innervated and C) denervated NMJs. D) Example of partially denervated control NMJs (top row) and a partially denervated NMJ 48 days post-VML (bottom row). Notably, denervated NMJs from control muscle often displayed an expected absent S100 staining, while denervated NMJs 48 days post-injury often had numerous tSCs and extensive fragmentation. Scale bars are 10μm.

Figure 4.

A) Western blot analyses for skeletal muscle NRG1 protein content, data represented as value normalized to control (p<0.001). Dotted line is the control average of 1.0. B) Representative chemiluminescent immunoblot image of the NRG1 bands at each timepoint, and corresponding stain-free blot used to quantify the total lane protein. Bands for NRG1 were identified at both 26 and 50 kDa. In each lane, intensity of both bands was quantified and summed for total intensity quantification and then normalized by the respective total lane protein. C) Protein content of BDNF as determined by ELISA (p<0.001). Values are normalized to total protein content of the tissue. Each dot represents an individual muscle. All data analyzed by one-way ANOVA with Tukey’s HSD post-hoc; significantly different than * control (C) or ‡ 21 days post-VML.

Figure 5.

Percent area of tSC A) coverage (p=0.004) and B) extension (p=0.014), quantified by co-localization of S100 and α-BTX staining. Each dot represents a single NMJ; each color represents a single animal. Data analyzed by unpaired t-test; significantly different than * control (C). C) Example of tSC coverage area (white) and extension area (blue). Images were cropped as close to the NMJ as possible without removing extended tSCs, and a custom script was created in Nikon Elements General Analysis 3 to quantify co-localization. Pixel threshold values were manually set for each NMJ by a blinded observer. Coverage area is positive for both S100 and α-BTX, whereas extended area is only positive for S100. Scale bar is 10μm.