Simultaneous Fluorometric Determination of Thiols and Disulfides by

samples by capillary zone electrophoresis. Nuran Ercal , Kang Le , Piyanee Treeratphan , Richard Matthews. Biomedical Chromatography 1996 10 (1), ...
8 downloads 0 Views 678KB Size
Anal. Chem. 1982, 5 4 , 1951-1955

TEA results from parallel cartridge no. 1which was preeluted with DCM. The results of this experiment not only support the possibility that the DCM wash removes compounds which suppress the TEA response but also refutes the unlikely possibility that the DCM preelution is responsible for artifact formation of NMOR. The TEA results for NMOR were consistently lower by a factor of 2 to 3 when the DCM wash was omitted. Although this difference is not considered serious it does point out the usefulness of the selective, rapid GC/MS method for doing confirination studies.

ACKNOWLEDGMENT The authors thank Michael D. Kelly for his technical assistance. LITERATURE CITED "Nitrosamines"; IARC 8oi Pubiicatlon: 1980; Voi. 18 and 19. Flne, D. H. "Monitoring loxic Substances"; Schuetzle, D., (Ed.: American Chemlcai Society: Washington, DC, 1979; ACS Symp. Ser. No. 94, p 247. Fine, D. H. presented at "Risk Assessment of N-nitroso Compounds for Human Health", Heidelberg, Germany, May 1979. Fine, D. H. Callf. Alr Envkon. 1980, 7 , 1-4. Rounbehler, D. P.; Reisch, J. W.; Coombs, J. R.; Fine, D. H. Anal. Chem. 1980, 52, 273-276. Maki, T. Bull. Environ. Confam. Toxicol. 1980, 25, 751-754. Fine, D. H.; Ross, R.; Rounbehler, D. P.; Silvergleid, A.; Song L. J. Agric. FoodChsm. 1978, 2 4 , 1069-1071.

1951

(8) Hartmetz, G.; Slemrova, J. Bull. Environ. Contam. Toxicol. 1980, 25, 106- 112. (9) Hansen, T. J.: Archer, M. C.; Tannenbaum, S . R. Anal. Chem. 1979, 51, 1526-1528. (IO) Walker, E. A,; Castegnaro, M. J. Chromafogr. 1980, 187, 229-231. (11) Webb, K. S.;Gough, T. A. J. Chromafogr. 1979, 177, 349-352. (12) Gough, T. A. Analyst (London) 1978, 103, 785-806. (13) Webb, K. S.;Gough, T. A.; Carrick, A.; Hazeiby, D. Anal. Chem. 1978. 51. 989-992. (14) Gough, T.'A.; Webb, K. S.; Pringuer, M. A.; Wood, B. J. J. Agrlc. Food Chem. 1977, 25, 663-667. (15) . . Kaufmann. H.: RaDD, . . U.: Mlrna, A.; Harada, K.Varlan MAT ADDlication .. Note No. 37, 1978. (16) Manufacturers Literature, Envirochem. Inc.: Kemblesvilie, Pa, 1979; Technical bulletins 080,084-087. (17) Wolf, M. H. ASL Bulletin 3334; Thermo Electron Corp.: Waltham, MA, 1980

(18) ThermoSorb/N Air Sampler Analysis Instructions Note DS-11, Thermo Electron Corp., Feb 1980. (19) Rainey, W. T.; Christee, W. H.; Lijinsky, W. Biomed. Mass. Specfrom. 1978, 5,395-408. (20) Rounbehler, D.P.; Reisch, J.; Fine, D. H. Food Cosmef. Toxlcol. 1980, 18, 147-151. (21) Yeager, F. W.; Van Guiick, N. N.; Lasoski, B. A. Am. Ind. Hyg. .. AssocTJ. 1980, 41, 148-150. (22) Fajen, J. M.; Carson, G. A.; Rounbehier, D. P.; Fan, T. V.; Goff, U. E.; Wolf, M. H.; Edwards, G. S.; Fine, D. H.; Reinhold, V.; Biemann, K. Sclence 1979, 205, 1262-1264. (23) Goff, E. U.; Coombs, J. R.; Balnes, T. M.; Flne, D. H. Anal. Chem. 1980, 52, 1833-1896.

RECEIVED for review July 27,1981. Resubmitted January 28, 1982. Accepted July 2, 1982.

