Evaluation of Selected Chemotypes in Coupled Cellular and

3606

J. Med. Chem. 1996, 39, 3606-3616

Expedited Articles Evaluation of Selected Chemotypes in Coupled Cellular and Molecular Target-Based Screens Identifies Novel HIV-1 Zinc Finger Inhibitors William G. Rice,*,† Jim A. Turpin,† Catherine A. Schaeffer,† Lisa Graham,† David Clanton,‡ Robert W. Buckheit, Jr.,§ Daniel Zaharevitz,⊥ Michael F. Summers,| Anders Wallqvist,∇ and David G. Covell∇ Laboratory of Antiviral Drug Mechanisms and Anti-AIDS Virus Drug Screening Laboratory, Developmental Therapeutics Program, Division of Cancer Treatment, Diagnosis and Centers, National Cancer Institute-Frederick Cancer Research and Development Center, SAIC Frederick, Frederick, Maryland 21702, Virology Research Group, Southern Research Institute-Frederick Research Center, 431 Aviation Way, Frederick, Maryland 21702, Developmental Therapeutics Program, Division of Cancer Treatment, Diagnosis and Centers, National Cancer Institute, Bethesda, Maryland 20852, Howard Hughes Medical Institute and Department of Chemistry and Biochemistry, University of Maryland, Baltimore County, Baltimore, Maryland 21228, and Division of Basic Sciences, National Cancer Institute-Frederick Cancer Research and Development Center, SAIC Frederick, Frederick, Maryland 21702 Received May 24, 1996X

Conservation of the Cys-Xaa2-Cys-Xaa4-His-Xaa4-Cys retroviral zinc finger sequences and their absolute requirement in both the early and late phases of retroviral replication make these chemically reactive structures prime antiviral targets. We recently reported that select 2,2′dithiobisbenzamides (DIBAs) chemically modify the zinc finger Cys residues, resulting in release of zinc from the fingers and inhibition of HIV replication. In the current study we surveyed 21 categories of disulfide-based compounds from the chemical repository of the National Cancer Institute for their capacity to act as retroviral zinc finger inhibitors. Aromatic disulfides that exerted anti-HIV activity tended to cluster in the substituted aminobenzene, benzoate, and benzenesulfonamide disulfide subclasses. Only one thiuram derivative exerted moderate antiHIV activity, while a number of nonaromatic thiosulfones and miscellaneous disulfide congeners were moderately antiviral. Two compounds (NSC 20625 and NSC 4493) demonstrated antiHIV activity in proliferating T cell cultures and in nonproliferating monocyte/macrophage cultures. The two compounds chemically modified the p7NC zinc fingers in two separate in vitro assays, and interatomic surface molecular modeling docked the compounds efficiently but differentially into the zinc finger domains. The combined efforts of rational drug selection, cell-based screening, and molecular target-based screening led to the identification of zinc finger inhibitors that can now be optimized by medicinal chemistry for the development of biopharmaceutically useful anti-HIV agents. Introduction In the search for novel therapeutic modalities against HIV-1 infection and AIDS, the retroviral zinc finger domain represents a rational target for antiviral chemotherapy. Two copies of the nonclassical Cys-Xaa2Cys-Xaa4-His-Xaa4-Cys (CCHC) retroviral zinc finger sequence are contained in the HIV-1 p7 nucleocapsid (p7NC) protein, which is a maturational proteolytic product of the Pr55gag precursor polyprotein.1-3 The Cys and His zinc-chelating residues and the spacing of the CCHC motif are absolutely conserved in all onco and lenti retroviruses.4,5 While still a part of the Pr55gag * Corresponding author. Mailing address: Laboratory of Antiviral Drug Mechanisms, Developmental Therapeutics Program, National Cancer Institute-Frederick Cancer Research and Development Center, SAIC Frederick, Building 431T-B, P.O. Box B, Frederick MD 217021201. Phone: (301) 846-5060. Fax: (301) 846-6846. Electronic mail address: [email protected]. † Laboratory of Antiviral Drug Mechanisms, National Cancer Institute-Frederick Cancer Research and Development Center. ‡ Anti-AIDS Virus Drug Screening Laboratory, National Cancer Institute-Frederick Cancer Research and Development Center. § Southern Research Institute-Frederick Research Center. ⊥ National Cancer Institute, Bethesda. | University of Maryland, Baltimore County. ∇ Division of Basic Sciences, National Cancer Institute-Frederick Cancer Research and Development Center. X Abstract published in Advance ACS Abstracts, August 15, 1996.

S0022-2623(96)00375-5 CCC: $12.00

(Gag) precursor, the fingers function in selection and packaging of viral RNA into budding virions.6-9 After the p7NC protein is fully processed from the Gag precursor in the mature virus, the fingers are required for completion of reverse transcription during the initial infection of target cells.10-13 Thus, the CCHC domain is highly conserved and performs essential roles in multiple phases of the HIV-1 replication cycle. Recently, we found that certain electrophilic reagents such as 3-nitrosobenzamide (NOBA),14 disulfide-substituted benzamides (DIBAs),15 and various other chemotypes16,17 can covalently modify the cysteine residues of retroviral zinc fingers, resulting in release of zinc from the fingers. NOBA and DIBAs entered intact HIV-1 virions, chemically modified the p7NC zinc finger cysteine residues, and inactivated virus infectivity.14,15 Through an electrophilic attack on the p7NC protein zinc fingers, the compounds blocked the ability of the virus to synthesize proviral DNA, even though there was no direct inhibition of reverse transcriptase enzymatic activity or virion binding to target cells.18,19 In addition, DIBAs chemically modified the zinc fingers of Gag precursors during viral synthesis, rendering the © 1996 American Chemical Society

