Prevention of chemotherapy-induced cachexia by ACVR2B ligand blocking has different effects on heart and skeletal muscle

This is an electronic reprint of the original article. This reprint may differ from the original in pagination and typographic detail. Author(s): Title: Year: Version: Please cite the original version: All material supplied via JYX is protected by copyright and other intellectual property rights, and duplication or sale of all or part of any of the repository collections is not permitted, except that material may be duplicated by you for your research use or educational purposes in electronic or print form. You must obtain permission for any other use. Electronic or print copies may not be offered, whether for sale or otherwise to anyone who is not an authorised user. Prevention of chemotherapy-induced cachexia by ACVR2B ligand blocking has different effects on heart and skeletal muscle Hulmi, Juha; Nissinen, Tuuli; Räsänen, Markus; Degerman, Joni; Lautaoja, Juulia; Hemanthakumar, Karthik Amudhala; Backman, Janne T.; Ritvos, Olli; Silvennoinen, Mika; Kivelä, Riikka Hulmi, J., Nissinen, T., Räsänen, M., Degerman, J., Lautaoja, J., Hemanthakumar, K. A., Backman, J. T., Ritvos, O., Silvennoinen, M., & Kivelä, R. (2018). Prevention of chemotherapy-induced cachexia by ACVR2B ligand blocking has different effects on heart and skeletal muscle. Journal of Cachexia, Sarcopenia and Muscle, 9(2), 417432. https://doi.org/10.1002/jcsm.12265 2018 Prevention of chemotherapy-induced cachexia by ACVR2B ligand blocking has different effects on heart and skeletal muscle Juha J. Hulmi 1,2 * † , Tuuli A. Nissinen 1† , Markus Räsänen 3 , Joni Degerman 3 , Juulia H. Lautaoja 1 , Karthik Amudhala Hemanthakumar 3 , Janne T. Backman 4 , Olli Ritvos 2 , Mika Silvennoinen 1 & Riikka Kivelä 3 * 1 Biology of Physical Activity, Neuromuscular Research Center, Faculty of Sport and Health Sciences, University of Jyväskylä, Jyväskylä, Finland; 2 Department of Physiology, Faculty of Medicine, University of Helsinki, Helsinki, Finland; 3 Wihuri Research Institute, Helsinki, Finland and Translational Cancer Biology Program, Research Programs Unit, Faculty of Medicine, University of Helsinki, Helsinki, Finland; 4 Department of Clinical Pharmacology, Faculty of Medicine, University of Helsinki and Helsinki University Hospital, Helsinki, Finland Abstract Background Toxicity of chemotherapy on skeletal muscles and the heart may significantly contribute to cancer cachexia, mortality, and decreased quality of life. Doxorubicin (DOX) is an effective cytostatic agent, which unfortunately has toxic effects on many healthy tissues. Blocking of activin receptor type IIB (ACVR2B) ligands is an often used strategy to prevent skeletal muscle loss, but its effects on the heart are relatively unknown. Methods The effects of DOX treatment with or without pre-treatment with soluble ACVR2B-Fc (sACVR2B-Fc) were investigated. The mice were randomly assigned into one of the three groups: (1) vehicle (PBS)-treated controls, (2) DOX-treated mice (DOX), and (3) DOX-treated mice administered with sACVR2B-Fc during the experiment (DOX + sACVR2B-Fc). DOX was administered with a cumulative dose of 24 mg/kg during 2 weeks to investigate cachexia outcome in the heart and skeletal muscle. To understand similarities and differences between skeletal and cardiac muscles in their responses to chemotherapy, the tissues were collected 20 h after a single DOX (15 mg/kg) injection and analysed with genome-wide transcriptomics and mRNA and protein analyses. The combination group was pre-treated with sACVR2B-Fc 48 h before DOX administration. Major findings were also studied in mice receiving only sACVR2B-Fc. Results The DOX treatment induced similar (~10%) wasting in skeletal muscle and the heart. However, transcriptional changes in response to DOX were much greater in skeletal muscle. Pathway analysis and unbiased transcription factor analysis showed that p53-p21-REDD1 is the main common pathway activated by DOX in both skeletal and cardiac muscles. These changes were attenuated by blocking ACVR2B ligands especially in skeletal muscle. Tceal7 (3-fold to 5-fold increase), transferrin receptor (1.5-fold increase), and Ccl21 (0.6-fold to 0.9-fold decrease) were identified as novel genes responsive to blocking ACVR2B ligands. Overall, at the transcriptome level, ACVR2B ligand blocking had only minor influence in the heart while it had marked effects in skeletal muscle. The same was also true for the effects on tissue wasting. This may be explained in part by about 18-fold higher gene expression of myostatin in skeletal muscle compared with the heart. Conclusions Cardiac and skeletal muscles display similar atrophy after DOX treatment, but the mechanisms for this may differ between the tissues. The present results suggest that p53-p21-REDD1 signalling is the main common DOX-activated pathway in these tissues and that blocking activin receptor ligands attenuates this response, especially in skeletal muscle supporting the overall stronger effects of this treatment in skeletal muscles. Keywords Myostatin; Activins; Transcriptome; p53; Doxorubicin; Ccl21 Received: 4 August 2017; Revised: 15 September 2017; Accepted: 12 October 2017 *Correspondence to: Juha Hulmi, Neuromuscular Research Center, Faculty of Sport and Health Sciences, University of Jyväskylä, Jyväskylä, Finland.Email:juha.hulmi@jyu.fi Riikka Kivelä, Translational Cancer Biology Program, Faculty of Medicine, University of Helsinki, Helsinki, Finland.Email:riikka.kivela@helsinki.fi †These authors contributed equally to the work. ORIGINAL ARTICLE © 2017 The Authors. Journal of Cachexia, Sarcopenia and Muscle published by John Wiley & Sons Ltd on behalf of the Society on Sarcopenia, Cachexia and Wasting Disorders Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 Published online 11 December 2017 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/jcsm.12265 This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. Introduction Cancer cachexia is associated with increased mortality. 