Our studies have relevance for patients who are suffering from muscle ischemia due to PAD, which affects a large number of patients and carries a significant threat of amputation in those who are not candidates for bypass or endovascular treatment (39). There are few effective medical treatments for this condition, and so alternative pharmacological attempts to improve perfusion and/or the response to ischemia may help attenuate the risk of limb loss in PAD patients. Hydroxychloroquine (HCQ) has been shown to have certain benefits in cardiovascular disease, and our previous studies showed similar potential protective effects in mouse FAL models. However, the exact mechanisms of action associated with CQ and HCQ effects is not clear.
Recently, CQ and HCQ have been proposed as a potential protective treatment against the novel coronavirus, SARS-CoV-2, which causes COVID-19 disease.(12) Since PAD affects a population primarily older than 65 years,(40) this is also the population at high risk for infectious complications of SARS-CoV-2. PAD along with associated cardiovascular disease present significant risks for infected patients contributing to the highest risk of death. Cytokine storm secondary to infectious activation of pathways including inflammasome and caspase-1/11 were shown to contribute to infectious complications of a similar coronavirus causing severe acute respiratory syndrome (SARS) in 2002-2003.(12) In this current study, as well as in our previous study, we show that CQ actually increases caspase-1 expression in muscle and this has both positive and detrimental effects. Our data suggest the need for caution using CQ and HCQ for COVID-19 patients without more fully understanding the pathways these drugs affect, particularly in end organ tissues such as liver and muscle.
Activation of both caspase-1 and caspase-11 can result in inflammatory cell death (pyroptosis) via cleavage of gasdermin D in monocytes and macrophages, which induces release of inflammatory cytokines (e.g. IL1β) and danger signals (41) (42). However, in non-immune cells such as hepatocytes caspase-1/11 activation is protective in ischemic injury and promotes cell survival pathways such as autophagy (9). Our data here suggest myocytes and hepatocytes share autophagy as an alternative function for caspase-1/11, with neither cell type being a major producer of caspase-1/11-mediated cytokines, IL1β and IL18, and both having excellent regenerative capacities.
Autophagy is likely to play multiple roles in recovery from muscle injury (43). Many studies suggest that autophagy is likely protective in skeletal muscle, preventing atrophy, and maintaining muscle mass (44). In inducible knockout mice lacking muscle-specific autophagy associated gene Atg7, muscle tissue was not protected from denervation-induced atrophy (44). CQ is known to inhibit autophagy by preventing autophagosome fusion with lysosomes, and it also effects lysosomal pH.(45) Knowing the likely protective effect of autophagy on muscle, we were initially surprised by our previous finding that CQ resulted in less fat replacement in muscle tissue, suggesting protection rather than increased injury (2). Fat replacement and fibrosis may result from inflammation, and since CQ is known to be anti-inflammatory (46) we hypothesized that CQ would reduce inflammatory markers, including those associated with inflammasome signaling, such as caspase-1. Instead, we found that CQ increased caspase-1 expression, cleavage and activity in C2C12 mouse myoblasts without increasing cell death (2). We therefore wondered whether caspase-1, which we knew can also induce autophagy and cellular protection (47) was influenced or influencing CQ effects in muscle related to autophagy, understanding that the effects would likely be both linked and complex. As caspase-1KO mice were not available to us, we used caspase1/11KO mice understanding the limitations of also knocking out caspase-11. We originally hypothesized that CQ-mediated increased caspase-1 in skeletal muscle would be protective, and that these protective effects of CQ would therefore be absent in caspase1/11KO mice. Our results demonstrated a more complex picture, with mixed effects of CQ on ischemic skeletal muscle, some of which were dependent on caspase-1/11 signaling, and some of which were independent.
In our previous report, we found that CQ disrupted autophagic flux resulting in accumulation of LC3IIB expression indicative of autophagosome formation, as well as decreased autophagic consumption of p62/SQTM (2). In our current study, CQ predictably resulted in an increase in autophagosomes in muscle, and also increased caspase-1 expression, including in between myofibers where capillaries and arterioles are found (48). In contrast to our hypothesis of protective effects of CQ in ischemic skeletal muscle, CQ reduced perfusion recovery after 21 days in WT mice but not caspase-1/11KO mice. Perfusion recovery in the murine model of hindlimb ischemia is dependent on arteriogenesis and angiogenesis (32, 49) suggesting CQ effects on perfusion were caspase-1/11-dependent. Others have shown that inhibition of caspase-1 is associated with improved perfusion in a mouse hind-limb ischemia model, strengthening the idea that a CQ-induced increase in caspase-1 might ultimately be detrimental to perfusion (50). We did not assess endothelial caspase-1 levels, and it is possible that these are important in the reperfusion effects of CQ. Further studies will be needed to elucidate the cell type responsible.
