Content uploaded by Albina Jablonka-Shariff
Author content
All content in this area was uploaded by Albina Jablonka-Shariff on Mar 24, 2021
Content may be subject to copyright.
BASIC SCIENCE RESEARCH ARTICLE
What is Normal? Neuromuscular junction reinnervation
after nerve injury
Bianca Vannucci BA
1
| Katherine B. Santosa MD, MS
2
| Alexandra M. Keane BA
1
|
Albina Jablonka-Shariff PhD
2
| Chuieng-Yi Lu MD
2
| Ying Yan MD, PhD
3
|
Matthew MacEwan MD, PhD
3
| Alison K. Snyder-Warwick MD, FACS
2
1
Division of Plastic Surgery, Department of
Surgery, Washington University School of
Medicine, St Louis, Missouri
2
Division of Plastic Surgery, Department of
Surgery, Washington University School of
Medicine, St Louis, Missouri
3
Department of Neurosurgery, Washington
University School of Medicine, St Louis,
Missouri
Correspondence
Alison K. Snyder-Warwick, MD, FACS,
Division of Plastic Surgery, Washington
University School of Medicine, 660 South
Euclid Avenue, Campus Box 8238, St Louis,
MO 63110.
Email: snydera@wustl.edu
Funding information
National Institute of Neurological Disorders
and Stroke of the National Institutes of Health,
Grant/Award Number: (F32NS098561 to
K.B.S. and K08NS096232 to A.K.S.W.)
Abstract
Introduction: In this study we present a reproducible technique to assess motor
recovery after nerve injury via neuromuscular junction (NMJ) immunostaining and
electrodiagnostic testing.
Methods: Wild-type mice underwent sciatic nerve transection with repair. Hindlimb
muscles were collected for microscopy up to 30 weeks after injury. Immunostaining
was used to assess axons (NF200), Schwann cells (S100), and motor endplates
(α-bungarotoxin). Compound motor action potential (CMAP) amplitude was used to
assess tibialis anterior (TA) function.
Results: One week after injury, nearly all (98.0%) endplates were denervated. At
8 weeks, endplates were either partially (28.3%) or fully (71.7%) reinnervated. At
16 weeks, NMJ reinnervation reached 87.3%. CMAP amplitude was 83% of naive mice
at 16 weeks and correlated with percentage of fully reinnervated NMJs. Morphological
differences were noted between injured and noninjured NMJs.
Discussion: We present a reproducible method for evaluating NMJ reinnervation.
Electrodiagnostic data summarize NMJ recovery. Characterization of wild-type rein-
nervation provides important data for consideration in experimental design and
interpretation.
KEYWORDS
motor endplate, motor recovery, nerve injury, neuromuscular junction, reinnervation
1|INTRODUCTION
Peripheral nerve injuries are debilitating conditions that affect more
than 20 million Americans.
1
Despite our current understanding of the
mechanisms of nerve regeneration and advances made in nerve recon-
struction, a majority of patients with nerve injury do not regain satisfac-
tory function after surgical intervention.
1
The extent to which motor
function can be regained after nerve repair depends on several factors,
including injury type and location, denervation duration, and patient
age.
2,3
If reinnervation does not occur within 12 to 18 months, neuro-
muscular junctions (NMJs) degenerate, precluding reinnervation.
1,4
Con-
sidering the suboptimal outcomes for many patients with nerve injuries,
more rigorous investigation is needed to elucidate the mechanisms
promoting motor recovery after injury, including evaluation at the end-
target muscle.
Abbreviations: AChR, acetylcholine receptor; CMAP, compound motor action potential;
DAPI, 40,60-diamidino-phenylindole; EDL, extensor digitorum longus; EHL, extensor hallucis
longus; NF, neurofilament; NMJ, neuromuscular junction; PBS, phosphate-buffered saline;
TA, tibialis anterior; tSC, terminal Schwann cell; WT, wild-type; α-BTX, alpha-bungarotoxin.
B.V. and K.B.S. contributed equally to this work.
Portions of the content of this study have been presented at the annual meeting of the
American Society of Peripheral Nerve on January 13, 2018, in Phoenix, Arizona.
Received: 9 December 2018 Revised: 31 July 2019 Accepted: 4 August 2019
DOI: 10.1002/mus.26654
604 © 2019 Wiley Periodicals, Inc. Muscle & Nerve. 2019;60:604–612.wileyonlinelibrary.com/journal/mus
Functional motor recovery after peripheral nerve injury depends on
two key factors: (1) nerve regeneration at the injury site; and (2) NMJ
reinnervation within target muscle. The NMJ represents the interface
of nerve and muscle and is composed of three main structures: the
nerve terminal, which contains acetylcholine vesicles to be released
across the synaptic cleft; the motor endplate covered in acetylcholine
receptors (AChRs); and 3 to 5 nonmyelinating terminal Schwann cells
(tSCs), or perisynaptic Schwann cells, that encase the nerve terminal
and synapse.
5,6
The NMJ is not a static structure as it undergoes con-
tinuous remodeling throughout the lifetime of the animal and after
injury.
