Communication THEJOURNAL OF BIOLOGICN.

CHEMISTRY
Vol. 255, No. 20, Issue of October 25, pp. 9530-9533. 1980
Printed in U.S.A.

Dynamic State of Glutathione in (2-6). Cells that have y-glutamyl transpeptidase (e.g. renal
cells) translocate glutathione to the membrane-bound en-
Blood Plasma* zyme, whereas cells that have relatively little transpeptidase
(Received for publication, June 11, 1980,and in revised form, activity (e.g. liver cells) translocate glutathione to the blood
August 6, 1980) plasma. Substantial amounts of glutathione are translocated
Mary E. Anderson and Alton Meister from the liver and probably other tissues to theblood plasma;
the glutathione levels of plasma remain low because plasma
From the Departmentof Biochemistry, Cornell
University Medical College, New York, New York 10021
glutathione is utilized by renal and extrarenal y-glutamyl
transpeptidase and possibly also by other metabolic mecha-
nisms (7, 8). These considerations, which indicate that the
Recent studies have shown that there is an inter-
organ cycle of glutathione metabolism in which gluta- glutathione of blood plasma turns over very rapidly, empha-
thione is translocated from certaincells into the blood size the importance of knowing more about the level and
plasma, and that plasma glutathione is utilized bycells oxidation-reduction state of plasma glutathione.
that have y-glutamyl transpeptidase. The present stud- In previous studies in this laboratory, in which it was shown
ies indicatethatthere is asignificantintravascular that translocation of intracellular glutathione is a discrete step
phase of glutathionemetabolism. The level of total in the y-glutamyl cycle(2), it was found that thelevel of total
glutathione (GSH +GSSG) in rat blood plasma was glutathione of the blood plasma of the mouse is 24.0 k 1.3
found to be 22 to 27 PM GSH equivalents, as determined p. In contrast, other investigators (8-10) reported that the
by the glutathione reductase recycling method. About level of total glutathione in rat blood plasma is in the range 3
85% of the total is in the form of GSH. These findings to 6 p ~ The . present work,whichbegan as an effort to
contrast with previous reportsof total levels of 3 to 6 investigate this apparent species difference, has shown that
p~ and 50 to 75% GSSG. We found that plasma allowed the level of total glutathione in rat blood plasma is 22 to 27
to stand at 23°C for 30 to 60 min has total glutathione p~ or possiblysomewhat higher, and that most of the total is
levels of 4 to 7 p ~ most
, (95%) of which is GSSG; after in the form of GSH rather thanGSSG, as previously reported.
treatment of this plasma (following deproteinization) The present studies, which elucidate the level and oxidation-
with KB-, levels of21 to 24 PM werefound. GSH reduction state of plasma glutathione, indicate that there is,
disappears rapidly from plasma, whereas GSSG disap- in addition to inter-organ translocation of glutathione (7), a
pears very slowly. y-Glutamyl transpeptidase does notsignificant intravascular phase of glutathione metabolism.
account for the loss of plasma GSH, nor does binding
to proteins account for more than a small fraction of EXPERIMENTALPROCEDURES
the GSH that disappears. Most of the GSH that disap-
pears can be found in the deproteinized samples after Materials-Glutathione was obtained from Sigma. Glutathione
treatment with KB-. The findings arein accord with of disulfide was prepared by passing oxygen through a solution (pH 7)
GSH a t 0°C. [U-GLu’*C]glutathionedisulfide was obtained from
the view that glutathioneis translocated to plasma in New England Nuclear. Potassium borohydride was obtained from
