Variability of 5-HT2C receptor cys23ser polymorphism among European populations and vulnerability to affective disorder


Substantial evidence supports a role for dysfunction of brain serotonergic (5-HT) systems in the pathogenesis of major affective disorder, both unipolar (recurrent major depression) and bipolar.1 Modification of serotonergic neurotransmission is pivotally implicated in the mechanism of action of antidepressant drugs2 and also in the action of mood stabilizing agents, particularly lithium carbonate.3 Accordingly, genes that code for the multiple subtypes of serotonin receptors that have been cloned and are expressed in brain,4 are strong candidates for a role in the genetic etiology of affective illness. We examined a structural variant of the serotonin 2C (5-HT2C) receptor gene (HTR2C) that gives rise to a cysteine to serine substitution in the N terminal extracellular domain of the receptor protein (cys23ser),5 in 513 patients with recurrent major depression (MDD-R), 649 patients with bipolar (BP) affective disorder and 901 normal controls. The subjects were drawn from nine European countries participating in the European Collaborative Project on Affective Disorders. There was significant variation in the frequency of the HT2CR ser23 allele among the 10 population groups included in the sample (from 24.6% in Greek control subjects to 9.2% in Scots, χ2?=?20.9, df 9, P?=?0.01). Logistic regression analysis demonstrated that over and above this inter-population variability, there was a significant excess of HT2CR ser23 allele carriers in patients compared to normal controls that was demonstrable for both the MDD (χ2?=?7.34, df 1, P?=?0.006) and BP (χ2?=?5.45, df 1, P?=?0.02) patients. These findings support a possible role for genetically based structural variation in 5-HT2C receptors in the pathogenesis of major affective disorder.


It is widely accepted that the genetic basis of affective disorders, both unipolar (recurrent major depression) and bipolar (manic depressive illness) is complex and is likely to involve several genes and also environmental factors.6,7 A principal aim of the European Collaborative Project on Affective Disorders (ECPAD) is to harness the statistical power potentially provided by the large numbers of subjects who can be recruited by the participating centers in order to elucidate the role of candidate genes and environmental factors in the pathogenesis of affective illness.8 In the course of the project a very large sample has been recruited and clinically evaluated, with blood samples obtained for DNA extraction. The method of recruitment of patients has been described in detail by Souery et al8 and is summarized below in the Methods section. Subjects were genotyped for the HT2CR cys23ser polymorphism in the participating laboratories according to a standardized protocol (see below). Clinical and genotype data were centralized in the ECPAD database in Brussels and were electronically transferred to BL and FM for statistical analysis. After removal of subjects whose geographical origin was unclear, whose genotype data were ambiguous or contradictory or who were missing key demographic variables (total number of subjects dropped?=?149), the subject groups available for the current analysis were generated. These consisted of 513 patients with a diagnosis of Major Depression: Recurrent (MDD-R) according to DSM-IV criteria,9 649 patients with DSM-IV Bipolar Disorder (BP) and 901 normal control subjects who were screened to exclude personal and family history of DSM-IV Axis I psychiatric disorders. The distribution of the 2063 subjects according to diagnosis, gender and population group is summarized in Table?1. Although nine countries participated in the current project (Belgium, Bulgaria, Croatia, Germany, Greece, Israel, Italy, Sweden and Scotland), there were 10 population groups in the analysis (Table?1). These were generated by combining individuals of the same geographical origin across national samples.

Table 1 Distribution of the sample according to diagnosis, gender and population

The initial focus of our analysis was on determining differences between the population samples in the distribution of the HT2CR cys23ser polymorphism. After that, the distribution of the polymorphism in relation to mood disorder diagnosis was examined with the population contribution to variability taken into account. MDD and BP diagnoses were considered separately in these analyses under the hypothesis that genetically based variability in HT2CR may contribute to vulnerability to these two disorders in different ways. As shown in Table?2, there was substantial variability in the frequency of the HT2CR ser23 allele among the population samples, ranging from 24.6% in the Greek control subjects to 9.2% in the Scottish controls (χ2?=?20.9, df 9, P?=?0.01). In previous studies of this polymorphism in mood disorders, reported frequencies of the ser23 allele in control groups have been in the range of 6% in Croatians,10 9% in Spaniards,11 20% in Israeli Ashkenazi Jews and 21% in Israeli Non-Ashkenazi Jews.12 It is clear that a direct comparison of the frequency of the ser23 allele in combined patient and control groups could lead to spurious results that reflect differences in the frequency of the allele among populations and not association with illness.

