G Protein β3 Gene
Structure, Promoter, and Additional Polymorphisms
Abstract—Recent studies have shown that a polymorphism (C825T) in the gene encoding the G protein β3 subunit (GNB3) is associated with hypertension and obesity. We characterized the entire GNB3 gene, which spans 7.5 kb and is composed of 11 exons and 10 introns. Its promoter lacks a TATA box but harbors GC-rich regions. The functional activity of the GNB3 promoter was verified with reporter gene assays that also demonstrated its inducibility by phorbol esters. A novel polymorphism in the promoter region A(−350)G occurred with frequencies (G allele) of 76%, 97%, and 61% in Africans, Chinese, and Germans, respectively. Reporter gene constructs with either the A or the G allele did not differ with regard to inducement of the reporter protein. A silent nucleotide exchange in the coding region (A657T) occurred with T allele frequencies ranging from 0.5% to 2.4%. Another polymorphism (G814A) results in the replacement of glycine by serine at position 272. In Germans, the A allele occurred at a frequency of 10%. Finally, a C1429T polymorphism in the 3′ untranslated region of GNB3 was identified that occurred at T allele frequencies of 38%, 17%, and 30% in Africans, Chinese, and Germans, respectively. Haplotype prediction indicated in Germans an almost complete association of GNB3 825T with 1429T, and vice versa. An analysis of these polymorphic loci in nonhuman primates revealed that the ancestral GNB3 gene harbored the (−350)G, 825C, and 1429C alleles. This is the first complete characterization of the human GNB3 gene and its promoter region, which will enable refined epidemiological and biochemical investigations of GNB3 in hypertension and obesity.
We recently described a C825T polymorphism in the gene GNB31 that encodes the Gβ3 subunit of heterotrimeric G proteins, which are key components of intracellular signal transduction that are present in all cells of the body (for reviews, see Neer2 and Hamm3 ). Independent studies have shown that the GNB3 825T allele (GNB3 825T) is associated with essential hypertension in whites1 4 5 6 and blacks.7 Furthermore, GNB3 825T is associated with obesity,8 9 10 which may also affect blood pressure variation.8 The association between obesity and GNB3 825T has been demonstrated, in both white and nonwhite populations.8 9 10 However, the effect of GNB3 825T on blood pressure variation is not yet clear in Asian and American Indian populations.11 12 Interestingly, the frequency of GNB3 825T is considerably higher in black Africans (fT ≈0.8) and Asians (fT ≈0.45)8 11 12 compared with whites (fT ≈0.3).8
GNB3 825T is associated with the occurrence of the splice variant Gβ3s, which, despite a deletion of 41 amino acids, is functionally active in reconstituted systems.1 Furthermore, there is a strong association among GNB3 825T, the occurrence of Gβ3s, and enhanced signal transduction via pertussis toxin–sensitive G proteins.1 13 This may cause enhanced vascular reactivity and increased proliferation of smooth muscle cells and cardiac myocytes, ultimately resulting in vascular and myocardial hypertrophy. In vivo studies confirmed this concept and demonstrated an enhanced vascular reactivity on the stimulation of coronary α2-adrenergic receptors in carriers of GNB3 825T.14 Likewise, neutrophils from carriers of GNB3 825T exhibit an increased chemotactic response.15 16
Gβ proteins belong to the superfamily of propeller proteins, and all Gβ proteins identified so far consist of 7 WD repeats (referring to the conserved amino acids aspartate and tryptophan) that form a regular torus-like structure.17 18 Gβ3s results from alternative splicing of GNB3 and lacks the equivalent of 1 entire WD domain (Figure 1A⇓).1 Because numerous other WD proteins with 4, 5, or 6 WD domains exist,19 a Gβ3s structure, as shown in Figure 1B⇓, has been proposed by analogy.1 20 21 In GNB3, the C825T polymorphism is located >1700 bp upstream of the alternative splice site, indicating that the affect of GNB3 825T on the splice process is a complex mechanism. However, such mechanisms have been proposed for other genes.22 23
Single nucleotide polymorphisms comparable to GNB3 C825T are found frequently in the human genome,24 and there are numerous examples of originally identified polymorphic markers that are in linkage disequilibrium with subsequently detected causal mutations. The cDNA of GNB3 was originally cloned from human retina,25 and the C825T polymorphism was detected on the study of cell lines from hypertensive subjects with inherited increased signal transduction.1 Ansari-Lari et al26 reported the genomic sequence of GNB3 without describing the promoter or exon/intron structure or reporting additional polymorphisms. In another report, the GNB3 gene structure was described only incompletely.27 Given the obvious involvement of the G protein β3 subunit in the pathogenesis of obesity/hypertension, an in-depth characterization of its entire gene structure and additional polymorphisms is a prerequisite for an understanding of the molecular mechanisms that contribute to these traits.