1 This may, in part, be related to increased toxicity of chemotherapy on skeletal muscles and the heart. 1–5 The maintenance of skeletal 1 and cardiac 6 muscle mass and function predicts better response to treatment and survival in diseases. Therefore, it is crucial to discover and develop effective strategies to counteract pathological skeletal and cardiac muscle loss. Doxorubicin (DOX, Adriamycin®) is an anthracycline cytostatic agent, which acts through arresting cell cycle, and thus blocks proliferation of malignant cells. 7 Unfortunately, DOX has deleterious effects also on many healthy tissues. Large doses of DOX have been shown for decades to induce cardiotoxicity through various mechanisms. 5,8,9 In addition, DOX induces adverse effects on skeletal muscle tissue including muscle weakness, fatigue, dysfunction, and atrophy. 10 Some studies have suggested that the heart may be more sensitive to DOX than skeletal muscle, 2,3 but the degree of skeletal muscle dysfunction can be comparable to or even higher than that of heart. 11,12 Skeletal and cardiac muscles are very similar in several aspects; both are striated and composed of myofibrils. However, cardiac cells are smaller, more circular, branched, and have junctions between cells called intercalated discs connecting cardiomyocytes (CMCs) together. This complexity in the heart may be an advantage, but also a disadvantage regarding regeneration, which has been thought to be among the weakest in the adult mammalian body. 13,14 In comparison to the heart, skeletal muscles have a remarkable regeneration capacity even after very severe injury. 15 Skeletal muscle size is negatively regulated by myostatin, GDF11, and activins, which belong to the TGF-βsuperfamily of proteins. 16–18 They exert their effects through binding to activin receptor type IIB (ACVR2B). An often used strategy to increase muscle size and to prevent muscle loss is to block these ACVR2B ligands by administration of a soluble ligand binding domain of ACVR2B (sACVR2B-Fc). 4,19–21 In addition, sACVR2B-Fc treatment has been found to prolong survival and to reverse cancer cachexia in mice. 22,23 Activin receptor signalling is important for heart growth regulation and homeostasis as well, 18,22,24–29 but the effects of blocking activin receptor ligands in the heart have not been investigated in chemotherapy-induced cardiac atrophy. Previous studies have evaluated the effects of DOX and ACVR2B blocking either on the heart or skeletal muscle. In the present study, we compared the effects of DOX-treatment with or without pre-treatment with sACVR2B-Fc on cardiac and skeletal muscles. We show here that cardiac and skeletal muscle masses were similarly decreased by DOX chemotherapy treatment. However, transcriptional changes in response to DOX were much greater in skeletal muscle. Furthermore, we show that blocking of activin receptor ligands had more pronounced effect on skeletal muscle than cardiac wasting. Materials and methods Animals C57BL/6J male mice (Envigo), aged 6–10 weeks were maintained under standard conditions (temperature 22°C, 12:12 h light/dark cycle) with free access to food and water. The protocols were approved by the National Animal Experiment Board, and all the experiments were carried out in accordance with the guidelines of the committee and the ethical standards of the Declaration of Helsinki. Experimental design The mice were randomly assigned into three groups: (1) vehicle (PBS)-treated controls (CTRL), (2) DOX hydrochloride treated mice (DOX), and (3) DOX-treated mice administered with sACVR2B-Fc intraperitoneally (DOX + sACVR2B) (Figure 1A). In the long-term treatment experiments, both DOX groups received four intraperitoneal injections of DOX (each 6 mg/kg in PBS), administered every third day during the first 2 weeks of the experiment. 4,5 Control mice were administered with an equal volume of PBS. The mice were euthanized at 2 weeks and at 4 weeks after the first DOX injection. Half of the DOX mice were injected with sACVR2B-Fc (5 mg/kg in PBS) twice a week during the first 2 weeks of the experiment and once a week after that. The DOX treatment protocol was designed to mimic the treatment of human patients with low DOX doses to induce cardiotoxicity but no treatment-related deaths. 30 In the acute experiment, a single intraperitoneal injection of DOX (15 mg/kg in PBS) or an equal volume of PBS was administered. sACVR2B-Fc-treated mice received a single intraperitoneal injection of sACVR2B-Fc (10 mg/kg in PBS) 48 h before DOX administration, as we have previously shown that sACVR2B-Fc increases muscle protein synthesis 48 h after its administration. 