IJC International Journal of Cancer

Transcription

1 IJC International Journal of Cancer Evidence for the contribution of insulin resistance to the development of cachexia in tumor-bearing mice Michelle L. Asp, Min Tian, Angela A. Wendel y and Martha A. Belury Department of Human Nutrition, College of Education and Human Ecology, The Ohio State University, Columbus, Ohio Cancer cachexia is a syndrome of unintentional weight loss that is characterized by wasting of both skeletal muscle and adipose tissue. Glucose intolerance and insulin resistance have been associated with cancer cachexia. However, it is unknown whether resistance to insulin has a role in the development of cachexia. In the present study, male CD2F1 mice with colon-26 adenocarcinoma tumors underwent an insulin tolerance test before the onset of weight loss. Compared to mice without tumors, mice with tumors had a profoundly blunted blood glucose response to insulin. Corroborating these findings, mice with tumors had decreased phosphorylation of Akt in quadriceps muscle and epididymal adipose tissue at the end of the study. Expression of Akt-regulated genes Atrogin-1, MuRF-1, and Bnip3 was increased in muscle, suggesting a role for decreased insulin signaling in the induction of both proteasomal proteolysis and autophagy in cachectic muscle. Rosiglitazone treatment increased serum adiponectin, insulin sensitivity, and body weight, and decreased Atrogin-1 and MuRF-1 expression in the skeletal muscle of tumor-bearing mice. In conclusion, insulin resistance is an early event in mice with cachexia induced by colon-26 tumors. Rosiglitazone improves insulin sensitivity and decreases early markers of cachexia. These data provide evidence that insulin resistance is not only present in cachexia, but also has a role in cachexia pathogenesis. Correction of insulin resistance may be a novel therapeutic target for the treatment of cancer cachexia. Cachexia is a syndrome of unintentional weight loss occurring in people with cancer, HIV/AIDS, congestive heart failure, and kidney failure. It is characterized by wasting of skeletal muscle and adipose tissue, weakness, fatigue and anorexia. 1 Cachexia is estimated to occur in 30% to nearly 90% of cancer patients depending on the type of cancer 2 and account for 22% of cancer deaths. 3 Cachexia decreases the efficacy and tolerability of cancer therapies, quality of life, and survival time. 2 Dysregulated glucose metabolism is associated with cancer cachexia 4 and has been noted in cancer patients since Correcting glucose intolerance with exogenous insulin therapy has resulted in variable outcomes, 6 11 likely due to differences between studies in experimental models, insulin dose, length of treatment and type of tumor. Despite some evidence that insulin can attenuate tumor-induced weight loss in rodents, several studies report an increase in tumor size. 6,9,10 The ability of insulin to act as a growth factor in some cancers is a probable mechanism for these observations. 12 In clinical studies, hyperinsulinemia increased the risk of death in cancers of both the breast 13 and the prostate. 14 Together, these findings suggest that increasing circulating insulin through exogenous insulin therapy may have limited clinical utility in patients with cachexia. Additionally, increasing insulin levels can only be Key words: colon-26 adenocarcinoma, rosiglitazone, muscle atrophy Abbreviations ITT: insulin tolerance test;; PBS : PBS treated mice without tumors; PBSþ: PBS treated mice with tumors; RGZþ: rosiglitazone-treated mice with tumors; DMSO: dimethyl sulfoxide; SEM: standard error of the mean; AUC: area under the blood glucose curve; PTEN: Phosphatase and Tensin Homolog Deleted on Chromosome Ten effective in attenuating cachexia if tissues are insulin sensitive. Insensitivity to insulin, or insulin resistance, has been observed in patients with cachexia. 15,16 Although some evidence exists for an association between insulin resistance and cachexia, it is unknown whether insulin resistance is a consequence of cachexia or whether it has a role in cachexia development. The objective of the current study was to understand the contribution of insulin resistance to the development of tumor-induced cachexia in mice. We hypothesized that insulin resistance is an early event in the development of cachexia and that increasing insulin sensitivity improves body weight, muscle and adipose weight, and markers of proteolysis. We found that insulin resistance occurred before the onset of weight loss, and correcting insulin resistance with rosiglitazone, an insulin sensitizer, attenuated early stages of cachexia. These data suggest a role for insulin resistance in