Gut microbiome species Levilactobacillus brevis regulates reproductive fitness in C. elegans.

N.J.B. and D.E.M. conceived and designed the experiments. N.J.B. performed the experiments and analyzed the data. N.J.B. and D.E.M. wrote the paper.

Since C. elegans are able to feed on singular bacterial species, we took advantage of the ability to establish cause-and-effect relationships between individual gut microbiota and host health. To understand how the human gut microbiome species L. brevis impacts C. elegans reproduction, non-transgenic worms were first allowed to hatch and develop on the standard laboratory food source of Escherichia coli (strain OP50). During the final developmental stage (the L4 stage), the worms were moved to new plates seeded with either E. coli, or L. brevis derived from a human fecal isolate. This timeline was chosen in order to avoid potential effects on worm development. On day 4 of adulthood, L. brevis -fed worms exhibited a striking increase in egg retention compared to E. coli -fed worms ( Fig 1 A-B). Specifically, L. brevis -fed worms retained significantly more eggs inside the uterus than E. coli -fed worms with an average of 25 eggs for L. brevis compared to 11 eggs for E. coli , representing over a 120 % increase in egg retention ( Fig 1 C). Consistent with this result, we observed that over the first four days of adulthood, L. brevis -fed worms had significantly smaller total brood size and laid significantly fewer eggs than E. coli -fed controls (Supp Fig 1 ). We next conducted a thorough examination of the morphological characteristics of the eggs retained within the uterus of day 4 L. brevis -fed worms. We found that L. brevis worsens both the organization and the appearance of the retained eggs, as evidenced by higher incidences of disorganized (non-stacked), irregularly shaped (non-oval), and abnormally small eggs compared to those of E. coli -fed worms ( Fig 1 D). Fig. 1 L. brevis -fed worms retain their eggs. (A) Representative images of eggs in non-transgenic N2 worms fed either E. coli (top) or L. brevis (bottom); white arrowhead depicts unfertilized oocytes, white arrows depict fertilized embryos (eggs), scale bar = 50 µm. (B) Scoring images for normal, mild, or severely abnormal number of eggs shows that L. brevis -fed worms have severe egg retention at day 4 of adulthood ( n = 16 per group). Chi-squared test. (C) Quantification using the egg-in-worm assay shows that L. brevis -fed worms ( n = 56) have significantly more retained eggs than E.coli -fed worms ( n = 60) on day 4 of adulthood. Unpaired t -test. (D) Scored morphologies of the retained eggs show significantly worsened egg organization and appearance in L. brevis -fed worms on day 4 of adulthood compared to E. coli -fed worms. Chi-squared test. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. Fig. 1 L. brevis -fed worms retain their eggs. (A) Representative images of eggs in non-transgenic N2 worms fed either E. coli (top) or L. brevis (bottom); white arrowhead depicts unfertilized oocytes, white arrows depict fertilized embryos (eggs), scale bar = 50 µm. (B) Scoring images for normal, mild, or severely abnormal number of eggs shows that L. brevis -fed worms have severe egg retention at day 4 of adulthood ( n = 16 per group). Chi-squared test. (C) Quantification using the egg-in-worm assay shows that L. brevis -fed worms ( n = 56) have significantly more retained eggs than E.coli -fed worms ( n = 60) on day 4 of adulthood. Unpaired t -test. (D) Scored morphologies of the retained eggs show significantly worsened egg organization and appearance in L. brevis -fed worms on day 4 of adulthood compared to E. coli -fed worms. Chi-squared test. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. To validate that the observed reproductive phenotype is not due to starvation in the event of L. brevis being a non-preferred food source, non-transgenic worms were raised in the absence of food (starved) and directly compared to both L. brevis -fed worms and E. coli -fed worms at day 1 of adulthood ( Fig 2 ). Oil Red O (ORO) staining is an established and reliable method for measuring neutral lipid stores in response to dietary deprivation ( An et al., 2023 ; Buis et al., 2019 ; Escorcia et al., 2018 ; Stuhr et al., 2022 ). Consistent with prior findings showing rapid depletion of fat stores upon bacterial deprivation ( Buis et al., 2019 ; Li et al., 2020 ), the starved worms displayed markedly and significantly reduced ORO staining compared to both E. coli and L. brevis -fed worms, with a 101 % decrease relative to E. coli and a 78 % decrease relative to L. brevis ( Fig 2 A-B). In contrast, worms fed L. brevis exhibited an intermediate phenotype in which ORO staining was reduced by 29 % compared to E. coli -fed worms, suggesting L. brevis provides lower nutritional value and leads to less lipid accumulation ( Fig 2 B). Overall, however, these data indicate that exposure to L. brevis