G polymorphism in the 5-hydroxytryptamine type 2A receptor gene affects promoter activity

The –1438A/G Polymorphism in the 5-Hydroxytryptamine Type 2A Receptor Gene Affects Promoter Activity Michael J. Parsons, Ursula M. D’Souza, Maria-Jesus Arranz, Robert W. Kerwin, and Andrew J. Makoff Background: The –1438A/G single nucleotide polymorphism (SNP) lies just upstream of two alternative promoters for the 5-hydroxytryptamine type 2A (5-HT2A) receptor gene (HTR2A) and is in strong linkage disequilibrium with the 102T/C SNP. Both SNPs are associated with numerous psychiatric disorders and related phenotypes. A possible functional affect of the –1438A/G SNP might underlie associations of both linked SNPs with these neuropsychiatric disorders. A prior investigation into affects of this SNP on promoter function, lacking the more downstream promoter, found no significant difference with a reporter gene assay. Methods: To investigate possible functional effects of –1438A/G on either promoter, two different reporter gene assays were used in three cell lines. Results: Promoter activity was consistently detected that, in the presence of the SV40 enhancer, was significantly greater in the presence of the A allele relative to the G allele but only in cell lines that express endogenous HTR2A, suggesting that transcriptional factor(s) and the presence of both promoters might be necessary to elicit this effect. Conclusions: These findings show that the –1438A/G SNP has the potential to modulate HTR2A promoter activity and might be the functional variant responsible for the associations of both SNPs with many neuropsychiatric phenotypes. Key Words: ⫺1438A/G polymorphism, 5-HT2A receptor, enhancer, promoter, psychiatric disorders, reporter gene assay

T

he 5-hydroxytryptamine type 2A (5-HT2A) receptor gene (HTR2A) was one of the first human serotonin receptor genes to be cloned (Chen et al 1992; Saltzman et al 1991; Sparkes et al 1991). The gene is located on the long arm of chromosome 13 and includes three exons and two introns, spanning 20 kilobases (kb) (Chen et al 1992). It seems to have two alternative promoters, with a silencer element just downstream of the second promoter element (Zhu et al 1995). Four transcription initiation sites have been described (Shih et al 1996; Zhu et al 1995), with similar initiation sites found in the mouse and rat homologues (Chen et al 1992; Ding et al 1993; Du et al 1994). Significant genetic variation exists within HTR2A. Of particular interest are two common single nucleotide polymorphisms (SNPs) that are in almost complete linkage disequilibrium (LD) with each other: a silent variant in exon 1 (102T/C) and a variant close to the promoter region (–1438A/G) (Spurlock et al 1998). An association has been found between the 102C allele and schizophrenia in some studies, and the association was supported by a meta-analysis, although the effect was small (Williams et al 1997). A stronger association has been found between the 102C allele and poor response to clozapine in schizophrenia and was also significant in a meta-analysis (Arranz et al 1998). Both SNPs have also been associated with other psychiatric disorders, including anorexia nervosa (Collier et al 1997; Sorbi et al 1998), bipolar disorder (Bonnier et al 2002; Chee et al 2001), and seasonal affective disorder (Enoch et al 1999). A possible functional effect of the –1438A/G SNP might From the Department of Clinical Neuropharmacology (MJP, M-JA, RWK, AJM) and Social, Genetic and Developmental Psychiatry Centre (UMD), Institute of Psychiatry, King’s College, London, United Kingdom. Address reprint requests to Dr. Michael J. Parsons, King’s College London, Clinical Neuropharmacology PO51, Division of P. Medicine, Institute of Psychiatry, Denmark Hill, London SE5 8AF, United Kingdom; E-mail: [email protected].

