Transforming berberine into its intestine-absorbable form by the gut microbiota

Abstract

The gut microbiota is important in the pathogenesis of energy-metabolism related diseases. We focused on the interaction between intestinal bacteria and orally administered chemical drugs. Oral administration of berberine (BBR) effectively treats patients with metabolic disorders. However, because BBR exhibits poor solubility, its absorption mechanism remains unknown. Here, we show that the gut microbiota converts BBR into its absorbable form of dihydroberberine (dhBBR), which has an intestinal absorption rate 5-fold that of BBR in animals. The reduction of BBR to dhBBR was performed by nitroreductases of the gut microbiota. DhBBR was unstable in solution and reverted to BBR in intestine tissues via oxidization. Heat inactivation of intestinal homogenate did not inhibit dhBBR oxidization, suggesting the process a non-enzymatic reaction. The diminution of intestinal bacteria via orally treating KK-Ay mice with antibiotics decreased the BBR-to-dhBBR conversion and blood BBR; accordingly, the lipid- and glucose-lowering efficacy of BBR was reduced. Conclusively, the gut microbiota reduces BBR into its absorbable form of dhBBR, which then oxidizes back to BBR after absorption in intestine tissues and enters the blood. Thus, interaction(s) between the gut microbiota and orally administrated drugs may modify the structure and function of chemicals and be important in drug investigation.

Figure 1. BBR is metabolized into dhBBR in the intestine ecosystem.

(a) In vivo BBR metabolites in SD rats; M17 (m/z 337) was identified to be dihydroberberine (dhBBR), a BBR metabolite detected in the feces only; the parentheses indicate the sample in which the compound was detected. (b) Excretion of BBR, dhBBR and other BBR metabolites in rat feces, urine and bile after 72 h. (c) Distribution of dhBBR in organs of SD rats orally treated with BBR (200 mg/kg); ND: not detectable. (d) Conversion of BBR into dhBBR by the gut microbiota in vitro; U & insert: in rat large intestinal bacteria (RLIB); M & insert: in rat small intestinal bacteria (RSIB); L & insert: in human intestinal bacteria (HIB) in vitro. The Y axis shows the percentage of the administered amount of BBR (n = 6).

Figure 2. Generation of dhBBR by the gut microbiota.

(a) BBR was converted in vitro into dhBBR by 14 intestinal bacteria strains [S. aureus 08-43 (1), E. faecium 13-01 (2), E. faecalis 13-01 (3), E. cloacae 13-12 (4), E. coli 06-05 (5), S. epidermidis 12-12 (6), Ps. aeruginosa 13-10 (7), K. pneumoniae 13-14 (8), P. mirabilis 13-01 (9), A. baumannii 13-02 (10), and L. casei (11), L. acidophilus (12), B. longum (13), and B. breve (14)]. L: dhBBR was produced at different levels by the 14 intestinal bacteria strains (*P < 0.05 and **P < 0.01, n = 6); R: Determination of nitroreductase in the 14 intestinal bacteria strains (***P < 0.001, *P < 0.05, n = 6). (b) Oral treatment with antibiotics generated pseudo germ-free (PGF) rats and resulted in a reduced conversion of BBR to dhBBR in the rat intestinal ecosystem; L: dhBBR production was delayed and decreased in the feces of PGF rats compared with conventional SD rats; L insert: oral treatment with antibiotics for 3 days reduced the bacterial colony numbers by 2 logs in the rat feces (***P < 0.001, n = 6); M: the accumulative excretion of BBR and dhBBR in PGF and conventional rats over a period of 72 h (**P < 0.01, *P < 0.05, n = 6); R: MIC of BBR and dhBBR on the 17 intestinal bacteria (24 h); The Y axis shows the percentage of the administered amount of BBR. (c) Molecular docking between BBR and nitroreductase, showing the chemical mechanism of BBR reduction by nitroreductase. (d) Nitroreductase-mediated BBR reduction resulted in a BBR-to-dhBBR conversion after 4 h; insert: BBR-to-dhBBR conversion was reduced in the presence of the nitroreductase inhibitor 2-IBA (100 μM) in the nitroreductase-containing cell-free incubation mixture or in the rat gut bacteria cultivation mixture (for 12 h); ND: not detectable.

Figure 3. DhBBR has a better intestinal absorption than BBR.

(a) Absorption of BBR and dhBBR in the Caco-2 cell model; L: The P_{app (AP-BL)} of dhBBR in Caco-2 cells was 11.9-fold higher than that of BBR (***P < 0.001); R: The efflux ratio of dhBBR in Caco-2 cells was significantly lower than that of BBR (1.58 vs. 32.39, ***P < 0.001). (b) Concentration-time curve for BBR or dhBBR in plasma after dhBBR oral administration (200 mg/kg) to rats (n = 3). (c) Concentration-time curve of BBR in plasma after the oral administration of dhBBR (200 mg/kg) or BBR (200 mg/kg) (n = 3). (d) Concentration-time curve of BBR in pseudo germ-free rats (generated by the oral administration of antibiotics for 3 days, curve 1) or in conventional SD rats (with no antibiotics, curve 2) after the oral administration of BBR (200 mg/kg, n = 3).

Figure 4. DhBBR reverted to BBR via oxidization in intestine tissue.

(a) dhBBR-to-BBR reversion in the rat small intestine homogenate (L) and in human intestine microsomes (HIMs, R) (the Y axis shows the percentage of administered amount of dhBBR); ND: not detectable. (b) DhBBR-to-BBR reversion by monoamine oxidase-B (MAO-B) in a cell-free system (insert) and homogenates of the rat duodenum, jejunum and ileum; the reversion was slightly inhibited by the MAO-B inhibitor deprenyl (100 μM) or pargyline (1 μM) (the Y axis shows the percentage of administered amount of dhBBR). (c) Addition of vitamin C into the homogenate almost terminated the oxidization reaction for reverting dhBBR to BBR, and the content of superoxide anion decreased subsequently (the left Y axis shows the percentage of administered amount of dhBBR; the right Y axis shows the percentage of superoxide anion of control). (d) Effect of metal ions (Fe³⁺, Cu²⁺, and Zn²⁺) on catalyzing the oxidization of dhBBR to BBR.

Figure 5. Gut microbiota modulates the therapeutic effect of BBR in vivo.

(a) Fasting blood glucose (Glu), triglyceride (TG) and cholesterol (CHO) in conventional KK-Ay mice or pseudo germ-free KK-Ay mice (n = 6) treated with (or without) BBR for 14 days; U: plasma glucose; M: plasma triglyceride; L: plasma cholesterol. (b) Number of intestinal bacteria colonies (in log scale) on day 14 (n = 6). (c) Plasma concentration of BBR in KK-Ay mice treated or untreated with antibiotics; U: C_max of BBR on day 14; L: folds of C_max[BBR/(BBR+antibiotics)] on days 1, 3, 6, 10 and 14 post-BBR oral administration. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 6. Proposed mechanism of BBR absorption in the intestine.