Eur. J. Biochem. 77, 529-534 (1977)
Sensitivity of Hemoglobin Thiol Groups within Red Blood Cells of Rat during Oxidation of Glutathione Nechama S. KOSOWER, Edward M. KOSOWER, and Rela L. KOPPEL Departments of Human Genetics and Chemistry, Tel-Aviv University (Received February 11, 1977)
The intracellular thiol-oxidising diazenes, diazenedicarboxylic acid bis-N,N-dimethylamide and diazenedicarboxylic acid bis-N'-ethylpiperazinide, have been used in the study of red cells. A difference in the consequences of diazene oxidant treatment between the human red cell and rat red cell has been found in respect to the quantity of oxidant needed for glutathione (GSH) oxidation, to the fate of GSH, and to the reactivity of hemoglobin. In the first place, significantly more oxidant is needed for GSH oxidation in the rat red cell than in the human cell. Secondly, in the human cell, all of the GSH is converted to glutathione disulfide (GSSG), from which GSH is regenerated. In the rat cell, GSH disappears without being converted to GSSG, and GSH is not regenerated. Thirdly, a decrease in rat hemoglobin thiol groups, but no change in human hemoglobin, is found. Sterically unhindered thiol groups in the rat hemoglobin are thought to react with the usual adduct intermediate in GSH oxidation by diazene (formed from RCON=NCOR GSH +RCON(SG)NHCOR) to produce mixed disulfides, from which GSH is not easily regenerated. The results support the idea that reduction of mixed disulfides of GSH and protein is not carried out directly by GSSG reductase but necessitates thiol transferase and GSH. The thiol-oxidising diazenes may be of use in mapping of exposed, reactive thiol groups in proteins.
+
The intracellular thiol-oxidising agents, diamide [11, diazenedicarboxylic acid bis-N'-methylpiperazi-
nide and diazenedicarboxylic acid bis-N'-ethylpiperazinide (compound I) [2,3], normally convert the major intracellular thiol, glutathione (GSH), to the corresponding disulfide (GSSG) without too much deviation from the expected stoichiometry [Eqn (l)] shown below for compound I [4]. A
CH CH N 3 2
u
N C C N = N C O P N C H ~ C H ~+
u
ZGSH
---
n n CH CH N"CON-NCON' ' N C H ~ C H ~t 3 * u H H I
GSSG
GSH can be regenerated from GSSG after such oxidations in the normal human red cell and in red cells of other species [l,2,5 - 81. Utilizing the discovery that diazenedicarboxylic acid bis-N'-methylpiperazinide penetrated human red blood cells at a modest rate at 1- 4 "C, we developed a procedure for measuring the Abbreviations. Diamide, diazenedicarboxylic acid bis-N,N-dimethylamide ; compound I, diazenedicarboxylic acid bis-"ethylpiperazinide.
transmembrane diffusion coefficients of diazenedicarboxylic acid bis-N'-methylpiperazinide and homologues like compound I [3]. The only information required is the initial rate of intracellular GSH oxidation, the partition coefficient between 2-octanol and water and the size of the red blood cell. In the course of applying this procedure to the red blood cells of many different mammalian species [9], we discovered and report here that rat red blood cells are unique in their behavior towards these reagents. In these cells, certain thiols of the native intracellular hemoglobin are oxidised at about the same rate as is glutathione, thus giving rise to a stoichiometry for the oxidation reaction different from that found with other red blood cells. In fact, the GSH of the rat red cell disappears in this procedure without being converted to GSSG. Furthermore, following complete oxidation, GSH is not regenerated upon incubation in the presence of glucose, suggesting that protein-glutathione mixed disulfides are formed, and that these are not reduced under these conditions within the intact rat erythrocyte. These results reflect what may be an important property of some thiol-containing proteins in the presence of oxidising agents and glutathione.
Rat Red-Cell Thiol Oxidation
530
MATERIALS AND METHODS Blood Samples
Blood was obtained from ether-anesthetized, Charles-River-derived albino rats (by cardiac puncture), and from healthy humans, mixed with heparin, centrifuged and buffy coat removed. The red blood cells were washed twice with NaCl 0.135 M-0.01 M/ phosphate buffer, pH 7.3, and suspended in the same buffer to a hemdtocrit of between 40 60 "/;',.
ether removed by gently passing NZ over the solution. The supernatant was then incubated with GSSG reductase (Sigma) (1 unit/ml incubation mixture), NADPH (Sigma, 1 mM) and phosphate buffer (50 mM, pH 7.3) at 37 "C for 10 min to permit reduction of any GSSG present to GSH. Samples were then analyzed for GSH in the usual way after the addition of metaphosphoric acid to denature the enzyme.
