ANALYTICAL

BIOCHEMISTRY

Determination Homocysteine

!I&

684-692 (1978)

of Cysteine in Plasma and Urine and in Plasma by High-Pressure Liquid Chromatography

ROLF SAETRE AND DALLAS L. RABENSTEIN Department

of Chemistry.

University

of Alhertu.

Edmonton,

Alhrrtu.

Cunadu

Received December 6. 1977 New methods are described for the determination of reduced and total cysteine in urine and the nonprotein fraction of plasma, and total homocysteine in the nonprotein fraction of plasma. The method for reduced cysteine involves separation by high performance liquid chromatography (hplc). followed by detection with a mercury-based electrochemical detector. The detector has a detection limit of ca. 10m6M cysteine. and is selective for sulfhydryl components of the biological fluids analyzed. Sample preparation involves centrifugation and filtration, and the hplc analysis time is approximately 8 min. Total cysteine and homocysteine are determined by electrolytic reduction of their disulfides to the sulfhydryl form prior to the hplc analysis. Results are presented which suggest that cysteine disulfide exchange reactions are a source of error in methods which employ derivatization with iodoacetate for the determination of cysteine in plasma.

Cysteine and cystine are present in measurable quantities in the plasma and urine of normal persons. The level of cysteine, which is produced by reduction of cystine formed by hydrolysis of dietary protein and which appears as a metabolite in the degradative pathway of methionine to sulfate, is about 10% of the cystine level. Since homocysteine also appears as a metabolite on the pathway from methionine to cysteine, it would be expected to be present in normal plasma although in low concentration. So far, however, neither homocysteine nor homocystine has been detected in the free state. The cysteine-homocysteine-mixed disulfide, on the other hand, has been reported to be present in normal plasma (1). Accurate and fast assays for free and total cysteine and homocysteine play a key role in the screening of selected populations for congenital disorders of sulfur metabolism such as cystinuria, cystinosis, and homocystinuria, and in the monitoring of their treatment. Sensitive and selective methods of analysis also are essential in the general study of the metabolism of thiols and disulfides and of extra- and intracellular disulfide exchange reactions. Several specific spectrophotometric assays have been described. For cysteine. these include methods based on the colored products formed by reaction with quinones (2), aldehydes (3), ninhydrin (4), 0003-2697/78/0902-0684$02.00/O Copyright 0 I!?78 by Academic Press, Inc. All rights of reproduction in any form reserved.

684

CYSTEINE

IN PLASMA

AND

URINE

68.5

and noradrenochrome (5), and for homocysteine a method based on the absorbance at 230 nm of its thioester(6). These methods, however, lack the necessary sensitivity or they are too susceptible to interferences to be suitable for the low levels normally encountered in plasma and urine. The method of choice at the present time appears to be the method introduced by Brigham ~t~11. (7), in which the sample is first reacted with iodoacetate to form the S-carboxymethyl derivative of cysteine and then analyzed by ion exchange chromatography on an amino acid analyzer with ninhydrin detection. In this paper, we describe a new method for the determination of reduced and total cysteine in plasma and urine and total homocysteine in plasma, which is based on high-performance liquid chromatography (hplc)’ with electrochemical detection (8). The detector is operated at a potential where it responds only to chloride and to compounds with thiol groups, so that the only requirement of the chromatographic step is that it separates these few compounds from each other. This separation considerably reduces the analysis time as compared to amino acid analyzer procedures which use ninhydrin detection. To obtain the total cysteine and homocysteine concentrations, the disulfides are electrolytically reduced at a mercury pool electrode (9) prior to the hplc analysis. We also present results which suggest that the disulfide exchange reactions between cysteine, cystine, and plasma albumin and its cysteine-mixed disulfides are a source of error in the iodoacetate method (7) for the determination of cysteine in plasma. MATERIALS