Novel HIV-1 Zinc Finger Inhibitors

Journal of Medicinal Chemistry, 1996, Vol. 39, No. 19 3607

Table 1. Summary of Evaluated Disulfides Total Table No. Moderate Active No. aromatic disulfides aminobenzene benzoate benzenesulfonamide nitrobenzene hydrophobic-substituted benzenes halogenated benzenes phenolic naphthyl quinolyl thiazole/benzothiazole imidazole benzonitrile pyridine pyrimidine aliphatic disulfides nonsymmetrical disulfides thiuram disulfides aromatic disulfoxides aromatic disulfones aromatic thiosulfones nonaromatic thiosulfones summary of active compounds

25 13 6 10 5 7 7 2 2 2 2 4 3 1 13 7 26 3 4 4 3 24

4 8 4 0 0 0 0 0 0 1 0 2 0 0 1 0 1 0 0 1 3

1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

2 3 4 5 6 6 6 7 7 7 7 7 7 7 8 9 10 11 11 11 11 12

viral replication assay that makes use of the proliferating CEM-SS T-cell line. Efficacious compounds were then tested for anti-HIV-1 activity in nonproliferating fresh human peripheral blood monocyte/macrophage cultures and for reaction with purified HIV-1 p7NC protein in vitro and resultant zinc ejection. Those compounds scoring active by all three criteria were then evaluated against various HIV-1 molecular targets to ensure target specificity. These efforts have identified several subclasses of disulfides as structures for optimization through medicinal chemistry, as well as two high-priority zinc finger inhibitors that will undergo advanced evaluation. Results and Discussion Anti-HIV-1 Activity of Selected Compounds Categorized by Structural Subclasses. Because we had previously demonstrated that a select group of disulfide compounds,16 including the 2,2′-dithiobisbenzamides (DIBAs),15,17 inhibit HIV-1 replication by acting on the retroviral zinc finger motif, we searched the NCI chemi-

Table 2. Inhibition of HIV-1 Replication by Amine-Derivatized Aromatic Disulfides

XTT assay NSC

R1

R2

R3

R4

R5

8186 38069 58950 83217 71033 62984 136554 165587 20625 52085 372003 220 38070 677447 677488 677549 677478 677551 677548 677550 677492 677496 677495 677498 677494

NH2 NH2 NH2 NH2 NH2 H H NHSO2-4′-MePh NHC(NH2)dNH H H H H NH2 NHC(O)NH-3Cl,4F-Ph NCH-2′-OHPh H H H H H H H H H

H H H H NO2 H H H H H H H H H H H H H H H H H H H H

H CO2H H SO2NH2 H NH2 NH2 H H NCS NdNOH NHC(O)(CH2)10Me NHCOCH2Cl Cl H H NHC(O)CHdCHCO2H NCHPh NCH-4′-ClPh NCH-2′-OHPh NCH-4′-MePh NCH-4′-N(Me)2Ph NCH-4′-NHACPh NCH-2′-OH-naphthyl NCHCHCHPh

H H SO2NH2 H OMe H OMe H H H H H H H H H H H H H H H H H H

H H H H H H H H H H H H H H H H H H H H H H H H H

precursors nonrecognizable by HIV-1 protease and preventing maturation of newly forming virion particles.20 The National Cancer Institute’s chemical repository represents a rich source of diverse chemical structures as potential zinc finger inhibitors. The repository contains over 400 000 compounds, approximately half of which are nonproprietary in their classification and can be utilized in studies of structure-activity relationships for selected molecular targets. To narrow the field of chemotypes to a manageable number of compounds for study, we initiated a limited search of the repository for disulfide-based compounds. Because many disulfide compounds are toxic to proliferating cells, the experimental compounds were initially evaluated for anti-HIV-1 activity with a cell-based

EC50 31.2 14.1 4.9 0.47

1.2

IC50 5.4 >200 >200 73.1 4.0 29.6 26.4 5.05 38.4 6.3 81.2 >200 0.8 12.1 12.2 8.1 >200 33.3 23.4 27.7 26.8 25.0 23.6 31.1 16.5

TI50 >6.4 >14.2 14.9 8.5

32

activity I M M M M I I I A I I I I I I I I I I I I I I I I

cal repository for other lead candidates from this general class of compounds. Compounds were selected using a substructure search in the NCI Drug Information System for the general structures Ar-S-S-Ar or C-S-SC, where Ar refers to five- or six-membered aromatic rings. Among the many disulfide compounds available in the NCI chemical repository, 142 nonproprietary compounds were selected for study. The various categories of disulfide-based compounds studied are listed in Table 1. This table can serve as a quick guide, as it contains the total number of compounds and the number of moderately and fully antiviral compounds in each category, as well as the table number in the text that describes each class of compounds. From a total of 25 aminobenzene-derivatized aromatic disulfides (Table 2), four demonstrated moderate anti-

3608

Journal of Medicinal Chemistry, 1996, Vol. 39, No. 19

Rice et al.

Table 3. Inhibition of HIV-1 Replication by Carboxyl/Carboxamide-Derivatized Aromtic Disulfides

XTT assay NSC

R1

R2

R3

R4

R5

EC50

405662 633231 633232 372683 210281 121513 5343 624187 211 994 38069 4492 4493

CO2H CO2H CO2H CO2H H H CO2(Me) CO2EtS(O)-OMe H H NH2 C(O)NHNH2 C(O)NHNCH-2′,4′-Cl2-Ph

H H Me H H Me H H H H H H H

H Me H OMe CO2H CO2H H H CH2CO2 CH3 CO2H H H

H H Me H H H H H H H H H H

H H H H H Me H H H H H H H

117.7 46.9 123.5 28.4 96.2 23.1 68.1 0.73 12.6

IC50

TI50

>200 >200 125 >200 >200 102.4 18.6 >200 >200 3.5 >200 3.5 >200

>1.7 2.7 >1.6 3.6 >2.1 >8.7 >2.9 5.0 >15.9

activity M I M I M M I M M I M M A

Table 4. Inhibition of HIV-1 Replication by Sulfonium-Derivatized Aromatic Disulfides

