Differential expression of genes
Immediately following exercise (T1
) two probes were significantly (P
< 0.05) differentially regulated. Four hours (T2
) after exercise 1,485 probes were differentially expressed with fold changes ranging from +4.8-fold to -2.9-fold. At T2
, 923 probes were up-regulated and 562 probes were down-regulated. At the chosen significance threshold (α = 0.05) 74 of these probes are likely to be false positives. The probes with the greatest changes in expression (> +1.5-fold) immediately post-exercise are shown in Table . The probes with the greatest changes in expression (> +1.5-fold or -1.5-fold) four hours post exercise are shown in Table (up-regulated) and Table (down-regulated). A full list of gene expression changes at T1
are available in additional files 2
. The equine cDNA microarray expression data generated was deposited in the NCBI Gene Expression Omnibus (GEO) repository with experiment series accession [GEO:GSE16235].
Genes ≥ +1.5-fold (up-regulated) differential expression immediately post-exercise compared to pre-exercise levels.
Genes ≥ +1.5-fold (up-regulated) differential expression four hours post-exercise compared to pre-exercise levels.
Genes ≥ -1.5-fold (down-regulated) differential expression four hours post-exercise compared to pre-exercise levels.
Among the probes with the greatest expression changes (> +1.5-fold) at T1were seven probes representing four genes: three probes representing FOS (v-fos FBJ murine osteosarcoma viral oncogene homolog gene; mean +1.9-fold, unadjusted P = 0.004, 0.003, 0.039); two probes representing HSPA1A (heat shock 70 kDa protein 1A gene; mean +2.7-fold, unadjusted P = 1.50E-07, 2.42E-05); one probe located ~ 2kb upstream of PFKFB3 (6-phosphofructo-2-kinase/fructose-2,6-biphosphatase 3 gene; +2.0-fold, unadjusted P = 4.71E-06) and one probe representing EGR1 (early growth response 1 gene; +1.6-fold, unadjusted P = 0.014).
The gene expression changes observed for the FOS
genes are consistent with previous mammalian studies that have shown increased expression of these genes in response to exercise [24
]. HSPA1A, FOS
are members of the immediate-early response (IER) gene family. These genes are early regulators of cell growth and differentiation signals, and are induced in response to a wide variety of stress stimuli [31
]. The heat shock protein Hsp70, encoded by the HSPA1A
gene, is known to protect skeletal muscle cells against the path physiological effects of oxidative stress. In transgenic mouse models this cytoprotection is brought about both through improvement in muscle function and decreased apoptosis [32
]. It has been suggested that the cytoprotective effects of the Hsp70 protein are related to an ability to assist with the refolding of denatured or partially degraded proteins [35
]. Hsp70 can also interact with proteins involved in the regulation of cellular redox balance and Ca2+
homeostasis, and thus reduce oxidative stress and Ca2+
overload in response to physiological stress [36
]. In addition Hsp70 protects against muscular degeneration and atrophy [37
] through inhibition of caspase activation [38
] and protein catabolism [37
] and Hsp70 protein levels have been shown to correlate with muscular regeneration following injury [39
]. Together these facts highlight the key role of Hsp70 in muscle protection following stress and as a modulator of muscular regeneration. The HSPA1A
gene displayed a further increase in transcript expression at T2
< 0.001), whereas the expression of FOS, EGR1
had returned to resting levels. This suggests that while FOS
responses may be immediate and transient, the HSPA1A
response likely contributes to long term adaptation.
The probe upstream of the PFKFB3
gene shares strong homology to mammalian homologues of the gene thus it is likely that it represents expression of this gene product. The product of the PFKFB3
gene is involved in various aspects of energy sensing and metabolism, but has not previously been shown to be increased due to exercise. However, studies have shown increased expression of PFKFB3
in response to glucose deprivation [40
] and hypoxia [41
], both stimuli associated with exercise. The PFKFB3 protein is a powerful activator of glycolysis [42
]. Surprisingly, in a panel of genes encoding glycolytic enzymes and other anaerobic metabolites, differential mRNA expression was not observed in this experimental cohort despite significant increases in plasma lactate concentrations [43
]. Similar observations of a lack of transcriptional activation of glycolytic genes have been made in human exercise studies [44
]. PFKFB3 is also involved in glucose-induced insulin secretion in pancreatic β cells [45
] and a SNP in the 3' untranslated region of the PFKFB3
gene is associated with obesity in humans [46
]. The PFKFB3
gene promoter contains hypoxic response elements necessary for transactivation by hypoxia-inducible factor-1 alpha (HIF-1α) in response to hypoxia [47
]. This is relevant considering the observed increase in HIF-1α protein in this cohort of horses immediately after exercise [43
There was some overlap among probes differentially expressed at T2 and those tending towards differential expression at T1. Among the 434 probes tending towards differential expression (unadjusted P < 0.05) at T1 154 were also among those at T2, which is more than twice as many expected by chance. Over 96% of the genes had both the same direction of regulation at both time-points and a greater magnitude of change at T2. Two genes had a greater magnitude of change at T1 and a different directionality was observed for four genes. The genes with the highest observed fold changes at both T1 and T2 included HSPA1A (heat shock 70 kDa protein 1A gene, T1: +2.6-fold (mean of two probes), unadjusted P = 1.22E-05; T2: +4.8-fold, P = 1.61E-05); CRTC2 (CREB regulated transcription coactivator 2 gene, T1: +1.3-fold, adjusted P = 0.001; T2: +1.7-fold, P = 0.003); and SLC16A3 (solute carrier family 16, member 13 gene, T1: +1.2-fold, adjusted P = 0.03; T2: +1.6-fold, P = 0.012).
The CRTC2 protein is a potent activator of PGC-1α (peroxisome proliferater-activated receptor gamma coactivator 1 alpha), the master regulator of mitochondrial biogenesis [48
] and is also involved in the modulation of gluconeogenesis [49
]. The SLC16A3 protein is found in greater abundance in fast twitch rather than slow twitch muscle [50
] and plays a direct role in lactate efflux out of skeletal muscle. Thoroughbred horses have a strikingly high proportion of fast to slow twitch muscle fibres [51
], which was also observed in this cohort of horses [43
]. Increased mRNA levels of SLC16A3
were observed in "race fit" compared to moderately conditioned Standardbred horses [52
]. SLC16A3 also plays a role in the transport of the performance enhancing drug gamma-hydroxybutyric acid (GHB) [53
]. GHB is an endogenous metabolite but can also be administered orally as a performance-enhancing drug; therefore it is reasonable to hypothesize that endogenous GHB metabolism is associated with natural athletic ability. This hypothesis is supported by the observation that the alcohol dehydrogenase iron-containing protein 1 gene (ADHFE1
), which is involved in GHB catabolism [54
] is located in a genomic region that has been a target for positive selection during four hundred years of Thoroughbred evolution [55
Overall, these data suggest that, in addition to a rapid and dramatic induction of a small number of stress response genes immediately after exercise, there are also more subtle early changes in gene expression that are difficult to detect but are functionally relevant. It is possible that many of the genes differentially expressed at T2 were also differentially expressed at T1, but show more gradual changes in gene expression and were not detectable at that time point.