Molecular Psychiatry (2003) 8, 275–287 & 2003 Nature Publishing Group All rights reserved 1359-4184/03 $25.00 www.nature.com/mp FEATURE ARTICLE Chromosomal abnormalities and mental illness DJ MacIntyre1, DHR Blackwood1,2, DJ Porteous2, BS Pickard1,2 and WJ Muir1,2 1Department of Psychiatry, University of Edinburgh, Edinburgh, Scotland, UK; 2Department of Medical Sciences, University of Edinburgh, Edinburgh, Scotland, UK Linkage studies of mental illness have provided suggestive evidence of susceptibility loci over many broad chromosomal regions. Pinpointing causative gene mutations by conventional linkage strategies alone is problematic. The breakpoints of chromosomal abnormalities occurring in patients with mental illness may be more direct pointers to the relevant gene locus. Publications that describe patients where chromosomal abnormalities co-exist with mental illness are reviewed along with supporting evidence that this may amount to an association. Chromosomal abnormalities are considered to be of possible significance if (a) the abnormality is rare and there are independent reports of its coexistence with psychiatric illness, or (b) there is colocalisation of the abnormality with a region of suggestive linkage findings, or (c) there is an apparent cosegregation of the abnormality with psychiatric illness within the individual’s family. Breakpoints have been described within many of the loci suggested by linkage studies and these findings support the hypothesis that shared susceptibility factors for schizophrenia and bipolar disorder may exist. If these abnormalities directly disrupt coding regions, then combining molecular genetic breakpoint cloning with bioinformatic sequence analysis may be a method of rapidly identifying candidate genes. Full karyotyping of individuals with psychotic illness especially where this coexists with mild learning disability, dysmorphism or a strong family history of mental disorder is encouraged. Molecular Psychiatry (2003) 8, 275–287. doi:10.1038/sj.mp.4001232 Keywords: schizophrenia; bipolar disorder; chromosomes; linkage; mental retardation; learning disability Introduction of an overlap between the genetic contributions to depression and anxiety.8 While the aetiology of most psychiatric disorders A further difficulty is the lack of clarity as to the remains obscure, there is convincing evidence (from mode of inheritance of psychiatric illness, and like family, twin and adoption studies) that inherited other common disorders the inheritance of psychia- factors are important in the pathogenesis of both tric illness is likely to be complex with Mendelian schizophrenia1 and major affective disorder.2 How- and non-Mendelian subsets.9 An alternative hypoth- ever, it has so far proved difficult to identify these esis proposes that the additive or interactive effects of factors, for a variety of reasons. Chromosomal several genes, each of small effect (the quantitative- abnormalities in those with mental illness are a trait model), results in the observed phenotype. valuable resource: they can help us redefine pheno- Indeed, it is likely that genes of small size effect types, identify candidate genes and refine areas of operate.3 This complexity may help explain why the linkage.3 One difficulty is the uncertain validity of results of linkage studies have sometimes been psychiatric diagnoses, despite the use of standardised difficult to interpret. The initial optimism was diagnostic criteria.4,5 The absence of reliable biologi- generated by positive schizophrenia linkage studies,10 cal or genetic markers specific for schizophrenia or and was tempered by lack of independent replica- affective disorders continues to call into question the tion.11 Although it should be possible to achieve validity of existing classification systems. Linkage to statistically robust linkage results using nonpara- the same chromosomal region has been reported for metric (‘model-free’) approaches, such as the affected both schizophrenia and bipolar disorder at several sib-pair method, the numbers of individuals studied loci, supporting the notion that some genetic risk may have to be increased to new and unprecedented factors give rise to phenotypes that cross the tradi- levels to achieve adequate power. The most recent tional diagnostic boundaries.6,7 There is also evidence genome scans, although not individually generating ‘significant’ results, have shown repeated support for Correspondence: Dr Walter J Muir, Department of Psychiatry, several loci in both schizophrenia12 and bipolar University of Edinburgh, Kennedy Tower, Royal Edinburgh disorder.13 Interestingly, some of these loci appear to Hospital, Morningside Park, Edinburgh EH10 5HF, Scotland, UK. 14,15 E-mail: [email protected] be shared by both disorders, While chromosomal Received 6 May 2002; revised 10 May 2002; accepted 11 July 2002 regions may be consistently identified as suggestive of Chromosome abnormalities and mental illness DJ MacIntyre et al 276 linkage, they are broad genomic regions (B20–30 cM) for linkage (Lod of 3.2) comes from a marker located and need to be further refined. within DISC1, making this one of the most likely Association studies do not rely on knowledge of the current candidates for a susceptibility gene for a mechanism of transmission of a disorder, can detect psychotic disorder. genes of small effect, and can identify narrower In a recent study of families near Barcelona with a regions of interest. However, at present, it is imprac- high prevalence of panic disorder, social phobia and tical to screen the entire genome by association and joint laxity, cytogenetic analysis revealed an inter- therefore candidate genes must be selected. Given our stitial duplication of chromosome 15q24–26 co- limited knowledge of the pathophysiology of mental segregating with illness. Eventually, a Lod score of illness, it is unfortunately possible to construct a 5.0 was generated when those with panic disorder, hypothesis that almost any brain-expressed sequence agoraphobia, social phobia and joint laxity were may be involved in a mental disorder. In any case, the included. These findings were then replicated in 70 human genome sequence is in draft form and at unrelated patients: the chromosomal duplication was present undergoing considerable annotation. It con- present in 97% of individuals with panic disorder/ tains numerous small gaps and, although there is agoraphobia but only in 7% of a control group of 189 preliminary information on gene position and struc- individuals.24 This is the first strong evidence defin- ture, there is little data on predicted function. ing a region of genetic susceptibility for a non- Examination of individuals with chromosomal psychotic psychiatric condition that may affect up abnormalities and mental illness may be a way of to 10% of adults at some time in their life.25 overcoming some of these difficulties. It has been Given the uncertain validity of psychiatric diag- established in many other medical conditions with a noses, the methodological difficulties of linkage and genetic basis that chromosomal aberrations, either by association studies, the proven usefulness of chromo- direct gene disruption or by positional effects, can somal aberrations in medical disorders, and the produce identical or similar phenotypes to those promising findings mentioned above, it seemed that caused by point mutations and their existence has an up-to-date review of reports of chromosomal greatly facilitated the physical mapping and cloning abnormalities that coexist with psychiatric illness of candidate genes.16,17 Unfortunately, chromosomal combined with an analysis of associated evidence, analysis is rarely undertaken in adults with psychia- supporting or otherwise, was worthwhile. In the few tric disorders, unless perhaps they have a concurrent papers published in this area, previous authors have learning disability (this is the UK synonym for mental concentrated on schizophrenia, bipolar disorder or retardation) or a physical dysmorphism. However, the the sex chromosomes.26–28 This report builds on those rate of chromosomal abnormality is substantially studies and takes a broad overview of all chromoso- increased in those with learning disability, and may mal abnormalities reported in relation to psychiatric be as high as 19% in those with mild learning illness. disability.18 There is a well-described and familial link between schizophrenia and mild learning dis- Method ability and an increased rate of chromosome abnorm- alities within this specific population.19 There is also Literature reports of chromosome abnormalities and evidence that the schizophrenia may be the primary psychiatric illness were gathered from Medline (1966 disorder in this link.20 When a chromosomal abnorm- to October 2001), the online database of Chromosomal ality is detected in someone with a psychiatric illness, Variation in Man (www.wiley.com/legacy/products/ it may be considered noncoincidental and related to subject/life/borgaonkar/), and from other related re- the illness if one or more of the following criteria are views.26–28 Medline searching used one or the other of met: (a) the chromosomal abnormality is rare and the following keywords: affective disorders, bipolar there are independent reports of the abnormality disorder, depression, manic depression, mental dis- being associated with psychiatric illness; (b) there is orders, mood disorders, paranoid disorders, psychotic colocalisation of the abnormality with a region of disorders,