6,7
Whereas nerve regeneration at the injury site has been the
primary subject of many investigations, the process of NMJ rein-
nervation after nerve injury is still incompletely characterized.
8,9
Transgenic mice are used for various experimental applications. In
experimental conditions, however, appropriate controls may not be
available for comparison. For example, transgenic models may differ
in background strain. As a result, wild-type (WT) mice may be needed
for experimental controls. Morphological differences may exist in
structures in transgenic mice compared with the same structures visu-
alized with immunostaining in WT mice. Transgenic mice may have
more intense fluorescence of structures of interest compared with
immunostaining of the same structures in WT mice. Because of less
intense fluorescence after immunostaining in WT mice, however, fine
details of structures may be better seen. Each model may offer advan-
tages and disadvantages, depending on the planned analysis. As such,
the two should not be compared as they differ at baseline. In this
study, we characterize NMJ reinnervation and motor recovery in WT
mice after injury.
The primary goals of this study were to comprehensively evaluate
NMJ reinnervation patterns and NMJ morphology at different time-
points after sciatic nerve injury in a murine model.
2|METHODS
2.1 |Mice
All surgical procedures were performed in 3-month-old adult WT
C57BL/6J mice. Sexes were mixed across all experimental groups
(t = 0, 1, 2, 3, 4, 8, 12, 16, and 30 weeks after nerve injury). NMJ mor-
phology was also compared with 3-month-old transgenic mice
(S100-GFP and Thy1-YFP; refer to Figure S1 online). Pre- and postop-
erative animals were housed in a central animal housing facility and
were maintained in strict accordance with the National Institutes of
Health guidelines and protocols approved by the institutional animal
care and use committee at the Washington University School of
Medicine.
2.2 |Sciatic nerve injury and repair
The sciatic nerve of the right hindlimb was exposed via a muscle-
splitting technique. The sciatic nerve was sharply transected 3 mm
proximal to the trifurcation. Immediately after nerve transection, the
nerve was repaired using microsuture and fibrin glue (Tisseel, Baxter,
Deerfield, Illinois). After coaptation, the muscle and skin were closed
appropriately. Animals were recovered on a heating pad, anesthesia
was reversed with atipamezole hydrochloride (1 mg/kg), and animals
were returned to the central animal housing facility within 12 hours
for close monitoring and postoperative care.
2.3 |Muscle harvest
At the assigned time-point (t = 0, 1, 2, 3, 4, 8, 12, 16, and 30 weeks
after nerve injury), animals were anesthetized, and an incision was
made over the anterior right leg extending from the dorsal foot to the
knee. After identification, the distal tendon of the tibialis anterior
(TA) muscle was transected to expose the extensor hallucis longus
(EHL) tendon, immediately posterior to the TA tendon. The TA was
then retracted and cut at the insertion at the knee to provide better
exposure of the EHL and extensor digitorum longus (EDL) tendons
and muscle bellies. The EHL was then carefully dissected proximally
to avoid tearing where the branch of the peroneal nerve inserts. The
EDL tendons were transected distally below the ankle and proximally
at the insertion at the knee. After dissection, the EHL and EDL mus-
cles were immediately placed in cold 0.1-mol/L phosphate-buffered
saline (PBS, pH 7.4) solution to remove any debris or hair. Animals
underwent cervical dislocation under deep anesthesia.
2.4 |Whole mount immunofluorescent NMJ staining
After harvest, EHL and EDL muscles were placed in 2% paraformalde-
hyde solution (Electron Microscopy Science, Hatfield, Pennsylvania) in
PBS for 15 minutes for EHL and 30 minutes for EDL muscles at room
temperature. After fixation of whole muscles, the four component
tendons of the EDL were dissected out in cold PBS solution. Immuno-
fluorescent staining was performed as described elsewhere.
10
All mus-
cles were washed in PBS and incubated in blocking buffer (5% normal
goat serum, 2% Triton X-100, 5% bovine serum albumin in PBS) for
1 hour at room temperature. Tissues were then stained with primary
antibodies, rabbit anti–neurofilament 200 (NF200; 1:500; Millipore
Sigma, St Louis, Missouri) or rabbit anti-S100b (1:1000; DAKO North
America, Via Real, Carpinteria, California), overnight at 4C. After rins-
ing in washing buffer (PBS with 0.2% Triton X-100) for 45 minutes,
muscles were incubated with goat anti-rabbit immunoglubulin
G–Alexa Fluor 488 (1:1000; Invitrogen-Molecular Probes, Carlsbad,
California) for 1 hour at room temperature. Muscles were further rinsed
in washing buffer for 45 minutes before incubating with Alexa Fluor
555–α-bungatotoxin (α-BTX; 1:1000; Molecular Probes, Eugene,
Oregon), to label AChRs, for 1 hour at room temperature. Muscles were
mounted using mounting medium with 40,60-diamidino-phenylindole
(DAPI; VectaShield, Vector Laboratories, Burlingame, California).
