| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Laboratory for Molecular Pharmacology, Department of Neuroscience and Pharmacology (K.L.E., B.H., P.S.P., T.W.S.) and Department of Biomedical Sciences (J.B.H.), University of Copenhagen, DK-2200 Copenhagen N, Denmark; Department of Neuroscience (J.M., T.H.), Karolinska Institutet, 171 77 Stockholm, Sweden; and 7TM Pharma A/S (T.W.S.), DK-2790 Hoersholm, Denmark
Address all correspondence and requests for reprints to: Thue W. Schwartz, Laboratory for Molecular Pharmacology, Department of Neuroscience and Pharmacology, The Panum Institute, University of Copenhagen, Blegdamsvej 3, DK-2200 Copenhagen, Denmark. E-mail: schwartz{at}molpharm.dk.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
and 4 (HNF-1 and -4) and specificity protein 1 (SP1) transcription factors as being important for the expression of GPR39. In vivo experiments in rats demonstrated that GPR39 is up-regulated in adipose tissue during fasting and in response to streptozotocin treatment, although its expression is kept constant in the liver from the same animals. GPR39-1a was expressed in white but not brown adipose tissue and was down-regulated during adipocyte differentiation of fibroblasts. It is concluded that the transcriptional control mechanism, the tissue expression pattern, and in vivo response to physiological stimuli all indicate that the GPR39 receptor very likely is of importance for the function of a number of metabolic organs, including the liver, gastrointestinal tract, pancreas, and adipose tissue. |
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Neuromedin U and neurotensin are regulatory peptides that, like ghrelin and motilin, both have been implicated in, for example, control of food intake and GI tract function (5, 6).
|
Originally, GPR39 was, by Northern blot analysis, shown to be widely expressed throughout the whole body (1), which probably dampened the interest in trying to deorphanize this receptor. However, this expression analysis, which was performed with full-length GPR39 cDNA probes, could very well have been complicated or even invalidated by the occurrence of a gene called LYPD1 (Ly-6/PLAUR domain containing 1), which, in a large-scale systematic search for secreted proteins, was identified to be encoded at the antisense DNA strand corresponding to the genomic locus of GPR39 (1, 12). Moreover, several members of the ghrelin receptor family have been shown to be expressed in different splice variant forms, i.e. corresponding to both a full-length 7TM form and a truncated five-transmembrane (5TM) form (4, 13), which also could have complicated the original expression analysis. Variations in the control of expression in response to physiological stimuli are particularly interesting for constitutively active receptors, because, for example, an increased signaling activity will be directly correlated to an increased receptor expression, independent of the hormone or transmitter. This has been demonstrated recently for the constitutively active ghrelin receptor and the CB1 cannabinoid receptors, which both are closely regulated and fluctuate with physiological stimuli such as fasting (14, 15, 16). Importantly, this was not the case for nonconstitutively active, "normal" 7TM receptors, in which it is the expression of their ligands, and not the receptor, that is regulated, i.e. cholecystokinin, neuropeptide Y, and orexin (16).
In the present study, we characterize the human GPR39 gene, its promoter, and its expression pattern in peripheral tissues and in the central nervous system (CNS) of rats and have studied changes in the expression of GPR39 in vivo in response to starvation and streptozotocin (STZ) treatment in liver and adipose tissue, as well as during white and brown adipocyte differentiation of mouse fibroblasts.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
). Interestingly, an intron is found at the same position also in the genes for the ghrelin receptor and the motilin receptor, which in both cases give rise to truncated, 5TM splice variant forms of these receptors, i.e. GHSR-1b and motilin receptor 1b (4, 13). Within the intron of the GPR39 gene, we identified a potential alternative polyadenylation site, indicating that GPR39, like GHSR and the motilin receptor, could possibly be expressed in two splice variant forms. On the antisense strand corresponding to the 3'exon encoding TM segments VI and VII of GPR39 an antisense gene, LYPD1, is expressed in concordance with a recent systematic search for secreted proteins (12).
|
). The nucleotide sequence of the previously unrecognized GPR39-1b has been submitted to Entrez GenBank under the accession nos. BK005609 (Homo sapiens), BK005610 (Rattus norvegicus), BK005608 (Mus musculus), and BK005612 (Gallus gallus).