The data presented here are the first to describe the promoter region of GNB3, including its regulation by a variety of hormones. Furthermore, we describe the functional organization of GNB3 with regard to exon/intron boundaries. Finally, we report novel polymorphisms in GNB3, their frequencies in different ethnicities, linkage disequilibrium with GNB3 825T, and occurrence in nonhuman primates.
The German sample consisted of 1855 healthy white individuals of either gender (age 18 to 60 years) who were recruited at the local Department for Transfusion Medicine, University Hospital Essen, as detailed previously8 and who represent a cross-sectional sample of healthy Germans.8 The Chinese population sample (n=368) was collected at the Jinan Technical College and the Tongji Medical University at Wuhan as described previously.8 The African population sample (n=706) was recruited by the National Blood Transfusion Service Zimbabwe and the South African Blood Transfusion Service Johannesburg as detailed previously.8 All individuals were genotyped for GNB3 C825T. For the analysis of the novel GNB3 polymorphisms, samples from African, Chinese, and German populations were taken at random from these pools.
Genotyping Studies in Humans
DNA extraction from blood was performed with the QiaAMP blood kit (Qiagen) as described.1 PCRs were performed with 4 μL of DNA solution in a final volume of 50 μL with 2.5 U Taq polymerase and the standard PCR buffer (MBI Fermentas). Oligonucleotide primers were routinely used at 0.4 μmol/L each, and a 0.2 μmol/L concentration of each dNTP was added. A first denaturation step of 95°C for 5 minutes was followed by 35 cycles at 94°C for 1 minute, at 60°C for 45 seconds, and at 72°C for 1 minute. The reaction was completed by a final extension step at 72°C for 7 minutes. PCR products and the respective restriction fragments were size-fractionated on 2.5% agarose gels that contained ethidium bromide and were visualized with UV transillumination.
Oligonucleotide primers 5′-TGACCCACTTGCCACCCGTG C-3′ (sense) and 5′-GCAGCAGCCAGGGCTGGC-3′ (antisense), encompassing the genomic sequence from nucleotide (nt) 5348 to 5615, were used. The numbers of the cDNA sequence and the polymorphisms within the cDNA [eg, C825T] refer to the original numbering of the cDNA25 sequence beginning with the translation start codon ATG=1. Numbers used to describe the gene structure and the promoter polymorphism (A−350G) refer to the genomic sequence transcription start site +1. Thus, the ATG of the cDNA corresponds to nt 1077 of the genomic sequence. Likewise, position 825 of the cDNA corresponds to nt 5500 of the genomic sequence. GenBank accession numbers for these sequences are U47924, M86525, and U72506. The GNB3 start codon is located in this latter clone at nt 53298. The C825T polymorphism (genomic localization at nt 5500) was diagnosed by restriction of the PCR amplicon with BseDI (MBI Fermentas), resulting in 2 fragments of 115 and 152 bp for the C allele and in an unrestricted fragment of 267 bp for the T allele.
This polymorphism (genomic localization at nt 5489) was analyzed by digestion with PstI (MBI Fermentas) of the same PCR amplicon used for the analysis of GNB3 C825T. For the A allele, 2 fragments of 126 and 141 bp were observed, whereas the amplicon remained unrestricted for the G allele.
Oligonucleotide primers 5′-CCATTTTGGCAGTGCCTTGTG GG-3′ (sense) and 5′-ATGTGTTGTGGGGAGTGTCGGG-3′ (antisense), encompassing the genomic sequence from nt 3554 to 3829, were used. The A657T polymorphism (genomic localization at nt 3725) was diagnosed by restriction of the PCR amplicon with Eam1105I (MBI Fermentas), resulting in 2 fragments of 104 and 171 bp for the T allele and in an unrestricted fragment of 275 bp for the A allele.
Oligonucleotide primers 5′-CAGCCTCTCCCTTAATGAGC-3′ (sense) and 5′-ACTACTCTGCTCAGAACTCC-3′ (antisense), encompassing the genomic sequence from nt 6918 to 7527, were used. The C1429T polymorphism (genomic localization at nt 7087) was diagnosed by restriction of the PCR amplicon with BshNI (MBI Fermentas), resulting in 3 fragments of 135, 169, and 305 bp for the C allele and in 2 fragments of 135 and 474 bp for the T allele.
Oligonucleotide primers 5′-AGAGGATGGTGGGGTTGG GAGG-3′ (upstream) and 5′-GAGGCTGTGAAAGCAGGG-GTCAG-3′ (downstream), encompassing the promoter region from nt −441 to −79, were used. The A(−350)G polymorphism was diagnosed by restriction of the PCR amplicon with TaqI (MBI Fermentas), resulting in 2 fragments of 91 and 270 bp for the G allele and in an unrestricted fragment of 361 bp for the A allele.
