Published ahead of print on February 16, 2006, doi:10.1165/rcmb.2005-0408OC
© 2006 American Thoracic Society DOI: 10.1165/rcmb.2005-0408OC Alternative Promoter Use and Splice Variation in the Human Histamine H1 Receptor GeneDivision of Therapeutics and Molecular Medicine, University of Nottingham, Queen's Medical Centre, Nottingham, United Kingdom Correspondence and requests for reprints should be addressed to Caroline Swan, Division of Therapeutics and Molecular Medicine, D Floor, South Block, University of Nottingham, Queen's Medical Centre, Nottingham NG7 2UH, United Kingdom. E-mail: Caroline.Swan{at}nottingham.ac.uk
Upstream gene structure and mRNA expression of the human histamine H1 receptor gene was investigated in cells relevant to the pathogenesis of asthma, (primary cultured human airway smooth muscle (HASM) cells, primary cultured human bronchial epithelial cells and bronchial epithelial cell line [BEAS2B]), and other tissues known to express histamine H1 receptors (placenta and brain). Splice variation of the 5' terminal exons gave three separate locations for novel promoters upstream of the detected transcription start sites. Further splice variants in the 5' untranslated region were also observed. Transient transfections of promoter/luciferase constructs showed these regions directed expression in HASM cells and BEAS2B cells. Polymorphism screening of the major regulatory regions identified a number of novel single-nucleotide polymorphisms. Expression of splice variants was confirmed by real-time PCR assays. Results showed one 5' terminal exon splice variant, comprising exons B/K, expressed preferentially in all tissues. Interestingly, the other 5' terminal exon splice variants showed tissue-specific patterns of expression, with variant F/K expressed negligibly (0.1%) in HASM cells, but accounting for 19.3% and 8.3% of total expression in BEAS2B cells and differentiated human bronchial epithelial cells, respectively. Splice variant A/K was second most highly expressed in differentiated human bronchial epithelial cells (23%), whereas its expression in BEAS2B and HASM cells was 1.7% and 4.4%, respectively. These data suggest the use of alternative promoters directing human H1 receptor gene expression, both within and between cell types.
Key Words: asthma • histamine H1 receptor gene • messenger RNA • promoter • splice-variation
In this article, we describe studies designed to investigate the gene structure of the human histamine H1 receptor (HRH1) gene in tissues relevant to the pathogenesis of asthma. Histamine, released from mast cells in the airways and also derived from circulating and infiltrating basophils, mediates most of its effects on the airway via HRH1s. These effects include many facets of the asthma phenotype: bronchoconstriction and proliferation of airway smooth muscle, vasodilation, plasma extravasation, increased cholinerginic nerve transmission, neuropeptide release, and sensory nerve activation (1, 2). HRH1s may also show constitutive signaling activity, contributing to the maintenance of airway tone (3) and to cross-talk with other receptors, potentially extending their sphere of influence and interaction beyond their cognate downstream signaling pathways (4–11). The molecular structure and potential genetic variation of the gene is key to understanding receptor expression and regulation at the levels of transcription and translation. The sequence of the HRH1 gene coding region was published in 1993 (12), followed by several studies of the regions upstream of the ATG start codon (13–15). However, preliminary work demonstrated that these gene structures were highly unlikely to be relevant in airway cells. In addition, information we obtained from the DataBase of Transcriptional Start Sites (16) suggested further 5' exons might be present. In this article, we describe splice variation of novel 5' terminal exons of the HRH1 gene and three separate novel promoters. In addition, using a combination of reporter gene and transcript-specific quantitative RT-PCR approaches, we demonstrate tissue-specific expression of HRH1 splice variant transcripts.
Tissue Culture Human airway smooth muscle (HASM) cells were isolated and cultured as previously described (17). Full ethical approval was obtained from the local regional ethical committee. BEAS2B cells were a kind gift from Dr. R. A. Penn, Wake Forrest, North Carolina. Primary human bronchial epithelial cells (HBECs) were cultured and differentiated at an air–fluid interface, as described by Danahay and colleagues (18).
RNA/cDNA Preparation
Rapid Amplification of cDNA Ends For exon-specific RACE, primers GAAGCAGCTCACTCCT CAGTCTGTTA and GTTCTCATCCTCTTACCCTGGCAAGA were used. 5' RACE products were not gel-isolated to allow detection of the widest range of transcripts and to avoid skewing the products obtained by size selection on a purification column. Plasmid DNA obtained from RACE PCR clones was prepared using the Qiagen DNA miniprep kit and directly sequenced with vector primers M13F and M13R using a Prism 337 DNA sequencer or a model 3100 Genetic Analyser (Applied Biosystems, Foster City, CA). Sequence data were analyzed using the Basic Local Alignment Search Tool (BLAST) 2 sequence alignment program available online at: www.ncbi.nlm.nih.gov/BLAST.
