A functional serotonin transporter (5-HTT) polymorphism is ... - Nature

Molecular Psychiatry (1998) 3, 328–332  1998 Stockton Press All rights reserved 1359–4184/98 $12.00

ORIGINAL RESEARCH ARTICLE

A functional serotonin transporter (5-HTT) polymorphism is associated with psychosis in neuroleptic-free schizophrenics AK Malhotra1, D Goldman2, C Mazzanti2, A Clifton1, A Breier1 and D Pickar1 1

Experimental Therapeutics Branch, National Institute of Mental Health, National Institutes of Health, Bethesda, Maryland 20892–1380; 2Laboratory of Neurogenetics, National Institute of Alcohol Abuse and Alcoholism, National Institutes of Health, Rockville, Maryland 20852, USA The neurotransmitter serotonin has been implicated in the pathophysiology of psychosis. The serotonin transporter (5-HTT) plays a critical role in regulation of serotonergic function. A recently identified polymorphism in the promoter region of the 5-HTT gene (5-HTTLPR) produces significant differences in 5-HTT expression and function and was found to be associated with anxiety-related traits in healthy volunteers. We investigated whether 5-HTTLPR is associated with psychosis in neuroleptic-free schizophrenic or schizoaffective patients. Fifty patients with schizophrenia or schizoaffective disorder by DSM-III-R criteria were genotyped at 5-HTTLPR and underwent double-blind Brief Psychiatric Rating Scale (BPRS) ratings while neuroleptic-free for approximately 4 weeks. Patients with the 5-HTTLPR ll genotype (n = 19) had significantly higher BPRS ratings for psychosis than patients with the ls (n = 25) or ss (n = 6) genotypes. Examination of individual items revealed a specific significant increase in intensity of hallucinations in patients with the 5-HTTLPR ll genotype. These data provide preliminary evidence for a role of serotonin in the pathophysiology of hallucinations and may represent the identification of an allelic variant that modifies the complex clinical presentation of schizophrenia. Keywords: psychosis; schizophrenia; serotonin; polymorphism; hallucinations; gene

Several lines of evidence suggest that abnormalities in serotonin (5-HT) function are involved in the pathophysiology of psychosis and, in particular, hallucinations. The 5-HT partial agonist lysergic acid diethylamide (LSD) produces psychotic symptoms such as paranoia and hallucinations in normal controls1–3 while the 5-HT agonist, m-chlorophenylpiperazine (mCPP), can exacerbate psychosis in schizophrenia.4,5 Post-mortem studies have revealed alterations in 5-HT2 receptor densities in limbic6,7 and prefrontal cortical8–11 brain regions of schizophrenic patients. Finally, clozapine, the most effective antipsychotic agent for treatment-resistant schizophrenia, has potent effects at several 5-HT receptor subtypes.12–14 5-HT re-uptake plays a critical role in the regulation of 5-HT neurotransmission. The 5-HT transporter (5HTT) terminates the synaptic actions of serotonin by sodium-dependent re-uptake into the presynaptic terminal. Recently, Lesch and colleagues have identified and functionally characterized a polymorphism of the 5-HTT transcriptional control (promoter) region designated 5-HTTLPR.15,16 The polymorphism consists of a Correspondence to AK Malhotra at his current address: Unit of Molecular Psychiatry, Hillside Hospital, Research, 75-59 263rd St, Glen Oaks, NY 11004, USA Received 23 January 1998; revised and accepted 16 March 1998

44-bp insertion (the l or long allele) or deletion (the s or short allele).15,16 Lymphoblast cell lines transfected with the l variant exhibit significantly greater basal transcriptional activity than the s allele.15 Moreover, mRNA concentrations, [125I]RTI-55 binding and [3H]5HT uptake are significantly higher in lymphoblast cell lines derived from human subjects with the ll genotype in comparison to those with either the ls or ss genotypes.16 These data indicate that the 5-HTTLPR ll genotype produces significant effects on 5-HT function in human cells. There are limited data on the effect of 5-HTTLPR on human behavior.16,17 Lesch and colleagues16 have reported that individuals with the ll genotype have significantly lower scores on ratings of an anxiety-related trait than subjects with the ls and ss genotypes. The clinical effects of 5-HTTLPR on putatively 5-HTrelated behavioral phenomena such as psychosis, however, remain to be elucidated. We therefore have determined 5-HTTLPR genotypes in 50 neurolepticfree schizophrenic or schizoaffective patients to investigate the influence of 5-HTTLPR on psychotic symptoms.

