A major challenge throughout the COVID-19 pandemic has been effectively
treating patients with SARS-CoV-2 due to the high clinical variability
observed. Why have some patients presented as completely asymptomatic, while
others have ultimately lost their lives to the virus? Across the globe,
researchers have been conducting genomic studies in large numbers of
COVID-19 patient and non-patient samples to see if the answer lies in our
DNA.
By reading and analyzing human genomes, we can search for differences, or
variants, in the DNA code (genotype) across populations that may contribute
to observed phenotypes. An example of a phenotype is an individual's
susceptibility to specific diseases, i.e., COVID-19.
In 2020, two separate genome-wide association studies (GWAS) by Ellinghaus
et al and Pairo-Castineira et al identified a particular region of DNA on
chromosome three that appeared to be associated with severe forms of
COVID-19. However, the mechanisms by which this region of DNA conferred the
increased risk were unclear from these initial studies. New research from
Professors James Davies and Jim Hughes at the University of Oxford’s
MRC – published in Nature Genetics – has used artificial intelligence (AI)
to shed some light.
What is a GWAS?
A GWAS study is a method used in genomics to scan for markers in DNA or
entire genomes that are associated with particular traits, such as disease.
The researchers say that the genetic signal on chromosome three has proven
difficult to analyze thus far because it impacts a part of the genome often
referred to as "dark matter" or "junk" DNA. This part of our DNA make-up
earned such names due to the fact it contains introns, genes that do not
encode proteins. For many years, the purpose of this non-coding region
remained elusive, however growing research is demonstrating its importance
in gene regulation, i.e., turning specific genes "on" and "off". Variants in
this region therefore lead to differences in the genes that are expressed in
specific cells.
Studying gene expression with spatial context
Davies and colleagues trained an AI system to analyze large amounts of
genetic data from hundreds of different cell types in the body, which
revealed that the signal is most likely impacting lung cells
"We found that the increased risk is not because of a
difference in gene coding for a protein, but because of a difference in the
DNA that makes a switch to turn a gene on. It’s much harder to detect the
gene which is affected by this kind of indirect switch effect," Hughes said
in a news release.
The gene being upregulated by the sequence on chromosome three is leucine
zipper transcription factor like 1 – or LZTFL1 – which was "surprising" to
the research team as it has not been largely studied in the past. They
conducted spatial transcriptomic analysis, which is a novel method that
measures gene activity in a specific tissue sample (in this case, lung
biopsies from patients with COVID-19). The analysis detected signals that
are associated with an infective response known as epithelial-mesenchymal
transition (EMT) that is upregulated by LZTFL1. This finding could explain
the link between the genetic signals on chromosome three and increased risk
of severe disease in carriers of the variant.
"Higher levels of LZTFL1 may delay the positive effects of an acute EMT
response, blocking a reduction in ACE2 and TMPRSS2 levels and/or through
slowing EMT-driven tissue repair. Further investigation of the potential
role of LZTFL1 and EMT in pulmonary pathogenesis is needed. Our findings
suggest that a gain-of-function variant in an inducible enhancer, causing
increased expression of LZTFL1, may be associated with a worse outcome," the
authors write in the paper.
In an official news release from the University of Cambridge, the
researchers said that they do not anticipate the variant causing any issues
in vaccine response, as it affects the cells lining the airways and the
lungs, and not the immune system.
Prioritizing individuals for vaccination
The findings of this study could have important implications for developing
novel treatments for COVID-19. "The genetic factor we have found explains
why some people get very seriously ill after coronavirus infection. It shows
that the way in which the lung responds to the infection is critical. This
is important because most treatments have focussed on changing the way in
which the immune system reacts to the virus," Davies said.
It also could help to predict those that are at an increased risk across the
globe. Fifteen percent of individuals with European ancestry carry the
high-risk version of the gene, vs. 60% of people with South Asian ancestry.
Frances Filter, professor emeritus at King's College London said, "This is a
very interesting publication. The discrepancy between the risk of serious
disease and death in different ethnic groups has previously been attributed
in part to socio-economic differences, but it was clear that this was not a
complete explanation."
She added, "Evidence that a relatively unstudied gene, LZTFL1, has emerged
as a candidate causal gene, which is potentially responsible for some of the
twofold increased risk of respiratory failure from COVID-19 in some
populations, provides a big step forward in our understanding of the
variable susceptibility of some individuals to serious disease and death."
This work might also be used to direct vaccination efforts. While we cannot
directly change our genetic code, we could potentially priorotize immunizing
those that carry the genetic signal to ensure that their increased risk is
counteracted by the vaccine. “Vaccine uptake has been high in South Asian
groups but this study reinforces the importance of taking the booster doses
now to maximise their protection and reduce their risk as immunity is now
waning," Dr. Raghib Ali, senior clinical research associate at the MRC
Epidemiology Unit at the University of Cambridge, said.
Reference:
Downes DJ, Cross AR, Hua P, et al. Identification of LZTFL1 as a candidate
effector gene at a COVID-19 risk locus. Nat Gen. 2021;53(11):1606-1615.
DOI: 10.1038/s41588-021-00955-3.