the form of GSH and that such GSH constitutes the Ventron. [as, 5S]cu-Amino-3-chloro-4,5-dihydro-5-isoxazoleacetic
major source of plasma thiol. The intravascular phase acid (AT-125)was obtained through the courtesy of Dr. L. J. Hanka
ofGSH metabolism seems to involve reduction of di- of Upjohn. t-y-Glutamyl-(0-carboxy)phenyhydrazidewas obtained
sulfide bonds of plasma constituents and mobilization as described (2). Male Sprague-Dawley rats (250 to 450 g) were
of compounds bound by disulfide linkage to plasma obtained from Charles River. [’4C]Glutathione disulfide was reduced
proteins to form GSSG andlow molecular weight deriv-by treating it with a &fold molar excess of KBH4 inthe presence of
atives of glutathione suchas disulfides. 1 m~ EDTA for 1 h at 23’C. After reduction, the solution was
acidified at pH 3 to 4.
Methods-Blood was obtained from rats after decapitation; under
these conditions, mainly arterial blood is obtained. Studies of samples
Glutathione occurs predominantly intracellularly in concen- of blood obtained directly from the heart and from the aorta gave
results that were similar to those obtained from blood samples ob-
trations that range from about 0.5 mM to about 10 m ~more ; tained after decapitation. The blood samples (about 10 ml) were
than 95% of intracellular glutathione is in the form of GSH rapidly collected in a beaker containing 0.1 ml of 500 m~ EDTA. The
(l).’On the other hand, the extracellular level of glutathione plasma was obtained after centrifugation of the blood for 1.5 min at
is in the micromolar range. Perhaps for this reason, relatively 10,OOO X g in a Beckman MicrofugeB. The plasma was rapidly
less attention has been given to the status of glutathione in deproteinized by mixing it vigorously with 0.5 volume of 10% 5-
blood plasma than to its level invarious tissues. Recent studies sulfosalicylic acid, followed by centrifugation at 10,OOO X g for 2 min.
Hemolysis was not detected by procedures capable of revealing 0.2%
indicate that thetranslocation of glutathione acrow cell mem- hemolysis (11, 12). Glutathione was determined by the glutathione
branes is a highly significant aspect of glutathione metabolism reductase-DTNB recycling procedure (lo),which measures both GSH
and GSSG. The modification of this procedure described by Grifflth
* This research was supported in partby grants from the National (13) in which 2-vinylpyridine is used to mask the sulfhydryl group of
Institutes of Health, the United States Public Health Service, and GSH was also used. In these studies, 0.10 ml of sample was mixed
the National Science Foundation. The costs of publication of this with 0.002 ml of 2-vinylpyridine and 0.005 ml of triethanolamine (final
article were defrayed in part by the payment of page charges. This pH 6 to 8) and allowed to stand at23°C for 1 h. Analyses were then
article must therefore be hereby marked “aduertisement” in accord- carried out by the glutathione reductase-DTNB recycling procedure
ance with 18 U.S.C. Section 1734 solely to indicate this fact. (10); in some cases, the samples were brought to pH 3 and chromat-
’ The term ?h GSSG denotes the concentration of glutathione ographed on an automated Dionex amino acid analyzer. Borohydride
disulfide in GSH equivalents. In this paper, “total glutathione” is treatment of samples was performed as follows. The sample (0.2 ml)
defined as GSH + GSSG and is expressed as GSH equivalents. The was treated with 5.4 mg of KBH, in the presence of 5 mM EDTA a t
abbreviation used is: DTNB, 5,5’-dithiobis(2-nitrobenzoic acid). pH 7. After standing for 1 h at 23”C, 0.015 ml of 50% 5-sulfosalicylic

