Synthetic oligonucleotides were purchased from Fasmac (Kanagawa, Japan). Restriction enzymes and shrimp alkaline phosphatase were from Takara Bio (Shiga, Japan). D-luciferin potassium salt was from Nacalai Tesque (Kyoto, Japan) and the reduced form of nicotinamide adenine dinucleotide phosphate (NADPH) was from Oriental Yeast (Tokyo, Japan). Pooled human liver microsomes (HLM) prepared 50 donors, and Supersomes™ expressing CYP3A4 and NADPH-P450 reductase (CPR) were from BD Gentest (Franklin Lakes, NJ, United States). All other reagents were of the highest quality commercially available.

pEZT-BM, a transfer plasmid for the Bac-Mam system, was a gift from Ryan Hibbs (Addgene plasmid # 74099; http://n2t.net/addgene:74099; RRID: Addgene_74099). The plasmid was based on pFastBac Dual (Thermo Fisher Scientific, Waltham, MA, United States), and the original insect polyhedrin promotor was replaced by a cytomegalovirus (CMV) promotor for expression of target gene in mammalian cells (Morales-Perez et al., 2016). The open reading frames (ORFs) of CYP3A4, UGT1A1, and UGT2B7 were amplified with each pair of primers as listed in Supplemental Table S1, and each pFastBac1-based construct as template (Ishii et al., 2014; Miyauchi et al., 2015). KOD-Plus-Neo DNA Polymerase (Toyobo Life Science, Osaka, Japan) was used for the polymerase chain reaction, and thermal cycling condition was as follows: initial denaturation, 94°C, 2 min; 40×amplification step, 98°C, 10 s; 52°C, 30 s; 68°C, 1 min; hold, 4°C. The PCR products were digested with DpnI, purified with a FastGene Gel/PCR Extraction Kit (NIPPON Genetics, Tokyo, Japan), and introduced into KpnI/XhoI digested pEZT-BM using an In-Fusion HD Cloning Kit (Takara Bio). The ORF of CYP3A4 was also subcloned into pcDNA3.1/Hygro (−), a mammalian cell expression vector, with KpnI/XhoI sites using Ligation high Ver.2 (Toyobo Life Science). Primers for this CYP3A4 subcloning are also listed in Supplemental Table S1. The nucleotide sequences of the constructs were confirmed by an Applied Biosystems 3130xl Genetic Analyzer (Thermo Fisher Scientific), using a BigDye Terminator version 3.1 Cycle Sequencing Kit (Thermo Fisher Scientific).

Recombinant Bac-Mam viruses were prepared with the Bac-to-Bac Baculovirus Expression system (Thermo Fisher Scientific). The pEZT-BM constructs were transfected into competent Escherichia coli (E. coli) DH10Bac. A positive single clone was selected after blue/white colony selection, and recombinant bacmid, a part of the baculoviral DNA, was prepared according to the user manual. Insect cells Sf9 were cultured in a 100 or 500 mL plastic Erlenmeyer flask (Corning, Lowell, MA, United States) containing Sf-900 III medium (Thermo Fisher Scientific) supplemented with 5% fetal bovine serum (FBS). To prepare recombinant Bac-Mam virus, the constructed bacmids were transfected into Sf9 cells as described previously (Miyauchi et al., 2015). The control bacmid, obtained from transfection of mock pEZT-BM, was also transfected to obtain control virus. pEZT-BM contains the p10 promotor followed by the ORF of green fluorescent protein (GFP), which enabled us to confirm viral generation in Sf9 cells by detecting GFP (Supplemental Figure S1). Culture media were collected as primary viruses 1 week after transfection. Baculoviral DNA were purified from portions of these media (200 μL) by NucleoSpin Blood (Macherey-Nagel, Düren, Germany), and their titer was determined using a BacPAK qPCR Titration Kit (Clontech, Mountain View, CA, United States). The Bac-Mam viruses were amplified 2–3 times until their titer reached over 5.0×10⁷ plaque-forming unit (PFU)/mL.

In this study, COS-1 cells were mainly utilized for expression of drug-metabolizing genes. COS cells (COS-1 cells, COS-3 cells, and COS-7 cells) were established by transformation of African monkey kidney cells (CV-1 cells) with a mutant of simian virus 40 (SV₄₀) (Gluzman, 1981) and have been widely used in research including the field of drug metabolism (Mackenzie, 1986; Clark and Waterman, 1991; Guengerich et al., 1997; Thomae et al., 2002). In case of P450, it is especially advantageous that COS cells have endogenous CPR, which is sufficient for catalytic activity of transiently expressed P450 (Clark and Waterman, 1991). COS-1, HEK293, and HepG2 cells were grown in Dulbecco’s modified Eagle medium (FUJIFILM Wako Pure Chemical, Osaka, Japan, Cat#043-30085) supplemented with 10% FBS. Cells were routinely maintained in a 10 cm dish and seeded to a 35 mm dish at 1.0×10⁶ cells/dish the day before infection. After removing the culture medium, recombinant Bac-Mam viruses were added to the dish. Viral amount, multiplicity of infection (MOI), ranged from 25 to 150 and finally fixed at 100. The infection was performed in a CO₂ incubator for 1 hour. Then virus was removed, and new culture medium containing sodium butyrate ranged from 0 to 4.5 mM (finally fixed at 3 mM) was added. Sodium butyrate is used as an inhibitor of histone deacetylase, enhancing expression of recombinant protein in mammalian cells including the Bac-Mam system (Gorman et al., 1983; Condreay et al., 1999; Morales-Perez et al., 2016). The infected cells were further cultured for 24–96 h. The culture time was finally fixed at 48 h.

