Preclinical insights into the gut‐skeletal muscle axis in chronic gastrointestinal diseases

Abstract Muscle wasting represents a constant pathological feature of common chronic gastrointestinal diseases, including liver cirrhosis (LC), inflammatory bowel diseases (IBD), chronic pancreatitis (CP) and pancreatic cancer (PC), and is associated with increased morbidity and mortality. Recent clinical and experimental studies point to the existence of a gut‐skeletal muscle axis that is constituted by specific gut‐derived mediators which activate pro‐ and anti‐sarcopenic signalling pathways in skeletal muscle cells. A pathophysiological link between both organs is also provided by low‐grade systemic inflammation. Animal models of LC, IBD, CP and PC represent an important resource for mechanistic and preclinical studies on disease‐associated muscle wasting. They are also required to test and validate specific anti‐sarcopenic therapies prior to clinical application. In this article, we review frequently used rodent models of muscle wasting in the context of chronic gastrointestinal diseases, survey their specific advantages and limitations and discuss possibilities for further research activities in the field. We conclude that animal models of LC‐, IBD‐ and PC‐associated sarcopenia are an essential supplement to clinical studies because they may provide additional mechanistic insights and help to identify molecular targets for therapeutic interventions in humans.


| INTRODUC TI ON
In developed countries, muscle wasting occurs largely in the setting of disease. It often accompanies chronic gastrointestinal diseases such as liver cirrhosis (LC), inflammatory bowel diseases (IBD) or pancreatic cancer (PC) and is among the defining criteria for sarcopenia, cachexia and disease-related malnutrition (DRM). 1,2 Systemic pro-inflammatory conditions often cause these catabolic processes 3 as seen in metastatic cancer, sepsis, tuberculosis or HIV. Sarcopenia is present in up to 70% of patients with LC, 4 approximately 17% of patients with chronic pancreatitis (CP) and ranges from 30% to 65% in patients with PC. 5 In German hospitals and in the Western world in general, every fourth patient is malnourished with a particularly high prevalence of >30% in gastroenterology. 6 The prevalence of malnutrition is estimated to be between 20% and 60% in LC patients, and up to 75% and 62% in patients with Crohn's disease and ulcerative colitis, respectively. 7 A recent large study from Spain reported a more conservative estimate of 16% prevalence of malnutrition in IBD patients. 8 Sarcopenia, that is muscle failure, is defined as a muscle disease rooted in adverse muscle changes that accrue across a lifetime. 1 The loss of muscle mass and function in sarcopenia is often associated with physical disability, poor quality of life, increased hospital admissions and increased mortality. 9 Sarcopenia is common among the elderly and was initially considered being a geriatric syndrome.
However, its definition has been expanded to include its development in younger patients due to malnutrition and also the existence of chronic disease and inflammation. 10 According to the current consensus on definition and diagnosis of sarcopenia, low muscle strength is the key characteristic, low muscle quantity and quality are used to confirm diagnosis and poor physical performance indicates severe sarcopenia. 1 Malnutrition may not only be caused by compromised intake or assimilation of nutrients but also by concomitant diseases specifying this type as DRM. 2 Pathophysiologically, DRM is often triggered by a sustained disease-specific inflammatory response of mild to moderate degree. The chronic state accompanied by a milder inflammatory response is also called cachexia, although cachexia is often incorrectly perceived as end-stage malnutrition. 2 The chronic inflammation contributes to malnutrition through associated anorexia and decreased food intake as well as an altered metabolism with elevation of resting energy expenditure and increased muscle catabolism. 2,11 Consequently, DRM leads to an altered body composition manifesting as a decrease in muscle mass and thereby is associated with adverse functional and clinical outcomes. 11 Malnutrition and sarcopenia often manifest clinically through a combination of decreased nutrient intake, inflammation and decreased body weight, along with a decrease in muscle mass, strength and/or physical function. 