Probiotic Lactobacillus spp. improves Drosophila memory by increasing lactate dehydrogenase levels in the brain mushroom body neurons

Figures & data

Figure 1. Flies fed with Lactobacillus spp. GMNL-680 or GMNL-185 exhibit enhanced 3-minute olfactory memory.

Wild-type flies were individually reared on food containing different Lactobacillus spp. for five days, whereas the control group was reared on food without Lactobacillus spp. After five days, the flies were transferred to a T-maze for olfactory memory assays. During the assay, the flies were exposed to a CS+ odor (OCT or MCH) paired with 12 instances of 60-volt electric shock within 1 minute, and their 3-minute memory was tested immediately after the training. Notably, only flies that were fed Lactobacillus spp. L. rhamnosus #3 (termed GMNL-680) or L. acidophilus #1 (termed GMNL-185) showed a significant increase in the 3-minute memory. Each bar represents the mean ± SEM (N = 7); *p < 0.05, analyzed using the Student’s t-test.

Figure 2. Live GMNL-185 and GMNL-680 strains in the fly digestive tracts after five days of feeding.

(a) Confocal imaging of the fly digestive tract. The samples were immunostained with the anti-βPS-integrin antibody (gray). Scale bar = 500 μm. The fly digestive tract can be subdivided into distinct regions, including the foregut, midgut, and hindgut. Midgut can be further subdivided into R1–R5 regions. (b) RFP-tagged GMNL-185 or GFP-tagged GMNL-680 strains were cultured in Lactobacillus-supplemented fly food medium. Wild-type flies were fed with food medium containing RFP-tagged GMNL-185 or GFP-tagged GMNL-680 for five days, and their digestive tracts were dissected and mounted on coverslips. RFP and GFP fluorescent signals were detected in the crop, R1, R2, R3, R4, R5, and hindgut sub-regions of their digestive tracts. In contrast, the control group flies were fed with regular food without Lactobacillus spp. for five days. No RFP and GFP signals were observed in the digestive tracts of this group. Scale bar = 50 μm. (c) After five days of feeding with a medium containing RFP-tagged GMNL-185 or GFP-tagged GMNL-680, the fly digestive tracts were isolated and homogenized in PBS. The resulting supernatants were spread on MRS agar plates and incubated at 37°C for 2–3 days. Colonies of RFP-tagged GMNL-185 (c₁) and GFP-tagged GMNL-680 (c₃) were cultured on MRS agar plates. Individual colonies were isolated from the cultured MRS agar plates and placed on coverslips for confocal microscopy. The cultured bacterial colonies exhibited distinct RFP fluorescence (c₂) and GFP fluorescence (c₄) under confocal microscopy. Scale bar = 20 μm.

Figure 3. Flies with GMNL-185 and GMNL-680 mixed feeding showed synergistic effects on olfactory memory.

(a) In the regular training protocol (12 instances of 60-volt electric shock with CS+ odor for 1 minute), flies showed enhanced 3-minute memory performance in the GMNL-185 or GMNL-680 feeding group. The mixed feeding group (GMNL-185 + GMNL-680) did not exhibit significant differences from the GMNL-185 or GMNL-680 feeding groups. Each bar represents the mean ± SEM (N = 8), *p < 0.05; one-way ANOVA followed by Tukey’s test. (b) Flies showed enhanced 1-hour memory performance in GMNL-185 or GMNL-680 feeding groups. The mixed feeding group showed no significant difference from the GMNL-185 or GMNL-680 feeding groups. Each value represents the mean ± SEM (N = 9), *p < 0.05; one-way ANOVA followed by Tukey’s test. (c) Flies did not show enhanced 24-hour memory performance in GMNL-185 or GMNL-680 feeding groups as compared to the control group; however, enhanced 24-hour memory performance was observed in the mixed feeding group. Each bar represents the mean ± SEM (N = 9‒15), *p < 0.05; one-way ANOVA followed by Tukey’s test. (d) In the milder training protocol (three times 40-volt  electric shock plus CS+ odor for 1 minute), flies showed enhanced 3-minute memory performance in GMNL-185 or GMNL-680 feeding groups, while the mixed feeding group showed a synergistic effect. Each value represents the mean ± SEM (N = 13‒18), *p < 0.05; one-way ANOVA followed by Tukey’s test. (e) In milder training conditions, flies showed enhanced 1-hour memory performance in GMNL-185 or GMNL-680 feeding groups as compared to the control group. The mixed feeding group exhibited a synergistic effect. Each value represents the mean ± SEM (N = 15‒22), *p < 0.05; one-way ANOVA followed by Tukey’s test. (f) In milder training conditions, flies did not show a significant enhancement of 24-hour memory performance in GMNL-185, GMNL-680, or mixed feeding groups as compared to the control group. Each value represents the mean ± SEM (N = 8‒9). Non-significant difference (ns), p > 0.05; one-way ANOVA.