Table 2 HT2CR Ser23 allele frequency according to diagnosis and population

To examine the association of HT2CR cys23ser with mood disorders, we first compared allele frequencies in the patient and control groups stratified by population group. Prior to that we performed power analyses. These were based on small effect sizes since previous reports have shown either small or no differences in the distribution of the HT2CR cys23ser polymorphism between affective patients and controls.10,11,12,13 The power analyses revealed that most of the individual samples, for MDD in particular, had less than 80% power to demonstrate a difference of 15% in the frequency of the ser23 allele between patients and controls. Thus, the separate analyses shown in Table 2 should be regarded as exploratory.

All the control samples were in Hardy–Weinberg equilibrium (females) except for the control groups from Croatia (χ2?=?5.24, P?=?0.02) and Greece (χ2?=?6.96, P?=?0.008). In both cases, there was a deficiency in the observed proportion of heterozygotes. Therefore, analyses involving all the populations were also carried out without the Croatian and Greek samples in order to avoid the possibility of spurious results.

As shown in Table?2, in six out of the 10 samples, there was an excess of the ser23 allele in the patients with MDD compared to the controls. This difference was highly variable and reached statistical significance only in the Bulgarian sample. In four of the samples the excess was of the order of 5% or more. However, in four samples the direction of the difference was opposite with the cys23 allele predominating; in one case by more than 5% but not to a statistically significant degree in any of the samples. Genotype distribution was also examined but this analysis was limited by the size of the individual case groups when separated by gender as well as population group and the results are thus not presented (available on request). Considering ser23 allele carriers (as homo- or heterozygous females or hemizygous males), in eight out of 10 groups there were more ser23 carriers among the MDD patients than controls (Figure?1, a), the difference reaching statistical significance in the Bulgarian and German samples. Considering the ser23 allele under a ‘recessive’ model did not yield a clear trend except for a significant absence of ser23 hemi- and homozygotes among the Italian patients with major depression (data not shown).

Figure 1

HT2CR ser23 allele carriage in patients with major depression (a) and patients with bipolar disorder (b) vs controls from the same population. Ser23 carrier status, MDD vs CON: χ2?=?7.29, df 1, P?=?0.007 (Bulgaria). Ser23 carrier status, MDD vs CON: χ2?=?4.25, df 1, P?=?0.04 (Germany). For all other comparisons, P?>?0.05.

In the comparison of BP and controls by population groups, an excess of ser23 allele in the BP patients was seen in six of the 9 groups (Table?2). The difference was 5% or more in five of the groups but was not statistically significant in any of them. Genotypic analysis is not presented because of the small sizes of the groups when separated by gender and population group (available on request). Considering ser23 allele carriers (Figure?1, lower panel), the rate was higher among the BP patients than the controls in six of the 9 groups but no differences were statistically significant.

In order to determine whether there is an association of HT2CR ser23 with affective disorder, over and above the difference in the frequency of the allele among the different populations, we performed stepwise logistic regression analyses. In these analyses, the response variable was ser23 allele carrier status (ser23–ser23 homozygous females, ser23 hemizygous males and cys23–ser23 heterozygous females vs cys23–cys23 homozygous females and cys23 hemizygous males). This approach was used because of the suggestive findings from the exploratory analyses reported above and also the fact that it enabled us to count each female subject once only as for the hemizygous males. Since the approach may be seen as implying a ‘dominant’ role for ser23, we also examined the alternative recessive hypothesis in a separate analysis, correcting the output of both analyses for multiple testing. With the stepwise logistic regression analysis we evaluated the effect of ‘population’ and ‘diagnosis’ as predictors of the response variable, ser23 allele carrier status; the likelihood ratio (LR) test of the two models (population?+?diagnosis and population alone) determines the significance of diagnosis, controlled for the potential confounding effect of population.

