Materials and Methods

Human Tissues.

Human livers were obtained from the Liver Tissue Procurement and Distribution System (Minneapolis, MN), Cooperative Human Tissue Network (Birmingham, AL), Dr. Urs Meyer (University of Basel), and Dr. Erin Schuetz (St. Jude Children's Research Hospital, Memphis, TN). Only livers that were histologically normal at the time of procurement and obtained through biopsy or within 1 h of “cross-clamp” (in the case of organ donors) were used in the analyses of P450 catalytic activities. For the majority of livers, complete drug history before liver collection was not available. Because catalytic activities were not found to differ between livers from individuals who were known to have received potent P450 inducers (e.g., glucocorticoids, phenytoin, and barbiturates) and those who were not, “induced” livers were included in all other comparisons. Microsomes were prepared and protein was measured as reported previously (Relling et al., 1992). Microsomal recovery (expressed as a percentage, %) was estimated using the following formula: total amount of microsomal protein/wet weight of frozen liver × 100.

P450 Catalytic Activities.

The substrates used in this study were teniposide (CYP3A4/5 substrate), ethoxycoumarin (CYP2E1 and other P450s substrate), midazolam (CYP3A4 and CY3A5 substrate), paclitaxel (CYP2C8), tolbutamide (CYP2C9), and ethoxyresorufin (CYP1A2 substrate). Maximal P450 catalytic activities were assessed by incubating 0.1 to 0.3 mg of microsomal protein in duplicate with prototypical substrates for specific P450s; not all activities were assayed in all livers. Details for teniposide (Relling et al., 1994), midazolam (Relling et al., 1994), paclitaxel (Sonnichsen et al., 1995), ethoxycoumarin (Evans and Relling, 1992), and tolbutamide (Relling et al., 1990) incubations and assays were reported previously. “Maximal” catalytic activities at 500-, 60-, 20-, 1000-, and 2000-μM substrate concentrations, respectively, were assessed. Ethoxyresorufin (5 μM) deethylation was assessed using a modification of a fluorimeter-based method (Lubet et al., 1985). Briefly, 0.2 mg of microsomal protein, 10 μl of NADPH regenerating system, in a total volume of 100 μl of 0.1 M TRIS-HCl buffer (pH 7.8) were incubated at 37°C for 5 min. The reaction was stopped by addition of 100 μl of cold methanol. After low-speed centrifugation, 50 μl were injected onto an HPLC system using a Bondapak Phenyl column (Waters, Milford, MA) with conditions essentially as described (Evans and Relling, 1992). Resorufin was quantitated using calibrators prepared exactly as described for unknowns but having resorufin spiked after the addition of methanol to microsomes and NADPH. All catalytic activities are expressed as nanomoles of metabolite formed per hour per milligram of microsomal protein.

Statistical Methods.

Wilcoxon rank-sum tests were used to determine whether or not the distributions of two groups of maximal catalytic activity assay results differed, whereas Kruskal-Wallis H tests were used to determine whether or not at least one of the distributions of k > 2 groups of catalytic activity assay results differed. Exact methods were used when necessary. Spearman's rank correlation coefficient was used to determine whether or not a relationship existed between age or maximal activities and microsomal recovery.

No adjustments have been made to the P values included in this report. However, Bonferroni adjustments were made to the significance levels to control for multiple testing within each catalytic activity. The overall significance level within each activity was controlled at α = 0.05.

All analyses were conducted using SAS Release 6.12 and StatXact-3 Version 3.0.2.

Results

Maximal catalytic activity assays were conducted in 52 normal livers; not all activities were assayed in each liver sample. The age range for the youngest group (<10 years, n = 3–13) was 0.5 to 9 years of age, for the intermediate group (>10 to <60 years, n = 7–19) was 10 to 59 years of age, and for the oldest group (>60 years) was 63 to 93 years of age. As most studies show little difference in clearance among children over 10 years of age and adults (Ellis et al., 1976; Crom et al., 1991), only those <10 years of age are denoted “children” in the following analyses.

Figure 1 shows the distributions of each cytochrome P450 catalytic activity in the three defined age groups. There were no differences between the three age groups in the oxidation of the following substrates: ethoxyresorufin (P = .83, CYP1A2 substrate), teniposide (P = .58, CYP1A4/5 substrate), ethoxycoumarin (P = .52, CYP2E1 and other P450s substrate), midazolam (P = .47, CYP3A4/3A5), or paclitaxel (P = .24, at the 17α position, CYP2C8). Tolbutamide hydroxylation tended to be lower in children than adults (P = .047, CYP2C9 substrate), but did not reach statistical significance after correcting for multiple comparisons. Additionally, the catalytic activity of each CYP was plotted as a function of the age for each of the individual liver samples for all six substrates analyzed (Fig. 2). These data fail to indicate a trend for age-related differences in maximal P450 oxidation rates.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Box plots depicting median, interquartile range, and absolute range of catalytic activities (nmol/mg/h) by age group for ethoxyresorufin deethylation (CYP1A2), teniposide O-demethylation (CYP1A4/5), ethoxycoumarin deethylation (CYP2E1 and others), midazolam 1′-hydroxylation (CYP3A4/3A5), paclitaxel 17α hydroxylation (CYP2C8), and tolbutamide hydroxylation (CYP2C9).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Catalytic activities (nmol/mg/h) for six P450 substrates versus age of liver donor.

⋄, ethoxycoumarin O-deethylation; ▵, midazolam 1′hydroxylation; ✳, ethoxy-resorufin O-deethylation; … ■… , teniposide demethylation; —■—, tolbutamide hydroxylation; ●, paclitaxel hydroxylation.

We also explored whether other factors may have accounted for the lack of a relationship between age and maximal P450 catalytic activities. There were no differences in any P450 activity related to sex or race (data not shown). Moreover, microsomal recovery from liver (average = 1.07%, 1 S.D. = 0.60) was not related to age (P = .98), nor was recovery related to any of the six catalytic activities (P = .094 for ethoxyresorufin,P = .83 for teniposide; P = .69 for ethoxycoumarin; P = .62 for midazolam;P = .70 for paclitaxel; and P = .40 for tolbutamide).

Discussion

In previous clinical pharmacokinetic studies, we proposed that greater overall catalytic activity of hepatic drug-metabolizing enzymes could have contributed to the high antipyrine clearance that we observed in younger children compared with adolescents (Murry et al., 1995). To test this hypothesis, we evaluated the catalytic activities of six phase I drug-metabolizing enzymes (CYP1A2, CYP2E1, CYP3A4/3A5, CYP2C8, and CYP2C9) from the microsomal fraction of normal human livers covering a wide age range (0.5–93 years). These enzymes constitute over 80% of the P450 enzymes present in human liver microsomes (Shimada et al., 1994).

Catalytic activities were determined using saturating substrate concentrations (zero order kinetics). Under these conditions, although differences in K_m are not assessed, the catalytic rate is a function of the amount of active enzyme present in the sample. However, we found no correlation between age and maximal P450 catalytic activity with any of the six P450 substrates, with ages encompassing both pediatric and elderly liver donors. Moreover, the absence of statistical differences between age groups is confirmed with the lack of age-related trends observed for each individual value versus age (see Fig. 2).

Heretofore, there have been relatively few data reported in non-neonatal children. A lack of relationship between age and hepatic microsomal P450 activity (Shimada et al., 1994) or NADPH cytochrome P450 reductase (Schmucker et al., 1990) has been found in studies, including only one sample from a donor younger than 10 years of age. Studies in samples from fetal and neonatal livers indicate that most P450s are expressed at close to adult levels by a few months of age (Treluyer et al., 1991; Cazeneuve et al., 1994; Vieira et al., 1996;Lacroix et al., 1997). Limited data suggest that CYP2E1 and CYP3A5 activities are comparable in children and adults (Vieira et al., 1996;Lacroix et al., 1997). Taken together with our findings, there are not data to support the hypothesis that P450 activities are higher in the 1- to 10-year age range compared with other ages. Because it is this group that tends to exhibit higher drug clearance clinically, it appears that other mechanism(s) must account for their higher clearance. Our findings that in vitro maximal catalytic activities do not mirror the increased drug clearance of children are consistent with and complementary to previous studies showing comparable maximal catalytic activities in adult versus elderly livers (Brodie et al., 1981) (Le Couteur and McLean, 1998).

There are several factors other than maximal P450 activity that could contribute to age-related changes in drug clearance, in vivo, and could account for the lack of correspondence between in vitro and clinical findings. Substantial reductions in blood flow and hepatic size contribute to reduced clearance in the elderly (Dawling and Crome, 1989). Many high intrinsic-clearance-drugs that are cytochrome P450 substrates, for example, erythromycin (Miglioli et al., 1990), nifedipine (Robertson et al., 1988), and hexobarbital (Chandler et al., 1988) have lower elimination in the elderly than in young adults (Kinirons and Crome, 1997). However, for P450 substrates that are low intrinsic-clearance-drugs, clearance is not significantly dependent on hepatic blood flow, and, thus, higher hepatic blood flow is unlikely to account for most higher P450 drug substrate clearances in children.

Another explanation that has been put forward to explain lower drug clearance in the elderly is the “oxygen limitation” theory (Le Couteur and McLean, 1998). According to this hypothesis, the hepatic oxygen supply for phase I enzymatic reactions appears to be compromised among the elderly. This effect may not have been observed in our microsomal model because oxygen supply is not as constrained in vitro as it might be in vivo. It is conceivable that transcellular uptake of oxygen, substrates, or cofactors that facilitate metabolism are more efficient in pediatric liver, and these uptake effects would not be observable with in vitro microsomal studies. Factors that affect enzyme-substrate affinity or endoplasmic reticulum availability of either substrate or enzyme could vary as a consequence of subtle age-related changes in the hepatocyte structure or enzyme conformation (e.g., glycosylation status, folding, etc.), and these factors would not have been appreciated in our analysis of maximal catalytic activity in microsomes. Moreover, if microsomal recovery had been reduced in pediatric relative to adult liver, maximal activities normalized to microsomal protein concentration would have been low relative to that normalized to total pediatric liver. However, we found no relationship between age and microsomal recovery, nor any relationship (nor was any expected) between recovery and maximal catalytic activity.

We conclude that increased maximal P450 activity, as a function of amount of hepatic microsomal protein, is unlikely to account for the higher clearance observed for most P450 drug substrates in children versus adults.