Number Of Stem Cell Divisions Research Articles

An improved epigenetic counter to track mitotic age in normal and precancerous tissues

The cumulative number of stem cell divisions in a tissue, known as mitotic age, is thought to be a major determinant of cancer-risk. Somatic mutational and DNA methylation (DNAm) clocks are promising tools to molecularly track mitotic age, yet their relationship is underexplored and their potential for cancer risk prediction in normal tissues remains to be demonstrated. Here we build and validate an improved pan-tissue DNAm counter of total mitotic age called stemTOC. We demonstrate that stemTOC’s mitotic age proxy increases with the tumor cell-of-origin fraction in each of 15 cancer-types, in precancerous lesions, and in normal tissues exposed to major cancer risk factors. Extensive benchmarking against 6 other mitotic counters shows that stemTOC compares favorably, specially in the preinvasive and normal-tissue contexts. By cross-correlating stemTOC to two clock-like somatic mutational signatures, we confirm the mitotic-like nature of only one of these. Our data points towards DNAm as a promising molecular substrate for detecting mitotic-age increases in normal tissues and precancerous lesions, and hence for developing cancer-risk prediction strategies.

Statistical Models for High-Risk Intestinal Metaplasia with DNA Methylation Profiling.

We consider the newly developed multinomial mixed-link models for a high-risk intestinal metaplasia (IM) study with DNA methylation data. Different from the traditional multinomial logistic models commonly used for categorical responses, the mixed-link models allow us to select the most appropriate link function for each category. We show that the selected multinomial mixed-link model (Model 1) using the total number of stem cell divisions (TNSC) based on DNA methylation data outperforms the traditional logistic models in terms of cross-entropy loss from ten-fold cross-validations with significant p-values 8.12×10-4 and 6.94×10-5. Based on our selected model, the significance of TNSC's effect in predicting the risk of IM is justified with a p-value less than 10-6. We also select the most appropriate mixed-link models (Models 2 and 3) when an additional covariate, the status of gastric atrophy, is available. When the status is negative, mild, or moderate, we recommend Model 2; otherwise, we prefer Model 3. Both Models 2 and 3 can predict the risk of IM significantly better than Model 1, which justifies that the status of gastric atrophy is informative in predicting the risk of IM.

Abstract 847: Association of methylation-based stem cell divisions with lifetime cancer risk and cancer progression

Abstract Cancer initiation arises from the accumulation of somatic mutations acquired during stem cell mitotic events mediated by carcinogenic risk factors. On an ecological level, Tomasetti and Vogelstein (2015) demonstrated that lifetime risk of solid tumor cancer types were strongly correlated with total number of stem cell divisions estimated from tissue-specific cell lines. Recently, several mitotic clocks (MC) have been developed to estimate stem cell division rates (SCDR) based on whole-genome methylation arrays. Here, we compare estimated SCDR for several tissue types with tissue-specific cancer incidence and lifetime risk estimates. Furthermore, we assess the link between SCDR with cancer progression and survival. Array-based methylation data derived from normal adjacent tissue in TCGA was used to estimate SCDR for 15 cancer types. A Pearson correlation test was performed to assess the association between SCDR and cancer-specific incidence and lifetime risk for each cancer type overall and stratified by self-reported race (e.g. White and Black). SCDR was estimated in tumors from TCGA for 4 common cancers and subtypes: breast (BRCA), prostate (PRAD), lung adenocarcinoma (LUAD), and lung squamous cell carcinoma (LUSC). A Spearman correlation test was used to identify potential trends between SCDR and pathological stage. Cox proportional hazard (PH) models adjusted for age at diagnosis and treatment status (e.g. surgery, radiation, etc.) were used to evaluate the association of time to death with SCDR terciles across different cancer stages. We observed a similar trend between SCDR and overall incidence (cor=0.61, p=0.01), aligning with earlier findings. Stratified-specific trends between cancer incidence and SCDR was consistent for both Black and White populations (cor=0.62 vs 0.66). Tumor analysis revealed a significant correlation between SCDR and pathological stage (e.g. stage 1, 2, and 3/4) for BRCA (p<0.001) and pathological stage (e.g. stage 2, 3 and 4) for PRAD (p<0.001), suggesting its potential as a marker of cancer progression. However, the correlation between LUAD, LUSC and pathological stage was insignificant. Based on Cox PH model, within stage III BRCA, SCDR significantly contributed to explaining variability in survival (p=0.04), emphasizing its potential prognostic value beyond cancer stage. Our study supports earlier findings applying methylation based SCDR estimates and shows that this association is consistent across racial groups. This implies that SCDR may act as a biological marker of a combination of various genetic and environmental risk factors underlying disease etiology. Further analyses will be extended to all cancers in TCGA. Lastly, we demonstrate that MC-based SCDR estimation has the potential to serve as a clinical tool to identify individuals at elevated risk of cancer initiation, track progression, and complement cancer stage for predicting survival. Citation Format: Yanning Wu, Cheryl Thompson, Fredrick Schumacher. Association of methylation-based stem cell divisions with lifetime cancer risk and cancer progression [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 847.

Role of DNA methylation-based mitotic ageing indices in oral cancer development and recurrence.

DNA methylation data can be used to derive mitotic indices from complex tissues. Here, we assessed if the DNA methylation-derived mitotic ageing indices are associated with oral squamous cell carcinoma (OSCC) development and recurrence-free survival (RFS). DNA methylation-based mitotic indices (MitoticAge, TNSC and hypoSC) were derived using algorithms "MitoticAge" and "epiTOC2" for the discovery [non-malignant (n = 22), premalignant (n = 22) and OSCC (n = 68) tissues] and validation datasets (GSE87053, GSE136704 and TCGA-HNSCC). Differences in mitotic indices between non-malignant, premalignant and OSCC tissues were assessed. Finally, the association between estimated mitotic indices and RFS was evaluated in OSCCs. In the discovery and validation datasets, increased mitotic ageing was observed in OSCC compared to non-malignant and premalignant oral tissues. HPV-positive HNSCCs had higher mitotic index TNSC. Mitotic age index hypoSC was associated with RFS in OSCC (p = 0.011, HR 2.61, 95% CI 1.24-5.48). DNA methylation-derived mitotic indices are associated with OSCC development and RFS. Thus, DNA methylation-derived mitotic indices may be a valuable research tool to reliably estimate the cumulative number of stem cell divisions in malignant and non-malignant oral tissues. Future research utilizing mitotic indices for predicting clinical outcomes in OSCC is warranted.

A mathematical model for cancer risk and accumulation of mutations caused by replication errors and external factors.

Replication errors influence mutations, and thus, lifetime cancer risk can be explained by the number of stem-cell divisions. Additionally, mutagens also affect cancer risk, for instance, high-dose radiation exposure increases lifetime cancer risk. However, the influence of low-dose radiation exposure is still unclear because this influence, if any, is very slight. We can assess the minimal influence of the mutagen by virtually comparing the states with and without mutagen using a mathematical model. Here, we constructed a mathematical model to assess the influence of replication errors and mutagens on cancer risk. In our model, replication errors occur with a certain probability during cell division. Mutagens cause mutations at a constant rate. Cell division is arrested when the number of cells reaches the capacity of the cell pool. When the number of cells decreases because of cell death or other reasons, cells resume division. It was assumed that the mutations of cancer driver genes occur stochastically with each mutation and that cancer occurs when the number of cancer driver gene mutations exceeds a certain threshold. We approximated the number of mutations caused by errors and mutagens. Then, we examined whether cancer registry data on cancer risk can be explained only through replication errors. Although the risk of leukemia was not fitted to the model, the risks of esophageal, liver, thyroid, pancreatic, colon, breast, and prostate cancers were explained only by replication errors. Even if the risk was explained by replication errors, the estimated parameters did not always agree with previously reported values. For example, the estimated number of cancer driver genes in lung cancer was larger than the previously reported values. This discrepancy can be partly resolved by assuming the influence of mutagen. First, the influence of mutagens was analyzed using various parameters. The model predicted that the influence of mutagens will appear earlier, when the turnover rate of the tissue is higher and fewer mutations of cancer driver genes were necessary for carcinogenesis. Next, the parameters of lung cancer were re-estimated assuming the influence of mutagens. The estimated parameters were closer to the previously reported values. than when considering only replication errors. Although it may be useful to explain cancer risk by replication errors, it would be biologically more plausible to consider mutagens in cancers in which the effects of mutagens are apparent.

