Anal. Chem. 1998, 70, 3505-3509
Quantitative Determination of Biological Sulfhydryl Groups by Postcolumn Derivatization and Elucidation of Microheterogeneity of Serum Albumins Tadashi Yasuhara† and Kiyoshi Nokihara*,‡
Junior College of Tokyo University of Agriculture, Sakuragaoka 1-1-1, Setagayaku, Tokyo 156, Japan, and Shimadzu Scientific Research Inc., Kanda-Nishikicho 1-3, Chiyodaku, Tokyo 101, Japan
A quantitative analytical system for biological sulfhydryl compounds has been developed using an ion-pair reagent with isocratic elution and an on-line postcolumn derivatization with Ellman-type reagents. As human or bovine serum albumin has 35 cysteinyl residues, one cysteinyl residue exists as a free sulfhydryl moiety, and this gives rise to the microheterogeneity in serum albumin. Here we report for the first time the quantitative characterization of the microheterogeneity of serum albumin. Cysteine was found to be the major molecule attached to the sulfydryl group of the serum albumins. Although glutathione could not be detected, the Cys-Gly element of glutathione was found. Freshly prepared human serum albumin from healthy volunteers contained 0.46 nmol of Cys/mL of serum, 0.24 nmol of Cys-Gly/mL of serum, and very small amounts of glutathione (0.02 nmol/mL). Cysteine residues in proteins and peptides play important roles in stabilizing the conformation of molecules through disulfide linkages in the process of protein folding and also occur as reactive sulfydryl groups that contribute to oxidation/reduction processes in vivo. Endogenous sulfhydryl compounds having small molecular weight bind to cysteinyl residue(s) of proteins or peptides and are considered to play a role in the folding process via disulfide interchange that can give rise to microheterogeneity of these cysteinyl residues in proteins. A sensitive and reproducible quantitative analytical system for small molecular sulfhydryl compounds is required in order to elucidate these biological pathways as well as to characterize the microheterogeneity. Human serum albumin (HSA)1 is a globular monomeric protein with a molecular weight of 66 500 and contains 35 cysteinyl * To whom correspondence should be addressed. Fax: +81-3-3219-5729. E-mail:
[email protected]. † Tokyo University of Agriculture. ‡ Shimadzu Scientific Research Inc. (1) Abbreviations used are as follows: BSA, bovine serum albumin; CE, capillary electrophoresis; GSH, reduced form of glutathione; GSSG, oxidized form of glutathione; HSA, human serum albumin; i.d., internal diameter; MALDI, matrix-assisted laser desorption/ionization; MES, 2-(N-morpholino)ethanesulfonic acid; MS, mass spectrometry; RP-HPLC, reversed-phase highperformance liquid chromatography; TOF, time-of-flight. S0003-2700(98)00263-7 CCC: $15.00 Published on Web 07/11/1998
© 1998 American Chemical Society
residues, of which 34 residues participate in 17 disulfide bridges.2 Indeed, plasma HSA contains 17 disulfides and one free cysteine at position 34 that is believed to be a mixture of a mercapto form and a non-mercapto form which has a modified cysteinyl residue. The sulfhydryl content was reported as 0.6-0.7 sulfhydryl/mol.3 Although Brown has reported the non-mercapto HSA as a mixed disulfide with cysteine and glutathione,2 no experimental details have been given. Recently, Ikegaya et al. have completely determined the 17 disulfide forms of recombinant HSA and indicated that HSA produced by yeast had a single cysteinyl residue at position 34, which is a free sulfhydryl moiety.4 This moiety is probably the site that causes microheterogeneity in plasma-derived HSA, as this has a similar sulfhydryl content. Small sulfhydryl compounds such as cysteine and glutathione might be attached to the free sulfydryl group on Cys-34 of HSA.5 Although this phenomenon is considered to be a part of the biological function of HSA, the biological and clinical significance of mercapto HSA has not yet been clarified. Sulfhydryl content should be related to disease and the change of sulfhydryl amounts of plasma HSA before and after renal filtration has been reported.6 However, only the sulfhydryl content was determined, and the types of molecules bound to the sulfhydryl group were not determined. The present paper describes a rapid, facile, and reproducible quantitative analytical method for biological sulfhydryl group determination by on-line postcolumn derivatization and its application to serum albumins that reveal their microheterogeneity. EXPERIMENTAL SECTION Materials. Plasma HSA (HSA, Pentex) was purchased from Miles Inc. Diagnostics Division (Kankakee, IL). Different preparations of bovine serum albumin (BSA) manufactured by Intergen (2) Brown, J. R. In Albumin Structure, Function and Uses; Rosenoer, V. M., Murray, O., Marcus, A. R., Eds.; Pergamon Press: New York, 1977; pp 2751. (3) Anderson, L.-O. In Plasma Proteins; Blombock, B., Hansen, L. P., Eds.; WileyInterscience: New York, 1979; pp 43-54. (4) Ikegaya, K.; Hirose, M.; Ohmura, T.; Nokihara, K. Anal. Chem. 1997, 69, 1986-1991. (5) Janatova, J.; Fuller, J. K.; Hunter, M. J. J. Biol. Chem. 1968, 243, 36123622. (6) Nishimura, K.; Harada, K.; Nakayama, M.; Sugii, A.; Uji, Y.; Okabe, H. J. Anal. Biosci. 1992, 15, 200-205.
