Identification of three novel natural product compounds that activate PXR and CAR and inhibit inflammation

Abstract

Purpose: To investigate the effects of three natural product compounds, carapin, santonin and isokobusone, on the activity of pregnane X receptor (PXR) and constitutive androstane receptor (CAR) in induction of drug-metabolizing enzymes and inhibition of inflammation.

Methods: The monkey kidney-derived fibroblast (CV-1) cells and human embryonic kidney HEK293 cells were used for transient transfection and luciferase reporter gene assays. Human primary hepatocytes and primary hepatocytes from wild type, PXR-/-, and hPXR transgenic mice were used to study the induction of drug-metabolizing enzymes and the implication of these compounds in inflammation.

Results: Carapin, santonin and isokobusone activated both PXR and CAR in transient transfection and luciferase reporter gene assays. Mutagenesis studies showed that two amino acid residues, Phe305 of the rodent PXR and Leu308 of the human PXR, are critical for the recognition of these compounds by PXR. Importantly, the activation of PXR and CAR by these compounds induced the expression of drug-metabolizing enzymes in primary human and mouse hepatocytes. Furthermore, activation of PXR by these compounds inhibited the expression of inflammatory mediators in response to lipopolysaccharide (LPS). The effects of these natural compounds on drug metabolism and inflammation were abolished in PXR-/- hepatocytes.

Conclusions: Our results show that carapin, santonin and isokobusone activate PXR and CAR and induce drug-metabolizing enzymes. In addition, these compounds inhibited the expression of inflammatory mediators in response to LPS through the activation of PXR.

Figure 1. Carapin, santonin and isokobusone activated the human and rat PXR

(A) Chemical structures of carapin, santonin and isokobusone, along with those of the typical PXR and CAR agonists rifampicin, PCN, TCPOBOP, and CITCO. (B and C) CV-1 cells were cotransfected with the tk-CYP3A23-Luc reporter gene, together with the wild-type and mutant hPXR (B) or rPXR (C). Transfected cells were treated with the vehicle or the indicated compounds (10 μM each) for 24 h before luciferase assay. Rifampicin (RIF) and prenenolon-16α-carbonitrile (PCN) were included as the positive control ligands for hPXR and rPXR, respectively. Results are shown as fold induction over vehicle control and represent the average from triplicate assays. *, P < 0.05, vs the vehicle control group of each WT and mutant receptors.

Figure 2. Carapin, santonin and isokobusone induced the expression of PXR target gene CYP3A in human and mouse primary hepatocytes

(A) Treatment with carapin, santonin and isokobusone induced the mRNA expression of CYP3A4 in primary human hepatocytes. (B) The induction of Cyp3a11 mRNA by carapin, santonin and isokobusone in primary hepatocytes was abolished in hepatocytes isolated from PXR−/− mice. In contrast, the inducibility was restored in hepatocytes isolated from hPXR humanized mice. The cells were treated with the compounds at the dose of 10 μM for 24 h before harvesting total RNA and real-time PCR analysis. RIF was included as positive control for CYP3A4 induction in human hepatocytes, whereas PCN was used as positive control for Cyp3a11 induction in mouse hepatocytes. Results normalized to GAPDH are shown as fold induction over vehicle control and represented the averages from triplicate assays. *, P < 0.05, vs respective vehicle controls of each genotype.

Figure 3. Carapin, santonin and isokobusone activated human and mouse CAR

(A and B) HEK293 cells were co-transfected with hCAR (hCAR1) or hCAR3 (A) and mCAR or mCAR3 (B), together with the tk-PBRE-Luc reporter gene. Transfected cells were treated with the vehicle or the indicated compounds (10 μM each) for 24 h before luciferase assay. CITCO (1 μM) and TCPOBOP (250 nM) were included as the positive control ligands for hCAR and mCAR, respectively. Results are shown as fold induction over vehicle control and represent the averages from triplicate assays. *, P < 0.05, vs vehicle control of each receptor or its mutant variant.

Figure 4. Carapin, santonin and isokobusone induced the expression of CAR target gene CYP2B in human and mouse primary hepatocytes

Human and mouse primary hepatocytes were treated vehicle DMSO), CITCO (1 μM), TCPOBOP (250 nM), carapin, santonin, or isokobusone (10 μM each) for 24 h. CYP2B6 and Cyp2b10 mRNA expression was measured by real-time PCR analysis. Data were expressed as fold change over the control, normalized to GAPDH expression. *, P < 0.05, vs the vehicle controls of each genotype.

Figure 5. Effects of carapin, santonin and isokobusone on CYP3A and CYP2B activities

(A and B) 6β-hydroxytestosterone level in primary human hepatocytes as well as wild-type and humanized mice (A) and 16β-hydroxytestosterone level in primary mouse hepatocytes (B). Cells were treated with test compounds at the concentration of 10 μM or vehicle for 24 h. Then, they were washed with medium followed by the addition of 200 μM testosterone for 6 h. Aliquots of medium were analyzed for 6β-and 16β-hydroxytestosterone by HPLC. RIF and PCN were included as the control ligands for hPXR and mPXR, respectively. CITCO (1 μM) and TCPOBOP (250 nM) were included as the positive control ligand for hCAR and mCAR, respectively. Data are expressed as means ± SD from triplicate experiments. *, P < 0.05, vs DMSO-treated cells of each genotype.

Figure 6. Carapin, santonin and isokobusone inhibited the LPS-responsive expression of pro-inflammatory genes in human primary hepatocytes

(A and B) Human primary hepatocytes were pre-treated with vehicle, RIF, carapin, santonin or isokobusone (10 μM each) for 24 h before the addition of 20 μg/ml LPS for 6 h. Total RNA was isolated and the expression of iNOS (A) and MCP-1 (B) was determined by real-time PCR. Data were normalized to cyclophilin expression and expressed as fold change over controls. *, P < 0.05, vs LPS alone treatment groups. N = 3 for each group.

Figure 7. Carapin, santonin and isokobusone suppressed the LPS-responsive expression of pro-inflammatory genes in mouse hepatocytes in a PXR-dependent manner

(A–C) Real-time PCR analysis of the expression of iNOS (A), MCP-1 (B) and IL-1β (C) in WT and PXR−/− mouse hepatocytes that were pre-treated with vehicle, PCN, carapin, santonin, or isokobusone (10 μM each) for 24 h before treatment with LPS (20 μg/ml) for 6 h. Results were normalized to cyclophilin mRNA levels and expressed as fold change over controls. *, P < 0.05, vs LPS alone treatment groups in WT mice; # P < 0.05, vs LPS alone treatment groups in PXR KO mice. N = 3 for each group.