Gene array and real time PCR analysis of the adrenal sensitivity to adrenocorticotropic hormone in pig

The hypothalamic-pituitary-adrenal (HPA) axis and more particularly the adrenal gland constitute a principal node of the mammalian endocrine system. The main function of the adrenal cortex is to produce glucocorticoids (cortisol in pig) and mineralocorticoids under the influence of pituitary adrenocorticotropic hormone (ACTH). Adrenal hormones, essential for survival, play important roles in stress responses, metabolism regulation, immunity, reproduction, water and salt balance and various brain functions [1].

Large variations in HPA axis activity (i.e. basal and in response to stress) have been related to genetic factors (in human [2], in mice [3], in rats [4–6], in pigs [7] and in birds [8–10]). Variations in HPA axis activity are also related to important metabolic differences. For example, human abdominal obesity has been associated with alterations in HPA axis functioning [11, 12]. In a recent experiment comparing five genetic stocks of pigs (Large White, Landrace, Duroc, Meishan and Piétrain), a positive relationship between cortisol levels in urine (basal and after transportation stress) and body fat content was found both within and across breeds [13]. In addition, Meishan pigs have been reported to exhibit higher basal cortisol levels, response to ACTH and body fat content than Large White pigs [7, 14]. Meishan and Large White lines of pigs thus constitute a valuable biological model to investigate the relationship between HPA axis activity and metabolic regulation.

Among the different genetic mechanisms involved in HPA axis activity variability [15], several experimental findings suggest that sensitivity to ACTH is a major target in human [16] and in rats [4, 17]. In the pig, genetic-based differences in cortisol secretion were shown in response to corticotropin-releasing hormone (CRH) although the ACTH response did not differ among individuals [18]. Moreover, metabolic clearance of cortisol bears no relationship to the cortisol response to ACTH [19]. Previous findings [14] indicate that the difference in HPA axis activity between LW and MS pigs may originate from the adrenal gland although these breeds also differ in corticosteroid-binding globulin (CBG) levels that influence circulating levels of cortisol [20]. In this study, we explore the molecular mechanism responsible for the difference in adrenal sensitivity to ACTH in MS and LW pigs.

The actions of ACTH in the adrenal cortex are mediated via two temporally distinct pathways. Acute and chronic regulation of steroidogenesis occur within minutes and hours, respectively [21, 22]. The acute response is initiated by the mobilization and delivery of the substrate, cholesterol, for steroid hormone biosynthesis from the outer to the inner mitochondrial membrane, where it is metabolized to pregnenolone by the cytochrome P450 cholesterol side chain cleavage enzyme (P450scc) [23]. The slower, long-lasting response to ACTH directs transcription of genes encoding the steroidogenic enzymes [22, 24]. Most studies of ACTH action have focused on the long term induction (several hours) of transcripts either by investigating transcripts with specific functions in steroidogenesis [22, 24] or by genome-wide analysis [25–27]. However, the acute response to ACTH stimulation has been reported to require de novo protein synthesis (for review see ref. [21]). The steroidogenic acute regulatory protein (StAR), that is responsible for the transfer of cholesterol from the outer to the inner mitochondrial membrane, has been proposed to be the rate-limiting and regulated step in steroidogenesis [23, 28]. Changes in Star gene expression induced by ACTH have been observed as early as 30 min and were maximally elevated between 1 and 3 h [29]. On the other hand, cortisol responses to ACTH injections in Meishan and Large White pigs have been reported to be at maximum levels at 1 hour [14, 30]. Taking together these findings, we hypothesized that transcriptional regulation may in fact take place in the acute response to ACTH, i.e. within 1 hour, in both lines of pigs. Differences in transcriptional regulation at the adrenal gland could already exist between both genotypes in basal conditions because basal cortisol levels are typically greater in MS pigs.

Considering the findings described above and the literature concerning acute regulation of transcripts by ACTH in adrenals, we undertook a microarray analysis of gene expression in the adrenal glands of MS and LW pigs under basal conditions and in response to acute stimulation by ACTH. The aim of this study was to investigate the molecular bases of the genetic differences in adrenal sensitivity to ACTH. We found that genotype and/or ACTH affected the expression of 211 genes which provide new potential candidate genes to explain genetic variations in the adrenal sensitivity to ACTH.

We used a comprehensive gene expression profiling by microarray analysis to identify groups of genes differentially expressed by genotype and/or by acute ACTH treatment. This is the first gene array analysis investigating in vivo adrenal response to an acute ACTH stimulation and exploring genetic variability at the adrenal level by using different breeds of pigs (i.e. an interesting biological model because pigs produce cortisol as humans). In microarrays, which included almost 8960 transcripts, the present results indicate that genotype and/or ACTH treatment affected the levels of 211 genes in adrenals. Moreover, although previous gene array analyses of ACTH action have been conducted in vitro and/or focused on the effects of chronic stimulation [25, 27], our experiments demonstrate in vivo that acute ACTH treatment affects a large number of transcripts.

