Abstract A075: A protective role for group 3 innate lymphoid cells in colitis-associated colorectal cancer

Chronic inflammation is recognized as a causative factor in the development of cancer and recent paradigms suggest that microbe-driven chronic inflammation is causally associated with the development and progression of cancer in the colon. Further, recent studies in mice have implicated innate lymphoid cells (ILC) as a key cell population that regulates intestinal inflammation. Group 3 innate lymphoid cells (ILC3) are essential regulators of immunity, inflammation and tissue homeostasis in the intestine, yet their role in cancer remains poorly defined. Here, we identify that ILC3 are associated with both human and mouse colon tumors. Tumor-associated ILC3 are selectively localized within lymphoid aggregates and exhibit a unique phenotypic profile as compared with nonmalignant tissue. Critically, mice with a selective deletion of ILC3-specific MHCII exhibit a striking increased susceptibility to intestinal tissue damage and develop highly invasive and flat colorectal tumors. These data demonstrate a protective role for antigen-presenting ILC3 in the context of cancer development and progression in the intestine, and suggest that further interrogation may lead to the development of novel immunotherapeutic strategies in colon cancer.

Citation Format: Jeremy Goc, Nick Bessman, Sheena Sahota, Flamar Anne Laure, Gregory Putzel, David Withers, Janelle Arthur, Manish Shah, Gregory Sonnenberg. A protective role for group 3 innate lymphoid cells in colitis-associated colorectal cancer [abstract]. In: Proceedings of the Fourth CRI-CIMT-EATI-AACR International Cancer Immunotherapy Conference: Translating Science into Survival; Sept 30-Oct 3, 2018; New York, NY. Philadelphia (PA): AACR; Cancer Immunol Res 2019;7(2 Suppl):Abstract nr A075.

Impact of antibiotic use on response to treatment with immune checkpoint inhibitors

143

Background: There is an emerging evidence that the gut microbiome may influence anti-tumor immune response during treatment with checkpoint inhibitors (CPIs). A number of factors can cause gut microbiome dysbiosis including the use of antibiotics (ABX). We hypothesized that the use of ABX during immunotherapy can adversely affect the efficacy of CPIs. We explored the prevalence of ABX use amongst patients (pts) using CPIs, and whether the use of ABX influences the response to CPIs. Methods: We performed an institutional retrospective review of all the pts treated with CPIs from 2/2015-3/2018. A patient was considered to have used ABX if he or she was prescribed ABX within 6 months before or after, initiating CPIs. Statistical analysis was done using logistic regression with overall response rate (ORR) (CR, PR and SD) as the outcome. 9 separate analyses were done: one for each temporality (30 days, 60 days, 6 months) and use-order (before, after, neither) combination. Odds ratios for ORR as univariate analysis, and adjusted for age and sex; and adjusted for age, sex, and tumor type were calculated. Results: Out of 242 pts, 111 were lung, 36 bladder, 35 renal, 16 gastrointestinal and 44 other cancers. 50% (121) pts received nivolumab, 28% (68) pembrolizumab and 21% (52) atezolizumab. 75%, 46% and 32% of the pts received ABX within 6 months, 60 days and 30 days of starting CPIs. Only ABX use in the first 30- or 60-days following CPI initiation was associated with inferior ORR (OR 0.42 [95% CI:0.23,0.76] p = 0.005 for 60-days). ABX use prior to initiation of CPI at any time point, or ABX use in the first 6 months of CPI use did not impact CPI efficacy, Table 1. Conclusions: ABX use within the first 60 days of starting CPI therapy is prevalent (32%). This study suggests that the use of ABX within 60 days following initiation of CPIs significantly negatively impacts the ORR. Unnecessary usage of ABX should be avoided, especially during the early phase of starting CPIs. [Table: see text]

Segmented filamentous bacteria induce systemic IgA responses to commensal bacteria (MUC4P.854)