Generation of Promoter/Luciferase Reporter Constructs For the promoter upstream of exon B, forward primers TTGTT GCTCGAGCTTAGGTGGTCCCTT and TTGTTGCTCGAGCTGC TTACCAGGGGCTTGAAAT were used with reverse primer GTT GTTGTTCAAGCGCGCTCTTCTCAA on a genomic DNA template derived from HASM cells to create 1 kb and 2 kb fragments, respectively. The forward primers had a 5' nonbinding extension of TGG TGG CTCGAG, which created the restriction site Xho1 in subsequent rounds of PCR amplification. The PCRx enhancer kit (Invitrogen) at 3x strength was found to be optimal for amplification of the promoter B PCR product. The part of the construct containing the 5' UTR was created by PCR with primers CGCCCAATACGCAAACCGCCT and TGGGGAGGCCCATGGGCGAAAGA on a RACE-derived clone using Biotaq (Bioline, London, UK). This clone contained the B 5' exon, transcription start site (TSS) and exon K to the HRH1 start ATG. The first PCR fragments (containing the promoter regions) were cut with Xho1 and an internal BstX1 site, and the second fragments with the same internal BstX1 site and a created Nco1 site. Fragments were ligated into pGL3Enhancer, which had been cut Xho1/Nco1. The completed construct therefore contained promoter sequence spliced to the 5' UTR region upstream of the luciferase coding region. Similarly, for promoter A (the region directly upstream of exon A), forward primers TTGTTGCTCGAGGGGAGAAAGAAGGAGA GG and TTGTTGCTCGAGTGCACCTCAGTGACCCT, and reverse primer GTTGTTCCAATCAGCCACCTCAG, were designed to amplify the same promoter lengths. Xho1 sites were included on the forward primers and the products cut with Xho1 and an internal Alw44I site. For the 1 kb and 2 kb products, the PCRx enhancer kit at 2x strength and the Expand High Fidelity PCR system (Roche, Lewes, UK) were used, respectively. The 5' UTR was amplified from another RACE clone containing the A exon, TSS, exon K, and HRH1 start site using the same primers and polymerase as before; the fragment was cut Alw44I/Nco1 and ligated as before. For promoter F constructs, forward primers TGGTGGCTCGA GAGTGCTGGGAAGTGCCAC and TGGTGGCTCGAGAATCC TTGCCCTGAAGACTG, and reverse primer GTTGTTATAAG CAAACAGGTCTACTCC, were used to generate 1 kb and 2 kb fragments from the region upstream of exon F, which were cut Xho1/Xma1. A RACE clone containing the F 5' exon, TSS, exon K, and HRH1 start site was used as a template, with the same primers and Taq polymerase as described before to amplify a fragment, which was cut Xma1/Nco1 and ligated as described previously here. Three constructs based on the previously published gene structure, and containing the previously published promoters, were made. Construct Hs contained upstream sequence, including exons I and H, of length 1,002 bp, plus exon K to the HRH1 start ATG. Construct H1k and construct H2k contained 1,002 bp or 2,160 bp of sequence 5' to exon H, respectively, and included exons H, I and exon K up to the HRH1 start ATG. A previously prepared construct (pH1ENLWT) was used as template for PCR to prepare constructs Hs and H1k with forward primers TGGTGGCTCGAGCTTTCTTCTGTTAGCAAA GT and TGGTGGCTCGAGCATCTGTAGTTCGGATTAAATC, and reverse primer TGGGGAGGCCCATGGGCGAAAGA. Products were cut Xho1/Xmn1. Construct H2k was prepared by PCR on HASM genomic DNA, with primers TGGTGGCTCGAGGATCTCCAGGCTG GTGTTG and CTGTAATTGAAGGTCTTCTCCA, and the PCR product was cut Xho1/Xmn1. The 5' UTR portion of the constructs was subcloned from a previously prepared construct (pH1EN#8) by cutting Xmn1/Nco1, to give a fragment containing the HRH1 start site plus preceding exons H and I. The fragments were then ligated into pGL3Enhancer, as previously described here. For short open reading frame (sORF) analysis, a construct containing 392 bp (1,639–1,994 and 7,780–7,817) of AJ000742 cloned upstream of the luciferase coding region in pGL3E was used. This region comprised exons H, I, and K to the HRH1 start ATG, with some 5' sequence as shown by DeBacker and colleagues (15). The Quickchange Site-directed mutagenesis kit (Stratagene) was used to change each upstream sORF ATG to CTG.
