Comparative Transcriptome and Methylome Analysis in Human Skeletal Muscle Anabolism, Hypertrophy and Epigenetic Memory

Daniel C Turner, Robert A Seaborne, Adam P Sharples, Daniel C Turner, Robert A Seaborne, Adam P Sharples

Abstract

Transcriptome wide changes in human skeletal muscle after acute (anabolic) and chronic resistance exercise (RE) induced hypertrophy have been extensively determined in the literature. We have also recently undertaken DNA methylome analysis (850,000 + CpG sites) in human skeletal muscle after acute and chronic RE, detraining and retraining, where we identified an association between DNA methylation and epigenetic memory of exercise induced skeletal muscle hypertrophy. However, it is currently unknown as to whether all the genes identified in the transcriptome studies to date are also epigenetically regulated at the DNA level after acute, chronic or repeated RE exposure. We therefore aimed to undertake large scale bioinformatical analysis by pooling the publicly available transcriptome data after acute (110 samples) and chronic RE (181 samples) and comparing these large data sets with our genome-wide DNA methylation analysis in human skeletal muscle after acute and chronic RE, detraining and retraining. Indeed, after acute RE we identified 866 up- and 936 down-regulated genes at the expression level, with 270 (out of the 866 up-regulated) identified as being hypomethylated, and 216 (out of 936 downregulated) as hypermethylated. After chronic RE we identified 2,018 up- and 430 down-regulated genes with 592 (out of 2,018 upregulated) identified as being hypomethylated and 98 (out of 430 genes downregulated) as hypermethylated. After KEGG pathway analysis, genes associated with 'cancer' pathways were significantly enriched in both bioinformatic analysis of the pooled transcriptome and methylome datasets after both acute and chronic RE. This resulted in 23 (out of 69) and 28 (out of 49) upregulated and hypomethylated and 12 (out of 37) and 2 (out of 4) downregulated and hypermethylated 'cancer' genes following acute and chronic RE respectively. Within skeletal muscle tissue, these 'cancer' genes predominant functions were associated with matrix/actin structure and remodelling, mechano-transduction (e.g. PTK2/Focal Adhesion Kinase and Phospholipase D- following chronic RE), TGF-beta signalling and protein synthesis (e.g. GSK3B after acute RE). Interestingly, 51 genes were also identified to be up/downregulated in both the acute and chronic RE pooled transcriptome analysis as well as significantly hypo/hypermethylated after acute RE, chronic RE, detraining and retraining. Five genes; FLNB, MYH9, SRGAP1, SRGN, ZMIZ1 demonstrated increased gene expression in the acute and chronic RE transcriptome and also demonstrated hypomethylation in these conditions. Importantly, these 5 genes demonstrated retained hypomethylation even during detraining (following training induced hypertrophy) when exercise was ceased and lean mass returned to baseline (pre-training) levels, identifying them as genes associated with epigenetic memory in skeletal muscle. Importantly, for the first time across the transcriptome and epigenome combined, this study identifies novel differentially methylated genes associated with human skeletal muscle anabolism, hypertrophy and epigenetic memory.

Conflict of interest statement

The authors declare no competing interests.