XTT assay NSC 58950 83217 8186 255089 677463 666724

R1 NH2 NH2 NH2 NO2 NO2 –SO2

N N N

R2

R3

R4

R5

EC50

IC50

H H H H H H

H SO2NH2 H SO2NH2 SO2Me Me

SO2NH2 H H H H Cl

H H H H H H

27.1 18.0

127.5 65.2 5.4 >200 34.5 97.6

34.6 6.2

TI50 4.7 3.6 >5.8 15.7

activity M M I M I M

N

HIV-1 activity. Simple unsubstituted aminobenzenebased disulfides (NSC 8186 and NSC 62984) were inactive, whether the amine group was located in the ortho (R1) or para (R3) position; but, substitution of an electron-withdrawing group, such as CO2H, SO2NH2, or NO2, onto the aminobenzene backbone yielded compounds with anti-HIV-1 activity (see NSC compounds 38069, 58950, 83217, and 71033). The only truly active compound in this series contained a derivatized amino group at the R1 position (NSC 20625). The derivatization produced an electron-withdrawing group at the R1 position, which would delocalize electrons away from the sulfur atoms and enhance their electrophilic potential. Thirteen carboxyl/carboxamide-derivatized aromatic disulfides were evaluated (Table 3), with eight demonstrating moderate anti-HIV-1 activity and one (NSC 4493) demonstrating more impressive in vitro efficacy in the XTT cytoprotection assay. In the series shown in Table 3, moderate antiviral activity correlated with the presence of the carboxyl electron-withdrawing group, provided there was no hydrophobic residue in the R3 position. Among the sulfonium-derivatized aromatic disulfides (Table 4), four compounds were moderately antiviral but none were considered as strong antiviral candidates. None of the compounds from the nitro-substituted aromatic disulfides shown in Table 5 demonstrated antiHIV-1 activity. The only nitro-substituted compound

to do so was NSC 71033, which was listed in Table 2 as a modified aminobenzene derivative. No compounds exerted anti-HIV-1 activity from the group of aromatic disulfides substituted with hydrophobic, halogen, or hydroxyl-based groups (Table 6). Three moderately active compounds were found in the thiazole- and nitrile-substituted aromatic disulfides (Table 7), but no actives were found in several other classes of miscellaneous aromatic disulfides. Typically, aliphatic disulfides are inactive against HIV-1; yet, NSC 28727 (Table 8) was moderately antiviral. This compound did not react chemically with the HIV-1 p7NC protein zinc finger in vitro (see Table 12), suggesting that the moderate antiviral activity of this compound was due to interference with an antiviral target other than the zinc fingers. None of the asymmetric disulfides tested were active against HIV-1 (Table 9), although this does not rule out the possibility that other compounds of this category might be active. The antiviral data for a variety of thiuram disulfide congeners are shown in Table 10 (A and B). Only NSC 20871 was moderately antiviral, but it was not considered as a viable candidate for advanced studies. Generally, the thiurams were too toxic to the proliferating T-cell cultures for there to be an observable antiviral effect, and the thiurams tested were also inactive against HIV-1 in nonproliferating monocyte/macrophage cultures (data not shown). Even though the compounds



Table 5. Inhibition of HIV-1 Replication by Nitro-Derivatized Aromatic Disulfides

Table 6. Inhibition of HIV-1 Replication by Aromatic Disulfides Substituted with Hydrophobic, Halogenated, or Hydroxyl Groups

XTT assay NSC

R1

R2

R3

R4

R5

IC50

activity

203 677489 677462 677463 677457 677455 677441 677442 677446 158411

NO2 NO2 NO2 NO2 NO2 NO2 H H H NO2

H H H H H H NO2 NO2 H H

H C(O)Ph Me SO2Me CF3 NO2 H Cl NO2 CO2Me

H H H H H H H H H H

H H H H H H Cl H H H

13.1 14.0 15.5 0.16 13.1 9.2 15.4 17.9 24.0 11.3

I I I I I I I I I I

were not antiviral, a number of them were reactive in vitro with the zinc fingers of purified p7NC protein in the high throughput TSQ assay that measures zinc ejection from the fingers (Figure 1). Thiuram congeners containing the cyclopentamethylene rings (NSC compounds 527035, 402538, and 403854) demonstrated rapid reaction rates with the p7NC zinc fingers. NSC 527035 elicited an apparent rate of reaction that was less than the other cyclopentamethylene derivatives. However, the actual rate of reaction with the fingers was greater than depicted, as the NSC 527035 also exhibited a competing reaction with the zinc:TSQ complex that caused an artificially low signal in the high throughput assay. The high rate of reactivity of this series of compounds with the zinc fingers suggests that they might offer a template for drug design. On the other hand, the highly reactive nature of the cyclopentamethylene-derivatized thiurams might also be generally more reactive with other biological substrates and as a consequence demonstrate greater nonspecific cytotoxicity. The thiuram compounds illustrate the benefit of performing coupled cellular and mechanism-based assays and not relying solely on the molecular targetbased assays. Exclusive dependence on the molecular target-based screen for zinc finger reactivity in the early stages of selection of lead compounds would have yielded false positives having no discernable in vitro antiviral activity. Having stated this, it is also possible that the in vitro toxicities do not accurately reflect the in vivo toxicity profiles, and it is prudent to examine the literature for known in vivo data for the compounds in question. As an example, NSC 25953 is also known as Antabuse, an FDA-approved drug for the treatment of alcoholism. The compound may not be useful as an antiHIV-1 agent, but it raises the possibility that a complete reliance on the in vitro toxicity data might be misleading in certain instances. Anti-HIV-1 Activity of Atypical Disulfide-Based Compounds. In addition to the various types of aromatic and nonaromatic disulfides presented thus far, several disulfoxides, disulfones, and thiosulfones were also evaluated (Table 11). Four thiosulfones demonstrated anti-HIV-1 activity (NSC compounds 342015, 112801, 76302, and 342031). Identification of the thiosulfones as active against HIV-1 revealed that sulfur-based compounds other than the classic disulfide