Abstract B012: Advanced epigenetic aging in non-AIDS-defining cancer patients living with HIV

Abstract Background. With widespread use of antiretroviral therapy, people with HIV (PWH) are living longer and are now at risk for age-related co-morbidities, including non-AIDS defining cancers (NADCs). Given this epidemiological trend, it is important to understand how prolonged HIV infection may relate to molecular features of aging that can impact cancer risk or outcomes. Methyl groups naturally accumulate on human DNA over time, and methylation patterns can be translated into metrics of accelerated aging. We compared epigenetic measures of aging between PWH receiving treatment for an NADC and matched HIV-uninfected NADC patients at Moffitt Cancer Center in Tampa, Florida. Methods. We identified PWH diagnosed with a non-AIDS defining solid tumor from 2004 to 2021 were identified. To be included, patients needed to have an archived specimen amenable for DNA extraction (whole blood, buffy coat, or PBMC) collected within 1 year of cancer diagnosis. HIV-uninfected matches were identified in a similar manner and matched by tumor site, reported age (+/- 5 years), tumor sequence (primary vs. non-primary), and treatment status. Extracted DNA was assayed using the Illumina MethylationEPIC BeadChip, which incorporates data from 850,000 CpG sites across the genome. Data were translated into a biological age using the epiTOC2 clock, which estimates the number of total number of stem cell divisions, and GrimAge, associated with mortality. Methylation data were also deconvoluted to determine immune cell composition (CD4+ T-cells, CD8+ T-cells, B-cells, neutrophils, natural killer cells, monocytes, basophils, eosinophils). Epigenetic age and immune cell proportions were compared across HIV status using Wilcoxan rank sum testing. Age acceleration was calculated as the difference (residuals from a linear model) in the observed age for PWH and a NADC diagnosis, compared to the predicted age based on HIV-uninfected NADC patients, using both unadjusted and immune-cell adjusted linear models. Results. DNA methylation was assessed in 65 NADC PWH and HIV and 64 HIV-uninfected NADC patients. Epigenetic age assessed using the epiTOC2 and GrimAge clocks was significantly elevated in PWH (p<0.001 and p=0.02, respectively). When methylation data was used to de-convolute immune cell composition, the proportion of CD4+ T-cells and neutrophils was significantly (P<0.05) lower in PWH, while the proportions of memory B-cells and CD8+ T-cells were higher. Epigenetic age was accelerated in PWH compared to HIV-uninfected patients using both epiTOC2 and GrimAge clocks, with the difference in the GrimAge clock remaining significant after adjustment for immune cell composition (P-value<0.01). Conclusions. The results of this study demonstrated age acceleration in PWH with a NADC diagnosis compared to their HIV-uninfected counterparts. As cancer is now the leading cause of non-AIDS death in PWH, research into the prognostic implications of these differences is warranted. Citation Format: Anna E. Coghill, Brittney Dickey, Ryan Putney, Anders Berglund. Advanced epigenetic aging in non-AIDS-defining cancer patients living with HIV [abstract]. In: Proceedings of the AACR Special Conference: Aging and Cancer; 2022 Nov 17-20; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2022;83(2 Suppl_1):Abstract nr B012.

Quantifying substantial carcinogenesis of genetic and environmental factors from measurement error in the number of stem cell divisions

BackgroundThe relative contributions of genetic and environmental factors versus unavoidable stochastic risk factors to the variation in cancer risk among tissues have become a widely-discussed topic. Some claim that the stochastic effects of DNA replication are mainly responsible, others believe that cancer risk is heavily affected by environmental and hereditary factors. Some of these studies made evidence from the correlation analysis between the lifetime number of stem cell divisions within each tissue and tissue-specific lifetime cancer risk. However, they did not consider the measurement error in the estimated number of stem cell divisions, which is caused by the exposure to different levels of genetic and environmental factors. This will obscure the authentic contribution of environmental or inherited factors.MethodsIn this study, we proposed two distinct modeling strategies, which integrate the measurement error model with the prevailing model of carcinogenesis to quantitatively evaluate the contribution of hereditary and environmental factors to cancer development. Then, we applied the proposed strategies to cancer data from 423 registries in 68 different countries (global-wide), 125 registries across China (national-wide of China), and 139 counties in Shandong province (Shandong provincial, China), respectively.ResultsThe results suggest that the contribution of genetic and environmental factors is at least 92% to the variation in cancer risk among 17 tissues. Moreover, mutations occurring in progenitor cells and differentiated cells are less likely to be accumulated enough for cancer to occur, and the carcinogenesis is more likely to originate from stem cells. Except for medulloblastoma, the contribution of genetic and environmental factors to the risk of other 16 organ-specific cancers are all more than 60%.ConclusionsThis work provides additional evidence that genetic and environmental factors play leading roles in cancer development. Therefore, the identification of modifiable environmental and hereditary risk factors for each cancer is highly recommended, and primary prevention in early life-course should be the major focus of cancer prevention.

Limited accumulation of high-frequency somatic mutations in a 1700-year-old Osmanthus fragrans tree.

