SUMMARY We define here the activity and mechanisms of action of a small molecule lead compound for cancer targeting. We show that the compound, BMH-21, has wide and potent antitumorigenic activity across NCI60 cancer cell lines and represses tumor growth in vivo. BMH-21 binds GC-rich sequences, which is present at high frequency in ribosomal DNA genes, and potently and rapidly represses RNA polymerase I (Pol I) transcription. Strikingly, we find that BMH-21 causes proteasome-dependent destruction of RPA194, the large catalytic subunit protein of Pol I holocomplex, and this correlates with cancer cell killing. Our results show that Pol I activity is under proteasome-mediated control, which reveals an unexpected therapeutic opportunity.

The metabolite (–)-lomaiviticin A, which contains two diazotetrahydrobenzo[b]fluorene (diazofluorene) functional groups, inhibits the growth of cultured human cancer cells at nanomolar–picomolar concentrations; however, the mechanism responsible for the potent cytotoxicity of this natural product is not known. Here we report that (–)-lomaiviticin A nicks and cleaves plasmid DNA by an ROS- and iron-independent pathway and that the potent cytotoxicity of (–)-lomaiviticin A arises from induction of DNA double-strand breaks (dsbs). In a plasmid cleavage assay, the ratio of single-strand breaks (ssbs) to dsbs is 5.3±0.6:1. Labeling studies suggest this cleavage occurs via a radical pathway. The structurally related isolates (–)-lomaiviticin C and (–)-kinamycin C, which contain one diazofluorene, are demonstrated to be much less effective DNA cleavage agents, thereby providing an explanation for the enhanced cytotoxicity of (–)-lomaiviticin A compared to other members of this family.

Manipulation of the activity of the p53 tumor suppressor pathway has demonstrated potential benefit in preclinical mouse tumor models and has entered human clinical trials. We describe here an improved, extensive small-molecule chemical compound library screen for p53 pathway activation in a human cancer cell line devised to identify hits with potent antitumor activity. We uncover six novel small-molecule lead compounds, which activate p53 and repress the growth of human cancer cells. Two tested compounds suppress in vivo tumor growth in an orthotopic mouse model of human B-cell lymphoma. All compounds interact with DNA, and two activate p53 pathway in a DNA damage signaling-dependent manner. A further screen of a drug library of approved drugs for medicinal uses and analysis of gene-expression signatures of the novel compounds revealed similarities to known DNA intercalating and topoisomerase interfering agents and unexpected connectivities to known drugs without previously demonstrated anticancer activities. These included several neuroleptics, glycosides, antihistamines and adrenoreceptor antagonists. This unbiased screen pinpoints interference with the DNA topology as the predominant mean of pharmacological activation of the p53 pathway and identifies potential novel antitumor agents.

Comparative mutagenesis of γ- or X-ray-induced tandem DNA lesions G[8,5-Me]T and T[5-Me,8]G intrastrand cross-links was investigated in simian (COS-7) and human embryonic (293T) kidney cells. For G[8,5-Me]T in 293T cells, 5.8% of progeny contained targeted base substitutions, whereas 10.0% showed semitargeted single-base substitutions. Of the targeted mutations, the G → T mutation occurred with the highest frequency. The semitargeted mutations were detected up to two bases 5′ and three bases 3′ to the cross-link. The most prevalent semitargeted mutation was a C → T transition immediately 5′ to the G[8,5-Me]T cross-link. Frameshifts (4.6%) (mostly small deletions) and multiple-base substitutions (2.7%) also were detected. For the T[5-Me,8]G cross-link, a similar pattern of mutations was noted, but the mutational frequency was significantly higher than that of G[8,5-Me]T. Both targeted and semitargeted mutations occurred with a frequency of ∼16%, and both included a dominant G → T transversion. As in 293T cells, more than twice as many targeted mutations in COS cells occurred in T[5-Me,8]G (11.4%) as in G[8,5-Me]T (4.7%). Also, the level of semitargeted single-base substitutions 5′ to the lesion was increased and 3′ to the lesion decreased in T[5-Me,8]G relative to G[8,5-Me]T in COS cells. It appeared that the majority of the base substitutions at or near the cross-links resulted from incorporation of dAMP opposite the template base, in agreement with the so-called “A-rule”. To determine if human polymerase η (hpol η) might be involved in the mutagenic bypass, an in vitro bypass study of G[8,5-Me]T in the same sequence was carried out, which showed that hpol η can bypass the cross-link incorporating the correct dNMP opposite each cross-linked base. For G[8,5-Me]T, nucleotide incorporation by hpol η was significantly different from that by yeast pol η in that the latter was more error-prone opposite the cross-linked Gua. The incorporation of the correct nucleotide, dAMP, by hpol η opposite cross-linked T was 3−5-fold more efficient than that of a wrong nucleotide, whereas incorporation of dCMP opposite the cross-linked G was 10-fold more efficient than that with dTMP. Therefore, the nucleotide incorporation pattern by hpol η was not consistent with the observed cellular mutations. Nevertheless, at and near the lesion, hpol η was more error-prone compared to a control template. The in vitro data suggest that translesion synthesis by another Y-family DNA polymerase and/or flawed participation of an accessory protein is a more likely scenario in the mutagenesis of these lesions in mammalian cells. However, hpol η may play a role in correct bypass of the cross-links.

