Major Affective Disorder 1

Bipolar affective disorder is a genetically heterogeneous complex trait. One susceptibility locus for bipolar disorder, MAFD1, has been mapped to chromosome 18p.

Other mapped loci include MAFD2 (309200) on chromosome Xq28, MAFD3 (609633) on chromosome 21q22, MAFD4 (611247) on chromosome 16p12, MAFD7 (612371) on chromosome 22q12, MAFD8 (612357) on chromosome 10q21, and MAFD9 (612372) on chromosome 12p13.

Epistatic interaction has been postulated for loci designated MAFD5 (611535) on chromosome 2q22-q24 and MAFD6 (611536) on chromosome 6q23-q24.

Depressive disorders represent a prevalent (1 to 2%) and major illness characterized by episodes of dysphoria that are associated with somatic symptoms. It may have a manic-depressive (bipolar) or purely depressive (unipolar) course. The most characteristic features of bipolar affective disorder (BPAD) are episodes of mania (bipolar I, BP I) or hypomania (bipolar II, BP II) interspersed with periods of depression (Goodwin and Jamison, 1990). If untreated, manic-depressive illness is associated with a suicide rate of approximately 20%.

Wright et al. (1984) studied binding of radioiodine-labeled hydroxybenzylpindolol to beta-adrenoceptors in lymphoblastoid cell lines from members of 5 families affected by manic-depressive disorder. Binding was reduced to less than half of control values in cell lines from 4 of 6 patients with manic-depressive disorder and only 1 of 18 unaffected relatives or controls. All the cell lines with reduced binding came from 3 families; members of 2 remaining families showed normal binding. The findings were interpreted as indicating genetic heterogeneity in manic-depressive disorder and a role played by a beta-adrenoceptor defect in genetic susceptibility to the disorder in some cases.

O'Reilly et al. (1994) presented a 2-generation family in which 8 members met DSM-III-R criteria for major depression. Four of the affected individuals failed to respond to standard therapeutic doses of tricyclic and new generation antidepressants but subsequently responded to the monoamine oxidase inhibitor tranylcypromine. O'Reilly et al. (1994) suggested that families demonstrating preferential response to a particular psychotropic drug may be a more homogeneous group in which to perform linkage analysis. Grof et al. (1994) also suggested that a helpful criterion for selecting more homogeneous groups of patients for linkage analysis would be the presence of a response to specific treatment. To that end, they studied 121 probands with primary affective disorders and 903 first-degree relatives and spouses. Seventy-one probands responded to lithium treatment and 50 were nonresponders. Study of the first-degree relatives of lithium responders revealed that 3.8% had bipolar disorder, whereas none of the relatives of the nonresponders were so affected. Schizophrenia was more common in the families of nonresponders (2.4% vs 0.3%).

In studies of the effect of lithium on Xenopus morphogenesis, Klein and Melton (1996) determined that lithium acts through inhibition of glycogen synthase kinase 3-beta (GSK3B; 605004) and not through inhibition of inositol monophosphatase (602064). They suggested that their observations may provide insights into the pathogenesis and treatment of bipolar disorder.

Using positron emission tomographic (PET) images of cerebral blood flow and rate of glucose metabolism to measure brain activity, Drevets et al. (1997) localized an area of abnormally decreased activity in the prefrontal cortex ventral to the genu of the corpus callosum in both familial bipolar depression patients and familial unipolar depression patients. The decrement in activity was at least partly explained by a corresponding reduction in cortical volume, as demonstrated by magnetic resonance imaging (MRI). This region had previously been implicated in the mediation of emotional and autonomic responses to socially significant or provocative stimuli, and in the modulation of the neurotransmitter systems targeted by antidepressant drugs.

Williams et al. (2002) demonstrated that lithium, carbamazepine, and valproic acid, all drugs used to treat bipolar affective disorder, inhibit the collapse of sensory neuron growth cones and increase growth cone area. These effects do not depend on glycogen synthase kinase-3 (see 606784) or histone deacetylase (see 601241) inhibition. Inositol, however, reverses the effects of the drugs on growth cones, thus implicating inositol depletion in their action. Moreover, the development of Dictyostelium is sensitive to lithium and to valproic acid, but resistance to both is conferred by deletion of the gene that codes for prolyl oligopeptidase (600400), which also regulates inositol metabolism. Inhibitors of prolyl oligopeptidase reversed the effects of all 3 drugs on sensory neuron growth cone area and collapse. Williams et al. (2002) concluded that their results suggest a molecular basis for both bipolar affective disorder and its treatment.

The role of genetic factors in bipolar disorder is indicated by concordance in monozygotic and dizygotic twins, respectively, of 57% and 14%, and the correlation between adopted persons and their biologic relatives (Cadoret, 1978).

Brent and Mann (2005, 2006) noted that in studies of adopted children and twins, familial concordance for suicidal behavior is explained by both genetic and environmental factors. Suicidal behavior that begins before 25 years of age is highly familial, and having a greater number of affected family members is associated with an earlier age.

Craddock and Sklar (2013) noted that strong evidence exists for a polygenic contribution to risk of bipolar disorder (i.e., many risk alleles of small effect). They stated that most cases of bipolar disorder involve the interplay of several genes or more complex genetic mechanisms, together with the effects of nongenetic (environmental) risk factors and stochastic factors.

