Glucuronidation is a major metabolic pathway in the biotransformation of many xenobiotics. Recent studies have shown that in humans, UDP-glucuronosyltransferase (UGT)-mediated glucuronidation plays a critical role in the detoxification of food-borne carcinogenic heterocyclic amines. 2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), the most abundant carcinogenic heterocyclic amine found in well-cooked meats, has been shown to be extensively glucuronidated in humans. To determine which UGT isozymes are involved in the biotransformation of PhIP and the cytochrome P4501A2-mediated reactive intermediate N-hydroxy-PhIP, microsomes expressing human UGT1A1, -1A4, -1A6 or -1A9 were incubated with PhIP and N-hydroxy-PhIP and the reaction products analyzed by HPLC and ESI-MS. Incubations containing N-hydroxy-PhIP and UGT1A1 expressing microsomes, with an apparent Km of 4.58 μM and a Vmax of 4.18 pmol/min/mg protein, had the highest capacity to convert N-hydroxy-PhIP to N-hydroxy-PhIP-N2-glucuronide. Microsomes expressing UGT1A9 produced N-hydroxy-PhIP-N3-glucuronide at the highest rate with an apparent Km and Vmax of 3.73 μM and 4.07 pmol/min/mg, respectively. A third previously undefined glucuronide accounted for 31% of the total glucuronides formed from the UGT1A4 expressing microsomes. No glucuronide conjugates were detected from microsomes expressing UGT1A6. Incubations containing PhIP as substrate formed direct PhIP-glucuronides in microsomes expressing UGT1A1, UGT1A4 and UGT1A9 but at levels averaging 53-fold lower than when N-hydroxy-PhIP was used as the substrate. Knowing the glucuronidation capacity of the specific UGT isozymes involved in PhIP and N-hydroxy-PhIP glucuronidation should help in determining the individual susceptibility to the potential cancer risk from exposure to PhIP.
Glucuronidation has been well established as a major conjugation reaction in the biotransformation of many xenobiotics. These conjugation reactions are catalyzed by numerous isoforms of UDP-glucuronosyltransferase (UGT) (1). These enzymes are primarily found in the endoplasmic reticulum of many tissues, with the liver being, quantitatively, the most significant site of glucuronidation. In addition to a wide tissue distribution of UGTs, different isozymes can be preferentially expressed in specific tissues. Furthermore, polymorphic expression of certain UGTs has also been observed (2). Three UGT families have been identified in humans; designated UGT1, UGT2 and UGT8 based on their evolutionary divergence (1). Of these three families UGT1 and UGT2 have been shown to catalyze the glucuronidation of a wide variety of xenobiotic substrates, with UGT1 being more active in the glucuronidation of amines (3,4). Aryl- and alkylamines, sulfonamides, heterocyclic amines and hydroxylated compounds have all been reported to undergo glucuronidation in many animal species and humans. The primary function of UGTs is to eliminate substrates from the body via urine and feces by catalyzing the formation of hydrophilic glucuronide conjugates (5).
Recent studies have shown that UGT-mediated glucuronidation appears to play a critical role in the bioactivation/detoxification of food-borne carcinogenic heterocyclic amines (6–9). 2-Amino-1-methyl-6-phenylimidazo[4,5-b]pyridine (PhIP), the most abundant carcinogenic heterocyclic amine found in well-done cooked beef and chicken (10,11), has been shown to be extensively glucuronidated in humans, both in vivo and in vitro (8,12,13). It has also been shown that fundamental differences in PhIP glucuronidation capacity exist between humans and animal species (8). These differences could be related to the differential expression of specific UGT isozymes.