6,10 In liver cirrhosis, malnutrition plays a central role in muscle loss, which is not restricted to a reduced energy intake. Malnourished LC patients were also shown to have an increase of the intestinal permeability 12 and changes in the gut microbiota. 13 It is assumed that a microbial translocation and the consecutive inflammation contribute to sarcopenia in LC patients. 4 Muscle wasting is associated with complications like infections, ascites or hepatic encephalopathy and with an impaired prognosis and higher mortality. 14 Besides malnutrition, an altered lipid and carbohydrate metabolism, inhibition of muscle growth, inflammation and inactivity contribute to muscle loss in LC. 15 Due to the multifactorial aetiology, muscle loss in LC is not adequately treated by nutrition and physical activity alone. 15 In patients with LC, the transjugular intrahepatic portosystemic shunt (TIPS) is widely used to treat portal hypertension, especially in the management of refractory variceal bleeding and refractory ascites. After successful treatment with TIPS, body weight gain and improvement in the nutritional status have been widely reported. In malnourished LC patients with hypermetabolism who underwent TIPS placement, an enhancement in the body cell and muscle mass and thereby an improvement in the body composition was described. 16 Further, TIPS was shown to reverse sarcopenia whereas failure to improve muscle area after TIPS was accompanied by higher mortality. 17 Other strategies to minimize or reverse muscle loss include branched chain amino acid, L-ornithine and L-aspartate supplementation as well as Rifaximin administration. [18][19][20] However, due to the multifactorial aetiology and the selection of the best possible therapy, it is important to understand the complex metabolic and hormonal changes associated with muscle wasting in LC.
Malnutrition is also highly prevalent among patients with inflammatory bowel disease, especially with Crohn's disease, 21 and occurs in active as well as remission phases of the disease. Further, it has been shown that malnourished patients more often suffer from sarcopenia as well. 7,22 It is assumed that the process of muscle wasting is reversible in IBD patients when the underlying disease is treated. 23 However, muscle loss in IBD is common, negatively impacts various clinical outcomes and is associated with increased hospitalizations, disease flares, the need for surgery and post-operative complications. 8 Recent studies have indicated that low muscle mass is an independent predictor of the surgical outcome and that sarcopenic IBD patients who require surgery are often in need of increased post-operative care. 22,23 It was shown that the prevalence of muscle wasting is related to a deficiency of vitamin D, linking the muscular failure to the nutritional status. 24 Further molecular mechanisms of muscle wasting in IBD remain to be elucidated.
In chronic pancreatitis, repeated episodes of inflammation induce fibrosis, 25 eventually resulting in the loss of pancreatic exocrine and endocrine function. Impairment of exocrine pancreatic function is also the single most important factor regulating intestinal microbiota composition and thus predisposing to gut dysbiosis. 26 Malnutrition and sarcopenia are frequent complications of CP.
Besides pancreatic exocrine insufficiency, small intestinal bacterial overgrowth, pain-induced anorexia and ongoing alcohol and nicotine consumption present presumed risk factors. 27,28 An elevated resting energy expenditure, found in 30%-50% of patients, may further contribute to loss of muscle mass and function. 29 For malnutrition and sarcopenia alike, the use of inconsistent definitions and diagnostic modalities makes accurate prevalence estimation difficult. However, reported prevalences of 8%-39% and 17%-52% for malnutrition based on body mass index and sarcopenia, respectively, indicate that both conditions are common manifestations in CP. 27,30,31 Because both malnutrition and sarcopenia have been associated with adverse health-related outcomes and increased mortality, 27,32 treatment can be considered mandatory. Although loss of muscle mass is an inherent overlap between the two conditions, it is still elusive whether identical therapy should be applied. Unfortunately, understanding of the underlying pathophysiology is still limited. Advanced insights into interrelations between malnutrition, sarcopenia, and CP therefore hold potential to benefit future therapeutic options.
