1.13.1. RGN: A Ca2+-Binding Protein
Regucalcin (RGN) was first identified as a cytosolic Ca2+-binding protein in rat liver (Yamaguchi and Yamamoto, 1978). It is unusual due to the absence of an EF hand motif in its structure, which is a structural motif that was once seen as an essential feature for a protein to able to bind Ca2+. Later on, it was demonstrated that this protein is capable of contributing to Ca2+ signalling, just like proteins such as CaM and calcineurin, despite its apparent major structural deficit. To date, it has been shown to be involved in calcium homeostasis in various organs, most notably the liver and kidneys (Yamaguchi, 2000a). This protein is also sometimes referred to as the
senescent marker protein 30 (SMP30) because its expression levels in a number of tissues appear to be age-dependent (Fujita, 1999).
1.13.2. The RGN Gene
RGN consists of 299 amino acids, has a molecular weight of approximately 33kDa, and the gene which encodes it has been located on the X-chromosome in rats (region Xq11.1-12) and humans (region Xp11.3-q11.2) (Fujita et al., 1995). The RGN gene has been found to exist in the genomes of a number of other mammalian organisms, including dogs and chickens, but has notably not been found in yeast (Shimokawa et al., 1995). Comparisons made between gene sequences from seven different
vertebrate species have shown the coding regions of the RGN gene have been highly conserved during evolution. At the protein level, the amino acid sequences
corresponding to these genes also demonstrate a high degree of conservation, with 69.9-91.3% similarity between all RGN proteins from the same seven different sources (Misawa and Yamaguchi, 2000a).
The RGN gene is made up of seven exons and six introns. Its 5’-flanking region is where transcription factors bind to alter RGN expression levels in response to changes in the concentration of certain factors, such as oral calcium administration (Murata and Yamaguchi, 1998). Important sequences within the 5’-flanking region for basal and regulated gene expression have been elucidated (Murata and
Yamaguchi, 1999). The promoter region of the RGN gene has been suggested to contain 28 transcription factor binding sites, and those that repress its expression (e.g. Sp1, C/EBP-β and SRY) are believed to bind within the region of -513 to -352
nucleotides from the transcriptional start site (Rath et al., 2008). In humans, it has been noted that the RGN gene can be transcribed to produce two transcripts of different lengths. Detailed analysis showed that only their 5’-untranslated regions differ, which may contribute to the synthesis of different transcripts at different times or stresses (Misawa and Yamaguchi, 2000b). Not all factors that affect RGN gene transcription influence Ca2+ homeostasis. A key example of this is the effect of
dexamethasone (a corticosteroid), which has been demonstrated to increase RGN mRNA levels significantly in rat kidney cortex but, at the same time, has no effect on the calcium content within this organ (Kurota and Yamaguchi, 1996).
1.13.3. Transcriptional Control of RGN Expression
Nuclear transcription factors that are most likely to influence RGN expression levels include NFI-A1 and AP-1 (Yamaguchi, 2005), as well as the regucalcin gene
promoter region-related protein (RGPR-p117) (Misawa and Yamaguchi, 2001). RGPR-p117 is expressed in a number of higher organisms (including humans, rats, dogs, cows and fish) and has a well conserved gene sequence (Misawa and
Yamaguchi, 2002). This transcription factor, when phosphorylated, is believed to bind the TTGGC sequence of the nuclear factor (NF1)-like motif in the regucalcin gene promoter sequence, and possibly promoter regions of other genes that also have this motif (Yamaguchi, 2009).
RGPR-p117 proteins in mammals consist of 1045-1060 amino acids and all have a conserved DNA-binding leucine zipper motif (Sawada and Yamaguchi, 2005). One RGPR-p117 splice variant of 625 amino acids has been found in humans and suggested to be involved in the formation of placental carcinomas (Yamaguchi, 2009). Post-translational modification of RGPR-p117 is likely to be essential for its function as a transcription factor because recombinant RGPR-p117 protein has been shown to be unable to bind the RGN gene promoter (Yamaguchi et al., 2003a) and analysis of its amino acid sequence has suggested it contains a number of sites for phosphorylation, N-glycosylation and amidation (Yamaguchi, 2009)
Unlike RGN, RGPR-p117 expression is unaffected by the aging process (Yamaguchi et al., 2003a), but it has been shown in rats to be sensitive to Ca2+ administration when expressed in the liver (Misawa and Yamaguchi, 2002). Over- expression studies in the rat kidney epithelial cell-line, NRK52E, have suggested RGPR-p117 is involved in the regulation of RGN expression when the latter is required for inhibiting DNA and protein synthesis (Tomono et al., 2007), and also in preventing apoptotic cell death by altering the mRNA expression levels of caspase proteins (Yamaguchi et al., 2007). Related experiments have shown RGPR-p117 over-expression alone cannot significantly alter RGN mRNA levels and that hormonal control is also required (Yamaguchi, 2009).
