Association study of T102C 5-HT(2A) polymorphism in schizophrenic patients: diagnosis, psychopathology, and suicidal behavior.

The objective of this study was to examine the association between the serotonin (5-HT)2A gene polymorphism (102T/C) and suicidal behavior in schizophrenic inpatients. We studied 129 subjects who met the diagnostic criteria for schizophrenia according to a structured clinical interview (MINI-PLUS). Patients underwent a semistructured interview to assess suicide attempt history and its characteristics. In addition, at least one close relative of the patient was interviewed to assess proband and family suicidal behavior. Healthy controls were students and hospital staff members free of psychiatric and medical illness. Genotypes were determined after polymerase chain reaction amplification of the region of 5-HT(2A)/T102C containing the polymorphic site and digestion with the restriction enzyme Hpall. We found no association between suicidal attempt history and suicide attempt characteristics and genotypic or allele frequencies. Suicidal behavior was also not associated with demographic or psychopathological characteristics. These results suggest that the 5-HT(2A) gene polymorphism (102T/C) is not involved in genetic susceptibility to suicidal behavior, but further studies in a larger sample are needed.

The study of the association between schlzophrenla and the gene encoding the serotonin (5-HT)2A receptor has been a matter of Interest In recent years. The 5-HT2A receptor gene (5-HTR has several polymorphisms in the general population, but the T102C polymorphism is probably the most extensively studied. This polymorphism does not result In an amino acid change in the receptor, as both alleles encode a serine at position 34[1] Nevertheless, there is some evidence showing that the C-allele form can be significantly less functional than the T-allele of 5-HTR both In healthy controls and schizophrenic patients, and this difference is even more prominent in schizophrenic patients.[2]

Additional evidence for a functional role of this mutation comes from a recent meta-analysis showing that, at least in European populations, the C allele Is associated with schizophrenia.[3]

Another point of Interest In the study of 5-HTR is the fact that there are compelling results showing that suicidal behavior is associated with serotonergic dysfunction, and that suicidal behavior is, at least partially, genetically determinated.[4] The 5-HT2A receptor may play a major role in this association.[5]

Since suicidal behavior is a rather frequent event in schizophrenia,[6] an important question to be answered is whether the T102C 5-HT2A polymorphism could be more closely associated with suicidal behavior than with schizophrenia itself.

Therefore, the aim of our work was to examine the association of the T102C polymorphism of the 5-HT2A gene with schizophrenia diagnosis, psychopathological characteristics, and history of suicidal behavior in a well-characterized sample of Brazilian schizophrenic inpatients.

Material and methods

Subjects

A total of 214 unrelated subjects were enrolled after a full explanation of this study and obtaining of informed consent. This study was approved by the University Ethics Committee.

One hundred and twentynine schizophrenic patients, consecutively admitted to Hospital Santa Maria, Belo Horizonte, meeting the Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) diagnostic criteria based on a structured interview (MINI-PLUS),[7,8] without any other current comorbid axis I disorders, were enrolled in this study Healthy controls were students and hospital staff members (n=85), all free of psychiatric and medical illness.

Psychopathology

The psychiatric symptoms of the patients were rated using the Brief Psychiatric Rating Scale (BPRS),[9] which has been validated in Brazil.[10] The principal component analyses of the symptoms assessed by the BPRS were also performed, and four factors emerged: delusion dimension (DD, items 4 + 12 + 15), hebephrenic dimension (DH, items 3 + 13 + 16), paranoid dimension (DP, items 10 + 11 + 14) and anxiety-depressive dimension (DA, items 2 + 5 + 9), as previously described.[5]

Suicide history

The suicide history was assessed using a semistructured interview,[5] a review of medical records, and a supplementary interview with at least one close relative in order to collect information about suicidal behavior in probands. The Lethality Rating Scale (LRS)[11] was used to measure the degree of medical damage of the most lethal lifetime suicide attempt. We used an anchored modified version[12] scored 0 to 8 and, as previously described, a lethality score of three or greater as a cutoff point of three in the LRS scores.[5] The suicide attempt methods, defined by the most lethal lifetime suicide attempt, were classified as nonviolent (drug overdose) or violent (cutting beyond a superficial scratch, jumping from a height, shooting, hanging) as defined elsewhere.[13]

Genotyping

Deoxyribonucleic acid (DNA) was isolated from lymphocytes using routine procedures. Polymerase chain reaction (PCR) amplification of the HT2A/T102C region containing the polymorphic site produced a 342 base-pair (bp) fragment. This was digested with the restriction enzyme HpalLThe uncut product corresponded to the 102T allele. Digested products with 216bp and 126 bp corresponded to the 102C allele.

