Regulation of Peripheral Inflammation by a Non-Viable, Non-Colonizing Strain of Commensal Bacteria

Abstract

The gastrointestinal tract represents one of the largest body surfaces that is exposed to the outside world. It is the only mucosal surface that is required to simultaneously recognize and defend against pathogens, while allowing nutrients containing foreign antigens to be tolerated and absorbed. It differentiates between these foreign substances through a complex system of pattern recognition receptors expressed on the surface of the intestinal epithelial cells as well as the underlying immune cells. These immune cells actively sample and evaluate microbes and other particles that pass through the lumen of the gut. This local sensing system is part of a broader distributed signaling system that is connected to the rest of the body through the enteric nervous system, the immune system, and the metabolic system. While local tissue homeostasis is maintained by commensal bacteria that colonize the gut, colonization itself may not be required for the activation of distributed signaling networks that can result in modulation of peripheral inflammation. Herein, we describe the ability of a gut-restricted strain of commensal bacteria to drive systemic anti-inflammatory effects in a manner that does not rely upon its ability to colonize the gastrointestinal tract or alter the mucosal microbiome. Orally administered EDP1867, a gamma-irradiated strain of Veillonella parvula, rapidly transits through the murine gut without colonization or alteration of the background microbiome flora. In murine models of inflammatory disease including delayed-type hypersensitivity (DTH), atopic dermatitis, psoriasis, and experimental autoimmune encephalomyelitis (EAE), treatment with EDP1867 resulted in significant reduction in inflammation and immunopathology. Ex vivo cytokine analyses revealed that EDP1867 treatment diminished production of pro-inflammatory cytokines involved in inflammatory cascades. Furthermore, blockade of lymphocyte migration to the gut-associated lymphoid tissues impaired the ability of EDP1867 to resolve peripheral inflammation, supporting the hypothesis that circulating immune cells are responsible for promulgating the signals from the gut to peripheral tissues. Finally, we show that adoptively transferred T cells from EDP1867-treated mice inhibit inflammation induced in recipient mice. These results demonstrate that an orally-delivered, non-viable strain of commensal bacteria can mediate potent anti-inflammatory effects in peripheral tissues through transient occupancy of the gastrointestinal tract, and support the development of non-living bacterial strains for therapeutic applications.

Keywords: CNS; Th1; Th17; Th2; bacteria; gastrointestinal tract; inflammation.

Figure 1

EDP1867 induces IL-10 in vitro and resolves inflammation in vivo. (A) Human Macrophages were stimulated with different anaerobic bacterial strains for 24h and flushed with 1% oxygen and supernatants were collected to test for cytokine levels by MSD. Data shown represent collective data from 6 independent human donors. (B) Human PBMCs, Dendritic Cells and Macrophages were stimulated with EDP1867 bacterial strains for 24h and flushed with 1% O₂ and supernatants were collected to test for cytokine levels by MSD. Data shown represent collective data from 6 independent human donors. (C) DTH response to KLH. C57BL/6 mice were immunized with KLH and CFA on day 0 s.c. and challenged in the ear i.d. 9 days later with KLH. Mice were orally dosed daily from the day after immunization through ear challenge with vehicle or EDP1867; TCC-2.84E+10. Ear inflammation was measured on day 9. Change in ear thickness (n = 5 mice/group) for groups dosed with EDP1867; TCC -7.8E+11 and other non-replicating Veillonella strains (TCC-3.8E+11- 1.03E+12). (D) DTH was performed as previously described. Mice were orally dosed daily from the day after immunization through ear challenge with vehicle or EDP1867 (TCC 2.16E+12). Ear inflammation was measured on day 9. Change in ear thickness (n = 5 mice/group) demonstrate dose dependent effects of EDP1867. (E) EDP1867 acts through IL-10R pathway to reduce ear inflammation. Mice in various groups were treated with IL-10R blocking antibody on days 2,4, and 6 as indicated (EDP1867 TCC 2.16E+12). Representative figure from n = 2 experiments with 5mice/group in each experiment; All data show mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns, not significant. as determined by unpaired Student’s t-test. DTH, delayed-type hypersensitivity; KLH, keyhole limpet hemocyanin; CFA, complete Freund’s adjuvant; TCC, total cell count; s.c., subcutaneous; i.p., intraperitoneal; i.d., intradermal.

Figure 2

EDP1867 is gut restricted and transits through GI tract within 24hrs. Biodistribution of EDP1867. Following a single oral administration of EDP1867 (TCC – 1E+9), exposure of the small intestine occurs by 10 mins and complete exposure in 1 hour. By 6 hours, EDP1867 is incorporated into fecal pellets and by 24 hours signal in GIT diminishes suggesting complete excretion of the microbe with no detectable persistence. No systemic exposure is observed at any timepoint and >99% of the total signal remains in the GIT. Left column– brightfield images, middle- free dye, right- EDP1867 overlapped with free dye GIT- gastrointestinal tract. n = 3 mice/timepoint/group. Data are represented as mean + SEM. p values are represented as **p < 0.01, ns > 0.1, as determined by 2-way ANOVA followed by Sidak’s multiple comparison test.