20 The mice were euthanized 20 h after DOX/PBS administration. To analyse the effects of sACVR2BFc alone, another experiment was conducted in which sACVR2B-Fc (10 mg/kg in PBS) or PBS were administered 48 h before sample collection into wild-type mice. 20 Tissue collection At the end of the experiment, the mice were anaesthetized with ketamine and xylazine and then euthanized by cardiac puncture followed by cervical dislocation. Hindlimb muscle tibialis anterior (TA), gastrocnemius, soleus, and the heart were immediately excised and weighed. The left TA and gastrocnemius muscles and part of the heart (apex) were snapfrozen in liquid nitrogen. Blood was fully drained from hearts prior to tissue weighing. All tissue weights were normalized to the length of the tibia (mm). TA muscle was used in 418 J.J. Hulmi et al. Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265 subsequent analyses, except when analysing the effects of sACVR2B-Fc alone 20 and DOX-measurements, in which gastrocnemius was used. Puromycin injection for protein synthesis measurement Puromycin incorporation assay 31 was conducted as earlier 4,20,32 with small modifications. In brief, mice were intraperitoneally injected with 0.040 μmol/g (21.78 mg/kg) puromycin (Calbiochem, Darmstadt, Germany) dissolved in 200 μLof PBS. At exactly 25 min after the injection of puromycin, mice were euthanized by cervical dislocation and heart was collected and snap-frozen at exactly 30 min after puromycin injection. sACVR2B-Fc production The recombinant fusion protein was produced and purified as described earlier in detail. 20 The ectodomain of human ACVR2B was fused with a human IgG1 Fc domain and expressed in Chinese hamster ovary cells grown in a suspension culture. The protein is similar but not identical to that originally generated by Lee and colleagues. 19 RNA analysis Total RNA was extracted from muscle and the heart with TRIsure reagent (Bioline) and further purified with NucleoSpin® RNA II columns. For quantitative polymerase chain reaction (qPCR), RNA was reverse transcribed to cDNA by using iScript ™ Advanced cDNA Synthesis Kit for real-time qPCR (Bio-Rad Laboratories) according to the manufacturer’s instructions. Real-time qPCR was performed according to standard procedures by using iQ SYBR Supermix (Bio-Rad Laboratories) and CFX96 Real-Time PCR Detection System (Bio-Rad Laboratories). Quantification was carried out by using standard curve or efficiency corrected ΔΔCt method. The relative mRNA expressions were normalized by using 36b4 as a reference gene, as it was the most stable (lowest intergroup and intragroup variances) from three candidate reference genes (36b4,Gapdh, and Rn18S). Primer sequences are listed in Online Resource 1: Supplementary Methods. Microarray analysis RNA from the TA and the heart samples of the acute experiment were analysed with Illumina Sentrix MouseRef-6 v2 Expression BeadChip containing 45 281 transcripts (Illumina Inc., San Diego, CA, USA) by the Functional Genomics Unit at Biomedicum Helsinki, University of Helsinki, Finland according to the manufacturer’s instructions. Five muscle and heart samples from control and DOX groups and five muscle and three heart samples from DOX + sACVR2B-Fc group were analysed. RNA was analysed for integrity and quality on Agilent Bioanalyser 2100. Illumina’s GenomeStudio software was used for initial data analysis and quality control. Raw data were normalized with quantile normalization (including log2-transformation of the data), data quality was assessed, and statistical analyses were performed by using Chipster software (IT Center for Science, Espoo, Finland). 33 Statistically significant differences in individual genes between the groups were tested by using Empirical Bayes statistics and the Benjamini-Hochberg algorithm controlling false discovery rate (FDR). FDR values of <0.05 with ≥1.2-fold change difference were considered significant. MIAME guidelines were followed during array data generation, pre-processing, and analysis. The complete data set is publicly available in the NCBI Gene Expression Omnibus (http://www.ncbi.nlm.nih. gov/geo/; accession no. GSE77745 and GSE97642). Heatmap Figure 1 (A) Experimental design. The study included acute as well as 2 and 4 week experiments. (B) Adjusted mass (control = 1, adjusted to tibial length) of skeletal muscle and the heart after 4 weeks of cumulative 24 mg/kg doxorubicin administration (mean ± SD). Skeletal muscle mass is a sum of tibialis anterior, gastrocnemius, and soleus masses. n= 15, 16, and 17 in skeletal muscle and n= 14, 16, and 16 in heart in CTRL, DOX, and in DOX + sACVR2B, respectively. General linear model analysis of variance with Bonferroni post hoc test was used. * or *** = significant (P<0.05 or P<0.001, respectively) difference to respective CTRL. ### = significant (P<0.001) difference to respective DOX. Prevention of chemotherapy-induced cachexia 419 Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265 illustrations were performed with GENE-E software (Broad Institute, Cambridge, USA). Transcription factor analysis Transcription factor (TF) analysis is usually conducted in cultured