Skeletal muscle function during the progression of cancer cachexia in the male Apc^Min/+ mouse

Abstract

While cancer-induced skeletal muscle wasting has been widely investigated, the drivers of cancer-induced muscle functional decrements are only beginning to be understood. Decreased muscle function impacts cancer patient quality of life and health status, and several potential therapeutics have failed in clinical trials due to a lack of functional improvement. Furthermore, systemic inflammation and intrinsic inflammatory signaling’s role in the cachectic disruption of muscle function requires further investigation. We examined skeletal muscle functional properties during cancer cachexia and determined their relationship to systemic and intrinsic cachexia indices. Male Apc^Min/+ (MIN) mice were stratified by percent body weight loss into weight stable (WS; <5% loss) or cachectic (CX; >5% loss). Age-matched C57BL/6 littermates served as controls. Tibialis anterior (TA) twitch properties, tetanic force, and fatigability were examined in situ. TA protein and mRNA expression were examined in the nonstimulated leg. CX decreased muscle mass, tetanic force (P_o), and specific tetanic force (_sP_o). Whole body and muscle fatigability were increased in WS and CX. CX had slower contraction rates, +dP/dt and −dP/dt, which were inversely associated with muscle signal transducer and activator of transcription 3 (STAT3) and p65 activation. STAT3 and p65 activation were also inversely associated with P_o. However, STAT3 was not related to _sP_o or fatigue. Muscle suppressor of cytokine signaling 3 mRNA expression was negatively associated with TA weight, P_o, and _sP_o but not fatigue. Our study demonstrates that multiple functional deficits that occur with cancer cachexia are associated with increased muscle inflammatory signaling. Notably, muscle fatigability is increased in the MIN mouse before cachexia development.

NEW & NOTEWORTHY Recent studies have identified decrements in skeletal muscle function during cachexia. We have extended these studies by directly relating decrements in muscle function to established cachexia indices. Our results demonstrate that a slow-fatigable contractile phenotype is developed during the progression of cachexia that coincides with increased muscle inflammatory signaling. Furthermore, regression analysis identified predictors of cancer-induced muscle dysfunction. Last, we report the novel finding that whole body and muscle fatigability were increased before cachexia development.

INTRODUCTION

Cachexia is a complex wasting disorder characterized by unintentional body weight loss secondary to chronic disease (24, 25, 29). Cachexia affects roughly 80% of cancer patients and is responsible for over 20% of all cancer-related deaths (7, 59). Cachexia-induced skeletal muscle mass loss is associated with reduced ability to perform basic daily tasks leading to a loss of functional independence (4, 8, 29, 30, 45, 65). More clinically relevant, cancer patients with decreased functional ability and increased fatigue, regardless of cachexia diagnosis, experience decreased quality of life and a poorer prognosis (3, 48). Cachectic patients and preclinical cachexia models exhibit decreased volitional activity, whole body weakness, and fatigue (10, 46, 52, 53, 61). Recently, specific decrements to skeletal muscle’s contractile quality have been identified in cachectic patients and tumor-bearing mice (22, 34, 38, 46, 53). However, the drivers of these functional deficits in skeletal muscle have not been clearly defined (13, 22, 50). To this end, several proposed cancer cachexia therapeutics have failed in clinical trials due to their inability to improve overall patient function (9, 13, 28). This inability of potential therapeutics to improve muscle function illustrates our lack of understanding related to the cachectic factors that induce these deficits. Critical gaps in our knowledge of this regulation are likely related to the dearth of studies examining both muscle function and cachectic factors in the same cohort of tumor-bearing mice. This lack of cohesive integration with these measurements has likely contributed to the plethora of unsuccessful cancer cachexia treatment paradigms (3, 10, 22, 46, 53, 66, 70).

Disrupted proteostasis and oxidative metabolism, linked to chronic inflammation, are integral to the pathology of cachectic skeletal muscle (6, 21, 66). Both systemic and muscle inflammation have been extensively examined for their regulation of muscle wasting with cancer (6, 8, 20, 25, 29, 49, 56, 59, 63). Widely examined systemic mediators of muscle wasting include proinflammatory cytokines interleukin-6 (IL-6), tumor necrosis factor-α (TNFα), transforming growth factor-β (TGF-β), interferon-γ (IFNγ), and others (20, 41, 43, 49, 51, 63). Although many systemic mediators of cancer cachexia appear to be associated with specific preclinical models, IL-6 has been implicated as a driver of cachexia in cancer patients and preclinical models (20). Interestingly, there is a significant inverse relationship between circulating IL-6 and muscle strength in heart failure patients and sarcopenia (60). Disrupted protein turnover and metabolic homeostasis in the Apc^Min/+ (MIN) mouse, an established cancer cachexia model, is dependent on elevated circulating IL-6 (20, 62, 66, 67, 69, 70). Muscle inflammatory signaling has also be investigated for the disruption of protein turnover and metabolism induced by cancer (1, 5, 56, 77). Chronic activation of signal transducer and activator of transcription 3 (STAT3), nuclear factor-κB (p65), mitogen-activated protein kinases (p38 and ERK1/2), and protein kinase B (Akt) signaling in skeletal muscle are all associated with muscle wasting (6, 16, 20, 40, 77). Interestingly, each of these signaling pathways can regulate muscle mass in healthy and atrophic conditions (16, 35, 69, 76). Moreover, STAT3 can disrupt cardiac myocyte contractility (74, 75). However, neither systemic IL-6 nor muscle inflammatory signaling have been directly associated with skeletal muscle function decrements during the progression of cancer cachexia.

