Development and validation of a cell-based assay system to assess human immunodeficiency virus type 1 integrase multimerization

Abstract

Multimerization of HIV-1 integrase (IN) subunits is required for the concerted integration of HIV-1 proviral DNA into the host genome. Thus, the disruption of IN multimerization represents a new avenue for intervening HIV-1 infection. Here, we generated a cell-based assay system to assess IN multimerization using a newly constructed bimolecular fluorescence complementation (BiFC-IN) system. BiFC-IN proteins were efficient in emitting fluorescence, and amino acid (AA) substitutions associated with IN multimerization attenuated fluorescence, suggesting that the BiFC-IN system may be useful for evaluating the profile of IN multimerization. A recently reported non-catalytic site IN inhibitor (NCINI), which allosterically induces IN over-multimerization/aggregation, significantly increased fluorescence in the BiFC-IN system. An IN's substitution, A128T, associated with viral resistance to NCINIs, decreased the NCINI-induced increase of fluorescence, suggesting that A128T reduces the potential for IN over-multimerization. Moreover, E11K and F181T substitutions known to inhibit IN tetramerization also reduced the NCINI-induced fluorescence increase, suggesting that NCINI-induced IN over-multimerization was more likely to occur from tetramer subunits than from dimer subunits. The present study demonstrates that our cell-based BiFC-IN system may be useful in elucidating the profile of IN multimerization, and also help evaluate and identify novel compounds that disrupt IN multimerization in living cells.

Keywords: BiFC; HIV-1; HIV-1 integrase multimerization; NCINIs.

Fig. 1.. BiFC-IN expression system.

(A) The genes encoding the Venus-N-terminus protein (VN, residues 1–172) or Venus-C-terminus protein (VC, residues 155–238) were added to the N-terminus of full-length IN derived from pNL_4-3 through a linker, 5′-GGG GSG GGG S-3′ (10GS). (B) The schematic illustrating BiFC-IN system. VN or VC linked to the N-terminus of the IN, VN-IN and VC-IN are shown on the left side. When VN-IN and VC-IN proteins are expressed and dimerize within the cells, the two BiFC fragments (VN and VC) reconstitution into a complete Venus protein, which emits fluorescence as shown on the right side. The crystal structures of Venus, VN, and VC are taken from PDB ID 1MYW.

Fig. 2.. Localization of Venus-IN and BiFC-IN proteins in living cells, and flow cytometry gating to define cells expressing Venus.

(A) pVenus-IN or (B) pVN-IN plus pVC-IN (pBiFC-IN) were transfected into COS7 cells, after 48 h Venus fluorescence emitting from Venus-IN and BiFC-IN proteins were observed under the laser scanning microscopy. White arrow head indicates the nucleus stained with SYTO62. White scale bar: 25 μm. (C) Approximately 60% of HEK293T cells were gated with SSC and FSC (Gate 1) into Panel D and E. The horizontal axes of Panels D, E, and F denote Venus’ fluorescence intensity. (D) Shown are the quadrilateral gate (Gate 2) in negative cells containing intrinsic fluorescence, non-expressing Venus, defined to be under 1% in the left side, and (E) atypical fluorescence distribution pattern in positive cells expressing Venus fluorescence of BiFC-IN^WT on the right side. (F) The histogram of the positive cells (BiFC-IN^WT) from gate 2 (E) comprises the number of Venus expressing cells and Venus fluorescence intensity.

Fig. 3.. Assessment of nonspecific assembly in BiFC-IN system.

(A) The linkers of no linker (0GS), GGG GS (5GS), GGG GSG GGG S (10GS), or GGG GSG GGG SGG GGS (15GS) are applied in the pBiFC-IN. The schemas show possible situations of VN and VC reconstitution in the BiFC-IN systems containing 0GS, 5GS, 10GS or 15GS linkers. Two NTDs in an IN dimer taken from PDB ID 1K6Y, when 2 NTD-CCD monomers dimerize, are shown as light blue and pink cartoons, and the linkers shown as gray and black spheres. The broken line in BiFC-IN (0GS) illustration denotes the distance between N-termini of 2 NTD-CCD monomers at 14.2 Å measured by PyMOL software, when they dimerized. (B) HEK293T and 293T cells were transfected with pBiFC-IN containing 0GS, 5GS, 10GS, or 15GS linkers. After 48 h, each AUC value and expression of them were compared. All of the data shown represent mean values (±SD) derived from three independent experiments. Statistical significance was examined using the nonparametric Mann-Whitney U test, * P < 0.05, ** P < 0.005.