XTT assay category

NSC

R1

R2

R3

R4 R5

IC50

activity

hydrophobic 158725 403320 994 158727 38073

Me Me H H H

H H H H H

H Me Me tBu CH2Ph

H H H H H

H Me H H H

3.6 28.0 3.5 12.4 27.2

I I I I I

halogenated 32025 202618 982 238936 992 677445 677461

H Cl Cl Cl H I H

H H H H H H Cl

Cl Cl H Cl Br H Me

H H Cl Cl H H H

H 5.2 H 6.8 H 13.7 H >200 H 3.4 H 8.8 H >200

I I I I I I I

H Cl CO2H H Cl

H H H OMe H

H H Br H H

H H H H H

63.4 34.7 122.4 20.6 17.5

I I I I I

H

H

Me H

8.7

I

OMe H

H H

17.2

I

phenolic

45168 39632 314654 995 39633

OH OH OH H OC(O)NHMe 39620 OC(O)NHMe 677472 OMe

moiety can serve as templates for medicinal chemistry efforts to identify more efficacious analogues. These thiosulfones, except for NSC 342031, demonstrated rapid kinetic zinc ejection from the zinc fingers of purified p7NC protein (see Table 12). Second Stage Evaluation of Antiviral Compounds. From the disulfide compounds tested in the proliferating T cell antiviral assay, 24 compounds demonstrated activity. Each was tested for the ability to inhibit replication of the monocytotropic HIV-1ADA in nonproliferating fresh human monocyte/macrophage cultures and to chemically modify and eject zinc from the HIV-1 p7NC protein in vitro using both the high throughput TSQ zinc fluorochrome assay and the intrinsic Trp37 fluorescence assay (Table 12). Compounds found not to inhibit HIV-1 replication in the monocyte/ macrophage model of HIV-1 infection were considered inactive (I). Compounds demonstrating anti-HIV-1 activity in both models but having relative therapeutic indices (TI50 ) IC50/EC50) of less than 10 in either model were considered as moderately active (M). The inactive and moderately active compounds listed in Table 12 were not pursued further due to insufficient antiviral activity, even though many of these compounds reacted rapidly with the p7NC protein zinc fingers. NSC 20625 was considered active in both anti-HIV-1 models and was reactive with the retroviral zinc finger domains by both assay methodologies. NSC 4493 was also considered as an active antiviral compound. Although NSC 4493 demonstrated zinc finger reactivity in the Trp37 assay for zinc ejection, it appeared relatively inactive in the high throughput TSQ assay. However, in more advanced concentration-dependency studies with the TSQ assay performed on a single sample basis we observed that NSC 4493 did indeed promote zinc ejection from the p7NC protein (data not shown). Together, these findings demonstrate that NSC 4493

3610


Rice et al.

Table 7. Inhibition of HIV-1 Replication by Miscellaneous Aromatic Disulfides

R-SS-R

and NSC 20625 represent lead candidates for further evaluation as potentially useful antivirals. Molecular Modeling of the HIV-1 Zinc Fingers and Ligand Docking. NSC 20625 and NSC 4493

clearly possess the appropriate chemical reactivities that allow them to mediate an electrophilic attack on the zinc-sulfur bonds of HIV-1 zinc fingers. However, no information to date has addressed the issue of how



Table 8. Inhibition of HIV-1 Replication by Aliphatic Disulfides (R1 ) R2)

R-SS-R

XTT assay NSC

R1 ) R2

677460 677439 677543 677465 677434 677454 677453 677468 677544 677452 677469 677471 28727

C(O)Ph CH2CH2CH(Me)2 C(NH2)NH CH2Ph tBu (CH2)5Me C(Me)2Pr C(Me)2C(Me)2CN CH2CH2CO2H CH2CO2Bu C(Me)2tBu CH2C(O)Ph CH2C(O)NH2

EC50

IC50

12.9

23.2 >200 100.7 12.4 >200 >200 >200 177 >200 62.4 >200 11.3 83.9

TI50

activity

6.5

I I I I I I I I I I I I M

such compounds can selectively and productively dock onto the zinc finger structure. To begin to broach this issue, we performed interatomic surface molecular modeling studies to map the reactive sites of the zinc finger domains and to dock NSC 20625 and NSC 4493 into those high-affinity ligand binding sites. For simplicity, we first focused our attention only on the N-terminal (or first) zinc finger domain of the p7NC protein. As illustrated in Figure 2, two possible ligand attachment sites for the compounds were found on finger one of the p7NC protein. The first attachment site involves what we will term the “proximal” side of the finger, where a relatively large crevice is formed by His23, Gly22, Ala30, and Pro31. Figure 2 displays R-carbon and atomic surface representations (panels A and B, respectively) of the first zinc finger motif, allowing for visualization of the proximal binding site. The second site involves the putative mRNA binding site as described by South and Summers21 on the “distal” side of the finger. This site consists of residues Ile12, Ile13, Phe16, Ile24, and Ala25, which form a narrow pocket for ligand binding. R-Carbon and atomic surface renderings of the distal binding site are shown in panels C and D of Figure 2, respectively. Attachment to either of these binding sites would position the ligand within close proximity to the zinccoordination locus. However, differences in the predicted strengths of ligand binding at these two sites were found for NSC 20625 versus NSC 4493. NSC 20625 is predicted to have comparable binding affinity to each site. In contrast, NSC 4493 is predicted to strongly favor binding at the putative mRNA binding site (distal site), with only a weak interaction at the site

