Bmal1 regulates circadian expression of cytochrome P450 3a11 and drug metabolism in mice

Bmal1 knockout mice were generated by deleting a portion of exon5 using the CRISPR/Cas9 technique (Supplementary Fig. 2), and validated by PCR genotyping and expression profiling (Fig. 1a−c and Supplementary Fig. 3). As expected, hepatic and intestinal Cyp3a11 showed a robust circadian expression in wild-type mice (Fig. 1d, e and Supplementary Fig. 4A). However, global deletion of Bmal1 decreased Cyp3a11 expression, and blunted its circadian rhythmicity in both liver and small intestine (Fig. 1d, e and Supplementary Fig. 4A). Bmal1-deficient mice showed reduced metabolic activities toward testosterone and midazolam (two known Cyp3a11 substrates) consistent with the expression changes of Cyp3a11 (Fig. 1d, e). Likewise, Bmal1 knockdown led to downregulation of Cyp3a11/CYP3A4 in various hepatoma and colon carcinoma cell lines (i.e., Hepa1-6, Hepa-1c1c7, CT26, HepG2, and Caco-2) (Fig. 1f, g). Additionally, overexpression of Bmal1 caused increases in Cyp3a11 mRNA and protein in Hepa1-6 cells (Fig. 1h and Supplementary Fig. 4B). Bmal1 induced the promoter activity of a −2.0 kb Cyp3a11 reporter in the luciferase reporter assays (Fig. 1i). Taken together, these data supported a critical role for Bmal1 in circadian regulation of Cyp3a11.

No canonical E-box element (a DNA motif for Bmal1 binding and transactivation) was found in the promoter region of mouse Cyp3a11 based on in silico sequence analysis (Genomatix program). Therefore, an indirect rather than direct mechanism was involved in Bmal1 regulation of Cyp3a11. Interestingly, of six known Cyp3a11 regulators, Dbp and Hnf4α mRNAs were decreased in mouse liver and small intestine as a result of Bmal1 deletion (Fig. 2a, b). In line with the mRNA changes, Bmal1 deficiency caused reductions in Dbp and Hnf4α proteins (Fig. 2c, d and Supplementary Fig. 5). In addition, Bmal1 knockdown resulted in reduced expressions of Dbp and Hnf4α in both Hepa-1c1c7 and CT-26 cells (Fig. 2e, f). These data suggested potential roles of Dbp and Hnf4α in Bmal1 regulation of Cyp3a11. It was noted that the expressions of the other two PAR bZip genes Tef and Hlf were unaffected by Bmal1 deletion (Supplementary Fig. 6).

Overexpression of Bmal1 or Dbp or Hnf4α led to an increased level of Cyp3a11 mRNA in Hepa-1c1c7 cells (Fig. 3a). Luciferase reporter assays revealed a Cyp3a11 promoter region of −220/+78 bp as a key determinant to Bmal1 responsiveness (Fig. 3b). A noncanonical E-box (−1488/−1482 bp) was found in Cyp3a11 promoter. However, mutation of this E-box did not alter the effect of Bmal1 on the Cyp3a11 reporter (−2.0 kb) activity (Fig. 3c). ChIP assays showed no recruitment of Bmal1 protein to this “E-box” (Fig. 3d), supporting that the noncanonical E-box was nonfunctional. In-depth sequence analysis of the −2.0 kb promoter identified a DR1 (−123/−110 bp) and a D-box (−45/−36 bp) motif to which Hnf4α and Dbp might bind, respectively (Fig. 3e). Recruitment of Hnf4α to the DR1 motif (−123/−110 bp) of Cyp3a11 was confirmed by a ChIP assay (Fig. 3f). Dbp enhanced the promoter (−2.0 kb) activity of Cyp3a11 in luciferase reporter assays (Fig. 3g). However, this activation effect was lost when the D-box of −45/−36 bp was mutated (Fig. 3g). Similarly, Hnf4α increased the Cyp3a11 reporter activity, and this effect was abolished when the DR1 of −123/−110 bp was mutated (Fig. 3h and Supplementary Fig. 7). By contrast, a mutation of D-box or DR1 alone attenuated but was unable to abrogate Bmal1-mediated activation of Cyp3a11 reporter (Fig. 3i). A dual mutation of D-box and DR1 eliminated the responsiveness of Cyp3a11 reporter to Bmal1 (Fig. 3i). Knockdown of Dbp or Hnf4α alone (by specific siRNA) caused a reduction in Cyp3a11 mRNA in Hepa-1c1c7 cells (Fig. 3j). The knockdown efficiencies of siDbp and siHnf4α were verified by western blotting (Supplementary Fig. 8). The extent of Cyp3a11 mRNA reduction was more significant when both were knocked-down (Fig. 3j). Moreover, the effect of Bmal1 on Cyp3a11 was attenuated when Dbp and/or Hnf4α were silenced (Fig. 3j). Also, Bmal1 ablation led to reductions in respective recruitments of Hnf4α and Dbp proteins to the DR1 (−123/−110 bp) and D-box (−45/−36 bp) elements of Cyp3a11 promoter (Fig. 3k), supporting the mediating roles of Hnf4α and Dbp in Bmal1 regulation of Cyp3a11. Taken together, Bmal1 regulated Cyp3a11 expression through Dbp and Hnf4α.