0006-3223/04/$30.00 doi:10.1016/j.biopsych.2004.06.020

underlie associations of both linked SNPs with these neuropsychiatric disorders. This SNP lies just upstream of the promoter region of HTR2A and might affect expression by altering promoter function. A postmortem brain study found that the 102C variant (in LD with the –1438G allele) is associated with lower messenger ribonucleic acid (mRNA) and lower protein expression than the T variant (Polesskaya and Sokolov 2002); however, a more recent postmortem study failed to replicate differences in mRNA expression (Bray et al 2004). The possible role of –1438A/G in promoter function was directly investigated with a reporter gene assay, but no significant difference was observed (Spurlock et al 1998); however, the constructs used in this study lacked the more downstream promoter and used only a single reporter gene vector. In view of these contradictory results, we have investigated promoter activity of both variants of –1438A/G in the presence and absence of an enhancer in three human cell lines, using two reporter gene assays.

Material and Methods Cell Culture SH-SY5Y cells (epithelial cells derived from neuroblastoma), HeLa cells (epithelial cells derived from adenocarcinoma), and IMR-32 cells (fibroblast cells derived from neuroblastoma) were used in this study. Cells were cultured as per the American Type Culture Collection cell culture protocols for the appropriate cell types. Endogenous expression of the 5-HT2A receptor (mRNA) was examined in the three cell types by reverse transcriptase polymerase chain reaction (PCR). Total RNA was isolated from the three cell types with the Trizol reagent protocol (SigmaAldrich, St. Louis, Missouri). Two micrograms of RNA were reacted with 200 U Moloney murine leukemia virus reverse transcriptase (Promega, Madison, Wisconsin), 25 units RNase inhibitor (Promega), and 50 ng oligo dT15 primer (Promega). Polymerase chain reactions were performed with an annealing temperature of 60°C and transintronic primers (to eliminate amplification of deoxyribonucleic acic [DNA]). The resulting PCR products were resolved on 3% agarose gels and visualized by ethidium bromide staining and ultraviolet transillumination. BIOL PSYCHIATRY 2004;56:406 – 410 © 2004 Society of Biological Psychiatry

BIOL PSYCHIATRY 2004;56:406 – 410 407

M.J. Parsons et al

Figure 1. Reporter gene vectors. (A) A diagram of the vectors used in this study, including the reporter gene (firefly luciferase reporter gene or chloramphenicol acetyltransferase), SV40 enhancer (only present in the enhancer vectors), and SV40 promoter (only found in the promoter vectors). The inserts were placed as indicated. (B) A diagram representing the inserts used in the above vectors. Note that the variation at the –1438 site is just upstream of the two promoters noted as P1 and P2.

Construct Design The promoter segment of both ⫺1438 variants of HTR2A from –1534 to –538 was amplified from human genomic DNA. Both products were cloned into pCR 2.1-TOPO plasmid with the TOPO cloning kit (Invitrogen, Carlsbad, California). These segments were cloned into the pCAT and pGL3 basic, enhancer, and promoter vectors (Promega) with Xho I and Bgl II restriction sites. All constructs were verified by sequencing. Transfection Cell lines were cotransfected with pCAT construct (2 ␮g) and 1 ␮g pSV-B-Gal with the lipofectAMINE 2000 reagent (Invitrogen). The pGL3 constructs were cotransfected with pRL-SV40 vector (Promega). Each of these constructs was transfected into each cell line in triplicate experiments, with each triplicate experiment being performed three times per cell line (n ⫽ 9). Reporter Gene Assays The chloramphenicol acetyltransferase (CAT) activity was assayed with a CAT enzyme-linked immunosorbent assay kit (Roche Applied Science, Indianapolis, Indiana) and normalized with respect to ␤-galactosidase activity (with a spectrophotometric assay) to control for transfection efficiency. Firefly luciferase reporter gene (Fluc) activity was normalized with respect to Renilla luciferase (Rluc) activity, both of which were measured with the dual luciferase reporter assay system (Promega) and the Fluoroskan Ascent FL (Labsystems, Helsinki, Finland). The data were analyzed by t test analysis.

Figure 2. Relative levels of chloramphenicol acetyltransferase (CAT) activity in cells transfected with pCAT vectors. Relative CAT values for the nine different constructs are indicated by the histograms for the three cell lines: (A) SH-SY5Y, (B) HeLa, and (C) IMR-32. The three different constructs, with or without each insert, are indicated at the bottom. CAT values are relative to pCAT-basic without insert. Error bars are ⫾ SEM. Comparisons were determined by t test (n ⫽ 9), with p values given where significant. Most p values refer to comparisons with the vector without insert; the p values in parenthesis refer to comparisons between the A and G variants.