Pancreatic cancer (PC) represents the fourth leading cause of cancer-related mortality. 33 Malnutrition is a common phenomenon in PC patients and negatively influences the quality of life, length of hospital stay and survival. 34,35 Specifically, failure of exocrine and endocrine functions can lead to malnutrition, resulting in a catabolic state. 35 Sarcopenia in PC patients is associated with a worse prognosis, 5 and preoperative sarcopenia was found to be associated with a prolonged hospital stay after pancreatic surgery. 36 It has been suggested that muscle loss is influenced by an imbalance of anabolic and catabolic pathways as well as systemic inflammation, 37 but the underlying molecular mechanisms are poorly understood yet.

| THE G UT-S K ELE TAL MUSCLE A XIS
A growing body of evidence suggests that gastrointestinal organs and skeletal muscles are directly linked to each other by a network of hormones, metabolites and mediators that together form the so-called gut-skeletal muscle axis. 38,39 In this chapter, we provide a short overview of the current knowledge in this field.
The gastrointestinal tract constitutes an important source of hormones that directly and indirectly affect morphology and function of skeletal muscles. Gut-derived hormones with potential anti-sarcopenic effects are glucagon-like peptide-1 (GLP-1) and −2 (GLP-2) as well as ghrelin. GLP-1 agonists, which are widely used for the treatment of diabetes mellitus, have recently been proposed to ameliorate muscle wasting by suppressing myostatin and muscle atrophic factors (such as atrogin-1 and muscle RING-finger protein-1 (MuRF-1)), and enhancing myogenic factors including MyoG and MyoD. 40 Pharmacologic GLP-2-stimulation promotes gut-specific growth, enhances intestinal epithelial function and alleviates malnutrition, likely secondary contributes to the gut-specific pro-absorptive effects. 41 Similarly, ghrelin has been suggested to exert protective effects on fasting-induced muscle atrophy in ageing mice through an enhanced expression of myogenic genes and decreased expression of degradation genes. 42 The effects of these hormones on muscle wasting associated with chronic gastrointestinal diseases have not been studied yet.
A plethora of nutritional components also plays specific roles in the maintenance and regulation of muscle metabolism. These substances include, but are not restricted to, individual proteins and amino acids, 43,44 vitamins (especially, vitamin D 45 ) and various lipid components, such as n-3 polyunsaturated fatty acids. 46,47 Since a detailed survey of this aspect is beyond the scope of this paper, we kindly refer the reader to a more dedicated recent review in the field. 39 The gut microbiome influences and modulates the gut-skeletal muscle axis through a variety of microbiota-derived metabolites and other molecules. Prominent examples include short-chain fatty acids (eg butyrate), secondary bile acids, water-soluble B-vitamins, the amino acid tryptophan, polyphenols and antioxidants (eg urolithines) as well as pathogen-associated molecular patterns, which, under pathophysiological conditions, trigger inflammatory responses by interacting with pattern-recognition receptors on cells of the innate immune system. 38,39 Disease-associated imbalances of the gut-skeletal muscle axis are frequently accompanied, mediated and/or enhanced by a low-grade systemic inflammation, characterized by increased levels of pro-inflammatory cytokines (such as tumour necrosis factor-α (TNF-α) and interleukin-6 (IL-6)), elevated circulating lipopolysaccharide (eg in gastrointestinal diseases) and a predominance of anti-vs pro-myogenic mediators (eg myostatin vs insulin-like growth factor-1 (IGF-1)). 39,48 In skeletal muscle cells, evolutionary conserved signalling cascades constitute the point of convergence for the action of many mediators mentioned above. To this end, the following signal transduction pathways have emerged as particularly important with respect to the induction or inhibition of muscle wasting in the context of gastrointestinal diseases: myostatin, a member of the transforming growth factor-β (TGF-β) protein family, exerts its inhibitory effects on skeletal muscle growth through its binding to the activin type 2B receptor and activation of mothers against decapentaplegic homolog (Smad) 2 and 3 transcription factors. 49 Subsequently, inhibition of the phosphatidylinositol 3-kinase (PI 3-K)/protein kinase B (AKT)/mammalian target of rapamycin (mTOR) pathway results in diminished protein synthesis and muscle atrophy. 