cells rather than actual tissue samples, and it focuses on a single a priori chosen TF at a time. When the TF is not known in advance, or when only gene expression profiling is available, regulatory relationships can be uncovered by reverse-engineering a gene regulatory network starting from the expression data. A-genome-wide ranking-and-recovery approach using iRegulon software 34 was used to detect enriched TF motifs and their optimal set of direct target genes. This analysis also links these candidate motifs to TFs by using motif2TF procedure. Pathway analysis Enrichment of functionally related genes in four different gene set collections was first performed by using a nonbiased method by gene set enrichment analysis software (GSEA; Version 2.0) 35 as previously done in our laboratory. 4,36,37 The collections used were the Canonical Pathways, Biocarta, KEGG, and Reactome (http://www.broadinstitute. org/gsea/msigdb/collections.jsp). The number of permutations by gene set was set to 1000 and gene sets with at least 10, and no more than 500 genes were taken into account in each analysis. The statistical significance was calculated by using FDR, and the level of significance was set at FDR<0.05. Doxorubicin measurement The DOX concentration from gastrocnemius muscle was measured with an Agilent 1100 HPLC system (Agilent Technologies, Waldbronn, Germany) coupled to an AB Sciex API 2000 tandem mass spectrometer (Framingham, MA), as previously described. 5 Tissue processing for the protein analysis Muscle and heart samples were homogenized and treated with proper inhibitors as previously reported. 4,5 One part of the heart homogenate was taken for the puromycin incorporation examination. For that purpose, the sample was centrifuged at 500 gfor 5 min to remove cell debris. For the analysis of individual proteins, the rest of the homogenate was centrifuged at 10 000 gfor 10 min. Total protein content was determined by using the bicinchonic acid protein assay (Pierce Biotechnology, Rockford, USA) with an automated KoneLab analyser (Thermo Scientific, Vantaa, Finland). Western immunoblot analyses Western immunoblot analyses were performed as previously reported. 4,5 Ponceau S staining and GAPDH were used as loading controls, and all the results are normalized to the mean of Ponceau S and GAPDH. The quantification of GAPDH normalized to Ponceau S was similar among the groups, indicating that GAPDH protein content remained stable under the experimental conditions. The antibodies used are listed in the Online Resource 1: Supplementary Methods. Immunohistochemistry Cardiac and skeletal muscle (TA) tissue sections were cut with cryomicrotome and fixed with ice-cold acetone. Masson trichrome staining was performed to analyse the amount of fibrosis in the tissues. Immunohistochemistry was performed to measure CMC cross-sectional area by using mouse-antidystrophin antibody (1:500 dilution, NCL-Dys 2, Novocastra). Rabbit-anti-Ki67 antibody (1:300, ab1667, Abcam) was used to evaluate the effects of DOX and sACVR2B-Fc on the cell proliferation. Sections were imaged with Zeiss Axioimager microscope, and CMC size was calculated using Cell Profiler software. Statistical analysis Multiple group comparisons except microarray (see details above) were conducted with general linear model analysis of variance followed by Bonferroni post hoc test or by nonparametric Kruskal–Wallis test followed by Holm-Bonferroni corrected Mann–Whitney U-test as post hoc when appropriate. For two-group comparisons, a two-tailed unpaired Student’sttest or non-parametric Mann–Whitney U-test was used. Data were checked for normality and for the equality of variances. The level of significance in these analyses was set at P<0.05. Data are expressed as means ± SEM if not otherwise mentioned. Statistical analyses were performed with IBM SPSS STATISTICS version 24 for Windows (SPSS, Chicago, IL). Results ACVR2B blocking can prevent chemotherapy-induced skeletal muscle but not cardiac atrophy At 4 weeks, chemotherapy-induced atrophy was almost identical between skeletal and cardiac muscles (Figure 1B). In skeletal muscle, sACVR2B-Fc treatment effectively prevented the loss of muscle mass and was able to even increase muscle mass (Figure 1B). However, sACVR2B-Fc was unable to fully block the cardiac atrophy, although the weight loss was 420 J.J. Hulmi et al. Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265 slightly less consistent in DOX + sACVR2B (P= 0.165) when compared with DOX alone (P= 0.030). Furthermore, CMC cross-sectional area was not increased by sACVR2B (Online Resource 2: Supplementary Figure S1A). Larger transcriptomic changes in skeletal muscle than in the heart in response to doxorubicin chemotherapy Whole-genome microarray analysis using FDR <0.05 and fold change ≥1.2 criteria showed that 485 and 40 annotated transcripts were up-regulated and 473 and 24 were downregulated by DOX in skeletal muscle and heart, respectively, at 20 h after a single DOX injection. Out of these genes, there were 21 and 6 genes that were up-regulated or down-regulated, respectively, by DOX in both muscle and the heart (Online Resource 3: Supplementary Figure S2A and S2B). In addition to having much larger number of genes altered, the genes with largest changes showed more