Partial GNB3 Sequences of Nonhuman Primates
DNA of common chimpanzees (Pan troglodytes) and orangutans (Pongo pygmaeus) was prepared from Epstein-Barr virus–immortalized lymphoblasts (gift from Dr H. Grosse-Wilde, European Collection of Biomedical Research, Essen, Germany) with the column extraction technique (Qiagen). DNA from pygmy/Bonobo chimpanzees (Pan paniscus) and gorillas (Gorilla gorilla) was a kind gift of Dr W. Schempp (Freiburg, Germany).
To sequence the alternative splice site region and the C825T polymorphic region in these species, PCR fragments were amplified with the oligonucleotide primers 5′-ACTGT(A/G)T-TTGTGGGACACAC-3′ (sense) and 5′-CCGACTC(G/A)TGGCCAGTGA AA-3′ (antisense; amplicon nt 532 to 683 of GNB3 cDNA) and the oligonucleotide primers 5′-GAGGCCA-TCTGCACGGGCTC-3′ (sense) and 5′-ACGCTCAG(A/C)CTTCATGG AGTC-3′ (antisense; nt 715 to 912 of GNB3 cDNA). The sequences encompassing the A(−350)G and the C1429T polymorphisms, respectively, were determined in these species with PCR fragments that were generated with the oligonucleotides indicated for the genotyping studies in humans. PCR fragments were purified with the QIAquick PCR purification kit (Qiagen) and PCR sequenced.
PCR Analysis of GNB3 Structure
For identification of exon/intron boundaries, overlapping fragments of the GNB3 gene were PCR amplified using human genomic DNA of 2 white donors (1 individual 825TT, the other 825CC) as template, and oligonucleotides derived from the Gβ3 cDNA. PCR amplicons were purified with spin columns (Qiagen), subcloned into the pGEM-T vector (Promega), sequenced, and aligned.
Screening of a Human Genomic λPS Library
To characterize the 5′-flanking region of human GNB3, 4×106 phage clones of a genomic library in λPS (MoBiTec) were lifted in duplicate onto nylon filters (Amersham) and hybridized to an α-32P-labeled probe derived by PCR amplification of an intron 3-fragment. Hybridization was performed in 6× SSC, 5× Denhardt’s solution, 1% SDS, and 0.1 mg/mL denatured salmon sperm DNA at 65°C for 24 hours. The filters were washed twice at room temperature in 2× SSC/0.1% SDS for 30 minutes, once at 60°C in 0.1× SSC/0.1% SDS for 20 minutes, and exposed to Kodak X-AR film at −70°C with intensifying screens. Positive clones were further purified in 3 consecutive rounds, and the resulting phages were used to infect Escherichia coli strain BNN132 (Cre+) cells to generate plasmid DNA by translocation that was finally sequenced.
Characterization of the GNB3 Promoter
Generation of Reporter Gene Constructs
Four presumptive promoter fragments containing artificial KpnI and XhoI sites (underlined here) were generated with PCR. These fragments extended from a constant 3′-primer (AP; 5′-CTCGAGGACAGGTCTGCCCCTA TTGTGG-3′, ending at nt 21) and 5′-primers starting at nt −221 (P4, 5′-GGTACCCAC CTGCTCCCCTCATGCAAATGACC-3′), nt −304 (P3, 5′-GGTACCGCTGGCCTGGGA GGAGACAGG-3′), nt −382 (P2, 5′-GGTACCGGTCGTGA-AGATCTCTCAGCC-3′), or nt −440 (P1, 5′-GGTACCAGAGGATGGTGGGGTTG-GGAGG-3′). PCR amplification consisted of an initial denaturation at 94°C for 3 minutes, followed by 35 cycles of 1 minute at 94°C, 45 seconds at 62°C, and 45 seconds at 72°C. Amplicons were agarose gel-purified (QiaEx system, Qiagen), digested with KpnI and XhoI, ligated into the corresponding sites of the pSEAP vector (Clontech) upstream of the reporter gene secretory alkaline phosphatase (SEAP), and sequenced. The resulting plasmids were transformed into E coli JM109 cells, and vector DNA was prepared using Qiafilter plasmid Maxi kits (Qiagen).