Lifespan varies greatly between and within species. Mutation accumulation is considered an important factor explaining this life-history trait. However, direct assessment of somatic mutations in long-lived species is still rare. In this study, we sequenced a 1700-year-old sweet olive tree and analysed the high-frequency somatic mutations accumulated in its six primary branches. We found the lowest per-year mutation accumulation rate in this oldest tree among those studied via the whole-genome sequencing approach. Investigation of mutation profiles suggests that this low rate of high-frequency mutation was unlikely to result from strong purifying selection. More intriguingly, on a per-branching scale, the high-frequency mutation accumulation rate was similar among the long-lived individuals such as oak, wild peach and sweet olive investigated here. We therefore suggest the possibility that the accumulation of high-frequency somatic mutations in very long-lived trees might have an upper boundary due to both the possible limited number of stem cell divisions and the early segregation of the stem cell lineage.

Dysregulation of DNA Methylation and Epigenetic Clocks in Prostate Cancer among Puerto Rican Men.

In 2021, approximately 248,530 new prostate cancer (PCa) cases are estimated in the United States. Hispanic/Latinos (H/L) are the second largest racial/ethnic group in the US. The objective of this study was to assess DNA methylation patterns between aggressive and indolent PCa along with ancestry proportions in 49 H/L men from Puerto Rico (PR). Prostate tumors were classified as aggressive (n = 17) and indolent (n = 32) based on the Gleason score. Genomic DNA samples were extracted by macro-dissection. DNA methylation patterns were assessed using the Illumina EPIC DNA methylation platform. We used ADMIXTURE to estimate global ancestry proportions. We identified 892 differentially methylated genes in prostate tumor tissues as compared with normal tissues. Based on an epigenetic clock model, we observed that the total number of stem cell divisions (TNSC) and stem cell division rate (SCDR) were significantly higher in tumor than adjacent normal tissues. Regarding PCa aggressiveness, 141 differentially methylated genes were identified. Ancestry proportions of PR men were estimated as African, European, and Indigenous American; these were 24.1%, 64.2%, and 11.7%, respectively. The identification of DNA methylation profiles associated with risk and aggressiveness of PCa in PR H/L men will shed light on potential mechanisms contributing to PCa disparities in PR population.

Folate Intake Alters Mutation Frequency and Profiles in a Tissue- and Dose-Specific Manner in MutaMouse Male Mice

While cancer is common, its incidence varies widely by tissue. These differences are attributable to variable risk factors, such as environmental exposure, genetic inheritance, and lifetime number of stem cell divisions in a tissue. Folate deficiency is generally associated with increased risk for colorectal cancer (CRC) and acute lymphocytic leukemia (ALL). Conversely, high folic acid (FA) intake has also been associated with higher CRC risk. Our objective was to compare the effect of folate intake on mutant frequency (MF) and types of mutations in the colon and bone marrow of mice. Five-week-old MutaMouse male mice were fed a deficient (0mg FA/kg), control (2mg FA/kg), or supplemented (8mg FA/kg) diet for 20 wk. Tissue MF was assessed using the lacZ mutant assay and comparisons made by 2-factor ANOVA. LacZ mutant plaques were sequenced using next-generation sequencing, and diet-specific mutation profiles within each tissue were compared by Fisher's exact test. In the colon, the MF was 1.5-fold and 1.3-fold higher in mice fed the supplemented diet compared with mice fed the control (P=0.001) and deficient (P=0.008) diets, respectively. This contrasted with the bone marrow MF in the same mice where the MF was 1.7-fold and 1.6-fold higher in mice fed the deficient diet compared with mice fed the control (P=0.02) and supplemented (P=0.03) diets, respectively. Mutation profiles and signatures (mutation context) were tissue-specific. Our data indicate that dietary folate intake affects mutagenesis in a tissue- and dose-specific manner in mice. Mutation profiles were generally tissue- but not dose-specific, suggesting that altered cellular folate status appears to interact with endogenous mutagenic mechanisms in each tissue to create a permissive context in which specific mutation types accumulate. These data illuminate potential mechanisms underpinning differences in observed associations between folate intake/status and cancer.

Synthetic lethality across normal tissues is strongly associated with cancer risk, onset, and tumor suppressor specificity.

Various characteristics of cancers exhibit tissue specificity, including lifetime cancer risk, onset age, and cancer driver genes. Previously, the large variation in cancer risk across human tissues was found to strongly correlate with the number of stem cell divisions and abnormal DNA methylation levels. Here, we study the role of synthetic lethality in cancer risk. Analyzing normal tissue transcriptomics data in the Genotype-Tissue Expression project, we quantify the extent of co-inactivation of cancer synthetic lethal (cSL) gene pairs and find that normal tissues with more down-regulated cSL gene pairs have lower and delayed cancer risk. Consistently, more cSL gene pairs become up-regulated in cells treated by carcinogens and throughout premalignant stages in vivo. We also show that the tissue specificity of numerous tumor suppressor genes is associated with the expression of their cSL partner genes across normal tissues. Overall, our findings support the possible role of synthetic lethality in tumorigenesis.