cachexia pathogenesis. Material and methods All study procedures were approved by the Institutional Animal Care and Use Committee at The Ohio State University. Mice were housed with a room temperature of C, a 12 h light/dark cycle, and free access to food and water. Mice were allowed to acclimate to the environment for approximately one week prior to the start of the study. Body weights were measured daily and food intake was measured every other day throughout the study. Colon-26 adenocarcinoma cell culture Colon-26 adenocarcinoma cells were cultured with RPMI 1640 þ L-glutamine medium (Sigma-Aldrich, St. Louis, MO) supplemented with 5% fetal bovine serum and 1% Penicillin- Int. J. Cancer: 126, (2010) VC 2009 UICC

2 Asp et al. 757 Streptomycin at 37 C and 5% CO 2. On study day 0, cells suspended in 100 ll PBS were injected into the right flank just under the skin of each mouse in the Tumor groups. An equal volume of PBS was injected into the No Tumor group. Insulin tolerance test Mice were fasted overnight for 12 hours and then injected intraperitoneally with 0.75 U/kg body weight insulin (Humulin VR R, Eli Lilly and Co., Indianapolis, IN). Glucose was measured from tail vein blood using a One Touch Ultra VR (LifeScan, Inc., Milpitas, CA) glucose meter immediately prior to insulin injection (time 0) and 15, 30, 45, 60, 90, and 120 minutes following the injection. Incremental area under the glucose curve (AUC) was calculated as the net area contained between individual baselines (set by the glucose value at time 0) and curves. 17 Necropsy Blood glucose was tested using the One Touch Ultra VR glucose meter and tail vein blood. Mice were anesthetized with isofluorane and blood collected via heart puncture. Mice were euthanized by cervical dislocation. Blood was allowed to clot and then centrifuged at 1500 g for 20 min to separate serum. Sera were frozen at 80 C until further analysis. Adipose tissue and muscle were extracted, weighed, snap frozen in liquid nitrogen, and stored at 80 C until further analysis. Study 1 Five-week old male CD2F1 mice (BALB/c DBA/2; Charles River Laboratories, Wilmington, MA) were housed five per cage and fed the AIN-93G purified diet in the form of pellets (Research Diets, New Brunswick, NJ). Mice were randomized by weight into either the Tumor group or No Tumor group and inoculated with colon-26 cells or vehicle (PBS). CD2F1 mice were chosen because they are highly sensitive to colon-26 inoculation for inducing muscle atrophy and may be a good model of human cachexia. 18 An insulin tolerance test (ITT) was conducted on days of the study, as body weight gain in the Tumor group leveled off but had not yet begun decreasing. Mice were sacrificed on days 18 19, in the fed state, when there was approximately a 20% difference in body weight between the Tumor and the No Tumor groups. Study 2 Eight week old male CD2F1 mice (BALB/c DBA/2; Harlan Laboratories, Indianapolis, IN) were housed four per cage and fed a high fat diet (23.6% by weight) in pellet form (Research Diets). Mice were randomized by weight to one of three groups: PBS treated mice without tumors (PBS ), PBS treated mice with tumors (PBSþ), or rosiglitazonetreated mice with tumors (RGZþ). Mice in the tumor groups were inoculated with colon-26 cells and treated daily with intraperitoneal injections of RGZ at 10 mg/kg body weight, or an equal volume of PBS. The RGZ (Cayman Chemical, Ann Arbor, MI) was solubilized in dimethyl sulfoxide (DMSO) as a stock solution and diluted in PBS just prior to use. Mice treated with RGZ received between 7 and 10 ll DMSO, or approximately 0.3 ll/g of body weight per day. An ITT was conducted on study day 8, before a difference in body weight was detected between the two groups. Mice were sacrificed on study day 15, after a 12 hour overnight fast, when there was approximately a 10% difference in body weight between the two groups. Western Blotting Quadriceps muscle and epididymal adipose tissue were homogenized in lysis buffer (20 mm trizma base, 1% triton- X100, 50 mm NaCl, 250 mm sucrose, 50 mm NaF, 5 mm Na 4 P 2 O 7 10H 2 O) with Complete Mini Protease Inhibitor Cocktail Tablets (Roche Diagnostics, Indianapolis, IN). Homogenates were incubated for 1 h and centrifuged at 16,000 g for 15 min. at 4 C. Supernatant protein concentration was determined with the BCA Protein Assay Kit (Pierce, Rockford, IL). Protein (80 lg/sample) was separated on 10% polyacrylamide gels and transferred to 0.45 lm nitrocellulose membranes. Ponceau S Solution (Sigma, St. Louis, MO) was used to qualitatively examine bands for equal loading of protein onto the gel and successful transfer of protein to the membrane. Membranes were blocked with 5% non-fat