Ca²⁺-mediated coupling between neuromuscular junction and mitochondria in skeletal muscle

Abstract

The volitional movement of skeletal is controlled by the motor neuron at the site of neuromuscular junction (NMJ) where the retrograde signals are also passed back from muscle to the motor neuron. As the normal function of muscle largely depends on mitochondria that determine the fate of a skeletal muscle myofiber, there must exist a fine-controlled functional coupling between NMJ and mitochondria in myofibers. This mini-review discusses recent publications that reveal how spatiotemporal profiles of intracellular free Ca²⁺ could couple mitochondrial function with the activity of NMJ in skeletal muscle myofibers.

The volitional movement of skeletal muscle is controlled by motor neurons at the site of the neuromuscular junction (NMJ), a synaptic interface between a skeletal muscle fiber (myofiber) and an axonal terminal of a motor neuron¹. When the action potential reaches the axonal terminal, the neurotransmitter acetylcholine (Ach) is released and open the nicotinic acetylcholine receptors (nAchR) on the post-synaptic site of the NMJ to initiate action potentials in a myofiber, subsequentially leading to Ca²⁺ release from the sarcoplasmic reticulum (SR) and triggering myofiber contraction, a process defined as excitation contraction (EC)-coupling². Muscle contraction is highly energy demanding. Mitochondria, occupying about 10–15% of the myofiber volume³, are a major source of ATP under physiological conditions. However, under certain pathological conditions, the energy crisis induced by decreased mitochondrial ATP production, the release of excessive reactive oxygen species (ROS) and proapoptotic factors from mitochondria can contribute to muscle atrophy^4–9. Thus, mitochondria are not only essential for energy supply but also determine the survival or death of myofibers. There must exist a fine-controlled coordination between NMJ activity and mitochondrial function in myofibers for normal muscle activities. Although the detailed mechanisms remain poorly explored, multiple clues indicate that intracellular Ca²⁺ signaling could be a crucial mediator of this coordination.