Transfection of Promoter/Reporter Constructs
Quantification of HRH1 Transcript Expression by TaqMan Real-Time PCR
The percentage primer efficiency (E) of the TaqMan primers was calculated from standard curves generated from assays performed on serial dilutions of plasmid DNA clones containing the target sequences (E = [10–1/slope–1] x 100, where the slope was derived from CT [cycle threshold] versus log dilution graphs). All primer/probe sets gave acceptable high efficiencies, allowing the comparative cycle threshold (
Statistical Analysis
Polymorphism Screening
Identification of HRH1 Splice Variants and Promoter Regions by 5' RACE Initial RACE experiments were designed to define HRH1 transcripts. Results showed three common splice variants: A/K, B/K, and F/K, with alternate 5' terminal exons. Exon sequences obtained by RACE were compared with AC083855, a GenBank clone containing the whole of the region, using BLAST2 sequence analysis to obtain intronic distances. Figure 1 shows schematic diagrams of genomic structure of the HRH1 gene (with the position of all exons detected shown) and splice variants detected by RACE. The most prevalent splice variant was B/K, accounting for 85% of clones sequenced in HASM, 36% in BEAS2B, and 95% in brain (Table 1). This splice variant consisted of 5' terminal exon B, with multiple TSSs being identified spread over a 51-base region, from 52,704 to 52754, spliced to exon K, the 3' terminal exon, which comprised 5' UTR, CDS, and 3' UTR, with a splice site at –35 bp relative to the HRH1 ATG start codon. Two further relatively common transcripts were observed: splice variant A/K comprised 5' terminal exon A spliced to exon K. This variant was seen in both BEAS2B and HASM, accounting for 5 and 9% of clones sequenced, respectively (Table 1). Splice variant F/K comprised 5' terminal exon F spliced to exon K. This variant was seen only in BEAS2B by standard RACE, accounting for 54% of clones obtained (Table 1). By using a reverse primer specific to the 3' end of exon F, clones containing this exon were also obtained from HASM, suggesting that some low-level expression of this transcript does occur in this cell type, a result which was later confirmed by TaqMan analysis (see below). 5' terminal exon D was seen only once, derived from brain cDNA. Additional splice variants for the HRH1 gene were seen: a total of 11 splice variants incorporating additional exons, but all with exon B as the most 5' exon, were detected—two from BEAS2B and nine from HASM (Table 1). Also, one splice variant with 5' terminal exon A and incorporating exon I was obtained from HASM. This variant has been assumed to splice to exon K, as it was detected using a primer specific to exon I, which bound 5' to the splice site at –35 bp. These less-common splice variants are also shown in Figure 1. Of note, exons H and I, which are congruent with upstream sequence reported by DeBacker and colleagues (15), and were previously believed to represent the most 5' exons present in HRH1 transcripts, were seen only with other exons splicing in a position 5' to them. Examination of the sequence 5' to the putative TSS previously shown (15) identified a consensus acceptor splice site. The upstream gene structure described by DeBacker and coworkers (15), shown for comparative purposes (Figure 1, dashed box), gave sequence corresponding to exons H and I as one exon and with the TSS at the start of exon H. Our data show the presence of an additional 364-bp intron between these two sections of sequence, but the same intronic size of 5.79 kb between exon I and exon K. The intronic sequence detected between exons H and I was submitted to GenBank as AF420434. All splice variants detected would encode identical peptides, as alternate splicing was limited to the 5' UTR, with all variant exons splicing to a common 3' terminal exon (K).
Data from 207 cDNA RACE clones analyzed by direct sequencing (Table 1) show that the B 5' terminal exon was present in more clones than any other variant, suggesting that the region upstream of this exon is the dominant promoter in HASM cells and brain. However, in BEAS2B cells, exon F was more common, contrasting with negligible expression of this exon in HASM or brain. Exon A was seen by RACE in a total of 15 clones, and was not detected in brain cDNA by this method.
