Decreased food intake contributes to elevated insulin-responsiveness in pre-clinical cancer cachexia

Cancer cachexia, muscle, insulin sensitivity, glucose metabolism, food restriction 13 Highlights 14 • C26 and KPC cancer-induced weight loss (cachexia) increases glucose tolerance in mice 15 • Insulin responsiveness is increased in cachectic, but not in non-cachectic, tumor–bearing mice. 16 • Lowered food intake drives elevated muscle insulin responsiveness in cachectic mice 17 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint

18 Purpose: Cancer cachexia is a life-threatening complication of advanced malignancies, driven 19 by anorexia and profound systemic metabolic reprogramming. Insulin action in skeletal muscle 20 is markedly impaired in patients with cancer and may contribute directly to cachexia pathogen-21 esis. However, the interplay between reduced nutrient intake and cancer -associated metabolic 22 rewiring in cachexia remains poorly defined. Clarifying this relationship is essential for iden-23 tifying the fundamental drivers of cachexia and for developing effective therapeutic strategies. 24

We assessed metabolic rewiring by glucose tolerance test and isotopic tracers to 25 determine muscle insulin-stimulated glucose uptake in male cachectic and non-cachectic C26- 26 and KPC-tumor-bearing, as well as mice towards C26 cachectic mice. 27

Cachectic C26 -tumor-bearing mice displayed reduced body weight, lean , and fat 28 mass, and food intake (-20%, -15%, -75%, -40%, respectively). Cachectic C26- and KPC-tu-29 mor mice showed improved glucose tolerance compared to non -cachectic mice, correlating 30 inversely with tumor size. Ex vivo insulin -stimulated glucose uptake was elevated in soleus 31 (+78%) and extensor digitorum longus (+35%) muscle from cachectic C26 -cancer mice com-32 pared to non-cachectic and control mice. This increase was associated with enhanced AKT 33 signaling. This was phenocopied in pair-fed non-tumor-bearing mice to match the food intake 34 of cachectic mice , where glucose tolerance, insulin-stimulated glucose uptake ex vivo , and 35 AKT signaling were all enhanced by food restriction. 36

Our findings suggest that enhanced skeletal muscle insulin responsiveness in 37 cachectic tumor-bearing mice is due to anorexia-induced adaptations, highlighting AKT sig-38 naling as a key node connecting nutrient status to muscle glucose metabolism in cancer ca-39 chexia. 40 41 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint 1. Introduction 42 Cancer cachexia is a multifactorial metabolic syndrome characterized by involuntary loss of 43 skeletal muscle mass, with or without concomitant fat loss, due to reduced appetite and sys-44 temic metabolic rewiring. Affecting up to 80% of patients with advanced cancer, cachexia con-45 tributes substantially to morbidity, treatment intolerance, mortality , and impairs quality of 46 life[1,2]. Systemic metabolic rewiring in cancer involves reduced i nsulin responsiveness in 47 skeletal muscle[3,4], possibly contributing to the pathogenesis of cachexia. However, the in-48 terplay between reduced food intake and cachexia -associated metabolic reprogramming re-49 mains poorly defined. Yet, it is central in delineating the mechanisms that drive cachexia and 50 in identifying treatment targets. 51 Skeletal muscle, being a major site of glucose disposal under insulin -stimulated conditions, is 52 likely to play a central role in the metabolic disturbances associated with cancer. Clinically, 53 cancer is often associated with systemic insulin resistance and impaired glucose tolerance in 54 patients[5–7] to a degree that is similar to the insulin resistance observed in patients with type 55 2 diabetes[8]. Accordingly, patients with cancer have a higher risk of developing type 2 diabe-56 tes after their diagnosis[9,10]. In the preclinical setting, weight-stable LLC-tumor-bearing mice 57 display impaired insulin tolerance and skeletal muscle insulin resistanc e[11–13]. Because in-58 sulin is an anabolic hormone, insulin resistance might contribute to cancer cachexia. This has 59 been indicated by some preclinical evidence showing that insulin resistance precedes cachexia 60 and that cachexia can be delayed by insulin sensitizers in a rat model of cachexia[14,15]. More-61 over, patients with cancer and diabetes display a greater weight loss compared to patients with 62 cancer but without diabetes[16]. Yet, cachectic patients did not exhibit insulin resistance , in 63 contrast to the general cancer patient population[7] and non-cachectic preclinical mouse tumor-64 bearing mice[11,12,17], as indicated by a recent systematic review and meta -analysis [18]. 65 Moreover, at basal conditions, cachectic mouse muscle was recently shown to have glucose 66 hypermetabolism[19]. Thus, it remains unclear whether cachexia is associated with altered in-67 sulin responsiveness, and potential mechanisms linking cancer cachexia with insulin action are 68 unresolved. 69 A key contributor to cachexia that could influence insulin action is the reduced food intake 70 associated with cancer cachexia . Anorexia in patients and tumor -bearing mice exacerbates 71 weight loss by limiting energy and nutrient availability, accelerating muscle and fat wasting , 72 but it is not the sole driver of cachexia [20–22]. However, reduced caloric intake itself can 73 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint significantly affect glucose metabolism[23–25], which could be a potential confounding factor 74 in interpreting metabolic outcomes in pre-clinical cancer cachexia models and in the clinic. 75 In this study, we aimed to investigate whole-body and skeletal muscle-specific glucose metab-76 olism across two models of pre-clinical cancer cachexia and how short-term cachexia-mimick-77 ing food restriction may be a confounding factor in the interpretation of the metabolic effect s 78 of cachexia. 79 2. Methods 80 2.1 Animal experiments 81 All experiments were approved by the Danish Animal Experimental Inspectorate (License: 82 2021-15-0201-01104 and 2021 -15-0201-01085). Mice were maintained under a 12:12 h 83 light/dark photocycle at ambient (22 ± 1 °C) temperature with nesting material. All mice re-84 ceived a rodent chow diet (Altromin no. 1324; Chr. Pedersen, Denmark) and water ad libitum. 