Studies have shown that the bioactivation of PhIP in humans is highly dependent upon the cytochrome P4501A2 (CYP1A2)-mediated N-hydroxylation of the parent amine to the corresponding 2-hydroxyamino-1-methyl-6-phenylimidazo[4,5-b]pyridine (N-hydroxy-PhIP) (14,15). N-Hydroxy-PhIP can be subsequently esterified by sulfotransferase and/or acetyltransferase which generates the highly electrophilic O-sulfonyl and O-acetyl esters, respectively (16,17). Alternatively, N-hydroxy-PhIP can form stable glucuronide conjugates at the N2 and N3 positions which can either be excreted through the urine or bile or be transported to extrahepatic tissue where further metabolism can occur (8,9). Furthermore, PhIP can form direct non-reactive glucuronides at the N2 and N3 positions (18). A recent study implicated the UGT1A subfamily of UGTs as being a major contributor to the biotransformation of N-hydroxy-PhIP (19). However, there is still some uncertainty as to which specific UGT1A isozymes are involved in N-hydroxy-PhIP glucuronidation. Understanding the N-glucuronidation of PhIP and N-hydroxy-PhIP in humans, and identifying the specific isozymes involved and their regulation is especially important because the failure to conjugate N-hydroxy-PhIP by glucuronidation could result in further activation by sulfotransferase and/or acetyltransferase, resulting in highly reactive esters which can bind DNA (16,17). Therefore, determining the glucuronidation capacity of the specific isozymes involved in PhIP and N-hydroxy-PhIP glucuronidation may help in evaluating the individual susceptibility to the potential cancer risk from exposure to PhIP.
The goal of the present study was to determine which UGT isozymes contribute to PhIP and N-hydroxy-PhIP glucuronidation. Four microsomal preparations, each expressing a different human UGT1A isozyme, were used to investigate the capacity of each isozyme to conjugate PhIP and N-hydroxy-PhIP. Each isozyme possessed different catalytic capabilities in conjugating both PhIP and N-hydroxy-PhIP.
Materials and methods
PhIP and 2-nitro-PhIP were purchased from Toronto Research Chemicals (Downsview, Ontario). N-Hydroxy-PhIP was synthesized from 2-nitro-PhIP by dissolving 1 mg in 1.0 ml tetrahydrofuran containing 8 mg 10% palladium on activated carbon at 0°C, with vigorous stirring under nitrogen. Hydrazine hydrate (20 μl) was added drop-wise and the reaction was allowed to progress for 15 min. The reaction was stopped by decanting the liquid into a clean vial and placing it onto an ethanol dry-ice slurry for 30 min. The N-hydroxy-PhIP product was purified by HPLC (isocratic at 40% methanol) to >95% purity. UDP-glucuronic acid (UDPGA), alamethicin and β-glucuronidase type VII-A from Escherichia coli were purchased from Sigma (St Louis, MO). All other reagents were of analytical grade or better.
To determine which UGT isoforms are involved in the glucuronidation of PhIP and N-hydroxy-PhIP, microsomes from the human lymphoblastoid cell line AHH-1 TK+/–, expressing human UGT1A1, -1A4, -1A6 or -1A9 (Gentest, Woburn, MA), were incubated with PhIP or N-hydroxy-PhIP. Incubations were prepared on ice in conical plastic tubes and consisted of 1.0 mg/ml microsomal protein, 5.0 mM MgCl2, 0.5 mM EDTA (N-hydroxy-PhIP incubations only), 6.0 mM UDPGA, 60 μg/ml alamethicin and 10 μl PhIP or N-hydroxy-PhIP (dissolved in dimethylsulfoxide at concentrations described in the figure legends) in 0.1 M Tris–HCl, pH 7.8, in a final volume of 0.5 ml. Samples were incubated for 3 h at 37°C with no agitation after initial mixing. The 3 h time point was determined to be the optimal incubation time for both substrates, and all UGT isoforms. After the incubation time, 1 vol ice-cold methanol was added to each sample to precipitate the proteins. The samples were allowed to stand on ice for 30 min. The protein was then removed by centrifugation in a microcentrifuge at maximum speed for 5 min. The methanolic extracts were placed in clean plastic tubes and stored at –70°C until HPLC analysis.
The aqueous–methanol extracts from the microsomal incubations were evaporated to dryness under nitrogen, then reconstituted in 50 μl HPLC starting mobile phase. The samples were centrifuged in a microcentrifuge for 1 min at maximum speed and the supernatant was injected into an Alliance HPLC system (Waters, Milford, MA) equipped with a 5 μm, 4.6×220 mm TSK-GEL ODS-80 TM column (TosoHass, Montgomeryville, PA) and a Waters 990 photodiode array detector. Metabolites were eluted at 1.0 ml/min using a gradient starting at 30% methanol/0.1% triethylamine, pH 6.0, up to 55% methanol/0.1% triethylamine, pH 6.0, at 8 min. The methanol concentration was maintained at 55% from 8 to 20 min. The direct quantification of the glucuronide metabolites was difficult due to the low yield of each metabolite; therefore, the metabolic products were quantified based on the molar extinction coefficient of PhIP. To confirm metabolite identity, each HPLC peak was collected from each injection and combined, based on retention time and concentrated under nitrogen for further analysis by mass spectrometry.