2.5 |NMJ imaging
Whole mount muscles were imaged using a fluorescent microscope
(AxioImager M2, Zeiss, Thornwood, New York). Sequential capture
was used to separate the green and red channels in order to prevent
VANNUCCI ET AL.605
crosstalk between dyes. Z-serial images were collected with 20×,40×,
and 63×objectives to allow analysis of three-dimensional structures.
Images were viewed and analyzed using NIH ImageJ (http://rsb.info.
nih.gov/iJ/). Figures were prepared using GraphPad Prism 8 (GraphPad
Software, La Jolla, California), Adobe Photoshop CC 2015, and Adobe
Illustrator CC 2015 (Adobe Systems, San Jose, California).
2.6 |NMJ evaluation and reinnervation
quantification
NMJ reinnervation in EDL and EHL muscles was evaluated at 0, 1, 2, 3,
4, 8, 12, and 16 weeks after sciatic nerve transection with immediate
repair. Endplate reinnervation was visualized and quantified with a com-
pound confocal microscope by multiple, independent, blinded reviewers.
Whenever possible, muscles were imaged immediately after mountingto
minimize potential for staining deterioration over time. Each slide was
assessed for staining quality and adequate antibody penetration. Rein-
nervation quantification was based exclusively on the most superficial
layer of the muscle, as described elsewhere.
10,11
NMJ reinnervation was
then defined and graded based on colocalization of α-BTX and NF200
staining. For each NMJ, motor endplates (α-BTX staining) lying en face
were visualized and overlap with NF200 staining was assessed. Rein-
nervation assessmentrequired multiple sequential images for each NMJ.
This methodensured that all motor endplates were evaluated. After each
endplate was assessed, a reference image was chosen for record keep-
ing, and a running endplate tally was recorded, classifying each endplate
as (1) denervated, (2) partially innervated, or (3) fully innervated
(Figure S2 online). Denervation was defined as the absence of α-BTX
and NF200 colocalization, despite insurance of adequate staining
(Figure S2A,A0online). Partial reinnervation was defined as partial and/or
FIGURE 1 NMJ reinnervation increases and axonal branch pattern morphology reconstitutes with time after motor nerve injury. (A-G)
Confocal images of adult WT EDL muscles at 1, 2, 3, 4, 8, 12, and 16 weeks after sciatic nerve transection and immediate repair. (E-G) Inset
images are representative higher magnification images of reinnervated NMJs at 8, 12, and 16 weeks after injury. (H) Quantification of NMJ
reinnervation is summarized (n = 3 mice per time-point, except n = 2 mice at 16 weeks; average 266 of NMJs evaluated per time-point).
Reinnervation classified as full, partial, or denervated. Graph shows the proportions of NMJs with the three categories of innervation at the
various experimental time-points. BTX, α-bungarotoxin (for acetylcholine receptors, red); DAPI, 40,60-diamidino-phenylindole (nuclear staining,
blue); EDL, extensor digitorum longus; NF200 Ab, anti-neurofilament antibody (for axons, green); NMJ, neuromuscular junction; WT, wild-type.
Scale bar = 20 μm [Color figure can be viewed at wileyonlinelibrary.com]
606 VANNUCCI ET AL.
interrupted colocalization of α-BTX and NF200 stains (Figure S2B,B0
online). Full reinnervation was defined as the complete and continuous
colocalization of α-BTX and NF200 stains in a superficially located
endplate (Figure S2C,C0online).
2.7 |Evaluation of polyinnervation and axonal
sprouting
Polyaxonal innervation is characteristic of NMJ development, but may
occur after reinnervation. Axonal sprouting is a characteristic feature
of nerve regeneration after injury. Rates of polyaxonal reinnervation
and axonal sprouting were assessed in naive EDL and EHL muscles
and in EDL and EHL muscles at 2, 3, 4, 8, 12, and 16 weeks after sci-
atic nerve transection with immediate repair. We defined polyaxonal
reinnervation as the presence of more than one axon reaching the
endplate and axonal sprouting as an axon branch extending away
from an endplate.
2.8 |Evaluation of motor endplate fragmentation
Motor endplate morphology was assessed in EDL and EHL muscles at
2, 3, 8, 12, 16, and 30 weeks after sciatic nerve transection with
immediate repair, and in contralateral uninjured muscles. Endplates
stained with α-BTX were imaged by blinded reviewers. Each endplate
lying en face was analyzed at high magnification with image editing
software (Windows Photo Viewer, Microsoft, Redmond, Washington)
and assessed for the presence of gray-black background interposed
between islands of α-BTX staining within the same endplate. The
reviewers then outlined these fragments with a bright color to facili-
tate counting. The numbers of fragments per endplate in each group
were then tallied.
2.9 |Electrodiagnostic testing
Functional capacity of naive TA muscles and functional recovery of TA
muscles at 2, 3, 4, 6, and 16 weeks after sciatic nerve transection and
immediate repair were assessed using compound motor action potential
(CMAP) amplitudes. Measurement of TA muscle activation (in millivolts),
evoked by electrical stimulationof the sciatic nerve at50 Hz, was utilized
to assess the functional capacity. The experimental technique and data
acquisition were performed as described elsewhere.