robust response in skeletal muscle as compared with the heart (Figure 2A–2D). Of the most up-regulated genes, a well-known cell-cycle inhibitor p21/Cdkn1a was highly up-regulated by DOX injection in both muscle and the heart (Figure 2A and 2C), and this was validated by qPCR (Figure 3A). Interestingly, this response was significantly decreased by sACVR2B-Fc treatment preceding DOX administration in skeletal muscle (Figure 3A). sACVR2B-Fc alone did not, however, decrease p21/Cdkn1a below healthy controls (Figure 3A). Moreover, DNA-damage response indicator Redd1/Ddit4 was up-regulated in both skeletal and heart muscle as published earlier, 4,5 and blocking ACVR2B ligands attenuated this response in skeletal muscle as published earlier 4 without an effect in the heart (Online Resource 2: Supplementary Figure S1B). Tceal7 and Ccl21 mRNAs are regulated by sACVR2B-Fc One hundred eighteen and 1 annotated transcripts were upregulated, and 84 and 2 transcripts were down-regulated in DOX + sACVR2B-Fc-treated mice when compared with DOX alone in muscle and the heart, respectively. In skeletal muscle, the gene with the highest increase by sACVR2B-Fc in microarray (Figure 2E) was Tceal7 [transcription elongation factor A], a protein involved in skeletal muscle development and regeneration, 38 and this finding was further confirmed by qPCR (Figure 3B). The expression level of Tceal7 in the heart was very low and could not be analysed reliably. The only significantly up-regulated gene by sACVR2B-Fc in the heart was Vsig4 (3.52-fold, FDR <0.001), which remained unchanged in skeletal muscle. Of the five annotated probes down-regulated by sACVR2B-Fc in the heart, four were probes for Ccl21 gene and the response of this gene was also validated by qPCR in both tissues (Figure 3C). sACVR2B-Fc treatment alone also increased the expression of Tceal7 and decreased the expression of Ccl21 (Figure 3B and 3C), confirming that these effects are due to the blocking of activin receptor type IIB ligands. Potassium voltage-gated channel, Isk-related subfamily, member 1 (Kne1) was another gene decreased by sACVR2B-Fc in the heart (0.55-fold, FDR = 0.01) without an effect in skeletal muscle. Transcripts, which showed the largest down-regulation by sACVR2B-Fc in muscle, are shown in Figure 2F. There were 22 and 17 genes that were up-regulated or down-regulated, respectively, by DOX in skeletal muscle, and whose expression was normalized by sACVR2B-Fc treatment (Online Resource 3: Supplementary Figure S2C and S2D). An interesting gene among these was transferrin receptor, as iron metabolism has been shown to be affected and to play a role in DOX-induced toxicity. 39 In both muscle and the heart transferrin receptor mRNA decreased by DOX and in both tissues, but especially in skeletal muscle, this was rescued by sACVR2B-Fc (Figure 3D). The increase in transferrin receptor by sACVR2B-Fc was translated into protein level as well, especially in skeletal muscle (Figure 3E). PGC-1 gene expression As microarray platforms do not have probes for most of the recently identified PGC-1αisoforms, we analysed them by qPCR. In skeletal muscle, Pgc-1αexon 1a (Ppargc1a exon 1a) and Pgc-1αexon 1c isoforms as well as Pgc-1β (Ppargc1b) mRNA decreased by DOX (P<0.05), and only the N-truncated Pgc-1αisoforms remained unchanged (Figure 4A–4D). In contrast, in the heart, there was an overall increase by DOX in Pgc-1αisoforms (Figure 4A–4D). No effect of sACVR2B-Fc was observed in either tissue type (Figure 4A–4D). Heart protein synthesis and ubiquitin ligases We recently reported that skeletal muscle protein synthesis was decreased by DOX and this could be restored by sACVR2B-Fc. 4 In this study, we analysed protein synthesis in the heart in response to DOX and sACVR2B-Fc. Unlike in skeletal muscle, there was no consistent effect of either DOX or sACVR2B-Fc on puromycin incorporation into proteins, a marker of protein synthesis, in the heart (Figure 5A). The level of ubiquitinated proteins was also unchanged in the heart (Figure 5B) similarly as previously published in skeletal muscle. 4 E3 ubiquitin ligase Atrogin1 mRNA increased by DOX in both tissues, but more robustly in the skeletal muscle (Figure 5C). Interestingly, sACVR2B-Fc prevented the increase in Atrogin1 mRNA in muscle (Figure 5C). Murf1 mRNA showed a small decrease by DOX in the heart (Figure 5C), while as previously published in skeletal muscle, Murf1 mRNA was unaltered by DOX, but decreased due to sACVR2B-Fc. 4 Prevention of chemotherapy-induced cachexia 421 Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265 Muscle doxorubicin content is unaltered by sACVR2B-Fc To analyse whether the strong effects of sACVR2B-Fc were simply due to attenuated DOX levels in the combination group, DOX concentration in gastrocnemius muscle after the acute DOX treatment was analysed. We found that DOX content at 20 h post-injection did not differ between the DOX and DOX + sACVR2B administered mice (Figure 5D), suggesting that sACVR2B-Fc effects are not due to altered tissue concentration of DOX. Pathway analysis shows similar responses in skeletal muscle and the heart To identify affected pathways, further analyses were conducted by using an unbiased gene clustering analysis with Figure 2 Top 10 genes with largest response to doxorubicin in skeletal muscle (A and B) and in the heart (Cand D). Top 10 genes with largest change by sACVR2B-Fc in skeletal muscle (E and F). C = CTRL, D = DOX, A = DOX + sACVR2B. *, **, or *** = significant (P<0.05, P<0.01, or P<0.001, respectively) adjusted difference (false discovery rate). n= 5 per group. 