Analytical Chemistry, Vol. 70, No. 16, August 15, 1998 3505
Figure 1. Schematic drawing of a system for characterization of sulfhydryl compounds by on-line postcolumn derivatization. Dual pump (1, 2), injector or autosampler (3), column oven (4), separation column (5), coil reactor (6), detector (7), eluent (8), reactant (9), and integrator (10).
(Purchase, NY) were generous gifts from Kokusai-Shiyaku Co. Ltd. (Kobe, Japan). BSA was also purchased from Sigma (St. Louis, MO). HSA and BSA were used as received. 5,5′-Dithiobis(2-nitrobenzoic acid) (DTNB) from Sigma, trifluoroacetic acid (TFA), acetonitrile, dithiothreitol (DTT), H3PO4, NaH2PO4‚2H2O, ethylenediamine tetraacetate (EDTA), NaOH, from Wako Pure Chemicals (Osaka, Japan), the reduced form of glutathione (GSH) and cysteinylglycine from Sigma, γ-glutamylcysteine from Kohjin (Tokyo, Japan), and tri-n-butylamine from Tokyo Kasei (Tokyo, Japan) were used without further purification. Human plasma was obtained from 30 healthy volunteers between 21 and 23 years old. MilliQ (Millipore, Bedford, MA) water was used. Sulfhydryl Determination. Sulfhydryl groups were determined according to the method of Ellman.7 Reduction of proteins was carried out with 0.5% DTT in 0.2 M NaH2PO4, 2 mM EDTA (pH 8.0). Residual DTT was not removed from the reduction mixture. Instruments. Reversed-phase high-performance liquid chromatography (RP-HPLC) was carried out using a model CCPM dual pump (Tosoh, Tokyo, Japan) for eluent and reactant transport, respectively, a Rheodyne or an autosampler, model AS8010 (Tosoh), for sample injection, and a detector, model UV8010 (Tosoh), for monitoring. The materials were separated on a TSKODS80TM column (4.6 mm i.d. × 250 mm, Tosoh). The above column and a coil reactor constructed with Teflon tubing (0.2 mm i.d. × 2 m) were placed in a column oven, model CO-8010 (Tosoh), and analysis was carried out at 40 °C under isocratic conditions. The system is indicated schematically in Figure 1. A model SC8020 (Tosoh) was used for system control. Sample Preparation and Analysis. Albumin was diluted to 20 nmol/100 µL with 0.1 M acetic acid, of which 10 µL was further diluted with water to 100 µL and reduced with 100 µL of 0.5% DTT in 0.2 M NaH2PO4 and 4.2 mM EDTA (pH 8.0) under a (7) Ellman, G. L. Arch. Biochem. Biophys. 1959, 82, 70-77.