When the GPR39 was originally cloned, the genomic structure shown in Fig. 2A was not yet available, and the interpretation of Northern blot analysis was that the orphan receptor was expressed throughout the body, including multiple areas of the brain (1). However, as determined by specific quantitative RT-PCR (QPCR), the full-length GPR39-1a is in fact not expressed to any significant degree in the CNS (Fig. 2B). Instead, GPR39-1a is highly expressed in the liver, the GI tract from the stomach to the colon, the pancreas, the kidney, as well as in adipose tissue, and to a lower degree in the spleen, thyroid, lung, and heart. The truncated GPR39-1b form was found to be expressed more widely but again especially in the stomach and the small intestine (Fig. 2B). In contrast, the antisense gene LYPD1 showed a very different expression pattern because it was found highly expressed in all brain regions tested, with the highest levels observed in amygdala and septum. Although LYPD1 evidently was expressed to a much higher degree in the CNS than anywhere else (Fig. 2B), the expression in peripheral tissues was large enough to be readily detected through QPCR, with the heart showing the highest expression outside the CNS (Fig. 2B). As shown in Fig. 3A, in which the QPCR analysis in selected tissues was performed with the three gene products against each other instead of individually against ß-actin, the expression of GPR39-1a is approximately 100-fold higher than the expression of LYPD1 in, for example, the liver and GI tract, whereas the expression of LYPD1 is between 100- and 1000-fold higher than the expression of GPR39-1a in various brain regions, thus revealing a ratio shift between the two genes of approximately 1 million in different tissues (Figs. 2B and 3A). It should be noted that the expression of GPR39-1a in some brain tissues is so low that it approaches the detection limit for the method. A similar analysis showed that the expression of the truncated GPR39-1b is approximately 10-fold lower than the expression of the full-length GPR39-1a in the liver and kidney but relatively similar in the stomach and small intestine, in which both gene products are high expressed. In contrast, the expression of GPR39-1b is approximately 10-fold higher than that of the 1a form in, for example, the hippocampus and amygdala (Fig. 3B).
|
). Strong labeling was found in the lateral septum, islets of Calleja, basolateral, basomedial, and medial, intercalated amygdala nuclei, CA1 of the hippocampus, pyramidal cells in layer IV and neurons in layer VI of the cortex, as well as the paraventricular nucleus of the thalamus. Other areas, including the caudate putamen, globus pallidus, bed nucleus of the stria terminalis, the arcuate nucleus, and several other hypothalamic nuclei, revealed a clear but moderate labeling for LYPD1. LYPD1 seems to be expressed in a subpopulation of neurons, although expression in non-neuronal cells can not be excluded. In the investigated brain areas, no specific labeling for GPR39-1a could be observed (Fig. 4).
|
In the U-138 cells, maximal promoter activity was observed in constructs encompassing the region from –673 to +14, whereas in the Hep G2 cells the construct corresponding to the region from –573 to +14 showed maximal promoter activity, indicating that the minimal promoter varies between the two cell lines (Fig. 5). The results also indicated that the region –673 to –573 holds negative regulatory elements that are used exclusively in Hep G2 cells and not in the U-138 cells, in which no negative regulations were observed because the transcriptional activity just continued at a plateau when the constructs exceed position –673 (Fig. 5).
|
, 5' rapid amplification of cDNA ends (RACE) PCR surprisingly identified two alternative transcriptional start sites located closer to the translational initiation site than expected, i.e. one at position –196, which was used in both Hep G2 and U-138 cells, and one at position –396, which was used exclusively in the Hep G2 cell line. The transcriptional start site proposed to be located in position –469 as part of the automated annotation performed on 21,243 genes (17) was not observed to be used in any of the cell lines.
|
), of which especially the sites for hepatocyte nuclear factor 1 (HNF-1) and HNF-4 appeared interesting in view of the QPCR-based tissue expression pattern for GPR39-1as presented above (Fig. 2). The HNF-1, HNF-4, and SP1 binding sites were characterized experimentally through mutagenesis in which nucleotides were mutated to disrupt the binding sites for the relevant transcription factor using the matrices from the TRANSFAC6 public database.
|
site around position –468 impaired the transcriptional activity of the GPR39 promoter significantly in both U-138 and Hep G2 cells, mutation of the HNF-4 site around position –1253 only impaired the promoter activity in the U-138 cells and not in the Hep G2 cells (Fig. 7B). Interestingly, in the gene for the HNF-4 transcription factor, a conserved aspartic acid in the DNA binding region of HNF-4 is mutated into an alanine in the Hep G2 cell line. The consequence of this mutation is somewhat unclear, but it has been suggested to increase the transactivation potential of HNF-4 (20). Nevertheless, this altered HNF-4 protein to some degree disqualifies the Hep G2 cell line as a tool to investigate the actions of HNF-4.