is not equivalent to complete food deprivation ( Fig 2 B). Consistent with this, worms readily ingested L. brevis , which was visually detected as individual rod-shaped bacterial cells stained with acridine orange within the worm’s buccal cavity, and intestinal colonization assays confirmed the presence of L. brevis within the digestive tract ( Fig 2 C-D). Fig. 2 L. brevis -fed worms are not starved. (A) Representative images of Oil Red O stained lipid stores in day 1 adult non-transgenic N2 worms fed either E. coli ( n = 23), L. brevis ( n = 36), or Starved ( n = 25), scale bar = 270 µm. (B) Quantification of ORO pigment intensity normalized to worm area. One-way ANOVA with Tukey’s post hoc test, A.U., arbitrary units. (C) C. elegans ingest L. brevis as visualized with acridine orange; arrow depicts bacterial cells in the buccal cavity of the worm. Scale bar = 5 µm. (D) Intestinal colonization of the digestive tract by L. brevis determined by Colony Forming Units (CFU) per worm at day 1 of adulthood ( n = 35). (E) Proportion of daf-16::GFP localization in day 1 adult worms fed either E. coli, L. brevis , or starved. Chi-square test. (F) Representative images of daf-16::GFP diffuse in the cytoplasm of E. coli ( n = 56) and L. brevis -fed ( n = 54) worms versus translocated to the nucleus upon starvation ( n = 56), scale bar = 100 µm. (G) No significant difference in lifespan between E. coli -fed ( n = 245) and L. brevis -fed ( n = 96) worms. Log-rank Mantel-Cox survival curve analysis. ns, not significant. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. Fig. 2 L. brevis -fed worms are not starved. (A) Representative images of Oil Red O stained lipid stores in day 1 adult non-transgenic N2 worms fed either E. coli ( n = 23), L. brevis ( n = 36), or Starved ( n = 25), scale bar = 270 µm. (B) Quantification of ORO pigment intensity normalized to worm area. One-way ANOVA with Tukey’s post hoc test, A.U., arbitrary units. (C) C. elegans ingest L. brevis as visualized with acridine orange; arrow depicts bacterial cells in the buccal cavity of the worm. Scale bar = 5 µm. (D) Intestinal colonization of the digestive tract by L. brevis determined by Colony Forming Units (CFU) per worm at day 1 of adulthood ( n = 35). (E) Proportion of daf-16::GFP localization in day 1 adult worms fed either E. coli, L. brevis , or starved. Chi-square test. (F) Representative images of daf-16::GFP diffuse in the cytoplasm of E. coli ( n = 56) and L. brevis -fed ( n = 54) worms versus translocated to the nucleus upon starvation ( n = 56), scale bar = 100 µm. (G) No significant difference in lifespan between E. coli -fed ( n = 245) and L. brevis -fed ( n = 96) worms. Log-rank Mantel-Cox survival curve analysis. ns, not significant. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. To further determine if L. brevis -fed worms are nutritionally compromised, we utilized a transgenic strain expressing a GFP reporter for the nutrient-sensitive insulin/IGF-1 signaling pathway, in which the daf-16/FOXO transcription factor is known to undergo translocation to the nucleus in response to nutrient deprivation or starvation, among other stressors ( Henderson and Johnson, 2001 ; Henderson et al., 2006 ; Weinkove et al., 2006 ). We found that daf-16 localization was predominantly cytoplasmic in L. brevis -fed worms which showed no significant differences from well-fed E. coli controls, whereas starved worms showed a robust nuclear localization of daf-16 consistent with activation of a starvation stress response ( Fig 2 E-F). Additionally, the observation that L. brevis does not alter lifespan compared with E. coli suggests there is no lifespan extension that is a well-known consequence of caloric restriction in C. elegans ( Buis et al., 2019 ; Hahm et al., 2019 ; Lee et al., 2006 ; Serna et al., 2020 ) ( Fig 2 G). While L. brevis may represent a low-quality bacterial food source, it is nevertheless sufficient to sustain C. elegans without eliciting a canonical starvation response. Moreover, we found that the abnormal egg morphologies associated with the egg retention phenotype in L. brevis -fed worms ( Fig 1 D) were not recapitulated by a lack of food, further confirming that the observed reproductive deficits are not a result of starvation (Supp Fig. 2 ). Egg-laying behavior in C. elegans is controlled by a relatively simple circuit consisting of six cholinergic Ventral Cord neurons (VCs) and two serotonergic Hermaphrodite Specific Neurons (HSNs), which synapse onto the surrounding vulval muscles whose contraction results in the expulsion of eggs ( Fig 3 A) ( Brewer et al., 2019 ; Collins et al., 2016 ; Kopchock et al., 2021 ). HSN activity and the release of serotonin excites the vulval muscles and the cholinergic VC neurons, with the VC activity directly linked to vulval muscle contractions ( Brewer et al., 2019 ; Collins et al., 2016 ; Kopchock et al., 2021 ). In order to investigate whether the egg retention