A Simple Method to Evaluate Reactivity of Acylglucuronides Optimized for Early Stage Drug Discovery

Abstract

Drugs containing the carboxylic functional group can be metabolized to form acylglucuronides believed to cause idiosyncratic drug toxicity when the acylglucuronide is unstable. Recent studies have shown that the half-life of an acylglucuronide in phosphate buffer is the best means for classifying acylglucuronides into safe, warning, and withdrawn drugs. However, it is difficult to halt the late stage development of new chemical entities due to the instability of their acylglucuronides. We report an optimized in vitro method for determining the half-lives of acylglucuronides in simple phosphate buffer without the need for authentic standards. The experiment was divided into two incubations. In the first incubation, acylglucuronide was synthesized by human liver microsomes, and in the second incubation, the degradation rate of acylglucuronide in phosphate buffer was determined. The degradation rate constants of acylglucuronides were determined from changes in the LC-MS/MS peak area and the half-lives were calculated. We evaluated the half-lives of 10 drugs: 3 safe drugs (telmisartan, gemfibrozil and flufenamic acid) and 7 withdrawn or warning drugs (zomepirac, diclofenac, furosemide, ibuprofen, S-naproxen, probenecid and tolmetin). The half-lives of the 3 safe drugs were 10.6 h or longer, whereas the half-lives of the 7 withdrawn or warning drugs were 4.0 h or shorter. Although authentic acylglucuronide standards were not used, we obtained half-lives of acylglucuronides in phosphate buffer similar to those reported previously. Using this method, the risk of reactivity caused by acylglucuronides can be evaluated in the early stages of drug discovery.

Acyl glucuronidation is a major metabolic pathway for many drugs containing carboxylic acid groups. These detoxification pathways typically employ phase II enzymes. However, some acylglucuronides are unstable under physiological conditions and undergo hydrolysis or intramolecular rearrangement, such as the migration of the drug moiety from the 1-O-β position to the C-2, C-3, or C-4 position on the glucuronic acid ring.^1,2) The resulting reactive acylglucuronides and isomers could potentially bind covalently to cellular macromolecules. Furthermore, it has been reported that covalent binding with proteins correlates to the risk of idiosyncratic drug toxicity (IDT).^3,4)

To assess chemical reactivity of an acylglucuronide, it is meaningful to evaluate its stability in simple buffer, buffer containing human serum albumin (HSA), or fresh human plasma. In a pioneering study, Benet et al. showed good correlation between the extent of covalent binding to HSA and the apparent first-order degradation rate of the acylglucuronides of six drugs in buffer containing HSA.⁵⁾ Subsequently, several groups determined the half-lives of these acylglucuronides in simple buffer, buffer containing HSA, or fresh human plasma.^2,6,7)

Recently, Sawamura et al. determined the half-lives of 21 acylglucuronides using authentic acylglucuronide standards in phosphate buffer, phosphate buffer containing 4% HSA, or fresh human plasma, and showed that the half-lives determined in simple buffer were most useful for classifying drugs into safe, warning or withdrawn drugs.⁸⁾ However, it is difficult to discontinue the clinical development of New Chemical Entities (NCEs) on the grounds of the instability of their acylglucuronides at a late stage of drug development. Furthermore, the preparation of authentic acylglucuronide is often the limiting factor for evaluating the highly reactive acylglucuronides of NCEs at an early stage of drug discovery. Bolze et al. developed an in vitro screening model to determine the reactivity of acylglucuronides derived from carboxylic acid and human liver microsomes,⁷⁾ and Chen et al. developed a simple in vitro model that did not require the use of β-glucuronidase.⁹⁾ However, these studies evaluated the stabilities of acylglucuronides in HSA solution or in fresh human plasma which is known to be less predictive for the risk of IDT. We here describe an optimized in vitro model to determine the half-lives of acylglucuronides in simple phosphate buffer without the need for authentic standards, and report the correlation between the half-lives determined using our method with those from previous studies which used authentic acylglucuronide standards.