Concerning the three SP1 sites, which were studied by mutational analysis, the one located around position –598 was clearly essential for the GPR39 promoter activity in both the U-138 and the Hep G2 cells (Fig. 7B). Promoter function was also significantly impaired by mutations of the SP1 sites located especially at position –514 but also at –522 in the U-138 cells. However, smaller or no effect was observed in the Hep G2 cell line with these mutations. Such differences between two cell lines is to be expected because of variance in the transcriptional factor milieu, which also is emphasized by the differences in transcriptional start sites (Fig. 6).
It is concluded that SP1, HNF-1, and HNF-4 are important transcription factors involved in the regulation of the expression of the GPR39 receptor.
In Vivo Regulation of GPR39 Expression
In an attempt to obtain preliminary data concerning the in vivo regulation of GPR39 expression, we performed QPCR analysis of its expression in liver and fat tissue in normally fed vs. 48-h-fasted rats, as well as in rats that had been made diabetic by exposure to STZ for 5 d.
In the liver, fasting led to an expected down-regulation of the expression of the glucose transporter GLUT2 but did not affect the expression of GPR39-1a (Fig. 8A). Similarly, no effect was observed in the expression of GPR39 in the STZ-treated animals despite the fact that an expected, strong up-regulation of the control gene peroxisome proliferator-activated receptor 
coactivator 1 (PGC-1) was observed in these animals (Fig. 8A).
|
). Very similar results were obtained in the STZ-treated animals, i.e. a 4- to 5-fold decrease in the expression of GLUT4 and an approximately 2.5-fold increase in the expression of GPR39-1a (Fig. 8B). Thus, in the liver, the high expression of GPR39 is apparently kept surprisingly stable because it is not affected by major alteration in the metabolism of the animal. In contrast, in adipose tissue, a clear up-regulation of GPR39 expression is observed in response to fasting and in response to STZ-induced diabetes.
GPR39 Expression in White vs. Brown Adipose Tissue
Based on the interesting expression and regulation of GPR39 in adipose tissue, we further investigated GPR39-1a expression in white adipose tissue (WAT) and brown adipose tissue (BAT). As shown in Fig. 9A, expression of GPR39-1a in BAT was close to the detection level of the method, whereas GPR39 was expressed at relative high levels in WAT (Fig. 2B). To gain insight into the regulation of GPR39 expression during adipocyte differentiation, we measured levels of GPR39 mRNA in two mouse embryo fibroblast (MEF) cell lines capable of differentiating into white or brown adipocytes (21). Surprisingly, GPR39-1a was expressed already in the undifferentiated state (designated d 0), and this was observed in both the retinoblastoma protein positive (Rb+/+) and negative (Rb–/–) fibroblasts, with expression levels being approximately 2-fold higher in the former (Fig. 9B). The Rb+/+ cells are capable of differentiating into white adipocytes, whereas Rb–/– cells differentiate into brown adipocytes, as was confirmed by measuring expression levels of the general adipocyte marker gene fatty acid binding protein 4 (FABP4) as well as the brown adipocyte-specific gene uncoupling protein 1 (UCP1) (Fig. 9B). Interestingly, in both cell lines, GPR39-1a expression was completely down-regulated during the differentiation process (Fig. 9B, compare d 0 with d 8).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
, HNF-4, and SP1 are important for the expression of the GPR39 receptor.
GPR39 Is Found in Two Splice Variant Forms
A characteristic feature of several members of the ghrelin receptor family is the occurrence of an intron located at a conserved site in the gene corresponding to the position in which where TM VI enters the membrane (Fig. 1). In GPR39, this intron is particularly large, and, for awhile, it led to the misconception that GPR39 was a pseudogene. In mice, GPR39 is still wrongly annotated in the Entrez GenBank (accession no. 71111), and this error was only recently corrected in the ensemble database in which it is now shown correctly (accession no. ENSRNOG00000021586). Interestingly the antisense gene LYPD1 is very conserved between all of these species, and, in fact, before the annotation of GPR39 was corrected in mice, etc., we identified the second exon of GPR39 by locating LYPD1.