Figure 3

EDP1867 alleviates skin inflammation in imiquimod-induced psoriasis. Imiquimod induced psoriasis. BALB/c mice were topically treated with 20mg 5% imiquimod for 7 days on the ear. Mice were orally dosed daily from day 1 through 7 with vehicle or EDP1867 (TCC – 7.8E+11). (A) Ear inflammation over the course of 7 days and area under the curve. (B) mRNA transcript levels for Il17a, Il17f and Defb3 measured by qPCR. Data are representative from 2 experiments with n = 5/group. All data show mean ± SEM. *p < 0.05, ***p < 0.0005, ****p < 0.0001, as determined by unpaired Student’s t-test.

Figure 4

EDP1867 displays efficacy treating neuroinflammation in a model for relapsing remitting MS. PLP driven EAE. EAE was induced in SJL mice by immunization with PLP 139-151 in CFA on day 0, hour 0 and PTX was administered on day 0, hour 2. (A) Prophylactic dosing – vehicle, EDP1867 (TCC- 8.46E+10) and fingolimod (1mg/kg) were orally dosed daily for 41 days. Cumulative EAE scores of mice. Clinical scores were assessed daily for the duration of the experiment. ****p ≤ 0.00005 by Unpaired t-test with Welch’s correction was used to calculate p-value in cumulative AUC for the EAE score. (B) Representative hematoxylin and eosin (H&E)-stained images of the brain of spinal cords of mice treated with EDP1867, Vehicle or Fingolimod. Spinal cord sections are enlarged at 289X magnification to show regions with inflammation and inflammatory loci. Data for Figures 5B, C are representative of 2 independent experiments (n = 15 mice per group). (C) Therapeutic dosing – vehicle, EDP1867 (TCC- 8.46E+10) and fingolimod (1mg/kg) were orally dosed from day 10 to day 41. Cumulative EAE scores of mice. Clinical scores were assessed daily for the duration of the experiment. ****p ≤ 0.00005 by Unpaired t-test with Welch’s correction was used to calculate p-value in cumulative AUC for the EAE score. (D) Representative Luxol blue-stained images of the brain of spinal cords of mice treated with EDP1867, Vehicle or Fingolimod. Data for Figures 5D, E are representative from n = 15 mice per group. *p ≤ 0.05 by unpaired Student’s t-test. PLP, proteolipid protein; EAE, experimental autoimmune encephalomyelitis; PTX, pertussis toxin; CFA, complete Freund’s adjuvant. ns, not significant.

Figure 5

EDP1867 resolves inflammation in a Th2 cell driven atopic dermatitis model. FITC- driven AD. BALB/c mice were topically sensitized with 0.5% FITC, on day 1 and 2 and 6 days later challenged with 0.5% FITC on the ear. Mice were dosed daily orally with vehicle or EDP1867 (TCC – 2.16E+12). Ear inflammation was measured 24h post ear challenge on day 7. (A) Change in ear thickness. (B) At termination on day 7, total cells from mesenteric lymph nodes and cervical lymph nodes were restimulated with PMA for 48h. and ear tissue homogenates were prepared. Cytokines were measured from supernatants and homogenates by multiplex ELISA. Data are representative of 3 independent experiments (n = 5 mice/group). *p < 0.05, ***p < 0.0005, as determined by unpaired Student’s t-test.

Figure 6

EDP1867 resolves inflammation in a delayed type hypersensitivity T cell driven disease model in vivo. KLH-DTH was induced as previously described. Mice were orally dosed daily from day 1 to 8 with a compressed tablet form of placebo or EDP1867. Ear inflammation was measured on day 9. (A) Change in ear thickness (n = 10 mice/group). (B) At termination on day 9 total cells from mesenteric lymph nodes and spleen were restimulated with PMA for 48h and total cells from ear draining lymph nodes were restimulated with KLH for 72h. Cytokines from supernatants were measured by multiplex ELISA. Data are representative from 2 experiments with n = 10/group. All data show mean ± SEM. *p < 0.05, **p < 0.001, ***p < 0.0005, ****p < 0.0001, as determined by unpaired Student’s t-test.

Figure 7

EDP1867 requires trafficking of gut immune cells to periphery to impact efficacy. Mice were treated with anti-LPAM1/CD62L blocking antibody or a rat IgG2a isotype antibody intraperitoneally on days 1,3, 5, and 7. Mice were challenged on day 14 and change in ear thickness was measured on day 15. All studies have n = 5 mice/group. Representative figure from n = 2 experiments. All data show mean ± SEM. **p < 0.01, ****p < 0.0001, ns, not significant as determined by ordinary One-Way ANOVA.

Figure 8

Adoptively transferred EDP1867-treated CD4 T cells mediate efficacy in DTH. Adoptive transfer of CD4+ T cells from mice treated with EDP1867. Mice with DTH were treated with EDP1867 for 4 days and then CD4 T cells from these treated animals were adoptively transferred into recipient immunized animals that were not dosed with EDP1867. Representative figure from n = 2 experiments with 5mice/group in each experiment; ****p < 0.0001, as determined by Ordinary one-way ANOVA.