phenotype induced by L. brevis may be caused by defects in specific components of the egg-laying circuit, worms were exposed to serotonin or levamisole, which are pharmacological agents previously shown to stimulate egg-laying ( Bany et al. 2003 ; Trent et al., 1983 ). On day 4 of adulthood, L. brevis -fed worms exposed to exogenous serotonin were able to lay eggs with no significant difference from the E. coli -fed controls ( Fig 3 B), indicating that the vulval muscle itself is intact and able to contract leading to successful egg-laying. qPCR analysis revealed no significant changes in the expression of the vulval muscle nicotinic receptor unc-63, or rsu-1, a vulval muscle gene required for functional egg-laying, further suggesting that the vulva is preserved in L. brevis -fed worms (Supp Fig 3 ). Since the application of exogenous serotonin bypasses the activity of the HSNs, these data offer the possibility that a lack of endogenous serotonin signaling from the HSNs may be responsible for the egg retention phenotype. Fig. 3 HSN structural and functional defects in worms fed L. brevis . (A) Simplified schematic of the C. elegans egg-laying circuitry. HSN, hermaphrodite-specific neurons. VC, Ventral Cord neurons. VM, vulval muscle. (B) Eggs laid in response to exogenous serotonin (5HT) or M9 treatment of day 4 adults ( n = 30 per group). Two-way ANOVA, data are mean ± SEM. ns, not significant. (C) Eggs laid in response to levamisole or M9 treatment of day 4 adults ( n = 30 per group). Two-way ANOVA, data are mean ± SEM. ns, not significant. (D) Representative images of the HSN in day 4 adults raised on E. coli (left, n = 10) or L. brevis (right, n = 10), with arrow depicting the cell body. Scale bar = 50 µm . (E-F) Quantification of HSN cell body area (E) and fluorescence intensity (F) . Unpaired t -tests, data are mean ± SEM. A.U., arbitrary units. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. Fig. 3 HSN structural and functional defects in worms fed L. brevis . (A) Simplified schematic of the C. elegans egg-laying circuitry. HSN, hermaphrodite-specific neurons. VC, Ventral Cord neurons. VM, vulval muscle. (B) Eggs laid in response to exogenous serotonin (5HT) or M9 treatment of day 4 adults ( n = 30 per group). Two-way ANOVA, data are mean ± SEM. ns, not significant. (C) Eggs laid in response to levamisole or M9 treatment of day 4 adults ( n = 30 per group). Two-way ANOVA, data are mean ± SEM. ns, not significant. (D) Representative images of the HSN in day 4 adults raised on E. coli (left, n = 10) or L. brevis (right, n = 10), with arrow depicting the cell body. Scale bar = 50 µm . (E-F) Quantification of HSN cell body area (E) and fluorescence intensity (F) . Unpaired t -tests, data are mean ± SEM. A.U., arbitrary units. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. To further test the function of the HSNs, we used the acetylcholine receptor agonist, levamisole, to stimulate egg-laying and found that in this case L. brevis -fed worms did not show enhanced egg-laying, in contrast to E. coli -fed worms ( Fig 3 C). Previous studies have suggested that serotonin from the HSN is essential to enhance VC and vulval muscle activity and to coordinate the process of egg-laying. In the absence of HSN-potentiation, the VC neurons alone are unable to stimulate the vulval muscles to the threshold required for laying eggs. Since the application of exogenous levamisole, effectively bypassing VC activity, was insufficient to enhance egg-laying in L. brevis -fed worms, this further supports that HSN neuronal activity is impaired. Together our findings indicate that L. brevis impacts the function of serotonin signaling by the HSNs, likely contributing to the egg retention phenotype observed in L. brevis -fed worms. This was further corroborated with fluorescence imaging of the HSNs, which revealed structural defects in the L. brevis -fed worms compared to E. coli controls at day 4 of adulthood ( Fig 3 D). Examination of the HSNs showed significantly smaller cell bodies in L. brevis -fed worms compared to E. coli controls, in conjunction with decreased HSN fluorescence intensity ( Fig 3 E-F). Interestingly, expression of the tryptophan hydroxylase tph-1 and a downstream effector of serotonin signaling, egl-19 , both remained unchanged by L. brevis, suggesting serotonin biosynthesis and calcium channel activity are intact (Supp Fig 3 ). These results suggest that impaired HSN signaling is driven by mechanisms other than expression levels of key genes in the egg-laying circuitry. Since the reproductive defect in L. brevis -fed worms not only involves pronounced egg retention and alterations in the egg-laying circuitry, but is also associated with severely abnormal egg morphologies ( Fig 1 D), we next asked if the health of the progeny is reduced. To assess the fitness of the progeny of hermaphrodite ‘mothers’ raised on L. brevis, we performed viability assays throughout the first four