Simultaneous Fluorometric Determination of Thiols and Disulfides by Liquid Chromatography with Modified Postcolumn Derivatizat ioti1 Hiroshl Nakamura" and Zenzo Tamura Department of Analytical Chemistry, Faculty of Pharmaceutical Sciences, University of Tokyo, 7-3- 1, Hongo, Bunkyo-ku, Tokyo 1 13, Japan

By modlflcatlon of our postcolumn derlvatlzatlon system for thlols, an HPLC method has been developed for the slmultaneous fluorometric determlnatlon of blogenlc thlols and dlsulfldes. They are separated by anion-exchange chromatography wlth gradlent elution,,reacted with sodlum sulflte at pH 6-7 and o-phthalaldehyde (OPA) at pH 9.4-10.5 In the presence of taurlne to produce fluorescent lsolndole fluorophores, and monitored with a fluorescence detector (A,, = 360 nm; A, > 405 nm). Under the optimized conditions for chromatography and the postcolumn derlvatizatlon, 250 pmol of L-cysteine (CySH), L-cystine (CySS), glutathlone (GSH), and oxldlzed glutathione (GSSG) and 750 pmol of 3'-dephosphocoenzyme A (DP-CoASH), coenzyme A (CoASH), and thelr oxldlzed forms (DP-CoASS and CoASS) can be determlned. The relatlve standard devlations ( n = 5 ) of the method are 1.35%, 4.09%, and 6.87% for the analyses of 5 nmol of CySS and GSSG and 8.5 nmol of CoASS, respectlvely.

The physiological and biochemical significance of thiol and disulfide functions has been well documented (1). For the determination of thiols, %varioushigh-performance liquid chromatographic (HPLC) methods combined with fluores-

Scheme I 5-R +

kNH2

+

-

RSH -2H20

~\

N

-

6

cence detection (2-11) or voltammetric detection (12-19) have been reported as sensitive and specific means for the analyses of biological materials. On the other hand, the previous HPLC methods for disulfides with UV detection (20-24) lack specificity. Although the postcolumn ligand-exchange reaction, reported by Frei et al. ( I O ) , between the nonfluorescent Pd(Wcalcein complex and disulfides liberating fluorescent calcein is sensitive, it is not specific for disulfides and many sulfur-containing compounds yield fluorescence upon reacting with the reagent. The HPLC method reported by Takahashi et al. (25),which involves the pretreatment of disulfides with potassium cyanide and successive derivatization of the resultant thiols with fluorogenic reagent N-(g-acridinyl)maleimide prior to separation, does not give the information on the original structure of disulfides, Le., whether they are mixed disulfides or symmetrical disulfides. From the biochemical point of view, a sensitive and specific analytical method which allows the simultaneous determination of thiols and disulfides is highly desired. Beales et al. (26) performed postcolumn reduction of D-penicillamine disulfide (PenSS) by passing the

