Characterization of human alcohol dehydrogenase 4 and aldehyde dehydrogenase 2 as enzymes involved in the formation of 5-carboxylpirfenidone, a major metabolite of pirfenidone.

Pirfenidone (PIR) is used to treatment of idiopathic pulmonary fibrosis. After oral administration, it is metabolized by cytochrome P450 1A2 to 5-hydroxylpirfenidone (5-OH PIR) and further oxidized to 5-carboxylpirfenidone (5-COOH PIR), a major metabolite excreted in the urine (90% of the dose). This study aimed to identify enzymes that catalyze the formation of 5-COOH PIR from 5-OH PIR in the human liver. 5-COOH PIR was formed from 5-OH PIR in the presence of NAD⁺ by human liver microsomes (HLM) more than by human liver cytosol (HLC), with the concomitant formation of the aldehyde form (5-CHO PIR) as an intermediate metabolite. By purifying enzymes from HLM, alcohol dehydrogenases (ADHs) were identified as candidate enzymes catalyzing 5-CHO PIR formation, although ADHs are localized in the cytoplasm. Among constructed recombinant ADH1-5 expressed in HEK293T cells, only ADH4 efficiently catalyzed 5-CHO PIR formation from 5-OH PIR with a K _m value (29.0 {plus minus} 4.9 µM), which was close to that by HLM (59.1 {plus minus} 4.6 µM). In contrast to commercially available HLC, in-house prepared HLC clearly showed substantial 5-CHO PIR formation, and ADH4 protein levels were significantly (rs = 0.772, P < 0.0001) correlated with 5-CHO PIR formation in 25 in-house prepared HLC samples. Some components of the commercially available HLC may inhibit ADH4 activity. Disulfiram, an inhibitor of aldehyde dehydrogenases (ALDH), decreased 5-COOH PIR formation and increased 5-CHO PIR formation from 5-OH PIR in HLM. ALDH2 knockdown in HepG2 cells by siRNA decreased 5-COOH PIR formation by 61%. Significance Statement This study clarified that 5-COOH PIR formation from 5-OH PIR proceeds via a two-step oxidation reaction catalyzed by ADH4 and disulfiram-sensitive enzymes, including ALDH2. Inter-individual differences in the expression levels or functions of these enzymes could cause variations in the pharmacokinetics of PIR.