~
Oxidation of G S H by diamide and by Compound 1
Oxidation of red cell GSH was carried out at 1-2 -C. Ice-cold solutions of diamide [(CH3)2NCON = NCON(CH3)zI or ofcompound I (freshly prepared!) made in the buffer mentioned above, were quickly added to 40 - 60 % red cell suspensions, usually of equivalent volume. while the cell suspensions was being agitated on a Vortex mixer. Mixing was continued for a few seconds, following which the mixtures were returned to the ice bath. Aliquots were removed at intervals, immediately added to 20- 20 volumes of ice-cold buffer, the dilute suspension centrifuged, the cells resuspended in buffer and the GSH determined as previously described [lo]. In some experiments with compound I, GSH solution was added to aliquots at intervals in order to arrest its penetration into the cell, while allowing quantitation of the oxidant remaining in the extracellular medium, through titration of the unconsumed GSH. In the case of the diamide, the reaction with GSH can be stopped by rapidly lowering the pH to 1 or less with metaphosphoric acid (0.1 M) and extracting the remaining diamide with dichloromethane, prior to returning the pH to 7.5 for GSH analysis. Lowering the pH does not decrease the rate of reaction of GSH with compound I sufficiently, nor can the doubly charged compound I molecule be extracted into organic solvents. For these reasons, dilution or reduction with GSH were used as described above. Regeneration of GSH from GSSG
Red cell suspensions, after treatment with diamide or compound I, were washed, resuspended in buffer to a hematocrit of 25 - 30 %, and glucose added to a concentration of 10 mM. The cell suspensions were incubated at 37 "C and aliquots taken at intervals for GSH determination. Determination of GSSG
Samples of red cell suspensions, following treatment with diamide or compound I, were deproteinized by the addition of trichloracetic acid (5 % final concentration). After centrifugation, the acid was removed from the supernatant by extraction with ether, and the
Determination of Hemoglobin Thiol Groups
Hemoglobin solution was prepared according to Garrick et al. [ ll] . Samples of red blood cell suspensions, some treated with oxidant and some not, were centrifuged, washed with NaCl/phosphate buffer, pH 7.3, centrifuged, washed with 0.1 M NaC1/0.1 M Tris buffer, pH 9.1, and lysed in 10-20 volumes of 5 mM Tris-HC1buffer, pH 8.4 containing 0.5 mM EDTA and centrifuged at 20000 x g for 15 min. The stroma-free hemolysate was diluted with Tris-HC1 buffer, pH 8.4, to give a final concentration of 5- 10 pM hemoglobin. Hemoglobin concentrations were determined spectrophotometrically as oxyhemoglobin, using an absorption coefficient of 56000 M-' cm-' at 540nm. Bis-(3-carboxy-4-nitrophenyl)disulfide[I21 was added to the stroma-free hemolysate to give a final concentration of 0.1 mM. After several minutes the hemolysate-disulfide reagent mixture was mixed with 1 vol. of ethanol followed by 1 vol. of chloroform. The thiolate anion derived by reduction of the disulfide bond of the reagent was determined spectrophotometrically in the hemoglobin-free upper layer. A full description of this new and simple procedure for hemoglobin - SH groups and its applications will be published separately (N. S. Kosower and R. L. Koppel, unpublished results).
RESULTS Oxidation of GSH in the Intact Red Cell
A comparison of the overall stoichiometry of oxidation of GSH within human red cells to that within rat red cells is made in Fig. 1. Approximately 2.5 times as much oxidant is required for each mole of GSH lost within a rat red cells as is necesary to produce the same loss within a human red cell. In order to test whether or not low membrane permeability to oxidant could be responsible for the different behavior of the rat red blood cell, the rate of intracellular GSH oxidation was compared to the rate of compound I disappearance from the extracellular medium. The rate of entry of compound I into the rat red cell is in fact higher than its rate of entry into the human red blood cell (Fig. 2), while the initial rates of GSH oxidation in the rat and human red cells are comparable in magnitude. The
531
N. S. Kosower, E. M. Kosower, and R. L. Koppel
c
4
1
/ X
-
- 100 c
-
-
-
I
2 3 Oxidant added ( p m o l / m l cells)
4
Fig. 1. A comparison of' the amount of glutathione ( G S H ) oxidised w5ithin red cells by diflerent amounts of oxidunt for human and rut red blood cells. Human red blood cell oxidant, compound I (A), diamide ( x ); rat red blood cell oxidant, compound I (O), diamide (0).Units are pmol/ml red blood cells. Oxidation carried out at 2 "C for 10 min, red cells then washed in cold buffer, resuspended and GSH determined as described in Materials and Methods
0
15
30 45 60 75 Time of incubation (min)
Fig. 3. Regeneration of' glutathione ( G S H ) in human and rat red cellsfollowing oxidation by compound I or diamide. Diazene oxidants were used at amounts of 4.4 pmol/ml red blood cells. GSH concentrations of untreated controls were: human, 2.08 pmol/ml red blood cells, rat, 1.97 pmol/ml red blood cells. The oxidants were added to a red cell suspension maintained at 2 " C , excess oxidant washed off after 10 min, glucose added and the suspensions incubated at 37 "C. Human red cells: (A) compound I; ( x ) diamide; rat red cells: (0)compound I; (0) diamide
Regeneration of GSH in the Rat and Human Red Blood Cells
Time (s)
Time ( s )
Fig. 2. Curves illustrating ( A ) the rates of' intracellular glutathione ( G S H ) oxidation by compound I and ( B ) the rates of disappearance of compound I f r o m the extracellular medium. Times are given in seconds. Oxidant concentrations used were: human, 0.94 pmol compound I/ml cells; rat, 0.91 pmol compound I/ml cells. (A) Human red cells; (0)rat red cells. Oxidation carried out at 2 T; reaction stopped at time intervals by the addition of GSH solution to aliquots of cell suspension. GSH was analysed both in the medium (allowing the titration of oxidant left) and in the cells to determine rates of intracellular GSH oxidation
intracellular oxidation in the rat red cell, however, terminates early in the course of the usual experiment, so that the amount of GSH oxidised under these conditions is far less than that oxidised in the human red blood cell. The total complement of intracellular GSH can, however; be oxidised provided sufficiently large amounts of oxidant are used, i.e. about 2-2.5 mol of oxidant per mol of contained GSH.
The normal human red blood cell can efficiently regenerate GSH from GSSG after complete oxidation by a diazene oxidant [1,2,5]. From 90-100% of the GSH is regenerated within 30 - 60 min after treatment with a diazene oxidant during incubation at 37 "C in the presence of glucose. Full regeneration of GSH after treatment with either diamide or compound I in excess is illustrated in Fig. 3 for the human red cell. The rat red cell behaves quite differently, and relatively little GSH is regenerated under these conditions (0 - 10 for diamide oxidant, 10-20% for compound I oxidant) (Fig. 3).