IgA is a component of the mucosal barrier surface and plays an important part in maintaining separation between the host and the commensal microbiota. Mucosal IgA responses are considered short-lived and restricted to mucosal sites with no systemic dissemination. However, there is little known about the regulation and maintenance of IgA secreting plasma cells (PCs) and how they compare to bone marrow PCs. Interestingly, we find that IgA secreting PCs make up a large proportion of the bone marrow PC pool in mice colonized with segmented filamentous bacteria (SFB), but not in germ-free or conventional SFB free mice. The increase in IgA PCs in the bone marrow correlates with a large increase in serum IgA. The co-housing of SFB free mice with SFB colonized mice results in significantly increased levels of IgA in the serum and higher numbers of IgA+ PCs in the bone marrow. This process is dependent upon T cells, because SFB colonized TCRβ-/-δ-/- mice have similar levels of IgA bone marrow PCs as wild-type SFB free mice. The SFB associated increase of IgA antibody titers coincides with the ability to detect commensal bacteria specific IgA in the serum. These data suggest that the induction of systemic IgA is influenced by the composition of the commensal microbiota. These results highlight the need for greater understanding of the role commensal bacteria play in systemic immunity, which could impact oral vaccine development and autoimmunity.

Innate lymphoid cells are critical regulators of host-microbiota interactions and intestinal homeostasis (P3152)

Mammals are colonized with trillions of commensal bacteria that are essential for normal intestinal physiology and immune cell development. However, dysregulated host-microbiota interactions have been causally linked with the pathogenesis and progression of numerous chronic infectious, inflammatory and metabolic diseases, such as HIV, cancer, obesity, viral hepatitis and inflammatory bowel disease (IBD). Innate lymphoid cells (ILCs) are an emerging family of immune cells that have recently been identified to play critical roles in regulation of immunity, inflammation and homeostasis in the intestine of mice and furthermore alterations in ILC populations have been observed in several chronic human diseases. However, whether ILCs are involved in the regulation of host-microbiota interactions is poorly understood. Here we show that ILCs are critical regulators of host immune responses to the microbiota through several distinct pathways. Surprisingly, abrogation of ILCs or their effector pathways resulted in pathologic host adaptive immune cell responses to resident intestinal microbiota and induction of spontaneous inflammation in the intestine and peripheral tissues. Collectively, these results identify an essential role for ILCs in regulating interactions between the mammalian adaptive immune system and commensal microbiota that dictate the state of intestinal health, and may represent a novel therapeutic target in chronic human disease.

TSLP elicits IL-33-independent innate lymphoid cell responses to promote skin inflammation (P3040)

Innate lymphoid cells (ILCs) are a recently identified family of immune cells that regulate inflammation and tissue repair at barrier surfaces such as the intestine and lung. Previous studies have demonstrated that group 2 ILCs found in the intestine or lung respond to the predominantly epithelial cell-derived cytokines IL-33 and IL-25 and produce the type 2 cytokines IL-5 and IL-13. However, whether group 2 ILCs are found in the skin, whether they respond to epithelial cell-derived cytokines and whether they contribute to skin inflammation has not been characterized. Here, we identified a population of skin-resident group 2 ILCs present in healthy human skin and enriched in lesional human skin isolated from atopic dermatitis (AD) patients. Group 2 ILCs were also found in normal murine skin, accumulated in the skin and skin-draining lymph nodes in a murine model of AD-like inflammation and were critical for the elicitation of type 2 cytokine responses and AD-like inflammation. Remarkably, ILC responses and AD-like inflammation were independent of IL-33 and IL-25 but critically dependent on thymic stromal lymphopoietin (TSLP). Collectively, these results demonstrate an essential role for IL-33- and IL-25-independent group 2 ILCs in promoting skin inflammation.

