A liver-specific nuclear factor interacts with the promoter region of the large surface protein gene of human hepatitis B virus.

The outer envelope of the 42-nm virion of the human hepatitis B virus (HBV) is composed of the large, the middle, and the major surface proteins. Whereas the middle and the major surface proteins are transcribed from the SPII promoter of the pre-S/S gene, the large surface protein is transcribed from the SPI promoter located upstream of SPII. We have previously shown that transcription of SPI (comprising nucleotides [nt] -380 to +17) occurs preferentially in differentiated hepatoma cell lines (H.K. Chang and L.P. Ting, Virology 170:176-183, 1989). In this report, we further demonstrated that a sequence of 95 base pairs in the upstream region of SPI (nt -95 to +17) was necessary and sufficient for such preferential expression in differentiated hepatoma cells. By analysis of the expression of the chloramphenicol acetyltransferase gene in a series of mutants with deletions at the 5' end of SPI, we identified a positive transcriptional cis-acting element mapping at nt -95 to -72 which appears to play a key role in the regulation of the expression of the large surface protein. This region shared a high degree of sequence homology with regulatory sequences of several liver-specific genes from human, mouse, and rat, with a consensus sequence (G/A)GTTA(A/C)TNNT(C/T)NNC(A/C). We further identified a nuclear factor present in the nuclear extracts of differentiated human hepatoma cell lines which interacted specifically with this element of the SPI promoter. This nuclear factor was similar to the rat liver-specific factor HNF-1, since an oligonucleotide containing the recognition sequence of HNF-1 could efficiently compete for the human factor in a footprinting assay. The sequence at nt -93 to -68 which was bound by this factor in SPI was termed the HNF-1-binding element. Activation of the SPI promoter by human differentiated hepatocyte nuclear factor 1, described in this report, probably explains, first, the formation of the 42-nm virion specifically in liver but not in several other tissues despite the synthesis of the middle and the major surface proteins in those tissues, and second, why only differentiated hepatoma cell lines are able to produce 42-nm-like virion particles on transfection by HBV DNA.