Young Investigator Award
The deleterious effects of Alzheimer’s disease pathology on skeletal muscle
2023 Volume 9 Pages 23-29
Details
2023 Volume 9 Pages 23-29
Cognitive dysfunction and skeletal muscle dysfunction tend to coexist, and the combined condition of MCI and physical frailty is defined as cognitive frailty. In this review, to understand the pathomechanism of skeletal muscle dysfunction in Alzheimer’s disease (AD), the most common cause of cognitive impairment, we first describe the physiological functions of Aβ peptide, the causative factor of AD, and its precursor, APP, in skeletal muscle. Next, the mechanism of skeletal muscle dysfunction associated with AD is discussed based on the findings of inclusion body myositis, which shows pathological conditions similar to AD, such as the accumulation of Aβ42. Then, recent findings on the effects of skeletal muscle secretory peptides on AD pathology will be presented and the physiological significance of the brain-skeletal muscle connection will be discussed. Finally, we will present our experimental results on the skeletal muscle phenotype of AD mouse models and discuss the effects of Alzheimer’s disease causative factors on skeletal muscle.
The relationship between physical and cognitive function has been explored in many clinical studies, and the coexistence of physical and cognitive decline has been identified as an emerging risk. In 2013, the International Academy on Nutrition and Aging (IANA) and the International Association of Gerontology and Geriatrics (IAGG) defined cognitive frailty (CF), as the coexistence of physical frailty and mild cognitive impairment, MCI (defined as Clinical Dementia Rating = 0.5). In addition, CF was defined as the absence of dementia, such as Alzheimer’s disease (AD)1). The definition of CF has drawn attention not only to the existence of common pathophysiological mechanisms in dementia and physical weakness, but also to the possibility that skeletal muscle dysfunction may occur before cognitive impairment manifests.
Cross-sectional studies have reported an association between physical and cognitive function, and cohort studies have showed both that cognitive decline was associated with the risk of physical decline2) and that the rate of concomitant cognitive impairment increased with the severity of physical frailty3, 4).
On the other hand, CF is reported to have a higher rate of progression to vascular dementia than AD5, 6), but prospective cohort studies have reported that the risk of developing AD is higher when there is a decline in physical function, such as muscle weakness or reduced ability to walk7, 8).
In addition, it has been reported that AD is associated with weight loss of more than 5% of normal weight9) and that weight loss begins gradually more than 10 years before the diagnosis of AD and accelerates especially 2–3 years before the onset of the disease10, 11). Furthermore, muscle mass loss is observed from the early stages of AD and continues to decline as the disease progresses12, 13). And, it has also been reported that balance functions such as the Functional Reach Test and one-leg stand time decline early in AD patients, and that walking ability and muscle strength decline with the progression of cognitive impairment14).
Thus, it has been suggested that loss of muscle mass and physical disability in AD may occur before the manifestation of dementia, and it has been hypothesized that skeletal muscle symptoms in AD patients are a manifestation of latent neurological deficits in patients who do not meet diagnostic criteria for dementia15, 16, 17).
Although there are many clinical studies on AD and skeletal muscle disorders, the pathophysiological mechanisms of why skeletal muscle disorders occur in AD and whether skeletal muscle disorders exacerbate the pathophysiology of AD are still unclear. In this review, we will first discuss the physiological role of AD pathogenic factors in skeletal muscle, inclusion body myositis (IBM), a skeletal muscle disease very similar to AD pa-thology, and the effects of skeletal muscle atrophy on AD pathology, followed by the results of our brief validation experiments using AD model mice and the resulting discussion.
In the brains of AD patients, senile plaques and neurofibrillary tangles occur in large numbers, and brain atrophy results from the associated neuronal cell death. Senile plaques are caused by the extracellular deposition of amyloid beta protein (Aβ), which is produced by the cleavage of amyloid precursor protein (APP), and neurofibrillary tangles are caused by the accumulation of phosphorylated tau protein in neurons. APP is a ubiquitous protein that is widely expressed in vivo outside the brain18), and in skeletal muscle, APP has been identified at the neuromuscular junction19). APP expression increases during neuronal differentiation and is highest during the formation of synaptic connections20, 21, 22, 23). And, overexpression of human wild-type APP in mice increases presynaptic terminals24), and increased APP mRNA expression has been reported in regenerating human muscle fibers from patients with neuromuscular diseases, suggesting that APP is important for synapse formation and maintenance in vivo25). Conversely, it has been reported that mice lacking APP have decreased body weight and reduced locomotor activity and forelimb grip strength compared to wild-type controls26, 27). Thus, APP expression is also essential in skeletal muscle to maintain and improve skeletal muscle function due to its ability to form synapses at the neuromuscular junction.