days of adulthood when C. elegans are most reproductively active ( Fig 4 A-D). Across days 1–4, the total number of live progeny per parent worm was significantly reduced in L. brevis -fed worms compared with E. coli -fed controls ( Fig 4 A). Throughout the experimental period, L. brevis -fed worms produced significantly fewer live (hatched) progeny on each day from days 1–3 ( Fig 4 B), as well as significantly greater unhatched embryos ( Fig 4 C) on each day from days 1–4, compared to E. coli -fed controls. Next, embryonic viability was assessed as the number of live progeny divided by the total progeny (live and unhatched) across the experimental period. E. coli -fed controls yielded an embryonic viability of 98 % while L. brevis -fed worms showed a substantial reduction to only 63 % ( Fig 4 D). These findings show that, in combination with the lower total number of eggs laid by L. brevis -fed worms (Supp Fig 1 ), overall L. brevis causes worms to produce fewer live progeny, and for a greater proportion of the eggs they lay to be unable to hatch. Fig. 4 Progeny of worms raised on L. brevis have hatching and developmental defects. (A) Total number of live progeny produced from the first 4 days of adulthood is significantly reduced in L. brevis -fed worms ( n = 29) compared with E. coli controls ( n = 30). Unpaired t -test, data are mean ± SEM. (B-C) Progeny viability assay over the first 4 days of adulthood, revealing that L. brevis -fed worms ( n = 29) have significantly fewer live progeny produced each day (B), and significantly more unhatched embryos produced each day (C) than E. coli controls ( n = 30). (D) Average embryonic viability was substantially decreased in the L. brevis -fed population ( n = 29) compared to E. coli ( n = 30) on day 4 of adulthood. Unpaired t -test, data are mean ± SEM. (E) Number of eggs retained after picking day 4 L. brevis -fed ( n = 35) or E. coli -fed ( n = 40) worms into a drop of bleach on E. coli -seeded plates. Unpaired t -test, data are mean ± SEM. (F-G) Percent of eggs hatched (F) and developed to L4 stage (G) after bleaching day 4 L. brevis- or E. coli -fed worms onto E. coli -seeded plates. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. Fig. 4 Progeny of worms raised on L. brevis have hatching and developmental defects. (A) Total number of live progeny produced from the first 4 days of adulthood is significantly reduced in L. brevis -fed worms ( n = 29) compared with E. coli controls ( n = 30). Unpaired t -test, data are mean ± SEM. (B-C) Progeny viability assay over the first 4 days of adulthood, revealing that L. brevis -fed worms ( n = 29) have significantly fewer live progeny produced each day (B), and significantly more unhatched embryos produced each day (C) than E. coli controls ( n = 30). (D) Average embryonic viability was substantially decreased in the L. brevis -fed population ( n = 29) compared to E. coli ( n = 30) on day 4 of adulthood. Unpaired t -test, data are mean ± SEM. (E) Number of eggs retained after picking day 4 L. brevis -fed ( n = 35) or E. coli -fed ( n = 40) worms into a drop of bleach on E. coli -seeded plates. Unpaired t -test, data are mean ± SEM. (F-G) Percent of eggs hatched (F) and developed to L4 stage (G) after bleaching day 4 L. brevis- or E. coli -fed worms onto E. coli -seeded plates. * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001. We further asked if the decline in progeny viability was due to the eggs being laid in the presence of L. brevis, possibly contributing to an unfavorable environment for hatching. To address this, C. elegans were raised as described previously to day 4 of adulthood before being picked into a drop of hypochlorite solution on an E. coli -seeded plate to “collect” the retained eggs. This allowed the eggs to hatch in a known favorable environment. After bleaching, L. brevis -fed worms retained 171 eggs compared to only 3 from E. coli -fed worms that were left to hatch on the E. coli -seeded plates ( Fig 4 E). 24 h later, 82 % of the eggs from parents raised on E. coli had hatched, while only 11 % of the eggs from parents raised on L. brevis had hatched ( Fig 4 F), suggesting that the L. brevis -induced decline in progeny viability is due to intrinsic factors of the eggs rather than the environment in which they are laid. Finally, we assessed the developmental rate of the progeny that had hatched on E. coli -seeded plates. At 20 °C, synchronized C. elegans will develop to the L4 larval stage within approximately 48 h. At 48 h after transferring the eggs to the E. coli -seeded plates by bleaching, 100 % of the progeny from parents raised on E. coli developed to the L4 stage while only 68 % of the L. brevis progeny developed to L4, indicating a developmental delay. Collectively, these data indicate that L. brevis -fed mothers have significantly reduced egg quality, resulting in a greater rate of embryonic lethality and delayed development of the progeny that do hatch.