Statistical analyses

A Chi-square test was used for categorical data and an analysis of variance (ANOVA) for continuous variables. The tests were two-tailed, and P was considered signifycant when P<0.05.

Results

Patients and controls were not different regarding demographic characteristics such as ethnicity and gender. Moreover, no differences were observed regarding genotype frequencies across these groups (Table I).

We further analyzed the association between the genotype and allele frequencies of the T102C polymorphism and total and factor scores of the BPRS. As shown in Table II, no association was found.

The history of a suicide attempt and the suicide attempt characteristics were not associated with genotype, allele frequencies, nor with psychopathological scores (Table III). Furthermore, the demographic characteristics were not statistically different between patients with a suicide attempt history (n=42; [32.5%], 23 were male, 19 were female; 28 were of Caucasian descent, 14 of African descent) and without such a history (n=87; [67.5%], 56 were male, 31 were female; 56 were of Caucasian descent, 31 were of African descent). In both groups the genotype frequencies were in Hardy-Weinberg equilibrium.

Discussion

Two previous meta-analyses had shown an association between the C allele of the T102C polymorphism and schizophrenia.[3,14] However, these meta-analyses yielded relatively slight odds ratios (ORs) of 1.18 and 1.1, respectively. An important question remains as to whether some characteristics present in schizophrenic patients, such as suicidality, rather than schizophrenia itself, could be related to the C allele.

There is much evidence to support this hypothesis. First, suicide is the leading cause of premature death in schizophrenic patients[15,16] and a substantial percentage of patients with schizophrenic patients also attempt suicide, with estimates of lifetime occurrence ranging from 18% to 55 %.[15] Second, a previous study showed an association between the C allele and suicidal thoughts in depressed patients,[17] and we may speculate whether this same finding could be observed in schizophrenia. Third, it is also known that the 5-HT2A receptor is, at least indirectly, involved in suicidal behavior in both depressed and schizophrenic patients.[6] Finally, it has been shown that the atypical antipsychotic clozapine, which acts on 5-HT2A, could have an antisuicidal action.[18,19]

To the best of our knowledge, there has been only one study that specifically analyzed the association between the 102T/C polymorphism of the 5-HTR2A gene and suicidality In schizophrenia, showing no positive correlation.[20] However, some methodological bias must be taken into consideration, because this study was based on a small sample of 65 schizophrenic and 6 schizoaffective patients. Thus, a type II error could not be reliably excluded. Another point of concern refers to suicide assessment. It has been proposed that suicidal behavior shows a continuum between suicide attempt and suicide completion. In other words, severe suicide attempts are biologically closer to suicide completion, and the seriousness of the suicide attempt might explain differences in serotonergic activity[4] Therefore, studying patients with suicidal ideation or suicide plans, as performed in this previous study, can be rather different from a biological point of view than studying patients with suicide attempts or suicide completion.

We were not able to observe any significant association between the T102C polymorphism and suicidal behavior in our sample. We investigated a fairly homogeneous sample of 129 schizophrenic inpatients, as assessed with structured instruments to evaluate diagnosis. Another strength of our study is the fact that suicidal behavior was assessed using a semistructured interview as well as a supplementary interview with at least one close relative, plus a review of medical records. This is rather important, since it has been previously shown that a significant degree of past suicidal behavior was not recorded during routine clinical assessment and, the use of a semistructured screening instrument may improve documentation and detection of lifetime suicidal behavior.[12] Indeed, a phenotypic characterization of suicide attempt, as performed in our study, could be of major interest since some categories of suicidal behavior (ie, more lethal or violent ones) could be more closely associated with a biological marker.[21]

We were also not able to observe an association between the T102C polymorphism and schizophrenia diagnosis. This result is in apparent contradiction with a recent metaanalysis.[3] However, the authors of this analysis showed that in East Asian countries, there was not a significant association with the C allele or CC homozygosity, indicating strong genetic differences and incompatibility between data from European and East Asian populations. They suggest that data from European and Asian samples should not be pooled when evaluating the involvement of this gene in schizophrenia. Interestingly, the frequency of the T allele was much higher in East Asian patients and controls (59.5% and 57.5%, respectively) than in European patients and controls (40% and 43.5%, respectively). In our sample, the frequency was intermediate between those values, since the T allele frequencies in patients and controls were 51% and 48.5%, respectively. In our sample we performed an ethnic stratification, and we did not observe a significantly different frequency in the studied genotypes between those of Caucasian and those of African descent. However as recently shown, race, as determined by physical evaluation, is a poor predictor of genomic African ancestry in Brazil.[22]

In conclusion, as in all casecontrol psychiatric genetic studies, we must be aware of false-positive or false-negative findings due to ethnic stratification and sample size. We attempted to control this by including a detailed demographic analysis, which demonstrated no significant differences between genotype frequencies and ethnicities between patients with a suicide attempt history and patients without such a history Despite this, further studies using a larger number of subjects should be carried out to firmly establish the role of the T102C polymorphism of the 5-HT2A gene in suicidal behavior in schizophrenia.

Table I.

Genotype distribution and demographic characteristics across schizophrenic patients and healthy controls. *: χ2=0.14, P=0.93; †: χ2=1,45, P=023; ‡: χ2=0.86, P=0.77

	Genotypes			Sex distribution		Ethnicity
	TT	TC	CC	Males	Females	Caucasian	African
Schizophrenic patients (n=129)	35 (27%)	62 (48%)	32 (25%)	79 (61 %)	50 (39%)	84 (65%)	45 (35%)
Healthy controls (n=85)	22 (25%)	40 (47%)	23 (27%)*	45 (53%)	40 (49%) †	57 (67%)	28 (33%) ‡

Table II.

BPRS total scores and factor scores across genotypes in schizophrenic patients. BPRS, Brief Psychiatric Rating Scale. *: F=1.06, P=0.4; †: F=11.32, P=0.2; ‡: F=1.26, P=0.3; ║: F=0.5, P=0.88

	TT genotype	TC genotype	CC genotype
BPRS Total score	38.7 ± 7.2	36.8 ± 7.2	39.6 ± 7.5 *
Delusion dimension	10.1 ± 3.8	9.5 ± 4.0	10.9 ± 3.9 †
Anxio-depressive dimension	5.4 ±2.0	5.3 ± 22	5.1 ± 2.6 ‡
Hebephrenic dimension	8,4 ± 3,2	8,b ± 3,0	7.9 ± 3.1 §
Paranoid dimension	5.9 ± 1.9	5.9 ± 1.9	6.3 ± 2.4 ║

Table III.

Distribution of the 5-HT2A receptor gene 102 TC polymorphism in suicidal and nonsuicidal schizophrenic patients. BPRS, Brief Psychiatric Rating Scale. *: χ2=0.48, P=0.79; †: χ2=0.49, P=0.78; ‡: χ2=2.1, P=0.36; §: χ2=0.28, P=0,6; ║: χ2=0.46, P=0.5; ¶: χ2=0.42, P=0.52; #: F=0,54; P=0.6; **: F=0.57; P=0.57; ††: F=0.36; P=0.72

	Suicide attempt history		Lethality		Type of suicide attempt
	No (n=87)	Yes (n=42)	<3 (n=23)	≥3 (n=19)	Nonviolent (n=22)	Violent (n=20)
TT	24 (275%)	11 (262%)	7 (305%)	4 (21%)	4 (18.2%)	7 (35%)
TC	43 (495%)	19 (45.2%)	10(43,4%)	9 (47.4%)	12 (54.5%)	7(35%)
CC	20 (23%)	12 (28.6%)*	6 (26.1%)	6 (316%) †	6 (27.3%)	6 (30%) ‡
T allele	91 (52.25%)	41 (48.8%)	24(52.2%)	17 (44.7%)	20 (45.5%)	21 (52.5%)
C allele	83 (47.75%)	43 (51.2%) §	22 (47.8%)	21 (55.3%) ║	24 (54.5%)	19 (47.5%) ¶
BPRS	38.5 ± 5.6	39.5 ± 5.9 #	38.4 ± 5.7	39.5 ± 6.4 **	38.7 ± 5.8	39.4 ± 6.2 ‡‡