dry milk and incubated overnight with primary antibody (1:1000) against phosphorylated Akt (Ser 473; Cell Signaling, Danvers, MA). Membranes were then incubated with secondary antibody (HRP-linked anti-rabbit IgG, Cell Signaling) for 1 h. Bands were visualized with chemiluminescense (Super Signal, Pierce, Rockford, IL) using Kodak Image Station 2000RT (Eastman Kodak, Rochester, NY). Membranes were then stripped and re-probed for total-akt (Cell Signaling). All bands on each membrane were normalized to a positive control that was the same for all membranes. This was done to correct for any variations between blots. The relative ratio of phosphorylated to total Akt was calculated from band densities using Kodak 1D 3.6 software. RT-PCR Muscle RNA was extracted with TRIzol (Invitrogen, Carlsbad, CA) according to the manufacturer s protocol. The RNA concentration was determined spectrophotometrically, and RNA quality was visualized on a 1% agarose gel. The RNA was then reversed transcribed to cdna using the High Capacity cdna Archive Kit (ABI, Foster City, CA). The cdna was amplified by real-time PCR with TaqMan Gene Expression Assays (ABI) using pre-designed and validated primers (FAM probes, ABI) under universal cycling conditions defined by ABI. Target gene expression was normalized to the endogenous control (VIC probe, mouse GAPD) amplified in the same reaction and expressed as 2 DDct relative to the control group. 19

3 758 Insulin resistance in the development of cachexia in tumor-bearing mice ELISA Insulin and adiponectin were measured from serum by ELISA (Millipore, Billerica, MA) according to the manufacturer s protocol. Statistics Data are presented as mean þ/ standard error of the mean (SEM). Differences between Tumor and No Tumor groups in Study 1 were detected by an unpaired Student s t-test using MINITAB 15 (State College, PA). Daily body weights and ITT glucose values were analyzed by an unpaired Student s t- test with the Bonferroni correction factor for repeated measurements using SYSTAT 12 (Chicago, IL). For Study 2, effect of group was detected by one-way ANOVA using MINITAB 15. When a group effect was found, post-hoc analysis was done using Tukey s test to determine differences between individual groups. Daily body weights and ITT glucose values were measured with one-way ANOVA using repeated measures analysis with SYSTAT 12. Differences of p < 0.05 were considered significant. Results Study 1 Colon-26 adenocarcinoma induces cachexia. Fourteen days after tumor inoculation, the Tumor mice weighed significantly less than the No Tumor mice (Fig. 1a) and continued to lose weight throughout the remainder of the study. At the end of the study, the Tumor mice had 22% less total body weight than the No Tumor mice. Compared to the No Tumor group, total muscle (sum of quadriceps, gastrocnemius, and tibialis anterior) and epididymal adipose weights from the Tumor mice were decreased by 29% and 73%, respectively. The differences in body and tissue weights between the Tumor and No Tumor groups indicate that cachexia was effectively induced in this model (Table 1). Insulin resistance is present before weight loss. An insulin tolerance test was administered to the mice on study days 11 12, before a significant difference in body weight was detected between Tumor and No Tumor groups (Fig. 1a). The Tumor group was insulin resistant compared to the No Tumor group as evidenced by smaller changes in blood glucose over a two hour period after insulin injection (Fig. 1b). Incremental area under the blood glucose curve (AUC) was also decreased in the Tumor group, indicating an overall decrease in insulin sensitivity days after tumor inoculation (Fig. 1c). Additional evidence for insulin resistance at study termination. Akt is a central signaling protein in the insulin signaling cascade. When in its phosphorylated and activated form, Akt signaling stimulates pathways leading to glucose and fatty acid uptake and protein synthesis, and inhibits pathways leading to proteolysis. Compared to the No Tumor group, the quadriceps muscle (Fig. 2a) and epididymal Figure 1. Body weight and insulin tolerance test. a) Body weights over time. Male CD2F1 mice were inoculated with either colon-26 adenocarcinoma cells (Tumor ¼ ~; N ¼ 15 20) or vehicle (No Tumor ¼ l; N ¼ 9) on study day 0. b) Insulin tolerance test (ITT). On study days 11 12, mice received an intraperitoneal injection of insulin (0.75 U/kg body weight) and blood glucose was measured over two hours. Tumor group (~; N ¼ 17); No Tumor group (l; N ¼ 9). C) Area between the baseline and blood glucose curves of the ITT. *The Tumor group is significantly different than the No Tumor group (p < 0.05).

