Gut microbiome modulates tacrolimus pharmacokinetics through the transcriptional regulation of ABCB1

Abstract

Background: Following solid organ transplantation, tacrolimus (TAC) is an essential drug in the immunosuppressive strategy. Its use constitutes a challenge due to its narrow therapeutic index and its high inter- and intra-pharmacokinetic (PK) variability. As the contribution of the gut microbiota to drug metabolism is now emerging, it might be explored as one of the factors explaining TAC PK variability. Herein, we explored the consequences of TAC administration on the gut microbiota composition. Reciprocally, we studied the contribution of the gut microbiota to TAC PK, using a combination of in vivo and in vitro models.

Results: TAC oral administration in mice resulted in compositional alterations of the gut microbiota, namely lower evenness and disturbance in the relative abundance of specific bacterial taxa. Compared to controls, mice with a lower intestinal microbial load due to antibiotics administration exhibit a 33% reduction in TAC whole blood exposure and a lower inter-individual variability. This reduction in TAC levels was strongly correlated with higher expression of the efflux transporter ABCB1 (also known as the p-glycoprotein (P-gp) or the multidrug resistance protein 1 (MDR1)) in the small intestine. Conventionalization of germ-free mice confirmed the ability of the gut microbiota to downregulate ABCB1 expression in a site-specific fashion. The functional inhibition of ABCB1 in vivo by zosuquidar formally established the implication of this efflux transporter in the modulation of TAC PK by the gut microbiota. Furthermore, we showed that polar bacterial metabolites could recapitulate the transcriptional regulation of ABCB1 by the gut microbiota, without affecting its functionality. Finally, whole transcriptome analyses pinpointed, among others, the Constitutive Androstane Receptor (CAR) as a transcription factor likely to mediate the impact of the gut microbiota on ABCB1 transcriptional regulation.

Conclusions: We highlight for the first time how the modulation of ABCB1 expression by bacterial metabolites results in changes in TAC PK, affecting not only blood levels but also the inter-individual variability. More broadly, considering the high number of drugs with unexplained PK variability transported by ABCB1, our work is of clinical importance and paves the way for incorporating the gut microbiota in prediction algorithms for dosage of such drugs. Video Abstract.

Keywords: Cytochrome P-450 CYP3A; Immunosuppressive therapy; Inter-individual pharmacokinetic variability; Intra-individual pharmacokinetic variability; Narrow therapeutic index; Therapeutic drug monitoring.

Fig. 1

Determination of tacrolimus (TAC) pharmacokinetics (PK) profile in mice. A Blood concentration of TAC according to the dose (n = 10/group). B TAC blood concentration–time curve at the selected dose of 3 mg/kg (24 mice providing each 3 measurements, n = 12 measurements/time points covering 6-time points, as described in the Supplementary Methods)

Fig. 2

Faecal microbiota composition evolves upon tacrolimus (TAC) treatment. A, B Principal component analysis (PCA) of the relative abundances of taxa in control (CTL) and TAC-treated mice. Data are presented for the genus A and the family B taxonomic levels before and after 2 and 5 days of oral gavage. C, D Relative abundance of genera (C) and families (D) identified as significantly impacted by TAC treatment (n = 7–8/group). **p < 0.01, *p < 0.05 (Mann–Whitney, CTL vs TAC). ^#p < 0.05 (Friedman test with Dunn post hoc test, TAC repeated measures)

Fig. 3

Antibiotic (ATB)-mediated gut microbiota depletion affects tacrolimus (TAC) pharmacokinetics (PK) and reduces TAC blood exposure. A TAC blood concentration in control (TAC) or ATB-treated mice (TAC + ATB). B TAC area under the curve (AUC) computed with the PK model for both mouse groups (n = 10/group). **p < 0.01

Fig. 4

Abcb1a expression in the small intestine correlates with tacrolimus (TAC) whole blood exposure. A Comparison of the mRNA expression of Abcb1a in the proximal, median and distal small intestine; in the colon; and in the liver of control and ATB-treated mice, with or without TAC treatment (n = 7–8/group). ^#significant ATB effect; ^$significant TAC effect; ***p < 0.001, **p < 0.01, and *p < 0.05. B Pearson’s correlation between TAC area under the curve (AUC) and Abcb1a mRNA expression level in the different tissues

Fig. 5

Abcb1a expression level is microbiota-dependent in the small intestine. Comparison of the mRNA expression of Abcb1a in the proximal, median and distal small intestine; in the colon; and in the liver of conventionalized (CVZ) and germ-free (GF) mice (n = 4–5/group). ***p < 0.001 and **p < 0.01

