Herpes simplex virus (HSV) encodes a DNA polymerase that is similar in several respects to the replicative mammalian DNA polymerase a. Recently, these and other DNA polymerases have been shown to share several regions of protein sequence similarity. Despite these similarities, antiviral drugs that mimic natural polymerase substrates specifically inhibit herpesvirus DNA polymerases. To study amino acids involved in substrate and drug recognition, we have characterized and mapped altered drug sensitivity markers of nine HSV pol mutants and sequenced the relevant portions of these mutants. The mutations were found to occur within four relatively small regions. One such region, which we designate region A, has sequence similarity only to DNA polymerases that are sensitive to certain antiviral drugs. The other three regions contain sequences that are similar among various DNA polymerases. The multiple mutations occurring within two of these regions make it likely that the regions interact directly with drugs and substrates. Our results lead us to favor a model in which protein folding allows interactions among the four regions to form the substrate and drug binding sites.DNA polymerases are central enzymes in DNA replication. In prokaryotic systems, studies of DNA polymerases have been greatly aided by model systems to which both genetic and biochemical methods can be applied. The herpes simplex virus (HSV) DNA polymerase, which is essential for HSV replication (1), provides such a model system for the study of eukaryotic DNA polymerases. This enzyme resembles the mammalian replicative DNA polymerase a (pol a) in several respects, including nuclear localization (2) and sensitivities to inhibitors such as aphidicolin (3-7) and dideoxythymidine triphosphate (4). Recently, HSV DNA polymerase (HSV pol), DNA polymerase I of yeast, a variety of animal and bacterial virus DNA polymerases, and pol a were shown to share six regions of striking sequence similarity (8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19). Of the viral polymerases sequenced to date, however, HSV pol appears most closely related to pol a (8). The six regions of sequence similarity have been designated regions I-VI by Wong et al. (8), with region I being most similar among the various polymerases and region VI being the least similar (Fig. 1). Such sequence similarities suggest conservation during evolution and important functional roles for these regions.Despite these similarities, HSV pol is more sensitive than pol a to a number of inhibitors and thus serves as the ultimate target for a number of selective antiviral drugs. These include the pyrophosphate (PP,) analogs phosphonoacetic acid (PAA) and phosphonoformic acid (20, 21), and the triphosphates of nucleoside analogs such as acyclovir (ACV), vidarabine, ganciclovir, and bromovinyldeoxyuridine (BVdU) (22-25). In addition, aphidicolin appears to inhibit HSV pol competitively with certain deoxynucleoside triphosphates (dNTPs) (5-7).The availability of viral pol mutants that exhibit altered se...

As part of our continuing search for potential anticancer candidates among 2-phenyl-4-quinolones and 2-phenyl-4-quinazolinones, two series of 6,7,2',3',4',5'-substituted 2-phenyl-4-quinazolinones and 6,2',3',4',5'-substituted 2,3-dihydro-2-phenyl-4-quinazolinones were synthesized and evaluated for cytotoxicity and as inhibitors of tubulin polymerization. In general, a good correlation was found between the two activities. Five of the 6-substituted heterocyclic 2-phenyl-4-quinozolinones (37-51) showed significant cytotoxicity against a panel of human tumor cell lines with EC(50) values in the low micromolar to nanomolar concentration ranges. Compound 38 was the most potent of these compounds, as well as the most potent inhibitor of tubulin polymerization in this series. The activity of 38 was in the same range as those of the antimitotic natural products, colchicine, podophyllotoxin, and combretastatin A-4. Substituted 2-phenyl-4-quinazolinones and 2, 3-dihydro-2-phenyl-4-quinazolinones also displayed highly selective cytotoxicity against the ovarian cancer 1A9 and P-gp resistant KB-VIN cell lines.

In a continuing study of curcumin analogues as potential drug candidates to treat prostate cancer at both androgen-dependent and androgen-refractory stages, we designed and synthesized over 40 new analogues classified into four series: monophenyl analogues (series A), heterocycle-containing analogues (series B), analogues bearing various substituents on the phenyl rings (series C), and analogues with various linkers (series D). These new compounds were tested for cytotoxicity against two human prostate cancer cell lines, androgen-dependent LNCaP and androgen-independent PC-3. Antiandrogenic activity was also evaluated in LNCaP cells and PC-3 cells transfected with wild-type androgen receptor. Ten compounds possessed potent cytotoxicity against both LNCaP and PC-3 cells, seven only against LNCaP, and one solely against PC-3. This study established an advanced structure-activity relationship (SAR), and these correlations will guide the further design of new curcumin analogues with better anti-prostate cancer activity.