12,13
2.10 |Statistical analysis
Data are reported as mean ± standard error of the mean. Quantifica-
tions were performed in a blinded fashion from at least three
FIGURE 2 Polyaxonal innervation and axonal sprouting increase in frequency with time after motor nerve injury in adult WT EDL and EHL
muscles. Representative confocal images of (A) normal monoaxonal innervation (asterisk), (B) polyaxonal innervation (arrowheads), and (C) axonal
sprouting (arrow) in EDL muscles. (D) Quantification of axonal patterns present in naive WT muscles (n = 3 mice), and in WT muscles at 2, 3, 4, 8,
12, and 16 weeks after nerve injury (n = 3 mice per time-point, except n = 2 mice at 16 weeks; average 140 NMJs evaluated per time-point) are
shown. Graph summarizes the proportions of NMJs with each of the three axonal patterns. Scale bar = 20 μm. BTX, α-bungarotoxin (for
acetylcholine receptors, red); DAPI, 40,60-diamidino-phenylindole (nuclear staining, blue); EDL, extensor digitorum longus; EHL, extensor hallucis
longus; NF200 Ab, anti-neurofilament antibody (for axons, green); NMJ, neuromuscular junction; WT, wild-type [Color figure can be viewed at
wileyonlinelibrary.com]
VANNUCCI ET AL.607
experimental groups, unless otherwise noted. Statistical analyses were
performed in either Microsoft Excel or GraphPad Prism 8. Two-tailed,
unpaired Student ttests (parametric data) or Mann–Whitney Utests
(nonparametric data) were utilized to assess statistical differences
between data sets. Correlation was calculated using the Pearson cor-
relation coefficient. Statistical significance was set at P< .05.
3|RESULTS
A total of 31 mice were used for analysis, including 19 females and
12 males. A total of 1513 NMJs were evaluated for reinnervation,
1350 NMJs for motor endplate fragmentation, and 978 NMJs for
polyaxonal innervation and sprouting.
3.1 |Reinnervation
Reinnervation results are displayed in Figure 1. One week after sciatic
nerve transection, 98% of EHL and EDL muscles were denervated
(Figure 1A,H). By 2 weeks, the majority of end plates were at least partially
reinnervated (Figure 1B,H). In addition, early evidence of axonal branch
pattern reorganization was present (Figure 1B). By 8 weeks, all NMJs
were partially (28.3%) or fully (71.7%) reinnervated, and the normal tree-
like axonal branching pattern was reconstituted (Figure 1E,H). Beyond
the 8-week time-point, the percentage of endplates classified as fully
reinnervated continued to increase (Figure 1F-H), and, by 16 weeks, the
vast majority (87.3%) of endplates were fully reinnervated (Figure 1G,H).
Although the percentage of reinnervated NMJs increased over time,
we noticed changes in regenerating axon configuration, including poly-
axonal reinnervation and nerve terminal sprouting from the endplate
(Figure 2). The percentage of NMJs with axonal sprouting increased over
time, with no sprouting observed 2 weeks after injury compared with a
maximum of 39% sprouting observed 16 weeks after injury (Figure 2D).
Polyaxonal reinnervation was less prevalent than axonal sprouting—a
maximumof~10%ofendplatesexhibitedpolyaxonalreinnervation
across all time-points (Figure 2D). Polyaxonal innervation and/or axonal
sproutingwere rarely noted in the absence of nerve injury.
3.2 |Motor endplate fragmentation evident despite
NMJ reinnervation
Although reinnervation rates increased steadily after injury, rein-
nervation did not represent a return to naive axon-endplate interactions.
FIGURE 3 Motor endplate fragmentation increases significantly with time after motor nerve injury compared with controls.
(A) Representative confocal image of a naive WT EDL motor endplate showing the “normal”pretzel-like configuration, with 7 BTX fragments.
(B) Representative image of a “fragmented”WT EDL motor endplate 30 weeks after sciatic nerve transection and immediate repair, with 17 BTX
fragments. (C) Graph shows the average number of fragments per endplate in WT EDL muscles at 2, 3, 8, 12, 16, and 30 weeks after sciatic nerve
transection with immediate repair (injured side), compared with contralateral control muscles (uninjured side) (n = 2 mice per time-point, except
n = 1 mouse at 30 weeks; average 225 endplates evaluated per time-point). At 2 and 3 weeks after injury, endplate fragmentation does not differ
between the injured and uninjured groups. At 8 weeks and beyond, injured groups display significantly more endplate fragmentation than
uninjured controls. Note that average number of endplate fragments in controls remains stable over time (3 or 4 fragments). Scale bar = 20 μm.