Determining cancer risk: the evolutionary multistage model or total stem cell divisions?

A recent hypothesis proposed that the total number of stem cell divisions in a tissue (TSCD model) determine its intrinsic cancer risk; however, a different model-the multistage model-has long been used to understand how cancer originates. Identifying the correct model has important implications for interpreting the frequency of cancers. Using worldwide cancer incidence data, we applied three tests to the TSCD model and an evolutionary multistage model of carcinogenesis (EMMC), a model in which cancer suppression is recognized as an evolving trait, with natural selection acting to suppress cancers causing a significant mean loss of Darwinian fitness. Each test supported the EMMC but contradicted the TSCD model. This outcome undermines results based on the TSCD model quantifying the relative importance of 'bad luck' (the random accumulation of somatic mutations) versus environmental and genetic factors in determining cancer incidence. Our testing supported the EMMC prediction that cancers of large rapidly dividing tissues predominate late in life. Another important prediction is that an indicator of recent oncogenic environmental change is an unusually high mean fitness loss due to cancer, rather than a high lifetime incidence. The evolutionary model also predicts that large and/or long-lived animals have evolved mechanisms of cancer suppression that may be of value in preventing or controlling human cancers.

High percentage of cancer stem cells in metastatic locations; a comment on a claim “variation in cancer risk among tissues can be explained by the number of stem cell divisions”

Metastasis is the major cause of cancer-related deaths. Cancer relapse and metastasis are associated with a part of cancer cells with stem cell properties. These cancer stem cells (CSCs) are resistant to treatments. In a recent survey, we observed that the population of cancer stem-like cells among metastatic tumor cells was significantly higher than that among the primary tumor cells. This high percentage can partly explain the reasons for chemoresistance and relapse in metastatic cancers. Analysis of the role of CSCs in metastasis has been mainly conceptual and speculative, and the reasons for a higher number of CSCs in the metastatic loci are questionable. Tomasetti and Vogelstein’s claim can partly answer the question. They postulated that the proliferation rate of normal stem cells in some tissue is greater than that of other tissues, and accordingly, the incidence of cancer in these tissues is high. In compliance with CSCs paradigm, resident normal stem cells of tissues are the most probable source of CSCs. After homing of metastatic cancer cells in a tissue with high rate of normal stem cell proliferation, there is a big opportunity for cancer cells to convert normal stem cells to cancer stem cells. This is the powerful effect of cancerous microenvironment on resident stem cells of tissue. Therefore, in metastatic cancers, the number of CSCs in primary tumor or in each metastatic location is relevant to the proliferation rate of resident normal stem cells of the location. This concept is a confirmation of Tomasetti and Vogelstein’s claim and can answer some fundamental questions about metastasis process.

A comparison of epigenetic mitotic-like clocks for cancer risk prediction

BackgroundDNA methylation changes that accrue in the stem cell pool of an adult tissue in line with the cumulative number of cell divisions may contribute to the observed variation in cancer risk among tissues and individuals. Thus, the construction of epigenetic “mitotic” clocks that can measure the lifetime number of stem cell divisions is of paramount interest.MethodsBuilding upon a dynamic model of DNA methylation gain in unmethylated CpG-rich regions, we here derive a novel mitotic clock (“epiTOC2”) that can directly estimate the cumulative number of stem cell divisions in a tissue. We compare epiTOC2 to a different mitotic model, based on hypomethylation at solo-WCGW sites (“HypoClock”), in terms of their ability to measure mitotic age of normal adult tissues and predict cancer risk.ResultsUsing epiTOC2, we estimate the intrinsic stem cell division rate for different normal tissue types, demonstrating excellent agreement (Pearson correlation = 0.92, R2 = 0.85, P = 3e−6) with those derived from experiment. In contrast, HypoClock’s estimates do not (Pearson correlation = 0.30, R2 = 0.09, P = 0.29). We validate these results in independent datasets profiling normal adult tissue types. While both epiTOC2 and HypoClock correctly predict an increased mitotic rate in cancer, epiTOC2 is more robust and significantly better at discriminating preneoplastic lesions characterized by chronic inflammation, a major driver of tissue turnover and cancer risk. Our data suggest that DNA methylation loss at solo-WCGWs is significant only when cells are under high replicative stress and that epiTOC2 is a better mitotic age and cancer risk prediction model for normal adult tissues.ConclusionsThese results have profound implications for our understanding of epigenetic clocks and for developing cancer risk prediction or early detection assays. We propose that measurement of DNAm at the 163 epiTOC2 CpGs in adult pre-neoplastic lesions, and potentially in serum cell-free DNA, could provide the basis for building feasible pre-diagnostic or cancer risk assays. epiTOC2 is freely available from https://doi.org/10.5281/zenodo.2632938