Titration of GSSG and of Hemoglobin Thiol Groups in Human and Rat Red Cells
The GSH and GSSG contents of human and rat red cells were measured following treatment with excess diazene oxidants but before any incubation procedure designed to regenerate GSH from GSSG. In human red cells, the total GSH content of the cell is found as GSSG in the non-protein fraction of the lysate after complete GSH oxidation. In contrast, none or very little GSSG is found in rat blood cells after complete GSH oxidation, either with diamide or with compound I. These results are summarized in Table 1. The - SH content of hemoglobin from human red cells treated with oxidants even in large excess shows almost no change from that found for hemoglobin of untreated red cells. The - SH content of hemoglobin
532
Rat Red-Cell Thiol Oxidation
Table 1. GSSG in red cells after oxidation of GSH Red cell suspensions were mixed with oxidants at 2 "C and kept for 10- 15 min; cell suspensions then washed, cells resuspended in buffer and aliquots used for determination of GSH and of GSSG. (For details see Materials and Methods). Cell origin
Oxidant
GSH, initial concentration
Oxidant added/GSH
pmol/ml red cells
GSH oxidised
GSH recovered as GSSG
%
Human
diamide diamide compound I
2.74 2.45 2.45
1.20 2.05 2.05
97 100 100
98 95 104
Rat
diamide diamide compound I compound I
1.88 2.10 2.10 2.28
2.08 2.38 2.38 2.20
92 100 100 98
9 6 11 14
Table 2. Hemoglobin - SHgroups in oxidant-treated red cells Red cell suspensions were mixed with solutions of the oxidants at 2 "C for 10- 15 min; after washing, the cells were hemolysed and prepared for the determination of hemoglobin -SH groups. (For details see Materials and Methods). Note that for hemoglobin - SH, GSH thiol groups have been subtracted from the total thiol group determination. Hemoglobin - SH was measured as pmol/pmol hemoglobin. It was assumed that 5 pmol hemoglobin were present in each ml of red blood cells. Variation in the hemoglobin content of the red cells might account for a part of the small variation observed in those experiments to which no oxidant was added Cell origin
Oxidant name
Hemoglobin - SH
Decrease in hemoglobin - SH
amount pmol/ml red cells
Human none diamide 5 compound1 5
30.5 30.5 30.5
0 0
Human none diamide diamide
4.8 8.0
29.25 29.0 28.5
0.25 0.75
5 6
40.0 37.5 35.0
2.5 5.0
none 5 diamide compound1 5
39.5 36.5 34.5
none compound1 5
43.5 39.5
Rat
Rat
Rat
none diamide diamide
3.0 5.0
4.0
from rat red cells diminishes significantly under similar conditions, as the data in Table 2 indicate. The addition of oxidants ( 5 pmol/ml red cells) leads to the oxidation of 2-2.5 pmol of GSH. (The theoretical maximum is 2 I S H groups oxidised per oxidant molecule).
diazene oxidant treatment between the human red cell and the rat red cell with respect to the fate of the GSH and GSSG, to the quantity of oxidant needed for GSH oxidation, and to the reactivity of hemoglobin. With regard to the fate of GSH and GSSG, in the human red cell, all of the GSH is converted to GSSG which can be reconverted to GSH under appropriate conditions within a short time. In the rat red cell, very little of the GSH is converted to GSSG and GSH is not regenerated from the oxidation product under the usual conditions for the regeneration of GSH from GSSG. With regard to the quantity of oxidant, significantly more oxidant is needed for GSH oxidation in rat red cell than in the human cell. With regard to hemoglobin reactivity, a decrease in rat hemoglobin -SH, but no change in human hemoglobin, is observed under these conditions. Diazene oxidants oxidise GSH in two stages, the first leading to an intermediate, and the second giving the final products. The intermediate is an adduct of GSH and the diazene group [Eqn ( 2 ) ] .The second stage involves the attack of a thiol nucleophile on the sulfur of the intermediate. In the usual red cell, the most reactive nucleophile is GSH [Eqn (3)]. (Details of the mechanism are given by Kosower and Kanety-Londner 141). It is evident that in the rat red cell another nucleophile can compete with GSH and the only candidates present in sufficient concentration are the thiol groups of hemoglobin, HbSH. The thiol groups could react with the diazene-SG intermediate [Eqn (4)] or directly with the oxidant [Eqns ( 5 , 6 ) ] . GSH
+ RCON=NCOR+RCON(SG)NHCOR (intermediate) ( 2 )
+ RCON(SG)NHCOR+GSSG + RCONHNHCOR HbSH + RCON(SG)NHCOR+HbSSG GSH
DISCUSSION The experiments described above demonstrate that there is a substantial difference in the consequences of
+RCONHNHCOR HbSH
(3) (4)
+ RCON = NCOR+RCON(SHb)NHCOR (5)
533
N. S. Kosower, E. M. Kosower, and R. L. Koppel
RCON(SHb)NHCOR
+ HbSH-HbSSHb
diazene oxidants : Thioredoxin, a dithiol protein, with ‘highly exposed’ thiol groups [19], is very reactive towards diamide (Holmgren, personal communicaR = (CH3)2N (diamide) or CH~CHZN(CHZCHZ)ZN tion). Formation of protein glutathione mixed di(compound I). These reactions would account for the sulfide has been found by Flohe et al. in isolated liver lack of GSSG, the decrease in hemoglobin SH and the cells treated with diamide [20]. The assembly of microamount of oxidant consumed in the rat red cell. tubules is inhibited in the presence of diamide, and The failure of regeneration within the rat red cell certain - SH groups in tubulin, which are involved in is not surprising in view of the absence of GSSG. the polymerization to microtubules, are especially Flohe and Gunzler have recently discussed the proreactive [21,22]. On the other hand, many other problems of deciding whether mixed disulfides are subject teins have thiol groups which are unreactive towards to direct reduction or