The use of stable isotopes in drug metabolism studies

The Use of Stable Isotopes in Drug Metabolism Studies Fred P. Abramson

Although there is a long history of stable isotopes use in drug metabolism research, it is appropriate to evaluate them in pregnancy drug studies in which safety takes highest priority. It is well established through a number of human and animal experiments that stable isotopes themselves rarely generate additional toxicities beyond the molecules to which they are attached. For the analysis of stable isotopes involved in metabolism studies, mass spectrometry plays the predominant role. Several mass spectrometry-based techniques now exist that enable the selective quantitative detection of stable isotopes with better sensitivity and better retention of chromatographic resolution than do in-line radioactivity monitors for 14C. Even mass balance studies can be performed by using stable isotopes, a type of experiment that still predominantly uses radioisotopes. Some of the newest developments in the use of stable isotopes involve biopolymers, in which fully isotope-labeled species can be generated from cells grown in isotopically labeled growth media. Having shown safety, sensitivity, specificity, and versatility, stable isotopes should play an important role in drug metabolism studies in pregnancy. Copyright 9 2001 by W.B. Saunders C o m p a n y

his article provides an orientation to the

T use of stable isotopes in drug metabolism to explain why this topic is so important in studies with pregnant women. Table 1 gives the stable isotopes most commonly used in these studies. Stable isotopes have a long history in drug metabolism research. Baillie 2 surveyed the field 20 years ago providing an orientation to the use in quantitative, qualitative, and mechanistic investigations. A Medline search for reviews of stable isotopes and pharmacology since then yields only 3 articles, 3-5 none more recent than 1989. Many more reviews exist regarding stable isotopes in biochemistry studies, where endogenous pathways, protein synthesis rates, are frequently performed. The major difference between pharmacological and biochemical studies is that the substances examined are exogenous, not endogenous; therefore, no unlabeled analogs exist in the body for drugs and their metabolites. In studies of drugs in pregnancy in which safety takes highest priority, confirming the freed o m from toxicity for stable isotopes is important. It is well established through a n u m b e r of h u m a n and animal experiments that stable isotopes themselves rarely generate additional toxicities beyond the molecules to which they are attached. The review by Koletzko et al6 empha-

sizes these qualities. Because of the natural abundance of stable isotopes, concerns about these species can be allayed by noting our natural body burden of stable isotopes. We already have 2,000 m g / k g of 13C in our body's carboncontaining molecules. Animals have been loaded up to 60% with 13C without apparent toxicities, and neonates have been loaded with 1.5 g / k g of 180 without harm. Special problems exist for deuterium, in which metabolic isotope effects can affect rates of reaction when the isotope is in a metabolically-labile position. One should remember that such isotope effects will be greater for the radioactive isotope of H or C, because such radioisotopes weigh 1 mass unit more than the nonradioactive component, causing yet greater isotope effects. A symposium was held in 1989 on "Stable Isotopes in Paediatric Nutritional and Metabolic Research" and its proFrom the Departments of Pharmacology and Chemistry, George Washington University, School of Medicine and Health Sciences, Washington, DC. The work has been supported by United States Public Health Service (USPHS) Grants NIH RO1-GM36143, RO1-GM58623, and RO1RR13300. Address reprint requests to Fred P. Abramson, PhD, Department of Pharmacology, RE6640, George Washington University School of Medicine, 2300 1 St NW,, Washington, DC 2003Z Copyright 9 2001 by W.B. Saunders Company 0146-0005/01/2503-0004535.00/0 doi:l O.1053/sper.2001.2 4568

Seminars in Perinatology, Vol 25, No 3 (June), 2001: pp 133-138

133

Fred P. Abramson

134

T a b l e l. Common Stable Isotopes for Drug

Metabolism Studies

Isotope 2H 13C z5N zsO

Natural Abundance* (%) 0.015 1.17 0.37 0.20

a d d e d label. O n e c a n n o t always know this when the metabolite may have a substantially different structure that leads to a different m o d e o f fragmentation. A hypothetical e x a m p l e o f the success a n d failure of the "twin-ion" technique is p r e s e n t e d in Figure 1. T h e alternative to the "twin-ion" technique is

