It is controversial whether dietary fiber protects against colorectal cancer because of conflicting results from human epidemiologic studies. However, these studies and mouse models of colorectal cancer have not controlled the composition of gut microbiota, which ferment fiber into short-chain fatty acids such as butyrate. Butyrate is noteworthy because it has energetic and epigenetic functions in colonocytes and tumorsuppressive properties in colorectal-cancer cell lines. We utilized gnotobiotic mouse models colonized with wild-type or mutant strains of a butyrate-producing bacterium to demonstrate that fiber does have a potent tumor-suppressive effect but in a microbiota- and butyrate-dependent manner. Furthermore, due to the Warburg effect, butyrate was metabolized less in tumors where it accumulated and functioned as an HDAC inhibitor to stimulate histone acetylation and affect apoptosis and cell proliferation. To support the relevance of this mechanism in human cancer, we demonstrate that butyrate and histone-acetylation levels are elevated in colorectal adenocarcinomas compared to normal colonic tissues.

SUMMARY Histones and their post-translational modifications influence the regulation of many DNA-dependent processes. Although an essential role for histone-modifying enzymes in these processes is well established, defining the specific contribution of individual histone residues remains a challenge because many histone-modifying enzymes have non-histone targets. This challenge is exacerbated by the paucity of suitable approaches to genetically engineer histone genes in metazoans. Here, we describe a facile platform in Drosophila for generating and analyzing any desired histone genotype, and we use it to test the in vivo function of three histone residues. We demonstrate that H4K20 is neither essential for DNA replication nor for completion of development, unlike conclusions drawn from analyses of H4K20 methyltransferases. We also show that H3K36 is required for viability and H3K27 is essential for maintenance of cellular identity during development. These findings highlight the power of engineering histones to interrogate genome structure and function in animals.

Concentrating factors in nuclear bodies is thought to promote efficient gene expression. Tatomer et al. show that the histone locus body (HLB) concentrates pre-mRNA processing factors at replication-dependent histone genes, resulting in optimal 3′ end formation of histone mRNAs coupled with transcription termination.

The scaffolding protein Symplekin is part of multiple complexes involved in generating and modifying the 3' end of mRNAs, including cleavage-polyadenylation, histone pre-mRNA processing and cytoplasmic polyadenylation. To study these functions in vivo, we examined the localization of Symplekin during development and generated mutations of the Drosophila Symplekin gene. Mutations in Symplekin that reduce Symplekin protein levels alter the efficiency of both poly A(+) and histone mRNA 3' end formation resulting in lethality or sterility. Histone mRNA synthesis takes place at the histone locus body (HLB) and requires a complex composed of Symplekin and several polyadenylation factors that associates with the U7 snRNP. Symplekin is present in the HLB in the early embryo when Cyclin E/Cdk2 is active and histone genes are expressed and is absent from the HLB in cells that have exited the cell cycle. During oogenesis, Symplekin is preferentially localized to HLBs during S-phase in endoreduplicating follicle cells when histone mRNA is synthesized. After the completion of endoreplication, Symplekin accumulates in the cytoplasm, in addition to the nucleoplasm, and localizes to tricellular junctions of the follicle cell epithelium. This localization depends on the RNA binding protein ypsilon schachtel. CPSF-73 and a number of mRNAs are localized at this same site, suggesting that Symplekin participates in cytoplasmic polyadenylation at tricellular junctions.