Similar to APP, β-site APP cleaving enzyme 1 (BACE1) and BACE228), the enzymes of β-secretase, and presenilins 1 and 2 (PS1 and PS2), the catalytic subunits of γ-secretase, have been found to be expressed in skeletal muscle29). In addition, Aβ accumulation and phosphorylated tau have been reported in peripheral tissues of AD patients, including skin and skeletal muscle30, 31, 32). Regarding the presence of Aβ peptide in skeletal muscle, quantification of Aβ40 and Aβ42 peptide levels in temporalis muscle of AD patients and controls showed that Aβ 40 tended to be higher in AD patients than in controls, but not significantly, while Aβ42 was significantly higher in AD patients than in controls30).
Next, we discuss the effects of Aβ peptide accumulation in skeletal muscle on skeletal muscle morphology and function, as suggested by studies of the pathogenesis of inclusion body myositis.
In the skeletal muscle diseases inclusion body myositis (IBM), polymyositis, and dermatomyositis, APP deposition and hyperphosphorylated tau are known to be present in skeletal muscle. Inclusion myositis is the most common age-related muscle disease in the elderly and is characterized by weakness of proximal and distal skeletal muscles33, 34). Histopathologically, the presence of atrophic myofibers, fringe vacuoles, and a variety of proteins including ubiquitin, Aβ protein, emerin, lamin A/C, valocin-containing protein (VCP), histones, 43 kDa TAR DNA binding protein (TDP-43), and p62 have been reported in skeletal muscle with IBM33, 35, 36). The similarity of these pathologic findings to neurodegenerative diseases such as AD and amyotrophic lateral sclerosis (ALS) is controversial, but the relationship between these diseases and IBM is not well understood37).
To investigate the influence of Aβ peptide accumulation on skeletal muscle, Sugarman et al. and Kitazawa et al. generated MCK-APP mice, an IBM mouse model in which a human Swedish APP mutation is overexpressed in skeletal muscle under the regulation of the muscle creatine kinase (MCK) promoter38), and MCK-APP/PS1 mice, double transgenic mice in which a mutation in the presenilin-1 (PS1) gene associated with familial AD was added to MCK-APP mice39). MCK-APP and MCK-APP/PS1 mice are models with increased production of Aβ40 and Aβ42, respectively, in skeletal muscle. MCK-APP/PS1 mice showed increased accumulation of phosphorylated tau compared to MCK-APP mice and exhibited histopathologic features similar to IBM, including a central core, intracellular accumulation of Aβ peptide, and increased inflammation around affected muscle fibers. Physical function was also more severely affected in MCK-APP/PS1 mice and developed at an earlier age39). These results suggest that Aβ42 accumulation in skeletal muscle, as in the brain, induces tau phosphorylation and induces cellular damage.
On the other hand, p62, which is abundant in IBM muscle fibers, is known to be deposited with neurofibrillary tangles in AD and Parkinson’s disease. Recently, p62 has been reported to function as an autophagy receptor protein, and autophagy pathways involving p62 are known to selectively degrade tau oligomers40). Deletion of p62 in a mouse model of AD (tau transgenic mouse: PS19 model) has been reported to increase the accumulation of tau oligomers in the hippocampus, leading to brain atrophy and increased inflammation41). Thus, in inclusion body myositis, as in AD, autophagy defects are thought to be one of the mechanisms of pathogenesis.
Neprilysin (NEP), a peptidase, efficiently degrades bioactive peptides such as cardioprotective BNP, antihypertensive bradykinin, and Aβ peptides. Regarding the physiological function of NEP in skeletal muscle, it has been reported that its expression is enhanced during skeletal muscle differentiation using primary human myoblasts and that NEP expression correlates with the degree of myofiber regeneration using IBM muscle biopsy samples, suggesting an important role in skeletal muscle regeneration42). In addition, studies using muscle biopsy samples from IBM patients showed that NEP protein expression was significantly increased compared to skeletal muscle from normal controls, and in particular, NEP protein was highly enriched in abnormal fibers with Aβ peptides42). This is thought to be a compensatory response to A β peptides accumulation. Interestingly, NEP gene transfer into skeletal muscle is also attracting attention as a new treatment for AD. It has been reported that introduction of an adeno-associated virus (AAV) expressing the mouse NEP gene into the hindlimb muscles of 3X-Tg-AD mice, a mouse model of AD, and overexpression of NEP in the muscles reduced brain soluble Aβ peptide levels and amyloid deposition43). Similar results have been observed using other techniques and AD mouse model. Introducing human NEP plasmid into the skeletal muscle of APP/PS1 mice, a mouse model of AD, using ultrasound combined with microbubbles, significantly reduced brain Aβ peptide levels after one month44). These results indicate that peripheral expression of NEP has a direct effect on the brain and demonstrate the efficacy of inducing NEP expression in muscle for the prevention and treatment of AD.
Physical inactivity has been cited as a factor strongly associated with the development of AD45), and there are several hypotheses about the mechanism by which exercise is effective in preventing AD. In this regard, it has recently been reported that brain-derived neurotrophic factor (BDNF), whose secretion is stimulated by exercise, is associated with brain volume46). Irisin released into the blood from skeletal muscle during exercise, is cleaved from fibronectin type III domain-containing protein 5 (FNDC5), a transmembrane precursor protein expressed in muscle under the control of PGC-1α. FNDC5/irisin is known to stimulate the expression of brain-derived neurotrophic factor (BDNF) in the hippocampus, and reduced levels of FNDC5/irisin have been reported in the hippocampus and cerebrospinal fluid in AD patients and in AD mouse models47).