For both the ghrelin receptor and the motilin receptor, splice variants corresponding to 5TM truncated receptor proteins have been described (4, 13). Through PCR analysis, we also found that GPR39 is expressed as an alternative splice protein, GPR39-1b. Interestingly, the 5TM form of the ghrelin receptor has been reported to be able to suppress the function of the full-length GHSR-1a receptor (22). A similar phenomenon has been observed in a few other cases in which 5TM splice variants are expressed, as for example in the GnRH receptor (23, 24). This could very well be related to the proposed two-domain function of 7TM receptors, in which TMs I–V, connected through loops of relatively well conserved length, constitute an A-domain that is separated by a very variable and often very long intracellular loop 3 from a B-domain consisting of the closely associated TM VI and TM VII (Fig. 1) (25, 26, 27, 28). According to the so-called global toggle switch model, 7TM receptor activation is associated with a vertical see-saw movement occurring in the kinked TM VI and TM VII of the B-domain in relation to the relatively rigid A-domain (29).
In the case of the GPR39-1b, we find that this splice form is rather widely expressed but with clearly higher levels observed in the stomach and small intestine, in which the full-length GPR39-1a receptor is also highly expressed (Fig. 2B). The fact that the splice site is surprisingly well conserved within the ghrelin receptor family indicates that these "5TM" proteins could serve a function in the in vivo setting. If the truncated 1b receptor form does affect the signaling of the full-length receptors, this could be particularly interesting for constitutive active receptors such as GPR39 and the ghrelin receptor, because modulation of the expression of the 1b form would be a novel, ligand-independent way of fine-tuning the level of constitutive activity of the receptor. For GPR39, the stomach and small intestine would then be tissues in which GPR39-1b could be expected to serve a physiological purpose.
Selective Expression of the Full-Length GPR39-1a Receptor in Metabolic Tissues
In the Northern blot analysis performed by McKee et al. in association with the original cloning of GPR39 (1), a cDNA construct containing the whole-coding region for GPR39 was used as a probe. This lead to the general deception that GPR39 was widely expressed in both the CNS and in multiple peripheral tissues because of the fact that this probe picked up not only the real receptor GPR39-1a but also the 5TM splice variant GPR39-1b, as well as the antisense gene LYPD1. Through QPCR analysis using specific probes for each of these three gene products, we demonstrate in the present study that the full-length, functional GPR39-1a receptor is mainly expressed in peripheral, metabolic organs, whereas the antisense gene proved to be a CNS-specific gene. The truncated GPR39-1b appeared to be more widely expressed, as discussed above. This picture of expression as well as the sizes of the individual mRNAs fit very well with the original Northern blot analysis, i.e. when it is interpreted with our current knowledge that the probe used could anneal to three individual transcripts giving rise to three distinct bands on the Northern blot, because the blots displayed not one but three bands conceivably corresponding to GPR39-1a, GPR39-1b, and LYPD1, respectively (1).
Our QPCR analysis of the tissue distribution of GPR39-1a expression is in very good agreement with a recently published QPCR analysis from Moechars et al. (30), who, according to the probes they used, very likely also detected GPR39-1a selectively. In both studies, the highest expression of GPR39 was found throughout the GI tract, including the liver and the pancreas, as well as in the kidney and with tissues such as lung, heart, and spleen showing clear but lower expression. We also found a high expression of GPR39-1a in adipose tissue and a clear expression in the thyroid, which however were not among the tissues tested by Moechars et al.
Debate Concerning GPR39 Expression in the CNS
In relation to the proposal that GPR39 could be the receptor for obestatin, a peptide fragment from the ghrelin precursor (see below), Zhang et al. (8) presented QPCR data indicating that GPR39 was expressed in the pituitary and hypothalamus at levels similar to those observed in the stomach and liver. However, subsequent QPCR analysis performed by two other groups showed very low, if any, expression of GPR39 in these areas as well as in the CNS in general, which is in agreement with the QPCR analysis for GPR39-1a of the present study (Fig. 2B) (9, 11, 30). It is in fact likely that there are major methodological differences between the approach of Zhang et al. and that of the other groups because Zhang et al. not only detected high levels of GPR39 expression in the hypothalamus and pituitary but also a low expression of GPR39 in, for example, the kidney and the colon (8), in which all the other groups detect high levels (9, 11, 30) (Fig. 2B).
Through in situ hybridization, GPR39 expression has been detected in the amygdala, hippocampus, and auditory cortex but explicitly not in the hypothalamus (31). In GPR39 knockout mice, who express LacZ instead of the GPR39 receptor, a signal was also detected in the septum-amygdala system in addition to throughout the GI tract, including the endocrine pancreas, but again not in the hypothalamus (30). It should be noted that, in the same paper, the QPCR analysis detected very low expression of GPR39 in septum and other brain areas in wild-type mice (30). It cannot at present be excluded that real GPR39-1a is expressed to a low degree in certain areas of the CNS such as perhaps specific nuclei in the septum, although it was not detected by our selective in situ hybridization (Fig. 4). However, it is likely that at least some of the GPR39 expression reported in various brain areas by use of either in situ hybridization or LacZ expression could be attributable to expression of the truncated, GPR39-1b form of the receptor (Fig. 2B). At least, all reports (except one) agree that GPR39 is not widely expressed in the CNS and in particular not in the hypothalamus.