thiol-disulfide exchange [13]. diamide, e.g. human normal hemoglobin in the intact Previously, Birchmeier et ul. [14] found that human cell, albumin, and papain [23,24]. It may thus be poshemoglobin glutathione mixed disulfides could not be sible to map exposed, especially reactive, thiol groups reduced by glutathione disulfide reductase. In many in certain proteins by the use of diazene oxidants; it cells, protein glutathione mixed disulfides are probably may also be possible to investigate the functional becleaved by reaction with a thiol transferase, leading to havior of proteins altered by diazene oxidants if the the thiol protein and GSSG, which is then reduced by thiol group is not required specifically for the functioglutathione disulfide reductase as has been definitively nal activity. Variations in the structure of the diazene demonstrated by the careful investigations of Manneroxidant [2S] provide a further tool for probing reacvik [lS, 161. tive protein thiols. Thiol transferase is apparently absent from the Support from the Chief Scientist’s Office, Ministry of Health, rat red cell [17]. Thus, our results are consistent with Government of Israel is acknowledged. A grant-in-aid from Mr the idea that mixed disulfides are not reduced directly, Arthur Blumenfeld was extremely helpful in this and other glutabut require GSH and a thiol transferase [13- 161, both thione work. of which are missing in the rat red cell treated with excess oxidant. Another rodent red cell, that of guinea pig (Cuviu REFERENCES porcellus) behaved similarly to the rat red cell with respect to the quantity of diazene required to oxidise 1. Kosower, N. S., Kosower, E. M., Wertheim, B. & Correa, W. (1969) Biochem. Biophys. Res. Commun. 37, 593- 596. a given quantity of GSH. Other types of animals, like 2. Kosower, E. M., Kosower, N. S., Kenety-Londner, H. & Levy, the cat and the rabbit, had red cells which behaved very L. (1974) Biochem. Biophys. Res. Commun. 59, 347- 351. much like the human red cell. The hemoglobin of the 3. Kosower, N. S . , Kosower, E. M., Saltoun, G. &Levy. L. (1975) cat has an especially high number of reactive thiol Biochem. Biophys. Res. Commun. 62, 98 - 102. 4. Kosower, E. M. & Kenety-Londner, H. (1976) J . Am. Chem. groups (8 - 10/mol as compared with 4/mol for rat SOC.98, 3001 3007. hemoglobin) so its failure to exhibit special reactivity 5. Kosower, N. S., Vanderhoff, G. A. & London, I. M. (1967) towards diazene oxidants is striking. The genetic signifBlood, 29, 313-319. icance of the behavior of rat red cells is not clear, nor 6. Smith, J. E. (1968) J. Lab. Clin. Med. 71, 826. is it known at present whether this special reactivity is 7. Benohr, H. Chr. & Waller, H. D. (1974) in Glutathione (Flohe et al., eds) pp. 184-191, Georg Thieme, Publishers, Stutta property of red cells of rodents other than those gart. cited, as well as of other species. That such a behavior 8. Agar, N. S. & Stephens, T. (1975) Comp. Biochem. Physiol. 52A, may occur in other species is indicated by the study of 605 - 606. Agar et al. [8] who found regeneration of GSH after 9. Kosower, N. S., Kosower, E. M. & Levy, L. (1975) Biochem. Biophys. Res. Commun. 65, 901 -906. diamide treatment of red cells of red kangaroo, but no 10. Kosower, N. S., Song, K. R. &Kosower, E. M. (1969) Biochim. GSH regeneration in red cells of grey kangaroo treated Biophys. Acfa, 192, 1-7. similarly. Preliminary experiments on lysis of rat red 11. Garrick, L. M., Sharma, V. S., McDonald, M. J. & Ranney, cells in the presence of GSH or dithiothreitol did not H. M. (1975) Biochem. J . 149,245-258. appear to change the degree of precipitation usually 12. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82,70-77. 13. Flohe, L. & Giinzler, W. A. (1976) in Glutathione: Metabolism observed at pH 7 111,181 so it is tentatively concluded and Function (Arias, I. M. & Jakoby, W. B., eds) pp. 17-34, that thiol oxidation does not contribute to this particKroc Foundation Series, vol. 6, Raven Press, New York. ular property. 14. Birchmeier, W., Tuchschmid, P. E. & Winterhalter, K. H. Reactivity of protein thiol groups is a function of (1973) Biochemistry, 12, 3667- 3672. 15. Mannervik, B. & Erikson, S. A. (1974) in Glutathione (Flohe their exposure to the external milieu, varying with locaet al., eds) pp. 120- 132, Georg Thieme, Stuttgart. tion in the three-dimensional structure of the molecule. 16. Mannervik, B. & Axelsson, K. (1975) Biochem. J . 149, 785Thus, the thiol groups of rat hemoglobin may be as 788. reactive as the thiol groups of GSH [18]. Some other 17. Jackson, R. C., Harrap, K. R. & Smith, C. A. (1968) Biochem. proteins have been shown to be reactive towards J . 110, 37P.
+ RCONHNHCOR
(6)
~
534
N. S. Kosower, E. M. Kosower, and R. L. Koppel: Rat Red-Cell Thiol Oxidation
18. Chua, C. G., Carrell, R. W. & Howard, B. H. (1975) Biochem. J . 149,259-269. 19. Holmgren, A,, Soderberg, B. O., Eklund, H. & Branden, C. I. (1975) Proc. Natl Acad. Sci. U.S.A. 72, 2305-2309. 20. Flohe, L. (1976) in Glutathione: Metabolism and Function (Arias, I. M. & Jakoby, W. B., eds) p. 156, Kroc Foundation Series, vol. 6, Raven Press, New York. 21. Mellon, M. G. & Rebhun, L. I. (1976) J . Cell Biol. 70,226238.
22. Oliver, J. M., Albertini, D. F. & Berlin, R. D. (1976) J . Cell Biol. 71, 921 -932. 23. Harris, J. W. & Biaglow, J. E. (1972) Biochem. Biophys. Res. Commun. 46,1743- 1749. 24. Kosower, E. M., Correa, W., Kimon, B. J. & Kosower, N. S. (1972) Biochim. Biophys. Acta, 264, 39-44. 25. Kosower, E. M. & Kosower, N. S. (1976) in Glutathione: Metabolism and Function (Arias, I. M. & Jakoby, W. A,, eds) pp. 139-157, Kroc Foundation Series, vol. 6, Raven Press, New York.
