Abstract

We have previously reported a linkage peak on 1q42 in a Finnish schizophrenia sample. In this study we genotyped 28 single nucleotide polymorphisms (SNPs) from 1q42 covering the three candidate genes TRAX, DISC1 and DISC2, using a study sample of 458 Finnish families ascertained for schizophrenia. Two-point and haplotype association analysis revealed a significant region of interest within the DISC1 gene. A common haplotype (HEP3) was observed to be significantly under-transmitted to affected individuals (P=0.0031). HEP3 represents a two SNP haplotype spanning from intron 1 to exon 2 of DISC1. This haplotype also displayed sex differences in transmission distortion, the under-transmission being significant only to affected females (P=0.00024). Three other regions of interest were observed in the TRAX and DISC genes. However, analysis of only those families with complete genotype information specifically highlights the HEP3 haplotype as a true observation. The finding of a common under-transmitted SNP haplotype might imply that this particular allele offers some protection from the development of schizophrenia. Analysis of component-traits of schizophrenia, derived from the Operational Criteria Checklist of Psychotic Illness (OCCPI), displayed association of HEP3 to features of the general phenotype of schizophrenia, including traits representing delusions, hallucinations and negative symptoms. This study provides further evidence for the hypothesis that the DISC1 gene is involved in the aetiology of schizophrenia, and implies a putative sex difference for the effect of the gene. Our findings would also encourage more detailed analyses of the effect of DISC1 on the component-traits of schizophrenia.

INTRODUCTION

Schizophrenia is a severe mental disorder affecting ∼1% of the worldwide population. In Finland, the lifetime prevalence of schizophrenia is 1.2%, however, within a young internal isolate the prevalence is higher, 2.2% (1). Schizophrenia is considered to be partly genetic in aetiology, with family, twin and adoption studies suggesting that there is a considerably high heritability to the disorder, as high as 83% based on the analysis of a Finnish twin cohort (2). Finland is one of the best characterized isolated populations and the frequency of disease alleles reflects the history of multiple bottlenecks (3). This population history and the presence of extensive genealogical records make the Finnish population valuable for studying genetic disorders. To date about 35 monogenic diseases have been identified as enriched in the population, and are often referred to as the Finnish disease heritage (4). The same population processes that led to this enrichment could also result in an increased homogeneity of the disease alleles underlying more complex disorders like schizophrenia (57).

In recent years, linkage studies of schizophrenia have suggested that the long arm of chromosome 1 is a major candidate region for susceptibility loci for this disorder. A number of research groups have reported some evidence of linkage or association at 1q21–24 (810), 1q32.2–41 (1113), or 1q42 (13,14, J. Ekelund et al., unpublished data). Although the data from all the suggested chromosomal loci for schizophrenia are not consistent, the findings for chromosome 1q are amongst the most replicated (15). Initially 1q42 was identified as a potential site for a schizophrenia susceptibility locus in a large Scottish family, containing a wide spectrum of major mental illnesses, where a balanced (1;11)(q42.1;q14.3) translocation was found to co-segregate with schizophrenia and other related psychiatric disorders (14). Two novel genes were found to be directly disrupted by this translocation, and were subsequently named Disrupted in Schizophrenia 1 and 2 (DISC1 and DISC2) (16). When other Scottish families ascertained for schizophrenia and bipolar disorder were used to monitor for association between single nucleotide polymorphisms (SNPs) in the DISC genes and the disorders, no significant association was observed (17). Later, linkage findings within a Finnish schizophrenia nuclear family sample confirmed the 1q42 locus and the DISC genes also as good positional candidates for susceptibility to schizophrenia, when a Zmax of 3.2 was observed for a microsatellite located intragenic of DISC1 (13). This linkage finding has recently been replicated in an independent Finnish nuclear family sample, giving further evidence that this region contains a schizophrenia susceptibility locus. The replication study also identified a three SNP haplotype spanning from exon 9 of DISC1 into intron 9, which was over-transmitted to affected individuals (J. Ekelund et al., unpublished data).

In addition to DISC1 and DISC2, there is a third positional candidate gene within this particular chromosomal region. The Translin-Associated Factor X (TRAX; TSNAX) is located centromeric and in the same orientation as DISC1 and was identified as a potential candidate gene when intergenic splicing was found to form fusion transcripts between TRAX, DISC1 and combinations of four intragenic exons located between these two genes (18). Such an event highlights that mutations affecting TRAX may also affect DISC1. It has also recently been shown that orthologs of DISC1 are highly conserved in genomic structure and in its location close to the TRAX orthologs on the mouse (19) and pufferfish (Fugu rubripes) (20) genomes. This would imply some functional significance for the physical vicinity of the TRAX and DISC1 genes. Furthermore, the genes represent good functional candidates for schizophrenia. Studies of DISC1 have established that it is expressed in neurons and glia, and probably interacts with proteins involved in signal transduction, glutamatergic neurotransmission and neuronal migration (21). The DISC2 gene is not known to encode a protein, but is thought to act through its RNA as a regulator of DISC1 (16). The TRAX protein is known to form a brain enriched complex with Translin, that can bind single-stranded DNA and RNA through which it is involved in protein regulation (22) and, consequentially, in development and function of the nervous system.

Here we have studied the 1q42 locus by analysing SNPs and the corresponding haplotypes across the three candidate genes. We monitored the transmitted and untransmitted alleles of TRAX, DISC1 and DISC2 in over 450 schizophrenia families. In addition to the end state diagnosis of schizophrenia, specific component-traits derived from the Operational Criteria Checklist for Psychotic Illness (OCCPI) were used to test for association to these candidate genes.

RESULTS

We genotyped 28 SNPs spanning the 600 kb region of 1q42 that contains the TRAX, DISC1 and DISC2 genes, in a study sample of 458 Finnish schizophrenia families. This represents a nationwide combined (Com) sample of the young isolate (IS) population (179 families) and the rest of Finland (AF) sample (279 families). The locations of the SNPs with reference to the candidate genes and previous linkage findings are illustrated in Figure 1.

