STRUCTURAL INVESTIGATION OF NONNUKLEOSIDE INHIBITORS OF HIV-1 REVERSE TRANSCRIPTASE

Grzegorz D. Bujacz 1, Natasza E. Ziólkowska1, Marshall L. Morningstar2 and Christopher J. Michejda2

1Institute of Technical Biochemistry, Technical University of Lódz, Stefanowskiego 4/10, 90-923 Lódz, Poland, gdbujacz@ck-sc.p.lodz.pl
2
Molecular Aspects of Drug Design Section, ABL, National Cancer Institute - Frederick Cancer Research and Development Center, P.O.Box B, Frederick, MD 21702, USA.

The enzymes encoded in the human immunodeficiency virus type 1 (HIV-1) genome have critical roles in the life cycle of the virus, which is the causative agent of AIDS. One such essential enzyme is reverse transcriptase (RT), an enzyme that contains both a DNA polymerase activity that can use either RNA or DNA as a template, and a ribonuclease-H activity [1]. These activities are crucial for the conversion of retroviral genomic RNA into the double-stranded linear viral DNA that subsequently integrates into the genome of the infected host cell. Inhibition of RT should therefore provide en effective means of blocking HIV-1 replication [2].

A number of inhibitors of HIV RT have been developed [3,4]. These inhibitors can be generally divided into two classes: nucleoside analogues and nonnucleoside RT inhibitors (NNRTI). As opposed to the nucleoside drugs, the NNRTIs do not function as chain terminators and do not bind at the dNTP binding site. The results of crystallographic studies suggest that most of these compounds share a common binding site located proximal to the RT polymerase active site [5-7]. It is characteristic that without the presence of NNRTI, the pocket responsible for its attachment does not exist. It is only upon binding an inhibitor that Tyr181, Thr188 and Thr228 move so as to create a hydrophobic pocket of sufficient volume to accommodate the inhibitor [6]. By binding to this allosteric site, the NNRTI increases the distance between the primer grip and the dNTP binding side in RT, suggesting this as a possible mechanism for the inhibition of polymerase activity [6]. The structural features that appear to be important for NNRTI binding include the presence of two planar -electron systems and the ability of those systems to adopt a conformation which has been designated as butterfly-like, with an orientation of the -systems such that they resemble the positoning of butterfly wings [7].

We utilized the known inhibitor of HIV-1 RT, 1-(2,6-difluorophenyl)-1H,3H- thiazolo[3,4-a]benzimidazole (TZB), as the lead structure for drug design, with the objective of making more potent inhibitors against the wild-type virus, but especially against mutant NNRTI's, the 1,2-disubstituted benzimidazoles [8]. We found that different substituents at the C4 position of the benzimidazole moiety had dramatically varied anti-viral activity against both the wild-type virus and many of the clinically derived resistant variants.

X = CH3
X = OCH3
X = NH2
X = NCH3COCH3

The best inhibitor, 1-(2,6-difluorobenzyl)-2-(2,6-difluorophenyl)-4-methoxybenzimidazole, (X =OCH3) has an IC50 = 7.3 nM against wild type RT in in vitro assays, and strong antiviral activity (EC50 = 4.6 nM) against wild type virus in cytopathic cell killing assays, while retaining broad activity against a number of clinically observed resistant HIV-1 strains.

Here we report on the crystallographic structures of four of the 4-substituted derivatives in an effort to try to understand the differences in that inhibitory.

Compound, X= CH3 OCH3 NCH3COCH3 NH2
Space group Pbcn P21/n C2/c P21/c
a (A) 32.130(9) 10.239(4) 18.106(4) 10.249(2)
b (A) 12.439(5) 9.482(2) 8.744(2) 14.117(3)
c (A) 8.766(2) 18.904(5) 27.202(5) 11.881(2)
b (o) 90.00 102.00(3) 108.33(3) 105.56(3)
V (A3) 3503.46 1795.22 4088.08 1656.00
No. of reflections 3596 3702 4210 2876
No. of parameters 294 310 340 297
R, F>4sF 0.0447 0.0453 0.0520 0.0399
t1 (o) 83.24(4) 78.93(3) 88.85(5) 69.48(4)
t2 (o) 67.24(4) 85.52(3) 52.21(4) 66.37(3)

t1-dihedral angle between benzimidazole ring and benzyl substituent
t2- dihedral angle between benzimidazole ring and phenyl substituent

One of the crucial geometrical parameters is the dihedral angle between the planes of the -electron systems (the angle between the butterfly's wings). There is a specific optimum for this parameter and only the compounds which meet it are potentially active . The active conformations determine the optimal torsion angles within the 2,6 difluorobenzyl and the benzimidazole ring systems, the components of the „butterfly". Structural investigation of a series of closely related compound provides useful information about the conformational preferences, important geometrical parameters and intramolecular interaction of -electron moieties. Crystal and molecular structure of NNRTI is a good base for modeling enzyme - inhibitor interactions. Using a single macromolecular structure of the RT-inhibitor complex, and X-ray data of series inhibitors, it is possible to check the interaction in a variety of complexes for different inhibitors and the RT-mutants. These structures will be compared with the structures of known RT/NNRTI complexes.

Research sponsored by the State Committee for Scientific Research (KBN) grant No 4P05F 03114 and by the National Cancer Institute, DHHS, under contract with ABL.

  1. R.A. Katz, A.M. Skalka, Annu. Rev. Biochem. 63 (1994) 133-73
  2. E. De Clerq, J. Med. Chem. 38 (1995) 2491-2517
  3. R.F. Schinazzi, Perspect. Drug. Discov. Des. 1 (1993) 151-180
  4. S.D. Young, Perspect. Drug Discov. Des. 1 (1993) 181-192
  5. C. Tantillo, J. Ding, A. Jacobo-Molina, R.G. Nanni, P.L. Boyer, S.H. Huges, R. Pauwels, K. Aries, P.A.J. Jansen, E. Arnold, J. Mol. Biol. 243 (1994) 369-387
  6. M.B.K. Smith, C.A. Rouzer, L.A. Taneyhill, N.A. Smith, S.H. Huges, P.L. Boyer, P.A.J. Jansen, H. Moereels, L. Koymans, E. Arnold, J. Ding, K. Das, W. Zhang, C.J. Michejda, R.H. Smith, Protein. Sci. 4 (1995) 2203-2222
  7. J. Ding, K. Das, H. Moereels, L. Koymans, K. Andries, , P.A.J. Jansen, S.H. Huges, E. Arnold, Nature. Struct. Biol. 2 (1995) 407-415
  8. .T. Roth, M.M. Morningstar, P.L. Boyer, S.H. Hughes, R.W. Buckheit, C.J. Michejda, J. Med. Chem. 40 (1997) 4199-4207.