0003-2700/82/0354-1951$01.25/0Q 1982 American Chemlcal Society

1952

ANALYTICAL CHEMISTRY, VOL. 54, NO. 12, OCTOBER 1982

Scheme I1 RSSR

+

Na2S03

====

RSS03Na

RSNa

sample through a column of dihydrolipoamide beads to form D-penicillamine (PenSH) followed by colorimetric detection with 5,5'-dithiobis(2-nitrobenzoic acid). Eggli and Asper (I7) introduced the column eluate to an electrode of amalgamated silver powder to reduce L-(-)-cystine (CySS) to L-cysteine (CySH) which was detected downstream amperometrically. In a previous paper (II), we reported a HPLC method for the determination of thiols using the postcolumn derivatization with o-phthalaldehyde (OPA) and taurine (Scheme I). In the present paper, we will report a modification of the previous paper (11) for the simultaneous determination of thiols and disulfides involving the separation of them by anion-exchange chromatography, postcolumn sulfitolysis of disulfides (Scheme 11), and successive derivatization of thiols with OPA and taurine into isoindole fluorophores. EXPERIMENTAL SECTION Materials. AS-Pellionex SAX (polystyrenematrix strong anion exchanger) and the packed columns (4.6 mm i.d. X 25 cm) of Partisil-10 SAX (10 pm, microparticulate silica-bonded strong anion exchanger) were purchased from Whatman (Clifton, NJ). D-Penicillamine disulfide (PenSS), DL-homocystine, oxidized coenzyme A (CoASS, lithium salt), and mixed disulfide between coenzyme A (CoASH) and glutathione (GSH) (CoASSG, sodium salt) were purchased from Sigma (St. Louis, MO). Oxidized glutathione (GSSG), thiamine disulfide nitrate, dithiodiglycolic acid, a,a'-dithiodipropionic acid, P,p-dithiodipropionic acid (Tokyo Kasei, Tokyo, Japan), 5,5'-dithiobis(2-nitrobenzoicacid) (Nakarai Chemicals, Kyoto, Japan), L-(-)-cystinylbis(g1ycine) (Vega Biochemicals, Tucson, AZ), and 2-hydroxyethyl disulfide (Aldrich, Milwaukee, WI) were used as received. Oxidized Nacetylcysteine (NAcCySS)and oxidized 3'-dephosphocoenzyme A (DP-CoASS) were prepared by dissolving the corresponding thiol NAcCySH or DP-CoASH at a concentration of 10 mM in 0.2 M sodium borate buffer (pH 10.0) and by exposing to atmospheric oxygen at 20 "C for 18 h. The sources of thiols and other chemicals used have been described previously (11). Preparation of Thiol and Disulfide Solutions. Ten or one millimolar solutions of thiols and disulfides were prepared separately with 0.2 M lithium citrate buffer (pH 4.0) supplemented with 1 mM EDTA. The buffer containing EDTA was used to prepare dilute solutions. Preparation of Postcolumn Reagents. Sulfite-Taurine Reagent. To 0.75 M NaH2P04-K,HP04buffer (pH 7.00) were added taurine and EDTA to make 25 mM and 1 mM solutions, respectively. The sulfite-taurine reagent was prepared daily by adding 15 mM NazSOaand 1 mM L-ascorbic acid to the above taurine solution. Both EDTA and L-ascorbic acid served as antioxida Its for sulfite ion. OPA Rdagent. The carbonate buffer was prepared by titrating 1M NaZCO3solution containing 1mM EDTA with solid NaHCOS to pH 10.50. The OPA reagent was prepared daily by diluting 250 mM OPA solution in methanol with the carbonate buffer to make 25 mM OPA solution. Standard Procedure of HPLC. The separation system used was essentially the same as described previously (11) except for the use of 10 mM citric acid containing 1 mM EDTA (solvent A) and 200 mM lithium citrate buffer (pH 4.66) containing 1mM EDTA (solvent B). The linear gradient elution was performed either by adding solvent B to solvent A at a rate of lO%/min (elution system I) or by replacing solvent A with solvent B in 1 h (elution system 11). The eluent was delivered at a flow rate of 0.7 mL/min through a Mini-micro pump (Type KHD-16; Kyowa Seimitsu, Tokyo, Japan). The column eluate was mixed in a three-way tee (Type KYS-16; Kyowa Seimitsu) with the sulfitetaurine reagent, delivered at a flow rate of 0.35 mL/min with a double plunger-type Mini-micro pump (Type WD-1; Kyowa Seimitsu). The outlet of the tee was connected to a 20-m length of tubing, which was immersed in a water bath at 80 OC in a reaction incubator (Type KWB-2R, Kyowa Seimitsu) and connected to a 38-cm length of tubing prior to a second three-way

MINUTES

Elution curve of a synthetic mixture of disulfides. Sample size consisted of a 5-ULaliauot Containing 6 nmol of CyNH,SS, 6 nmol of cyss 5 nmol of GSSG; 15 nmol ofbP-CoASS, and-20 nmoi of CoASS. Figure 1.

W

z

j

0

3

MINUTES

Separation of selected disulfides and thiols. Sample size consisted of a 20-pL aliquot containing 10 nmol of CyNH,SS, 8 nmol of CySS, 6 nmol of GSH, 8 nmol of &/3'-dithidipropionic acid, 8 nmol of GSSG, 10 nmol of NAcCySS, 30 nmol of DP-CoASH, 50 mol of DP-CoASS, 40 nmol of CoASH, and 20 nmol of CoASS. Figure 2.