We performed analysis of 1q42, using the Pseudomarker software (23), which combines the power of linkage and association. This showed that one SNP (2879A>G), with a minor allele frequency of 2%, located on DISC1 exon 13 displayed consistent suggestive association with schizophrenia within the Com sample. Within the liability classes (LC, see Methods section) this SNP displayed a P-value of 0.0021 (chi square=9.44, 1 d.f.) for LC 2, but also displayed P-values <0.01 for the other liability classes (Table 1). When the population was split into the IS and AF samples, no association was observed for this SNP. There was no significant difference between the two-point analyses of Pseudomarker and TRANSMIT, except for the SNPs with rare alleles which TRANSMIT ignored.

We performed haplotype analysis for the whole region using the broadest phenotype (LC 4) and the Com sample, in order to include all possible phase information and statistical power of our sample. We used the TRANSMIT program (24) to monitor pairs and triplets of SNPs in sliding windows across the 1q42 region. Only one haplotype was observed to be significantly over-transmitted [P=0.00045, expected transmissions (ET), 203; observed transmissions (OT), 214; frequency in sample (FS), 11.7%; frequency in affected (FA), 12.5%; frequency in unaffected (FU), 11.1%] to affected individuals, a three SNP haplotype 10 kb in length spanning from exon 9 of DISC1 into intron 9 (HEP1). This haplotype carries the T, C and G alleles of the SNPs 1872C>T, rs3890280, and rs1000731, respectively, and is the haplotype identified in the replication of the original linkage finding (J. Ekelund et al., unpublished data). Further, three other haplotypes displayed significant under-transmission to affected individuals (Fig. 2). The first significantly under-transmitted haplotype (P<0.00001; ET, 89; OT, 64; FS, 6.0%; FA, 5.3%; FU, 6.6%) consists of three SNPs spanning 10 kb of intron 4 of the TRAX gene (HEP2). The second haplotype (P=0.0031; ET, 145; OT, 125; FS, 8.8%; FA, 8.5%; FU, 9.0%) contains two SNPs that span 62 kb from intron 1 to exon 2 of DISC1 (HEP3). We also observed that the HEP3 haplotype displays significant evidence for under-transmission only to affected females (P=0.00024; ET, 60; OT, 44; FS, 7.9%; FA, 7.7%; FU, 7.9%) compared with under-transmission to affected males (P=0.38; ET, 85; OT, 81; FS, 9.7%; FA, 9.1%; FU, 10.3%). The third under-transmitted haplotype (P=0.00057; ET, 1604; OT, 1595; FS, 98.0%; FA, 97.9%; FU, 98.1%) is the result of two SNPs in exon 13 of DISC1 located 761 bp apart (HEP4). The TRANSMIT program predicted that these SNPs would only produce two haplotypes. Consequently, the alternative haplotype was over-transmitted with a significant P-value of 0.00057, but was ignored by the TRANSMIT program due to it having a sample frequency below 3%. Looking at the other two SNP haplotypes constructed using SNPs in DISC1, exon 13 showed that there are a number of haplotypes incorporating rare SNPs that are significantly over-transmitted (P<0.005), but are below 3% in frequency for this sample.

We performed two validation tests for these findings using the three common haplotypes (HEP1, HEP2 and HEP3) under LC 4, to ensure that these findings are true, and do not represent linkage or errors caused by estimation of the missing data. Initially we analysed the transmission distortion of these haplotypes in only those families where both parents were genotyped (n=147 nuclear families). Only HEP3 continued to display the statistical significance already observed (P=0.00084; ET, 42; OT, 29; FS, 7.0%; FA, 5.5%; FU, 8.2%), but although evidence for sex differences in transmission distortion remained, this difference was not equally impressive as in the complete study sample of 458 families (female, P=0.0020—ET, 19; OT, 12; FS, 5.9%; FA, 4.6%; FU, 6.9%; male, P=0.018—ET, 23; OT, 17; FS, 8.0%; FA, 6.2%; FU, 9.4%). We also performed tests randomly taking only one affected offspring per family. This was carried out 20 times for each haplotype taking the median value of these as the average. This technique was shown to be sufficient in the identification of an associating haplotype within the Dysbindin gene and schizophrenia (25). Only HEP1 displayed no significance (P=0.17; ET, 103; OT, 110; FS, 11.7%; FA, 12.5%; FU, 11.1%), suggesting that this haplotype may be an artefact of linkage. Both HEP2 (P=0.0040; ET, 51; OT, 40; FS, 6.0%; FA, 5.3%; FU, 6.6%) and HEP3 (P=0.035; ET, 77; OT, 66; FS, 8.8%; FA, 8.5%; FU, 9.0%) still showed some association with the LC 4 phenotype, with HEP3 still displaying sex-dependant differences in transmission (female, P= 0.0065—ET, 45; OT, 34; FS, 7.9%; FA, 7.7%; FU, 7.9%; male, P=0.38—ET, 85; OT, 81; FS, 9.7%; FA, 9.1%; FU, 10.3%). Performing such analysis greatly reduces the power available within our large study sample due to decreasing the number of affected individuals being analysed. In the test using randomly selected trios the power is reduced due to the drop from 871 affected individuals being analysed to 498, and in the analysis of only the complete families this reduction in power is indicated by the fact that the number of affected individuals being tested drops further to 266 individuals. Although both tests serve to validate the original finding as a truly observed trend, the fact that they might not have had sufficient power to detect association with the less frequent haplotype, HEP2, should be taken into consideration.

Incorporating additional neighbouring SNPs into the haplotypes failed to reveal any significant extended segregating haplotypes. The two common under-transmitted haplotypes (HEP2 and HEP3) were tested to see if they are independent of each other. This was done by studying the frequencies of the haplotypes constructed by TRANSMIT from the five SNPs involved. Of the resulting haplotypes the one corresponding to HEP2 and HEP3 was predicted to occur in this sample with a frequency of about 1%, whereas HEP2 has a frequency of 6.0% and HEP3 a frequency of 8.8%. This would suggest that these haplotypes could represent two independent DNA regions associating with schizophrenia.