Figure 4. Flies with GMNL-185 and GMNL-680 mixed feeding showed increased calcium responses to training odor in the mushroom body β and γ lobes.

(a) The GCaMP6 responses to the training odor were assessed in distinct mushroom body lobes of GMNL-185 and GMNL-680 mixed feeding group flies. The calcium responses of GCaMP6 were measured 1 hour following a milder training protocol (consisting of 40 volts and 3 episodes of electric shocks within 1 minute) in each mushroom body lobe. The GMNL-185 and GMNL-680 mixed feeding group flies exhibited enhanced calcium responses to the training odor (CS+) compared to the non-training odor (CS−), with a notable increase observed in the β and γ lobes. This increase in calcium responses to the CS+ odor was not observed in the control group (without mixed GMNL-185/GMNL-680 feeding). (b) Quantification of the GCaMP6 responses to the CS+ odor relative to the CS− odor at 1 hour following a milder training protocol. The different regions of the mushroom body lobes were analyzed, respectively. The logarithmic ratios of the CS+ response to the CS− response were computed using the peak response amplitudes. Each value represents the mean ± SEM (N = 12‒20). *p < 0.05; non-significant differences (ns), p > 0.05, analyzed using Student’s t-test.

Figure 5. GMNL-185 and GMNL-680 mixed feeding flies showed an increased LDH level in the brain but not in the body.

(a) Scheme of enzymes involved in converting lactate into acetylcholine (b) Ldh mRNA levels were significantly increased in the heads of GMNL-185 and GMNL-680 mixed feeding group flies (Student t-test, *p < 0.05). Each value represents the mean ± SEM (N = 4). (c) Ldh mRNA levels were not significantly changed in the body of GMNL-185 and GMNL-680 mixed feeding group flies. Each value represents the mean ± SEM (N = 6‒8). (d) Quantification of LDH protein level in the whole fly brain after GMNL-185 and GMNL-680 mixed feeding. The brain structures were immunostained with anti-LDH antibody (magenta). Flies showed increased brain LDH expression in the GMNL-185 and GMNL-680 mixed feeding group (Student’s t-test, *p < 0.05). Each value represents the mean ± SEM (N = 8‒9). The fly brains were immunostained with (e) anti-LDH antibody, (f) anti-PDH antibody, and (g) anti-ChAT antibody (magenta). There were significant increased LDH, PDH, and ChAT immunostaining signals in mushroom body lobes in GMNL-185 and GMNL-680 mixed feeding flies as compared to the control flies. The intensity of immunostaining signals in mushroom body lobes from anti-LDH, anti-PDH, or anti-ChAT antibodies were normalized to the intensity of immunostaining signals in subesophageal ganglion (SOG) from the same antibody. Each value represents the mean ± SEM (N = 8‒10); Student’s t-test, *p < 0.05.

Figure 6. Pan-neuronal knockdown of Ldh impairs GMNL-185/GMNL-680 mixed-feeding-induced memory enhancement.

