Influential Ideas and Experimental Progress in Schizophrenia Genetics Research

Schizophrenia is a devastating disorder affecting 1% of the population worldwide.¹ Typically presenting in adolescence or young adulthood, schizophrenia is characterized by major disruptions of thinking (delusions, disorganization), perception (hallucinations), mood, and behavior.² Symptoms tend to persist without treatment. The elucidation of the biology of schizophrenia will constitute a development of great medical and historic importance. The study of familial schizophrenia was instrumental in opening the field of psychiatry to genetic inquiry, and together with twin and adoption studies helped forge the field of psychiatric genetics. Over the past century, studies have consistently shown that both genetic and nongenetic factors play a significant role in the etiology of schizophrenia.

Currently, a myriad of molecular and genomic database tools, an understanding of complex genetics, and a convergence of results from genetic mapping of several chromosomal regions all contribute to optimism that genes involved in the pathogenesis of schizophrenia will be characterized in the near future. Identification of susceptibility genes, their products, and interacting proteins will likely illuminate pathways to illness, provide more specific pharmacological targets, and lead to improved understanding of environmental contributions to susceptibility. It is hoped that this knowledge will further advance clinical progress in treatment and even prevention of schizophrenia.

During the first part of the 20th century, family studies challenged Kraepelin's early conception of schizophrenia (dementia praecox) as a unitary disorder.³ These studies, which showed that a single gene could not explain the transmission of schizophrenia, contributed to our understanding of the illness. Unfortunately, Nazi geneticists and psychiatrists used this information to justify involuntary sterilization, mass murder, and unscrupulous experimentation upon the mentally ill and others.^{4 - 5} These practices, and involuntary sterilization of the mentally ill on eugenic grounds to a lesser extent in many other countries,⁵ cast a shadow of distrust over the field of behavioral genetics. As a consequence, during the 1950s and early 1960s research in psychiatric genetics was greatly decreased, and purely psychological theories regarding the transmission of schizophrenia became popular.^{6 - 7}

This shift to psychological theories of schizophrenia pathogenesis contributed to delaying the general use of available effective medications and was detrimental to the relationship between the field of psychiatry and the families of patients with schizophrenia.^{6 - 7} Psychological theories regarding the etiology of schizophrenia were rejected when landmark twin and adoption studies^{8 - 9} demonstrated the heritability of schizophrenia, thus inducing a rebirth of scientific interest in psychiatric genetics. The lessons of the past with regard to genetic research have induced a rigorous reassessment of ethical stances of research, exemplified recently by the efforts of the Ethical, Legal, and Social Implications of Human Genetics Research Program in the United States¹⁰ and the German Human Genome Project.¹¹

Familial clustering has been demonstrated in schizophrenia. The rate of schizophrenia in families of schizophrenic probands (first identified ill family member) is increased 4-fold over the rate of 1% observed in the general population.¹² A familial association between schizophrenia, other psychotic disorders, and some personality disorders has been established.¹³ Although most children of a parent with schizophrenia will not have schizophrenia, each child's empirical risk is elevated approximately 10-fold over the general population risk.¹⁴ The concordance rate for schizophrenia is higher in monozygotic than dizygotic twins, with a common ratio being approximately 3:1 or 4:1, and heritability estimates (percentage of variation in presence of disease explained by genetic factors) are approximately 80%.⁸ Family studies have also demonstrated that as the proportion of genes inherited in common from a schizophrenic proband decreases, the risk of developing schizophrenia decreases more rapidly than would be predicted by a single major locus model.¹⁴

Since the concordance rates for schizophrenia among identical twins have been found to be about 40% to 50%, environmental factors must play an important role. There are multiple biological nongenetic hypotheses regarding the etiology of schizophrenia, each with limited and variable amounts of experimental support. Obstetrical complications,^{15 - 16} prenatal infections,¹⁷ and prenatal starvation¹⁸ are several examples of proposed etiological factors. Environmental hypotheses, including biological hypotheses, do not necessarily exclude genetic mechanisms. To separate the contributions of genetic, prenatal, and perinatal events from postnatal environment as contributors to illness, researchers often rely on adoption studies, which have shown that the increased risk of developing schizophrenia observed in children of schizophrenic probands travels largely through genetic pathways.⁹ Segregation analyses, which attempt to fit the observed distribution of illness in families with known models of genetic transmission, suggest a model of genetic transmission of schizophrenia involving few (oligogenic) to many (polygenic) susceptibility genes rather than a simple mendelian inheritance model.¹⁹

Major current molecular genetic strategies used to study the etiology of schizophrenia include systematic linkage analyses or genome-wide scans, association studies, and functional genomics. Linkage analyses exploit the fact that during meiotic recombination, or crossing over between homologous chromosomes, adjacent loci are unlikely to be separated; these loci are usually inherited together and therefore are genetically linked. Proof that genetic linkage analysis can uncover genes for complex disorders was provided by the linkage mapping to chromosome 19q of late-onset Alzheimer disease²⁰ and ensuing positional candidate gene analysis with association studies implicating apolipoprotein E.²¹

Study of the genetic basis of schizophrenia is made particularly challenging by the lack of a standardized phenotypic definition. Schizophrenia and schizoaffective disorder (a combination of manic and/or depressive episodes along with central features of schizophrenia, both for substantial proportions of the overall illness)² are commonly included in disease model definitions as the most severe, or core, phenotypes as a consequence of family studies showing coaggregation of schizophrenia and schizoaffective disorder.²² Some researchers also include diagnoses such as schizotypal personality disorder that are further removed from the core phenotypes but, nevertheless, are within the group of likely alternative expressions of the illness.¹³ Heritable biological traits associated with schizophrenia, which are presumptively closer to the underlying genetic defect and simpler to study, have also been investigated by some schizophrenia genetics researchers. The neurophysiological traits of sensorimotor gating abnormalities²³ and dysfunction of smooth pursuit eye movements²⁴ have been notable foci of such investigations.

Studies using genetic data from multiple independently collected clinical samples show support for linkage of schizophrenia to several chromosomal areas: 1q,²⁵ 6p,²⁶ 6q,²⁷ 8p,²⁸ 10p,^{29 - 31} 13q,³² and 22q.³³ In the largest collaboration on testing candidate areas for schizophrenia to date, candidate regions (30-50 centimorgans) on chromosomes 5q, 6q, 10p, and 13q were investigated with identical genetic markers in more than 1000 affected sibling pairs from 8 independent family collections.³⁴ The data were most supportive for linkage on 6q; logistic regression analysis of linkage allowing for inter-sample heterogeneity produced an empirical P<.001. Some support for linkage was also observed for 10p, with logistic regression analysis of linkage producing an empirical P = .045. Additionally, previous combined analyses have supported 6p,³⁵ 8p,³⁵ and 22q.³⁶ Whether all the chromosomal regions showing linkage support from multiple samples represent definitive linkages (ie, they harbor susceptibility genes for schizophrenia) should be considered unproven until the pertinent susceptibility gene(s) are cloned. Whether current technology, clinical characterization, and size of the linked clinical samples will allow for the cloning of such genes is uncertain.

Association studies are based on linkage disequilibrium, meaning that as generations and meioses proceed over time, an illness gene mutation initially found with specific variants at nearby loci remains statistically associated because the extreme proximity of the variants markedly reduces the occurrence of recombinations between them. The positional cloning of genes for such complex genetic illnesses as late-onset Alzheimer disease^{20 - 21} and type 2 diabetes mellitus^{37 - 38} was accomplished by using linkage disequilibrium strategies to follow up linkage findings with robust supporting evidence. Investigators hope that these strategies can prove useful in schizophrenia genetics research as well.

A growing number of candidate genes, which encode proteins thought to be involved in schizophrenia pathogenesis, are being subjected to genetic analysis by means of association studies. However, in the absence of well-supported and specific pathophysiological hypotheses, the pool of plausible candidate genes is large (in theory, as large as the number of genes expressed in the brain) and their validity is difficult to assess. The dopaminergic hypothesis of schizophrenia,^{39 - 40} which posits that the brains of individuals with schizophrenia demonstrate excessive dopaminergic activity, remains the most tested hypothesis in the field by association and many other means. Although dopamine receptor genes such as dopamine receptor D₂⁴¹ were cloned and examined first, researchers investigating the dopaminergic hypothesis have considered multiple genes as pathophysiological candidates, depending primarily on the extent of the knowledge of the biochemical pathways at the time the candidate genes were selected. After almost a decade of intense scrutiny, no definitive demonstration of an association between a dopamine receptor gene variant and schizophrenia has been made, although the hypothesis has not been definitively disproved.

In the future, testing candidate genes on the basis of pathophysiological theories should improve as a result of the increasing number of single nucleotide polymorphisms that allow for haplotypic analyses,⁴² the availability of genomic sequence from the Human Genome Project, the identification of promoter regions, and continuously increasing sample sizes. The combination of multiple clinical samples will certainly allow for more definitive analyses of proposed associations. To decrease the number of false-positives, investigators should adhere to stringent criteria for statistical significance.⁴³ Some have proposed extending association studies to examine the whole genome,⁴⁴ although this is not yet feasible.

Functional genomics is a new alternative used to search for disease-specific differences in expression in all genes in a disease model (a cell line or an animal model). This approach has limitations; the causative link between altered expression of genes and heritable alterations can be weak, and the weakness of this link is further magnified in complex genetic illnesses. Moreover, neither a cellular phenotype for schizophrenia nor a definitive animal model has been demonstrated. Still, with the advent of gene chips and other microarray technology, this area of research is expanding and shows promise.⁴⁵

As the pace of discovery in schizophrenia genetics is likely to increase, it is hoped that the isolation of susceptibility genes for schizophrenia can be applied to advance research efforts and clinical practice. The identification of genes involved in the susceptibility to and pathogenesis of schizophrenia may provide a more specific definition of inherited phenotypes, advance the study of interactions between genetic and environmental factors, enable detection of disease before the development of overt psychosis, lead to improvement of current antipsychotic medications, and expand pharmacological targets for treatment.