Therapy-induced APOBEC3A drives evolution of persistent cancer cells.

Acquired drug resistance to anticancer targeted therapies remains an unsolved clinical problem. Although many drivers of acquired drug resistance have been identified1-4, the underlying molecular mechanisms shaping tumour evolution during treatment are incompletely understood. Genomic profiling of patient tumours has implicated apolipoprotein B messenger RNA editing catalytic polypeptide-like (APOBEC) cytidine deaminases in tumour evolution; however, their role during therapy and the development of acquired drug resistance is undefined. Here we report that lung cancer targeted therapies commonly used in the clinic can induce cytidine deaminase APOBEC3A (A3A), leading to sustained mutagenesis in drug-tolerant cancer cells persisting during therapy. Therapy-induced A3A promotes the formation of double-strand DNA breaks, increasing genomic instability in drug-tolerant persisters. Deletion of A3A reduces APOBEC mutations and structural variations in persister cells and delays the development of drug resistance. APOBEC mutational signatures are enriched in tumours from patients with lung cancer who progressed after extended responses to targeted therapies. This study shows that induction of A3A in response to targeted therapies drives evolution of drug-tolerant persister cells, suggesting that suppression of A3A expression or activity may represent a potential therapeutic strategy in the prevention or delay of acquired resistance to lung cancer targeted therapy.

Human activation-induced deaminase lacks strong replicative strand bias or preference for cytosines in hairpin loops.

Activation-induced deaminase (AID) is a DNA-cytosine deaminase that mediates maturation of antibodies through somatic hypermutation and class-switch recombination. While it causes mutations in immunoglobulin heavy and light chain genes and strand breaks in the switch regions of the immunoglobulin heavy chain gene, it largely avoids causing such damage in the rest of the genome. To help understand targeting by human AID, we expressed it in repair-deficient Escherichia coli and mapped the created uracils in the genomic DNA using uracil pull-down and sequencing, UPD-seq. We found that both AID and the human APOBEC3A preferentially target tRNA genes and transcription start sites, but do not show preference for highly transcribed genes. Unlike A3A, AID did not show a strong replicative strand bias or a preference for hairpin loops. Overlapping uracilation peaks between these enzymes contained binding sites for a protein, FIS, that helps create topological domains in the E. coli genome. To confirm whether these findings were relevant to B cells, we examined mutations from lymphoma and leukemia genomes within AID-preferred sequences. These mutations also lacked replicative strand bias or a hairpin loop preference. We propose here a model for how AID avoids causing mutations in the single-stranded DNA found within replication forks.

Genome-wide mapping of regions preferentially targeted by the human DNA-cytosine deaminase APOBEC3A using uracil-DNA pulldown and sequencing.

Activation-induced deaminase (AID) and apolipoprotein B mRNA-editing enzyme catalytic subunit (APOBEC) enzymes convert cytosines to uracils, creating signature mutations that have been used to predict sites targeted by these enzymes. Mutation-based targeting maps are distorted by the error-prone or error-free repair of these uracils and by selection pressures. To directly map uracils created by AID/APOBEC enzymes, here we used uracil-DNA glycosylase and an alkoxyamine to covalently tag and sequence uracil-containing DNA fragments (UPD-Seq). We applied this technique to the genome of repair-defective, APOBEC3A-expressing bacterial cells and created a uracilation genome map, i.e. uracilome. The peak uracilated regions were in the 5'-ends of genes and operons mainly containing tRNA genes and a few protein-coding genes. We validated these findings through deep sequencing of pulldown regions and whole-genome sequencing of independent clones. The peaks were not correlated with high transcription rates or stable RNA:DNA hybrid formation. We defined the uracilation index (UI) as the frequency of occurrence of TT in UPD-Seq reads at different original TC dinucleotides. Genome-wide UI calculation confirmed that APOBEC3A modifies cytosines in the lagging-strand template during replication and in short hairpin loops. APOBEC3A's preference for tRNA genes was observed previously in yeast, and an analysis of human tumor sequences revealed that in tumors with a high percentage of APOBEC3 signature mutations, the frequency of tRNA gene mutations was much higher than in the rest of the genome. These results identify multiple causes underlying selection of cytosines by APOBEC3A for deamination, and demonstrate the utility of UPD-Seq.

Distinguishing preferences of human APOBEC3A and APOBEC3B for cytosines in hairpin loops, and reflection of these preferences in APOBEC-signature cancer genome mutations.

