Distribution of Kaposi Sarcoma–Associated Herpesvirus/Human Herpesvirus 8 in Maternal Saliva and Breast Milk in Zambia: Implications for Transmission

Background. The seroprevalence of Kaposi sarcoma-associated herpesvirus (KSHV)/human herpesvirus 8 (HHV-8) in sub-Saharan Africa suggests that multiple routes of transmission exist. In the present study, we examined 2 possible routes of mother-to-child transmission, through breast milk and saliva, during the first 6 months after delivery. Methods. The prevalence of HHV-8 DNA in the breast-milk cells (n75), milk supernatant (n =56), colostrum (n =2), and saliva cells (n =65) of HHV-8-seropositive mothers who recently gave birth was examined. Polymerase chain reaction analysis was performed for the detection of HHV-8 in cross-sectional samples isolated at 2, 4, and 6 months after delivery. Results. None of the 75 breast-milk samples but 2 of the colostrum samples that were analyzed contained HHV-8 DNA at a limit of detection of ∼1 HHV-8 copy/104 cellular genomes, whereas Epstein-Barr virus DNA and HIV-1 DNA were detected in 16 and 22 samples, respectively. Analysis of 65 saliva cell samples, which were obtained from mothers who also provided milk samples, revealed that 19 of the samples had detectable HHV-8 DNA. Viral DNA was found at all time points, but the presence of viral DNA in saliva was independent of maternal HIV-1 serostatus (X2 = 0.33; P =.57). Conclusions. Our findings demonstrate the lack of HHV-8 DNA in the breast milk of seropositive mothers, and they suggest that contact with breast milk is not a likely source of horizontal transmission of virus to infants in sub-Saharan Africa. Human herpesvirus 8 (HHV-8) was first identified, in 1994, in tissue from a patient with AIDS who had Kaposi sarcoma (KS) [1]. In the United States and Western Europe, HHV-8 infection is uncommon in the general population, but it is considered to be common in regions of South America, southern Italy, and Africa [2-4]. After the onset of the AIDS pandemic, several sub-Saharan countries, including Zambia, reported a significant increase in the number of cases of KS among both adults and children [5, 6]. In the United States and Western Europe, similar increases in KS were observed, but these increases were observed almost exclusively among HIV-1-infected men. This suggests that different routes of transmission of HHV-8 or different rates of infection may be occurring in these populations, or that different genetic, cultural, or environmental factors influence transmission of HHV-8. The modes of transmission of HHV-8 remain poorly understood, despite the existence of significant data on virus seroprevalence. In areas in which seroprevalence is low, HHV-8 appears to be predominantly acquired through sexual contact. Infection and seroconversion have been associated with male homosexual practices and the number of HIV- 1-positive sex partners, and they also have been associated with individuals at risk for STD [7-9]. In the United States, saliva has been implicated as a source for transmission of HHV-8, because viral DNA has been detected in as many as 70% of oral cavity samples in studies involving HIV- 1-infected men with KS [10]. Very little is known regarding the presence of HHV-8 in saliva and the mechanisms of transmission of HHV-8 in regions of endemicity. A cumulative increase in the seroprevalence of HHV-8 in children, from birth through adolescence, has been observed in regions of endemicity; this finding suggests that multiple routes, including perinatal transmission, are involved [4, 11]. Perinatal transmission of HHV-8 may occur as a result of breast-feeding, as has been previously suggested for cytomegalovirus (CMV) [12], but it fails to account for the continued acquisition of HHV-8 after breast-feeding has ceased. CMV, as well as other herpesviruses, has been detected in milk and predominantly localizes to the breast-milk supernatant (liquid) fraction [13]. Whether HHV-8 can be detected in breast milk from infected mothers is not known. We previously documented HHV-8 infection that occurred in Zambian children by 12 months of age, using serological and HHV-8 DNA detection methods [14, 15]. Although children born to HHV-8-infected mothers can be congenitally infected, the frequency of such transmission is apparently low and does not adequately account for the increasing seroprevalence subsequent to gestation and delivery [4, 14, 15]. Therefore, horizontal transmission is a likely route for a large proportion of HHV-8 infections that occur in early childhood. As part of our continuing cohort study of the epidemiology of HHV-8 infection among Zambian mother-infant pairs, the major goal of the present study was to examine whether transmission of HHV-8 to infants can occur via breast milk. In the present study, we determined the prevalence of virus in breast milk and maternal saliva. Establishing the presence or absence of HHV-8 in these maternal body fluids will extend our understanding of how HHV-8 is spread to children and will thereby enable the development of strategies to prevent transmission of HHV-8 and pediatric KS. Subjects and specimen collection. Analyses were performed for specimens obtained from participants in a longitudinal cohort study of transmission of HHV-8 performed at the University of Zambia Teaching Hospital (UTH; Lusaka, Zambia). Approval for the study was obtained from the Ministry of Health of Zambia, the Research and Ethics Committee of the UTH, and the institutional review boards of the University of Nebraska-Lincoln and the University of Miami (Miami, FL). Written, informed consent was obtained from all study participants. In the present study, mothers who showed no clinical signs of KS were recruited, at delivery, in the labor ward of the UTH, and they were followed through the first 6 months after delivery. Whole breast-milk and saliva samples were obtained and were spun for 10 min at 500 g, to separate the liquid and cell portions; each sample was stored at 80°. Extraction of DNA from breast milk and saliva. Extraction of thawed cell pellets was performed using the Gentra Systems DNA isolation kit with proteinase (0.1 mg/mL), according to the manufacturer's instructions [14]. For extraction of viral DNA from the liquid portion of breast milk, samples were spun for 5 min at 10,000 g, and 150 , of supernatant was decanted and used for viral DNA extraction, which was performed as described above. All DNA extractions were segregated from polymerase chain reaction (PCR) amplification and plasmid manipulation and were performed in a dedicated laboratory, with the use of an isolated hood with UV light; stringent conditions and controls were used to minimize contamination between samples. Conventional HHV-8, Epstein-Barr virus (EBV), and HIV-1 PCR. DNA extracted from the samples was subjected to PCR specific for human -globin, by use of primers described elsewhere [16]. The -globin PCR-positive samples were subsequently analyzed for HHV-8 by use of primers for the minor capsid (ORF26) and gB (N-terminus [gBN]) genes, as described elsewhere [1, 15, 17]. Samples were defined as being positive for HHV-8, either by first-round PCR for both ORF26 and gB, with confirmation by Southern blot analysis, or by nested PCR that consistently detected positive samples by both ORF26 and gBN PCR performed in 2 separate amplification reactions. PCR analysis for EBV was performed using primers specific for the EBV thymidine kinase (TK) gene: TK1 (5'-GTGGGATCCA-TGGCTGGATT-3') and TK2 (S'-GCTACCCGGAGAGTTTCC-AGT-3'). Thermal-cycling parameters for EBV-TK PCR were as follows: 1 cycle for 5 min at 95°; 38 cycles for 30 s at 95°, for 30 s at 56°, and for 30 s at 72°; and 1 cycle for 3 min at 72°. All reactions were subjected to Southern blot analysis by use of a digoxigenin-labeled probe, as described elsewhere [15]. HIV-1 PCR was used to detect a short sequence of the 5'viral long-terminal repeat in all breast-milk samples obtained from HIV-1-infected mothers, as described elsewhere [18]. Real-time PCR for the detection of HHV-8. Real-time PCR for the detection of HHV-8 and human -globin was used for all samples demonstrated to be positive by conventional PCR, to quantitate the number of copies of HHV-8 and the number of cellular equivalents present in each sample [16, 19]. For the present study, real-time PCR was performed according to the manufacturer's suggested conditions, by use of TaqMan chemistry (Applied Biosystems). The HHV-8 reaction, which used an amplicon within the ORF73 gene, contained 300 nmol/ L and TaqMan The for HHV-8 was using a sequence of the ORF73 in the as according to the manufacturer's instructions was by and the number was from the of of the plasmid were in a of human from cells A similar was for human -globin PCR by known copies of DNA in DNA were by the of DNA by use of a and by this by a between the number of copies and the use of the cycle from the PCR amplification of the HHV-8 and was used to a for each amplicon for of HHV-8 and -globin present in the known samples. For each reaction, the for HHV-8 and a similar for the human were in with saliva samples. Samples were in and the was and used for of the and analysis of HHV-8 PCR amplification of the was performed on saliva samples, as well as the HHV-8 infected and cell by use of several For first-round PCR amplification of primers and were used as described elsewhere PCR was performed on samples by use of primers and both of which were on the samples were subjected to PCR that used primers and or PCR were and were and according to the manufacturer's conditions were used to limit PCR as in the HHV-8, virus (EBV), and HIV-1 and or were and in to as were using the and the DNA (Applied according to the manufacturer's were by use of and were using general time with analysis using for were performed for analysis of PCR detection between The was used for of the number of HHV-8 of HHV-8 in breast milk and saliva. examine the horizontal transmission of HHV-8 from to via breast milk saliva in Zambia, we first determined the of PCR performed on both body DNA with known HHV-8 as determined by real-time PCR, was to