Peroxisome proliferator-activated receptor γ coactivator 1α (PPARGC1A) is a transcriptional coactivator with many diverse functions and has a pivotal role in fat and energy metabolism. This cold- and exercise-inducible gene is crucial to adaptive thermogenesis and is an essential regulator of adipogenesis, adipocyte differentiation and mitochondrial biogenesis/respiration [1–4]. Recently, it has been shown that it is also involved in angiogenesis [5]. It exerts its function through a whole range of nuclear hormone receptors and other transcription factors, and is primarily expressed in tissues with high energy demands [6]. Besides having an important influence on the regulation and composition of the body weight, it also is an important factor in determining muscle fibre type composition [7–9]. It has been shown that PPARGC1A increases the amount of oxidative muscle fibres, and that it also is expressed at a higher level in these muscle fibres.

For several reasons, porcine PPARGC1A is an interesting candidate gene for meat quality, an economically important and complex characteristic which is composed of many different traits. Associations have been found between mutations in the coding region of PPARGC1A and certain fat characteristics in the pig [10–12]. Other interesting findings are that PPARGC1A is the only candidate gene so far that was located in the QTL region for leaf fat weight and backfat on chromosome 8p21 [11, 13], and that other candidate genes for meat quality, like GLUT4, are regulated by PPARGC1A [14].

As explained above, PPARGC1A has many functions which can strongly differ between tissues. It has been shown that there is a variation in mRNA expression in the pig, not only between tissues, but also between different locations within the longissimus dorsi muscle [15]. Only very little is known about this multifunctional gene in the pig and its possible use as a selection marker in the pig industry. In order to get a better understanding of the regulation of the many functions of PPARGC1A, exon-spanning primers were used to construct a detailed transcription profile of its presence in 32 different tissues and several embryonic developmental stages in the pig. The aim was to identify possible splice variants, because they could provide an explanation for the regulation of the tissue-dependent functions of PPARGC1A.

Tissue samples were collected from a freshly slaughtered female, commercial, hybrid pig and immediately submerged in RNA later (Sigma-Aldrich, Bornem, Belgium), according to the instructions manual. Testis was collected from a similar male pig. Total RNA was extracted with the Aurum Total RNA Fatty and Fibrous Tissue Kit (Bio-Rad, Nazareth, Belgium), according to the manufacturer's protocol which included an on-column DNase treatment. Ovaries were collected at a local slaughterhouse from pigs at slaughter age, and used for in vitro embryo production as described in Bijttebier et al. [16]. RNA extraction from embryonic samples (for the 2–4 cell, 8 cell, morula and blastocyst stage respectively 15, 12, 8 and 6 pooled embryos were used) was performed with the PicoPure RNA Isolation Kit (Arcturus, Mountain View, USA), according to the instructions manual, after which a DNase treatment was carried out with RQ1 RNase-free DNase (Promega, Leiden, The Netherlands). Both DNase treatments were verified by a minus reverse transcription (RT) control PCR and RNA integrity was checked, as described in Erkens et al. [15]. Also, RNA purity and concentration were measured with the ND-1000 Spectrophotometer (NanoDrop, Wilmington, USA). Next, the iScript cDNA Synthesis Kit (which contains both oligo dT and random primers; Bio-Rad, Nazareth, Belgium) was used to convert approximately 1 μg of RNA from each sample to cDNA, according to the manufacturer's protocol. This RT step was verified by a control PCR [15], in which a no-template control was included to check for DNA contamination. Ready-to-use human cDNA from kidney and liver was provided by Prof. Vandesompele (Department of Pediatrics and Medical Genetics, Ghent University), to verify whether the detected splice variants also occur in human tissues.

The amplicons from all primer pairs, except PGC1A+/-Ex7,9, were sequenced by direct sequencing. Because the use of PGC1A+/-Ex7,9 resulted in multiple amplicons, the GENECLEAN II Kit (Qbiogene, Brussels, Belgium) was used to first isolate and purify the separate amplicons from the agarose gel, before sequencing. Sequencing of the amplicons was conducted on an Applied Biosystems 3730xl DNA Analyser with the BigDye Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Lennik, Belgium), according to the manufacturer's protocol.