on the proximal side of the finger. Figure 2 displays the orientation of NSC 20625 when bound to the proximal (panels A and B) and distal (panels C and D) sites. Binding of NSC 20625 at the mRNA binding site (distal) involves deep penetration of the terminal guanidine group into the base of the binding site (see Figure 2D). This bound configuration places the nitrogen of the guanidine group at 3.3 Å from the sulfur of Cys15. This group is also positioned within hydrogen-bonding distance to the carbonyl oxygen of His23 and the amide nitrogen of Ala25. A nonnegligible portion of the predicted binding strength of NSC 20625 to this site also involves a favorable interaction between the benzyl ring of the compound and the benzyl ring of Phe16 on the p7NC backbone. NSC 4493 also binds to the proximal site, but in an orientation nearly perpendicular to the long axis of NSC 20625 in its bound configuration at this site (not shown). The axis of the docked NSC 4493 is oriented quite similar to that of the mRNA strand, as modeled into this site by South and Summers.21 This configuration places a sulfur and carbonyl oxygen of NSC 4493 equidistant from the Cys15 sulfur (5.5 Å), with near superimposition of one benzyl ring with that on the docked NSC 20625 orientation. This arrangement suggests the importance of a hydrophobic interaction with Phe15, as well as the sulfur on Cys15. NSC 20625 is predicted to bind with equal strength at the proximal site opposite that of the distal mRNA binding site. The predicted binding position places the molecule nearly parallel with the amino acid segment 22-24 of the p7NC. Close examination of the “mode” of binding reveals that this somewhat unexpected orientation actually produces atomic interactions similar to those identified at the distal mRNA site. As in the former case, a terminal guanidine group of NSC 20625 is oriented in close proximity to a sulfur atom, now associated with Cys28, and one benzyl group of NSC 20625 is oriented in a near planar fashion with His23. These observations suggested that a minimal ligand must possess a strong hydrophobic and hydrophilic component for interaction with either of these predicted binding sites, and that these “minimal” interactions can be achieved from at least two directions, involving opposite sides of the first zinc binding motif of the p7NC molecule. Conclusions In previous studies we determined that the ability of DIBA (disulfide benzamides or 2,2′-dithiobisbenza-

Table 9. Inhibition of HIV-1 Replication by Nonsymmetrical Disulfides (R * R2)

R-SS-R

3612


Table 10

Rice et al.



Figure 1. Reactivities of thiuram disulfides with the p7NC protein zinc fingers. The thiuram disulfide compounds shown in Table 10A were evaluated for their ability to eject zinc from the purified HIV-1 p7NC protein in a 96-well TSQ metal binding assay, as described in the Experimental Section. Data presented represent the findings in a single representative experiment. Table 11. Inhibition of HIV-1 Replication by Disulfoxides, Disulfones, and Thiosulfones

mides) congeners to eject zinc from the p7NC protein in vitro correlated with the cellular anti-HIV activity.15 In an attempt to identify more effective zinc finger inhibitors having superior in vivo properties, a rational selection of specific chemotypes from the NCI chemical repository for various categories of disulfide-based compounds was initiated. With these compounds we

performed coupled cell-based screening for anti-HIV-1 activity and molecular target-based screening for in vitro zinc ejection from purified p7NC protein zinc fingers. Over 800 compounds from the chemical repository were selected, and findings with 142 nonproprietary compounds from this list were reported herein. Two “hits” (NSC 4493 and NSC 20625) were discovered that

3614


Rice et al.

Table 12. Evaluation of Active Compounds against HIV-1 in Proliferating and Nonproliferating Cells and against the p7NC Protein Zinc Fingersa XTT cytoprotection assay

Mo/Mφ

NSC

EC50(µM)

IC50(µM)

TI50

EC50

IC50

211 4492 4493 20625 20871 28727 38069 58950 71033 76302 83217 112801 121513 210281 255089 327174 342015 342031 405662 624187 633232 666140 666141 666724

23.1 0.73 12.6 1.2 79.0 12.9 31.2 14.1 0.47 127.4 4.9 76.5 28.4 123.5 34.6 37.6 9.4 247.3 117.7 96.2 46.9 3.4 21.6 6.2

>200 3.5 >200 38.4 210.5 83.9 >200 >200 4.0 >200 73.1 329 102.4 >200 >200 >200 129 >457 >200 >200 125 10.9 74.3 97.6

>8.7 5.0 >15.9 32 2.6 6.5 >6.4 >14.2 8.5 >1.6 14.9 4.3 3.6 >1.6 >5.8 >5.3 13.7 >1.8 >1.7 >2.1 2.7 3.2 3.4 15.7

>100 21 3.5 10 >100 >100 >100 >100 7.5 >100 >100 37.2 32 >100 42 >100 71 >100 57.5 (NA) (NA) >100 >100 80

>100 96 >100 >100 >100 >100 >100 97 >100 >100 100 >100 >100 >100 >100 >100 >100 >100 >100

>100 >100 >100

p7NCZnFn Trp37

TSQ

judgement

>1.7

51.5 80.9 31.3 74.9 37.7 2.1 79.3 52.1 80.5 82.5 81.8 9.5 76.1 62.5 71.8 77.8 91.2 26.8 6.0

10.6 28.5 1.6 25.6 10.6 1.2 10.9 26.0 0 32.8 24.4 0.5 15.7 22.3 6.7 8.0 21.1 2.6 9.8

>1.3

78.1 24.8 92.1

0.3 0.3 1.5

I M A A I M I I M M I M M M M M M M M M M M M M

TI50 4.7 >28.6 >10

>13.3

>2.7 3.1 >2.4 >1.4

a

Data from the Trp37 assay are expressed as the percent decrease in RFU after treatment of p7NC protein with 25 µM compound for 10 min. Data for the 96-well TSQ assay are expressed as the increase in RFU after treatment of p7NC protein with 10 µM compound for 60 min.

