Genetics of Psychosis and Bipolar Disorder Updates PDF

**Key: Prep Slides - Genetics of Schizophrenia and Bipolar Disorder** **Nature/Nurture Debate** - does psychosis run in families? ***Family, twin and adoption studies show that in psychosis:*** - Heritability estimates **\~80%** *(Cardno et al, 2002; McGuffin, 2007; Sullivan et al, 2013)*...

**Key: Prep Slides - Genetics of Schizophrenia and Bipolar Disorder** **Nature/Nurture Debate** - does psychosis run in families? ***Family, twin and adoption studies show that in psychosis:*** - Heritability estimates **\~80%** *(Cardno et al, 2002; McGuffin, 2007; Sullivan et al, 2013)* - Risk increases with genetic relatedness to an affected person - Tells us there are things that are not genetic as risk is still only 50% with identical twins - The increased risk is due to genetic factors, not from sharing the same family environment. **Risk Factors for Psychosis Across the Lifespan** Caused by both environmental and genetic. **The Priest & The Pea -** Gregor Mendel - Reverend Gregor Mendel discovered the basic laws of genetics while studying pea plants in a monastery in the Czech Republic. - His 1866 paper \"Experiments in Plant Hybridisation", established the Mendelian **laws of inheritance**. - Examples of "Mendelian" diseases: Huntington's disease, cystic fibrosis, sickle-cell anaemia\... - In these, inheritance patterns are very predictable - ***But psychotic disorders do NOT follow Mendelian laws!*** **Schizophrenia, Bipolar Disorder and other Psychoses are COMPLEX GENETIC DISEASE and...** - are caused by many genetic risk factors - as well as by multiple environmental factors - and interactions of genetic & environmental influences - do not follow Mendelian patterns of inheritance *Austin et al, 2020* Other complex traits/diseases: - Height, weight, blood pressure, cognitive traits - Diabetes, heart disease - Many common cancers - Many other common diseases\... **Why is Genetics of Psychosis Such a Tough Job?** *Manolio et al. Nature 2009* Can be described via frequency and effect size. In psychosis we have **common variants** that are very common in the population but have very small disease risk. We also have **rare variants** of big effect (called **Copy Number Variants**) **Technology: Arrays & Sequencing** **Genome Wide Association Studies (GWAS)** - Examine \>1 million genetic markers (Single Nucleotide Polymorphisms SNPs) across the whole genome - Very successful in identifying CNVs - Platforms available Illumina and Affymetrix **Whole Genome / Exome Sequencing** - Examines the entire DNA sequence - Exome sequencing studies all segments of DNA sequence coding for proteins (exons). **What can GWAS & Sequencing do?** - No a priori hypothesis needed - Successful identifying loci/genetic variants for many common illnesses with complex genetics like diabetes, Crohn's disease, Parkinson's disease (*Burton et al, 2009; Kraft et al, 2009; Manolio et al, 2008)* **Common Genetic Variation in Psychosis** **Genome wide association study on schizophrenia** - 40+ countries - 150+ institutions - 400,000+ samples **Manhattan Plot Showing Schizophrenia Association** *Trubetskoy et al. Nature 2022* - Sample: \>69,300 people w/schizophrenia and \>236,600 controls - More than 270 loci show strong evidence of genome-wide significant association - 130 genes likely explain these associations **GWAS for Bipolar Disorder** *Mullins et al. Nature Genetics, 2021* **Manhattan Plot - Bipolar Disorder** - 41,900 people with bipolar disorder & \>371,500 controls - x-axis = genomic position (chromosomes 1-22 and X - y-axis = statistical significance (-log10\[*P* value) - Red line = the genome-wide significance threshold (*P* \< 5 x 10\^-8) - SNPs in genome-wide significant loci \~30 - previously associated with bipolar = green - novel associations = yellow - Genes labelled = prioritised by integrative eQTL analyses or notable genes in novel loci (MHC, CACNB2, KCNB1) Example of Key Findings **Implicated gene (name of gene/product)** **Increases the risk for** **Notes** ------------------------------------------------------------------------------------ ---------------------------------- ---------------------------------------------------------------------------------------------------- **DRD2** **(Dopamine D2 Receptor)** Schizophrenia Key target of antipsychotic medications. GWAS now shows significant association with schizophrenia **MHC (Major Histocompatibility Complex)** Schizophrenia & Bipolar Disorder Important