Pathogenetic Diversity of Cytomegalovirus Genotypes: Implications for Ocular DiseaseDavid Navarro1,2,31Microbiology Service, Clinic University Hospital, INCLIVA Health Research Institute, Valencia, Spain.2CIBER de Enfermedades Infecciosas, Instituto de Salud Carlos III, Madrid, Spain.3Department of Medicine, School of Medicine, University of Valencia, Valencia, Spain.*Correspondence:David Navarro, Microbiology Service, Hospital Clínico Universitario, Instituto de Investigación INCLIVA, Av. Blasco Ibáñez 17, 46010 Valencia, Spain. Phone: 34(96)1973500; Fax: 34(96)3864173; Email: david.navarro@uv.es.Cytomegalovirus (CMV) is a prototypical β-herpesvirus that establishes persistent infection in the human host and typically causes severe clinical conditions (end-organ disease) in immunocompromised individuals, particularly transplant recipients, in whom it is associated with high morbidity and mortality.¹ While CMV usually causes asymptomatic or mild infection in immunocompetent individuals, several recently published studies suggest that CMV may be responsible for an increasing number of anterior segment ocular infections in this population, including anterior uveitis and corneal endotheliitis.2-5 These conditions may lead to severe complications such as corneal decompensation and glaucomatous optic neuropathy, both of which can result in vision loss. Intriguingly, most cases of CMV-associated anterior uveitis have been reported in East Asian populations. Despite being a DNA virus, CMV exhibits substantial genetic diversity. Although sequenced CMV strains are approximately 95% homologous at both the DNA and protein levels, polymorphic sequences are distributed across both coding and non-coding regions of the viral genome. Notably, the greatest variability is found in genes encoding viral glycoproteins, including gB, gH, gN, gO, and gL, all of which are highly immunogenic and play a critical role in viral entry.6 Glycoprotein B (gB), encoded by ORF UL55, plays a critical role in viral entry by mediating fusion between the viral and cellular membranes, as well as cell-to-cell viral transmission.⁷ To date, five major gB genotypes -gB1, gB2, gB3, gB4, and gB5- have been identified based on nucleotide polymorphisms surrounding the furin cleavage site (codons 448-481), a key domain involved in the conformational transition from the prefusion state to the functional fusogenic form of the protein. The existence of potential pathogenetic differences among CMV gB genotypes has long been suspected.6In this context, several studies have investigated associations between gB genotypes and clinical outcomes -such as the development of end-organ disease or viremia- in diverse settings, including congenital CMV infection, allogeneic hematopoietic stem cell or solid organ transplantation, and human immunodeficiency virus infection, yielding inconsistent results. Previous reports have suggested a potential association between specific CMV gB genotypes and clinical outcomes in patients with anterior uveitis, corneal endotheliitis, or iridocyclitis.5,8,9 In a recent issue of theJournal of Medical Virology , Lestari and colleagues provide evidence supporting the hypothesis that genetic variations in the CMV gB gene may influence the pathogenesis of virus-induced ocular diseases.10 The authors first investigated whether there was an association between CMV gB genotypes and ocular tropism. To this end, they analyzed CMV PCR-positive aqueous or vitreous humor specimens from immunosuppressed patients with CMV retinitis -most of whom had undergone allogeneic hematopoietic stem cell transplantation (allo-HCT)- as well as from immunocompetent individuals with anterior uveitis and from patients with chronic retinal necrosis (CRN) exhibiting mild to moderate immunosuppression, either systemic or local (intravitreal triamcinolone acetonide injection). Blood samples from immunosuppressed patients, predominantly allo-HCT recipients with CMV pp65 antigenemia, were used as controls. CMV gB genotyping was performed using a multiplex sequence-specific PCR assay, and for some specimens, Sanger sequencing of amplicons spanning the furin cleavage site region of the UL55 gene. A differential distribution of CMV gB genotypes was observed between ocular and blood samples, with CMV gB3 genotype predominating in ocular specimens followed by gB1 (both accounted for 83% of genotypes detected in ocular fluid specimens) regardless of the type of ocular disease and, indirectly, the level of immunocompetence. In contrast, the CMV gB2 genotype was the most prevalent in blood specimens. Notably, mixed CMV gB genotypes were more frequently detected in blood than in ocular simples. While these data pointed to gB3 genotype, and gB1 to a lesser extent, displaying a higher tropism for ocular tissues their interpretation should be cautious as only three patients (all with retinitis) had paired blood and ocular specimens analyzed.An interesting observation was that, among patients with anterior uveitis, CMV DNA loads in ocular fluids were significantly higher in gB1- and gB3-positive specimens than in those harboring gB2. This led the authors to hypothesize that gB1 and gB3 genotypes may possess enhanced replicative capacity, facilitating hematogenous dissemination and seeding of ocular tissues. However, since viral loads were not normalized to the interval between disease onset and sample collection, this assumption should be interpreted with caution. In addition, unfortunately only 3 specimens contained the gB2 genotype. In line with this hypothesis, a single amino acid substitution (K518R) within gB domain III, located near the top of the gB core, was detected in most gB1- and gB3-positive ocular fluid samples but not in gB2 specimens, which were predominantly from blood. In silico modeling predicts that this mutation may enhance gB stability, potentially increasing viral entry efficiency and facilitating cell-to-cell spread. Interestingly, the K518R mutation appears to be quite common among Asian CMV strains and rare in CMV strains of European origin. The relevance of these findings should nevertheless be interpreted cautiously, given the complex nature of viral entry into host cells -a process in which a wide array of viral glycoproteins plays a crucial role.1Ultimately, further in vitro experiments using animal models or polarized epithelial cell culture systems should provide definitive insights into genotype-specific replicative behavior and ocular pathogenicity. Enhanced immunoscape capacity of gB1 and gB3 genotypes compared to gB2 may also help explain differences in viral loads. Using predictive binding algorithms (NetMHCIIpan-4.3) alongside a flow-cytometry-based peptide-HLA II binding assay, the authors identified a peptide pair, gB190-204, that was identical between gB1 and gB3 but differed from the corresponding gB2 sequence, exhibiting more than a two-fold difference in predicted binding affinity between gB genotypes (lower in gB1/3) on two HLA II allotypes present in the cohort (DPB1*0402 and DQB1*0301). Importantly, the gB190–204 sequence was present in ocular specimens as shown by amplicon-targeted sequencing. Although the data is interesting, adaptive immune responses against CMV are complex and redundant and the matrix protein pp65 and the immediate-early-protein-1 (IE-1) are immunodominant in most individuals.1 In summary, the observations made by Lestari and colleagues further reinforce the idea that pathogenetic differences may certainly exist across CMV strains that may impact on the risk and outcome of ocular diseases and by extension other clinical conditions linked to CMV infection both in immunocompetent and immunocompromised patients.