85 Male mice were group -housed for cancer studies, and single -housed for the food -restriction 86 study. 87 2.2 Tumor-bearing mouse models 88 12-week-old male Balbc/J mice (Janvier, Denmark) were used for Colon26-cancer (C26) stud-89 ies, and 14-week-old male C57BL/6JBomTac (Taconic, Denmark) were used for KPC studies. 90 All mice were group-housed in pairs and randomly assigned to either the cancer or control 91 group. C26 adenocarcinoma cells ( kind gift from Dr. Adam Rose ) were maintained at 37°C, 92 5% CO2, in RPMI 1640 medium (Gibco, #11875093) supplemented with 10 % fetal bovine 93 serum (FBS, Gibco #F0804) and 1% antibiotic-antimycotic (ThermoFisher, #15140122). KPC 94 cells (CancerTools, #153474) were maintained at 37°C, 5% CO 2, in DMEM (Gibco, 95 #11965092) supplemented with 10 % FBS (Gibco #F0804) and 1% antibiotic -antimycotic 96 (ThermoFisher, #15140122). All mice were shaved on the right flank the day before cancer 97 inoculation. On the day of inoculation, C26 or KPC cells were trypsinized and wash ed twice 98 with PBS (centrifuged for 3 min at 1,600 g), followed by resuspension in PBS. Next, all mice 99 were injected subcutaneously into the flank with PBS with or without C26 or KPC cells for a 100 final concentration of 5x105 cells or 1x106, respectively. Mice were terminated when reaching 101 humane endpoints following the tumor guidelines from the Danish Animal Experimental In-102 spectorate (tumors >14 mm - average length and width or ulcerations >5 mm). C26-cancer 103 mice were divided into non-cachexia (tumor with no weight loss) or cachexia (tumor with 104 weight loss) groups. 105 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint 2.3 Food restriction, pair-fed to cachectic C26 tumor-bearing mice 106 12-week-old male Balbc/J mice (Janvier, Denmark) were used for the pair-feeding study. All 107 mice were single-housed, and food intake was recorded daily during the last week of the ex-108 periment. The last three days, the food intake was reduced by approximately 30%, correspond-109 ing to what was observed in cachectic C26-cancer mice. 110 2.4 Body composition 111 Body mass composition was assessed by nuclear magnetic resonance using an EchoMRI ™ 112 (USA) or a Bruker LF90II Body Composition Analyser (Bruker) for total body, lean, and fat 113 mass. All mice were MRI scanned before cancer cell inoculation and the day before termina-114 tion. The tumor mass from dissections was subtracted from the total body weight and lean 115 mass. 116 2.5 Grip strength 117 Grip strength was assessed in the week before termination. Grip strength assessment was per-118 formed using a grip meter (Bioseb). The mice were held by the base of the tail and briefly left 119 to hover over the grid and allowed to grab with all four paws . Once grabbed, the mice were 120 gently pulled back in a horizontal position. This was repeated three times for each mouse, and 121 the maximal recording was used as final force readout. 122 2.6 Glucose tolerance test 123 All mice were fasted for 4 hours before the glucose tolerance test (GTT). D-mono-glucose (2g 124 kg−1 bw) was administered via intraperitoneal injections, and blood was collected from the tail 125 vein. Blood glucose levels were determined at timepoints 0, 20, 40, 60, 90, and 120 minutes 126 using a glucometer (Bayer Contour, Bayer, Switzerland). Blood aliquots from timepoints 0 and 127 20 min were centrifuged at 13,000 rpm for 5 minutes at 4 °C, and plasma was collected and 128 frozen in liquid nitrogen. Insulin levels were measured using the Mouse Ultrasensitive Insulin 129 ELISA (#80-INSMSU-E01ALPCO Diagnostics, USA). 130 2.7 Ex vivo insulin-stimulated glucose uptake 131 Soleus and extensor digitorum longus (EDL) muscles were rapidly isolated from anesthetized 132 mice, and non-absorbable 4–0 silk suture loops (Look SP116, Surgical Specialties Corporation) 133 were attached at both ends as previously described[12]. In brief, the muscles were placed in a 134 DMT Myograph system (820MS; Danish Myo Technology, Hinnerup, Denmark) and placed 135 under resting tension (4 -5 mN at 30°C) and oxygenated (95% O 2, 5% CO 2) Krebs -Ringer 136 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint buffer. Isolated muscles were pre-incubated in Krebs-Ringer buffer for 5 minutes, followed by 137 20 minutes of maximal insulin stimulation (60 nM). During the last 10 minutes, 0.75 µCi/mL 138 [3H]-2DG and 0.225 µCi/mL [14C] -Mannitol were added (including insulin) . Following the 139 insulin stimulation, muscles were immediately rinsed in ice-cold saline and snap-frozen. Mus-140 cle-specific 2DG-uptake was measured on tissue lysates[26]. 141 2.8 Lysate preparation and immunoblotting 142 Muscle tissues were homogenized in a modified GSK3 -buffer and prepared for immunoblot-143 ting as previously described in detail[17]. 144 2.9 Statistical analysis 145 Data are presented as mean ± SEM and individual data points (when applicable) and analyzed 146 using GraphPad Prism 10. Statistical tests were performed using paired/non‐paired t‐tests or 147 repeated/no‐repeated two‐way analysis of variance (ANOV A) as applicable. Multiple‐repeated 148 two‐way ANOV As were performed in analyses, including all experimental groups testing for 149 the effect of C26, KPC, or food restriction. Sidak's post hoc test was performed when ANOV A 150 revealed significant main effects and interactions. The significance level was set at α = 0.05. 151 3. Results 152 3.1 C26-cachectic mice exhibit improved glucose tolerance and elevated skeletal muscle 153 insulin-responsiveness ex vivo 154 We first determined classical manifestations of cachexia in the C26 -cancer model to establish 155 the effects of cachexia on glucose metabolism. Mice were divided into a control group (n=12), 156 a C26 non-cachectic group (n=6, 10% 157 weight loss) based on the weight loss at termination (Fig. 1A). The cachectic mice had a 20% 158 reduction in body mass 22 days after inoculation, while non-cachectic mice maintained their 159 body mass. Control mice had a minor body mass gain of 5% (Fig. 1B). Tumor size was approx-160 imately 80% larger in cachectic mice compared to non-cachectic mice (Fig. 1B). Similarly, 161 cachectic mice displayed a 15% reduction in lean body mass, whereas both non-cachectic and 162 control mice maintained lean body mass (Fig. 1C). Cachectic mice exhibited a 75% reduction 163 in fat mass, while non-cachectic (p=0.0510) and control mice exhibited a ~20% increase in fat 164 mass (Fig. 1D). 