tee. The mixture of eluate and the sulfite-taurine reagent was reacted in the second tee, which was immersed in a running tap water bath, with the OPA reagent delivered at a flow rate of 0.35 mL/min with the double plunger-type pump. The outlet of the second tee was connected to a 1-m reaction coil. The detection and recording of fluorescent peaks were conducted as described previously (11). Quantitation of Thiols and Disulfides. Typically, to 50 pL of sample solution was added 50 p L of a given concentration of @,/3'-dithiodipropionicacid which was used as an internal standard, the solution was vortex-mixed, and an aliquot was injected. The amount of compound was calculated from the working curve of the peak height ratio t o the internal standard vs. picomoles of the compound. RESULTS AND DISCUSSION Separation of Biogenic Disulfides and Thiols. The conditions for the separation of biogenic disulfides were first examined by using unoptimized postcolumn derivatization. The disulfides were found to be well separated on the column of Partsil-10 SAX with linear gradient elutions at a flow rate of 0.7 mL/min, using 10 mM citric acid containing 1 mM EDTA (solvent A) and 200 mM lithium citrate buffer containing 1 mM EDTA, final pH 4.66 (solvent B). The separation condition was essentially the same as that previously reported for thiols (11);however, the solvents were modified in this work by adding EDTA to avoid the possible oxidation of organic sulfur compounds. As shown in Figure 1,cystamine (CyNH,SS), CySS, GSSG, DP-CoASS, and CoASS were separated in about 30 min by adding solvent B to solvent A at a rate of lO%/min (elution system I). CoASSG coeluted with DP-CoASS. Since CoASS was most strongly retained on the column among biogenic thiols and disulfides tested, elution system I was convenient for rapidly scanning the entire profile of detectable compounds. Figure 2 shows the separation of seven disulfides and three thiols by using elution system I1 described in the Experimental

I

WI w

ar 0

2

cyss

cyss P

'\i

I

w 0 w z

-

O

Y

o

\

L ;,;::." ;B \

/'

0

""'%

b,:,p, 4

/ COASS

.LS=~-'

E

6

,

,

IO

I2

PH

NazSOg ( m M )

Figure 3. Effect of pH on the sulfitolysis of disulfides. Fifty microliters each of 10 mM disulfide, 10 mM EDTA, and 100 mM sodium sulfite and 350 p L of 0.2 M buffer were mixed and Incubated at 40 OC for 30 mln. A 1 0 - ~ Laliquot of the reaction mixture was assayed for thiol.

Figure 4. Effect of sulflte concentration on the sulfitolysis of disulfides. Fifty microliters each of 10 mM disulfide, 10 mM EDTA, and various concentrations of sodium sulfite and 350 pL of 0.5 M phosphate buffer (pH 7.0) were mixed and incubated at 40 OC for 10 min. A 10-pL

Buffers used are citric acid-monopotassium citrate (pH 2 and 3), monopotassium citrate-NaOH (pH 4, 5, and 6), KH2P04-Na,HP0, (pH 7 and 8), boric acid-Na0H (pH 9 and and Na,HPO,-NaOH (pH 1 1 and 12).

aliquot of the mixture was assayed for thiol.