All flies were fed a mixture of GMNL-185 and GMNL-680 for five days. Elav-GAL4 > UAS-Ldh^RNAi flies exhibited impaired performance in 3-minute memory and 1-hour memory compared to Elav-GAL4 > + and UAS-Ldh^RNAi > + flies. Elav-GAL4 > UAS-Ldh^RNAi flies. Elav-GAL4 > UAS-Ldh^RNAi flies did not exhibit impaired performance in 24-hour memory compared to Elav-GAL4 > + and UAS-Ldh^RNAi > + flies. Each value represents the mean ± SEM (N = 8). *p < 0.05; non-significant differences (ns), p > 0.05, analyzed using one-way ANOVA followed by Tukey’s test.

Figure 7. Ldh knockdown specifically in the mushroom body impairs GMNL-185/GMNL-680 mixed-feeding-induced memory enhancement.

All flies were fed with a mixture of GMNL-185 and GMNL-680 for five days before the behavioral assays. (a) R13F02-GAL4 > UAS-Ldh^RNAi flies exhibited disrupted performance in the 3-minute memory (N = 13) and 1-hour memory (N = 11) tests compared to R13F02-GAL4 > + and UAS-Ldh^RNAi > + flies. Moreover, 24-hour memory was not affected in R13F02-GAL4 > UAS-Ldh^RNAi flies compared to that in R13F02-GAL4 > + and UAS-Ldh^RNAi > + flies (N = 8). Each value represents the mean ± SEM. *p < 0.05; one-way ANOVA followed by Tukey’s test. (b) Adult stage-specific knockdown of Ldh in the mushroom body disrupts memory. Acute knockdown of Ldh expression in the entire mushroom body neurons using R13F02-GAL4 impaired the 3-minute memory. Adult flies were reared at 18°C and then shifted to 30°C for seven days before training to recover from tub-GAL80^ts inhibition of the GAL4 activity. Each value represents the mean ± SEM (N = 6), *p < 0.05; non-significant difference (ns), p > 0.05, determined using one-way ANOVA followed by Tukey’s test.

Figure 8. Ldh knockdown abolished the GMNL-185/GMNL-680 mixed-feeding-induced increased responses to training odor in the mushroom body β and γ lobes.

(a) All flies were fed with GMNL-185/GMNL-680 for five days before behavioral assays. The calcium responses of GCaMP6 were measured 1 hour following a milder training protocol in each mushroom body lobe. Recordings were made from optical sections in regions of the horizontal (β’, β, and γ lobes) of the mushroom body. (b) Flies bearing UAS-GCaMP6m/+; R13F02-GAL4/+ showed significantly increased calcium responses to CS+ odor in the mushroom body β and γ lobes after milder training protocol, but no changes were observed in flies carrying UAS-GCaMP6m/+; R13F02-GAL4/UAS-Ldh^RNAi. (c) Quantification of the GCaMP6 responses to the CS+ odor relative to the CS− odor at 1 hour following a milder training protocol. The different regions of the mushroom body lobes were analyzed, respectively. The logarithmic ratios of the CS+ response to the CS− response were computed using the peak response amplitudes. Each value represents the mean ± SEM (N = 7‒9). *p < 0.05; non-significant difference (ns), p > 0.05, determined using Student’s t-test.

Figure 9. Proposed model showing the effect of GMNL-185/GMNL-680 mixed feeding on Drosophila memory.

Flies that were fed with a mixture of GMNL-185 and GMNL-680 exhibited elevated Ldh levels within the brain and mushroom body neurons, along with increased calcium responses in the mushroom body β and γ lobes. Consequently, flies exhibited enhanced olfactory memory after GMNL-185 and GMNL-680 feeding, which was abolished by silencing Ldh expression in the mushroom body. Our results demonstrate potent probiotic effects of the two Lactobacillus spp. which improves the memory performance in flies through the gut-brain axis and remarkably enhances neuronal LDH levels in the brain.

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article and its Supplementary Material. The datasets generated for this study can be found in the https://reurl.cc/G4ANrG