https://orcid.org/0000-0002-8152-6113
Figures
Abstract
Glucocorticoids (GCs) are steroid hormones predominately produced by the adrenal glands and secreted into the blood, but are also produced locally within some target tissues. During early development in mice, postnatal day (PND) 2–12, local GC levels in lymphoid organs are higher than blood GC levels. Interestingly, studies in germ-free (GF) and antibiotic-treated mice demonstrate that the intestinal microbiome modulates circulating and local GC levels, possibly through production of short-chain fatty acids (SCFAs). Importantly, GF mice lack all microbes, and antibiotic treatment is a very powerful perturbation that eliminates most intestinal microbes; the effects of more subtle microbiome manipulations on neonatal blood and local GC levels have yet to be determined. To begin to fill this gap, we established two groups of C57BL/6J mice with distinct microbiomes M1 and M2. We then bred adult mice from M1 and M2 and collected blood, bone marrow, thymus, and spleen from neonatal animals at PND5. We measured progesterone, 11-deoxycorticosterone, corticosterone, and 11-dehydrocorticosterone in the blood and lymphoid organs via liquid chromatography tandem mass spectrometry. We also measured nine SCFAs in the blood to determine a possible connection between microbiome and GCs. The adult microbiomes differed in community composition, but not overall community diversity. Mice with M1 had increased corticosterone and 11-dehydrocorticosteorne levels in blood and lymphoid organs. Mice with M1 also had increased levels of six SCFAs in blood, indicating a possible connection between SCFA signaling and GC levels. These data further our understanding of intestinal microbiome modulation of GC physiology by demonstrating that even subtle differences in microbiome composition can influence systemic and local GC levels.
Citation: Hamden JE, Ainsworth-Cruickshank G, Gray KM, Ye J, Kachkovski GV, Forys BJ, et al. (2026) The parental intestinal microbiome modulates systemic and local glucocorticoid levels in neonatal mice. PLoS One 21(9): e0357469. https://doi.org/10.1371/journal.pone.0357469
Editor: Assoc. Prof. Phakkharawat Sittiprapaporn, Mae Fah Luang University School of Anti Aging and Regenerative Medicine, THAILAND
Received: December 17, 2025; Accepted: August 14, 2026; Published: September 23, 2026
Copyright: © 2026 Hamden et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: Data are publicly available at: https://borealisdata.ca/previewurl.xhtml?token=9e24477a-bb3a-4a2a-9887-2a34ac834ba5.
Funding: This work was supported by Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery Grant (RGPIN-2019-04837) and Discovery Accelerator Supplement (RGDAS-2019-00033) and a Canada Foundation for Innovation Grant (32631) to KKS, a Michael Smith Foundation for Health Research and Canadians for Leading Edge Alzheimer’s Research Post-doctoral fellowship to JEH, NSERC CGS-D to BF, NSERC CGS-M and CGS-D fellowships to MS. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have no conflicts of interest to disclose.
Introduction
Glucocorticoids (GCs) are produced by the adrenal cortices and secreted into circulation. Adrenal GC production is tightly regulated by the hypothalamic-pituitary-adrenal axis. Circulating GC levels fluctuate on a circadian cycle and increase in response to stressors. Mice experience a period of adrenal quiescence from postnatal day (PND) 2–12, termed the stress hyporesponsive period (SHRP) [1–4]. During the SHRP, total circulating GC levels are 10–100 fold lower than in adult mice and show very small increases to most stressors, relative to adult animals (humans have an analogous period from ~6 months to 5 years) [5–8].
Some target organs, including lymphoid organs (e.g., bone marrow, thymus, and spleen), can locally produce GCs. Evidence for lymphoid production of GCs includes expression of steroidogenic enzymes, in vitro production of corticosterone, high local levels of GCs, and studies of knockout mice [5–7,9–13]. Specifically, within the SHRP, corticosterone levels are elevated in bone marrow, thymus, and spleen, relative to blood [6,12]. In lymphoid organs, GCs modulate B- and T-cell development. For example, during T-cell development, GCs alter thymocyte selection by promoting survival of more reactive T-cells [14]. Less is known about GCs in B-cell development, but B-cells express glucocorticoid receptor (GR), are sensitive to GC-induced apoptosis, and GCs inhibit B-cell receptor signaling in vitro, together suggesting a role for GCs in B-cell development [15–18].
Adrenal GC production is regulated by the intestinal microbiome and microbial metabolites such as short-chain fatty acids (SCFAs) [19–21]. Relative to control mice with an intact microbiome, germ-free (GF) mice have higher baseline serum adrenocorticotropic hormone and corticosterone levels, higher serum corticosterone levels in response to stress, and show reduced anxiety- and depression-like behaviors [22–24]. Interestingly, treatment of GF mice with Bifidobacterium infantis, a common bacterium in the murine intestine, reduces serum corticosterone to levels in controls [23]. Further, mice treated with antibiotics show increased systemic GC levels compared to controls, and co-housing antibiotic-treated mice with control mice reduces corticosterone levels [20]. Treating conventional mice with Lactobacillus rhamnonsus or a cocktail of SCFAs reduces stress-induced corticosterone levels and anxiety- and depression-like behaviours [21,25]. Finally, GF mice treated with an anti-CD3 antibody show decreased local production of corticosterone in the small intestine compared to their SPF counterparts, demonstrating that the intestinal microbiome can modulate local steroid production [26]. Together, these studies demonstrate that very powerful manipulations of the microbiome alter systemic and local corticosterone levels in mice. However, to our knowledge, no study has examined the effects of more subtle intestinal microbiome manipulations, for example, comparing two intact microbiomes that differ in community composition on systemic or local GC levels.