Figure 2. Orientation of NSC 20625 when docked into the proximal and distal binding sites of the amino-terminal zinc finger of the HIV-1 p7NC protein. Panels A (top left) and C (bottom left) display the R-carbon backbone of the p7NC zinc finger, and panels B (top right) and D (bottom right) display the molecular surface of the all-atom p7NC zinc finger. All panels display zinc and its coordinating atoms. Surface renderings suggest that NSC 20625 docks into pockets on the p7NC zinc finger surface. These docked positions suggest that NSC 20625 interactions can be achieved from two directions, lying on opposite faces of the amino-terminal finger of the p7NC molecule.

promote zinc ejection from the HIV-1 p7NC protein zinc fingers and that express anti-HIV-1 activity in cell cultures. Both NSC 20625 and NSC 4493 demonstrated con-

centration-dependent in vitro zinc ejection from purified p7NC protein, but neither compound was found to inhibit any of the other classical anti-HIV-1 targets tested.19 Molecular modeling studies identified two



high-affinity binding sites (distal and proximal) on the amino-terminal finger of the p7NC molecule. NSC 20625 efficiently docked into both ligand binding sites, while NSC 4493 only docked into the distal site. These multiple levels of simultaneous evaluation have rapidly identified prime compounds for more advanced mechanistic and in vivo evaluation and for optimization through medicinal chemistry efforts. Currently, the clinically available drugs include the nucleoside and nonnucleoside inhibitors of reverse transcriptase and the peptidomimetic protease inhibitors. Unfortunately, the mutational plasticity of HIV-1 reverse transcriptase and protease threaten to undermine the long-term effectiveness of existing therapies.22-25 We therefore propose that, in addition to the classical antiviral targets (reverse transcriptase, protease, integrase, and virus binding), the retroviral zinc finger domain should be added to the list of rational targets. Moreover, the current study sets forth a rational pathway for identification of HIV-1 zinc finger inhibitors and for the optimization of lead candidates through molecular modeling.

prepared at 20 µg/mL in 10 mM sodium phosphate buffer (pH 7.0) and treated with 25 µM of each compound in a total volume of 1.0 mL, and then after the indicated time intervals the samples were diluted 1/10 in 10 mM sodium phosphate buffer (pH 7.0) and the fluorescence intensity measured. The excitation and emission wavelengths utilized with the Shimadzu RF5000 spectrofluorimeter were 280 and 351 nm, respectively. The dilution step was performed to prevent the compounds from introducing any artifactual quenching effects. In an alternate high-throughput zinc finger assay, the zincselective fluorescent probe N-(6-methoxy-8-quinolyl-)-p-toluenesulfonamide (TSQ, Molecular Probes, Eugene, OR) originally described by Frederickson28 was utilized to measure zinc released from the p7NC protein in vitro. Briefly, 4 µM recombinant HIV-1 p7NC protein in 10 mM sodium phosphate buffer, pH 7.0, 10% glycerol, was treated with 10 µM of each compound (200 µL total volume in 96-well plates) in the presence of 25 µM TSQ, and the time-dependent increase in fluorescence was measured on a Labsystems Fluoroskan II (360 nm excitation filter and 460 nm emission filter) over a period of 90 min. The purified p7NC protein (0.5 mL at a concentration of 1 mg/mL) was centrifuged at 13000g in a Spin-X spin column (3000 MWCO, Costar Scientific Corp., Cambridge, MA) for 45 min at 25 °C prior to usage in the TSQ assay and then diluted back to 0.5 mL. This removes unbound zinc from the p7NC protein preparation, enhancing the sensitivity of detection of released zinc. For precise concentration-dependent inhibition analyses the TSQ assay was performed in the Shimadzu RF5000 spectrofluorimeter using 360 and 460 nm as the excitation and emission wavelengths, respectively. The assay utilized 0.5 µM p7NC protein and 25 µM TSQ in buffering conditions described above. The relative fluorescence units were determined at 5 s intervals over a 5 min period after addition of various concentrations of compounds to the p7NC protein. Molecular Modeling and Docking of Experimental Compounds. A previously developed method for docking rigid ligands to their targets was used in this study,29 and a brief summary of the salient features of this approach is provided. This docking scheme utilizes both a surface complementarity screen as well as an energetic criterion based on surface area burial. The goal of docking is to produce a complex between ligand and target that optimizes geometric and chemical complementarity. A two-step method is applied.29 First, a set of geometrically compatible docks are determined. These initial dockings are accomplished using a hierarchical search of geometrically compatible triplets of surface normals on each molecule. A pruned tree of possible bound configurations is built using successive consideration of larger and larger triplets. The best scoring configurations are then passed to the second stage of docking. In this stage, further docking improvements are made, based on optimization of contacting atomic surfaces within the binding interface. A preference-based free-energy screen30 is used in this second step to select the lowest energy member of the set. The preference-based free-energy model used in this second step is derived from observations of surface burial by atom pairs across the interface of known enzyme/ inhibitor complexes. When tested against 20 rigid enzyme/ inhibitor complexes, starting from arbitrarily separated positions for ligand and target molecules, this method produced an average all-atom rms (root mean square) deviation of 1.0 Å from the native position of each ligand in the crystallographic complex.29 The target molecule in this analysis is the HIV-1 p7 nucleocapsid protein structure (p7NC) determined by South and Summers21 using nuclear magnetic resonance spectroscopy (Protein Data Bank entry 1AAF31). The averaged coordinates of the first zinc finger in p7NC, as presented in the Brookhaven database entry, were used for docking. In vacuo molecular dynamics followed by standard minimization procedures were used to determine the initial conformational geometries of NSC 20625 and NSC 4493. These conformations were subsequently docked into the p7NC target. The results for the best docks are presented.