50 Two major antagonists of myostatin are follistatin and IGF-1, which interfere with its effects by direct interaction (follistatin), 51 or at the level of PI 3-K/ AKT/mTOR through a (re)-activation of the signalling cascade. 52 Pro-inflammatory mediators such as TNF-α, but also hyperammonemia, activate nuclear factor 'kappa-light-chain-enhancer' of activated B cells (NF-κB) signalling and induce, in part mediated by myostatin, downstream target genes including MuRF-1 and atrogin-1 (MAFbx) that promote muscle protein degradation through the proteasomal pathway. 53,54 Protein degradation can also be triggered by increased levels of reactive oxygen species and subsequent activation of the transcription factor FoxO. 55 Additional molecular processes that have been implicated in disease-associated muscle wasting are enhancement of autophagy (eg via LC3/p62-SQSTM1) 54,56 and induction of mitochondrial dysfunction. 39 These pathophysiological processes may also result in cell death by apoptosis (rather than necrosis), which contributes to the progress of sarcopenia. 56

| ANIMAL MODEL S OF MUSCLE WA S TING IN CHRONI C G A S TROINTE S TINAL DISE A SE S
Animal models of LC, IBD and PC represent an important resource for preclinical studies on disease-associated muscle wasting and have provided significant pathomechanistic insights. In this chapter, we review frequently used animal models of gastrointestinal diseases associated with muscle wasting and focus on their specific advantages but also their limitations.

| Rodent models of chronic liver diseaseassociated muscle wasting
The most widely used rodent models of muscle wasting in the context of chronic liver disease comprise the models of portocaval shunts (PS), bile duct ligation (BDL) and carbon tetrachloride (CCl 4 ) administration in rats or mice ( Table 1).
The PS model has been established in rats [57][58][59] and shares with the BDL model the need for microsurgical skills and equipment. In 6-weeks-old male Sprague-Dawley rats, a portocaval shunt, within 4 weeks, leads to nutritional and metabolic changes similar to those that occur in human cirrhosis. Notably, it also causes muscle wasting and skeletal muscle atrophy, which are accompanied by lower skeletal muscle mass and grip strength. 57 Muscle wasting in this model has been shown to be the consequence of impaired satellite cell proliferation and differentiation, and linked to lower IGF-1/IGF-1-receptor expression as well as higher mRNA and protein levels of myostatin, activin type 2B receptor and cyclin-dependent kinase inhibitor p21 (CDKN1A). 58 Together, these findings are compatible with the concept of an uninhibited proteolysis and a reduced rate of muscle protein synthesis. Interestingly, application of follistatin, a potent myostatin antagonist, was sufficient to largely reverse muscle wasting in this model. 59 As a key mechanistic factor in PS-triggered myopenia, hyperammonemia has been proposed. 53,60 Specifically, hyperammonemia was found to stimulate myostatin expression in a NF-κB-dependent manner 53 and to activate autophagy in skeletal muscle. 60 Autophagy could be particularly relevant for degradation of nitrated muscle proteins, which are increased both in patients and rats with LC. 60 Transjugular intrahepatic portosystemic shunt improves survival in well selected LC patients. 61 However, TIPS may even worsen liver function 62 and, conversely, can be associated with muscle wasting. 59 Furthermore, hyperammonemia, as described in the PS model, is also frequently observed in patients with TIPS and may cause complications such as enhanced hepatic encephalopathy. Hyperammonemia is one of the major causes of muscular growth inhibition. 53  develop progressive cirrhosis associated with ascites. 64 In the meantime, the BDL model has been widely used for studies in the field of cholestatic liver disease, liver fibrosis and cirrhosis, and shown to be applicable in mice as well. 65  observed. 54 Furthermore, in BDL rat models 6 weeks after surgery, protein synthesis was decreased in gastrocnemius muscle but not in other organs. 67 To this end, the molecular basis of myopenia in bile duct-ligated rodents remains to be fully elucidated. As one key mechanism, increased degradation of muscle proteins through the ubiquitin-proteasome pathway has been proposed and locally acting TNF-α implicated in the mediation of these effects. 66 In FVB mice, decreased protein synthesis, in addition to increased protein degradation, has been suggested as the principal cause of skeletal muscle atrophy.
Blocking of protein synthesis could be attributed to increased levels of myostatin, which inhibited the PI 3-K/Akt/mTOR pathway. In addition, myostatin might trigger autophagy, thereby further enhancing the development of muscle atrophy in bile duct-ligated FVB mice. 54 In the mouse model, neither enhanced muscle tissue levels of TNF-α nor of IL-6 were observed. 54 Furthermore, to the best of our knowledge, no evidence for a pivotal role of systemic inflammation in myopenia of bile duct-ligated rodents has been reported, rendering the model less suitable for the analysis of this important aspect of the human disease. Additionally, the relatively short time course to disease onset is in contrast to the rather slow progression of chronic liver disease in humans. 68 The advantages of this model are the long availability, the therefore well-studied structural and functional changes which are induced by BDL, as well as its reproducibility. 68 The CCl 4 model is considered prototypical for rodent models of LC that rely on the application of hepatotoxins. Induction of LC requires the repeated application of CCl 4 over a period of 2-3 months (eg twice a week). Susceptibility to CCl 4 liver injury and fibrogenesis in mice largely depends on the genetic background, 69 with BALB/c as the most sensitive strain. 70 CCl 4 exerts consistent hepatotoxicity through its reactive metabolites generated by cytochrome P-450 enzymes, primarily CYP2E1, expressed in perivenular hepatocytes. 71 Studies on cirrhosis-associated muscle wasting have been performed both in mice and rat. 54 Morphological and functional alterations of skeletal muscles in In summary, each of the three models described above is characterized by a unique pathophysiology that resembles some but not all aspects of human chronic liver disease. Inhibition of protein synthesis and/or enhancement of protein degradation represent common features that can be studied in each of the models, although the underlying molecular mechanisms are to some degree variable. The two surgical models may not be applicable to the analysis of inflammation-associated processes of muscle wasting, whereas the toxicity and variable reproducibility of its effects limit the use of CCl 4 . Table 1 summarizes key features of the three models.
All three rodent models of chronic liver disease lead to body weight loss, indicative of malnutrition in these animals. Malnutrition has not been investigated systematically in these models and most of the literature was published before the year 2000. To the best of our knowledge, the gut-skeletal muscle axis in rodent models of chronic liver disease has not been surveyed yet. Intriguingly, CCl 4 -induced liver cirrhosis can be improved by oral administration of S boulardii, which reduces the intestinal permeability and modulates the gut microbiome composition. 74 Also, protective effects against bacterial antigen translocation via oral application of B pseudocatenulatum CECT7765 have been demonstrated in experimental cirrhosis. 75 Effects on the underlying disease most likely also have an impact on the disease-related muscle wasting.

| Models of muscle wasting in inflammatory bowel diseases
Even in patients with Crohn's disease in clinical remission, prevalence of muscle wasting is as high as 60%. 76  The gut-skeletal muscle axis has not been addressed in studies of experimental colitis. It has been shown that DSS colitis alters the expression of neurotrophins in smooth muscle cells 81 and TNBS colitis leads to dysfunction of calcium channels in smooth muscle cells. 82 Subsequently, gut dysmotility can lead to an impaired barrier function, 82 which might contribute to muscle wasting. However, a link to skeletal muscle cells has not been firmly established yet.