In this study, we compared mice housed side-by-side on the same rack in our animal facility but with two different intestinal microbiomes (M1 and M2), and examined SCFA and steroid levels at PND5. We established two groups of C57BL/6J mice with two distinct microbiomes. We bred the adult mice and collected blood and lymphoid organs from the offspring at PND5 (in the SHRP), to determine the effects of the parental intestinal microbiome composition on neonatal SCFA levels in blood and GC levels in the blood, bone marrow, thymus, and spleen. SCFA levels were measured by gas chromatography mass spectrometry and steroid levels were measured by liquid chromatography mass spectrometry (LC-MS/MS).
Methods
Subjects
Subjects were male and female C57BL/6J mice, and PND0 was defined as the first day that pups were present in the cage. 6-week old male and female mice were used for microbiome manipulation and then breeding, neonatal subjects were PND5. PND5 was selected because it is within the SHRP, when lymphoid organ GCs are higher and more sensitive to stressors than blood GC levels. Sex of PND5 mice was determined via genotyping at the University of British Columbia Genotyping Facility.
Mice were housed in a specific pathogen-free colony in the Modified Barrier Facility at the University of British Columbia. Colony rooms were maintained between 20–22°C with 40–70% relative humidity. Mice were housed under a 14:10 light:dark cycle (lights on 0600–2000h) in ventilated Ehret polysulfone Type IIL cages with a CO2 membrane, beta-chip bedding, and free access to water (purified by reverse osmosis and sterilized by chlorination) and food (Harland Teklad Global diet 2919 for breeders and diet 2918 after weaning). A red translucent hut and nestlet were placed in each cage for enrichment, and additional crinkle paper was supplied to breeder cages. All procedures were in compliance with the Canadian Council on Animal Care, and protocols were approved by the University of British Columbia Animal Care Committee (A14-0364; A14-0353).
Microbiome manipulation
Adult (6-weeks) C57BL/6J mice were purchased from the Jackson Laboratory. Upon arrival, mice were housed in soiled cages from mice already housed in this facility with two distinct previously established microbiomes. Mice were exposed to either Microbiome 1 (M1) or Microbiome 2 (M2) for two weeks prior to pairing for breeding. After pairing, mice were always housed in clean cages.
Tissue collection
PND5 mice were rapidly and deeply anesthetized with 5% isoflurane (2 L/min) and euthanized by rapid decapitation. Euthanasia was completed within 3 min of initial cage disturbance to avoid stress-induced changes in steroids. Trunk whole blood (hereafter “blood”) was collected in 1.5 mL polypropylene microcentrifuge tubes (VWR, Edmonton, AB) and snap frozen on dry ice. Blood was used for steroid measurements to facilitate comparisons between circulating and local steroid levels because serum and plasma overestimate circulating steroid levels [6]. Lymphoid organs were dissected and snap frozen on powdered dry ice. Blood and lymphoid organs were stored at −70°C.
DNA extraction and 16S sequencing
Feces were collected from adult mice (>PND80) and snap frozen on dry ice. DNA was extracted using the DNeasy Powersoil Kit (QIAGEN, Cat. 12888−100). Briefly, fecal pellets (adults) or entire large intestine, cecum, and contents (neonates) were homogenized in solution C1, 500 µL of supernatant was collected and mixed with solution C2, 600 µL supernatant was collected and mixed with solution C3, 750 µL was collected and mixed with solution C4, 675 µL was collected and added the spin column, flow through was discarded, columns were washed with solution C5, and DNA was eluted using solution C6. 16S amplicons were prepared from DNA by targeting the V4 (adult) and V4-B5b (neonates) regions (Table 1). Amplicons were sequenced on an Illumina Miseq (2x300bp sequencing chemistry) at the Dalhousie IMR (imr.bio) as originally described [27] and updated [28]. Importantly, neonatal samples could not be analyzed due to the relatively low number of bacterial ASVs sequenced from the neonatal intestine, so only adult extracts were amplified and analyzed.