Bmal1 induced Dbp transcription in luciferase reporter assays (Fig. 4a). ChIP experiments showed significant Bmal1 recruitment to the E-box site of Dbp (Fig. 4b). These data confirmed Bmal1 as a transcriptional activator of Dbp²⁶. Chromatin immunoprecipitation sequencing (ChIP-seq) revealed that Bmal1 bound to the distal region (from −6.1 to −6.0 kb) of the P1 promoter of Hnf4α (Fig. 4c). In luciferase reporter assays, Bmal1 efficiently induced Hnf4α P1 promoter (−6.1 kb) activity (Fig. 4d). The induction effect was lost when the distal region (from −6.1 kb to −6.0 kb) was truncated (Fig. 4d). Sequence analysis identified three potential E-box motifs (named E1, E2 and E3) within the 100-bp length of P1 promoter. EMSA assay further indicated direct binding of both E1 and E3 to Bmal1 protein in Bmal1-overexpressing Hepa1-6 cells (Fig. 4e and Supplementary Fig. 9). However, no DNA−protein complex bands were generated when using nontransfected and Bmal1-lacking cells instead (Fig. 4e, f). ChIP assays confirmed that Bmal1 bound to the distal region of Hnf4α P1 promoter (Fig. 4g). Overall, our data support Bmal1 as a transcriptional activator of both Dbp and Hnf4α (Fig. 4h).

Aconitine is a bioactive (and toxic) component of Aconitum carmichaelii Debeaux, a widely used herb medicine. Cyp3a11/CYP3A4 is the main enzyme responsible for aconitine metabolism and detoxification, generating four metabolites (Fig. 5a and Supplementary Fig. 10A/B)²⁷. A high metabolic activity of Cyp3a11 toward aconitine was also confirmed using recombinant Cyp3a11 enzyme (Fig. 5b). Interestingly, the toxicity of aconitine (i.p. at 0.3 mg/kg or p.o. at 1.8 mg/kg) was dosing-time-dependent with a higher toxicity at daytime (ZT2/8) than at night (ZT14/20) (Fig. 5c, d). It was noted that a lower value of PR interval derived from ECG recordings is associated with high risks for preexcitation syndrome and supraventricular tachycardia²⁸. This agreed well with the circadian protein expression and microsomal activity (Fig. 1d, e) of Cyp3a11 (higher levels at ZT14/20 and lower levels at ZT2/8). However, the time dependence of aconitine toxicity ceased to exist in Bmal1-deficient mice (Fig. 5e) consistent with blunted rhythms of Cyp3a11 expression and activity (Figs. 1d, e and 5f). Moreover, compared with wild-type, Bmal1-deficient mice showed higher levels of toxicity irrespective of dosing-time (Fig. 5e). This was probably associated with reduced metabolism of aconitine in Bmal1-deficient mice based on in vitro microsomal metabolism assays and pharmacokinetic analyses (Fig. 5f, g).

Triptolide is a bioactive (and toxic) component of Tripterygium wilfordii Hook F, an herbal medicine for treatment of leukemia and inflammation. Cyp3a11/CYP3A4 is the major enzyme contributing to triptolide metabolism and detoxification (Fig. 6a and Supplementary Fig. 10C&D)²⁹. We confirmed that Cyp3a11 was highly active in catalyzing triptolide metabolism, generating two hydroxylated metabolites (Fig. 6b). Similar to aconitine, triptolide toxicity showed dosing-time dependence with a more severe toxicity at daytime (ZT2/8) than at night (at ZT14/20). The diurnal variation in triptolide toxicity (Fig. 6c, d) was correlated with the circadian expression of Cyp3a11 (Fig. 1d, e). Bmal1 ablation resulted in a higher level of toxicity, and blunted its rhythmicity (Fig. 6c, d). This was accompanied by reduced levels of and dampened rhythm of triptolide metabolism based on microsomal metabolism assays and pharmacokinetic analysis (Fig. 6e, f). Taken together, Bmal1 modulates the chronotoxicity of aconitine and triptolide through regulation of rhythmic Cyp3a11.