Muscle force production, fatigue, and twitch characteristics are functional properties that define skeletal muscle quality. Life quality associated with aging, muscular dystrophy, COPD, and cancer patients is directly related to these muscle parameters (3, 11, 58, 72, 73). Furthermore, muscle fatigability (distinct from central fatigue), myofibrillar protein expression, and calcium handling characteristics are used to define muscle phenotype (23, 39, 57). A myofiber's response to catabolic stimuli has been shown to be regulated by phenotype; slow-oxidative fibers are more susceptible to disuse atrophy, whereas fast-glycolytic muscles are more sensitive to cachectic stimuli (1, 17, 53, 71). While understanding the biochemical regulation that drives skeletal muscle wasting is of great importance, decrements in muscle function directly impact cancer patient quality of life and health status (7, 22), and several proposed cancer cachexia therapeutics have failed in clinical trials due to their inability to improve muscle function (55). Therefore, determining the regulation of impaired muscle function during cancer cachexia progression is imperative to understanding this complex wasting disorder. We previously showed that MIN mice demonstrate decreased voluntary wheel running and cage activity before significant weight loss (10). However, it is not known whether this decreased volitional activity is associated with decrements in skeletal muscle function. To this end, we examined skeletal muscle functional properties during cancer cachexia and determined their relationship to systemic and intrinsic cachexia indices. We hypothesized that cachectic skeletal muscle would exhibit decreased strength and increased fatigability compared with weight-stable and healthy controls and that these decrements would be inversely related to increased muscle inflammatory signaling. To test this hypothesis, male MIN mice were stratified by percent body weight loss into weight stable (MIN-WS; <5% BW loss) or cachectic (MIN-CX; >5% BW loss). Age-matched male C57BL/6 (WT) littermates served as controls. The tibialis anterior (TA) was stimulated in situ and analyzed for properties of muscle function. Protein and mRNA expressions related to the cachectic phenotype were determined from the contralateral TA to determine relationships to muscle function.

METHODS

Animals.

Male WT and MIN mice were purchased from Jackson Laboratories and were bred at the University of South Carolina Animal Resources Facility. All animals were group housed and kept on a 12:12-h light-dark cycle. Body weights were measured weekly, and animals were monitored for signs of distress. Animals were given food and water ad libitum throughout the duration of the study. All animals were fasted 5 h before tissue collection. Mice were anesthetized with a ketamine-xylazine-acepromazine cocktail, and hindlimb muscles and selected organs were carefully dissected and snap-frozen in liquid nitrogen and stored at −80°C until further analysis. All animal experiments were approved by the University of South Carolina Institutional Animal Care and Use Committee.

Analysis of muscle function.

At ~20 wk of age, mice were anesthetized with 2% isoflurane inhalation and kept anesthetized at 1.5% isoflurane throughout the duration of the procedure (~1 h). Muscle function analysis of the TA in situ, which maintains an intact nerve and host blood supply, has been previously described (15, 27, 46). Briefly, the distal tendon of the left TA was isolated and tied to a force transducer (Aurora Scientific, Ontario, Canada) using 5-0 silk sutures. The mouse was place on the apparatus maintained at 37°C throughout the entirety of the procedure (15). The sciatic nerve was exposed proximal to the knee and maintained using warmed mineral oil. The sciatic nerve was then subjected to a single stimulus to determine the optimal length (L_o). Once L_o was obtained, a force-frequency curve was generated, and maximal tetanic force was determined. After a 5-min rest, the TA was subjected to a 4-min intermittent fatigue protocol consisting of 1-s maximal stimulation every 4 s (46). Maximal tension was again measured 5 and 10 min after completion of the fatigue protocol. Specific tension was determined using TA muscle cross-sectional area (CSA) calculated by muscle mass/(L_f × 1.06), where L_f represents fiber length determined by multiplying L_o by 0.6, the predetermined TA muscle length-to-fiber length ratio, and 1.06 represents the skeletal muscle density (17).

Western blot analysis.

Western blot analysis was performed as described previously (52). Briefly, the right proximal TA was homogenized, and protein concentration was determined using the standard Bradford protein assay. Homogenates were fractionated on 7–10% SDS-polyacrylamide gels and transferred to a polyvinylidene difluoride membrane. Primary antibodies for phosphorylated (p) and total (t) p65, STAT3, p38, and ERK1/2 (Cell Signaling Technology) were incubated 1:1,000 overnight at 4°C in 5% TBST milk. Anti-rabbit IgG-conjugated secondary antibodies (Cell Signaling Technology) were incubated 1:2,000 for 1 h at room temperature in 5% TBST milk. Enhanced chemiluminescence was used to visualize antibody-antigen interaction and captured using the Syngene: G-Box. Blots were analyzed by determining the integrated optical density of each band using ImageJ (NIH software).

RNA isolation and RT-PCR.

RNA isolation, cDNA synthesis, and real-time PCR were performed as previously described (47), using reagents from Applied Biosystems (Foster City, CA). Primers for GAPDH, IL-6, IL-1β, TNFα, SOCS3 (suppressor of cytokine signaling 3), SERCA1 (sarcoplasmic/endoplasmic reticulum calcium 1), RyR1 (ryanodine receptor 1), and calsequestrin were purchased from IDT (Coralville, IA) and run using SYBR Green PCR buffer (47). Data were analyzed using the 2^−ΔΔCT method.