4 Asp et al. 759 Table 1. Study 1 tissue weights, serum metabolites and food intake Absolute weight Percent body weight No tumor Tumor No tumor Tumor Change in Body Weight (g) * * Quadriceps Muscle (mg) * Gastrocnemius Muscle (mg) * Tibialis Anterior Muscle (mg) * * Epidydimal Adipose (mg) * * Tumor (g) Blood Glucose (mmol/l) * Serum Insulin (ng/ml) * Food Intake (g/g body weight/day) Weights of muscles, adipose and tumor are expressed as absolute weight and as percentage of body weight (including tumor weight for the Tumor group). Muscle weight values are the mean of the left and right muscles. Blood glucose and serum insulin were analyzed at the end of the study with mice in the fed state. Food intake represents the average intake over the course of the study. N ¼ 9 for the No Tumor group and N ¼ 15 for the Tumor group. *p < adipose tissue (Fig. 2b) of the Tumor group had a 41% and 83% decrease in the ratio of phosphorylated to total Akt, respectively. Although insulin resistance was clearly present in the Tumor group, it is likely that other factors also contributed to the blunted Akt activation in adipose and muscle. Serum insulin and glucose were both decreased in the Tumor group at the end of the study. Additionally, food intake over the course of the study was 20% lower in the Tumor group compared to the No Tumor group (Table 1). Atrogin-1, MuRF-1, and Bnip3 gene expression are negatively regulated by Akt through its phosphorylation and sequestration of the FoxO transcription factor in the cytosol Atrogin-1 and MuRF-1 are E3 ligases integral to the process of proteasomal proteolysis and Bnip3 is involved in lysosomal proteolysis. In the quadriceps muscle of the Tumor mice, gene expression of Atrogin-1, MuRF-1, and Bnip3 was increased 159, 22, and 51 fold compared to the No Tumor group, respectively (Fig. 2c). Together, these data suggest that the decreased activation of Akt diminished its inhibitory effect on the expression of proteolysis-inducing genes, resulting in muscle wasting in the Tumor group. Study 2 RGZ prevents decreased body weight from tumor. The body weight gain over the course of Study 2 was smaller in the PBS(þ) group compared to the PBS( ) group. Treatment with RGZ significantly improved body weight gain in the RGZ(þ) group. The PBS(þ) group also had decreased adipose tissue weight compared to the PBS( ) group, and this was restored by RGZ treatment. Neither muscle nor tumor weights were significantly different between groups (Table 2). Insulin sensitivity is increased with RGZ. In Study 2, an insulin tolerance test was performed at an earlier time point, 8 days after tumor inoculation. No differences in body weight were present (Fig. 3a). Comparing only the PBS( ) and PBS(þ) groups, mice with tumors were more resistant to insulin compared to mice without tumors, which is evidenced by a smaller AUC in the PBS(þ) group (Fig. 3b and c). The RGZ(þ) group was more insulin sensitive than the PBS(þ) group as determined by the improved blood glucose response to insulin in the ITT (Fig. 3b) and larger AUC (Fig. 3c). Insulin sensitivity was not different between the RGZ(þ) group and the PBS( ) group. Adiponectin is an insulin-sensitizing hormone secreted by the adipose tissue. The PBS(þ) group had a 2.1-fold decrease in serum adiponectin compared to the PBS( ) group at the end of the study. Treatment with RGZ completely restored adiponectin levels in tumor-bearing mice (Fig. 3d). RGZ alters expression of proteolysis-promoting genes. Study 2 was terminated at an earlier stage of cachexia and thus muscle wasting was not yet apparent in these mice. Despite this, differences in gene expression of atrophy-promoting Atrogin-1 and MuRF-1 were already present between the two groups in the quadriceps muscle. Compared to the PBS(þ) group, treatment with RGZ decreased Atrogin-1 by 60% and MuRF-1 by 70% (Fig. 4). Atrogin-1 and MuRF-1 in the RGZ(þ) group were not significantly different than the PBS( ) group. The RGZ(þ) group had an 89% decrease in Bnip3 compared to the PBS(þ) group, however this was not statistically significant due to the high variability in Bnip3 expression among the PBS(þ) group (Fig. 4). Discussion Cancer cachexia occurs in a large percentage of advanced cancer patients and decreases strength, effectiveness of cancer treatments, and quality of life. Due to the metabolic aberrations present in this syndrome, increasing food intake does not effectively resolve cachexia. The objective of the present research was to understand the role of insulin resistance in the pathogenesis of tumor-induced cachexia. CD2F1 mice with colon-26 adenocarcinoma tumors are a widely utilized