Transient Transfection in COS-7 Cells
COS-7 cells were routinely maintained in Dulbecco’s modified Eagle’s medium supplemented with 10% FBS and 100 U/mL penicillin and 100 μg/mL streptomycin (medium constituents from Sigma Chemical Co) in a 5% CO2 atmosphere at 37°C. Approximately 2×105 COS-7 cells were subcloned onto 6-well plates, grown to 60% to 70% confluency, and transiently transfected by lipofection with 4 μg pSEAP-promoter plasmid, complexed with 4 μg DAC30 (Eurogentec), essentially as recommended by the manufacturer. After transfection, COS-7 cells were cultivated in 1.5 mL growth medium for 48 hours, detached from the dishes, and pelleted by centrifugation (12 000g, 20 seconds). Supernatants were further concentrated with Centricon spin columns (Amicon). Samples were heated to 65°C for 30 minutes and subsequently chilled on ice to inactivate residual constitutive alkaline phosphatase. Thereafter, thermostable SEAP reporter gene activity was determined with the SEAP reporter system 2 (Clontech) by mixing 150 μL of the supernatant samples with 350 μL of dilution buffer and 500 μL of assay buffer supplied from the test kit. Samples were thermoequilibrated to room temperature for 5 minutes, the chemiluminescent substrate CSPD was added to a final concentration of 1.25 mmol/L, and the reaction mixtures were incubated at 37°C for 10 minutes. The emerging chemiluminescence, which was stable for ≥60 minutes under these conditions, was determined in a PICA luminescence aggregometer (Chronolog). Each reporter gene experiment was performed with COS-7 cells that were transfected in parallel with a positive control (pSEAP vector containing the SV40 early promoter) and a baseline control (pSEAP vector harboring the respective Gβ3 promoter fragment in inverted orientation). To analyze for potential GNB3 promoter-activating properties of several hormones, COS-7 cells were serum-starved beginning 24 hours after transfection for an additional 24 hours in the presence of the respective compound. The supernatants were harvested and treated as described earlier. Serum-depleted COS-7 cells cultivated in the absence of any agonist served as controls. All experiments were performed in triplicate.
For the prediction of promoter sequences, the TSSW: Recognition of Human PolII Promoter Region and Start of Transcription program was used (Gene Finder Program, Baylor College of Medicine; http://kiwi.imgen.bcm.tmc.edu:8088). Consensus sequences for transcription factors were sought with the program MatInspector (www.gsf.de/BIODV/matinspector.html). Haplotype analysis was conducted with the public domain program EH by Jürg Ott (http://linkage.rockefeller.edu/ott/eh.htm).
Structure of GNB3
The first aim of the present study was to extend the analysis of the GNB3 structure from the previous characterization of the novel splice site1 to the entire gene. We generated, sequenced, and aligned overlapping GNB3 PCR fragments with human DNA from 2 white individuals (1 GNB3 825TT and the other GNB3 825CC) as templates and primers derived from the cDNA of Gβ3. Portions of the 5′-flanking sequence of GNB3 were further identified by screening a human λPS library. During this study, we became aware that a huge cosmid clone had already been sequenced and published, which contained, among several other genes, GNB3.26 In parallel, a third group had incompletely characterized the GNB3 gene.27 Figure 1C⇑ and Table 1⇓ summarize the results of these independent studies. Thus, GNB3 is a gene of 7.5 kb that consists of 11 exons and 10 introns. The sizes of the exons range from 39 to 601 bp, whereas the sizes of introns range from 78 to 1607 bp (Table 1⇓). All intron/exon splice junctions follow the GT-AG rule established for eukaryotic genes, with flanking 5′-splice donor and 3′-acceptor sequences closely related to established consensus sequences.28 Four of the splice junctions separate codons, whereas the remainder of junctions are located within codons. The entire exons 1 and 2, as well as the first 30 bp of exon 3, encode the 5′-UTR of the human Gβ3 transcript, whereas the complete 3′-UTR is encoded in exon 11. During this analysis of the GNB3 structure and its flanking region, we detected 2 single nucleotide polymorphisms, A(−350)G and T657A, which were further characterized.
Identification and Characterization of the GNB3 Promoter
To define the promoter region of GNB3, we used the promoter prediction program TSSW and analyzed the 5′-flanking region GNB3 locus from nt −1720 to 1077 flanking the transcription start site. One single putative promoter region was identified at nt −242 (LDF score 5.07). A series of reporter gene constructs were prepared in which various stretches of this putative promoter region were ligated upstream of the reporter gene SEAP in the promoterless pSEAP vector. On expression of the promoterless pSEAP vector or the pSEAP vectors that contained the inverted GNB3 promoter region in COS-7 cells, SEAP activity was almost absent. However, on expression of the chimeric pSEAP vectors that harbored the putative GNB3 promoter region, SEAP activity increased strongly (Figure 2A⇓). Promoter construct 2 (encompassing nt −381 to 22) most potently increased SEAP activity by 37±6-fold above basal levels (n=3 independent experiments; mean±SD; P=0.0005). Truncations of this construct to motifs encompassing nt −303 to 22 (construct 3) and nt −220 to 22 (construct 4) resulted in 16±2-fold (mean±SD; P=0.0002) and 7±3-fold (mean±SD; P=0.026) increases in SEAP activity, respectively (Figure 2A⇓). The longest promoter construct investigated encompassing nt −440 to 22 (construct 1) exhibited a 26±4-fold (mean±SD; P=0.0004) increase in SEAP activity and thus was less potent than construct 2. For comparison, the strong SV40 early promoter induced a ≈200-fold increase in SEAP activity.