Persistent enteric murine norovirus infection is associated with functionally suboptimal virus-specific CD8 T cell responses (P3187)

Norovirus (NV) gastroenteritis is a major contributor to global morbidity and mortality, yet little is known about immune mechanisms leading to NV control. Previous studies using the murine norovirus (MNV) model have established a key role for T cells in MNV clearance. Despite these advances, important questions remain regarding the magnitude, location, and dynamics of the MNV-specific T cell response. To address these questions, we identified MNV-specific MHC class I immunodominant epitopes by conducting an overlapping peptide screen. Using MHC class I-peptide tetramers, we tracked MNV-specific CD8 T cells in lymphoid and mucosal sites during infection with two strains with distinct biological behavior: acutely-cleared CW3 and persistent CR6. Here, we show that enteric MNV infection elicited robust T cell responses primarily in the intestinal mucosa, and that MNV-specific T cells dynamically regulated the expression of surface molecules associated with activation, differentiation, and homing. Furthermore, compared to MNV-CW3, chronic infection with MNV-CR6 resulted in fewer and less functional CD8 T cells. Finally, MNV-specific CD8 T cells were capable of reducing viral load in persistently infected Rag1-/- mice, consistent with a crucial role for these cells in vivo. Collectively, these data provide fundamental new insights into the adaptive immune response to NV and bring us closer to understanding the correlates of protective antiviral immunity in the intestine.

Innate lymphoid cells promote airway epithelial repair through the amphiregulin-EGFR pathway (P3250)

Innate lymphoid cells (ILCs) are a recently described group of immune cells that regulate immunity and inflammation at barrier surfaces. While recent studies have implicated Group 2 ILCs as initiators of pathologic Type 2 airway inflammation, the tissue-protective roles of ILCs in the respiratory tract remain poorly understood. Using genome-wide transcriptional profiling, we identify that naïve lung-resident group 2 ILCs express multiple genes associated with tissue repair pathways, including the epidermal growth factor family member amphiregulin. Amphiregulin was increased in a model of influenza virus-induced lung injury and ILC depletion during infection resulted in decreased amphiregulin expression, reduced lung function and impaired airway remodeling. IL-33 promoted expression of amphiregulin in lung ILCs and administration of amphiregulin effectively restored airway tissue remodeling in ILC-depleted mice, suggesting that ILC-intrinsic amphiregulin is a central mediator of epithelial repair. Supporting this, chemical or genetic inhibition of the EGFR-amphiregulin pathway resulted in severely impaired lung function and a failure to restore airway tissue homeostasis, leading to increased host mortality. Taken together, these data indicate a critical role for group 2 ILCs in mediating lung tissue homeostasis though an IL-33-amphiregulin-EGFR axis and suggest that ILCs serve as a key link between the signaling pathways of Type 2 inflammation and tissue repair.

The genetic basis of plasma variation in adiponectin, a global endophenotype for obesity and the metabolic syndrome

Here we present the first genetic analysis of adiponectin levels, a newly identified adipocyte-derived protein. Recent work has suggested that adiponectin may play a role in mediating the effects of body weight as a risk factor for coronary artery disease. For this analysis we assayed serum levels of adiponectin in 1100 adults of predominantly northern European ancestry distributed across 170 families. Quantitative genetic analysis of adiponectin levels detected an additive genetic heritability of 46%. The maximum LOD score detected in a genome wide scan for adiponectin levels was 4.06 (P = 7.7 x 10(-6)), 35 cM from pter on chromosome 5. The second largest LOD score (LOD = 3.2; P = 6.2 x 10(-5)) was detected on chromosome 14, 29 cM from pter. The detection of a significant linkage with a quantitative trait locus on chromosome 5 provides strong evidence for a replication of a previously reported quantitative trait locus for obesity-related phenotypes. In addition, several secondary signals offer potential evidence of replications for additional previously reported obesity-related quantitative trait loci on chromosomes 2 and 10. Not only do these results identify quantitative trait loci with significant effects on a newly described, and potentially very important, adipocyte-derived protein, they also reveal the emergence of a consistent pattern of linkage results for obesity-related traits across a number of human populations.