Figures

Figure 1
Figure 1
(A) Venn Diagram demonstrating out of 866 genes upregulated after acute RE (p ≤ 0.01) in the pooled transcriptomic analysis, 270 of these genes were significantly hypomethylated in the methylome analysis. (B) Venn Diagram demonstrating out of 936 genes downregulated after acute RE (p ≤ 0.01) in the pooled transcriptomic analysis, 216 of these genes were significantly hypermethylated in the methylome analysis. Note: 5752/4604 total hypo/hypermethylated CpGs after acute RE is different than the original paper reporting 9,153/8,212 hypo/hypermethylated CpGs. This is due to the number of CpG’s that resided on the shared list of 14,992 annotated genes by ‘gene symbol’ across the pooled transcriptomic studies for acute RE, depicted above in 1A, in order to provide a direct comparison of CpG sites on the same genes from the pooled transcriptomic data set. (C) ‘Cancer’ pathway genes upregulated (GREEN bars) in pooled transcriptomic studies (p ≤ 0.01) and hypomethylated (BLUE bars) in (p ≤ 0.05) after acute RE. In skeletal muscle, the majority of these genes (13 out of 23) are associated with matrix / actin structure or remodelling and mechano-transduction in skeletal muscle (MSN THBS1, TIMP3, FLNB, LAMA5, CRK, COL4A1, ITGA2, ITGB3, CD63, CTTN, RASSF5, F2RL3) and 3 genes with TGF- Beta signalling (SMAD3, FOS, WNT9A), 2 genes with calcium signalling (ITPR3, ADCY3), 1 gene with IL-6 signalling (STAT3), and protein synthesis (GSK3B) and retinoic acid signalling (RARA). D, ‘Cancer’ genes downregulated (RED) in pooled transcriptomic studies and hypermethylated (YELLOW bars) in after acute RE.
Figure 2
Figure 2
(A) Venn Diagram analysis demonstrating out of 2,018 genes upregulated after chronic RE (p ≤ 0.01) in the pooled transcriptomic analysis, 592 of these genes were significantly hypomethylated in the methylome analysis after chronic RE. (B) Venn Diagram demonstrating out of 430 genes downregulated after chronic RE (p ≤ 0.01) in the pooled transcriptomic analysis, 98 of these genes were significantly hypermethylated in the methylome analysis after chronic RE. Note: 5262/4853 total hypo/hypermethylated CpGs after chronic RE is different than the original paper reporting 8,891/8,636 hypo/hypermethylated CpGs. This is due to the number of CpG’s that resided on the shared list of 15,317 annotated genes by ‘gene symbol’ across the pooled transcriptomic studies for chronic RE, depicted above in 2A, in order to provide a direct comparison of CpG sites on the same genes from the pooled transcriptomic data set. (C) ‘Cancer’ genes upregulated (GREEN bars) in pooled transcriptome studies (p ≤ 0.01) and hypomethylated (BLUE bars) (p ≤ 0.05) in methylome analysis after chronic RE. In skeletal muscle, the majority of these genes are associated with matrix and actin structure/remodelling COL4A2, HSPG2, ITGA6, TIAM1, CTTN, GNA12, ADCY4, BCR, PLCG2, FN1, FLNB, PLAUR, EZR), mechanotransduction (PTK2/FAK and PLD1) and TGF-Beta signalling (TGFB3, TGFBR2, LEF1, MECOM). (D) ‘Cancer’ genes downregulated (RED) in pooled transcriptomic studies and hypermethylated (YELLOW bars) in methylome analysis after chronic RE.
Figure 3
Figure 3
(A) Fold change in gene expression (normalised to 0) from pooled transcriptome analysis after acute and chronic RE and methylome analysis after acute and chronic RE, detraining and retraining: MYH9, SRGAP1, SRGN, ZMIZ1 and FLNB (left to right) demonstrated increased gene expression in both the transcriptome analysis after both acute and chronic RE and also demonstrated hypomethylation in these conditions in the methylome analysis. Importantly, these genes also demonstrated retained hypomethylation even during detraining (following training induced hypertrophy) when exercise was completely cessed and lean mass in the original study returned to baseline (pre-training) levels. (B) New analysis of FLNB gene expression via rt-qRT-PCR (ΔΔCt gene expression value normalised to 0) demonstrated that the gene significantly increased in expression after acute RE vs. pre/baseline (P = 0.01). The average fold change in FLNB was also increased after chronic RE (training) and remained elevated vs. pre/baseline after detraining (unloading) and retraining (reloading) However, these increases failed to reach statistical significance. In the methylation array data FLNB methylation demonstrated an inverse association with the gene expression where the gene was hypomethylated after acute and chronic RE, with sustained hypomethylation during detraining and retraining.