Subsequently, we searched this GNB3 region for transcription factor consensus binding sites. As shown in Figure 3⇓, the GNB3 promoter region lacks TATA or CAAT elements, has a 61% G/C content, and contains 1 repeat of the consensus Sp1 binding site. A number of potential regulatory elements, including activator protein-1 (AP-1)–, nuclear factor (NF)-κB–, interleukin-6 responsive element (IL6RE)–, and TCF-binding sites, as well as 5 E-boxes, were found in the regulatory region of GNB3 (Figure 3⇓). The presence of these elements suggests that control of Gβ3 expression involves the interactions of various cis-acting elements and trans-acting factors.
Next, we analyzed the inducibility of the GNB3 promoter by reporter gene analysis in COS-7 cells. Treatment with saturating concentrations of progesterone, estradiol, testosterone, aldosterone, dexamethasone, cAMP, l-thyronine, l-thyroxine, and retinoic acid and exposure to UV light did not significantly affect SEAP activity. However, treatment with the phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA; 250 nmol/L) increased reporter gene activity from 37±6- to 50±3-fold (P=0.043) above control levels (Figure 2B⇑). This result is in accordance with the presence of consensus sites for the transcription factors AP-2 and NF-κB in the putative GNB3 promoter region.
Screening of the GNB3 Gene for Additional Sequence Variants
GNB3 825T is stringently associated with the expression of Gβ3s, although the alternative splice mechanism is not completely understood.1 As mentioned, 2 additional polymorphisms were detected during characterization of the GNB3 structure and the identification of the promoter region. In addition, we screened GNB3 for additional nucleotide variants by sequencing the entire Gβ3 cDNA and the promoter region from several white individuals with different GNB3 825 alleles. Furthermore, we conducted an extensive search in the GenBank database. These procedures led to the identification and characterization of 4 novel GNB3 polymorphisms.
GNB3 G(−350)A Polymorphism
We detected a sequence variance in the promoter region leading to a G→A substitution at position −350 (Figure 3⇑). In a German sample of 616 individuals, 37% carried the GG genotype, 48% carried the AG genotype, and 15% carried the AA genotype, respectively, which resulted in an allele frequency fG of 61% (Table 2⇓). The genotype distribution in the German population was in accordance with Hardy-Weinberg equilibrium. The functional significance of the nucleotide exchange is unclear at the present. In the G allele, the consensus binding motif for an E-box of the MyoD transcription factor is destroyed. We analyzed in a reporter gene assay whether this nucleotide exchange affects GNB3 promoter activity. Reporter gene constructs corresponding to construct 2 described earlier were generated for both alleles, sequenced, and expressed in COS-7 cells. Both constructs exhibited functional promoter activity, but in 3 independent experiments, we observed no significant differences in SEAP activity between the 2 allelic constructs (Figure 2C⇑).
In black Africans, fG amounted to 76%, and in Chinese individuals, almost all are homozygous carriers of the (−350)G allele (Table 2⇑). Interestingly, all nonhuman primates were homozygous for the G allele, which is, therefore, most likely the ancestral state.
Physically, the GNB3 polymorphisms A(−350)G and C825T are separated by 5850 bp. Haplotype analysis, however, revealed a significant linkage disequilibrium between both loci for Africans (allelic association χ2=8.4, 3 parameters, P=0.0384) and Germans (allelic association χ2=58.06, 3 parameters, P<0.00001). Because of the low frequency of the (−350)A allele in the Chinese population, a further haplotype analysis was omitted. For both the African and the German human samples, the haplotypes G–T and A–C are significantly favored at the expense of the haplotypes A–T and G–C, respectively (Table 3⇓).
GNB3 A657T Polymorphism
Next, we identified an A657T polymorphism in GNB3 with a frequency of the 657A allele of 1.4% in the German population (n=290; Table 2⇑). This A224T exchange is a silent polymorphism located in the open reading frame of Gβ3 that does not change the amino acid composition of the protein. Allele frequencies for the 657A allele were 0.5% and 2.4%, respectively, in black Africans and Chinese. A further haplotype analysis was not possible due to the low number of 657A allele carriers in these populations.
GNB3 G814A Polymorphism
This polymorphism leads to a single nucleotide exchange at position 814 in the cDNA of Gβ3, which causes the replacement of the amino acid glycine at position 272 by serine. Thus, a highly conserved amino acid motif (Ile–Ile–Cys–Gly–Ile–Thr–Ser–Val) is affected, which connects the outer strand of the “propeller” blade 5 (domain 5d, according to Clapham and Neer17 and Sondek et al18 ) with the inner strand of the “propeller” blade 6 (domain 6a) in Gβ3. An identical protein motif is found in Gβ1, Gβ2, Gβ3, and Gβ4.
The frequency of the A allele in the German population amounts to 10% (Table 2⇑). In black Africans and Chinese, GNB3 814A is rare, with allele frequencies of only 2.0% and 0.5%, respectively (Table 2⇑). Haplotype analysis in the German sample indicates a significant association of the 814A allele with the 825C allele (P=0.0041). The G allele is also conserved in nonhuman primates (Figure 4C⇓).
GNB3 C1429T Polymorphisms
Finally, we identified a C1429T polymorphism in the distant 3′-UTR of Gβ3 that occurs frequently in the German population. This nucleotide exchange is located 1587 bp downstream of the C825T polymorphism on GNB3. In an analysis of 362 German individuals, we identified 179 subjects who carried the 1429CC genotype, 146 subjects who carried the 1429TC genotype, and 37 individuals who carried the 1429TT genotype. This distribution of genotypes was in accordance with Hardy-Weinberg equilibrium and resulted in a 1429T allele frequency of 30% (Table 2⇑). In the African and the Chinese human samples, we determined 1429T allele frequencies of 38% and 17%, respectively (Table 2⇑). In all populations investigated, the distribution of genotypes was in accordance with a model of strong linkage disequilibrium between the polymorphisms GNB3 C825T and GNB3 C1429T (Table 4⇓). In the German population, there is an almost complete association of GNB3 825T with GNB3 1429T and of GNB3 825C with GNB3 1429C. The haplotypes GNB3 825C–1429T and GNB3 825T–1429C occur only with very low frequencies (Table 4⇓). A similar pattern is also observed in the black Africans and Chinese, in whom the haplotypes GNB3 825T–1429T and GNB3 825C–1429C occur more frequently than expected. The haplotype GNB3 825C–1429T is almost absent, and in contrast to the German population, there is a considerable portion of the haplotype GNB3 825T–1429C (Table 4⇓).
Finally, we sequenced this region of the GNB3 gene from several nonhuman primates. Interestingly, the GNB3 1429C allele occurred in all of these species, which is, therefore, most likely the ancestral variant (Figure 4⇑).
We characterized the GNB3 locus, identified its promoter region, and defined new polymorphisms in the gene. Thus, the present report is the first detailed characterization of a mammalian Gβ subunit gene with regard to the gene structure and its promoter. It provides new insights but raises several important questions.
Gene Structure of GNB3
The GNB3 gene is located on a gene-rich cluster of chromosome 12p,26 spanning 7.5 kb and composed of 10 introns and 11 exons of usual sizes. Its gene product, the G protein subunit Gβ3, is a propeller protein that consists of 7 regular WD domains (Figure 1B⇑). Gβ3 is the first Gβ subunit for which the gene structure has been elucidated. There is no direct correlation between WD domains and the exon structure. For example, exon 6 codes for only 1 β-strand of a WD domain, whereas exons 9 and 10 define 7 individual β-strands each. Alternative splicing, which results in Gβ3s, is confined to exon 9, the longest exon in the open reading frame (Figure 1C⇑).
Identification and Characterization of the GNB3 Promoter
The human Gβ3 promoter belongs to the group of promoters without a TATA box but harbors GC-rich regions. The GC-rich region contains 1 putative binding site for the transcription factor SP1 and 3 putative binding sites for the transcription factor AP-1 (Figure 3⇑). Furthermore, consensus sites of NF-κB, NF-1, CBP (CREB-binding protein), IL6RE, TCF (T-cell factor), and 5 E-boxes were found in a 0.5-kb region that exhibited maximal promoter activity on expression in COS-7 cells (Figure 3⇑).
Expression of the longest promoter fragment (construct 1) resulted in a 26±4-fold increase in the reporter gene activity (Figure 2A⇑). Shortening of this fragment (plasmid 2) increased promoter activity to 37±6-fold (Figure 2A⇑). This may indicate the presence of a silencing element or elements located in this region. However, extensive computer-aided searches revealed no consensus sequence for various known transcriptional suppressors within this region. The human GNB3 promoter activity was not significantly induced by treatment with aldosterone, dexamethasone, retinoic acid, testosterone, progesterone, estradiol, l-thyronine, or cAMP, but it was inducible by the phorbol ester TPA. Several potential transcription factor binding sites that could mediate this activation are located within this region; in particular, the NF-κB site may be responsible for upregulation of gene expression after TPA treatment. On deletion of this site, the effect of TPA vanished (construct 4; 7±3-fold in the absence versus 6±2-fold increase in SEAP activity in the presence of TPA; data not shown).
