Figures & data
Figure 1. Maternal opioid exposure during gestation alters maternal gut microbiota.
(a) Generation of prenatal opioid-exposed (MET) or control (CTRL) offspring from opioid exposed (methadone mom) or control (control mom) dams. (b) Alpha diversity in control (n=`11) or methadone (n=6) exposed dams represented by Shannon indices using pairwise Wilcoxon rank sum test (p>.05 for measured index) (c) Principal coordinate analysis (PCoA) of beta diversity in control mom (n=11) and methadone mom (n=6) samples based on Bray-Curtis dissimilarity shows distinct clustering (q<0.05) between groups. (d) Linear discriminant analysis effect size (LEfSe) analysis of top discriminative bacteria genera between gut samples from control mom (n=11) and methadone mom (n=6) samples (e) BugBase prediction of the relative abundance of aerobic bacteria, biofilm-forming bacteria, gram negative bacteria and gram-positive bacteria (p<.05 for all pairwise comparisons). (f) KEGG pathway of gut microbiota predicted using phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt). Data are presented in a bar plot with 95% confidence intervals and P values between gut samples from control Moms (n=11) and methadone Moms (n=6). POE, prenatal opioid exposure; HYD, hydromorphone; SAL, saline; PG, pregestation; E, estrous day; P, postnatal day
Figure 2. Maternal opioid exposure during gestation alters neonatal gut microbiota.
(a) Alpha diversity in CTRL (n=`10) or MET (n=8) 3-week-old male offspring represented by Shannon index using pairwise Wilcoxon rank sum test. P>.05 for all measured indices. (b) Principal coordinate analysis (PCoA) of beta diversity in CTRL (n=10) and MET (n=8) samples based on Bray-Curtis dissimilarity shows distinct clustering (q<0.05) between groups. (c) Relative abundance of significantly differentially abundant genera in CTRL (n=10) vs MET (n=8) exposed offspring assessed by Wilcoxon rank sum test (d) BugBase prediction of the relative abundance of aerobic bacteria, anaerobic bacteria, bacteria containing mobile elements, facultative bacteria, biofilm forming bacteria, gram negative bacteria, gram positive bacteria, potentially pathogenic bacteria, and stress tolerant bacteria in CTRL (n=10) vs MET (n=8) exposed offspring.
Figure 3. Prenatal opioid exposure increases sensitivity to thermal and mechanical pain in 3-week-old male offspring.
(a) Latency time to flick in CTRL (n=14) or MET (n=8) 3-week-old offspring. (b) Mechanical threshold in CTRL (n=12) or MET (n=12) 3-week-old offspring assessed using manual von Frey apparatus (c) Schematic for fecal microbial transplantation of prenatally opioid-exposed (MET) or saline-exposed (CTRL) feces into recipient 3-week-old naïve male C57 BL/6 animals. (d) Latency time to flick in CTRL-assigned (n=6), opioid-assigned (n=6), CTRL-FMT (n=6), or opioid-FMT (n=6) 3-week-old offspring. % change in tail flick latency displayed for CTRL-FMT and opioid-FMT groups.
Figure 4. Supplementation with VSL#3 cocktail in dams alters maternal gut microbiome at weaning.
(a) Schematic of maternal VSL #3 probiotic administration in opioid or control dams. (a) Alpha diversity in methadone moms (n=11) or methadone moms receiving probiotics (MethaMompro) (n=6) represented by Shannon indices using pairwise Wilcoxon rank sum test (P>.05 for measured index) (c) PCoA of beta diversity in methadone mom (n=11) and MethaMomPro (n=6) samples based on Bray-Curtis dissimilarity shows distinct clustering (q<0.05) between groups. (d) LEfSe analysis of top discriminative bacteria genera between gut samples from methadone mom (n=11) and MethaMomPro mom (n=6) samples (e) BugBase prediction of the relative abundance of potentially pathogenic, aerobic, bacteria containing mobile elements, facultative bacteria, biofilm-forming bacteria, gram negative, gram positive, and stress tolerant bacteria (p<0.05 for all shown comparisons). (f) KEGG pathway of gut microbiota predicted using PICRUSt. Data are presented in a bar plot with 95% confidence intervals and P values between gut samples from methadone moms and MethaMomPro. (g) alpha diversity in control moms (n=6) or control moms receiving probiotics (ControlMompro) (n=5) represented by Shannon indices using pairwise Wilcoxon rank sum test (P>.05 for measured index) (h) PCoA of beta diversity in control mom (n=6) and ControlMomPro (n=5) samples based on Bray-Curtis dissimilarity shows distinct clustering (q<0.05) between groups. (i) LEfSe analysis of top discriminative bacteria genera between gut samples from control mom (n=6) and ControlMomPro mom (n=5) samples (j) BugBase prediction of the relative abundance of potentially pathogenic, aerobic, bacteria containing mobile elements, facultative bacteria, biofilm-forming bacteria, gram negative, gram positive, and stress tolerant bacteria (p<.05 for all shown comparisons). (k) KEGG pathway of gut microbiota predicted using PICRUSt. Data are presented in a bar plot with 95% confidence intervals and P values between gut samples from control moms and ControlMomPro.
Figure 5. Supplementation with VSL#3 cocktail in dams alters neonatal gut microbiome and rescues hypersensitivity to thermal and mechanical pain in 3-week-old methadone-exposed offspring.
Abundance of (a) Lactobacillus or (b) Bifidobacterium in 3-week-old control or methadone exposed offspring after maternal administration of VSL #3 probiotic cocktail (CTRL+PRO (n=10) or MET+PRO (n=10)). (c) PCoA of beta diversity in CTRL, MET, CTRL+PRO, or MET +PRO samples based on Bray-Curtis dissimilarity shows distinct clustering (q<0.05) among all groups. (d) LEfSe analysis of top discriminative bacteria genera between gut samples from CTRL+PRO or MET+PRO samples. (e) KEGG pathway of gut microbiota predicted using PICRUSt. Data are presented in a bar plot with 95% confidence intervals and P values between gut samples from CTRL+PRO or MET+PRO samples. (f) BugBase prediction of the relative abundance of aerobic, anaerobic, bacteria containing mobile elements, facultative bacteria, biofilm-forming bacteria, gram negative, gram positive, and potentially pathogenic bacteria (p>.05 for all comparisons). (g) Latency time to flick in 3-week-old CTRL or MET exposed offspring whose mothers were gavaged with vehicle (CTRL (n=11), MET (n=9)) or probiotics ((CTRL+PRO (n=15) or MET +PRO (n=13)) assessed by two-way ANOVA, with p<.05 used for statistical significance. (h) Mechanical threshold (per manual von Frey apparatus) in 3-week-old CTRL or MET exposed offspring whose mothers were gavaged with vehicle (CTRL (n=12) or MET (n=12)) or probiotics (CTRL+PRO (n=12) or MET +PRO (n=12)) assessed by two-way ANOVA, with p<.05 used for statistical significance.
Figure 6. Prenatal opioid exposure increases IL-17 in systemic circulation, which is attenuated by maternal probiotic administration.
(a) Serum concentration of IL-17a (pg/mL) in 3-week-old CTRL or MET exposed offspring whose mothers were gavaged with vehicle (CTRL (n=10) or MET (n=10)) or probiotics (CTRL+PRO (n=10) or MET +PRO (n=8)) assessed using the ProcartaPlex mouse cytokine and Chemokine Panel 1, 26-plex. Significance assessed using two-way ANOVA, with p<.05 used for statistical significance. (b) intestinal permeability measured by the concentration of FITC-dextran in serum of adult CTRL (n=5), CTRL (n=5) or MET (n=5) 3-week-old offspring per ANOVA adult control mice represent naïve 12-week-old male mice. P<.05 between adult CTRL vs CTRL and adult CTRL vs MET. (c) Representative H&E- stained section of mouse large intestine from 7 mice/treatment group. (d) Histological score of large sections in CTRL (n=7), MET (n=7), CTRL+PRO (n=7) and MET+PRO (n=7) treated mice. Significance assessed using two-way ANOVA, with p<.05 used for statistical significance.