The log of the likelihood ratio for the diagnosis of major depression in predicting ser23 allele carrier status was 3.724 (χ2?=?7.45, df 1, P?=?0.006) (Table?3). For bipolar disorder the log of the likelihood ratio was 2.72 (χ2?=?5.45, df 1, P?=?0.02). (Table?4). Both these differences remain significant when corrected for the fact that the ser23 recessive model (ser23 homozygous females and hemizygous males) was also tested. The recessive model did not reveal a significant relationship with major depression or bipolar disorder. An additional, conservative analysis without the Croatian and Greek samples, due to the departure of their control groups from HWE, confirmed the findings (MDD, χ2?=?8.80, df 1, P?=?0.003; BP, χ2?=?5.48, df 1, P?=?0.02) and showed that the increased risk for mood disorders associated with ser23 allele carrier status cannot be attributed to population distorting forces.

Table 3 Backward stepwise logistic regression for variables predicting HT2CR serine23 allele carrier status in patients with major depression and normal control subjects from 10 European population groups
Table 4 Backward stepwise logistic regression for variables predicting HT2CR serine23 allele carrier status in patients with bipolar disorder and normal control subjects from nine European population groups

Having demonstrated an association of major depression and bipolar disorder with HT2CR ser23 allele carrier status, we sought to determine whether there are specific clinical correlates of carrying this allele in the context of either of the two diagnoses. This was determined by comparing ser23 allele carriers with ser23 non-carriers (cys23 hemi- or homozygotes) on a series of clinical variables for which data had been entered into the ECPAD database. These variables included age at first depressive episode, frequency of depressive episodes (corrected for years of illness) and history of suicide attempts, in both diagnostic groups, and age at first manic episode and frequency of manic episodes (corrected for years of illness) in the bipolar patients. None of these comparisons emerged as statistically significant (data available on request).

The results of this study should be considered on the background of previous reports of the HT2CR cys23ser polymorphism in major affective disorders. These are summarized in Table?5. We could identify only one previous study in major depression.12 This was conducted in Israel and showed no significant difference between patients with MDD and controls among Ashkenazi or Non-Ashkenazi Jews. However, the study did not have sufficient power to identify a difference in ser23 allele carriers of 15% in the Ashkenazi sample and of 25% in the non-Ashkenazi sample (at alpha 5% and with a power of 80%). This observation is also applicable to the previous study of HT2CR cys23ser in bipolar disorder by Oruc et al,10 while those of Gutierrez et al11 and Vincent et al13 had somewhat stronger power. Nevertheless, Oruc et al10 and also Gutierrez et al11 found evidence for an association of HT2CR ser23 with bipolar disorder, although this was limited to females and was marginal in the Oruc et al10 data. In the Gutierrez et al11 study the finding was demonstrable with ser23 as recessive, in contradistinction to our study which did not show evidence for a recessive hypothesis. The study of Vincent et al13 did not find any significant differences between bipolar patients and controls in the distribution of the HT2CR cys23ser polymorphism.

Table 5 HT2CR cys23ser polymorphism in bipolar disorder and major depression: previous studies

Spurious findings due to population effects are considered a major problem of case control studies in psychiatric genetics,14 although there are views that the magnitude of the problem is overwrought and that it can be dealt with by appropriate matching and by statistical means.15 Family-based association studies effectively address the potential problem of spurious findings due to population stratification.16 On the other hand, such samples are more difficult to recruit than population-based case control samples leading to difficulties in achieving adequate statistical power. A further limitation is that in the family-based design it is more difficult to evaluate environmental variables in conjunction with genetic variables.15,17 This is a significant limitation in studying the genetics of complex disorders. Our study demonstrates the power of a case control design employing very large samples to demonstrate small genetic effects. It is also shows that population stratification which can lead to spurious findings,18 may be accounted for in the context of an appropriate statistical model. Use of genomic controls is another approach that may be implemented for this purpose and a comparison of the approaches would be of interest.19