****
P< .0001. Data are expressed as mean ± standard error of the mean. BTX, α-bungarotoxin (for acetylcholine receptors, gray); EDL, extensor
digitorum longus; WT, wild-type [Color figure can be viewed at wileyonlinelibrary.com]
608 VANNUCCI ET AL.
Differences in endplate morphology and axonal configuration were
apparent between injured and uninjured groups (Figure 3). After nerve
injury, motor endplates were increasingly fragmented over time
(Figure 3B), as opposed to the more typical “pretzel-like”conformation
(Figure 3A). At 2 and 3 weeks, both injured and uninjured conditionshad
an average of 3 or 4 fragments per endplate. At 8 weeks, we observed
6.8 ± 3.8 fragments per endplatein the injured condition, compared with
3.4 ± 2.2 in the uninjured group. By 12 weeks, the number of fragments
per endplate in the injured group remained significantly different,
more than twice those of the uninjured group. At 16 weeks, we observed
8.4 ± 4.7 fragments per endplate in the injured condition, significantly
more than the 3.1 ± 1.5 in the uninjured condition. At our final time-
point of 30 weeks, there were 12.1 ± 6.3 fragments per endplate in
the injured group, compared with 3.8 ± 3.4 in the uninjured group
(P< .0001). Notably, endplate fragmentation was not accompanied by
a loss of innervation, nor the loss of the one-to-one axon-to-endplate
relationship. Rather, we observed axons splitting into smaller branches,
thereby covering each fragment constituting an endplate.
3.3 |Functional recovery progresses over time
and correlates with full NMJ reinnervation
At 2 weeks after injury, TA CMAP amplitude was significantly reduced
to 21% of naive function (Figure 4). TA CMAP amplitude demonstrated
progressive functional recovery over 16 weeks postoperatively. By
4 weeks, CMAP amplitude measured 40% of naive function and, by
6 weeks, 54% of naive TA. At 16 weeks after injury, TA CMAP ampli-
tude reached 83% of naive function; however, TA function was still sig-
nificantly reduced compared with naive (P<.01).
We observed a strong correlation between the percentage of naive
CMAP amplitude and percentage of fully reinnervated NMJs after nerve
injury and repair, with a Pearson correlation coefficient of 0.835
(Figure 5). At 2 weeks after nerve injury and repair, 12.9% of NMJs were
fully reinnervated, and average CMAP amplitude was 21% that of naive
muscle. At 3 weeks, 28.5% of NMJs were fully reinnervated, and CMAP
amplitude was 23% of naive muscle. By 4 weeks, full reinnervation was
seen in 40.2% of NMJs, and average CMAP amplitude was 40% of naive.
By 16 weeks, 87.3% of NMJs were fully reinnervated, and average
CMAP amplitude was 83% of naive.
4|DISCUSSION
Investigation of nerve regeneration and recovery after injury is important
to better understand reinnervation and to identify innovative therapeu-
tic targets for clinical translation. Most scientific investigations employ
comparison of experimental conditions to controls; the mouse provides
an excellent model for nerve research, and WT mice are commonly used
as a control group. In this study, NMJ reinnervation was partially or fully
complete in all studied NMJs by 8 weeks after nerve repair, and full
FIGURE 4 TA muscle function steadily recovers with time after
motor nerve injury, but remains significantly less at 16 weeks
compared with naive mice. Evoked CMAP amplitudes were recorded
from naive WT TA muscles (n = 3 mice) and from WT TA muscles 2, 3,
4, 6, and 16 weeks after sciatic nerve transection with immediate
repair (n = 3 mice per time-point). CMAP amplitude in naive WT TA
measured 8.55 ± 0.76 mV. At 2 weeks after injury, TA CMAP
amplitude was 1.77 ± 2.08 mV (21% of naive function). By 3 weeks,
CMAP amplitude measured 1.98 ± 0.11 mV (23% of naive); by
4 weeks, 3.44 ± 0.63 mV (40% of naive); and, by 6 weeks,
4.63 ± 1.17 mV (54% of naive). At 16 weeks after injury, TA CMAP
amplitude reached 83% of naive function (7.13 ± 1.75 mV); however,
TA function was still significantly reduced compared with naive. Data
are expressed as mean ± standard error of the mean.
****
P< .0001;
**
P< .01. CMAP, compound motor action potential; TA, tibialis
anterior; WT, wild-type [Color figure can be viewed at
wileyonlinelibrary.com]
2 wks 3 wks 4 wks 16 wks
-20
0
20
40
60
80
100
120
Time after Nerve Injury
Percent
CMAP amplitude
Fully innervated
FIGURE 5 There is a strong relationship between the percentage
of fully reinnervated endplates and CMAP amplitude after nerve
injury. Line graph displays the percentage of fully reinnervated
endplates in EDL muscles (presented in Figure 1) and the percentage
CMAP amplitude recovery relative to naive TA muscle CMAP
amplitude (derived from Figure 4) at 2, 3, 4, and 16 weeks after sciatic
nerve transection and immediate repair. The data at each time-point
are highly correlated, with a Pearson coefficient of 0.835; P< .001.
Data are expressed as mean ± standard error of the mean. CMAP,
compound motor action potential; EDL, extensor digitorum longus;
TA, tibialis anterior; WT, wild-type [Color figure can be viewed at
wileyonlinelibrary.com]
VANNUCCI ET AL.609
reinnervation was present in 87% of NMJs by 16 weeks after repair.
