1. List the possible applications of genetics to psychiatry.

2. Explain why bipolar disorders are considered part of a spectrum of related mental illnesses.

3. Describe current research into the genetics of bipolar-spectrum disorders and into treatment response.

4. Discuss the pharmacogenetics of bipolar-spectrum disorders.

5. Incorporate genetic information into clinical practice.

John R. Kelsoe, MD
Professor of Psychiatry, University of California, San Diego, School of Medicine, La Jolla, CA

Introduction: pharmacologic treatment of bipolar disorder associated with challenges; limited or incomplete response in 30% to 50% of patients—insufficient knowledge of pathophysiology limits rational drug design; current medications access only limited number of disease targets in brain; limited predictive value of diagnosis—limited longitudinal stability of diagnostic criteria; limited ability to predict treatment response; probably bears limited relationship to underlying biologic disorder; individual variation in treatment response—currently, no effective method to predict response; treatment process based on trial and error; antidepressant-induced mania

Possible applications of genetics to clinical practice: susceptibility genes—elucidate disease mechanisms; find novel targets for drugs that have different mechanisms of action; mechanism-based diagnostic system— treatment more specific to diagnosis; may “cut across” (ie, complement and supplement) current syndromic diagnoses; pharmacogenomics—genetic component to individual variation in response to treatment; identification of gene variants involved in drug response may enable panel of DNA tests to predict drug response; personalized psycho-pharmacology

Bipolar disorder is diagnostic challenge: survey shows 69% of patients misdiagnosed initially; time to accurate diagnosis averages 7 yr from onset of symptoms; during that time, patients see average of 4 physicians and receive average of 3.5 inaccurate diagnoses (most commonly depression, followed by anxiety disorder, schizophrenia, borderline personality disorder (BPD), and antisocial personality disorder)

What is the bipolar spectrum? bipolar disorder appears to have relationship to many other psychiatric disorders, including attention-deficit/hyperactivity disorder (ADHD), panic and anxiety disorders, major depression, migraine, BPD, and schizophrenia, but those relationships unclear; relationships between bipolar II disorder, bipolar I disorder, and major depression suggest quantitative aspect to illness

Is bipolar disorder genetic? familial (studies show that family members of person with bipolar disorder are at 7-fold increased risk of having affective illness, compared to general population); ≈65% of identical twins share bipolar disorder (concordance rate in fraternal twins, 15%); however, environment also contributes, suggesting disorder not wholly genetic (currently estimated that bipolar disorder 65% genetic); data show genetic overlap between subforms of bipolar disorder (Danish study showed bipolar-bipolar and unipolar-unipolar patterns most common in twins, “but there was a significant occurrence of bipolar-unipolar”; spectrum disorders are partially genetically distinct

Genetics of bipolar II disorder: what is relationship between bipolar II and bipolar I disorders? bipolar II occurs more often than bipolar I in families of bipolar II probands, suggesting partial genetic overlap between these disorders; literature also suggests that other forms of bipolar disorder may be somewhat genetically distinct, eg, study found elevated rate of psychotic mania in families of probands with psychotic mania; bipolar disorder with anxiety and suicidality also appear to run in families; different forms of illness and presentations, as well as different degrees of severity of illness

Genetics of bipolar temperament: questionnaire found more dysthymia, cyclothymia, hyperthymia, and anxiety in bipolar II than in bipolar I; questionnaire scores of unaffected relatives approximately midway between scores of ill relatives and scores of unrelated controls; controls scored higher in hyperthymia, which may reflect cultural biases favoring attributes such as getting by with less sleep, and being talkative, charming, and charismatic

Bipolar spectrum partially overlaps with other psychiatric disorders: according to study, relatives of probands with schizophrenia have increased rates of schizophrenia and bipolar disorder; relatives of probands with mania have increased rates of mania and schizophrenia; studies in twins also point to genetic overlap between bipolar disorder and schizophrenia

Genetic features of bipolar spectrum: family members have 7-fold increased risk for illness; only ≈65% of what causes bipolar disorder inherited (“reduced penetrance”); milder forms of bipolar phenotype frequently found in relatives of bipolar probands; some aspects of bipolar spectrum more like quantitative traits, while other traits of bipolar spectrum are, in part, genetically distinct; bipolar spectrum overlaps with other psychiatric disorders

Models of genetic transmission: single major locus effect—one gene of large effect transmits trait; Mendelian patterns of transmission; heterogeneity (different genes in different families); polygenic or multifactorial transmission—many genes each contribute small effect; traits are quantitative, additive, and/or epistatic; mixed model—“almost certainly what’s going to play out”

Polygenic quantitative traits, multifactorial threshold model: spectrum of illness ranges (in order of severity) from affective temperament to major depression to bipolar II disorder to bipolar I disorder to schizoaffective disorder to schizophrenia; the more bipolar genes an individual inherits, the farther they move along this spectrum; in practice, however, model only partially fits observed data

Why does bipolar disorder exist? if deleterious to species, would be expected to eventually disappear; however, rate of ≈1% “remarkably constant around the world”; additionally, vulnerability factors very old, dating far back in human evolution; theory of balanced selection—posits that some of these alleles have both advantages and disadvantages; therefore, retained in given determined frequency that is optimal within environment; if model true, then most genes for bipolar disorder carried in population by individuals who do not have illness, ie, these genes modulate something else, and bipolar disorder probably uncommon side effect; speaker’s observation—bipolar disorder is “social,” ie, many of traits related to human–human interaction; speaker posits that there is advantage to social group to have certain mixture of personality styles (“if everybody were manic they’d all kill each other; but if everybody were depressed, then nothing would ever happen”)

The universe of bipolar genes: according to one hypothesis, large set of genes predisposes to depression, which is ≈5 times more common than bipolar disorder; also, genes that predispose to bipolar I, bipolar II, or bipolar III disorder (variously defined as unipolar disorder that for some reason resembles bipolar disorder); instead of asking what these disorders have in common, research must ask how they differ; theories include allelic heterogeneity (one gene with 2 mutations; one mutation results in bipolar disorder, other in schizophrenia), gene–gene interactions (“one gene screws up your brain in a relatively nonspecific way”), and gene–environment interactions (nonspecific set of genes that results in bipolar disorder in one type of environment and in schizophrenia in different type of environment)

How do we find those genes? Human Genome Project has sequenced all 3 million base pairs in human genome, establishing that humans have ≈18,000 genes; second stage of Human Genome Project attempting to identify which base pairs vary among human population; simplest type of genetic variant single nucleotide polymorphism (SNP), >20 million of which have been identified; HapMap Project (haplotype of human genome) has found that variation not random across genome, but tends to be bunched together in haplotype blocks; first step in genetic mapping involves linkage (determines which regions of chromosomes inherited along with illness; resolution 5- 20 million base pairs); association studies determine whether variation occurs more often in people who have specific illness (resolution 5-25 thousand base pairs); with further study, resolution eventually narrowed to 1 base pair and allows for functional testing; success achieved when variant found that affects function of gene and that is associated with illness

Mapping gene for bipolar disorder on chromosome 22: genomes of 20 large families with bipolar disorder screened, with ≈400 markers spaced evenly across all chromosomes; evidence for gene on chromosome 22; candidate gene GRK3 (G protein receptor kinase 3), which mediates homologous desensitization, identified as suspicious; after sequencing, SNPs found in promoter of GRK3; this mutation occurred more frequently in people who had bipolar disorder than in those who did not, and affects function of GRK3 (but in manner opposite from that expected); speaker’s group trying to develop these findings in animal model

Genome-wide association studies (GWAS): association has greater power to detect genes with small effect; common disease, common variant (CDCV) model asserts that if disease common, then alleles that cause it also probably common (accuracy of model questionable); screen entire genome with high density of markers (microarray technologies make this very inexpensive; large number of SNPs identified by HapMap Project; one million SNPs now routinely screened); technique has little power if many rare SNPs of strong effect involved; results to date— limited success, possibly due to inadequate sample size (goal to screen 500,000 markers in 10,000 patients; only 4000 patients screened to date); CDCV model may be wrong model (are there, instead, numerous rare strong mutations?); uncertain what role copy-number variation plays; meta-analysis of all bipolar GWAS data worldwide now under way

Summary of bipolar genetics: genes explain ≈65% of cause of bipolar disorder; affective temperaments and bipolar-spectrum disorders genetically related to bipolar disorder; some bipolar genes likely play role in other psychiatric disorders; numerous genes likely interact to produce spectrum of bipolar-related traits; GWAS promises discovery of many new genes, which may lead to new system of diagnosis and new, more specific treatments

Pharmacogenetics: science of identifying genes that influence drugs; lithium—currently under study because it has best clinical predictors of response; first mood-stabilizing medication for bipolar disorder; strongest evidence for efficacy; evidence for reduction of suicide risk; inexpensive; side effects include weight gain, acne, polydipsia, and hypothyroidism; large number of patients have excellent response to lithium; largely supplanted by val- proic acid in United States; great need in clinical practice for method that identifies good responders to lithium

Clinical predictors of response to lithium: family history of bipolar disorder; predictors of poor response to lithium include mixed state or dysphoric or irritable mania

Lithium response may be genetically distinct: interindividual variation in response to mood stabilizers; response associated with positive family history; response to lithium may be familial; lithium-responsive bipolar disorder may represent genetically distinct form of illness

Speaker’s study: found association between lithium response and variation within gene NTRK2; people with dysphoric mania had poor response (30%) to lithium; those with euphoric mania with 1 copy of variant had ≈60% response, and those with 2 copies had 80% response; speaker concludes that in patients with euphoric mania, number of mutations in NTRK2 gene robustly predicts probability of response to lithium

What is NTRK2? codes for tyrosine kinase receptor Trkb (receptor for brain-derived neurotrophic factor [BDNF]); BDNF involved in brain development and neuronal maintenance; antidepressants and lithium increase BDNF; BDNF necessary for antidepressant and lithium action in BDNF-knockout mouse; hypothesis that increased glucocorticoids during stress toxic to neurons, and antidepressants and lithium work by increasing BDNF and repairing neurons

Bipolar clinic of tomorrow: personalized psychopharmacology; patient provides DNA in sputum or blood for chip-based assay of hundreds of SNPs, which may result in biology-based diagnosis and prognosis and prediction of response to medication

Clinical and treatment implications: new understanding of basic pathophysiology of bipolar disorder; reduction of stigma associated with mental illness; new methods of diagnosis, with new classification of diseases based on pathophysiology (which may subdivide or cut across behavioral definition of diseases); new medications with novel targets; pharmacogenomics can be used to guide treatment decisions; gene therapy has not lived up to its promise; ethical issues related to risk testing