AND METHODS

Apparatus. The hplc system was constructed from a Milton Roy minipump, tubing, connectors, valves, and columns from the Cheminert Division of Laboratory Data Control, and a mercury-based electrochemical detector (8). The hplc system has been described in detail (8,9). The detector was operated at +O. 1 V rs the saturated calomel electrode (SCE). Separations were performed on a 50 x 0.2-cm column dry-packed with Zipax SCX strong cation exchange resin, obtained from DuPont, Wilmington, Delaware. Aliquots of 10 ,ul were injected onto the column with a syringe. For the urine samples a 1% H,PO, solution was used as eluent. and a pH 2.5 phosphate/citrate buffer of ionic strength 0.03 was used for the plasma samples. Both eluents were deaerated with oxygen-free nitrogen. A flow rate of 0.5 mlimin was used. The electrolysis cell used for electroreduction of the disulfides has been described previously (9). Chemicals. Cysteine and homocysteine (Nutritional Biochemicals Corp.) and cystine (Aldrich) were used as received. The purity of the ’ Abbreviations electrode; CSH, protein-cysteine-mixed

used: hplc, high-pressure cysteine: CSSC. cystine; disulfides.

liquid chromatography; SCE, saturated calomel PrSH, protein sulfhydryl groups: and PrSSC.

686

SAETRE

AND

RABENSTEIN

cysteine was determined by titration with coulometrically-generated Br, in 1 M HCl using biamperometric end point detection. Five replicate titrations gave a purity of 95.0 ? 0.5%. The purity of the homocysteine was determined to be 99.0% by reaction with iodoacetamide followed by titration with NaOH (10). The purity of the cystine was determined by first electroreducing it and then comparing it by chromatography with a standard cysteine solution. A purity of 101 t 2% was obtained. The assay values were used in the preparation of standard solutions. Doubly distilled water was used throughout. Procedure for urine. Approximately 30 ml of urine was collected directly in a beaker containing 3.0 ml of 2 M HCl. The volume was determined by weighing the beaker with and without the urine. If a precipitate formed, about 5 ml of the urine was centrifuged for 5 min at 2500 rpm. A small aliquot of the clear urine was then diluted 1: 1 with 0.2 M HCl and 10 ~1 was injected onto the column. The technique of standard addition was used for quantitation. Total cysteine was determined by electrolyzing for 15 min an aliquot of the urine diluted 2:25 with 0.2 M HCl. All standard cysteine and cystine solutions were prepared in 0.2 M HCl. Recovery studies of cystine from urine were done by addition of 1.0 ml of a 5.00 x 10m4 M cystine solution to 4.0 ml of diluted urine. Roston has shown that small amounts of cysteine disappear rapidly in nonacidified urine in contact with air (5). It is essential, therefore, to acidify the urine immediately after delivery. In this work it was found that making the urine 0.2 M in HCl inhibited cysteine oxidation for at least several hours at room temperature. Procedure for plasma. Whole blood was collected in a Vacutainer (Fisher Scientific) containing a K,HEDTA solution. The tube was centrifuged immediately at 2500 rpm for 5 min. The plasma, 0.8 ml, was withdrawn with a l-ml tuberculin syringe and added to a 10 x 75-mm test tube containing 0.2 ml of 200 g/liter metaphosphoric acid. The contents were mixed well and, after about 10 min, the tube was centrifuged at 2500 rpm for 10 min. The supernatant was then filtered through a 0.45~pm TABLE CYSTEINE

AND

TOTAL

CYSTEINE

[CW Sample

I 2 3 4 u Mean

1

IN URINE

[C=-Uro,

(M

X

lo”)

(M

3.5 3.6 2.1 3.8

2 t + +

0.1 0.1 0.1 0.2

27.6 37.0 20.8 22.6

5 SD for replicate

analyses.

X

t k + r

OF FOUR

MALE

ICSHlm

1w)

lC=l

0.6 0.8 0.6 0.4

7.9 10.3 9.9 6.0

ADULTS”

Recovery cystine

of

(%) 91 101 99 100

k t L +

4 5 4 4

CYSTEINE

IN PLASMA

AND

687

URINE

Millipore filter. Standard cysteine solutions were prepared in 40 g/liter metaphosphoric acid, and a calibration curve procedure was used. To obtain the total cysteine, 0.25 ml of the filtered plasma was diluted with 0.25 ml of 0.4 M HCI. This solution was electrolyzed for 10 min (9). Recovery studies of cystine from plasma were done by addition of 0.05 ml of 5.00 x IO-”

M

cystine to 0.45 ml of a second plasma aliquot diluted I: I

with 0.4 M HCl. Standard cysteine and cystine solutions were prepared in 0.2 M HCI. RESULTS Cystrine

in Urinr

The reduced and total cysteine levels found in the urine of four normal adult males are listed in Table 1. Also listed in Table I are the results of recovery studies for cystine. A standard addition procedure was used for quantitation of cysteine in urine rather than a calibration curve because the retention time of the cysteine, and thus the peak shape, is affected by the urine matrix. The linearity of the detector response to cysteine concentration is dem+3c I-

I

I

I

I

+20

-F

.i P / p”

+10

0

P

I

I

,/,’,/p1

/I/

.//’ -10

I 5 -log

FIG. I Response of the electrochemical Log I is the logarithm of the peak current moles per liter.