Results A 1-kb segment of HTR2A was inserted upstream of the CAT gene in three reporter gene vectors: pCAT-basic, pCAT-enhancer, and pCAT-promoter (Figure 1). In the pCAT-basic vector, no other promoter or enhancer element was present, so that expression of CAT is a measure of basal promoter activity in the HTR2A segment. In the other two vectors, either the SV40 enhancer or the SV40 promoter were present, so that CAT activity is a measure of enhanced promoter function in the former, or of enhancer or silencer activity in the latter. SH-SY5Y and HeLa cells express endogenous HTR2A (Zhu et al 1995) and are therefore likely also to contain trans-acting elements necessary for full HTR2A promoter activity. In preliminary experiments, we confirmed expression of endogenous HTR2A by reverse transcriptase PCR in both of these cell lines but not in IMR-32 cells (data not shown), which therefore serves as a control cell line. In SH-SY5Y cells transfected with pCAT-basic containing the HTR2A segment from either the A variant (1.7 ⫾ .3-fold over vector, p ⫽ .0003) or G variant (1.4 ⫾ .2-fold, p ⫽ .0008), a small but significant stimulation of CAT activity was observed (Figure 2A), indicating some promoter activity. The small difference seen www.elsevier.com/locate/biopsych

408 BIOL PSYCHIATRY 2004;56:406 – 410

M.J. Parsons et al proximately 30 –1000 times greater sensitivity than CAT assays (Pazzagli et al 1992). Both variants of the same 1-kb segment of HTR2A were cloned upstream of the Fluc gene into three equivalent pGL3-luciferase reporter gene vectors (Figure 1). Each construct was transfected into the same three human cell lines used before. The cells were co-transfected with pRL-SV40 vector, (which encodes Renilla luciferase, Rluc), enabling Fluc to be normalized with respect to Rluc to control for transfection efficiency in all the cell lines. The results with Fluc as a reporter gene strongly supported those with the pCAT constructs. In both SH-SY5Y (Figure 3A) and HeLa cells (Figure 3B), promoter activity was highly significant in both pGL3-basic and pGL3-enhancer. Furthermore, unlike in pCAT-basic, in pGL3-basic there was a small but significant increase in promoter activity of the A over the G variant (p ⫽ .01 in SH-SY5Y and p ⫽ .02 in HeLa cells), although these differences were far smaller and less significant than in pGL3-enhancer (p ⬍ 10⫺6 for both cell lines). In IMR-32 cells (Figure 3C), there was a small and just significant HTR2A promoter activity in both pGL3-basic and pGL3-enhancer. In all three cell lines, the strong SV40 promoter activity in pGL3promoter was affected by neither variant.

Discussion

Figure 3. Relative levels of firefly luciferase reporter gene (Fluc) activity in cells transfected with pGL3 vectors. Relative Fluc values for the nine different constructs are indicated as in Figure 2.

between the A and G variants was not significant (p ⫽ .09). When SH-SY5Y cells were transfected with each derivative of pCATenhancer, both HTR2A promoter segments exhibited greater stimulation of CAT activity (Figure 2A): A (4.0 ⫾ .8-fold, p ⬍ 10⫺6) and G (2.6 ⫾ .5-fold, p ⫽ 7 ⫻ 10⫺6), indicating stronger promoter activity. The difference between the two variants was significant for this vector (p ⫽ .006), indicating differential effects of the two variants on promoter activity. Transfection of SH-SY5Y cells with pCAT-promoter, with or without insert, resulted in increased CAT activity due to the SV40 promoter, but there was no effect of either HTR2A variant (Figure 2A). Similar results were obtained in HeLa cells (Figure 2B). Both the A variant (1.5 ⫾ .2-fold, p ⫽ .006) and G variant (1.2 ⫾ .1-fold, p ⫽ .0004) exhibited weak promoter activity in pCATbasic and strong promoter activity in pCAT-enhancer (3.0 ⫾ .4-fold, p ⬍ 10⫺6 and 2.1 ⫾ .2-fold, p ⬍ 10⫺6 respectively). Again, there was a significant difference between the A and G variants in pCAT-enhancer (p ⫽ .0003) but not in pCAT-basic (p ⫽ .06). Again, there was no significant difference with either insert in pCAT-promoter. In IMR-32 cells, the cell line lacking endogenous HTR2A mRNA expression, no significant effect on promoter activity was exhibited by the HTR2A segment in any vector (Figure 2C). Comparison of pCAT-promoter with pCAT-basic shows that the presence of the SV40 promoter stimulated CAT activity as efficiently in IMR-32 cells (Figure 2C) as in SH-SY5Y (Figure 2A) and HeLa cells (Figure 2B). We then repeated the study using the more sensitive firefly luciferase reporter gene (Fluc), because these assays have apwww.elsevier.com/locate/biopsych