Mass spectral analysis
Collected HPLC peaks were dissolved in starting mobile phase and injected into a Micro-Tech Scientific, Ultra Plus μLC system (Sunnyvale, CA) coupled to a Finnigan LCQ mass spectrometer through a Finnigan electrospray interface (Finnigan, San Jose, CA). Samples were separated on a YMC ODS-A column (3.0×250 mm) (Waters, Milford, MA) and eluted at a flow rate of 200 μl/min initially using a solvent concentration of 95% A (solvent A: 2% methanol/97% water/1% acetic acid; solvent B: 95% methanol/4% water/1% acetic acid) for 2 min. This was followed by a linear gradient up to 100% B at 25 min. The methanol concentration was then held at 100% B for 10 min. Full scan and collision induced dissociation (CID) mass spectra were obtained for each metabolite. Mass spectrometer conditions are described elsewhere (13).
To confirm the formation of glucuronides, individual metabolites were treated with 3000 U/ml β-glucuronidase (type VII-A from Escherichia coli) in 25 mM potassium phosphate buffer containing 7.5 mM dithiothreitol, pH 7.0, and incubated for 1 h at 37°C. After the incubation time, samples were placed on dry ice to stop the reaction. The samples were then analyzed for their susceptibility to enzymatic cleavage by HPLC as described above.
Glucuronidation of N-hydroxy-PhIP
Chromatographic analysis of the reaction products from the samples containing microsomes expressing UGT1A1 (Figure 1A) and UGT1A9 (Figure 1C) revealed two metabolite peaks eluting at 10.2 and 11.6 min. Microsomes expressing UGT1A4 produced an additional peak at 6.7 min (Figure 1B). The peak at 18.6 min was not a glucuronide metabolite because it was present in control incubations lacking UDP-glucuronic acid. All three glucuronide compounds were formed in a concentration dependent manner. No metabolite peaks were observed from the UGT1A6 expressing microsomes (data not shown).
Based on mass spectral analysis, susceptibility to enzymatic cleavage and HPLC co-elution with PhIP metabolite standards, the peaks eluting at 10.2 and 11.6 min have been identified as N-hydroxy-PhIP-N2-glucuronide and N-hydroxy-PhIP-N3-glucuronide, respectively. Electrospray interface-MS analysis of each compound produced a primary molecular ion at m/z 417 [M+H]+ and upon MS/MS analysis produced fragment ions at m/z 241 [M+H]+ and m/z 225 [M+H]+ indicating a loss of the glucuronide moiety revealing fragments corresponding to N-hydroxy-PhIP and PhIP. Only the peak at 11.6 min was susceptible to β-glucuronidase cleavage, yielding a compound with chromatographic properties identical to N-hydroxy-PhIP. This is characteristic of N-hydroxy-PhIP-N3-glucuronide (8). In addition, the chromatographic properties of the peaks at 10.2 and 11.6 min were identical to N-hydroxy-PhIP-N2-glucuronide and N-hydroxy-PhIP-N3-glucuronide standards, respectively.
Mass spectral analysis of the undefined peak, eluting at 6.7 min, from microsomes expressing UGT1A4 revealed a primary molecular ion at m/z 417 [M+H]+ and, upon CID, produced a fragment ion at m/z 241 [M+H]+. This peak was susceptible to β-glucuronidase treatment, which is similar to what was seen for N-hydroxy-PhIP-N3-glucuronide. It is believed that this peak represents a previously unknown N-hydroxy-PhIP glucuronide since it was formed in a concentration dependent manner and was not present in control incubations lacking UDP-glucuronic acid or N-hydroxy-PhIP. Further analysis is needed to fully characterize this compound.
UGT1A1 expressing microsomes had the highest capacity to convert N-hydroxy-PhIP to N-hydroxy-PhIP-N2-glucuronide. This metabolite accounted for 86% of the total N-hydroxy-PhIP-glucuronides formed from this isozyme (Figure 2A). UGT1A9 expressing microsomes produced 1000 pmol N-hydroxy-PhIP-N3-glucuronide, accounting for 80% of the total glucuronides formed. In the microsomes expressing UGT1A4, the unknown glucuronide accounted for 31% of the total glucuronides formed while N-hydroxy-PhIP-N3-glucuronide accounted for only 7%.