Polyaxonal NMJ reinnervation was present throughout the study, axonal
sprouting was noted at higher rates at later time-points, and endplate
fragmentation increased with time. Interestingly, we found that func-
tional recovery, as measured by CMAP amplitude, correlated closely with
the percentage of fully, but not partially, reinnervated NMJs throughout
all assessed time-points. Functional recovery did not, however, achieve
that of naive muscle by 16 weeks after nerve injury andrepair. This study
provides an important reference for both morphological and functional
comparisonfor interpretation of experimental results.
4.1 |NMJ reinnervation rates
The reinnervation rates noted in this study are consistent with those
from previous studies. We found no NMJs remained denervated
beyond 8 weeks after nerve repair. Similarly, no remaining denervated
NMJs were reported in WT (C57BL/6J) mice at 8 weeks after facial
nerve cut and repair.
14
In the neck, sternomastoid reinnervation after
spinal accessory nerve cut and repair was reported to occur between
postoperative days 14 and 30,
15
and after 1 month for nerve cut and
no repair.
16
Sternomastoid reinnervation is slightly faster given the
proximity of nerve injury to the muscle. Reinnervation of the biceps
brachii was noted 3 weeks after nerve cut and repair in C57BL/6J
mice.
17
4.2 |Advantages of systematic assessment of NMJ
reinnervation
Closely assessing NMJ reinnervation histologically is a surrogate of
functional recovery after injury. Reinnervation is commonly reported
as an all-or-none phenomenon. In this study, reinnervation was specif-
ically characterized as the degree of colocalization of the nerve termi-
nal with motor endplates (AChRs). Our assessment method of NMJ
reinnervation has several advantages. First, it allows for visualization
of two major NMJ components: the nerve terminal and motor
endplates. Second, it is reproducible with simple, defined parameters.
Third, our protocol makes use of the EHL and EDL muscles as
opposed to thicker muscles, such as TA. Both the EHL and EDL mus-
cles are located in the anterior compartment of the lower leg, are fast-
twitch muscles,
18
and are innervated by the deep peroneal branch of
the sciatic nerve. In addition, both muscles are ideally suited for imag-
ing using whole mount confocal microscopy due to their thin muscle
profile and EDL's ability to be split into its four tendinous compo-
nents. Thus, the use of EHL and EDL muscles facilitates better
endplate visualization such that reinnervation can be comprehensively
and systematically assessed.
4.3 |Polyinnervation, axonal sprouting, and endplate
fragmentation occur with reinnervation
Despite a high percentage of NMJ reinnervation via immunohistochem-
istry by week 16 and similar functional recovery by CMAP in our study,
morphological NMJ differences persisted after injury. Polyinnervation
and axonal sprouting progressed with time from injury and repair. These
phenomena are frequently noted during NMJ development
19-21
and
early in reinnervation,
16,22-24
rather than at later time-points. Similar
changes, however, including polyinnervation and axonal sprouting, have
been reported in aged rats beyond 8 weeks after nerve crush.
25
The
presence of increasing motor endplate fragmentation was unexpected,
particularly in the setting of progressive NMJ reinnervation. Previous
studies from our lab and others have shown endplate fragmentation
occurs with aging.
10,26-28
In the present study, age did not contribute to
endplate fragment number in uninjured mice; the average number of
fragments in uninjured mice was 3 or 4 at all time-points, and all mice
were under 10 months of age. Motor endplates, however, constantly
turnover in homeostatic conditions.
15
After NMJ denervation, endplate
turnover rates accelerate due to synthesis of more rapidly degrading
AChRs in denervated muscle. In the sternomastoid muscle, motor
endplate half-lives decreased from 8 to 12 days to 3 days when nerves
were prevented from regenerating, but half-lives returned to baseline
levels after reinnervation.
15,29,30
Thus, after denervation, two AChR
populations exist—the original population that degrades more slowly and
the more rapidly degenerating population. Once several half-lives pass
after reinnervation, the more rapidly degrading AChRs become sequen-
tially fewer in number, and homeostatic turnover rates are restored.
29
The fragmented endplates noted in our study may reflect this shift in
endplate population to the more rapidly degrading AChRs. The reason
for persistence of this morphological difference as well as the functional
implications, however, are unknown. Importantly, fragmented endplates
have been shown to have no decline in synaptic transmission in aged
mice.
28
Any of these morphological changes could easily be interpreted
as abnormal in different experimental conditions or as a result of treat-
ment effects. Here, we see these morphological differences in WT mice
after nerve cut and repair.
4.4 |CMAP amplitude provides a functional
correlate of NMJ recovery
Functional TA muscle recovery increased throughout the study period,
but surprisingly, even at 16 weeks after nerve cut and repair, CMAP
amplitude had not achieved levels of naive mice. This timeline is impor-
tant when planning and interpreting experiments. English et al
31
described similar results of low evoked gastrocnemius muscle maximal
response amplitudes, at 10% to 30% of baseline, in C57BL/6J mice at
8 weeks after sciatic nerve cut and repair.
