Much evidence suggests that the major immediate early (IE) transactivator o
f human cytomegalovirus (HCMV), IE-2, is likely to be critical for efficien
t viral replication; however, the lack of an IE-2 mutant HCMV has precluded
an experimental test of this hypothesis. As an initial step toward charact
erizing an IE-2 mutant, we first cloned the HCMV Towne genome as a bacteria
l artificial chromosome (BAC) and analyzed the ability of transfected Tonne
-BAC DNA (T-BACwt) to produce plaques following introduction into permissiv
e human fibroblasts. Like Towne viral DNA, transfected T-BACwt DNA was infe
ctious in permissive cells, and the resulting virus stocks were indistingui
shable from Towne virus. We then used homologous recombination in Escherich
ia coli to delete the majority of UL122, the open reading frame encoding th
e unique portion of IE-2, from T-BACwt. From this deleted BAG, a third BAC
clone in which the deletion was repaired with wild-type UL122 was created.
In numerous transfections of permissive human foreskin fibroblast cells wit
h these three BAC DNA clones, the rescued BAC and T-BACwt consistently yiel
ded plaques, while the UL122 mutant BAC never generated plaques, even after
4 weeks. Protein and mRNA of other IE genes were readily detected from tra
nsfected UL122 mutant BAC DNA; however, reverse transcription-PCR failed to
detect mRNA expression from any of five early genes examined. The generali
zed failure of this mutant to express early genes is consistent with expect
ations from in vitro assays which have demonstrated that IE-2 transactivate
s most HCMV promoters. These experiments provide the first direct demonstra
tion that IE-2 is required for successful HCMV infection and indicate that
virus lacking IE-2 arrests early in the replication cycle.