Hepa1-6 and HepG2 cells were purchased from American Type Culture Collection (Manassas, VA). Hepa1c1c7, CT-26, and Caco-2 cells were purchased from the cell bank of Chinese Academy of Sciences (Shanghai, China). The biochemical assay kits for CK, MDA, ALT and AST were purchased from Jiancheng Bioengineering Institute (Nanjing, China). BCA assay kit, cytoplasmic/nuclear protein extraction kit, and EMSA kit were purchased from Beyotime (Shanghai, China). Midazolam and NADPH were purchased from Sigma-Aldrich (St. Louis, MO). JetPrime transfection kit was purchased from POLYPLUS Transfection (Ill kirch, France). ChIP kit was purchased from Cell Signaling Technology (Beverly, MA). RNAiso Plus reagent and PrimeScript RT Master Mix were purchased from Takara (Shiga, Japan). ChamQ Universal SYBR qPCR Master Mix was purchased from Vazyme (Nanjing, China). Dual-Luciferase® Reporter Assay System was purchased from Promega (Madison, WI). Aconitine was purchased from MCE (MedChem Express, New Jersey). Quercetin, ketoconazole, vitamin K3, quinidine, 4-methylimidazole, testosterone and triptolide were purchased from Aladdin (Shanghai, China). pGL4.10 (see supplementary Fig. 1A for map) and pRL-TK vectors (see supplementary Fig. 1B for map) were purchased from Promega (Madison, WI). Cyp3a11-overexpressing HEK293 (named HEK3a11) cells, Dbp (2 kb)-Luc, pcDNA 3.1, pcDNA-Bmal1, pcDNA-Dbp, pcDNA-Hnf4α P1, and pcDNA-BMAL1 were obtained from Biowit Technologies (Shenzhen, China). The antibodies used in Western blotting were as follows: anti-Bmal1 (ab3350, Abcam, Cambridge, MA), anti-Cyp3a11 (ab3572, Abcam, Cambridge, MA), anti-Dbp (ab22824, Abcam, Cambridge, MA), anti-Hnf4α (TA321134, OriGene Technologies, Rockville, MD), and anti-Gapdh (ab9484, Abcam, Cambridge, MA). All the primary antibodies were diluted with 5% BSA at the ratio of 1:1000. The final concentration of antibodies was 1 μg/ml. The secondary antibody was purchased from Huaan Biotechnology and diluted with 5% skim milk at the ratio with 1:5000 (Hangzhou, China). For ChIP assay, anti-Bmal1 antibody was purchased from Abcam (Cambridge, MA) and normal rabbit IgG from Cell Signaling Technology (Beverly, MA). siRNA and negative control (siNC) were synthesized by Transheep Technologies (Shanghai, China).

Cyp3a11 (2 kb)-Luc, Cyp3a11 (1.34 kb)-Luc, Cyp3a11 (0.85 kb)-Luc, Cyp3a11 (0.22 kb)-Luc, Hnf4α-P1 (6.1 kb)-Luc, Hnf4α-P1 (5.9 kb)-Luc, Hnf4α-P1 (2.0 kb)-Luc, Hnf4α-P2 (2.0 kb)-Luc, Cyp3a11 (2 kb)-Luc containing mutated E-box, D-box or DR1 site, and Cyp3a11 (2 kb)-Luc containing double mutated D-box and DR1 sites were obtained from Transheep Technologies (Shanghai, China). The gene promoters were amplified by PCR reactions using specific primers (Supplementary Table 1). PCR amplification procedures consisted of an initial denaturation at 95 °C for 3 min, followed by 35 cycles of denaturation at 95 °C for 15 s, annealing at 60 °C for 30 s, and extension at 72 °C for 30 s, and a final extension at 72 °C for 5 min. Cyp3a11 promoter sequences were cloned into XhoI and BglII sites of pGL4.1 vector, and Hnf4α promoters into the XhoI and HindIII sites of pGL4.1 vector. All plasmids were verified by DNA sequencing. After transformed into E. coli JM109, plasmids were isolated using EasyPure HiPure Plasmid MiniPrep kits (TransGen Biotech, Beijing, China) according to the manufacturer’s instructions.