for acquired immunity. Epidemiological studies suggest immune dysregulation in schizophrenia. This provides genetic support for this hypothesis. **CACNA1C** **(L-type calcium channel α subunit, type 1c)** Schizophrenia & Bipolar Disorder Important in neuronal function. Mutations cause Timothy Syndrome and Brugada Syndrome. **GRIN2A** Schizophrenia & Bipolar disorder Glutamate receptor subunit. Involved in glutamatergic neurotransmission and synaptic plasticity **ANK3** Bipolar Disorder Codes for protein ankyrin-3 found in nodes of Ranvier of CNS and peripheral neurons. **Key Points on Common Genetic Variants and Psychosis** - Thousands of common genetic variants (SNPs) are associated with schizophrenia and/or bipolar disorder - They are common → \>1% of people carry the risk allele - The Odds Ratio = small (typically 1.1 to 1.2) - Therefore, these SNPs constitute only genetic risk factors for these diseases - There is genetic overlap between schizophrenia and bipolar disorder with many loci increasing the risk for both. - Important for nosology. - Other SNPs are specific to either Schizophrenia or Bipolar Disorder **Rare Genetic Variation → AKA Copy Number Variants** **The Role of Rare Genetic Variation in Psychosis** **CNVs increasing Schizophrenia Risk** *Mowry and Gratten, 2013; Kirov et al, 2014; Rees et al, 2016; Marshall et al, 2017* \| **Locus** \| **CNV type** \| **Gene/s**  **disrupted** \| **OR for schizophrenia** \| **also increases the risk for** \| \| \-\-- \| \-\-- \| \-\-- \| \-\-- \| \-\-- \| \| 1q21.1\* \| Deletion \| several \| 3-8 \| ASD, LD \| \| 2p16.3\* \| Deletion \| NRXN1 \| 8-14 \| ASD, LD \| \| 3q29\* \| Deletion \| many \| 17 \| ASD, LD \| \| 7q36.3 \| DUPLICATION \| VIPR2 \| 4 \| \| \| 7q11.2\* \| DUPLICATION \| \| 16 \| \| \| 15q11.2 \| Deletion \| CYFIP1 and other \| 3 \| ASD, Prader-Willi/Angelman Sd \| \| 15q13.3\* \| Deletion \| CHRNA7 and other \| 10-15 \| ASD, LD, Epilepsy \| \| 15q11.2-13.1 \| DUPLICATION \| many \| 7 \| \| \| 16p11.2 proximal\* \| DUPLICATION \| many \| 9-12 \| ASD, LD, DD \| \| 16p11.2 distal\* \| deletion \| \| 20 \| \| \| 16p13.1 \| DUPLICATION \| several \| 3 \| ASD, LD \| \| 17p12 \| Deletion \| many \| 10 \| Hereditary Neuropathy \| \| 22q11.2\* \| Deletion \| many \| \>35 \| VCFS \| - Loci with genome-wide significant evidence (*Marshall et al, Nature Genetics 2017*) - None of the CNVs are necessary or sufficient to cause schizophrenia - but their OR are large - The same CNV can increase risk for a range of neurodevelopmental conditions: Learning Disability (LD), Autism Spectrum Disorder (ASD), Developmental Delay (DD), ADHD and Schizophrenia - as well as others. **CNVs and Bipolar Disorder** Children with 22q11.2 deletion → have increased risk for bipolar disorder (compared to the population) - However, in bipolar disorder, the role of other copy number variants (CNVs) is not clear - more research needed. CNVs (copy number variants) = part of normal genetic variations - We all carry CNVs that have neutral effects However, CNVs affecting genomic regions, can lead to disease - they can be pathogenic. **Some CNVs confer risk for psychiatric disorders** - Pathogenic CNVs are rare - Frequency ranges 0.001% - 0.1% - (pathogenic = increased disease risk) - CNVs = the strongest known risk factors contributing to schizophrenia - CNVs = not fully penetrant - Meaning that carrying these CNVs increases disease risk, but not everyone who is a carrier develops the disease. **A CNV increasing the risk of Schizophrenia: Velo-Cardio-Facial Syndrome (VCFS) - 22q11 deletion** This CNV affects about 1 in 3,000 people - Not every child with have every feature and the severity varies Common features: - Heart defects (e.g. tetralogy of Fallot, interrupted aortic arch, ventricular septal defects, vascular rings) - Hearing loss - Genitourinary anomalies (absent or malformed kidney) - Hypocalcaemia (low blood calcium levels) - Microcephaly (small head) - Severe immunologic dysfunction (causing frequent infections - DiGeorge Syndrome) - Facial Features (cleft lip/palate, asymmetries, hooded eyelids, small ears, mouth, chin, and side areas of the nose tip) - Learning Disability (usually borderline to mild) - IQ generally in the 70-90s - Some criteria's of 22q11.2 deletion are diagnosed with ASD - Carriers have a substantial increase in risk of developing schizophrenia (OR \>35) **Key Findings and Implications of CNVs in Psychosis** - Pathogenic CNVs = rare - When present → have large ORs for Schizophrenia (ranging