br E coli inside CRC tumors
E. coli, inside CRC tumors (Figure 6J). No significant differences in bacteria adherent to the colonic epithelium in naive mice were
(G) qRT-PCR for bacterial infiltration with primers specific for 16S rRNA from indicated bacteria, normalized to mouse housekeeping gene (RpL32), N = 18 *p < 0.05.
(H) Scheme of experiment for antibiotic-mediated microbiota depletion. CD11bCre+Il1rf/f or control BM-chimeric CDX2ERT-Apcf/f mice were injected with tamoxifen and put on broad-spectrum 150493-34-8 or regular water for 4 weeks.
(I) Representative images of CRC bearing colons from mice treated or untreated with antibiotics.
(J) Tumors upon necroscopy after 4 weeks of antibiotics treatment, N R 6.
Figure 5. Myeloid IL-1R Signaling Prevents Tumor Associated Dysbiosis
(A–I) CDX2ERT-Apcf/f mice were reconstituted with BM from LysMCre+Il1rf/f or Cre- (control) mice and CRC was induced with 2 injections of tamoxifen.
(A) Macroscopic tumor multiplicity, size, and load analysis upon necropsy 6 weeks after last tamoxifen injection, for mice which are either co-housed or housed separately according to BM genotype. N R 5, p < 0.05.
(B) Representative images of CRC bearing colons of BM transplanted mice
(D) qRT-PCR analysis of CRC tumor lysates for specific bacterial content indicated; results are normalized to RpL32 expression N = 20, *p < 0.05.
(E) IF images of adhesive bacteria (white arrows) associated with tumors from LysMCre+Il1rf/f or control mice. YoYo-1 and phalloidin staining of whole colon tissue from indicated mice. Representative of 3 independent experiments, each of at least 5 tumors.
(legend continued on next page)
detected (Figure S7A) Thus, increased infiltration of CRC tumors by bacteria in the absence of IL-1R signaling in neutrophils results in enhanced bacteria-driven inflammation and CRC tumorigenesis.
IL-1R Signaling Controls Anti-Bacterial Functions of Neutrophils
Neutrophil recruitment into the tumors was not affected by myeloid specific or neutrophil specific IL-1R ablation, while donor BM neutrophils exhibited profound deletion of exon 5 of the Il1r1 gene indicating efficient cell recruitment and cell spe-cific gene deletion (Figures S7B–S7D). (Figures 7A and 7B and S7B–S7D). We did not find any significant changes in the pres-ence of neutrophil extracellular traps (NET’s) in CRC tumors of LysMCre+Il1r1f/f and controls (Figure S7E). Also, ex vivo experi-ments did not detect significant differences in intracellular levels of bactericidal ROS in neutrophils (Figure S7F), although for these experiments intratumoral neutrophils undergo mechanical isolation possibly resulting in their activation. CD11b+Ly6G+ neu-trophils, enriched or sorted from spleens, bone marrow, thiogly-collate elicited peritoneal exudate or from CRC tumors, were able to kill various bacteria, including lab strain of E. coli and bulk fecal bacteria, upon IL-1b treatment. That ability was impaired when neutrophils were rendered IL-1R deficient (Fig-ures 7C–7E). We further found that the absence of functional IL-1R signaling led to a decreased ability of neutrophils to ingest zymosan particles or GFP-labeled Salmonella typhimurium bac-teria (Figures 7F–7I). Expression of antimicrobial peptides was also reduced in the absence of IL-1R in neutrophils (Figure 7J).
Overall our data suggested that neutrophils lacking IL-1R signaling have defective anti-bacterial functions. Together, these results underscore the importance of direct IL-1R signaling in neutrophils preventing tumor associated dysbiosis, over inflam-mation and CRC development.
Inflammatory cytokines are emerging as regulators of the tumor microenvironment through multiple mechanisms of action. Cyto-kines are also increasingly attractive targets in cancer because they are amenable to antibody-mediated neutralization without the toxicities of common chemotherapies. As more cytokines are found to be essential for tumor development, at least in pre-clinical models, it is expected that the number of cytokine blockers tested in cancer clinical trials will increase. However, whether the efficacy of anti-cytokine therapy might be subopti-mal because of cytokine signals operating within a particular spatio-temporal context, remains unexplored.
Previous studies have demonstrated that genetic or pharma-cologic blockade of a single pro-tumorigenic cytokine can cause an unambiguous reduction in CRC, because these TEI cytokines act on epithelial and cancer cells and drive tumor growth (Chae
et al., 2010; Grivennikov et al., 2009; Wang et al., 2014). Based on a vast literature implicating IL-1 signaling in the regulation of autoimmunity, inflammation, and specific induction of IL-17 pro-duction and differentiation of Th17 cells (Carmi et al., 2011; Na-kae et al., 2003; Shaw et al., 2012), we hypothesized that IL-1 driven signals would be essential for controlling TEI and thereby facilitating CRC growth. We found that whole body and hemato-poietic cell-specific IL-1R1 ablation did decrease expression of TEI cytokines, but unexpectedly caused very limited, if any, decrease in CRC. This presented us with a conundrum of why inactivation of IL-1R signaling in the hematopoietic system did not tame CRC, despite a reduction in pro-tumorigenic cytokines whose inactivation reproducibly reduces CRC in a variety of experimental conditions (Blatner et al., 2012; Grivennikov et al., 2012; Kirchberger et al., 2013). Given the multifaceted roles of IL-1R1 and the fact that IL-1R1 is expressed by virtually all cell types in the body, it was conceivable that IL-1 signaling may have opposing roles in different cell types. In this work we showed that in the CRC tumor microenvironment, IL-1 signaling plays distinct roles, depending on cell type, and established the cellular and molecular mechanisms governing how this particular cytokine regulated multipronged processes during intestinal tumorigenesis.