9530

This is an Open Access article under the CC BY license.

 Dynamic State of Glutathione in Blood Plasma 9531
acid (or 0.02 ml of 6 M HC1) was added, andthe mixture was allowed
to stand for 20 min. Analysis for glutathione wasthen carried out by
the glutathione reductase-DTNB recycling method(10) using appro-
priatestandards. Some sampleswere also analyzedbyautomated
chromatography on a Dionex analyzer after treatment of the samples
withborohydrideand EDTA as described above (or with 33 m~
dithiothreitol and 5 m~ EDTA for 1 h at 23OC at pH 7) followed by
treatment with 2-vinylpyridineas described above.

RESULTS

The values of total plasma glutathione (GSH + GSSG)

obtained 3 to 6 min after bleeding were between 23 and 27
PM GSH equivalents. When the plasma was allowed to stand
at 23°C for short periods, much lower values were obtained. x) 40 60 80 loo Ix) 180240360
Thus, as shown in Fig.1, plasma analyzed 20 min after MINUTES
bleeding had about 13 p~ total glutathione. The datagiven in FIG. 1. Determinations of total glutathione in rat blood
Fig. 1 suggest that thelevel of total glutathione present at the plasma. The experimental details are given in the text.
time of bleeding is probably higher than 30 phf. Analyses by
automated column chromatography gave values for initial (3
TABLEI
to 5 min) total glutathione levels that agree closely with those
obtained by the enzymatic procedure shown in Fig. 1.2 Table Level and oxidation-reduction state of rat bloodplasma
glutathione
I summarizes data on the oxidation-reduction state of plasma
Glutathione
glutathione. The glutathione found in plasma analyzed 3 min
Time after GSH
after bleeding was about 85% in the form of GSH, whereas bleeding
GSSG"
plasma analyzed 60 min after bleeding was about 95% in the
form of GSSG. min w
Studies were carried out in which GSH and GSSG were 3 24.8 3.821.0 85 15
separately added to plasma that had previously been allowed 60 7.5 0.4 5 7.1 95
to stand at 23°C for 30 min after bleeding. Fig. 2 shows that Expressed as GSH equivalents.
GSSG added to such plasma disappears slowly, ie. at a rate
of about 0.08 to 0.10 p ~ / m i n On
. the other hand, when GSH
was added (Fig. 3 B ) , it disappeared rapidly at a rate compa-
rable to that found with freshly obtained plasma (Fig. 1);
subsequently, there was a much slower rate of disappearance,
similar to that seen when GSSG was added to the plasma
(Fig. 2). Analysis of the glutathione found 90 min after addi-
tion of GSH in the experiments described in Fig. 3B by the 2-
vinylpyridine method indicated that about 95% of the gluta-
thione present is GSSG (Fig. 3, open circle).Thus, aboutone-
half of the GSH added was converted to GSSG and the
remainder disappeared as determined by the glutathione re-
ductase-DTNB recycling method. The glutathione added in 20 40 60 80 1 0 0 I 2 0 140 160 1 8 0
the experiments described in the legend to Fig. 3 was labeled MINUTES
with "C in the glutamyl moiety, making it possible to estimate FIG. 2. Addition of GSSG to rat blood plasma. The plasma
the amount of glutathione that might be bound to theproteins wasallowed to stand at 23°C for 30 min, at which time 17 PM
precipitated by addition of sulfosalicylic acid.As indicated in glutathione equivalents of GSSG wereadded.Determinations of
curve C (Fig. 3 ) , the radioactivity found in the supernatant glutathione and two further additions of GSSG were made as indi-
solution after removal of precipitated proteins declined only cated.
slightly. Thus, between the 30- and 60-min points (Fig. 3 ) , the
glutathione level decreased by about 5036, whereas the loss of itated proteins, and cf, some of the GSH that disappears is
radioactivity during this period was equivalent to about 10% converted to GSSG, and most of the remainder is converted
of the labeled glutathione found initially (i.e. at the 30-min to products which are not detected as GSH or GSSG by the
point). glutathione reductase-DTNB recycling procedure. The exper-
The findings described above indicate that (a)the level of iments described below wereundertaken in an effort to explore
total glutathione in the plasma is about 25 PM or somewhat the disappearance of GSH.
higher, ( b ) at least 85% of the total is GSH, ( c ) GSH disap- Since y-glutamyl transpeptidase catalyzes a major pathway
pears rapidly from plasma, ( d ) GSSG disappears relatively of glutathione utilization, we examined the plasma for this
slowly, (e) a small amount of glutathione binds to theprecip- enzyme activity but could not detect it under conditions in
'The possibility that hemolysis contributessigrufkantly to the which an activity of 1 nmol/min/ml of plasma could have
values found for plasma glutathione was carefully considered. How- been detected. Furthermore, we found that addition to the
ever, as stated above, less than 0.2%hemolysis was found; about l to plasma of transpeptidase inhibitors (L-y-glutamyl-(0-car-
1.5% hemolysis wouldberequired to accountfor the glutathione boxy)phenylhydrazide (2.4 m ~ (2) ) and AT-125 (4 m ~ (14, )
levels observed.Furthermore, other studies in this laboratoryin 15)) did not affect the rateof disappearance of glutathione. It
which blood plasma was obtained and prepared for analysis in the may therefore be concluded that theobserved disappearance
same manner BS done here have shown that much lower levels of
of glutathione is not mediated by y-glutamyl transpeptidase.
plasma glutathione (2 to 8 PM) are consistently found after adminis-
tration of an inhibitorof glutathione synthesis (2) and whenthe blood In an experiment similar to that described in the legend to
is obtained fromthe renal vein (M. E. Anderson,R. J. Bridges and A. Fig. 3, [14C]glutathione (918cpm/nmol) was added to plasma
Meister, unpublished observations). (previously placed at 23°C for1h) to give a final concentration
9532 Dynamic State of Glutathione in Blood Plasma
TABLEI1
Effectof borohydride treatment onplasma glutathioneLevels
Experiment y2de Glutathione”
original After KBH,*
min Pf
1 4 18.0 21.4
60 6.0 15.0
2 4 18.9 21.6
1860
.3 6.0
3 23.6 3 20.7
23.8 60 7.8
Total glutathione(GSH + GSSG); as GSH equivalents.
iri
-$
+ 20
[ * The deproteinized plasma samples were treated withKBHI (see
“Methods”).