DNA intercalation is a major therapeutic modality for cancer therapeutic drugs. The therapeutic activity comes at a cost of normal tissue toxicity and genotoxicity. We have recently described a planar heterocyclic small molecule DNA intercalator, BMH-21, that binds ribosomal DNA and inhibits RNA polymerase I (Pol I) transcription. Despite DNA intercalation, BMH-21 does not cause phosphorylation of H2AX, a key biomarker activated in DNA damage stress. Here we assessed whether BMH-21 activity towards expression and localization of Pol I marker proteins depends on DNA damage signaling and repair pathways. We show that BMH-21 effects on the nucleolar stress response were independent of major DNA damage associated PI3-kinase pathways, ATM, ATR and DNA-PKcs. However, testing a series of BMH-21 derivatives with alterations in its N,N-dimethylaminocarboxamide arm showed that several derivatives had acquired the property to activate ATM- and DNA-PKcs -dependent damage sensing and repair pathways while their ability to cause nucleolar stress and affect cell viability was greatly reduced. The data show that BMH-21 is a chemically unique DNA intercalator that has high bioactivity towards Pol I inhibition without activation or dependence of DNA damage stress. The findings also show that interference with DNA and DNA metabolic processes can be exploited therapeutically without causing DNA damage.

Human XPA is an important DNA damage recognition protein in nucleotide excision repair (NER). We previously observed that XPA binds to DNA lesion as a homodimer (1). Herein we report that XPA recognized undamaged DNA doublestrand/ single-strand (ds-ssDNA) junctions containing ssDNA branches with binding affinity (K d = 49.1±5.1 nM) much higher than its ability to bind to DNA damage. The recognized DNA junction structures include Y-shape junction (with both 3′-and 5′-ssDNA branches), 3′-overhang junction (with a 3′-ssDNA branch), and 5′-overhang junction (with a 5′-ssDNA branch). Using gel filtration chromatography and gel mobility shift assays, we showed that the highly efficient binding appeared to be carried out by the XPA monomer and that the binding was largely independent of RPA. Furthermore, XPA efficiently bound to six-nucleotide mismatched DNA bubble substrates with or without DNA adducts including C8 guanine adducts of AF, AAF, and AP, and the T[6,4] T photo products. Using a set of defined DNA substrates with varying degrees of DNA bending, we also found that the XPC-HR23B complex recognized DNA bending, whereas neither XPA nor the XPA-RPA complex could bind to bent DNA. We propose that besides DNA damage recognition, XPA may also play a novel role in stabilizing, via its high affinity to ds-ssDNA junctions, the DNA strand opening surrounding the lesion for stable formation of pre-incision NER intermediates. Our results provide a plausible mechanistic interpretation for the indispensable requirement of XPA for both global genome and transcription-coupled repairs. Since ds-ssDNA junctions are common intermediates in many DNA metabolic pathways, the additional potential role of XPA in cellular processes is discussed.