of and the limit of detection in breast milk and saliva was determined by PCR amplification of the ORF26 DNA extracted from milk and saliva samples, at a of cellular was used to DNA in HHV-8 from 1 to viral use of the viral HHV-8 DNA was consistently detected using conventional PCR with first-round ORF26 primers at 1 viral in a of milk cell equivalents and at 10 viral copies in saliva cell equivalents In saliva a was observed with the but detection was In HHV-8 PCR detect the presence of 10 cells in milk, by use of the extraction and PCR conditions not PCR was the conditions, by use of DNA extracted from a herpesvirus cell of human herpesvirus 8 (HHV-8) in breast milk and saliva, by use of conventional and polymerase chain reaction ORF26 PCR and human PCR were performed on DNA HHV-8 as determined by real-time and on breast-milk and saliva DNA samples, to the of detection of 8 in each body DNA and DNA from the cell were in as of first-round ORF26 PCR and human PCR performed on DNA in breast-milk and saliva cell samples obtained from 6 with DNA used as a positive for real-time PCR for HHV-8. The used for of the HHV-8 from 5 independent and is The as of HHV-8 are the cycle with 2 the of a between copies and the cycle of HHV-8 in breast milk. Although HHV-8 has been detected in a of and body fluids from infected the presence of HHV-8 in breast milk has The cross-sectional prevalence of HHV-8 in breast milk from seropositive mothers in Zambia was HHV-8-infected were 2 and as defined elsewhere [14]. DNA extracted from colostrum and breast-milk cell pellets obtained to the time of birth or at 2, 4, or 6 months after delivery breast was first analyzed for DNA and for the presence of by use of human samples were used for subsequent detection of HHV-8. PCR analysis of 2 colostrum samples, as well as and breast-milk samples from HIV-1-infected mothers and mothers, to consistently detect HHV-8 DNA in sample milk specimens were determined to be positive for HHV-8 DNA they were with nested ORF26 but detection was and not be using primers from either the gB or the the of the these specimens were considered to be for HHV-8. of a performed with 6 breast-milk samples obtained from HHV-8-seropositive mothers are in breast-milk samples were also subjected to PCR specific for which is known to be present in breast milk to that herpesvirus DNA be in the present EBV was detected in 16 of 75 breast-milk cell samples, with the presence of EBV on the serostatus of the as in PCR amplification of a of the HIV-1 revealed that 22 of breast-milk samples obtained from in 1 contained HIV-1 DNA whether HHV-8 was present in the liquid portion of breast milk, DNA was extracted from the supernatant from milk. supernatant samples were from the breast-milk cell samples analyzed and reported in None of these samples was found to be positive for HHV-8 by ORF26 PCR, by use of both first-round and nested ORF26 PCR a finding that is similar to was observed for the cellular fraction the other PCR revealed positive samples. of human herpesvirus 8 Epstein-Barr virus (EBV), and by polymerase chain reaction (PCR) analysis of cross-sectional breast-milk cell samples obtained from HHV-8-seropositive mothers during the first 6 months after delivery of human herpesvirus 8 (HHV-8) and virus by polymerase chain reaction analysis of breast-milk supernatant samples obtained from mothers during the first 6 months after delivery. of HHV-8 in saliva. HHV-8 was detectable in breast-milk cells from HHV-8-seropositive mothers, and because breast milk is to be a source of transmission of HHV-8 to we examined the prevalence of HHV-8 in cross-sectional saliva cell samples obtained from the of seropositive For DNA from maternal saliva cells obtained at 2, 4, or 6 months after delivery, PCR analysis was performed using the for positive samples that were for breast-milk 3 the of HHV-8 PCR analysis of saliva cell samples obtained from 65 HHV-8-seropositive The saliva cell samples of 19 mothers were positive for HHV-8 with positive for and samples obtained at 2, 4, and 6 respectively. the of a a saliva cell sample that was positive at a time was independent of the HIV-1 serostatus of the in 1 in = 0.33; P =.57). A of ORF26 PCR performed on saliva samples obtained from the 6 mothers who had breast-milk samples analyzed for HHV-8 in in whether mothers breast-milk samples were for HHV-8 had detectable viral DNA in saliva mothers with breast-milk and saliva cell samples obtained at the had samples A of mothers with specimens were with to HHV-8 PCR each time HHV-8 was detected in saliva samples, but not in breast-milk samples, obtained from the patient of human herpesvirus 8 (HHV-8) by polymerase chain reaction analysis of cross-sectional saliva cell samples obtained from HHV-8-seropositive mothers during the first 6 months after delivery. of human herpesvirus 8 (HHV-8) polymerase chain reaction analysis performed on breast-milk and saliva cell samples obtained from HHV-8-seropositive mothers at different time after delivery The HHV-8 infection of the infants of a of the mothers saliva