Experimental Section Materials. All experimental compounds were obtained from the Developmental Therapeutics Program (DTP), Division of Cancer Treatment, Diagnosis and Centers (DCTDC), National Cancer Institute (NCI). Compounds were selected using a substructure search in the NCI Drug Information System. The initial search was for Ar-S-S-Ar, where Ar refers to a five- or six-membered aromatic ring. This search yielded 176 nonproprietary compounds, and 92 with sufficient material available were chosen for testing (Tables 2-7 and 11). The search was expanded for any compound having the general structure of C-S-S-C, which yielded 637 new testable compounds, of which 50 were selected for testing (Tables 8, 9, 10A, and 10B). CEM-SS cells were obtained from the NIH AIDS Research and Reference Reagent Program (Bethesda, MD) and were maintained in RPMI 1640, 10% FCS, 20 µg/mL gentamicin, 200 mM L-glutamate. Anti-HIV-1 Assays. Anti-HIV screening with a human T-cell line was performed by the Antiviral Evaluation Branch of DTP, DCTDC, NCI, with CEM-SS cells and HIV-1RF (MOI ) 0.01) using the XTT cytoprotection assay as previously described.26 Effective antiviral concentrations providing 50% cytoprotection (EC50) and cellular growth inhibitory concentrations causing 50% cytotoxicity (IC50) were calculated. 3′-Azido2′,3′-dideoxythymidine (AZT) and dextran sulfate were utilized as positive control compounds for anti-HIV activity. Monocytes (>95% positive by nonspecific esterase staining) were cultured for 7 days in AIM V medium (GIBCO BRL, Rockville, MD) supplemented with 1% pooled human AB+ serum (Sigma Chemical Co., St. Louis, MO), after which the monocyte/macrophage (M/M) cultures were infected with the monocytotropic ADA viral strain (MOI ) 0.1-0.01) as previously described.27 After a 2 h virus adsorption, the spent virus supernatant was removed and fresh medium containing the indicated concentration of reagent was added. Cultures were fed every 3 days by 1/2 volume replacement with replenishment of test reagent. At 12 days of infection the cells were analyzed for cell viability by XTT dye reduction and microscopic evaluation, and the level of virus-associated p24 in cell-free supernatants was quantitated. Effective antiviral concentrations providing 50% reduction in p24 production (EC50) and causing 50% cytotoxicity (IC50) were calculated. The p24 assays were performed using p24 ELISA kits obtained through the common services of the AIDS Vaccine Program (NCI-FCRDC, Frederick, MD). Zinc Finger Assays. Fluorescence measurements of the Trp37 residue in the C-terminal zinc finger of the recombinant HIV-1 p7NC protein (kindly provided by L. O. Arthur of the AIDS Vaccine Program, NCI-FCRDC, Frederick, MD) were performed as previously described.15 The p7NC protein was

3616


Rice et al.

Acknowledgment. This research was supported by the National Cancer Institute, Contract N01-C0-56000. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products or organizations imply endorsement by the U.S. Government. We thank Dr. Louis E. Henderson and Dr. Larry O. Arthur for their generous sharing of reagents and for their many helpful conversations. We also thank Ms. Judy Duears for her assistance in preparation of the manuscript and Ms. Terry Williams for preparation of figures.

(15) Rice, W. G.; Supko, J. G.; Malspeis, L.; Buckheit, R. W., Jr.; Clanton, D.; Bu, M.; Graham, L.; Schaeffer, C. A.; Turpin, J. A.; Domagala, J.; Gogliotti, R.; Bader, J. P.; Halliday, S. M.; Coren, L.; Sowder, R. C., II; Arthur, L. O.; Henderson, L. E. Inhibitors of HIV Nucleocapsid Protein Zinc Fingers as Candidates for the Treatment of AIDS. Science 1995, 270, 1194-1197. (16) Rice, W. G.; Turpin, J. A.; Arthur, L. O.; Henderson, L. E. Highly Conserved Retroviral Zinc Fingers as Targets for HIV-1 Therapeutic Management. Int. Antiviral News 1995, 3, 87-89. (17) Rice, W. G.; Arthur, L. O.; Henderson, L. E.; Turpin, J. A. Discovery and Development of Disulfide-Substituted Benzamides as Candidate Anti-HIV Drugs. Int. Antiviral News 1996, 4, 3-6. (18) Rice, W. G.; Schaeffer, C. A.; Graham, L.; Bu, M.; McDougal, J.; Orloff, S. L.; Villinger, F.; Young, M.; Oroszlan, S.; Fesen, M. R.; Pommier, Y.; Mendeleyev, J.; Kun, E. The Site of Action of 3-Nitrosobenzamide on the Infectivity Process of Human Immunodeficiency Virus in Human Lymphocytes. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 9721-9724. (19) Turpin, J. A.; Rice, W. G. Unpublished results. (20) Turpin, J. A.; Terpening, S. J.; Schaeffer, C. A.; Yu, G.; Glover, C. J.; Felsted, R. L.; Sausville, E. A.; Rice, W. G. Inhibitors of HIV-1 Zinc Fingers Prevent Normal Processing of Gag Precursors and Result in the Release of Noninfectious Virus Particles. J. Virol. 1996, 70, 6180-6189. (21) South, T. L.; Summers, M. F. Zinc and Sequence-dependent Binding to Nucleic Acids by the N-terminal Zinc Finger of the HIV-1 Nucleocapsid Protein: NMR Structure of the Complex with the Psi-site Analog, dACGCC. Protein Sci. 1993, 2, 3-19. (22) Coffin, J. M. HIV Population Dynamics In Vivo: Implications for Genetic Variation, Pathogenesis and Therapy. Science 1995, 267, 483-489. (23) Mellors, J. W.; Larder, B. A.; Schinazi, R. F. Mutations in HIV-1 Reverse Transcriptase and Protease Associated with Drug Resistance. Int. Antiviral News 1995, 3, 8-13. (24) St. Clair, M. H.; Martin, J. L.; Tudor-Williams, G.; Bach, M. C.; Vavro, C. L.; King, D. M.; Kellam, P.; Demp, S. D.; Larder, B. A. Resistance to ddI and Sensitivity to AZT Induced by a Mutation in HIV-1 Reverse Transcriptase. Science 1991, 245, 1557-1559. (25) Mellors, J. W.; Dutschman, G. E.; Im, G. J.; Tramontano, E.; Winkler, S. R.; Cheng, Y. C. In Vitro Selection and Molecular Characterization of Human Immunodeficiency Virus-1 Resistant to Non-Nucleoside Inhibitors of Reverse Transcriptase. Mol. Pharmacol. 1992, 41, 446-451. (Erratum 1993, 42, 174.). (26) Weislow et al., New Soluble-Formazan Assay for HIV-1 Cytopathic Effects: Application to High Flux Screening of Synthetic and Natural Products for AIDS-Antiviral Activity. J. Natl. Cancer Inst. 1989, 81, 577, 1989. (27) Turpin, J. A.; Schaeffer, C. A.; Bu, M.; Graham, L.; Buckheit, R. W., Jr.; Clanton, D.; Rice, W. G. Human Immunodeficiency Virus Type 1 (HIV-1) Replication is Unaffected by Human Secretory Leukocyte Protease Inhibitor. Antiviral Res. 1996, 29, 269-277. (28) Frederickson, C. J.; Kasarskis, E. J.; Ringo, D.; Frederickson, R. E. A Quinoline Fluorescence Method for Visualizing and Assaying the Histochemically Reactive Zinc (Bouton Zinc) in the Brain. J. Neurosci. Meth. 1987, 20, 91-103. (29) Wallqvist, A.; Covell, D. G. Docking Enzyme-Inhibitor Complexes Using a Preference Based Free-Energy Surface. Proteins: Struct., Funct. Genet., in press. (30) Wallqvist, A.; Jernigan, R. L.; Covell, D. G. A Preference-Based Free-Energy Parametrization of Enzyme-Inhibitor Binding. Applications to HIV-1 Protease Inhibitor Design. Protein Sci. 1995, 4, 1881-1903. (31) Bernstein, F. C.; Koetzle, T. F.; Williams, G. J. B.; Meyer, E. F., Jr.; Brice, M. D.; Rogers, J. R.; Kennard, O.; Shimanouchi, T.; Tasumi, M. The Protein Data Bank: A Computer Based Archival File for Macromolecular Structures. J. Mol. Biol. 1977, 112, 535-542.

