Phage display-directed discovery of LEDGF/p75 binding cyclic peptide inhibitors of HIV replication

Abstract

The interaction between the human immunodeficiency virus (HIV) integrase (IN) and its cellular cofactor lens epithelium-derived growth factor (LEDGF/p75) is crucial for HIV replication. While recently discovered LEDGINs inhibit HIV-1 replication by occupying the LEDGF/p75 pocket in IN, it remained to be demonstrated whether LEDGF/p75 by itself can be targeted. By phage display we identified cyclic peptides (CPs) as the first LEDGF/p75 ligands that inhibit the LEDGF/p75-IN interaction. The CPs inhibit HIV replication in different cell lines without overt toxicity. In accord with the role of LEDGF/p75 in HIV integration and its inhibition by LEDGINs, CP64, and CP65 block HIV replication primarily by inhibiting the integration step. The CPs retained activity against HIV strains resistant to raltegravir or LEDGINs. Saturation transfer difference (STD) NMR showed residues in CP64 that strongly interact with LEDGF/p75 but not with HIV IN. Mutational analysis identified tryptophan as an important residue responsible for the activity of the peptides. Serial passaging of virus in the presence of CPs did not yield resistant strains. Our work provides proof-of-concept for direct targeting of LEDGF/p75 as novel therapeutic strategy and the CPs thereby serve as scaffold for future development of new HIV therapeutics.

Figure 1

Phage display biopanning strategy to identify lens epithelium-derived growth factor (LEDGF)/p75 ligands. (a) Schematic representation of domains organization of LEDGF/p75 (gray) and integrase proteins (blue). The PWWP domain at the N-terminus of LEDGF/p75 provides a critical chromatin recognition function. The C-terminus (in yellow), represented as LEDGF_325–530, contains the integrase-binding domain (IBD). HIV-1 integrase (IN) (blue) is composed of three domains: the N-terminal domain (NTD), the catalytic core domain (CCD) and the C-terminal domain (CTD). (b) Phage display peptide libraries were preabsorbed using paramagnetic beads coated with wild-type HIV-1 integrase (WT IN) to deplete binders to beads and to WT IN (purple, blue phages; preabsorption). Supernatant with residual phages was subsequently used to isolate specific ligands of LEDGF_325–530 immobilized on beads (red, yellow phages; positive selection). Bound phages were eluted via pH shift, amplified and used for another round of selection. Overall three rounds of selections were performed. HIV, human immunodeficiency virus.

Figure 2

Saturation transfer difference (STD) NMR analysis for the lens epithelium-derived growth factor (LEDGF)/p75 interacting cyclic peptides (CPs). The recorded STD spectra plotted for each CP. At the top, a reference 1D ¹H NMR spectrum is shown for each CP displaying very narrow resonances typical for short peptides. A corresponding control STD NMR experiment with peptides only showed no signal indicating that the impurity contained in the spectrum can effectively be subtracted and therefore does not give rise to false positive signals in the difference spectrum obtained in the presence of the target proteins. STD NMR spectra for (a) CP64 and (c) CP65 showed specific signals only in the presence of LEDGF_325–530 but not with HIV-1 integrase (IN) catalytic core domain (CCD) demonstrating that CPs bind specifically to LEDGF_325–530. In contrast, (b) CP64m and (d) CP65m bind neither to LEDGF_325–530 nor to IN CCD. HIV, human immunodeficiency virus.