N. S. Kosower and R. L. Koppel, Department of Human Genetics, Sackler School of Medicine, Tel-Aviv University, Ramat-Aviv, Tel-Aviv, Israel 61390 E. M. Kosower, Department of Chemistry, Tel-Aviv University, Rdmat-Aviv, Tel-Aviv, Israel 61390
Sensitivity of Hemoglobin Thiol Groups within Red Blood Cells of Rat during Oxidation of Glutathione Nechama S. KOSOWER, Edward M. KOSOWER, and Rela L. KOPPEL Departments of Human Genetics and Chemistry, Tel-Aviv University (Received February 11, 1977)
The intracellular thiol-oxidising diazenes, diazenedicarboxylic acid bis-N,N-dimethylamide and diazenedicarboxylic acid bis-N'-ethylpiperazinide, have been used in the study of red cells. A difference in the consequences of diazene oxidant treatment between the human red cell and rat red cell has been found in respect to the quantity of oxidant needed for glutathione (GSH) oxidation, to the fate of GSH, and to the reactivity of hemoglobin. In the first place, significantly more oxidant is needed for GSH oxidation in the rat red cell than in the human cell. Secondly, in the human cell, all of the GSH is converted to glutathione disulfide (GSSG), from which GSH is regenerated. In the rat cell, GSH disappears without being converted to GSSG, and GSH is not regenerated. Thirdly, a decrease in rat hemoglobin thiol groups, but no change in human hemoglobin, is found. Sterically unhindered thiol groups in the rat hemoglobin are thought to react with the usual adduct intermediate in GSH oxidation by diazene (formed from RCON=NCOR GSH +RCON(SG)NHCOR) to produce mixed disulfides, from which GSH is not easily regenerated. The results support the idea that reduction of mixed disulfides of GSH and protein is not carried out directly by GSSG reductase but necessitates thiol transferase and GSH. The thiol-oxidising diazenes may be of use in mapping of exposed, reactive thiol groups in proteins.
+
The intracellular thiol-oxidising agents, diamide [11, diazenedicarboxylic acid bis-N'-methylpiperazi-
nide and diazenedicarboxylic acid bis-N'-ethylpiperazinide (compound I) [2,3], normally convert the major intracellular thiol, glutathione (GSH), to the corresponding disulfide (GSSG) without too much deviation from the expected stoichiometry [Eqn (l)] shown below for compound I [4]. A
CH CH N 3 2
u
N C C N = N C O P N C H ~ C H ~+
u
ZGSH
---
n n CH CH N"CON-NCON' ' N C H ~ C H ~t 3 * u H H I
GSSG
GSH can be regenerated from GSSG after such oxidations in the normal human red cell and in red cells of other species [l,2,5 - 81. Utilizing the discovery that diazenedicarboxylic acid bis-N'-methylpiperazinide penetrated human red blood cells at a modest rate at 1- 4 "C, we developed a procedure for measuring the Abbreviations. Diamide, diazenedicarboxylic acid bis-N,N-dimethylamide ; compound I, diazenedicarboxylic acid bis-"ethylpiperazinide.
transmembrane diffusion coefficients of diazenedicarboxylic acid bis-N'-methylpiperazinide and homologues like compound I [3]. The only information required is the initial rate of intracellular GSH oxidation, the partition coefficient between 2-octanol and water and the size of the red blood cell. In the course of applying this procedure to the red blood cells of many different mammalian species [9], we discovered and report here that rat red blood cells are unique in their behavior towards these reagents. In these cells, certain thiols of the native intracellular hemoglobin are oxidised at about the same rate as is glutathione, thus giving rise to a stoichiometry for the oxidation reaction different from that found with other red blood cells. In fact, the GSH of the rat red cell disappears in this procedure without being converted to GSSG. Furthermore, following complete oxidation, GSH is not regenerated upon incubation in the presence of glucose, suggesting that protein-glutathione mixed disulfides are formed, and that these are not reduced under these conditions within the intact rat erythrocyte. These results reflect what may be an important property of some thiol-containing proteins in the presence of oxidising agents and glutathione.
Rat Red-Cell Thiol Oxidation
530
MATERIALS AND METHODS Blood Samples
Blood was obtained from ether-anesthetized, Charles-River-derived albino rats (by cardiac puncture), and from healthy humans, mixed with heparin, centrifuged and buffy coat removed. The red blood cells were washed twice with NaCl 0.135 M-0.01 M/ phosphate buffer, pH 7.3, and suspended in the same buffer to a hemdtocrit of between 40 60 "/;',.
ether removed by gently passing NZ over the solution. The supernatant was then incubated with GSSG reductase (Sigma) (1 unit/ml incubation mixture), NADPH (Sigma, 1 mM) and phosphate buffer (50 mM, pH 7.3) at 37 "C for 10 min to permit reduction of any GSSG present to GSH. Samples were then analyzed for GSH in the usual way after the addition of metaphosphoric acid to denature the enzyme.