Materials and Methods

The experiment was divided into two incubations. The first incubation (biosynthesis phase) was conducted to synthesize specific acylglucuronides by human liver microsomes. The second incubation (“reactivity” phase) was conducted to determine the degradation rate of each acylglucuronide in phosphate buffer.

Chemicals and Reagents

Telmisartan, gemfibrozil, diclofenac, furosemide, ibuprofen probenecid, tolmetin, flufenamic acid, zomepirac, S-naproxen, niflumic acid, reserpine and alamethicin were purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.). Niflumic acid was used as an internal standard (IS) for the acylglucuronides detected by negative ionization mode, and with regard to an IS on positive ionization mode, reserpine was used as an IS. Pooled human liver microsomes prepared by combining the liver microsomal fractions from 50 donors (20 mg protein/mL) were purchased from Xenotech (Lenexa, KS, U.S.A.). UDP-glucuronic acid was purchased from Nacalai Tesque (Kyoto, Japan). All other reagents and solvents were of analytical grade or special grade.

Categorization of Carboxylic Acid-Containing Drugs

Ten test compounds were classified as “Safe” and “Withdrawn or Warning” based on previous reports.⁸⁾ The “Safe” category included three drugs: flufenamic acid, gemfibrozil and telmisartan. The “Withdrawn or Warning” category included seven drugs: zomepirac, diclofenac, furosemide, ibuprofen, naproxen, probenecid and tolmetin (Table 1).

Table 1. Lists of Tested Drugs and MS/MS Condition for Their Acylglucuronides Detection

Drug IDT risk	Toxicity*	Maximum daily dose* (mg)	MS/MS condition for acylglucuronide detection
Withdrawn or warning drugs
Zomepirac	Anaphylaxis, DILIa)	600	Positive	Q1: 468.1	Q3: 292.1
Tolmetin	Anaphylaxis, DILI, SJSb)	1800	Positive	Q1: 434.0	Q3: 258.0
Probenecid	Anaphylaxis	1000	Negative	Q1: 460.0	Q3: 284.0
Diclofenac	Anaphylaxis, DILI, SJS	150	Negative	Q1: 469.9	Q3: 193.0
Naproxen	Anaphylaxis, SJS	1000	Negative	Q1: 405.0	Q3: 193.0
Ibuprofen	Anaphylaxis, SJS	3200	Negative	Q1: 381.0	Q3: 205.0
Furosemide	Neutropenia, Thrombocytopenia	80	Negative	Q1: 504.9	Q3: 328.9
Safe drugs
Flufenamic acid		750	Negative	Q1: 455.9	Q3: 279.9
Telmisartan		80	Negative	Q1: 689.2	Q3: 513.2
Gemfibrozil		1200	Negative	Q1: 425.1	Q3: 249.1

a) DILI: Drug-induced liver injury, b) SJS : Stevens–Johnson syndrome, * Obtained from Sawamura et al.⁸⁾

In Vitro Biosynthesis of Acylglucuronide

The ten test compounds (Fig. 1) were incubated at 10 µM for 60 min at 37°C with pooled human liver microsomes (1.0 mg protein/mL) in 100 mM Tris buffer, pH 7.4, containing 3 mM MgCl₂, 25 µg/mL alamethicin, and 5 mM UDP-glucuronic acid. The volume of each reaction mixture was 100 µL. The reaction was stopped by precipitating the protein by adding an equal volume of acetonitrile/methanol (9/1) containing 1% formic acid and the internal standard (0.1 µM niflumic acid and reserpine). The reaction mixture was centrifuged at 3000 rpm for 20 min at 4°C, then the supernatants were collected and stored at −80°C until use in the second incubation.

Fig. 1. Structures of the Compounds Used in This Study

A) Safe drugs, B) Withdrawn or warning drugs.

Chemical Stability of Acylglucuronides in Buffer

The supernatants were transferred to new reaction tubes and mixed with 9 volumes of 250 mM potassium phosphate buffer, pH 7.4. The samples were incubated at 37°C and aliquots were removed at 0, 60, 120 and 240 min, except for telmisartan and gemfibrozil; for these drugs, aliquots were removed at 0, 16 and 24 h. The reaction was stopped by adding 0.4 volume of 20% acetic acid and one volume of 0.1% formic acid/acetonitrile (4/1). Samples were stored at −80°C until analysis.