We wanted to test to some extent the role of the three putatively associated haplotypes further by analysing trait components as well as the end diagnosis, using the remaining three liability classes, the four factor groups, and the 24 OCCPI items in all three sample populations. HEP1 was tested for over-transmission to both sexes combined and was found not to show further associations to any other phenotype or sub-sample (not shown). HEP2 was tested for under-transmission to both sexes combined, whereas HEP3 was only tested for under-transmission to affected females (Table 2). This showed that haplotypes mainly associate to component-traits representing delusions, hallucinations and negative symptoms. These three haplotypes were also monitored for transmissions to those offspring of unknown phenotype but not currently affected within liability class 4, with no significant over or under-transmissions being observed (not shown).

DISCUSSION

Here we have monitored the transmission of SNPs on 1q42 and their corresponding haplotypes within schizophrenia families, and have identified a restricted genomic region that shows significant association with schizophrenia and its component-traits, with three other regions displaying putative associations.

A common under-transmitted haplotype was identified as independently associating to schizophrenia. This haplotype (HEP3) is located within DISC1, highlighting the gene as contributing to the aetiology of schizophrenia. The HEP3 haplotype was significantly associated only within the AF population, with no suggestive results being seen in the IS sample. This is consistent with our previous linkage findings that have found linkage with the AF population to 1q42, while the IS population displays linkage to 1q32–41 (11,13) using the liability class-based phenotypes. Interestingly, when the OCCPI factors and items were used as alternative phenotypes it was found that HEP3 displays consistently more significant association to certain factors and their component-traits, delusions, hallucinations and negative symptoms, compared with the other component-traits of manic and depressive symptoms, although for a number of these component traits there is a large reduction in power due to the lower frequency of these traits in our sample. It is well characterized that greater negative symptoms in schizophrenics correlates with poorer cognitive ability (26), suggesting that this haplotype should be investigated with respect to the finding that 1q41 displays linkage and association to spatial working memory (27). Association findings for these component-traits were also restricted to the AF population.

HEP3 also displayed sex differences, the haplotype showing significantly distorted transmission only to affected females. This might imply that this genetic region could contribute to the sex differences found in the schizophrenia phenotype (28). There are a number of possible affects on known mechanisms for sex differences in gene expression, such as imprinting (29) and hormonal effects (30). Analysis of only those families with complete genotype information and trios strengthens the evidence that HEP3 and its sex differences is a genuine finding.

Initially it is necessary to determine if HEP3 is under-represented in the entire population regardless of affection status, as this would imply that the haplotype affects the general fitness. By monitoring the transmissions of the haplotype to those individuals currently unaffected under liability class 4, we could not show deviation from the expected haplotype frequency. It could be expected that the under-transmitted haplotype would be over-transmitted to people unaffected with schizophrenia. This is not observed in our study, but this test should ideally be performed in a sample with no ascertainment bias. A number of other interesting hypotheses can be proposed to explain the observed under-transmission of DISC1 alleles. Under-transmitted alleles could offer protection from the disorder, either directly by being beneficial to that gene's function, or indirectly by counteracting a susceptibility causing alteration in another gene. Based on the ‘protective locus’ hypothesis the under-transmitted haplotype would have arisen later in the Finnish population history, and become enriched during the expansion of the population, putatively as a result of even a modest positive selection.

Three other restricted regions were observed to associate with schizophrenia. HEP2, a second under-transmitted common haplotype located within intron 4 of the TRAX gene, was identified. HEP2 displays levels of significance similar to that of the HEP3 finding, and in the same pattern for all tests except one. When HEP2 was tested in only those families with complete genotype information no significant association was observed. Despite this the other association findings with HEP2 indicate that it is still worth investigating further. Within the six SNPs located in DISC1 exon 13, associations were observed for a single SNP (2879A>G) and a number of two-SNP haplotypes, but as all associations seen in this region involve rare polymorphisms (sample frequency <3%), conclusions should be drawn with caution. In the region of DISC1 intron 9 a three-SNP haplotype was observed to be over-transmitted (HEP1). This haplotype represents the haplotype observed in the replication of the original linkage finding on 1q42 in the Finnish schizophrenia sample (Ekelund et al., unpublished data). When this haplotype was analysed within the separate populations, and using the component-traits, no other associations were observed. HEP1 also failed to show association when testing for association using complete families or randomly selected trios, suggesting that this finding may be due to the previously observed linkage findings in this localized region.

The region indicated by the HEP3 haplotype highlights some interesting genomic features that warrant further investigation. The haplotype not only spans exon 2 of the DISC1 gene and the intragenic exon D, but also covers six regions of between 50 and 300 bp that are conserved between the human and mouse genomes under stringent conditions (UCSC genome browser, tight mouse annotation track, November 2002 assembly). Non-coding regions of such evolutionary conservation most probably have some functional significance since they have survived the 80 million years separating mouse and man in evolutionary history, therefore these six regions might represent potential regulatory elements.

These findings on DISC1, and putatively in TRAX, in addition to the linkage results for the 1q42 region, strengthen the evidence that these genes would play a role in the aetiology of schizophrenia. These findings primarily highlight the exon 2 region of DISC1 and shows that it might contribute to the sex differences observed in the schizophrenia phenotype. This finding also highlights certain component-traits of schizophrenia such as delusions, hallucinations and negative symptoms, which should be investigated further with regard to the 1q42 locus.

MATERIALS AND METHODS

Study sample

The study sample is an extension of the samples previously used to perform linkage analysis for schizophrenia in Finland (11,13,3133). The full details of the sample identification and collection have been addressed in previous papers (31,34). The research was approved by the Ministry of Social Affairs and Health (Finland) and the appropriate institutional review boards, and informed consent was obtained from all subjects.

The sample now totals 458 families consisting of 498 nuclear families that contain 2756 individuals, of which 2059 have been genotyped (Table 3). Of these genotyped individuals, 931 are classified as affected using increasingly inclusive liability classes (LC) derived from the Diagnostic and Statistical Manual of Mental Disorders, forth edition (DSM-IV) (35). LC 1 constituted schizophrenia only, LC 2 added those individuals affected with schizoaffective disorder, LC 3 added individuals with schizophrenia spectrum disorder (36), and LC 4 added individuals with bipolar disorder or major depressive disorder (Fig. 3).

The entire sample described above (Com) could be divided into two sub-samples that are defined by the geographical origin of the family. The first sub-sample contains the families obtained from the young internal isolate population (IS) used in the genome wide scan that showed linkage on chromosome 1q32–41 (11). This sample now consists of 179 families, made up of 1137 individuals. The second sub-sample contains families obtained from the rest of Finland (AF), exclusive of the internal isolate, that displayed linkage on chromosome 1q42 (13). This sample now consists of 279 families, made up of 1619 individuals. The numbers of affected offspring in each liability class for each of the samples is shown in brackets in Table 1.

Diagnostic assessment

The diagnostic assessment used for the extension of the sample population remained the same, and has previously been described in detail in other articles (11,13,3133). The process is based on the analysis of patient records for those individuals with a register-based diagnosis of schizophrenia between 1969 and 1998. In addition to the consensus diagnosis made according to the DSM-IV criteria, one reviewer completed the Operational Criteria Checklist for Psychotic Illness (OCCPI). The OCCPI checklist consists of 90 items of psychopathology, pre-morbid functioning, and personal history (37,38). Factor analysis has been performed for 30 of the 90 OCCPI items from 190 patients with schizophrenia from the IS population and 466 affected sib-pairs from the AF sample. It was found that 24 of these 30 items segregate into four factors: factor 1, ‘delusions and hallucinations’; factor 2, ‘manic’; factor 3, ‘negative’; and factor 4, ‘depressive’ (39) (Fig. 3). These factor structures were used to determine qualitative trait phenotypes for use in our study, where at least one of the items loading >0.5 had to be present in the individual to count them as positive for that factor. If any of the 24 individual OCCPI items had more than two possibilities for classification, then these were also converted into qualitative traits by taking any score >0 as meaning the individual is positive for this item.

Laboratory methods

The SNPs were either identified from public databases, from contig alignments, from EST alignments or by sequencing. Potential SNPs not located by sequencing were verified in 12 Finnish controls. The resulting 28 SNPs were genotyped in the entire sample using TaqMan (Applied Biosystems), Mini-sequencing (40) or by using array based genotyping (41). The SNPs that are not located in the dbSNP database are located on GenBank sequence AF222980.

Statistical methods

The genotypes of the SNPs were corrected for Mendelian errors using PedCheck (42). Two-point analysis was then performed using the program Pseudomarker, which performs joint linkage and linkage disequilibrium (LD) analysis on a mixture of pedigrees and singletons. Pseudomarker is able to combine the power of linkage analysis with that of association, and can test for LD in general pedigrees conditional on linkage. The latter is used when it is known that the sample already displays linkage to a particular region being analysed for association (23). Once two-point analysis had been performed on the SNPs the analysis of haplotypes was performed, using the TRANSMIT program. This software is able to test for transmission of a haplotype even when phase is unknown and when parental genotypes are not completely known. The TRANSMIT program is also able to compensate for the presence of linkage when using family data by the calculation of a robust variance estimate (24). Haplotypes below a sample frequency of 3% were aggregated and counted as one haplotype when calculating the global P-value, but were ignored as being too rare as individual haplotypes. TRANSMIT performed 100 000 bootstrap tests for all analyses, from which it derived the empirical P-values.

Pseudomarker analysis was performed for all SNPs in all LC phenotypes using all three samples. However, TRANSMIT analysis was performed originally in the Com sample using only LC 4 to locate any significant haplotypes (P<0.05), which were then tested in all of the phenotypes with all samples. In this study we have analysed multiple tightly linked SNPs and their haplotypes, and used multiple phenotypic models, many of which were derived from other phenotypes being analysed. This means that calculating a simple Bonferroni correction for multiple testing might result in an overly conservative test, as neither the genotypes (43) nor the phenotypes (33) (Fig. 3) are completely independent of each other. Therefore, we report only the uncorrected P-values provided by Pseudomarker and TRANSMIT.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the contributions of Ms M. Marttila, M. Levander and P. Ellonen for assistance in the laboratory, Drs H. Juvonen, M. Muhonen, J. Suokas, K. Suominen and J. Suvisaari for their diagnostic work with this sample and Assistant Professor J.D. Terwilliger for assistance with statistics. We would also like to thank T. Hiekkalinna for assistance with novel software, and P. Haimi and Ms M. Schreck for their work with data managing. This work was partly funded by Wyeth Pharmaceuticals Inc., Millennium Pharmaceuticals Inc., and Academy of Finland, Center of Excellence in Disease Genetics and research grants for M.K. and L.P. Dr Peltonen is the Gordon and Virginia MacDonald Distinguished Chair in Human Genetics at UCLA.

*

To whom correspondence should be addressed at: Biomedicum/KTL, MLO, PL 104, 00251, Helsinki, Finland. Tel: +358 94744-8393; Fax: +358 947448480; Email: leena.peltonen@ktl.fi

Figure 1. Schematic diagram of the 1q42 region showing exonic structure of TRAX (▪) and the DISC genes (▪), as well as the location of the intergenic exons (□). The analysed SNPs are also shown with respect to the exonic structure of these genes. Bold markers above the genes indicate the previous linkage findings. Bold markers below the genes indicate SNPs that are involved in the association findings reported here, with their respective haplotypes (HEP1-4) indicated. The SNPs that are not located in the dbSNP database are located on GenBank sequence AF222980.

Figure 1. Schematic diagram of the 1q42 region showing exonic structure of TRAX (▪) and the DISC genes (▪), as well as the location of the intergenic exons (□). The analysed SNPs are also shown with respect to the exonic structure of these genes. Bold markers above the genes indicate the previous linkage findings. Bold markers below the genes indicate SNPs that are involved in the association findings reported here, with their respective haplotypes (HEP1-4) indicated. The SNPs that are not located in the dbSNP database are located on GenBank sequence AF222980.

Figure 2. TRANSMIT results with sliding windows of two SNP haplotypes (above) or three SNP haplotypes (below), using the entire sample and liability class 4. Below each significant under-transmitted haplotype (P<0.05) is the P-value for that individual haplotype in the entire sample and by sex, the haplotype frequency, and the global P-value derived by evaluation of the transmissions of all haplotypes at once. P-values less than 0.001 are highlighted with bold text, and P-values of 0 are below 0.00001. When a single haplotype is not significant (NS) NA (not applicable) replaces the individual haplotype P-value. Intervals are reported here in base pairs (illustration style adapted from 25).

Figure 2. TRANSMIT results with sliding windows of two SNP haplotypes (above) or three SNP haplotypes (below), using the entire sample and liability class 4. Below each significant under-transmitted haplotype (P<0.05) is the P-value for that individual haplotype in the entire sample and by sex, the haplotype frequency, and the global P-value derived by evaluation of the transmissions of all haplotypes at once. P-values less than 0.001 are highlighted with bold text, and P-values of 0 are below 0.00001. When a single haplotype is not significant (NS) NA (not applicable) replaces the individual haplotype P-value. Intervals are reported here in base pairs (illustration style adapted from 25).

Figure 3. Schematic illustration of the relationship between the increasingly inclusive liability classes and the factor analysis phenotypes. Numbers are the amount of offspring displaying the phenotype. The illustration shows that, although each phenotype is distinct, it is not independent of the other phenotypes. This is observed by the numbers of offspring who display all four factor phenotypes. It also shows that these factor phenotypes are in part dependent on the end state diagnosis, highlighted by the large number of offspring having both factors 1 and 3.

Figure 3. Schematic illustration of the relationship between the increasingly inclusive liability classes and the factor analysis phenotypes. Numbers are the amount of offspring displaying the phenotype. The illustration shows that, although each phenotype is distinct, it is not independent of the other phenotypes. This is observed by the numbers of offspring who display all four factor phenotypes. It also shows that these factor phenotypes are in part dependent on the end state diagnosis, highlighted by the large number of offspring having both factors 1 and 3.

Table 1.

Results from pseudomarker LD analysis of SNP 2879A>G using liability class phenotypes, for the combined sample and the two sub-samples

Phenotype Pseudomarker dominant LD for 2879A>G 
 Coma AFa IS 
Liability classes    
LC 1 (schizophrenia) 0.0077 (607) 0.058 (439) 0.29 (168) 
LC 2 (adds schizoaffective disorder) 0.0021 (+111) 0.17 (+88) 0.13 (+33) 
LC 3 (adds schizophrenia spectrum diagnosis) 0.0035 (+82) 0.20 (+67) 0.21 (+15) 
LC 4 (adds bipolar or major depressive disorder) 0.0075 (+61) 0.25 (+33) 0.86 (+28) 
Phenotype Pseudomarker dominant LD for 2879A>G 
 Coma AFa IS 
Liability classes    
LC 1 (schizophrenia) 0.0077 (607) 0.058 (439) 0.29 (168) 
LC 2 (adds schizoaffective disorder) 0.0021 (+111) 0.17 (+88) 0.13 (+33) 
LC 3 (adds schizophrenia spectrum diagnosis) 0.0035 (+82) 0.20 (+67) 0.21 (+15) 
LC 4 (adds bipolar or major depressive disorder) 0.0075 (+61) 0.25 (+33) 0.86 (+28) 

Numbers in brackets are the numbers of offspring affected with each increasingly inclusive liability class for each population.

aDenotes that LD conditioned for the presence of linkage was used as linkage was previously reported to this region for these samples.

Table 2.

TRANSMIT analysis for the two under-transmitted haplotypes within the combined sample and the two sub-samples using all phenotypes

Phenotype Number of affected offspring TRANSMITHEP2a HEP3b 
 Com AF IS Com AF IS Com AF IS 
Liability classes          
LC 1 (schizophrenia) 607 439 168 0.00089 0.0018 0.089 0.00032 0.000030 0.90 
LC 2 (adds schizoaffective disorder) +121 +88 +33 0.000070 0.00023 0.075 0.000020 0 0.62 
LC 3 (adds schizophrenia spectrum diagnosis) +82 +67 +15 0 0.000010 0.039 0.00037 0.00022 0.57 
LC 4 (adds bipolar or major depressive disorder) +61 +33 +28 0 0 0.018 0.00024 0.00064 0.10 
Factor analysis          
Factor 1 (delusions and hallucinations) 700 481 219 0.000010 0.000090 0.076 0.00030 0.00060 0.18 
Factor 2 (manic) 452 315 137 0.000010 0.00048 0.13 0.00056 0.0028 0.078 
Factor 3 (negative) 677 466 211 0 0.000030 0.029 0.0018 0.0025 0.35 
Factor 4 (depressive) 589 387 202 0.000010 0.000070 0.027 0.00087 0.0012 0.27 
Factor 1 delusions and hallucinations          
Persecutory delusions 584 407 177 0.00013 0.00024 0.14 0 0.000080 0.15 
Well organized delusions 160 74 86 0.94 0.78 0.65 0.15 0.0090 0.78 
Delusions of influence 469 305 164 0.00064 0.00038 0.27 0 0 0.22 
Widespread delusions 405 295 110 0.012 0.012 0.26 0.000030 0.000060 0.036 
Delusions and hallucinations last for one week 606 423 183 0.000050 0.00010 0.045 0.00030 0.00036 0.52 
Persecutory/jealous delusions and hallucinations 471 316 155 0.0015 0.00057 0.26 0.000040 0.00038 0.061 
Thought insertion 76 46 30 0.97 0.63 0.82 0.045 0.018 0.25 
Abusive/accusatory/persecutory voices 414 302 112 0.000070 0.00036 0.013 0.018 0.019 0.45 
Factor 2 manic          
Pressured speech 159 109 50 0.043 0.070 0.33 0.0020 0.073 0.057 
Thoughts racing 145 99 46 0.19 0.025 0.64 0.000020 0.0031 0.090 
Elevated mood 125 81 44 0.038 0.023 0.46 0.043 0.056 0.48 
Irritable mood 377 266 111 0 0.00011 0.024 0.000090 0.000070 0.21 
Increased self esteem 124 77 47 0.015 0.099 0.42 0.059 0.043 0.75 
Grandiose delusions 172 112 60 0.067 0.19 0.18 0.10 0.060 0.99 
Factor 3 negative          
Catatonia 100 66 34 0.0015 0.15 0.11 0.49 0.49 0.88 
Speech difficult to understand 464 281 183 0.00010 0.00036 0.04 0.00075 0.00027 0.36 
Positive formal thought disorder 443 310 133 0.00031 0.00080 0.28 0.000030 0.000040 0.079 
Negative formal thought disorder 413 306 107 0.0017 0.0010 0.30 0.048 0.041 0.81 
Inappropriate affect 553 399 154 0.00019 0.000090 0.36 0.00025 0.00031 0.32 
Factor 4 depressive          
Slowed activity 362 230 132 0.00029 0.00075 0.082 0.0059 0.0014 0.67 
Loss of energy/tiredness 376 249 127 0.000010 0.051 0.043 0.0011 0.0026 0.10 
Loss of pleasure 274 146 128 0.015 0.72 0.36 0.0052 0.0032 0.29 
Excessive self reproach 161 108 53 0.37 0.52 0.89 0.042 0.078 0.29 
Suicidal ideation 382 262 120 0.00019 0.0011 0.23 0.0058 0.0073 0.35 
Phenotype Number of affected offspring TRANSMITHEP2a HEP3b 
 Com AF IS Com AF IS Com AF IS 
Liability classes          
LC 1 (schizophrenia) 607 439 168 0.00089 0.0018 0.089 0.00032 0.000030 0.90 
LC 2 (adds schizoaffective disorder) +121 +88 +33 0.000070 0.00023 0.075 0.000020 0 0.62 
LC 3 (adds schizophrenia spectrum diagnosis) +82 +67 +15 0 0.000010 0.039 0.00037 0.00022 0.57 
LC 4 (adds bipolar or major depressive disorder) +61 +33 +28 0 0 0.018 0.00024 0.00064 0.10 
Factor analysis          
Factor 1 (delusions and hallucinations) 700 481 219 0.000010 0.000090 0.076 0.00030 0.00060 0.18 
Factor 2 (manic) 452 315 137 0.000010 0.00048 0.13 0.00056 0.0028 0.078 
Factor 3 (negative) 677 466 211 0 0.000030 0.029 0.0018 0.0025 0.35 
Factor 4 (depressive) 589 387 202 0.000010 0.000070 0.027 0.00087 0.0012 0.27 
Factor 1 delusions and hallucinations          
Persecutory delusions 584 407 177 0.00013 0.00024 0.14 0 0.000080 0.15 
Well organized delusions 160 74 86 0.94 0.78 0.65 0.15 0.0090 0.78 
Delusions of influence 469 305 164 0.00064 0.00038 0.27 0 0 0.22 
Widespread delusions 405 295 110 0.012 0.012 0.26 0.000030 0.000060 0.036 
Delusions and hallucinations last for one week 606 423 183 0.000050 0.00010 0.045 0.00030 0.00036 0.52 
Persecutory/jealous delusions and hallucinations 471 316 155 0.0015 0.00057 0.26 0.000040 0.00038 0.061 
Thought insertion 76 46 30 0.97 0.63 0.82 0.045 0.018 0.25 
Abusive/accusatory/persecutory voices 414 302 112 0.000070 0.00036 0.013 0.018 0.019 0.45 
Factor 2 manic          
Pressured speech 159 109 50 0.043 0.070 0.33 0.0020 0.073 0.057 
Thoughts racing 145 99 46 0.19 0.025 0.64 0.000020 0.0031 0.090 
Elevated mood 125 81 44 0.038 0.023 0.46 0.043 0.056 0.48 
Irritable mood 377 266 111 0 0.00011 0.024 0.000090 0.000070 0.21 
Increased self esteem 124 77 47 0.015 0.099 0.42 0.059 0.043 0.75 
Grandiose delusions 172 112 60 0.067 0.19 0.18 0.10 0.060 0.99 
Factor 3 negative          
Catatonia 100 66 34 0.0015 0.15 0.11 0.49 0.49 0.88 
Speech difficult to understand 464 281 183 0.00010 0.00036 0.04 0.00075 0.00027 0.36 
Positive formal thought disorder 443 310 133 0.00031 0.00080 0.28 0.000030 0.000040 0.079 
Negative formal thought disorder 413 306 107 0.0017 0.0010 0.30 0.048 0.041 0.81 
Inappropriate affect 553 399 154 0.00019 0.000090 0.36 0.00025 0.00031 0.32 
Factor 4 depressive          
Slowed activity 362 230 132 0.00029 0.00075 0.082 0.0059 0.0014 0.67 
Loss of energy/tiredness 376 249 127 0.000010 0.051 0.043 0.0011 0.0026 0.10 
Loss of pleasure 274 146 128 0.015 0.72 0.36 0.0052 0.0032 0.29 
Excessive self reproach 161 108 53 0.37 0.52 0.89 0.042 0.078 0.29 
Suicidal ideation 382 262 120 0.00019 0.0011 0.23 0.0058 0.0073 0.35 

Bold results are P-values <0.001. P-values of 0 are values below 0.00001.

aHEP2 (CAC haplotype for rs1615344, rs1615409 and rs766288) was tested using both sexes combined.

bHEP3 (TA haplotype for rs751229 and rs3738401) was tested using females only.

Table 3.

Division of the 2059 genotyped individuals according to affection status and family structure

 Number of genotyped parents Number of genotyped offspring 
Affected  60 871 
Unaffected 420 708 
 Number of genotyped parents Number of genotyped offspring 
Affected  60 871 
Unaffected 420 708 

References

1
Hovatta, I., Terwilliger, J.D., Lichtermann, D., Mäkikyrö, T., Suvisaari, J., Peltonen, L. and Lönnqvist, J. (
1997
) Schizophrenia in the genetic isolate of Finland.
Am. J. Med. Genet.
 ,
74
,
353
–360.
2
Cannon, T.D., Kaprio, J., Lönnqvist, J., Huttunen, M. and Koskenvuo, M. (
1998
) The genetic epidemiology of schizophrenia in a Finnish twin cohort.
Arch. Gen. Psychiat.
 ,
55
,
67
–74.
3
Pastinen, T., Perola, M., Ignatius, J., Sabatti, C., Tainola, P., Levander, M., Syvanen, A.C. and Peltonen, L. (
2001
) Dissecting a population genome for targeted screening of disease mutations.
Hum. Mol. Genet.
 ,
10
,
2961
–2972.
4
Norio, R., Nevanlinna, H.R. and Perheentupa, J. (
1973
) Hereditary diseases in Finland; rare flora in rare soul.
Ann. Clin. Res.
 ,
5
,
109
–141.
5
Peltonen, L., Jalanko, A. and Varilo, T. (
1999
) Molecular genetics of the Finnish disease heritage.
Hum. Mol. Genet.
 ,
8
,
1913
–1923.
6
Peltonen, L., Palotie, A. and Lange, K. (
2000
) Use of population isolates for mapping complex traits.
Nat. Rev. Genet.
 ,
1
,
182
–190.
7
Boehnke, M. (
2000
) A look at linkage disequilibrium.
Nat. Genet.
 ,
25
,
246
–247.
8
Brzustowicz, L.M., Hodgkinson, K.A., Chow, E.W.C., Honer, W.G. and Bassett, A.S. (
2000
) Location of a major susceptibility locus for familial schizophrenia on chromosome 1q21–q22.
Science
 ,
288
,
678
–682.
9
Gurling, H.M., Kalsi, G., Brynjolfson, J., Sigmundsson, T., Sherrington, R., Mankoo, B.S., Read, T., Murphy, P., Blaveri, E., McQuillin, A. et al. (
2001
) Genomewide genetic linkage analysis confirms the presence of susceptibility loci for schizophrenia, on chromosomes 1q32.2, 5q33.2, and 8p21–22 and provides support for linkage to schizophrenia, on chromosomes 11q23.3–24 and 20q12.1–11.23.
Am. J. Hum. Genet.
 ,
68
,
661
–673.
10
Chowdari, K.V., Mirnics, K., Semwai, P., Wood, J., Lawrence, E., Bhatia, T., Deshpande, S.N., Thelma, B.K., Ferrell, R.E., Middleton, F.A. et al. (
2002
) Association and linkage analyses of RGS4 polymorphisms in schizophrenia.
Hum. Mol. Genet.
 ,
11
,
1373
–1380.
11
Hovatta, I., Varilo, T., Suvisaari, J., Terwilliger, J.D., Ollikainen, V., Arajärvi, R., Juvonen, H., Kokko-Sahin, M.-L., Väisänen, L., Mannila, H. et al. (
1999
) A genomewide screen for schizophrenia genes in an isolated Finnish subpopulation, suggesting multiple susceptibility loci.
Am. J. Hum. Genet.
 ,
65
,
1114
–1124.
12
Bulayeva, K., Devlin, B., Bacanu, S.-A., Bennett, P., Roeder, K. and Byerley, W. (
2000
) Linkage analysis of schizophrenia in isolated Daghestanian kindreds. The American Society of Human Genetics.
Am. J. Hum. Genet.
 ,
67
,
309
.
13
Ekelund, J., Hovatta, I., Parker, A., Paunio, T., Varilo, T., Martin, R., Suhonen, J., Ellonen, P., Chan, G., Sinsheimer, J.S. et al. (
2001
) Chromosome 1 loci in Finnish schizophrenia families.
Hum. Mol. Genet.
 ,
10
,
1611
–1617.
14
St Clair, D., Blackwood, D., Muir, W., Carothers, A., Walker, M., Spowart, G., Gosden, C. and Evans, H.J. (
1990
) Association within a family of a balanced autosomal translocation with major mental illness.
Lancet.
 ,
336
,
13
–16.
15
Waterworth, D.M., Bassett, A.S. and Brzustowicz, L.M. (
2002
) Recent advances in the genetics of schizophrenia.
Cell. Mol. Life. Sci.
 ,
59
,
331
–348.
16
Millar, J.K., Wilson-Annan, J.C., Anderson, S., Christie, S., Taylor, M.S., Semple, C.A.M., Devon, R.S., St Clair, D.M., Muir, W.J., Blackwood, D.H.R. et al. (
2000
) Disruption of two novel genes by a translocation co-segregating with schizophrenia.
Hum. Mol. Genet.
 ,
9
,
1415
–1423.
17
Devon, R.S., Anderson, S., Teague, P.W., Burgess, P., Kipari, T.M.J., Semple, C.A.M., Millar, J.K., Muir, W.J., Murray, V., Pelosi, A.J. et al. (
2001
) Identification of polymorphisms within Disrupted In Schizophrenia 1 and Disrupted In Schizophrenia 2, and an investigation of their association with schizophrenia and bipolar affective disorder.
Psychiatr. Genet.
 ,
11
,
71
–78.
18
Millar, J.K., Christie, S., Semple, C.A.M. and Porteous, D.J. (
2000
) Chromosomal location and genomic structure of the human translin-associated factor X gene (TRAX; TSNAX) revealed by intergenic splicing to DISC1, a gene disrupted by a translocation segregating with schizophrenia.
Genomics
 ,
67
,
69
–77.
19
Ma, L., Liu, Y., Ky, B., Shughrue, P.J., Austin, C.P. and Morris, J.A. (
2002
) Cloning and characterization of Disc1, the mouse ortholog of DISC1 (Disrupted-in-Schizophrenia 1).
Genomics
 ,
80
,
662
–672.
20
Taylor, M.S., Devon, R.S., Millar, J.K. and Porteous, D.J. (
2003
) Evolutionary constraints on the Disrupted in Schizophrenia locus.
Genomics
 ,
81
,
67
–77.
21
Millar, J.K., James, R., Christie, S., Taylor, M.S., Devon, R.S., Hogg, G., Muir, W.J., Blackwood, D.H.R. and Porteous, D.J. (
2002
) Functional characterisation of DISC1, a candidate susceptibility gene for major psychiatric illness. Xth World Congress of Psychiatric Genetics.
Am. J. Med. Genet.
 ,
114
,
748
.
22
Finkenstadt, P.M., Kang, W.-S., Jeon, M., Taira, E., Tang, W. and Baraban, J.M. (