165 At termination, we weighed selected muscles and adipose tissues. Only cachectic mice had 166 lower weights for gastrocnemius (G ast, -17%), tibialis anterior (T . anterior , -18%), and 167 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint perigonadal adipose tissue (pWAT, -95%) compared to control mice. Heart tissue weights were 168 reduced in both tumor-bearing groups by 8% and 14% for non-cachectic and cachectic mice, 169 respectively. Brown adipose tissue (BAT) was 17% and 47% lower for non-cachectic and ca-170 chectic mice, respectively, compared to control mice. The spleen was similarly increased in 171 both tumor-bearing groups by approximately 100% (Fig. 1E), indicative of similar systemic 172 inflammatory burden despite marked differences in tumor mass and cachectic phenotype. De-173 spite no significant reduction in muscle weights, we detected a 17% reduction in grip strength 174 in non-cachectic mice, while cachectic mice showed a reduction of 27% compared to control 175 mice (Fig. 1F). Food intake was assessed three days prior to termination and a reduction (~40% 176 vs. controls) was observed exclusively in cachectic mice (Fig. 1G). 177 Considering the reports of glucose intolerance in mice with cancer[15], we performed a glucose 178 tolerance test 14 days after cancer inoculation to assess the effects of cachexia. Dividing C26 179 tumor-bearing mice into non-cachectic mice (tumor volume 700 mm³), we were surprised to observe increased glucose tolerance in the 181 cachectic mice compared to control and non-cachectic mice (Fig. 1H), contrasting previous 182 findings[11,27]. Glucose excursions during the GTT were inversely correlated with tumor vol-183 ume, potentially reflecting cachexia-associated metabolic remodeling or tumor glucose seques-184 tration (Fig. 1I). The observed changes in glucose tolerance were independent of glucose-stim-185 ulated plasma insulin (Fig. 1J). 186 Since skeletal muscle is the primary site for glucose disposal, we proceeded to determine insu-187 lin-stimulated glucose uptake in soleus and EDL ex vivo. In the soleus muscles, insulin led to 188 increased glucose uptake in all groups (Fig. 1K). However, cachectic mice showed ~80% 189 higher response to insulin non-cachectic and control mice (Fig. 1K). Similarly, in the EDL 190 muscle, cachectic mice shoed an 35-40% elevation in insulin response compared to non-ca-191 chectic and control mice (Fig. 1L). Thus, cachexia was associated with elevated insulin respon-192 siveness towards glucose uptake in mouse muscle. 193 Together, our results demonstrate that cachexia in C26 tumor-bearing mice was associated with 194 improved whole-body glucose tolerance and muscle insulin-stimulated glucose uptake concur-195 rently with muscle and fat mass loss. 196 3.3 The cachectic KPC tumor-bearing mice also display markedly elevated glucose toler-197 ance 198 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint Having observed elevated glucose tolerance in the C26 cachectic model, we next assessed glu-199 cose tolerance in another pre-clinical cachectic model, the KPC model (Fig. 1M). KPC tumor-200 bearing mice had an 8% lower body mass at termination compared to control mice, indicative 201 of cachexia (Fig. 1 N). Similarly, fat mass was 25% lower in cachectic mice at termination 202 compared to control mice (Fig. 1 O). Measurements of tissues at termination revealed that ca-203 chectic KPC mice had lower muscle and adipose tissue weight compared to control mice (Gast 204 -10% p=0.081, T. anterior -9%, quadriceps -11%, pWAT -30%, BAT -28%), while the spleen 205 weight was increased by 75% in cachectic mice (Fig. 1 P). On day 30 after inoculation, we 206 performed a glucose tolerance test, and similar to the observation in C26 tumor-bearing mice, 207 and in line with previous literature of the KPC model [28], KPC tumor-bearing mice had a 208 marked improvement in glucose tolerance (Fig. 1Q). The change in glucose tolerance was in-209 dependent of glucose-stimulated plasma insulin (Fig. 1R). Thus, two different pre-clinical ca-210 chectic mouse models exhibit ed elevated glucose tolerance , which could be, at least in part, 211 driven by elevated insulin-responsiveness in skeletal muscle of cachectic mice. 212 3.4 Elevated Akt signaling coincides with higher insulin-stimulated glucose uptake in C26 213 cachectic mice 214 After demonstrating an evidently higher muscle insulin responsiveness, we proceede d to in-215 vestigate intramyocellular insulin signaling (Fig. 2A). We found a reduction in protein contents 216 of Glucose transporter 4 (GLUT4) (-20%), Glycogen Synthase (-13%, p=0.057), Akt2 (-22%, 217 p=0.057), p70S6K ( -17%), and rS6 ( -29%, p=0.068) in C26 cachectic compared to control 218 mouse soleus muscle. In contrast, Hexokinase II, Pyruvate Dehydrogenase, TBC1D4, GSK3β, 219 and PRAS40 were unaffected (Fig. 2B). It is well known that insulin acts as an anabolic hor-220 mone to activate the Akt signaling pathway stimulating glucose uptake and glycogen synthesis 221 in skeletal muscle[29]. Thus, we investigated Akt phosphorylation and downstream targets in 222 response to insulin. In control and C26 non-cachectic mice, we observed a 2-fold and 2.5-fold 223 increase in phosphorylated (p) AKTthr308, respectively, in the insulin-stimulated soleus muscle. 224 Yet, in cachectic mice, there was a n even higher 4.4-fold increase by insulin compared to the 225 non-stimulated muscle (Fig. 2C). Correspondingly, pAKTser473 was increased by 3.8-fold upon 226 insulin stimulation in control and non-cachectic mice, while cachectic mice exhibited a remark-227 able 8.5 -fold increase (Fig. 2 D). Yet we observed a similar i ncrease (3.3- and 3.9 -fold) in 228 pTBC1D4thr642, a downstream target of AKT, in insulin-stimulated soleus muscle from non-229 cachectic and cachectic mice, respectively, compared to a 2.4 -fold increase in control mice 230 (Fig. 2E). 231 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint Thus, the elevated insulin-responsiveness in cachectic muscle exists despite a reduction of the 232 main glucose transporter (GLUT4), likely circumvented by elevated AKT signaling. 