samples were analyzed for the presence of HHV-8 (n was determined at 12 months after delivery. the detection of maternal the of the infants were at different time after delivery, by use of analysis, as performed elsewhere [14]. infants had determined at birth and at 6 months after delivery, whereas infants had determined at birth and at 12 months after infants who had increasing from birth to either 6 or 12 months of were considered to be In the of mother-infant pairs, of the mothers were found to be positive for HHV-8 DNA in saliva, by PCR, whereas were found to be for HHV-8 by PCR. analysis revealed that of infants were seropositive for HHV-8 by 12 months after delivery. these seropositive 3 were born to mothers who had detectable HHV-8 DNA in saliva and were born to mothers who had no HHV-8 DNA in saliva at the time that saliva cells were The of an being seropositive for HHV-8 was independent of the HHV-8 PCR of saliva, as at 1 time (X2 = P of HHV-8 in saliva. that several mothers in our study had detectable HHV-8 in saliva, we to the number of copies of HHV-8 in saliva by use of real-time PCR use of plasmid from 10 to copies of the ORF73 a was observed between plasmid copies and the = this finding that detection of HHV-8 was of use of real-time PCR, and cells were found to and HHV-8 not The cellular and viral were and samples that contained cellular equivalents and copies of HHV-8 were reported as the number of HHV-8 cellular of several of our saliva samples analysis of all mothers found to be positive for HHV-8 by conventional PCR. samples obtained from the 12 mothers who were found to be positive for HHV-8 by conventional PCR mothers with samples from 1 time and mothers with samples from 2 time the for samples are in large in the number of HHV-8 cells was observed the mothers who had samples from 2 time 3 had samples that showed between time mothers had samples with at time but low at the other time whereas 1 had a between time whether HIV-1 infection increases the HHV-8 in saliva the number of HHV-8 cells was examined in both HIV-1-infected mothers and mothers all time The the and the first and are in for HIV- 1-infected mothers and who were infected with HIV-1 to have a number of HHV-8 but the of not a significant between HIV-1-infected and mothers by on the of of a number of samples (n analysis of the of The HHV-8 has been to regions of 1 and 2 and these regions have been used to HHV-8 of the of present in the we used this to whether the HHV-8 detected in the saliva samples in the present study were to other known In we to the of PCR contamination by or HHV-8 the of the of for 2 saliva samples that were found to be positive for HHV-8 and from samples isolated from 3 Zambian with KS analyzed at the University of Nebraska-Lincoln and and other all and of HHV-8 on The saliva are from each other at 3 and they from the and used in our they but are from including and and the other Zambian and analysis of the sequence from the 2 saliva samples in the present study, with the used in the is in to the the analysis that the 2 saliva from Zambian mothers are to to from other and analysis of saliva samples. of the first of 1 from and saliva samples from and 3 cell samples obtained from with KS in Zambia and analyzed in our and other HHV-8 from of the sequence by with analysis using for all samples in with the sequence used as an and are to the of each the 2 saliva samples from the present United The seroprevalence of HHV-8 in several regions of the the that multiple factors influence transmission and that multiple modes of transmission may 14, Our on transmission of HHV-8 suggested that in or transmission was not a major to the of HHV-8 infection in the first of [14, 15]. has also that many infants have seroconversion to an during the first after the of the present analysis was to whether HHV-8 be detected in maternal fluids with for virus transmission, such as breast milk and saliva. A number of herpesviruses, including human herpesvirus CMV, and virus have been detected in breast milk, the of transmission via this of infants who were by mothers have reported rates of infection of with for infants by mothers DNA was detected in infants in infants 2 studies reported the absence of viral DNA in the breast milk of mothers, with the result that of the infants of these mothers were infected after birth and HHV-8 are known to and both of which are found in the cellular of colostrum and milk of findings from studies of suggest that HHV-8 also be found in breast milk and cells within breast milk. In the present study, we examined a number of cross-sectional breast-milk samples obtained from infected mothers within the first 6 months after delivery, but we to detect HHV-8 DNA at of detection of 1 HHV-8 copy/104 performed in breast milk suggest that HHV-8-infected cells can be detected in breast milk, In the present study, to detect HHV-8 have been the result of being cells present in our specimen to for detection of HHV-8 as has been suggested for a of breast-milk samples contained a of cellular as determined by PCR This that of