Abstract P2-05-02: Age-related genetic and epigenetic changes in estrogen-receptor positive breast cancer

Abstract Background: Age is the main risk factor for developing breast cancer (BC) in women. The increase in incidence with age is dominated by estrogen receptor positive (ER+) BC despite the >90% reduction of estrogen levels during menopause and later life. Recent models suggest the lifetime risk of BC is correlated with the number of stem cell divisions with random mutations arising during each division and driving cancer development. However, only two genes are mutated in more than 20% of BC while several recurrent copy number alterations (CNAs) occur in more than 50%. New models of age-related risk factors need to be developed that incorporate mutations, CNAs, epigenetics, and other biological factors. As data of this type is limited in normal breast tissue, factors correlated with age of diagnosis in ER+ BC were identified to understand which variables changed with age and might impact risk of developing BC. Methods: In silico analysis of public databases was used to identify factors in ER+ primary BC and normal adjacent tissue (NT) that are correlated with age of diagnosis including DNA mutations, CNAs, DNA methylation and gene/protein expression (TCGA n=599ER+,113NT; METABRIC n=1435ER+,144NT). Results: DNA mutations accumulated with age in ER+ BC with the median mutation count below 28 in primary tumors diagnosis in patients <50yr and rising to more than 43 in patients >80yr. However, the two most frequently mutated genes (PIK3CA and TP53) showed no significant correlation with age. As previously reported, GATA3 mutation rate was nearly twice as high in younger patients (<50yr=0.25,>80yr=0.12). There was little evidence of a general accumulation of CNAs with age, but there were higher rates of gain in 16p (<50yr=0.42,>80yr=0.51) and loss of 1p (<50yr=0.18,>80yr=0.37) in older patients and lower rates of loss in 6q (<50yr=0.34,>80yr=0.24). The most significant correlations with age related to ESR1 including expression of ESR1 mRNA (rho=0.39,P=1.2e-23 Spearman), ERα protein (rho=0.35,P=1.7e-15 Spearman), and demethylation of ESR1 promoter (rho=-0.36,P=1.6e-13 Spearman). The levels of the activated form of ERα (pS118) also correlated with age but to a much lesser extent than total ERα (rho=0.12,P=0.01 Spearman). Demethylation and expression of ESR1 were highly correlated (rho=-0.50,P<1e-16 Spearman). Analysis of data from NT revealed correlation of age of diagnosis, expression of ESR1 (rho=0.34,P=0.001 Spearman) and demethylation of the ESR1 promoter (rho=-0.22,P=0.03 Spearman) but not with ERα expression (rho=0.32,P=0.19 Spearman) although only 19 NT had protein data. Conclusions: Analysis of DNA mutations and CNAs fit previous theoretical models with both the result of stochastic processes. In these models, the low impact on fitness from individual mutations allows the accumulation over time while the high fitness costs of CNAs prevent accumulation and are likely acquired in a single catastrophic event. The ability of ER+ BC to progress in the presence of very low postmenopausal estrogen levels may be partly explained by demethylation leading to higher ESR1 expression and maintenance of ERα activation. In theory, changes in expression and demethylation can be targeted to reduce the age-related increase in risk for developing BC. Citation Format: Schuster EF, Dowsett M. Age-related genetic and epigenetic changes in estrogen-receptor positive breast cancer [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P2-05-02.

A 30-s exposure to ethanol 20% is cytotoxic to human keratinocytes: possible mechanistic link between alcohol-containing mouthwashes and oral cancer.

To provide mechanistic evidence for the epidemiological link between long-term use of alcohol-containing mouthwashes and oral cancer. Human epithelial keratinocytes were exposed for 30s to concentrations of ethanol commonly present in mouthwashes. After a recovery period, cell viability was assessed with the MTT assay. A marked cytotoxic effect was observed for ethanol concentrations of 20% and above. The cytotoxicity of ethanol may explain the epidemiological association between mouthwash use and oral cancer. Evidence suggests that the risk of developing cancer in a tissue is strongly determined by the number of stem cell divisions accumulated by the tissue during a person's lifetime; cell division is a major source of mutations and other cancer-promoting errors. Since cell death activates the division of stem cells, the possible cytotoxicity of ethanol on the cells lining the oral mucosa will promote the division of the stem cells located in deeper layers to produce new cells to regenerate the damaged epithelium. If we regularly use mouthwashes containing cytotoxic concentrations of ethanol, the stem cells of the oral cavity may need to divide more often than usual and our risk of developing oral cancer may increase. Many mouthwashes contain percentages of ethanol above 20%. Because ethanol is not crucial to prevent and reduce gingivitis and plaque, members of the dental team should consider the potential risk of oral cancer associated with frequent use of alcohol-containing mouthwashes when advising their patients.

Abstract IA13: Extrinsic control of the breast cancer spectrum

Abstract Breast cancer consists of multiple subtypes characterized by genetic, transcriptional, histologic, and clinical differences that confer distinct prognoses. While epidemiology associates distinct classes with certain groups of women, it is unclear what aspects of biology determine the spectrum of breast cancer observed in different populations. Ionizing radiation is a well-established human carcinogen, particularly of breast. Of particular concern is that girls who have been successfully treated with radiation for childhood cancers have a relative risk of subsequently developing breast cancer similar to that conferred by BRCA1 germline mutation. Moreover, recent studies specifically found that the invasive breast cancer arising in previously irradiated breast tissue is more likely to be triple-negative (TNBC, i.e., negative for hormone receptors and HER2 amplification) compared with age-matched sporadic invasive breast cancer (1). Cancer risk following radiation exposure is often considered exclusively through the prism of the cell-intrinsic effects of DNA damage and mutations that might initiate transformation in long-lived cells. For example, a model proposed by Vogelstein and Tomasetti attributes cancer risk to the frequency of tissue stem cells and the 3 factors that increase their mutation load, hereditary defects, environmental mutagens, and replication; the latter that occur stochastically have been dubbed the bad luck mutations (2). Although radiation is a mutagen, the cell-intrinsic paradigm cannot explain the high rate of TNBC in radiation-preceded breast cancer. Over the last decade, research has “zoomed out” to not only include intrinsic genetic and epigenetic alterations of cancer cells, but to also incorporate tumor microenvironment constituents as essential partners in cancer development (3). We postulated that the effect of ionizing radiation on tumor type and age dependence could act through local and systemic effects. To test how host biology shapes the course of carcinogenesis, we used a genetic-mammary chimera in which Trp53 null epithelium is transplanted to wild-type syngeneic hosts. The Trp53 null epithelium gives rise to an apparently normal ductal outgrowth with a high rate of malignant transformation that generates carcinomas over the course of 12-18 months. These murine mammary cancers are biologically and genomically diverse in a manner very like that of human breast cancer (4, 5). To assess the contribution of radiation effects on stromal-epithelial-immune interactions during carcinogenesis, we irradiated mice prior to Trp53 fragment transplantation (i.e., a radiation-genetic chimera). These experiments show that radiation exposure acts via the microenvironment to promote fast-growing, estrogen receptor (ER)-negative cancers (6, 7). Notably, the signature of Trp53 null tumors arising in irradiated mice clusters cancers obtained from irradiated people and sporadic TNBC (4). Our recent focus is the basis for the increased growth rate of tumors arising in irradiated mice. New studies show that aggressive, rapidly growing ER-negative tumors arising in irradiated mice are enriched in immunosuppressive myeloid cells and depleted of tumor-infiltrating cytotoxic T cells compared to matched cancers arising in nonirradiated mice. Treating irradiated mice with a nontoxic compound that reduces inflammation eliminates these effects of host irradiation but has no evident effect on nonirradiated mice, suggesting that low-grade chronic inflammation may be a target for preventing or delaying breast cancer in women post-irradiation. Our extensive transcriptomic and histologic characterization of Trp53 null mammary cancers show that the distribution of subtypes is modulated by host genetic background, prior radiation exposure, type of radiation, and age. These studies lead us to propose that the breast cancer spectrum results from dominant extrinsic processes that likely influence selection of sporadic genomic mutations.