| Models of muscle wasting in pancreatic disorders
With a prevalence of approximately 17%, sarcopenia represents a frequent complication of CP. 27 Muscle wasting in CP patients is closely associated with pancreatic exocrine insufficiency and has been linked to adverse health-related outcomes. 27,83,84 Acknowledging the high clinical relevance of chronic pancreatic disorders, several animal models for CP have been established. These include repetitive caerulein injections leading to recurrent acute pancreatitis, a well-known cause for chronic pancreatitis. 85 Caerulein is a cholecystokinin analogue which causes acute pancreatitis when given in supraphysiological concentrations, 86 which is mediated by the balance between trypsin-activating and trypsin-degrading lysosomal enzymes. 87,88 The caerulein model can be modified experimentally ranging from mild oedematous, to necrotizing and to chronic pancreatitis. 89 Intraperitoneal injections over 10 weeks twice per week lead to repetitive bouts of acute pancreatitis and ultimately leading to chronic injury of the organ. Another approach is based on an obstruction of the pancreatic duct that is performed by a ligation followed by a single caerulein injection, leading to severe necrotizing pancreatitis and subsequent fibrosis. 90 This model is also F I G U R E 1 Mechanisms of sarcopenia addressed by animal models of gastrointestinal diseases. Rodent models of chronic liver diseases, inflammatory bowel diseases and pancreatic disorders including cancer induce sarcopenia through a variety of molecular and cellular processes. Increased transcription of myostatin, Murf-1 and/or atrogin-1 leads to an increased protein degradation (mostly via proteasomal degradation) and decreased protein synthesis. Additionally, NF-κB can trigger myostatin, Murf-1 and atrogin-1. Myostatin furthermore inhibits PI 3-K/Akt/mTOR signalling, which in turn leads to autophagy. Molecular processes can also be influenced by hyperammonemia or cytokines such as IL-6, TNF-α, TGF-β or activin, which circulate in the blood system. Especially hyperammonemia and the pro-inflammatory cytokines IL-6 and TNF-α stimulate autophagy and NF-κB-dependent protein degradation. Models of chronic liver disease (portocaval shunt, bile duct ligation, CCl 4 ), inflammatory bowel disease (TNBS and DSS colitis), chronic pancreatitis and pancreatic cancer (GEMM, transplant models) address different aspects of the complex system that can lead to sarcopenia. CCl 4 , carbon tetrachloride; DSS, dextran sulphate sodium; GEMM, genetically engineered mouse models; IL-6, interleukin 6; MyD88, myeloid differentiation factor; Murf-1, muscle RINGfinger protein-1; NF-κB, nuclear factor 'kappa-light-chain-enhancer' of activated B cells; PI3-K/Akt/mTOR, phosphatidylinositol 3-kinase/ protein kinase B/mammalian target of rapamycin; TGF-β, transforming growth factor-β; TNBS, trinitrobenzene sulphonic acid; TNF-α, tumour necrosis factor-α; ZIP4, zinc transporter

| E XPERIMENTAL MODEL S -FUTURE DIREC TIONS
Recent advancements in the field of omics technologies have also opened the avenue for novel approaches to elucidate the gut-skel-

| CON CLUS IONS
In chronic gastrointestinal diseases such as LC, IBD, CP and PC, an incomplete pathophysiological understanding of the gut-skeletal muscle axis hampers the development of specific anti-sarcopenic therapies. Rodent models are useful to gain further mechanistic insights into disease-associated muscle wasting and to identify promising molecular targets for therapeutic interventions. Since none of the established animal models completely resembles the human disease, care needs to be taken when choosing the appropriate model to address dedicated scientific questions. Further progress in the field can be expected from recent advancements of omics technologies (eg for microbiome and metabolome studies), and from the rapidly increasing availability of well-defined genetically engineered mouse models.