The APOBEC3 enzymes convert cytosines in single-stranded DNA to uracils to protect against viruses and retrotransposons but can contribute to mutations that diversify tumors. To understand the mechanism of mutagenesis, we map the uracils resulting from expression of APOBEC3B or its catalytic carboxy-terminal domain (CTD) in Escherichia coli. Like APOBEC3A, the uracilomes of A3B and A3B-CTD show a preference to deaminate cytosines near transcription start sites and the lagging-strand replication templates and in hairpin loops. Both biochemical activities of the enzymes and genomic uracil distribution show that A3A prefers 3 nt loops the best, while A3B prefers 4 nt loops. Reanalysis of hairpin loop mutations in human tumors finds intrinsic characteristics of both the enzymes, with a much stronger contribution from A3A. We apply Hairpin Signatures 1 and 2, which define A3A and A3B preferences respectively and are orthogonal to published methods, to evaluate their contribution to human tumor mutations.

Mesoscale DNA features impact APOBEC3A and APOBEC3B deaminase activity and shape tumor mutational landscapes.

Antiviral DNA cytosine deaminases APOBEC3A and APOBEC3B are major sources of mutations in cancer by catalyzing cytosine-to-uracil deamination. APOBEC3A preferentially targets single-stranded DNAs, with a noted affinity for DNA regions that adopt stem-loop secondary structures. However, the detailed substrate preferences of APOBEC3A and APOBEC3B have not been fully established, and the specific influence of the DNA sequence on APOBEC3A and APOBEC3B deaminase activity remains to be investigated. Here, we find that APOBEC3B also selectively targets DNA stem-loop structures, and they are distinct from those subjected to deamination by APOBEC3A. We develop Oligo-seq, an in vitro sequencing-based method to identify specific sequence contexts promoting APOBEC3A and APOBEC3B activity. Through this approach, we demonstrate that APOBEC3A and APOBEC3B deaminase activity is strongly regulated by specific sequences surrounding the targeted cytosine. Moreover, we identify the structural features of APOBEC3B and APOBEC3A responsible for their substrate preferences. Importantly, we determine that APOBEC3B-induced mutations in hairpin-forming sequences within tumor genomes differ from the DNA stem-loop sequences mutated by APOBEC3A. Together, our study provides evidence that APOBEC3A and APOBEC3B can generate distinct mutation landscapes in cancer genomes, driven by their unique substrate selectivity.

Distinguishing preferences of human APOBEC3A and APOBEC3B for cytosines in hairpin loops, and reflection of these preferences in APOBEC-signature cancer genome mutations.

The APOBEC3 family of enzymes convert cytosines in single-stranded DNA to uracils thereby causing mutations. These enzymes protect human cells against viruses and retrotransposons, but in many cancers they contribute to mutations that diversify the tumors and help them escape anticancer drug treatments. To understand the mechanism of mutagenesis by APOBEC3B, we expressed the complete enzyme or its catalytic carboxy-terminal domain (CTD) in repair-deficient Eschericia coli and mapped the resulting uracils using uracil pull-down and sequencing technology. The uracilomes of A3B-full and A3B-CTD showed peaks in many of the same regions where APOBEC3A also created uracilation peaks. Like A3A, the two A3B enzymes also preferred to deaminate cytosines near transcription start sites and in the lagging-strand template at replication forks. In contrast to an earlier report that A3B does not favor hairpin loops over linear DNA, we found that both A3B-full and A3B-CTD showed a strong preference for cytosines in hairpin loops. The major difference between A3A and A3B was that while the former enzyme prefers 3 nt loops the best, A3B prefers loops of 4 nt over those of other lengths. Furthermore, within 5 nt loops, A3A prefers cytosine to be in the penultimate position, while A3B prefers it to be at the 3' end of the loop. We confirmed these loop size and sequence preferences experimentally using purified A3A and A3B-CTD proteins. Reanalysis of hairpin loop mutations in human tumors using the size, sequence and position preferences of the two enzymes found that the tumors displayed mutations with intrinsic characteristics of both the enzymes with a stronger contribution from A3A.

Mesoscale DNA Features Impact APOBEC3A and APOBEC3B Deaminase Activity and Shape Tumor Mutational Landscapes.

Antiviral DNA cytosine deaminases APOBEC3A and APOBEC3B are major sources of mutations in cancer by catalyzing cytosine-to-uracil deamination. APOBEC3A preferentially targets singlestranded DNAs, with a noted affinity for DNA regions that adopt stem-loop secondary structures. However, the detailed substrate preferences of APOBEC3A and APOBEC3B have been fully established, and the specific influence of the DNA sequence on APOBEC3A APOBEC3B deaminase activity remains to be investigated. Here, we find that APOBEC3B selectively targets DNA stem-loop structures, and they are distinct from those subjected deamination by APOBEC3A. We develop Oligo-seq, a novel in vitro sequencing-based to identify specific sequence contexts promoting APOBEC3A and APOBEC3B activity. Through this approach, we demonstrate that APOBEC3A an APOBEC3B deaminase activity is strongly regulated by specific sequences surrounding the targeted cytosine. Moreover, we identify structural features of APOBEC3B and APOBEC3A responsible for their substrate preferences. Importantly, we determine that APOBEC3B-induced mutations in hairpin-forming sequences within tumor genomes differ from the DNA stem-loop sequences mutated by APOBEC3A. Together, our study provides evidence that APOBEC3A and APOBEC3B can generate mutation landscapes in cancer genomes, driven by their unique substrate selectivity.