3506 Analytical Chemistry, Vol. 70, No. 16, August 15, 1998
Figure 2. Standard separation profiles of various biological sulfhydryl compounds (500 pmol each). Cysteinylglycine (1), cysteine (2), homocysteine (3), GSH (4), γ-glutamylcysteine (5), mercaptoethanol (6), mercaptoacetic acid (7), mercaptopropionic acid (8), N-acetylcysteine (9), DTT (10). Column, TSK-ODS80TM (4.6 mm i.d. × 250 mm); monitored at 412 nm; operation temperature, 40 °C; elution solution (isocratic conditions with a flow rate of 0.8 mL/min), 0.1% tri-n-butylamine, 50 mM phosphate buffer (pH 2.3); reactant (transferred at a flow rate of 1.0 mL/min to the coil reactor), 0.025% DNTB, 2 mM EDTA, 0.1 M phosphate buffer (pH 7.6).
stream of nitrogen at room temperature in the dark with stirring for 60 min. The resulting mixture (20 µL) was injected into the HPLC. Ethanol (400 µL) was added to the human plasma (100 µL) and allowed to stand at 0 °C overnight. Precipitate obtained by centrifugation (10 000 rpm at 0 °C for 20 min) was suspended in 80% ethanol (1 mL) and recentrifuged as above. The resulting precipitate was again suspended in 70% ethanol and centrifuged as above. This protein was suspended in water (0.5 mL), and the supernatant after centrifugation as above was used as the albumin fraction. Next, 0.5% DTT in 0.2 M NaH2PO4 and 4.2 mM EDTA (pH 8.0) was added to 100 µL of this fraction contained in a Eppendorf tube (500 µL) under a stream of nitrogen and allowed to stand at room temperature for 1 h. The resulting mixture (20 µL) was injected into the HPLC. Analysis was carried out with 0.1% (v/v) tri-n-butylamine, 50 mM phosphate buffer, pH 2.30, as an eluent under isocratic conditions at a flow rate of 0.8 mL/min and monitored at 412 nm. Reactant consisting of 0.025% DNTB, 2 mM EDTA, and 0.1 M phosphate buffer (pH 7.6) was transferred at a flow rate of 1.0 mL/min to the above coil reactor, maintained at 40 °C. The column could be used for more than 50 analyses by using an autosampler and can easily be regenerated by washing with 0.1% TFA and 0.1% TFA acetonitrile in the conventional manner of the gradient elution. RESULTS AND DISCUSSION A quantitative analytical system for biological small molecular sulfhydryl compounds was constructed by on-line postcolumn derivatization with Ellman-type reagents. The system is illustrated in Figure 1, consisting of a HPLC pump operating with isocratic elution, an injector, and a column for RP-HPLC, followed by a coil reactor, contained in an oven, and a UV detector. Ellman reagentcontaining reactant for postcolumn labeling was continuously injected into the coil. By the use of tri-n-butylamine solution as an ion-pair reagent, biological sulfhydryl compounds such as cysteinylglycine, cysteine, homocysteine, GSH, γ-glutamylcysteine,
Table 1. Characterization of Sulfhydryl Compounds in Various Commercial Serum Albumin Preparations (mole %)
Figure 3. Calibration curves of sulfydryl compounds related to GSH.
mercaptoethanol, mercaptoacetic acid, mercaptopropionic acid, N-acetylcysteine, and DTT could be efficiently and quantitatively analyzed within 25 min. The flow rates of elution and of the reactant were changed to give optimal resolution and retention times. A standard separation profile of these biological sulfhydryl compounds is shown in Figure 2. GSH and its components could be determined within 10 min. Figure 3 indicates the calibration curve of sulfydryl compounds generated from glutathione. The present method allows determinations in the range from 10 pmol to 1 nmol. Hence, the difference slopes for the compounds seemed to reflect the reactivity in the postcolumn derivatization. GSH metabolism is one of the most important biological pathways. Meister reviewed the formation of sulfhydryl compounds from GSH by γ-glutamyltransferase and aminopeptidase M or dipeptidase to cysteinylglycine and cysteine.8 Thus, the analysis of GSH components can provide a tool for the elucidation of the biological processing. It can be predicted that GSH bound to sulfhydryl groups of proteins may partially decompose and cause heterogeneity in the material. Postcolumn derivatization offers the advantage that various labeling reagents can be used and collection of intact target material in preparative amounts is much easier; this material can then be subjected to nuclear magnetic resonance spectroscopy and/or mass spectrometry (MS). When 2,2′-dithiodipyridine was used as labeling reagent and monitored at 343 nm, only half the sensitivity was achieved (data not shown). Labeling with the fluorescent reagents can also be employed and may be expected to give a higher sensitivity, although it is not practical because of the high cost in running continuous infusion of expensive reagents to the coil reactor. The present sensitivity of DTNB with monitoring at 412 nm is a practical compromise and shows sufficient sensitivity with a low background. In addition, the stability of the DTNB solution was adequate for routine analysis using an autosampler. The present system allows isocratic separation with a high reproducibility and a stable baseline during analytical runs. The pH value of the eluent in Figure 2 is 2.30. Efficient analyses could be conducted between 2.0 and 3.5, although the pH of the eluent influences the retention time; for (8) Meister, A. Trends Biochem. Sci. 1981, 6, 231-234.
