Recognition of commensal bacterial peptidoglycans defines Drosophila gut homeostasis and lifespan

Abstract

Commensal microbes in animals have a profound impact on tissue homeostasis, stress resistance, and ageing. We previously showed in Drosophila melanogaster that Acetobacter persici is a member of the gut microbiota that promotes ageing and shortens fly lifespan. However, the molecular mechanism by which this specific bacterial species changes lifespan and physiology remains unclear. The difficulty in studying longevity using gnotobiotic flies is the high risk of contamination during ageing. To overcome this technical challenge, we used a bacteria-conditioned diet enriched with bacterial products and cell wall components. Here, we demonstrate that an A. persici-conditioned diet shortens lifespan and increases intestinal stem cell (ISC) proliferation. Feeding adult flies a diet conditioned with A. persici, but not with Lactiplantibacillus plantarum, can decrease lifespan but increase resistance to paraquat or oral infection of Pseudomonas entomophila, indicating that the bacterium alters the trade-off between lifespan and host defence. A transcriptomic analysis using fly intestine revealed that A. persici preferably induces antimicrobial peptides (AMPs), while L. plantarum upregulates amidase peptidoglycan recognition proteins (PGRPs). The specific induction of these Imd target genes by peptidoglycans from two bacterial species is due to the stimulation of the receptor PGRP-LC in the anterior midgut for AMPs or PGRP-LE from the posterior midgut for amidase PGRPs. Heat-killed A. persici also shortens lifespan and increases ISC proliferation via PGRP-LC, but it is not sufficient to alter the stress resistance. Our study emphasizes the significance of peptidoglycan specificity in determining the gut bacterial impact on healthspan. It also unveils the postbiotic effect of specific gut bacterial species, which turns flies into a "live fast, die young" lifestyle.

Fig 1. A. persici-conditioned diet shortens lifespan.

(A) An overview of preparing the “bacteria conditioned diet (BacD)”. (B) Colony forming units (CFU) of BacD (before addition of antibiotics). For the statistics, one-way ANOVA with Holm-Šídák’s multiple comparison was used. (C) The experimental scheme of the ageing experiments using BacD. (D)(E) Lifespan of female w^Dah (D) and Canton-S (E) flies with the BacD. A log-rank test was used to compare between control (Ctrl) and each BacD. (F) Phospho-histone H3-positive cell numbers in the midgut of Canton-S female flies fed a conventional diet (left, Day 10 vs Day 41) or 25 days of BacD (right, Day 30). For the statistics, unpaired two-tailed Student’s t test (left figure) or one-way ANOVA with Holm-Šídák’s multiple comparisons (right figure) was used. (G) The climbing ability of Canton-S male flies with either a conventional diet (Left, Day 9 vs Day 22 or Day 41) or BacD (Center, Day 6 or Right, Day 35) assessed by the negative geotaxis assay. For the statistics, unpaired two-tailed Student’s t test (left figure) or one-way ANOVA with Holm-Šídák’s multiple comparison (right figures) was used. The control diet followed the same procedure for BacD but it has only MRS broth in place of bacterial isolates, resulting in the antibiotics-contained diet. Sample sizes (n) and P values are in each figure.

Fig 2. A. persici-conditioned diet enhances resistance to paraquat and Pe oral infection but susceptibility to starvation.

(A)(B) Survival curve of female Canton-S (A) and w^Dah (B) flies during 10 mM paraquat feeding after 5 days of a bacteria-conditioned diet (BacD). (C)(D) Survival curve of female Canton-S (C) and w^Dah (D) flies during starvation stress after 5 days of BacD. (E) Survival curve of female Canton-S flies orally infected with Pseudomonas entomophila (Pe) after 5 days of BacD. A log-rank test was used to compare between control (Ctrl) and each BacD. (F) Food consumption of female flies by CAFE assay. Flies were treated with BacD for five days. Each point shows data of 7 or 8 flies. For the statistics, One-way ANOVA with Holm-Šídák’s multiple comparison was used. The control diet followed the same procedure for BacD but it has only MRS broth in place of bacterial isolates, resulting in the antibiotics-contained diet. Sample sizes (n) and P values are in each figure.

Fig 3. Transcriptional regulation of distinct Imd target genes by bacterial species.

(A) The experimental overview of RNA-seq analysis of female w^Dah gut after 24 hours of BacD. (B) RNA-seq read count data of antimicrobial peptide (AMP) genes and Imd negative regulator genes in w^Dah female fly guts after 24 hours of BacD shown in fold change. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. (C)(D) Quantitative RT–PCR of DptA and PGRP-SC1a in w^Dah (C) and Canton-S (D) female fly guts after 24 hours of BacD. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. (E) Quantitative RT–PCR of DptA and PGRP-SC1a in yw and yw-backcrossed Dredd^B118 female fly guts after 24 hours of BacD. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. The control diet followed the same procedure for BacD but it has only MRS broth in place of bacterial isolates, resulting in the antibiotics-contained diet. Sample sizes (n) and P values are in each figure.

Fig 4. Differential expression of Imd target genes in the anterior and posterior midgut.

(A) Quantitative RT–PCR of DptA and PGRP-SC1a (A) in female w^Dah anterior and posterior gut after 24 hours of BacD. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. (B) Quantitative RT–PCR of caudal in female Canton-S anterior and posterior gut after 24 hours of BacD. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. (C) Quantitative RT–PCR of DptA and PGRP-SC1a in the gut of female 5966^GS > caudal-RNAi flies after 24 hours of BacD. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. The control diet followed the same procedure for BacD but it has only MRS broth in place of bacterial isolates, resulted in the antibiotics-contained diet. Sample sizes (n) and P values are in each figure.

Fig 5. Distinct receptor PGRPs regulate different Imd target genes in gut.

(A)(B) Quantitative RT–PCR of the genes DptA (A) and PGRP-SC1a (B) in female w^Dah, Rel^E20, PGRP-LC^E12, and PGRP-LE¹¹² fly guts after 24 hours of BacD. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. The control diet followed the same procedure for BacD but it has only MRS broth in place of bacterial isolates, resulted in the antibiotics-contained diet. Sample sizes (n) and P values are in each figure.

Fig 6. Heat-killed A. persici shortens lifespan and increases the host defence but does not enhance paraquat resistance.

(A) An overview of the method to prepare the heat-killed bacteria (HK) diet. (B) Lifespan of female Canton-S flies fed the HK diet. A log-rank test was used to compare between control (Ctrl) and each HK diet. (C) Phospho-histone H3-positive cell numbers in the midgut of Canton-S female flies fed the HK diet for 40 days (Day45). For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. (D) Quantitative RT–PCR of the genes DptA and PGRP-SC1a in the female Canton-S fly gut after 24 hours of HK diet. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. (E)-(H) Resistance to oral infection with Pseudomonas entomophila (Pe) in female and male Canton-S flies fed the HK diet for five days (E, F), and resistance to 10 mM paraquat (G), or starvation (H) in female Canton-S flies with HK diet for five days in female Canton-S flies. A log-rank test was used to compare between control (Ctrl) and each HK diet. The control diet is the antibiotics-contained diet. Sample sizes (n) and P values are in each figure.

Fig 7. Feeding L. plantarum Lsi and A. persici Ai purified peptidoglycans.

(A) Quantitative RT–PCR of the genes DptA and PGRP-SC1a in female fly Canton-S guts after 24 hours of purified peptidoglycan diet with the same concentration (measured and calculated by OD₂₅₄). For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. (B) Quantitative RT–PCR of the genes DptA and PGRP-SC1a in female fly Canton-S guts after 24 hours of A.persici Ai purified peptidoglycan (1 or 100 times dilution) diet with (measured and calculated by OD₂₅₄). For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. The control diet is the antibiotics-contained diet. Sample sizes (n) and P values are in each figure.

Fig 8. PGRP-LC in the gut promotes ageing by A. persici.

(A) Phospho-histone H3-positive cell numbers in the midgut of 4-week-old female flies with BacD. The flies used were Su(H)GBE-LacZ; esg-Gal4, tub-Gal80^ts, UAS-GFP > LacZ-RNAi, Rel-RNAi, PGRP-LC-RNAi, and PGRP-LE-RNAi, and the flies were reared at 30°C from their adult day2-4. Unpaired two-tailed Student’s t test was used. (B) Quantitative RT–PCR of DptA in female 5966^GS > Rel-RNAi fly gut after 24 hours of BacD with ethanol or 10 μM RU. For the statistics, ANOVA with Holm-Šídák’s multiple comparison was used. (C) Lifespan of female 5966^GS > Rel-RNAi flies treated with BacD and ethanol or 10 μM RU. A log-rank test was used to compare between control (Ctrl) and BacD with ethanol or 10μM RU. (D) Lifespan of female 5966^GS > PGRP-LC-RNAi flies with heat-killed (HK) diet and ethanol or 10 μM RU. A log-rank test was used to compare between control (Ctrl) and HK diet with ethanol or 10μM RU. (E) Lifespan of female esg^ts > w^CS and esg^ts > PGRP-LC-RNAi flies with HK diet. Flies were reared at 30°C during their adult period. A log-rank test was used to compare between control (Ctrl) and HK diet in each genotype. The 50% survival rate is compared in (D) and (E). The control diet in (A-C) followed the same procedure for BacD but it has only MRS broth in place of bacterial isolates, resulting in the antibiotics-contained diet. The control diet in (D) and (E) is the antibiotics-contained diet. Sample sizes (n) and P values are in each figure.

Fig 9. Graphical summary of this study.