102 # M [14C]-GSH phosphate buffer (pH 7.4) for 18 h at 4°C; after dialysis, no
glutathione could be detected. When 150 p~ glutathione was
added to the dialyzed plasma, it disappeared rapidly in a
L
manner similar to that shown in Fig. 3B with undialyzed
20 40 60 80 100 120
MINUTES
plasma. In such experiments, about one-half of the GSH was
Ffc. 3. Disappearance ofGSH added to plasma. A, glutathione converted to GSSG and theremainder disappeared as deter-
levels foundinitially (6min) and after standing for 30 and 120 min at mined by the enzymatic recycling procedure; however, after
23°C. 8,labeled GSH (102phf) was added to theplasma after it had treatment with borohydride and EDTA, about 90% of the
been allowed to stand for 30 min; the level of total glutathione was initially added glutathione was found by this assay procedure
followed for 90min as indicated. C,decrease in the radioactivity found and also by quantitative chromatography of the 2-vinylpyri-
in the supernatant solution (expressed in terms of glutathione con- dine derivative on a Dionex analyzer. Treatment with 33 m~
centration).
dithiothreitol and EDTA gave similar results in studies in
which glutathione was determined as the 2-vinylpyridine de-
of 99PM. After 1 h, the plasma was deproteinized and a portion
rivative. After addition of 200to 2 5 0 GSH
~ ~ to either dialyzed
was subjected to chromatography on a column (37 X 1 cm) of or undialyzed plasma, further addition of GSH led only to its
Sephadex G-10 (equilibrated with a buffer containing 100 m~ rapid conversion to GSSG. These studies indicate that non-
sodium acetate, pH 5.0). The flow rate was 15 ml/h, and dialyzable components of the plasma play a major role in the
fractions of 1 ml were collected. In this experiment, 0.225 ml utilization of GSH, and that theamount of GSH that can be
of the sulfosalicylic acid supernatant containing 17,000 cpm converted to form(s) unreactive in the enzymatic assay is
was placed on the column. The radioactivity and glutathione equivalent to about 100 p ~ .
were eluted together in a symmetrical peak. The fractions
were combined, concentrated, and added to the top of another DISCUSSION
column (103 X 1.4 cm) of Sephadex G-10 which was developed The present fiiding that most of the glutathione in blood
in the same manner as the previous column. The radioactivity plasma is in the form of GSH supports the conclusion that
eluted in two distinct but overlapping peaks. These were glutathione is translocated out of cells mainly in the form of
separately combined and analyzed for radioactivity and glu- GSH. The observation that GSH (in contrast to GSSG) dis-
tathione. The first peak contained 7,400 cpm and 5.99 nmol of appears rapidly in plasma is sigruficant in relation to GSH
glutathione; the second peak contained 8,200 cpm and 2.0 translocation. It thus appearsthat GSH is continuously trans-
nmol of glutathione. The two separately combined fractions ported into the plasma and used in reactions which lead to
were concentrated to low volume and reduced with KBH, in formation of GSSG and other products. Studies in which
the presence of EDTA as described under “Methods.” After plasma was incubated with 150 PM glutathione for 1 h and
reduction, the fractions were foundto contain 8.0 and 5.9 nmol then deproteinized and treated with KBH4 (or dithiothreitol)
of glutathione, respectively. Thus, a total of 13.9 nmol of and EDTA followed by treatment with 2-vinylpyridine did
glutathione was recovered after reduction, compared to 17.4 not reveal detectable amounts of the 2-vinylpyridine deriva-
nmol of glutathione added initially. Studies in which longer tives of cysteine, y-glutamylcysteine, or cysteinylglycine. Sam-
columns were used suggestthat thesecond peak may contain ples of such plasma that were oxidizedwith performic acidas
more than one product. These experiments indicate that the described (16) showed only glutathione sulfonic acid and no
