Background

There is consistent evidence pointing the immune system as playing an important role in both etiology and pathophysiology of schizophrenia [1, 2]. A key element of this immune theory is the significant increase in the levels of some Th1 (T helper cell type 1) cytokines such as IL-1, IL-6 or TNF (tumor necrosis factor) in plasma, or even cerebrospinal fluid, found in schizophrenic patients, and abnormal levels of Th2 (T helper cell type 2) cytokines such as IL-10 [3]. It is well documented that some cytokines production is regulated at the transcriptional level, which may indicate that the immune alterations in schizophrenia could have a genetic origin [4].

This occurs with IL-10 in which three gene promoter single nucleotide polymorphisms (SNPs), -1082G > A, -819C > T and -592C > A, define three haplotypes in Caucasians, GCC, ACC, ATA, associated to different IL-10 production rates [5]. Both IL-10 SNPs and haplotypes have been reported to be in linkage disequilibrium with schizophrenia in some populations [3, 6, 7].

On the other hand, there is evidence of sex-specific association of certain genes with many complex diseases including schizophrenia [8] where sex differences have been described for genes such as RELN [9], XBP1 [10]Nogo [11] and IL-10 [6].

Thus, the aims of this study were to verify the linkage of the IL-10 gene promoter SNPs and haplotypes with schizophrenia and to explore a putative sex-specific genetic association of these polymorphisms in a Spanish schizophrenic population.

Methods

Subjects

241 Caucasian (64% males) unrelated Spanish schizophrenic adult patients, mean age (± Standard Deviation-SD-) 45 ± 13 years, were enrolled in the study. DSM-IV diagnosis was provided using a brief structured psychiatric interview, the Spanish version of the Mini International Neuropsychiatric Interview version 4.4 (MINI 4.4) [12].

435 unrelated adult (47% males), mean age ( ± SD) 38 ± 18 years, Caucasian, Spanish healthy volunteers, with no clinical history of chronic illness, were enrolled in the study from 3 different centers.

The study protocol was reviewed and approved by Ethics Committees of all participating centers and conducted according to the tenets of the Declaration of Helsinki. Before enrolment, all participating subjects signed an informed consent form after the study objectives and procedures were fully explained.

DNA extraction and genotyping

Genomic DNA was extracted from 7 ml of peripheral blood samples using both automatic DNA extractors (BioRobot EZ1, QIAGEN, Hilden, Germany; MagNA Pure System, Roche Applied Science) and conventional salting-out methods. DNA concentration and quality of samples were assessed spectrophotometrically (NanoDrop ® ND-1000 Spectrophotometer, Wilmington DE, USA).

Genotyping of -1082G > A SNP was performed through a pharmacogenetic tool, PHARMAchip ® (Progenika Biopharma, Spain) that includes genetic variants involved in therapeutic outcome, which methodology has been described elsewhere [13].

IL-10 -592C > A was genotyped with a TaqMan® Pre-Designed SNP genotyping assay (c__747363_10) according to the TaqMan Allelic Discrimination technology [14] using the ABI Prism 7000 (Applied Biosystems, Foster City, CA, USA).

Since there were two different genotyping methods, both results were further confirmed by genotyping 100 randomly selected samples for the three IL-10 promoter SNPs by direct DNA sequencing, using a BigDye Terminator Cycle Sequencing Kit and an ABI Prism 3130xl DNA sequencer ( Applied Biosystems, Foster City, CA, USA). Therefore, the -819C > T SNP was also assessed in these samples. Primers used for the PCR reaction were 5'GACAACACTACTAAGGCTTCTTTGG3'-forward- and 5'TGTAGGAAGCCAGTCTCTGGA3'-reverse- obtaining a PCR product of 540 bp. Cycling conditions were 95ºC for 10 minutes, followed by 35 cycles of 95ºC for 30s, 59ºC for 30s and 72ºC for 40s, with a final elongation of 10 minutes at 72ºC.