for OR = 2 to 30) - CNVs defy diagnostic boundaries - Can increase the risk for a range of neurodevelopmental disorders, incl. 1. Learning disabilities 2. ASD 3. ADHD 4. Schizophrenia - This is known as pleiotropy - Some CNVs increasing schizophrenia risk also increase risk for epilepsy and for some rare genetic syndromes - 22q11.2 deletion cause = velo-cardio-facial syndrome - Incl. intellectual disabilities - Increased risk of developing schizophrenia or bipolar disorder when they grow up **Notes: Genetics of Schizophrenia and Bipolar Disorder** **Schizophrenia Exome Sequencing Meta-Analysis (SCHEMA) Consortium Study** *Singh et al, Nature 2022; Trubetskoy et al, Nature 2022* **Rare Genetic Variants** **Copy Number Variants (CNVs)** - Several rare CVNs found to substantially increase schizophrenia risk - ORs ranging from 2-30 - *(Stefansson et al., 2008; International Schizophrenia Consortium, 2008; Marshall et al., 2017; Li et al., 2020)* - These CNVs represent larger structural variations in the genome. Recent international studies have made significant progress in identifying genetic variants associated with schizophrenia and bipolar disorder: **Common Genetic Variants** - Researchers identified 270 loci containing common genetic variants that individually confer a subtle increase in risk for schizophrenia and/or bipolar disorder - *(Ripke et al., 2014; Ruderfer et al., 2018; Lam et al., 2019; SWGPGC et al., 2020; Mullins et al., 2021)* - Findings come from large-scale GWAS conducted by international consortia. **Single Nucleotide Variants (SNVs)** - The Schizophrenia Exome Sequencing Meta-Analysis (SCHEMA) consortium, *included contributions from UCL*, recently identified SNVs in 10 genes - Showing significant associations with Schizophrenia **Ultra-Rare Protein-Coding Variants in 10 Genes, Which Confer Substantial Risk for Schizophrenia** **Exome Sequencing for Bipolar Disorder** Samples of Bipolar is not as big as Schizophrenia so less genetic variation identified. *Palmer et al, Nature Genetics 2022* - Evidence that variation in this gene increases risk of bipolar disorder **Can We Predict Who Will Develop Psychosis?** Predictions are not possible → but we can estimate a person's Polygenic Risk Score - being done in lots of clinical fields to help alert for later illness - before developing an disease **Polygenic Risk Scores (PRS) Measure "Genetic Predisposition"** *Purcell et al, 2009; Visscher and Wray, 2015* 1. From the Psychiatric Genomics Consortium GWAS data 1. We obtain a panel of SNPs that convey risk for Schizophrenia / Bipolar Disorder → we calculate the OR and p value for each SNP 1. For each participant in our independent study, we calculate their PRS for Schizophrenia & Bipolar Disorder *By adding all risk SNPs weighted by the OR, number of alleles and p value.* **Schizophrenia Polygenic Risk Scores** Highly significant group differences in scores (p=6.3x10\^-40) *very small p value* - The score explains 9% of the variance in Schizophrenia risk - Only taking into account common genetic variants - Low end = low risk - High end = high risk **Bipolar Disorder Polygenic Risk Scores** PRS for Bipolar Disorder in the same sample (1,473 control & 1,170 cases) - Highly significant group differences in scores (p=7.3x10\^-11) - The score explains 2% of the variance in disease risk - More overlap than Schizophrenic PRS → as fewer genes have been identified - The better the dataset, the better the PRS will be **Polygenic Risk Scores: Interpretation and Meaning** PRS can distinguish patients vs controls very well. - The PRS for Schizophrenia does a better job - Due to the larger Schizophrenia sample in the PGC study (more data) - PRS are NOT accurate enough for clinical use but great research tool **What we don't know yet...** - 100s of common genetic variants, several CNVs and Single gene mutations identified, with compelling evidence - though only about 10 genes in schizophrenia **Models with Polygenic Risk Scores ROC Curves** *Calafato et al British J Psychiatry 2018* - Receiver Operating Characteristic (ROC) curves measure the accuracy of a diagnostic test. - A poor test will classify correctly 50% of the time (grey line). - The area under the curve (AUC) measure accuracy. - A good screening test should get it right \>85-90% of the time. - The PRS is NOT accurate enough to use as screening. **How Do We Follow Up The Current Genetic Findings?