5 760 Insulin resistance in the development of cachexia in tumor-bearing mice Figure 2. Akt activation and downstream gene expression. a) Quadriceps Akt activation. The ratio of phosphorylated to total Akt was determined. Representative blots are pictured, and bar graphs illustrate the mean ratio of phosphorylated to total Akt in each group, measured by relative band densities. Tumor group (N ¼ 14); No Tumor group (N ¼ 9). b) Epididymal Akt activation. Tumor group (N ¼ 12); No Tumor group (N ¼ 7). c) Quadriceps gene expression. Real-time RT PCR was utilized to analyze transcript levels of Akt-regulated Atrogin-1 (Tumor group N ¼ 14; No Tumor group N ¼ 9), MuRF-1 (Tumor group N ¼ 7; No Tumor group N ¼ 5), and Bnip3 (Tumor group N ¼ 18; No Tumor group N ¼ 10). Black bars represent the No Tumor group and gray bars represent the Tumor group. *The Tumor group is significantly different than the No Tumor group (p < 0.05). pre-clinical model of cancer cachexia. However, to our knowledge, insulin sensitivity and insulin signaling data have not been reported for these mice. We found that, compared to mice without tumors, mice with tumors had a remarkably blunted blood glucose response to insulin during an insulin tolerance test, which is a measure of insulin resistance. 23 Importantly, this occurred before a difference in body weight was detected between mice with and without tumors. Body weight loss of >5% has been suggested as indicative of overt cachexia. 24 In the present study, this 5% loss had not yet been achieved at the times of the ITT, giving further evidence that insulin resistance is an early event in the development of cachexia. Additionally, treatment with the insulin-sensitizer rosiglitazone improved insulin sensitivity and attenuated skeletal muscle protein degradation in tumor-bearing mice. These data provide evidence that in mice bearing colon-26 tumors, insulin resistance may be involved in the development of cachexia rather than occur as a result of cachexia. Akt activation at the end of Study 1 was decreased in both quadriceps muscle and epididymal adipose tissue of mice with tumors, corroborating our findings of insulin resistance from the ITT. We also found increased transcription of three atrophy-promoting genes, Atrogin-1, MuRF-1, and Bnip3, which are negatively regulated by Akt signaling. To our knowledge, upregulation of autophagy-regulator Bnip3 has not been reported in a cancer cachexia model. These data suggest that blunted insulin signaling may contribute to increased protein degradation and skeletal muscle wasting, both seen in cachexia. Although insulin resistance and decreased Akt signaling were the focus of the present study, numerous mechanisms of proteolysis are likely involved, including the action of counter-regulatory hormones such as cortisol 25 and epinephrine. 26 Plasma cortisol was increased in mice with tumors, and RGZ significantly attenuated this increase (data not shown). Neither the specific contributions of various mechanisms of proteolysis nor the causes of insulin resistance were determined in this study, and they should be the focus of future investigation. Although insulin resistance likely played a role in decreased insulin signaling, it cannot be concluded that insulin resistance is fully responsible for the decreased ratio of phosphorylated to total Akt in these tissues. Because of decreased food intake over the course of the study, it is possible that the tumor-bearing mice were in less of a fed state at the end of the study than the mice without tumors. Decreased food intake could have contributed to decreased blood glucose and serum insulin and a subsequent blunting of Akt phosphorylation. Additonally, serum insulin was lower in fed tumor-bearing mice in Study 1 than fasted tumorbearing mice in Study 2, suggesting the possibility of decreased pancreatic insulin secretion in severely cachectic mice with colon-26 tumors. Finally, IGF-1 is another activator of Akt and was not measured in this study. Despite these factors, insulin resistance likely has a role in decreased Akt