~
Oxidation of G S H by diamide and by Compound 1
Oxidation of red cell GSH was carried out at 1-2 -C. Ice-cold solutions of diamide [(CH3)2NCON = NCON(CH3)zI or ofcompound I (freshly prepared!) made in the buffer mentioned above, were quickly added to 40 - 60 % red cell suspensions, usually of equivalent volume. while the cell suspensions was being agitated on a Vortex mixer. Mixing was continued for a few seconds, following which the mixtures were returned to the ice bath. Aliquots were removed at intervals, immediately added to 20- 20 volumes of ice-cold buffer, the dilute suspension centrifuged, the cells resuspended in buffer and the GSH determined as previously described [lo]. In some experiments with compound I, GSH solution was added to aliquots at intervals in order to arrest its penetration into the cell, while allowing quantitation of the oxidant remaining in the extracellular medium, through titration of the unconsumed GSH. In the case of the diamide, the reaction with GSH can be stopped by rapidly lowering the pH to 1 or less with metaphosphoric acid (0.1 M) and extracting the remaining diamide with dichloromethane, prior to returning the pH to 7.5 for GSH analysis. Lowering the pH does not decrease the rate of reaction of GSH with compound I sufficiently, nor can the doubly charged compound I molecule be extracted into organic solvents. For these reasons, dilution or reduction with GSH were used as described above. Regeneration of GSH from GSSG
Red cell suspensions, after treatment with diamide or compound I, were washed, resuspended in buffer to a hematocrit of 25 - 30 %, and glucose added to a concentration of 10 mM. The cell suspensions were incubated at 37 "C and aliquots taken at intervals for GSH determination. Determination of GSSG
Samples of red cell suspensions, following treatment with diamide or compound I, were deproteinized by the addition of trichloracetic acid (5 % final concentration). After centrifugation, the acid was removed from the supernatant by extraction with ether, and the
Determination of Hemoglobin Thiol Groups
Hemoglobin solution was prepared according to Garrick et al. [ ll] . Samples of red blood cell suspensions, some treated with oxidant and some not, were centrifuged, washed with NaCl/phosphate buffer, pH 7.3, centrifuged, washed with 0.1 M NaC1/0.1 M Tris buffer, pH 9.1, and lysed in 10-20 volumes of 5 mM Tris-HC1buffer, pH 8.4 containing 0.5 mM EDTA and centrifuged at 20000 x g for 15 min. The stroma-free hemolysate was diluted with Tris-HC1 buffer, pH 8.4, to give a final concentration of 5- 10 pM hemoglobin. Hemoglobin concentrations were determined spectrophotometrically as oxyhemoglobin, using an absorption coefficient of 56000 M-' cm-' at 540nm. Bis-(3-carboxy-4-nitrophenyl)disulfide[I21 was added to the stroma-free hemolysate to give a final concentration of 0.1 mM. After several minutes the hemolysate-disulfide reagent mixture was mixed with 1 vol. of ethanol followed by 1 vol. of chloroform. The thiolate anion derived by reduction of the disulfide bond of the reagent was determined spectrophotometrically in the hemoglobin-free upper layer. A full description of this new and simple procedure for hemoglobin - SH groups and its applications will be published separately (N. S. Kosower and R. L. Koppel, unpublished results).
RESULTS Oxidation of GSH in the Intact Red Cell
A comparison of the overall stoichiometry of oxidation of GSH within human red cells to that within rat red cells is made in Fig. 1. Approximately 2.5 times as much oxidant is required for each mole of GSH lost within a rat red cells as is necesary to produce the same loss within a human red cell. In order to test whether or not low membrane permeability to oxidant could be responsible for the different behavior of the rat red blood cell, the rate of intracellular GSH oxidation was compared to the rate of compound I disappearance from the extracellular medium. The rate of entry of compound I into the rat red cell is in fact higher than its rate of entry into the human red blood cell (Fig. 2), while the initial rates of GSH oxidation in the rat and human red cells are comparable in magnitude. The
531
N. S. Kosower, E. M. Kosower, and R. L. Koppel
c
4
1
/ X
-
- 100 c
-
-
-
I
2 3 Oxidant added ( p m o l / m l cells)
4
Fig. 1. A comparison of' the amount of glutathione ( G S H ) oxidised w5ithin red cells by diflerent amounts of oxidunt for human and rut red blood cells. Human red blood cell oxidant, compound I (A), diamide ( x ); rat red blood cell oxidant, compound I (O), diamide (0).Units are pmol/ml red blood cells. Oxidation carried out at 2 "C for 10 min, red cells then washed in cold buffer, resuspended and GSH determined as described in Materials and Methods
0
15
30 45 60 75 Time of incubation (min)
Fig. 3. Regeneration of' glutathione ( G S H ) in human and rat red cellsfollowing oxidation by compound I or diamide. Diazene oxidants were used at amounts of 4.4 pmol/ml red blood cells. GSH concentrations of untreated controls were: human, 2.08 pmol/ml red blood cells, rat, 1.97 pmol/ml red blood cells. The oxidants were added to a red cell suspension maintained at 2 " C , excess oxidant washed off after 10 min, glucose added and the suspensions incubated at 37 "C. Human red cells: (A) compound I; ( x ) diamide; rat red cells: (0)compound I; (0) diamide
Regeneration of GSH in the Rat and Human Red Blood Cells
Time (s)
Time ( s )
Fig. 2. Curves illustrating ( A ) the rates of' intracellular glutathione ( G S H ) oxidation by compound I and ( B ) the rates of disappearance of compound I f r o m the extracellular medium. Times are given in seconds. Oxidant concentrations used were: human, 0.94 pmol compound I/ml cells; rat, 0.91 pmol compound I/ml cells. (A) Human red cells; (0)rat red cells. Oxidation carried out at 2 T; reaction stopped at time intervals by the addition of GSH solution to aliquots of cell suspension. GSH was analysed both in the medium (allowing the titration of oxidant left) and in the cells to determine rates of intracellular GSH oxidation
intracellular oxidation in the rat red cell, however, terminates early in the course of the usual experiment, so that the amount of GSH oxidised under these conditions is far less than that oxidised in the human red blood cell. The total complement of intracellular GSH can, however; be oxidised provided sufficiently large amounts of oxidant are used, i.e. about 2-2.5 mol of oxidant per mol of contained GSH.
The normal human red blood cell can efficiently regenerate GSH from GSSG after complete oxidation by a diazene oxidant [1,2,5]. From 90-100% of the GSH is regenerated within 30 - 60 min after treatment with a diazene oxidant during incubation at 37 "C in the presence of glucose. Full regeneration of GSH after treatment with either diamide or compound I in excess is illustrated in Fig. 3 for the human red cell. The rat red cell behaves quite differently, and relatively little GSH is regenerated under these conditions (0 - 10 for diamide oxidant, 10-20% for compound I oxidant) (Fig. 3).