The Antisense Gene LYPD1 Is Highly Expressed in the CNS
As shown in Fig. 2, LYPD1 is encoded on the antisense strand corresponding to the exon encoding TM VI and TM VII and the C-terminal segment of GPR39. In fact, GPR39 and LYPD1 have been used as a model system in a study on the use of RNA interference to control the expression of antisense genes (18). In the present study, we find that LYDC1 and GPR39 have a reciprocal or inverse expression pattern, because LYPD1 is expressed highly selectively in the CNS and because the in situ hybridization analysis reveals a distinct cellular, presumably neuronal localization in distinct brain nuclei (Fig. 4). In contrast, GPR39, as discussed above, is expressed selectively in metabolic tissues in the periphery (Fig. 2B). In certain peripheral tissues, there appears to be some overlap in the expression pattern, for example, the heart and the lung (Fig. 2B). However, it remains to be demonstrated whether this apparent coexpression really is a true coexpression in the same cells or whether LYPD1 may be expressed, for example, in neurons in these tissues. For many, but not all, pairs of antisense genes, it has been found that expression of one gene excludes the expression of the other or at least that they are expressed at opposite levels in the cell (32, 33). This is the pattern that we observe for GPR39 and LYPD1 in a large number of tissues throughout the body (Fig. 2B).
LYPD1 appears to encode a highly interesting protein. An amino-terminal signal peptide indicates that the LYPD1 gene product is a secreted protein. The signal peptide is followed by a sequence that apparently encodes an Ly-6 domain or module characterized by a highly conserved pattern of Cys residues forming up to five disulfide bridges in the folded protein structure (34). Ly-6 domains or modules are found, for example, in the urokinase-type plasminogen activator receptor, in CD59, in C4.4A, in RGTR430, in Ly-6 neurotoxin-like 1c, and in SLURP1 (secreted Ly-6/PLAUR domain containing 1) (34). In LYPD1, the Ly-6 domain is followed by a potential proteolytic cleavage site (rich in Pros and with multiple basic residues), which again is followed by a potential glycosylphosphatidylinositol anchor site. Thus, it is likely that LYPD1 is expressed both as a glycosylphosphatidylinositol-modified, membrane-associated protein and as a secreted protein that could serve some signaling purpose.
Transcription Factors Controlling GPR39 Expression
The mutational analysis of the promoter constructs using the luciferase reporter assay clearly indicated that HNF-1, HNF-4, and SP1 are involved in the control of GPR39 expression (Fig. 7). This also fits well with the observed expression pattern for the receptor in the GI tract, including the liver and pancreas, etc. HNF-1 and HNF-4 are strongly linked to early-onset type 2 diabetes especially maturity-onset diabetes of the young (MODY), in which mutation in HNF-1 results in MODY3 whereas mutations in HNF-4 results in MODY1 (35, 36). Thus, among carriers of a mutation in HNF-1, approximately 63% will have diabetes by the age of 25 yr and 95% by the age of 55 yr.
In a systematic analysis performed by microarrays and chromatin immunoprecipitation on 13,000 genes, it was found that, in hepatocytes, 1.7% of the genes are targeted by HNF-1 whereas 12% are targeted by HNF-4 (37). GPR39 was not discussed in the paper, but the receptor is found in the table listing the 158 genes, which are characterized by potentially being targeted both by HNF-1 and by HNF-4 in a feedforward, multicomponent loop manner (37). Interestingly, in this group of genes, GPR39 and the somatostatin receptor SSTR-1 were the only 7TM receptors listed (37). In the present study, we provide functional data that substantiates that the expression of GPR39 is in fact regulated by both HNF-1 and HNF-4. This obviously also fits well with the expression pattern for GPR39 in the liver, pancreas, and GI tract as such (Fig. 2B) and point to a potential important role for GPR39 in metabolism.
What Is the Physiological Role of GPR39?