6 Asp et al. 761 Table 2. Study 2 tissue weights, serum metabolites and food intake Absolute weight Percent body weight PBS (2) PBS (1) RGZ (1) PBS (2) PBS (1) RGZ (1) Change in Body Weight (g) a,b a b a b a Quadriceps Muscle (mg) a,b a b Gastrocnemius Muscle (mg) Tibialis Anterior Muscle (mg) Epidydimal Adipose (mg) a,b a b Inguinal Adipose (mg) a,b a b Brown Adipose (mg) a,b a b a,b a b Tumor (g) Blood Glucose (mmol/l) a b c Serum Insulin (ng/ml) a b b Food Intake (g/g body weight/day) Weights of muscles, adipose and tumor are expressed as absolute weight and as percentage of body weight (including tumor weight for tumor groups). Muscle weights are the mean of the left and right muscles. Blood glucose and serum insulin were analyzed at the end of the study with mice in the fasted state. Food intake represents the average intake over the course of the study. N ¼ 4 for the PBS( ) group and N ¼ 12 for both the PBS(þ) group and the RGZ(þ) group. Values for each measurement that have different letters are significantly different (p < 0.05) Figure 3. Effect of rosiglitazone on insulin sensitivity. a) Body weights over time. Tumor-bearing mice were treated daily with either rosiglitazone [RGZ(þ) ; N ¼ 12] or PBS [PBS(þ) ; N ¼ 12], and mice without tumors were treated with PBS [PBS( ) ; N ¼ 4]. b) Insulin tolerance test (ITT). On study day 8, mice received an intraperitoneal injection of insulin (0.75 U/kg body weight) and blood glucose was measured over two hours. PBS( ) group ( ; N ¼ 4); PBS(þ) group ( ; N ¼ 12); and RGZ(þ) group ( ; N ¼ 12). c) Area between the baseline and blood glucose curves of the ITT. d) Serum adiponectin at the end of the study. N ¼ 12 for both PBS(þ) and RGZ(þ) groups; N ¼ 4 for PBS( ) group. Different letters represent statistically significant differences (p < 0.05). *Significantly different from PBS( ) using unpaired Student s t-test (p < 0.05). Int. J. Cancer: 126, (2010) VC 2009 UICC

7 762 Insulin resistance in the development of cachexia in tumor-bearing mice Figure 4. Effect of rosiglitazone on markers of proteolysis. Quadriceps Atrogin-1, MuRF-1, and Bnip3 gene expression. Realtime RT PCR was utilized to analyze transcript levels of Akt-regulated genes. PBS( ) group (black bars; N ¼ 4); PBS(þ) group (light gray bars; N ¼ 12); RGZ(þ) group (dark gray bars; N ¼ 12). Different letters represent statistically significantly differences (p < 0.05). signaling because of the profound blunting of insulin-stimulated glucose disposal in the insulin tolerance tests. It is most logical that blunted signaling through Akt is a cumulative result of many of these potential contributing factors. In Study 2, rosiglitazone was used as a clinically relevant means to enhance insulin sensitivity in tumor-bearing mice. Rosiglitazone, a PPAR-gamma agonist, is a potent insulin sensitizer through its ability to induce adipocyte differentiation and partition lipid away from non-adipose tissues and into adipose. 27 This adipogenic effect was apparent in our study, in which the RGZ(þ) group had 53% more adipose mass than the PBS(þ) group. PPAR-gamma activation also increases adiponectin, an important insulin-sensitizing adipokine involved in cross-talk between insulin-responsive tissues Consistent with the insulin-sensitizing effects of RGZ in other mouse models, 27,28 in the present study RGZ increased insulin sensitivity in tumor-bearing mice, and also increased serum adiponectin levels. We suspect Akt phosphorylation would also be improved in muscle and adipose tissue with RGZ treatment, however this was not measured in Study 2 because mice were in a fasted state at the time of sacrifice. Tumor-bearing mice treated with RGZ had decreased early markers of cachexia compared to tumor-bearing mice without RGZ. RGZ-treated mice had a higher body weight change over the course of the study than PBS treated mice with tumors. They also had larger epididymal, inguinal, and brown adipose depots. Because overt cachexia was not yet present at the end of Study 2, muscle weight was not different between mice with and without RGZ treatment by the References end of the study. However, both Atrogin-1 and MuRF-1 gene expression were decreased with RGZ, suggesting a blunted activation of proteolysis. These E3 ligases have been shown to be necessary and sufficient for muscle atrophy and are upregulated early in the wasting process, before changes in muscle weight are detected. 