albumin (lot no.)a
Cys(SH)-Gly
Cys(SH)
sulfhydryl contents
(1) HSA (42) (2) BSA (M64606) (3) BSA, Cohn Fr. V, pH 5 (P7812) (4) BSA, Cohn Fr. V, pH 7 (N77069) (5) BSA, Modified Cohn Fr. V (P21075) (6) BSA, pH 7 (RT93606) (7) BSA (RT12801) (8) BSA, standard, pH 5 (M22204) (9) BSA, standard, pH 7 (R40102) (10) BSA (PT88705) (11) BSA (L58807) (12) BSA (P65806) (13) BSA (84H0341) (14) BSA (75H0894) (15) BSA (34H0275) (16) BSA (14H0246)
9.0 6.1 7.7
29.9 9.5 14.8
60 70 66
6.3
14.5
71
6.0
11.9
61
11.1
19.9
55
10.6 9.9
18.4 21.8
55 53
10.0
21.8
60
7.8 10.7 11.3 6.3 4.8 4.8 6.0
14.9 20.3 19.4 17.1 12.5 14.2 15.4
56 56 59 59 56 59 61
a Key: (1) Pentex, Miles; (2-5) Intergen, manufactured by the cold ethanol process; (2) microbiological grade; (6-12) Intergen, manufactured by the heat shock process; (6) biotechnology grade; (7) endotoxin reduced biotechnology grade; (10) fatty-acid-free biotechnology grade; (11) clinical reagent grade; (12) protease-free powder; (13-16) Sigma.
instance, when a buffer of pH 2.75 was used, cysteinylglycine eluted at 3.5 min and DTT at 25 min. More than 97% recovery of each sulfhydryl compound used in this study was achieved, as calculated from the UV absorption. Recently, Russell and Rabenstein reported a sensitive analysis of biological thiols and disulfides by capillary electrophoresis (CE).9 As CE requires several nanomole per liter (20 µL) samples, the minimum amount of sample necessary for analysis is similar in both methods. However, the present method has several advantages. In the CE method, peaks derived from reagents are also detected, while in the present method peaks are observed only for sulfhydryl compounds. In the CE method, derivatization (approximately 20 min) must be carried out prior to the analysis, while here this is not necessary; hence, the analysis is more simple and rapid. The sensitivity in the CE method is sample dependent, while here a similar sensitivity is found for all target compounds; therefore, it is suitable for following the time course of GSH. Although the CE method is not relevant for the analysis of sulfhydryl compounds bound through disulfide linkages, the present method can be used for such compounds after release from proteins by reduction with DTT, etc. Finally, HPLC is superior with respect to preparative separation. Both plasma and recombinant HSA have one cysteinyl residue at position 34 with a free sulfhydryl moiety, although sulfhydryl determination according to Ellman7 was 0.6-0.7,4 and this gives rise to the microheterogeneity of HSA. The present method was applied to reveal this heterogeneity in the different serum albumin (9) Russell, J.; Rabenstein, D. L. Anal. Biochem. 1996, 242, 136-144.
Analytical Chemistry, Vol. 70, No. 16, August 15, 1998
3507
Table 2. Sulfhydryl Compounds in the Freshly Prepared HSA Fraction of 30 Healthy Volunteers (nmol/mL Serum)
Figure 4. Profiles of two analytical examples. Background was subtracted. (a) BSA (3) in Table 1; (b) HSA (1) in Table 1. HPLC conditions are the same as in Figure 2. The large peak eluted after 20 min is DTT used in the sample preparation.