C. elegans strains were maintained at 20 °C under standard conditions ( Brenner, 1974 ). Hypochlorite alkaline-bleaching solution was used to synchronize worms for experiments ( Brenner, 1974 ). Briefly, synchronization was performed by exposing gravid hermaphrodites to the hypochlorite solution to collect eggs, followed by thorough washing of the collected eggs in M9 buffer and dispensing onto high growth medium (HG) plates seeded with OP50 E. coli . For experiments, standard nematode growth medium (NGM) plates were seeded with 1 mL of either E. coli in LB medium or L. brevis in MRS medium, for ad libitum feeding. The following strains were used: wild-type N2 Bristol strain, LX975 vsls13 [ lin-11::pes-10::GFP + lin-15(+) ]; vsIs97 [ tph-1p::DsRed2 + lin-15(+) ]; vsIs100 [ myo-3p::CFP + lin-15(+) ], and MQD1543 [ daf-16(hq23[daf-16::GFP ]) I]. The L. brevis strain used in experiments was obtained from Microbiologics (Catalog No. 01262 L) originally derived from ATCC 14869 isolated from human fecal samples. For these experiments, L. brevis was cultivated aerobically in deMan, Rogosa, and Sharpe (MRS) medium at 35 °C for 24–48 h and stored at 4 °C. OP50 E. coli was cultivated aerobically at room temperature overnight as per standard protocols ( Brenner, 1974 ). Day 4 adult hermaphrodites were placed individually into 10μL of hypochlorite solution and eggs were counted after the body of the worm dissolved (roughly 10 min). Protocol was followed as described previously ( Gardner et al., 2013 ). Three biological replicates were completed with 20 worms per group. Worms were mounted on 2 % agarose pads with M9 and sodium azide and imaged on a Leica M165 FC fluorescent stereomicroscope with a Leica K5 microscope camera. Brightfield images were used to characterize the morphology of eggs retained within the uterus. Egg morphology was evaluated using a categorical scoring system adapted from a previous characterization of oocyte morphology by Templeman et a (2020) . Four phenotypic categories were assessed: Retention, Stacking, Shape, and Size. For each category, worms were assigned a score of normal, mild, or severe according to predefined criteria: • Retention: Normal = 0–15 eggs at mixed developmental stages; Mild = 16–20 eggs with some fully developed embryos; Severe = >20 retained eggs with nearly all fully developed embryos or evidence of matricidal hatching. • Stacking: Normal = eggs arranged in an orderly, stacked orientation; Mild = occasional overlap or misalignment of eggs; Severe = disorganized arrangement with crowding throughout the uterus. • Shape: Normal = eggs uniformly oval with no detectable abnormalities; Mild = ≤25 % of eggs exhibiting irregular or slightly deformed morphology; Severe = >25 % of eggs exhibiting irregular or damaged morphology. • Size: Normal = no apparent size irregularities; Mild = ≤25 % of eggs visibly smaller than average; Severe = >25 % of eggs noticeably reduced in size compared to the cohort. Retention: Normal = 0–15 eggs at mixed developmental stages; Mild = 16–20 eggs with some fully developed embryos; Severe = >20 retained eggs with nearly all fully developed embryos or evidence of matricidal hatching. Stacking: Normal = eggs arranged in an orderly, stacked orientation; Mild = occasional overlap or misalignment of eggs; Severe = disorganized arrangement with crowding throughout the uterus. Shape: Normal = eggs uniformly oval with no detectable abnormalities; Mild = ≤25 % of eggs exhibiting irregular or slightly deformed morphology; Severe = >25 % of eggs exhibiting irregular or damaged morphology. Size: Normal = no apparent size irregularities; Mild = ≤25 % of eggs visibly smaller than average; Severe = >25 % of eggs noticeably reduced in size compared to the cohort. To minimize bias, all scoring was performed under the same magnification with consistent lighting conditions, and worms were evaluated in random order by a single trained experimenter to ensure consistency across replicates. Two biological replicates were analyzed. Worms were mounted on 2 % agarose pads with M9 and sodium azide and imaged on a Zeiss Axioplan 2 Fluorescence Microscope with differential interference contrast (DIC) at 10x magnification. Three biological replicates were completed with 8–10 worms per group. This assay was used to track egg-laying and progeny-hatching in adult worms over the first four days of adulthood. On Day 0, synchronized L4 stage worms were individually plated onto 35 mm NGM plates seeded with either E. coli or L. brevis . The assay involved moving individually plated worms to new plates daily and counting the number of eggs laid, live hatched progeny, and unhatched embryos until the initial worm reached day 4 of adulthood. Worms were censored due to matricide (progeny hatched inside the hermaphrodite parent), vulval rupture (intestinal expulsion), or if lost. Censored animals were not included in the data. The average number of live progeny per worm was calculated by summing the number of progeny produced and dividing by the number of parent worms. Embryonic viability was calculated as the number of live progeny divided by the total progeny (live and unhatched) across the experimental period. Brood size was calculated the same as the average of live progeny with the inclusion of unhatched embryos to account for total brood size produced by the parent worms per biological replicate. Modified from ( Kwah and Jaramillo-Lambert, 2023 ). Embryonic Viability = [ # live progeny # live progeny + # unhatched