30 Besides increased proteolysis, decreased protein synthesis may also contribute to net muscle loss. Similar to proteolysis-inducing genes, key protein synthesis proteins are also regulated through the Akt signaling pathway 31 and will therefore be examined in future studies. Adipose tissue generally begins to atrophy earlier than muscle in this model. 32 Therefore, the increase in adipose mass found in RGZ-treated mice may delay the onset of muscle wasting by providing an additional source of energy in order to spare muscle protein. Rosiglitazone is part of the thiazolidinedione class of drugs, which has been noted to have potential anti-tumor activity. Rosiglitazone upregulates expression of the tumor suppressor gene Phosphatase and Tensin Homolog Deleted on Chromosome Ten (PTEN), subsequently inhibiting Akt activation. 33 This effect of RGZ appears to be tumor cell specific as RGZ has been found to decrease PTEN expression in normal cells. 34 Although it is possible that RGZ has an inhibitory effect on colon-26 tumors, tumor sizes were not different between groups in Study 2. Furthermore, final body weight was not correlated to tumor size (data not shown), indicating that the effects of RGZ to attenuate cachexia were not tumor size dependent. On the other hand, it is worth noting that RGZ also did not lead to enhanced tumor growth. The lack of association between RGZ and tumor size suggests that insulin sensitization may not increase tumor size, unlike exogenous insulin administration has been shown to do. 6,9,10 In conclusion, we found that insulin resistance is present before the onset of overt cachexia in mice with colon-26 tumors. Additionally, treatment with rosiglitazone improved insulin sensitivity and attenuated skeletal muscle protein degradation. Future research will focus on mechanisms by which insulin resistance develops and contributes to cachexia, and how rosiglitazone exerts its anti-cachectic effects. The data presented here provide evidence that insulin resistance is an early event in cachexia development, and improving insulin sensitivity with rosiglitazone is associated with decreased skeletal muscle protein degradation. Acknowledgements The authors thank Bethanie Combs for her work on the second animal study, Michael Stout and Denis Guttridge for scholarly discussions about the data, and Julia Richardson for editorial assistance with the manuscript. 1. Argiles JM, Almendro V, Busquets S, Lopez-Soriano FJ. The pharmacological treatment of cachexia. Curr Drug Targets 2004;5: Dewys WD, Begg C, Lavin PT, Band PR, Bennett JM, Bertino JR, Cohen MH, Douglass HO, Jr., Engstrom PF, Ezdinli

8 Asp et al. 763 EZ, Horton J, Johnson GJ, et al. Prognostic effect of weight loss prior to chemotherapy in cancer patients. Am J Med 1980;69: Warren S. The immediate causes of death in cancer. Am J Med Sci 1932;184: Tayek JA. A review of cancer cachexia and abnormal glucose metabolism in humans with cancer. J Am Coll Nutr 1992;11: Rohdenburg GL, Bernhard A, Krehbiel O. Sugar tolerance in cancer. J Am Med Assoc 1919;72: Beck SA, Tisdale MJ. Effect of insulin on weight loss and tumour growth in a cachexia model. Br J Cancer 1989;59: Lazarus DD, Kambayashi T, Lowry S, Strassmann G. The lack of an effect by insulin or insulin-like growth factor-1 in attenuating colon-2-mediated cancer cachexia. Cancer Lett 1996;103: Moley JF, Morrison SD, Norton JA. Insulin reversal of cancer cachexia in rats. Cancer Res 1985;45: Moley JF, Morrison SD, Norton JA. Preoperative insulin reverses cachexia and decreases mortality in tumor-bearing rats. J Surg Res 1987;43: Moley JF, Morrison SD, Gorschboth CM, Norton JA. Body composition changes in rats with experimental cancer cachexia: improvement