Ca²⁺ plays a multifaceted role in mitochondrial function. Mitochondrial Ca²⁺ influx (transient) has been first observed in vivo during skeletal muscle contraction induced by motor nerve stimulation¹⁰ and later quantified in isolated individual myofibers during EC-coupling^{11, 12}. During muscle contraction, Ca²⁺ uptake into mitochondria activates multiple enzymes related to tricarboxylic acid (TCA) cycle and oxidative phosphorylation, promoting ATP synthesis to meet the energy demands^13–15. Various disease conditions, such as spinal cord injury (denervation), neuromuscular diseases, or aging, etc. can lead to long-term muscle disuse. It is well known that prolonged muscle inactivity is associated with an increased level of resting Ca²⁺ in cytosol^{16, 17}, which could overload mitochondria with Ca²⁺, stimulating ROS production and increasing the release of proapoptotic proteins through mPTP opening (mitochondrial permeability transition pore)⁴. However, the cause of the elevated intracellular Ca²⁺ level in inactivated myofibers remained elusive for a long time. A series of studies from Saez’s group have made an important discovery of a potential mechanism underlying this phenomenon^18–20. They found that after 7-days of denervation, fast (but not slow) myofibers showed enhanced membrane permeability revealed with increased Evans Blue dye uptake and elevated inflammatory response evidenced by increased NF-ĸB phosphorylation and TNF-α, IL-1β expression. Their pharmacological and genetic assessments indicate that the de novo expression of connexin 43 and 45 (Cx43/Cx45) gap junction hemichannels (HCs) and purinergic receptors P2X₇Rs in the sarcolemma of denervated myofibers could be a potential upstream cause of the enhanced membrane permeability and inflammatory response¹⁸. Both Cx HCs and P2X₇Rs are known to be permeable to Ca²⁺ upon opening and both have been implicated in elevating cytosolic Ca²⁺ in excitable cells^21–27. As the free extracellular Ca²⁺ level [Ca²⁺]_ex is around 10,000-fold higher than cytosolic free Ca²⁺ ([Ca²⁺]_in), a slight offset of the membrane permeability could lead to excessive entry of Ca²⁺ from extracellular space into the cytosol. Meanwhile ATP leaked out from Cx43/Cx45 HCs can activate P2X₇Rs to promote Ca²⁺ influx²⁵, creating a feedforward mechanism of [Ca²⁺]_in elevation¹⁸, although in an amyotrophic lateral sclerosis (ALS) mouse model overexpressing human SOD1^G93A, pharmacological stimulation of P2X₇Rs rescued denervation-related muscle atrophy, arguing against the pathological role of P2X₇Rs after denervation²⁸. Moreover, a systemic analysis on myofibers post denervation (PD) by Cisterna et al revealed the temporal sequence of the following events: the increase of membrane permeability (day 3 PD) proceeded the elevation of intracellular Ca²⁺ and Na⁺ signals (day 5 PD), followed by a fall in myofiber cross-section area (day 7 PD). Remarkably, all above denervation-induced alterations were drastically reduced in myofibers derived from the mice with muscle specific knockout of Cx43/Cx45 HCs²⁰. Their data strongly suggest that denervation-induced Cx43/Cx45 expression initiates [Ca²⁺]_in elevation in myofibers. Their recent study further explored the potential mechanism that connected the NMJ activity to the expression of Cx HCs in myofibers¹⁹. They first validated the cultured mouse flexor digitorum brevis (FDB) myofibers as an in vitro denervated cellular model and found that Ach analog repressed denervation-induced Cx43/Cx45 expression and cellular alterations in cultured FDB myofibers. The application of pyridostigmine bromide, an acetylcholinesterase inhibitor, significantly prevented the denervation-induced fiber size reduction in mice. The protective effect of Ach was diminished by the application of pancuronium (Pcu), a competitive blocker of the nAChR, demonstrating a specific role of nAChR in regulating Cx HCs expression in myofibers. Although they suggest that PKC activation is downstream of nAChR, the detailed molecular mechanism linking it to the post-transcriptional regulation of Cx43/Cx45 HCs in fast myofibers (as their mRNAs showed no significant changes after denervation) remains unknown. This study demonstrated an unforeseen role of nAChRs in controlling the sarcolemmal permeability of myofibers, and for the first time, revealed a molecular mechanism that connects the NMJ inactivity to the elevated steady-state Ca²⁺ level in cytosol ([Ca²⁺]_in) of a denervated myofiber. Interestingly, denervation-induced pathological changes are more likely to occur in fast myofibers than slow fibers¹⁸, making it worthwhile to compare the transcriptomics and mitochondrial signaling changes in different myofiber types after denervation or in neuromuscular degenerative disease models.

While prolonged muscle denervation or disuse is known to increase mitochondrial ROS production^{4, 29, 30}, the initial trigger of mitochondrial ROS production in inactivated myofibers remains elusive³¹. Absence of NMJ activity eliminates EC-coupling and cytosolic Ca²⁺ transients. What is the immediate response of mitochondria to the cease of cytosolic Ca²⁺ transients? Through examining ROS-related mitoflash activities in live myofibers within 24 hours after denervation, Karam et al (2017) observed drastically increased CypD-dependent mitoflash activities, accompanied by elevated MitoSOX Red signals, indicating that short-term denervation is sufficient to promote mitochondrial ROS production³². Remarkably, mimicking physiological Ca²⁺ transients by a brief (350ms) train of electric stimulation immediately diminished the mitoflash activity, reduced the mitochondrial ROS production, and stopped the repetitive mPTP opening in the denervated muscle fibers within a minute³². The alleviating effect of electric stimulation on mitochondrial ROS generation and mitoflash activities could be blocked by RU360, an inhibitor of mitochondrial Ca²⁺ uptake through mitochondrial Ca²⁺ uniporter (MCU)^{33, 34}. Thus, loss of physiological Ca²⁺ influx in mitochondria is likely an initial trigger for excessive mitochondrial ROS production in myofibers following denervation. These results suggest that the temporal dynamics of [Ca²⁺]_mito induced by EC-coupling following NMJ activation are vital to keep ROS level in check³⁵, either through accelerating turnover of the respiratory chain or desensitizing mPTP via an unidentified Ca²⁺ responding domain, which is discussed in detail by a recent review article³⁶. This study provides another evidence of Ca²⁺-mediated coupling between NMJ activity and mitochondrial function in skeletal muscle.