References

    1. Sharples AP, et al. Longevity and skeletal muscle mass: the role of IGF signalling, the sirtuins, dietary restriction and protein intake. Aging cell. 2015;14:511–523. doi: 10.1111/acel.12342.
    1. Sharples AP, Stewart CE, Seaborne RA. Does skeletal muscle have an ‘epi’-memory? The role of epigenetics in nutritional programming, metabolic disease, aging and exercise. Aging cell. 2016;15:603–616. doi: 10.1111/acel.12486.
    1. MacNeil LG, Melov S, Hubbard AE, Baker SK, Tarnopolsky MA. Eccentric exercise activates novel transcriptional regulation of hypertrophic signaling pathways not affected by hormone changes. PLoS ONE. 2010;5:e10695. doi: 10.1371/journal.pone.0010695.
    1. Murton AJ, et al. Transient transcriptional events in human skeletal muscle at the outset of concentric resistance exercise training. J Appl Physiol (1985) 2014;116:113–125. doi: 10.1152/japplphysiol.00426.2013.
    1. Raue U, et al. Transcriptome signature of resistance exercise adaptations: mixed muscle and fiber type specific profiles in young and old adults. J Appl Physiol. 2012;112:1625–1636. doi: 10.1152/japplphysiol.00435.2011.
    1. Vissing K, Schjerling P. Simplified data access on human skeletal muscle transcriptome responses to differentiated exercise. Scientific data. 2014;1:140041. doi: 10.1038/sdata.2014.41.
    1. Lundberg TR, Fernandez-Gonzalo R, Tesch PA, Rullman E, Gustafsson T. Aerobic exercise augments muscle transcriptome profile of resistance exercise. Am J Physiol Regul Integr Comp Physiol. 2016;310:R1279–1287. doi: 10.1152/ajpregu.00035.2016.
    1. Liu D, et al. Skeletal muscle gene expression in response to resistance exercise: sex specific regulation. BMC Genomics. 2010;11:659. doi: 10.1186/1471-2164-11-659.
    1. Phillips BE, et al. Molecular networks of human muscle adaptation to exercise and age. PLoS genetics. 2013;9:e1003389. doi: 10.1371/journal.pgen.1003389.
    1. Thalacker-Mercer A, et al. Cluster analysis reveals differential transcript profiles associated with resistance training-induced human skeletal muscle hypertrophy. Physiol Genomics. 2013;45:499–507. doi: 10.1152/physiolgenomics.00167.2012.
    1. Bogdanovic O, Veenstra GJ. DNA methylation and methyl-CpG binding proteins: developmental requirements and function. Chromosoma. 2009;118:549–565. doi: 10.1007/s00412-009-0221-9.
    1. Seaborne RA, et al. Human Skeletal Muscle Possesses an Epigenetic Memory of Hypertrophy. Scientific Reports (Nature) 2018;8:1898. doi: 10.1038/s41598-018-20287-3.
    1. Egner IM, Bruusgaard JC, Eftestøl E, Gundersen K. A cellular memory mechanism aids overload hypertrophy in muscle long after an episodic exposure to anabolic steroids. The Journal of Physiology. 2013;591:6221–6230. doi: 10.1113/jphysiol.2013.264457.
    1. Bruusgaard JC, Johansen IB, Egner IM, Rana ZA, Gundersen K. Myonuclei acquired by overload exercise precede hypertrophy and are not lost on detraining. Proc Natl Acad Sci USA. 2010;107:15111–15116. doi: 10.1073/pnas.0913935107.
    1. Sharples AP, et al. Skeletal muscle cells possess a ‘memory’ of acute early life TNF-alpha exposure: role of epigenetic adaptation. Biogerontology. 2016;17:603–617. doi: 10.1007/s10522-015-9604-x.
    1. Li Q, Birkbak NJ, Gyorffy B, Szallasi Z, Eklund AC. Jetset: selecting the optimal microarray probe set to represent a gene. BMC Bioinformatics. 2011;12:474–474. doi: 10.1186/1471-2105-12-474.
    1. Gravendeel LAM, et al. Gene expression profiles of gliomas in formalin-fixed paraffin-embedded material. British Journal of Cancer. 2012;106:538–545. doi: 10.1038/bjc.2011.547.
    1. Brannon AR, et al. Meta-analysis of clear cell renal cell carcinoma gene expression defines a variant subgroup and identifies gender influences on tumor biology. European urology. 2012;61:258–268. doi: 10.1016/j.eururo.2011.10.007.