5-HTT polymorphism and psychosis AK Malhotra et al

Methods Subjects Fifty patients (36 M, 14 F, mean age = 36.4 ± 8.0 yrs) from the Experimental Therapeutics Branch of the National Institute of Mental Health (NIMH) who met DSM-III-R criteria for schizophrenia or schizoaffective disorder granted written informed consent to participate in the study. Subjects were studied under a human research protocol approved by an NIMH Institutional Review Board. The ethnic composition was 88% white, 10% black, and 2% Asian. As part of a double-blind, placebo-controlled antipsychotic drug trial, patients underwent a neuroleptic-free period of at least 2 weeks. Behavioral rating Psychotic symptoms were assessed with the 18-item Brief Psychiatric Rating Scale (BPRS).18 Each item in the BPRS is rated from 1 (‘not present’) to 7 (‘severe’). BPRS ratings were administered after patients were neuroleptic-free for at least 2 weeks (mean neurolepticfree days = 26.5 ± 6.0) by raters blind to the patient’s neuroleptic status and genotype. Intraclass correlations between all raters were ⬎0.80. These ratings are conducted to reflect the patient’s symptomatology for the week prior to the rating session. The empirically derived BPRS factors19 for thought disturbance (also designated the psychosis or positive symptom factor), withdrawal–retardation (also designated the negative symptom factor), hostility–suspiciousness and anxiety–depression were each analyzed. Genotyping Genomic DNA was extracted from lymphocytes (Puregene, Minneapolis, MN, USA). Polymerase chain reaction (PCR) amplification was accomplished using two flanking primers: U: 5⬘ GGCGTTGCCGC TCTGAATGC 3⬘ and L: 5⬘ GAGGGACTGAGCTGG ACAACCAC 3⬘. The PCR reaction mixture contained a total volume of 20 ␮l with: 100 ng genomic DNA, 0.1 ␮M primers, 40 ␮M dNTPs, 20 ␮M deaza dGTP, and 1 U AmpliTaq with the appropriate buffer (Perkin Elmer, Norwalk, CT, USA). To enhance the specificity and sensitivity of PCR amplification, TaqStart Antibody (Clontech, Palo Alto, CA, USA), a neutralizing monoclonal antibody to Taq DNA polymerase, was used. Cycling conditions were: one denaturing cycle at 95°C for 5 min, two cycles with touch down annealing temperature at 62–63°C for 30 s, and 38 cycles with annealing temperature at 61°C. Final DNA elongation was at 72°C for 10 min. The bands were visualized using a prestained (0.4 ␮g ml−1 of ethidium bromide) 2% agarose gel. The gel was run for 1 h at 120 V. Statistical analysis For the purpose of statistical comparisons, we assumed that 5-HTTLPR acts in a dominant-recessive manner and grouped the ls and ss genotypes. This grouping was based upon: (1) previous functional data demonstrating that mRNA concentrations, [125I]RTI-55 binding, and [3H]5-HT uptake are significantly higher in

lymphoblast cell lines derived from human subjects with the ll genotype compared to either the ls or ss genotypes;16 (2) the previous behavioral association between anxiety-related traits and 5-HTTLPR using this grouping;16 and (3) because only six individuals were identified with the ss genotype. To evaluate the effect of ll genotype on BPRS score, we compared subjects with the ll genotype to subjects with either the ls or ss genotype with Mann–Whitney U tests. The BPRS factors for thought disturbance, withdrawal–retardation, hostility–suspiciousness and anxiety–depression were analyzed. Individual items within BPRS factors found to be significantly affected by genotype were also analyzed with Mann–Whitney U tests. Bonferroni corrections were applied by dividing 0.05 by the number of comparisons to determine the P-value necessary to achieve significance. Student’s t-tests were conducted to analyze the patients’ demographic data. All data are presented ± standard deviation. A qualitative analysis was performed on the BPRS item for hallucinations with a score of 3 representing a threshold between patients with hallucinations rated as ‘often’ or ‘pervasive’ vs those with ratings of ‘not present’ or ‘occasionally’. A chi-squared test was utilized to analyze these data.