The rate of formation of the N-hydroxy-PhIP-glucuronides from N-hydroxy-PhIP followed Michaelis–Menton kinetics (Figure 3). When incubations were normalized to milligrams of microsomal protein, apparent Km values for the formation of N-hydroxy-PhIP-N2-glucuronide were 4.58, 39.6 and 22.1 μM for microsomes expressing UGT1A1, UGT1A4 and UGT1A9, respectively (Table I). The strong binding affinity and the relatively high Vmax for the UGT1A1 expressing microsomes produced the highest level of N-hydroxy-PhIP-N2-glucuronide. Even though the reaction rate of the microsomes expressing UGT1A4 was similar to the UGT1A1 expressing microsomes, the high Km value of the UGT1A4 isoform resulted in a lesser amount of N-hydroxy-PhIP-N2-glucuronide formed, indicating the UGT1A1 protein possesses a much higher capacity to conjugate N-hydroxy-PhIP in these microsomal preparations.
Microsomes expressing UGT1A9 had a stronger binding affinity than UGT1A4 but a much slower reaction rate, which subsequently produced the least amount of N-hydroxy-PhIP-N2-glucuronide. The opposite was true for N-hydroxyPhIP-N3-glucuronide formation. The rate of formation of N-hydroxy-PhIP-N3-glucuronide was faster in the UGT1A9 expressing microsomes compared with the UGT1A1 or UGT1A4 isozymes (Figure 3). In UGT1A1 expressing microsomes a strong binding affinity towards N-hydroxy-PhIP-N3-glucuronide was observed, but the reaction rate was so slow that only a relatively small amount of metabolite was formed (Figure 3A). Overall, microsomes expressing UGT1A1 had the strongest substrate binding affinity of the three UGTs tested, whereas UGT1A4 expressing microsomes had the weakest. The reaction rates (Vmax) were similar in both UGT1A1 and UGT1A4 isozymes for both N-hydroxy-PhIP-N2-glucuronide and N-hydroxy-PhIP-N3-glucuronide. The reaction rate for UGT1A9, however, was faster for N-hydroxy-PhIP-N3-glucuronide than for the N2-glucuonide, which was in contrast to what was observed in the UGT1A1 and UGT1A4 expressing microsomes (Table I).
Glucuronidation of PhIP
The formation of direct PhIP-glucuronides was investigated in microsomes expressing UGT1A1, UGT1A4, UGT1A6 and UGT1A9. PhIP-glucuronides were identified by mass spectrometry, susceptibility to β-glucuronidase and HPLC co-elution with authentic standards. Microsomes expressing UGT1A1 and UGT1A9 produced two metabolite peaks eluting at 8.1 and 9.9 min (data not shown). Mass spectral analysis of each peak revealed a primary molecular ion at m/z 401 [M+1]+. Subsequent MS/MS analysis produced ions at m/z 225 [M+1]+, indicating the loss of the glucuronide moiety. The peak eluting at 8.1 min co-eluted with a PhIP-N3-glucuronide standard, and the peak at 9.9 min co-eluted with a PhIP-N2-glucuronide standard. Microsomes expressing UGT1A4 produced a third metabolite eluting at 5.2 min. Mass spectral analysis of this peak also produced a primary molecular ion at m/z 401 [M+1]+, and upon MS/MS the peak fragmented to m/z 225 [M+1]+. Under the conditions used none of the isolated metabolites were susceptible to β-glucuronidase treatment. Further analysis is needed to fully characterize the metabolite that eluted at 5.2 min. No metabolite peaks were observed from the UGT1A6 expressing microsomes.
At a substrate concentration of 50 μM, the overall ability of UGT1A1, UGT1A4 and UGT1A9 to form direct PhIP-glucuronides was 70-, 7- and 83-fold lower, respectively, than the overall ability of these isozymes to form N-hydroxy-PhIP-glucuronides. This clearly indicates that the formation of direct PhIP-glucuronides is a minor reaction compared with N-hydroxy-PhIP conjugation reactions with these isozymes (Figure 2A and B). Interestingly, UGT1A4 expressing microsomes were most active in forming PhIP-N3-glucuronide, whereas in samples containing N-hydroxy-PhIP as substrate, the UGT1A4 protein was the least active in forming the N-hydroxy-PhIP-N3-glucuronide.