422 J.J. Hulmi et al. Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265 GSEA. 35 In GSEA analysis, 72 and 51 gene sets were upregulated and 17 and 0 down-regulated (FDR <0.05) in skeletal muscles and hearts of DOX-injected mice, respectively (Online Resource 4: Supplementary Figure S3A and S3B). It was evident that p53/p63 pathways/gene sets were activated in both skeletal muscle and the heart (Figure 6A and 6B). Also, pathways related to RNA polymerase and transcription were up-regulated in both tissues (Figure 6A and 6B). In skeletal muscle the most down-regulated pathways were related to extracellular proteins, collagens, and their regulation (Figure 6A). Same pathways tended to be decreased also in the heart after DOX-injection, but less significantly than in skeletal muscle (FDR >0.05). No sign of fibrosis was observed with Masson trichrome staining either in skeletal muscle or in the heart at 4 weeks (Online Resource 2: Supplementary Figure S1C). Correspondingly, nine gene sets were up-regulated and six down-regulated (FDR <0.05) in DOX + sACVR2B when compared with DOX alone in skeletal muscle without any affected gene sets in the heart (Figure 6C). In muscle, the most activated gene sets were related to translation capacity and efficiency, and the most down-regulated ones to extracellular proteins, and especially proteoglycans and extracellular matrix glycoproteins. Of the gene sets that were down-regulated by DOX in skeletal muscle, p38-MAPK pathway was significantly increased by DOX + sACVR2B when compared with DOX alone (Figure 6C and Online Resource 4: Supplementary Figure S3C and S3D). Figure 3 Doxorubicin alters the gene expression of (A)Cdkn1a (p21), while sACVR2B-Fc increases (B)Tceal7 in skeletal muscle and attenuates (C)Ccl21 mRNA levels. (D) Doxorubicin decreases transferrin receptor (Tfrc) mRNA in both tissues while its mRNA and protein (E) levels are increased by sACVR2B-Fc. Tceal7 was expressed in heart only very weakly, and thus, it was not analysed. The values are presented as fold changes compared with the control group. For multiple group comparisons, general linear model analysis of variance with Bonferroni post hoc test (Aand B) or Kruskal-Wallis test with Holm-Bonferroni corrected Mann–Whitney Upost hoc test (C–E) were used. For two-group comparisons, the Student’st-test (A–C) or nonparametric Mann–Whitney U-test (E) were used. *, **, or *** = significant (P<0.05, P<0.01, or P<0.001, respectively) difference to respective CTRL. # or ### = significant (P<0.05 and P<0.001, respectively) difference to respective DOX. n=7–9 per group in the doxorubicin experiment and n=5–6 per group in the sACVR2B-Fc alone vs. PBS experiment. Prevention of chemotherapy-induced cachexia 423 Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265 As sACVR2B prevented the DOX-induced increase in the cell cycle inhibitor p21, we stained muscle and heart sections with Ki67 antibody. DOX treatment had no significant effect on the number of Ki67-positive nuclei, but surprisingly, sACVR2B significantly increased Ki67-positivie nuclei in both muscle and the heart (Online Resource 2: Supplementary Figure S1D). Transcription factor analysis Next, to gain insight into the key mediators of the effects of DOX and sACVR2B-Fc, a TF analysis was conducted. For this analysis, the gene lists of FDR <0.05 up-regulated or down-regulated genes with at least 1.5-fold change were loaded to iRegulon software. As a common TF up-regulated by DOX in both the muscle and the heart, p53 was again identified (Online Resource 5: Supplementary Table S1). mRNA expression changes of genes containing p53 targeted motifs in microarray (>1.5-fold increase DOX vs. Control) are shown in Figure 7A and 7B. p21/Cdkn1a and Redd1/Ddit4 rankings were high in both tissues suggesting that p53-REDD1-p21 pathway is the main common pathway activated by DOX in skeletal and cardiac muscles. p53 itself is mainly posttranscriptionally regulated, and indeed, its protein content was increased in both tissues after DOX (Figure 7C). However, sACVR2B-Fc completely blocked this response in both tissues (Figure 7C). This is in line with the expression changes downstream of p53, as p21/Cdkn1a mRNA up-regulated in response to DOX was attenuated by sACVR2B-Fc, especially in skeletal muscle (Figure 3A). Other notable TFs affected by DOX were MyoD in skeletal muscle and Stat3 in the heart. MyoD binding site was enriched in iRegulon by DOX (Online Resource 5: Supplementary Table S1). This result was associated with increased Myod1 mRNA by DOX in both microarray and in qPCR (Online Resource 6: Supplementary Figure S4A). On the other hand, Stat3 binding site was strongly enriched in the heart, but not in skeletal muscle (Online Resource 5: Supplementary Table S1) without changes in the phosphorylation of Stat3 (Online Resource 6: Supplementary Figure S4B). Gene expression of ACVR2B and its ligands in muscle and in heart and in response to doxorubicin and sACVR2B-Fc The difference in the response to ACVR2B ligand blocking between skeletal muscle and the heart could be explained by different expression level of activin receptor IIB or its ligands or their responses to DOX. First, we compared the mRNA levels in muscle and heart tissues of young healthy mice Figure 4 (A–D) Doxorubicin differentially alters the gene expression of Ppargc1 (PGC-1) mRNA isoforms in muscle and the heart. Notice that in (C) Ppargc1a exon 1c, the * in S. muscle with lines depicts the doxorubicin effect of both doxorubicin groups pooled when compared with control without treatments. General linear model analysis of variance with Bonferroni post hoc test was used. *, **, or *** = significant (P<0.05, P<0.01, or P<0.001, respectively) difference to respective CTRL. n=6–9 per group. 424 J.J. Hulmi et al. Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265 Hippo signaling, and capillary density are altered by blocking of myostatin and activins. Am J Physiol Endocrinol Metab 2013;304:E41–E50. 21. Rahimov F, King OD, Warsing LC, Powell RE, Emerson CP Jr, Kunkel LM, et al. Gene expression profiling of skeletal muscles treated with a soluble activin type IIB receptor. Physiol Genomics 2011;43:398–407. 22. Zhou X, Wang JL, Lu J, Song Y, Kwak KS, Jiao Q, et al. Reversal of cancer cachexia and muscle wasting by ActRIIB antagonism leads to prolonged survival. Cell 2010;142:531–543. 23. Toledo M, Busquets S, Penna F, Zhou X, Marmonti E, Betancourt A, et al. Complete reversal of muscle wasting in experimental cancer cachexia: additive effects of activin type II receptor inhibition and beta-2 agonist. Int J Cancer 2016;138:2021–2029. 24. Oh SP, Li E. The signaling pathway mediated by the type IIB activin receptor controls axial patterning and lateral asymmetry in the mouse. Genes Dev 1997;11:1812–1826. 25. Heineke J, Auger-Messier M, Xu J, Sargent M, York A, Welle S, et al. Genetic deletion of myostatin from the heart prevents skeletal muscle atrophy in heart failure. Circulation 2010;121:419–425. 26. Breitbart A, Auger-Messier M, Molkentin JD, Heineke J. Myostatin from the heart: local and systemic actions in cardiac failure and muscle wasting. Am J Physiol Heart Circ Physiol 2011;300:H1973–H1982. 27. Bish LT, George I, Maybaum S, Yang J, Chen JM, Sweeney HL. Myostatin is elevated in congenital heart disease and after mechanical unloading. PLoS One 2011;6:e23818. 28. Chen JL, Walton KL, Qian H, Colgan TD, Hagg A, Watt MJ, et al. Differential effects of IL6 and activin A in the development of cancer-associated cachexia. Cancer Res 2016;76:5372–5382. 29. Morissette MR, Cook SA, Foo S, McKoy G, Ashida N, Novikov M, et al. Myostatin regulates cardiomyocyte growth through modulation of Akt signaling. Circ Res 2006;99:15–24. 30. Vejpongsa P, Yeh ET. Prevention of anthracycline-induced cardiotoxicity: challenges and opportunities. J Am Coll Cardiol 2014;64:938–945. 31. Schmidt EK, Clavarino G, Ceppi M, Pierre P. SUnSET, a nonradioactive method to monitor protein synthesis. Nat Methods 2009;6:275–277. 32. Goodman CA, Mabrey DM, Frey JW, Miu MH, Schmidt EK, Pierre P, et al. Novel insights into the regulation of skeletal muscle protein synthesis as revealed by a new nonradioactive in vivo technique. FASEB J 2011;25:1028–1039. 33. Kallio MA, Tuimala JT, Hupponen T, Klemela P, Gentile M, Scheinin I, et al. Chipster: user-friendly analysis software for microarray and other high-throughput data. BMC Genomics 2011;12: 507-216412-507. 34. Janky R, Verfaillie A, Imrichova H, Van de Sande B, Standaert L, Christiaens V, et al. iRegulon: from a gene list to a gene regulatory network using large motif and track collections. PLoS Comput Biol 2014;10:e1003731. 35. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005;102:15545–15550. 36. Kivela R, Silvennoinen M, Lehti M, Rinnankoski-Tuikka R, Purhonen T, Ketola T, et al. Gene expression centroids that link with low intrinsic aerobic exercise capacity and complex disease risk. FASEB J 2010;24:4565–4574. 37. Kainulainen H, Papaioannou KG, Silvennoinen M, Autio R, Saarela J, Oliveira BM, et al. Myostatin/activin blocking combined with exercise reconditions skeletal muscle expression profile of mdx mice. Mol Cell Endocrinol 2015;399:131–142. 38. Shi X, Garry DJ. Myogenic regulatory factors transactivate the Tceal7 gene and modulate muscle differentiation. Biochem J2010;428:213–221. 39. Ichikawa Y, Ghanefar M, Bayeva M, Wu R, Khechaduri A, Naga Prasad SV, et al. Cardiotoxicity of doxorubicin is mediated through mitochondrial iron accumulation. J Clin Invest 2014;124:617–630. 40. Yi X, Bekeredjian R, DeFilippis NJ, Siddiquee Z, Fernandez E, Shohet RV. Transcriptional analysis of doxorubicin-induced cardiotoxicity. Am J Physiol Heart Circ Physiol 2006;290:H1098–H1102. 41. Shum AM, Fung DC, Corley SM, McGill MC, Bentley NL, Tan TC, et al. Cardiac and skeletal muscles show molecularly distinct responses to cancer cachexia. Physiol Genomics 2015;47:588–599. 42. Olson EN. Regulation of muscle transcription by the MyoD family. The heart of the matter. Circ Res 1993;72:1–6. 43. Megeney LA, Kablar B, Garrett K, Anderson JE, Rudnicki MA. MyoD is required for myogenic stem cell function in adult skeletal muscle. Genes Dev 1996;10:1173–1183. 