Results 5-HTTLPR allele frequencies in our schizophrenia sample were l: 63%, s: 37% (n = 100). Genotype frequencies were: ll: 38%, ls: 50%, ss: 12% (n = 50). These 5-HTTLPR allele (␹2 = 1.39, d.f. = 1, P = 0.24) and genotype (␹2 = 1.65, d.f. = 2, P = 0.44) frequencies were not significantly different from a group of 221 NIMH healthy volunteers with a similar ethnic background (␹2 = 2.0, d.f. = 1, P = 0.15),16 consistent with prior studies demonstrating no significant relationship between the 5-HTT gene and increased risk for schizophrenia.20,21 No significant differences in demographic data were observed between schizophrenic patients with the ll genotype vs patients with the ls or ss genotypes (Table 1). Duration of the neuroleptic-free period between patients with the ll and the ls/ss genotypes was also not significantly different (Table 1). Analysis of the BPRS factors by genotype revealed significantly higher thought disturbance ratings in patients with the ll genotype compared to patients with either the ls or ss genotypes (Table 2). There were no significant differences in BPRS ratings for hostility– suspiciousness, withdrawal–retardation, or anxiety– depression by genotype (Table 2) after Bonferroni correction for the four comparisons (P ⬍ 0.013). Analysis of the three individual items within the BPRS thought disturbance factor revealed that hallucinations (Figure 1) were significantly more frequent and severe in patients with the ll genotype compared to those with the ls/ss genotypes (Table 2) after Bonferroni correction for the three comparisons (P ⬍ 0.017). Qualitative analysis of the effect of the ll genotype on hallucinations revealed that the ll genotype was sig-

329

5-HTT polymorphism and psychosis AK Malhotra et al

330

Table 1 Demographic data (s.d.) from 50 neuroleptic-free schizophrenics with 5-HTTLPR genotypes Genotype ll (n = 19)

ls/ss (n = 31)

35.9 (5.7)

36.7 (9.3)

0.72

12 M, 7 F

24 M, 7 F

0.28

19.6 (4.9)

21.4 (6.7)

0.34

11.5 (11.5)

15.1 (13.9)

0.36

17 W, 2B

27 W, 3B, 1A

0.85

27.9 (5.3)

25.6 (6.3)

0.19

Age Gender Age-of-onset Months hospitalized Ethnicity

P

Neuroleptic-free days

W, white; B, black; A, Asian.

nificantly associated with ratings of ‘often’ or ‘pervasive’ hallucinations rather than with ratings of ‘not present’ and ‘occasionally’ (Table 3).

Discussion This study provides preliminary evidence that a polymorphism that alters serotonin transporter function, 5HTTLPR, is associated with psychosis and, in particular, hallucinations in neuroleptic-free schizophrenic patients. Quantitative analysis revealed that the BPRS factor score for thought disturbance, a measure of positive symptoms, was significantly higher in patients with the ll genotype, in comparison to patients with either the ls or ss genotypes. Moreover, analysis of individual items within the thought disturbance factor demonstrated that the ll genotype was specifically associated with the frequency of hallucinations in neuroleptic-free patients. This association was highly significant with both the quantitative analysis of BPRS score by genotype and the qualitative analysis utilizing Table 2