Figure 7. Maternal opioid exposure alters gene expression profile in whole brain of 3-week-old offspring.
(a) Principal component analysis of CTRL (n=5) and MET (n=5) samples. (b) Volcano plot of all expressed genes with threshold of differential expression with p-value <0.05. 3516 total genes were identified, of which 1671 were upregulated and 1845 were downregulated. (c) Expression heatmap of all differentially expressed genes |log2FoldChange| ≥ 1.5 & p-value ≤.05 (d) Select canonical pathways predicted by Qiagen Ingenuity pathway analysis for significant pairwise differential gene set with |log2FoldChange| ≥ 0.5 & p-value ≤.05. (E) correlation network between significant bacterial taxa between CTRL vs MET with DEG |log2FoldChange| ≥ 0.5 & p-value ≤.05 in pathways belonging to opioid signaling, neuroinflammation signaling, or neuropathic pain signaling in dorsal horns. Ovals indicate taxa (genus or species) and rectangles indicate genes. Upregulated genes or taxa are in red while green represents downregulated genes. Red edges represent a positive correlation and blue represent a negative correlation.
Figure 8. Neonatal brain transcription networks altered by maternal probiotic administration.
(a) Principal component analysis of CTRL (n=5), MET (n=5), CTRL+PRO (n=5), and MET+PRO (n=6) samples. (b) Venn diagram of differentially expressed genes in CTRL, MET, CTRL+PRO, and MET+PRO samples. MET vs CTRL (total DEG: 986: upregulated=585, downregulated=401); MET. PRO vs MET (total DEG: 141: upregulated=64, downregulated=77); CTRL. PRO vs CTRL (total DEG: 476: upregulated=373, downregulated=103); MET.PRO vs CTRL PRO (total DEG: 1468: upregulated=786, downregulated=682). (c) Expression heatmap of differentially expressed genes between CTRL and MET offspring, significantly altered by maternal probiotic treatment in opioid-exposed animals, using |log2FoldChange| ≥ 1.5 & p-value ≤.05 for multigroup analysis. (d) Canonical pathways predicted by Qiagen Ingenuity pathway analysis for significant differential gene set of genes rescued in prenatally opioid exposed offspring with maternal probiotic administration using |log2FoldChange| ≥ 0.5 & p-value ≤0.05.
Figure 9. Cross-fostering opioid-exposed offspring to control dams also rescues hypersensitivity to thermal pain in methadone-exposed offspring.
(a) schematic showing cross fostering strategy. (b-e) alpha diversity in CTRL FCM (n=10), MET FMM (n=8), MET FCM (n=11), or CTRL FMM (n=9) offspring represented by Shannon index using pairwise Wilcoxon rank sum test with p<.05 for statistical significance. P>.05 for all pairwise comparisons, except for CTRL FCM vs CTRL FMM (B) by Simpson index alone (p<.05). (f-i) PCoA of B diversity in CTRL FCM (n=10), MET FMM (n=8), MET FCM (n=11), or CTRL FMM (n=9) samples based on Bray-Curtis dissimilarity show distinct clustering (q<0.05) for all pairwise comparisons. (j) latency time to flick in CTRL FCM (n=10), MET FMM (n=8), MET FCM (n=11), or CTRL FMM (n=9) 3-week-old offspring assessed by two-way ANOVA, with p<.05 used for statistical significance. CTRL FCM, CTRL offspring cross fostered to saline-treated dams, MET FMM, prenatally methadone exposed offspring cross fostered to methadone-treated dams; MET FCM, prenatally opioid exposed offspring cross fostered to control (saline-treated) mothers; CTRL FMM, CTRL offspring cross fostered to methadone-treated dams.
Supplemental material