SCFA extraction and measurement
30 µl blood samples were mixed with 700 μl of isopropyl alcohol containing 2-ethylbutyric acid at a concentration of 0.01% v/v, which was used as an internal standard. The samples were vortexed at maximum speed for 1 min, then left at room temperature for 15 min. Following incubation, samples were centrifuged at 15,100 × g for 10 min at 4°C. The supernatant was carefully transferred to a gas chromatography vial for subsequent analysis. 1 μl aliquot of the cleared supernatant was injected into a Trace 1300 Gas Chromatograph (Thermo Fisher Scientific) equipped with a flame-ionization detector (FID) and an AI/AS 1310 auto sampler (Thermo Fisher Scientific), operating in splitless mode. Chromatographic separation was achieved using a fused silica FAMEWAX column (30 m × 0.32 mm i.d., 0.25 μm film thickness; Restek, Bellefonte, PA, USA), with helium (Praxair) as the carrier gas at a flow rate of 1.8 mL/min. The oven temperature program was as follows: initial temperature of 80°C, held for 5 min, increased to 90°C at 5°C/min, then increased from 90°C to 105°C at 0.9°C/min, followed by an increase to 240°C at 20°C/min, and held for 5 min. The FID and injection port temperatures were set at 240°C and 230°C, respectively. The flow rates of hydrogen, air, and nitrogen (used as makeup gas) were 30 mL/min, 300 mL/min, and 20 mL/min, respectively. Data were acquired and analyzed using Chromeleon software, version 7.2 (Thermo Fisher Scientific).
A serial dilution of the Volatile Free Acid Mix (CRM46975, Sigma) was prepared to generate standards with the following concentrations: 1.11, 0.37, 0.123, and 0.041 mg/mL. The calibration curve was generated by running each standard under the same conditions described for the samples. Peak areas were extracted from the chromatographic data, and a standard curve was generated by plotting the concentration of the standards against the corresponding peak areas.
Steroid extraction and measurement
Blood (5 µl; 5.25 mg) was placed in 2 mL polypropylene bead ruptor tubes with 5 zirconium oxide beads. Steroids were extracted from blood and brain tissue via liquid-liquid extraction as previously described. Calibration curves, blanks, and quality controls were processed alongside samples. The calibration curve range was 0.05–1000 pg for all steroids and prepared in 50:50 HPLC-grade methanol:MilliQ water.
Steroid analysis by LC-MS/MS was conducted as previously described. Total progesterone, 11-deoxycorticosterone (DOC), corticosterone, 11-dehydrocorticosterone (DHC), 11-deoxycortisol, cortisol, and cortisone were measured using multiple reaction monitoring with 2 unique mass transitions with unique retention times for each analyte. Steroid concentrations were acquired with a Sciex 6500 Qtrap triple quadrupole tandem mass spectrometer (Sciex LLC, Framingham, MA) in positive electrospray ionization mode for all steroids. The lower limit of quantification was 0.05 pg per sample. Pooled mouse serum was used as an inter-assay control to allow for comparison across all sample runs (n = 3/run, 3 runs total). Deuterated internal standards (progesterone-d9, corticosterone-d8, and cortisol-d4, C/D/N Isotopes Inc., Pointe-Claire, Canada) were included in all standards, blanks, quality controls and samples to correct for losses and matrix effects.
Data analysis
For PND5 animals, there was no effect of Sex and no interaction between Sex and any other variable for both SCFA and steroid levels (p ≥ 0.12 in all cases) so data from males and females were pooled as before [5,6,9,12,29,30]. One animal was excluded from each group for being greater than 2 standard deviations from the mean. Data are publicly available at: https://doi.org/10.5683/SP3/JNRLAR.
Microbiome processing/taxonomy.
Demultiplexed fastq files were processed in Rstudio (v.4.4.2) using the dada2 package (v1.24.0). Dada2 processing consisted of files being primer trimmed (cutadapt v4.1)), quality filtered, dereplicated, and chimera removal. In order to remove amplicon sequence variants (ASVs) — unique, single-nucleotide variations in DNA sequences used to determine bacterial taxonomy — with low reads counts and singletons, ASVs with a relative abundance less than 0.1% were removed. Taxonomy was assigned according to the SILVA v138 database using the dada2 pipeline. Post dada2 processing was done in Rstudio using the phyloseq package (v1.40.0). ASVs identified as chloroplast, mitochondria and eukaryotic as well as sequences unassigned at the domain level were filtered out. Additionally, ASV with count of two less per sample were filtered out to minimize the potential effect of barcode switching and ASVs of less than 250 reads across all samples were filtered out. The filtered phyloseq object was rarefied to the minimum sample ASV count (58011) for community-level analyses. Non-community level analyses were conducted using the non-rarefied phyloseq object.
Predicted metabolic pathways.
Biochemical pathways identified from the non-rarefied dataset were annotated using the PICRUSt2 methods (version 2.3.0 beta) and using an available GitHub pipeline [31,32]. Filtering the predicted biochemical pathway phyloseq object was conducted by removing classified pathways with counts less than two per sample and by removing low sequence samples ≤ 250 bp. For community-level analyses, the predicted biochemical pathway dataset was rarefied to the minimum predicted pathway read count per sample (3975518).
Microbiome analysis.
Statistical analysis and visualizations were conducted through R. Studio (v.4.2.0) using the base plot package and ggplot2 (v.3.36) (n = 5/group). Community diversity (α-diversity) as measured by Shannon Index (Richness and Evenness) was compared between M1 and M2 using a one-way t-test. To compare differences in bacterial community (β-diversity) as measured by Bray-Curtis dissimilarity index (presence and absence-based), permutational multivariate ANOVA (PERMANOVA) was performed with 999 sample permutations. Non-metric multidimensional scaling (NMDS) plots of Bray-Curtis dissimilarity index were used to visualize community composition. Diversity indexes calculations and PERMANOVA were conducted via the vegan package (v.2.6−2). Relative abundance plots of the top 15 genera were made to visualize differences in specific taxa between groups and DESeq2 analysis (v.1.36.0) was performed to statistically detect differences in specific taxa between groups at the genus level. Analyzing predicted biochemical pathway composition and differences in specific predicted biochemical pathways followed the same methodology as when analyzing taxa.