The TSS positions observed demonstrated extensive minor variation for all variants. For example, the TSSs for exon B fell into three main clusters with exon lengths of 62, 27, and 24 bp, respectively: these together accounting for 82% of the clones seen. The TSSs for exons A and F spread over a region of 280 and 299 bp, respectively (Table 2). To ensure that this was not an artifact of the RACE methods used, in parallel experiments, RACE clones were also obtained for the human
It was noted that the 13 most 5' bases of the previously described HRH1 clone, AJ000742, were artifacts of library cloning. PCR with primers designed 5' and 3' to the published sequence, as shown by GenBank produced bands, and subsequent direct sequencing confirmed these artifactual bases were not present, with the GenBank sequence continuing into the published sequence. To eliminate the possibility that some of the clones observed originated from unprocessed transcripts, RACE was repeated on cytoplasmic RNA from one HASM donor: the same splice variants were detected. Sequences around each intron/exon boundary conformed extremely closely to consensus splice sequences (i.e., AG GTRAGT for the donor sequence and YnNYAG G for the acceptor sequence [21]), even for rarer transcripts, giving high confidence that these represent real gene structures. In keeping with this observation, when splice variation was seen, the splice positions were still conserved.
Characterization of the 3' UTR of the HRH1
Transient Transfection of Promoter/Reporter Constructs
sORFs in 5' UTR of HRH1 Gene When the published gene structure of the HRH1 gene was examined, it was noted that exons H and I together contained two sORFs. In transient transfection experiments performed with exons H and I cloned into the pGL3 enhancer vector upstream of a luciferase reporter gene, removal of both these sORFs by site-directed mutagenesis led to a 2.26 ± 0.07–fold increase when compared with the wild-type construct in BEAS2B cells (n = 8; P < 0.01 for all changes by one-way ANOVA with Bonferroni's correction). A similar trend was seen in parallel experiments performed using the same constructs in primary cultures of HASM: 2.18 ± 0.33–fold, although changes did not reach statistical significance (22). These data suggest that the presence of sORFs in exons H and I can exert a negative regulatory effect on downstream gene expression. When the 5' UTR sequences of the most commonly expressed transcripts (A/K, B/K, and F/K) were studied, it was evident that they too had varying numbers of sORFs before the main HRH1 coding region ATG start codon. The 5' UTR of B/K contains no upstream ATGs, whereas, in marked contrast, the A/K 5' UTR contained six and the F/K 5' UTR contained five putative sORFs. Further investigation of these sORFs will be necessary to establish if they have a regulatory function; however, it is interesting to note that use of the most 3' TSSs detected for transcripts A/K and F/K would bypass 3/6 and 2/5 of the sORFs, respectively, and that the most commonly expressed B/K transcript has a simple 5' UTR structure containing no sORFs.
Quantification of HRH1 Transcript Expression
Polymorphism Screening of the HRH1 Screening for polymorphic variation by direct sequencing in 20–32 individuals revealed multiple novel SNPs, as well as previously identified SNPs in the HRH1 gene coding region, intron, 5' UTR, and putative promoter regions (see Table 3). All positions are numbered relative to AC083855. Five novel SNPs were seen in promoter A, and we confirmed the presence of two SNPs already present in the National Center for Biotechnology Information SNP database. Promoter B showed only one novel SNP (a "C" insertion at 52662 and a "GC" insertion before 52704 were seen, but were present in all individuals screened), and promoter F showed three novel and two database SNPs. The portion of sequence spliced out in the cells studied here, but previously described as the HRH1 promoter, showed seven SNPs: three novel and four previously described in the database and elsewhere (23). Sequencing of the coding region of the gene confirmed the presence of two previously described SNPs (24). One SNP, previously described by Sasaki and colleagues (25) was seen in the 5' UTR of exon K at –17 bp relative to the ATG start codon.
Species Sequence Comparison Exon B is a highly conserved region that is present upstream of the mouse H1 receptor (H1R) gene with the same splice sites conserved (the mouse has an additional exon between this exon and the 3' terminal exon comprising positions 65–211 of NM_008285). This conserved region is also present in contig NW_047696.1 genomic DNA, which aligns to rat chromosome 4, a region containing the rat H1R gene homolog. The region is also highly conserved in porcine cDNA (BP154008.1), where it splices to an equivalent of the human exon K containing a porcine H1R coding region homolog with 88% amino acid identity to the human receptor over the 231 amino acids available for comparison. The B exon also has high homology to a region of canine genomic DNA (NW_139888) found on the same chromosome (20) as the canine H1R homolog, and is a total match for sequence from chimp genomic DNA on NW_104865, which aligns to the same chromosome (3) area as the chimp H1R gene (see Figure 4).
