Please can you introduce yourself, tell us about your background in medical informatics, and what inspired your latest research into identifying independent variants that significantly predispose to critical COVID-19?
My name is Erola Pairo-Castineira, and I am a postdoctoral researcher specializing in statistical genetics in Professor Baillie's lab at the University of Edinburgh.
I'm interested in the differences in DNA sequence between people and how this affects how sick you get when infected with SARS-CoV-2. Most people are asymptomatic or have mild symptoms when infected, while some people need hospitalization and a few need to be admitted into ICU to receive mechanical ventilation. Although certain factors like age and sex affect your predisposition to critical COVID-19, there are differences even in people with similar age, sex, and other variables affecting COVID-19 severity.
Our research aims to find the genetic factors that may predispose to critical COVID-19 and learn how COVID-19 affects cells in the body. The main objective of the project is to find which biological pathways influence critical COVID-19 to be able to find druggable targets that may benefit everyone.
What is meant by the term 'critical COVID-19', and what mechanisms cause it?
People suffering from critical COVID-19 have been admitted into an intensive care unit. They will have all the classic symptoms of COVID-19, such as a cough and fever, but also very low blood oxygen. The mechanism causing it is inflammation in the lung, where the virus is replicating.
Previous studies have found associations between particular genetic variations and critical COVID-19. What did these findings reveal about genetic predisposition to COVID-19?
In our first study, we found five genetic signals related to critical COVID-19 related to host antiviral defense mechanisms and mediators of inflammatory organ damage. Other studies since then have found signals related to susceptibility to infection with SARS-CoV-2 and hospitalization and critical illness with COVID-19.
In your latest research, you conducted Whole Genome Sequencing. What advantages does this provide over microarray genotyping?
Microarray genotyping reads a set of variants in the genome, and from them, and using information about genomes of known populations, we infer the rest of the genome. This is very difficult to do for rare variants, which will have low accuracy in the inference, and it is also affected by the quality of the reference panel used. Even in the best cases, not all variants can be retrieved with high accuracy, and some of them have to be filtered. Whole-genome sequencing reads every base in our genome, providing high accuracy for all variants.
How did you carry out your genome-wide association study, and what new insights did the current study provide?
To carry out the analysis, we compared the genome of people who had critical COVID-19, focusing on people in intensive care units in UK hospitals, with individuals that had only mild symptoms during SARS-CoV-2 infection, and with the general population. The comparison highlights variants of the genome that affect your probability of becoming critically ill with COVID-19.
In this study, we found 16 new regions associated with critical COVID-19. We were also able to identify some of these variants as variants that affect the function of some proteins (for example, a protein called IFNA10). In other cases, we could show that changes in gene expression would affect the probability of having severe COVID-19.
Five critical COVID-19-associated variants had direct roles in interferon signaling. What is interferon signaling, and what do these associations suggest about critical COVID-19?
Interferons are a large family of proteins with roles in antiviral defense. They are produced when the cell recognizes a viral intruder, and they propagate defensive and offensive signals to help destroy the virus and protect nearby cells. Interferons can induce inflammation by activating the transcription of a number of different genes, and if they are not tightly controlled during an immune reaction, they may cause hyper-inflammation.
There are different types of interferon, called Interferon Type-I, Interferon Type-II, and Interferon Type-III. The genes we have found are mainly associated with Type-I Interferon signaling. Interestingly, one of these genes (TYK2) has a protein product that a widely available drug, baricitinib, can inhibit. It was recently shown in the RECOVERY trial (a large-scale randomized control trial for many potential COVID-19 treatments in the UK) that administration of baricitinib can lower mortality in COVID-19.
Your results provided evidence in support of causal roles for myeloid cell adhesion molecules and coagulation factor F8, all of which are potentially druggable targets. How do you see your findings influencing future therapeutic targets?
In the case of baricitinib, it was lucky that it was already available for rheumatoid arthritis, so it had been proven safe and effective. Usually, the process is much longer when a potential gene target is identified. The most important first step is confirming the associations we see in the DNA data in cells and tissues - in vitro experiments will be required to ensure the target is worth pursuing. An appropriate drug must be identified, tested, and shown to be safe. It can be a long process; however, our analysis can narrow down the most likely drug targets, which is a major step forward.
What were the limitations of your study, and how may they be improved upon in future studies of this kind?
The study's main limitation is that we recruited cases in the middle of a pandemic, so we didn't have genotypes for mild cases to use as controls. Instead of mild COVID-19 cases, we used controls from different studies of the general population. This was technically challenging since the genotyping or whole-genome sequencing tools used differed, and we had to account for these differences. We now have genotypes from mild COVID-19 cases generated using the same pipelines as our critical COVID-19 cases.
Your work was in partnership with Genomics England. How important is collaboration within this field and to this work specifically?
Collaboration is vital to this research. People with very different expertise are needed to do an analysis like this. To carry out this analysis, we needed people to work in many different roles. Administrative staff organized paperwork and established sites, doctors and nurses recruited patients, research technicians sequenced DNA, then informaticians processed this DNA information in a computer and analyzed it to find the DNA regions related to COVID-19. Finally, it's rarely true that the results immediately make sense, so there were a number of groups of researchers working closely together to interpret the biological meaning of the results.
What is next for you and your research?
We're still working on critical COVID-19. We have recruited more individuals and will soon start a new analysis with 15,000 critical cases and 15,000 mild controls (people who had mild or asymptomatic COVID-19). We also have gene expression data from hospitalized patients that we can link to our analysis outputs which will likely help us address some of the limitations mentioned above.
Where can readers find more information?
About Erola Pairo-Castineira
My name is Erola Pairo-Castineira I am currently a postdoctoral researcher at the Roslin Institute at the University of Edinburgh. Over the last two years, I have been working on host genetic studies of COVID-19. I have conducted several analyses that have led to the discovery of new genetic associations of COVID-19 critical illness. Before my current position, I was a postdoctoral researcher in the MRC-HGU, working on the integration of omics and genotype data to identify causal genes and pathways underlying pigmentation traits.