In other ways, however, our hypothesis of CQ protection was supported by our data. CQ reduced fat replacement and fibrosis in ischemic skeletal muscle in WT mice, with CQ effects dependent on caspase1/11 signaling in ischemic muscle at least in regards to histologic recovery. Additionally, CQ resulted in smaller myocytes in WT mice but not caspase1/11KO mice compared with PBS controls, which were potentially reflected in the fusion ability of MuSC (51), which was notably diminished in both CQ and caspase1/11KO mice. Decreased myofiber size may also reflect myofiber typing differences, and our data suggested a significant role for caspase-1/11 in determining fiber type after ischemic injury. Specifically, loss of caspase-1/11 signaling resulted in the near absence of slow twitch fibers, which was partially reversed by the addition of CQ. This is the first report describing skeletal muscle fiber-typing dependency on the presence of caspase1/11 signaling and is an intriguing finding that will lead to future studies.
Our fiber-typing suggested that caspase-1/11KO mice lacked slow twitch fibers except in the presence of CQ, and these data were supported peak force measurements after tetanic stimulation (52). CQ resulted in significantly lower peak force during tetanic stimulation in caspase-1/11KO mice, but not in WT mice. Preservation of slow twitch fibers is optimal for sustained exercise like maintenance of posture, and walking (53) and these are the main fibers affected in muscular dystrophies (54, 55). More study is required to understand if there are therapeutic benefits of preserving one fibertype over the other in ischemic disease, but our data clearly show potential effects of autophagy on fibertype and downstream functionality.
Different myofiber types have different metabolic profiles, and as caspase-1 has been shown to degrade glycolytic enzymes (38), this may have played a role in the differences seen in our model. Our studies were performed on tibialis anterior which is predominantly fast-twitch and glycolytic. Indeed, most of the regenerating fibers in WT stained positively for fast-twitch specific myosin heavy chain. However, addition of CQ in the caspase-1/11KO mice was associated with increased in slow twitch fibers over the baseline in the KO mice, which may indicate protective effects of caspase-1/11 for preserving slow twitch fibers. Differences in fiber-typing between WT and KO mice, matched metabolic status of MuSC in our model. Both CQ and absence of caspase-1/11 reduced parameters associated with mitochondrial respiration, which suggests damaged or malfunctioning mitochondria unable to be removed by mitochondrial autophagy.
Our studies have relevance for patients who are suffering from muscle ischemia due to PAD, which affects a large number of patients and carries a significant threat of amputation in those who are not candidates for bypass or endovascular treatment (39). There are few effective medical treatments for this condition, and so alternative pharmacological attempts to improve perfusion and/or the response to ischemia may help attenuate the risk of limb loss in PAD patients. Hydroxychloroquine (HCQ) has been shown to have certain benefits in cardiovascular disease, and our previous studies showed similar potential protective effects in mouse FAL models. However, the exact mechanisms of action associated with CQ and HCQ effects is not clear.
Recently, CQ and HCQ have been proposed as a potential protective treatment against the novel coronavirus, SARS-CoV-2, which causes COVID-19 disease.(12) Since PAD affects a population primarily older than 65 years,(40) this is also the population at high risk for infectious complications of SARS-CoV-2. PAD along with associated cardiovascular disease present significant risks for infected patients contributing to the highest risk of death. Cytokine storm secondary to infectious activation of pathways including inflammasome and caspase-1/11 were shown to contribute to infectious complications of a similar coronavirus causing severe acute respiratory syndrome (SARS) in 2002-2003.(12) In this current study, as well as in our previous study, we show that CQ actually increases caspase-1 expression in muscle and this has both positive and detrimental effects. Our data suggest the need for caution using CQ and HCQ for COVID-19 patients without more fully understanding the pathways these drugs affect, particularly in end organ tissues such as liver and muscle.
Activation of both caspase-1 and caspase-11 can result in inflammatory cell death (pyroptosis) via cleavage of gasdermin D in monocytes and macrophages, which induces release of inflammatory cytokines (e.g. IL1β) and danger signals (41) (42). However, in non-immune cells such as hepatocytes caspase-1/11 activation is protective in ischemic injury and promotes cell survival pathways such as autophagy (9). Our data here suggest myocytes and hepatocytes share autophagy as an alternative function for caspase-1/11, with neither cell type being a major producer of caspase-1/11-mediated cytokines, IL1β and IL18, and both having excellent regenerative capacities.
Autophagy is likely to play multiple roles in recovery from muscle injury (43). Many studies suggest that autophagy is likely protective in skeletal muscle, preventing atrophy, and maintaining muscle mass (44). In inducible knockout mice lacking muscle-specific autophagy associated gene Atg7, muscle tissue was not protected from denervation-induced atrophy (44). CQ is known to inhibit autophagy by preventing autophagosome fusion with lysosomes, and it also effects lysosomal pH.(45) Knowing the likely protective effect of autophagy on muscle, we were initially surprised by our previous finding that CQ resulted in less fat replacement in muscle tissue, suggesting protection rather than increased injury (2). Fat replacement and fibrosis may result from inflammation, and since CQ is known to be anti-inflammatory (46) we hypothesized that CQ would reduce inflammatory markers, including those associated with inflammasome signaling, such as caspase-1. Instead, we found that CQ increased caspase-1 expression, cleavage and activity in C2C12 mouse myoblasts without increasing cell death (2). We therefore wondered whether caspase-1, which we knew can also induce autophagy and cellular protection (47) was influenced or influencing CQ effects in muscle related to autophagy, understanding that the effects would likely be both linked and complex. As caspase-1KO mice were not available to us, we used caspase1/11KO mice understanding the limitations of also knocking out caspase-11. We originally hypothesized that CQ-mediated increased caspase-1 in skeletal muscle would be protective, and that these protective effects of CQ would therefore be absent in caspase1/11KO mice. Our results demonstrated a more complex picture, with mixed effects of CQ on ischemic skeletal muscle, some of which were dependent on caspase-1/11 signaling, and some of which were independent.
In our previous report, we found that CQ disrupted autophagic flux resulting in accumulation of LC3IIB expression indicative of autophagosome formation, as well as decreased autophagic consumption of p62/SQTM (2). In our current study, CQ predictably resulted in an increase in autophagosomes in muscle, and also increased caspase-1 expression, including in between myofibers where capillaries and arterioles are found (48). In contrast to our hypothesis of protective effects of CQ in ischemic skeletal muscle, CQ reduced perfusion recovery after 21 days in WT mice but not caspase-1/11KO mice. Perfusion recovery in the murine model of hindlimb ischemia is dependent on arteriogenesis and angiogenesis (32, 49) suggesting CQ effects on perfusion were caspase-1/11-dependent. Others have shown that inhibition of caspase-1 is associated with improved perfusion in a mouse hind-limb ischemia model, strengthening the idea that a CQ-induced increase in caspase-1 might ultimately be detrimental to perfusion (50). We did not assess endothelial caspase-1 levels, and it is possible that these are important in the reperfusion effects of CQ. Further studies will be needed to elucidate the cell type responsible.
In other ways, however, our hypothesis of CQ protection was supported by our data. CQ reduced fat replacement and fibrosis in ischemic skeletal muscle in WT mice, with CQ effects dependent on caspase1/11 signaling in ischemic muscle at least in regards to histologic recovery. Additionally, CQ resulted in smaller myocytes in WT mice but not caspase1/11KO mice compared with PBS controls, which were potentially reflected in the fusion ability of MuSC (51), which was notably diminished in both CQ and caspase1/11KO mice. Decreased myofiber size may also reflect myofiber typing differences, and our data suggested a significant role for caspase-1/11 in determining fiber type after ischemic injury. Specifically, loss of caspase-1/11 signaling resulted in the near absence of slow twitch fibers, which was partially reversed by the addition of CQ. This is the first report describing skeletal muscle fiber-typing dependency on the presence of caspase1/11 signaling and is an intriguing finding that will lead to future studies.
Our fiber-typing suggested that caspase-1/11KO mice lacked slow twitch fibers except in the presence of CQ, and these data were supported peak force measurements after tetanic stimulation (52). CQ resulted in significantly lower peak force during tetanic stimulation in caspase-1/11KO mice, but not in WT mice. Preservation of slow twitch fibers is optimal for sustained exercise like maintenance of posture, and walking (53) and these are the main fibers affected in muscular dystrophies (54, 55). More study is required to understand if there are therapeutic benefits of preserving one fibertype over the other in ischemic disease, but our data clearly show potential effects of autophagy on fibertype and downstream functionality.
Different myofiber types have different metabolic profiles, and as caspase-1 has been shown to degrade glycolytic enzymes (38), this may have played a role in the differences seen in our model. Our studies were performed on tibialis anterior which is predominantly fast-twitch and glycolytic. Indeed, most of the regenerating fibers in WT stained positively for fast-twitch specific myosin heavy chain. However, addition of CQ in the caspase-1/11KO mice was associated with increased in slow twitch fibers over the baseline in the KO mice, which may indicate protective effects of caspase-1/11 for preserving slow twitch fibers. Differences in fiber-typing between WT and KO mice, matched metabolic status of MuSC in our model. Both CQ and absence of caspase-1/11 reduced parameters associated with mitochondrial respiration, which suggests damaged or malfunctioning mitochondria unable to be removed by mitochondrial autophagy.