Human cytomegalovirus glycoprotein N genotypes in AIDS patients

Abstract

Human cytomegalovirus (HCMV) clinical isolates display genetic polymorphisms that are supposed to be implicated in HCMV-induced immunopathogenesis. Open reading frame UL73, encoding for the envelope glycoprotein N, shows four distinct genomic variants (gN-1, 2, 3, 4). This report analysed the distribution of glycoprotein N genotypes in 29 AIDS patients and 50 solid-organ transplant recipients compared with 111 immunocompetent individuals. HIV patients exhibited a prevalence of HCMV strains with the gN-1 genotype, which seem to replicate favourably in immunocompromised hosts. Human cytomegalovirus (HCMV) is a ubiquitous betaherpesvirus associated with a wide spectrum of diseases, particularly if acquired in utero and in immunocompromised individuals, and is one of the most common opportunistic infections in patients with AIDS. However, immunocompromised patients can develop different kinds of HCMV-associated disease, suggesting that the immunological status of the individual influences the clinical outcome of HCMV infection [1]. In addition, virus strains can vary in virulence and differ in their tropism and ‘pathogenic potential', and these differences seem to be related to genetic variability exhibited by ‘key genes’ and detected among HCMV isolates [2]. Genomic variants have been detected for a variety of HCMV genes, such us UL4 (gp48 [3]), UL144 [4], gH [5] and gB [1,6]. The most interesting polymorphic genes discovered to date are those encoding for viral envelope glycoproteins (gB, gH), mainly because their products are targets for neutralizing antibodies, often produced with a strain-specific pattern [7], and involved in virus entry and cell-to-cell virus spread [8,9]. gpUL73-gN, encoded by the open-reading frame UL73, is another recently investigated HCMV surface polymorphic glycoprotein. It is a component of the envelope gC-II complex, the major heparin-binding component [10], in association with gM [11–13] and is able to induce a neutralizing antibody response in the host [14]. Like other HCMV envelope glycoproteins, gpUL73-gN displayed a polymorphic N-terminal region, which allowed the identification of four main glycoprotein N genomic variants, denoted as genotypes gN-1, gN-2, gN-3 and gN-4 [15]. The aims of this study were to assess the distribution and prevalence of the different glycoprotein N genomic variants in the immunocompromised host (AIDS patients and solid-organ transplant recipients) and in the immunocompetent population in order to investigate the potential association of genotypes with conditions of the host immune system. This report retrospectively analysed the distribution of the glycoprotein N genotypes in 29 HCMV isolates obtained from AIDS patients and in 50 stains collected from solid-organ transplant recipients, and compared the glycoprotein N genotypic frequencies with those obtained from 111 immunocompetent individuals with HCMV infection, mainly pregnant women and congenitally infected infants. Samples of urine and saliva were inoculated onto confluent human embryonic lung fibroblasts and cultured to allow virus replication before DNA extraction, which was peformed using the IsoQuick Nucleic Acid Extraction Kit (ORCA Research Inc., WA, USA). The Instagene Whole Blood Kit (Biorad, CA, USA) was used to collect DNA directly from blood samples or polymorphonuclear leukocytes. DNA was polymerase chain reaction amplified using a pair of primers selected in the conserved region of UL73 [15]. The chosen primers were gN-up 5′–TGGTGTGATGGAGTGGAAC–3′ and gN–lw 5′–TAGCCTTTGGTGGTGGTTGC–3′. Genotyping was performed according to Pignatelli et al. [15] and allowed the assignment of HCMV isolates to one of the four glycoprotein N genotypes. The distributions of the glycoprotein N variants in the immunocompromised and immunocompetent populations were analysed and compared using contingency tables and the chi-square test. The results are reported in Table 1. Comparing the immunocompromised hosts with the immunocompetent population, we detected a statistically significant difference between the two populations (P = 0.023) because of the comparison between the gN-1 and gN-4 variants, gN-1 being overrepresented in the immunocompromised population (35.4 versus 20.7%).Table 1: Distribution of glycoprotein N genotypes among cytomegalovirus strains from immunocompetent individuals and immunocompromised patients (HIV-positivbe and solid-organ transplant recipients).Further detailed analysis was performed by partitioning the chi-square into degrees of freedom. When immunocompetent individuals and solid-organ transplant recipients were compared, the association between glycoprotein N types was statistically significant (P = 0.045). This significance was determined by the comparison between groups gN-1 and gN-2, the gN-2 group being well represented among the transplant recipients (4/50 against 1/111 of the immunocompetent individuals). Comparing the AIDS patients with the immunocompetent population, we noted that the association of the glycoprotein N variants with HIV-positive infected subjects was caused by the comparison between the gN-1 and gN-4 groups, the gN-1 genotype being the most represented in this immunocompromised subpopulation (P = 0.028). In contrast, we did not find any significant correlation among cytomegalovirus strains in HIV-infected and transplanted patients (P = 0.155). A potential relationship between glycoprotein N genotypes and immunocompromised conditions in the host was thus observed after the analysis of AIDS patients, who displayed a prevalence of the gN-1 variant with respect to findings in transplant recipients and in the immunocompetent population. HCMV strains with a gN-1 genotype could represent a less virulent viral phenotype, especially considering that this variant is typical of the laboratory-adapted strain AD169, the strain immunologically furthest away from HCMV clinical isolates [16]. This hypothesis is in agreement with data reported on glycoprotein B genotypes in AIDS patients, who showed a high prevalence of the AD169-like glycoprotein B type (gB-2) [9,17,18]. However, the significant difference in the prevalence of the glycoprotein N genomic variants between the immunocompetent and the immunocompromised populations suggests that the host's immune system plays a primary role in selecting the genotypes. It remains unclear why the glycoprotein N type association differed when the single immunocompromised populations, AIDS patients, and solid-organ transplant recipients were compared with the immunocompetent population. The main factor that influences clinical outcome is the immune status of the host, and clearly the above two groups are immunologically distinct. The immune system of organ transplant recipients can suppress HCMV more efficiently than that of bone marrow transplant recipients and AIDS patients, and the differences in virulence of glycoprotein N types may be less relevant in a more immunocompetent host, as proposed for glycoprotein B [19,20]. Our results do not exclude the possibility that specific glycoprotein N genotypes could influence the clinical outcome of AIDS-related cytomegalovirus diseases, such as encephalitis or retinitis, or post-transplantation fatal or non-fatal HCMV diseases, such as graft-versus-host disease or pneumonia. Those specific glycoprotein N associations should be analysed in larger populations in a prospective study with detailed clinical data available for each patient to determine whether glycoprotein N genomic variants can provide additional prognostic or predictive information on HCMV disease progression. However, the assignment of a virus strain and its ‘pathogenic potential’ to a certain genotype on the basis of a single gene region may be misleading. Our results on glycoprotein N, together with data previously reported for the glycoprotein B and UL144 genotypes [21,22], suggest that HCMV strains’ ‘potential of infectivity’ and ability to produce severe disease is not related to a single polymorphic gene, but is probably determined by a cooperation of gene variants.