a threshold score. Moreover, variables such as ethnicity, age, or duration of the neuroleptic-free period did not significantly differ between patients with the different 5-HTTLPR genotypes. In this study, we assumed that 5-HTTLPR has a dominant-recessive effect and grouped the ls and ss genotypes. If the action of the l and s alleles are codominant-additive, the statistical comparison should be conducted across all three genotypes. For thought disturbance, the effect of genotype remains significant (Kruskal-Wallis H = 10.92, d.f. = 2, P = 0.004), with no significant difference between the ls and ss genotypes. For hallucinations, the effect of genotype remains significant (Kruskal-Wallis H = 8.71, d.f. = 2, P = 0.013), with no significant difference between the ls and ss genotypes. Therefore, our data suggest that 5-HTTLPR influences psychosis regardless of whether the genetic model is assumed to be dominant-recessive or codominant. We found that 5-HTTLPR influences the clinical presentation of schizophrenia but is not associated with increased vulnerability to disease. These data are consistent with affected sibling pair studies demonstrating that the genetic predisposition to hallucinations in schizophrenia is independent of the genetic predisposition to illness. Kendler et al22 recently reported a specific lack of concordance for hallucinations in 256 sibling pairs concordant for schizophrenia, consistent with previous studies demonstrating lack of concordance for hallucinations in schizophrenic sibling pairs.23–25 However, it may be difficult to extrapolate heritability estimates for symptoms such as hallucinations from sibling pair studies as these studies reflect assessments of lifetime symptomatology, are primarily retrospective, the majority of patients are not neuroleptic-free and the potent effects of antipsychotic drug treatment on intensity of hallucinations are not controlled for. Whether 5-HTTLPR modifies the presentation of other illnesses in which psychosis is present is unknown. Studies of 5HTTLPR’s role in disorders such as Huntington’s disease, bipolar disorder, and Alzheimer’s disease would also be of interest.

5-HTTLPR genotype and mean BPRS scores (s.d.) in 50 neuroleptic-free schizophrenics

BPRS factor

Genotype ll (n = 19)

Thought disturbance Hallucinations Unusual thought Conceptual disorg. Hostility–suspiciousness Withdrawal–retardation Anxiety–depression

12.1 4.1 4.1 3.9 9.7 10.1 8.5

(4.5) (1.8) (1.6) (1.7) (4.7) (3.9) (2.7)

ls (n = 25) 8.0 2.4 2.8 2.8 7.9 8.5 7.1

(3.8) (1.7) (1.4) (1.3) (3.8) (2.7) (3.2)

*Significance determined by Mann–Whitney U-tests, ll vs ls/ss genotypes. **Significant after Bonferroni correction.

U

P*

146 151 177 196 239 212 190

0.003** 0.004** 0.018 0.048 0.267 0.097 0.036

ss (n = 6) 9.8 2.5 3.7 3.7 7.2 9.2 6.2

(1.9) (1.5) (0.5) (1.0) (2.8) (2.3) (1.6)

5-HTT polymorphism and psychosis AK Malhotra et al

331

Figure 1 5-HTTLPR genotype and BPRS ratings of hallucinations in 50 neuroleptic-free schizophrenics. Data points represent individual patient BPRS ratings. Mean ratings are indicated by the dark lines with the standard deviations represented by the shaded region. Mann–Whitney U-test comparing BPRS ratings between the ll and the ls/ss genotypes was significant (U = 151, P = 0.004).

Table 3 5-HTTLPR genotype and severity of hallucinations in 50 neuroleptic-free schizophrenics* Genotype

ll (n = 19) ls (n = 25) ss (n = 6)

Hallucinations Not present or occasional

Often or pervasive

6 (31.6%) 18 (72.0%) 5 (83.3%)

13 (68.4%) 7 (28.0%) 1 (16.7%)

*␹2 test between ll and ls/ss genotypes (␹2 = 8.78, P = 0.003).