1.13.4. Hormonal Regulation of RGN Expression
Insulin treatment of food-deprived rats has been demonstrated to increase RGN mRNA levels in rat liver (Yamaguchi et al., 1995). Furthermore, it has also been shown that RGN over-expression can cause insulin resistance in the rat hepatoma cell
line, H4-II-E, by repressing the expression of proteins involved in insulin signalling (e.g. insulin receptor and phosphatidylinositol 3-kinase (PI3K)) (Nakashima and Yamaguchi, 2007). Parathyroid hormone (PTH) can also increase RGN mRNA expression levels, which has been demonstrated in the mouse osteoblastic cell line, MC3T3-E1 (Yamaguchi et al., 2008a). This effect of PTH on RGN could be related to its stimulatory effect on RGPR-p117 mRNA expression levels, which has been
observed in NRK52E cells (Yamaguchi, 2009). As mentioned above, RGN mRNA levels have also been observed to be affected by dexamethasone (a glucocorticoid) in the kidney cortex of rats (Kurota and Yamaguchi, 1996).
Studies on different rat tissues have suggested 17-β-estradiol (an estrogen) can either increase (e.g. liver) (Yamaguchi and Oishi, 1995) or decrease (e.g. mammary gland and prostate) RGN mRNA levels (Maia et al., 2008). Thyroid hormone (T3) can
have varied effects that is time-dependent, whereby treatment of rats with T3 to
induce hyperthyroidism can initially result in higher than normal RGN mRNA levels in liver but this effect decreases with prolonged (5.5 days) treatment before eventually leading to the opposite effect (Sar et al., 2007).
1.13.5. RGN Distribution in Mammalian Cells
RGN mRNA is not expressed widely among all tissues in organisms that have the gene, though the findings that can be found in the literature so far often conflict with each other as to which tissues do express RGN. In rat, RGN protein has been detected in samples of liver and kidney at abundant levels, both with the use of ELISA and Western blotting (Yamaguchi et al., 1991, Yamaguchi and Isogai, 1993, Yamaguchi et al., 2002b). On the other hand, there has been a conflict between data from these two methods of analysis, as well as from immunohistochemistry experiments, for samples from spleen, testis, skeletal muscle, lung, duodenum, colon and heart of rat (Yamaguchi et al., 1991, Yamaguchi and Isogai, 1993, Yamaguchi et al., 2002b). RGN expression in rat brain has in some studies been shown to not exist at the protein level (Yamaguchi et al., 2002b) and in other studies has been detected successfully (Yamaguchi et al., 1991, Yamaguchi et al., 1999). Mice have been shown to endogenously express detectable levels of RGN mRNA in liver, kidney, lung and cerebrum (Mori et al., 2004). This same study on mice also showed mice express RGN mRNA in testis and lung, which conflicts with the findings from Western blotting and immunohistochemistry experiments stated above for rats. Sex-dependent
differences in RGN expression have also been observed in rat stomach tissue for endogenous expression, as well as in a number of other tissues when transgenically expressed (Yamaguchi et al., 2002b).
RGN has been suggested to play a role in bone resorption/loss from experiments with transgenic mice (Yamaguchi et al., 2002a, Uchiyama and Yamaguchi, 2004, Yamaguchi et al., 2004b, Yamaguchi et al., 2005). Lower than normal RGN protein levels have been suggested to be associated with X-linked muscular dystrophy from experiments with mice diaphragm, and the same study has also demonstrated RGN expression in mice heart and limb muscle (Doran et al., 2006). However, the expression of RGN in rat heart is questionable due to conflicting outcomes of attempts to detect it, whereby some have shown it cannot be detected (Yamaguchi et al., 1991, Yamaguchi and Isogai, 1993, Yamaguchi et al., 2002b) and some have shown it can be detected (Yamaguchi and Nakajima, 2002). Discrepancy in RGN expression in rat heart may or may not be due to differences in age of the animals used for study (Akhter et al., 2007, Lim et al., 2009). Mammary glands and prostates in rats have been identified as other tissues where RGN is expressed (Maia et al., 2008, Maia et al., 2009).
1.13.6. The Protein Structure of RGN
RGN from mouse has been shown to share 60-66% amino acid sequence similarity with yeast (residues 136-186) and bacterial (residues 142-192) RNA polymerases, but it shares little homology with mammalian RNA polymerase (Ishigami et al., 2003). Circular dichroism experiments (Yamaguchi, 1988) have shown rat liver RGN has an α-helical content of 34% in the absence of Ca2+, but this percentage decreases when Ca2+ is present in its surrounding medium and this suggests RGN binds Ca2+ by loosening its structure. The same study also calculated that one RGN protein is capable of binding up to seven Ca2+, however, a later study disputed this finding by suggesting RGN has no Ca2+-binding activity (Kondo et al., 2004). The isoelectric point of rat RGN is 5.20 (Yamaguchi, 1988, Yamaguchi, 2005). Two attempts have been made to obtain a crystal structure of human RGN, with the most recent one successfully demonstrating the ability of this protein to bind one Ca2+ (Warizaya et
al., 2004) (Chakraborti and Bahnson, 2010). Figure 1.13.6.1 shows early predictions of secondary structure features within the rat RGN amino acid sequence, alongside the completed crystal structure of human RGN published in 2010. The main contrast
Figure 1.13.6.1. The Predicted & Solved Structures of RGN Protein
(A) Predicted secondary structure features of the rat RGN amino acid sequence, which contains 299 amino acids and the numbers indicate the positions of these residues from the N-terminus (left) to the C-terminus (right) (Yamaguchi, 2005). (B) The solved crystal structure of human RGN, shown to bind to a single Ca2+ in
the centre of its six-β-propeller barrel fold (Chakraborti and Bahnson, 2010)
A
between these two structures, along with data from circular dichroism experiments, is the crystal structure shows far less α-helical content in the RGN protein structure.