The interaction between the metabolic activities of the intestinal microbiome and its host forms an important part of health. The basis of this interaction is in part mediated by the release of microbially-derived metabolites that enter the circulation. These products of microbial metabolism thereby interface with the immune, metabolic, or nervous systems of the host to influence physiology. Here, we review the interactions between the metabolic activities of the microbiome and the systemic metabolism of the host. The concept that the endocrine system includes more than just the eukaryotic host component enables the rational design of exogenous interventions that shape human metabolism. An improved mechanistic understanding of the metabolic microbiome-host interaction may therefore pioneer actionable microbiota-based diagnostics or therapeutics that allow the control of host systemic metabolism via the microbiome.

The microbiome has recently joined the club of endocrine entities of the human body that are involved in homeostasis and disease. Microbiome characterizations are now typically included in longitudinal and cross-sectional population studies, associations with microbiome features have been made for almost any human disease, and the molecules by which the microbiome functionally contributes to host physiology are being elucidated. The leverage of these efforts for human health, however, is still rather modest. In this Perspective, we summarize some of the challenges that need to be overcome in order to make microbiome studies as informative for human health as genetic studies. Focusing on the role of the microbiome in host metabolism and inflammation, we also outline potential strategies that can be employed to achieve the next milestones in the journey toward microbiome-informed human health assessment and action.

Background Individual colorectal polyp risk factors are well characterized; however, insights into their pathway-specific interactions are scarce. We aimed to identify the impact of individual risk factors and their joint effects on adenomatous (AP) and serrated polyp (SP) risk. Methods We collected information on 363 lifestyle and metabolic parameters from 1597 colonoscopy participants, resulting in over 521,000 data points. We used multivariate statistics and machine-learning approaches to assess associations of single variables and their interactions with AP and SP risk. Results Individual factors and their interactions showed common and polyp subtype-specific effects. Abdominal obesity, high body mass index (BMI), metabolic syndrome, and red meat consumption globally increased polyp risk. Age, gender, and western diet associated with AP risk, while smoking was associated with SP risk. CRC family history was associated with advanced adenomas and diabetes with sessile serrated lesions. Regarding lifestyle factor interactions, no lifestyle or dietary adjustments mitigated the adverse smoking effect on SP risk, whereas its negative effect was exacerbated by alcohol in the conventional pathway. The adverse effect of red meat on SP risk was not ameliorated by any factor, but was further exacerbated by western diet along the conventional pathway. No modification of any factor reduced the negative impact of metabolic syndrome on AP risk, whereas increased fatless fish or meat substitutes’ intake mitigated its effect on SP risk. Conclusions Individual risk factors and their interactions for polyp formation along the adenomatous and serrated pathways are strongly heterogeneous. Our findings may facilitate tailored lifestyle recommendations and contribute to a better understanding of how risk factor combinations impact colorectal carcinogenesis.

Imipridone ONC201 is a first-in-class dopamine receptor D2 (DRD2) antagonist and mitochondrial protease ClpP agonist that is well tolerated and induces durable tumor regressions in H3 K27M-mutant glioma patients. ONC206, a chemical derivative of ONC201 currently in Phase I trials for central nervous system tumors, is also a DRD2 and ClpP dual targeting imipridone with differentiated target engagement and nanomolar potency. We explored molecular differentiation and predictive biomarker studies for ONC201 and ONC206 using gene expression profiling, development of resistance clones and RNA-seq/proteomics in large cancer cell line panels including the 1000-cell line GDSC panel. Gene expression profiling revealed ONC206 and ONC201 induce partially overlapping signatures in U87 glioblastoma cells. Similarly, T98G glioblastoma cells with acquired resistance to ONC201 or ONC206 revealed partial cross resistance and combinatorial efficacy, supporting distinct functional effects. Next, we evaluated a curated 35-gene panel corroborated from prior literature on ONC201 including DRD2- (DRD2, DRD5, TH, EGFR), ClpP- (ClpP, NDUFS7, NDUFA12, SDHA, SDHB, POLD2, POLDIP2), epigenetic- (EZH2, EED, EZHIP, KDM1A, KDM3A, KDM4A, KDM6A, KDM6B), hypoxia- (HIF1, HIF2, HIF3, VHL, VEGFA, HES1, TGM2, CELF2, P4HA2, CEBPB) and downstream signaling-related (ATF3, ATF4, MYC, MYCN, GNA11, GNA15) genes as predictive biomarkers for ONC201/ONC206 efficacy (IC50, IC90, AUC and mRNA fold change) in the GDSC panel using CCLE RNA-seq data. Optimized thresholds for predictive power revealed that the top molecular correlatives prioritized for ONC201 (ClpP, TGM2, NDUFS7, EZH2, EGFR, POLD2, HIF2A, CEBPB, HES1) and ONC206 (ClpP, TGM2, NDUFS7, EZH2, EGFR, HIF1, VHL, ATF4, MYC, KDM6B) in GDSC partially overlap. Biomarker combinations with the prioritized markers improved predictive power compared to single markers. Dual (both markers must be co-expressed) or independent combinations involving EGFR, NDUFS7, EZH2 and ClpP were most predictive for ONC206 while those involving ClpP, EGFR and EZH2 were most predictive for ONC201. Results from GDSC for ONC201 were further confirmed using the PRISM dataset with RNA-seq and proteomics that revealed similar results. Hypoxic tumor cell cultures confirmed GDSC findings that elevated tumor cell hypoxia can impart ONC201 or ONC206 resistance. Thus, ONC206 is a distinct agent that may be uniquely poised to address tumors that are not addressed by or have developed acquired resistance to ONC201. Our results indicate that for accurate prediction of ONC201 and ONC206 clinical benefit, a curated combinatorial biomarker approach for each tumor type, using diverse detection approaches such as protein expression, genomics, and transcriptomics, may be used in ongoing clinical trials. Citation Format: Sara Morrow, Kirti Nath, Yiqun Zhang, Mathew J. Garnett, Ultan McDermott, Cyril H. Benes, Joel Basken, Wafik S. El-Deiry, Joshua E. Allen, Varun V. Prabhu. Predictive biomarker evaluation and molecular differentiation for imipridones ONC201 and ONC206 [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 393.