Design, synthesis, and antifungal activity of novel cinnamon-pyrazole carboxamide derivatives.

Hit, Lead & Candidate Discovery To discover succinate dehydrogenase inhibitors with a novel structure, we introduced cinnamic acid structure to optimize the lead structure 1 and synthesized four series of cinnamon-pyrazole carboxamide derivatives. The bioassay data showed that compounds (E)-N-(1-[4-chlorophenyl]-4-cyano-1H-pyrazol-5-yl)-3-(2-fluorophenyl) acrylamide (5III-d) and (E)-3-(2-chlorophenyl)-N-(1-[4-chlorophenyl]-4-cyano-1H-pyrazol-5-yl) acrylamide (5III-f) showed the significant antifungal activity against three fungi. In addition, 5III-d and 5III-f exhibited the excellent inhibitory effect against succinate dehydrogenase (SDH) enzymes with IC50 values ranging from 19.4 to 28.7 µM. The study demonstrates that the chlorine substituent group is present on both the phenyl and pyrazole rings that have a very good effect on the antifungal effect, and the compounds 5III-d and 5III-f can act as potential SDH inhibitors (SDHI) and throw a sprat for a new generation of SDHI.

Microwave-assisted and continuous flow multistep synthesis of 4-(pyrazol-1-yl)carboxanilides.

A series of 4-(pyrazol-1-yl)carboxanilides active as inhibitors of canonical transient receptor potential channels were synthesized in an efficient three-step protocol using controlled microwave heating. The general synthetic strategy involves condensation of 4-nitrophenylhydrazine with appropriate 1,3-dicarbonyl building blocks, followed by reduction of the nitro group to the amine, which is then amidated with carboxylic acids. Compared to the conventional protocol a dramatic reduction in overall processing time from ~2 days to a few minutes was achieved, accompanied by significantly improved product yields. In addition, the first two steps in the synthetic pathway were also performed under continuous flow conditions providing similar isolated product yields. As an alternative to the three-step protocol, a novel two-step route to the desired 4-(pyrazol-1-yl)carboxanilides was devised involving condensation of 4-bromophenylhydrazine with appropriate 1,3-dicarbonyl building blocks, followed by Pd-catalyzed Buchwald-Hartwig amidation with carboxylic acid amides.

Carboxin and its major metabolites residues in peanuts: Levels, dietary intake and chronic intake risk assessment.

We developed an ultra-performance liquid chromatography-tandem mass spectral method to determine the fungicide carboxin and its metabolites, oxycarboxin and carboxin sulfoxide in peanut samples. The method was used to detect the concentration of the analytes in the samples from fields and markets. The total residue quantities in peanut kernels were used to evaluate the chronic dietary risk of total carboxin upon peanut consumption. The estimated dietary intake of carboxin from peanuts whose seeds had been treated with carboxin at the recommended dose was between 0.020% and 0.344% of acceptable daily intake and the risk was found to be negligible. The chronic dietary risk assessment from markets and commercial field samples for various groups of humans indicated that the group with the greatest degree of exposure was 45 to 75-year-old women who lived in rural areas. However, their acceptable daily intake percentage was 0.006%, meaning that their health risk was extremely small.

The role of genotypic heterogeneity in wild type virus populations on the selection of nonnucleoside reverse transcriptase inhibitor-resistant viruses.

Virus populations were selected in cell culture using two widely used protocols in order to evaluate the role of selection methodology on the genotype and phenotype of nonnucleoside reverse transcriptase inhibitor resistant viruses. Selection was performed by serial passage of virus in the presence of gradually increasing concentrations of antiviral compound or passage in the presence of a constant high concentration of compound. Using the CEM-SS cell line, the IIIB strain of HIV-1, and identical nonnucleoside reverse transcriptase inhibitors, resistant viruses were obtained and their phenotypic and genotypic properties were defined. Resistant virus populations containing the Y181C amino acid change in the reverse transcriptase were predominantly selected with each of the tested compounds. Several of the compounds selected secondary amino acid changes using both methods. A comparison of the resistant viruses selected in our laboratory using each of the two protocols with viruses reported by a second laboratory employing one of the two methods suggests that genotypic differences in the selected virus isolates may most likely result from the variation in the genetic composition of the respective wild type virus pools, rather than the specific selection methodology employed. These results imply that HIV may select a wide variety of amino acid changes to avoid the inhibitory effects of the nonnucleoside reverse transcriptase inhibitors and the selection of compounds for clinical use in combination with agents possessing non-overlapping resistance phenotypes will require evaluation of the agents against virus isolates possessing each of the mutations known to confer drug resistance.

Activity of various thiocarboxanilide derivatives against wild-type and several mutant human immunodeficiency virus type 1 strains.

A large variety of carboxanilide derivates in which the original oxathiin moiety present in the prototype compound UC84 was replaced by a non-cyclic lipophilic entity has been evaluated for their inhibitory effect against wild-type human immunodeficiency virus type 1 (HIV-1/IIIB) and several mutant viruses derived thereof (i.e. HIV-1/138-Lys, HIV-1/181-Cys, HIV-1/106-Ala and HIV-1/100-IIe). Isopropoxy was the most favorable substituent resulting in molecules that were markedly inhibitory to the wild-type (EC50 0.004-0.04 microgram/ml) as well as the mutant HIV-1 strains (EC50 0.06-0.75 microgram/ml). In this respect, they proved superior to several other HIV-1-specific non-nucleoside reverse transcriptase inhibitors (NNRTIs) that are currently the subject of clinical trials. One of the most potent HIV-1 inhibitors among the thiocarboxanilide derivatives, namely UC38, selected for a mutant virus strain in which Lys at position 101 and Gly at position 190 of the reverse transcriptase was replaced by Glu.

Liquid chromatographic analysis in mouse, dog and human plasma; stability, absorption, metabolism and pharmacokinetics of the anti-HIV agent 2-chloro-5-(2-methyl-5,6-dihydro-1,4-oxathiin-3-yl carboxamido) isopropylbenzoate (NSC 615985, UC84).

NSC 615985 (UC 84) has demonstrated anti-HIV activity in the NCI-AIDS antiviral screen and was under consideration as an anti-AIDS drug. The compound was subsequently shown to be a non-nucleoside reverse transcriptase inhibitor (NNRTI). An HPLC method was developed for the analysis of NSC 615985 in mouse, dog and human plasma; and was used to study its stability in plasma and blood as well as its absorption and metabolism in mice. The method involved precipitation of plasma protein with three volumes of methanol followed by HPLC analysis of the supernatant. The HPLC analysis was carried out on a reversed-phase Nova-Pak C18 column with a mobile phase of KH2PO4 (0.01 M; pH 4.8)-acetonitrile (52:48, v/v) at a flow rate of 1 ml min-1 and quantification with a UV detector set at 259 nm. The lower limit of quantitation was 0.05 microgram ml-1 in 1 ml of dog or human plasma or 0.1 microgram ml-1 in 0.5 ml of mouse plasma. NSC 615985 was more stable in dog and human plasma than in mouse plasma, and was less stable in blood than in plasma of the three species investigated. Following bolus intravenous (i.v.) administration at 10 mg kg-1 to male CDF1 mice, NSC 615985 elimination followed biexponential kinetics with half-lives of 1 and 7 min, and was extensively metabolized. NSC 615985 was very poorly absorbed following oral (PO) administration as a suspension in water or in 20% lipid emulsion (Liposyn II). Following bolus subcutaneous (s.c.) administration of [14C]NSC 615985 at 10 mg kg-1, relatively low concentrations of the parent compound were observed in three of 36 mice. One metabolite was tentatively identified in plasma of both the i.v.- and s.c.-treated animals as the sulfoxide of the parent compound. No parent compound was detected in the urine of NSC 615985 dosed mice. At least seven metabolites were present in urine; one metabolite (constituting 8-14% of urinary radioactivity) was tentatively identified as the carboxylic acid resulting from the hydrolysis of the isopropyl group from the parent compound. In summary, NSC 615985 was poorly absorbed following oral administration and extensively metabolized and eliminated following i.v. or s.c. administration. This unfavorable pharmacokinetic profile of NSC 615985 as well as its pattern of activity against NNRTI-resistant strains of HIV-1 precluded its progression to clinical trial; however, other members of the general chemical class are currently being evaluated by the NCI.

Designs and syntheses of oxathiin carboxanilide analogues and their antiviral activities.