LC-MS/MS Analytical Method

The above samples were analyzed by liquid chromatography-tandem mass spectrometry. HPLC was conducted using a LC-20A system (Shimadzu, Kyoto, Japan) equipped with a Sunniest C18 HT analytical column (2.0 µm, 2.1 mm×100 mm; ChromaNik Technologies Inc., Osaka, Japan). The acylglucuronides were separated from other isomers and metabolites using an elution gradient. The mobile phase was a mixture of 10 mM ammonium acetate containing 0.5% acetic acid (solvent A) and neat acetonitrile (solvent B). The column was eluted at a flow rate of 0.35 mL/min at 40°C. The elution program was: linear gradient from 10% to 73% solvent B over 22.0 min; increase from 73% to 95% B over 0.1 min; hold at 95% B until 26.5 min; decrease B from 95% to 10% over 0.1 min; hold B constant until 35.0 min. Detection and semi-quantification of acylglucuronides were performed by tandem mass spectrometry using a 4000QTRAP (AB SCEX, Framingham, MA, U.S.A.).

Identification and Detection of Acylglucuronides

MS/MS conditions for the detection of acylglucuronides were optimized for each aglycone following previous report.¹⁰⁾ The acylglucuronides detected by positive ionization mode, zomepirac-acylglucuronide and tolmetin-acylglucronide, were monitored using two MS/MS conditions: 1) Q1:Parent Q1+176 Da Q3:Parent Q1, and 2) Q1:Parent Q1+176 Da Q3:Parent Q3 (Fig. 2 A). Other acylglucuronides detected by negative ionization mode were monitored using three MS/MS conditions: 1) Q1:Parent Q1+176 Da Q3:Parent Q1, 2) Q1:Parent Q1+176 Da Q3:Parent Q3, and 3) Q1:Parent Q1+176 Da Q3:193 Da (Fig. 2 B). We chose the most sensitive and selective condition for data analysis. If several peaks were detected in the supernatant from the first incubation, the most unstable peak was assigned as 1-O-acylglucuronide. The identity of the 1-O-acylglucuronide was determined by treatment with β-glucuronidase, if necessary. After the first incubation, glucuronidation was stopped by the addition of UDP (10 mM final concentration) and the sample was incubated with 1000 units of helix β-glucuronidase at 37°C for 2 h to hydrolyze the 1-O-acylglucuronides.

Fig. 2. Examples of Proposed Major Fragmentation Pathway for Acyl Glucuronides Using LC-MS/MS

A) Tolmetin acylglucuronide in positive ionization mode. B) Diclofenac acylglucuronide in negative ionization mode.

After the second incubation, the areas of LC-MS/MS peaks arising from acylglucuronides were monitored with time and changes in the peak areas were evaluated.

Data Analysis

The degradation rate constant (K) of each acylglucuronide was determined from the LC-MS/MS peak area versus time curve by linear regression of the semi-logarithmic plot.

The half-lives (T_1/2) were calculated from K by the following equation:

  

Results

Identification of Acylglucuronides

Biosynthesis of acylglucuronides was observed for all tested drugs. Acylglucuronide detected by MS/MS was assigned as 1-O-acylglucuronide and its stability in buffer was evaluated. If several peaks arising from acylglucuronides were detected by MS/MS, the most unstable peak in buffer was assigned as 1-O-acylglucuronide. In addition, these peaks were confirmed as 1-O-acylglucuronide using β-glucuronidase (data not shown).

Half-lives of Acylglucuronides in Buffer and Correlation between Observed Half-lives and Previously Reported Half-lives

We evaluated the half-lives of 10 drugs (3 safe drugs: telmisartan, gemfibrozil and flufenamic acid; 7 withdrawn or warning drugs: zomepirac, diclofenac, furosemide, ibuprofen, S-naproxen, probenecid and tolmetin).