All wild-type (C57BL/6J) and Bmal1^−/− mice (on a C57BL/6J background) were maintained on a12 h L:12 h D cycle (light on 7:00 a.m. to 7:00 p.m.), with free access to food and water. Bmal1^−/− mice were generated by deleting a portion of exon 5 (Supplementary Fig. 2). Genotype of Bmal1^−/− mice was verified in our previous publication⁴⁹. Bmal1^−/− mice showed a disrupted rhythmicity in wheel-running activity. The genetically modified mice were reproduced through breeding of heterozygous animals at expected Mendelian frequencies with no gender bias, and phenotypically normal in terms of body weight and survival at an age of ≤12 weeks. Experiments were planned and performed in accordance with current “3Rs” and ARRIVE guidelines and approved by Jinan University Institutional Animal Care and Use committee. We also applied the principles of the ARRIVE guideline for data management and interpretation, and all efforts were made to minimize the number of animals used and their suffering. Mice were randomly allocated into equal experimental groups based on body weight. Sample size was determined according to preliminary experiments, and experimenters were blinded for analysis purposes. For the analyses of mRNA, proteins and microsomal CYP enzyme activity, wild-type or Bmal1^−/− mice (8–12 weeks of age, male, n = 5 per group) were sacrificed every 4 h (at ZT2, ZT6, ZT10, ZT14, ZT18 and ZT22, whereby ZT0 is defined as sunrise), and livers and small intestines were harvested.

Toxicological studies for aconitine and triptolide were performed with wild-type and Bmal1^−/− mice as previously described⁵⁰. The doses for aconitine and triptolide were 0.3 and 0.6 mg/kg, respectively. Plasma MDA (malonaldehyde), CK (creatine kinase), ALT (alanine aminotransferase), and AST (aspartate aminotransferase) levels were measured using the biochemical assay kits (Jiancheng Bioengineering Institute, Nanjing, China). Electrocardiogram (ECG) recordings of C57BL/6 mice treated with aconitine (i.p. at 0.3 mg/kg) at ZT2, ZT8, ZT14 or ZT20 were monitored using Dual Bio Amp/Stimulator (AD instruments, Sydney, Australia).

For pharmacokinetic studies, the drug (acontine (0.3 mg/kg) and triotplide (0.6 mg/kg)) was administered to wild-type or Bmal1^−/− mice (8–12 weeks of age, male) by intraperitoneal injection at ZT2 or ZT14. At each time point, five mice were rendered unconscious with isoflurane for blood sampling. After centrifugation, the plasma samples were collected and subjected to LC/MS/MS analysis.

Hepa1-6, Hepa-1c1c7, CT-26, HepG2 and/or Caco-2 cells were used in the gene overexpression and knockdown experiments. Luciferase reporter, electrophoretic mobility shift and chromatin immunoprecipitation assays were based on Hepa1-6 cells. Hepa-1c1c7, CT26 and Caco-2 cells were cultured in DMEM medium supplemented with 10%FBS, Hepa1-6 cells in RPMI 1640 supplemented with 10%FBS, and HepG2 cells in the α-MEM medium supplemented with 10%FBS. Once reaching a confluence of 60–70%, the cells were transfected with indicated plasmids using a JetPrime transfection kit according to the manufacturer’s protocol. Six hours later, the culture medium was replaced with fresh medium. On the next day, the cells were collected for further experiments.

DNA was extracted from mouse tail (1–2 mm). PCR was performed with 400 ng DNA template. The products were subjected to the 2% agarose gel electrophoresis, and the bands were imaged using the Omega LumTM G imaging system (Aplegen). The primers are listed in Supplementary Table. Bands of 677 bp and 573 bp indicate wild-type and homozygous, respectively. 2

qPCR assays were performed as previously described¹³. In brief, total RNA was extracted from mouse tissues and cells using RNAiso Plus reagent according to the manufacturer’s instructions. 500 ng RNA was used as a template for cDNA synthesis using the PrimeScript RT Master Mix. The sequences of all primers applied to qPCR are listed in Supplementary Table 2. PCR amplification reactions consisted of an initial denaturation at 95 °C for 5 min, followed by 40 cycles of denaturation at 95 °C for 15 s, annealing at 60 °C for 30 s, and extension at 72 °C for 30 s. Mouse 18s RNA or human GAPDH was used as an internal control. Gene expression was calculated using the 2^−ΔΔCT method. The primers (listed in Supplementary Table 2) were obtained from Ruiboxingke Biological Technology (Beijing, China). Primer specificity was checked by BLAST analysis (http://www.ncbi.nlm.nih.gov/BLAST/↗) and by performing a melting curve analysis. All qPCR assays were performed according to the MIQE (Minimum Information for Publication of qPCR experiments) guideline⁵¹.