* Data from rcf 1. <

~0

ceedings published in 1990. 7 These articles clearly show the r a n g e o f uses of stable isotopes in a n o t h e r p o p u l a t i o n where radioisotopes would n o t be used for n o n t h e r a p e u t i c purposes. O n e can imagine similar studies being perf o r m e d in p r e g n a n t women, with follow-up analyses in their newborns. For the analysis o f stable isotopes involved in m e t a b o l i s m studies, mass s p e c t r o m e t r y plays the p r e d o m i n a n t role. N u c l e a r m a g n e t i c reson a n c e assists in the determination o f metabolite structures, but currently lacks the sensitivity, selectivity, a n d speed to c o m p e t e with mass spectrometry for analysis o f metabolites f r o m biological matrices separated by gas c h r o m a t o g r a p h y or high p e r f o r m a n c e liquid c h r o m a t o g r a p h y (HPLC). Two strategies exist for the recognition o f d r u g metabolites: the use of isotope patterns a n d isotope ratio monitoring. Much o f the research to identify d r u g metabolites uses a version o f isotope pattern known as the "twin-ion" technique, a m e t h o d popularized by K n a p p et al. 8 T h e y synthesized nortriptyline containing 3 atoms o f deuterium, generating a species whose molecular weight was 3 masses above the n o r m a l species. After administering a dose containing 50% n o r m a l and 50% trideuterated nortriptyline, they observed a series of comp o u n d s in their mass-spectrometric analyses that were of equal height but 3 masses apart. No naturally occurring isotope cluster has that appearance, so finding equal peak heights for masses 3 U apart was diagnostic of the presence of n0rtriptyline or 1 o f its metabolites. T h e "twin-ion" technique d e p e n d s on 2 assumptions that are n o t always fulfilled, thus allowing a degree of uncertainty regarding the c o m p r e h e n siveness of that m e t h o d . First, one has to have the drug or its metabolite sufficiently free o f interference that the twin ions can be found. Second, each metabolite has to have m a j o r components o f its mass spectrum that retain the

TARGE'TION + 3

A

so 40 2o

LIJ O z
0~0

IIIIIIIIIIIIIIHH]IIIIIItlIIIIIIIIIIIIIII 37o

3gO

I]llllllllllll[lll[lllllllllllll]ll]lllllll[] 410

MASS/CHARGE 100

s

< TARGETION + 3

z m <=:

80

LU

40

I.-

20

B

>_

5 w

n.-

II[tlllflllJIIllll[lllItlllllllllllllll[rlll[llllllllll[llrlli IIIIIIIIIIl[lll[lllllllllrlll[I ~

37O

~0

410

43O

~0

MASS/CHARGE

a0

FRAGMENTATION NOT INCLUDING LABELED MOIETY? INTERFERING MATRIX PEAKS? TARGET]ON * 3

C 0 3SO

37O

~0

410

,~0

4~0

MA,~HARGE

Figure 1. The "twin-ion" technique. These 3 graphs simulate possible situations when using the "twin-ion technique. (A) The molecular ions for 2 forms of a drug: an unmodified one having MW = 400 along with 1 having 3 heavy atoms substituted, MW = 403. The unique pattern of equally abundant ions spaced 3 masses apart is obvious. (B) A monohydroxylated metabolite with the expected shift of 16 mass units caused by the new atom of oxygen. Again, the appearance of equally abundant ions spaced 3 masses apart relates this species to the parent species. (C) The mass spectrum of the N-acetylated metabolite. Here, the MWs are 442 and 445, but these ions are not prominent because of other components in this spectrum. The possibility of interferences from the biological matrix overwhelming the spectrum of'the metabolite, and the possibility that this particular chemically-modified molecule has a mode of fragmentation that does not retain the site containing the three heavy atoms both have to be considered. It is likely that this metabolite would not be identified from this sample.