cells and DNA were present in the breast-milk samples, the number of infected cells present was the limit of detection of our In of this we detected HHV-8 nested PCR in a breast-milk but such detection was and not be using a of is also possible by first samples at 2 we the for the detection of viral DNA in breast milk. this we examined colostrum obtained from 2 HHV-8-seropositive is known to the number of cells during the number of cellular DNA equivalents in the colostrum samples, both of the samples were PCR for HHV-8. is an that breast milk, which several and virus or cells from breast milk our suggest that HHV-8-infected cells may be present in the breast tissue and milk, and that breast milk is not a likely route of transmission of HHV-8 to the to detect HHV-8 in breast-milk samples, EBV and HIV-1 DNA were detected in the samples analyzed for HHV-8. are similar to of studies of EBV and HIV-1 in breast milk, which reported of for EBV and for HIV-1 the presence of EBV DNA in breast milk at the time of was on the being infected with HIV-1 at the time of delivery. The of HIV-1 infection of the with to the presence of EBV DNA in breast milk, to be We that maternal saliva be an of transmission because HHV-8 is not present in detectable in breast milk and infection appears to occur early in by other in or HHV-8 is detected in samples obtained from the oral of the from which they are this finding suggests that HHV-8 be via saliva horizontal transmission has been suggested previously among men who have sex with men and within is known that individuals and with KS all have of HHV-8 in the oral with of HHV-8 in the and this suggests that the oral cavity may be a of viral and reported that HHV-8 DNA was detected in saliva in of of a patient with KS. that similar viral were detected in the saliva of the of transmission via the oral In the present study, of HHV-8-infected mothers during the first 6 months after delivery of infants that of the mothers were positive for HHV-8 in the cells from saliva. This finding is in with studies of HIV-1-infected men in the United States The HHV-8 serostatus of the at 12 months after delivery was independent of the presence of HHV-8 in the saliva cells of the at a time infants were born to mothers with detectable viral whereas infants were born to mothers with viral DNA in saliva cells at the time of of the 3 mothers with HHV-8-infected children have for of viral in saliva cells by use of real-time PCR, which detection of a between virus and infection of the infants of these The lack of a between the presence of HHV-8 DNA in the saliva cells of a and the of HHV-8 infection in is not The of HHV-8 in the oral cavity has been to time in a cohort of in the United States In of longitudinal samples from the mothers in the present study suggests that the presence of HHV-8 in saliva is between time the time at which saliva from the mothers was obtained for analysis may not with the time at which transmission of HHV-8 to infants and routes of other and routes by other likely in this population, as has been suggested for similar In a study of with the prevalence of HHV-8 was among the and children of the among Although the presence of HHV-8 in saliva cells was independent of the HIV-1 infection of the to be a number of HHV-8 cells in HIV-1-infected mothers by use of the of was no significant in the of HHV-8 cells between HIV-1-infected mothers and Although this finding is that studies of HIV-1-infected men and individuals with KS have observed an of HHV-8 in HIV-1-infected the lack of a observed in the present study is likely the result of the number of samples for analysis and the of the HHV-8 in saliva with the present study other studies have suggested for regions HHV-8 is that through infected saliva is likely for the observed increase in seroprevalence through [4, 11]. We the first study of breast milk for the presence of HHV-8 DNA in a of Africa HHV-8 is this study is a cross-sectional analysis of specimens obtained from mothers, a longitudinal analysis, it is possible that the presence of HHV-8 in breast milk at other time have been the of the present study suggest that breast milk does not HHV-8 and is not a of transmission of HHV-8. detection of HHV-8 in saliva suggests that contact with saliva that virus may be a of transmission of HHV-8 to infants and children in regions of endemicity. We and our Zambian study for sample and We are to and for and We are also to for with the we to the in our study and infants for to Real-time polymerase chain reaction (PCR) analysis of the 16 saliva cell samples obtained from 12 who were found to be positive for human herpesvirus 8 (HHV-8) by conventional PCR. The of human and HHV-8 copies were and the of HHV-8 cells analyzed are for with the and 12 had real-time PCR performed at 2 time The time the saliva specimens were obtained are of the number of HHV-8 copies in saliva cell samples obtained from HIV-1-infected mothers and and as well as the and are samples obtained at 2 samples obtained at and samples obtained at 6 The the of between the first and of