From chemical stability studies of these acylglucuronides, their half-lives in buffer were calculated (Fig. 3). All acylglucuronides, except zomepirac- and probenecid-acylglucuronide, showed single phase degradation in buffer. Because zomepirac- and probenecid-acylglucuronide degradated following biphasic kinetics, its degradation rate constant was calculated from the first degradation phase. All 3 safe drugs had half-lives of 10.6 h or longer, whereas the half-lives of all 7 withdrawn or warning drugs were 4.05 h or shorter (Table 2). Although no authentic acylglucuronide standard was used, our obtained half-lives for acylglucuronides in phosphate buffer were similar to those reported in previous studies^2,8) and could be used to distinguish between ‘safe’ and ‘withdrawn or warning’ drugs. In the previous study, the classification value of the half-life in phosphate buffer that separated the ‘safe’ drugs from the ‘withdrawn’ drugs was calculated to be 3.6 h, and the half-lives of warning drugs (except mefenamic acid) were shorter than 3.6 h. The half-lives of acylglucuronides in the ‘safe’ category were 7.2 h (the half-life of flufenamic acid) or longer, whereas the half-life of the ‘withdrawn’ category was 3.2 h (the half-life of furosemide) or shorter (except for mefenamic acid). These results indicate that it may be possible to distinguish ‘safe’ drugs from ‘withdrawn or warning’ drugs by using flufenamic acid and furosemide as benchmarks.

Fig. 3. Stability of Acylglucuronides in Potassium Phosphate Buffer, pH 7.4

A) Withdrawn or warning drugs, B) Safe drugs.

Table 2. Half-lives of Acylglucuronides in Potassium Phosphate Buffer, pH 7.4

Compounds	Half-lives (h)
	This study	Previous study 1	Previous study 2
Withdrawn or warning drugs
Zomepirac	0.643±0.02	0.45	0.4
Tolmetin	0.460±0.02	0.26	0.4
Probenecid	0.450±0.16	0.4	0.3
Diclofenac	0.783±0.06	0.51	0.7
S-Naproxen	2.89±0.09	1.8	2.2
Ibuprofen	4.05±0.70	3.3	2.7
Furosemide	4.00±0.53	5.3	3.2
Safe drugs
Flufenamic acid	10.6±1.21	7	7.2
Telmisartan	61.8±4.86	26	45.6
Gemfibrozil	74.3±5.89	44	71.4

    * Obtained from Ebner et al.²⁾, ** Obtained from Sawamura et al. ⁸⁾

Discussion

Assessing the stability of acylglucuronides requires authentic standards. These standards can be prepared synthetically, isolated from the bile of animals following dosing with the acidic drug, or isolated following incubation of the drug with microsomes. The preparation of acylglucuronides is often the limiting factor in the early detection of reactive acylglucuronides of NCEs. To help circumvent this, methods not requiring authentic acylglucuronide standards were developed by Bolze et al.⁷⁾ and Chen et al.⁹⁾ However, these methods require the incubation of acylglucuronides in phosphate buffer with HSA or in fresh human plasma to evaluate their chemical stabilities. As previously reported by Sawamura et al.⁸⁾ the half-lives of acylglucuronides in simple phosphate buffer most accurately distinguished ‘safe’ drugs from ‘warning’ drugs, with the exception of mefenamic acid, and ‘withdrawn’ drugs. In light of these studies, we developed a new method that does not require authentic acylglucuronide standards and evaluated the stabilities of acylglucuronides in simple phosphate buffer. The half-lives of acylglucuronides obtained using our method correlated well with those obtained by Sawamura et al.⁸⁾ (Fig. 4).

Fig. 4. Zone Classifications Determined from Half-lives of Acylglucuronides in Potassium Phosphate Buffer, pH 7.4

Bolze et al. reported an excellent correlation between the extent of drug covalent binding to albumin and the aglycone appearance rate constant weighted by the percentage of isomerization.⁷⁾ Therefore, of the percentage of isomerization can be an appropriate indicator to differentiate ‘safe’ drugs and ‘warning’ drugs.