233 3.5 Glycogen and mTORC1 signaling are retained in insulin -stimulated soleus muscle 234 from C26 cachectic mice 235 Akt phosphorylates several downstream targets. In the insulin-stimulated soleus muscle from 236 cachectic mice, we found a higher increase in pGSKser9 (3.3-fold, Fig. 2F) and pGSKser21 (2.6-237 fold, Fig. 2G ) compared to control and non-cachectic mice (2.0 - and 1.8-fold, respectively). 238 PRAS40 is a direct downstream target of AKT during insulin stimulation[30,31]. We detected 239 an elevated abundance of pPRAS40 thr246 in insulin-stimulated cachectic soleus muscles (4.8 -240 fold), compared to both control and non-cachectic mice (3.3-fold) (Fig. 2H). Insulin-induced 241 activation of the mTORC1 signaling pathway has been widely implicated in the promotion of 242 muscle growth. Despite the presence of cachexia, mTORC1 signaling remained intact, shown 243 by increased p-p70S6Kthr389 (3.3-fold) and prS6ser235-236 (1.6-fold), which contrast previous re-244 sults in the cachectic ApcMin/+ mouse model showing reduced mTORC1 signaling after a glu-245 cose challenge[32]. Control mice exhibited increased p -p70S6Kthr389 (2.5-fold), but not 246 prS6ser235-236 (Fig. 2I and J). The parallel activation of PRAS40, p70S6K, and rS6 suggests that 247 anabolic signaling downstream of mTOR remained partially intact despite muscle wasting, in-248 dicating that cachectic muscle retains the capacity for insulin-induced anabolic responses under 249 these conditions. 250 3.6 Three days of food restriction lowers body weight, muscle mass, and fat mass in mice 251 We next sought to delineate the underlying contributors to enhanced insulin responsiveness . 252 Anorexia is a key contributor to the manifestations of cachexia, and lowered food intake is 253 known to impact metabolism[23–25]. We therefore executed a 3-day food restriction study ( -254 30%, mimicking the reduced food intake in C26 cachetic mice) in male BALB/cJ mice to assess 255 whether reduced food intake could explain the improved insulin responsiveness observed in 256 cachectic mice (Fig. 3A). First, we measured body weight daily and discovered that food-re-257 stricted mice lost 1.2g (-3,9%) body weight on average after 3 days, compared to control mice 258 who gained 0.7g (2.1%) body weight (Fig. 3B). We measured tissue weights and observed a 259 reduction in tissue weights of all skeletal muscles and adipose tissues upon food restriction (TA 260 -4%, Gast -7%, Quad -7%, iWAT -34%, pWAT -22%, BAT -23%, spleen -16%), while the heart 261 weight was similar (Fig. 3C). Despite lower muscle weights in food -restricted mice, grip 262 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint strength was unaffected (Fig. 3D) , indicating that anorexia may underlie some body weight 263 loss but does not compromise muscle function in tumor-bearing mice. 264 3.7 Glucose tolerance, muscle ex vivo insulin-stimulated glucose uptake and Akt signaling 265 in increased upon food restriction 266 To investigate if anorexia may underlie some of the improvements in glucose tolerance and 267 muscle insulin responsiveness, observed in cachetic C26 mice, we determined glucose toler-268 ance and muscle insulin reonsiveness in 3-day food-restricted Balb/c mice. Interestingly, food-269 restriction markedly improved glucose tolerance with a 56% reduction in iAUC compared to 270 ad libitum fed control mice (Fig. 3D). We then proceeded to determine ex vivo insulin-stimu-271 lated glucose uptake in the soleus muscles. We were intrigued to see that food -restricted mice 272 had a higher insulin response (30%) compared to control mice (Fig. 3F) . In line with our ob-273 servations in C26 cachectic mice, AKT signaling was elevated in food-restricted mice. This 274 was evidenced by, pAKTthr308 increasing 8-fold in response to insulin in food-restricted mice 275 compared to 5-fold in control mice (Fig. 3G). Likewise, pAKTser473 increased by 7-fold in re-276 sponse to insulin in food-restricted mice compared to 4-fold in control mice (Fig. 3H). In ad-277 dition, we found a 6-fold increase by insulin in pTBC1D4thr642 in food-restricted mice, which 278 was only increased by 3-fold in control mice (Fig. 3I). 279 Taken together, our findings show that reduced food intake is a potential driver of enhanced 280 glucose tolerance and skeletal muscle insulin responsiveness in cachectic C26 -cancer mice, 281 associated with enhanced Akt signaling. 282 4. Discussion 283 In this study, we investigated glucose metabolism and skeletal muscle insulin responsiveness, 284 using two distinct tumor models , C26 and KPC, and in pair-fed food-restricted mice. Our re-285 sults highlight three key discoveries. First, cachexia was associated with increased glucose tol-286 erance in both C26 - and KPC tumor-bearing mice. Second, cachectic, but not non -cachectic, 287 C26 mice exhibit elevated skeletal muscle insulin responsiveness associated with increased Akt 288 signaling. Third, food restriction , mimicking the lower food intake in C26 cachectic mice, 289 caused a metabolic phenotype phenocopying cachexia. These findings suggest that d ecreased 290 food intake contributes to elevated insulin -responsiveness in cancer cachexia, illuminating an 291 intriguing interplay between food intake and cachexia-associated metabolic reprogramming. 292 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint Our first key finding demonstrates that cachectic C26-tumor-bearing mice exhibited improved 293 glucose tolerance compared to non-cachectic mice, which was corroborated in cachectic KPC-294 tumor-bearing mice. Here, we identified a relationship between tumor burden and glucose tol-295 erance, with cachectic mice harboring significantly larger tumors. It is plausible that larger 296 tumors dispose more glucose during hyperglycemia, contributing to improvements in glucose 297 tolerance. A marked metabolic change in glucose utilization is supported by a recent study, 298 using isotopic tracers in C26 cachectic mice, showing that one-carbon metabolism was a tissue-299 overarching pathway characterized in wasting leading to glucose hypermetabolism[19]. How-300 ever, our second key finding highlights that cancer-associated cachexia is also accompanied by 301 elevated skeletal muscle insulin responsiveness . Thus, these data suggest that changes within 302 the skeletal muscle likely contribute to the enhanced glucose tolerance in cachectic mice. 303 In the soleus and EDL muscles, insulin responsiveness was markedly greater in cachectic com-304 pared to both control and non-cachectic tumor-bearing mice. This was observed despite a re-305 duction in protein content of the key glucose transporter during insulin stimulation, 306 GLUT4[33]. Yet, increased GLUT4 translocation to the muscle membrane was indicated by a 307 substantially elevation of phosphorylation AKT (Thr308 and Ser473) in response to insulin . 