It is obviously far too early to answer this question. However, in the present study, we found that GPR39 is up-regulated in adipose tissue during fasting and in response to STZ-induced diabetes in rats (Fig. 8B), which support the notion that this receptor is a metabolically interesting target. Somewhat surprisingly, in the same animals, the expression of GPR39 in the liver was held very stable. Although these are early and preliminary studies, it could indicate that GPR39 has some other, nonmetabolic function at least in the liver. The observation that GPR39 is expressed in MEFs and that its expression changes during adipocyte differentiation suggest that the receptor could be expressed specifically on preadipocytes and possibly play a role in adipocyte differentiation. One speculative possibility would be that GPR39 could play a similar role in, for example, the liver and the endocrine pancreas and perhaps the GI tract being involved also in tissue regeneration and differentiation.
Recently, the two first reports were published concerning GPR39 knockout mice (9, 30). These mice breed normally and do not display any overt phenotypic characteristics. Moechars et al. (30) focused on the GI function and presented evidence that gastric emptying and GI tract passage as such is enhanced in the GPR39–/– mice conceivably mainly as a result of an almost 3-fold increase in gastric fluid secretion. The GPR39–/– mice had normal body weight and food intake until wk 20, whereas very old animals showed increased body weight. Most biochemical parameters were normal, but the GPR39–/– animals did show increased levels of plasma cholesterol (30). Tremblay and coworkers focused on the food intake and body weight issues, in which they found that the GPR39–/– animals behaved just like the wild-type littermates in all respects. Both groups reported normal plasma glucose and insulin levels for the GPR39–/– animals (9, 30). Although the GPR39–/– mice in many ways appear to be rather normal, besides especially the gastric secretion, it will be interesting to see how the GPR39–/– mice respond to various challenges of the function of the liver, pancreas, intestine, adipocytes, etc.
Current Status for GPR39
GPR39 is a constitutively active 7TM receptor that is activated and or modulated by zinc ions and potentially by another as yet unidentified ligand. The receptor is expressed throughout the GI tract, including the liver and pancreas as well as in adipose tissue, and its expression is controlled in a rather unique way, at least among 7TM receptors, by transcription factors such as HNF-1, HNF-4a, and SP1, which points to an interesting function of GPR39 in these metabolically important organs. The genomic arrangement of GPR39 together with the membrane-associated/secreted neuronally expressed LYPD1 protein as a sense-antisense gene pair points to interesting possibilities in both relation to function and connection with regulatory mechanisms such as genomic imprinting (32, 33).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
PCR
PCR for constructs analyzing the promoter region was performed directly on chromosomal DNA, which were extracted from human blood using the QIAamp DNA Blood Mini kit (Qiagen, Hilden, Germany). The PCR program was 94 C for 3 min, followed by 25 cycles at 94 C for 45 s, 53 C for 45 s, 72 C for 2 min. Primers used are listed in Table 1.
|
.
5'RACE PCR was performed using the SMART RACE cDNA Amplification kit (BD Biosciences, San Jose, CA) for investigation of the transcription start site. Total RNA was isolated from the cell lines Hep G2 and U-138, using the RNeasy Lipid Tissue Mini kit (Qiagen) including a treatment with deoxyribonuclease I. The PCR program was 94 C for 3 min, then a 20-cycle touchdown, 94 C for 45 s, 63 C for 45 s decreasing 0.5 C per cycle, 72 C for 2 min, followed by 25 cycles at 94 C for 45 s, 53 C for 45 s, 72 C for 2 min. This was used for both the initial PCR and the nested PCR. Primers used are listed in Table 1.
All of the above PCRs were run on Eppendorf Mastercycler machines using the Pfu DNA polymerase (Promega, Madison, WI) and further analyzed on a 1.2% agarose gel. Products were extracted using the NucleoSpin Extract kits (Macherey-Nagel, Düren, Germany), restricted by the enzymes Kpn1 and HindIII (New England Biolabs, Beverly, MA), and ligated into the pGL3-basic vector (Promega) using the TaKaRa (Tokyo, Japan) DNA Ligation kit. The constructs were transformed into the XL-1 Blue Competent Cells (Stratagene, La Jolla, CA) and plated on LB agar containing 100 µg/ml ampicillin. One clone was selected, and plasmid DNA was extracted using the Qiagen Plasmid Maxi kit. Sequencing was done by MWG Biotech (Ebersberg, Germany).