Data analysis

-1082G > A variant results were translated into genotypes by Progenika. Similarly, -592C > A results were processed by the ABI Prism 7000 Allelic Discrimination Analysis software. Since the three IL-10 gene promoter SNPs are in linkage disequilibrium in Caucasians and only three haplotypes have been described (GCC, ACC, ATA), they were constructed according to SNPs at positions -1082 and -592 as previously described by Ortiz et al [15].

Allele, genotype and haplotype (allelic and genotypic) frequencies for both cases and controls were calculated and Hardy-Weinberg equilibrium was estimated. A sex-stratified case-control analysis was performed to find out the possible sex-specific association of IL-10 gene promoter SNPs to schizophrenia.

Statistical analysis

Student T test was performed to assess for possible significant differences in age between cases and controls globally and in each sex group. Pearson's chi squared test was used to compare allele, genotype or haplotype frequencies between schizophrenics and controls in male and female populations. Test for equality of proportions was used to find possible differences in the incidence of ATA between schizophrenic and healthy females. The odds ratio (OR), with its 95% confidence interval (95% CI), was the measure of the strength of association between alleles, genotypes and haplotypes with schizophrenia in the two groups. Statistical analysis was carried using R.2.10.0 software (from R Project for Statistical Computing -http://www.r-project.org-).

Results

241 patients and 435 controls were finally genotyped for SNP -1082G > A. The totality of DNA control samples could not be genotyped for -592C > A. Thus, IL-10 -592C > A variant and IL-10 gene promoter haplotypes could be determined in 241 patients and 381 controls. The results of -1082G > A and -592C > A sequencing yielded a 98% of concordance, since there were two discrepancies between methods for both variants. -1082G > A, -592C > A genotypes and haplotype distributions were in Hardy-Weinberg equilibrium at the 5% significant level.

Table 1 summarizes the results of the -1082G > A and -592C > A allelic and genotypic frequencies in the sex-stratified analysis and the mean age of each population. Mean age between cases and controls resulted significantly different globally (p = 0.000) and stratifying by sex (p = 0.002 in female case-control group and p = 0.000 in male case-control group) Neither allelic nor genotypic frequencies reached the statistical significance in the association in the male population while female schizophrenics had a statistically significant higher incidence of allele -1082A than the controls as well as genotype -1082A/A.

Table 1 Allelic and genotypic frequencies of -1082G > A and -592C > A IL-10 gene promoter variants
Full size table

  • 592C > A allelic and genotypic distribution was similar between schizophrenic and healthy males while, again, there was a statistically significant association of genotype A/A with schizophrenia in the females.

Table 2 shows the analysis of haplotypes stratified by sex. While allelic frequencies were similar in cases and controls in both groups, when comparing genotypic distribution, it differed significantly in schizophrenic females. Significant trends were observed for genotypes ATA/ATA and GCC/ATA. Frequencies of carrying 2, 1 or no copies of ATA in females, and the resultant p-value of the test for equality of proportions, are summarized in table 3 which shows a statistically significant higher incidence of 2 ATA in the schizophrenic females compared to the healthy ones. The ORs of association of 2 ATA versus 1 or no copies of ATA and, also, of 1 ATA versus 0 ATA (table 4) points that carrying 2 copies of ATA was the higher statistically significant risk factor when comparing to 1 or no copies.

Table 2 Allelic and genotypic frequencies of IL-10 gene promoter haplotypes in schizophrenic and healthy males and females
Full size table
Table 3 ATA frequencies in the female subgroup
Full size table
Table 4 ORs of ATA as risk factor for schizophrenia in females
Full size table

Discussion

At the genetic level, many components of the immune system have been linked to schizophrenia, through candidate-gene [3, 6], pathway-based [2] and genome-wide association [1] approaches that have yielded solid evidence of the immune system involvement in schizophrenia. Specifically the recent report by Sun highlights the IL-10 signaling as a candidate pathway in schizophrenia [2].