31
After tibial nerve cut and
repair in the NMRI mouse, in vitro evoked tetanic force in the soleus was
noted to be low (54% of the uninjured side) at 4 weeks postinjury, but
with nearly complete recovery (98% of the uninjured side) by 2 months
postinjury.
32
These differing results suggest variability in motor recovery,
which may be related to muscle fiber type. The TA is 59% type IIB fibers
and the gastrocnemius is 54% type IIB fibers in the C57BL/6J mouse,
compared withthe soleus composition of 37% typeI fibers and 38% type
IIA fibers.
33
Interestingly, we found that functional recovery of the TA
correlated with the percentage of fully reinnervated NMJs noted histo-
logically. Although nearly all endplates were occupied by nerve by
3 weeks after nerve injury and repair, functional recovery was poor at
610 VANNUCCI ET AL.
that time-point (23% of naive) and only correlated with full NMJ rein-
nervation rather than partial reinnervation. This observation highlights
the importance of fully restored synaptic transmission for overall muscle
function. These data also suggest that CMAP amplitude provides a sensi-
tive assessment of complete NMJ recovery. Measurement of functional
recovery in rodent nerve injury models can be challenging and fraught
with limitations. Challenges associated with commonly used functional
metrics,such as walking trackanalysis, include:(1) low sensitivity;(2) diffi-
culty discriminating between the effects of the specific nerve injury com-
pared with the general effects of surgery; (3) training time required for
animals to learn motor tasks; and (4) the potential need for expensive
and special equipment.
24,34
Our study supports the use of elec-
trodiagnostic testingas a functional correlate of NMJ recovery.
Although the use of specific transgenic mice facilitates visualiza-
tion and quantification of reinnervation, those mice are often ill-suited
to serve as controls for studies in which experimental mice are not of
the same transgenic line or background strain. For this reason, we
believe a comprehensive description of WT NMJ reinnervation offers
an important reference to the literature.
In conclusion, these data highlight the importance of complete
NMJ reinnervation for functional recovery and provide timelines for
expected reinnervation in a murine model. Interestingly, functional
recovery did not reach levels of uninjured mice, even at 16 weeks
postinjury, but CMAP amplitude correlated closely with the percent-
age of fully reinnervated NMJs. We have proposed a reproducible
method of reinnervation assessment and provided a reference for
reinnervation in WT mice.
CONFLICTS OF INTEREST
The authors declare no potential conflicts of interest.
ETHICAL PUBLICATION STATEMENT
We confirm that we have read the Journal's position on issues
involved in ethical publication and affirm that this report is consistent
with those guidelines.
ORCID
Alison K. Snyder-Warwick https://orcid.org/0000-0002-9882-4213
REFERENCES
1. Grinsell D, Keating CP. Peripheral nerve reconstruction after injury: a
review of clinical and experimental therapies. Biomed Res Int. 2014;
2014:698256.
2. Moore AM, Novak CB. Advances in nerve transfer surgery. J Hand
Ther. 2014;27:96-104; quiz 105.
3. Robinson LR. Traumatic injury to peripheral nerves. Muscle Nerve.
2000;23:863-873.
4. Boyd KU, Nimigan AS, Mackinnon SE. Nerve reconstruction in the
hand and upper extremity. Clin Plast Surg. 2011;38:643-660.
5. Bloch-Gallego E. Mechanisms controlling neuromuscular junction sta-
bility. Cell Mol Life Sci. 2015;72:1029-1043.
6. Wilson MH, Deschenes MR. The neuromuscular junction: anatomical
features and adaptations to various forms of increased, or decreased
neuromuscular activity. Int J Neurosci. 2005;115:803-828.
7. Deschenes MR. Motor unit and neuromuscular junction remodeling
with aging. Curr Aging Sci. 2011;4:209-220.
8. Cattin AL, Burden JJ, Van Emmenis L, et al. Macrophage-induced
blood vessels guide Schwann cell–mediated regeneration of periph-
eral nerves. Cell. 2015;162:1127-1139.
9. Cattin AL, Lloyd AC. The multicellular complexity of peripheral nerve
regeneration. Curr Opin Neurobiol. 2016;39:38-46.
10. Snyder-Warwick AK, Satoh A, Santosa KB, Imai SI, Jablonka-Shariff A.
Hypothalamic Sirt1 protects terminal Schwann cells and neuromuscu-
lar junctions from age-related morphological changes. Aging Cell.
2018;e12776.
11. Arbour D, Tremblay E, Martineau E, Julien JP, Robitaille R. Early and
persistent abnormal decoding by glial cells at the neuromuscular junc-
tion in an ALS model. J Neurosci. 2015;35:688-706.
12. Gamble P, Stephen M, MacEwan M, Ray WZ. Serial assessment of
functional recovery following nerve injury using implantable thin-
film wireless nerve stimulators. Muscle Nerve. 2016;54:1114-
1119.
13. MacEwan MR, Gamble P, Stephen M, Ray WZ. Therapeutic electrical
stimulation of injured peripheral nerve tissue using implantable thin-
film wireless nerve stimulators. J Neurosurg. 2018;1-10.