In addition, we detected a novel polymorphism in the promoter region of GNB3. The substitution of (−350)A by the nucleotide G disrupts the consensus binding sequence for the MyoD transcription factor, which raises the possibility that differences in the transcription levels of Gβ3 and Gβ3s may contribute to the hypertensive phenotype, especially because a significant linkage disequilibrium exists between the 825 locus and the promoter polymorphism (Table 3⇑). However, expression of reporter gene constructs with the different GNB3 (−350) alleles resulted in almost identical promoter activities (Figure 2C⇑). Furthermore, on Western blot analysis, we did not observe gross differences in Gβ3 expression levels between cells harboring the GNB3 825T or 825C allele.1 Nevertheless, future studies that involve the use of more precise methods, including RNase protection assays, will again address the question of whether Gβ3 expression levels are influenced by the respective genotypes.
Significance of Additional Polymorphisms in GNB3
There is strong evidence that the GNB3 825T allele favors the generation of the splice variant Gβ3s, which in turn causes increased signal transduction and, ultimately, hypertension and obesity.1 However, the molecular mechanisms involved have not been elucidated so far. Therefore, we asked whether additional nucleotide exchanges in linkage disequilibrium with GNB3 825T may be necessary to cause the alternative splicing of the gene. As shown in Figure 1⇑, there is a distance of ≈1700 bp between the alternative splice site and the GNB3 C825T polymorphism. Nevertheless, there are examples that single distant nucleotide exchanges, not related to conserved splice branch, donor, and acceptor sites, can cause such alternative splicing.21 22
In addition to the promoter polymorphism A(−350)G discussed earlier, we characterized 3 additional polymorphisms in GNB3. Two of them, GNB3 A657T and GNB3 G814A, occur at low frequencies in the 3 populations investigated so far (Table 2⇑). Interestingly, GNB3 G814A leads to an amino acid exchange at a highly conserved protein motif in Gβ proteins. At the present, we have no evidence of any biochemical significance of this amino acid substitution. Expression and functional analysis of this Gβ3 variant are required to understand whether it affects Gβ3 processing, its interaction with Gα and Gγ proteins, and, ultimately, the effect of such a Gβ3γ dimer on interaction with receptors and effectors. Because this allele occurs at a frequency of ≈10% in the white population and it is strongly associated with GNB3 825C, association studies with hypertension, obesity, or both are required in the future.
The GNB3 1429T allele occurs with high frequencies in whites, black Africans, and Asians (Table 2⇑). It is located in the 3′-UTR of the Gβ3 mRNA and in tight linkage disequilibrium to the GNB3 825T allele. Thus, in whites, there is such a strong linkage disequilibrium between these loci at positions 825 and 1429 that in praxi we observed only 2 haplotypes: GNB3 825C–1429C and GNB3 825T–1429T. This leads to the question of whether GNB3 1429T is necessary (together with GNB3 825T) or possibly sufficient (without GNB3 825T) for the generation of Gβ3s. It is possible that a concerted action of GNB3 825T and GNB3 1429T is required, to favor an hnRNA structure necessary for the splicing of the Gβ3s gene product. Because our preceding studies on the association of GNB3 825T with the expression of Gβ3s and essential hypertension included only German subjects,1 we cannot answer this question at the present. Quantification of Gβ3s transcripts and the analysis of Gβ3s protein levels in different ethnic groups will help to resolve the question of how these haplotypes may affect alternative splicing of GNB3.
In nonwhites, the frequency of the GNB3 1429T allele is significantly lower than that of GNB3 825T. In these populations, the haplotypes GNB3 825C–1429C, 825T–1429C, and 825T–1429T occur frequently, whereas the haplotype 825C–1429T is rare. Several independent studies have confirmed the association of GNB3 825T with essential hypertension in whites1 4 5 6 and blacks7 ; however, in Asian and American Indian populations, the association of GNB3 825T with hypertension was absent11 or complex.12 It is possible that a certain genetic or ethnic background is necessary for the expression of the hypertensive phenotype in carriers of GNB3 825T. Hence, detailed epidemiological studies in nonwhite populations for both GNB3 825T and GNB3 1429T and their potential interaction with hypertension are necessary to resolve this issue. Although this question is open for hypertension, the presence of GNB3 825T appears sufficient for an increased risk of overweight in all populations investigated so far.8 9 10
Finally, we characterized the GNB3 polymorphisms and the sequence motifs decisive for the splicing of Gβ3s in nonhuman primates (Figure 4⇑). In all species, the sequence motifs necessary for the alternative splicing leading to Gβ3s (ie, the alternative branch and acceptor sites) are conserved. However, at loci GNB3 825 and GNB3 1429, the C alleles are present, suggesting that the C is the ancestral state at these positions. Therefore, it is likely that the mutation from C to T at these positions occurred since the split from humans and chimpanzees 5 million years ago.
