Measurement of Homocysteine Concentrations ... - ACS Publications

Michael J. MacCoss, Naomi K. Fukagawa, and Dwight E. Matthews* ...... Paolo Tessari , Edward Kiwanuka , Anna Coracina , Michela Zaramella , Monica Vet...
0 downloads 0 Views 92KB Size
Anal. Chem. 1999, 71, 4527-4533

Measurement of Homocysteine Concentrations and Stable Isotope Tracer Enrichments in Human Plasma Michael J. MacCoss, Naomi K. Fukagawa, and Dwight E. Matthews*

Departments of Chemistry and Medicine, University of Vermont, Burlington, Vermont 05405

Elevated levels of plasma homocysteine have been established as an independent risk factor for cardiovascular disease. Homocysteine is in low concentration in plasma (5-15 µM) and is bound to other thiols (e.g., cysteine in plasma proteins) via disulfide bonds. Existing methods for measuring homocysteine have difficulty in reducing and maintaining the reduction of homocysteine for measurement. We describe a GC/MS method that first reduces the disulfides in the physiological sample matrix and then immediately alkylates the free thiols with 4-vinylpyridine to prevent the reformation of the disulfide bonds. We use a deuterated internal standard, [3,3,3′,3′,4,4,4′,4′-2H8]homocystine to account for losses associated with the isolation, derivatization, and measurement of the natural homocysteine. The amino acids are separated and derivatized to form the tertbutyldimethylsilyl derivatives. This method requires only 50 µL of plasma to measure homocysteine concentrations to 5 µM. Total homocysteine concentrations in plasma can be measured routinely from 0.5-mL samples with relative intra- and interday precisions of 1.3 and 4.0%, respectively. This method is sensitive enough to determine tracer enrichments of [1-13C]homocysteine with a detection limit of 70% of homocysteine is bound to other thiols (e.g., plasma cysteine and cysteine in plasma proteins) via the formation of covalent disulfide bonds.1 The measurement of total homocysteine requires the reduction of the disulfide bond between homocysteine and other thiols or plasma proteins before analysis. Presently, homocysteine is measured using high-pressure liquid chromatography (HPLC),5-8 immunoassays,9 and gas chromatography/mass spectrometry (GC/MS).10-12 All of the methods are limited by their ability to reduce and maintain the reduction of homocysteine for measurement. Stabler et al. reported difficulty in maintaining disulfide bridge reduction when 2-mercaptoethanol was used as the reducing agent.10 They observed that homocysteine formed mixed disulfide bonds with 2-mercaptoethanol and other free thiols present in plasma after disulfide bridge reduction. Stabler used [3,3,3′,3′,4,4,4′,4′-2H8]homocystine as an internal standard. Homocystine is the disulfide form of homocysteine and, when reduced, yields 2 molar equiv of homocysteine. The deuterated homocystine was used to mimic the losses associated with the reduction, sample preparation, and measurement of the homocysteine in plasma. The use of a deuterated internal standard improves precision and enhances the repeatability of the analysis but requires the use of mass spectrometry.1,10,11 (2) Stampfer, M. J.; Malinow, M.; Willett, W. C.; Newcomer, L.; Upson, B.; Ullman, D.; Tishler, P.; Hennekens, C. JAMA, J. Am. Med. Assoc. 1992, 268, 877-881. (3) Nygard, O.; Nordrehaug, J. E.; Refsum, H.; Ueland, P. M.; Farstad, M.; Vollset, S. E. N. Engl. J. Med. 1997, 337, 230-236. (4) McCully, K. S.; Ragsdale, B. D. Am. J. Pathol. 1970, 61, 1-11. (5) Araki, A.; Sako, Y. J. Chromatrogr. 1987, 422, 43-52. (6) Refsum, H.; Ueland, P. M.; Svardal, A. M. Clin. Chem. 1989, 35, 19211927. (7) Fiskerstrand, T.; Refsum, H.; Kvalheim, G.; Ueland, P. M. Clin. Chem. 1993, 39, 263-271. (8) Pastore, A.; Massoud, R.; Motti, C.; Russo, A. L.; Fucci, G.; Cortese, C.; Federici, G. Clin. Chem. 1998, 44, 825-832. (9) Shipchandler, M. T.; Moore, E. G. Clin. Chem. 1995, 41, 991-994. (10) Stabler, S. P.; Marcell, P. D.; Podell, E. R.; Allen, R. H. Anal. Biochem. 1987, 162, 185-196. (11) Pietzsch, J.; Julius, U.; Hanefeld, M. Clin. Chem. 1997, 43, 2001-2004. (12) Sass, J. O.; Endres, W. J. Chromatrogr., A 1997, 776, 342-347.

Analytical Chemistry, Vol. 71, No. 20, October 15, 1999 4527

Another reason for using mass spectrometry to measure homocysteine is for the measurement of stable isotopically labeled homocysteine tracers. Stable isotopes are useful tracers of metabolism in vivo and are commonly used to measure production and disposal rates of a variety of amino acids.13 We have used these tracers and GC/MS to measure tracer enrichments simultaneously for metabolite concentration and kinetics.14,15 Stable isotope tracers have been used for the measurement of methionine and cysteine kinetics16-19 but have not been used for homocysteine. Homocysteine measurement by existing GC/MS methods has adequate sensitivity to measure homocysteine but not necessarily an isotopically labeled tracer that is typically 10-100-fold less abundant than unlabeled homocysteine.16,17 We have developed a novel method of reducing and isolating homocysteine in blood for subsequent quantitation using GC/ MS. The method covalently modifies homocysteine after reduction such that it cannot re-form disulfide bonds. This modification improves recovery, derivatization, and measurement by GC/MS. The improvement increases sensitivity, allows us to measure the concentration of homocysteine relative to a deuterated internal standard, and has the necessary sensitivity and precision to measure stable isotope enrichments in homocysteine for use in kinetic studies of sulfur amino acid metabolism in humans. EXPERIMENTAL SECTION Materials. 4-Vinylpyridine, L-methionine, D,L-homocystine, and dithiothreitol (DTT) were obtained from Sigma Chemical Co. (St. Louis, MO); N-methyl-N-(tert-butyldimethylsilyl)trifluoroacetamide (MTBSTFA) was obtained from Pierce Chemical Co. (Rockford, IL). N,N′-Dimethyl-N,N′-bis(mercaptoacetyl)hydrazine (DMH) was prepared via the method of Singh and Whitesides.20 [3,3,3′,3′,4,4,4′,4′-2H8]Homocystine ([3,3,4,4-2H4]homocysteine when reduced, 96% 2H4) was purchased from Cambridge Isotope Laboratories (Andover, MA), and L-[1-13C]methionine (99% 13C) was obtained from Mass Trace (Woburn, MA). Infusion Protocol. Three healthy, normal weight for height adults were studied at the University of Vermont Clinical Research Center (CRC). The subjects were instructed of the purpose, benefits, and risks of the study and gave written consent in accordance with protocols approved by the University of Vermont Institutional Review Board and by the CRC Scientific Advisory Committee. The subjects were admitted to the CRC and given an evening meal before 7 p.m. Thereafter, only water was consumed until completion of the infusion study the following day. At 6 a.m. on the infusion day, a subject was awakened and an intravenous catheter was placed in a forearm vein for infusion of the tracer and a hand vein for sampling “arterialized” venous blood using (13) Matthews, D. E. In Modern Nutrition in Health and Disease; Shils, M. E., Olson, J. A., Shike, M., Ross, A. C., Eds.; Williams and Wilkins: Baltimore, 1998; Chapter 2, pp 11-48. (14) Darmaun, D.; Manary, M. J.; Matthews, D. E. Anal. Biochem. 1985, 147, 92-102. (15) Gilker, C. D.; Pesola, G. R.; Matthews, D. E. Anal. Biochem. 1992, 205, 172-178. (16) Storch, K. J.; Wagner, D. A.; Burke, J. F.; Young, V. R. Am. J. Physiol. 1988, 255, E322-E331. (17) Storch, K. J.; Wagner, D. A.; Burke, J. F.; Young, V. R. Am. J. Physiol. 1990, 258, E790-E798. (18) Hiramatsu, T.; Fukagawa, N. K.; Marchini, J. S.; Cortiella, J.; Yu, Y. M.; Chapman, T. E.; Young, V. R. Am. J. Clin. Nutr. 1994, 60, 525-533. (19) Young, V. R.; Ajami, A. M. J. Parenter. Enteral Nutr. 1999, 23, 175-194. (20) Singh, R.; Whitesides, G. M. J. Org. Chem. 1991, 56, 2332-2337.

4528

Analytical Chemistry, Vol. 71, No. 20, October 15, 1999

the heated-hand box technique. Both catheters were kept patent with a slow infusion of sterile saline. At 7 a.m., a priming dose of [1-13C]methionine (1.6 µmol/kg) was administered intravenously, and an infusion of [1-13C]methionine (2.4 µmol/kg per hour) was begun and continued for 8 h using an infusion pump. Blood samples were taken just prior to the start and at intervals during the tracer infusion. Blood was placed in vials containing EDTA and immediately centrifuged, and the plasma fraction was frozen at -60 °C until analysis. Sample Preparation. The rubber septa of 5-mL evacuated blood collection tubes containing no additive (Vacutainer, Rutherford, NJ) were removed, and a 0.5-mL aliquot of plasma was added to each collection tube. A 20-µL aliquot of 0.8 mM (16 nmol) of [2H8]homocystine and 75 µL of either DMH or DTT solution (5 mg/mL in water) were added to each sample tube. The septa were replaced and the tubes were reevacuated for 2 min via a needle through the septum attached to a vacuum line with a rotary pump. The evacuated tubes were heated at 50 °C for 1 h. Using an HPLC syringe with a fixed needle (Hamilton, Reno, NV), 50µL of undiluted 4-vinylpyridine was added to each evacuated tube through the septum to alkylate the reduced thiols from the previous step. The 4-vinylpyridine was allowed to react with the thiols in the sample tubes at room temperature for 2 h. The samples were then acidified with 2 mL of 1 M acetic acid and centrifuged. The supernatant was removed from the precipitate and poured directly into disposable columns containing 0.5 mL of thoroughly washed cation-exchange resin (AG 50W-X8 100200 mesh, hydrogen form, Bio-Rad Laboratories, Richmond, CA). The cation resin was washed twice with 5 mL of distilled water, and the amino acids (including the reacted thiols) were eluted with 2 mL of 3 M NH4OH. The amino acid fraction was then dried under a constant stream of N2. Next 50 µL of MTBSTFA/ acetonitrile (1:1 v/v) was added to form the tert-butyldimethylsilyl (tBDMS) derivatives. Included in this fraction are S-4-ethylpyridine bis-tBDMS homocysteine (Figure 1) and bis-tBDMS methionine derivatives. Gas Chromatography/Mass Spectrometry. A 1-µL aliquot of each derivatized sample was injected into a Hewlett-Packard 5971A GC/MS (Palo Alto, CA) operated in splitless mode (injection temperature 265 °C). Freshly silanized glass injectors and Silcosteel-treated inlet seals (Restek, Bellefonte, PA) were used to prevent absorption and peak-tailing of the thiol-containing amino acids. Chromatographic resolution of S-4-ethylpyridine bistBDMS homocysteine and bis-tBDMS methionine was accomplished using a 30-m × 0.25-mm-i.d., 0.25-µm film, Rtx-1 capillary column (Restek, Bellefonte, PA). The GC oven temperature was held constant at 220 °C for 1 min after the sample injection, before increasing at 40 °C/min to a final temperature of 280 °C for 5 min. The samples were measured by electron impact ionization at 70 eV. The [M - 57]+ fragment resulting from the loss of a tertbutyl group was measured by selected-ion monitoring (SIM) for the unlabeled homocysteine (m/z ) 411), 13C-enriched homocysteine (m/z ) 412), and the reduced internal standard [2H4]homocysteine (m/z ) 415). The integrated areas (Am/z) of the monitored ions (m/z ) 411, 412, and 415) were used to produce the ion intensity ratios of R1/0 ) A412/A411 and R0/4 ) A411/A415. Likewise the [M - 57]+ fragment for the unlabeled methionine

Figure 1. Three-step reduction and derivatization procedure for total homocysteine. (1) DMH is used to reduce all disulfide-bound homocysteine to the free thiols. (2) The resulting free thiols are alkylated with 4-vinylpyridine. (3) The amino acids are derivatized with MTBSTFA to form S-4-ethylpyridine bis-tBDMS homocysteine. The method also reduces and derivatizes all disulfide-bound cysteine in the same fashion as homocysteine (not shown).

(m/z ) 320) and 13C-enriched methionine (m/z ) 321) were also measured, and the integrated areas (Am/z) of the monitored ions were used to produce the ion intensity ratio of R1/0 ) A321/A320. [1-13C]Homocysteine Standard Preparation. Because [1-13C]homocysteine is not commercially available, enrichment standards of known [1-13C]homocysteine content were prepared by the decomposition of [1-13C]methionine standards to [1-13C]homocysteine using a modified method described previously.21 First, [1-13C]methionine standards of known enrichments of the range of 2-8 mol % excess (mpe) were prepared. Next, 500-µL aliquots of each enrichment standard (10 mM) were added to 2-mL conical bottom, screw cap glass reaction vials, and the solvent was evaporated to dryness under a constant stream of N2 gas. Each reaction vial received 50 µL of 12 M HCl, was purged with argon, capped, and heated at 105 °C for 15 h. The HCl was dried under N2, and the samples were reconstituted in 500 µL of distilled water and transferred to evacuated blood collection tubes. These [1-13C]homocysteine standards were prepared for measurement by GC/ MS as described above for the plasma samples. Yield of [1-13C]homocysteine was 25-34%. Data Analysis. The ion intensity ratios determined by selectedion-monitoring GC/MS were converted into molar ratios using the previously described approach.15 The ion intensity ratio (Ra/b) for two isotopic species, a and b, measured by GC/MS is a linear function of the molar ratio of a/b, in the form

Ra/b ) Ra/b(b) + ka/b(na/nb)

na/nb ) [Ra/b - Ra/b(b)]/ka/b

(2)

The molar ratio of na/nb can be used either to quantitate a as a function of added b (in the case of b being an internal standard) or to calculate the enrichment of a in b (e.g., determination of tracer/tracee ratio and tracer enrichment). Homocysteine concentration was determined from the ion intensity ratio of unlabeled homocysteine to [2H4]homocysteine (R0/4):

R0/4 ) R0/4(4) + k0/4(n0/n4)

(3)

where n0 is the amount of homocysteine (nmol), n4 is the amount of [2H4]homocysteine added to the sample (2 times the amount of [2H8]homocystine), R0/4 is the measured A411/A415 ion intensity ratio, and R0/4(4) is the A411/A415 ion intensity ratio when pure [2H4]homocysteine is injected. The amount of homocysteine in individual samples was determined by rearrangement of eq 3 to solve for n0:

n0 ) n4[R0/4 - R0/4(4)]/k0/4

(4)

(1)

where na and nb are the amounts of a and b in moles, ka/b is the molar response factor of the GC/MS system (ideally equal to unity), and Ra/b(b) is the na/nb ratio measured by the GC/MS system when pure b is injected. Ra/b(b) and ka/b are determined by the measurement of Ra/b for standard samples with a known na/ (21) Butz, L. W.; DuVigneaud, V. J. Biol. Chem. 1932, 99, 135-142.

nb molar ratio. Equation 1 can be rearranged to measure an unknown molar ratio as a function of an observed ratio:

The enrichment of [1-13C]homocysteine or [1-13C]methionine in a sample was determined from the A412/A411 ion intensity ratio for homocysteine or the A321/A320 ion intensity ratio for methionine:

R1/0 ) R1/0(0) + k1/0(n1/n0)

(5)

where n1/n0 is the 13C-tracer/unlabeled tracee ratio for either homocysteine or methionine, R1/0 is the measured A412/A411 or Analytical Chemistry, Vol. 71, No. 20, October 15, 1999

4529

Figure 3. GC/MS ion chromatograms for m/z ) 411 and 415 of a human plasma sample to which [2H8]homocystine has been added. The sample was prepared and measured as described in the Experimental Section. The [2H8]homocystine becomes [2H4]homocysteine during the sample reduction and is measured as [2H4]homocysteine. Figure 2. Mass spectra of S-4-ethylpyridine bis-tBDMS homocysteine and [2H4]homocysteine acquired by EI-GC/MS at 70 eV.

A321/A320 ion intensity ratio for [13C]homocysteine or [13C]methionine, respectively, and R1/0(0) is the ion intensity ratio when pure unlabeled material is injected. The mole ratio of 13C to unlabeled homocysteine (the tracer/tracee ratio, n1/n0) was calculated for individual samples by rearranging eq 5:

n1/n0 ) [R1/0 - R1/0(0)]/k1/0

(6)

Enrichments (E), expressed as mpe, were computed from the mole ratio determined in eq 6 using the equation

E ) 100(n1/n0)/[1 + n1/n0]

(7)

RESULTS Mass Spectrometry. The reaction scheme for the reduction of homocysteine in physiological samples, reaction with 4-vinylpyridine to form a unique modified species, and the derivatization to form the tBDMS derivative is shown in Figure 1. The final derivatized species produces multiple fragment ions by EI mass spectrometry. The EI mass spectra of the S-4-ethylpyridine bistBDMS homocysteine and [2H4]homocysteine are shown in Figure 2. The base peak in each mass spectrum was m/z ) 106, corresponding to a fragment consisting of the 4-ethylpyridine group, but not the homocysteine moiety. As characteristic of tBDMS-derivatized amino acids, the molecular ion was hardly visible for homocysteine or [2H4]homocysteine. There was a significant fragment ion, [M - 57]+, at m/z ) 411, resulting from the loss of a tert-butyl group from the molecule. Also present were fragment ions resulting from the loss of CO2tBDMS at m/z ) 309 and 313 for the natural and [2H4]homocysteine, respectively. The ions at m/z ) 204 and 278 have significant intensity, and contain all four deuterium atoms, but do not contain the amino acid carboxylic acid group. The [M - 57]+ ion is derived from a simple one-bond break and is at a higher mass than the mass fragments at m/z ) 204 or 278. For these reasons we have used 4530 Analytical Chemistry, Vol. 71, No. 20, October 15, 1999

the [M - 57]+ fragment for both quantitative and tracer/tracee measurements. Selected ion-chromatograms of a plasma sample measuring homocysteine and added [2H4]homocysteine internal standard are shown in Figure 3. The chromatographic conditions allow homocysteine to elute in