In this article, we have, for the first time, provided a comprehensive characterization of the structure of the human H1R gene and provided a description of the transcripts observed in a range of cell types. We demonstrate that tissue-specific expression of transcripts occurs in physiologically relevant cell types. This is likely to be driven by the selective usage of the three main HRH1 promoters identified in different cell types. In particular, airway epithelial cells (including the BEAS2B cell line, and both differentiated and undifferentiated HBEC cultures) have increased expression of transcript variant F/K, which is expressed at barely detectable levels in airway smooth muscle, and HBECs also have high expression of variant A/K. Evidence for this differential expression profile of HRH1 transcripts comes both from extensive RACE experiments and from quantitative PCR experiments. In the latter work, we were also able to estimate the probable contribution of the identified transcripts to the total coding region transcripts by comparing the abundance of the different 5' UTR transcripts with total HRH1 coding region mRNA: these experiments suggest that in each cell type studied, we were able to account for over 65% of all transcripts present, indicating that, although some rarer transcripts may not have been identified, the majority of transcript species contributing to expression of the HRH1 were found. The final piece of evidence supporting the identity of the novel exons that we found is the high degree of evolutionary conservation present within these regions between species for exon B, which also extends to the main splice sites identified. These data contrast with a previous study that describes an upstream gene structure of the HRH1 gene (15). In this article, a TSS was identified at the 5' end of exon H. We believe that this probably represents a truncated splice variant, based upon our inability to identify this transcript except with additional 5' exons spliced to exons H and I (see Figure 1). Further evidence against this TSS being active in the different cell types studied comes from reporter gene expression studies, in which only small increments in luciferase expression were induced by constructs containing regions 5' to this putative TSS. Murata and colleagues have recently shown that the region upstream of exon H was able to direct luciferase reporter gene expression in a granulocyte-macrophage colony–stimulating factor stimulated U937 human monocytic cell line (26), so it is possible that this region is active as a promoter in other tissues, or that it functions in a regulatory manner at the transcriptional level before mRNA processing occurs. In our studies, attempts to identify HRH1 transcripts in PBMN cells demonstrated only low levels of expression under unstimulated conditions. Although the HRH1 is similar to some other G protein–coupled receptors in having an intronless coding region, the complexity of the upstream gene structure suggests that active transcriptional regulation is critical in achieving tissue-specific expression profiles.
There is evidence for regulated expression of the genes encoding HRH1s by a range of mediators in several cell types and model systems. These include IL-13, which has been shown to upregulate HRH1 mRNA expression (27), granulocyte-macrophage colony–stimulating factor, shown to cause HRH1 mRNA to increase in human monocytic cell line U937 (26), IL-4, reported to upregulate HRH1 mRNA in human rheumatoid synovial fibroblasts (28), and human umbilical vein endothelial cells (29), platelet-derived growth factor BB, shown to increase HRH1 mRNA expression in cultured human aortic intimal smooth muscle cells (30), and platelet activating factor, which has been shown to induce mRNA upregulation of the rat H1 gene in trigeminal nerve ganglions (31). Also, retinoic acid, used to cause differentiation in a neuroblastoma cell line, caused upregulation of human H1Rs (32). Insulin and 4- Our findings that the HRH1 gene shows splice variation, and therefore has at least three putative promoter regions, may begin to explain some of the differences seen in the effects of cellular mediators on histamine H1 gene expression between cell types and species, although further studies will be needed to define the mechanisms underlying the effects of different mediators.
In the complexity of mRNA transcripts arising from this locus, the HRH1 gene is similar to other airway G protein–coupled receptors, such as the muscarinic M2 receptor, CHRM2 (35), although other receptors with intronless coding regions do have simpler upstream structures. A good example of the latter is the human In addition to defining the main promoter regions used by the HRH1 in a range of relevant cell types, we have undertaken a polymorphism search of the coding region, the region containing the previously studied promoter, and the novel promoters that we identified. The whole genomic region is relatively rich in polymorphisms, and a number of novel SNPs were identified; some previously identified SNPs were also found. In summary, the upstream gene structure of the HRH1 gene is considerably more complex than previously believed, with three main promoters being present. The mechanisms underlying tissue-specific expression of different transcripts require further study. The evidence provided here for active, cell type–specific regulation of transcript species suggests an important physiologic role for transcriptional regulation at this locus in governing response to stimuli that release histamine in the airways and other tissues.
The authors thank Amanda Wheatley for preparation of genomic DNA used in this study and Samuel J. Wadsworth for HBEC cell culture.
This work was supported in part by Asthma UK. This article has an online supplement, which is accessible from this issue's table of contents at www.atsjournals.org Originally Published in Press as DOI: 10.1165/rcmb.2005-0408OC on February 16, 2006 Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript. Received in original form November 4, 2005 Accepted in final form January 19, 2006
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
2 adrenergic receptor, used as a positive control, were 
1kb and 2kb). Three additional constructs with the promoter and 5' UTR (previously published) were also studied for comparative purposes (Hs, H1k, H2k). The vector pGL3Control (pGL3C) was used as a positive control.
B/NF-