glutathione that disappears is converted to a form or forms of taurine, indicating the absence of cysteamine. We must thus
molecular weight not markedly different from that of gluta- consider the possibility that other compounds, not containing
thione disullide. That glutathione appears after treatment free amino groups, may be involved in reactions with GSH.
with KB& and EDTA is consistent with the conclusion that GSH may beoxidized to GSSG by mechanisms involving
the glutathione that disappears in the plasma is converted molecular oxygen.For example, cysteinylglycineand cysteine
into disulfide form;formation of thiol esters is also possible. catalyze conversion of GSH to GSSG by undergoing nonen-
The results described above led us to determine the effect zymatic oxidation with 02 to form the corresponding disul-
of treating deproteinized plasma samples with KBH4 and fides; transhydrogenation of these with GSH yields GSSG
EDTA. Thus, plasma that had been allowedto standat 23OC (17). In the present work, several studies of the disappearance
for 1 h was deproteinized and the sulfosalicylic acid superna- of GSH from plasma were carried out under nitrogen. The
tant was treated with KBH,. As indicated in Table11, almost results, which were the same as found when the experiments
all of the glutathione that disappeared from the plasma on were done in air, suggest that reactions involving molecular
standing for 1 h was found after treatmentof the deproteinized oxygen are not responsible forGSSG formation; however,
plasma samples with potassium borohydride, as described since complete exclusion of oxygen is difficult to achieve, we
under “Methods.” cannot rule out the occurrence of such reactions unequivo-
Plasma was dialyzedagainst large volumes of 0.1 M sodium cally.
 Dynamic State of Glutathione in Blood Plasma 9533
Glutathione reductase occurs only intracellularly and there 4. Meister, A., Griffith, 0. W., Novogrodsky, A., and Tate, S . S.
does not appear to be an extracellular mechanism for GSSG (1980)Ciba Found. Symp. 72, 135-161
5. Bartoli, G. M., and Sies, H. (1978)FEBS Lett. 86,89-91
reduction. It thus appears that translocation of GSH consti- 6. Bannai, S., and Tsukeda, H. (1979)J. Bwl. Chem. 284,3444-3450
tutes a major source of thiol for the plasma. We may tenta- 7. Griffith, 0.W., and Meister, A. (1979)Proc. Natl. Acad. Sci. U.
tively consider the idea that the functions of plasma GSH S. A . 76,5606-5610
include reduction of the disulfide bonds of certain plasma 8. Haberle, D., Wahllander, A., and Sies, H. (1979)FEBS Lett. 108,
constituents and the mobilization of compounds that are 335-340
bound by disulfide linkage to proteins. The interesting possi- 9. Bartoli, G.M., Haberle, D., and Sies, H. (1978)in Functions of
Glutathione in Liver and Kidney (Sies, H., and Wendel, A.,
bility that GSH reacts to form other types of glutathione eds) pp. 27-31,Springer-Verlag, Heidelberg, West Germany
derivatives must also be considered. The nature andmetabolic 10. Tietze, F. (1969)Anal. Bwchem. 27,502-522
fate of the derivative(s) of GSH which are formed in plasma 11. Oser, B. L. (1965)Hawk’s Physiological Chemistry, 14th Ed, p.
need to be studied. 1096, McGraw-Hill Publications, New York
12. Wintrobe, M. M. (1967)Clinical Hematology, 6th Ed, p. 429, Lea
and Febiger, Philadelphia, Pa.
REFERENCES 13. Griffith, 0.W. (1980)Anal. Biochern.,in press
14. Allen, L., Merck, R., and Tunis, A. (1980)Res. Commun. Chem.
1. Meister, A., and Tate,S. S. (1976)Annu. Rev. Bwchem. 45,559- Pathol. Phannacol. 27, 175-182
604 15. Griffith, 0.W., and Meister, A. (1980)Proc. Natl. Acad. Sci. U.
2. Griffith, 0.W., and Meister, A. (1979)Proc. Natl. Acad. Sci. U. S. A . 77,3384-3387
S. A. 76,268-272 16. Tabor, C. W., and Tabor, H. (1977)Anal. Biochern. 78,543-553
3. G r B t h , 0 W., Novogrodsky, A., and Meister, A. (1979)Proc. 17. Griffith, 0.W., and Tate, S. S. (1980)J. Bwl. Chem. 258, 5011-
Natl. Acad.Sci. U.S. A. 76, 2249-2252 5014