2000
) Somatodendritic localization of translin, a component of the translin/TRAX RNA binding complex.
J. Neurochem.
 ,
75
,
1754
–1762.
23
Göring, H.H.H. and Terwilliger, J.D. (
2000
) Linkage analysis in the presence of errors IV: joint pseudomarker analysis of linkage and/or linkage disequilibrium on a mixture of pedigrees and singletons when the mode of inheritance cannot be accurately specified.
Am. J. Hum. Genet.
 ,
66
,
1310
–1327.
24
Clayton, D. (
1999
) A generalization of the transmission/disequilibrium test for uncertain-haplotype transmission.
Am. J. Hum. Genet.
 ,
65
,
1170
–1177.
25
Straub, R.E., Jiang, Y., MacLean, C., Ma, Y., Webb, B.T., Myakishev, M.V., Harris-Kerr, C., Wormley, B., Sadek, H., Kadambi, B. et al. (
2002
) Genetic variation in the 6p22.3 gene DTNBP1, the human ortholog of the mouse dysbindin gene, is associated with schizophrenia.
Am. J. Hum. Genet.
 ,
71
,
337
–348.
26
Sharma, T. and Antonova, L. (
2003
) Cognitive function in schizophrenia. Deficits, functional consequences, and future treatment.
Psychiatr. Clin. N. Am.
 ,
26
,
25
–40.
27
Gasperoni, T.L., Ekelund, J., Huttunen, M., Palmer, C.G., Tuulio-Henriksson, A., Lonnqvist, J., Kaprio, J., Peltonen, L. and Cannon, T.D. (
2003
) Genetic linkage and association between chromosome 1q and working memory function in schizophrenia.
Am. J. Med. Genet.
 ,
116B
,
8
–16.
28
Hafner, H. (
2003
) Gender differences in schizophrenia.
Psychoneuroendocrinology
 ,
28
,
17
–54.
29
Reik, W. and Walter, J. (
2001
) Evolution of imprinting mechanisms: the battle of the sexes begins in the zygote.
Nat. Genet.
 ,
27
,
255
–256.
30
Ostrer, H. (
2001
) Genome and hormones: gender differences in physiology. Invited review: sex-based differences in gene expression.
J. Appl. Physiol.
 ,
91
,
2384
–2388.
31
Hovatta, I., Lichtermann, D., Juvonen, H., Suvisaari, J., Terwilliger, J.D., Arajärvi, R., Kokko-Sahin, M.-L., Ekelund, J., Lönnqvist, J. and Peltonen, L. (
1998
) Linkage analysis of putative schizophrenia gene candidate regions on chromosome 3p, 5q, 6p, 8p, 20p and 22q in a population-based sampled Finnish family set.
Mol. Psychiat.
 ,
3
,
452
–457.
32
Ekelund, J., Lichtermann, D., Hovatta, I., Ellonen, P., Suvisaari, J., Terwilliger, J.D., Juvonen, H., Varilo, T., Arajärvi, R., Kokko-Sahin, M.-L. et al. (
2000
) Genome-wide scan for schizophrenia in the Finnish Population: evidence for a locus on chromosome 7q22.
Hum. Mol. Genet.
 ,
9
,
1049
–1057.
33
Paunio, T., Ekelund, J., Varilo, T., Parker, A., Hovatta, I., Turunen, J.A., Rinard, K., Foti, A., Terwilliger, J.D., Juvonen, H. et al. (
2001
) Genome-wide scan in a nationwide study sample of schizophrenia families in Finland reveals susceptibility loci on chromosomes 2q and 5q.
Hum. Mol. Genet.
 ,
10
,
3037
–3048.
34
Lichtermann, D., Hovatta, I., Terwilliger, J.D., Peltonen, L. and Lönnqvist, J. (
1998
) Concordance for sex and the pseudoautosomal gene hypothesis revisited: no evidence of increased sex concordance in a nationwide Finnish sample of siblings with paternally derived schizophrenia.
Am. J. Psychiat.
 ,
155
,
1365
–1375.
35
American Psychiatric Association. (
1994
)
Diagnostic and Statistical Manual of Mental Disorders
 , 4th edn. American Psychiatric Press, Washington, DC.
36
Kendler, K.S., McGuire, M., Gruenberg, A.M., O'Hare, A., Spellman, M. and Walsh, D. (
1993
) The Roscommon family study. III. Schizophrenia-related personality disorders in relatives.
Arch. Gen. Psychiat.
 ,
50
,
781
–788.
37
Williams, J., Farmer, A.E., Ackenheil, M., Kaufmann, C.A. and McGuffin, P. (
1996
) A multicentre inter-rater reliability study using OPCRIT computerized diagnostic system.
Psychol. Med.
 ,
26
,
775
–783.
38
McGuffin, P., Farmer, A.E. and Harvey, I. (
1991
) A polydiagnostic application of operational criteria in studies of psychotic illness. Development and reliability of the OPCRIT system.
Arch. Gen. Psychiat.
 ,
48
,
764
–770.
39
Arajarvi, R., Haukka, J., Varilo, T., Suokas, J., Juvonen, H., Suvisaari, J., Muhonen, M., Suominen, K., Tuulio-Henriksson, A., Schreck, M. et al. (
2003
) Clinical phenotype of schizophrenia in a Finnish isolate.
Schizophr. Res.
  (in press).
40
Syvänen, A.C., Aalto-Setala, K., Harju, L., Kontula, K. and Soderlund, H. (
1990
) A primer-guided nucleotide incorporation assay in the genotyping of apolipoprotein E.
Genomics
 ,
8
,
684
–692.
41
Pastinen, T., Raitio, M., Lindroos, K., Tainola, P., Peltonen, L. and Syvanen, A.C. (
2000
) A system for specific, high-throughput genotyping by allele-specific primer extension on microarrays.
Genome Res.
 ,
10
,
1031
–1042.
42
O'Connell, J.R. and Weeks, D.E. (
1998
) PedCheck: a program for identification of genotype incompatibilities in linkage analysis.
Am. J. Hum. Genet.
 ,
63
,
259
–266.
43
Lee, W.C. (
2002
) Testing for candidate gene linkage disequilibrium using a dense array of single nucleotide polymorphisms in case-parents studies.
Epidemiology
 ,
13
,
545
–551.