Inheritance of chromosomally integrated viral DNA?

To the editor:

Inheritance of chromosomally integrated viral DNA?

In their recent report, Daibata et al1 describe the intriguing finding of human herpesvirus 6 (HHV-6) genome integration in 3 family members (mother, father, daughter). Using fluorescence in situ hybridization (FISH), the authors found HHV-6 DNA sequences integrated both into the long arm of chromosome 22 (22q13) in mother and daughter, and at chromosome locus 1q44 in father and daughter. For the 3 family members and the healthy control donors, FISH and polymerase chain reaction (PCR) studies were performed exclusively on peripheral blood cells. We fully agree with the authors that their observations demonstrate that HHV-6 genome can be integrated at specific sites, as they previously reported,2,3 but we object to their conclusion that the only way to explain the observation that the daughter has HHV-6 DNA integrated at both chromosome 22 (22q13) and chromosome locus 1q44 in blood cells is that she inherited integrated HHV-6 genome with her parents' chromosomes.

Indeed, inheritance of chromosomally integrated HHV-6 DNA is one possibility that should have been tested, as the formal proof can be obtained relatively easily. In order for the daughter to inherit HHV-6 from her parents' chromosomes, HHV-6 genome must have been integrated into the DNA of both the father's spermatocytes and the mother's oocytes. Thus, FISH analysis should reveal HHV-6 DNA integration at both 22q13 and 1q44 loci in all the daughter's cells, not only blood lymphocytes. FISH analysis of the daughter's other cells (fibroblasts, for example, from the inside of the cheek, or skin) should prove this hypothesis or not. In addition, the father's spermatozoids should be positive for HHV-6 DNA (PCR could prove that).

There is another explanation, however, perhaps as likely as chromosomal viral inheritance. Intrauterine viral transmission is established for HHV-6,4-6 so the daughter could have been infected during intrauterine life with her mother's virus, and HHV-6 DNA integrated at 22q13 in some/most of the daughter's cells (blood lymphocytes being likely host cells). After birth, the daughter could have been infected with her father's HHV-6 virus, again likely to infect blood lymphocytes, but this time integrating at locus 1q44. Again, this hypothesis can be verified by FISH analysis: if the hypothesis is correct, the daughter's blood cells should not be all positive for HHV-6 DNA integrated at both loci, which is actually what the authors observed,1 and more importantly, the daughter's other cells (fibroblasts) would very likely be negative for HHV-6 DNA integration at locus 1q44. Lastly, the father's sperm would most likely be negative for HHV-6 DNA.

In summary, we believe that Daibata et als conclusion, being unsufficiently documented, is premature, but a few additional and relatively simple experiments should enable Daibata et al to definitely prove whether chromosomal inheritance of HHV-6 DNA occurs, a finding that would be truly remarkable.

Response:

Presence of human herpesvirus 6 DNA in somatic cells

We previously detected chromosomally integrated human herpesvirus 6 (HHV-6) DNA in peripheral blood mononuclear cells (PBMCs) of family members of 3 generations at the identical chromosome site, and proposed the chromosomal transmission of HHV-6.1-1 In a recent report,1-2 we found another family in which chromosomal transmission of HHV-6 was suggested by the detection of integrated HHV-6 in PBMCs at the same chromosome loci. In agreement with Hermouet and Minvielle, to prove this novel mode of viral transmission, it seemed important to demonstrate the germline integration of the viral genome. To address this issue, we looked for the presence of HHV-6 DNA in skin fibroblasts from an asymptomatic 52 year-old man who had integrated HHV-6 in PBMCs at chromosome band 1q44.1-1 After having informed consent, pieces of skin tissue were biopsied from his left arm and were cultured in RPMI 1640 medium with 10% fetal bovine serum. Two weeks later, we harvested the sheet of fibroblasts, extracted DNA, and performed polymerase chain reaction (PCR) analysis with 7 distinct primer sets that target different regions of HHV-6 genome including the U4, U31, U57, U67, U89, U91, and U94 genes. These HHV-6 sequences were consistently detected by single-step PCR. We next performed fluorescence in situ hybridization (FISH) on metaphase chromosomes of the skin fibroblasts. We used 7 distinct viral probes that hybridize with different regions of HHV-6 (pH6Z-101, -109, -204, -205, -207, -606, -802, supplied by Dr P.E. Pellett, Center for Disease Control and Prevention, Atlanta, GA).1-3 All probes gave rise to symmetrical doublet signals on both chromatids of single homologue of chromosome 1 (1q44), showing that HHV-6 was integrated in the fibroblasts at the chromosome site identical to that of the PBMCs (Figure 1). In parallel experiments, none of them showed any hybridization signals in control PBMCs from 3 healthy adults.

Unfortunately, we could not obtain informed consent from the second family members1-2 to biopsy their skin for the study of the integrated HHV-6. Hermouet and Minvielle hypothesized that the daughter's lymphocytes were first infected with the mother's virus during pregnancy and then with her father's virus after birth. This hypothesis is less likely. Although HHV-6 can persist in PBMCs after primary infection, the amount of viral copies is invariably too low to be detectable by FISH. Macrophages are suggested to be a potential reservoir for latent infection.1-4 We believe that latent integration of the HHV-6 genome into the chromosomal DNA should be distinct from the latency after primary infection.

The present findings would provide convincing evidence that HHV-6 is constitutionally integrated and chromosomally transmitted. The mechanism of how HHV-6 enters cells of the germline lineage remains to be elucidated.