The Ca²⁺-mediated link between NMJ and abnormal mitochondrial function is also observed in earlier studies of myofibers derived from ALS SOD1^G93A mice, in which the withdrawal of motor neuron innervation from skeletal muscle is the pathological hallmark of this disease^{11, 34}. Mitochondria with depolarized inner membrane potential were revealed in the fiber segment surrounding NMJ of ALS myofibers, where enhanced Ca²⁺ activities induced by osmotic shock, including Ca²⁺ sparks and propagating Ca²⁺ waves also occured³⁴. It was speculated that the collapse of the mitochondrial inner membrane potential reduced the driving force for mitochondrial Ca²⁺ uptake, thus leading to the hyperactive cytosolic Ca²⁺ activity at the site of NMJ. This hypothesis was supported by the observation that application of FCCP (to uncouple mitochondrial inner membrane potential) or RU360 (to block mitochondrial Ca²⁺ uptake) expanded the hyperactive Ca²⁺ activities from the NMJ area to the whole SOD1^G93A myofiber³⁴. A later study by Yi et al further supported this hypothesis in a more physiological and quantitative manner. In SOD1^G93A myofibers, mitochondria in the NMJ segment with collapsed membrane potential exhibits a reduced Ca²⁺ uptake transient ([Ca²⁺]_mito) during an electrical stimulation induced Ca²⁺ release. The simultaneous recording revealed that this reduced mitochondrial Ca²⁺ uptake was mirrored by a ~10% increase in the amplitude of cytosolic Ca²⁺ transients at the fiber segment harboring the NMJ, when compared to other regions of the SOD1^G93A myofiber following an electrical stimulation¹¹. These results demonstrated that mitochondrial Ca²⁺ uptake can shape the spatiotemporal profiles of cytosolic Ca²⁺ transients, and loss of mitochondrial Ca²⁺ uptake leads to uncontrolled hyperactive Ca²⁺ activity at the NMJ. It is well known that retrograde signals can be conducted from muscle back to motor neuron at the NMJ^{37, 38}. The mitochondrial lesion at the NMJ region of the SOD1^G93A muscle appears before the axonal withdrawal revealed in the same ALS mouse model^34,39. It is possible that mitochondrial lesion could exacerbate the degeneration of NMJ in ALS muscle via promoting the uncontrolled hyperactive Ca²⁺ activity at the site of NMJ³⁴. Additionally, the neuron-myofiber co-culture experiments performed by Cisterna et al., also demonstrated that decreasing Cx43/Cx45 expression in denervated myofibers somehow promoted their reinnervation by axons, supporting the presence of retrograde signals from myofibers to neurons¹⁹.

In summary, the multifaceted roles of NMJ have gradually been revealed. It not only controls volitional muscle movement, but also prevents pathological changes leading to muscle loss. The functional coupling between NMJ and mitochondria is essential for maintaining normal muscle function. The spatiotemporal profiles of the cytosolic free Ca²⁺ most likely perform a critical role to tie the NMJ activity with mitochondrial function. Dysregulated cytosolic Ca²⁺ signal could decouple the functional coordination between NMJ and mitochondria by altering the mitochondrial Ca²⁺ signaling that could result in abnormal mitochondrial Ca²⁺ retention capacity, mPTP status, ATP synthesis, and ROS production. These factors can contribute to membrane permeabilization, hyperactive cytosolic Ca²⁺ signaling and finally muscle degeneration. Loss of the functional coupling between NMJ and mitochondria could be a cause of muscle degeneration and may also promote the degeneration of motor axonal terminals in various neuromuscular disease conditions.

Highlights:

• Bidirectional communication between skeletal muscle and the motor neuron occurs at the site of neuromuscular junction.

• The fate of skeletal muscle myofibers depends on mitochondria.

• There must exist a fine-controlled functional coordination between neuromuscular junction and mitochondria in a myofiber.

• The spatiotemporal profiles of intracellular Ca²⁺ could couple mitochondrial function with the neuromuscular junction activity in a myofiber.