All exon-spanning primer pairs in all tissues resulted in one amplicon of the expected length (Table 1), except for PGC1A+/-Ex7,9. With this last primer pair, 3 different amplicons were detected and the results were tissue-dependent (Table 2). The forward and reverse primer of PGC1A+/-Ex7,9 are located in exon 7 and exon 9, respectively, which means its amplicon completely contains exon 8. Besides the expected fragment of 1017 bp (which incorporates all 916 bp of exon 8), 2 new splice variants of PPARGC1A in the pig were detected. Sequencing revealed that in these splice variants exon 8 was either partly (the last 66 bp of exon 8 were conserved) or completely spliced out, which resulted in amplicons of 167 and 101 bp, respectively (Figure 1, 2). Additional PCRs with longer amplicons (exon 4–9 and exon 4–12) confirmed the existence of both exon 8 splice variants and indicate that the other exons were preserved in the rest of the transcript. Taken together with the fact that no pseudogenes for PPARGC1A are described in any species and the use of a genomic DNA control, this makes it unlikely that our observations are the result of pseudogenes.

As can be seen from Figure 2, both splice variant boundaries have GC-AG splice sites, instead of following the more common GT-AG rule. Table 2 shows that one or both splice variants were found in almost every tissue that was tested. They could also be detected in the pre-implantation embryonic stages, which implies PPARGC1A is involved in early development. An interesting finding however, is the fact that the complete amplicon (1017 bp) was not detected in any of the 4 embryonic stages. This could indicate certain functions of PPARGC1A are switched off or altered during early development, but without a functional analysis of the splice variants it is not possible to discuss the effects on its functionality or to give an explanation for these findings. In testis, a PCR artefact (221 bp) was detected with a 92% identity to human outer dense fibre of sperm tails 2 (ODF2) and was deposited in GenBank as an EST [GenBank:EY122774].

Only very little is known about the existence of splice variants of PPARGC1A in any species, and up until now nothing was known about it in the pig. There have been previous reports suggesting the existence of splice variants in which exon 8 was possibly spliced out, in rat skeletal muscle and brown adipose tissue, but their sequence was not determined [21, 22]. However, this is the first study giving a detailed description of the actual presence and sequence of 2 splice variants in a whole range of porcine tissues.

The 2 newly identified splice variants possibly give rise to a much shorter protein of respectively 359 and 337 aa, depending on whether exon 8 is partly or completely spliced out (Figure 3). This is much shorter than the complete protein with 796 aa. The first 291 aa of both variants are identical to the complete protein, but the rest of the aa sequence is completely different. It can be expected that this will have an important influence on the function of the produced protein (Figure 3). In the study by Baar et al. [22] an increase of a smaller PPARGC1A protein was detected after exercise in rats. This was consistent with the increase of the smaller cDNA band they detected, although it was not established if that protein originated from the shorter mRNA and if this had any functional significance.

The PPARGC1A protein can generally be divided into 3 regions (Figure 3a). The N-terminal region consists of a transcriptional activation domain which contains an essential LXXLL motif, and is involved in the activation of many transcription factors. A less distinct, central region contains both an inhibitory domain and several interaction domains (PPARγ, NRF-1, MEF2C). At the C-terminal end, an RNA processing domain is located, which comprises 2 serine-arginine-rich domains (SR) and an RNA recognition motif (RRM) [1, 23–25]. Figure 3 shows that in the putative protein from both splice variants the N-terminal activation domain is conserved. The central region however is only partly conserved. The aa in the first part of the inhibitory and NRF-1 domain are conserved, but the aa of the complete PPARγ interaction region are altered. The second part of the inhibitory domain and NRF-1 interaction region is either altered or absent. The RNA processing domain at the C-terminal end of the complete protein is completely absent in the splice variants. These results suggest that the putative proteins from both splice variants show some remarkable alterations and this is likely to have a large impact on the function of PPARGC1A.

Currently, human medicine shows a great interest in PPARGC1A, because of the important role this gene plays in the worldwide problems concerning obesity, insulin resistance and correlated diseases, such as type 2 diabetes mellitus [26, 27]. It has also been shown recently that a lower expression of PPARGC1A is involved in the onset of multiple neurodegenerative diseases, like Parkinson's, Alzheimer's and Huntington's disease [28]. Because of its significance, human kidney and liver tissue were also tested for the presence of the newly detected splice variants in the pig. This showed that both splice variants were also found in human liver and only the shortest splice variant in human kidney. In porcine kidney, both splice variants were detected, indicating the existence of possible species differences. The discovery of these new splice variants could therefore be of importance for the human research regarding PPARGC1A as well.

The results from this study contribute to a better understanding of this complex gene and are of possible use not only for research in the pig industry regarding meat quality and carcass composition, but also for human research. Considering the functional domains of the PPARGC1A protein, it is very likely these splice variants considerably affect the function of the protein and alternative splicing could be one of the mechanisms by which the diverse functions of PPARGC1A are regulated.