We have obtained consistent results between the two HTR2A promoter variants in expressing either reporter gene, with large differences in the presence of the downstream enhancer and smaller differences in its absence. Reporter gene assays can be limited by the absence of unknown trans factors and cis elements compared with the normal physiologic situation. In our experiments, significant promoter activity, even in the absence of the SV40 enhancer, was observed in SH-SY5Y and HeLa cells lines, both of which express endogenous 5-HT2A receptor gene mRNA. In contrast, in IMR-32 cells, which do not express the 5-HT2A receptor gene mRNA, the strongest activity for the HTR2A promoter was of borderline significance and was only observed under the most favorable conditions. Yet IMR-32 cells were just as able as the other two cell lines to support the promoter activity of the SV40 promoter (Figures 2 and 3). These findings suggest that SH-SY5Y and HeLa cells contain at least some of the trans-acting factors necessary for HTR2A promoter function, but some of which are absent in IMR-32 cells. The differential effects of the –1438A/G variants on HTR2A promoter activity was therefore detected only in the presence of specific transcription factors. At first sight, our finding of increased promoter function of the –1438-A allele does not seem to support a previous study (Spurlock et al 1998), which failed to find any significant difference between the A and G variants in luciferase reporter gene activity in HeLa cells; however, this study differs from ours in two important respects: 1) it included only the more upstream of the two HTR2A promoters; and 2) it did not include the SV40 enhancer. Our results clearly show the importance of the enhancer, but because we observed a small difference between the two variants in the pGL3-basic vector, this suggests that the more downstream promoter might play a role in the differential effects on promoter activity. As in the previous study, we omitted the area downstream of the two promoters because it contains a silencer element. There are also inconsistencies in postmortem brain studies, all of which have investigated the 102T/C SNP, which is in almost