** **Endophenotypes** Required for a marker of genetic risk, all is required: - A trait associated with the disease of interest - Found in unaffected relatives of cases - A heritable trait **Intermediate/alternative phenotypes are:** - Quantitative measures of brain function/structure - Associated with the disease - Heritable - Not invasive and inexpensive **The Psychosis Endophenotypes - International Consortium** - EEG measures - Symptom measures **Biomarkers Characterising Psychosis** **Polygenic Risk Scores for Schizophrenia: Association with Endophenotypes** - Changes in EEG, Cognition and Brain volumes - Found and association with block design and the PRS for Schizophrenia - Doing less well at the block design performance - Interpret: a high PRS for Schizophrenia can lead to problem with cognitive development later on **CNV Influences on Cognition** *Thygesen et al, Molecular Psychiatry, 2020* CNV-carrier patients had poorer spatial awareness and verbal learning than non-carrier patients. The same occurred amongst controls (control CNV carriers had poorer cognition than non-carrier controls, not shown in graph). Illustrate the impact of these CNVs on cognitions. **Endophenotypes Could Help:** - Understand disease mechanisms - What is the role of certain SNPs or CNVs in brain functions/structure? - Improve disease classifications (nosology) - Stratify patients for personalised treatment **The Clinical Implications of Genetics in Psychosis** **What do genetic findings mean for disease classification?** - Some SNPs confer risk specifically for Schizophrenia or Bipolar Disorder - Several SNPs increase risk of both - Genetic data supports "dimensional" view of psychosis - Current diagnostic systems (ICD-10, DSM-V) still based on phenomenology, not genetics - Creates discrepancy between genetic findings and clinical classification **If the odds ratios for common variants are so small, why do they matter?** - SNPs with small effects can be valuable targets for drug development in other diseases - E.g. DRD2 (dopamine receptor D2) - key target of antipsychotic drugs - The success of DRD2 as a drug target, demonstrates the importance of understanding the underlying biology - Gaining knowledge of the biological mechanisms = crucial for developing new medications **Why is this important?.** ***Because Families Want to Know About Risks and Ways to Reduce Them*** **Genetic Counselling:** Promoting early intervention and detection. A communication process which deals with human problems associated with the occurrence, or the risk of occurrence, of a genetic disorder in a family *American Society of Human Genetics, 1975* **Aims to help individuals or families to:** - Understand information about a genetic condition - Appreciate the inheritance pattern and risk of recurrence - Understand treatment options available - Make decisions appropriate to their personal and family situation - Make the best possible adjustment to the disorder or risk - Advise on changes that could help to reduce risk **Not the same as genetic testing** - In the UK, there is no genetic testing for psychosis in routine clinical practice. Genetic counselling is needed in mental health - can be done before and/or after testing - There is demand from patients and their families - Specialised services are available in Canada (*Austin et al, 2020)* - And Wales → AWMGS - All Wales Psychiatric Genomics Service **Genetics influences how we respond to Treatment** **Why Pharmaco-Genetics?** - 20 different antipsychotics - Treatment is empirical (drug selection & doses - trial and error) - Genetics influences **drug response** and **side effects** - *De Leon & Spina, 2016; Walden et al, 2015* **Genotype to Phenotype** Genetic variation → Enzyme Activity → Metabolic Phenotype - Can make predictions to how someone will respond to a medication. - Saving time and money for the health care system **Optimising Psychosis Treatments by Including Genetics - OPT in Genetics** GEMS - Genetic and Environment in Mental Health Study (UK) **Early findings** - 430 patients recruited - 35% self-identify as BAME - 93% of participants have at least one genetic variant to inform prescribing - 41.9% = Two actionable variants - 31.5% = One actionable variants - 17.6% = Three actionable variants - 6.8% = No actionable variants - 2.2% = Four actionable variants **To Conclude:** - The **vast majority** of people with psychosis **get better** with **medication, psychological** and **social interventions** - We have a **range of effective drugs** - **Understanding the biology is key to develop new treatments**