Figure 3

The cyclic peptides (CPs) do not interfere with cellular functions of lens epithelium-derived growth factor (LEDGF)/p75. (a) Chromatin-binding analysis performed as described before.^, A representative western blot of the different cellular fractions is shown using antibodies against the indicated proteins. Extracts were made from (i) CP64-mRFP-, (ii) CP64m-mRFP-, (iii) CP65-mRFP-, (iv) CP65m-mRFP-, (v) mRFP-expressing cells, or (vi) the A3 (LEDGF/p75 depleted) cell line. T, total cell lysate; S1, Triton-soluble cellular fraction; P1, Triton-insoluble cellular fraction; S2, DNase/(NH₄)₂SO₄-soluble cellular fraction; P2, DNase/(NH₄)₂SO₄-insoluble cellular fraction. (b) Fluorescence correlation spectroscopy (FCS) experiment with transiently expressed wild-type (WT) eGFP-LEDGF/p75 in each of the different CP-mRFP cell lines. The autocorrelation plots reveal that only the PWWP double-mutant K56D-R74D (eGFP-LEDGF/p75 K56D-R74D), displays impaired chromatin binding and scanning as described before. Data represent mean values of five independent measurements for each condition. GFP, green fluorescent protein.

Figure 4

Antiviral activity of the lens epithelium-derived growth factor (LEDGF)/p75 interacting cyclic peptides (CPs). (a,b) HeLaP4 cells stably expressing CP63-, CP64-, CP64m-, CP65-, CP65m-mRFP, or mRFP alone were infected with (a) HIV-1_NL4.3 or (b) HIV-2_ROD and the p24 level was quantified in the supernatant at the indicated time points. (c) Similar experiments with HIV-1_NL4.3 were performed in PM1 cells expressing the above peptides and p24 was measured in the supernatant at the indicated time points. (d) NIH3T3 cells stably expressing the above fusion proteins were challenged with replication competent MLV_GFP-MOV strain and green fluorescent protein (GFP) expression was monitored by fluorescent-activated cell sorter (FACS) at the indicated time points. HIV, human immunodeficiency virus.

Figure 5

The antiviral activity of lens epithelium-derived growth factor (LEDGF)/p75 interacting peptides is dependent on the conserved tryptophan residue. The antiviral activity of the peptides is dependent on the tryptophan residue and its orientation in the common motif. A scrambled peptide with conserved tryptophan position (CP65s¹) retains its antiviral activity while the random scrambled peptide (CP65s²) lost its activity completely as evaluated by NL4.3 viral breakthrough in HeLaP4 cells expressing CP64-, CP65-, CP65m-, CP65s¹-, CP65s²-mRFP, or mRFP alone. Mean values ± s.d. from at least duplicate measurements of representative experiment is shown.

Figure 6

Mechanism of action studies for CP64 and CP65. (a–c) Show the cross-resistance profile of CP64 and CP65. Activity against raltegravir resistant strains, (a) HIV-1_G140S/Q148H and (b) HIV-1_N155H. (c) Antiviral activity against a strain resistant to transdominant inhibition by IBD, HIV-1_A128T/E170G. (d) Human immunodeficiency virus (HIV) DNA quantification by quantitative PCR (qPCR) analysis to determine the effect of cyclic peptides (CPs) on the levels of late reverse transcriptase (RT) product (total HIV DNA) at 10 hours postinfection. Integrated copies were quantified after 7 days to ensure that only the integrated DNA species was measured. The data is presented as percentage with respect to cells expressing only mRFP (control). AZT and raltegravir are included as controls for inhibition of RT and integration. (e) Effect of CPs on virus production from the NL4.3 plasmid transfected in HeLaP4 cells that express the respective peptides as quantified by p24 enzyme-linked immunosorbent assay (ELISA). Raltegravir and ritonavir were included as early and late stage inhibitor controls. (f) Lens epithelium-derived growth factor (LEDGF)/p75 interacting CPs significantly reduces the infectivity of viral progeny. 48 hours post-transfection, the supernatants from cells expressing CP65 or mRFP were transferred to HeLaP4 cells expressing the mRFP alone (control) and the amount of p24 was quantified 72 hours postinfection. As control, NL4.3 was produced in HeLaP4 cells expressing mRFP in the presence of either raltegravir (30 nmol/l) or ritonavir (300 nmol/l) and the infectivity of the virus was assessed in HeLaP4 control cells. Data represent mean values ± s.d. from three independent experiments.