LONDON – Computational biology specialist Precisionlife Ltd. has used UK Biobank data to find sepsis risk genes that are present specifically in patients who suffer severe COVID-19 infections and shown that 13 of those genes are known druggable targets.
Based on that, a list of 59 compounds has been assembled with potential for repurposing as treatments for COVID-19 patients who develop sepsis, while other previously unknown genes could form the basis for de novo discovery.
“Ours is the first study looking at host genomics and opportunities to treat later severe disease where host immune processes take over,” said Steve Gardner, co-founder and CEO of Precisionlife.
As yet, there is insufficient genetic data available from confirmed COVID-19 patients to link genomics to disease severity and outcome, co-morbidities and treatments, meaning a direct study of the impact of host genomics on the course of the illness is not possible.
Instead, Precisionlife applied its combinatorial technology to analyze a population of 6,820 people in the UK Biobank who had experienced sepsis, as a surrogate. The available genotype data included 542,245 single nucleotide polymorphisms (SNPs).
Rather than a standard genomewide association study (GWAS) looking for single SNPs that occur more frequently in patients than controls, the Precisionlife technology finds and statistically validates combinations of three to 10 SNPs in combination, that together are strongly associated with a specific disease diagnosis.
Those sepsis disease signatures are then mapped to the human reference genome to identify the disease-associated genes.
Mapping the highest-scoring SNPs uncovered 70 genes that are strongly associated with the risk of developing sepsis. Of those, several previously have been implicated with sepsis pathology, which Precisionlife said validates its approach.
However, the findings significantly increase the number of sepsis-related genes, with a recent meta-analysis of published studies pinpointing only 23 genes that have been shown to have a significant association with disease susceptibility.
Sepsis has been observed in up to 59% of COVID-19 patients who were ill enough to require hospital treatment. To assess the relevance of the genes uncovered through analysis of the historical sepsis cohort, Precisionlife cross-referenced them with data from 572 volunteers in the UK Biobank who have been hospitalized with COVID-19, and 97 volunteers who tested positive in the community.
A standard GWAS analysis on the UK Biobank dataset revealed no significant SNPs. That is in line with the first meta-analysis reported last week by the international COVID-19 Host Genetics Initiative, based on data from 917 COVID-19 patients.
When the UK Biobank COVID-19 dataset was analyzed with the Precisionlife datamining tools, 33 sepsis disease signatures were found in the 572 hospitalized patients that were not found in the COVID-19 patients showing mild symptoms, or in any sepsis controls.
“If you do a GWAS on a sepsis cohort or a COVID-19 cohort, you don’t find any features that distinguish severe sepsis,” Gardner told BioWorld. “Our high resolution genomic analysis tools have allowed us to develop new insights.”
Initial investigations have linked those COVID-19 SNP disease signatures to neutrophil degranulation, leukocyte activation and immune effector processes. Long Hanborough, U.K.-based Precisionlife is now conducting an in-depth study of the functional impact of the SNP signatures found in severely affected COVID-19 patients.
Moving research forward
The 70 risk genes found in the UK Biobank sepsis cohort fall into four categories, all of which can be linked to known sepsis pathology.
As to be expected, there are a number of genes relating to the innate and adaptive immune response, indicating patients who develop sepsis may have inherent aberrations in how their immune systems responds to infectious agents.
For example, one gene encodes for a T-cell differentiation antigen that binds to key components of gram-positive and gram-negative bacteria, causing agglutination and inhibiting their virulence factors. Patients with variants in that gene are likely to have defective pathogen binding abilities and greater susceptibilities to infection by sepsis-causing bacteria, according to Precisionlife researchers.
A second group of genes is involved in endothelial inflammation, a mechanism known to contribute to sepsis pathology, while a third group is associated with apoptosis-related pathways known to be important in regulation of the immune system. Meanwhile, the fourth group of genes is associated with neuronal signaling pathways that play a role in regulating the immune system and in cytokine expression.
Among the sepsis and COVID-19 risk genes, there are 13 which are known to be druggable. The researchers went on to identify 59 compounds and drugs that are known to be active against those 13 targets.
Precisionlife published its finding on the preprint server Biorxiv. Gardner said the company will be working with international collaborators to take the research forward. “We are reaching out to key opinion leaders and other groups working in sepsis that have existing models and assays we can test our findings in,” he said.
As more COVID-19 patient data become available in the UK Biobank and other patient data sources, Precisionlife said it will analyze the clinical impact of the disease signatures in a larger group of patients. The company also is making its tools available to other research groups working on COVID-19 host genomics.