Short-chain fatty acid data analysis.
Data were analyzed for an effect of Microbiome for each SCFA by Mann-Whitney U test (n = 9/group).
Steroid data analysis.
A steroid was considered below the lower limit of quantification (LLOQ) if the quantifier and qualifier transitions were not present. For groups with at least 20% of measurements above the LLOQ, the measurements below the LLOQ were replaced with LLOQ/√2. For bone marrow, progesterone was only detectable in 1 animal per group and 11-deoxycorticosterone (DOC) was non-detectable in all animals (Table 2). Thus, bone marrow was excluded from the progesterone and DOC analysis. Data were analyzed for effects of Tissue and Microbiome and for a Tissue x Microbiome interaction for each steroid separately by mixed-effects model (α = 0.05 for all statistical tests). To ensure homogeneity of variance, data were log transformed prior to analysis when necessary. All graphs are presented using non-transformed data as mean ± standard error of the mean (n = 8–9/group). All data were analyzed using GraphPad Prism 10 (version 10.4.2).
Results
Adult microbiomes
To determine if the fecal microbiome differed between M1 and M2, community level and non-community level analyses were performed (n = 5/group). The overall bacterial diversity (α-diversity) in adult mice from M1 and M2 did not significantly differ (Shannon: t(5.84) = 0.46, p = 0.65, n1 = 5, n2 = 5; Fig 1A). However, a significant difference in community compositions (β-diversity) was detected (Bray-Curtis: Pseudo-F = 4.6, df = 1, R2 = 0.36, p = 0.01; Fig 1B). The differences in community composition were further demonstrated by the relative abundance of the top 15 genera across all microbiome samples in both groups (Fig 1C). We performed DEseq2 to identify what taxa were driving the differences in community composition between groups (Fig 1D). Fifteen genera were differentially abundant between the two groups, with 7 genera higher in M2 (Helicobacter, Alloprevotella, Tyzerella, Mucispirillum, Anaerotruncus, Lachnospiraceae_NK4A136_group, and Odoribacter) and 8 genera higher in M1 (Bilophila, Lachnospiraceae_UCG-006, Christensenellaceae_UN, UCG-010_un, UCG-009, A2, RF39_un, and Turicibacter). Notably, most differences between groups were due to a complete absence of the genera in one group. Only Lachnospiraceae_NK4A136_group and Odoribacter overlapped between the 15 most abundant taxa across all microbiome samples across both groups and those identified as significantly different (Fig 1C and 1D). Predicted biochemical pathway overall composition did not significantly differ between groups (Bray-Curtis: Pseudo-F = 1.03228, df = 1, R2 = 0.11434, p = 0.355; Fig 1E). However, 12 specific biochemical pathways were significantly different between groups (PWY-7254 (TCA cycle VII – acetate producers), PPGPPMET-PWY (ppGpp metabolism), PWY-7373 (superpathway of demethylmenaquinol-6 biosynthesis II), PWY-5747 (2-methylcitrate cycle II), PWY0–42 (2-methylcitrate cycle I), PWY0–1338 (polymyxin resistance), P125-PWY (superpathway of (R,R)-butanediol biosynthesis), PWY-6396 (chimeric pathway: superpathway of 2,3-butanediol biosynthesis), PWY-1622 (formaldehyde assimilation I – serine pathway), GALACTARDEG-PWY (D-galactarate degradation I), GLUCARGALACTSUPER-PWY (superpathway of D-glucarate and D-galactarate degradation), and PWY0–1533 (methylphosphonate degradation I) (Fig 1F).
(A-D) Taxonomic analysis. (E and F) Predicted biochemical pathway analysis. (A) α-diversity. Boxplot showing median Shannon (evenness and richness) index to describe the most abundant taxa across both groups, each point represents an individual mouse. (B and E) Taxa (B) and Biochemical pathways (E). β-diversity. Non-metric multidimensional scaling (NMDS) plot in two dimensions of Bray-Curtis dissimilarity (presence and absence-based) index. Each point represents an individual mouse and ellipsoids represent the 95% confidence interval for each group. (C) Relative Abundance barplot of the top 15 genera across samples. (D and F) Taxa (D) and Biochemical pathways (F). Left: DESeq2 results of significant genera that differ between facilities. Color of dot indicates decrease or increase in Microbiome 1 mice compared to Microbiome 2 mice. Size of dot is proportional to the log2fold change in differential abundance. Right: Bubble plot indicating the presence and abundance of each significant genera from DESeq2 across samples. p ≤ 0.05; *, ns; non-significant.
Neonatal short-chain fatty acids
Nine short chain fatty acids were quantifiable in the blood of PND5 mice (n = 9/group) (Table 3). There was a significant effect of Microbiome for acetic acid (p = 0.0149), isobutyric acid (p < 0.0001), butyric acid (p < 0.0001), hexanoic acid (p = 0.0018), n-heptanoic acid (p = 0.0015), and isovaleric acid (p = 0.0149). There was no significant effect of Microbiome for formic acid (p = 0.0984), propionic acid (p = 0.4755), or isocaproic acid (p = 0.7448).