M.J. Parsons et al complete LD with –1438A/G. In two recent studies, the relative allelic ratio within 102T/C heterozygotes was used to control for variability in HTR2A expression between individuals (in particular, variability due to postmortem artifacts). One of these, in which a restriction fragment length polymorphism-based assay was used (Polesskaya and Sokolov 2002), found that 102T was significantly more expressed than the C allele. Although this difference is in the same direction as predicted by our results with –1438A/G, it is also in the direction predicted for incomplete restriction enzyme digestion, a possible artifact in such studies. In the other study, in which a primer extension-based assay (robust to such artifacts) was used, this difference was not found (Bray et al 2004). In the former study, however, differences between individuals were also investigated, and significant differences were found in both mRNA and protein expression between the three 102T/C genotypes (TT ⬎ TC ⬎ CC) (Polesskaya and Sokolov 2002) that support our data. In another study, significant differences in ketanserin binding activity in postmortem brains were found between different genotypes (Turecki et al 1999). These differences are in the same direction as in our study and in the study by Polesskaya and Sokolov (2002), although such differences were not seen in all binding studies (Hrdina and Du 2001; Kouzmenko et al 1997, 1999). It is possible that the conflicting results across the radioligand binding studies could be limited by the limited specificity of ketanserin for the 5-HT2A receptor (Filip et al 2001; Ismaiel et al 1995; Leysen et al 1981), possibly masking any real effects. How can we assess the conflicting data from both reporter gene and postmortem studies? Interestingly, other promoter systems have yielded similarly discrepant results, for example, monoamine oxidase A promoter (Balciuniene et al 2002; Sabol et al 1998). Although reporter gene studies are limited by possible missing trans factors and cis elements, postmortem studies are not without their limitations as well. Although they provide us with insight into what is happening in human brain tissue, they are in essence investigations in the chemical state of dead rather than living cells. Some of these postmortem studies are vulnerable to changes in the integrity of the mRNA both during and after the death of the patient, although this only applies to comparisons between subjects and not to comparisons within each heterozygote. Moreover, these studies have examined only selected brain areas and have failed to examine HTR2A expression within the developing nervous system. Such studies cannot address any subtle effects that occur only transiently and under certain conditions, such as elicited by a transcription factor under regulation. It is possible that the effects of the –1438A/G SNP on promoter function might only be significant during induced promoter activity, which was mimicked in SH-SY5Y and HeLa cells by the presence of the SV40 enhancer. Under conditions of basal activity, which is likely to occur in postmortem cells, the effects might be more difficult to detect. In summary, we have found that the HTR2A –1438A/G SNP might be a functional SNP that affects promoter activity. The presence of the A allele significantly increases promoter activity but only within cell lines that express HTR2A, suggesting that transcriptional factor(s) might be necessary to elicit this effect. Both promoter elements present in the –1534 to –538 region also seem to be required. These findings suggest that the –1438A/G SNP might have functional effects on expression of the 5-HT2A receptor in the brain and might be responsible for the associations of both –1438A/G and the strongly linked 102T/C SNP with many neuropsychiatric phenotypes.