An extended APOBEC3A mutation signature in cancer.

APOBEC mutagenesis, a major driver of cancer evolution, is known for targeting TpC sites in DNA. Recently, we showed that APOBEC3A (A3A) targets DNA hairpin loops. Here, we show that DNA secondary structure is in fact an orthogonal influence on A3A substrate optimality and, surprisingly, can override the TpC sequence preference. VpC (non-TpC) sites in optimal hairpins can outperform TpC sites as mutational hotspots. This expanded understanding of APOBEC mutagenesis illuminates the genomic Twin Paradox, a puzzling pattern of closely spaced mutation hotspots in cancer genomes, in which one is a canonical TpC site but the other is a VpC site, and double mutants are seen only in trans, suggesting a two-hit driver event. Our results clarify this paradox, revealing that both hotspots in these twins are optimal A3A substrates. Our findings reshape the notion of a mutation signature, highlighting the additive roles played by DNA sequence and DNA structure.

Spectrum of Mechanisms of Resistance to Crizotinib and Lorlatinib in ROS1 Fusion-Positive Lung Cancer.

PURPOSE: Current standard initial therapy for advanced, ROS proto-oncogene 1, receptor tyrosine kinase fusion (ROS1)-positive (ROS1+) non-small cell lung cancer (NSCLC) is crizotinib or entrectinib. Lorlatinib, a next-generation anaplastic lymphoma kinase/ROS1 inhibitor, recently demonstrated efficacy in ROS1+ NSCLC, including in crizotinib-pretreated patients. However, mechanisms of lorlatinib resistance in ROS1+ disease remain poorly understood. Here, we assessed mechanisms of resistance to crizotinib and lorlatinib. EXPERIMENTAL DESIGN: Biopsies from patients with ROS1 + NSCLC progressing on crizotinib or lorlatinib were profiled by genetic sequencing. RESULTS: From 55 patients, 47 post-crizotinib and 32 post-lorlatinib biopsies were assessed. Among 42 post-crizotinib and 28 post-lorlatinib biopsies analyzed at distinct timepoints, ROS1 mutations were identified in 38% and 46%, respectively. ROS1 G2032R was the most commonly occurring mutation in approximately one third of cases. Additional ROS1 mutations included D2033N (2.4%) and S1986F (2.4%) post-crizotinib and L2086F (3.6%), G2032R/L2086F (3.6%), G2032R/S1986F/L2086F (3.6%), and S1986F/L2000V (3.6%) post-lorlatinib. Structural modeling predicted ROS1L2086F causes steric interference to lorlatinib, crizotinib, and entrectinib, while it may accommodate cabozantinib. In Ba/F3 models, ROS1L2086F, ROS1G2032R/L2086F, and ROS1S1986F/G2032R/L2086F were refractory to lorlatinib but sensitive to cabozantinib. A patient with disease progression on crizotinib and lorlatinib and ROS1 L2086F received cabozantinib for nearly 11 months with disease control. Among lorlatinib-resistant biopsies, we also identified MET amplification (4%), KRAS G12C (4%), KRAS amplification (4%), NRAS mutation (4%), and MAP2K1 mutation (4%). CONCLUSIONS: ROS1 mutations mediate resistance to crizotinib and lorlatinib in more than one third of cases, underscoring the importance of developing next-generation ROS1 inhibitors with potency against these mutations, including G2032R and L2086F. Continued efforts are needed to elucidate ROS1-independent resistance mechanisms.

A novel class of chemicals that react with abasic sites in DNA and specifically kill B cell cancers.

Most B cell cancers overexpress the enzyme activation-induced deaminase at high levels and this enzyme converts cytosines in DNA to uracil. The constitutive expression of this enzyme in these cells greatly increases the uracil content of their genomes. We show here that these genomes also contain high levels of abasic sites presumably created during the repair of uracils through base-excision repair. We further show that three alkoxyamines with an alkyne functional group covalently link to abasic sites in DNA and kill immortalized cell lines created from B cell lymphomas, but not other cancers. They also do not kill normal B cells. Treatment of cancer cells with one of these chemicals causes strand breaks, and the sensitivity of the cells to this chemical depends on the ability of the cells to go through the S phase. However, other alkoxyamines that also link to abasic sites- but lack the alkyne functionality- do not kill cells from B cell lymphomas. This shows that the ability of alkoxyamines to covalently link to abasic sites is insufficient for their cytotoxicity and that the alkyne functionality may play a role in it. These chemicals violate the commonly accepted bioorthogonality of alkynes and are attractive prototypes for anti-B cell cancer agents.