Integrative genetics of behavior with high throughput technologies

National Institute On Alcohol Abuse And Alcoholism

Investigators

Linked publications, trials & patents

Abstract

Identification of functional loci is key to the Maps to Mechanisms to Medicine (M2M2M) paradigm articulated by the International Common Disease Alliance and that has been LNG's focus since 1992. Alcohol Use Disorder (AUD) and Substance Use Disorders (SUDs) are moderately heritable but most of the variance is unexplained. We discovered functional loci with roles in AUD via founder populations, people at risk, model organisms and intermediate phenotypes closer to gene action. Hypothesis-free omic approaches including genome wide association studies (GWAS), transcriptomics, methylome and exome sequencing identified targets for functional studies, and elucidated genetic architecture and transmission. Via methylome, aging was observed to be accelerated by alcohol, stress and pain (Jung et al, 2022; Spagnolo et al, 2023). Via GWAS, we found loci for alcohol response (Lai et al, 2020), AUD (Lai et al, 2021), AUD treatment response (Biernacka, 2021), and alcohol cirrhosis (Schwantes-An et al, 2020, Whitfield et al, 2022). We found 3 exome wide significant genes in Alcohol Hepatitis (Yuan et al, 2024). We found that induced hepatocyte like cells (iHLCs) from Alcohol Cirrhosis patients have intrinsic metabolic and transcriptomic disturbances (Mukhopadhyay et al, 2024). Via transcriptomics we advanced understanding of glia in SUDs (De Biase, 2017), cirrhosis (Guillot et al, 2021), developmental effects of nicotine (Jung, 2016) and Fetal Alcohol Effects (Sambo et al, 2022;2023). Via transcriptomics we identified pathways and genes contributing to intrinsic differences in ovarian steroid response in Premenstrual Dysphoric Disorder, Perimenopausal Depression and Perimenopausal Depression (Dubey, 2016; Rudzinskas, 2020;2022;2023;2023;2024 Li et al, 2021). Via polygenic scores, in the representative NESARCIII sample, we found AUD is cross-inherited with nicotine use disorder whereas inheritance of opioid use disorder is largely independent (Jung et al 2021). We made progress in identifying functional loci contributing to AUD. OPRM1 Asn40Asp (Bergen et al, 1997) was found to be functional and linked to AUD naltrexone treatment response (Anton et al, 2008). We described a common, functional SNP in the serotonin transporter (Hu et al, 2007) enhancing understanding of OCD, neuroimaging responses to emotion, and gene x stress interaction. Common HTR2C Ser23Cys (Lappalainen et al, 1999, Okada et al, 2004) and HTR2A Asn452His variants (Ozaki et al, 1997) were found, shown functional, and linked to behavior. We found that a DRD2 promoter variant influences antipsychotic response (Lencz et al, 2006). We traced linkages of functional haplotypes of NPY (Zhu et al, 2008), GCH1 (Tegeder et al, 2006) and DISC1 (Hodgkinson et al, 2004) to emotion, schizophrenia, and pain. The low expression Neuropeptide Y (NPY) haplotype increased anxiety and emotionality but had stronger effects on NPY RNA and protein and brain imaging responses to emotion and pain/stress (Zhou et al, Nature, 2008). Via in vitro cell models, namely lymphoblastoid cell lines (LCLs), iPSC's and cells differentiated from iPSCs such as induced Hepatocyte-Like Cells (iHLCs) we found intrinsic differences in cellular function in both genetically influenced alcohol liver disease and reproductive mood disorders (RMDs). The liver studies, collaborative with Arun Sanyal (VCU), revealed that Alcohol Cirrhosis iHLCs formed fat droplets due to inherent metabolic impairments detected by several methods including transcriptomics and oxygen consumption rate measured after specific inhibitors of mitochondrial respiration (Mukhopadyay, Nature Communications). The RMD studies, collaborative with Peter Schmidt (NIMH) revealed that women with RMDs had alterations in specific pathways including the ESC/EZ (polycomb repressor)complex of genes (Rudzinskas et al, various). We proposed the Addictions Neuroclinical Assessment (Kwako et al, 2017), recovered the theorized 3 factor structure in the intramural natural history sample (Kwako et al,2019) and validated the ANA framework prospectively (Gunawan et al, 2024). The imaging genetics paradigm we helped initiate (Heinz et al, 2000; Hariri et al, 2002; Egan et al, 2001; 2003; Zubieta et al, 2003) led to groundbreaking findings. CHRNA5 Asn398, an allele altering nicotine addiction risk, weakened connectivities including Dorsal Anterior Cingulate/Ventral Striatal connectivity predictive of craving (Hong et al, 2010). Clinical subphenotyping enabled linkage of HTR1B to antisocial AUD (Lappalainen et al, 1998), serotonin