Impact of prolonged sepsis on neural and muscular components of muscle contractions in a mouse model

Abstract

Background: Prolonged critically ill patients frequently develop debilitating muscle weakness that can affect both peripheral nerves and skeletal muscle. In-depth knowledge on the temporal contribution of neural and muscular components to muscle weakness is currently incomplete.

Methods: We used a fluid-resuscitated, antibiotic-treated, parenterally fed murine model of prolonged (5 days) sepsis-induced muscle weakness (caecal ligation and puncture; n = 148). Electromyography (EMG) measurements were performed in two nerve-muscle complexes, combined with histological analysis of neuromuscular junction denervation, axonal degeneration, and demyelination. In situ muscle force measurements distinguished neural from muscular contribution to reduced muscle force generation. In myofibres, imaging and biomechanics were combined to evaluate myofibrillar contractile calcium sensitivity, sarcomere organization, and fibre structural properties. Myosin and actin protein content and titin gene expression were measured on the whole muscle.

Results: Five days of sepsis resulted in increased EMG latency (P = 0.006) and decreased EMG amplitude (P < 0.0001) in the dorsal caudal tail nerve-tail complex, whereas only EMG amplitude was affected in the sciatic nerve-gastrocnemius muscle complex (P < 0.0001). Myelin sheath abnormalities (P = 0.2), axonal degeneration (number of axons; P = 0.4), and neuromuscular junction denervation (P = 0.09) were largely absent in response to sepsis, but signs of axonal swelling [higher axon area (P < 0.0001) and g-ratio (P = 0.03)] were observed. A reduction in maximal muscle force was present after indirect nerve stimulation (P = 0.007) and after direct muscle stimulation (P = 0.03). The degree of force reduction was similar with both stimulations (P = 0.2), identifying skeletal muscle, but not peripheral nerves, as the main contributor to muscle weakness. Myofibrillar calcium sensitivity of the contractile apparatus was unaffected by sepsis (P ≥ 0.6), whereas septic myofibres displayed disorganized sarcomeres (P < 0.0001) and altered myofibre axial elasticity (P < 0.0001). Septic myofibres suffered from increased rupturing in a passive stretching protocol (25% more than control myofibres; P = 0.04), which was associated with impaired myofibre active force generation (P = 0.04), linking altered myofibre integrity to function. Sepsis also caused a reduction in muscle titin gene expression (P = 0.04) and myosin and actin protein content (P = 0.05), but not the myosin-to-actin ratio (P = 0.7).

Conclusions: Prolonged sepsis-induced muscle weakness may predominantly be related to a disruption in myofibrillar cytoarchitectural structure, rather than to neural abnormalities.

Keywords: Biomechanics; Muscle contraction; Muscle weakness; Neuropathy; Sepsis.

Figure 1

Electromyography (EMG) measurements of the dorsal caudal tail nerve–tail muscle and sciatic nerve–gastrocnemius muscle complexes. In vivo EMG measurements were performed before randomization at Day 0 and before sacrifice at Day 5. Data are presented as the delta of both measurements. (A) Compound muscle action potential (CMAP) latency in the dorsal caudal tail nerve–tail and (B) in the sciatic nerve–gastrocnemius muscle. (C) CMAP amplitude in the dorsal caudal tail nerve–tail and (D) in the sciatic nerve–gastrocnemius muscle. (E) Intensity required to achieve supramaximal stimulation in the dorsal caudal tail nerve–tail and (F) in the sciatic nerve–gastrocnemius muscle. (healthy n = 14, sepsis n = 30) *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Figure 2