308 Thus, the current data support the idea of cancer cachexia being associated with elevated glu-309 cose tolerance and insulin responsiveness, although the molecular mechanisms remain to be 310 defined. 311 Our last key finding was that reduced food intake, mimicking food intake in cachexia, could 312 contribute to the increased insulin responsiveness in our cachectic mice as a compensatory 313 adaptation to nutrient limitation arising from anorexia and tissue loss. In our C26 model, only 314 cachectic mice had significantly reduced food intake (~40 %) with severe lean and fat mass 315 loss. In this context, improved insulin-stimulated glucose uptake in muscle may reflect height-316 ened responsiveness to insulin when exogenous glucose availability is limited, a phenomenon 317 reminiscent of the known effects of caloric restriction[34–36]. Notably, while improved insulin 318 sensitivity may initially represent a compensatory mechanism to preserve energy efficiency 319 during anorexia, chronic activation of this pathway in the face of cataboli c conditions such as 320 cachexia fails to prevent muscle loss. Underscoring this is our finding that cachexia-mimicking 321 food restriction alone did not alter muscle strength measured by grip strength in the current 322 study. Decreased grip strength was only observed in tumor-bearing mice and in both cachectic 323 and non-cachectic mice, in line with another study showing decreased muscle function in C26 324 before cachexia [37]. Accordingly, it is also known that decreased energy intake in cancer is 325 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint not the sole driver of cachexia [38,39]. Thus, while decreased food intake likely affects the 326 systemic metabolism in cancer cachexia, the pathology of cachexia is complex and involves 327 multiple factors[38]. 328 5. Conclusion 329 In summary, our study reveals that improved glucose tolerance and enhanced skeletal muscle 330 insulin responsiveness in cachectic C26-cancer mice are likely affected by anorexia-induced 331 adaptations. Our data highlight enhanced Akt signaling as a key node linking nutrient status to 332 skeletal muscle glucose metabolism during cancer progression. These findings challenge the 333 assumption that cachexia universally associates with insulin resistance and underscore the need 334 to consider both cancer type, cancer progression, and food intake when interpreting metabolic 335 outcomes in cancer. 336

337 We thank our colleagues at the Molecular Metabolism in Cancer and Aging Group, Faculty of 338 Health and Medical Sciences, University of Copenhagen, for profitable discussions on this 339 topic. We acknowledge the Rodent Metabolic Phenotyping Platform (RMPP) , Novo Nordisk 340 Foundation Center for Metabolic Research, University of Copenhagen, for the use of their fa-341 cilities. We also acknowledge Simon Bech Petersen, Department of Biomedical Sciences, Fac-342 ulty of Health and Medical Sciences, University of Copenhagen, for his assistance. Illustrations 343 were generated using Biorender.com (RRID:SCR_018361) and ©Inkscape. Graphs were gen-344 erated with GraphPad PRISM (RRID:SCR_002798). 345 CRediT: 346 Conceptualization: EF, LS, SHR 347 Data curation: EF, KWP, ZKJO, PP, JRK, LS, SHR 348 Formal analysis: EF, SHR 349 Funding acquisition: LS, SHR 350 Investigation: EF, KWP, ZKJO, PP, JRK, LS, SHR 351 Methodology: LS, SHR 352 Project administration: EF, LS, SHR 353 Resources: LS, SHR 354 Supervision: LS, SHR 355 Validation: EF, SHR 356 Visualization: EF, LS, SHR 357 Writing – original draft: EF, LS, SHR 358 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint Writing – review and editing: all authors 359 360 Funding: 361 362 SHR was funded by Independent Research Fund Denmark (2030 -00007A), the Lundbeck Foundation 363 (R380-2021-1451) and the Novo Nordisk Foundation ( NNF0101703). TCPP was funded by the 364 Lundbeck Foundation (R449-2023-1468) and the Novo Nordisk Foundation (NNF24OC0088823). JRK 365 was funded by Novo Nordisk Foundation ( 17SA0031406) and Independent Research Fund Denmark 366 (9058-00047B). 367 Competing interests: 368 KWP, JRK and LS are founders of Hercu ApS; no competing interests are declared. 369 Declaration of generative AI and AI-assisted technologies in the writing process. 370 Statement: During the preparation of this work, the authors used ChatGPT to do minor corrections and 371 shorten sentences. After using this tool/service, the authors reviewed and edited the content as needed 372 and we take full responsibility for the content of the published article. 373 Figure 1: C26 and KPC tumor-bearing mice exhibit improved glucose tolerance despite lowered 374 muscle and fat mass. A) Study design of the C26-cancer study. Mice were divided into three groups. 375 1) Control (n=12), 2) pre-cachectic (tumor, no weight loss, n=6), and 3) cachectic (tumor, weight loss, 376 n=9). B) Body mass and tumor mass. C) Magnetic Resonance Imaging -derived lean body mass. D) 377 Magnetic Resonance Imaging-derived fat mass. E) Tissue weights at termination from gastrocnemius 378 (Gast), tibialis anterior (T A), heart, perigonadal white adipose tissue (pWAT), brown adipose tissue 379 (BAT), and spleen. F) Grip strength measured on day 19-22 before termination. G) Food intake the last 380 three days before termination. H) Glucose tolerance test on day 14 after inoculation. C26-tumor-bearing 381 mice were divided into non-cachectic small tumors (700 mm3). 