transporter (SLCA4) to anxiety (Mazzanti et al, 1998; Hariri et al, 2002), BDNF Val66Met to episodic memory (Egan et al, cell, 2003), and GTP cyclohydrolase to chronic pain and pain response (Tegeder et al, 2006). Frontal cognitive deficit is a risk factor in AUD and other diseases, with dopamine generally enhancing prefrontal cortical efficiency. Met158, a common COMT variant that leads to four-fold reduction in COMT activity showed an allele-dosage relationship to frontal cognitive function and better cortical efficiency (Egan et al, 2001). The COMT effect on cognition was seen in populations differing in cognitive function: schizophrenia, head injury (Lipsky et al, 2005, 2011), and controls (Malhotra et al, 2002, 2009). We proposed that Met158 has a counter-disadvantage. In two populations Met158 predicted anxiety in women and decreased frontal EEG coherence (Enoch et al, 2003), and Met158 was associated with lower resiliency to pain/stress (Zubieta et al, Science, 2003; Diatchenko et al, 2005, 2006). We extended knowledge of GxE effects of stress related genes such as COMT, NPY, SLC6A4 and SLC6A4 in studies using both behavioral and endocrine outcomes (Lovallo et al 2015, 2016). A functional polymorphism of Fatty Acid Amide Hydrolase (FAAH) predicted placebo response to pain (Pecina et al, 2014). The Chr 4 GABAA gene cluster was implicated in AUD by family linkage (Long et al, 1998), an effect that appeared anxiety-modulated (Enoch et al, 2006). The GABAA gene cluster on Chr 5 was implicated in AUD and alcohol response (Radel et al, 2005). GABRA6 has a missense variant associated with alcohol dependence and response to alcohol and diazepam (Iwata et al, 2005). Family linkage yielded genome-wide significant linkage of CRF-BP to an AUD-associated EEG trait as was supported in two populations (Enoch et al, 2008). In GWAS of EEG (Hodgkinson et al, PNAS, 2010) in Plains Indians we detected genome-wide significant loci (Hodgkinson et al, PNAS 2010). Via GWAS, a schizophrenia risk locus was identified (Lencz et al, Nature Commumications, 2013). Early life stress and heavy alcohol exposure individually led to epigenetic age acceleration of about two years each and as was additive (Jung et al, Biol Psych). In mice, embryonic alcohol exposure predicted accelerated (or premature) pace of neuronal differentiation and as such could account for smaller brain and head sizes that are a hallmark of fetal alcohol spectrum disorder (Sambo et al). Founder, phenotypically extreme, and exposed populations enhance power to detect genes. An African American SUD sample revealed GxE of childhood trauma and HTTLPR in suicidality (Roy et al, 2007). Stress and poverty, but not ancestry, predicted SUDs (Ducci et al, 2009). An MAOA VNTR previously linked to dyscontrol via stress interaction was linked to AUD and ASPD in Native American women, half of whom were sexually abused as children (Ducci et al, 2008). A strong interaction between an MAOA VNTR and testosterone on aggression was observed (Sjoberg et al, 2008). We found an HTR2B stop codon that is common in Finns, absent in others, and associated with, but not determinant for, impulsive behavior. It cosegregated with ASPD and AUD in families and the mouse Htr2b knockout was hyperaroused (Bevilacqua et al, Nature, 2010). In Native Americans, recurrent CNVs (rCNVs) were discovered in two populations, shown to be ancient, and in several instances shown to lead to haploinsufficiency or excess of expression of genes within the CNV. An rCNV of the Chr 12 Velocardiofacial Syndrome region was linked to AUD and "any psychiatric disorder." In animal models, we discovered a stop codon in metabotropic glutamate receptor 2 (Grm2) that leads to alcohol preference in the alcohol preferring P rat (Zhou et al, 2013). The Grm2 Stop codon is common or fixed in parental Wistar rats, depending on source. It leads to uncompensated changes in glutamate function, and concerns about neuroscience studies in which this stop codon is an uncontrolled variable. We identified loci altering drug preference in rat and Drosophila. We re-discovered the same Grm2 stop codon in a Wistar-derived cocaine model. We found 7 genome wide significant loci accounting for 2/3rds of variance in novelty induced hyperlocomotion, a hyperarousal model predicting drug use (Zhou et al, PNAS, 2019). In wsp/wsr mice, we discovered that genes implicated in epilepsy and neuroexcitability are critical in alcohol withdrawal (Zhou et al, 2022). In Drd2 iMSN partial knockout mice, we found that deficiency of D2 receptors drives gene networks related to GABA, cAMP, growth and neuroinflammation (Guerri et al, Biol Psych, GoS in press), helping to explain why DRD2 is linked to SUDs in people.