I

I

4

3

[CYSI

detector as a function of cysteine concentration. in nanoamperes. and cysteine concentration is in

688

SAETRE

AND

RABENSTEIN

onstrated in Fig. I for cysteine concentrations which is about the limit of detection. Cysteine and Homocysteine

in the range lo-” to lO-‘j M,

in Plasma

The procedures we have developed for plasma samples permit the determination of the levels of reduced cysteine, total cysteine, which includes cystine and any other nonprotein disulfide forms of cysteine, and total homocysteine. which has been reported to be present in normal plasma in the mixed disulfide form with cysteine (I). A representative chromatogram of an electrolyzed plasma sample is shown in Fig. 2. With

I

5 nA

I

I

I

I

I

0

2

4

6

8

Minutes FIG. 2. Representative chromatogram of an electrolyzed plasma sample. The sharp unretained peak at about 1.5 min is due to changes in double-layer capacitance at the mercury surface as well as to the detector response to chloride present in the sample. The cysteine peak appears at 4 min and the homocysteine peak appears at about 6.5 min. Between 0.5 and 2 min recorder sensitivity was changed to keep the unretained peak on scale.

CYSTEINE

IN PLASMA TABLE

CYSTEINE,

Sample I 2 3 4 5

TOTAL

CYSTEINE.

AND TOTAL

AND 2

HOMOCYSTEINE

IN PLASMA

[CSHI

[CSWn,t

[CW,,,

(M X loj)

(M X 10’)

[CSHI

1.09 0.90 1.08 1.09 0.76

? t k t +

0.04 0.03 0.04 0.03 0.03

12.9 10.3 13.2 15.3 10.6

? 2 + 4 -+

0.3 0.2 0.3 0.4 0.2

689

URINE

11.8 11.3 12.2 14.0 14.0

OF FIVE ADULTS [Homocysteine],,,, (M X 10") 1,s 1.1 1.8 1.2 0.9

the pH 2.5 phosphate/citrate buffer as eluent, k’ for cysteine is I .4 and k’ for homocysteine is 3.2, where k’ is the capacity factor for the particular component. k’ is defined as (t, - tJt,,, where I, is the elution time for the component and f0 is the elution time for unretained components. The peaks in Fig. 2 were identified by comparison with standard solutions containing 40 g/liter metaphosphoric acid and 0.2 M HCl. Under these conditions, ergothioneine has a k’ of 2.4. The results obtained for the plasma from five normal adults are presented in Table 2. During the development of the procedure, the contents of one Vacutainer were left exposed to air for about 5 min after centrifugation to determine the extent to which cysteine is lost by oxidation before the sample can be acidified. A decrease of approximately 5% from the initial cysteine concentration was observed, indicating this not to be a significant source of error. Recovery studies were done for cysteine and cystine. For cystine, recoveries of 96 ? 4 and 99 + 5% were obtained. Several different recovery studies were done for cysteine. In one study, cysteine was added to the metaphosphoric acid solution used to precipitate the plasma proteins. One aliquot of the plasma was then acidified with metaphosphoric acid solution and another with the metaphosphoric acid to which had been added cysteine. Two separate experiments gave recoveries of 97 and 98%, which verify the procedures used. In another cysteine recovery study, the Vacutainers were opened before blood collection, deaerated with O,-scrubbed nitrogen, and 0.10 ml of a standard cysteine solution in 0.9% NaCl was added. The Vacutainers were then evacuated on a vacuum line. Blood was then collected, acidified, centrifuged, and analyzed as above. As indicated in Table 3, recoveries of approximately 15% were obtained for additions of cysteine calculated to increase the plasma cysteine concentration by 0.72 x IO-“. 0.93 x IO-“, and 1.40 x lo-” M. In these studies, the volume of blood collected was obtained by the difference in the weight of the tube before and after collection, and the plasma volume was calculated from the measured hematocrit. In the calculations. it was assumed that a negligible amount of