lo),

Section. L-(-)-Cystinylbis(g1ycine)eluted between CyNHzSS and CySS. 2-Hydroxyethyl disulfide (oxidized form of 2mercaptoethanol) appeared just after CySS. Dithiodiglycolic acid eluted between NAcCySS and DP-CoASH, though its peak height was much smaller compared to the latter two on a molar basis. Also appearing in this region of the chromatogram were the unresolved pairs: CyNH2SH/CyNH2SS, CySHjCySS, NAcCyEZHjGSSG, and CoASSGjCoASH. However, resolution of these pairs was not always necessary as only the peaks due to disulfides completely disappeared by omitting sodium sulfite from the sulfite-taurine reagent. The relative standard deviations (n = 5) of the retention times were 1.69, 1.15, 0.92, and 1.14% for CySS, GSSG, DPCoASS, and CoASS with elution system I and 2.70, 0.65, and 0.48% for CySS, GSSG, and CoASS with elution system 11. Determination of Conditions for the Postcolumn Sulfitolysis of Disulfides. In preliminary experiments for the production of thiols from disulfides, various nucleophilic reactions for disulfides including cyanolysis with KCN, reduction with sodium arsenite or tri-n-butylphosphine, and sulfitolysis with Na2S03were examined. The use of cyanide and phosphine resulted in high background due to the reaction of the OPA reagent with the nucleophilic reagents. The reduction of disulfides with sodium arsenite was considered to be inadequate for the above purpose because of its slow reaction and the toxicity of arsenic. Consequently, we tentatively chose the sulfitolyriis reaction for producing thiols from disulfides. The reaction conditions were studied with some biogenic disulfides (CyNH2SS,CySS, GSSG, and CoASS) by measuring the amounts of thiols produced by flow injection analysis. Unless otherwise specified, solvent A was used as the eluent (0.7 mL/mirn) and 25 mM OPA in 50% methanol and then 25 mM taurinu? in 1 M carbonate buffer (pH 10.5) were separately pumped at a flow rate of 0.35 mL/min. As shown in Figure 3, the sulfitolysis reaction of disulfides proceeded most effectivelly at around pH 7 except for the case of CyNH2SS. When 1 m M disulfides and various concentrations of sulfite were reacted at pH 7 and 40 OC in the presence of 1 mM EDTA, the reaction mixtures containing ca. 5 mM sulfite gave the most intense fluorescende upon reacting with OPA and taurine (Figure 4). The use of higher concentrations of sulfite caused decreased fluorescence due probably to the consumption of OPA by coexistent bisulfite ion (HS03-) a t pH 7 since HS03- exhibited the similar reactivity as thiols in some cases (12, 27, 28). When 10 nmol of disulfide was injected and reacted downstream with 5 mM sulfite a t pH 7 and 40 "C in various lengths of coil, CyNHzSS and CySS gave maximal peak heights after the sulfitolysis reaction in coils shorter than 10 m. Also peak heights de-

TEMPERATURE ("C )

Figurn 5. Effect of temperature on the sulfitolysis of disulfides as studied by flow Injection analysis. A 1 0 - ~ Laliquot containing a disulfiie (5 nmloi for CySS and 10 nmol for others) was injected to the HPLC postcolumn derivatization system from which the columns were removed. Eluents used are (a) solvent A and (b) solvent 8.

creased with increasing length of coil. However, the peak heighits of GSSG and CoASS were maintained maximal when the coil lengths were 5-30 and 20-40 m, respectively. As a compromise of these different results for optimal length of coil, the length of the sulfitolysis coil was set a t 20 m. Although the peak height of CoASS was small in the sulfitolysis at 40 OC, it became large by carrying out the sulfitolysis at higher temperatures (Figure 5). The sulfitolysis reaction a t elevaled temperatures decreased the peak heights of CyNHzSS and CySS when solvent B was used as the carrier solvent for the flow injection analysis. Since these two compounds and GSSG eluted essentially with solvent A and the elevated temperature did not decrease their peak heights, 80 "C was chosen for the sulfotolysis reaction. The efficiencies of the sulfitolysis of these disulfides under the conditions were found to be 80-100% by comparing the peak heights of disulfides to those of corresponding thiols using flow injection analyses. Optimization of Conditions for the Postcolumn Derivatization with OPA and Taurine. As in the previous paper ( I I ) , taurine was used as the primary amine reagent for thie fluorogenic reaction for thiols. T o determine the optimal pH for the fluorogenic reactions, a 10-pL aliquot contaaning 10 nmol of a thiol was injected into the stream of solvent A (0.7 mL/min) and mixed with 25 mM taurine in 0.2 M buffer with various pHs and successively 25 mM OPA in 50% methanol, both delivered at a flow rate of 0.35 mL/min. As shown in Figure 6, CyNH,SH, GSH, and CoASH fluoresced most intensely at around pH 9-10. The fluorescence-pH profile of CySH was different from those of other thiols and the fluorescence intensity rapidly increased above pH 9, which is not explicable. Since OPA and taurine gradually reacted to produce a red color ( I I ) , they could not be used in a mixture. However, taurine and sulfite showed no

1954

ANALYTICAL CHEMISTRY, VOL. 54, NO. 12, OCTOBER 1982

V

Z

w

V

w

a

0

Flgure 8. Effect of pH on the OPA fluorogenlc reactlon of thiols. The buffers used are the same as give in Figure 3.