The extent of covalent binding has been an indicator for IDT.¹¹⁾ The risk of IDT can be ranked using the extent of drug covalent binding and daily dose.^12,13) Similarly, for acylglucuronide, both the degradation ratio and the daily dose must be considered in order to accurately assess the risk from acylglucuronides. Furthermore, the contribution ratio of acylglucuronidation should be considered for all metabolic pathways, since some parent compounds of acylglucuronides could be metabolized to other metabolites, including reactive metabolites.

In conclusion, we have established an optimized method to evaluate the stability of acylglucuronides, without the need for authentic standards. This method allows the risk of acylglucuronides to be evaluated at an early stage of drug discovery, reducing the potential risk to discontinue the clinical development of NCEs at a late stage of drug development.

REFERENCES

• 1)  Spahn-Langguth H, Benet LZ. Acyl glucuronides revisited: is the glucuronidation process a toxification as well as a detoxification mechanism? Drug Metab. Rev., 24, 5–47 (1992).
• 2)  Ebner T, Heinzel G, Prox A, Beschke K, Wachsmuth H. Disposition and chemical stability of telmisartan 1-O-acylglucuronide. Drug Metab. Dispos., 27, 1143–1149 (1999).
• 3)  Uetrecht J. Prediction of a new drug’s potential to cause idiosyncratic reactions. Curr. Opin. Drug Discov. Devel., 4, 55–59 (2001).
• 4)  Zhou SF, Xue CC, Yu XQ, Li C, Wang G. Clinically important drug interactions potentially involving mechanism-based inhibition of cytochrome P450 3A4 and the role of therapeutic drug monitoring. Ther. Drug Monit., 29, 687–710 (2007).
• 5)  Benet LZ, Spahn-Langguth H, Iwakawa S, Volland C, Mizuma T, Mayer S, Mutschler E, Lin ET. Predictability of the covalent binding of acidic drugs in man. Life Sci., 53, PL141–PL146 (1993).
• 6)  Boelsterli UA. Xenobiotic acyl glucuronides and acyl CoA thioesters as protein-reactive metabolites with the potential to cause idiosyncratic drug reactions. Curr. Drug Metab., 3, 439–450 (2002).
• 7)  Bolze S, Bromet N, Gay-Feutry C, Massiere F, Boulieu R, Hulot T. Development of an in vitro screening model for the biosynthesis of acylglucuronide metabolites and the assessment of their reactivity toward human serum albumin. Drug Metab. Dispos., 30, 404–413 (2002).
• 8)  Sawamura R, Okudaira N, Watanabe K, Murai T, Kobayashi Y, Tachibana M, Ohnuki T, Masuda K, Honma H, Kurihara A, Okazaki O. Predictability of idiosyncratic drug toxicity risk for carboxylic acid-containing drugs based on the chemical stability of acyl glucuronide. Drug Metab. Dispos., 38, 1857–1864 (2010).
• 9)  Chen Z, Holt TG, Pivnichny JV, Leung K. A simple in vitro model to study the stability of acylglucuronides. J. Pharmacol. Toxicol. Methods, 55, 91–95 (2007).
• 10)  Bolze S, Lacombe O, Durand G, Chaimbault P, Massière F, Gay-Feutry C, Bromet N, Hulot T. Standardization of a LC/MS/MS method for the determination of acyl glucuronides and their isomers. Curr. Sep., 20, 55–59 (2002).
• 11)  Evans DC, Watt AP, Nicoll-Griffith DA, Baillie TA. Drug–protein adducts: an industry perspective on minimizing the potential for drug bioactivation in drug discovery and development. Chem. Res. Toxicol., 17, 3–16 (2004).
• 12)  Nakayama S, Atsumi R, Takakusa H, Kobayashi Y, Kurihara A, Nagai Y, Nakai D, Okazaki O. A zone classification system for risk assessment of idiosyncratic drug toxicity using daily dose and covalent binding. Drug Metab. Dispos., 37, 1970–1977 (2009).
• 13)  Usui T, Mise M, Hashizume T, Yabuki M, Komuro S. Evaluation of the potential for drug-induced liver injury based on in vitro covalent binding to human liver proteins. Drug Metab. Dispos., 37, 2383–2392 (2009).