382 I) Pearson correlation analysis of tumor size and incremental area under the curve. J) Plasma insulin 383 levels from time -point 0 and 20 min during the glucose tolerance test. Of not, the sample number is 384 lower due to lack of plasma collected during the tolerance test. K) Ex vivo insulin-stimulated (60 nM) 385 glucose uptake in the soleus muscle. L) Ex vivo insulin-stimulated (60 nM) glucose uptake in the EDL 386 muscle. M) Study design of the KPC-cancer study. N) Body mass and tumor mass. O) Magnetic Reso-387 nance Imaging-derived fat mass. P) Tissue weights at termination from Gast, TA, quadriceps (Quad), 388 pWAT, BAT, and spleen. Q) Glucose tolerance test on day 30 after inoculation. R) Plasma insulin levels 389 from time-point 0 and 20 min during the glucose tolerance test. Abbreviation: C26 = Colon26, Con = 390 control, CC = cancer cachexia, Int = interaction, ME = main effect, AUC = area under the curve. Values 391 are shown as mean ± SEM, including individual values, and as mean ± SD when individual values are 392 not shown. Effect of C26 or KPC: *= p<0.05, **=p<0.01, ***=p<0.001. 393 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint Figure 2: Akt signaling likely drives improvements in ex vivo insulin-stimulated glucose transport. 394 A) Insulin-stimulated Akt signaling in skeletal muscle. B) Total protein content in the soleus muscle of 395 GLUT4 (Thermo Fisher Scientific , #PA1-1065), Hexokinase II ( Cell Signaling Technology, #2867), 396 Pyruvate dehydrogenase (D.G. Hardie (University of Dundee, Scotland)), Glycogen synthase (Gift from 397 Prof. Oluf Pedersen), Akt (Cell Signaling Technology, #3063), TBC1D4 (Abcam, ab189890), GSK3β 398 (BD Bioscience , # 610202), PRAS40 ( Cell Signaling Technology, #2691 ), p70S6K ( Cell Signaling 399 Technology, #2708), rS6 (Cell Signaling Technology, #2217). Protein content is displayed in arbitrary 400 units (AU). Immunoblotting of phosphorylated (p) proteins in the soleus muscle of: C) pAKTThr308 (CST, 401 cat#9271), D) pAKTSer473 (CST, cat#9271), E) pTBC1D4thr642 (Cell Signaling Technology, #8881), F) 402 pGSK3ser9 (Cell Signaling Technology, #9331), G) pGSK3ser21 (Cell Signaling Technology, #9331), H) 403 pPRAS40thr246 (Cell Signaling Technology, #2997 ), I) p-p70S6Kthr389 (Cell Signaling Technology, 404 #9205), J) prS6ser235-236 (Cell Signaling Technology, #2211). Representative blots are represented with 405 Coomassie staining as a loading control. V alues are shown as mean ± SEM, including individual values. 406 Effect of C26: *= p<0.05, **=p<0.01, ***=p<0.001. 407 Figure 3: Three days of food restriction in mice improves glucose tolerance, ex vivo insulin-stim-408 ulated glucose transport in soleus, and enhances Akt signaling. A) Study design of the food re-409 striction study. B) Body weight measured during the food restriction period. C) Tissue weights at ter-410 mination from TA, Gast, Quad, heart, inguinal white adipose tissue (iWAT), pWA T, BAT, and spleen. 411 D) Grip strength measured after 3 days of food restriction. G) Glucose tolerance test and incremental 412 area under the curve (iAUC) after three days of food restriction. F) Ex vivo insulin-stimulated (60 nM) 413 glucose uptake in the soleus muscle. Immunoblotting of phosphorylated proteins in the soleus muscle 414 of: C) pAKTThr308 (Cell Signaling Technology, cat#9271), D) pAKTSer473 (Cell Signaling Technology, 415 cat#9271), E) pTBC1D4thr642 (Cell Signaling Technology, #8881). Values are shown as mean ± SEM, 416 including individual values, and as mean ± SD when individual values are not shown. Effect of 3-days 417 food restriction: *= p<0.05, **=p<0.01, ***=p<0.001. 418 419 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint

420 [1] Fearon, K., Strasser, F., Anker, S.D., Bosaeus, I., Bruera, E., Fainsinger, R.L., et al., 2011. Defi 421 nition and classifi cation of cancer cachexia: an international consensus. Www.Thelan-422 cet.Com/Oncology 12: 489–95, Doi: 10.1016/S1470. 423 [2] Baracos, V .E., Martin, L., Korc, M., Guttridge, D.C., Fearon, K.C.H., 2018. Cancer-associated 424 cachexia. Nature Reviews Disease Primers, Doi: 10.1038/nrdp.2017.105. 425 [3] Dev, R., Bruera, E., Dalal, S., 2018. Insulin resistance and body composition in cancer pa-426 tients. Annals of Oncology 29: ii18–26, Doi: 10.1093/annonc/mdx815. 427 [4] Màrmol, J.M., Carlsson, M., Raun, S.H., Grand, M.K., Sørensen, J., Lang Lehrskov, L., et al., 428 2023. Insulin resistance in patients with cancer: a systematic review and meta-analysis. Acta 429 Oncologica (Stockholm, Sweden) 62(4): 364–71, Doi: 10.1080/0284186X.2023.2197124. 430 [5] Tayek, J.A., 1992. A review of cancer cachexia and abnormal glucose metabolism in humans 431 with cancer. Journal of the American College of Nutrition 11(4): 445–56, Doi: 432 10.1080/07315724.1992.10718249. 433 [6] Chiefari, E., Mirabelli, M., La Vignera, S., Tanyolaç, S., Foti, D.P., Aversa, A., et al., 2021. In-434 sulin resistance and cancer: In search for a causal link. International Journal of Molecular Sci-435 ences, Doi: 10.3390/ijms222011137. 436 [7] Màrmol, J.M., Carlsson, M., Raun, S.H., Grand, M.K., Sørensen, J., Lang Lehrskov, L., et al., 437 2023. Insulin resistance in patients with cancer: a systematic review and meta-analysis. Acta 438 Oncologica 62(4): 364–71, Doi: 10.1080/0284186X.2023.2197124. 439 [8] Tam, C.S., Xie, W., Johnson, W.D., Cefalu, W.T., Redman, L.M., Ravussin, E., 2012. Defining 440 insulin resistance from hyperinsulinemic-euglycemic clamps. Diabetes Care 35(7): 1605–10, 441 Doi: 10.2337/DC11-2339. 442 [9] Sylow, L., Grand, M.K., von Heymann, A., Persson, F., Siersma, V ., Kriegbaum, M., et al., 443 2022. Incidence of New-Onset Type 2 Diabetes After Cancer: A Danish Cohort Study. Diabe-444 tes Care 45(6): e105–6, Doi: 10.2337/DC22-0232. 445 [10] Hwangbo, Y ., Kang, D., Kang, M., Kim, S., Lee, E.K., Kim, Y .A., et al., 2018. Incidence of 446 Diabetes After Cancer Development: A Korean National Cohort Study. JAMA Oncology 4(8): 447 1099, Doi: 10.1001/JAMAONCOL.2018.1684. 448 [11] Han, X., Raun, S.H., Carlsson, M., Sjøberg, K.A., Henriquez-Olguín, C., Ali, M., et al., 2020. 449 Cancer causes metabolic perturbations associated with reduced insulin-stimulated glucose up-450 take in peripheral tissues and impaired muscle microvascular perfusion. Metabolism: Clinical 451 and Experimental 105, Doi: 10.1016/j.metabol.2020.154169. 