NEW YORK (GenomeWeb) – Based on genome-wide association study data, researchers from the University of Hong Kong have developed a list of drugs that could be re-purposed to treat psychiatric conditions.
In their study, the researchers imputed transcriptomes from GWAS data and compared those transcriptomes to a database of drug-induced gene expression profiles. As they reported in Nature Neuroscience today, Hong Kong's Pak-Chung Sham and his colleagues used this approach to unearth existing drugs that could potentially treat half a dozen psychiatric conditions.
"In this study, we developed a framework for drug repositioning by linking two apparently disparate sources of information," the authors wrote, adding that they "identified a number of interesting candidates for repositioning, many of which are supported by animal or clinical studies."
The researchers downloaded eight sets of GWAS summary statistics for studies of seven psychiatric conditions, including schizophrenia, bipolar disorder, and major depressive disorder. Using this variant data, they imputed expression profiles for 10 brain regions. They noted that such imputed data, unlike expression data captured through RNA-sequencing or microarray analysis, isn't affected by whether people in the cohort were already taking medications and that it bypassed the need for post-mortem brain tissue samples.
At the same time, the researchers obtained drug-induced transcriptomes from the Broad Institute's Connectivity Map database. They compared the two sets to look for reverse expression patterns to find drugs that might work for each disorder and brain region.
For schizophrenia, the top hits Sham and his colleagues uncovered included antipsychotics like thioproperazine, droperidol, and triflupromazine. In addition, one of the top 15 hits, idazoxan, had been shown in a clinical study to possibly improve the negative symptoms of schizophrenia.
Likewise, for bipolar disorder, the top hits included antipsychotics as well as drugs that were known or suspected antidepressants. Additionally, cardiovascular drugs like simvastatin and metformin were among the top drugs for bipolar disorder, which the researchers said was in line with the finding that people with bipolar disorder are at increased risk of developing cardiovascular disease.
Meanwhile, for major depressive disorder, the researchers included two GWAS datasets, one from the Psychiatric Genomics Consortium and one from the CONVERGE Consortium. For the PGC set, the top hits included the selective serotonin reuptake inhibitor fluoxetine and antipsychotics like sulpiride, promazine, and perphenazine. At the same time, the CONVERGE set's top hits included the SSRI zimelidime as well as the norepinephrine-dopamine reuptake inhibitor nomifensine and the monoamine oxidase inhibitor isocarboxazid.
The researchers suggested that the lack of overlap in the drugs between the two samples could be attributed to the CONVERGE cohort being made up of women of Chinese ancestry while the PGC cohort included men and women of mostly European ancestry.
Overall, the researchers noted that the drugs suggested for schizophrenia were enriched for antipsychotics and that the drugs suggested for bipolar disorder were enriched for antipsychotics and antidepressants. Since these classes of drugs are already used to treat those disorders, the researchers said that result was "encouraging" and supported the validity of their approach.
"We present a list of repositioning opportunities for each disorder, which we believe will serve as a useful resource for preclinical and clinical researchers pursuing further studies," the researchers added.
Sham and his colleagues also noted that combining their approach with other drug re-positioning methods or studies of what drugs can cross the blood-brain barrier may help whittle down the list of candidate drugs.