PO-503 The cohesin stromal antigen 1 (SA-1) modulates colonic and colorectal cancer (CRC) stem cells: mechanism for racial disparities

IntroductionCRC disproportionately impacts African-Americans (incidence and mortality increased by ~25% and~50%, respectively). While mechanisms remain unclear, Vogelstein posited that the number of stem cell divisions determine CRC risk Science 2015. CRC stem cells may impact mortality via chemoresistance. LGR5, aldehyde dehydrogenase (ALDH1a3) and DCAMKL1 are markers of both intestinal and CRC stem cells. We have noted loss of SA-1 (a chromatin remodeler) occurred during colonic field carcinogenesis was markedly accentuated in Blacks (Cancer Prev Res 2016) via specific SNPs (Neoplasia 2018). SA-1 loss was also associated with poorer CRC prognosis. We hypothesised that SA-1 loss leads to stem cell induction and hence CRC disparities.Material and methodsRectal biopsies were obtained from endoscopically normal mucosa from ~200 patients undergoing screening colonoscopy with an IRB approved protocol. SA-1 was assessed by RT-PCR normalised to β-actin. We modulated SA-1 in human CRC cell line HT29 and tested efficacy of chemotherapy 5 fluorouracil (5-FU) and oxaliplatin via annexin V apoptosis assay.Results and discussionsAdenoma-harbouring subjects had ~50% increase in LGR5, ALD1a3 and DCAMKL1 (p<0.05) with concomitant suppression of SA-1. Causality was indicated by demonstrating that SiRNA SA-1 knockdowns (KD) in HT29 cells caused stem cell marker induction (LGR5=380%, ALD1a3=30% and DCAMKL1=85%, p<0.05). SA-1 overexpression resulted in reciprocal effects downregulation of all 3 stem cell markers. Functionally, SA-1 KD suppressed 5-FU and oxaliplatin induced apoptosis by 56% and 72% respectively versus scramble vector (p<0.0001). Underscoring racial disparities relevance was the observation that Blacks have a 31% greater SA1 loss vs. Whites (p<0.0004) which mirrored a 35% higher upregulation in LGR5 and ALDH1a3 (p<0.05). CRISPR editing of RKO cells to have the SNP rs34149860 (found only in Blacks) resulted in SA1 loss (41% loss, p<0.005) and concomitant ~45% LGR5 and ALDH1a3 upregulation (p<0.05).ConclusionThis novel finding that the proneoplastic effects of SA-1 loss may be transduced through intestinal/colonic stem cell (CRC incidence) and also augmenting CRC stem cells resulting in chemotherapy resistance (CRC mortality). Future studies may mitigate CRC disparities in Blacks through development of effective biomarkers and therapeutic.

Does risk of brain cancer increase with intracranial volume? A population-based case control study.

Glioma is the most common primary brain tumor and is believed to arise from glial stem cells. Despite large efforts, there are limited established risk factors. It has been suggested that tissue with more stem cell divisions may exhibit higher risk of cancer due to chance alone. Assuming a positive correlation between the number of stem cell divisions in an organ and size of the same organ, we hypothesized that variation in intracranial volume, as a proxy for brain size, may be linked to risk of high-grade glioma. Intracranial volume was calculated from pretreatment 3D T1-weighted MRI brain scans from 124 patients with high-grade glioma and 995 general population-based controls. Binomial logistic regression analyses were performed to ascertain the effect of intracranial volume and sex on the likelihood that participants had high-grade glioma. An increase in intracranial volume of 100 mL was associated with an odds ratio of high-grade glioma of 1.69 (95% CI: 1.44‒1.98; P < 0.001). After adjusting for intracranial volume, female sex emerged as a risk factor for high-grade glioma (odds ratio for male sex = 0.56, 95% CI: 0.33‒0.93; P = 0.026). Intracranial volume is strongly associated with risk of high-grade glioma. After correcting for intracranial volume, risk of high-grade glioma was higher in women. The development of glioma is correlated to brain size and may to a large extent be a stochastic event related to the number of cells at risk.

Cancer etiology: Variation in cancer risk among tissues is poorly explained by the number of gene mutations.

Recent evidence indicates that the risk of being diagnosed with cancer in a tissue is strongly correlated (0.80) with the number of stem cell divisions accumulated by the tissue. Since cell division can generate random mutations during DNA replication, this correlation has been used to propose that cancer is largely caused by the accumulation of unavoidable mutations in driver genes. However, no correlation between the number of gene mutations and cancer risk across tissues has been reported. Because many somatic mutations in cancers originate prior to tumor initiation and the number of cell divisions occurring during tumor growth is similar among tissues, I use whole genome sequencing information from 22086 cancer samples and incidence data from the largest cancer registry in each continent to study the relationship between the number of gene mutations and the risk of cancer across 33 tissue types. Results show a weak positive correlation (mean = 0.14) between these 2 parameters in each of the 5 cancer registries. The correlation became stronger (mean = 0.50) when gender-related cancers were excluded. Results also show that 1003 samples from 29 cancer types have zero mutations in genes. These data suggest that cancer etiology can be better explained by the accumulation of stem cell divisions than by the accumulation of gene mutations. Possible mechanisms by which the accumulation of cell divisions in stem cells increases the risk of cancer are discussed.