On the other hand, Nagase et al. examined the effects of skeletal muscle atrophy on cognitive function in 5XFAD mice, a mouse model of AD48). By inducing muscle atrophy in 12-week-old 5XFAD mice by immobilizing their hind limbs in a cast prior to the onset of memory impairment, memory performance was normal in mice without induced muscle atrophy, but memory deficits developed in mice with muscle atrophy. A comprehensive study of molecules secreted by atrophied skeletal muscle showed that hemopexin, a glycoprotein that binds free heme in the blood, is particularly elevated, and the amount of hemopexin is increased in skeletal muscle, blood, and the hippocampus of the brain in mice with muscle atrophy48). In addition, continuous administration of hemopexin directly into the ventricles of younger, 6–7 week old AD model mice for 2 weeks resulted in memory impairment, suggesting that skeletal muscle atrophy may increase the synthesis and secretion of hemopexin from muscle and promote cognitive dysfunction in AD model mice.
To investigate the effects of AD pathology on skeletal muscle, we evaluated skeletal muscle morphology and strength in APP23 mice, a mouse model of AD, and compared them to wild-type mice. APP23 mice are transgenic for APP with a Swedish mutation in the Thy-1 promoter, and the formation of intracerebral amyloid plaques and cognitive dysfunction usually appear around 6 to 12 months of age.
In male APP23 mice and wild-type littermates, we evaluated skeletal muscle function and cognitive function at 3 months of age. The grip strength test we developed was used to assess skeletal muscle function49), and the passive avoidance test and open field test were used to assess cognitive function. Then, as factors related to both muscle strength and cognitive function, glucose tolerance was assessed by the ipGTT test and voluntary locomotion was assessed by placing a running wheel in the cage over a 3-day period. After these experiments, skeletal muscle was harvested for histologic evaluation. As a result, grip strength, body weight, and some skeletal muscle weights of APP23 mice were lower than those of wild-type mice (Fig.1 A, B). The cross-sectional area of gastrocnemius muscle fibers was also significantly smaller in APP23 mice than in wild-type mice (Fig.1 C). There were also no significant differences between APP23 and wild-type mice in food intake or assessment of glucose tolerance using the ipGTT test. Results of cognitive function tests using the passive avoidance test and the open field test showed no significant differences between APP23 mice and wild-type mice. However, APP23 mice showed significantly more voluntary movements than wild-type mice during the three days of free activity as measured by the running wheel (Fig.2). These findings suggest that skeletal muscle dysfunction is already apparent in APP23 mice at the early stage of cognitive dysfunction. In addition, the results of the ipGTT test suggest that this skeletal muscle dysfunction is independent of glucose intolerance.
Comparison of muscle strength (A), wet weight of gastrocnemius and soleus muscles (B), and myofiber cross- sectional area of gastrocnemius muscle (C) in APP23 and wild-type mice.
WT, Wild-type
Comparison of voluntary locomotor activity between APP23 and wild-type mice over three days.
Assessed by the number of revolutions of the running wheel.
Furthermore, although APP23 mice showed increased voluntary exercise, skeletal muscle function and skeletal muscle mass were rather decreased, suggesting that the process of exercise-induced muscle hypertrophy is impaired in APP23 mice.
In addition, APP23 mice are known as a model for overexpression of mutant APP specifically in the central nervous system, but we confirmed gene expression of mutant APP in skeletal muscle. A recent study reported that brain Aβ 42 affects skeletal muscle, as a model in which Aβ42 was injected into the dorsal hippocampus of rats also showed reduced muscle strength and skeletal muscle mass50). Therefore, it is possible that the skeletal muscle impairment in APP23 mice is also dependent on the presence of Aβ42, but more detailed studies are needed.
This review first discussed the possibility that skeletal muscle dysfunction precedes cognitive dysfunction, based on findings from several clinical trials and the definition of cognitive frailty. Next, the negative effects of AD pathogenic factors, such as Aβ peptide, on skeletal muscle were discussed based on findings from IBM’s pathophysiology studies. This series of discussions assumes that AD pathogenic factors such as Aβ42 affect skeletal muscle even before AD pathology becomes apparent. However, the only reports of Aβ peptide quantification in skeletal muscle are studies using temporalis muscle, and little is known about the detailed phenotype of skeletal muscle in AD patients. This is due to the invasive nature of muscle biopsy, which makes it difficult to obtain a muscle biopsy sample unless there is an obvious skeletal muscle disease. For this reason, it is and will remain nearly impossible to test for skeletal muscle Aβ peptide in human muscle biopsy samples prior to the MCI stage. Therefore, it is necessary to further verify the fluctuation of Aβ peptide in skeletal muscle before cognitive dysfunction becomes apparent using experimental animal methods, and this will be our future study topic.
Disclosures: The author declare no conflicts of interest associated with this manuscript.
