Design and Synthesis of HIV Integrase as Potential Anti-AIDS Drugs

Division Of Basic Sciences - Nci

Investigators

Linked publications & trials

Abstract

Inhibitors of HIV-1 integrase (IN) inhibitors represent the most recent anti-AIDS drugs. Merck's raltegravir (RAL) (October 2007) and Gilead's elvitegravir (EVG) (August 2012) were the first two IN inhibitors to be approved by the FDA. These agents are members of a class of drugs called IN strand transfer inhibitors (INSTIs), due to their ability to preferentially block the enzyme's strand transfer (ST) reaction as compared to the enzymes 3'-processing (3'-P) reaction. Treatment with RAL and EVG selects for resistant forms of HIV and there is considerable cross-resistance to these two drugs. GlaxoSmithKline's dolutegravir (DTG) was approved by the FDA in August of 2013 as a 2nd-generation INSTI having improved efficacies against RAL and EVG-resistant strains of HIV. However, DTG also selects for resistant strains of HIV, emphasizing the need for continued development of agents that can overcome resistant strains of IN, including the emerging DTG-resistant strains. Utilizing my laboratory's design and synthetic capabilities, we have teamed with pharmacologists (Dr. Yves Pommier, NCI), virologists (Dr. Hughes, NCI) and structural biologists (Dr. Cherepanov, London Research Institute) to develop new IN inhibitors. DTG is characterized by an extended tricyclic system, in which the third ring is known to be important, but the underlying reasons for this are poorly understood. We elaborated the core of our bicyclic 7,8-dihydroxy-3,4-dihydroisoquinolin-1(2H)-one based HIV-1 integrase inhibitors platform into the region IN catalytic site that is occupied by DTG's third ring and we found that appropriately tethering oxygen functionality at the 6-position improved inhibitory profiles. The potencies of these compounds equal or exceed those of DTG across an entire large panel of 18 single, 11 double and 6 triple HIV mutants, which represent the major INSTI-resistant forms. In collaboration with Dr. Cherepanov we obtained co-crystal structures of our best inhibitors bound to the prototype foamy virus (PFV) intasome (multimeric integrase with DNA substrate and metal cofactor). We found that there are overlaps in aspects of the 6-side chains that coincide with both unprocessed viral DNA in a pre-3'-P complex and the target DNA in a ST complex. We realized that this represents an unusual form of bi-substrate mimicry. It exemplifies the more general concept of substrate envelope, originally used to explain why some HIV protease inhibitors are able to retain efficacy against multiple resistance mutations. We found similar components of bi-substrate mimicry in other potent INSTIs having good profiles against resistant mutant forms of IN. This work has led us to postulate the general concept that substrate mimicry and substrate envelope may be broadly applicable in the design of INSTIs endowed with an ability to retain efficacy against resistant mutant forms of IN. Of note: This provides general guiding rationale for designing INSTIs that circumvent resistant mutant forms of IN. Most recently, we have identified additional substituents at the 6-position that are highly effective, with the best compound retaining better efficacy against the broad panel of known INSTI resistant mutants than any analogs we have previously described. Separately, we have begun investigating inhibitors of the polymerase (Pol) and RNase H domains of HIV- 1 reverse transcriptase (RT). Both RT and IN are members of a superfamily of polynucleotidyl transferases that use divalent metal cofactors. RT has two catalytic centers, a DNA polymerase (Pol) and an RNase H that cleaves RNA if it is part of an RNA-DNA duplex. These two activities collaborate to produce, from single-stranded genomic RNA, a double-stranded DNA that is the substrate for IN. At this time, there are no FDA-approved HIV-1 RNase H inhibitors and it remains as one of the last unutilized targets for AIDS therapeutics. Collateral cytotoxicity is a significant limitation of current RNase H inhibitors and achieving potent inhibition without concomitant cytotoxicity remains a major challenge to be addressed. Similar to the active site of IN, the active sites for both the Pol and the RNase H domains of RT have two bound Mg2+ ions and utilizing metal chelation may allow us to develop new classes of RT inhibitors directed at the Pol and RNase H active sites. This leaves open the possibility of developing metal-chelating inhibitors of Pol catalysis that may function in multiple ways. This idea is supported by the fact that metal chelation underlies the activity of INSTIs. In addition, there are metal-chelating RNase H inhibitors, although the compounds described in the literature have limited potency, and most are relatively toxic. As a consequence, no RNase H inhibitors have been approved by the FDA. We are exploring the use of metal-chelating molecules to inhibit RNase H as well as the Pol domains of RT. The original design of our metal-chelating INSTIs was derived from pyridopyrimidinone-containing motifs present in RNase H inhibitors developed by Merck. We have shown that structurally-related compounds inhibit both the Pol and RNase H activities of RT in biochemical assays and that the potency of these compounds is not affected, in a viral replication assay, by several mutations in RT that affect the potency of non-nucleoside reverse transcriptase inhibitors (NNRTIs). The relatively low cytotoxicity of the compounds is notable, because many of the RNase H inhibitors that have been described in the literature show significant toxicity. Based on our promising early results, we prepared derivatives that lack the moiety that specifically interacts with the viral DNA at the IN active site. Our intent was to obtain compounds that would retain affinity at either the polymerase or the RNase H active sites, but would decrease affinity at the IN active site. Although it is not a simple matter to correlate data from in vitro biochemical assays with efficacy against the virus, it appears that the potency of our best polymerase inhibitor is within an order of magnitude of nevirapine, an approved non-nucleoside RT inhibitor (NNRTI) in an in vitro assay. We have preliminary data showing that the antiviral potencies of the most potent compounds are not significantly affected by the G140S/Q148H double mutation in IN, which is frequently associated with INSTI resistance, nor by a panel of RT mutants (V106A, L100I/K103N, Y188L, Y181C, M184V, or K70R), which are broadly associated with resistance to NNRTIs. Importantly, the best of these compounds show considerable potency (IC50 of 200 nM) with only modest toxicity in sensitive cell-based assay that detects changes in the ATP level in the cells [therapeutic index (TI) 70 as compared with Merck's reference RNase inhibitor (TI = 4)].