For the vast majority of affected transcripts, the changes were less than 2-fold excepted for some, probably of great interest, for which changes were as much as 4.5-fold. While some studies reported that small changes may be due to the tendency of microarray analysis to underestimate fold changes in transcripts accumulation [31], comparison of fold changes between microarray and real-time PCR analysis showed the accuracy of nylon microarrays used, even if fold-changes higher than 2-fold seemed to be somewhat underestimated. Our results suggest that genotypic difference and ACTH action may produce relatively small changes in transcript accumulation but these small changes could well be of physiological significance [32].

The dose of ACTH used in our study has been shown to maximally activate cortisol production within one hour [14, 30] and thus, may produce maximum effects on steroidogenesis. Cortisol concentrations measured in our study were consistent with previous results reporting that cortisol levels induced by ACTH were higher in Meishans than in Large Whites, as well as basal cortisol levels [7, 14]. However, ACTH did not affect gene expression of steroidogenesis enzymes (i.e. Cyp11a1, Cyp17, 3βhsd, Cyp21, Cyp11b). This result is consistent with the fact that acute stimulation by ACTH had little effect on adrenal P450s and steroidogenic enzymes while in contrast, long-term ACTH treatment provokes profound changes in the mRNA levels of many adrenal steroidogenic enzymes [22, 29, 33]. Interestingly, no differences in expression of steroidogenesis genes were found between Meishan and Large White pigs. These results suggest that the difference of corticosteroidogenesis between Meishan and Large White pigs is not triggered by changes in gene expression of adrenal P450 and 3βHSD enzymes under basal state or following acute ACTH stimulation. Nevertheless, we cannot exclude that steroidogenic activity might be higher in Meishan pigs than in Large White pigs. Indeed, expression of several genes (Mdh2, Sdha and Suclg2) involved in the tricarboxylic acid (TCA) pathway was greater in MS pigs. The main catalytic function of TCA cycle is to provide reducing equivalents to the respiratory complexes [34], for example, steroid hydroxylation [35]. Moreover, the TCA cycle also contributes to the synthesis of heme [34], necessary for the prosthetic groups of the steroidogenic cytochrome P450s [36]. In this respect, it is worth noting that Alas1, the rate-limiting enzyme in heme biosynthesis [37], shows a greater expression in MS than in LW pigs (see additional file 1). These mechanisms may together result in a higher steroidogenic activity by supplying more reducing equivalents and heme to steroidogenic enzymes. This hypothesis is supported by previous observations indicating that heme availability limited adrenal corticosteroid biosynthesis [38] and by recent data showing that acute stimulation of steroid production by ACTH was significantly increased when heme oxygenase activity was inhibited [39].

The effects of ACTH are mediated through the ACTH receptor (MC2R) belonging to the melanocortin receptor family (MCRs). The binding of ACTH to its cognate G protein-coupled receptor promotes the activation of protein kinase A and MAPK-dependent signaling cascades that collectively initiate adrenal-specific steroidogenic transcriptional programs [21, 22, 40]. Findings from numerous in vitro studies support the notion that ACTH is a positive regulator of ACTH-R mRNA expression [33, 41, 42]. More particularly, Winnay and Hammer [43] demonstrated in vitro that ACTH stimulation acutely activates the Mc2r gene promoter (i.e. within 80 min). In our study, we showed that ACTH did not affect Mc2r gene expression within 1 hour and that this gene was not differentially expressed between MS and LW pigs. Similarly, no difference in gene expression encoding key regulator proteins of the ACTH signaling pathway (i.e. G protein, Adenylate Cyclase, Protein Kinase A and MAPK ERK1, ERK2) was found in our study. On the other hand, the levels of Crem (cAMP response element modulator), a cAMP-dependent transcription factor that functions to activate genes involved in steroidogenesis [44], were increased in presence of ACTH and higher in MS than LW pigs. Moreover, trophic hormone stimulation of steroidogenic cells has been shown to result in the activation of G proteins that stimulate adenylate cyclase activity and produce increased intracellular levels of cAMP [21]. Thus, high levels of Crem in MS compared to LW and in response to ACTH may result to a larger increase of signal transduction induced by ACTH.