BIOL PSYCHIATRY 2004;56:406 – 410 409 This work was supported by funding from the Medical Research Council and the National Science Foundation. We thank Dr. Rachel Flomen for the kind gift of SH-SY5Y and IMR-32 cells. Arranz MJ, Munro J, Sham P, Kirov G, Murray RM, Collier DA, et al (1998): Meta-analysis of studies on genetic variation in 5-HT2A receptors and clozapine response. Schizophr Res 32:93–99. Balciuniene J, Emilssson L, Oreland L, Petterson U, Jazin EE (2002): Investigation of the functional effect of monoamine oxidase polymorphisms in human brain. Hum Genet 100:1–7. Bonnier B, Gorwood P, Hamon M, Sarfati Y, Boni C, Marie-Christine HB (2002): Association of 5-HT2A receptor gene polymorphism with major affective disorders: The case of a subgroup of bipolar disorder with low suicide risk. Biol Psychiatry 51:762–765. Bray NJ, Buckland PR, Hall H, Owen MJ, O’Donovan MC (2004): The serotonin-2A receptor gene locus does not contain common polymorphism affecting mRNA levels in adult brain. Mol Psychiatry 9:109 –114. Chee IS, Lee SW, Kim JL, Wang SK, Shin YO, Shin SC, et al (2001): 5-HT2A receptor gene promoter polymorphism –1438A/G and bipolar disorder. Psychiatr Genet 11:111–114. Chen K, Yang W, Grimsby J, Shih JC (1992): The human 5HT2A receptor is encoded by a multiple intron-exon gene. Brain Res Mol Brain Res 14:20 – 26. Collier DA, Arranz MJ, Li T, Mupita D, Brown N, Treasure J (1997): Association between 5-HT2A gene promoter polymorphism and anorexia nervosa. Lancet 350:412. Ding G, Toth M, Zhou Y, Parks C, Hoffman BJ, Shenk T (1993): Glial cellspecific expression of the serotonin 2 receptor gene: Selective reactivation of a repressed promoter. Brain Res Mol Brain Res 20:181–191. Du YL, Wilcox BD, Teitler M, Jeffery JJ (1994): Isolation and characterization of the rat 5-hydroxytryptamine type 2 receptor promoter constitutive and inducible activity in myometrial smooth muscle cells. Mol Pharmacol 45:1125–1131. Enoch MA, Goldman D, Barnett R, Sher L, Mazzanti CM, Rosenthal NE (1999): Association between seasonal affective disorder and the 5-HT2A promoter polymorphism, ⫺1438A/G. Mol Psychiatry 4:89 –92. Filip M, Nowak E, Papla I (2001): On the role of serotonin 2A/2C receptors in the sensitization of cocaine. J Physiol Pharmacol 52:471– 481. Hrdina PD, Du L (2001): Levels of serotonin receptor 2A higher in suicide victims? Am J Psychiatry 158:147–148. Ismaiel AM, Arruda K, Teitler M, Glennon RA (1995): Ketanserin analogies: The effect of structural modification on 5-HT2 serotonin receptor binding. J Med Chem 38:1196 –1202. Kouzmenko AP, Hayes WL, Pereira AM, Dean B, Burnet PWJ, Harrison PJ (1997): 5-HT2A receptor polymorphism and steady state receptor expression in schizophrenia. Lancet 349:1815. Kouzmenko AP, Scaffidi A, Pereira AM, Hayes WL, Copolov DL, Dean, B (1999): No correlation between A(-1438)G polymorphism in 5-HT2A receptor gene promoter and the density of frontal cortical 5-HT2A receptors in schizophrenia. Hum Hered 49:103–105. Leysen JE, Awouters F, Kennis L, Laduron PM, Vandenberk J, Janssen PA (1981): Receptor binding profile of Ketanserin (R 41 468), a novel antagonist at 5-HT2 receptors. Life Sci 28:1015–1022. Pazzagli M, Devine JH, Peterson DO, Baldwin TO (1992): Use of bacterial and firefly luciferases as reporter genes in DEAE-dextran-mediated transfection of mammalian cells. Anal Biochem 204:315–323. Polesskaya OO, Sokolov BP (2002): Differential expression of the “C” and “T” alleles of the 5-HT2A receptor gene in the temporal cortex of normal individuals and schizophrenics. J Neurosci Res 67:812– 822. Sabol SZ, Hu S, Hamer D (1998): A functional polymorphism in the monoamine oxidase A gene promoter. Hum Genet 103:273–279. Saltzman AG, Morse B, Whiteman MM, Ivanshchenko Y, Jaye M, Felder S (1991): Cloning of the human serotonin 5-HT2 and 5-HT1C receptor subtypes. Biochem Biophys Res Commun 181:1469 –1478. Shih JC, Zhu Q, Chen K (1996): Determination of transcription initiation sites and promoter activity of the human 5-HT2A receptor gene. Behav Brain Res 73:59 – 62. Sorbi S, Nacmias B, Tedde A, Ricca V, Mezzani B, Rotella CM (1998): 5-HT2A promoter polymorphism in anorexia nervosa. Lancet 351:1785.

www.elsevier.com/locate/biopsych

410 BIOL PSYCHIATRY 2004;56:406 – 410 Sparkes RS, Lan N, Klisak I, Mohanda T, Diep A, Kojis T, et al (1991): Assignment of a serotonin 5HT-2 receptor gene to human chromosome 13q14q21 and mouse chromosome 14. Genomics 9:461– 465. Spurlock G, Heils A, Holmans P, Williams J, Dsouza UM, Cardno A, et al (1998): A family based association study of T102C polymorphism in 5HT2A and schizophrenia plus identification of new polymorphisms in the promoter. Mol Psychiatry 3:42– 49. Turecki G, Briere R, Dewar K, Antonetti T, Lesage AD, Seguin M, et al (1999):

www.elsevier.com/locate/biopsych

M.J. Parsons et al Prediction of level of serotonin 2A receptor binding by serotonin receptor 2A genetic variation in postmortem brain samples from subjects who did or did not commit suicide. Am J Psychiatry 156:1456 –1458. Williams J, McGuffin P, Nothen M, Owen MJ, Spurlock G (1997): Meta-analysis of association between the 5-HT2a receptor T102C polymorphism and schizophrenia. Lancet 349:1221. Zhu Q, Chen K, Shih JC (1995): Characterization of the human 5HT2A receptor gene promoter. J Neurosci 15:4885– 4895.