Measurement of plasma homocysteine may be of value in several clinical conditions including homocystinuria, atherosclerosis, thrombophilia, and folate/vitamin B12 deficiency. The increasing interest in measuring total homocysteine in plasma has led to the development of several different methods (1). A widely used technique for measuring total plasma homocysteine is reversed-phase HPLC with fluorescence detection after derivatization of plasma thiols with ammonium 7-fluorobenzo-2-oxa-1,3-diazole-4-sulfonate (SBD-F) (2)(3). Most published methods use external calibration alone for quantitation of homocysteine because of the difficulty in selecting an internal standard. We have modified this method by adding cysteamine hydrochloride as an internal standard to the plasma or homocysteine calibrator to compensate for variations in thiol derivatization and sample injection procedures. HPLC was carried out by an isocratic system with fluorescence detection (SFM 25 spectrofluorometer), autosampler (SA 360), and HPLC pump (325) supplied by Kontron Instruments. Chemicals were obtained from Sigma. The method has been adapted from that of Ubbink et al. (2) on the basis of the chemical description provided by Araki and Sako (3). The plasma or homocysteine calibrator (150 μL) was incubated with 100 mL/L tri-n-butylphosphine in dimethylformamide (15 μL) for 30 min at 4 °C to reduce and release protein-bound thiols. Deproteinization was achieved by the addition of 100 g/L trichloroacetic acid (150 μL) and centrifugation. An aliquot of the supernatant (50 μL) was mixed with sodium hydroxide (10 μL, 1.55 mol/L), borate buffer (125 μL, 0.125 mol/L, pH 9.5, containing 4 mmol/L EDTA), and SBD-F (50 μL, 1 g/L) and incubated for 60 min at 60 °C. The SBD-F derivative from the supernatant (20-μL aliquot) was eluted isocratically from the Spherisorb ODS2 [4.6 mm (i.d.), 5-μm particles] analytical column (Jones Chromatography). The mobile phase was 0.1 mol/L KH2PO4, pH 2.0, containing 40 mL/L acetonitrile at a flow rate of 0.8 mL/min. Homocysteine calibrators (dl form) were prepared in the following matrices: deionized water, pooled plasma (anticoagulated with EDTA), phosphate buffer (0.1 mol/L, adjusted to pH 9.5), and borate buffers (0.1 mol/L, pH 9.5, with and without 2 mmol/L EDTA). Cysteamine hydrochloride (2-mercaptoethylamine) was used as an internal standard, which was added to the plasma or homocysteine calibrator to achieve a final concentration of 10.0 μmol/L (30 μL of 50.0 μmol/L cysteamine plus 120-μL sample). Plasma samples from healthy adults (volunteer blood donors) were obtained from whole blood collected into evacuated EDTA tubes (Vacutainer Tubes, Becton Dickinson) and cooled on ice, and the plasma was separated by centrifugation within 60 min of venipuncture. Plasma was stored for a maximum of 3 months at −70 °C before assay. This sampling was approved by the Hospital Ethical Committee. Chromatography of a plasma sample with added cysteamine produced five peaks (see Fig. 1 ) that were identified as cysteine, cysteamine, cysteinyl glycine, homocysteine, and glutathione by comparison of retention times with those of pure compounds run separately and confirmed by coelution when plasma was supplemented with these compounds. In the absence of cysteamine no peak appears at that position for plasma samples. The mean recovery of homocysteine dissolved in water added to plasma to final concentrations of 2.5–40.0 μmol/L was 101.2% (n = 8; range = 98.3–106.3%). In the presence of cysteamine, the mean recovery of added homocysteine to plasma to final concentrations of 2.0–16.0 μmol/L was 100.3% (n = 10; range = 88.8–105.5%). The response of the fluorometric detector was linear for homocysteine calibrators from 0 to 50.0 μmol/L for all matrices used. When homocysteine was diluted with either water or pooled plasma, no difference in the calibration slopes was obtained. However, homocysteine calibrators in other matrices resulted in different calibration slopes (Fig. 2 , top) in the following order, starting with the largest peaks: water = pooled plasma > phosphate buffer (0.1 mol/L, pH 9.5) > borate buffer (0.1 mol/L, pH 9.5) > borate buffer (0.1 mol/L, pH 9.5, containing 2 mmol/L EDTA). The calibration slopes for homocysteine diluted in different matrices were indistinguishable from each other when cysteamine was included as an internal standard and the calibration slopes were calculated with homocysteine/cysteamine peak area ratios (Fig. 2 , bottom). The within-batch imprecision (CV) with and without internal standards for low control plasma (7.8 μmol/L) was 2.8% (n = 10) and 8.3% (n = 10), respectively, and for high control plasma (23.6 μmol/L) was 2.2% (n = 10) and 5.3% (n = 10), respectively. The between-batch CVs of the assay with and without internal standards for low control plasma were 6.9% (n = 75) and 11.0% (n = 142), respectively, and for high control plasma were 7.9% (n = 75) and 