    1. Nygaard V, Rodland EA, Hovig E. Methods that remove batch effects while retaining group differences may lead to exaggerated confidence in downstream analyses. Biostatistics (Oxford, England) 2016;17:29–39. doi: 10.1093/biostatistics/kxv027.
    1. Seaborne, R. A. et al. Methylome of human skeletal muscle after acute & chronic resistance exercise training, detraining & retraining. Scientific data 5, 180213, 10.1038/sdata.2018.213 (2018).
    1. Maksimovic J, Gordon L, Oshlack A. SWAN: Subset-quantile within array normalization for illumina infinium HumanMethylation450 BeadChips. Genome Biol. 2012;13:R44. doi: 10.1186/gb-2012-13-6-r44.
    1. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes. Nucleic Acids Res. 2000;28:27–30. doi: 10.1093/nar/28.1.27.
    1. Kanehisa M, Furumichi M, Tanabe M, Sato Y, Morishima K. KEGG: new perspectives on genomes, pathways, diseases and drugs. Nucleic Acids Res. 2017;45:D353–d361. doi: 10.1093/nar/gkw1092.
    1. Kanehisa M, Sato Y, Kawashima M, Furumichi M, Tanabe M. KEGG as a reference resource for gene and protein annotation. Nucleic Acids Res. 2016;44:D457–462. doi: 10.1093/nar/gkv1070.
    1. Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T) method. Nature protocols. 2008;3:1101–1108. doi: 10.1038/nprot.2008.73.
    1. Barres R, et al. Acute exercise remodels promoter methylation in human skeletal muscle. Cell Metab. 2012;15:405–411. doi: 10.1016/j.cmet.2012.01.001.
    1. Guzzoni V, et al. Effect of Resistance Training on Extracellular Matrix Adaptations in Skeletal Muscle of Older Rats. Frontiers in Physiology. 2018;9:374. doi: 10.3389/fphys.2018.00374.
    1. Xu Q, et al. The next hotspot in skeletal research? Journal of genetics and genomics = Yi chuan xue bao. 2017;44:335–342. doi: 10.1016/j.jgg.2017.04.007.
    1. Hyldahl RD, et al. Extracellular matrix remodeling and its contribution to protective adaptation following lengthening contractions in human muscle. Faseb j. 2015;29:2894–2904. doi: 10.1096/fj.14-266668.
    1. Kanazawa Y, et al. Effects of aging on basement membrane of the soleus muscle during recovery following disuse atrophy in rats. Exp Gerontol. 2017;98:153–161. doi: 10.1016/j.exger.2017.08.014.
    1. Sinanan AC, Machell JR, Wynne-Hughes GT, Hunt NP, Lewis MP. Alpha v beta 3 and alpha v beta 5 integrins and their role in muscle precursor cell adhesion. Biol Cell. 2008;100:465–477. doi: 10.1042/bc20070115.
    1. Nazari H, et al. Cortactin, an actin binding protein, regulates GLUT4 translocation via actin filament remodeling. Biochemistry. Biokhimiia. 2011;76:1262–1269. doi: 10.1134/s0006297911110083.
    1. Rayagiri SS, et al. Basal lamina remodeling at the skeletal muscle stem cell niche mediates stem cell self-renewal. Nature Communications. 2018;9:1075. doi: 10.1038/s41467-018-03425-3.
    1. Goel HL, Dey CS. PKC-regulated myogenesis is associated with increased tyrosine phosphorylation of FAK, Cas, and paxillin, formation of Cas-CRK complex, and JNK activation. Differentiation; research in biological diversity. 2002;70:257–271. doi: 10.1046/j.1432-0436.2002.700604.x.
    1. Kawauchi K, et al. p130Cas-dependent actin remodelling regulates myogenic differentiation. Biochem J. 2012;445:323–332. doi: 10.1042/bj20112169.
    1. Puntschart A, et al. Expression of fos and jun genes in human skeletal muscle after exercise. Am J Physiol. 1998;274:C129–137. doi: 10.1152/ajpcell.1998.274.1.C129.
    1. Chen Y-W, et al. Response of rat muscle to acute resistance exercise defined by transcriptional and translational profiling. The Journal of Physiology. 2002;545:27–41. doi: 10.1113/jphysiol.2002.021220.