Syntheses of new analogues of oxathiin carboxanilide (UC84) and their antiviral activities were described. The heterocyclic carboxylic acids including oxathiins (4), thiazines (9) and dithiins (13) in which the methyl was replaced either by lipophilic trifluoromethyl- or bulky phenyl-group were synthesized starting from beta-keto esters (5). Reaction of 4, 9 and 13 with thionyl chloride followed by treatment of the substituted aniline 22 gave the corresponding carboxanilides (24a-24f). The carboxanilides were subjected to Laweson's reagent the corresponding thiocarboxanilides (24g-24k). The antiviral activities of the synthesized compounds against human immunodeficiency virus type 1 (HIV-1), poliovirus type 1 (PV-1), coxsackie B virus type 3 (CoxB-3), vesicular stomatitis virus (VSV) and herpes simplex virus type 1 (HSV-1) were presented. The antiviral activity against HIV-1 of dithiin carboxanilide (24e) was similar with that of UC84 (24a). The corresponding thiocarboxanilides (24g-24k) showed higher inhibitory activity against HIV-1 than the carboxanilides (24a, 24b, 24d, 24e). The compounds in which ether the lipophilic trifluoromethyl substituents (24d, 24f, 24i, 24k) or bulky phenyl substituent is present in the heterocyclic compounds showed lower inhibitory activity than that of the methyl substituents is present in the compounds against the HIV-1. But the trifluoromethylated dithiin (24f) showed higher inhibitory activity against PV-1 and CoxB-3 virus than commercial antiviral agents, ribavirin (RV).

The rate of penetration of leaf tissues by the systemic fungicides benodanil (2-iodobenzanilide) and oxycarboxin (2,3-dihydro-6 methyl-5 phenylcarbamoyl-1,4-oxathin-4,4-dioxide).

Leaf washing after application of benodanil and oxycarboxin reduced the protectant effect more than the eradicant effect. Oxycarboxin penetrated more rapidly than did benodanil . Residues of benodanil (WP) on the leaf surfaces can be absorbed again later in the presence of water.

Chemical ionization and high resolution electron impact mass spectra of 5,6-dihydro-2-methyl-1,4-oxathiin-3-carboxanilide and three of its metabolites.

The electron impact and the hydrogen and methane chemical ionization mass spectra of 5,6-dihydro-2-methyl-1,4-oxathin-3-carboxanilide, the sulfoxide and sulfone derived therefrom, and 2-(2-hydroxyethylthio)-acetoacetanilide enol have been determined. All four compounds show abundant molecular ions in the electron impact spectra and abundant [MH]+ ions in the methane chemical ionization spectra (along with the expected [M + C2H5]+ and [M + C3H5]+ ions), but relatively low [MH]+ ion signals in the hydrogen chemical ionization spectra. From high resolution mass measurements and metastable transitions determined by metastable ion refocusing, electron impact fragmentation mechanisms have been established. For 5,6-dihydro-2-methyl-1,4-oxathiin-3-carboxanilide, 5,6-dihydro-2-methyl-1,4-oxathiin-3-carboxanilide-4-oxide and 5,6-dihydro-2-methyl-1,4-oxathiin-3-carboxanilide-4,4-dioxide the major fragmentation mode involves loss of the anilino radical from [M]+, followed by loss of C2H4. Fragmentation to form the aniline molecular ion increases in importance with increasing oxidation state of the sulfur. In the chemical ionization of these three compounds fragmentation of [MH]+ proceeds in a similar fashion by loss of neutral aniline and by formation of protonated aniline. The electron impact mass spectrum of 2-(2-hydroxyethylthio)acetoacetanilide is dominated by the molecular ion and the aniline molecular ion. However, in the chemical ionization mass spectra characteristic fragment ions involving loss of water and loss of aniline from [MH]+, as well as protonated aniline, are observed and serve to characterize the compound.

Chemical barriers to human immunodeficiency virus type 1 (HIV-1) infection: retrovirucidal activity of UC781, a thiocarboxanilide nonnucleoside inhibitor of HIV-1 reverse transcriptase.

UC781, a thiocarboxanilide nonnucleoside inhibitor of human immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT), inhibited RT DNA polymerase activity in vitro with marked potency. Significant inhibition was noted at a 1:1 molar ratio of UC871 to RT, characteristic of a tight-binding inhibitor. Infectivity of the HIV-1(IIIB) laboratory strain was eliminated in a concentration-dependent manner following short exposure of isolated virion particles to UC781. Neither nevirapine nor certain other carboxanilide nonnucleoside inhibitors were effective in this manner. Endogenous reverse transcription in UC781-treated virus particles was markedly reduced. Treatment of chronically HIV-1-infected H9 cells with UC781 did not alter virus production, but the infectivity of the virus produced by the cells during drug exposure was markedly reduced. Moreover, the infectivity of nascent virus produced by the UC781-treated H9 cells after removal of exogenous drug was dramatically attenuated. Similarly, pretreatment of peripheral blood lymphocytes isolated from HIV-infected patients abolished the infectivity of virus produced by these cells after removal of exogenous drug, as measured by coculture experiments with uninfected cord blood mononuclear cells, indicating the utility of UC781 against a variety of clinical HIV samples. Importantly, preincubation of uninfected MT2 cells with UC781 rendered these cells refractory to subsequent HIV infection in the absence of extracellular drug, an effect that persisted for several days following removal of exogenous drug. These unique properties of UC781 indicate that this nonnucleoside inhibitor may have considerable promise for use in retrovirucidal formulations to minimize the spread of HIV from infected to noninfected individuals.

Identification of novel thiocarboxanilide derivatives that suppress a variety of drug-resistant mutant human immunodeficiency virus type 1 strains at a potency similar to that for wild-type virus.

A large variety of carboxanilide and thiocarboxanilide derivatives in which the original oxathiin or aliphatic moieties present in the prototype compounds UC84 and UC38 were replaced by an (un) substituted furanyl, thienyl, phenyl, or pyrrole entity have been evaluated for activity against wild-type human immunodeficiency virus type 1 strain IIIB [HIV-1 (IIIB)] and a series of mutant virus strains derived thereof. The mutant viruses contained either the Leu-100-->Ile, Lys-103-->Asn, Val-106-->Ala, Glu-138-->Lys, Tyr-181-->Cys, or Tyr-188-->Leu mutation in their reverse transcriptase. Several 3-(2-methylfuranyl)- and 3-(2-methylthienyl)-thiocarboxanilide ester, (thio)ether, and oxime ether derivatives showed exquisitely potent antiviral activity against wild-type HIV-1 (50% effective concentration, 0.009 to 0.021 microM). The pentenylethers of the 2-methylfuranyl and 2-methylthienyl derivatives (i.e., 313, N-[4-chloro-3-(3-methyl-2-butenyloxy)phenyl]- 2-methyl-3-furancarbothioamide or UC-781, and 314, N-[4-chloro-3-(3-methyl-2-butenyloxy)phenyl] -2-methyl-3-thiophenecarbothioamide or UC-82) proved virtually equally inhibitory for wild-type and the Ile-100, Ala-106, and Lys-138 mutant virus strains (50% effective concentration, 0.015 to 0.021 microM). Their inhibitory effect against the Asn-103 and Cys-181 reverse transcriptase mutant virus strains was decreased only four- to sevenfold compared with wildtype virus. UC-781 and UC-82 should be considered potential candidate drugs for the treatment of HIV-1-infected individuals.

Segmental analysis of molecular surface electrostatic potentials: application to enzyme inhibition.

We have recently shown that the anti-HIV activities of reverse transcriptase inhibitors can be related quantitatively to properties of the electrostatic potentials on their molecular surfaces. We now introduce the technique of using only segments of the drug molecules in developing such expressions. If an improved correlation is obtained for a given family of compounds, it would suggest that the segment being used plays a key role in the interaction. We demonstrate the procedure for three groups of drugs, two acting on reverse transcriptase and one on HIV protease. Segmental analysis is found to be definitely beneficial in one case, less markedly so in another, and to have a negative effect in the third. The last result indicates that major portions of the molecular surfaces are involved in the interactions and that the entire molecules need to be considered, in contrast to the first two examples, in which certain segments appear to be of primary importance. This initial exploratory study shows that segmental analysis can provide insight into the nature of the process being investigated, as well as possibly enhancing the predictive capability.

Degradation of carboxin (Vitavax) and oxycarboxin (Plantvax) by Pseudomonas aeruginosa isolated from soil.

Pseudomonas aeruginosa capable of utilizing carboxin and oxycarboxin as sole sources of carbon and nitrogen was isolated from red sandy loam soil perfused with the solutions of these fungicides. The bacterium hydrolyzed oxycarboxin via the intermediate compound 2-(vinylsulphonyl) acetanilide liberating 2- (2-hydroxyethylsulphonyl) acetic acid and aminophenol, whereas carboxin was first oxidized to its sulphoxide and then to its sulphone before hydrolysis. Further hydrolysis of aminophenol by the organism resulted in the accumulation of ammonium which was partly oxidized to nitrite. Nitrite accumulated in the medium without further oxidation to nitrate.