Effect of sepsis on markers of neuromuscular junction (NMJ) denervation, axonal degeneration, and demyelination. (A) Innervation of NMJs, quantified in the tibialis muscle as the percentage of NMJs with co‐localization of neurofilament light chain (NF‐L)/synaptophysin (neural synapses) and α‐bungarotoxin (neuromuscular endplate) staining. Scale bars are 20 μm. (healthy n = 12, sepsis n = 23) (B) The number of axons per 100 μm², quantified as the median number per animal. (healthy n = 9, sepsis n = 20) (C) Histogram of the axon size and fitted curve. Bars are mean ± standard error of the mean. Grey line = healthy animals, black line = septic animals. (D) Quantification of the median axon size per animal. (healthy n = 9, sepsis n = 20) (E) The g‐ratio in relation to the axon diameter. Grey crosses + fitted grey line = healthy animals, black dots + fitted black line = septic animals. Insert graph shows the median g‐ratio per animal. (healthy n = 8, sepsis n = 20) (F) Myelin abnormalities (abnormal myelin thickness and in‐/out‐foldings) indicated with yellow arrows. Scale bar is 10 μm. (G) Quantification of the percentage of axons with myelin abnormalities. (healthy n = 8, sepsis n = 20) All axon/myelin analyses were performed in toluidine blue stained axons of the sciatic nerve. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Figure 3

In situ extensor digitorum longus (EDL) muscle force measurements. Representative force traces and quantification of the absolute maximal tetanic force generated by the EDL muscle after (A) indirect stimulation of the deep peroneal nerve and (B) after direct EDL stimulation. Grey line = healthy mice, black line = septic mice. (C) Delta of the absolute maximal tetanic force measurements after indirect nerve and direct muscle stimulations. (healthy n = 8, sepsis n = 26) *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Figure 4

Effect of sepsis on myofibre calcium sensitivity of the contractile apparatus. (A) Representative force traces showing increasing force for each decreasing pCa step. Grey curve = healthy controls, black curve = septic mice. (B) Mean pCa–force group curves with reconstructed Hill fits. Data presented as mean ± standard error of the mean. White circles + grey curve = healthy controls, black circles + black curve = septic mice. (C) Group analyses of pCa₅₀ and (D) Hill coefficients. (healthy n = 23 myofibres from five mice, sepsis n = 45 myofibres from 10 mice) *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Figure 5

Effect of sepsis on sarcomere organization and muscular myosin/actin content. (A) Sarcomeres were visualized by label‐free multiphoton second harmonic generation (SHG) microscopy. Representative images of SHG measurements in healthy controls and septic mice. Scale bars are 20 μm. (B) Cosine angle sum (CAS) measurements of the SHG signal, indicating the variability in angular myofibrils. (healthy n = 256 from 86 myofibres from six mice, sepsis n = 433 from 145 myofibres from 10 mice) (C) Representative silver staining of myosin and actin in the gastrocnemius muscle. H lane represents a healthy animal, and S a septic one. (D) Muscular myosin and actin protein content as percentage of controls, and myosin‐to‐actin ratio. (healthy n = 15 mice, sepsis n = 30 mice) *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Figure 6

Effect of sepsis on myofibre elasticity. Myofibres were passively stretched to 140% of their resting length (L ₀) to assess passive elasticity. (A) Representative recordings of resting length–tension (RLT) curves showing force in relation to percentage of strain. (B) Survival analysis of myofibres subjected to the stretching protocol. Numbers indicate the amount and percentage of ruptured myofibres per 10% strain. (grey line: healthy n = 23 myofibres from four mice, black line: sepsis n = 51 myofibres from 10 mice) Elastic properties of ruptured (grey) and non‐ruptured myofibres (black), evaluated as (C) axial compliance and (D) Young's modulus. (statistic symbols represent difference between ruptured and non‐ruptured myofibres; myofibres from healthy mice: ruptured n = 10, non‐ruptured n = 13; myofibres from septic mice: ruptured n = 37, non‐ruptured n = 14) Properties of myofibres that survived the stretching protocol were assessed as the (E) force at 140% stretch of L ₀, (F) axial compliance, and (G) Young's modulus. (statistic symbols represent difference between healthy and septic mice; grey full triangles: healthy n = 13 myofibres, black open triangles: sepsis n = 14 myofibres) Axial compliance and Young's modulus data (C, D, F, G) are shown as mean ± standard error of the mean. (H) Maximal myofibrillar force, produced at the lowest pCa step, shown in relation to myofibre rupturing in the stretching protocol. (healthy mice: ruptured n = 10, non‐ruptured n = 11; septic mice: ruptured n = 22, non‐ruptured n = 13) (I) Relative mRNA expression of sarcomeric protein titin in the whole muscle. (healthy n = 14 mice, sepsis n = 33 mice) *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.