embryos ] × 100 Average # live progeny = ∑ ( # live progeny ) # parent hermaphrodites Average Brood Size = ∑ ( # live progeny + # unhatched embryos ) # parent hermaphrodites For lifespan assays, as previously outlined ( Kenyon et al., 1993 ), groups of 12–15 L4 stage hypochlorite-synchronized worms were initially placed on individual NGM plates, seeded with either E. coli or L. brevis . Worms were transferred to freshly seeded plates every two days while producing progeny, and every 3 days thereafter. Worms were considered dead when they no longer responded to touch stimuli and were censored in the event of matricide (progeny hatched inside the hermaphrodite parent), vulval rupture (intestinal expulsion), or if lost. Censored animals were not included in the data. Three biological replicates were completed with 120 initial worms per group. Oil Red O staining was performed as described previously ( Stuhr et al., 2022 ). Day 1 adult worms were collected and washed in PBS + 0.01 % Triton X-100 (PBST) before being fixed in 60 % isopropanol. Following fixation 600uL of 60 % ORO staining solution was added and worms were incubated for two hours rotating at room temperature. Worms were then de-stained in PBST for 30 min before they were mounted on 2 % agarose pads and imaged on an Echo Revolve Microscope at 10x magnification. Fiji ImageJ software was used for quantitative image analysis of the ORO pigment intensity ( Schindelin et al., 2012 ). L. brevis cells were collected into pellets by centrifugation (4000 g x 10 mins). The pellet was washed in PBS three times before being resuspended in acridine orange for 15 min. 500μL of the acridine orange-stained bacteria was seeded on an NGM plate and allowed to dry before worms were placed into the bacteria. After one hour, the worms were mounted on 2 % agarose pads with M9 and sodium azide and imaged at 40x magnification with a GFP filter. Intestinal bacterial colonization was adapted from previously described methods ( Palominos and Calixto, 2020 ; Ayala et al., 2017 ). Sterile conditions were maintained for the entirety of the assay. 35 age-synchronized worms per replicate were picked into 500μL of M9 containing 0.25 mM levamisole to induce temporal paralysis, preventing pharyngeal pumping and defecation to preserve intestinal bacterial contents. Tetracycline was added at 1:500 from a 12 mg/mL stock solution to sterilize the exterior of the worms and incubated for 15–20 min. Worms were then gently pelleted and washed 3x with M9. After the final wash the worms were resuspended in 200μL of M9 and transferred to 1.5 mL screw-cap RINO Navy tubes containing beads and homogenized using a Bullet Blender at speed 8 for 2 min. In a laminar flow hood, a series of tenfold dilutions from 10 –1 to 10 –6 in M9 were made using the whole worm lysates. 50μL of each dilution was pipetted onto agar plates and spread evenly around. Whole worm lysate dilutions from L. brevis were plated onto MRS agar and incubated overnight at 35 °C. Colonies were then counted, from plates with 30–300 colonies, multiplied by the dilution factor and divided by the number of worms to produce CFU/worm. Three biological replicates were performed for each bacterial food source. Day 4 adult worms expressing dsRED in HSN neurons (strain LX975) were mounted on 2 % agarose pads with M9 and sodium azide and imaged on a Nikon A1R MP+ multiphoton/confocal microscope at 60x magnification. Maximum intensity projections were generated, and a region of interest was drawn around each cell body to measure cell body area (μm 2 ) and fluorescence intensity. LUTS settings were standardized across images. Fiji ImageJ software was used for quantitative image analysis of the HSN cell body area and fluorescence intensity ( Schindelin et al., 2012 ). Day 1 adult worms expressing daf-16 ::GFP (strain MQD1543) were mounted on 2 % agarose pads with M9 and sodium azide and imaged on a Nikon A1R MP+ multiphoton/confocal microscope at 10x magnification. LUTS settings were standardized across images. Fiji ImageJ software was used to generate maximum intensity projections for qualitative image analysis to characterize daf- 16 localization as nuclear, cytoplasmic, or intermediate. ( Schindelin et al., 2012 ) Day 4 adult hermaphrodites were placed in individual microtiter wells containing 50μL of a solution of either serotonin (5mg/mL) or levamisole (0.1mg/mL) dissolved in M9 buffer and the number of eggs released in each well were counted after 60 mins. 10 worms were tested per drug per experiment with three biological replicates. Day 4 adult worms were collected and homogenized in Trizol LS reagent using a Bullet Blender at speed 12 for 5 min. RNA was isolated using a chloroform-isopropanol extraction method and samples were DNase treated (Qiagen). cDNA was synthesized with an oligo dT primer and Superscript III First-Strand Synthesis System (Thermo Fisher Scientific). SYBR Green PCR Master Mix (Thermo Fisher Scientific) was mixed with cDNA and primers for tph-1, rsu-1, unc-63, and egl-19. pmp-3 was used as a reference gene. qPCR was performed on a Bio-Rad CFX96 Real-Time PCR Detection System, and gene expression was quantified using the ΔΔCt method. Statistical analyses were performed using GraphPad Prism 10.2. An unpaired two-tailed Student’s t -test was used for all comparisons between two groups. For comparisons between multiple groups, One-Way ANOVA or Two-Way ANOVA (for repeated measures or two-variable analyses) was performed with post-hoc testing as indicated.

While the gut microbiome consists of a diverse community of over 1500 bacterial species across the human population that contribute to nutrient metabolism, pathogen defense, and intestinal barrier integrity ( Tang et al., 2023 ), its complex interactions with distant organs and systems, such as the immune, cardiovascular, and reproductive systems ( An et al., 2018 ; Ashonibare et al., 2024 ; Ghosh et al., 2022 ; Wiertsema et al., 2021 ), suggest that its influence extends far beyond digestion. These microbial communities engage in bidirectional communication with the host through neuronal, endocrine, and immune signaling pathways ( Brown and Hazen, 2015 ; Wiertsema et al., 2021 ). Such interactions underscore the microbiome’s essential role in maintaining systemic homeostasis, and it is increasingly recognized that microbial shifts in the gut, whether due to dietary, environmental, or genetic factors, can have profound effects on reproductive health. Gut microbiota dysbiosis is thought to be a contributing factor in multiple reproductive pathologies in humans, including PCOS, endometriosis, and chronic anovulation ( Chadchan et al., 2022 ; Qi et al., 2021 ; Sasaki et al., 2019 ). While C. elegans are anatomically far simpler than humans, it is precisely their simplicity and highly controllable gut microbiome that can help unlock fundamental causal relationships by which the gut influences the reproductive system. Our study provides new insights into the causal effects of the human gut isolate L. brevis on C. elegans reproduction, revealing significantly reduced reproductive fitness, including increased egg retention, reduced egg-laying, and health defects in both embryos and progeny. Though in C. elegans , the bacterial food source is both a dietary component and potential gut microbiome species, we were able to extract live L. brevis from the digestive tract using colonization assays, suggesting that L. brevis is not solely a dietary food source. Our results therefore have implications for understanding dysbiosis of the gut microbiome in humans and its effects on reproductive health, particularly when L.A.B. species are altered. In terms of nutrition, L. brevis appears to be a lower quality food source for C. elegans than the standard E. coli diet, yet the absence of canonical nutrient deprivation stress response suggests that the observed reproductive phenotypes arise from factors beyond simple nutrient insufficiency. Moreover, C. elegans raised on L. brevis are distinct from animals that undergo adult reproductive diapause (ARD), a starvation-induced hypometabolic state characterized by gonad shrinkage, rapid lipid depletion, lifespan extension, and nuclear localization of daf-16 ( Angelo and Van-Gilst, 2009 ; Baugh and Hu, 2020 ; Carranza-Garcia and Navarro, 2019 ). Unlike in ARD, worms fed L. brevis retain substantial lipid stores, have predominantly cytoplasmic daf-16 , and do not show alterations in lifespan. Thus, the reproductive defects in L. brevis -fed worms cannot be attributed to starvation or ARD. Mechanistically, we show that L. brevis disrupts reproductive physiology through both structural and functional deficits in the serotonergic HSN neurons of the egg laying circuitry. While the serotonergic signaling components tph-1 and egl-19 were not affected at the transcriptional level, it is possible that post-transcriptional regulation, and/or synaptic remodeling of the circuit may play a role in the serotonergic defects. While additional impairment of the VC cholinergic neurons cannot be ruled out, we provide evidence that nicotinic acetylcholine receptor expression on vulval muscles is not affected, and that the vulva is functionally intact. Future studies can further elucidate changes in the egg laying circuitry that underlie L. brevis -induced reproductive outcomes. Candidate microbial factors mediating the negative effects on reproduction include secreted metabolites such as short-chain fatty acids, bacterial neurotransmitters, or other small molecules capable of modulating neuronal signaling or the metabolic state of the worm ( Gadenne et al., 2022 ; O’Donnell et al., 2020 ; Silva et al. 2020 ; Strandwitz, 2018 ). Previous studies have demonstrated that microbial metabolites can influence host neuromodulation and reproduction ( Gadenne et al., 2022 ; O’Donnell et al., 2020 ; Silva et al. 2020 ; Strandwitz, 2018 Sengupta et al., 2024 , Shi and Murphy, 2016 ). Future investigations utilizing bacterial mutants, metabolite supplementation, and/or metabolically inactivated bacteria will be imperative to dissect the contributions of dietary components to host physiology. Our finding that L. brevis reduced reproductive fitness in C. elegans aligns with a previous study that observed low reproductive and developmental rates in C. elegans fed 35 different strains of L.A.B. ( Chelliah et al., 2018 ). Several studies across different model systems support the idea that L.A.B. can influence various aspects of reproduction. For instance, L.A.B. have been shown to alleviate symptoms of PCOS in rat models by modulating the gut microbiota through the regulation of sex hormones ( He et al., 2020 ). Similarly, a study in zebrafish demonstrated that feeding L. casei and L. rhamnosus improved both fecundity and immunity. However, these beneficial effects were lost once the probiotics were removed from the diet ( Qin et al., 2013 ). Additionally, research in Drosophila highlighted both direct and transgenerational effects of the gut microbiota on reproduction ( Morimoto et al., 2017 ). Furthermore, germ-free mice showed an increase in reproductive capacity after exposure to bacteria ( Shimizu et al., 1998 ), reinforcing the idea that the gut microbiota plays a crucial role in regulating reproductive health. These studies together with our findings underscore the need to better understand the specific effects of individual bacterial species, including L. brevis , on host physiology. Our study provides compelling evidence that L. brevis disrupts host reproductive function in C. elegans through altered serotonergic HSN function. These findings lay the groundwork for future studies to explore the role of specific microbiota in regulating reproductive health and the broader implications of gut dysbiosis for human disease. Given the growing evidence of the gut microbiome’s influence on both local and distal bodily functions, restoring microbial balance could emerge as a promising avenue for mitigating disease and promoting overall health across the lifespan.