COS-1 cells were seeded in a 35 mm dish the day before transfection. Transfection of pcDNA3.1/Hygro (−)_CYP3A4 was conducted as described previously (Miyauchi et al., 2020b). In short, 2 μg plasmid DNA was mixed with 8 μg of Polyethylenimine HCl MAX (PEI; Polysciences, Warrington, PA, United States) in Opti-MEM I Reduced Serum Media (Thermo Fisher Scientific) and incubated at room temperature for 10 min. The DNA-PEI complex was added to COS-1 cells, and the cells were cultured for 48 h.

Cells were collected in homogenate buffer containing 10 mM Tris-HCl (pH7.4), 250 mM sucrose, and 10% glycerol, and then sonicated with an AS38A ultrasonic cleaner (AS ONE, Osaka, Japan). Protein concentration was determined using Quick Start Bradford 1×Dye Reagent (Bio-Rad Laboratories, Hercules, CA, United States) with bovine serum albumin as a standard.

Cell homogenates (20 μg) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and electroblotted onto a polyvinylidene difluoride membrane (FUJIFILM Wako Pure Chemical, Cat#034-25663). The blots were washed with Tris-buffered saline containing 0.1% Tween 20 (TBS-T) and incubated in TBS-T containing 5% skim milk at room temperature for 30 min. After blocking, the blots were incubated with primary antibody diluted 2,000-fold with TBS-T containing 5% skim milk at 4°C overnight. The following primary antibodies were utilized: mouse anti-CYP3A4 antibody, rabbit anti-UGT1A antibody (respective Cat# sc-53850 and sc-25847; Santa Cruz Biotechnology, Dallas, TX, United States), rabbit anti-UGT2B7 antibody, and rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody (respective Cat# 16661-1-AP and 10494-1-AP, Proteintech, Rosemont, IL, United States). The blots were extensively washed with TBS-T and further incubated with secondary antibody diluted 10,000-fold with TBS-T containing 5% skim milk at room temperature for 1 hour. The following secondary antibodies labeled with horseradish peroxidase (HRP) were utilized: HRP-goat anti-mouse IgG (H + L) antibody and HRP-goat anti-rabbit IgG (H + L) antibody (respective Cat# SA00001-1 and SA00001-2, Proteintech). After extensive washing with TBS-T, EzWestLumi plus (ATTO, Tokyo, Japan) was used as a substrate of HRP, and the signal was visualized and analyzed using iBright Imaging System (Thermo Fisher Scientific). To quantify expressed CYP3A4 in 20 μg homogenates, Supersomes™ (BD Gentest) containing 1 pmol CYP3A4 were applied as a standard.

CYP3A4 activity was measured with P450-Glo CYP3A4 Assay and Screening System (Luciferin-IPA; Promega, Madison, WI, United States). We used 100 μM NADPH (final concentration). Substrate luciferin-IPA concentration varied from 0.3 to 15 μM (nine points of concentration). Chemiluminescence was measured using a microplate reader, Infinite M200 Pro (Tecan Group, Männedorf, Switzerland). As enzyme sources, 20 μg HLM, Supersomes™, or COS-1 homogenates were utilized. In kinetic experiments, data were fitted to the Michaelis-Menten model defined by the equation below: V = V m a x × S K m + S Where V is the reaction rate, S is the substrate concentration, Vmax is the maximum enzyme velocity, and Km is the Michaelis constant, which is equal to the concentration of substrate for half-maximal velocity. The kinetic parameters were determined with GraphPad Prism 5.04 software (GraphPad software, La Jolla, CA, United States).

The drug-metabolizing enzymes were transiently expressed using the Bac-Mam system in COS-1 cells and their expressions were confirmed by immunoblotting (Figure 1). Infection conditions were fixed as below: MOI = 100; culture time, 24 h; sodium butyrate, 3 mM. With every enzyme, specific bands were observed that were not detectable in control virus-infected cells. In addition, the enzymes showed the same mobility as those in the positive control, HLM. The antibody that we used in detection of UGT1A1 can react with other UGT1A isoforms in HLM, which resulted in a slight difference of observed bands between HLM and COS-1 homogenates. These results indicated that the Bac-Mac system is suitable for transient expression of the major drug-metabolizing enzymes. In this experiment, we loaded 20 μg homogenates and 10 μg HLM. CYP3A4 showed the closest expression to that in HLM among the three targets, so we optimized infection conditions by monitoring CYP3A4 expression level in subsequent experiments.

We evaluated infection conditions to obtain maximum CYP3A4 expression using the recombinant Bac-Mam virus in COS-1 cells. P450 can be quantified by measuring its CO difference spectrum (Omura and Sato, 1964), but a previous study suggested that the amount of P450 that can be overexpressed in COS-1 cells was below the amount required for detection using the CO difference spectrum assay (Clark and Waterman, 1991). Therefore, we quantified expressed CYP3A4 by immunoblotting in the current study. First, we tested varying the amount of virus. In our previous study, CYP3A4, CPR, and some UGT isoforms were expressed in insect Sf9 cells using an original baculovirus expression system. The viral amounts in the experiments were set at MOI = 0.01–1, which was enough to express the enzymes in Sf9 cells (Ishii et al., 2014; Miyauchi et al., 2015; Miyauchi et al., 2020a). In contrast, more than 10-fold larger amounts of virus seem to be necessary in the Bac-Mam system since the Bac-Mam virus cannot amplify in mammalian cells, and so secondary infection is not expected (Condreay et al., 1999). Hence, COS-1 cells were infected with the prepared Bac-Mam virus at MOI = 0 (not infected), 25, 50, 75, 100, and 150. Culture time and concentration of sodium butyrate were fixed at 24 h and 3 mM, respectively. Immunoblotting and its analyzed result are shown in Figures 2A,B, respectively. CYP3A4 expression tended to increase in a MOI-dependent manner and reach a peak around MOI = 75–100. Hence, MOI was fixed at 100 in the later experiments.

Next, we tested varying culture time after viral infection. Viral amount and concentration of sodium butyrate were fixed at MOI = 100 and 3 mM, respectively. Bac-Mam virus-infected cells began to float from dishes 48 h after infection, and almost all the cells were dead at 96 h. Thus, we prepared cell homogenates between 24 and 72 h post infection and quantified CYP3A4 expressions (Figures 2C,D). Expression levels increased in a time-dependent manner and reached a maximum at 48–72 h after infection. Given that the cells began to die, we concluded that 48 h is the best culture time after infection.

As cell death was also observed in the absence of viral infection, we predicted that sodium butyrate was the cause of this toxicity. We next varied the concentration of sodium butyrate from 0 to 4.5 mM under the fixed condition: viral amount, MOI = 100; culture time, 48 h. As explained, sodium butyrate enhances target gene expression in the Bac-Mam system. Hence, CYP3A4 was not detectable under 1.5 mM while significant expression was observed at three and 4.5 mM (Figures 2E,F). In accordance with our hypothesis, however, not only significant CYP3A4 expression but also cell toxicity was observed, and COS-1 cells seemed to be more sensitive to sodium butyrate than HEK293 and HepG2 cells. We determined that 3 mM is the best concentration of sodium butyrate for use. Taken together, we concluded that the optimized infection condition was: viral amount, MOI = 100; culture time after infection, 48 h; concentration of sodium butyrate, 3 mM.

According to the optimized infection condition, COS-1 homogenates were prepared, and the activity of expressed CYP3A4 was measured with luciferin-IPA as a substrate, which is converted to luciferin by CYP3A4 but shows minimal cross-reactivity with CYP3A5 and CYP3A7. Kinetic analysis was conducted with the COS-1 homogenates, HLM, and Supersomes™ as enzyme sources, and obtained parameters were compared (Figure 3 and Table 1). CYP3A4 activity was confirmed in the prepared homogenates, which supported our view that the Bac-Mam system can be applied for expression of drug-metabolizing enzymes. The highest CYP3A4 activity was observed when Supersomes™ was used as the enzyme source, followed by HLM and homogenates. In comparison with HLM and Supersomes™, the Vmax value of the Bac-Mam expressed CYP was one 30th of those of HLM/Supersomes™ although the Km value was comparable.

Transfection of recombinant plasmid with a transfection reagent is one of the major methods to introduce a target gene into mammalian cells and is also utilized in research of drug-metabolizing enzymes. We compared the expressed CYP3A4 levels between the Bac-Mam system and the widely used transfection method (Figures 4A,B). With the Bac-Mam system, 2–3 times higher expression of CYP3A4 was observed compared to that expressed using the transfection method.

We have applied the Bac-Mam system to introduce genes of CYP3A4 and some UGTs into COS-1 cells and confirmed their expression and CYP3A4 function. Further, we utilized the Bac-Mam system to introducing CYP3A4 into other kinds of cells (HEK293 and HepG2 cells) under the optimized condition determined with COS-1 cells. Although these cells are widely used in research of drug-metabolizing enzymes, there was a large difference in the expressed CYP3A4 levels (Figures 4C,D). COS-1 cells showed the highest CYP3A4 level, almost 4- and 20-times higher than that in HepG2 and HEK293 cells, respectively.