A constant supply of cholesterol is required within adrenal cells for steroidogenesis. The rate-limiting step for steroidogenesis is the movement of unesterified cholesterol into mitochondria where it can then be metabolized by CYP11A1 and other enzymes in the steroidogenic pathway (for review see ref. [45]). The movement of cholesterol into the mitochondria is mediated by steroidogenic acute regulatory protein (StAR) [23] and other partners such as peripheral-type benzodiazepine receptor or translocator protein [46]. Interestingly, greater expression of the Star gene in Meishan than in Large White pigs found in our study suggests that enhanced cholesterol transport into mitochondria may contribute to the higher corticosteroid biosynthesis found in MS pigs compared to LW pigs. All studies on StAR function agree that this enzyme mediates acute stimulation of steroidogenesis in response to ACTH administration and requires de novo protein synthesis (for review see ref. [21]). However, we did not find changes in transcript levels of Star 1 h following ACTH treatment while changes in Star gene expression induced by ACTH have been previously observed as early as 30 min and levels were maximally elevated between 1 and 3 h in the rat [29]. Nevertheless, phosphorylation of more preexisting StAR protein in MS pigs, a mechanism involved in the acute response to ACTH stimulation [47], could contribute to the higher cortisol levels induced in response to ACTH in MS than in LW pigs.

The unesterified cholesterol needed for steroidogenesis can be derived from several different sources (for review see ref. [45]). In our study we did not find differences in gene expression of enzymes involved in endogenous cellular cholesterol synthesis, such as Sqle or Hmgcr, between genotypes or in response to ACTH while those genes have been reported to be regulated by ACTH [25, 27]. On the other hand, cellular cholesterol delivery for steroidogenesis includes uptake of lipoprotein-derived cholesterol via low density lipoprotein (LDL) receptor mediated endocytic pathways and SRB1 (Scavenger Receptor class B, type1) mediated "selective" pathways (for review see ref. [45]). Interestingly we found higher Ldlr expression in MS than in LW pigs and in response to ACTH but no changes were observed in Scarb1 expression.

Changes in Ldlr expression in response to ACTH found in our study are consistent with Ldlr up-regulation reported in vitro following 24 h ACTH treatment in Y1 mouse adrenal cells [25]. Conversely, while Scarb1 have also been reported to be up-regulated by 24 h ACTH treatment in Y1 mouse adrenal cells [25], and to a larger extent than Ldlr, we did not find change in Scarb1 expression. Our results indicate that receptor-mediated endocytic uptake of LDL-cholesterol may be a more important source of cholesterol for adrenal steroidogenesis in pigs as is the case in humans [48], while it appears to play a negligible role in mouse [45]. Moreover, this is the first evidence indicating in vivo that the acute response to ACTH may involve cellular cholesterol supply for steroidogenesis via Ldlr regulation.

A large number of genes found to be differentially expressed in our study encode transcription factors. Most of them have not been yet described as requisite in transcription networks involved in adrenal steroidogenesis. Nevertheless, they are potential interesting candidates, particularly those that were affected by both genotype and ACTH treatment. We were particularly interested in peripheral clock genes (such as Bmal1, Per2, Cry1), because recent studies reported that in the adrenal they regulate a large number of genes involved in general cellular processes (e.g. protein synthesis) as well as in pathways related to major organ-specific function (e.g. corticosteroid biosynthesis) and probably adjust adrenal sensitivity to ACTH [49–51]. Interestingly, while Per2 and Cry1 were not differentially expressed in our study, Bmal1 gene showed less expression in MS pigs. Thus, we can not exclude that differences in the expression level of some clock genes may be involved in differences in basal cortisol levels and adrenal reactivity to ACTH between MS and LW pigs. Further studies are needed to investigate other clock genes and to clarify how clock-controlled transcriptional rhythms in adrenals could contribute to the differences observed between both lines of pigs.

Phosphorylation and dephosphorylation mechanisms might be also involved in the difference of adrenal function between LW and MS pigs because a large number of differentially expressed genes encode diverse protein kinases and protein phosphatases. Among them Snf1lk (SIK1 protein) constitute a valuable candidate since it was reported to be an important regulator in the early phase of ACTH-signaling in the adrenal cortex [52].

In conclusion, we report differential gene expression in adrenal in two lines of pigs in basal conditions and following acute ACTH treatment. Some of the genes have been already reported to be implicated in adrenal physiology, but the majority has not been documented as directly involved in steroidogenesis regulation or as acutely ACTH-responsive. Although their contributions to adrenal function merit further investigation, we may speculate on the involvement of a few of them. The higher cortisol levels in basal state and in response to ACTH in MS than in LW pigs was probably not triggered by changes in gene expression of known key regulator proteins of the ACTH signaling pathway and steroidogenic enzymes. However, a higher TCA cycle activity in MS pigs than in LW pigs may explain the higher steroidogenic activity by supplying more reducing equivalents and heme to steroidogenic enzymes. Alternately, differences in the adrenal function between MS and LW pigs involve likely mechanisms requisite for cholesterol supply to steroidogenesis. The genes described in this report are thus excellent potential candidates to mediate the genetic differences in adrenal steroidogenesis, particularly those affected by genotype and ACTH. Because dysregulation of glucocorticoid production results in diverse diseases, elucidation of the function of these genes in adrenals will lead to better understanding the molecular basis of such pathological conditions.