Effects of Human Cytomegalovirus Infection on Colonic Epithelial Morphology and Signaling

Abstract

Human cytomegalovirus (CMV) is a beta-herpes virus that infects 40%-90% of the population, depending on geographical region and socio-economic status. While in most individuals, the virus remains latent, reactivation of CMV can cause severe illness. Reactivation is most common in immunocompromised individuals such as transplant recipients, HIV patients, and cancer patients. Further, patients with ulcerative colitis (UC) are at high risk for CMV reactivation, resulting in CMV colitis, which can ultimately lead to colectomy or death. Since the virus only infects primary cultures of human cells, studies of how CMV impacts the intestinal epithelium and its signaling and barrier function can only be done in vitro. In this study, we use a human colonoid monolayer to study the impact of viral infection on epithelial cell morphology, tight junction structure, and signaling. Colonocytes show evidence of infection as early as two days after exposure to the virus by observing immediate early (IE) gene expression and identifying enlarged nuclei, a notable feature of CMV infection. Subsequent experiments reveal that the expression of soluble GFP, marking infected cells, coincides with the viral replication process, as indicated by the presence of the unique long region protein (UL44) of CMV. These experiments provide context as to how cells in colonoids infected by CMV behave since most studies report CMV infection in fibroblasts and endothelial cells. Overall, studies were performed over a 2, 4, 6, and 8-day time course using a virus that expresses soluble GFP, which is controlled by the SV40 promoter. To assess monolayer integrity, we use E-cadherin to mark cell borders, as E- cadherin is not affected by CMV infection. As previously reported, at 4 days, zonula occludins-1 (ZO-1) localization is compromised by CMV infection. In addition, we found that the leaky claudin, claudin-2, is substantially upregulated by CMV as early as 4 days after infection. We also note claudin-2 localization to an intracellular compartment in CMV-infected cells. Further, both claudin-1 and occludin are disrupted by infection. Overall, we show how CMV impacts the morphology and barrier of the intestinal epithelium in colonoids. These changes can lead to persistent infection and inflammation in the colon, often resulting in progressively worsening outcomes. In addition, because YAP activity has been shown to be modulated by CMV infection and YAP is essential for intestinal epithelial repair and integrity, we examined YAP distribution in infected monolayers. We observed a decrease in nuclear YAP in CMV-infected cells. This suggests that CMV could be replicating in greater numbers or more efficiently because of this decreased nuclear YAP. This study highlights the impact of CMV infection on cells at the mucosal surface and may provide insight into how CMV reactivation in immunosuppressed patients drives CMV colitis.