Table I. Relative Fluorescence Intensities of Various Disulfides Obtained by Flow Injection Analysisa disulfide homocystine oxidized glutathione L-(-)-cystinylbis( glycine) L-(-)-cystine oxidized coenzyme A cystamine DL-oelipoic acid thiamine disulfide D-penicillamine disulfide 5,5'-dithiobis(2-nitrobenzoic acid)

solvent Ab

solvent B C

108

104

lOOd

lOOd

98.6 56.1 46.9 3.59 0.42 0.03

Q.03 nil

101 85.6 29.4 21.5 0.62 0.06 nil nil

a Apparatus: the HPLC postcolumn derivatization system from which columns are removed. Sample size: a 10-pL ali uot containing 10-100 nmol of a test compound. ? 10 mM citric acid containing 1 mM EDTA. 0.2 M lithium citrate buffer (pH 4.66) containing 1mM EDTA. Oxidized glutathione is arbitrarily taken as 100.

reaction and they were combined as the sulfite-taurine reagent. To adjust the pH of the sulfitolysisreaction at around 7, we prepared the sulfite-taurine reagent with 0.75 M phosphate buffer (pH 7.00). When the eluate was mixed with the sulfite-taurine reagent at a flow ratio of 2 to 1, the pH of the mixture was between 7.0 and 6.0 depending on the gradient using solvents A and B. To perform the fluorogenic reaction at pH 9-10, OPA was dissolved in ca. 1 M carbonate buffer (pH 10.50). The apparent pH of the waste was initially 10.5 and gradually decreased with the gradient elution to reach a constant value of 9.4 where the eluent was 100% of solvent B. This demostrated the effectiveness of the carbonate buffer for the adjustment of pH. Under the established conditions described in the Experimental Section, the OPA reaction was rapid at room temperature. A 1-m length of the OPA reaction coil, which corresponded to a residence time of 8.4 s, gave the most intensely fluorescent peaks for CyNH2SH,CySH, GSH, and CoASH; the use of shorter or longer reaction coils resulted in the decrease in peak height in all cases. On the basis of the above findings, the HPLC system for the postcolumn derivatization of biogenic thiols and disulfides was established as described in the Experimental Section. The relative fluorescence responses of various disulfides using these conditions are summarized in Table I. When elution system I1 (Figure 2) was used, the lower limits of determination were 250 pmol for CySH, CySS, GSH, and GSSG and 750 pmol for DP-CoASH, DP-CoASS, CoASH, and CoASS as injected amounts. The precision of the method was evaluated by analyzing the mixture of CySS, GSSG, and CoASS five times. The relative standard deviations were

MINUTES Flgure 7. Chromatogram of the perchloric acid extract of rat liver. To 7 g of fresh llver from a male Wistar rat (220 g) was added 14 mL of 0.4 M perchloric acid containing 25 mM EDTA, homogenized with a

Potter-Elvehjem homogenizer in an ice bath, and centrifuged at 150009 for 5 min. The supernatant (500 pL) was neutralized with 3 M KHCO, (50 pL). After centrifugation, the supernatant (IO pL) was analyzed by the HPLC system using elution system 11.

1.35% for 5 nmol of CySS, 4.09% for 5 nmol of GSSG, and 6.87% for 8.5 nmol of CoASS, respectively. The present HPLC method permitted the simultaneous determination of thiols and disulfides. Since neither the oxidation of thiols to disulfides nor the exchange reaction of thiols with disulfides was observed during chromatography, the present method seems to be extremely suited for the characterization of thiols and disulfides, especially mixed disulfides in a mixture. Actually, we found in our preliminary tests that the method was applicable to the simultaneous determination of thiols and disulfides in biological materials. As an example, the chromatogram of perchloric acid extract of rat liver is shown in Figure 7. GSH and minute amounts of GSSG and CySH were detected. A trace amount of CySS was contained in the peak of CySH. The OPA fluorogenic reaction employed in this study is highly specific for thiols and therefore the present HPLC assay based on sulfitolysis and the OPA reaction is specific only for thiols and disulfides. The characterization of thiols and disulfides was easily achieved by comparing the retention times of samples with authentic compounds and by omitting sulfite from the sulfite-taurine reagent. This resulted in complete disappearance of disulfide peaks while thiol peaks were virtually unchanged. Although CyNHzSH and CySH coeluted with their oxidized disulfides in our separation systems, differential determination was possible by omitting sulfite from the postcolumn reagent. Actually, the two-run differential approach was used to for quantitative analysis of disulfides in the presence of coeluting thiols. Another approach is to employ a chromatographic system used for the separation of primary amino compounds in the brain (29) without any modification of the present postcolumn derivatization system. Although there are, in the presence of coeluting thiols or disulfides, compounds with primary amino groups, e.g., amino acids and peptides, which may form interfering isoindole products with OPA, the concentrations of coeluting amino compounds are usually extremely low in comparison with that of the taurine reagent. The fact that the lower limits of determination of thiols and the corresponding disulfides were almost the same suggests the complete sulfitolysis of disulfides under the conditions employed in this work. However, the sensitivities of detection of thiols in the present method are about one-tenth compared with those we reported previously for thiols (11). This decreased sensitivity is due to the use of more concentrated