Stable Isotopes in Drug Metabolism

called isotope-ratio monitoring. In this scheme, all analytes are d e c o m p o s e d so that a given elem e n t and thereby its isotopic a b u n d a n c e is recognized in a way that is i n d e p e n d e n t o f the structure from which it derived. For example, any c a r b o n - c o n a i n i n g species might be cornbusted to generate CO 2 and the e n r i c h m e n t of a~C would be d e t e r m i n e d by monitoring the ratio of 1~CO2/12CO2 at m/z 45 and 44, respectively. In this manner, n o matter what the structure of the metabolite was, n o matter what its mass was, and n o matter what its m o d e o f flagmentation might be if directly analyzed by mass spectrometry, the m/z 4 5 / 4 4 ratio would indicate e n r i c h m e n t and confirm that species as related to the lSC-labeled starting material. T h e first example of isotope-ratio monitoring in pharmaceutical research was published in 19769 with many examples continuing to the present time. We have b e e n particularly interested in exploring the use o f isotope-ratio monitoring in pharmacology for a technique called CRIMS (chemical reaction interface mass spectrometry) that has many features that assist drug metabolism studies, a0 CRIMS uses a microwaveinduced helium plasma to decompose all analyres and adds a reactant gas to complete oxidation, reduction, or fluorination reactions. With the use o f CRIMS, or o t h e r isotope-ratio monitoring methods, the data can be manipulated to generate chromatograms that show only enriched species while ignoring all o t h e r backg r o u n d components. This is accomplished by a subtraction. T h e natural a b u n d a n c e of the min o r isotope required by the intensity o f the maj o r isotope is subtracted from the mass channel monitoring the e n r i c h e d isotope. For 13CO2 this means that 0.0119 times the intensity o f 12CO~ at m/z 44 is subtracted from the intensity o f m/z 45. In this way, all analytes having the natural 0.0119 a b u n d a n c e of 13CO2 will disappear and only those chromatographic peaks that have more than that abundance will remain visible. This process is illustrated in Figure 2. Carbon-14 has been the mainstay o f drug metabolism studies. Radiotracers, no matter how low the dose, will not be preferred in p r e g n a n t women, but researchers have legitimate concerns about whether a sable-isotope-based m e t h o d has the requisite sensitivity and selectivity to accomplish their goals. Two recent studies have compared 14C with s a b l e isotopes in drug

135

metabolism studies. 11,a2 Abramson et a111 gave a dose of the neurosteroid tirilazad containing unlabeled drug, 13C, X~N-labeled drug, and laGlabeled drug to monkeys, collected bile, and examined it for the pattern o f metabolites. HPLC separations were p e r f o r m e d with either an in-line radioactivity m o n i t o r (RAM) or a CRIMS system as detectors. T h e data regarding signal/noise, sensitivity, and ability to maintain chromatographic resolution all favored CRIMS and s a b l e isotopes c o m p a r e d with RAM and X4C. The 8 largest peaks in the metabolite pattern were quantified by each m e t h o d and the correlation coefficients (r 2) comparing 13C with RAM and 15N with RAM were 0.90 and 0.86 respectively. Goldthwaite et al a2 showed a good correlation in the m e t a b o l i t e / p a r e n t drug ratios for buspirone between H P L C / R A M for 14C and HPLC/CRIMS for I~N. These studies support the use o f s a b l e isotopes in drug metabolism studies where radioisotopes have previously been desired. Even mass balance studies can be p e r f o r m e d by using s a b l e isotopes, a type of experiment that still predominantly uses radioisotopes. The most recent attempt to p e r f o r m this type o f experiment was r e p o r t e d by C h e n et al.a3 References to the three earlier studies are cited there. Rats were dosed with two forms o f azidothymidine (AZT), one containing 13C, and the o t h e r containing 14C. Urine and feces were collected and analyzed both by scintillation counting and by flow-injection analysis using a chemical reaction interface o n a multicollector isotope-ratio mass spectrometer (IRMS). G o o d agreement was f o u n d between 13C and X4C. Chen et al a~ conclude that as little as 1% excretion can be d e t e r m i n e d by their m e t h o d for doses of 1 to 2 mg/kg. This level o f p e r f o r m a n c e will include many drugs, although highly p o t e n t agents will not be detectable at the 1% level. Another important use of s a b l e isotopes for drug disposition involves bioavailablity studies. The idea of using an intravenous dose of a stable-isotope labeled drug with the simultaneous administration o f an unlabeled oral dose o f a drug was first i n t r o d u c e d by Strong et ala4 in 1975. From this protocol, blood samples are oba i n e d from which both the concentration of the labeled intravenous (IV) form o f the drug, and the unlabeled oral f o r m of the drug are determined, based o n the mass difference between

136

Fred P. Abramson

z

o_ t-LU

z

X1

L

/

m/z 44 CHROMATOGRAM

(GENERAL120)

Xl O0

A

m/z 45 CHROMATOGRAM (GENERAL 13C)

L

THE EQUATION M/Z 4 5 - 0 . 0 1 1 9 * M / Z 44 GIVES A

TIME

>

13C - selective CRIMS CHROMATOGRAM

Figure 2. Generating an enrichment-only chromatogram. These 3 simulated chromatograms show how isotoperatio monitoring data can be manipulated to suppress unenriched species and thereby generate a chromatogram 9 showing only enriched species. The top trace shows all carbon-12-containing species at m/z 44, 1 2 CO 2. The 13 -middle trace shows all carbon-13-containing species at m/z 45, CO 2 with a lesser amount of 1 2 C 1 6 O 1 7 O. The natural abundance of lSco2 and 1 2 C 1 6 0 1 7 ( ~ ) compared to 12CO2 is 0.01199 When an amount equal to 09 9 m/z 44 abundance is subtracted from the abundance of the m/z 45 trace, all unenriched peaks disappear. In the bottom trace only those 4 peaks containing more 13C than natural abundance remain. The area under each of these peaks is quantitatively related to the amount of e x c e s s 1 3 C in the original analyte species. the labeled a n d unlabeled species. T h e advantage h e r e is that b o t h d r u g profiles c o m e f r o m the same patient at the same times, so that interexperimental variation is minimized. In this way, the n u m b e r of patients n e e d e d is greatly r e d u c e d c o m p a r e d with the usual scheme o f dosing s o m e patients IV, o t h e r subjects orally, a n d making the g r o u p comparisons. Stable isotopes also have an extensive history o f use to evaluate mechanisms of metabolism. Substituting a heavy isotope for o n e or m o r e atoms o f hydrogen, carbon, oxygen, or nitrogen may be useful in u n d e r s t a n d i n g pathways, reaction intermediates. D e u t e r i u m isotope effects are particularly effective tools because o f the substantial difference in rates of C-H vs. C-D b o n d dissociations. Baillie~ covers m u c h of the p i o n e e r i n g work in this field.

T h e simplest, yet probably the most widely practiced, use o f stable isotopes in p h a r m a c o l ogy is as internal standards for quantitation. A stable isotope-labeled version o f any molecule is the ideal internal standard, because its chemical a n d physical behavior is closer to the analyte species than any o t h e r possible choice o f standard. T h e molecular weight of the isotopic analog will be higher than the unlabeled analyte, a n d the masses of each can be d e t e r m i n e d with a mass spectrometer. Analytical precision with such techniques can be better than + 1%. More than 100 isotopically-labeled drug standards are commercially available for this purpose. Most o f these d r u g standards are labeled with deuterium, which does n o t create any p r o b l e m s regarding isotope effects because these are only analytical, not metabolic, standards. Stable-iso-

137

Stable Isotopes in Drug Metabolism

INITIAL OCTAPEPTIDE A-B-C-D-E-F-Y-z

Figure 3. Possible metabolites of a singly-labeled peptide. The products resulting from each possible combination of peptide proteolyses are displayed. The labeled tyrosine residue is indicated by its larger and bolder font.