Fig. 6

ABCB1A inhibition reverses the gut microbiota effect on tacrolimus (TAC) pharmacokinetics (PK). A TAC blood concentration in control (TAC) or antibiotic (ATB)-treated (TAC + ATB) mice under ABCB1 inhibitor, zosuquidar (ZSQ). ***p < 0.001. B TAC area under the curve (AUC) derived from the PK model in TAC + ZSQ or TAC + ATB + ZSQ mice (n = 10–11/group). ***p < 0.001

Fig. 7

Modulation of ABCB1A efflux activity by the gut microbiota explains differences in tacrolimus (TAC) blood exposure. TAC area under the curve (AUC) derived from the PK model in mice under control (TAC), zosuquidar (TAC + ZSQ), antibiotics (TAC + ATB), or ZSQ with ATB (TAC + ATB + ZSQ) conditions (n = 10–36/group). ***p < 0.001, **p < 0.01; ^###p < 0.001 (TAC + ZOSU vs TAC + ATB + ZOSU); ^$$$p < 0.001 (TAC + ATB vs TAC + ATB + ZOSU)

Fig. 8

Polar bacterial metabolites impact the transcriptional regulation of ABCB1, but not directly its functionality. A ABCB1 substrate (rhodamine 123) intracellular accumulation (in relative fluorescent units, r.f.u.) with or without a 15 min pre-exposition to mouse faecal water (FW) in control plasmid or in ABCB1-transfected cells (n = 6/group, N = 2). ***p < 0.001. B Comparison of the mRNA expression of ABCB1 in control Caco-2 cells (PBS) or cells exposed to mouse FW for 48 h (n = 3/group, N = 5). ***p < 0.001. C Comparison of the mRNA expression of ABCB1 in control Caco-2 cells (PBS) or cells exposed for 48 h to FW from untreated mice (FWctl), or to FW from ATB-treated mice (FWatb), or to antibiotics (ATB) (n indicated on the figure, for FW n = 7 biological replicates). *p < 0.05 and.^###p < 0.001 (paired t-test between FWctl and Fwatb)

Fig. 9

The nuclear receptor CAR may mediate the effect of the microbiome on ABCB1 expression. Whole transcriptome analysis of control Caco-2 cells (PBS) or cells exposed for 48 h to FW from untreated mice (FWctl), or to FW from ATB-treated mice (FWatb), or to antibiotics (ATB) (n = 4–7/group). A Analysis of the transcription factors (TF) known to have inductive (INDUCERS, yellow box), variable (white box) or repressive (REPRESSORS, orange box) effects on ABCB1 expression. On the left, rho value is represented for Spearman’s correlation between the expression of ABCB1 and each of the TF. ***p < 0.001, **p < 0.01, *p < 0.05. On the right, heatmap showing the log twofold change (LogFC) for each condition compared to the control (PBS). *adjusted p < 0.05. B Comparison of the mRNA expression of CAR in these different conditions. *p < 0.05, ^###p < 0.001 (paired t-test between FWctl and Fwatb). C Spearman’s correlation between CAR and ABCB1 mRNA expression levels in these conditions

Fig. 10

mRNA expression of CAR correlates with Abcb1a levels in mouse intestine. In mice, the transcription factor CAR is encoded by the Nr1i3 gene. A Comparison of the mRNA expression of Nr1i3 in the proximal, median and distal small intestine; in the colon; and in the liver of control and ATB-treated mice, with or without TAC treatment (n = 7–8/group). ^#significant ATB effect; ^$significant TAC effect; ***p < 0.001, **p < 0.01, and *p < 0.05. B Spearman’s correlation between Nr1i3 and Abcb1a mRNA expression levels in the different tissues

Fig. 11

Schematic summary of the bidirectional interactions between tacrolimus (TAC) and the gut microbiota. After oral administration, TAC diffuses passively in the small intestine but the efflux transporter ABCB1A (encoded by Abcb1a in mice) limits its absorption. Some of the drug reaches the blood circulation and is then metabolized in the liver by CYP3A enzymes (encoded by Cyp3a11 and Cyp3a13 in mice). In the lumen, the gut microbiota produces compounds/metabolites. Some of them have a repressive effect on Abcb1a expression (red arrow), resulting in less efflux activity. Moreover, TAC impacts the gut microbiota composition (blue arrow). Created with BioRender.com