1955

Anal. Chem. 1982, 5 4 , 1955-1959

Scheme I11

(2) Lankmayr, E. P.; Budna, K. W.; Muller, K.; Nachtmann, F. Fresenius’ Z.Anal. Chem. 1979,295, 371-374. (3) Lankmayr, E. P.; Budna, K. W.; Muller, K.; Nachtmann, F. J. Chromatogr. 1981,222, 249-255. (4) Takahashl, H.; Nara, Y.; Meguro, H.; Tuzlmura, K. Agric. Biol. Chem. 1979,43, 1439-1445. (5) Anzai, N.; Kimura, T.; Chida, S.; Tanaka, T.; Takahashi, H.; Meguro, H. Yakugaku Zasshll981, 101, 1002-1009. (6) Jarrott, 6.; Anderson, A,; Hooper, R.; Louis, W. J. J. Pharm. Sci. 1081. ..- ., 7 . 0 ,. 665-667. ... ... .

buffer solutions in which the OPA reaction may be incomplete and/or the efficiencies1 of excitation and emission may also partially be suppressed. Some disulfides gave little or no fluorescence after the postcolumn derivatization reactions adopted in this work. The low response of D-penidamine disulfide (PenSS) is thought to be due to the steric hindrance around the disulfide bond which lies between two carbon atoms with dimethyl groups. Although steric hindrance caused by the methyl groups has been observed in the OF’A reaction of Dpenicillamine (PenSH) ( I I ) , the disturbance of the sulfitolysis reaction of PenSS by the bulky methyl groups would be more serious as in the case of the reduction of penncillamine-S-sulfonic acid by dithiothreitol (30,31). Little to faint fluorescence of &!,a‘-dithiodipropionic acid in the present assay, compared with the intense fluorescence of /3,/3’-dithiodipropionic acid, would be another example for the steric hindrance by methyl groups on the carbon atom bearing the sulfur. The reduced form of thiamine disulfide is considered to be easily transformed into thiamine losing its thiol group, which may be responsible for the practically nonfluorescent property of the disulfide in the present assay. DL-&!-Lipoicacid did not give any substantial fluorescence as presented in Table I, suggesting the absence of sulfitolysis with the cyclic disulfide. By analogy with the thiol-dithiothreitol mixed disulfide whose equilibrium constant for the cyclization reaction is reported to be about lo4 (32),the equilibrium of the reaction in Scheme I11 should extremely incline towaird the left side. In fact, we observed that DL-a-lipOiC acid redluced by sodium borohydride, N&H4 (33), fluoresced intensely upon reacting with OPA in the presence of taurine. The use of Na13H4 instead of NazSOBfor the postcolumn reduction of disulfides to thiols is being studied in our laboratory.

LITERATURE CITED (1) Jocelyn, P. C. “Biochemlstry of the SH Group”; Academic Press: New York, 1972.

