Activation of PXR causes drug interactions with Paxlovid in transgenic mice

Abstract

Paxlovid is a nirmatrelvir (NMV) and ritonavir (RTV) co-packaged medication used for the treatment of coronavirus disease 2019 (COVID-19). The active component of Paxlovid is NMV and RTV is a pharmacokinetic booster. Our work aimed to investigate the drug/herb-drug interactions associated with Paxlovid and provide mechanism-based guidance for the clinical use of Paxlovid. By using recombinant human cytochrome P450s (CYPs), we confirmed that CYP3A4 and 3A5 are the major enzymes responsible for NMV metabolism. The role of CYP3A in Paxlovid metabolism were further verified in Cyp3a-null mice, which showed that the deficiency of CYP3A significantly suppressed the metabolism of NMV and RTV. Pregnane X receptor (PXR) is a ligand-dependent transcription factor that upregulates CYP3A4/5 expression. We next explored the impact of drug- and herb-mediated PXR activation on Paxlovid metabolism in a transgenic mouse model expressing human PXR and CYP3A4/5. We found that PXR activation increased CYP3A4/5 expression, accelerated NMV metabolism, and reduced the systemic exposure of NMV. In summary, our work demonstrated that PXR activation can cause drug interactions with Paxlovid, suggesting that PXR-activating drugs and herbs should be used cautiously in COVID-19 patients receiving Paxlovid.

Keywords: CYP3A4/5; Drug/herb–drug interactions; Nirmatrelvir; Paxlovid; Pregnane X receptor; Rhynchophylline; Rifampicin; Ritonavir.

Graphical abstract

Figure 1

Metabolomic profiling of NMV metabolism in human liver microsomes (HLM). (A) A score plot showing the differentiation of the three incubation groups of NMV with or without NADPH. (B) A loading S-plot highlighting ions that were NMV metabolites. (C) A trend plot showing that M4 was only produced in the incubation group with NMV, HLM, and NADPH. The Y axis is the relative abundance of M4, quantified in peak areas. (D) The chromatogram of M4. The Y axis is the percentage of M4 when normalized to the maximum response in the window. (E) The MS/MS spectrum and chemical structure of M4. The X axis is mass to charge ratio (m/z) and the Y axis is the percentage of MS fragments when normalized to the maximum response in the window.

Figure 2

Roles of CYPs in NMV metabolism. (A–D) The formation of NMV metabolites M1 (A), M2 (B), M3 (C), and M4 (D) in the incubation with recombinant CYPs. NMV metabolites were detected by UPLC–Q-TOF-MS. The data in CYP3A4 group are set as 100% and expressed as mean ± SD (n = 3); ∗∗∗∗P < 0.0001 vs. control. (E) A metabolic map of NMV.

Figure 3

Impact of CYP3A deficiency on the pharmacokinetics (PK) of Paxlovid (NMV/r). hPXR/CYP3A and hPXR/Cyp3a-null mice were treated orally with NMV and RTV. Blood samples were collected from 0 to 8 h after NMV/r treatment (n = 3 or 4 at each time point). (A, D) Impact of CYP3A deficiency on the concentration–time curve of RTV (A) and NMV (D) in serum. Data are expressed as mean. (B, E) The area under the curve (AUC) of RTV (B) and NMV (E) in serum. (C, F) The maximum concentrations (C_max) of RTV (C) and NMV (F) in serum. Data are expressed as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001 vs. hPXR/CYP3A group.

Figure 4

RIF-mediated PXR activation induces CYP3A4/5 expression and accelerates NMV metabolism. hPXR/CYP3A mice were treated with RIF for 1 week. (A, B) The mRNA expression of CYP3A4 (A) and 3A5 (B) in the liver of hPXR/CYP3A mice after RIF treatment. (C) The protein expression of CYP3A in the liver of hPXR/CYP3A mice after RIF treatment. (D–G) Pretreatment with RIF increased the production of NMV metabolites M1 (D), M2 (E), M3 (F), and M4 (G). NMV was incubated in the liver S9 of hPXR/CYP3A mice pretreated with RIF. NMV metabolites were detected by UPLC–Q-TOF-MS. Data are expressed as mean ± SD (n = 3); ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001 vs. control.

Figure 5

RCP-mediated PXR activation induces CYP3A4/5 expression and accelerates NMV metabolism. hPXR/CYP3A and Pxr-null/CYP3A mice were treated with RCP for 4 days. (A, B) The mRNA expression of CYP3A4 (A) and 3A5 (B) in the liver of hPXR/CYP3A mice after RCP treatment. (C) The protein expression of CYP3A in the liver of hPXR/CYP3A and Pxr-null/CYP3A mice after RCP treatment. (D–G) Pretreatment with RCP increased the production of NMV metabolites M1 (D), M2 (E), M3 (F), and M4 (G). NMV was incubated in the liver S9 of hPXR/CYP3A mice pretreated with RCP. NMV metabolites were detected by UPLC–Q-TOF-MS. Data are expressed as mean ± SD (n = 3); ∗P < 0.05 vs. control.

Figure 6

Effects of PXR activators on the PK of NMV in hPXR/CYP3A mice receiving Paxlovid (NMV/r). hPXR/CYP3A mice were pretreated with PXR activators RIF or RCP followed by NMV/r. Blood samples were collected from 0 to 8 h after NMV/r treatment (n = 3 or 4 at each time point). (A, D) Effects of RIF (A) and RCP (D) on the concentration–time curve of NMV. Data are expressed as mean. (B, E) Effects of RIF (B) and RCP (E) on the AUC of NMV in serum. (C, F) Effects of RIF (C) and RCP (F) on the C_max of NMV in serum. Data are expressed as mean ± SD; ∗∗P < 0.01 and ∗∗∗P < 0.001 vs. control group. (G) A diagram showing the role of PXR and CYP3A in drug- and herb-interactions with Paxlovid. Drug- or herb-mediated PXR activation upregulates CYP3A4 and 3A5 expression, which in turn accelerates NMV metabolism and reduces the systemic exposure of NMV. PXR RE, PXR response element. The diagram was created with BioRender.