Cytochrome c oxidase activity.

Extensor digitorum longus (EDL) muscle samples were homogenized in extraction buffer (0.1 M KH₂P0₄-Na₂HP0₄, 2 mM EDTA, pH 7.2). Cytochrome c oxidase (COX) enzyme activity was determined by measuring the rate of oxidation of fully reduced cytochrome c at 550 nm using a CYTOCOX1 Sigma Aldrich Kit and a spectrophotometer (Eppendorf) (37).

Treadmill run to fatigue.

A separate cohort of WT (n = 6) and MIN (n = 15) mice was run on a treadmill to examine time to exhaustion as a measure of whole body fatigue, as previously described (64). Three days before the fatigue test, mice were acclimated to the treadmill by running at a 5% grade for a total of 20 min, with 5 m/min incremental increases in speed starting at 5 m/min and finishing at 20 m/min. After acclimation, the mice were run on a treadmill until complete exhaustion, determined by 2 min of resistance to hand prodding. The fatigue test consisted of 5 min at 5, 10, and 15 m/min and 30 min at 20 m/min, and then increased to the final speed of 25 m/min.

Statistical analysis.

Values are presented as means ± SE. Student's t-tests were performed to determine differences between genotypes. A one-way ANOVA was used to determine differences in muscle function and inflammatory signaling when MIN mice were stratified by age or body weight loss. Post hoc analyses were performed with Student-Newman-Keuls methods. A Bartlett’s test was used to determine significantly different standard deviations (P < 0.05). If a significant difference was observed between group standard deviations, a nonparametric Kruskal-Wallis one-way ANOVA was used. A Pearson correlation was used to determine correlations between inflammatory genes and proteins with muscle function properties in MIN mice. Stepwise linear regression models were used to identify predictors (cachexia indices) of the eight measured outcomes related to skeletal muscle function (SAS). Significance was set at P < 0.05.

RESULTS

Animal characteristics.

Age-matched WT (n = 10) and MIN (n = 16) mice were examined between 18 and 22 wk of age (Table 1). MIN mice had decreased hindlimb muscle mass (−22%) and fat mass (−76%) without changes in developmental growth, indicated by similar tibia lengths. To examine the effects of cachexia progression, mice were stratified into MIN-WS (n = 7, <5% BW loss) and MIN-CX (n = 9, >5% BW loss) based on body weight loss from peak body weight. As expected, MIN-CX demonstrated reduced hindlimb muscle mass (−37%) and fat mass (−100%) compared with MIN-WS (Table 2). Both MIN-WS and MIN-CX had enlarged spleens, indicative of elevated systemic inflammation due to the presence of intestinal tumors.

Table 1.

Characteristic data of ~20-wk-old male MIN and WT mice

	WT	MIN
n	10	16
Age, wk	20.3 ± 0.8	20.9 ± 0.5
Peak BW, g	26.7 ± 0.5	25.4 ± 0.5
Final BW, g	26.7 ± 0.5	22.8 ± 1.0
BW change from peak, %	0.0 ± 0	−10.6 ± 2.3*
Hindlimb muscle, mg	206 ± 5	161 ± 11*
Testes, mg	196 ± 5	152 ± 12*
Epididymal fat, mg	396 ± 18	93 ± 29*
Spleen, mg	67 ± 3	423 ± 34*
Tibia length, mm	16.8 ± 0.1	16.9 ± 0.0

Values are means ± SE. Body weights (BW) given in grams (g). Percent (%) loss determined from peak weight to weight before euthanasia. All tissue weights expressed in milligrams (mg). WT, C57BL/6; MIN, Apc^Min/+.

^*Significant from WT, P < 0.05 (t-test).

Table 2.

Characteristic data of male MIN mice stratified by body weight loss

	MIN-WS	MIN-CX
n	7	9
Age, wk	21.8 ± 0.6	20.0 ± 0.6
Peak BW, g	27.0 ± 0.6	24.1 ± 0.4*
Final BW, g	26.8 ± 0.6	19.7 ± 0.4*
BW change from peak, %	−0.9 ± 0.4	−18.3 ± 0.9*
Hindlimb muscle, mg	201 ± 5	127 ± 9*
Testes, mg	194 ± 5	118 ± 13*
Epididymal fat, mg	212 ± 20	0 ± 0*
Spleen, mg	384 ± 65	453 ± 21
Tibia length, mm	16.9 ± 0.0	16.8 ± 0.1

Values are means ± SE. Body weights (BW) given in grams (g). %Loss determined from peak weight to weight before euthanasia. All tissue weights expressed in milligrams (mg). WS, weight stable; CX, cachexia. Male Apc^Min/+ (MIN) mice were stratified by weight loss: 0–5% loss (MIN-WS), 5–20% loss (MIN-CX).

^*Significant from MIN-WS, P < 0.05 (t-test).

Cachexia indices in tumor-bearing mice.

