Title: A search for human herpesvirus 8 (HHV-8) in HIV-1 infected mothers and their infants does not suggest vertical transmission of HHV-8
Abstract: Human herpesvirus 8 (HHV-8), a novel member of the gamma herpesviruses closely related to herpesvirus saimiri and rhesus monkey rhadinovirus, has been consistently identified in all epidemiological forms of Kaposi's sarcoma (KS), in primary effusion lymphomas, and in some forms of Castleman's disease (reviewed by Schulz, 1998). Several epidemiological and molecular findings show that sexual transmission represents the predominant route for HHV-8 infection in non-endemic areas (Martin et al., 1998). In KS patients, glandular epithelial cells in limited areas of the prostate appear to be the virus reservoir as latency-associated viral transcripts have been found to be more abundant than lytic phase-associated transcripts in this compartment (Diamond et al., 1998); therefore, intermittent virus shedding in the prostate may account for the low frequency of HHV-8 detection in semen samples from these patients, as well as subjects from endemic areas (reviewed by Blackbourn and Levy, 1997). Moreover, in our previous studies HHV-8 specific DNA was rarely detected in the cell fraction of cervico-vaginal secretions from HHV-8 seropositive women either HIV-1 infected or not, and its detection appeared to be favored by concomitant inflammatory lesions of the genital tract (Calabrò et al., 1999). That HHV-8 infection also circulates among children living in endemic areas (Goedert et al., 1997; Calabrò et al., 1998; Mayama et al., 1998) suggests an additional mode of transmission; indeed, saliva was found to contain infectious HHV-8 viral particles (Vieira et al., 1997). Gessain et al. (1999) have studied the prevalence of HHV-8 in children from Cameroon, Central Africa, and have demonstrated that infection occurs progressively during childhood, and by 15 years of age seroprevalence (48%) is almost similar to that observed in adults from the same area. This study also included the analysis of sera from 32 mother-child pairs, which revealed that infants born to HHV-8 seropositive mothers were also seropositive mainly during their 5 first months of life. As HIV-1 infection represents a risk factor for the acquisition of HHV-8 infection (Martin et al., 1998; Calabrò et al., 1999), we explored the potential for mother-to-child transmission of HHV-8 in 32 HIV-1 infected mothers and their children. Twenty-nine mothers were from Northern Italy, and 3 from Africa (2 from Ghana, and 1 from Liberia); 3 mothers gave birth to twins. Plasma samples from mothers at delivery, and from the 35 infants at 3 and 12 months of age were analysed for HHV-8 antibodies to a latency-associated nuclear antigen (LANA) and to a lytic phase-related structural protein encoded by ORF65, as previously described (Calabrò et al., 1998). Peripheral blood mononuclear cells (PBMC) were analyzed for the presence of HHV-8 DNA sequences by nested-polymerase chain reaction (PCR), using 2 sets of nested primers specific for 2 different regions of HHV-8 genome, ORF25 and ORF26, as previously reported (Calabrò et al., 1999). HIV-1 infectious status in the infants was assessed by virus culture and PCR, as previously described (De Rossi et al., 1992); HIV-1 viral load was determined in plasma using a quantitative PCR (Amplicor Monitor, Roche Diagnostic Systems, Branchburg, NJ, version 1.5). Serological findings are reported in the Table. Nine mothers transmitted HIV-1 infection to their children, and HIV-1 viral load was significantly higher in transmitting than in non-transmitting mothers (3.8 log10 vs. 3.2 log10 HIV-1 RNA copies/ml; p<0.025 Kolmogorov-Smirnov test). Eleven of the 32 (34.3%) HIV-1 seropositive mothers had antibodies to at least 1 of the 2 HHV-8 antigens under study; 3 of them (2 from Ghana and 1 from Italy) had antibodies to both antigens. The rate of HHV-8 seropositive women was higher among the HIV-1 transmitting (5/9) than the non-transmitting mothers (6/23); no correlation was found between HHV-8 seropositivity and HIV-1 RNA molecules in plasma, since the mean HIV-1 viral load in HHV-8 seropositive and seronegative women did not significantly differ (3.52 log10 vs. 3.48 log10 RNA copies/ml). All infants born to HHV-8 seropositive mothers were HHV-8 seroreactive, and at 3 months of age, 5 of 9 HIV-1 infected, and 6 of 26 HIV-1 uninfected children had antibodies to LANA; only 1 of these last 6 children also had antibodies to ORF65 protein. The HIV-1 infected and uninfected children showed a different antibody clearance; indeed, at 12 months of age, only 1 of 26 (3.8%) uninfected, but 4 of 9 (44.4%) HIV-1 infected still exhibited HHV-8 LANA antibodies (p=0.01, Fisher's test). Nested-PCR analysis at 12 months of age did not show the presence of HHV-8 sequences in the PBMC of these 5 HHV-8 seroreactive infants. Evaluation of LANA antibody titers in HHV-8 seropositive mothers did not reveal any differences among the mothers whose infants showed prolonged seroreactivity to HHV-8 antigens. A 24 month follow-up of the 4 HHV-8 seroreactive HIV-1 infected disclosed no evidence of HHV-8 antibodies or HHV-8 sequences in their PBMC. In agreement with Gessain et al. (1999), our findings suggest that maternal HHV-8 antibodies are passively transmitted. Indeed, mother-to-child HHV-8 vertical transmission appears very unlikely because all infants who were HHV-8 seropositive at 3 months of age reverted to seronegativity by 24 months of age; in addition, no HHV-8 sequences were detected in their PBMC. On the other hand, the rate of HIV-1 vertical transmission was higher in HHV-8 seropositive (45%) than in HHV-8 seronegative mothers (19%); since no significant correlation emerged between HHV-8 seropositivity and HIV-1 viral load, it could be argued that HHV-8 infection in itself might be a risk factor for HIV-1 transmission. Passively transferred maternal HHV-8 antibodies persisted longer in HIV-1 infected vs. uninfected children; this clearance delay was not dependent on a higher LANA antibody titer in the mothers. This last finding suggests a need for caution in the serological diagnosis of HHV-8 infection in infants living in HIV-1 endemic areas. Maria Luisa Calabrò [email protected], Paola Gasperini, Massimo Barbierato, Lucia Ometto, Marisa Zanchetta, Anita De Rossi, Luigi Chieco-Bianchi