Association of HT2CR with major depression and bipolar disorder, demonstrated in this study, does not permit the conclusion that the polymorphism examined (cys23ser) plays a causative role in either disorder. A brief consideration of the role of 5-HT2C receptors in these disorders is nevertheless warranted. Serotonin receptors of the 5-HT2C subtype have a very characteristic distribution in brain, which is quite similar in all mammalian species. The highest concentration of 5-HT2C receptors is on the epithelial cells of the choroid plexus and it is hypothesized that they may regulate the composition and volume of cerebrospinal fluid.20 5-HT2C receptors have also been identified in human basal ganglia, particularly substantia nigra, globus pallidus and caudate, by autoradiographic techniques,20 and by radioligand binding with 3H-mesulergine in these areas and also hippocampus, amygdala, hypothalamus and prefrontal cortex.21 Studies in ‘knockout’ mice lacking functional 5-HT2C receptors have suggested the possible importance of these receptors in the serotonergic control of food intake and in the tonic inhibition of neuronal excitability.22,23 The efficacy of anticonvulsant drugs as mood stabilizers24,25 is interesting to note in the context of the latter correlate of 5-HT2C receptor function. Studies with 5-HT2C receptor agonists in humans, specifically m-chloro-phenylpiperazine (mCPP), indicate that direct stimulation of these receptors induces a spectrum of anxiety manifestations in psychiatric patients and normal controls.26

Neuroendocrine challenge studies with d and dl fenfluramine have consistently shown blunted release of prolactin in depressed patients and it has been reported that this blunting is a trait phenomenon which is also observed in the well state.27 Fenfluramine-induced prolactin release has been shown to be mediated by 5-HT2C receptors.28 The HT2CR cys23ser polymorphism examined in this study may have functional effects on mCPP binding in vitro.5 In a study of normal female subjects, Quested et al29 found a reduced hypophagic response to mCPP in ser23 carriers, but no effect on endocrine or thermic responses. Thus, the cys23ser polymorphism may have functional correlates in humans, but this needs to be further examined.

In conclusion, we have found that both major depression and bipolar disorder are associated with HT2CR ser23 allele carrier status and that this effect is demonstrable over and above considerable inter-population variability in the frequency of this allele. Replication of this finding in a sample of sufficient statistical power is important. If it is confirmed that the ser23 allele is associated with major depression and bipolar disorder, further studies will be needed in order to establish whether ser23 itself is the risk factor or another polymorphism in HT2CR or in a nearby gene that is in linkage disequilibrium with cys23ser. Functional studies will be needed in order to determine the mechanism whereby 5-HT2C receptor mutation may contribute to the pathogenesis of major depression and bipolar disorder.


The patients for this study were consecutively recruited in the nine participating centers. Diagnoses of major depression: recurrent and bipolar disorder were according to DSM-IV criteria9 and were based on semi-structured interview using the Schedule for Affective Disorders and Schizophrenia: Lifetime version (SADS-L) (New York State Psychiatric Institute, New York, 1978) or, for the Bulgarian patients, the Schedules for Clinical Assessment in Neuropsychiatry (SCAN) (World Health Organisation, Geneva, 1992). Control subjects were screened to exclude a personal or family history of major psychiatric disorder. Background data and specific clinical characteristics of the subjects were centralized in Brussels and entered into the ECPAD database.

DNA was available from all subjects after being extracted from fresh whole blood by standard methods. Genotyping was performed in the participating laboratories by PCR-based restriction analysis, according to a standardized protocol that was provided by the reference laboratory (University of Antwerp, CVB). Primers, PCR conditions and restriction digestion were as described in Xie et al.30 An error checking procedure was implemented to insure standardization of genotyping by typing three identical CEPH individuals in each participating group which served as reference genotypes for intergroup genotyping and scoring.

The statistical analyses employed Stata Statistical Software (Release 6, Stata Corporation; 1999). For categorical analyses, maximum likelihood chi-square statistics were employed. For bivariate comparisons of parametric data, the student's t-test was used. To evaluate the relationship of more than one independent variable to genotype, a stepwise logistic regression procedure was implemented. The outcome (‘response’) variable of the logistic regression was the presence of the HT2CR ser23 allele. The covariates of the model were the clinical phenotype of interest ie UP vs control in the first model, BP vs control in the second model and ethnicity, given a possible population effect on the frequency of HT2CR ser23 allele. In this case, the phenotype of interest is a predictor of the outcome and ethnicity a potential confounder. Ethnicity was tested generating a set of dummy variables equal to the number of populations included in the sample in order to evaluate the specific effect of any population. Results from the model are presented as Likelihood Ratio Tests. P values of <0.05 (two-tailed) were regarded as significant in all analyses. The data were also evaluated with a stratified analysis for 2 × 2 tables using both the Cornfield and the Mantel–Haenszel methods to control for the population effect as more standard ways to look for possible confounders in a case-control study. With this approach, as in the logistic regression analysis, we checked for the homogeneity of the odds ratios across populations and tested whether the combined effect of the odds ratio showed an equivalent significant effect of ser23 allele carrier status in increasing the risk for the disorder, ie major depression and bipolar disorders respectively. These results may be found at our website: http://www.