Titration of GSSG and of Hemoglobin Thiol Groups in Human and Rat Red Cells
The GSH and GSSG contents of human and rat red cells were measured following treatment with excess diazene oxidants but before any incubation procedure designed to regenerate GSH from GSSG. In human red cells, the total GSH content of the cell is found as GSSG in the non-protein fraction of the lysate after complete GSH oxidation. In contrast, none or very little GSSG is found in rat blood cells after complete GSH oxidation, either with diamide or with compound I. These results are summarized in Table 1. The - SH content of hemoglobin from human red cells treated with oxidants even in large excess shows almost no change from that found for hemoglobin of untreated red cells. The - SH content of hemoglobin
532
Rat Red-Cell Thiol Oxidation
Table 1. GSSG in red cells after oxidation of GSH Red cell suspensions were mixed with oxidants at 2 "C and kept for 10- 15 min; cell suspensions then washed, cells resuspended in buffer and aliquots used for determination of GSH and of GSSG. (For details see Materials and Methods). Cell origin
Oxidant
GSH, initial concentration
Oxidant added/GSH
pmol/ml red cells
GSH oxidised
GSH recovered as GSSG
%
Human
diamide diamide compound I
2.74 2.45 2.45
1.20 2.05 2.05
97 100 100
98 95 104
Rat
diamide diamide compound I compound I
1.88 2.10 2.10 2.28
2.08 2.38 2.38 2.20
92 100 100 98
9 6 11 14
Table 2. Hemoglobin - SHgroups in oxidant-treated red cells Red cell suspensions were mixed with solutions of the oxidants at 2 "C for 10- 15 min; after washing, the cells were hemolysed and prepared for the determination of hemoglobin -SH groups. (For details see Materials and Methods). Note that for hemoglobin - SH, GSH thiol groups have been subtracted from the total thiol group determination. Hemoglobin - SH was measured as pmol/pmol hemoglobin. It was assumed that 5 pmol hemoglobin were present in each ml of red blood cells. Variation in the hemoglobin content of the red cells might account for a part of the small variation observed in those experiments to which no oxidant was added Cell origin
Oxidant name
Hemoglobin - SH
Decrease in hemoglobin - SH
amount pmol/ml red cells
Human none diamide 5 compound1 5
30.5 30.5 30.5
0 0
Human none diamide diamide
4.8 8.0
29.25 29.0 28.5
0.25 0.75
5 6
40.0 37.5 35.0
2.5 5.0
none 5 diamide compound1 5
39.5 36.5 34.5
none compound1 5
43.5 39.5
Rat
Rat
Rat
none diamide diamide
3.0 5.0
4.0
from rat red cells diminishes significantly under similar conditions, as the data in Table 2 indicate. The addition of oxidants ( 5 pmol/ml red cells) leads to the oxidation of 2-2.5 pmol of GSH. (The theoretical maximum is 2 I S H groups oxidised per oxidant molecule).
diazene oxidant treatment between the human red cell and the rat red cell with respect to the fate of the GSH and GSSG, to the quantity of oxidant needed for GSH oxidation, and to the reactivity of hemoglobin. With regard to the fate of GSH and GSSG, in the human red cell, all of the GSH is converted to GSSG which can be reconverted to GSH under appropriate conditions within a short time. In the rat red cell, very little of the GSH is converted to GSSG and GSH is not regenerated from the oxidation product under the usual conditions for the regeneration of GSH from GSSG. With regard to the quantity of oxidant, significantly more oxidant is needed for GSH oxidation in rat red cell than in the human cell. With regard to hemoglobin reactivity, a decrease in rat hemoglobin -SH, but no change in human hemoglobin, is observed under these conditions. Diazene oxidants oxidise GSH in two stages, the first leading to an intermediate, and the second giving the final products. The intermediate is an adduct of GSH and the diazene group [Eqn ( 2 ) ] .The second stage involves the attack of a thiol nucleophile on the sulfur of the intermediate. In the usual red cell, the most reactive nucleophile is GSH [Eqn (3)]. (Details of the mechanism are given by Kosower and Kanety-Londner 141). It is evident that in the rat red cell another nucleophile can compete with GSH and the only candidates present in sufficient concentration are the thiol groups of hemoglobin, HbSH. The thiol groups could react with the diazene-SG intermediate [Eqn (4)] or directly with the oxidant [Eqns ( 5 , 6 ) ] . GSH
+ RCON=NCOR+RCON(SG)NHCOR (intermediate) ( 2 )
+ RCON(SG)NHCOR+GSSG + RCONHNHCOR HbSH + RCON(SG)NHCOR+HbSSG GSH
DISCUSSION The experiments described above demonstrate that there is a substantial difference in the consequences of
+RCONHNHCOR HbSH
(3) (4)
+ RCON = NCOR+RCON(SHb)NHCOR (5)
533
N. S. Kosower, E. M. Kosower, and R. L. Koppel
RCON(SHb)NHCOR
+ HbSH-HbSSHb
diazene oxidants : Thioredoxin, a dithiol protein, with ‘highly exposed’ thiol groups [19], is very reactive towards diamide (Holmgren, personal communicaR = (CH3)2N (diamide) or CH~CHZN(CHZCHZ)ZN tion). Formation of protein glutathione mixed di(compound I). These reactions would account for the sulfide has been found by Flohe et al. in isolated liver lack of GSSG, the decrease in hemoglobin SH and the cells treated with diamide [20]. The assembly of microamount of oxidant consumed in the rat red cell. tubules is inhibited in the presence of diamide, and The failure of regeneration within the rat red cell certain - SH groups in tubulin, which are involved in is not surprising in view of the absence of GSSG. the polymerization to microtubules, are especially Flohe and Gunzler have recently discussed the proreactive [21,22]. On the other hand, many other problems of deciding whether mixed disulfides are subject teins have thiol groups which are unreactive towards to direct reduction or thiol-disulfide exchange [13]. diamide, e.g. human normal hemoglobin in the intact Previously, Birchmeier et ul. [14] found