Fig. 4.. Expression and degradation of BiFC-IN proteins in cells.

(A) AUC values and expression of BiFC-IN^WT proteins were confirmed 12, 24, 36, and 48 h after the transfection. (B) HEK 293T and 293T cells were transfected with pBiFC-IN^WT, 24 h after the transfection DMSO and MG132 were added at various concentrations (10, 1, and 0.1 μM), and 48 h after the transfection fluorescence and the expression of BiFC-IN^WT proteins by FCM and Western blot were examined. The AUC ratios in the presence of MG132 were calculated compared to the mean of the AUC values in DMSO control. (C) Multimerization profiles of BS3 cross-linked BiFC-IN proteins. Following transfection of COS7 cells with each plasmid, the presumably multimerized BiFC-IN proteins were cross-linked by BS3, and Multimerized VN-IN, VC-IN, and VN-IN plus VC-IN proteins were detected by Western blot. VC-IN protein is shown in Lanes 1, 2, and 3; VN-IN protein in Lanes 4, 5, and 6; VN-IN plus VC-IN proteins in Lanes 7, 8 and 9. The concentrations of BS3 were at 0.1 mM (Lanes 2, 5, and 8), 0.25 mM (Lanes 3, 6, and 9), and no cross-linked samples (Lanes 1, 4, and 7). (D) The AUC ratios of BiFC-IN^E11K, -IN^K186E, -IN^E11K+K186E, and -IN^D116A were compared with that of BiFC-IN^WT. Shown in the middle panel are expression of the BiFC-IN proteins extracted from 293T cells using the anti-IN antibody and polyclonal anti-HA antibody. The monoclonal anti-IN antibody reacts with a peptide of the NTD (containing the E11 residue). Shown at the bottom are the amounts of beta-actin in each transfected 293T cell. Data shown represent mean values (±SD) derived from three independent experiments. Statistical significance was examined using the nonparametric Mann-Whitney U test, * P < 0.05.

Fig. 5.. BiFC-IN multimers can interact with endogenous LEDGF/p75.

(A) 293T cells were transfected with pVenus-IN, pBiFC-IN^WT, pBiFC-IN^V165A and pVenus. After 48 h, the expression of Venus-IN, BiFC-IN^WT, BiFC-IN^V165A and Venus proteins were examined by Western blot using anti-IN or anti-HA antibody. (B) (C) 293T cell lysates transfected with no plasmids, and pVenus-IN, pBiFC-IN^WT, pBiFC-IN^V165A and pVenus, were incubated with Protein G Mag Sepharose coupled to anti-LEDGF or anti-IN antibody. The eluted samples were examined by Western blot using the same anti-LEDGF or anti-IN antibody. All data are from two independent experiments and the representative results are shown.

Fig. 6.. IN over-multimerization induced by NCINIs.

(A)–(C) Multimerization of IN was examined in the presence of NCINI-1, NCINI-2, and NCINI-3 in the BiFC-IN system. The AUC ratios of BiFC-IN^WT (in closed columns), BiFC-IN^A128T (in gray columns), in the presence of the 3 NCINIs are shown. Note that Venus expressed alone (in open columns) confirms that the test compounds did not damage the cells or affect the Venus fluorescence. (D) Multimerization of BiFC-IN^E11K and -IN^F181T induced by NCINI-3 were examined compared to that of BiFC-IN^WT. The AUC ratios of BiFC-IN^WT (in closed columns), BiFC-IN^E11K (in dark gray columns), and BiFC-IN^F181T (in light gray columns) in the presence of NCINI-3 are shown. Data shown represent mean values (±SD) derived from three independent experiments. Statistical significance between the AUC ratios of BiFC-IN^WT in DMSO and the presence of various concentration of the NCINIs (A)–(C), and between those of BiFC-IN^WT and BiFC-IN^E11KorF181T in NCINI-3 (D) was determined using Student t-test.