Reaction site of carboxanilides and of thenoyltrifluoroacetone in complex II.

Oxathiin carboxanilides are systemic fungicides that inhibit the oxidation of succinate by interrupting electron transport between succinate dehydrogenase [succinate:(acceptor) oxidoreductase, EC 1.3.99.1] and coenzyme Q. Kinetic and electron paramagnetic resonance studies have established that the specific binding site of carboxanilides and of thenoyltrifluoroacetone responsible for the inhibition is the same. Although the binding of carboxanilides to membrane preparations of the dehydrogenase is very tight (Ki = 0.01-0.1 microM), it is noncovalent. Identification of the membrane component(s) to which specific binding occurs has therefore required the introduction of a photoaffinity label onto the carboxanilide molecule. By using [G-3H]3'-azido-5,6-dihydro-2-methyl-1,4-oxathiin-3-carboxanilide, it was found, in accord with earlier data with other carboxanilides, that unresolved complex II specifically binds about 0.6 mol of the inhibitor per mol of succinate dehydrogenase in equilibrium dialysis experiments. The resolved components of the complex, succinate dehydrogenase and the two binding peptides CII-3 and CII-4, failed to bind the inhibitor; however, when these were recombined with reconstitution of coenzyme Q reductase activity, the initial binding titer was restored. Azidocarboxanilide-inhibited complex II was irradiated to generate covalent linkages with the binding site, and the components of the complex were separated on polyacrylamide gel. Most of the specifically bound inhibitor was found in the low molecular weight binding peptides and phospholipids.

Oxathiin carboxanilide, a potent inhibitor of human immunodeficiency virus reproduction.

Oxathiin carboxanilide (OC), NSC 615985, a compound originally synthesized as a potential fungicide, was demonstrated to be highly active in preventing human immunodeficiency virus (HIV)-induced cell killing and in inhibiting HIV reproduction. Virus-infected CD4+ lymphocytes were completely protected by 0.5 microM OC, whereas no toxicity was observed at concentrations below 50 microM OC. Production of infectious virus, viral p24 antigen, and virion reverse transcriptase were reduced by OC at concentrations that prevented viral cell killing. A variety of CD4+ T-cell lines were protected by OC from HIV cytopathicity, and OC inhibited two distinct strains of HIV-1. However, HIV-2 infections were unaffected by OC. OC had no direct effect on virions of HIV or on the enzymatic activities of HIV reverse transcriptase or HIV protease. Time-limited treatments of cells with OC before, during, or after exposure of cells to virus failed to protect cells from the eventual cytopathic effects of HIV, and OC failed to inhibit the production of virus from cells in which infection was established or from chronically infected cells. We conclude that the highly active OC has a reversible effect on some early stage of HIV-1 reproduction and cytopathicity. Pilot in vivo experiments showed that circulating concentrations of OC exceeding 1 microM could be achieved and sustained in hamsters for at least a week with no remarkable toxicological sequelae. OC represents a new class of anti-HIV agents that are promising candidates for drug development.

Highly potent oxathiin carboxanilide derivatives with efficacy against nonnucleoside reverse transcriptase inhibitor-resistant human immunodeficiency virus isolates.