The 5-HTTLPR l allele has been found to result in enhanced serotonin re-uptake. Our data indicate that this allele is associated with frequency of hallucinations. These data are consistent with a recent finding that the serotonin re-uptake inhibitors, fluoxetine, sertraline and paroxetine, block LSD-induced hallucinations in drug abusers26 and a study reporting that fluoxetine treatment resulted in a 30% reduction in hallucinations in schizophrenic patients.27 The mech-

anism by which increased serotonin re-uptake leads to increased frequency of hallucinations is unclear. We speculate that chronically reduced synaptic 5-HT levels may stimulate post-synaptic receptor upregulation or alterations in signal transduction pathways in individuals with schizophrenia. The consequence of these ‘downstream’ alterations may be a hypersensitive 5-HT system that is preferentially activated by neurochemical or environmental perturbation. Some important caveats need to be considered in the interpretation of our data. First, our behavioral assessment was based upon a single BPRS rating in each patient and does not provide detailed information on chronic symptomatology. As the intensity of hallucinations can vary across time, it is possible that our results represent a false positive secondary to chance. However, it should be noted that the BPRS ratings were conducted to reflect symptoms occurring over a time period of one week in a neuroleptic-free patient, and are not strictly a measure of symptoms at one instance in time. Moreover, intra-individual variation in symptoms would be expected to introduce statistical ‘noise’ into the analysis and would bias the study against a

5-HTT polymorphism and psychosis AK Malhotra et al

332

significant association between symptoms and genotype. Second, ethnic variation in allele frequencies between study groups may be a factor in case-control association studies. We are, however, unaware of any relationship between intensity of hallucinations and ethnic differences in schizophrenia. Nevertheless, replication in independent neuroleptic-free study groups and/or studies utilizing haplotype relative risk or transmission disequilibrium test designs (in which parental DNA provides an internal control for ethnic variation) are needed to confirm this finding. In conclusion, 5-HTTLPR appears to influence a cardinal clinical feature of schizophrenia, the positive symptoms and hallucinations. Further studies with other neuroleptic-free patients and with populations in which treatment and drug response are well-defined are needed to replicate and explore the mechanism of the effect of 5-HTT on psychosis. Such studies may yield the information necessary to dissect the heterogeneity of this devastating illness.

9

10

11

12

13

14

15

16

Acknowledgements The authors gratefully acknowledge Longina Akhtar for lymphocyte and DNA preparation, Lisa Picken for statistical assistance, and the staff and fellows of the 4-East Nursing Unit for their expert clinical care of neuroleptic-free schizophrenic patients.

17

18 19

References 1 Gaddum JH. Drugs antagonistic to 5-hydroxytryptamine. In: Wolstenholme GW (eds). Ciba Foundation Symposium on Hypertension. Little Brown and Company: Boston, 1954, pp 75–77. 2 Wooley DW, Shaw E. A biochemical and pharmacological suggestion about certain mental disorders. Proc Natl Acad Sci USA 1954; 40: 228–231. 3 Fischman LG. Dreams, hallucinogenic drug states, and schizophrenia: a psychological and biological comparison. Schizophr Bull 1983; 9: 73–94. 4 Krystal JH, Seibyl JP, Price LH, Woods SW, Heninger GR, Aghajanian GK et al. m-Chlorophenylpiperazine effects in neuroleptic-free schizophrenic patients. Arch Gen Psychiatry 1993; 50: 624–635. 5 Iqbal N, Asnis GM, Wetzler S, Kahn RS, Kay SR, van Praag HM. The MCPP challenge test in schizophrenia: hormonal and behavioral responses. Biol Psychiatry 1991; 30: 770–778. 6 Joyce JN, Shane A, Lexow N, Winokur A, Casanova MF, Kleinman JE. Serotonin uptake sites and serotonin receptors are altered in limbic system of schizophrenics. Neuropsychopharmacology 1993; 8: 315–336. 7 Hashimoto T, Nishino N, Nakai H, Tanaka C. Increase in serotonin 5-HT1A receptors in prefrontal and temporal cortices of brains from patients with chronic schizophrenia. Life Sci 1991; 48: 355– 363. 8 Bennett JP, Enna SJ, Bylund DB, Gillin JC, Wyatt RJ, Snyder SH.