with exogenous insulin. Cancer Res 1988;48: Peacock JL, Norton JA. Impact of insulin on survival of cachectic tumorbearing rats. J Parenter Enteral Nutr 1988; 12: Vona-Davis L, Howard-McNatt M, Rose D. Adiposity, type 2 diabetes and the metabolic syndrome in breast cancer. Obes Rev 2007;8: Goodwin P, Ennis M, Pritchard K, Trudeau M, Koo J, Madarnas Y, Hartwick W, Hoffman B, Hood N. Fasting insulin and outcome in early-stage breast cancer: results of a prospective cohort study. J Clin Oncol 2001;20: Ma J, Li H, Giovannucci E, Mucci L, Qiu W, Nguyen P, Gaziano J, Pollak M, Stampfer M. Prediagnostic body-mass index, plasma C-peptide concentration, and prostate-specific mortality in men with prostate cancer: a long-term survival analysis. Lancet Oncol 2008;9: Cersosimo E, Pisters PW, Pesola G, Rogatko A, Vydelingum NA, Bajorunas D, Brennan MF. The effect of graded doses of insulin on peripheral glucose uptake and lactate release in cancer cachexia. Surgery 1991;109: Dodesini AR, Benedini S, Terruzzi I, Sereni LP, Luzi L. Protein, glucose and lipid metabolism in the cancer cachexia: A preliminary report. Acta Oncol 2007;46: Wolever TM, Jenkins DJ, Jenkins AL, Josse RG. The glycemic index: methodology and clinical implications. Am J Clin Nutr 1991; 54: Acharyya S, Butchbach ME, Sahenk Z, Wang H, Saji M, Carathers M, Ringel MD, Skipworth RJ, Fearon KC, Hollingsworth MA, Muscarella P, Burghes AH, et al. Dystrophin glycoprotein complex dysfunction: a regulatory link between muscular dystrophy and cancer cachexia. Cancer Cell 2005;8: Livak KJ, Schmittgen TD. Analysis of relative gene expression data using realtime quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 2001;25: Mammucari C, Milan G, Romanello V, Masiero E, Rudolf R, Del PP, Burden SJ, Di LR, Sandri C, Zhao J, Goldberg AL, Schiaffino S, et al. FoxO3 controls autophagy in skeletal muscle in vivo. Cell Metab 2007;6: Sandri M, Sandri C, Gilbert A, Skurk C, Calabria E, Picard A, Walsh K, Schiaffino S, Lecker SH, Goldberg AL. Foxo transcription factors induce the atrophyrelated ubiquitin ligase atrogin-1 and cause skeletal muscle atrophy. Cell 2004;117: Stitt TN, Drujan D, Clarke BA, Panaro F, Timofeyva T, Kline WO, Gonzalez M, Yancopoulos GD, Glass DJ. The IGF-1/ PI3K/Akt pathway prevents expression of muscle atrophy-induced ubiquitin ligases by inhibiting FOXO transcription factors. Mol Cell 2004;14: Wendel AA, Purushotham A, Liu LF, Belury MA. Conjugated linoleic acid fails to worsen insulin resistance but induces hepatic steatosis in the presence of leptin in ob/ob mice. J Lipid Res 2008;49: Fearon KC, Voss AC, Hustead DS. Definition of cancer cachexia: effect of weight loss, reduced food intake, and systemic inflammation on functional status and prognosis. Am J Clin Nutr 2006;83: Tanaka Y, Eda H, Tanaka T, Udagawa T, Ishikawa T, Horii I, Ishitsuka H, Kataoka T, Taguchi T. Experimental cancer cachexia induced by transplantable colon 26 adenocarcinoma in mice. Cancer Res 1990;50: Ryall JG, Lynch GS. The potential and the pitfalls of beta-adrenoceptor agonists for the management of skeletal muscle wasting. Pharmacol Ther 2008;120: Lehrke M, Lazar MA. The many faces of PPARgamma. Cell 2005;123: Guerre-Millo M. Adiponectin: an update. Diabetes Metab 2008;34: Musi N, Goodyear LJ. Insulin resistance and improvements in signal transduction. Endocrine 2006;29: Gomes MD, Lecker SH, Jagoe RT, Navon A, Goldberg AL. Atrogin-1, a musclespecific F-box protein highly expressed during muscle atrophy. Proc Natl Acad Sci USA2001;98: Proud CG. Regulation of protein synthesis by insulin. Biochem Soc Trans 2006;34: Acharyya S, Ladner KJ, Nelsen LL, Damrauer J, Reiser PJ, Swoap S, Guttridge DC. Cancer cachexia is regulated by selective targeting of skeletal muscle gene products. J Clin Invest 2004;114: Patel L, Pass I, Coxon P, Downes CP, Smith SA, Macphee CH. Tumor suppressor and anti-inflammatory actions of PPARgamma agonists are mediated via upregulation of PTEN. Curr Biol 2001;11: Kim KY, Cho HS, Jung WH, Kim SS, Cheon HG. Phosphatase and tensin homolog deleted on chromosome 10 suppression is an important process in peroxisome proliferator-activated receptor-gamma signaling in adipocytes and myotubes. Mol Pharmacol 2007;71: