Early vulnerability to ischemia/reperfusion injury in motor terminals innervating fast muscles of SOD1-G93A mice

Abstract

In mouse models of familial amyotrophic lateral sclerosis (fALS), motor neurons are especially vulnerable to oxidative stresses in vitro. To determine whether this increased vulnerability also extends to motor nerve terminals in vivo, we assayed the effect of tourniquet-induced ischemia/reperfusion (I/R) injury on motor terminals innervating fast and slow hindlimb muscles in male G93A-SOD1 mice and their wild-type littermates. These mice also expressed yellow fluorescent protein (YFP) in motor neurons. We report that in SOD1-G93A/YFP mice the motor terminals innervating two predominantly fast muscles, extensor digitorum longus (EDL) and plantaris, were more vulnerable to I/R injury than motor terminals innervating the predominantly slow soleus muscle. The mean duration of EDL ischemia required to produce a 50% reduction in endplate innervation in SOD1-G93A/YFP mice was 26 min, compared to 45 min in YFP-only mice. The post-I/R destruction of EDL terminals in SOD1-G93A mice was rapid (<2 h) and was not duplicated by cutting the sciatic nerve at the tourniquet site. The increased sensitivity to I/R injury was evident in EDL muscles of SOD1-G93A/YFP mice as young as 31 days, well before the onset of motor neuron death at approximately 90 days. This early vulnerability to I/R injury may correlate with the finding (confirmed here) that in fALS mice motor nerve terminals innervating fast hindlimb muscles degenerate before those innervating slow muscles, at ages that precede motor neuron death. Early vulnerability of fast motor terminals to I/R injury thus may signal, and possibly contribute to, early events involved in motor neuron death.

Fig. 1

An ischemia/reperfusion (I/R) stress denervates most endplates in the EDL of SOD1-G93A/YFP (B,D), but not YFP-only mice (A,C). One hindlimb in each of 2 male mice (P33 YFP-only; P31 SOD1-G93A/YFP) was subjected to 30 min of tourniquet-induced ischemia followed by 6 hr of reperfusion in vivo. The non-stressed contralateral hindlimb served as a control. The upper, low-power micrographs (Aa-b, Ba-b) are digital overlays of the fluorescence of YFP (green, delineating motor axons and terminals) and Alexa 594-labelled α-BgTx (red, marking endplate ACh receptors) in stressed and non-stressed EDLs. Innervated endplates are yellow-orange; denervated endplates are red. Middle and lower rows show at higher magnification the separate α-BgTx (Ac-d, Bc-d) and YFP (Ae-f, Be-f) images corresponding to the regions indicated by the dotted boxes in Aa-b and Ba-b. Endplates marked by α-BgTx staining were outlined (dotted lines) and the motor terminal innervating each endplate was outlined in the YFP image, to determine endplate occupancy (see Materials and Methods). Histograms of endplate occupancy in Ca,b and Da,b show results pooled from EDLs of all analyzed YFP-only and SOD1-G93A/YFP mice (P31–P101) subjected to 30 min ischemia followed by 6–24 hr reperfusion. Each histogram is based on measurements from 152–291 terminals in 5–8 mice. Analysis of cumulative histograms compiled from these data (Cc, Dc) indicated that the I/R-induced reductions in endplate occupancy were significant for both SOD1-G93A/YFP and YFP-only mice (p<0.01 or better, Mann-Whitney 2-tailed test; Kruskal-Wallis test). We found no significant differences in vulnerability to I/R stress among endplates associated with the 4 different heads of the EDL muscle, perhaps because EDL’s fiber type composition is quite uniform (98% type IIb or type IIx).

Fig. 2

In SOD1-G93A/YFP mice, the increased vulnerability of EDL motor terminals to I/R injury begins within the first month; soleus (SOL) motor terminals are more resistant to this stress. Graphs plot mean endplate occupancy as a function of age for YFP-only and SOD1-G93A/YFP EDL (A) and soleus (B) muscles. Filled circles indicate the post-I/R limb; open circles indicate the non-stressed (control) limb. Mean endplate occupancy decreased slowly with age in non-stressed EDLs of SOD1-G93A/YFP mice (dashed line, slope = −0.004 ± 0.0007, significantly different from zero, linear trend test). In these mice the I/R stress greatly decreased EDL occupancy beginning at ~P30; the exponential fit to these points (solid line) had a time constant of 9.8 days. In contrast, in SOD1-G93A/YFP soleus and in all YFP-only muscles, the slopes of fitted linear regression lines (continuous for stressed, dashed for non-stressed) were not significantly different from zero, indicating maintenance of innervation throughout the sampled age range. In all cases the ischemic stress lasted 30 min; reperfusion times varied from 6–24 hr. Each symbol indicates the mean endplate occupancy computed from 30–80 endplates in the indicated muscle from an individual mouse. Data came from 11 SOD1-G93A/YFP and 16 YFP-only mice aged 24–121 days.

Fig. 3

Effect of ischemia duration on endplate occupancy in EDL (A) and soleus (B) of SOD1-G93A/YFP and YFP-only mice following I/R. Each data set was fitted with a sigmoidal equation (see Materials and Methods) to calculate the ischemia duration at which mean endplate occupancy would decrease to 50% of maximal (ID₅₀). ID₅₀ values (± SEM) for EDL were 45 ± 0.5 min in YFP-only vs. 26 ± 1 in SOD1-G93A/YFP; comparable values for soleus were 44 ± 0.4 and 48 ± 1. All of the following differences were significant at p<0.05 or better (t-test): SOD1-G93A/YFP vs. YFP-only for both EDL and soleus, and YFP-only EDL vs. YFP-only soleus. All curves used maximal and minimal occupancy values of 0.88 and 0, respectively, and a slope of -8.0; the R² for all fits exceeded 0.98. Values for 0 min ischemia came from non-stressed muscles. Each point was derived from measurements in 2–14 muscles, except for the 55 min point from a single animal in which the tourniquet was removed after 45 min ischemia, but limb blood flow did not return until 10 min later. More mice were used for ischemia durations that straddled the transitions between maintained innervation and complete denervation, where variability was greatest. In some cases the SEM was smaller than the plotting symbol.

Fig. 4

Axotomy does not duplicate the effects of I/R on EDL endplate occupancy in a SOD1-G93A/YFP mouse. Cumulative histograms of endplate occupancies were measured in EDL muscles of a P85 mouse in which one hindlimb was subjected to 30 min ischemia followed by 6 hr reperfusion (I/R, filled circles), and the other hindlimb underwent axotomy at the sciatic notch (open circles), which is near the upper hindlimb region where tourniquets were applied. 30–37 endplates were analyzed in each muscle. Comparable measurements for soleus in both axotomized and post-I/R limbs overlapped with those for post-axotomy EDL (not shown).

Fig. 5

SOD1-G93A/YFP EDL motor terminal degenerates rapidly following post-ischemic reperfusion/reoxygenation. Fluorescent micrographs of motor terminals from P76 YFP-only (A) and P69 SOD1-G93A/YFP (B) EDLs that were subjected to 30 min ischemia in vivo. After tourniquet removal, muscles were dissected and placed in physiological saline equilibrated with 95% O₂/5% CO₂ at 32–34 °C. Images taken 35 min and 100 min following conclusion of the ischemic stress show that the YFP-only terminal remained intact (Aa,b), whereas the SOD1-G93A/YFP terminal broke apart (Ba,b). Dotted lines in Ba outline the preterminal axon at 35 min; this region had disappeared by 100 min (Bb). Arrowheads in B indicate a section of the terminal that was continuous at 35 min, but became fragmented by 100 min. Bc, filled circles show the time course of fluorescence decay in the SOD1-G93A/YFP terminal; open circles show that the fluorescence of a neighboring background region did not change. The plot begins at 30 min, the time required to dissect and mount the muscle. Note that this graph measures total YFP fluorescence (arbitrary units); YFP measurements made for all other figures included only those sections of the motor terminal that remained connected to the preterminal axon. Similar results were obtained in 2 additional post-I/R SOD1-G93A/YFP terminals and 3 additional post-I/R YFP-only terminals.

Fig. 6

EDL endplates in a SOD1-G93A/YFP mouse are reinnervated 10 days following a 30 min I/R stress. Micrographs are overlays of α-BgTx-labelled endplates (red) and YFP-labelled motor axons (green) from non-stressed (left) and stressed (right) hindlimb. Ischemia was applied at P65. Arrowheads indicate terminal sprouting that extended beyond labelled endplates.