MIN-CX mice had increased plasma IL-6 levels compared with MIN-WS and WT mice (Fig. 1A). MIN-CX mice had significantly reduced TA weight (35%) compared with MIN-WS (Fig. 1B). We then examined intrinsic muscle inflammatory signaling proteins in the nonstimulated TA (Fig. 1, C and D). Activation of signaling proteins was determined from the phosphorylated-to-total ratio of the expressed protein in muscle. STAT3 activation increased 24-fold in MIN-CX compared with MIN-WS and 19-fold compared with WT mice. p65 activation increased 1.5-fold in MIN-CX compared with MIN-WS and 1.6-fold compared with WT. ERK1/2 activation increased 3.4-fold in MIN-CX compared with MIN-WS and 3.3-fold compared with WT. p38 activation was increased 2.2-fold in MIN-CX compared with MIN-WS and 2.0-fold compared with WT. We then examined muscle inflammatory gene expression in the nonstimulated TA (Fig. 1E). IL-6 mRNA was increased in MIN-CX 9.1-fold compared with MIN-WS and 9.0-fold compared with WT. SOCS3 mRNA was increased in MIN-CX 15-fold compared with MIN-WS and 16-fold compared with WT. IL-1β mRNA was increased in MIN-CX 7.1-fold compared with MIN-WS and 4.8-fold compared with WT. Last, TNFα mRNA was increased in MIN-CX 1.3-fold compared with MIN-WS, but was not different from WT. EDL COX enzyme activity was reduced 50% in MIN-CX compared with MIN-WS and 52% compared with WT (Fig. 1F). There was a negative relationship between STAT3 activation and TA weight in all MIN mice (Fig. 1G). Similarly, a negative relationship was observed with SOCS3 mRNA and TA weight (Fig. 1H). Many indices of cachexia had a negative relationship with body weight loss in all MIN mice (Table 3); SOCS3 expression and STAT3 activation exhibited strong negative relationships with percent body weight loss. Interestingly, only p38 activation and COX activity were related to muscle mass in all MIN mice.

Fig. 1.

Indices of cachexia in Apc^Min/+ (MIN) mice. A: plasma levels of IL-6 in pg/ml. B: tibialis anterior (TA) weight in mg. Gray shading represents SE of WT. C: ratio of phosphorylated (p) to total (t) protein expression of key inflammatory signaling proteins, signal transducer and activator of transcription 3 (STAT3), p65, extracellular signal-related kinase (ERK)1/2, and p38 in TA from MIN weight-stable (MIN-WS) and cachectic MIN (MIN-CX) normalized to WT mice. D: representative Western blot images of phosphorylated and total protein expression of inflammatory proteins in TA. Dotted lines indicate where blots were cropped for representation. E: mRNA expression of IL-6, STAT3, IL-1β, and tumor necrosis factor-α (TNFα) in TA from MIN-WS and MIN-CX normalized to WT. F: cytochrome c oxidase (COX) activity assay in extensor digitorum longus (EDL). Gray shading represents SE of WT. G: Pearson’s correlation between TA weight and STAT3 activation in all ~20-wk-old MIN mice. H: Pearson’s correlation between TA weight and suppressor of cytokine signaling 3 (SOCS3) mRNA in all ~20-wk-old MIN mice. Values are means ± SE. *Significant from MIN-WS; ^significant from WT. Significance was set at P < 0.05.

Table 3.

Relationship between BW loss, muscle mass, and cachexia indices in ~20-wk-old male MIN mice

Cachexia Indices	TA Weight	BW Loss
Plasma IL-6	−0.53 (see Fig. 1)	−0.46
STAT3		−0.78*
p65	−0.44	−0.55*
ERK1/2	−0.49	−0.47*
p38	−0.77* (see Fig. 1)	−0.74*
SOCS3 mRNA		−0.76*
IL-6 mRNA	−0.28	−0.44
IL-1β mRNA	−0.41	−0.42
TNFα mRNA	−0.37	−0.52*
COX activity†	0.68*	0.64*

Values are expressed as Pearson's correlation (R). Values used for STAT3, p65, ERK1/2, and p38 are ratios of phosphorylated to total protein expression. TA, tibialis anterior; BW, body weight; STAT3, signal transducer and activator of transcription 3; ERK1/2, extracellular signal-related kinase-1/2; SOCS3, suppressor of cytokine signaling 3; TNFα, tumor necrosis factor-α; COX, cytochrome c oxidase.

^*Significant correlation, P < 0.05;

^†analysis done in the extensor digitorum longus muscle.

Whole body fatigability in tumor-bearing mice.

We (10) have previously reported decreased voluntary wheel running and grip strength throughout the progression of cachexia in MIN mice. We have extended these prior findings by measuring time to exhaustion as a measure of whole body fatigability, using a graded treadmill exercise test in MIN and age-matched WT mice. Both MIN-CX (>5% BW loss, n = 7) and MIN-WS (<5% BW loss, n = 5) mice had decreased (77 and 43%, respectively) exercise capacity compared with WT controls (Fig. 2A). There was no difference between 12-wk Apc^Min/+ (MIN-12) mice and age-matched WT, indicating that the deficit in fatigue was not inherent in the genotype (Fig. 2A).

Fig. 2.

Effect of cachexia on whole body and muscle specific fatigue. A: treadmill time to fatigue in weight-stable Apc^Min/+ (MIN-WS), cachectic MIN (MIN-CX), and 12-wk MIN (MIN-12) vs. WT mice. Gray shading represents SE of WT. B: in situ muscle fatigability shown as %maximal tension after 2 min of an intermittent fatigue protocol. Gray shading represents SE of WT. C: absolute force over duration of the intermittent fatigue protocol. D: relative (%) force over duration of the intermittent fatigue protocol. E: Pearson’s correlation between muscle fatigability and signal transducer and activator of transcription 3 (STAT3) activation in all ~20-wk-old MIN mice. F: Pearson’s correlation between muscle fatigability and suppressor of cytokine signaling 3 (SOCS3) mRNA in all ~20-wk-old MIN mice. Values are means ± SE. *Significant from MIN-WS; ^significant from WT; ^@significant from MIN-12. Significance was set at P < 0.05.