B-C-D-E-F-Y-z

A-B-C-D-E-F-Y

B-C-D-E-F-Y

C-D-E-F-Y-z

A-B-C-D-E-F

D-E-F-Y-z

C-D-E-F-Y

B-C-D-E-F

A-B-C-D-I

E-F-Y-z

D-E-F-Y

C-D-E-F

B-C-D E

A-B-C-D

F-Y-Z

E-F-Y

I~E-F

C-D-E

B-C-D

A-B-C

Y-Z

F-Y

E-F

D-E

C-D

B-C

Z

Y

F

E

D

C

tope internal standards are essential in the highest quality quantitative drug analyses. Some of the newest developments in the use of stable isotopes involve biopolymers, where fully isotope-labeled species can be generated from cells grown in isotopically labeled growth media. The reason for going through this procedure is that biopolymers may cleave anywhere along their backbone, so that strategies that do not involve uniform labeling cannot reliably detect all possible metabolites. Figure 3 gives an example showing how specific labeling of a biopolymer is limited when monitoring its degradation products. Here, we imagine all the possible proteolytic metabolites of an octapeptide that is singly labeled in its one tyrosine. Of the 35 products, only 13 would be detectable through the tyrosine label. Clearly, if we had started with A-B-C-D-E-F-Y-Z, rather than A-B-C-D-E-F-Y-Z, all 35 products would have been labeled and could then have been recorded. Osborn and Abramson 15 have generated uniformly asC-labeled rat growth hormone by incubating a rat pituitary adenoma cell line in fully asC-labeled growth medium. This substance has been used to study both metabolism 16 and kineticsJ 7 With shown safety, and mass-spectrometric methods that generate sensitivity, specificity, and versatility, stable isotopes should play an important role in drug metabolism studies in pregnancy. Stable isotope-labeled drugs are ideal for this purpose. Obtaining pharmaceutical substances containing the appropriate label at the appropriate positions has been a major difficulty because few independent medicinal chemists can synthesize drugs in their labeled form. The pharmaceutical manufacturers h a v e t o help in this effort. With ever more new drugs generated from biological, rather than chemical, methods,

A-H B

obtaining labeled species will be easier. Adding suitable labeled nutrients to the medium in which cells are grown will readily produce protein and peptide drugs with stable isotopic substitutions.

References 1. McLafferty FW: Interpretation of Mass Spectra (ed 3). Mill Valley, CA, University Science Books, 1980 2. Balllie TA: The use of stable isotopes in pharmacological research. Pharmacol Rev 33:81-132, 1981 3. Pons G, Rey E, BadoualJ, Olive G: Use of the labelling of drugs with stable isotopes in clinical pharmacology in children. Arch Francaises Pediatrie 46:223-228, 1989 4. Browne TR: Stable isotopes in pharmacology studies: present and future. J Clin Pharmacol 26:485-489, 1986 5. Browne TR, Van Langenhove A, Costello CE, et al: Applications of stable isotope methods to studying the clinical pharmacology of antiepileptic drugs in newborns, infants, children, and adolescents. Therap Drug Monitoring 6:3-9, 1984 6. Koletzko B, Sauerwald T, Demmehnair H: Safety of stable isotope use. Eur. J. Pediatr. 156:S12-S17, 1997 7. Chapman TE, Berger R, Reijngoud DJ, et al: Stable Isotopes in Paediatric Nutritional and Metabolic Research. Andover, England, Intercept Limited, 1990 8. Knapp DR, Gaffney TE, McMahon RE, et ah Studies of human urinary and biliary metabolites of nortriptyline with stable isotope labeling. J. Pharmacol Exp Ther 180: 784-790, 1972 9. Sano M, Yotsui Y, Abe H, et al: A new technique for the detection of metabolites by the isotope aaC using mass fragmentography. Biomed Mass Spectrom 3:1-3, 1976 10. Abramson FP: CRIMS: Chemical reaction interface mass spectrometry. Mass Spectrom Rev 13:341-356 1994 11. Abramson FP, Teffera Y, KnsmierzJ, et al: Replacing 14C with stable isotopes in drug metabolism studies. Drug Metab Dispos 24:697-701 1996 12. Goldthwaite CA, Jr, Hsieh F-Y, Womble SW, et al: Liquid chromatography/chemical reaction interface mass spectrometry as an alternative to radioisotopes for quantitative drug metabolism studies. Anal Chem 68:2996-3001 1996 13. Chen P, Teffera Y, Black GE, et al: Flow injection with chemical reaction interface-isotope ratio mass spectrum-

138

Fred P. Abramson

etry: An alternative to off-line combustion for detecting low levels of enriched lSc in mass balance studies. J Am Soc Mass Spectrom 10:153-158 1999 14. StrongJM, DutcherJs, Lee W-K, et al: Absolute bioavailahility in man of N-acetylprocalnamide determined by a novel stable isotope method. Clin Pharmacol Ther 18:613-622, 1975 15. Osborn BL, Abramson FP: Production of uniformly stable isotope labeled proteins from mammalian cells. J Labelled Cpds Radiopharm 39:935-953, 1997

16. Osborn BL, Abramson FP: Metabolism and kinetics of fully-labeled proteins using chemical reaction interface mass spectrometry, in Heys JR, Melillo DG (eds): Synthesis and Applications of Isotopically Labelled Compounds, 1997, J o h n Wiley, Chichester, 1998 pp 119-122 17. Osborn BL, Abramson FP: Pharmacokinetic and metabolism studies using uniformly stable isotope labeled proteins with HPLC/CRIMS detection. Biopharm Drug Dispos 19:439-444, 1998