Abstract

The reverse transcriptases (RTs) from human immunodeficiency viruses types 1 and 2 (HIV-1 and HIV-2, respectively) are relatively highly related yet there are several significant differences in their catalytic activities. In an attempt to relate these functional dissimilarities to the differences in amino acid sequences, we have employed a novel approach of constructing chimeric molecules composed of complementary amino acid sequences derived from the two HIV RTs. These recombinant proteins were analyzed for their enzymatic activities and for their sensitivity to tetrahydroimidazo[4,5,1-jk][1,4]benzodiazepin-2[1H]-one and thione (TIBO), which selectively inhibits only HIV-1 RT. The active chimeric RTs were used to map the TIBO binding site on the HIV-1 RT molecules and to localize the putative sequences responsible for the high RNase H activity of HIV-1 RT relative to that of HIV-2 RT. The results suggest that TIBO interacts with amino acid residues located around residue 200 within the DNA polymerase domain of HIV-1 RT which shows a relatively low similarity to HIV-2 RT. The difference in the RNase H activity maps to a position in the DNA polymerase domain rather than to the RNase H domain. Out of the 12 chimeric RTs generated, four were either fully active or hyperactive, three others lost most of their catalytic activities, and the rest were totally inactive. The pattern of catalytic activities of these hybrid proteins can be explained by a model for the initial folding of HIV RTs, which entails the formation of three distinct and independently folded regions. Each region can be formed by amino acid sequences derived exclusively from either HIV-1 RT or HIV-2 RT.