Neonatal steroids
Progesterone was above the LLOQ in only 2 bone marrow samples and DOC was below the LLOQ in all bone marrow samples, and thus bone marrow was excluded from analyses for progesterone and DOC (Table 2). For progesterone and DOC, there were significant main effects of Tissue only (p < 0.0001 for both steroids) (Fig 2). For corticosterone, there were significant main effects of Tissue (p < 0.0001) and Microbiome (p = 0.0095), but no Tissue x Microbiome interaction (p = 0.3267). For DHC, there were significant main effects of Tissue (p < 0.0001) and Microbiome (p = 0.0113), but no Tissue x Microbiome interaction (p = 0.7579). 11-deoxycortisol, cortisol, and cortisone were non-detectable in all samples.
(A) progesterone, (B) 11-deoxycorticosterone (DOC), (C) corticosterone, (D) 11-dehydrocorticosterone (DHC) in whole blood, bone marrow, thymus, spleen of postnatal day 5 mice with Microbiome 1 or Microbiome 2. Data were analyzed by Mixed-effects model for all steroids. Data are shown as mean ± SEM. n = 8-9 for all groups.
Discussion
In this study, we established two groups of C57BL/6J mice (in the same room) with two distinct microbiomes, profiled the adult microbiome, and measured neonatal blood SCFA levels and blood and lymphoid GC levels. It has been demonstrated that GF and antibiotic-treated mice have higher circulating and intestine GC levels than mice with an intact microbiome, but this is the first study to demonstrate that mice with microbiomes that differ in beta diversity have different blood and lymphoid GC levels. This is also the first study to show that the parental microbiome modulates GC levels in early postnatal development, a time of critical importance for lymphocyte development, which is strongly modulated by GCs [33].
Unfortunately, we could not assess the neonatal microbiome in this study because of the relatively low bacterial load in neonatal mice compared to adults. Due to the lack of neonatal microbiome data, all relationships between the adult microbiome and neonatal SCFA and GC levels are associative. Several studies have attempted to profile the microbiome in neonatal mice from birth to weaning. Pantoja-Feliciano et al. (2013) demonstrated that 1 day after birth, neonatal colonic microbiome more closely resembles the dam vaginal microbiome than the dam fecal microbiome [34]. The PND1 microbiome is primarily composed of Streptococcaceae and various taxa from the Proteobacteria phylum. By PND3 and PND9, neonatal colonic microbiome diversity decreases sharply, becoming dominated by Lactobacillus, Streptococcus, and Pasteurella. By PND21, colonic microbiome diversity increases to more closely mirror that of the dams, which is thought to occur as the pups become coprophagic. A study by van Best et al. (2020) similarly found a dominance of Lactobacillus at PND7, with microbiome composition becoming more “adult-like” by PND21 [35].
Until recently, the dynamics and source of the microbes in the neonatal microbiome were not well understood. A recent study by Kennedy et al. (2025) sequenced pup fecal samples from PND4-PND9, then at PND11, 13, 15, 21, and into adulthood [36]. Similar to previous studies, the PND4–14 microbiome is dominated by Ligilactobacillus (a newly described genus that was previously included in the Lactobacillus genus) and Streptococcus, which are rarely found in maternal fecal samples and other body sites. When comparing the pup and dam microbiome composition, similarity was lowest before PND10, and by PND15–20, the pup microbiome was more similar to its own dam’s microbiome than other dams [36]. Overall, the neonatal mouse gut microbiome is compositionally and functionally distinct from the adult mouse gut microbiome. The dominance of Ligilactobacillus and Lactobacillus in the neonatal mouse microbiome is widely thought to be due to their ability to ferment breast milk sugars such as lactose [37]. Although there is some evidence that specific strains of Lactobacillus spp. can produce SCFAs, they are not typically considered canonical SCFA producers [38].
The two adult microbiomes that we established in this study did not differ in alpha diversity, but did differ in beta diversity, a measure of community composition. In adult microbiomes, only two significantly different taxa between M1 and M2 were among the 15 most abundant taxa in both microbiomes, suggesting that differences in genera abundance are driven largely by lower abundance taxa and by presence in one microbiome only, rather than shifts in abundance between microbiomes. For example, taxa from Helicobacter, Alloprevotella, Tyerella, and Mucispirillum were present in all M2 samples and absent in all M1 samples. Interestingly, all genera, excluding Biophilia, that were present in M1 but absent in M2 (Christensenellaceae_Un, UCG-010_Un, UCG-009, A2, RF39_Un, and Turicibacter) belong to the phylum Firmicutes, which are among the main SCFA producers in the murine intestine. Predicted metabolic pathway analysis revealed no overall differences in pathway composition; however, specific pathways varied between M1 and M2. While these results are predictive, they suggest a functional shift in the microbiome, corresponding to observed taxonomic changes. Direct assessment of the microbiome’s metabolomic profile, using shotgun metabolomics, alongside blood and brain metabolite quantification may further elucidate the relationship between the microbiome, SCFAs, and GCs.
Compared to M2 neonates, M1 neonates had significantly higher blood levels of six SCFAs, including acetate and butyrate. Interestingly, acetate, butyrate, and propionate decreased corticosterone levels in stressed adult mice when given in drinking water [21]. As treatment with SCFAs decreased corticosterone levels in adults, it might be expected that neonates with higher blood SCFA levels would have lower blood corticosterone levels. However, M1 neonates had both higher SCFA and GC levels. The discrepancy between previous reports and the present data may be due to a difference in age. Previous studies have conducted experiments in adult animals, and to our knowledge, these are the first data examining the relationship between SCFA and GC levels in neonatal animals. Alternatively, experiments that treat animals with exogenous SCFAs artificially elevate circulating SCFA levels, and experiments in GF and antibiotic-treated animals dramatically reduce circulating SCFA levels. It is possible that the corticosterone-suppressive effects of SCFAs are less pronounced at physiological SCFA levels. However, more mechanistic study is needed to determine a causal relationship between SCFAs and GC levels in neonatal animals. Compared to SCFA levels in adult mice, these neonatal blood levels are considerably lower and less variable across SCFAs [39]. This large difference in neonatal SCFA levels compared to adult levels is likely due to reduced SCFA-producing bacteria in the neonatal intestine. Interestingly, microbiota-produced maternal SCFAs can reach embryos during pregnancy and are delivered through breast milk during neonatal development, SCFAs can be synthesized by neonates through breakdown of carbohydrates present in breast milk [40–42]. Thus, the higher blood SCFA levels in M1 neonates may be due to maternal microbiome.
Blood GC levels in M1 and M2 neonates were low, consistent with the SHRP, but higher in M1 than M2 [5–10,12]. Lymphoid GC levels were also similar to previously reported levels in neonatal mice [6,7,10,12] and also higher in M1 than M2. Additionally, GC levels were higher in the thymus than the bone marrow or spleen in both groups. These are the first data demonstrating that the intestinal microbiome can modulate local GC levels within lymphoid organs. GCs affect the immune system and lymphocyte development [43–47]. However, these studies used repeated stressors that produce large increases in GC levels (e.g., lipopolysaccharide) or used in vitro models that cannot be easily compared to in vivo GC levels. It is unclear how more subtle changes in GC levels during early postnatal development alter lymphoid system development in neonatal animals and function in adult animals. However, increased GC levels during lymphoid system and cell development generally lead to a more reactive lymphoid cell repertoire [47].
Studies in GF and antibiotic-treated mice indicate that the intestinal microbiome can modulate blood and intestinal GC levels, possibly through microbial SCFAs. The present data also support the idea that the parental intestinal microbiome modulates neonatal GC levels. These are the first data to demonstrate that more subtle changes in the intestinal microbiome (compared to antibiotic treatment) modulate both blood and lymphoid organ GC levels in neonatal animals. Interestingly, animals from M1 and M2 also had different neonatal blood SCFA levels. These differences in circulating SCFA levels provide a possible mechanism for the intestinal microbiome to modulate neonatal GC levels.
References
- 1. Meaney MJ, Sapolsky RM, McEwen BS. The development of the glucocorticoid receptor system in the rat limbic brain. I. Ontogeny and autoregulation. Brain Res. 1985;350(1–2):159–64. pmid:3986609
- 2. D’Amato FR, Cabib S, Puglisi-Allegra S, Patacchioli FR, Cigliana G, Maccari S, et al. Effects of acute and repeated exposure to stress on the hypothalamo-pituitary-adrenocortical activity in mice during postnatal development. Horm Behav. 1992;26(4):474–85. pmid:1478632
- 3. Schmidt MV. Stress-Hyporesponsive Period. Stress: Physiology, Biochemistry, and Pathology. Elsevier. 2019. p. 49–56.
- 4. Schmidt MV, Enthoven L, van der Mark M, Levine S, de Kloet ER, Oitzl MS. The postnatal development of the hypothalamic-pituitary-adrenal axis in the mouse. Int J Dev Neurosci. 2003;21(3):125–32. pmid:12711350
- 5. Hamden JE, Gray KM, Salehzadeh M, Soma KK. Isoflurane stress induces region-specific glucocorticoid levels in neonatal mouse brain. J Endocrinol. 2022;255(2):61–74. pmid:35938697
- 6. Hamden JE, Salehzadeh M, Gray KM, Forys BJ, Soma KK. Isoflurane stress induces glucocorticoid production in mouse lymphoid organs. J Endocrinol. 2021;251(2):137–48. pmid:34432644
- 7. Salehzadeh M, Hamden JE, Li MX, Bajaj H, Wu RS, Soma KK. Glucocorticoid Production in Lymphoid Organs: Acute Effects of Lipopolysaccharide in Neonatal and Adult Mice. Endocrinology. 2022;163(2):bqab244. pmid:34864986
- 8. Hamden JE, Salehzadeh M, Bajaj H, Li MX, Soma KK. Lipopolysaccharide differentially alters systemic and brain glucocorticoid levels in neonatal and adult mice. J Neuroendocrinol. 2025;37(2):e13481. pmid:39694531
- 9. Hamden JE, Gray KM, Salehzadeh M, Kachkovski GV, Forys BJ, Ma C, et al. Steroid profiling of glucocorticoids in microdissected mouse brain across development. Dev Neurobiol. 2021;81(2):189–206. pmid:33420760
- 10. Hamden JE, Salehzadeh M, Jalabert C, O’Leary TP, Snyder JS, Gomez-Sanchez CE, et al. Measurement of 11-dehydrocorticosterone in mice, rats and songbirds: Effects of age, sex and stress. Gen Comp Endocrinol. 2019;281:173–82. pmid:31145891
- 11. Taves MD, Otsuka S, Taylor MA, Donahue KM, Meyer TJ, Cam MC, et al. Tumors produce glucocorticoids by metabolite recycling, not synthesis, and activate Tregs to promote growth. J Clin Invest. 2023;133(18):e164599. pmid:37471141
- 12. Taves MD, Plumb AW, Sandkam BA, Ma C, Van Der Gugten JG, Holmes DT, et al. Steroid profiling reveals widespread local regulation of glucocorticoid levels during mouse development. Endocrinology. 2015;156(2):511–22. pmid:25406014
- 13. Taves MD, Plumb AW, Korol AM, Van Der Gugten JG, Holmes DT, Abraham N, et al. Lymphoid organs of neonatal and adult mice preferentially produce active glucocorticoids from metabolites, not precursors. Brain Behav Immun. 2016;57:271–81. pmid:27165988
- 14. Van Laethem F, Baus E, Andris F, Urbain J, Leo O. A novel aspect of the anti-inflammatory actions of glucocorticoids: inhibition of proximal steps of signaling cascades in lymphocytes. Cell Mol Life Sci. 2001;58(11):1599–606. pmid:11706987
- 15. Gruver-Yates AL, Quinn MA, Cidlowski JA. Analysis of glucocorticoid receptors and their apoptotic response to dexamethasone in male murine B cells during development. Endocrinology. 2014;155(2):463–74. pmid:24196358
- 16. Garvy BA, Telford WG, King LE, Fraker PJ. Glucocorticoids and irradiation-induced apoptosis in normal murine bone marrow B-lineage lymphocytes as determined by flow cytometry. Immunology. 1993;79(2):270–7. pmid:7688342
- 17. Trottier MD, Newsted MM, King LE, Fraker PJ. Natural glucocorticoids induce expansion of all developmental stages of murine bone marrow granulocytes without inhibiting function. Proc Natl Acad Sci U S A. 2008;105(6):2028–33. pmid:18250324
- 18. Cain DW, Bortner CD, Diaz-Jimenez D, Petrillo MG, Gruver-Yates A, Cidlowski JA. Murine Glucocorticoid Receptors Orchestrate B Cell Migration Selectively between Bone Marrow and Blood. J Immunol. 2020;205(3):619–29. pmid:32571841
- 19. Vagnerová K, Vodička M, Hermanová P, Ergang P, Šrůtková D, Klusoňová P, et al. Interactions Between Gut Microbiota and Acute Restraint Stress in Peripheral Structures of the Hypothalamic-Pituitary-Adrenal Axis and the Intestine of Male Mice. Front Immunol. 2019;10:2655. pmid:31798585
- 20. Scheer S, Medina TS, Murison A, Taves MD, Antignano F, Chenery A, et al. Early-life antibiotic treatment enhances the pathogenicity of CD4+ T cells during intestinal inflammation. J Leukoc Biol. 2017;101(4):893–900. pmid:28034915
- 21. van de Wouw M, Boehme M, Lyte JM, Wiley N, Strain C, O’Sullivan O, et al. Short-chain fatty acids: microbial metabolites that alleviate stress-induced brain-gut axis alterations. J Physiol. 2018;596(20):4923–44. pmid:30066368
- 22. Luo Y, Zeng B, Zeng L, Du X, Li B, Huo R, et al. Gut microbiota regulates mouse behaviors through glucocorticoid receptor pathway genes in the hippocampus. Transl Psychiatry. 2018;8(1):187. pmid:30194287
- 23. Sudo N, Chida Y, Aiba Y, Sonoda J, Oyama N, Yu X-N, et al. Postnatal microbial colonization programs the hypothalamic-pituitary-adrenal system for stress response in mice. J Physiol. 2004;558(Pt 1):263–75. pmid:15133062
- 24. Huo R, Zeng B, Zeng L, Cheng K, Li B, Luo Y, et al. Microbiota Modulate Anxiety-Like Behavior and Endocrine Abnormalities in Hypothalamic-Pituitary-Adrenal Axis. Front Cell Infect Microbiol. 2017;7:489. pmid:29250490
- 25. Bravo JA, Forsythe P, Chew MV, Escaravage E, Savignac HM, Dinan TG, et al. Ingestion of Lactobacillus strain regulates emotional behavior and central GABA receptor expression in a mouse via the vagus nerve. Proc Natl Acad Sci U S A. 2011;108(38):16050–5. pmid:21876150
- 26. Ergang P, Vagnerová K, Hermanová P, Vodička M, Jágr M, Šrůtková D, et al. The Gut Microbiota Affects Corticosterone Production in the Murine Small Intestine. Int J Mol Sci. 2021;22(8):4229. pmid:33921780
- 27. Comeau AM, Douglas GM, Langille MGI. Microbiome Helper: a Custom and Streamlined Workflow for Microbiome Research. mSystems. 2017;2(1):e00127-16. pmid:28066818
- 28.