452 [12] Raun, S.H., Knudsen, J.R., Han, X., Jensen, T.E., Sylow, L., 2022. Cancer causes dysfunc-453 tional insulin signaling and glucose transport in a muscle-type-specific manner. FASEB Jour-454 nal 36(3), Doi: 10.1096/fj.202101759R. 455 [13] Counts, B.R., Halle, J.L., Carson, J.A., 2022. Early Onset Physical Inactivity and Metabolic 456 Dysfunction in Tumor-bearing Mice Is Associated with Accelerated Cachexia. Medicine and 457 Science in Sports and Exercise 54(1): 77, Doi: 10.1249/MSS.0000000000002772. 458 [14] Asp, M.L., Tian, M., Kliewer, K.L., Belury, M.A., 2011. Rosiglitazone delayed weight loss and 459 anorexia while attenuating adipose depletion in mice with cancer cachexia. Cancer Biology & 460 Therapy 12(11): 957–65, Doi: 10.4161/cbt.12.11.18134. 461 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint [15] Asp, M.L., Tian, M., Wendel, A.A., Belury, M.A., 2010. Evidence for the contribution of insu-462 lin resistance to the development of cachexia in tumor-bearing mice. International Journal of 463 Cancer 126(3): 756–63, Doi: 10.1002/ijc.24784. 464 [16] Chovsepian, A., Prokopchuk, O., Petrova, G., Gjini, T., Kuzi, H., Heisz, S., et al., 2023. Diabe-465 tes increases mortality in patients with pancreatic and colorectal cancer by promoting cachexia 466 and its associated inflammatory status. Molecular Metabolism 73, Doi: 467 10.1016/j.molmet.2023.101729. 468 [17] Raun, S.H., Ali, M.S., Han, X., Henríquez-Olguín, C., Pham, T.C.P., Meneses-Valdés, R., et 469 al., 2023. Adenosine monophosphate-activated protein kinase is elevated in human cachectic 470 muscle and prevents cancer-induced metabolic dysfunction in mice. Journal of Cachexia, Sar-471 copenia and Muscle 14(4): 1631–47, Doi: 10.1002/jcsm.13238. 472 [18] Sørensen, J., Hammershøi, A., Màrmol, J.M., Lehrskov, L.L., Nørgaard, O., Sylow, L., 2025. 473 Revisiting insulin resistance in human cancer cachexia – a systematic review and meta-analy-474 sis, Doi: 10.1101/2025.03.28.25324822. 475 [19] Morigny, P ., V ondrackova, M., Ji, H., Brejchova, K., Krakovkova, M., Makris, K., et al., 2026. 476 Multi-omics profiling of cachexia-targeted tissues reveals a spatio-temporally coordinated re-477 sponse to cancer. Nature Metabolism, Doi: 10.1038/s42255-025-01434-3. 478 [20] Yeom, E., Yu, K., 2022. Understanding the molecular basis of anorexia and tissue wasting in 479 cancer cachexia. Experimental and Molecular Medicine: 426–32, Doi: 10.1038/s12276-022-480 00752-w. 481 [21] Ezeoke, C.C., Morley, J.E., 2015. Pathophysiology of anorexia in the cancer cachexia syn-482 drome. Journal of Cachexia, Sarcopenia and Muscle: 287–302, Doi: 10.1002/jcsm.12059. 483 [22] Amitani, M., Asakawa, A., Amitani, H., Inui, A., 2013. Control of food intake and muscle 484 wasting in cachexia. International Journal of Biochemistry and Cell Biology: 2179–85, Doi: 485 10.1016/j.biocel.2013.07.016. 486 [23] Fu, J., Liu, S., Li, M., Guo, F., Wu, X., Hu, J., et al., 2024. Optimal fasting duration for mice as 487 assessed by metabolic status. Scientific Reports 14(1), Doi: 10.1038/s41598-024-72695-3. 488 [24] Anson, R.M., Guo, Z., De Cabo, R., Iyun, T., Rios, M., Hagepanos, A., et al., 2003. Intermit-489 tent fasting dissociates beneficial effects of dietary restriction on glucose metabolism and neu-490 ronal resistance to injury from calorie intake. Proc Natl Acad Sci 100(10), Doi: 491 10.1073/pnas.1035720100. 492 [25] Kalant, N., Stewart, J., Kaplan, R., 1988. Effect of diet restriction on glucose metabolism and 493 insulin responsiveness in aging rats. Mechanisms of Ageing and Development 46(1–3): 89–494 104, Doi: 10.1016/0047-6374(88)90117-0. 495 [26] Raun, S.H., Henriquez-Olguín, C., Karavaeva, I., Ali, M., Møller, L.L. V ., Kot, W., et al., 2020. 496 Housing temperature influences exercise training adaptations in mice. Nature Communications 497 11(1): 1560, Doi: 10.1038/s41467-020-15311-y. 498 [27] Irazoki, A., Frank, E., Cam Phung Pham, T., Braun, J.L., Ehrlich, A.M., Haid, M., et al., 2025. 499 Journal of Cachexia, Sarcopenia and Muscle Housing Temperature Impacts the Systemic and 500 Tissue-Specific Molecular Responses to Cancer in Mice. Journal of Cachexia, Sarcopenia and 501 Muscle 16: 13781, Doi: 10.1002/jcsm.13781. 502 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint [28] Pasquale, V ., Dugnani, E., Liberati, D., Marra, P ., Citro, A., Canu, T., et al., 2019. Glucose me-503 tabolism during tumorigenesis in the genetic mouse model of pancreatic cancer. Acta Diabeto-504 logica 56(9): 1013–22, Doi: 10.1007/s00592-019-01335-4. 505 [29] Sylow, L., Tokarz, V .L., Richter, E.A., Klip, A., 2021. The many actions of insulin in skeletal 506 muscle, the paramount tissue determining glycemia. Cell Metabolism: 758–80, Doi: 507 10.1016/j.cmet.2021.03.020. 508 [30] Sancak, Y ., Thoreen, C.C., Peterson, T.R., Lindquist, R.A., Kang, S.A., Spooner, E., et al., 509 2007. PRAS40 Is an Insulin-Regulated Inhibitor of the mTORC1 Protein Kinase. Molecular 510 Cell 25(6): 903–15, Doi: 10.1016/j.molcel.2007.03.003. 511 [31] Haar, E. Vander., Lee, S., Bandhakavi, S., Griffin, T.J., Kim, D.-H., 2007. Insulin signalling to 512 mTOR mediated by the Akt/PKB substrate PRAS40. Nature Cell Biology 9(3): 316–23, Doi: 513 10.1038/ncb1547. 514 [32] White, J.P., Puppa, M.J., Gao, S., Sato, S., Welle, S.L., Carson, J.A., 2013. Muscle mTORC1 515 suppression by IL-6 during cancer cachexia: a role for AMPK. American Journal of Physiol-516 ogy-Endocrinology and Metabolism 304(10): E1042–52, Doi: 10.1152/ajpendo.00410.2012. 517 [33] Huang, S., Czech, M.P., 2007. The GLUT4 Glucose Transporter. Cell Metabolism 5(4): 237–518 52, Doi: 10.1016/j.cmet.2007.03.006. 519 [34] Silvestre, M.F.P., Viollet, B., Caton, P.W., Leclerc, J., Sakakibara, I., Foretz, M., et al., 2014. 520 The AMPK-SIRT signaling network regulates glucose tolerance under calorie restriction con-521 ditions. Life Sciences 100(1): 55–60, Doi: 10.1016/j.lfs.2014.01.080. 522 [35] Sequea, D.A., Sharma, N., Arias, E.B., Cartee, G.D., 2012. Calorie restriction enhances insu-523 lin-stimulated glucose uptake and akt phosphorylation in both fast-twitch and slow-twitch 524 skeletal muscle of 24-month-old rats. Journals of Gerontology - Series A Biological Sciences 525 and Medical Sciences 67(12): 1279–85, Doi: 10.1093/gerona/gls085. 526 [36] Gazdag, A.C., Dumke, C.L., Kahn, C.R., Cartee, G.D., 1930. Calorie Restriction Increases In-527 sulin-Stimulated Glucose Transport in Skeletal Muscle From IRS-1 Knockout Mice. vol. 48. 