Mouse tissues or cells were lysed in RIPA buffer containing 1 mM PMSF. Protein concentrations were detected using a BCA assay kit. The samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (10% acrylamide gels), and transferred onto PVDF membranes. After blocking with 5% skimmed milk for 1 h, the membranes were sequentially probed with the primary antibodies and horseradish peroxidase-conjugated secondary antibody. Protein bands were detected by enhanced chemiluminescence and imaged by the Omega Lum^TM G imaging system (Aplegen). The densities of protein bands were analyzed by Fluorchem 5500 software. Uncropped scans of representative blots are provided in Supplementary Fig. 14.

Mouse liver/intestine microsomes and HEK3a11 cell lysate were prepared as previously described^13,52. In vitro metabolic activity was determined using the published incubation procedures⁵³. After 2-h-incubation at 37 °C, the reactions were terminated by adding ice-cold acetonitrile. The enzymatic activity was determined by LC/MS/MS analysis of the generated metabolites from substrates. Incubation conditions are shown in Supplementary Table 3. Preliminary experiments were performed to ensure that the rates of metabolism were determined under linear conditions with respect to incubation time and protein concentration.

Drugs and their metabolites were quantified using an LC/MS/MS system as previously described⁵⁰.

Luciferase reporter assays were performed as previously described¹⁴. In brief, the cells were transfected with 250 ng of (Hnf4α or Cyp3a11) luciferase reporter plasmid, 50 ng of pRL-TK vector, and/or 250 ng (Bmal1) overexpression plasmid. After 6 h, the medium was changed to fresh medium. On the next day, the medium was removed and the cells were collected. The cells were lysed with 45 μl passive lysis buffer (Promega) and the luciferase activities were detected using the Dual-Luciferase® Reporter Assay system. Relative luciferase activity was expressed as the ratio of firefly over renilla luciferase activity.

ChIP assays were performed using the Enzymatic Chromatin IP kit as previously described¹³. In brief, mouse livers were fixed in 1% formaldehyde for 20 min at room temperature. The reactions were terminated by the addition of glycine. After 5 min centrifugation, the livers were collected and homogenized by using a Dounce homogenizer. Cells were incubated with ChIP buffer containing micrococcal nuclease. After addition of 0.5 M EDTA, the precipitate was suspended in ChIP buffer and subjected to sonication. An aliquot of sheared chromatin was used as Input sample, whereas the remaining was immunoprecipitated with anti-Bmal1, anti-Hnf4α or anti-Dbp overnight at 4 °C. On the next day, the protein-DNA complex was incubated with protein G-coupled Dynabeads (30 μl) at 4 °C for 2 h. The immune complex was separated from beads and eluted in ChIP buffer. The immunoprecipitated chromatin was decross-linked at 65 °C for 4 h and protein was digest with proteinase K. The purified DNA was amplified by real-time PCR with specific primers (Supplementary Table 4).

Nuclear proteins from the cells were first extracted using a cytoplasmic/nuclear protein extraction kit. EMSA assays were performed with a chemiluminescent EMSA kit as previously described¹⁴. In brief, 6-μg nuclear proteins were mixed with EMSA binding buffer in the presence or absence of 12.5 nmol unlabeled probes (or unlabeled mutated probes). The mixture was incubated at room temperature for 10 min. Then, the biotin-labeled probes (0.5 nmol) were added into the mixture and incubated at room temperature for 20 min. The DNA−protein complexes were loaded onto 4% nondenaturing polyacrylamide gels. After 30-min electrophoresis, the products were transferred onto Hybond-N + membranes for 45 min. The membranes were visualized by using enhanced chemiluminescence and Omega Lum G imaging system (Aplegen, Pleasanton, CA). The primers are listed in Supplementary Table 5.

Data are presented as mean ± SD (standard deviation). Statistical analyses on time-dependent data were performed using a two-way analysis of variance (ANOVA), followed by a post-hoc Bonferroni test (a post-hoc test was run only when a significant omnibus F-test was obtained) using GraphPad Prism software version 7 (GraphPad Software Inc., San Diego, CA). The homogeneity of variance assumption for ANOVA was verified by Levene’s test. Statistical differences between two groups were analyzed by Student’s t test. Circadian parameters were obtained through cosine curve fitting using the cosinor periodogram program (Boise State University, Boise, ID). The level of significance was set at P < 0.05 (*).

Further information on research design is available in thelinked to this article. Nature Research Reporting Summary

The data that support the findings of this study are available from the authors on reasonable request. The source data underlying plots are presented in Supplementary Data. Full blots are shown in Supplementary Information. 1