    1. Trenerry MK, Carey KA, Ward AC, Cameron-Smith D. STAT3 signaling is activated in human skeletal muscle following acute resistance exercise. J Appl Physiol (1985) 2007;102:1483–1489. doi: 10.1152/japplphysiol.01147.2006.
    1. Zhang Y, Feng XH, Derynck R. Smad3 and Smad4 cooperate with c-Jun/c-Fos to mediate TGF-beta-induced transcription. Nature. 1998;394:909–913. doi: 10.1038/29814.
    1. Tando T, et al. Smad2/3 Proteins Are Required for Immobilization-induced Skeletal Muscle Atrophy. J Biol Chem. 2016;291:12184–12194. doi: 10.1074/jbc.M115.680579.
    1. Lessard SJ, et al. JNK regulates muscle remodeling via myostatin/SMAD inhibition. Nat Commun. 2018;9:3030. doi: 10.1038/s41467-018-05439-3.
    1. Zhang, W. et al. Synergistic effects of TGFbeta2, WNT9a, and FGFR4 signals attenuate satellite cell differentiation during skeletal muscle development. Aging cell, e12788, 10.1111/acel.12788 (2018).
    1. Powell JA, et al. IP3 receptors and associated Ca2+ signals localize to satellite cells and to components of the neuromuscular junction in skeletal muscle. J Neurosci. 2003;23:8185–8192. doi: 10.1523/JNEUROSCI.23-23-08185.2003.
    1. Ding Q, et al. Raf kinase activation of adenylyl cyclases: isoform-selective regulation. Molecular pharmacology. 2004;66:921–928.
    1. Li W, et al. Folic acid supplementation alters the DNA methylation profile and improves insulin resistance in high-fat-diet-fed mice. The Journal of nutritional biochemistry. 2018;59:76–83. doi: 10.1016/j.jnutbio.2018.05.010.
    1. Guadagnin, E., Mazala, D. & Chen, Y. W. STAT3 in Skeletal Muscle Function and Disorders. Int J Mol Sci19, 10.3390/ijms19082265 (2018).
    1. Begue G, et al. Early activation of rat skeletal muscle IL-6/STAT1/STAT3 dependent gene expression in resistance exercise linked to hypertrophy. PLoS ONE. 2013;8:e57141. doi: 10.1371/journal.pone.0057141.
    1. Perez-Schindler J, et al. Overload-mediated skeletal muscle hypertrophy is not impaired by loss of myofiber STAT3. Am J Physiol Cell Physiol. 2017;313:C257–c261. doi: 10.1152/ajpcell.00100.2017.
    1. Leger B, et al. Akt signalling through GSK-3beta, mTOR and Foxo1 is involved in human skeletal muscle hypertrophy and atrophy. J Physiol. 2006;576:923–933. doi: 10.1113/jphysiol.2006.116715.
    1. El Haddad M, et al. Retinoic acid maintains human skeletal muscle progenitor cells in an immature state. Cell Mol Life Sci. 2017;74:1923–1936. doi: 10.1007/s00018-016-2445-1.
    1. Collins BC, et al. Deletion of estrogen receptor alpha in skeletal muscle results in impaired contractility in female mice. J Appl Physiol (1985) 2018;124:980–992. doi: 10.1152/japplphysiol.00864.2017.
    1. Chen HH, et al. Estrogen/ERR-alpha signaling axis is associated with fiber-type conversion of upper airway muscles in patients with obstructive sleep apnea hypopnea syndrome. Scientific reports. 2016;6:27088. doi: 10.1038/srep27088.
    1. Yamashita Y, et al. Perlecan, a heparan sulfate proteoglycan, regulates systemic metabolism with dynamic changes in adipose tissue and skeletal muscle. Scientific reports. 2018;8:7766. doi: 10.1038/s41598-018-25635-x.
    1. You GY, et al. Tiam-1, a GEF for Rac1, plays a critical role in metformin-mediated glucose uptake in C2C12 cells. Cellular signalling. 2013;25:2558–2565. doi: 10.1016/j.cellsig.2013.08.018.
    1. von Holleben M, Gohla A, Janssen KP, Iritani BM, Beer-Hammer S. Immunoinhibitory adapter protein Src homology domain 3 lymphocyte protein 2 (SLy2) regulates actin dynamics and B cell spreading. J Biol Chem. 2011;286:13489–13501. doi: 10.1074/jbc.M110.155184.
    1. Sorensen JR, Skousen C, Holland A, Williams K, Hyldahl RD. Acute extracellular matrix, inflammatory and MAPK response to lengthening contractions in elderly human skeletal muscle. Exp Gerontol. 2018;106:28–38. doi: 10.1016/j.exger.2018.02.013.
    1. DiPasquale DM, et al. Urokinase-type plasminogen activator and macrophages are required for skeletal muscle hypertrophy in mice. Am J Physiol Cell Physiol. 2007;293:C1278–1285. doi: 10.1152/ajpcell.00201.2007.
    1. Deshmukh A, et al. Exercise-induced phosphorylation of the novel Akt substrates AS160 and filamin A in human skeletal muscle. Diabetes. 2006;55:1776–1782. doi: 10.2337/db05-1419.
    1. Ulbricht A, et al. Induction and adaptation of chaperone-assisted selective autophagy CASA in response to resistance exercise in human skeletal muscle. Autophagy. 2015;11:538–546. doi: 10.1080/15548627.2015.1017186.
    1. Franchi MV, et al. Regional regulation of focal adhesion kinase after concentric and eccentric loading is related to remodelling of human skeletal muscle. Acta physiologica (Oxford, England) 2018;223:e13056. doi: 10.1111/apha.13056.
    1. Fluck M, von Allmen RS, Ferrie C, Tevaearai H, Dick F. Protective effect of focal adhesion kinase against skeletal muscle reperfusion injury after acute limb ischemia. European journal of vascular and endovascular surgery: the official journal of the European Society for. Vascular Surgery. 2015;49:306–313. doi: 10.1016/j.ejvs.2014.11.011.
    1. Klossner S, Li R, Ruoss S, Durieux AC, Fluck M. Quantitative changes in focal adhesion kinase and its inhibitor, FRNK, drive load-dependent expression of costamere components. Am J Physiol Regul Integr Comp Physiol. 2013;305:R647–657. doi: 10.1152/ajpregu.00007.2013.
    1. Teng S, et al. Phospholipase D1 facilitates second-phase myoblast fusion and skeletal muscle regeneration. Mol Biol Cell. 2015;26:506–517. doi: 10.1091/mbc.E14-03-0802.
    1. Jaafar R, et al. Phospholipase D regulates the size of skeletal muscle cells through the activation of mTOR signaling. Cell communication and signaling: CCS. 2013;11:55. doi: 10.1186/1478-811x-11-55.
    1. O’Neil TK, Duffy LR, Frey JW, Hornberger TA. The role of phosphoinositide 3-kinase and phosphatidic acid in the regulation of mammalian target of rapamycin following eccentric contractions. J Physiol. 2009;587:3691–3701. doi: 10.1113/jphysiol.2009.173609.
    1. Aloysius, A., DasGupta, R. & Dhawan, J. The transcription factor Lef1 switches partners from beta-catenin to Smad3 during muscle stem cell quiescence. Science signaling11, 10.1126/scisignal.aan3000 (2018).
    1. Kurokawa M, et al. The oncoprotein Evi-1 represses TGF-beta signalling by inhibiting Smad3. Nature. 1998;394:92–96. doi: 10.1038/27945.
    1. Liang X, Kiru S, Gomez GA, Yap AS. Regulated recruitment of SRGAP1 modulates RhoA signaling for contractility during epithelial junction maturation. Cytoskeleton (Hoboken, N.J.) 2018;75:61–69. doi: 10.1002/cm.21420.
    1. Wong K, et al. Signal transduction in neuronal migration: roles of GTPase activating proteins and the small GTPase Cdc42 in the Slit-Robo pathway. Cell. 2001;107:209–221. doi: 10.1016/S0092-8674(01)00530-X.
    1. Hjorth, M. et al. The effect of acute and long-term physical activity on extracellular matrix and serglycin in human skeletal muscle. Physiological reports3, 10.14814/phy2.12473 (2015).
    1. Li X, Thyssen G, Beliakoff J, Sun Z. The novel PIAS-like protein hZimp10 enhances Smad transcriptional activity. J Biol Chem. 2006;281:23748–23756. doi: 10.1074/jbc.M508365200.
    1. Lee J, Beliakoff J, Sun Z. The novel PIAS-like protein hZimp10 is a transcriptional co-activator of the p53 tumor suppressor. Nucleic acids research. 2007;35:4523–4534. doi: 10.1093/nar/gkm476.
    1. Sharma M, et al. hZimp10 is an androgen receptor co-activator and forms a complex with SUMO-1 at replication foci. Embo j. 2003;22:6101–6114. doi: 10.1093/emboj/cdg585.

Source: PubMed

Sponsors et collaborateurs

Les conditions médicales

Interventions en matière de drogue

3
S'abonner