Skeletal muscle glycolysis, oxidation, and storage of an oral glucose load

Although muscle is considered to be the most important site for postprandial glucose disposal, the metabolic fate of oral glucose taken up by muscle remains unclear. We, therefore, employed the dual isotope technique (intravenous, [6-3H]-glucose; oral, [1-14C]glucose), indirect calorimetry, and forearm balance measurements of glucose, lactate, alanine, pyruvate, O2, and CO2 in nine normal volunteers to determine the relative importance of muscle glycogenic, glycolytic, and oxidative pathways in disposal of an oral glucose load. During the 5 h after glucose ingestion (1 g/kg), 37 +/- 3% (24.9 +/- 2.3 g) of the load was oxidized and 63 +/- 3% (42.8 +/- 2.7 g) was stored. At least 29% (19.4 +/- 1.3 g) was taken up by splanchnic tissues. Muscle took up 26% (17.9 +/- 2.9 g) of the oral glucose coincident with a 50% reduction in its oxidation of fat. 15% of the oral glucose taken up by muscle (2.5 +/- 0.9 g) was released as lactate, alanine, or pyruvate; 50% (8.9 +/- 1.4 g) was oxidized, and 35% (6.4 +/- 2.3 g) was available for storage. We conclude that muscle and splanchnic tissues take up a comparable percentage of an oral glucose load and that oxidation is the predominant fate of glucose taken up by muscle, with storage in muscle accounting for less than 10% of the oral load. Thus, contrary to the prevailing view, muscle is neither the major site of storage nor the predominant site of disposal of an oral glucose load.

Incidence and management of severe hypoglycemia in 434 adults with insulin-dependent diabetes mellitus

The risk of severe hypoglycemia associated with the particular therapeutic approach of two University hospitals was assessed in 96% of all patients with insulin-dependent diabetes mellitus (IDDM) who had been admitted during a period of almost 3 yr to the diabetic wards of two hospitals and who participated in a structured teaching and treatment program. During a mean follow-up period of 18 mo, 10% of the conventionally treated patients (N = 384; age 30 +/- 13 yr; duration of diabetes 12 +/- 9 yr) and 9% of the CSII-treated patients (N = 50, age 28 +/- 7 yr, duration of diabetes 13 +/- 7 yr, total follow-up period 1093 patient-mo) experienced at least one severe hypoglycemic episode per year, and a total of 123 severe hypoglycemic episodes occurred. In a subgroup of 169 conventionally treated patients, mean glycosylated hemoglobin values decreased from 10.5 +/- 1.9% before participation in the program to 9.2 +/- 2.0% (P less than 0.001) 18 +/- 4 mo thereafter. For the CSII-treated patients, glycosylated hemoglobin values were 9.7 +/- 1.9% before initiation of pump therapy and remained at the upper normal range from 3 mo thereafter throughout the study. There was no relationship between glycosylated hemoglobin levels and the occurrence of severe hypoglycemic episodes. Fifty-three severe hypoglycemic episodes were treated with glucagon injections by the patients' relatives (all but one effectively), 30 were managed by assisting physicians, and 44 led to hospitalization. Thus, successful attempts to improve glycosylated hemoglobin values in an unselected group of patients with IDDM were not associated with an unduly high risk of severe hypoglycemia when compared with the scarce data from the literature.

Absorption kinetics of semisynthetic human insulin and biosynthetic (recombinant DNA) human insulin

Absorption kinetics of subcutaneously injected human insulin (recombinant DNA) and semisynthetic human regular insulin were investigated and compared to the respective porcine insulin preparations in normal volunteers. The absorption of all regular human insulin preparations tested was accelerated as compared to porcine insulins. The clinical relevance of these findings in regard to the treatment of diabetes mellitus appears to be doubtful. As for mixtures of porcine regular and intermediate-acting insulin preparations, the absorption of human regular insulin is not altered when mixed with intermediate-acting human insulin.