20

21

22

23

24 25 26

27

Neurotransmitter receptors in frontal cortex of schizophrenics. Arch Gen Psychiatry 1979; 36: 927–934. Arora RC, Meltzer HY. Serotonin2 (5-HT2) receptor binding in the frontal cortex of schizophrenic patients. J Neural Transm Gen Sect 1991; 85: 19–29. Mita T, Hanada S, Nishino N, Kuno T, Nakai H, Yamadori T et al. Decreased serotonin S2 and increased dopamine D2 receptors in chronic schizophrenics. Biol Psychiatry 1986; 21: 1407–1414. Laruelle M, Abi-Dargham A, Casanova MF, Toti R, Weinberger DR, Kleinman JE. Selective abnormalities of prefrontal serotonergic receptors in schizophrenia. Arch Gen Psychiatry 1993; 50: 810–818. Meltzer HY, Matsubara S, Lee J-C. Classification of typical and atypical antipsychotic drugs on the basis of dopamine. J Pharmacol Exp Ther 1989; 251: 238–246. Roth BL, Ciaranello RD, Meltzer HY. Binding of typical and atypical antipsychotic agents to transiently expressed 5-HT1C receptors. J Pharmacol Exp Ther 1992; 260: 1361–1365. Watling KJ, Beer MS, Stanton JA, Newberry NR. Interaction of the atypical neuroleptic clozapine with 5-HT3 receptors in the cerebral cortex and superior cervical ganglion of the rat. Eur J Pharmacol 1990; 182: 465–472. Heils A, Teufel A, Petri S, Sto¨ber G, Riederer P, Bengel D et al. Allelic variation of human serotonin transporter gene expression. J Neurochem 1996; 66: 2621–2624. Lesch KP, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S et al. Association of anxiety-related traits with a polymorphism in the serotonin transporter gene regulatory region. Science 1996; 274: 1527–1531. Ebstein RP, Gritsenko I, Nemanov L, Frisch A, Osher Y, Belmaker RH. No association between the serotonin transporter gene regulatory region polymorphism and the Tridimensional Personality Questionnaire (TPQ) temperament of harm avoidance. Mol Psychiatry 1997; 2: 224–226. Overall JE, Gorham DP. The brief psychiatric rating scale. Psychol Rep 1962; 10: 799–812. Hedlund JL, Vieweg BW. The brief psychiatric rating scale (BPRS): a comprehensive review. J Oper Psychiatr 1980; 11: 48–64. Collier DA, Arranz MJ, Sham P, Battersby S, Vallada H, Gill P et al. The serotonin transporter is a potential susceptibility factor for bipolar affective disorder. Neuroreport 1996; 8: 1675–1679. Bonnet-Brilhault F, Laurent C, Thibaut F, Campion D, Chavand O, Samolyk D et al. Serotonin transporter gene polymorphism and schizophrenia: an association study. Biol Psychiatry 1997; 42: 634–636. Kendler KS, Karkowski-Shuman L, O’Neill FA, Straub RE, MacLean CJ, Walsh D. Resemblance of psychotic symptoms and syndromes in affected sibling pairs from the Irish Study of HighDensity Schizophrenia Families: evidence for possible etiologic heterogeneity. Am J Psychiatry 1997; 154: 191–198. DeLisi LE, Goldin LR, Gershon ES. Studies of biological factors associated with the inheritance of schizophrenia: a selective review. J Psychiat Res 1987; 21: 507–513. Slater E. General cases of schizophrenic symptoms. Monatsschrift fu¨r Psychiatrie und Neurologie 1947; 113: 50–58. Slater E. Psychotic and Neurotic Illnesses in Twins. Her Majesty’s Stationery Office: London, 1953. Bonson KR, Buckholtz JW, Murphy DL. Chronic administration of serotonergic antidepressants attenuates the subjective effects of LSD in humans. Neuropsychopharmacology 1996; 14: 425–436. Goff DC, Brotman AW, Waites M, McCormick S. Trial of fluoxetine added to neuroleptics for treatment-resistant schizophrenic patients. Am J Psychiatry 1990; 147: 492–494.