No abstract

It is controversial whether dietary fiber protects against colorectal cancer because of conflicting results from human epidemiologic studies. These studies have been complicated by the participants’ genetic heterogeneity and differences in the composition of microbiota within their gastrointestinal tracts. To eliminate these confounding variables, we utilized a gnotobiotic mouse model of colorectal cancer. Our experiments were designed to investigate the function of butyrate because it is a short-chain fatty acid produced by bacterial fermentation of fiber in the colon at high (mM) levels and has potent energetic and epigenetic properties in host colonocytes. Here, we report that fiber did, in fact, have a chemoprotective effect but in a microbiota- and butyrate-dependent manner. The incidence, number, size, and histopathologic progression of AOM/DSS-induced colorectal tumors were significantly diminished when BALB/c mice were provided a high-fiber diet only if they were colonized with defined microbiota that included a butyrate-producing bacteria. This chemoprotective effect was attenuated when mice were colonized with the same microbiota except that the wild-type butyrate producer was replaced by a mutant strain with a 0.8-kb deletion in the butyryl-CoA synthesis operon. To confirm that butyrate is a causal factor, the chemoprotective effect was recapitulated in mice without any butyrate-producing bacteria if they were provided a butyrate-fortified diet. Our data support a general mechanism that includes microbial fermentation of fiber rather than fiber exclusively speeding colonic transit to minimize the exposure of colonocytes to ingested carcinogens. Our data also support a molecular mechanism that is metaboloepigenetic. Normal colonocytes utilize butyrate as their preferred energy source, whereas cancerous colonocytes rely on glucose because of the Warburg effect. Due to this metabolic difference, butyrate accumulated in tumors (as measured by LC-MS) and functioned as an HDAC inhibitor to increase histone acetylation levels and apoptosis. To support the applicability of this model to human cancer, we demonstrate that butyrate also accumulates at higher levels in human colorectal tumors than in normal colonic tissue, and this is associated with higher levels of histone acetylation in tumors. These results link diet and microbiota to a common metabolite that influences epigenetics and cancer predisposition. Citation Format: Dallas Donohoe, Darcy Holley, Leonard Collins, Stephanie Montgomery, Alan Whitmore, Andrew Hillhouse, Kaitlin Curry, Sarah Renner, Alicia Greenwalt, Elizabeth Ryan, Virginia Godfrey, Mark Heise, Deborah Threadgill, James Swenberg, David Threadgill, Scott Bultman. Dietary fiber protects against colorectal tumorigenesis in a microbiota- and butyrate-dependent manner. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr SY04-02. doi:10.1158/1538-7445.AM2014-SY04-02

<p>Supplementary Figure 1. Overall experimental design including the rationale behind specific details. Supplementary Figure 2. Confirmation of the ASF strains and B. fibrisolvens by PCR. Supplementary Figure 3. Formulations of the 3 diets provided to gnotobiotic mice in these experiments. Supplementary Figure 4. A combination of a high-fiber diet and a butyrate-producing bacterium (Butyrivibrio fibrisolvens) yields higher levels of butyrate in the lumen of the colon than either factor in isolation. Supplementary Figure 5. Detection of b-catenin mutations in tumor cells. Supplementary Figure 6. Analysis of b catenin nuclear localization in control and experimental tumors. Supplementary Figure 7. Characterization of squamous cell tumors. Supplementary Figure 8. Analysis of wild-type and mutant B. fibrisolvens strains. Supplementary Figure 9. Expression of Warburg effect markers. Supplementary Figure 10. Butyrate oxidation and uptake in colorectal tumors and normal colonic tissues. Supplementary Figure 11. Representative IHC images showing cleaved caspase 3 positive cells (arrows) in normal colonic tissue and tumors from mice in control and experimental conditions. Supplementary Figure 12. Flow cytometric analysis of immune cell populations from BALB/c mice colonized with the 4 ASF and B. fibrisolvens in gnotobiotic isolators. Supplementary Figure 13. Serum and colonic cytokine levels (pg/mL) from BALB/c mice colonized with the 4 ASF and B. fibrisolvens in gnotobiotic isolators based on Luminex assays. Supplementary Figure 14. Minimal inflammation based on histologic assessment of H&E-stained sections of colorectal tumor sections from 7-10 mice of each treatment group 5 months after exposure to the 5 AOM/3 DSS regimen. Supplementary Figure 15. Control experiment demonstrating that inflammation plays a minor role in colorectal tumorigenesis in the gnotobiotic model where BALB/c mice were polyassociated with 4 ASF plus B. fibrisolvens and then treated with AOM/DSS. Supplementary Figure 16. Relative Il18 mRNA levels normalized to Gapdh levels in normal colonic tissue and tumors based on RT-qPCR experiments. Supplementary Table 1. Luminal Concentrations (μM) of Other Two Major SCFAs (Propionate and Acetate). Supplementary Table 2. Clinical Information Regarding Human Adenocarcinomas. Supplementary Table 3. Statistical tests used for each figure panel.</p>

No abstract