Activation of the p53 pathway has been considered a therapeutic strategy to target cancers. We have previously identified several p53-activating small molecules in a cell-based screen. Two of the compounds activated p53 by causing DNA damage, but this modality was absent in the other four. We recently showed that one of these, BMH-21, inhibits RNA polymerase I (Pol I) transcription, causes the degradation of Pol I catalytic subunit RPA194, and has potent anticancer activity. We show here that three remaining compounds in this screen, BMH-9, BMH-22, and BMH-23, cause reorganization of nucleolar marker proteins consistent with segregation of the nucleolus, a hallmark of Pol I transcription stress. Further, the compounds destabilize RPA194 in a proteasome-dependent manner and inhibit nascent rRNA synthesis and expression of the 45S rRNA precursor. BMH-9-and BMH-22-mediated nucleolar stress was detected in ex vivo-cultured human prostate tissues indicating good tissue bioactivity. Testing of closely related analogues showed that their activities were chemically constrained. Viability screen for BMH-9, BMH-22, and BMH-23 in the NCI60 cancer cell lines showed potent anticancer activity across many tumor types. Finally, we show that the Pol I transcription stress by BMH-9, BMH-22, and BMH-23 is independent of p53 function. These results highlight the dominant impact of Pol I transcription stress on p53 pathway activation and bring forward chemically novel lead molecules for Pol I inhibition, and, potentially, cancer targeting. Mol Cancer Ther; 13(11); 2537-46. Ó2014 AACR.

Nucleotide excision repair (NER) plays an important role in maintaining the integrity of DNA by removing various types of bulky or distorting DNA adducts in both prokaryotic and eukaryotic cells. In Escherichia coli, the excision repair proteins UvrA, UvrB, and UvrC recognize and incise the bulky DNA damages induced by UV light and chemical carcinogens. In this process, when a putative lesion in DNA is identified initially by UvrA, a subsequent strand opening is carried out by UvrB that not only ensures that the distortion is indeed due to a damaged nucleotide but also recognizes the chemical structure of the modified nucleotides with varying efficiencies. UvrB also recruits UvrC that catalyzes both the 3'- and the 5'-incisions. Herein, we examined the interaction of UvrABC with a DNA substrate containing a single G[8,5-Me]T cross-link and compared it with T[6,4]T (the 6-4 pyrimidine-pyrimidone photoproduct) and the C8 guanine adduct of N-acetyl-2-aminofluorene (AAF). The intrastrand vicinal cross-link G[8,5-Me]T containing a covalent bond between the C8 position of guanine and the 5-methyl carbon of the 3'-thymine is formed by X-radiation, while T[6,4]T is a vicinal cross-link induced by the UV light. We also selected the AAF adduct for comparison because it represents a highly distorting monoadduct containing a covalent linkage at the C8 position of guanine. The dissociation constants (K(d)) for UvrA protein binding to DNA substrates containing the G[8,5-Me]T, T[6,4]T, and AAF adducts, as determined by gel mobility shift assays, were 3.1 +/- 1.3, 2.8 +/- 0.9, and 8.2 +/- 1.9, respectively. Although UvrA had a considerably higher affinity for G[8,5-Me]T than for the AAF adduct, the G[8,5-Me]T intrastrand cross-link was incised by UvrABC much less efficiently than the T[6,4]T intrastrand cross-link and the AAF adduct. Similar incision results also were obtained with the DNA substrates containing the adducts in a six-nucleotide bubble, indicating that the inefficient incision of G[8,5-Me]T cross-link by UvrABC was probably due to the lack of efficient recognition of the adduct by UvrB at the second step of DNA damage recognition in the E. coli NER. Indeed, as compared to T[6,4]T and AAF substrates, which clearly showed UvrB-DNA complex formation, very little UvrB complex was detectable with the G[8,5-Me]T substrate. Our result suggests that G[8,5-Me]T intrastrand cross-link is more resistant to excision repair in comparison with the T[6,4]T and AAF adducts and thus will likely persist longer in E. coli cells.