Integrating GWAS with bulk and single-cell RNA-sequencing reveals a role for LY86 in the anti-Candida host response

Candida bloodstream infection, i.e. candidemia, is the most frequently encountered life-threatening fungal infection worldwide, with mortality rates up to almost 50%. In the majority of candidemia cases, Candida albicans is responsible. Worryingly, a global increase in the number of patients who are susceptible to infection (e.g. immunocompromised patients), has led to a rise in the incidence of candidemia in the last few decades. Therefore, a better understanding of the anti-Candida host response is essential to overcome this poor prognosis and to lower disease incidence. Here, we integrated genome-wide association studies with bulk and single-cell transcriptomic analyses of immune cells stimulated with Candida albicans to further our understanding of the anti-Candida host response. We show that differential expression analysis upon Candida stimulation in single-cell expression data can reveal the important cell types involved in the host response against Candida. This confirmed the known major role of monocytes, but more interestingly, also uncovered an important role for NK cells. Moreover, combining the power of bulk RNA-seq with the high resolution of single-cell RNA-seq data led to the identification of 27 Candida-response QTLs and revealed the cell types potentially involved herein. Integration of these response QTLs with a GWAS on candidemia susceptibility uncovered a potential new role for LY86 in candidemia susceptibility. Finally, experimental follow-up confirmed that LY86 knockdown results in reduced monocyte migration towards the chemokine MCP-1, thereby implying that this reduced migration may underlie the increased susceptibility to candidemia. Altogether, our integrative systems genetics approach identifies previously unknown mechanisms underlying the immune response to Candida infection.
Author summary Candida albicans is a fungus that can cause a life-threatening infection in individuals with an impaired immune system. To improve the prognosis and treatment of patients with such an infection, a better understanding of an individual’s immune response against Candida is required. However, small patient group sizes have limited our ability to gain such understanding. Here we show that integrating many different data layers can improve the sensitivity to detect the effects of genetics on the response to Candida infection and the roles different immune cell types have herein. Using this approach, we were able to prioritize genes that are associated with an increased risk of developing systemic Candida infections. We expand on the gene with the strongest risk association, LY86, and describe a potential mechanism through which this gene affects the immune response against Candida infection. Through experimental follow-up, we provided additional insights into how this gene is associated with an increased risk to develop a Candida infection. We expect that our approach can be generalized to other infectious diseases for which small patient group sizes have restricted our ability to unravel the disease mechanism in more detail. This will provide new opportunities to identify treatment targets in the future.