Comeau AM, Kwawukume A. Preparing multiplexed 16S/18S/ITS amplicons for the Illumina MiSeq. 2023.
- 29. Spinedi E, Chisari A, Pralong F, Gaillard RC. Sexual dimorphism in the mouse hypothalamic-pituitary-adrenal axis function after endotoxin and insulin stresses during development. Neuroimmunomodulation. 1997;4(2):77–83. pmid:9483198
- 30. McCormick CM, Kehoe P, Kovacs S. Corticosterone release in response to repeated, short episodes of neonatal isolation: evidence of sensitization. Int J Dev Neurosci. 1998;16(3–4):175–85. pmid:9785114
- 31. Douglas GM, Maffei VJ, Zaneveld JR, Yurgel SN, Brown JR, Taylor CM, et al. PICRUSt2 for prediction of metagenome functions. Nat Biotechnol. 2020;38(6):685–8. pmid:32483366
- 32. Douglas GM. PICRUSt2 Tutorial (v2.3.0 beta). 2021. https://github.com/picrust/picrust2/wiki/PICRUSt2-Tutorial-(v2.3.0-beta)
- 33. Taves MD, Hamden JE, Soma KK. Local glucocorticoid production in lymphoid organs of mice and birds: Functions in lymphocyte development. Horm Behav. 2017;88:4–14. pmid:27818220
- 34. Pantoja-Feliciano IG, Clemente JC, Costello EK, Perez ME, Blaser MJ, Knight R, et al. Biphasic assembly of the murine intestinal microbiota during early development. ISME J. 2013;7(6):1112–5. pmid:23535917
- 35. van Best N, Rolle-Kampczyk U, Schaap FG, Basic M, Olde Damink SWM, Bleich A, et al. Bile acids drive the newborn’s gut microbiota maturation. Nat Commun. 2020;11(1):3692. pmid:32703946
- 36. Kennedy EA, Weagley JS, Kim AH, Antia A, DeVeaux AL, Baldridge MT. Bacterial community assembly of specific pathogen-free neonatal mice. Microbiome. 2025;13(1):46. pmid:39920864
- 37. Schwab C, Gänzle M. Lactic acid bacteria fermentation of human milk oligosaccharide components, human milk oligosaccharides and galactooligosaccharides. FEMS Microbiol Lett. 2011;315(2):141–8. pmid:21175746
- 38. Nami Y, Shaghaghi Ranjbar M, Modarres Aval M, Haghshenas B. Harnessing Lactobacillus-derived SCFAs for food and health: Pathways, genes, and functional implications. Curr Res Microb Sci. 2025;9:100496. pmid:41230178
- 39. Colombo AV, Sadler RK, Llovera G, Singh V, Roth S, Heindl S, et al. Microbiota-derived short chain fatty acids modulate microglia and promote Aβ plaque deposition. Elife. 2021;10:e59826. pmid:33845942
- 40. Kimura I, Miyamoto J, Ohue-Kitano R, Watanabe K, Yamada T, Onuki M, et al. Maternal gut microbiota in pregnancy influences offspring metabolic phenotype in mice. Science. 2020;367(6481):eaaw8429. pmid:32108090
- 41. Xi M, Yan Y, Duan S, Li T, Szeto IM-Y, Zhao A. Short-chain fatty acids in breast milk and their relationship with the infant gut microbiota. Front Microbiol. 2024;15:1356462. pmid:38440144
- 42. Gao Y, Davis B, Zhu W, Zheng N, Meng D, Walker WA. Short-chain fatty acid butyrate, a breast milk metabolite, enhances immature intestinal barrier function genes in response to inflammation in vitro and in vivo. Am J Physiol Gastrointest Liver Physiol. 2021;320(4):G521–30. pmid:33085904
- 43. Avitsur R, Sheridan JF. Neonatal stress modulates sickness behavior. Brain, Behavior, and Immunity. 2009;23(7):977–85.
- 44. Hodgson DM, Knott B, Walker FR. Neonatal endotoxin exposure influences HPA responsivity and impairs tumor immunity in Fischer 344 rats in adulthood. Pediatr Res. 2001;50(6):750–5. pmid:11726735
- 45. Shanks N, Windle RJ, Perks PA, Harbuz MS, Jessop DS, Ingram CD, et al. Early-life exposure to endotoxin alters hypothalamic-pituitary-adrenal function and predisposition to inflammation. Proc Natl Acad Sci U S A. 2000;97(10):5645–50. pmid:10779563
- 46. Walker AK, Nakamura T, Hodgson DM. Neonatal lipopolysaccharide exposure alters central cytokine responses to stress in adulthood in Wistar rats. Stress. 2010;13(6):506–15. pmid:20666652
- 47. Taves MD, Ashwell JD. Glucocorticoids in T cell development, differentiation and function. Nat Rev Immunol. 2021;21(4):233–43. pmid:33149283