Gut Bacteria Regulate the Pathogenesis of Huntington's Disease in Drosophila Model

Abstract

Changes in the composition of gut microbiota are implicated in the pathogenesis of several neurodegenerative disorders. Here, we investigated whether gut bacteria affect the progression of Huntington's disease (HD) in transgenic Drosophila melanogaster (fruit fly) models expressing full-length or N-terminal fragments of human mutant huntingtin (HTT) protein. We find that elimination of commensal gut bacteria by antibiotics reduces the aggregation of amyloidogenic N-terminal fragments of HTT and delays the development of motor defects. Conversely, colonization of HD flies with Escherichia coli (E. coli), a known pathobiont of human gut with links to neurodegeneration and other morbidities, accelerates HTT aggregation, aggravates immobility, and shortens lifespan. Similar to antibiotics, treatment of HD flies with small compounds such as luteolin, a flavone, or crocin a beta-carotenoid, ameliorates disease phenotypes, and promotes survival. Crocin prevents colonization of E. coli in the gut and alters the levels of commensal bacteria, which may be linked to its protective effects. The opposing effects of E. coli and crocin on HTT aggregation, motor defects, and survival in transgenic Drosophila models support the involvement of gut-brain networks in the pathogenesis of HD.

Keywords: Huntington’s disease; crocin (PubChem CID: 5281233); gut-brain; microbiota; neurodegeneration.

FIGURE 1

Microbiota modulate HTT aggregation, locomotor behavior, and lifespan of Drosophila HD models. (A) qPCR amplification of total bacteria from DNA extracted from different batches of HTTex1-expressing offspring. Data are shown as fold change over the value for one batch of WT HTTex1. (B) Western blot detection of mutant HTTex1 aggregates in the lysates of untreated or rifaximin-treated Ex1-HD larvae, using anti-HTT (PHP1) antibody. Each lane represents 10 pooled larvae (N = 10) isolated from a different fly vial. (C) Representative confocal images of brains (N = 3) from the Ex1-HD larvae treated with rifaximin and immunolabelled with anti-HTT (PHP1, green). DAPI (blue) was used to stain the nuclei. (D) SDD-AGE and WB analysis of lysates (50 μg each) derived from female N-586 HD flies untreated (control), or colonized with E. coli or E. coli expressing curli. PHP1 antibody was used to detect HTT aggregates. For each condition 10 larvae were pooled together for analysis (N = 10). (E) Female N-586 HD flies were fed curli-producing/deficient E. coli or L. rhamnosus. A climbing assay was performed at day 20 after eclosion as described in M&M. Data are reported as mean ± SEM and were analyzed by one-way ANOVA with Tukey’s post hoc test. ^***p < 0.001, n = 4 groups of 10 flies. (F) The percentage of flies surviving at day 20 was calculated and plotted for each experimental condition. Data are represented as mean ± SEM and were analyzed by one-way ANOVA with Tukey’s post hoc test. ^***p < 0.001; ^**p < 0.01, n = 4 groups of 10 flies.

FIGURE 2

E. coli enhances the seeding activity of mutant HTT assemblies in female N-586 HD flies. (A) SDS-PAGE analysis of brain lysates followed by WBs of ∼15-day old non-transgenic (DA) or N-586 HD flies treated with E. coli probed with anti-HTT PHP1. Details are provided in the section “Materials and Methods.” (B) SDD-AGE of equal amounts of seeds from lysates in part A added to 50 μg of human neuronal lysates before or after treatment with proteinase-K (PK) (0.5 μg/ml of PK at RT for 30 min and subsequently inactivated by heating at 75°C (Chongtham et al., 2020). Seeds mixed with neuronal lysates were incubated at RT for 4 h and subsequently examined by SDD-AGE. Products were probed with PHP1 antibody.