  1. 1

    Maes M, Meltzer HY . The serotonin hypothesis of major depression. In: Bloom FE, Kupfer DE (eds) Psychopharmacology: The Fourth Generation of Progress Raven Press: New York 1995 pp?933–944

    Google Scholar 

  2. 2

    Newman ME, Lerer B, Shapira B . 5-HT-1a receptor mediated effects of antidepressants Prog Neuropsychopharmacol Biol Psychiatry 1993 17: 1–19

    Article  CAS  Google Scholar 

  3. 3

    Price LH, Charney DS, Delgado PL, Heninger GR . Lithium and serotonergic function: implications for the serotonergic hypothesis of depression Psychopharmacology 1990 100: 3–12

    Article  CAS  Google Scholar 

  4. 4

    Barnes NM, Sharp T . A review of central 5-HT receptors and their function Neuropharmacology 1999 38: 1083–1152

    Article  CAS  Google Scholar 

  5. 5

    Lappalainen J, Zhang L, Dean M, Oz M, Ozaki N, Yu D-H et al. Identification, expression, and pharmacology of a Cys23-Ser23 substitution in the human 5-HT2C receptor gene (HTR2C) Genomics 1995 27: 174–279

    Article  Google Scholar 

  6. 6

    Segman RH, Lerer B . Genetic factors in the etiology of bipolar disorder. In: Soares J, Gershon S (eds). Bipolar Disorders: Basic Mechanisms and Therapeutic Implications Marcel Dekker: New York 2000

    Google Scholar 

  7. 7

    Tsuang MT, Faraone SSV . The Genetics of Mood Disorders Johns Hopkins University Press: Baltimore 1990

    Google Scholar 

  8. 8

    Souery D, Lipp O, Mahieu B, Serretti A, Cavallini C, Ackenheil M et al. European collaborative project on affective disorders: interactions between genetic and psychosocial vulnerability factors Psychiatr Genet 1998 8: 197–205

    Article  CAS  Google Scholar 

  9. 9

    American Psychiatric Association . Diagnostic and Statistical Manual of Mental Disorders (4th edn) APA: Washington, DC 1994

    Google Scholar 

  10. 10

    Oruc L, Verheyen GR, Furac I, Jakovljevic M, Ivezic S, Raeymaekers P et al. Association analysis of the 5-HT2C receptor and 5-HT transporter genes in bipolar disorder Am J Med Genetics (Neuropsychiat Genet) 1997 74: 504–506

    Article  CAS  Google Scholar 

  11. 11

    Gutierrez B, Fananas L, Arranz MJ, Valles V, Guillamat R, van Os J et al. Allelic association analysis of the 5-HT2C receptor gene in bipolar affective disorder Neurosci Lett 1996 212: 65–67

    Article  CAS  Google Scholar 

  12. 12

    Frisch A, Postilnick D, Rockah R, Michaelovsky E, Postilnick S, Birman E et al. Association of unipolar major depressive disorder with genes of the serotonergic and dopaminergic pathways Mol Psychiatry 1999 4: 389–392

    Article  CAS  Google Scholar 

  13. 13

    Vincent JB, Masellis M, Lawrence J, Choi V, Gurling H, Phil M et al. Genetic association analysis of serotonin system genes in bipolar affective disorder Am J Psychiatry 1999 156: 136–138

    Article  CAS  Google Scholar 

  14. 14

    Malhotra AK, Goldman D . Benefits and pitfalls encountered in psychiatric genetic association studies Biol Psychiatry 1999 45: 544–550

    Article  CAS  Google Scholar 

  15. 15

    Jorm AF, Easteal S . Assessing candidate genes as risk factors for mental disorders: the value of population-based epidemiological studies Soc Psychiatry Psychiatr Epidemiol 2000 35: 1–4