QPCR was performed using the Mx3000P (Stratagene) and the SYBR Premix Ex Taq (TaKaRa). Cycle threshold (Ct) values were obtained using Stratagene Mx3000P software, ß-actin was used as a reference gene, and a calibrator sample was included in each round of QPCR to normalize between runs. The data were normalized by setting the maximum expression value to 1, thus showing the relative expression of the gene. To show the direct fold change of the ratio of two genes (gene_A/gene_B), a direct 
Ct method was used, basically using one of the genes as reference gene giving the following ratio: gene_A/gene_B = 2–(Ctgene_A–Ctgene_B). This ratio was arbitrary set to 1 for the heart in which all three gene products are expressed to a reasonable degree and thereby clearly show the ratio change in the various other tissues. RNA from tissue was extracted using the RNeasy Lipid Tissue Mini kit (Qiagen), followed by cDNA synthesis using the ImProm-II Reverse Transcriptase (Promega). Primers and probes used are listed in Table 1, and the specificity of the primers was evaluated through both melting curve analyzes and sequencing.
In Situ Hybridization
A 200-bp cDNA fragment corresponding to a 200-bp region of the 3' untranslated region of GPR39-1a and likewise a 200-bp region of the 3' region of LYPD1 were amplified using the primers Fr39/LYPD1 and Rr39/LYPD1 listed in Table 1 and ligated into the pSTP18 vector (Roche, Nutley, NJ). The plasmid containing this cDNA fragment was linearized with appropriate restriction enzymes. [35S]Uridine-5'-triphosphate-labeled RNA antisense (GPR39-1a) and sense (LYPD1) probes were transcribed in vitro using the appropriate template and phage RNA polymerase, purified using Probe Quant G-50 microcolumns (GE Healthcare, Little Chalfont, UK), and checked in denaturing acrylamide gel.
Expression of GPR39-1a and LYPD1 mRNA in the rat brain was investigated using in situ hybridization technique according to Diaz Heijtz and Castellanos (39). Briefly, brains were rapidly dissected and frozen on dry ice. Coronal sections (14 µm) were prepared on a cryostat and stored at –80 C until used. The frozen tissue sections were fixed in cold 4% paraformaldehyde [in 0.1 M PB (pH 7.4)], deproteinated with 0.1 M HCl, incubated in 0.25% acetic anhydride in 0.1 M triethanolamine (pH 8.0), and dehydrated. Sections were air dried and prehybridized [50% formamide, 50 mM Tris-HCl (pH 7.6), 25 mM EDTA (pH 8.0), 20 mM NaCl, 0.25 mg/ml yeast tRNA, and 2.5x Denhardt’s solution] for 4 h at 55 C followed by overnight (14–16 h) hybridization in a humidified chamber at 55 C. For hybridization, labeled probe was diluted to a final concentration of 0.5 x 106 cpm/200 µl containing 50% deionized formamide (pH 5), 0.3 M NaCl, 20 mM Tris-HCl (pH 7.6), 5 mM EDTA (pH 8.0), 10 mM PBS, 0.2 mM dithiothreitol, 0.5 mg/ml yeast tRNA, 0.1 mg/ml poly-A-RNA, 10% dextran sulfate, and 1x Denhardt’s solution. After hybridization and rinsing, the sections were then treated with 1 µg ribonuclease A (Roche) in ribonuclease buffer [0.5 M NaCl, 10 mM Tris-HCl, and 5 mM EDTA (ph 8.0)] for 1 h at 37 C, dehydrated in ascending alcohol series, and air dried. Sections were placed against ®-Max film (GE Healthcare) and stored at room temperature for 2–3 d.
Cell Cultures and Promoter Assay
The glioblastoma cell line U-138 (American Type Culture Collection, Manassas, VA), which has been reported to express GPR39 (18), and the hepatocellular carcinoma cell line Hep G2 (kindly provided by Prof. Jens Høiriis Nielsen, University of Copenhagen, Copenhagen, Denmark), were both grown in GIBCO DMEM/F-12 with glutamine and HEPES (Invitrogen, Carlsbad, CA). The following were added to the media: nonessential amino acids (0.1 mM), fetal bovine serum (5%), and penicillin-streptomycin (50 U/ml). Lipofectamine Reagent (Invitrogen) was used for transfection, and the promoter activity was analyzed using the Firelite Dual Luminescence Reporter Gene Assay System (PerkinElmer, Wellesley, MA). This includes a measurement of Renilla luciferase to normalize for transfection efficiency, which was obtained by cotransfection with a pCDNA3.1(+) (Invitrogen) construct carrying Renilla luciferase. The luminescence was measured in a TopCount (PerkinElmer). Luminescences from all constructs were normalized to the empty vector pGL3-basic, thus giving a relative measurement for transcriptional activity. All assays were performed at least three times in five replicates.
Rb+/+ and Rb–/– MEFs were propagated and differentiated as described previously (21). Three dishes were harvested for each cell line at each time point and further investigated for gene expression by QPCR.