690

SAETRE

AND

RABENSTEIN

TABLE RECOVERY

STUDIES

3

OF CYSTEINE

ADDED

[CSHh111

[CSHI (M x 105)

Sample A A II a L r text

Initial l.IO? 0.02 1.10 2 0.02 I.16 2 0.02

Calculated after addition 1.82 2.03 2.56

T O PLASMA

(M x 105) Found after addition 1.20 f 0.02 1.25 k 0.03 1.41 k 0.03

Recovery (%) I4 16 I8

Found after addition

Initial 13.3 2 0.2 13.3 2 0.2 I I.4 + 0.2

Sample electrolytically reduced prior to hplc analysis. Percentage of added cyst&e not recovered in reduced form. Independent values calculated from the [CSHI and [CSH],,, for details.

found

K’

14.7 z 0.3 15.0 2 0.3 13.5 t 0.3

before

addition

ReCOV+ m 233 + 83 218 k 64 183 ? 44

and found

Initial

After addition

9.2 9.2 7.4

9.7 9.6 7.5

after addition.

See

cysteine enters the blood cells. To determine if the amount of EDTA in the Vacutainer has any effect on the recovery, a recovery study was done using a standard solution of cysteine which also contained 1 g/liter of EDTA. Again the recovery was about 15%. Finally, the plasma samples to which standard cysteine solutions had been added were electrolyzed and the total cysteine was determined. As can be seen from the results presented in Table 3, after electrolysis the recoveries were approximately 200%. The significance of the 200% recovery is discussed below. DISCUSSION

The levels obtained for reduced cysteine in urine (Table 1) agree well with those reported by Brigham et al. (7). The total cysteine levels in Table 1, however, are significantly higher than the sum ofthe cysteine and cystine levels reported by Brigham et al. (7). but are in good agreement with other results (1 1 - 13). The levels obtained for reduced cysteine in plasma (Table 2) lie in a rather narrow concentration range, which is considerably lower than the range reported by others (7,14,15), all of whom used iodoacetate to make the S-carboxymethyl derivative of cysteine. However, the total cysteine levels are all higher than the sum of the reported normal cysteine and cystine levels (7,16). The results from the second set of plasma cysteine recovery studies (Table 3) suggest that these differences result from relatively fast disulfide exchange equilibria in the plasma which affect the results obtained by methods which employ S-carboxymethyl or other S-methyl cysteine derivatives. Taken together, the results of the various plasma recovery studies indicate that the added cysteine is not lost by oxidation due to the presence of oxygen. If so, the recovery after electrolysis would be 100% rather than the 200% observed. Instead, they are consistent with a rather fast disulfide exchange equilibrium between cysteine (CSH), cystine (CSSC), protein

CYSTEINE

IN PLASMA

AND

URINE

sulfhydryl groups (PrSH), and protein-cysteine-mixed as represented by Eq. [l]. CSH + PrSSC * CSSC + PrSH

691

disulfides (PrSSC), [II

The added CSH reacts according to Eq. [I] to give CSSC. which is not detected when CSH is determined. However. when the total cysteine is determined, the CSSC is reduced giving 2 mol of CSH for each mole added and a 200% recovery. Evidence for protein-cysteine-mixed disulfides in plasma was first obtained by King (17), and a normal concentration of about 300 pM has been indicated (IX). The serum protein sulfhydryl content of normal persons is quite stable, and rarely fluctuates below the 400 to 600 pM range (19). When these values and the measured concentrations of cysteine and cystine are used. an equilibrium constant as defined by Eq. [2]

K = [CSSCI PrSHl [CSH]

[PrSSC]