(7) Fahey, R. C.; Newton, 0. L.; Dorlan, R.; Kosower, E. M. Anal. Biochem. 1981. 1 1 1 . 357-365. (8) Newton, G. L.; Dorian, R.; Fahey, R. C. Anal. Biochem. 1981, 114, 383-387. (9) Takahashi, H.; Yoshlda, T.; Meguro, H. Bunsekl Kagaku 1981, 30, 339-341. (IO) Werkhoven-Goewie, C. E.; Niessen, W. M. A,; Brinkman, U. A. Th.; Frei, R. W. J. Chromatogr. 1981,203, 165-172. (11) Nakamura, H.; Tamura, 2. Anal. Chem. 1981,53, 2190-2193. (12) Shimada, K.; Tanaka, M.; Nambara, T.; Imai, Y.; Abe, K.; Yoshinaga, K. J. Chromatogr. 1982,227, 445-451. (13) Rabensteln, D. L.; Saetre, R. Anal. Chem. 1977, 49, 1036-1039. (14) Saetre, R.; Rabenstein, D. L. Anal. Chem. 1978,50, 276-280. (15) Rabenstein, D. L.; Saetre, R. Clin. Chem. (Winston-Salem, N . C . ) 1978,24, 1140-1143. (16) Saetre, R.; Rabenstein, D. L. J. Agric. Food Chem. 1978, 26, 982-983. (17) Eggll, R.; Asper, R Anal. Chim. Acta 1978, 101, 253-259. (18) Kreuzig, F.; Frank, J. J. Chromatogr. 1981,218, 615-620. (19) Bergstrom, R. F.; Kay, D. R.; Wagner, J. G. J. Chromatogr. 1981, 222, 445-452. (20) Halvorsen, 0.; Skrede, S. Anal. Biochem. 1980, 107,103-108. (21) Howard, S. C.; McCormlck, D. B. J. Chromatogr. 1981, 208, 129-131. (22) Ingebretsen, 0. C.; Farstad, M. J. Chromatogr. 1980,202, 439-445. (23) Jones, D. P.; MoldBus, P.; Stead, A. H.; Ormstad, K.; Jornvali, H.; Orrenlus, S. J. 8/01. Chem. 1979,254, 2787-2792. (24) Reed, D. J.: Babson, J. R.; Beatty, P. W.; Brodie, A. E.; Ellis, W. W.; Potter, D. W. Anal. Biochem. 1980. 106, 55-62. (25) Takahashi, H.; Narrr, Y.; Yoshida, T.; Tuzimura, K.; Meguro, H. Agric. Biol. Chem. 1981,45, 79-85. (26) Beales, D.; Flnch, R.; McLean, A. E. M.; Smith, M.; Wilson, I.D. J. Chromatogr. 1981,226, 498-503. (27) Nakamura, H.; Tamura, 2. Chem. Pharm. Bull. 1975, 23, 1261-1270. (28) Nakamura, H.; Tamura, 2 . Chem. Pharm. Bull. 1974, 22, 1632-1 638. (29) Nakamura, H.; Zlmmerman, C. L.; Pisano, J. J. Anal. Biochem. 1979, 93, 423-429. (30) Nakamura, H.; Tamura, 2. J. Chromatogr. 1975, 104, 389-398. (31) Nakamura, H.; Tamura, 2 . Chem. Pharm. Bull. 1977, 2 5 , 2369-2377. (32) Cleland, W. W. Biochemistry 1984,3, 480-482. (33) Gunsaius, I . C.; Barton, L. S.; Gruber, W. J. Am. Chem. SOC. 1956, 78, 1763-1766.

RECEIVED for review May 5 , 1982. Accepted July 16, 1982. Presented in part at the 102nd Annual Meeting of the Pharmaceutical Society of Japan, Osaka, April 3-5,1982. The authors are grateful to the Iatrochemical Institute Foundation for partial support of this work.

Time Optimization in Thin-Layer Chromatography Davld Nurok,* Rose WI. Becker, and Karen A. Sassic Department of Chemistty, Indiana University-Purdue University at Indianapolis, Post Office Box 647, Indianapolis,

A method Is descrlbed for predlctlng the optlmum comblnatlon of solvent path length arid binary solvent composltlon for obtaining a glven separatlon between a pair of compounds In the mlnimum the by contlnuous development thln-layer chromatography. The contlnuoue development of a palr of dyes and a pair of steroids is presented to Illustrate the agreement between predlcted and cexperlmental analysls tlme and spot separation.

Thin-layer chromatog.raphy (TLC) is a very attractive technique for routine analyses. The precision and sensitivity

-..diana 462,

I

attainable with modern scanning densitometers are comparable with those obtained in high-performance liquid chromatography. When working with multiple samples, the analysis time per sample is far lower for TLC than for HPLC. The combination of state-of-the-art techniques in TLC is often referred to as high-performance TLC (HPTLC) and has recently been discussed ‘both in a monograph (I)and in an article (2) in the A pages of Analytical Chemistry. The obvious disadvantage of TLC is that its separating power is in general lower than that of HPLC. This can often be overcome by using multiple development techniques, but these are generally time-consuming. An alternative approach

0003-2700/82/0354-1955$01.25/00 1982 American Chemical Society