The human gut microbiome has gained significant appreciation in recent years for its influence on a wide range of host physiological processes. Beyond its well-known role in digestion, the gut microbiome actively participates in regulating immune, metabolic, and endocrine functions, shaping host physiology across lifespan ( An et al., 2018 ; Ghosh et al., 2022 ; Tang et al., 2023 ; Walter, 2008 ). Emerging evidence suggests that disruptions in the gut microbiome, referred to as dysbiosis, are linked to a variety of pathologies, including neurodegenerative diseases, mental health disorders, and reproductive dysfunctions ( Chadchan et al., 2022 ; Keshavarzian et al., 2015 ; Li et al., 2017 ; Lin et al., 2019 ; Romano et al., 2021 ; Xiong et al., 2023 ; Zhao et al., 2018 ). Furthermore, gut dysbiosis is increasingly recognized as a contributing factor to reproductive health disorders, including polycystic ovary syndrome (PCOS), endometriosis, and gynecologic cancers ( Chadchan et al., 2022 ; Qi et al., 2021 ). Recent studies suggest that microbial imbalances in the gut may influence hormonal regulation, immune responses, and metabolic pathways that are critical for maintaining reproductive function ( Fu et al., 2021 ; Qi et al., 2021 ). Dysbiosis may thus serve as a potential underlying factor in the pathogenesis of these disorders, with gut microbiota serving as both a modulator and a potential target for interventions. However, the mechanisms through which the microbiome affects these processes remain poorly understood. Lactic acid bacteria (L.A.B.) are known inhabitants of the human gut microbiome and are among the bacterial species that undergo diversity changes contributing to gut dysbiosis observed with numerous health conditions ( An et al., 2018 ). L.A.B. are widely used in consumable probiotics and fermented foods under the general assumption that they are uniformly beneficial; however, the lack of current knowledge focusing on specific L.A.B. species leaves a major gap in understanding the interactions of these ubiquitous gut microbiota with the host. One such species, Levilactobacillus brevis ( L. brevis ), has been found to be elevated in the gut microbiome of Parkinson's disease patients ( Hasegawa et al., 2015 ). In other studies, supplementation with a mix of L.A.B. strains including L. casei, L. acidophilus , and L. brevis caused improvements in mood, behavior, and symptoms of depression in humans, potentially through the production of γ-aminobutyric acid (GABA) ( Daliri and Oh, 2016 ; Rout et al., 2018 ). Understanding how specific microbial shifts, such as those in L.A.B., influence host physiology across various organs and systems is crucial for developing therapeutic strategies aimed at restoring microbiome balance and mitigating disease. Yet, the complexity of the mammalian gut microbiome in tandem with low tractability and slow aging of rodent model systems has made it difficult to establish clear cause-and-effect relationships between specific L.A.B. species and host physiology or disease state. The small model organism, C. elegans , is an excellent system in which to explore host-microbiota interactions given that worms can survive on singular bacterial species allowing for a highly controlled, reductionist approach to determining the causal roles of individual species on host outcomes. In the wild, C. elegans maintain a natural, but minimal, gut microbiome consisting of the bacterial species they feed on, including L.A.B. ( Samuel et al., 2016 ). While the C. elegans digestive tract is anatomically simple, it offers a streamlined and controlled system with which to uncover species-specific mechanisms of host-microbiota interactions that are not feasible in higher order model organisms. C. elegans is also an exceptional model system for the study of host healthspan and offers the unique ability to conduct rapid, mechanistic studies due to having a short lifespan of 2–4 weeks and high genetic tractability. Although human reproduction is complex and relies on an intricate biological system, fundamental evolutionary processes, such as oogenesis, oocyte differentiation, and critical genetic pathways regulating reproduction, like MAPK and FoxO/Insulin signaling, are conserved in C. elegans ( Athar and Templeman, 2022 ; Edmonds et al., 2010 ; Marco et al., 2024 ). Moreover, the well-defined spatiotemporal layout of the C. elegans’ reproductive system combined with the transparency of its tissues, make it a powerful system for the study of reproduction and development within the germline ( Luo and Murphy, 2011 ). In this study, we show that C. elegans raised on L. brevis bacteria from the onset of adulthood manifest severe reproductive deficits, including elevated egg retention, an increased proportion of unhatched eggs among those that are laid, and progeny with developmental delay. Furthermore, we find that the serotonergic hermaphrodite-specific neurons (HSNs) of the egg-laying circuitry are both structurally and functionally impaired in L. brevis -fed worms, suggesting a critical role of the gut microbiome in regulating neuronal and reproductive health.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.