44. Kurabayashi M, Jeyaseelan R, Kedes L. Antineoplastic agent doxorubicin inhibits myogenic differentiation of C2 myoblasts. J Biol Chem 1993;268:5524–5529. 45. Ellisen LW, Ramsayer KD, Johannessen CM, Yang A, Beppu H, Minda K, et al. REDD1, a developmentally regulated transcriptional target of p63 and p53, links p63 to regulation of reactive oxygen species. Mol Cell 2002;10:995–1005. 46. Deng C, Zhang P, Harper JW, Elledge SJ, Leder P. Mice lacking p21CIP1/WAF1 undergo normal development, but are defective in G1 checkpoint control. Cell 1995;82:675–684. 47. Kurz EU, Douglas P, Lees-Miller SP. Doxorubicin activates ATM-dependent phosphorylation of multiple downstream targets in part through the generation of reactive oxygen species. J Biol Chem 2004;279:53272–53281. 48. Zhu W, Soonpaa MH, Chen H, Shen W, Payne RM, Liechty EA, et al. Acute doxorubicin cardiotoxicity is associated with p53-induced inhibition of the mammalian target of rapamycin pathway. Circulation 2009;119:99–106. 49. Patel AK, Tripathi AK, Patel UA, Shah RK, Joshi CG. Myostatin knockdown and its effect on myogenic gene expression program in stably transfected goat myoblasts. In Vitro Cell Dev Biol Anim 2014;50:587–596. 50. Bullwinkel J, Baron-Luhr B, Ludemann A, Wohlenberg C, Gerdes J, Scholzen T. Ki-67 protein is associated with ribosomal RNA transcription in quiescent and proliferating cells. J Cell Physiol 2006;206:624–635. 51. von Walden F, Liu C, Aurigemma N, Nader GA. mTOR signaling regulates myotube hypertrophy by modulating protein synthesis, rDNA transcription, and chromatin remodeling. Am J Physiol Cell Physiol 2016;311: C663–C672. 52. Shoshani T, Faerman A, Mett I, Zelin E, Tenne T, Gorodin S, et al. Identification of a novel hypoxia-inducible factor 1responsive gene, RTP801, involved in apoptosis. Mol Cell Biol 2002;22:2283–2293. 53. Tidball JG, Villalta SA. Regulatory interactions between muscle and the immune system during muscle regeneration. Am J Physiol Regul Integr Comp Physiol 2010;298:R1173–R1187. 54. Velez JM, Miriyala S, Nithipongvanitch R, Noel T, Plabplueng CD, Oberley T, et al. p53 Regulates oxidative stress-mediated retrograde signaling: a novel mechanism for chemotherapy-induced cardiac injury. PLoS One 2011;6:e18005. 55. Fox DK, Ebert SM, Bongers KS, Dyle MC, Bullard SA, Dierdorff JM, et al. p53 and ATF4 mediate distinct and additive pathways to skeletal muscle atrophy during limb immobilization. Am J Physiol Endocrinol Metab 2014;307:E245–E261. 56. Yndestad A, Finsen AV, Ueland T, Husberg C, Dahl CP, Oie E, et al. The homeostatic chemokine CCL21 predicts mortality and may play a pathogenic role in heart failure. PLoS One 2012;7:e33038. 57. Finsen AV, Ueland T, Sjaastad I, Ranheim T, Ahmed MS, Dahl CP, et al. The homeostatic chemokine CCL21 predicts mortality in aortic stenosis patients and modulates left ventricular remodeling. PLoS One 2014;9: e112172. 58. Kotamraju S, Chitambar CR, Kalivendi SV, Joseph J, Kalyanaraman B. Transferrin receptor-dependent iron uptake is responsible for doxorubicin-mediated apoptosis in endothelial cells: role of oxidantinduced iron signaling in apoptosis. J Biol Chem 2002;277:17179–17187. 59. Barrientos T, Laothamatas I, Koves TR, Soderblom EJ, Bryan M, Moseley MA, et al. Metabolic catastrophe in mice lacking transferrin receptor in muscle. EBioMedicine 2015;2:1705–1717. 60. Martinez-Redondo V, Pettersson AT, Ruas JL. The hitchhiker’s guide to PGC-1alpha isoform structure and biological functions. Diabetologia 2015;58:1969–1977. 61. Mutikainen M, Tuomainen T, Naumenko N, Huusko J, Smirin B, Laidinen S, et al. Peroxisome proliferator-activated receptorgamma coactivator 1 alpha1 induces a Prevention of chemotherapy-induced cachexia 431 Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265 cardiac excitation-contraction coupling phenotype without metabolic remodelling. J Physiol 2016;594:7049–7071. 62. Kunisada K, Negoro S, Tone E, Funamoto M, Osugi T, Yamada S, et al. Signal transducer and activator of transcription 3 in the heart transduces not only a hypertrophic signal but a protective signal against doxorubicin-induced cardiomyopathy. Proc Natl Acad Sci U S A 2000;97:315–319. 63. Biesemann N, Mendler L, Wietelmann A, Hermann S, Schafers M, Kruger M, et al. Myostatin regulates energy homeostasis in the heart and prevents heart failure. Circ Res 2014;115:296–310. 64. Yndestad A, Ueland T, Oie E, Florholmen G, Halvorsen B, Attramadal H, et al. Elevated levels of activin A in heart failure: potential role in myocardial remodeling. Circulation 2004;109:1379–1385. 65. Hulmi JJ, Oliveira BM, Silvennoinen M, Hoogaars WM, Pasternack A, Kainulainen H, et al. Exercise restores decreased physical activity levels and increases markers of autophagy and oxidative capacity in myostatin/activin-blocked mdx mice. Am J Physiol Endocrinol Metab 2013;305: E171–E182. 66. Morine KJ, Bish LT, Selsby JT, Gazzara JA, Pendrak K, Sleeper MM, et al. Activin IIB receptor blockade attenuates dystrophic pathology in a mouse model of Duchenne muscular dystrophy. Muscle Nerve 2010;42:722–730. 67. von Haehling S, Morley JE, Coats AJ, Anker SD. Ethical guidelines for publishing in the Journal of Cachexia, Sarcopenia and Muscle: update 2015. J Cachexia Sarcopenia Muscle 2015;6:315–316. 432 J.J. Hulmi et al. Journal of Cachexia, Sarcopenia and Muscle 2018; 9: 417–432 DOI: 10.1002/jcsm.12265