Skeletal muscle fatigability in tumor-bearing mice.

TA muscle fatigability was determined using an intermittent fatigue protocol in situ. TA fatigability was increased in the MIN compared with WT mice (Fig. 2B). Interestingly, when MIN mice were stratified by body weight loss, no difference was observed in muscle fatigability between MIN-CX and MIN-WS mice (Fig. 2B). We also observed no difference in muscle fatigability between MIN-12 and age-matched WT mice, indicating that the deficit in muscle fatigability was not inherent to the genotype. Plasma IL-6 demonstrated a trend (P = 0.07) for a relationship with muscle fatigue in all MIN mice; however, muscle fatigability was not significantly related to any other measured cachexia indices (Tables 4 and 5). Interestingly, COX activity, a surrogate for oxidative metabolism, was not related to skeletal muscle fatigability, indicating that fatigue may be present before decrements in muscle oxidative metabolism. However, it should be noted that COX activity analysis was performed in the EDL instead of the TA.

Table 4.

Relationship between force and fatigability of the TA and indices of cachexia in ~20-wk-old male MIN mice

	Fatigue	Po	sPo	−dP/dt	+dP/dt
Cachexia indices
Plasma IL-6	−0.56	−0.48	−0.10	−0.39	−0.33
BW loss	0.01	0.93*	0.60*	0.86*	0.87*
Spleen/BW	0.26	−0.62*	−0.36	−0.48	−0.64*
TA/BW	−0.18	0.72*	0.01	0.63	0.66*
Testes	0.11	0.85*	0.21	0.75*	0.85*
Cachexia indices in skeletal muscle
STAT3	Figure 2	Figure 3	Figure 4	−0.64*	−0.73*
p65	0.28	−0.62*	−0.71*	−0.66*	−0.51*
ERK1/2	−0.48	−0.44	−0.09	−0.35	−0.42
p38	0.05	−0.73*	−0.34	−0.62*	−0.73*
IL-6 mRNA	−0.08	−0.41	−0.48	−0.41	−0.26
SOCS3 mRNA	Figure 2	Figure 3	Figure 4	−0.72*	−0.58*
IL-1β mRNA	0.01	−0.41	−0.14	−0.43	−0.20
TNFα mRNA	−0.03	−0.47*	−0.41	−0.47	−0.46*
COX activity†	−0.08	0.65*	0.20	0.54*	0.63*

Values are Pearson's correlation (R). Values used for STAT3, p65, ERK1/2, and p38 are ratios of phosphorylated to total protein expression. absolute force (P_o). specific force (_sP_o). rate of relaxation (−dP/dt). rate of rise (+dP/dt). TA, tibialis anterior; BW, body weight; STAT3, signal transducer and activator of transcription 3; ERK1/2, extracellular signal-related kinase-1/2; SOCS3, suppressor of cytokine signaling 3; TNFα, tumor necrosis factor-α; COX, cytochrome c oxidase.

^*Significant correlation P < 0.05;

^†analysis done in the extensor digitorum longus.

Table 5.

Regression analysis of force and fatigability of TA and indices of cachexia in ~20-wk old male MIN mice

	β	SE	Partial R2	Model R2	P Value
Fatigue, %
Po, mN
BW loss, %	2,850	340.8	0.87	0.87	<0.001
TA/BW, mg/g	498	158.1	0.06	0.93	0.008
sPo, kN/m2
p65	−51.2	13.5	0.51	0.51	0.002
−dP/dt, mN/ms
BW loss, %	38,462	6,189	0.73	0.73	<0.001
+dP/dt, mN/ms
TA weight, mg	324	47.9	0.77	0.77	<0.001

β, estimated regression coefficient; SE, standard error of β. BW, body weight; TA, tibialis anterior; p65, phosphorylated-to-total ratio of p65; P_o, absolute force; sP_o, specific force; −dP/dt, rate of relaxation; +dP/dt, rate of rise. Functional outcomes (dependent variable) are italicized. Significant (P < 0.05) predictors (independent variables) obtained from stepwise linear regression models that included 16 variables are shown in boldface for each functional outcome. Other variables that were not significant were removed from the model.

Skeletal muscle force through the progression of cachexia.

TA maximal tension (P_o) was reduced 41% in MIN-CX compared with MIN-WS and 42% compared with WT mice (Fig. 3C). Interestingly, a loss in force production was observed at lower frequencies (e.g., 80 Hz; Fig. 3A). Furthermore, there was an increase in the percent maximal tetanic force at 10, 30, and 50 Hz, which is indicative of a slower contractile phenotype in MIN-CX (Fig. 3B). The reduction in P_o was related strongly to STAT3 activation in MIN mice (Fig. 3D). P_o also had a strong relationship with SOCS3 mRNA in MIN mice (Fig. 3E). Due to the loss of muscle mass, we examined specific tension (_sP_o) to correct for alterations in muscle size. _sP_o was reduced 14% in MIN-CX compared with MIN-WS and 21% compared with WT mice (Fig. 4C). MIN-CX mice also displayed alterations in _sP_o at both low (50 Hz) and higher stimulation frequencies (200–300 Hz; Fig. 4A). Similarly to P_o, percent maximal specific tetanic force was increased at 10, 30, and 50 Hz in MIN-CX (Fig. 4B). The negative relationship between STAT3 activation and P_o was not present when adjusted for muscle CSA (Fig. 4D); however, the negative relationship between SOCS3 and P_o remained when adjusting for muscle CSA (Fig. 4E).

Fig. 3.

Effect of cachexia on maximal tetanic force. A: in situ force-frequency curve of tibialis anterior (TA) in weight-stable Apc^Min/+ (MIN-WS), cachectic MIN (MIN-CX) mice. B: force-frequency curve relative to maximal force. C: absolute maximal tetanic force (P_o) of TA in MIN-WS and MIN-CX. Gray shading represents SE of WT. D: Pearson’s correlation between P_o and STAT3 activation in all ~20-wk-old MIN mice. E: Pearson’s correlation between P_o and suppressor of cytokine signaling 3 (SOCS3) mRNA in all ~20-wk-old MIN mice. Values are means ± SE. *Significant from MIN-WS; ^significant from WT. Significance was set at P < 0.05.

Fig. 4.

Effect of cachexia on specific tetanic force. A: in situ force-frequency curve of tibialis anterior (TA) corrected for TA cross-sectional area (CSA) in weight-stable Apc^Min/+ (MIN-WS) and cachectic MIN (MIN-CX) mice (kN/m²). B: force-frequency curve relative to maximal force. C: specific tetanic force (_sP_o) of TA in MIN-WS and MIN-CX. Gray shading represents SE of WT. D: Pearson’s correlation between _sP_o and STAT3 activation in all ~20-wk-old MIN mice. E: Pearson’s correlation between _sP_o and cytokine signaling 3 (SOCS3) mRNA in all ~20-wk-old MIN mice. Values are means ± SE. *Significant from MIN-WS; ^significant from WT. Significance was set at P < 0.05.

We then examined the relationship between indices of cachexia and maximal force (Table 4). We observed a strong relationship between body weight loss and both P_o and _sP_o as well as a strong relationship between testes mass (68) and P_o, but this was lost when we corrected for muscle CSA. In addition to the relationships between P_o with STAT3 activation and SOCS3 mRNA, P_o was negatively related to p65 and p38 activation, TNFα mRNA, and EDL COX activity (Table 4). Notably, the negative correlation between p65 activation and _sP_o was strengthened when corrected for muscle CSA (Table 4). While several indices of cachexia had strong relationships to reduced P_o (Table 4), stepwise regression analysis determined that percent body weight loss and the relative TA mass were both significant predictors of reduced P_o in MIN mice (Table 5). Interestingly, stepwise regression analysis also identified p65 activation as a significant predictor for reduced _sP_o in MIN mice (Table 5).

Cachectic skeletal muscle contractile properties.

Twitch and tetanic contractile characteristics were examined to better determine cachectic skeletal muscle contractile phenotype. One-half relaxation time (1/2 RT) was increased 19% in MIN-CX compared with MIN-WS and 19% compared with WT (Fig. 5A). Furthermore, the rate of relaxation from tetanus, -dP/dt, was reduced 49% in MIN-CX compared with MIN-WS and 67% compared with WT (Fig. 5C). There was a trend for a relationship between ½ RT and STAT3 activation (P = 0.07) in all MIN mice (Table 6). Interestingly, SOCS3 mRNA expression was not related to ½ RT (Table 6). Stepwise regression analysis demonstrated that percent body weight loss and SOCS3 mRNA expression were significant predictors of reduced ½ RT in MIN mice (Table 7). There was a strong negative relationship with –dP/dt and STAT3 as well as SOCS3 (Table 4); however, percent body weight loss was the only significant predictor of reduced –dP/dt (Table 5). Time to peak twitch (TPT) was increased 15% in MIN-CX compared with MIN-WS and 15% compared with WT (Fig. 5B). TPT was related to STAT3 activation (Table 6); however, only TNFα emerged as the only predictor for the decrements to TPT (Table 7). The rate of rise to tetanus, +dP/dt, was reduced 35% in MIN-CX compared with MIN-WS and 55% compared with WT (Fig. 5D). A negative relationship existed between +dP/dt and STAT3 activation (Table 4) as well as SOCS3 mRNA expression (Table 4). Stepwise regression analysis identified TA weight as a significant predictor of reduced +dP/dt in MIN mice (Table 5). We found no difference in twitch tension (_tP_o) between any groups (data not shown). Similarly, neither STAT3 activation nor SOCS3 mRNA was related to _tP_o (Table 6).

Fig. 5.

Effect of cachexia on muscle force rates. A: ½ relaxation time of tibialis anterior (TA) muscle twitch at L_o in weight-stable Apc^Min/+ (MIN-WS) and cachectic MIN (MIN-CX) mice. Gray shading represents SE of WT. B: time to peak twitch of TA muscle twitch at L_o in MIN-WS and MIN-CX. Gray shading represents SE of WT. C: rate of relaxation (−dP/dt) of TA muscle at 200 Hz. Gray shading represents SE of WT. D: rate of rise (+dP/dt) of TA muscle at 200 Hz. *Significant from MIN-WS; ^significant from WT. Significance was set at P < 0.05.

Table 6.

Relationship between twitch properties of TA and indices of cachexia in ~20-wk-old male MIN mice

	1/2 RT	TPT	tPo
Cachexia indices
Plasma IL-6	0.27	0.44	0.27
BW loss	−0.52*	−0.63*	−0.52*
Spleen/BW	0.34	0.49*	0.34
TA/BW	−0.22	−0.18	−0.22
Testes	−0.44	−0.43	0.24
Cachexia indices in skeletal muscle
STAT3	0.45	0.57*	−0.35
p65	0.22	0.43	0.22
ERK1/2	0.26	0.32	0.26
p38	0.37	0.48	0.37
IL-6 mRNA	−0.25	−0.15	−0.25
SOCS3 mRNA	0.02	0.19	−0.44
IL-1β mRNA	−0.13	0.04	−0.13
TNFα mRNA	0.38	0.64*	0.38
COX activity†	−0.04	−0.23	0.51*

Values are Pearson's correlation (R). Values used for STAT3, p65, ERK1/2, and p38 are ratios of phosphorylated to total protein expression. 1/2 RT, 1/2 relaxation time; TPT, time to peak twitch; _tP_o peak twitch; TA, tibialis anterior; BW, body weight; STAT3, signal transducer and activator of transcription 3; ERK1/2, extracellular signal-related kinase-1/2; SOCS3, suppressor of cytokine signaling 3; TNFα, tumor necrosis factor-α; COX, cytochrome c oxidase.

^*Significant correlation, P < 0.05;

^†analysis done in the extensor digitorum longus.

Table 7.

Regression analysis of twitch properties of TA and indices of cachexia in ~20-wk old male MIN mice

	β	SE	Partial R2	Model R2	P Value
1/2 RT, ms
BW loss, %	−27.8	5.89	0.29	0.29	0.033
SOCS3 mRNA	−0.19	0.05	0.35	0.63	0.004
TPT, ms
TNFα mRNA	6.90	2.20	0.41	0.41	0.007
tPo, mN
COX activity	17.9	7.99	0.26	0.26	0.042

β, estimated regression coefficient. standard error of β (SE). 1/2 RT, 1/2 relaxation time; TPT, time to peak twitch; _tP_o peak twitch; TA, tibialis anterior; BW, body weight; SOCS3, suppressor of cytokine signaling 3; TNFα, tumor necrosis factor-α; COX, cytochrome c oxidase. Functional outcomes (dependent variable) are italicized. Significant (P < 0.05) predictors (independent variables) obtained from stepwise linear regression models that included 16 variables are shown in boldface for each functional outcome.

Calcium handling gene expression.

On the basis of the observations demonstrating a slower contractile phenotype (Figs. 3C, 4C, 5, A and B, and 6, A and B), we investigated the gene expression of key calcium handling proteins, SERCA1, RyR1, and calsequestrin (Fig. 6). SERCA1 gene expression was increased 1.4-fold in MIN-CX muscle compared with both MIN-WS and WT. RyR1 gene expression was increased 1.3-fold in MIN-CX compared with MIN-WS and 1.6-fold compared with WT. Last, calsequestrin gene expression was increased in MIN-CX (3.1-fold) and MIN-WS (2.2-fold) compared with WT; however, no significant difference between MIN-CX and MIN-WS was found.

Fig. 6.

Effect of cachexia on calcium handling gene expression. mRNA expression of sarco(endo)plasmic reticulum calcium ATPase-1 (SERCA1), ryanodine receptor 1 (RyR1), and calsequestrin in tibialis anterior (TA) muscle from weight-stable Apc^Min/+ (MIN-WS) and cachectic MIN (MIN-CX) mice normalized to WT. *Significant from MIN-WS; ^significant from WT. Significance was set at P < 0.05.

DISCUSSION

The current study aimed to improve our understanding of cachectic muscle’s functional decrements and illuminate potential contributing factors. Although cachectic skeletal muscle force, fatigability, and inflammatory signaling have all been studied, the lack of a cohesive integration of these measurements has limited the possible conclusions in understanding their role in the pathogenesis of muscle dysfunction with cancer (3, 10, 22, 46, 53, 66, 70). Therefore, the primary goal of this study was to examine the functional properties of skeletal muscle during the development of cachexia and to determine their relationship to the systemic and intrinsic regulators of the cachectic muscle phenotype. This study demonstrates that a slow-fatigable contractile phenotype develops during the progression of cachexia similar to what has been previously reported in other cachexia models (34, 38, 46, 53). However, we have extended these findings by demonstrating that functional changes have a strong relationship to intrinsic muscle inflammatory signaling. We also report the novel finding that muscle fatigability was increased before significant cachexia development. Unexpectedly, we found no relationship between increased muscle fatigability and systemic IL-6 levels in male MIN mice.

Along with muscle mass, fatigue is a strong predictor of survival in cancer patients (22, 65). Determining whether this fatigue is due to neural, humoral, or musculoskeletal alterations remains a consistent barrier in cancer and cachexia research. The present study provides evidence to support a direct role for skeletal muscle in cancer fatigue development before the onset of cachexia. Cachectic skeletal muscle has been reported to have disrupted metabolic homeostasis, which may contribute to muscle-specific fatigue. Our laboratory has previously found disrupted mitochondrial quality control (decreased content and biogenesis, disrupted dynamics, etc.) in MIN muscle (21, 70), which coincides with decreased physical activity (10). Decrements in mitochondrial biogenesis and mitochondrial content have been strongly associated with increased fatigue in skeletal muscle (18, 42). The results of this study further extend this work by showing decreased COX activity in cachectic MIN muscle. Unexpectedly, MIN mice that were not cachectic exhibited increased whole body and muscle fatigue that preceded elevated inflammatory signaling. Ex vivo muscle function has also demonstrated increased fatigability in C26 mice (53), but the relationship with weight loss or inflammation was not determined. However, additional research is needed to determine the contribution of alterations in the neuromuscular junction and muscle blood flow to functional decrements with cachexia, as these can occur with aging and chronic disease (54). To this end, we demonstrated that tumor-bearing mice, independently of weight loss, experienced increased fatigue measured by whole body exercise and in situ muscle function, which was not related to elevated circulating IL-6, activated muscle inflammatory signaling, or COX activity.

Examining skeletal muscle’s contractile properties provides important information on alterations to muscle function and phenotype with aging, disease, and disuse (11, 12, 17, 31, 32, 71, 73). Skeletal muscle twitch characteristics depend on neural excitation, sarcomeric structure, calcium handling, and myosin isoform expression (23, 39, 57). One potential explanation for the slower twitch properties we report with cachexia could be a shift in fiber type distribution. While it is generally accepted that preferential atrophy of type II fibers occurs during cachexia, we have previously demonstrated that IIB, IIX, and IIA fiber types all decrease cross-sectional area in TA muscle from cachectic MIN mice. This fiber atrophy coincided with a decrease in the percentage of TA myofibers expressing high succinate dehydrogenase enzyme activity (36). Thus, atrophy of all fiber types and decreased muscle oxidative enzyme activity could contribute to the altered contractile phenotype during severe cachexia. Classically, twitch relaxation is dependent on the rate of calcium reuptake by the SERCA after stimulation (12, 23). The modifications of SERCA expression and function by cachexia remain a largely untapped area of study. Cachexia can increase SERCA1 expression in EDL and gastrocnemius muscles from tumor-bearing mice (33). Additionally, leaky calcium channels resulting in aberrant calcium handling have been hypothesized to contribute to cachexia-induced mitochondrial dysfunction, excitation-contraction uncoupling, and fatigue (21, 26, 33). Consistent with this notion, ½ relaxation time (RT) is increased in the EDL and soleus in tumor-bearing mice (38, 53). We extend these findings by demonstrating that cachectic MIN have a decreased rate of relaxation (−dP/dt) and increased ½ RT, which was related to the development of cachexia. Moreover, the decline in −dP/dt was associated with body weight loss, muscle mass, and the activation of STAT3 and p38. We also report that the mRNA expression for SERCA1, RyR, and calsequesterin were induced in cachectic TA muscle, which would normally be associated with a shift to a faster phenotype (39). Future investigations are warranted to determine whether cachexia causes disruptions in sarcoplasmic reticulum calcium release and reuptake that alter muscle function.

A fundamental characteristic of skeletal muscle is the capacity to produce force, which is related to muscle size and cross-sectional area (14, 17, 23). The specific tension produced by muscle, force per unit area, has been a principal parameter for establishing muscle structural integrity (11, 17, 19). Whereas reduced specific tension has been reported previously in cachectic muscle from LLC tumor-bearing mice (53), others have reported no change in C26 tumor-bearing mice (46). These equivocal results are likely related to the muscle tested, degree and duration of cachexia, and heterogeneity related to the model and tumor type. Specific tension can be negatively impacted by expansion of the extracellular matrix, which involves connective tissue deposition and/or edema. We have previously reported increased noncontractile tissue in MIN-CX, which did not appear to be associated with regeneration or degeneration (36). Additionally, edema serves to decrease muscle protein concentration, which has not been found in MIN-CX mice (44). However, we found that indices of muscle inflammation were associated with reduced specific tension in cachectic muscle. Circulating inflammatory factors are widely investigated for their role in cancer cachexia; however, in non-tumor-bearing mice TNFα overexpression was demonstrated to decrease muscle force through the activation of muscle RING-finger protein-1 (MuRF1) through p65 (2). While elevated MuRF1 and p65 have been reported, elevated circulating TNFα has not been shown in the MIN (44). Furthermore, we have previously demonstrated that repeated bouts of eccentric muscle contraction can attenuate the induction of muscle noncontractile tissue in MIN mice that have initiated wasting (36). There is a strong justification for further examination of muscle inflammatory signaling’s role in the induction of noncontractile tissue to decrease force independently of mass in the cachectic muscle and to further understand the beneficial effects of muscle contractions or physical activity on these processes.

In summary, our study demonstrates that multiple functional deficits that occur with cancer cachexia are associated with increased muscle inflammatory signaling and that this includes the development of a slow-fatigable contractile phenotype. Although the mechanistic regulators of this phenotype require further investigation, our findings provide a rationale for examining specific inflammatory signaling pathway regulation of muscle function. Notably, muscle fatigability is increased in the MIN mouse before cachexia development. Additional research extending our mechanistic understanding of the underpinnings of cachectic muscle’s functional decrements should provide valuable information for the development of efficacious treatments for this wasting syndrome.

GRANTS

The research described in this report was supported by R01 CA-121249 (National Cancer Institute).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

B.N.V., J.P.H., and D.K.F. performed experiments; B.N.V., J.P.H., and D.K.F. analyzed data; B.N.V. and J.A.C. interpreted results of experiments; B.N.V. and J.A.C. prepared figures; B.N.V. drafted manuscript; B.N.V., J.P.H., D.K.F., and J.A.C. edited and revised manuscript; B.N.V., J.P.H., D.K.F., and J.A.C. approved final version of manuscript; J.A.C. conceived and designed research.