[21

was estimated for the above reaction. As can be seen from Table 3, the calculated values for K from results obtained with and without the addition of cysteine are approximately the same, which indicates that the disulfide exchange equilibrium is established within the time scale of these experiments (approximately 10 min). The presence of a rather fast equilibrium between serum albumin and thiols or disulfides has also been observed by others (20,21). In those methods which use the S-carboxymethyl or other S-methyl cysteine derivatives, the formation of the derivative may cause the position of the equilibrium represented by Eq. [ I] to shift, depending on the relative rates of reaction of the derivatization reagent with the cysteine-SH and protein-SH groups. If so, the results for cysteine will be in error; for example, if the derivatization reagent reacts more rapidly with cysteine-SH, the results for cysteine will be high and those for cystine will be low. As mentioned above, these are exactly the differences observed when the results obtained by the hplc method with electrochemical detection are compared with the results from methods employing S-methyl derivatives. In the hplc method with electrochemical detection, the position of equilibrium is frozen upon acidification of the plasma. These considerations suggest,that the results in Table 2 more nearly represent normal plasma cysteine and cystine levels than do the currently accepted values (7,14-16). The hplc methods described in this paper are fast and sensitive, and both the reduced and total cysteine levels and the total homocysteine level can be determined. The speed and sensitivity are due largely to the characteristics of the mercury-based electrochemical detector. In this determination, the detector is operated at +O. IV ~5sthe SCE, and signals

SAETRE

692

AND RABENSTEIN

are obtained only as those compounds which are electroactive at this potential are eluted through the detector. Since the thiols and chloride are the only components of plasma and urine which are electroactive at this potential, this reduces considerably the requirements and the time of the chromatographic step as compared to amino acid analyzer procedures using ninhydrin detection. ACKNOWLEDGMENTS This research has been supported by the University of Alberta and the National Research Council of Canada. Financial support to R. S. by a University of Alberta scholarship is gratefully acknowledged.

REFERENCES 1. Schneider, .I. A., Bradley, K. H., and Seegmiller. J. E. (1968) J. Lob. C&n. Med. 71, 122-125. 2. Femandez. A. 0.. and Henry. R. (1965) Anul. Biochem. 11, 190-198. 3. Wronski. M. (1967) Biochem. J. 104, 627-633. 4. States, B., and Segal, S. (1973) Clin. Chim. Actu 43, 49-53. 5. Roston, S. (1963) Anul. Biochem. 6, 486-490. 6. Racker, E. ( 1955) J. Biol. Chem. 217, 867-874. 7. Brigham, M. P., Stein. W. H.. and Moore. S. (1960) J. C/in. Invrsr. 39, 1633- 1638. 8. Rabenstein, D. L.. and Saetre, R. (1977) Awl. Chrm. 49, 1036-1039. 9. Saetre, R., and Rabenstein, D. L. (1978) Anal. Chem. 50, 276-280. IO. Benesch, R.. and Benesch. R. (1957) Biochim. Biophys. Acto 23, 643-644. Il. Sullivan, M. X.. and Hess, W. C. (1936)5. Biol. Chem. 116, 221-232. 12. Reed, G. (1942)J. Bid/. Chum. 142,61-64. 13. Camien, M. N.. and Dunn. M. S. (1950) J. Biol. Chum. 183, 561-568. 14. London, D. R., and Foley, T. H. (1965) C/in. Sci. 29, 129-141. 15. Rosenberg, L. E., Dumont. J. L., and Holland, J. M. (1965) Ne\c, Engl. J. Med. 273, 1239-1245. 16. Crawhall, J. C.. Lietman, P. S., Schneider, J. A., and Seegmiller, J. E. (1968)Amer. J. Mc)d.

44, 330-339.

17. King, T. P. (1961)5. Biol. Chem. 236, PC5. 18. Jocelyn, P. C. (1972) Biochemistry of the SH Group, p. 253. Academic Press, New York. 19. Lorber, A., Pearson, C. M.. Meredith, W. L.. and Gautz-Mandell, L. E. (1964) Ann. Intern.

Med.

61, 423-434.

20. Isles, T. E.. and Jocelyn, P. C. (1963) Biochem. J. 88, 84-88. 21. Lorber. A., Chang, C. C., Masuoka. D., and Meacham. J. (1970) Biochem. 19, 1551-1560.

Pharmacol.

Determination of cysteine in plasma and urine and homocysteine in plasma by high-pressure liquid chromatography.

ANALYTICAL BIOCHEMISTRY Determination Homocysteine !I& 684-692 (1978) of Cysteine in Plasma and Urine and in Plasma by High-Pressure Liquid Chrom...
515KB Sizes 0 Downloads 0 Views