528 [37] Delfinis, L.J., Bellissimo, C.A., Gandhi, S., DiBenedetto, S.N., Garibotti, M.C., Thuhan, A.K., 529 et al., 2022. Muscle weakness precedes atrophy during cancer cachexia and is linked to mus-530 cle-specific mitochondrial stress. JCI Insight 7(24), Doi: 10.1172/jci.insight.155147. 531 [38] Berriel Diaz, M., Rohm, M., Herzig, S., 2024. Cancer cachexia: multilevel metabolic dysfunc-532 tion. Nature Metabolism 6(12): 2222–45, Doi: 10.1038/s42255-024-01167-9. 533 [39] Tisdale, M.J., 1997. Biology of Cachexia. JNCI Journal of the National Cancer Institute 534 89(23): 1763–73, Doi: 10.1093/jnci/89.23.1763. 535 536 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint BA C D F GE IH M N O P Q R J K L Con Non-CC CC 0 2 4 6 Food intake Last 3 days prior to terminationg ### Con Non-CC CC 0 1 2 3 Grip strength N (m/s 2) # ### Pre Post Pre Post Pre Post 0 1 2 3 4 Pre Post Pre Post Pre Post 0 1 2 3 4 Fat mass g Con Non-CC CC *** *** ### Pre Post Pre Post Pre Post 0 20 25 30 35 Pre Post Pre Post Pre Post 0 20 25 30 35 Body mass g (excl. tumor) Con Non-CC CC *** ### *** 0 1 2 3 Tumor g # Non-CC CC Gast TA Heart pWAT BAT Spleen 0 200 400 600 800 Tissue weight mg # ### ### ### ### ### # ### # Pre Post Pre Post Pre Post 0 15 20 25 30 Pre Post Pre Post Pre Post 0 15 20 25 30 Lean body mass g (excl. tumor) Con Non-CC CC ### *** 400 800 1200 0 100 200 300 400 500 Pearson correlation Tumor Volume (mm3) incremental AUC r2=0.5173 p=0.01906 Gast Quad pWAT BAT Spleen 0 200 400 600 800 Tissue weight mg p=0.086 # # # ## ## TA Gram Pre Post Pre Post 0 2 4 6 8 Fat mass g Int, KPC x Time: p>0.01 ME, KPC: p=0.078 ### *** Body weight Gram Pre Post Pre Post 0 25 30 35 40 Body mass g (excl. tumor) ME, KPC: p=0.078 # KPC 0 500 1000 1500 Tumor mg 0 20 40 60 80 100 120 0 5 10 15 20 Glucose tolerance Time (min) mM Control KPC### ### ## Int, KPC x time: p<0.001 ME, KPC: p<0.01 Control Balb/cJ C26 or vehicle Con Non-CC CC Cachexia Non- cachexia Tumor, no weight loss Tumor, weight loss No tumor Control C57bl6J KPC or vehicle Con KPCCachexia Tumor, weight loss No tumor 0 10 20 30 40 50 Bas Ins Bas Ins Bas Ins 0 10 20 30 40 50 Soleus µmol/g/h Con Non-CC CC *** ### * ** ###Int, Group x Ins: p<0.001 ME, Ins: p<0.001 0 10 20 30 Bas Ins Bas Ins Bas Ins 0 10 20 30 EDL µmol/g/h *** ## ****** ### # Con Non-CC CC ME, Group: p<0.01 ME, Ins: p<0.001 Body weight Gram 0 min 20 min 0 min 20 min 0.0 0.5 1.0 1.5 2.0 Plasma insulin ng/ml ME, Time: p<0.001 ** * Non-CC CC 0 min20 min 0 min20 min 0 min20 min 0 1 2 3 4 5 0 min 20 min 0 min 20 min 0 min 20 min 0 1 2 3 4 5 Plasma Insulin ng/ml Con 0 0 4 8 12 Glucose tolerance Time (min) mM Control Non-cachectic (average tumor: 623mm3) 9020 40 60 ME, C26: p<0.01 Int, C26 x time: p<0.001 ## ## # Cachectic (average tumor: 1044mm3) Insulin-stimulated glucose uptake Ex vivo ME, time: p<0.01 preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint BA C D F G E Bas Ins Bas Ins Bas Ins 0 2 4 6 8 10 Bas Ins Bas Ins Bas Ins 0 2 4 6 8 10 pAKTthr308 Band intensity (AU) Con Pre-CC CC *** ### ** ** ### Bas Ins Bas Ins Bas Ins 0 2 4 6 8 10 Bas Ins Bas Ins Bas Ins 0 2 4 6 8 10 pTBC1D4thr642 Band intensity (AU) Con Pre-CC CC *** *** *** ### Bas Ins Bas Ins Bas Ins 0 5 10 15 Bas Ins Bas Ins Bas Ins 0 5 10 15 pAKTser473 Band intensity (AU) Con Pre-CC CC *** ### ***** ### Bas Ins Bas Ins Bas Ins 0 2 4 6 8 10 Bas Ins Bas Ins Bas Ins 0 2 4 6 8 10 pGSK3ser9 Band intensity (AU) Con Pre-CC CC *** ### ***** ### Bas Ins Bas Ins Bas Ins 0 1 2 3 4 5 Bas Ins Bas Ins Bas Ins 0 1 2 3 4 5 pGSK3ser21 Band intensity (AU) Con Pre-CC CC *** ## ***** ### H Bas Ins Bas Ins Bas Ins 0 2 4 6 8 10 Bas Ins Bas Ins Bas Ins 0 2 4 6 8 10 pPRAS40thr246 Band intensity (AU) Con Pre-CC CC *** ## ****** ### I Bas Ins Bas Ins Bas Ins 0 2 4 6 8 Bas Ins Bas Ins Bas Ins 0 2 4 6 8 p-p70S6Kthr389 Band intensity (AU) Con Pre-CC CC *** *** p=0.0503 GLUT4 Hexokinase II Pyruvate dehydrogenase Glycogen Synthase AKT2 TBC1D4 GSK3β PRAS40 p70S6K rS6 0 1 2 3 Total protein content soleus muscle # p=0.057 p=0.075 # p=0.068 Band intensity (AU) J Bas Ins Bas Ins Bas Ins 0 1 2 3 4 Bas Ins Bas Ins Bas Ins 0 1 2 3 4 prS6ser235-236 Band intensity (AU) Con Pre-CC CC ** Glucose uptake Glycogen synthesis Protein synthesis Insulin stimulation thr246 thr389 thr642 ser235-236 ser9 ser21 ser473thr308 mTORC1 Akt p70S6K TBC1D4 rpS6 GSK3 PRAS40 kDa pAKT Ser473 pAKT Thr308 β-Ser9 α-Ser21 Coomassie Stain (loading control) pTBC1D4 Thr642 AKT2 GSK3 GSK3β TBC1D4 60 50 50 60 60 160 160 Con BAS INS Pre-CC BAS INS Con BAS INS CC BAS INS kDa Coomassie Stain (loading control) Glycogen Synthase Hexokinase II Pyruvate dehydrogenase GLUT4 90 100 40 50 Con BAS INS Pre-CC BAS INS Con BAS INS CC BAS INS kDa PRAS40 pPRAS40 Ser246 p-rS6 Ser235 Coomassie Stain (loading control) p-p70S6K Thr389 p70S6K rS6 40 40 70 70 35 35 Con BAS INS Pre-CC BAS INS Con BAS INS CC BAS INS preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint 0 1 2 3 -2 -1 0 1 Body mass Day g AL 3-day FR ### ### ### ** Int, FR x time: p<0.01 ME, FR : p<0.001 *** *** ****** * 3-day FR Food intake 30% reduction AL 0 2 4 6 8gram/day 0 20 40 60 80 100 0 5 10 15 20 Glucose tolerance Min mM AL 3-day FR ### ### ### ### Int, FR x time: p<0.01 ME, FR : p<0.001 AL 3-day FR 0 250 500 750 1000 iAUC AU ## AL 3-day FR 0 1 2 3 Grip Strength N (m/s2) NS GastTA Quad Heart iWAT pWAT BAT Spleen 0 10 20 30 Tissue weight mg tissue/BM at Day 0 AL 3-day FR NS## ## ### ### ### ## ### Control 3-day FRBAS INS BAS INS 0 25 50 75 100 Soleus Glucose transport µmol/g/h *** *** ## Int, FR x Ins: p<0.01 ME, Ins: p<0.001 BAS INS BAS INSBAS INS BAS INS 0 6 12 18 Soleus pAKT Thr308 Band Intensity (AU) *** *** # Int, FR x Ins: p<0.05 ME, Ins: p<0.001 BAS INS BAS INSBAS INS BAS INS 0 6 12 18 Soleus pAKT Ser473 Band Intensity (AU) *** *** p=0.058 Int, FR x Ins: p=0.073 ME, Ins: p<0.001 BAS INS BAS INSBAS INS BAS INS 0 4 8 12 Soleus pTBC1D4 Thr642 Band Intensity (AU) *** *** # Int, FR x Ins: p<0.05 ME, Ins: p<0.001 kDaBAS pAKT Thr308 pAKT Ser473 AKT2 pTBC1D4 Thr642 TBC1D4 Coomassie Stain (loading control) 3-day FR 60 60 60 160 160 Ad Lib INS BAS INS A B C D E F G H I Ad libitum (AL) food intake 3-day food restriction (FR) (-30%), pair-fed to C26 cachectic mice preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted March 15, 2026. ; https://doi.org/10.64898/2026.03.12.711318doi: bioRxiv preprint