FIGURE 3

Gut bacteria regulate the motor function and lifespan of female FL-HD Drosophila. (A) Newly eclosed FL-HD flies or controls expressing full length HTT with 25Qs (WT-HTT) were treated with penicillin-streptomycin (ABX) for 15 days to eliminate gut bacteria. A climbing assay was performed at days 1, 5, 10, and 15 post-treatment. The data are graphed as mean ± SEM, two-way ANOVA with Tukey’s multiple-comparisons test. ***p < 0.001, n = 9 groups of 10 flies. (B) Bacteria-deficient adult FL-HD flies were fed different strains of E. coli, A. senegalensis, and L. rhamnosus for 15 days. Locomotive behavior for each condition was quantified as in part (A). The data are reported as mean ± SEM, two-way ANOVA with Tukey’s multiple-comparisons test. ***p < 0.001; **p < 0.01, n = 6 groups of 10 flies. Part (C) shows the survival curve for FL-HD flies treated with two E. coli strains. The percent of flies that survived over time was calculated at different time points. The data are represented as mean ± SEM, two-way ANOVA with Tukey’s multiple-comparisons test. ***p < 0.001; **p < 0.01; *p < 0.05, n = 6 groups of 10 flies. (D) Immunostaining of the GI tract of control or E. coli (curli) treated FL-HD flies demonstrating bacterial colonization of fly gut. A monoclonal antibody generated to E. coli was used for detection. DAPI was used to stain the nuclei.

FIGURE 4

Crocin ameliorates E. coli-induced motor defects and mortality in female FL-HD flies. (A) Freshly eclosed FL-HD flies were treated with rifaximin, luteolin, and crocin for 15 days and their climbing ability was evaluated. Data are reported as mean ± SEM and were analyzed by one-way ANOVA with Tukey’s post hoc test. ***p < 0.001, n = 6 groups of 10 flies. (B) Climbing assay was performed to monitor the motor function of flies colonized with crocin, E. coli or E. coli plus crocin. Untreated FL-HD flies were used as control. Data are represented as mean ± SEM and were analyzed by one-way ANOVA with Tukey’s post hoc test. ***p < 0.001; **p < 0.01, n = 6 groups of 10 flies. Part (C,D) show the percentage of FL-HD flies, which survived over time (days) under different treatments. Temperature was elevated to 25°C to accommodate E. coli growth. The data are represented as mean ± SEM, two-way ANOVA with Tukey’s multiple-comparisons test. ***p < 0.001; **p < 0.01; *p < 0.05, n = 6 groups of 10 flies. (E) Representative confocal images of the GI tract of untreated (control) FL-HD flies or those treated with E. coli (curli) or E. coli plus crocin for 15 days showing bacterial colonization and suppression by crocin treatment. Immunostaining was performed using a monoclonal antibody reactive to E. coli DAPI was used to stain the nuclei.

FIGURE 5

Crocin and antibiotics reduce HTTex1 aggregation and mortality. (A) SDD-AGE and WB analysis of lysates generated from the Ex1-HD larvae treated with crocin or penicillin/streptomycin (ABX). PHP1 antibody was used to detect aggregates. Elav-Gal4 larvae were used as a negative control. For each condition, ten larvae (n = 10) were pooled and analyzed. (B) Representative confocal images of brains from Ex1-HD larvae untreated (Control) and treated with crocin or ABX, and immunolabelled with PHP1 (green), anti-Elav antibody (red), and DAPI (blue). Part (C) shows the percent of pupal survival of Ex1-HD larvae treated with crocin or ABX. The data are represented as mean ± SEM, one-way ANOVA with Tukey’s post hoc test. n = 5 independent crosses for each condition, ***p < 0.001.

FIGURE 6

FL-HD female flies have lower abundance of Lactobacilli but elevated levels of Acetobacter. Microbial counts were determined by serial dilution plating of fly homogenates on MRS or Mannitol agar plates (section “Materials and Methods”). (A,B) Time course of the relative abundance of colony forming units (CFU) of Lactobacilli and Acetobacter, respectively. The data are represented as mean ± SEM, two-way ANOVA and Sidak’s multiple comparison Test. n = 4 agar plates for each condition, ^***p < 0.001.

FIGURE 7

Relative abundance of Lactobacilli strains and A. senegalensis in the gut of female HD flies by 16S rRNA gene sequence analysis. (A) Comparative analysis of Da-Gal4 flies with those expressing WT-HTT, (B) Da-Gal4 with FL-HD flies, and (C) WT-HTT with FL-HD. Parts (D–F) show the effects of crocin on the abundance of bacteria in Da-Gal4, WT-HTT, and FL-HD flies, respectively. All analyses were performed as described in the section “Materials and Methods” by Zymogen’s computational biologists. Lactobacillus-NA strain was not detected in the data base.