In VitroCharacterization of CYP3A4 Inhibition and Induction of Nevirapine Analogs as Substitutes in HIV Treatment

Abstract

<b>Abstract ID 21354</b> <b>Poster Board 379</b> Nevirapine (NVP) is an HIV non-nucleoside reverse transcriptase inhibitor used both in the US and developing countries to treat and prevent HIV transmission in women and infants. NVP is generally well tolerated, but idiosyncratic drug reactions such as hepatotoxicity have been reported. NVP can cause drug-drug interactions (DDI) through both inhibition and induction of the cytochrome P450 (CYP) enzymes. NVP is metabolized by several CYPs, such as CYP3A4 and CYP2B6 and it also induces their expression. NVP is a mechanism-based inactivator (MBI) of CYP3A4, although the exact mechanism is still debated in the literature. One prominent hypothesis is that NVP forms a reactive quinone methide metabolite, through the 12-OH-NVP oxidation pathway, which can form adducts to CYP3A4. Previous data from our lab has demonstrated several NVP analogs have similar efficacy against the target HIV-1 reverse transcriptase compared to NVP. Our hypothesis is that a structural analog of NVP with similar efficacy but reduced toxicity could be a candidate to replace NVP by decreasing the formation of the reactive quinone methide species through the 12-OH pathway, thereby reducing DDI resulting from MBI of CYP3A4. <b>Methods:</b> We analyzed three NVP analogs and NVP for their CYP induction of mRNA and activity in primary human hepatocytes. mRNA induction was analyzed via qPCR and activity was determined using probe substrate reactions and analyzed via LC-MS. We characterized the inactivation of CYP3A4 by the top candidate NVP-d3 versus NVP. IC₅₀ shift assays, K_I/k_inact assays, and GSH incubations were performed with recombinant CYP3A4 and analyzed via LC-MS. <b>Results:</b> NVP and NVP-d3 demonstrated similar mRNA induction of CYP3A4 and CYP2B6. CYP3A4 activity was reduced from 25 μM to 100 μM for NVP, despite the increased mRNA levels. NVP-d3 demonstrated increased activity at 100 μM compared to 25 μM. The CYP3A4 IC₅₀ values with a 30 minute pre-incubation were 30.0 ± 2.8 μM (- NADPH) and 5.3 ± 1.0 μM (+NAPDH) for NVP, and 9.4 ± 1.6 μM (- NADPH) and 2.9 ± 0.4μM (+NAPDH) for NVP-d3, resulting in a shift of 5.7x for NVP and 3.2x for NVP-d3. The MBI assay resulted in a K_I of 142.4 ± 10.9 μM and k_inact of 0.103 ± 0.003 min^-1 for NVP, and a K_I of 128.0 ± 18.1 μM and k_inact of 0.079 ± 0.004 min^-1 for NVP-d3. No GSH adducts were found for NVP or NVP-d3 at the 12 position, but one likely adduct was identified at the 3 position. There was a significant difference in the metabolites produced by CYP3A4. With NVP-d3, the fold change of 12-OH metabolite formation was 0.16, while 2-OH and GSH metabolites increased by about 2 fold each. <b>Discussion and Conclusions:</b> These results demonstrate that both NVP and NVP-d3 are MBIs of CYP3A4 with important differences in metabolite profiles. NVP-d3 demonstrated lower CYP3A4 IC₅₀ and K_I values, although the difference may not be significant. Deuteration decreased the 12-OH metabolite from being formed and increased the GSH adduct formation at the 3 position, indicating reduced formation of the quinone methide reactive species at the 12 position. In rats, it was previously reported that NVP-d3 is cleared faster than NVP, indicating the induction effect may overcome inhibition, as supported by our data here. NVP-d3 may still be a candidate to replace NVP in the future, but pharmacokinetics need to be evaluated if decreasing the MBI of CYP3A4 is leading to increased clearance. This work was supported by the NIH, NIAID Award Number R01 AI150494 (JNL).