Due to the limited number of data points and the low yield of PhIP-glucuronides the Km and Vmax could not be accurately determined. Therefore, the rate of PhIP-glucuronide formation was based on the observed reaction rates. In microsomes expressing UGT1A1, PhIP-N2-glucuronide was generated at a rate ~50-fold faster than PhIP-N3-glucuronide, whereas in the UGT1A4 isoform, PhIP-N3-glucuronide was formed at a much faster rate than PhIP-N2-glucuronide (Figure 4A and B). In microsomes expressing UGT1A9, the rate of formation of PhIP-N2- and PhIP-N3-glucuronide was essentially the same (Figure 4C). In both UGT1A1 and UGT1A9 expressing microsomes the reaction rate began to plateau at a substrate concentration of ~100 and 250 μM PhIP, respectively. No saturation point was observed up to 500 μM PhIP using UGT1A4 expressing microsomes (Figure 4).
Glucuronidation has been established as a major conjugation reaction in the detoxification of many xenobiotics, including PhIP (8,9,20). In humans, N-hydroxy-PhIP-N2-glucuronide was the most prominent metabolite found in urine after administration of PhIP (12), or after consuming a meal of well-cooked meat (13). Little has been reported, however, on the specific UGTs involved in PhIP and N-hydroxy-PhIP glucuronidation. Of the four UGTs tested, UGT1A1, UGT1A4 and UGT1A9 showed catalytic activity towards PhIP and N-hydroxy-PhIP to some degree. The relative overall ability of these three UGTs to conjugate N-hydroxy-PhIP was on average 53-fold greater than their capacity to N-glucuronidate PhIP. This difference clearly indicates that N-hydroxy-PhIP is a much better substrate for these three UGTs compared with PhIP. These findings indicate that CYP1A2-mediated N-hydroxylation is a critical first step in the metabolic process in order for significant glucuronidation to occur. Since the formation of N-hydroxy-PhIP is also the first step in bioactivation, it can be postulated that, in vivo, detoxification via glucuronidation would be competing with phase II bioactivation pathways, such as acetylation and sulfation. The difference between bioactivation and detoxification would then depend on the level of enzyme expression and the binding affinity of each enzyme towards the N-hydroxy-PhIP substrate. The binding efficiency for N-hydroxy-PhIP with acetyltransferase or sulfotransferase is unknown, but if it is greater than the UGT binding efficiency, activation would prevail over detoxification. In the present study, 20–30% of N-hydroxy-PhIP was converted to N-hydroxy-PhIP-N-glucuronides. Theoretically, this suggests that up to 80% of the N-hydroxy-PhIP substrate would be available for further activation. However, the contribution of other UGTs towards inactivation cannot be discounted.
Each of the three active UGTs demonstrated very different catalytic capabilities and specificity. Based on total N- glucuronide formation, the results suggest the following relative ranking of transferase capacity to conjugate N-hydroxy-PhIP: UGT1A1 > UGT1A9 > UGT1A4 >> UGT1A6. The relative transferase ranking for PhIP-N-glucuronide formation was UGT1A4 > UGT1A1 > UGT1A9 >> UGT1A6. The reason for the difference in UGT rank order between N-hydroxy and N-glucuronide formation is, at present, unknown but could be due to differences at the enzyme substrate binding site, or differences in the expression levels of the UGTs in the microsomal preparations. It should be noted that in order to obtain direct PhIP-glucuronides at comparable levels with N-hydroxy-PhIP-glucuronide levels, the PhIP concentration needed to be up to 150-fold higher (depending on UGT isoform) than the N-hydroxy-PhIP levels. It is unclear why N-hydroxy-PhIP was preferentially conjugated by these UGT isozymes over PhIP, but it is most likely due to differences in the interactions between the UGT enzymes and the substrates. This low level of PhIP-N-glucuronidation does not account for the relatively high levels of PhIP-N-glucuronides that have been detected in human urine from individuals exposed to PhIP (12,13). In humans, PhIP-N-glucuronides accounted for 5–10% of all PhIP urinary metabolites, whereas N-hydroxy-PhIP-glucuronides accounted for 55–65% of the urinary metabolites. This difference between the in vitro and in vivo results suggests that other UGTs may be involved in PhIP glucuronidation or that the kinetics are significantly different in vivo.