14. Kiryakova S, Sohnchen J, Grosheva M, et al. Recovery of whisking
function promoted by manual stimulation of the vibrissal muscles
after facial nerve injury requires insulin-like growth factor 1 (IGF-1).
Exp Neurol. 2010;222:226-234.
15. Salpeter MM, Cooper DL, Levitt-Gilmour T. Degradation rates of ace-
tylcholine receptors can be modified in the postjunctional plasma
membrane of the vertebrate neuromuscular junction. J Cell Biol.
1986;103:1399-1403.
16. Rich MM, Lichtman JW. In vivo visualization of pre- and postsynaptic
changes during synapse elimination in reinnervated mouse muscle.
J Neurosci. 1989;9:1781-1805.
17. Crews LL, Wigston DJ. The dependence of motoneurons on their tar-
get muscle during postnatal development of the mouse. J Neurosci.
1990;10:1643-1653.
18. Zimowska M, Kasprzycka P, Bocian K, et al. Inflammatory response
during slow- and fast-twitch muscle regeneration. Muscle Nerve.
2017;55:400-409.
19. Darabid H, Perez-Gonzalez AP, Robitaille R. Neuromuscular syn-
aptogenesis: coordinating partners with multiple functions. Nat Rev
Neurosci. 2014;15:703.
20. Lee YI, Li Y, Mikesh M, et al. Neuregulin1 displayed on motor axons
regulates terminal Schwann cell-mediated synapse elimination at
developing neuromuscular junctions. Proc Natl Acad Sci USA. 2016;
113:E479-E487.
21. Park SY, Jang SY, Shin YK, et al. The scaffolding protein,
Grb2-associated binder-1, in skeletal muscles and terminal Schwann
cells regulates postnatal neuromuscular synapse maturation. Exp Neu-
robiol. 2017;26:141-150.
22. Gorio A, Carmignoto G, Finesso M, Polato P, Nunzi MG. Muscle
reinnervation—II. Sprouting, synapse formation and repression. Neu-
roscience. 1983;8:403-416.
23. Gorio A, Marini P, Zanoni R. Muscle reinnervation—III. Motoneuron
sprouting capacity, enhancement by exogenous gangliosides. Neuro-
science. 1983;8:417-429.
24. Magill CK, Tong A, Kawamura D, et al. Reinnervation of the tibialis
anterior following sciatic nerve crush injury: a confocal microscopic
study in transgenic mice. Exp Neurol. 2007;207:64-74.
25. Kawabuchi M, Zhou CJ, Wang S, Nakamura K, Liu WT, Hirata K. The spa-
tiotemporal relationship among Schwann cells, axons and postsynaptic
VANNUCCI ET AL.611
acetylcholine receptor regions during muscle reinnervation in aged rats.
Anat Rec. 2001;264:183-202.
26. Balice-Gordon RJ. Age-related changes in neuromuscular innervation.
Muscle Nerve. 1997;5(suppl):S83-S87.
27. Pannerec A, Springer M, Migliavacca E, et al. A robust neuromuscular
system protects rat and human skeletal muscle from sarcopenia.
Aging. 2016;8:712-729.
28. Willadt S, Nash M, Slater CR. Age-related fragmentation of the motor
endplate is not associated with impaired neuromuscular transmission
in the mouse diaphragm. Sci Rep. 2016;6:24849.
29. Shyng SL, Salpeter MM. Effect of reinnervation on the degradation
rate of junctional acetylcholine receptors synthesized in denervated
skeletal muscles. J Neurosci. 1990;10:3905-3915.
30. Xu R, Salpeter MM. Acetylcholine receptors in innervated muscles of
dystrophic mdx mice degrade as after denervation. J Neurol. 1997;17:
8194-8200.
31. English AW, Liu K, Nicolini JM, Mulligan AM, Ye K. Small-molecule
trkB agonists promote axon regeneration in cut peripheral nerves.
Proc Natl Acad Sci USA. 2013;110:16217-16222.
32. Badke A, Irintchev AP, Wernig A. Maturation of transmission in
reinnervated mouse soleus muscle. Muscle Nerve. 1989;12:580-586.
33. Augusto V, Padovani CR, Campos GER. Skeletal muscle fiber types in
C57BL6J mice. Braz J Morphol Sci. 2004;21:89-94.
34. Nichols CM, Myckatyn TM, Rickman SR, Fox IK, Hadlock T,
Mackinnon SE. Choosing the correct functional assay: a comprehen-
sive assessment of functional tests in the rat. Behav Brain Res. 2005;
163:143-158.
SUPPORTING INFORMATION
Additional supporting information may be found online in the
Supporting Information section at the end of this article.
How to cite this article: Vannucci B, Santosa KB, Keane AM,
et al. What is Normal? Neuromuscular junction reinnervation
after nerve injury. Muscle Nerve. 2019;60:604–612. https://
doi.org/10.1002/mus.26654
612 VANNUCCI ET AL.