that human cell, albumin, and papain [23,24]. It may thus be poshemoglobin glutathione mixed disulfides could not be sible to map exposed, especially reactive, thiol groups reduced by glutathione disulfide reductase. In many in certain proteins by the use of diazene oxidants; it cells, protein glutathione mixed disulfides are probably may also be possible to investigate the functional becleaved by reaction with a thiol transferase, leading to havior of proteins altered by diazene oxidants if the the thiol protein and GSSG, which is then reduced by thiol group is not required specifically for the functioglutathione disulfide reductase as has been definitively nal activity. Variations in the structure of the diazene demonstrated by the careful investigations of Manneroxidant [2S] provide a further tool for probing reacvik [lS, 161. tive protein thiols. Thiol transferase is apparently absent from the Support from the Chief Scientist’s Office, Ministry of Health, rat red cell [17]. Thus, our results are consistent with Government of Israel is acknowledged. A grant-in-aid from Mr the idea that mixed disulfides are not reduced directly, Arthur Blumenfeld was extremely helpful in this and other glutabut require GSH and a thiol transferase [13- 161, both thione work. of which are missing in the rat red cell treated with excess oxidant. Another rodent red cell, that of guinea pig (Cuviu REFERENCES porcellus) behaved similarly to the rat red cell with respect to the quantity of diazene required to oxidise 1. Kosower, N. S., Kosower, E. M., Wertheim, B. & Correa, W. (1969) Biochem. Biophys. Res. Commun. 37, 593- 596. a given quantity of GSH. Other types of animals, like 2. Kosower, E. M., Kosower, N. S., Kenety-Londner, H. & Levy, the cat and the rabbit, had red cells which behaved very L. (1974) Biochem. Biophys. Res. Commun. 59, 347- 351. much like the human red cell. The hemoglobin of the 3. Kosower, N. S . , Kosower, E. M., Saltoun, G. &Levy. L. (1975) cat has an especially high number of reactive thiol Biochem. Biophys. Res. Commun. 62, 98 - 102. 4. Kosower, E. M. & Kenety-Londner, H. (1976) J . Am. Chem. groups (8 - 10/mol as compared with 4/mol for rat SOC.98, 3001 3007. hemoglobin) so its failure to exhibit special reactivity 5. Kosower, N. S., Vanderhoff, G. A. & London, I. M. (1967) towards diazene oxidants is striking. The genetic signifBlood, 29, 313-319. icance of the behavior of rat red cells is not clear, nor 6. Smith, J. E. (1968) J. Lab. Clin. Med. 71, 826. is it known at present whether this special reactivity is 7. Benohr, H. Chr. & Waller, H. D. (1974) in Glutathione (Flohe et al., eds) pp. 184-191, Georg Thieme, Publishers, Stutta property of red cells of rodents other than those gart. cited, as well as of other species. That such a behavior 8. Agar, N. S. & Stephens, T. (1975) Comp. Biochem. Physiol. 52A, may occur in other species is indicated by the study of 605 - 606. Agar et al. [8] who found regeneration of GSH after 9. Kosower, N. S., Kosower, E. M. & Levy, L. (1975) Biochem. Biophys. Res. Commun. 65, 901 -906. diamide treatment of red cells of red kangaroo, but no 10. Kosower, N. S., Song, K. R. &Kosower, E. M. (1969) Biochim. GSH regeneration in red cells of grey kangaroo treated Biophys. Acfa, 192, 1-7. similarly. Preliminary experiments on lysis of rat red 11. Garrick, L. M., Sharma, V. S., McDonald, M. J. & Ranney, cells in the presence of GSH or dithiothreitol did not H. M. (1975) Biochem. J . 149,245-258. appear to change the degree of precipitation usually 12. Ellman, G. L. (1959) Arch. Biochem. Biophys. 82,70-77. 13. Flohe, L. & Giinzler, W. A. (1976) in Glutathione: Metabolism observed at pH 7 111,181 so it is tentatively concluded and Function (Arias, I. M. & Jakoby, W. B., eds) pp. 17-34, that thiol oxidation does not contribute to this particKroc Foundation Series, vol. 6, Raven Press, New York. ular property. 14. Birchmeier, W., Tuchschmid, P. E. & Winterhalter, K. H. Reactivity of protein thiol groups is a function of (1973) Biochemistry, 12, 3667- 3672. 15. Mannervik, B. & Erikson, S. A. (1974) in Glutathione (Flohe their exposure to the external milieu, varying with locaet al., eds) pp. 120- 132, Georg Thieme, Stuttgart. tion in the three-dimensional structure of the molecule. 16. Mannervik, B. & Axelsson, K. (1975) Biochem. J . 149, 785Thus, the thiol groups of rat hemoglobin may be as 788. reactive as the thiol groups of GSH [18]. Some other 17. Jackson, R. C., Harrap, K. R. & Smith, C. A. (1968) Biochem. proteins have been shown to be reactive towards J . 110, 37P.
+ RCONHNHCOR
(6)
~
534
N. S. Kosower, E. M. Kosower, and R. L. Koppel: Rat Red-Cell Thiol Oxidation
18. Chua, C. G., Carrell, R. W. & Howard, B. H. (1975) Biochem. J . 149,259-269. 19. Holmgren, A,, Soderberg, B. O., Eklund, H. & Branden, C. I. (1975) Proc. Natl Acad. Sci. U.S.A. 72, 2305-2309. 20. Flohe, L. (1976) in Glutathione: Metabolism and Function (Arias, I. M. & Jakoby, W. B., eds) p. 156, Kroc Foundation Series, vol. 6, Raven Press, New York. 21. Mellon, M. G. & Rebhun, L. I. (1976) J . Cell Biol. 70,226238.
22. Oliver, J. M., Albertini, D. F. & Berlin, R. D. (1976) J . Cell Biol. 71, 921 -932. 23. Harris, J. W. & Biaglow, J. E. (1972) Biochem. Biophys. Res. Commun. 46,1743- 1749. 24. Kosower, E. M., Correa, W., Kimon, B. J. & Kosower, N. S. (1972) Biochim. Biophys. Acta, 264, 39-44. 25. Kosower, E. M. & Kosower, N. S. (1976) in Glutathione: Metabolism and Function (Arias, I. M. & Jakoby, W. A,, eds) pp. 139-157, Kroc Foundation Series, vol. 6, Raven Press, New York.
N. S. Kosower and R. L. Koppel, Department of Human Genetics, Sackler School of Medicine, Tel-Aviv University, Ramat-Aviv, Tel-Aviv, Israel 61390 E. M. Kosower, Department of Chemistry, Tel-Aviv University, Rdmat-Aviv, Tel-Aviv, Israel 61390