preparations. Characterization of sulfydryl compounds in the various commercial preparations (mole %) is listed in Table 1. As examples, the chromatograms of two analytical runs are shown in Figure 4. Albumin amounts were calculated from UV absorption at 279 nm,10 and sulfhydryl content was determined according to the method of Ellman.7 Although details of the downstream processing of commercially manufactured serum albumin were not disclosed, different protocols were employed for different purposes. Intergen’s BSA is manufactured in two distinct processes, by the cold ethanol precipitation and by the heat shock process. However, there were no significant differences in the present analytical results. In all commercial preparations, neither GSH nor γ-glutamylcysteine was found. The other biological sulfhydryl derivatives, not related to GSH, such as homocysteine, mercaptoethanol, mercaptoacetic acid, mercaptopropionic acid, and N-acetylcysteine, were also not found. Hence, cysteine, molecular mass 120.9 Da, was confirmed by MALDI-TOF-MS in eluate without derivatization, although Cys-Gly could not be identified by MS, since the eluate contained a high concentration of salts. Consequently, the commercial preparations of BSA contain 55-70% sulfhydryl molecules, and the major substituted material is 10-22% cysteine and 6-10% cysteinylglycine. The rest is considered to be dimer as reported by Janatova et al.5 On the other hand, HSA is 29.9% cysteine, 9.0% cysteinylglycine and contained almost no dimer. The amount of cysteine in BSA was (10) Peters, T. Jr. In Plasma Proteins; Putnam, F. W., Ed.; Academic Press: New York, 1975; Vol. 1, p 147.
3508 Analytical Chemistry, Vol. 70, No. 16, August 15, 1998
no.
Cys(SH)-Gly
Cys(SH)
GSH
γGlu-Cys(SH)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
0.82 0.27 0.24 0.27 0.24 0.31 0.22 0.26 0.22 0.24 0.22 0.29 0.24 0.22 0.24 0.24 0.24 0.20 0.22 0.22 0.24 0.22 0.18 0.20 0.18 0.13 0.16 0.18 0.16 0.16
0.87 0.42 0.42 0.40 0.38 0.45 0.36 0.40 0.36 0.47 0.45 0.47 0.49 0.44 0.56 0.51 0.47 0.44 0.49 0.40 0.44 0.42 0.40 0.42 0.49 0.40 0.42 0.45 0.51 0.44
0.21 0.03 0.03
0.16
av
0.24
0.46
0.02
0.03 0.03
0.03 0.03 0.06 0.03 0.03
0.03 0.03
Figure 5. Profile of a representative chromatogram. Background was subtracted. HPLC conditions are the same as in Figure 2.
less than that of HSA, and the difference seemed to be due to dimer. Sulfhydryl compounds in freshly prepared HSA fractions from the sera of 30 healthy volunteers were also determined. The HSA fraction was separated by ethanol precipitation of plasma and was carried out after uptake of blood without long-term storage. The results are summarized in Table 2. Figure 5 is a representative profile. In the microheterogeneity of freshly prepared HSA, sulfhydryl compounds bound to HSA are individual-specific, although cysteine was (0.46 nmol/mL of serum) the major substituent, together with cysteinyl glycine (0.24 nmol/mL). GSH, which did not occur in purified commercial HSAs, was found in very small amounts (0.02 nmol/mL), but no γ-glutamyl cysteine
was present. Sulfhydryl contents should be related to the status of some diseases,6 and hence characterization of such molecules is interesting and may be performed with the present system. CONCLUSION We have developed a rapid, sensitive, and reproducible quantitative analytical system for biological small molecular sulfhydryl compounds using an on-line postcolumn derivatization with Ellman-type reagents. With this method, sulfhydryl compounds bound to cysteinyl residue(s) of proteins or peptides can be easily determined. The system has been applied to serum albumins, and we have succeeded for the first time in quantitatively characterizing their heterogeneity. Hence, with regard to the sulfhydryl moiety, significant microheterogeneity was demonstrated. Cysteine was found to be the major molecule attached to the sulfydryl group of the cysteinyl residue of serum albumin. Although glutathione could not be detected, peptidyl elements of
the glutathione molecule were found. The system provides a useful method for quality control not only for plasma-derived proteins but also for recombinant proteins. The characterization and investigation of the relationship between diseases and sulfhydryl compounds will also be of clinical interest. ACKNOWLEDGMENT A part of the present study was supported by the Sunbor grant. We thank Dr. V. Wray, Gesellschaft fu¨r Biotechnologische Forschung, Braunschweig, Germany, for discussion and linguistic assistance with the manuscript.
Received for review March 9, 1998. Accepted May 29, 1998. AC9802630
Analytical Chemistry, Vol. 70, No. 16, August 15, 1998
3509
