Inferring the in vivo cellular program of developing bovine skeletal muscle from expression data Nicholas J. Hudson a,⇑ , Russell E. Lyons a , Antonio Reverter a , Paul L. Greenwood b , Brian P. Dalrymple a a Computational and Systems Biology Group, CSIRO Food Futures and CSIRO Animal, Food and Health Sciences, 306 Carmody Road, St. Lucia, QLD 4072, Australia b New South Wales Department of Primary Industries, University of New England, Armidale, NSW 2351, Australia article info Article history: Received 5 April 2012 Received in revised form 4 February 2013 Accepted 6 February 2013 Available online 16 February 2013 Keywords: Muscle development Transcription factors Mitochondria Epigenetic abstract We outline an in vivo cellular program of bovine longissimus muscle development inferred from expres- sion data from 60 days post conception to 3 months postnatal. Analytic challenges included changes in cellular composition, ambiguous ‘diagnostic’ markers of cell type and contrasts between cattle human and mouse myogenesis. Nevertheless, the expression profiles of the myosin isoforms support slow and fast muscle fibres emanating from primary and secondary myogenesis respectively, while expression of the prenatal myosin subunits is down regulated prior to birth. Of the canonical pro-myogenic tran- scription factors (TF), MYF6 and MYF5 are negatively co-expressed, with MYF6 displaying higher expres- sion in the post-natal samples and MYF5, MYOG, HES6 and PAX7 displaying higher expression in early development. A set of TFs (SIX1, EYA2 and DACH2) considered important in undifferentiated murine cells were equally abundant in differentiated bovine cells. An examination of mammalian regulators of fibre composition, muscle mass and muscle metabolism, underscored the roles of PPARGC1A, TGFb signalling and the NHR4 Nuclear Hormone Receptors on bovine muscle development. Enriched among the most variably expressed genes from the entire data set were molecules regulating mitochondrial metabolism of carbohydrate (PDK4), fat (UCP3), protein (AGXT2L1) and high energy phosphate (CKMT2). The dramatic increase in the expression of these transcripts, which may enable the peri-natal transition to metabolic independence critical for new-born herbivores, provides surprising evidence for substantial developmen- tal remodelling of muscle mitochondria and reflects changes in nutrient availability. Overall, despite dif- ferences in size, metabolism and physiology, the muscle structural subunit expression program appears very similar in ruminants, rodents and humans. Crown Copyright Ó 2013 Published by Elsevier B.V. All rights reserved. We are interested in determining the mammalian molecules that play key roles in muscle mass, muscle fibre composition and muscle metabolism. Gaining this developmental understanding in laboratory rodent models presents technical challenges, mainly because the early prenatal muscles cannot be reliably identified. On the other hand, isolated in vitro work, such as that on C2C12 mouse monolayer cells (Berry et al., 2002; Joulia et al., 2003; Taylor et al., 2001) is experimentally convenient, but has the disadvantage of over-simplifying a 3-dimensional organ that is functionally con- nected to other components of living systems such as the brain and liver. We focus our analysis on in vivo developmental samples from bovine epaxial m. longissimus, a large back skeletal muscle that spans the entire thoracic and lumbar regions enabling head flexion and vertebral column extension. From a comparative perspective, domestic bovids offer unique insights into skeletal muscle devel- opment and physiology. Relative to rodents, they possess a body mass four orders of magnitude larger (Hoppeler and Fluck, 2002), they are thermally inert (Vaughan et al., 1999), possess an excep- tional capacity to burn volatile fatty acids (Bell, 1979) and are ‘‘pre- cocious’’ developers (Cronje, 2000), capable of supporting themselves against gravity almost immediately post-partum. These combined factors place particular constraints on, and also present certain challenges to skeletal muscle development in cattle. Compared to similarly-sized horses, cattle are sedentary and unathletic, reflected by a Field Metabolic Rate (FMR) less than twice Basal Metabolic Rate (BMR) (Havstad and Malechek, 1982) and a low VO 2MAX (Hoppeler and Fluck, 2002). Phenotypic adjust- ments include reductions in both muscle microvasculature density and mitochondrial mass (Hoppeler and Fluck, 2002). It has been ar- gued that these reductions in spare capacity increase livestock metabolic efficiency through engineering considerations of ‘eco- nomic design’ (Hudson, 2009). Paying the biological construction and maintenance costs for physiological capacity one does not need is energetically wasteful (Weibel et al., 1991). Arguably, artificial selection for extreme phenotypes relating to muscle and metabolism (Diamond, 2002) makes production 1567-133X/$ - see front matter Crown Copyright Ó 2013 Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.gep.2013.02.001 ⇑ Corresponding author. Tel.: +61 7 3214 2212; fax: +61 7 3214 2200. E-mail address: nick.hudson@csiro.au (N.J. Hudson). Gene Expression Patterns 13 (2013) 109–125 Contents lists available at SciVerse ScienceDirect Gene Expression Patterns journal homepage: www.elsevier.com/locate/gep