The variation in the catalytic activity among the three active UGTs suggests that each isozyme interacts with the substrate in very specific ways. The finding that the UGT1A1 expressing microsomes favored the formation of N-hydroxy-PhIP-N2-glucuronide, whereas UGT1A9 expressing microsomes preferentially formed N-hydroxy-PhIP-N3 glucuronide demonstrates the specificity of these isozymes. The difference in substrate interaction within a specific UGT isozyme further suggests differences in isozyme specificity. These results demonstrate the importance of the differential expression of specific UGT isozymes in various tissues. Tissues that express high levels of a certain isozyme can potentially produce very different metabolites than tissues that express high levels of a different isozyme.
The findings from this present study are in contrast to previous reports showing no ability for human UGT1A1 to catalyze the N-glucuronidation of primary or secondary amines, and that N-hydroxy-amines were substrates for UGT1A6 (reviewed in ref. 3). In the present study UGT1A1 readily metabolized N-hydroxy-PhIP, and UGT1A6 showed no activity towards PhIP or N-hydroxy-PhIP. Also, previous reports have shown that N-glucuronidation of 4-aminobiphenyl, β-naphthylamine and benzidine, and its analogs, was catalyzed by UGT1A4 with high efficiency (21,22), whereas in the present study UGT1A4 had a low efficiency towards the heterocyclic amine substrate compared with the other UGTs tested. Furthermore, Nowell et al. (19) showed no capacity for UGT1A1 or UGT1A4 to catalyze the glucuronidation of N-hydroxy-PhIP. The disparity between these studies and the present one cannot presently be explained, but it could be due to the source of the UGT proteins and/or the incubation conditions. Galijatovic et al. (23) showed that UGT1A1 could catalyze the N-glucuronidation of N-hydroxy-PhIP when using UGT1A1 expressing microsomes obtained from the same source as the present study. Furthermore, the data presented in this study provide significant evidence to support the conclusion that N-hydroxy-PhIP is a substrate for UGT1A1, UGT1A4 and UGT1A9 catalyzed N-glucuronidation. Based on these and previous results (3,22,24) it appears that UGT1A isozymes are highly substrate specific. To determine if the specificity of these UGT isozymes towards other heterocyclic amines of the aminoimidazoazaarene class are similar to PhIP, experiments using other food-borne heterocyclic amines as substrates for these UGT isozymes are ongoing in our laboratory.
Since N-hydroxy-PhIP-N2-glucuronide was the predominant metabolite detected in humans in vivo (12,13) and UGT1A1 was the most active transferase in catalyzing the formation of N-hydroxy-PhIP-N2-glucuronide in this present study, UGT1A1 would appear to be an important enzyme in human metabolism of N-hydroxy-PhIP. This implies that tissues with a high level of expression of UGT1A1 will have a high capacity to detoxify N-hydroxy-PhIP. Previous studies have determined that the liver is the predominant site for UGT1A1 activity (25), although the colon has also shown appreciable levels of UGT1A1 (26). In addition, polymorphic distribution of UGT1A1 has been observed in human gastric epithelium (2). Therefore, individuals with a high expression of UGT1A1 in these tissues should have a higher capacity to detoxify N-hydroxy-PhIP by glucuronide conjugation than individuals with low UGT1A1 expression. In contrast, individuals who express high levels of UGT1A9 (also found in liver and colon) would be predisposed to form higher quantities of N- hydroxy-PhIP-N3-glucuronide. This is important since the N3-glucuronides can be de-conjugated by bacterial β-glucuronidase (found in human intestine and colon) to N-hydroxy-PhIP which is again available for bioactivation. By understanding which UGT isozymes are involved in PhIP and N-hydroxy-PhIP glucuronidation, and the tissue specificity of each isozyme, a better understanding of risk from PhIP exposure can be accomplished. Individuals with high levels of specific UGTs will be at less risk than individuals with low levels, since their capacity to detoxify PhIP will be increased.
|Km (μM)||Vmaxa||Km (μM)||Vmaxa||Km (μM)||Vmaxa|
|aVmax is expressed as pmol/min/mg protein.|
|bND, not detected.|
|Km (μM)||Vmaxa||Km (μM)||Vmaxa||Km (μM)||Vmaxa|
|aVmax is expressed as pmol/min/mg protein.|
|bND, not detected.|
This work was performed under the auspices of the US Department of Energy by the University of California, Lawrence Livermore National Laboratory under contract no. W-7405-Eng-48 and supported by NCI grant CA55861.