The structure-activity relationships of a series of compounds related to the nonnucleoside reverse transcriptase (RT) inhibitor (NNRTI) oxathiin carboxanilide have been described (R. W. Buckheit, Jr., T. L. Kinjerski, V. Fliakas-Boltz, J. D. Russell, T. L. Stup, L. A. Pallansch, W. G. Brouwer, D. C. Dao, W. A. Harrison, R. J. Schultz, J. P. Bader, and S. S. Yang, Antimicrob. Agents Chemother. 39:2718-2727, 1996). From these studies, the furanyl-containing analog UC10 was identified as the most potent inhibitor of human immunodeficiency virus type 1 (HIV-1) replication and a promising candidate for further development. Three new UC analogs (UC040, UC82, and UC781) have been determined to inhibit laboratory-derived and low-passage-number, primary virus isolates at low nanomolar concentrations in both established and fresh human cells. Each of the compounds synergistically interacted with the nucleoside analogs zidovudine, dideoxyinosine, dideoxycytosine, and lamivudine to inhibit HIV-1 replication. As a group, the UC compounds were found to be less active against viruses with the L100I, K103N, and Y181C amino acid changes in the RT and, upon in vitro selection, yielded resistant virus with the Y181C mutation in the RT. The most potent of the three new compounds, UC781, contains a furanyl side chain, similar to UC10, but differs in having an extended ether side chain instead of an oxime chain. The broad therapeutic index of UC781 (>62,000) resulted in effective inhibition of NNRTI-resistant virus isolates at high nanomolar concentrations. Furthermore, UC781 and the NNRTI costatolide were able to synergistically inhibit HIV-1 replication when used in combination, suggesting that UC781 may interact with the RT differently than the other UC analogs. The favorable anti-HIV properties of the UC compounds suggest they should be considered for further clinical development.

Crystal structures of HIV-1 reverse transcriptase in complex with carboxanilide derivatives.

The carboxanilides are nonnucleoside inhibitors (NNIs) of HIV-1 reverse transcriptase (RT), of potential clinical importance. The compounds differ in potency and in their retention of potency in the face of drug resistance mutations. Whereas UC-84, the prototype compound, only weakly inhibits many RTs bearing single point resistance mutations, inhibition by UC-781 is little affected. It has been proposed that UC-38 and UC-781 may form quaternary complexes with RT at a site other than the known binding pocket of other NNIs. X-ray crystal structures of four HIV-1 RT-carboxanilide complexes (UC-10, UC-38, UC-84, and UC-781) reported here reveal that all four inhibitors bind in the usual NNI site, forming binary 1:1 complexes with RT in the absence of substrates with the amide/thioamide bond in cis conformations. For all four complexes the anilide rings of the inhibitors overlap aromatic rings of many other NNIs bound to RT. In contrast, the second rings of UC-10, UC-84, and UC-781 do not bind in equivalent positions to those of other "two-ring" NNIs such as alpha-APA or HEPT derivatives. The binding modes most closely resemble that of the structurally dissimilar NNI, Cl-TIBO, with a common hydrogen bond between each carboxanilide NH- group and the main-chain carbonyl oxygen of Lys101. The binding modes differ slightly between the UC-10/UC-781 and UC-38/UC-84 pairs of compounds, apparently related to the shorter isopropylmethanoyl substituents of the anilide rings of UC-38/UC-84, which draws these rings closer to residues Tyr181 and Tyr188. This in turn explains the differences in the effect of mutated residues on the binding of these compounds.

Structure-activity and cross-resistance evaluations of a series of human immunodeficiency virus type-1-specific compounds related to oxathiin carboxanilide.

A series of compounds related to the nonnucleoside reverse transcriptase (RT) inhibitor (NNRTI) oxathiin carboxanilide (UC84) were evaluated for activity against the human immunodeficiency virus (HIV) to determine structural requirements for anti-HIV activity. Twenty-seven compounds representative of the more than 400 Uniroyal Chemical Company (UC) compounds were evaluated for structure-activity relationships. Several of the compounds evaluated were highly active, with 50% effective concentrations in the nanomolar range and therapeutic indices of > 1,000. Highly synergistic anti-HIV activity was observed for each compound when used in combination with 3'-azido-3'-deoxythymidine; additive to slightly synergistic interactions were observed with the compounds used in combination with dideoxycytidine. In combination with the NNRTI costatolide, only UC38 synergistically inhibited HIV type 1. Residues in the RT which, when mutated, impart resistance to the virus isolates selected in cell culture, against virus variants with site-directed mutations, and against RTs containing defined single amino acid changes. The mutations included changes in RT amino acids 100, 101, 103, 106, 108, and 181. The results with isolates selected in cell culture indicate that the carboxanilide compounds interact with the RT at two vulnerable sites, selecting UC-resistant virus isolates with the Y-to-C mutation at position 181 (Y181C) or the L100I substitution. A resistant virus isolate containing both Y181C combination with calanolide A, an NNRTI which retains activity against virus with the single Y181C mutation, UC10 rapidly selected a virus isolate with the K103N mutation. The merits of selecting potential candidate anti-HIV agents to be used in rational combination drugs design as part of an armamentarium of highly active anti-HIV compounds are discussed.