Gut Microbes Egested during Bites of Infected Sand Flies Augment Severity of Leishmaniasis via Inflammasome-Derived IL-1β

Abstract

Leishmania donovani parasites are the cause of visceral leishmaniasis and are transmitted by bites from phlebotomine sand flies. A prominent feature of vector-transmitted Leishmania is the persistence of neutrophils at bite sites, where they protect captured parasites, leading to enhanced disease. Here, we demonstrate that gut microbes from the sand fly are egested into host skin alongside Leishmania parasites. The egested microbes trigger the inflammasome, leading to a rapid production of interleukin-1β (IL-1β), which sustains neutrophil infiltration. Reducing midgut microbiota by pretreatment of Leishmania-infected sand flies with antibiotics or neutralizing the effect of IL-1β in bitten mice abrogates neutrophil recruitment. These early events are associated with impairment of parasite visceralization, indicating that both gut microbiota and IL-1β are important for the establishment of Leishmania infections. Considering that arthropods harbor a rich microbiota, its potential egestion after bites may be a shared mechanism that contributes to severity of vector-borne disease.

Keywords: IL-1β; Leishmania; disease severity; gut microbiota; inflammasome; neutrophil; parasite dissemination; sand fly; skin inflammatory response; vector-borne disease.

Figure 1. The inflammatory response at the bite site of vector-transmitted L. donovani parasites

(A) Parasite burden determined by qPCR in individual mice ears 3h after exposure to 20 infected sand flies (n ≥ 6 mice ears per experiment). (B) Parasite burden in individual mice spleens was determined by serial dilution at five and 30 weeks after exposure to 20 infected sand flies (IS) belonging to the same group or to an intradermal injection of 10⁵ metacyclic parasites (LI). Data are pooled from two independent experiments (n ≥ 13 for IS, and ≥ 5 mice for LI). (C–I) Mice ears 3–48h after exposure to 20 IS, 20 uninfected sand flies (US), intradermal co-inoculation with LI plus extract of one salivary gland (LG), and LI. (C) Mice ear sections stained with anti-Gr1 antibody. Pictures are representative of four independent experiments (n ≥ 4 mice ears per condition per timepoint). Scale bar indicate 20µm. (D–F) Cells recovered from individual mice ears were stained for flow cytometry. Data are representative of two independent experiments (n = 6 mice ears per condition per timepoint). (D) Representative plots of Ly6G and Ly6C expression distinguish neutrophils (square gate) and inflammatory monocytes (oval gate) graphed in (E) and (F), respectively. (G) Degranulated mast cells in at least six fields counted from ear sections stained with Toluidine blue. Data are representative of four independent experiments (n ≥ 5 mice ears per condition per timepoint). (H and I) mRNA expression of inflammatory mediators produced in mice ears determined by qPCR. Data are pooled from more than three independent experiments (n ≥ 7 mice ears per condition per timepoint). (H) Heat map of the global inflammatory skin response. (I) Expression profile of distinct genes relevant to the inflammatory response after IS. Statistical significance *P < 0.05; **P < 0.01; ***P < 0.001, ****P < 0.0001 was calculated using the Wilcoxon ranked sum test (A and B), the one-way ANOVA followed by a Holm-Sidak multiple comparisons test (E–G and I). Error bars in (A, B) indicate the geometric mean with 95% CI. Error bars in (E–G and I) indicate the mean ± SEM. See also Figure S1, S2 and S3.

Figure 2. L. donovani-infected sand fly bites activate the NLRP3 inflammasome in neutrophils

(A, C–G and I) Mice ears processed 6h after exposure to 20 infected (IS) or uninfected (US) sand flies. (A) Ex-vivo IL1β protein levels from ear lysates measured by ELISA. Data are representative of two independent experiments (n ≥ 7 mice ears per condition). Dotted line represents the mean + SEM of endogenous IL1β levels in naïve samples. (B) Mice ear sections stained with either anti-IL1β antibody targeting the mature protein or its isotype control (CTL) at 3–18h after IS or US. Pictures are representative of two independent experiments (n = 2 mice ears per condition per timepoint). Scale bars indicate 20µm. Arrows indicate IL1β secreting cells. (C–E) Cells recovered from mice ears after IS were stained for flow cytometry. Representative plots of IL1β⁺ cells (C) back-gated (red) onto LY6G⁺ / CD11b⁺ neutrophils (D). (E) The number of IL1β⁺ neutrophils per 10⁵ ear cells from two independent experiments (n = 6 mice ears per condition). (F) Representative Western blot of NLRP3, Caspase 1 and IL1β protein levels after IS and US (n = 4 mice ears per condition). (G) Representative Western blot of NLRP3 protein levels after immunoprecipitation using anti-ASC antibody (n = 2 mice ears per condition). (H) NLRP3 and IFNβ mRNA expression determined by qPCR at 3–18h after IS or US. Data are pooled from three independent experiments (n ≥ 7 mice ears per condition per timepoint). (I) Mice ear sections stained with TUNEL. Pictures are representative of two independent experiments (n = 3 mice ears per condition). Boxed area is magnified to highlight apoptotic cells (arrow). Scale bars indicate 20µm. (J) Mature IL1β and IL-10 cytokine levels in cell culture supernatant after in vitro stimulation of bone marrow-derived macrophages with L. donovani (Ld) stationary parasites in the presence or absence of lipopolysaccharide (LPS). Med, medium. Data are representative of two to three independent experiments (n = 2 replicates per condition). Statistical significance *P < 0.05, **P < 0.01, ***P < 0.001 calculated by the unpaired two-tailed t test (A and H), the Mann-Whitney test (E), and the one-way ANOVA followed by a Holm-Sidak multiple comparisons test (J). Error bars indicate the mean ± SEM. See also Table S1.

Figure 3. Identification of gut microbes egested during infected sand fly bites

(A) Bacterial colonies growing on LB/Agar after exposure to L. donovani-infected sand flies. Inset shows the mesh imprint of the feeder containing the sand flies. Pictures representative of five independent experiments. (B) Representative Real-Time PCR amplification curves (red lines) for Leishmania minicircle kDNA using bacterial colonies picked from agar plates as template. Green line, Threshold. Blue Lines, Negative controls (water or agar). (C) Bacteria identified from Leishmania-positive colonies. I, Real-time PCR C_T values for amplified Leishmania. II, Taxonomy of bacteria ordered by phylum, family, and genus. *, colonies depicted in (B). (D) Bacteria identified from the midgut of infected sand fly groups used in (A). III, Taxonomy of bacteria ordered by phylum, family, and genus. See also Video S1.

Figure 4. Diminishing sand fly gut microbiota mitigates inflammasome activation, abrogates neutrophil recruitment and compromises dissemination of Leishmania donovani

(A–C and F–L) Sand flies harboring mature infections with L. donovani were left untreated (IS) or were provided a cocktail of three antibiotics for 36h prior to parasite transmission to mice ears (IS+ATB). (A) Bacterial colony growth on LB/Agar after plating10 midguts. (B) Survival of sand flies without (red) and after antibiotic treatment (green). (C) Parasite burden and percent metacyclics in sand flies prior to transmission. (A–C) Data are representative of four independent experiments (n ≥ 9 sand flies per condition per timepoint). (D and E) In vitro invasion and multiplication of anterior gut residing parasites from IS or IS+ATB in bone marrow-derived macrophages (BMDM). Data are representative of two independent experiments. (D) BMDM 48h after infection. Arrows indicate intracellular Leishmania amastigotes. (E) Percent of infected BMDM and the number of parasites per cell. (F) Feeding score after parasite transmission to mice ears. Data representative of four independent experiments (n = 16 sand flies per condition). (G–J) Mice ear lysates after exposure to 20 IS or IS+ATB. (G) Parasite burden determined by qPCR from individual mice ears 2h after bites. (H) Ex-vivo IL1β protein levels measured by ELISA 6h after bites. (G and H) Data are pooled from two independent experiments (n ≥ 15 mice ears per condition). (I) Representative Western blot of NLRP3, Caspase 1 and IL1β protein levels (n = 2 mice ears per condition). (J) Representative Western blot of NLRP3 protein levels after immunoprecipitation using anti-ASC antibody (n = 2 mice ears per condition). (K and L) Mice ears were exposed to IS or IS+ATB. (M and N) Mice were left untreated or were treated with anakinra 12h and immediately before exposure to IS (IS+Ana). (K and M) Mice ear sections stained with anti-Gr1 antibody 6h after exposure to IS. Data are representative of two independent experiments (n = 4–5 mice ears per condition). Scale bars indicate 50µm. (L and N) Parasite burden in spleens of individual mice determined by serial dilution three weeks after exposure to IS or IS+ATB. Data are pooled from two independent experiments distinguished by solid and clear symbols (n ≥ 8 mice per condition). Statistical significance *P < 0.05, **P < 0.01was calculated using the unpaired two-tailed t test (H) and the Wilcoxon ranked sum test (L and N). Error bars indicate the geometric mean with 95% CI for parasite burden (C), (G), (L) and (N), and the mean ± SEM for (B), percent metacyclics (C), (E), (F) and (H). See also Figure S4 and Table S1.