Since first described by Bochio Chiavetto [3] the linkage between -1082G of the IL-10 gene promoter variant with schizophrenia in Italians, several studies have tried to replicate this association in various populations with conflicting results. While confirmed in a recent study [6], this trend has not been observed in other populations [16] or contradictory findings have been reported [7, 17].

This study addresses the significant association of IL-10 -1082A allele with schizophrenia females. These results contrast to those reported in Italians [3] and Polish [6] case-control studies that found the G allele significantly increased in the schizophrenia group. Regarding the report of Bochio Chiavetto [3], the allelic frequencies of the population selected as the reference are significantly different to data published in Caucasians (TSI or CEU populations of HapMap) suggesting that the association may be due to a biased selection of the control population (most likely an incorrect matching with the cases). In the second report only paranoid schizophrenics were included in the study and significant association was only found in males [6].

Considering IL-10 gene promoter haplotypes, only GCC in Italians [3] or GTA in Turkish [18], both associated with high IL-10 production, have been linked to the disease and although high plasma levels of this cytokine have been reported in some studies associated to schizophrenia, it has not always been replicated [19]. Furthermore, in a recent meta-analysis in which the effect of 10 cytokines plasma levels in schizophrenia was assessed, including IL-10, only significant trends have been observed for IL-1RA, sIL-2R and IL-6, providing evidence of an inflammatory syndrome in schizophrenia [20].

This study also describes for the first time the association of IL-10 gene promoter ATA/ATA low IL-10 producing genotype linked to schizophrenia in Spanish females. Our results indicate that being homozygote for ATA is the main risk factor for schizophrenia in Spanish females compared to ATA heterozygosity or other genotypes not including ATA.

There is evidence of the role of IL-10 in the neurodevelopmental abnormalities found in schizophrenia [21]. IL-10 is an anti-inflammatory cytokine that regulates the inflammatory response, by inhibiting pro-inflammatory cytokine production [22], and it is constitutively expressed during fetal brain development in humans [23]. Meyer et al suggest that the genetically determined differences in IL-10 production could lead to behavioral abnormalities in the adulthood after prenatal immune challenge or innate immune imbalances. They also pointed out that not only an excess of pro-inflammatory cytokines but also an imbalance between both classes of cytokines during development might alter normal brain functions in adult life [21].

Our findings are also supported by a recent report by Sharma et al, which describe a linkage between the ATA haplotype and differential repression of IL-10 production, under infectious conditions, in human trophoblast [24].

On the other hand, there are several evidences of gender differences in schizophrenia regarding incidence, age of onset, disease course, therapeutic response, social and intellectual functioning and brain abnormalities [25]. Additionally, a sex-specific risk of some genes to certain diseases, including schizophrenia [8] with a number of loci involved, has been described [6, 10, 11]. There are various possible mechanisms for sex differences in gene expression, including imprinting or hormonal effects [26]. In the case of IL-10, or genes encoding cytokines in general, estrogen receptors are found in certain immune cells [27] responsible for IL-10 production. Direct estrogen-mediated modulation of this immune cell activity leads to changes in cytokine production [28] and can explain the gender differences of IL-10 as risk factor in females.

Conclusions

Although further exploration of immune system involvement in schizophrenia is needed, our results highlight the previously described hypothesis of an imbalance towards a pro-inflammatory syndrome as the immune abnormality of schizophrenia [20]. Anyway, it should be noted that immune abnormalities are found only in a relatively small subgroup of patients [21]. A better understanding of the involvement of the immune system in schizophrenia would imply the search of immune abnormalities in what has been called endophenotypes, intermediate phenotypes between the clinical entity and susceptibility genes and so, presumably closer to relevant genes [29] in whose sex and ethnicity may be differential factors. Additionally, the results of this study reinforce the need of performing complex gene studies based on multiple cytokine SNPs, including anti- and pro-inflammatory, to clarify the immune system abnormalities direction in the etiology of schizophrenia.