References (1) Henderson, L. E.; Copeland, T. D.; Sowder, R. C.; Smythers, G. W.; Oroszlan, S. Primary Structure of the Low Molecular Weight Nucleic Acid-Binding Proteins of Murine Leukemia Viruses. J. Biol. Chem. 1981, 256, 8400-8406. (2) South, T. L.; Blake, P. R.; Sowder, R. C., II; Arthur, L. O.; Henderson, L. E.; Summers, M. F. The Nucleocapsid ProteinIsolated From HIV-1 Particles Binds Zinc and Forms RetroviralType Zinc Fingers. Biochemistry 1990, 29, 7786-7789. (3) Summers, M. F.; Henderson, L. E.; Chance, M. R.; Bess, J. W., Jr.; South, T. L.; Blake, P. R.; Sagi, I.; Perez-Alvarado, G.; Sowder, R. C., II; Hare, D. R.; Arthur, L. O. Nucleocapsid Zinc Fingers Detected in Retroviruses: EXAFS Studies of Intact Viruses and the Solution-State Structure of the Nucleocapsid Protein From HIV-1. Protein Sci. 1992, 1, 563-574. (4) Covey, S. N. Amino Acid Sequence Homology in Gag Region of Reverse Transcribing Elements and the Coat Protein Gene of Cauliflower Mosaic Virus. Nucleic Acids Res. 1986, 14, 623633. (5) Green, L. M.; Berg, J. M. A Retroviral Cys-Xaa2-Cys-Xaa4-HisXaa4-Cys Peptide Binds Metal Ions: Spectroscopic Studies and a Proposed Three-Dimensional Structure. Proc. Natl. Acad. Sci. U.S.A. 1989, 86, 4047-4051. (6) Aldovini, A.; Young, R. A. Mutations of RNA and Protein Sequences Involved in Human Immunodeficiency Virus Type 1 Packaging Result in Production of Non Infectious Virus. J. Virol. 1990, 64, 1920-1926. (7) Dupraz, P.; Oertle, S.; Meric, C.; Damay, P.; Spahr, P.-F. Point Mutations in the Proximal Cys-His Box of Rous Sarcoma Virus Nucleocapsid Protein. J. Virol. 1990, 64, 4978-4987. (8) Gorelick, R. J.; Nigida, S. M., Jr.; Bess, J. W., Jr.; Arthur, L. O.; Henderson, L. E.; Rein, A. Noninfectious Human Immunodeficiency Virus Type 1 Mutants Deficient in Genomic RNA. J. Virol. 1990, 64, 3207-3211. (9) Gorelick, R. J.; Henderson, L. E.; Hanser, J. P.; Rein, A. Point Mutations of Moloney Murine Leukemia Virus That Fail to Package Viral RNA: Evidence for Specific RNA Recognition by a “Zinc Finger-Like” Protein Sequence. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 8420-8424. (10) Darlix, J.-L.; Vincent, A.; Gabus, C.; De Rocquigny, H.; Roques, B. Trans-Activation of the 5′ to 3′ Viral DNA Strand Transfer by Nucleocapsid Protein During Reverse Transcription of HIV-1 RNA. C. R. Acad. Sci. III 1993, 316, 763-771. (11) Rodriguez-Rodriguez, L.; Tsuchihashi, Z.; Fuentes, G. M.; Bambara, R. A.; Fay, P. J. Influence of Human Immunodeficiency Virus Nucleocapsid Protein on Synthesis and Strand Transfer by the Reverse Transcriptase In Vitro. J. Biol. Chem. 1995, 270, 15005-15011. (12) Tranchou, V.; Gabus, C.; Rogemond, V.; Darlix, J.-L. Formation of Stable and Functional HIV-1 Nucleoprotein Complexes In Vitro. J. Mol. Biol. 1995, 252, 563-571. (13) You, J. C.; McHenry, C. S. Human Immunodeficiency Virus Nucleocapsid Protein Accelerates Strand Transfer of the Terminally Redundant Sequences Involved in Reverse Transcription. J. Biol. Chem. 1994, 269, 31491-31495. (14) Rice, W. G.; Schaeffer, C. A.; Harten, B.; Villinger, F.; South, T. L.; Summers, M. F.; Henderson, L. E.; Bess, J. W., Jr.; Arthur, L. O.; McDougal, J. S.; Orloff, S. L.; Mendeleyev, J.; Kun, E. Inhibition of HIV-1 infectivity by zinc-ejecting aromatic C-nitroso compounds. Nature 1993, 361, 473-475.

JM960375O

Evaluation of Selected Chemotypes in Coupled Cellular and

Recommend Documents