Novel Reductant for Determination of Total Plasma Homocysteine

Homocysteine (Hcy), increasingly being recognized as a risk factor for vascular disease (1), is found primarily in plasma in the form of homocystine and mixed disulfides, both protein-bound and unbound; total Hcy (tHcy) is the sum of all Hcy species obtained after quantitative reduction. Several methods are currently used to measure tHcy in plasma, including GC/MS, ion-exchange chromatography, and HPLC (2). In one of the most popular HPLC techniques, any homocystine and homocysteine-mixed disulfides present are reduced with tri-butyl phosphine (TBP) to Hcy, which is subsequently derivatized with ammonium 7-fluorobenzo-2-oxa-1,3-diazole-4-sulfonate (SBD-F) to produce a fluorescent product (3)(4). TBP has a highly disagreeable odor and is poorly soluble in water so that it must be dissolved in dimethylformamide for use, both of which properties hinder the routine use of TBP in a clinical laboratory. Here, we call attention to a newer phosphine reagent, tris(2-carboxylethyl) phosphine (TCEP), which is nonvolatile, stable, and soluble in aqueous solution and thus is more suitable for routine use. We have therefore modified the method of Ubbick and Vermaak (4) as follows. First, add 9 μL of 100 g/L TCEP (aq) (Pierce Chemical Co., Rockford, IL) to 90 μL of plasma, mix, and allow to react for 30 min at room temperature. Then, add 90 μL of 100 g/L trichloroacetic acid containing 1 mmol/L EDTA. After centrifuging the sample for 10 min at 13 000g, add 100 μL of the supernatant to a tube containing 20 μL of 1.55 mol/L NaOH, 250 μL of 0.125 mol/L borate buffer containing 4 mmol/L EDTA, pH 9.5, and 100 μL of SBD-F (Wako Chemicals USA, Richmond, VA), 1 g/L in the borate buffer. After mixing, incubate the solution for 1 h at 60 °C, let cool to room temperature, and inject 20 μL of this onto a C18 column to run at a flow rate of 1.5 mL/min (4). Figure 1 shows the relation between the Hcy determined after reduction with TCEP as a function of that determined after reduction with TBP in aliquots of the same plasma samples (5). All plasma samples were obtained in accordance with the guidelines of the research ethics board at our institution. For the Hcy determinations with TCEP, Hcy calibrators were prepared in isotonic saline (NaCl 9 g/L), the assigned concentration being based on a molar absorptivity for Hcy of 397 L mol−1 cm−1 at 244 nm. These saline calibrators were chromatographed and used to prepare a calibration curve from which the Hcy concentrations in the plasma samples were calculated according to peak heights. When Hcy was added to plasma samples, recovery was 100% (results not shown). Hcy determinations in plasma with TBP were performed at the Institut de Recherches Cliniques de Montreal (by L.-J. Fortin, whom we thank for these determinations). For the Hcy determinations done with TBP, Hcy calibrators were prepared in plasma, and the Hcy concentrations were based on peak areas (5). The correlation coefficient between the results obtained with the two different reductants is 0.988, although the TCEP method yields values ∼21% higher. This may reflect simple differences in methodology—e.g., the calculation of results (peak height vs peak area), the calibration matrix (saline vs plasma), or quantification of the calibrators (molar absorptivity vs dry weight, respectively)—or differences in the efficiency of reduction with TCEP, and suggests the necessity of developing an interlaboratory standardization procedure. The 5th, 50th, and 95th percentiles for plasma tHcy determined with TCEP reduction for a population of 30 individuals (15 men, 15 women; mean ± SD ages, 36.9 ± 10.6 years) were 5.2, 9.8, and 14.9 μmol/L, respectively. Usable values can be produced from as little as 15 μL of plasma. The TCEP solution can be stored at −20 °C and is stable to freezing and thawing. author for correspondence: fax 514-843-1499, e-mail mcbg@musica.mcgill.ca. Total plasma homocysteine measured after reduction with TCEP as calculated by using peak height vs the concentration measured after reduction with TBP as calculated by using peak area in aliquots of the same plasma sample. The equation of the line is y = −0.088 + 1.21x, r = 0.988. Sy|x, the standard deviation of residuals for this line, is 0.50. We thank Min-Wen Hsiung, Angela Hosack, and Sally Lue-Shing for their expert technical assistance. This work was supported by a grant from the Program Conjoint, FRSQ/Hydro-Québec.