In conclusion, we have presented the organization of the GNB3 gene, and we characterized its promoter. This and the identification of additional polymorphisms will enable the future analysis of the processes that generate the splice variant Gβ3s. The combined analysis of the novel single nucleotide exchanges with respect to hypertension and obesity in different ethnic groups could help to provide an understanding of the scenario in which GNB3 825T in concert with the other variants influences the observed variation in blood pressure and body weight. This could ultimately lead to an enhanced precision in the prediction of the risk for hypertension and obesity in individuals with these nucleotide exchanges in GNB3.
This work was supported by grants from the Deutsche Forschungsgemeinschaft and the University Hospital Essen IFORES program. We would like to acknowledge Sabine Phillips, Birgit Real, and Gerlinde Siffert for their expert technical assistance.
- Received January 26, 2000.
- Revision received February 4, 2000.
- Accepted February 9, 2000.
Hamm HE. The many faces of G protein signaling. J Biol Chem. 1998;273:669–672.
Schunkert H, Hense H-W, Döring A, Riegger GA, Siffert W. Association between a polymorphism in the G protein β3-subunit gene and lower renin and elevated diastolic blood pressure levels. Hypertension. 1998;32:510–513.
Benjafield AV, Jeyasingam CL, Nyholt DR, Griffiths LR, Morris BJ. G-protein β3 subunit gene (GNB3) variant in causation of essential hypertension. Hypertension. 1998;32:1094–1097.
Beige J, Hohenbleicher H, Distler A, Sharma AY. G-protein β3 subunit C825T variant and ambulatory blood pressure in essential hypertension. Hypertension. 1999;33:1049–1051.
Dong Y, Zhu H, Sagnella GA, Carter ND, Cook DG, Cappuccio FP. Association between the C825T polymorphism of the G protein β3-subunit gene and hypertension in blacks. Hypertension. 1999;34:1193–1196.
Siffert W, Forster P, Jöckel K-H, Mvere DA, Brinkmann B, Naber C, Crookes R, Heyns AdP, Epplen JT, Fridey J, Freedman BI, Müller N, Stolke D, Sharma AM, al Moutaery K, Grosse-Wilde H, Buerbaum B, Ehrlich T, Ahmad HR, Horsthemke B, du Toit ED, Tilikainen A, Ge J, Wang Y, Yang D, Hüsing D, Rosskopf D. Worldwide ethnic distribution of the G protein β3 subunit 825T allele and its association with obesity in Caucasians, Chinese, and black African individuals. J Am Soc Nephrol. 1999;10:1921–1930.
Hegele RA, Anderson C, Young TK, Connelly PW. G-protein β3 subunit gene splice variant and body fat distribution in Nunavut Inuit. Genome Res. 1999;9:972–977.
Kato N, Sugiyama T, Morito H, Kurihara H, Yamori Y, Yazaki Y. G protein β3 subunit variant and essential hypertension in Japanese. Hypertension. 1998;32:935–938.
Hegele, RA, Harris SB, Hanley AJG, Cao H, Zinman B. G protein β3 subunit gene variant and blood pressure variation in Canadian Oji-Cree. Hypertension. 1998;32:688–692.
Siffert W, Rosskopf D, Moritz A, Wieland T, Kaldenberg-Stasch S, Kettler N, Hartung K, Beckmann S, Jakobs KH. Enhanced G protein activation in immortalized lymphoblasts from patients with essential hypertension. J Clin Invest. 1995;96:759–766.
Baumgart D, Naber C, Haude M, Oldenburg O, Erbel R, Heusch G, Siffert W. G protein β3 subunit 825T allele and enhanced coronary vasoconstriction on α2-adrenoceptor activation. Circ Res. 1999;85:965–969.
Virchow S, Ansorge N, Rosskopf D, Rübben H, Siffert W. The G protein β3 subunit splice variant Gβ3-s causes enhanced chemotaxis of human neutrophils in response to interleukin-8. Naunyn-Schmiedeberg’s Arch Pharmacol. 1999;369:27–32.
Stallings-Mann ML, Ludwiczak RL, Klinger KW, Rottman F. Alternative splicing of exon 3 on the human growth hormone receptor is the result of an unusual genetic polymorphism. Proc Natl Acad Sci U S A. 1996;93:12394–12399.
Levine MA, Smallwood PM, Moen PT Jr, Helman LJ, Ahn TG. Molecular cloning of β3 subunit, a third form of the G protein β-subunit polypeptide. Proc Natl Acad Sci U S A. 1990;80:2329–2333.
Ansari-Lari MA, Muzny DM, Lu J, Lu F, Lilley CE, Spanos S, Malley T, Gibbs RA. A gene-rich cluster between CD4 and triosephosphate isomerase genes at human chromosome 12p13. Genome Res. 1996;6:314–326.
Lewin B. Nuclear splicing. In: Lewin B. Genes. New York. NY: Oxford University Press; 1997:885–913.