Sequence Analyzes and Software Programs
Vector NTI Advance 10 (Invitrogen) was used for general handling of sequence information. The TRANSFAC 6.0 public database was used for identifying consensus of transcription factor binding sites, using the program Match (BIOBASE). The consensus of transcription factor binding sites were also used for designing disrupting point mutations. The program ConSite (Center for Genomics and Bioinformatics, Karolinska Institute) was used to locate conserved binding sites between H. sapiens and M. musculus. PRISM version 4 (GraphPad Software, San Diego, CA) was used for data handling and generation of graphs. The data were normalized to show fold induction relative to one chosen sample, and a t test (Mann-Whitney) was used to investigate significant variances.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Disclosure Statement: The authors have nothing to disclose.
First Published Online May 8, 2007
Abbreviations: BAT, Brown adipose tissue; CNS, central nervous system; Ct, cycle threshold; FABP4, fatty acid binding protein 4; GHSR, GH secretagogue receptor; GI, gastrointestinal; GPR, G protein-coupled receptor; HNF, hepatocyte nuclear factor; MEF, mouse embryo fibroblast; MODY, maturity-onset diabetes of the young; PGC-1, peroxisome proliferator-activated receptor coactivator 1; QPCR, quantitative RT-PCR; RACE, rapid amplification of cDNA ends; Rb, retinoblastoma protein; SP1, specificity protein 1; STZ, streptozotocin; 5TM, five transmembrane; 7TM, seven transmembrane; UCP1, uncoupling protein 1; WAT, white adipose tissue.
Received for publication January 26, 2007. Accepted for publication April 30, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
gene impair the function of the transcription factor to a varying degree. Nucleic Acids Res 28:430–437This article has been cited by other articles:
 |
|
 |
|
  P. S. Petersen, D. P. D. Woldbye, A. N. Madsen, K. L. Egerod, C. Jin, M. Lang, M. Rasmussen, A. G. Beck-Sickinger, and B. Holst In Vivo Characterization of High Basal Signaling from the Ghrelin Receptor Endocrinology, November 1, 2009; 150(11): 4920 - 4930. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Holst, K. L. Egerod, C. Jin, P. S. Petersen, M. V. Ostergaard, J. Hald, A. M. E. Sprinkel, J. Storling, T. Mandrup-Poulsen, J. J. Holst, et al. G Protein-Coupled Receptor 39 Deficiency Is Associated with Pancreatic Islet Dysfunction Endocrinology, June 1, 2009; 150(6): 2577 - 2585. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Tremblay, A.-M. T. Richard, S. Will, J. Syed, N. Stedman, M. Perreault, and R. E. Gimeno Disruption of G Protein-Coupled Receptor 39 Impairs Insulin Secretion in Vivo Endocrinology, June 1, 2009; 150(6): 2586 - 2595. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Zhang, Y. Liu, X. Huang, X. Liu, B. Jiao, Z. Meng, P. Zhu, S. Li, H. Lin, and C. H K Cheng Two alternatively spliced GPR39 transcripts in seabream: molecular cloning, genomic organization, and regulation of gene expression by metabolic signals J. Endocrinol., December 1, 2008; 199(3): 457 - 470. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. V. Zhang, H. Jahr, C.-W. Luo, C. Klein, K. Van Kolen, L. Ver Donck, A. De, E. Baart, J. Li, D. Moechars, et al. Obestatin Induction of Early-Response Gene Expression in Gastrointestinal and Adipose Tissues and the Mediatory Role of G Protein-Coupled Receptor, GPR39 Mol. Endocrinol., June 1, 2008; 22(6): 1464 - 1475. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Dittmer, M. Sahin, A. Pantlen, A. Saxena, D. Toutzaris, A.-L. Pina, A. Geerts, S. Golz, and A. Methner The Constitutively Active Orphan G-protein-coupled Receptor GPR39 Protects from Cell Death by Increasing Secretion of Pigment Epithelium-derived Growth Factor J. Biol. Chem., March 14, 2008; 283(11): 7074 - 7081. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. K Samson, G. L C Yosten, J.-K. Chang, A. V Ferguson, and M. M White Obestatin inhibits vasopressin secretion: evidence for a physiological action in the control of fluid homeostasis J. Endocrinol., March 1, 2008; 196(3): 559 - 564. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
; HNF-4
; and SP1, 
. At least three independent assays were performed, and the luciferase activity was normalized to the construct pGL3-F7 encompassing the wild-type (WT) sequence containing no point mutations. PAX2, Paired box gene 2; C/EBPß, CCAAT/enhancer-binding protein ß.
