Human immunodeficiency virus type 1 integrase: arrangement of protein domains in active cDNA complexes

Abstract

Early steps of retroviral replication involve reverse transcription of the viral RNA genome and integration of the resulting cDNA copy into a chromosome of the host cell. The viral-encoded integrase protein carries out the initial DNA breaking and joining reactions that mediate integration. The organization of the active integrase-DNA complex is unknown, though integrase is known to act as a multimer, and high resolution structures of the isolated integrase domains have been determined. Here we use site-specific cross-linking based on disulfide bond formation to map integrase-DNA contacts in active complexes. We establish that the DNA-binding C-terminal domain of one integrase monomer acts with the central catalytic domain from another monomer at each viral cDNA end. These data allow detailed modeling of an integrase tetramer in which pairs of trans interactions link integrase dimers bound to substrate DNA. We also detected a conformational change in integrase- DNA complexes accompanying cleavage of the viral cDNA terminus.

Fig. 1. Oligodeoxynucleotide modifications used for site-specific cross-linking. (A) Structure of the modified adenine residue in an A–T base pair. (B) Structures of the different modified thio groups used for cross-linking. The relative reactivity increases from the bottom to the top of the figure.

Fig. 2. Derivatives of the viral cDNA end used in this study. (A) The blunt-ended substrate S(Bl) mimics the wild-type viral cDNA at the U5 end. The site of cleavage by integrase is denoted by an arrow. Sites of modification are as illustrated. (B) The types of integrase substrates used in this study.

Fig. 3. IN mutants used in cross-linking reactions. (A) Illustration of the positions of the mutation sites in the two-domain (50–270) HIV-1 IN structure (J.C.-H.Chen et al., 2000) (1EX4). (B) Disintegration assay. (C) 3′-processing activity assay. The DNA substrates and products are as illustrated beside the gel.

Fig. 4. Cross-linking of IN mutants to the viral cDNA end S(Bl) containing different tether arms. The positions of the tether arm and the arm length are indicated above the gels. (A) IN mutants (wild type, R262C, S230C, E246C and I191C) cross-linked to S(Bl) carrying the most reactive DTNB-modified tether arm [S(Bl)-C2(TNB)]. (B) IN wild type and mutants cross-link to DNA substrates carrying the intermediate reactivity SH tether arm (S(Bl)-C2(SH)). (C) IN mutants cross-linked to DNA substrates carrying the least reactive SS tether arm [S(Bl)-C2(SS) or S(Bl)-C3(SS)].

Fig. 5. Effects of salt concentration (A) and βME concentration (B) on cross-linking of E246C with S(Bl)-7C3(SS). Wild-type IN is also included as a control. Concentrations are indicated above the gel.

Fig. 6. Cis–trans test of integrase function. (A) Schematic illustration of the cis–trans analysis. The dumbbell disintegration substrate S(Db)-7C3(SS) was pre-incubated with the indicated integrase derivatives to allow cross-linking, then the disintegration was initiated by addition of Mn²⁺. The predictions for cis or trans action with IN E246C/D64R and wild-type (WT) proteins are shown. See the text for details. (B) Autoradiogram illustrating the results of the complementation assay. Whether proteins were added sequentially or together in a pre-incubation step is specified above the gel. The structure of the products formed is illustrated beside the gel. (C) Staging of the cross-linking to D64R/E246C and addition of wild-type IN. The S(Db)-7C3(SS) substrate was cross-linked to D64R/E246C, and then subjected to further treatments: either no treatment (lane 3), exposure to 1 M NaCl, then reduction of the salt to 0.1 M NaCl (lane 2), or exposure to 1 M NaCl, addition of wild-type IN, then reduction of the salt to 0.1 M NaCl (lane 4). For comparison, a reaction in which D64R/E246C was pre-mixed with wild-type IN is shown in lane 1. Incubation with 1 M NaCl was for 30 min, then the salt concentration was reduced by dilution in buffer, the solution was adjusted to 5 mM MnCl₂ and the samples incubated an additional 20 min at 37°C. Samples were concentrated using a microcon YM-3 unit prior to electrophoresis.

Fig. 7. Biochemical assay of conformation change in different steps of integration. (A) Cross-linking assay of IN S230C with the indicated substrates. (B) Cross-linking assay of IN E246C with the indicated substrates. In all cases, the C2(SH) tether arm was used.

Fig. 8. A candidate model for the IN–DNA complex based on cross-linking and structural studies. (A) Illustration of the modeled interaction between one IN two-domain dimer and DNAs (target DNA, gray; viral cDNA ends, pink). The position of the active site in the catalytic domain is shown in red. The green diamond indicates the position of the E246–A7 cross-link. (B) Illustration of a complex containing two dimers of the two-domain IN bound to the DNA substrates. (C) Organization of an IN tetramer bound to DNA, illustrating a possible location for the N-terminal domain. (D) Overview of the complex model with integrase shown as the calculated Van der Waal’s surface (shaded light blue and blue, yellow and brown for the four monomers). Viral cDNA ends are indicated as pink ribbons, target DNA (bent) is indicated by the gray ribbons. Coordinates for the model are available upon request.