Anticipation

In 34 families unilineal for bipolar affective disorder, McInnis et al. (1993) compared age of onset and disease severity between 2 generations. They found that the second generation experienced onset 8.9 to 13.5 years earlier and illness 1.8 to 3.4 times more severe than did the first generation. They concluded that genetic anticipation was demonstrated and they suggested that genes with expanding trinucleotide repeats might be involved in the genetic etiology of the disorder.

Gelernter (1995) reviewed the genetics of BPAD and other behavioral disorders and discussed specifically the difficulties associated with possible anticipation in bipolar affective disorder as discussed by McInnis et al. (1993). McInnis (1996) traced the history of anticipation. He suggested that it has its roots in the social chaos of the French Revolution, as well as in the medical observations of the French 'alienist' (an early term for psychiatrist) (Morel, 1857) and in the theories of atavism advocated by the Italian psychiatrist of the late 19th century (Lombroso, 1911). As to the validity of anticipation in bipolar disorder and schizophrenia, McInnis (1996) quoted Hodge and Wickramaratne (1995) as concluding that 'in psychiatric disorders, bias of ascertainment is pervasive and that there is no simple way to circumvent it.'

Parent-of-Origin Effect

McMahon et al. (1995) tested for parent-of-origin effect on the transmission of bipolar affective disorder, which might reflect either imprinting or mitochondrial inheritance. They examined the frequency and risk of BPAD among relatives in a sample of 31 families ascertained through treated probands with BPAD and selected for the presence of affected phenotypes in only 1 parental lineage. They observed a higher than expected frequency of affected mothers (p less than 0.04), a 2.3- to 2.8-fold increased risk of illness for maternal relatives (P less than 0.006), and a 1.3- to 2.5-fold increased risk of illness for the offspring of affected mothers (p less than 0.017). In 7 large pedigrees, fathers repeatedly failed to transmit the affected phenotype to daughters or sons. Taken together, these findings were interpreted as indicating a maternal effect in the transmission in BPAD susceptibility and suggested that molecular studies of mtDNA and imprinted DNA are warranted in patients with BPAD.

A number of family studies had reported increased morbid risk to the mothers, relative to the fathers, of probands with bipolar affective disorder. An excess of mother-offspring pairs had also been reported. These observations suggested that bipolar affective disorder may be caused by mitochondrial DNA mutations. Kirk et al. (1999) sequenced the mitochondrial genome in 25 bipolar patients with family histories of psychiatric disorder that suggested matrilineal inheritance. No polymorphism identified more than once in this sequencing showed any significant association with bipolar affective disorder in association studies using 94 cases and 94 controls. To determine whether their sample of patients showed evidence of selection against the maternal lineage, Kirk et al. (1999) determined genetic distances between all possible pairwise comparisons within the bipolar and control groups, based on multilocus mitochondrial polymorphism haplotypes. These analyses revealed fewer closely related haplotypes in the bipolar group than in the matched control group, suggesting selection against maternal lineages in this disease. Such selection was considered compatible with recurrent mitochondrial mutations, which are associated with slightly decreased fitness.

Linkage to Chromosome 18 (MAFD1)

In the course of a systematic genomic survey, Berrettini et al. (1994) examined 22 manic-depressive (bipolar) families for linkage to 11 chromosome 18 pericentromeric marker loci, under dominant and recessive models. Overall lod score analysis for the pedigrees was not significant under either model, but several families yielded scores consistent with linkage under dominant or recessive models. Affected sib pair analysis of these data yielded evidence for linkage (p less than 0.001) with D18S21. Affected pedigree member analysis also suggested linkage. The results were interpreted as suggesting a susceptibility gene in the pericentromeric region of chromosome 18, with a complex mode of inheritance.

Pursuant to the report by Berrettini et al. (1994) of a BPAD susceptibility locus on chromosome 18 and the report of a parent-of-origin effect by McMahon et al. (1995), Stine et al. (1995) undertook a linkage study in 28 nuclear families selected for apparent unilineal transmission of the BPAD phenotype. They used 31 polymorphic markers spanning chromosome 18. The study was distinguished by relatively small, densely affected families with apparent unilineal transmission and direct clinical evaluation of family members by psychiatrists. Evidence for linkage was tested with affected-sib-pair and lod score methods under 2 definitions of the affected phenotype. The affected-sib-pair analyses indicated excess allele sharing for markers on 18p within the region reported previously. The greatest sharing was at D18S37. In addition, excess sharing of the paternally, but not maternally, transmitted alleles was observed at 3 markers on 18q. The evidence for linkage to loci on both 18p and 18q was strongest in the 11 paternal pedigrees, i.e., those in which the father or one of the father's sibs was affected. The results were interpreted by Stine et al. (1995) as providing further support for linkage of BPAD to chromosome 18.

Prompted by the report by Berrettini et al. (1994) of a bipolar susceptibility locus in the region of the centromere of chromosome 18, Pauls et al. (1995) studied markers from this region in Old Order Amish families. Although linkage findings were replicated in 1 previously studied Amish pedigree containing 4 affected individuals, linkage to this region was excluded in the larger sample. Pauls et al. (1995) concluded that if a susceptibility locus for bipolar disorder is located in this region of chromosome 18, it is of minor significance among the Amish.

In a genetically isolated population of the central valley of Costa Rica (CVCR), Freimer et al. (1996) undertook a linkage study of severe bipolar disorder in 2 pedigrees. The 2.6 million residents of the CVCR are descended mainly from a small group of Spanish and Amerindian founders who lived in the 16th and 17th centuries; by the beginning of the 18th century, the CVCR had a single population that then grew rapidly, without subsequent immigration, for almost 200 years (Escamilla et al., 1996). Freimer et al. (1996) found strongest evidence for a specific locus on chromosome 18q22-q23 where 7 of 16 markers yielded peak lod scores over 1.0. This localization was supported by marker haplotypes shared by 23 of 26 affected individuals studied. As a continuation of that study, McInnes et al. (1996) performed a complete genome screen for genes predisposing to severe bipolar disorder. They considered as affected only individuals with bipolar mood disorder and screened the genome for linkage with 473 microsatellite markers. They used a model for linkage analysis that incorporated a high phenocopy rate and a conservative estimate of penetrance. They suggested on the basis of their results that 18q, 18p, and 11p deserve further study; in these regions suggestive lod scores were observed for 2 or more contiguous markers. Isolated lod scores that exceeded the threshold in 1 or both families studied also occurred on 10 other chromosomes. Additional linkage studies on 2 extended BP I pedigrees from the CVCR implicated a candidate region on 18p11.3 (Escamilla et al., 1999).

Knowles et al. (1998) could find no evidence for significant linkage between bipolar affective disorder and chromosome 18 pericentromeric markers in a large series of multiplex extended pedigrees. This was one of the largest samples reported to date: 1,013 genotyped individuals in 53 unilineal multiplex pedigrees. Ten highly polymorphic markers and a range of parametric and nonparametric analyses were used. Not only was there no evidence for linkage, but there was also no evidence for significant parent-of-origin effect.

McInnes et al. (2001) further investigated the 18p11.3 region by creating a physical map and developing 4 new microsatellite and 26 single-nucleotide polymorphism (SNP) markers for typing in the Costa Rican pedigree and population samples. The results of fine-scale association analyses in the population sample, as well as evaluation of haplotypes in 1 of the large pedigrees, suggested a candidate region containing 6 genes but also highlighted the complexities of linkage disequilibrium mapping of common disorders.

To clarify the issue of genetic linkage between bipolar affective disorder and 18q, McMahon et al. (2001) analyzed the relationship between clinical features and allele sharing. Relatives ascertained through a proband who had BP I disorder were interviewed by a psychiatrist, assigned an all-sources diagnosis, and genotyped with 32 markers on 18q21-q23. The authors found that paternal allele sharing on 18q21 was significantly associated with a diagnostic subtype, and was greatest in pairs where both sibs had BP II. Paternal allele sharing across 18q21-q23 was also significantly greater in families with at least 1 sib pair in which both had BP II. In these families, multipoint affected sib-pair linkage analysis produced a peak paternal lod score of 4.67 versus 1.53 in all families. Thus, affected sib pairs with BP II discriminated between families who showed evidence of linkage to 18q and families who did not. Families with a BP II sib pair produced an increased lod score and improved linkage resolution. These findings strengthened the evidence of genetic linkage between BPAD and 18q, and provided preliminary support for BP II as a genetically valid subtype of BPAD. Bipolar type II disorder is characterized by hypomania that is so brief or so slight as to cause no significant problems in functioning. BP I, on the other hand, is the diagnosis attached to anyone with a significantly problematic manic state (extreme symptoms of grandiosity, poor social judgment or functional impairment due to distractibility at work, etc.).

Linkage to Chromosome 2q22-q24

See MAFD5 (611534) for a discussion of linkage of susceptibility to bipolar disorder to chromosome 2q22-q24.

Linkage to Chromosome 3p

Etain et al. (2006) conducted a genomewide search with 384 microsatellite markers using nonparametric linkage (NPL) analysis in 87 sib pairs ascertained as part of the European Collaborative Study of Early Onset Bipolar Affective Disorder. Early-onset patients (age at onset of 21 years or below) were studied because age of onset may help to define homogeneous bipolar affective disorder subtypes. The 3p14 region showed the most significant linkage in the first phase of analysis with an NPL score of 3.51. Additional linkage analysis with increased marker density revealed an NPL score of 3.83 at chromosome 3p14.

Linkage to Chromosome 4

Blackwood et al. (1996) carried out a linkage study in Scotland in 12 bipolar families. In a single family, a genome search using 193 markers indicated linkage on 4p where D4S394 generated a 2-point lod score of 4.1 under a dominant model of inheritance. With 3-point analyses using neighboring markers, they obtained a maximum lod score of 4.8. (Eleven other bipolar families were typed using D4S394 and in all families combined there was evidence of linkage with heterogeneity with a maximum 2-point lod score of 4.1 (theta = 0.0; alpha = 0.35).)

Ginns et al. (1998) reported a different approach to linkage study of BPAD in the Old Order Amish. To determine whether there could be protective alleles that prevent or reduce the risk of developing BPAD, similar to what is observed in other genetic disorders, they used 'mental health wellness' (absence of any psychiatric disorder) as the phenotype in their genomewide linkage scan of several large multigeneration Old Order Amish pedigrees exhibiting a high incidence of BPAD. They found strong evidence for linkage of mental health wellness to a locus on 4p, designated MHW1 (603663), at D4S2949; maximum nonparametric linkage score = 4.05, p = 5.22 x 10(-4). They also found suggestive evidence for a locus on 4q, designated MHW2 (603664), at D4S397; maximum nonparametric linkage score = 3.29, p = 2.57 x 10(-3). Findings were consistent with the hypothesis that certain alleles can prevent or modify the clinical manifestations of BPAD and perhaps other related affective disorders.

Ekholm et al. (2003) performed a genomewide scan for susceptibility loci in bipolar disorder in 41 Finnish families with at least 2 affected sibs. They identified a distinct locus on 16p12 (see MAFD4, 611247) and observed 3 additional loci with a 2-point lod score greater than 2.0, at markers on 4q32, 12q23, and Xq25. After fine mapping these chromosomal regions and genotyping additional family members, 4q32 provided significant evidence of linkage for the 3-point analyses (maximum lod = 3.6 between D4S3049 and D4S1629).

Linkage to Chromosome 5

Garner et al. (2001) used an algorithm that permitted nonparametric linkage analysis of large, complex pedigrees with multiple inbreeding loops to reanalyze the genome-screen data from the Costa Rican kindred segregating severe bipolar disorder (Freimer et al., 1996; McInnes et al., 1996). The results were consistent with previous linkage findings on chromosome 18 and also suggested a novel locus on chromosome 5 that was not identified using traditional linkage analysis.

Hong et al. (2004) performed linkage analysis using 74 individuals from the Costa Rican pedigree and found evidence for a 3.2-Mb region between markers D5S1480 and D5S2090 on chromosome 5q31-q33. The authors suggested that conflicting haplotype data reflected incomplete penetrance, phenocopies, or locus/allelic heterogeneity. The authors noted that Freimer et al. (1996) described a conserved haplotype on 18q22-q23 in the same kindred. Hong et al. (2004) found that 12 of 20 affected individuals shared both haplotypes, suggesting that both loci are important in conferring disease risk.

Coon et al. (1993) carried out an extensive linkage analysis in 8 moderate-sized families with manic-depression. When autosomal dominant inheritance was assumed, 273 DNA markers gave lod scores less than -2.0 at theta = 0.0, 174 DNA loci produced lod scores less than -2.0 at theta = 0.05, and 4 DNA marker loci yielded lod scores greater than 1. Of the markers giving lod scores greater than 1, only D5S62 continued to show evidence for linkage when the affected-pedigree-member method was used. D5S62 maps to distal 5q, a region containing neurotransmitter receptor genes for dopamine (e.g., 126449), gamma-aminobutyric acid (e.g., 137160, 137164), glutamate (e.g., 138248), and norepinephrine (e.g., 109690, 104219, 104220).

Linkage to Chromosome 6p

Smeraldi et al. (1978) first suggested linkage between HLA on chromosome 6p21.3 and affective disorders on the basis of the finding that pairs of affected sibs shared HLA haplotypes more often than would be predicted by chance. Weitkamp et al. (1981) likewise found evidence of a susceptibility gene or genes linked to HLA. Neither group subdivided the depressive disorders into bipolar and unipolar subtypes. Stronger evidence of linkage might be found in 1 subtype, or it may turn out that both are linked to HLA, suggesting that they are different forms of the same illness. One of Weitkamp's study families was that reported earlier by Pardue (1975)--in fact, Pardue's own kindred (Wingerson, 1982). Weitkamp et al. (1981) found that HLA haplotype identity in pairs of affected sibs and in pairs of unaffected older sibs deviated markedly from expected (p less than 0.005). Perhaps surprisingly, no increase in HLA haplotype identity was found in sibships with more than 2 affected members. When parents had a difference in load of genes for susceptibility (as estimated by the occurrence of affective illness in themselves and their relatives), HLA haplotypes were randomly transmitted to unaffected or affected children from the affected, 'high-load' parent, but not randomly from the unaffected, 'low-load' parent (p less than 0.001), suggesting a recessive effect, i.e., greater chance of illness in homozygotes.

Stancer et al. (1988) published data apparently confirming the relationship between HLA and manic depression. When combined with their previous data, the total number of families analyzed was 117. As in the previous study, the increase in HLA haplotype sharing over random expectation was greater if 'high-load' sibships, i.e., sibships with 3 or more affected sibs, were omitted from the analysis. Weitkamp (1981, 1983) suggested that the extent of HLA haplotype sharing among affected sib pairs should decrease as the number of parental HLA haplotypes containing susceptibility genes increases from 1 to 4. Thus, he reasoned that there may actually be less HLA haplotype sharing among sibs when the parents have maximum genetic susceptibility ('high load') compared with families in which the genetic susceptibility that could be contributed by either parent is limited to the genes in 1 of the 2 HLA haplotypes in that parent. If an increased number or variety of affective disorder susceptibility genes in a person results in a greater probability of illness, then nuclear families with a higher proportion of affected family members are likely to have a greater number or variety of affective disorder susceptibility genes than families with a low proportion of affected members. Weitkamp and Stancer (1989) suggested that the HLA effect may be greater in unipolar than in bipolar disorders and more apparent in families with few affected members than in 'high-load' families.

Schulze et al. (2004) extended the study of Dick et al. (2003) to test for robustness of the linkage to differing analysis methods, genotyping error, and gender-specific maps; for parent-of-origin effects; and for interaction with markers within the schizophrenia linkage region on chromosome 6p (see SCZD3; 600511). Members of 245 families ascertained through a sib pair affected with bipolar I or schizoaffective-bipolar disorder were genotyped with 18 markers spanning chromosome 6, and nonparametric linkage analysis was performed. Linkage to 6q was robust to analysis methods, gender-specific map differences, and genotyping error. The locus conferred a 1.4-fold increased risk. Affected sibs shared the maternal more often than paternal chromosome (p = 0.006), which could reflect a maternal parent-of-origin effect. There was a positive correlation between family-specific linkage scores on 6q and those on 6p22.2 (p less than 0.0001). Linkage analysis for each locus conditioned on evidence of linkage to the other increased the evidence for linkage at both loci (p less than 0.0005). Lod scores increased from 2.26 to 5.42 on 6q and from 0.35 to 2.26 on 6p22.2. The results supported linkage of bipolar disorder to 6q, revealed a maternal parent-of-origin effect, and demonstrated an interaction of this locus with a locus on chromosome 6p22.2 linked to schizophrenia.

Linkage to Chromosome 6q22

Middleton et al. (2004) performed a linkage analysis on 25 extended multiplex Portuguese families, including 12 families previously reported by Pato et al. (2004), segregating for bipolar disorder using a high-density SNP genotyping assay with a 0.21-Mb intermarker spacing. The analysis revealed genomewide significance with a maximum NPL of 4.20 and a maximum lod score of 3.56 at 6q22 (125.8 Mb).

Linkage to Chromosome 8

Ophoff et al. (2002) performed a genomewide association study of severe bipolar disorder in the patients from the central valley of Costa Rica. They observed LD with severe bipolar disorder on several chromosomes; the most striking results were in proximal 8p, a region that had previously shown linkage to schizophrenia. Ophoff et al. (2002) suggested that this region could be important for severe psychiatric disorders rather than for a specific phenotype.

Cichon et al. (2001) conducted a complete genome screen with 382 markers in a sample of 75 BPAD families of German, Israeli, and Italian origin. Parametric and nonparametric linkage analysis was performed. The highest 2-point lod score was obtained on 8q24 (D8S514; lod score = 3.62), and the authors confirmed a putative locus on 10q25-q26 (D10S217; lod score = 2.86). By analyzing the autosomal genotype data, putative paternally imprinted loci were identified in chromosomal regions 2p24-p21 and 2q31-q32; maternally imprinted susceptibility genes may be located on 14q32 and 16q21-q23.

Park et al. (2004) genotyped 373 individuals from 40 extended pedigrees with high density bipolar disorder and found evidence for significant linkage for psychotic bipolar disorder (genomewide p less than 0.05) to chromosomes 9q31 (lod = 3.55) and 8p21 (lod = 3.46). Nine other sites obtained lod scores supportive of linkage. The highest lod scores occurred in the subgroup of families with the largest concentration of psychotic individuals. Seven of the loci identified in this study had previously been implicated in schizophrenia, suggesting that psychosis is a potentially useful phenotype in bipolar disorder for genetic studies.

Linkage to Chromosome 9

Sherrington et al. (1994) performed linkage analysis on 5 multigenerational families with bipolar and unipolar affective disorder, using highly polymorphic microsatellite markers from the ABO-AK1-ORM region at 9q34. The dopamine beta-hydroxylase locus (223360) is also at 9q34 and was considered to be a candidate gene. Their analyses provided strong evidence against a major susceptibility allele in this region, in contradistinction to the findings of Hill et al. (1988), Tanna et al. (1989), and Wilson et al. (1989, 1991).

Venken et al. (2005) conducted a genomewide scan to identify susceptibility loci for affective spectrum disorder (bipolar disorder and recurrent unipolar depression) in 9 families from an isolated population in Vasterbotten in northern Sweden. A region on chromosome 9q showed the highest 2-point and multipoint lod scores. A common ancestral haplotype was inherited by 18 of 21 patients from 3 families linked to 9q, which reduced the candidate region to 1.6 Mb on 9q31-q33. Further analysis identified the shared haplotype in 4.2% of 182 unrelated patients with bipolar disorder from the Vasterbotten isolate, but not in 182 control individuals. Venken et al. (2005) concluded that a susceptibility locus for affective disorder is located on chromosome 9q31-q33.

Park et al. (2004) genotyped 373 individuals from 40 extended pedigrees with high density bipolar disorder and found evidence for significant linkage for psychotic bipolar disorder (genomewide p less than 0.05) to chromosomes 9q31 (lod = 3.55) and 8p21 (lod = 3.46). Nine other sites obtained lod scores supportive of linkage. The highest lod scores occurred in the subgroup of families with the largest concentration of psychotic individuals. Seven of the loci identified in this study had previously been implicated in schizophrenia, suggesting that psychosis is a potentially useful phenotype in bipolar disorder for genetic studies.

Linkage to Chromosome 10q21

Ferreira et al. (2008) tested 1.8 million variants in 4,387 cases of bipolar disorder and 6,209 controls from 3 independent samples and identified a region of strong association with SNP rs10994336 in the ankyrin G gene (ANK3; 600465) on chromosome 10q21, with a p value of 9.1 x 10(-9). See MAFD8 (612357).

Linkage to Chromosome 11

A form of manic-depressive disorder in the Old Order Amish of Lancaster County, Pennsylvania, was thought by Egeland et al. (1987) to be tightly linked to INS (176730) and HRAS1 (190020). In linkage studies using RFLPs related to these genes on the tip of 11p, the maximum lod score was 4.5 at theta = 0.0. Of interest is the description by Joffe et al. (1986) of cosegregation of thalassemia and affective disorder in a non-Amish pedigree. Egeland et al. (1987) suggested that the tyrosine hydroxylase (TH; 191290) gene, which maps to 11p, should be considered as a candidate gene because this enzyme catalyzes an important step in the dopamine synthesis pathway. Gill et al. (1988) ruled out tight linkage between manic-depressive psychosis and the 11p markers HRAS1 and INS, however. Two other groups failed to find linkage of 11p markers to manic-depressive illness (Neiswanger et al., 1990). From another extension of the study of the original Amish pedigree, Pauls et al. (1991) likewise excluded linkage to 11p markers. Pakstis et al. (1991) found no evidence of linkage after screening 185 marker loci in the Old Order Amish. They estimated that roughly 23% of the autosomal genome had been excluded. Law et al. (1992) determined the INS and HRAS1 genotypes of 81 persons in this pedigree and excluded that region of chromosome 11 as the site of the gene, which they symbolized BAD (for bipolar affective disorder).

In 5 Icelandic pedigrees, Holmes et al. (1991) could find no evidence of linkage of manic depression to the dopamine D2 receptor (DRD2; 126450) or other markers in its vicinity on 11q.

Linkage to Chromosome 12p13

Ferreira et al. (2008) tested 1.8 million variants in 4,387 cases of bipolar disorder and 6,209 controls from 3 independent samples and identified association with SNP rs1006737 in the CACNA1C gene (114205) on chromosome 12p13, with a p value of 7.0 x 10(-8). See MAFD9 (612372).

Linkage to Chromosome 12q

By linkage analysis in 2 Danish families with bipolar affective disorder, Ewald et al. (1998) found that the microsatellite marker D12S1639 gave a significant lod score of 3.37. Earlier, Craddock et al. (1994) had suggested linkage between affective disorder and Darier disease (124200), which maps to 12q23-q24.1. Linkage results from independent studies in Canadian families (Morissette et al., 1999) supported the existence of a susceptibility locus on 12q23-q24. To take advantage of isolated populations for genetic mapping and disease gene identification, Degn et al. (2001) investigated a possible chromosomal segment shared among distantly related patients with bipolar affective disorder in the Faroe Islands, using 17 microsatellite markers covering 24 cM in the 12q24 region. The region of most interest contained the primary region suggested by the previously reported haplotypes in the 2 Danish families studied by Ewald et al. (1998).

Ewald et al. (2002) reported a genomewide scan for risk genes involved in bipolar disorder in 2 Danish Caucasian families with affected members in several generations. Ewald et al. (2002) used 613 microsatellite markers in a 2-stage approach. Linkage was obtained at 12q24.3 (D12S1639) with a 2-point parametric lod score of 3.42 (empirical P-value 0.00004, genomewide P-value 0.0417) in both families tested. The multipoint parametric lod score at D12S1639 was 3.63 (genomewide P-value 0.0265). At 1p22-p21 (D1S216), a parametric, affecteds-only 2-point lod score of 2.75 (empirical P-value 0.0002, genomewide P-value 0.1622) was found. A 3-point lod score of 2.98 (genomewide p value = 0.1022) was found at D1S216, and a multipoint nonparametric analysis yielded a maximum nonparametric linkage (NPL)-all score of 17.60 (p value = 0.00079) at D1S216.

In 2 cohorts of patients with bipolar affective disorder from Germany and Russia totaling 883 patients and 1,300 controls, Cichon et al. (2008) observed an association between disease and the minor alleles of 3 SNPs in haplotype 1 of the TPH2 gene on chromosome 12q21, (rs11178997, rs11178998, and rs7954758; odds ratio of 1.6, p value of 0.00073). Haplotype 1 covers part of the 5-prime regulatory region and exons 1 and 2 of the TPH2 gene. Cichon et al. (2008) also observed an association between bipolar disorder and a nonsynonymous SNP in the TPH2 gene (P206S; 607478.0003).

Linkage to Chromosome 16

See MAFD4 (611247) for a discussion of linkage of susceptibility to bipolar disorder to chromosome 16p12.

Linkage to Chromosome 17

Dick et al. (2003) performed genomewide linkage analyses on 1,152 individuals from a new sample of 250 families segregating for bipolar disorder and related affective illnesses, ascertained at 10 sites in the United States through a proband with BP I affective disorder and a sib with BP I or schizoaffective disorder, bipolar type. Suggestive evidence for linkage was found on chromosome 17q (peak maximum lod score = 2.4) at marker D17S928, and on 6q (peak maximum lod score = 2.2) near marker D6S1021. Suggestive evidence of linkage was observed in 3 other regions, on chromosomes 2p, 3q, and 8q. This study, based on a linkage sample for bipolar disorder larger than any previously analyzed, indicated that several genes contribute to bipolar disorder.

Linkage to Chromosome 20

In 9 Australian pedigrees, Le et al. (1994) excluded close linkage of bipolar disorder to the gene encoding the alpha subunit of the stimulatory form of G protein (139320), previously mapped to chromosomal region 20q13.2.

Radhakrishna et al. (2001) studied a large Turkish pedigree segregating apparently autosomal dominant BPAD, which contained 13 affected individuals. The age of onset ranged from 15 to 40 years with a mean age of 25 years. The phenotypes consisted of recurrent manic and major depressive episodes, including suicide attempts. There was usually full remission with lithium treatment. A genotyping of 230 highly informative polymorphic markers throughout the genome and subsequent linkage analysis using a dominant mode of inheritance showed strong evidence for a BPAD susceptibility locus on chromosome 20p11.2-q11.2. The highest 2-point lod score of 4.34 (theta = 0.0) was obtained with markers D20S604, D20S470, D20S836, and D20S838 (100% penetrance). Haplotype analysis using informative recombinants enabled the mapping of the BPAD locus in this family between markers D20S186 and D20S109 in a region of approximately 42 cM. The authors noted that the chromosome 20 BPAD susceptibility locus had not been identified in previous studies of common 'polygenic' small pedigrees, which could be explained by an absence of common deleterious mutations of the chromosome 20 BPAD locus in those pedigrees and/or by the presence of a severe mutation in the Turkish pedigree that by itself confers susceptibility to BPAD.

Linkage to Chromosome 21q22

In a preliminary genome screen of 47 bipolar disorder families, Straub et al. (1994) detected one in which a lod score of 3.41 was demonstrated for linkage with the PFKL (171860) locus on 21q22.3. Largely positive lod scores were obtained also with 14 other markers in 21q22.3 in this family. In a linkage analysis with an 'affecteds-only' method, Aita et al. (1999) found linkage to the 21q22 region, corroborating the findings of the earlier study by the same group (Straub et al., 1994).

Following up on the work of Straub et al. (1994) suggesting a susceptibility locus for bipolar affective disorder on the long arm of chromosome 21, Smyth et al. (1997) studied 23 multiply affected pedigrees collected from Iceland and the U.K., using the markers PFKL, D21S171, and D21S49. Positive lod scores were obtained with 3 Icelandic families. Affected sib pair analysis demonstrated increased allele sharing. The same set of pedigrees had previously been typed for a tyrosine hydroxylase gene (TH; 191290) polymorphism at 11p15 and had shown some evidence for linkage. When information from TH and the 21q markers was combined in a 2-locus admixture analysis, an overall admixture lod of 3.87 was obtained using the bipolar affection model. Thus the data of Smyth et al. (1997) were compatible with the hypothesis that a locus at or near TH influences susceptibility of some pedigrees, while a locus near D21S171 is active in others.

Age at onset (AAO) is a potential clinical marker of genetic heterogeneity in BP (Bellivier et al., 2001). Rates of comorbidity and clinical indicators of severity (e.g., suicide attempt) vary across different AAO subgroups, and AAO subgroups aggregate in families such that affected relatives typically have similar AAOs. Therefore, Lin et al. (2005) sought to incorporate AAO as a covariate in linkage analysis of BP using 2 different methods in genomewide scans of 150 multiplex pedigrees with 874 individuals. The LODPAL analysis identified 2 loci: one on 21q22.13 (MAFD3; 609633) and the other on 18p11.2 (MAFD1) for early onset (AAO = 21 years or younger) and later onset (AAO = older than 21 years), respectively. The finding on 21q22.13 was significant at the chromosome-wide level, even after correction for multiple testing. Moreover, a similar finding was observed in an independent sample of 65 pedigrees (lod = 2.88). The finding on 18p11.2 was only nominally significant and was not observed in the independent sample. However, 18p11.2 emerged as one of the strongest regions in the ordered-subset analysis (OSA) with a lod of 2.92, in which it was the only finding to meet chromosomewide levels of significance after correction for multiple testing. These results suggested that 21q22.13 and 18p11.2 may harbor genes that increase the risks for early-onset and later-onset forms of BP, respectively. Lin et al. (2005) suggested that previous inconsistent linkage findings may have been due to differences in the AAO characteristics of the samples examined.

Linkage to Chromosome 22q12

For information on linkage of susceptibility to bipolar disorder to chromosome 22q12, see MAFD7 (612371) and Kelsoe et al., 2001.

Other Genomewide Linkage Studies

In an 'Old Order Amish revisited' study, Ginns et al. (1996) performed a genomewide linkage analysis in the Lancaster County group. In addition to the so-called pedigree 110, which was used for reporting the original genetic linkage data by Egeland et al. (1987), 2 pedigrees closely related to pedigree 110 and 2 other pedigrees, 210 and 310, were studied; all 5 pedigrees traced back to a founder couple who immigrated to the U.S. around 1750. The diagnoses were broken down into BP I (bipolar disorder with mania) and BP II (bipolar disorder with hypomania). Ginns et al. (1996) found evidence that regions on chromosomes 6, 13, and 15 harbor susceptibility loci for bipolar affective disorder, suggesting to them that bipolar affective disorder in the Old Order Amish is inherited as a complex trait.

LaBuda et al. (1996) reported progress of a full genome screen for loci predisposing to affective disorder in the Old Order Amish. To the previously reported lod score results published by Gerhard et al. (1994), they added lod score results for an additional 367 markers distributed throughout the genome, along with allele- and haplotype-sharing analyses on those chromosomes sufficiently saturated with markers. No statistically significant lod scores resulted. Some degree of allele sharing was found at 74 loci, and 3.8% of all markers analyzed passed more stringent significance criteria suggestive of linkage. Although genomic areas were highlighted for further exploration, the studies of LaBuda et al. (1996) identified no region clearly involved in the etiology of affective disorder in this population.

Risch and Botstein (1996) reviewed 19 linkage studies in manic-depressive illness; the studies purported to identify loci on 10 different autosomes, including both the short arm and the long arm of chromosome 18; linkage in 3 different regions of distal Xq had been proposed.

Segurado et al. (2003) applied the rank-based genome scan metaanalysis (GSMA) method (Levinson et al., 2003) to 18 bipolar disorder genome scan datasets in an effort to identify regions with significant support for linkage in the combined data. No region achieved genomewide statistical significance by several simulation-based criteria. The most significant p values (less than 0.01) were observed on chromosomes 9p22.3-p21.1, 10q11.21-q22.1, and 14q24.1-q32.12. Nominally significant p values were observed in several other chromosomal regions.

In a study of Ashkenazi Jewish families, Fallin et al. (2004) identified 4 regions suggestive of linkage to bipolar disorder on chromosomes 1, 3, 11, and 18.

Pato et al. (2004) conducted a genomewide scan of 16 extended families from a Portuguese genetic isolate with bipolar disorder and identified 3 regions on chromosomes 2, 11, and 19 with genomewide suggestive linkage and several other regions, including chromosome 6q, that approached suggestive levels of significance. This research replicated the finding of an elevated lod score near marker D6S1021 on chromosome 6q (peak NPL at D6S1021 = 2.02; p = 0.025). Higher density mapping provided additional support for this locus (NPL = 2.59; p = 0.0068) and another marker, D6S1639 (NPL = 3.06; p = 0.0019). On chromosome 11, linkage was found to D11S1883 (NPL = 3.15; p = 0.0014).

Middleton et al. (2004) performed a linkage analysis on 25 extended multiplex Portuguese families, including 12 families previously reported by Pato et al. (2004), segregating for bipolar disorder using a high-density SNP genotyping assay with a 0.21-Mb intermarker spacing. The analysis revealed genomewide significance with a maximum NPL of 4.20 and a maximum lod score of 3.56 at 6q22 (125.8 Mb). Several other areas had suggestive linkage: 2 regions on chromosome 2 (57 Mb, NPL = 2.98; 145 Mb, NPL = 3.09), chromosome 4 (91 Mb, NPL = 2.97), chromosome 11 (45-68 Mb, NPL = 2.51), chromosome 16 (20 Mb, NPL = 2.89), and chromosome 20 (60 Mb, NPL = 2.99).

McQueen et al. (2005) hypothesized that combining original genotype data on linkage of bipolar disorder would provide benefits of increased power and control over sources of heterogeneity that outweigh the difficulty and potential pitfalls of the implementation. Thus, they conducted a combined analysis using the original genotype data from 11 bipolar disorder genomewide linkage scans comprising 5,179 individuals from 1,067 families. Heterogeneity among studies was minimized in the analyses by using uniform methods of analysis and a common, standardized marker map. They demonstrated that combining original genome-scan data is a powerful approach for the elucidation of linkage regions underlying complex disease. Their results established genomewide significant linkage to BP on chromosomes 6q and 8q, and provided solid information to guide future gene-finding efforts that rely on fine mapping and association approaches. McQueen et al. (2005) observed the most significant result for 'narrow' BP (BP type I-only phenotype) on chromosome 6q. When the analysis was expanded to include BP II, the linkage signal on 6q was attenuated, despite the increase in the number of affected relative pairs (ARPs). In contrast, removal of the individuals with BP II from the analysis reduced the evidence of linkage on 8q.

Maziade et al. (2005) performed a dense genome scan to identify susceptibility loci shared by schizophrenia and bipolar disorder. They used the same ascertainment, statistical, and molecular methods for 480 members from 21 multigenerational families from Eastern Quebec affected by schizophrenia, bipolar affective disorder, or both. Five genomewide significant linkages with maximized lod scores over 4.0 were observed: 3 for bipolar disorder (15q11.1, 16p12.3, 18q12-q21) and 2 for the shared 'common locus' phenotype (15q26, 18q12-q21). Nine maximized lod scores exceeded the suggestive threshold of 2.6: 3 for bipolar disorder (3q21, 10p13, 12q23), 3 for schizophrenia (6p22, 13q13, 18q21), and 3 for the combined locus phenotype (2q12.3, 13q14, 16p13). Maziade et al. (2005) noted that all of the linkage signals overlapped formerly reported susceptibility regions except the signal at 15q26.

Cheng et al. (2006) conducted a 9-cM genomewide scan in a large bipolar pedigree sample from the National Institute of Mental Health Genetics Initiative (1,060 individuals from 154 multiplex families). Parametric and nonparametric analyses using both standard diagnostic models and comorbid conditions thought to identify phenotypic subtypes were conducted. Genomewide significant linkage was observed on chromosomes 10q25, 10p12, 16q24, 16p13, and 16p12 using standard diagnostic models, and on 6q25 (suicidal behavior), 7q21 (panic disorder), and 16p12 (psychosis) using phenotypic subtypes. Several other regions were suggestive of linkage including 1p13 (psychosis), 1p21 (psychosis), 1q44, 2q24 (suicidal behavior), 2p25 (psychosis), 4p16 (psychosis, suicidal behavior), 5p15, 6p25 (psychosis), 8p22 (psychosis), 8q24, 10q21, 10q25 (suicidal behavior), 10p11 (psychosis), 13q32 and 19p13 (psychosis).

Kimmel et al. (2005) reported a large family in which bipolar disorder appeared to cosegregate with autosomal dominant medullary cystic kidney disease. Of the 7 members with kidney disease, 5 had bipolar I disorder, one had unipolar depression, and 1 had a hyperthymic phenotype. The authors noted that the 2 known loci of medullary cystic kidney disease are in regions of chromosome 1 (MCKD1; 174000) and 16 (MCDK2; 603860) had previously been linked to bipolar disorder and schizophrenia.

Exclusion Studies

Although corticotropin-releasing hormone (CRH; 122560) and its function in the hypothalamic-pituitary-adrenal axis had been implicated in depression (Stratakis and Chrousos, 1995), Stratakis et al. (1997) could demonstrate no linkage between the CRH gene and bipolar affective disorder.