    Article  CAS  Google Scholar 

  16. 16

    Hamer D, Sirota L . Beware the chopsticks gene Mol Psychiatry 2000 5: 11–13

    Article  CAS  Google Scholar 

  17. 17

    Wacholder S, Silverman DT, McLaughlin JK, Mandel JS . Selection of controls in case-control studies. II. Types of controls Am J Epidemiol 1992 135: 1029–1041

    Article  CAS  Google Scholar 

  18. 18

    Risch NJ . Searching for genetic determinants in the new millenium Nature 2000 405: 847–856

    Article  CAS  Google Scholar 

  19. 19

    Devlin B, Roeder K . Genomic control for association studies Biometrics 1999 55: 997–1004

    Article  CAS  Google Scholar 

  20. 20

    Prazos A, Probst A, Palacios JM . Serotonin receptors in the human brain. III. Autoradiographic mapping of serotonin-1 receptors Neuroscience 1987 21: 97–122

    Article  Google Scholar 

  21. 21

    Marazziti D, Rossi A, Giannaccini G, Zavaglia KM, Dell'Osso L, Lucacchini A et al. Distribution and characterization of [3H]mesulergine binding in human brain postmortem Eur Neuropsychopharmacol 1999 10: 21–26

    Article  CAS  Google Scholar 

  22. 22

    Tecott LH, Sun LM, Akana SF, Strack AM, Lowenstein DH, Dallman MF et al. Eating disorder and epilepsy in mice lacking 5-HT2C serotonin receptors Nature 1995 374: 542–546

    Article  CAS  Google Scholar 

  23. 23

    Brennan TJ, Seeley WW, Kilgard M, Schreiner CE, Tecott LH . Sound-induced seizures in serotonin 5-HT2C receptor mutant mice Nature Genet 1997 16: 387–390

    Article  CAS  Google Scholar 

  24. 24

    Lerer B, Moore N, Meyendorff E, Cho SR, Gershon S . Carbamazepine versus lithium in mania: a double blind study J Clin Psychiatry 1987 48: 89–93

    PubMed  CAS  Google Scholar 

  25. 25

    Bowden CL . New concepts in mood stabilization: evidence for the effectiveness of valproate and lamotrigine Neuropsychopharmacology 1998 19: 194–199

    Article  CAS  Google Scholar 

  26. 26

    Pigott TA, Zohar J, Hill JL, Bernstein EB, Grover GN, Zohar-Kadouch R et al. Metergoline blocks the behavioral and neuroendocrine effects of orally administered m-chlorophenlypiperazine in patients with obesessive-compulsive disorder Biol Psychiatry 1991 29: 418–426

    Article  CAS  Google Scholar 

  27. 27

    Newman ME, Shapira B, Lerer B . Evaluation of central serotonergic function in affective and related disorders by the fenfluramine challenge test: a critical review Int J Neuropsychopharmacol 1998 1: 49–70

    Article  CAS  Google Scholar 

  28. 28

    Coccaro EF, Kavoussi RJ, Oakes M, Cooper TB, Hauger R . 5-HT-2a/2c receptor blockade by amesergide fully attenuates prolactin response to d-fenfluramine challenge in physically healthy human subjects Psychopharmacology 1996 126: 24–30

    Article  CAS  Google Scholar 

  29. 29

    Quested DJ, Whale R, Sharpley AL, McGavin CL, Crossland N, Harrison PJ et al. Allelic variation in the 5-HT2C receptor and functional response to the 5-HT2c receptor antagonist mCPP Psychopharmacol (Berl) 1999 144: 306–307

    Article  CAS  Google Scholar 

  30. 30

    Xie E, Zhu L, Zhao L, Chang LS . The human serotonin 5-HT2C receptor: complete cDNA, genomic structure, and alternatively spliced variants Genomics 1996 35: 551–561

    Article  CAS  Google Scholar 

Download references


This project was supported by the European Community Biomed 2 Grant (No. BMH4-97-2307).

Author information



Corresponding author

Correspondence to B Lerer.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lerer, B., Macciardi, F., Segman, R. et al. Variability of 5-HT2C receptor cys23ser polymorphism among European populations and vulnerability to affective disorder. Mol Psychiatry 6, 579–585 (2001).

Download citation


  • bipolar disorder
  • major depression
  • unipolar disorder
  • affective disorder
  • 5-HT2C receptor gene
  • HT2CR
  • genetic polymorphism
  • genetic association
  • population genetics