11.3% (n = 125), respectively. The distribution of homocysteine concentrations of plasma from healthy adult blood donors was nongaussian with a moderate skew towards higher values. In men (n = 295, mean age 38.4 years, range 20–69 years), the median plasma homocysteine was 8.5 μmol/L with reference range 5.3–16.9 μmol/L (2.5 and 97.5 percentiles). In women (n = 348, mean age 34.3 years, range 19–64 years), the median plasma homocysteine was 7.2 μmol/L with reference range 4.5–13.9 μmol/L. This SBD-F method for homocysteine estimation gives chromatograms with well-defined peaks and stable baselines. The addition of cysteamine has considerably improved the within- and between-assay precision. Homocysteine and cysteamine derivatives behave in a predictable way in the assay both in water and plasma, in that the response of the fluorometric detector is linear for both compounds up to at least 50 μmol/L. External homocysteine calibration is markedly matrix-dependent. The use of different buffers and the presence of EDTA has an overall effect on the thiol derivatization and its detection. Because total plasma homocysteine values were obtained from interpolation on the calibration curve, spuriously low calibrators result in falsely high homocysteine values. Few details have been given in published methods for homocysteine assays as to how the homocysteine calibrators were constituted (1)(2)(3)(4)(5)(6). In the original HPLC SBD-F-based method of Araki and Sako (3), homocysteine calibrators were dissolved in borate buffer (0.1 mol/L, pH 9.5, containing 2 mmol/L Na2EDTA). In the method of Ubbink et al. (2), based on the same principle, water is used instead of borate buffer [J. B. Ubbink, personal communication]. Vester and Rasmussen (4) reported that the calibration slopes of homocysteine calibrators diluted in borate buffer (0.1 mol/L, pH 9.5, containing 2 mmol/L EDTA) were 15% less than in the pooled plasma calibration, which is similar to the effect reported here. They suggested that the differences could result from the presence of proteins, ionic strength, or other species affecting the derivatization step, and recommended plasma-based calibrators. Because the data reported here show no differences in the calibration slopes between water and pooled plasma, it is simpler and equally effective to dilute homocysteine calibrators in water. The use of cysteamine as an internal standard was originally suggested by te Poele-Pothoff et al. (5) but has not been fully characterized for this purpose. HPLC assays for homocysteine incorporating mercaptopropionylglycine and N-acetylcysteine as internal standards have also been described (4)(6). Unlike cysteamine, these compounds elute later than the homocysteine peak, which prolongs the run time. Cysteamine (NH2–CH2–CH2–SH) has sufficient chemical similarities to homocysteine to make it suitable as an internal standard. The concentration of naturally occurring cysteamine in human plasma as reported by previous workers is extremely low (<0.1 μmol/L) (7). This is confirmed by the failure to detect a peak at the cysteamine position in the experiments reported here when the internal standard is omitted from the reaction. The negligible concentration of naturally occurring cysteamine is therefore unlikely to interfere with cysteamine as an internal standard at a concentration of 10.0 μmol/L. Results of the recovery study have shown that cysteamine does not affect the behavior of homocysteine present in either plasma or aqueous matrices. The addition of cysteamine also compensates for the matrix dependency of external homocysteine calibration. This indicates that whatever the reason for variation in peak areas of homocysteine because of different matrices, this effect also operates for cysteamine and thus is compensated for when the results are calculated as ratios of peak area. Use of an internal standard can also compensate for variation in other aspects of the assay such as minor differences in sample injector volume or detector sensitivity. Several advantages are associated with inclusion of an internal standard, which is reflected in the improved precision achieved. Chromatogram of human plasma sample with cysteamine as an internal standard (homocysteine 23.6 μmol/L). Calibration slopes for homocysteine calibrators constituted in different matrices calculated by reference to the external calibrator only (top) or by using an internal standard (bottom). ♦, water; □, pooled plasma; ▴, 0.1 mol/L phosphate (pH 9.5); •, 0.1 mol/L borate (pH 9.5); X, 0.1 mol/L borate (pH 9.5, plus 2 mmol/L EDTA). This work is funded in part by a project grant from the British Heart Foundation. Part of this work has been published in abstract form (J Inherit Metab Dis 1996;19(Suppl 1):28).
No takes yet. Share an insight, caveat, or question.
Kuo et al. (1997) studied this question.
Synapse has enriched one closely related paper. Consider it for comparative context: