- Email: [email protected]
31P-NMR measurements of ATP, ADP, 2,3-diphosphoglycerate and Mg2+ in human erythrocytes
Biochimica et Biophysica Acta, 1035 (1990) 169-174
169
Elsevier BBAGEN 23350
31p-NMR measurements of ATP, ADP, 2,3-diphosphoglycerate and Mg 2+ in human erythrocytes A q q a l u k P e t e r s e n 1,2, S o r e n R i s o m K r i s t e n s e n and Mogens Horder 1
1, J e n s
Peter Jacobsen 3
Department of Clinical Chemistry, Odense University Hospital, Odense, 2 Department of Clinical Chemistry, Rigshospitalet, Copenhagen and 3 Department of Chemistry, Odense Unioersity, Odense (Denmark)
(Received 13 September1989) (Revised manuscript received11 January 1990)
Key words: ADP; ATP; 2,3-Diphosphoglycerate;31p-NMR visibility; Free Mg2+; (Human erythrocyte)
Absolute 3tP-NMR measurements of ATP, ADP and 2,3-diphosphoglycerate (2,3-DPG) in oxygenated and partly deoxygenated human erythrocytes, compared to measurements by standard assays after acid extraction, show that ATP is only 65% NMR visible, ADP measured by NMR is unexpectedly 4 0 ~ higher than the enzymatic measurement and 2,3-DPG is fully NMR visible, regardless of the degree of oxygenation. These results show that binding to hemoglobin is unlikely to cause the decreased visibility of ATP in human erythrocytes as deoxyhemoglobin binds the phosphorylated metabolites more tightly than oxyhemoglobin. The high ADP visibility is unexplained. The levels of free Mg 2+ ([Mg2+]fre~) in human erythrocytes are 225 pmol/I at an oxygen saturation of 98.6% and instead of the expected increase, the level decreased to 196 pmol/I at an oxygen saturation of 38.1% based on the separation between the aand fl-ATP peaks. [Mg2+]fr~ in the erythroeytes decreased to 104/~mol/! at a high 2,3-DPG concentration of 25.4 m m o l / l red blood cells (RBC) and a normal ATP concentration of 2.05 m m o l / i RBC. By increasing the ATP concentration to 3.57 mmol/I RBC, and with a high 2,3-DPG concentration of 24.7 m m o l / l RBC, the 31p-NMR measured [Mg2+lf~ decreased to 61 pmol/I. These results indicate, that the 3~p-NMR determined [Mg2+]free in human erythrocytes, based solely on the separation of the a- and fl-ATP peaks, does not give a true measure of intracellular free Mg 2+ changes with different oxygen saturation levels. Furthermore the measurement is influenced by the concentration of the Mg 2+ binding metabolites ATP and 2,3-DPG. Failure to take these factors into account when interpreting 3tp-NMR data from human erythrocytes may explain some discrepancies in the literature regarding [Mg2+lfr~.
Introduction 32p-NMR measurements of phosphorylated metabolites [1-12] and intracellular free Mg 2÷ [13-18] ([Mg2+]tr~) in human erythrocytes have commonly been reported. In spite of this, absolute measurements of phosphorylated metabolites by this technique have only been attempted a few times [2,8]. Recently, we reported decreased 31P-NMR visibility of ATP in human erythro-
Abbreviations: 2,3-DPG, 2,3-diphosphoglycerate; HBSS, Hepes buffered saline solution; MDP, methylenediphosphonic acid; [Mg2+ ]free, free intracellular Mg2÷; Pi, inorganic phosphate; RBC, red blood cells. Correspondence: A. Petersen, Department of Clinical Chemistry, KK3011, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark.
cytes ( N M R measured ATP concentration lower than a value measured by perchloric acid extraction and a standard enzymatic method) [12]. In this study we have confirmed the decreased visibility as well as reporting on the visibility of ATP, ADP and 2,3-diphosphoglycerate (2,3-DPG) in human erythrocytes at two levels of oxygen saturation in order to investigate the influence of hemoglobin binding. By using the separation of the a-ATP and fl-ATP peaks and knowing the dissociation constant of MgATP, the [MgE+]free can be estimated by 31p-NMR in human erythrocytes [13,14]. [MgE*]fr¢e in human erythrocytes have been shown by this method to be lowered in untreated essential hypertension [14] and in erythrocytes from hypertensive rats compared to erythrocytes from control rats [19]. A more recent study failed to confirm the association between [Mg2+]rr¢~ in human erythrocytes and hypertension [18]. In addition, the
0304-4165/90/$03.50 © 1990 ElsevierSciencePublishers B.V. (BiomedicalDivision)
170 [Mg2+]free has been shown to decrease in stored erythroctyes [15-17]. An additional object of this study is to investigate the influence of oxygen saturation and the concentration of the major Mg2+-binding metabolites ATP and 2,3-DPG o n [Mg2+]fre e in order to assess the sensitivity of the [Mg2+]fr, , on these factors. Materials and Methods Erythrocytes
Erythrocytes, isolated from venous blood drawn from healthy donors, were washed and resuspended in an oxygenated Hepes (25 mM) buffered saline solution (HBSS) with 5 m M glucose (pH 7.40), essentially as described previously [20]. In some cases whole blood was concentrated by centrifugation for 31p-NMR measurements. The erythrocytes were partially deoxygenated to oxygen saturation levels found in human venous blood by incubation in a tonometer (Eschweiler, Kiel, F.R.G.) with pure nitrogen saturated with water vapor at 25 o C. Variable levels of 2,3-DPG were induced by preincubating the erythrocytes in a medium containing 75 mM NaCI, 10 m M inosine, 10 m M pyruvate and 50 m M N a H 2 P O 4, p H = 7 . 4 0 at 3 7 ° C [21]. ATP (and 2,3-DPG) were increased by preincubating the erythrocytes in the same medium, except that adenosine was substituted for inosine [22]. After the preincubation the erythrocytes were washed once in the oxygenated HBSS before concentration for 31p-NMR measurements. The mean hematocrit value was 70 + 2% (S.E.), n = 36. Care was taken to ensure near complete oxygenation of the preincubated erythrocytes. Oxygen saturation was measured as previously described [10]. Enzymatic measurements
Aliquots of the erythrocyte suspensions both before and after the N M R measurement were prepared for enzymatic analysis as previously described [12]. A T P and A D P were determined by a standard luciferinluciferase method [23], and 2,3-DPG was determined enzymatically as previously described [10].
and 20 s. This reflects exactly the expected saturation of the 3-phosphorus peak of 2,3-DPG ( T 1 = 2.25 s). The concentration of ATP was measured using the area of the fl-ATP peak while the concentration of A D P was estimated by subtracting the/3-ATP from the combined fl-ADP and "y-ATP peaks [12]. The validity of the N M R measured concentrations were controlled by simultaneously measuring known amounts of ATP in a Hepes buffered saline solution (pH 7.3) by both N M R and the luciferin-luciferase method. The N M R measured ATP value was 0.80 + 0.10 (S.D.) mmol/1, n = 3, while the enzymatically determined value was 0.82 ___0.02 (S.D.) mmol/1, n = 3. 125 m m o l / 1 methylenediphosphonic acid (MDP) in 50 m m o l / l Tris-HC1 (pH 9.0) served as external reference (16.90 p p m relative to 85% H3PO4) as well as a concentration standard (1.09 _+ 0.03 (S.D.) mmol/1, n = 6 relative to known amounts of Pi). For the N M R determined values of both the ATP control and the Pi calibration measurements, a pulse angle of 90 ° (here 12 kts), an acquisition time of 0.5 s and a relaxation delay of 29.5 s were used. For intracellular p H determinations the chemical shifts of 3phosphorus of 2,3-DPG and Pi were used, based on calibration curves obtained on oxygenated hemolysates at 25 o C, essentially as previously described [20]. Estimation o f [ M g 2
+]free
The basis for the estimation of [Mg2+]free by 3:p_ N M R spectroscopy is defined the following equilibrium: M g A T P ~- ATP r + Mg 2 +
( 1)
Where ATPf denotes all A T P species not complexed to Mg 2+. The chemical shift difference between the a- and /3-ATP peaks for MgATP is 2.5 p p m lower than for ATP [13]. As there is a rapid exchange on the N M R time-scale between the ATPf and MgATP, the measured chemical shift difference between the a- and /3-ATP peaks is a weighted average for intermediate degrees of complexation. The fraction of ATPf to ATP t, where ATP t is total ATP, is readily obtained from the 31p_ N M R spectrum as [13,14]:
3JP - N M R measurements
31p-NMR measurements were performed at 101.3 M H z in a Bruker AC 250 M H z instrument using a 15 m m probe. The samples were spun with 20 Hz and 125 scans were accumulated with 90 ° pulses and gated decoupling and a repetition time of 7 s at 25 o C. The 90 ° pulse was determined to 14 /~s in erythrocyte suspensions by finding the 180 ° pulse to 28 /~s. The repetition time used ensured complete relaxation of the ATP (and ADP) peaks ( T 1 = 1.2-1.4 s), while the 3phosphorus peak of 2,3-DPG was partially saturated, for which a correction factor of 1.04 was obtained by accumulating spectra alternately at repetition times of 7
= [ATP¢ ] / [ A T P t ] = (6obs _ ~ M g A T P ) / / ( ~ A T P ~MgATP)
(2)
Where 6 is the chemical shift difference between the aand jS-ATP peaks. [Mg2+]eree is then obtained by: [Mg 2+ ]r,e* = KD (dP- 1 _ 1)
(3)
where K D is the dissociation constant for M g A T P [13,14]. The value for the dissociation constant used in this study is 50 ktmol/1 at 2 5 ° C [24]. This uncorrected value was used in experiments with normal levels of ATP and 2,3-DPG, but the p H lowering in erythrocytes
171 with high levels of 2,3-DPG and A T P necessitated a p H correction as described in Ref. 15. Total erythrocyte Mg 2+ was checked using a K o d a k Ectaderm 400 analyzer [25] in a series of experiments where the erythrocytes were preincubated in the Mg 2÷ free inosine-pyruvate-phosphate medium. A value of 2 . 1 6 _ 0.11 (S.E.) m m o l / 1 RBC, n = 6, was obtained, irrespective of incubation time up to 5 h, indicating that total erythrocyte Mg 2÷ does not change during the course of the experiments and confirming that Mg 2+ transport across the erythrocyte membrane is very slow [261.
A
2,3-DPG
MDP
~
ATP
C
Results
A 31p-NMR spectrum at 101.3 M H z of erythroctyes in the oxygenated HBSS is depicted in Fig. 1A, while Fig. 1B shows a spectrum of erythrocytes deoxygenated to an oxygen saturation of 31%. Note the global downfield shifts of the resonances and the linebroadening especially evident in the 2,3-DPG peaks, Fig. 1B [27]. Fig. 1C is a 31p-NMR spectrum of erythrocytes preincubated in the inosine-pyruvate-phosphate medium for 4.5 h in order to induce supraphysiological levels of 2,3-DPG, while the ATP concentration remains at the physiological level [21]. Preincubation of erythrocytes in adenosine-pyruvate-phosphate medium induces high levels of both ATP and 2,3-DPG [22] as is evident in Fig. 1D, which is a 31p-NMR spectrum of erythrocytes preincubated in this medium for 4 h. The accumulation of organic phosphates results in a Donnan-mediated lowering of intraceUular pH, which in the spectra results in upfield shifts of the 2,3-DPG peaks and the 7-ATP peak, Fig. 1C and D [10]. At the same time the separation between the a- and fl-ATP peaks is increased as the level of [Mg2+]fre e is decreased, Fig. 1C and D [13].
Measurements of ADP, A TP and 2,3-DPG Quantitative measurements of ADP, ATP and 2,3D P G in human erythrocytes at two levels of oxygen saturation are shown in Table I. Measurements by 31p-NMR spectroscopy are compared to enzymatic measurements on the same samples. 2,3-DPG is fully 31p-NMR visible, both in the fully oxygenated and partially deoxygenated erythrocytes, Table I. Deoxygenation causes an increase in the concentration of triose-phosphates [28,29] and this may explain the slightly higher N M R determined 2,3-DPG value in the partially deoxygenated state, as the chemical shift of 3-phosphorus of 2,3-DPG is very close to that of the triose phosphates. Contrary to the full visibility of 2,3D P G , the 31P-NMR visibility of A T P is decreased to about 65%, again regardless of oxygen saturation, Table I. The differences between the 3tp-NMR measured values and the values determined by the luciferin-luciferase method are 0.74 to 0.78 m m o l / 1 cell water in absolute
D
I
20
'
10
'
0 ppm
--
'
-10
I
-20
Fig. 1. 31P-NMR spectra of human erythrocytes at 101.3 MHz and 25 ° C. An exponential multiplication factor resulting in a linebroadening of 20 Hz was applied to the FIDs before Fourier transformation. The spectra have been normalized to the reference peak of methylenediphosphonic acid (MDP). The peak assignments are: ATP, adenosinetriphosphate; 2,3-DPG, 2,3-diphosphoglycerate and Pi, inorganic phosphate. The top spectrum (A) is of oxygenated erythrocytes while the next spectrum (B) is of erythroctyes deoxygenated to an oxygen saturation of 31%. Note the global downfield shifts induced by deoxygenation. Spectrum (C) is of erythrocytes preincubated in inosine-pyruvate-phosphatemedium for 4.5 h in order to increase the concentration of 2,3-DPG. The bottom spectrum (D) is of erythrocytes preincubated for 4 h in adenosine-pyruvate-phosphatemedium thereby increasing the concentration of both 2,3-DPG and ATP. Note the upfield shifts of 2,3-DPG in (C) and (D), mainly caused by a lower intraceUular pH. In addition the chemical shift difference between the a-ATP and fl-ATP peaks is increased in both (C) and (D) due to a lower concentration of free intraceUular Mg2 + .
terms, Table I. Surprisingly the 31p-NMR determined A D P concentrations are about 400% higher than the values determined by the luciferin-luciferase method and corresponds to a difference of 0.3 m m o l / 1 cell water in absolute terms, regardless of oxygen saturation, Table I. This difference is comparable to the difference found in our previous study [12], although the absolute concentrations of A D P are lower in this study.
Measurements of M,_2 g +1]free The [ M g > ] freewas measured using the chemical shift difference between the a- a n d / 3 - A T P peaks [13,14]. The
172 TABLE I ADP, ATP and 2,3-DPG concentrations in human erythrocytes at two different levels of oxygen saturation
The erythrocytes were washed and resuspended in in Hepes buffered saline solution [20]. 31p-NMR determined concentrations are compared to determinations by luciferin-luciferase(ADP and ATP) and by an enzymatic method (2,3-DPG) on samples extracted by perchloric acid both before and after the NMR measurement. For 31P-NMR measurement of ATP, the fl-ATP peak was used, while the concentration of ADP was obtained by subtracting the fl-ATP peak from the combined fl-ADP and 3,-ATP peaks. The 3-phosphorus peak of 2,3-DPG was used to estimate the concentration of 2,3-DPG. The concentration is given in retool/1 cell water 5: S.E. (n = 8) assuming the water fraction of the erythrocytes to be 0.70. Oxygen saturation
ATP (mmol/1)
ADP (mmol/1)
2,3-DPG (mmol/1)
(%)
NMR
luciferase
NMR
luciferase
NMR
enzymatic
38.1 + 1.9
1.49 + 0.13 a
2.27 + 0.05
0.39 + 0.07 b
0.081 + 0.005
7.91 + 0.26
7.35 _+0.09
98.6+0.3
1.24 + 0.11d
1.98+0.10
0.41 +0.10 d
0.127+0.003
7.01 5 : 0 . 1 9
7.245:0.13
a p < 0.001, b p < 0.005, c p < 0.0005 and ,t p < 0.05 by paired two-tailed t-tests as compared to the values determined by luciferin-hiciferase.
results o b t a i n e d are s u m m a r i z e d in T a b l e II, where [Mg/+]rre ~ in whole b l o o d a n d in fully oxygenated a n d partially deoxygenated erythrocytes are shown. The resuits o n [Mg2+]fr~¢ m e a s u r e m e n t s are from the same erythroctyes as used in the q u a n t i f i c a t i o n of A D P , A T P a n d 2 , 3 - D P G depicted in T a b l e I. The 3 a p - N M R measured c o n c e n t r a t i o n s of A T P a n d 2 , 3 - D P G are also shown in the case of the isolated erythrocytes, T a b l e II. I n whole blood, the [Mg2+]er~e is 221 # m o l / 1 a n d c o m p a r a b l e to the 225 /~mol/1 in oxygenated erythro-
TABLE II 3~P-NMR measurements of intracellular free Mg 2+ in human erythrocytes
The intracellular free Mg 2÷ was calculated from the 31P-NMR spectra using the chemical shift difference between the a- and fl-ATP peaks [13,14] and assuming a dissociation constant of MgATP of 50 ttmol/1 at 25°C [24]. The results shown are from whole blood and isolated erythrocytes in a Hepes buffered saline solution [20] at two different levels of oxygen saturation (ox. sat.). In addition, results from erythrocytes preincubated in adenosine-pyruvate-phosphate medium (High ATP/2,3-DPG) and in inosine-pyruvate-phosphate medium (High 2,3-DPG) are shown. The concentrations of ATP and 2,3-DPG shown were determined by 31p-NMR and are expressed in mmol/l erythrocytes, while [Mg 2+ ]tree is expressed as /xmol/1. The results are expressed as means + S.E. with the number of determinations in parentheses.
cytes, T a b l e II. These results are similar to previously published values of [Mg2+]free in erythrocytes [1315,18]. I n contrast to the expected increase of [Mg2+]tree in deoxygenated erythrocytes [13], the result o b t a i n e d in erythrocytes partially d e o x y g e n a t e d to an oxygen s a t u r a t i o n of 38% is 196 # m o l / 1 , T a b l e II a n d Fig. 1A a n d B. 2 , 3 - D P G a n d A T P are q u a n t i t a t i v e l y the most imp o r t a n t Mg2+-complexing metabolites in the h u m a n erythrocyte [31]. By p r e i n c u b a t i n g the erythrocytes in i n o s i n e - p y r u v a t e - p h o s p h a t e m e d i u m [21] the c o n c e n t r a tion of 2 , 3 - D P G can be elevated 4 - 5 - f o l d , a n d b y p r e i n c u b a t i o n in a d e n o s i n e - p y r u v a t e - p h o s p h a t e m e d i u m the c o n c e n t r a t i o n s of b o t h 2 , 3 - D P G a n d A T P c a n be elevated 2 - 5 - t i m e s physiological levels [22]. These exp e r i m e n t a l procedures were used to produce erythrocytes with elevated c o n c e n t r a t i o n s of 2 , 3 - D P G a n d A T P in short term i n c u b a t i o n s ( < 5 h). A t high A T P a n d 2 , 3 - D P G levels the [Mg2+]fre e decreases to 6 1 / ~ m o l / l as expected, while the c o n c e n t r a t i o n of [Mg2+]fre ~ is lowered to 104 /~mol/1 in erythrocytes with n o r m a l levels of ATP, b u t 5-fold elevated levels of 2 , 3 - D P G (Table II a n d Fig. 1C a n d D).
Discussion 31p-NMR measurements of ADP, A TP and 2,3-DPG
ATP (mmol/l)
2,3-DPG (mmol/1)
[Mg2+ ]free (/~mol/1)
Whole blood
-
-
221 + 14 (10)
Erythrocytes ox. sat. =99%
0.89-t-0.11(8)
4.91+0.19 (8)
225+ 8 (8)
Erythrocytes ox. sat. = 38%
1.04+0.05 (8)
5.54+0.26 (8)
196+13 (8)
Erythrocytes High ATP/ 2,3-DPG
3.57+0.10(5)
24.7 +2.1 (5)
61+10 (5)
Erythrocytes High2,3-DPG
2.05+0.13(5)
25.4 +1.2 (5)
104_+ 7 (5)
Despite n u m e r o u s 3 1 p - N M R investigations of h u m a n erythrocytes, absolute q u a n t i f i c a t i o n s of A T P a n d 2,3D P G have only b e e n reported a few times [2,8]. Recently we reported that the 31p-NMR visibility of A T P is 78% when c o m p a r i n g the N M R m e a s u r e m e n t with a s t a n d a r d enzymatic m e t h o d [12]. I n order to investigate the influence of b i n d i n g of 2 , 3 - D P G a n d A T P to hemoglobin, we have repeated the m e a s u r e m e n t s in fully oxygenated a n d in partially d e o x y g e n a t e d erythrocytes as these p h o s p h o r y l a t e d metabolites are b o u n d preferentially to d e o x y h e m o g l o b i n [13,31,32]. 2 , 3 - D P G is fully 3 a p - N M R visible, while we c o n f i r m the decreased visibility of A T P , regardless of the degree
173 of oxygen saturation (Table I). In this study we find that the ATP visibility is decreased to 65% or slightly less than previously reported [12]. The mechanism behind this decreased visibility of ATP remains unclear as a systematic underestimation by the 31P-NMR procedure can be excluded by the full visibility of 2,3-DPG, Table I. Recently, the occurrence of 31P-NMR invisible ATP was reported in ischaemic rat liver, possibly related to mitochondrial ATP [33] but obviously this explanation cannot be used in the case of the erythrocyte. The fl-ATP peak used to quantitate ATP may be broadened by slow exchange of Mg 2÷ [34], although this explanation seems more unlikely because of the temperature and field dependence of the slow exchange situation [34] and by the fact that most ATP in the human erythrocyte is complexed to Mg 2+ [13]. The most plausible biochemical explanation for the decreased visibility is binding of ATP to the membrane of the erythrocyte, although further speculation is not warranted without experimental evidence. The 31p-NMR determined ADP value is unexpectedly 400% higher than the value determined by the luciferin-luciferase method, Table I. In absolute terms, the N M R determined value is 0.3 mmol/1 cell water higher, Table I, and this difference is of the same magnitude as previously reported, albeit at higher ADP concentrations [12]. The reason for the higher N M R determined ADP value is not clear although it might reflect an underestimation of the fl-ATP peak by 0.3 mmol/1, as ADP is determined by subtracting the flATP peak from the combined ~,-ATP and fl-ADP peaks [12]. This of course would increase the visibility of ATP, but only to about 80%. Measurements o f [ M g 2 +] free
Information regarding [Mg2+]free can be extracted from the 3~p-NMR spectrum by using the chemical shift difference between the a- and fl-ATP peaks [13,14]. Using this technique an association between lowered [Mg2+]free in erythrocytes and essential hypertension has been reported both in human erythroctyes [14] and in erythrocytes from spontaneously hypertensive rats [19]. Using the same method a recent study failed to confirm the association between essential hypertension and [Mg2÷]fr~ in human erythrocytes [18]. Furthermore, [Mg2+]fr~ has been reported to be lowered in stored erythrocytes [15-17], although the mechanism remains unclear. By decreasing oxygen saturation, [Mg2+]fr¢~ is expected to increase due to differential binding of the Mg 2+ complexed and uncomplexed ATP and 2,3-DPG to deoxyhemoglobin [13,31,32]. As shown in Table II, this expected increase does not occur when the measurement is based solely on the separation between the a- and fl-ATP peaks. In fact as shown in Table III, where the results of this study are compared to calculated levels from previously published
TABLE III Effect of deoxygenation of human erythrocytes on the 31P-NMR determined value of [Mg 2+]/tee
The 31p-NMR determined [Mg2+]fr¢¢ concentrations in human erythrocytesfrom three studies at different degreesof oxygenationare shown. In the study of Labotka [27] and of Gupta et al. [13] [Mg2+ ]free was calculated from spectral data given, and solely by use of the chemical shift difference between the a- and fl-ATP peaks without taking the hemoglobinbinding of ATP into account. As these studies were conducted at 37o C, the dissociation constant for MgATP used were 38 ttmol/l [13]. This study was conducted at 25°C and a dissociation constant for MgATP of 50/~mol/1 was used [24]. Oxygen saturation Oxygenated Deoxygenated
[Mg2+ ]free (#mol/l) 193 97
Reference 27 27
Oxygenated Deoxygenated
258 163
13 13
Oxygenated Ox. sat. = 38%
225 196
this study this study
spectra, deoxygenation tends to decrease the measured [Mg2+]rr~. This unexpected finding is probably due to the equation for q~ used (see Materials and Methods) where q , = [ A T P f ] / [ A T P t ] . Here ATP t denotes all species of ATP not complexed to Mg 2÷ and ATP t the total concentration of ATP [13]. Included in the free species of ATP is hemoglobin bound ATP, and as deoxygenation increases the hemoglobin binding of ATP relative to MgATP [13,31,32] the above fraction is increased. [Mg2+]free is thereby decreased when using Eqn. 3 to calculate the level, as it is assumed that all ATP is available for binding to Mg 2÷. Using a more complete analysis, where the differential binding of MgATP and ATP to hemoglobin is included, one may obtain an increase in [MgE+]fre e by deoxygenation using the spectral data and appropriate dissociation constants for ATP-hemoglobin complexes [13]. These results indicate that [Mg2+]free measurements in human erythrocytes using Eqns. 2 and 3 are biased by ATP binding to hemoglobin. A number of factors, such as oxygen saturation of hemoglobin, 2,3-DPG concentration and intracellular pH, are likely to influence the binding of ATP to hemoglobin [13,31,32]. A change of these factors may alter q~ and thereby the N M R measured [Mg2+]fr~ without a corresponding change in true [Mg2+]free or more likely a change that is in the opposite direction of the N M R measured. This may explain the contradictory results regarding an association between essential hypertension and [Mg2+]fr~ in human erythrocytes [14,18] and may even explain the apparently lowered [Mg2+]free in stored erythrocytes which so far lacks explanation [15-17]. In addition we have manipulated the concentrations of the major Mg2+-binding metabolites ATP and 2,3-
174 D P G to investigate the sensitivity of the measured [Mg2÷]free on these metabolites. As shown in Table II, increasing the 2,3-DPG concentration 5-fold induces a decrease in [Mg2+]fre e from 225 to 104/~mol/1, while a simultaneous increase of A T P 2-fold and 2,3-DPG 5-fold induces the [Mg2+]fre e to go down as low as 61/~mol/1. Although the perturbations of 2,3-DPG and A T P in this study are highly unphysiological, one may expect the [Mg2+]rree tO be dependent on the concentrations of ATP and 2,3-DPG at more physiological levels at any given total Mg 2÷ level.
Acknowledgements Anne Marie Jacobsen is thanked for technical assistance, while Soren Wamberg is thanked for help with equipment and advice. This research was partly supported by a junior research fellowship to A.P. from the Danish Research Academy.
References 1 Henderson, T.O., Costello, A.J.R. and Omachi, A. (1974) Proc. Natl. Acad. Sci. USA 71, 2487-2490. 2 Labotka, R.J., Glonek, T., Hruby, M.A. and Honig, G.R. (1976) Biochem. Med. 15, 311-329. 3 Kagimoto, T., Hayashi, F., Yamasaki, M., Morino, Y., Akasaka, K. and Kishimoto, S. (1978) Experimentia 34, 1092-1093. 4 Lam, Y.-F., Lin, A.K.-L.C. and Ho, C. (1979) Blood 54, 196-209. 5 Labotka, R.J. and Honig, G.R. (1980) Am. J. Hematol. 9, 55-65. 6 Sarpel, G., Lubansky, H.J., Danon, M.J. and Omachi, A. (1981) Arch. Neurol. 38, 271-274. 7 Swanson, M.S., Angle, C.R., Stohs, S.J., Wu, S.T., Salhany, J.M., Eliot, R.S. and Markin, R.S. (1983) Proc. Natl. Acad. Sci. USA 80, 169-172. 8 Zwerling, H.K., Diamond, J.N., Levy, G.C. and Clark, D.A. (1986) Magn. Res. Med. 4, 10-14. 9 Ambruso, D.R., Hawkins, B., Johnson, D.L., Fritzberg, A.R., Klingensmith III, W.C. and McCabe, E.R.B. (1986) Biochem. Med. Metabol. Biol. 35, 376-383. 10 Petersen, A., Jacobsen, J.P. and Horder, M. (1987) Scand. J. Clin. Lab. Invest. 47, 233-237.
11 Monti, J.P., Martinez, S., Braguer, D., Courtina, C., Francois, G. and Crevat, A. (1987) Clin. Chem. 33, 2284-2285. 12 Petersen, A., Pedersen, E.J. and Quistorff, B. (1989) Biochim. Biophys. Acta 1012, 267-271. 13 Gupta, R.J., Benovic, J.L. and Rose, Z.B. (1978) J. Biol. Chem. 253, 6172-6176. 14 Resnick, L.M., Gupta, R.J. and Laragh, J.H. (1984) Proc. Natl. Acad. Sci. USA 81, 6511-6515. 15 Bock, J.L., Wenz, B. and Gupta, R.J. (1985) Blood 65, 1526-1530. 16 Bock, J.L., Wenz, B. and Gupta, R.J. (1987) Biochim. Biophys. Acta 928, 8-12. 17 Bock, J.L. and Yusuf, Y. (1988) Biochim. Biophys. Acta 941, 225-231. 18 Woods, K.L., Walmsley, D., Haegerty, A.M., Turner, D.L. and Lian, L.-Y. (1988) Clin. Sci. 74, 513-517. 19 Matuura, T., Kohno, M., Kanayama, Y. et al. (1987) Biochem. Biophys. Res. Commun. 143, 1012 1017. 20 Petersen, A., Jacobsen, J.P. and Horder, M. (1987) Magn. Res. Med. 4, 341-350. 21 Deuticke, B., Duhm, J. and Dierkesmann, R. (1971) Pfliigers Arch. 326, 15-34. 22 Zachara, B. (1977) Transfusion 17, 628-634. 23 Lundin, A., Hasenson, M., Persson, J. and Pousette, ,~. (1986) in Bioluminescence and Chemiluminescence (Deluca, M. and McElroy, W., eds.), part B, pp. 27-42, Academic Press, New York. 24 Gupta, R.J., Gupta, P., Yushok, W.D. and Rose, Z.B. (1983) Biochem. Biophys. Res. Commun. 117, 210-216. 25 Bonnay, M.M., Legac, E.M., Brunier, A.P. and Bohuon, C.J. (1988) Clin. Chem. 34, 2154-2155. 26 F6ray, J.-C. and Garay, R. (1986) Biochim. Biophys. Acta 856, 76-84. 27 Labotka, R.J. (1984) Biochemistry 23, 5549-5555. 28 Flatman, P.W. and Lew, V.L. (1980) J. Physiol. 305, 13-30. 29 Hamasaki, N., Asakura, T. and Minamaki, S. (1970) J. Biochem. 68, 157-161. 30 Rapoport, I., Berger, H., Rapoport, S.M., Eisner, R. and Gerber, G. (1976) Biochim. Biophys. Acta 428, 193-204. 31 Bunn, H.F., Ransil, B.J. and Chao, A. (1972) J. Biol. Chem. 246, 5273-5279. 32 Gupta, R.J., Benovic, J.L. and Rose, Z.B. (1978) J. Biol. Chem. 253, 6165-6171. 33 Murphy, E., Gabel, S.A., Funk, A. and London, R.E. (1988) Biochemistry 27, 526-528. 34 Sontheimer, G.M., Kuhn, W. and Kalbitzer, H.R. (1986) Biochem. Biophys. Res. Commun. 134, 1379-1386.
169
Elsevier BBAGEN 23350
31p-NMR measurements of ATP, ADP, 2,3-diphosphoglycerate and Mg 2+ in human erythrocytes A q q a l u k P e t e r s e n 1,2, S o r e n R i s o m K r i s t e n s e n and Mogens Horder 1
1, J e n s
Peter Jacobsen 3
Department of Clinical Chemistry, Odense University Hospital, Odense, 2 Department of Clinical Chemistry, Rigshospitalet, Copenhagen and 3 Department of Chemistry, Odense Unioersity, Odense (Denmark)
(Received 13 September1989) (Revised manuscript received11 January 1990)
Key words: ADP; ATP; 2,3-Diphosphoglycerate;31p-NMR visibility; Free Mg2+; (Human erythrocyte)
Absolute 3tP-NMR measurements of ATP, ADP and 2,3-diphosphoglycerate (2,3-DPG) in oxygenated and partly deoxygenated human erythrocytes, compared to measurements by standard assays after acid extraction, show that ATP is only 65% NMR visible, ADP measured by NMR is unexpectedly 4 0 ~ higher than the enzymatic measurement and 2,3-DPG is fully NMR visible, regardless of the degree of oxygenation. These results show that binding to hemoglobin is unlikely to cause the decreased visibility of ATP in human erythrocytes as deoxyhemoglobin binds the phosphorylated metabolites more tightly than oxyhemoglobin. The high ADP visibility is unexplained. The levels of free Mg 2+ ([Mg2+]fre~) in human erythrocytes are 225 pmol/I at an oxygen saturation of 98.6% and instead of the expected increase, the level decreased to 196 pmol/I at an oxygen saturation of 38.1% based on the separation between the aand fl-ATP peaks. [Mg2+]fr~ in the erythroeytes decreased to 104/~mol/! at a high 2,3-DPG concentration of 25.4 m m o l / l red blood cells (RBC) and a normal ATP concentration of 2.05 m m o l / i RBC. By increasing the ATP concentration to 3.57 mmol/I RBC, and with a high 2,3-DPG concentration of 24.7 m m o l / l RBC, the 31p-NMR measured [Mg2+lf~ decreased to 61 pmol/I. These results indicate, that the 3~p-NMR determined [Mg2+]free in human erythrocytes, based solely on the separation of the a- and fl-ATP peaks, does not give a true measure of intracellular free Mg 2+ changes with different oxygen saturation levels. Furthermore the measurement is influenced by the concentration of the Mg 2+ binding metabolites ATP and 2,3-DPG. Failure to take these factors into account when interpreting 3tp-NMR data from human erythrocytes may explain some discrepancies in the literature regarding [Mg2+lfr~.
Introduction 32p-NMR measurements of phosphorylated metabolites [1-12] and intracellular free Mg 2÷ [13-18] ([Mg2+]tr~) in human erythrocytes have commonly been reported. In spite of this, absolute measurements of phosphorylated metabolites by this technique have only been attempted a few times [2,8]. Recently, we reported decreased 31P-NMR visibility of ATP in human erythro-
Abbreviations: 2,3-DPG, 2,3-diphosphoglycerate; HBSS, Hepes buffered saline solution; MDP, methylenediphosphonic acid; [Mg2+ ]free, free intracellular Mg2÷; Pi, inorganic phosphate; RBC, red blood cells. Correspondence: A. Petersen, Department of Clinical Chemistry, KK3011, Rigshospitalet, Blegdamsvej 9, DK-2100 Copenhagen, Denmark.
cytes ( N M R measured ATP concentration lower than a value measured by perchloric acid extraction and a standard enzymatic method) [12]. In this study we have confirmed the decreased visibility as well as reporting on the visibility of ATP, ADP and 2,3-diphosphoglycerate (2,3-DPG) in human erythrocytes at two levels of oxygen saturation in order to investigate the influence of hemoglobin binding. By using the separation of the a-ATP and fl-ATP peaks and knowing the dissociation constant of MgATP, the [MgE+]free can be estimated by 31p-NMR in human erythrocytes [13,14]. [MgE*]fr¢e in human erythrocytes have been shown by this method to be lowered in untreated essential hypertension [14] and in erythrocytes from hypertensive rats compared to erythrocytes from control rats [19]. A more recent study failed to confirm the association between [Mg2+]rr¢~ in human erythrocytes and hypertension [18]. In addition, the
0304-4165/90/$03.50 © 1990 ElsevierSciencePublishers B.V. (BiomedicalDivision)
170 [Mg2+]free has been shown to decrease in stored erythroctyes [15-17]. An additional object of this study is to investigate the influence of oxygen saturation and the concentration of the major Mg2+-binding metabolites ATP and 2,3-DPG o n [Mg2+]fre e in order to assess the sensitivity of the [Mg2+]fr, , on these factors. Materials and Methods Erythrocytes
Erythrocytes, isolated from venous blood drawn from healthy donors, were washed and resuspended in an oxygenated Hepes (25 mM) buffered saline solution (HBSS) with 5 m M glucose (pH 7.40), essentially as described previously [20]. In some cases whole blood was concentrated by centrifugation for 31p-NMR measurements. The erythrocytes were partially deoxygenated to oxygen saturation levels found in human venous blood by incubation in a tonometer (Eschweiler, Kiel, F.R.G.) with pure nitrogen saturated with water vapor at 25 o C. Variable levels of 2,3-DPG were induced by preincubating the erythrocytes in a medium containing 75 mM NaCI, 10 m M inosine, 10 m M pyruvate and 50 m M N a H 2 P O 4, p H = 7 . 4 0 at 3 7 ° C [21]. ATP (and 2,3-DPG) were increased by preincubating the erythrocytes in the same medium, except that adenosine was substituted for inosine [22]. After the preincubation the erythrocytes were washed once in the oxygenated HBSS before concentration for 31p-NMR measurements. The mean hematocrit value was 70 + 2% (S.E.), n = 36. Care was taken to ensure near complete oxygenation of the preincubated erythrocytes. Oxygen saturation was measured as previously described [10]. Enzymatic measurements
Aliquots of the erythrocyte suspensions both before and after the N M R measurement were prepared for enzymatic analysis as previously described [12]. A T P and A D P were determined by a standard luciferinluciferase method [23], and 2,3-DPG was determined enzymatically as previously described [10].
and 20 s. This reflects exactly the expected saturation of the 3-phosphorus peak of 2,3-DPG ( T 1 = 2.25 s). The concentration of ATP was measured using the area of the fl-ATP peak while the concentration of A D P was estimated by subtracting the/3-ATP from the combined fl-ADP and "y-ATP peaks [12]. The validity of the N M R measured concentrations were controlled by simultaneously measuring known amounts of ATP in a Hepes buffered saline solution (pH 7.3) by both N M R and the luciferin-luciferase method. The N M R measured ATP value was 0.80 + 0.10 (S.D.) mmol/1, n = 3, while the enzymatically determined value was 0.82 ___0.02 (S.D.) mmol/1, n = 3. 125 m m o l / 1 methylenediphosphonic acid (MDP) in 50 m m o l / l Tris-HC1 (pH 9.0) served as external reference (16.90 p p m relative to 85% H3PO4) as well as a concentration standard (1.09 _+ 0.03 (S.D.) mmol/1, n = 6 relative to known amounts of Pi). For the N M R determined values of both the ATP control and the Pi calibration measurements, a pulse angle of 90 ° (here 12 kts), an acquisition time of 0.5 s and a relaxation delay of 29.5 s were used. For intracellular p H determinations the chemical shifts of 3phosphorus of 2,3-DPG and Pi were used, based on calibration curves obtained on oxygenated hemolysates at 25 o C, essentially as previously described [20]. Estimation o f [ M g 2
+]free
The basis for the estimation of [Mg2+]free by 3:p_ N M R spectroscopy is defined the following equilibrium: M g A T P ~- ATP r + Mg 2 +
( 1)
Where ATPf denotes all A T P species not complexed to Mg 2+. The chemical shift difference between the a- and /3-ATP peaks for MgATP is 2.5 p p m lower than for ATP [13]. As there is a rapid exchange on the N M R time-scale between the ATPf and MgATP, the measured chemical shift difference between the a- and /3-ATP peaks is a weighted average for intermediate degrees of complexation. The fraction of ATPf to ATP t, where ATP t is total ATP, is readily obtained from the 31p_ N M R spectrum as [13,14]:
3JP - N M R measurements
31p-NMR measurements were performed at 101.3 M H z in a Bruker AC 250 M H z instrument using a 15 m m probe. The samples were spun with 20 Hz and 125 scans were accumulated with 90 ° pulses and gated decoupling and a repetition time of 7 s at 25 o C. The 90 ° pulse was determined to 14 /~s in erythrocyte suspensions by finding the 180 ° pulse to 28 /~s. The repetition time used ensured complete relaxation of the ATP (and ADP) peaks ( T 1 = 1.2-1.4 s), while the 3phosphorus peak of 2,3-DPG was partially saturated, for which a correction factor of 1.04 was obtained by accumulating spectra alternately at repetition times of 7
= [ATP¢ ] / [ A T P t ] = (6obs _ ~ M g A T P ) / / ( ~ A T P ~MgATP)
(2)
Where 6 is the chemical shift difference between the aand jS-ATP peaks. [Mg2+]eree is then obtained by: [Mg 2+ ]r,e* = KD (dP- 1 _ 1)
(3)
where K D is the dissociation constant for M g A T P [13,14]. The value for the dissociation constant used in this study is 50 ktmol/1 at 2 5 ° C [24]. This uncorrected value was used in experiments with normal levels of ATP and 2,3-DPG, but the p H lowering in erythrocytes
171 with high levels of 2,3-DPG and A T P necessitated a p H correction as described in Ref. 15. Total erythrocyte Mg 2+ was checked using a K o d a k Ectaderm 400 analyzer [25] in a series of experiments where the erythrocytes were preincubated in the Mg 2÷ free inosine-pyruvate-phosphate medium. A value of 2 . 1 6 _ 0.11 (S.E.) m m o l / 1 RBC, n = 6, was obtained, irrespective of incubation time up to 5 h, indicating that total erythrocyte Mg 2÷ does not change during the course of the experiments and confirming that Mg 2+ transport across the erythrocyte membrane is very slow [261.
A
2,3-DPG
MDP
~
ATP
C
Results
A 31p-NMR spectrum at 101.3 M H z of erythroctyes in the oxygenated HBSS is depicted in Fig. 1A, while Fig. 1B shows a spectrum of erythrocytes deoxygenated to an oxygen saturation of 31%. Note the global downfield shifts of the resonances and the linebroadening especially evident in the 2,3-DPG peaks, Fig. 1B [27]. Fig. 1C is a 31p-NMR spectrum of erythrocytes preincubated in the inosine-pyruvate-phosphate medium for 4.5 h in order to induce supraphysiological levels of 2,3-DPG, while the ATP concentration remains at the physiological level [21]. Preincubation of erythrocytes in adenosine-pyruvate-phosphate medium induces high levels of both ATP and 2,3-DPG [22] as is evident in Fig. 1D, which is a 31p-NMR spectrum of erythrocytes preincubated in this medium for 4 h. The accumulation of organic phosphates results in a Donnan-mediated lowering of intraceUular pH, which in the spectra results in upfield shifts of the 2,3-DPG peaks and the 7-ATP peak, Fig. 1C and D [10]. At the same time the separation between the a- and fl-ATP peaks is increased as the level of [Mg2+]fre e is decreased, Fig. 1C and D [13].
Measurements of ADP, A TP and 2,3-DPG Quantitative measurements of ADP, ATP and 2,3D P G in human erythrocytes at two levels of oxygen saturation are shown in Table I. Measurements by 31p-NMR spectroscopy are compared to enzymatic measurements on the same samples. 2,3-DPG is fully 31p-NMR visible, both in the fully oxygenated and partially deoxygenated erythrocytes, Table I. Deoxygenation causes an increase in the concentration of triose-phosphates [28,29] and this may explain the slightly higher N M R determined 2,3-DPG value in the partially deoxygenated state, as the chemical shift of 3-phosphorus of 2,3-DPG is very close to that of the triose phosphates. Contrary to the full visibility of 2,3D P G , the 31P-NMR visibility of A T P is decreased to about 65%, again regardless of oxygen saturation, Table I. The differences between the 3tp-NMR measured values and the values determined by the luciferin-luciferase method are 0.74 to 0.78 m m o l / 1 cell water in absolute
D
I
20
'
10
'
0 ppm
--
'
-10
I
-20
Fig. 1. 31P-NMR spectra of human erythrocytes at 101.3 MHz and 25 ° C. An exponential multiplication factor resulting in a linebroadening of 20 Hz was applied to the FIDs before Fourier transformation. The spectra have been normalized to the reference peak of methylenediphosphonic acid (MDP). The peak assignments are: ATP, adenosinetriphosphate; 2,3-DPG, 2,3-diphosphoglycerate and Pi, inorganic phosphate. The top spectrum (A) is of oxygenated erythrocytes while the next spectrum (B) is of erythroctyes deoxygenated to an oxygen saturation of 31%. Note the global downfield shifts induced by deoxygenation. Spectrum (C) is of erythrocytes preincubated in inosine-pyruvate-phosphatemedium for 4.5 h in order to increase the concentration of 2,3-DPG. The bottom spectrum (D) is of erythrocytes preincubated for 4 h in adenosine-pyruvate-phosphatemedium thereby increasing the concentration of both 2,3-DPG and ATP. Note the upfield shifts of 2,3-DPG in (C) and (D), mainly caused by a lower intraceUular pH. In addition the chemical shift difference between the a-ATP and fl-ATP peaks is increased in both (C) and (D) due to a lower concentration of free intraceUular Mg2 + .
terms, Table I. Surprisingly the 31p-NMR determined A D P concentrations are about 400% higher than the values determined by the luciferin-luciferase method and corresponds to a difference of 0.3 m m o l / 1 cell water in absolute terms, regardless of oxygen saturation, Table I. This difference is comparable to the difference found in our previous study [12], although the absolute concentrations of A D P are lower in this study.
Measurements of M,_2 g +1]free The [ M g > ] freewas measured using the chemical shift difference between the a- a n d / 3 - A T P peaks [13,14]. The
172 TABLE I ADP, ATP and 2,3-DPG concentrations in human erythrocytes at two different levels of oxygen saturation
The erythrocytes were washed and resuspended in in Hepes buffered saline solution [20]. 31p-NMR determined concentrations are compared to determinations by luciferin-luciferase(ADP and ATP) and by an enzymatic method (2,3-DPG) on samples extracted by perchloric acid both before and after the NMR measurement. For 31P-NMR measurement of ATP, the fl-ATP peak was used, while the concentration of ADP was obtained by subtracting the fl-ATP peak from the combined fl-ADP and 3,-ATP peaks. The 3-phosphorus peak of 2,3-DPG was used to estimate the concentration of 2,3-DPG. The concentration is given in retool/1 cell water 5: S.E. (n = 8) assuming the water fraction of the erythrocytes to be 0.70. Oxygen saturation
ATP (mmol/1)
ADP (mmol/1)
2,3-DPG (mmol/1)
(%)
NMR
luciferase
NMR
luciferase
NMR
enzymatic
38.1 + 1.9
1.49 + 0.13 a
2.27 + 0.05
0.39 + 0.07 b
0.081 + 0.005
7.91 + 0.26
7.35 _+0.09
98.6+0.3
1.24 + 0.11d
1.98+0.10
0.41 +0.10 d
0.127+0.003
7.01 5 : 0 . 1 9
7.245:0.13
a p < 0.001, b p < 0.005, c p < 0.0005 and ,t p < 0.05 by paired two-tailed t-tests as compared to the values determined by luciferin-hiciferase.
results o b t a i n e d are s u m m a r i z e d in T a b l e II, where [Mg/+]rre ~ in whole b l o o d a n d in fully oxygenated a n d partially deoxygenated erythrocytes are shown. The resuits o n [Mg2+]fr~¢ m e a s u r e m e n t s are from the same erythroctyes as used in the q u a n t i f i c a t i o n of A D P , A T P a n d 2 , 3 - D P G depicted in T a b l e I. The 3 a p - N M R measured c o n c e n t r a t i o n s of A T P a n d 2 , 3 - D P G are also shown in the case of the isolated erythrocytes, T a b l e II. I n whole blood, the [Mg2+]er~e is 221 # m o l / 1 a n d c o m p a r a b l e to the 225 /~mol/1 in oxygenated erythro-
TABLE II 3~P-NMR measurements of intracellular free Mg 2+ in human erythrocytes
The intracellular free Mg 2÷ was calculated from the 31P-NMR spectra using the chemical shift difference between the a- and fl-ATP peaks [13,14] and assuming a dissociation constant of MgATP of 50 ttmol/1 at 25°C [24]. The results shown are from whole blood and isolated erythrocytes in a Hepes buffered saline solution [20] at two different levels of oxygen saturation (ox. sat.). In addition, results from erythrocytes preincubated in adenosine-pyruvate-phosphate medium (High ATP/2,3-DPG) and in inosine-pyruvate-phosphate medium (High 2,3-DPG) are shown. The concentrations of ATP and 2,3-DPG shown were determined by 31p-NMR and are expressed in mmol/l erythrocytes, while [Mg 2+ ]tree is expressed as /xmol/1. The results are expressed as means + S.E. with the number of determinations in parentheses.
cytes, T a b l e II. These results are similar to previously published values of [Mg2+]free in erythrocytes [1315,18]. I n contrast to the expected increase of [Mg2+]tree in deoxygenated erythrocytes [13], the result o b t a i n e d in erythrocytes partially d e o x y g e n a t e d to an oxygen s a t u r a t i o n of 38% is 196 # m o l / 1 , T a b l e II a n d Fig. 1A a n d B. 2 , 3 - D P G a n d A T P are q u a n t i t a t i v e l y the most imp o r t a n t Mg2+-complexing metabolites in the h u m a n erythrocyte [31]. By p r e i n c u b a t i n g the erythrocytes in i n o s i n e - p y r u v a t e - p h o s p h a t e m e d i u m [21] the c o n c e n t r a tion of 2 , 3 - D P G can be elevated 4 - 5 - f o l d , a n d b y p r e i n c u b a t i o n in a d e n o s i n e - p y r u v a t e - p h o s p h a t e m e d i u m the c o n c e n t r a t i o n s of b o t h 2 , 3 - D P G a n d A T P c a n be elevated 2 - 5 - t i m e s physiological levels [22]. These exp e r i m e n t a l procedures were used to produce erythrocytes with elevated c o n c e n t r a t i o n s of 2 , 3 - D P G a n d A T P in short term i n c u b a t i o n s ( < 5 h). A t high A T P a n d 2 , 3 - D P G levels the [Mg2+]fre e decreases to 6 1 / ~ m o l / l as expected, while the c o n c e n t r a t i o n of [Mg2+]fre ~ is lowered to 104 /~mol/1 in erythrocytes with n o r m a l levels of ATP, b u t 5-fold elevated levels of 2 , 3 - D P G (Table II a n d Fig. 1C a n d D).
Discussion 31p-NMR measurements of ADP, A TP and 2,3-DPG
ATP (mmol/l)
2,3-DPG (mmol/1)
[Mg2+ ]free (/~mol/1)
Whole blood
-
-
221 + 14 (10)
Erythrocytes ox. sat. =99%
0.89-t-0.11(8)
4.91+0.19 (8)
225+ 8 (8)
Erythrocytes ox. sat. = 38%
1.04+0.05 (8)
5.54+0.26 (8)
196+13 (8)
Erythrocytes High ATP/ 2,3-DPG
3.57+0.10(5)
24.7 +2.1 (5)
61+10 (5)
Erythrocytes High2,3-DPG
2.05+0.13(5)
25.4 +1.2 (5)
104_+ 7 (5)
Despite n u m e r o u s 3 1 p - N M R investigations of h u m a n erythrocytes, absolute q u a n t i f i c a t i o n s of A T P a n d 2,3D P G have only b e e n reported a few times [2,8]. Recently we reported that the 31p-NMR visibility of A T P is 78% when c o m p a r i n g the N M R m e a s u r e m e n t with a s t a n d a r d enzymatic m e t h o d [12]. I n order to investigate the influence of b i n d i n g of 2 , 3 - D P G a n d A T P to hemoglobin, we have repeated the m e a s u r e m e n t s in fully oxygenated a n d in partially d e o x y g e n a t e d erythrocytes as these p h o s p h o r y l a t e d metabolites are b o u n d preferentially to d e o x y h e m o g l o b i n [13,31,32]. 2 , 3 - D P G is fully 3 a p - N M R visible, while we c o n f i r m the decreased visibility of A T P , regardless of the degree
173 of oxygen saturation (Table I). In this study we find that the ATP visibility is decreased to 65% or slightly less than previously reported [12]. The mechanism behind this decreased visibility of ATP remains unclear as a systematic underestimation by the 31P-NMR procedure can be excluded by the full visibility of 2,3-DPG, Table I. Recently, the occurrence of 31P-NMR invisible ATP was reported in ischaemic rat liver, possibly related to mitochondrial ATP [33] but obviously this explanation cannot be used in the case of the erythrocyte. The fl-ATP peak used to quantitate ATP may be broadened by slow exchange of Mg 2÷ [34], although this explanation seems more unlikely because of the temperature and field dependence of the slow exchange situation [34] and by the fact that most ATP in the human erythrocyte is complexed to Mg 2+ [13]. The most plausible biochemical explanation for the decreased visibility is binding of ATP to the membrane of the erythrocyte, although further speculation is not warranted without experimental evidence. The 31p-NMR determined ADP value is unexpectedly 400% higher than the value determined by the luciferin-luciferase method, Table I. In absolute terms, the N M R determined value is 0.3 mmol/1 cell water higher, Table I, and this difference is of the same magnitude as previously reported, albeit at higher ADP concentrations [12]. The reason for the higher N M R determined ADP value is not clear although it might reflect an underestimation of the fl-ATP peak by 0.3 mmol/1, as ADP is determined by subtracting the flATP peak from the combined ~,-ATP and fl-ADP peaks [12]. This of course would increase the visibility of ATP, but only to about 80%. Measurements o f [ M g 2 +] free
Information regarding [Mg2+]free can be extracted from the 3~p-NMR spectrum by using the chemical shift difference between the a- and fl-ATP peaks [13,14]. Using this technique an association between lowered [Mg2+]free in erythrocytes and essential hypertension has been reported both in human erythroctyes [14] and in erythrocytes from spontaneously hypertensive rats [19]. Using the same method a recent study failed to confirm the association between essential hypertension and [Mg2÷]fr~ in human erythrocytes [18]. Furthermore, [Mg2+]fr~ has been reported to be lowered in stored erythrocytes [15-17], although the mechanism remains unclear. By decreasing oxygen saturation, [Mg2+]fr¢~ is expected to increase due to differential binding of the Mg 2+ complexed and uncomplexed ATP and 2,3-DPG to deoxyhemoglobin [13,31,32]. As shown in Table II, this expected increase does not occur when the measurement is based solely on the separation between the a- and fl-ATP peaks. In fact as shown in Table III, where the results of this study are compared to calculated levels from previously published
TABLE III Effect of deoxygenation of human erythrocytes on the 31P-NMR determined value of [Mg 2+]/tee
The 31p-NMR determined [Mg2+]fr¢¢ concentrations in human erythrocytesfrom three studies at different degreesof oxygenationare shown. In the study of Labotka [27] and of Gupta et al. [13] [Mg2+ ]free was calculated from spectral data given, and solely by use of the chemical shift difference between the a- and fl-ATP peaks without taking the hemoglobinbinding of ATP into account. As these studies were conducted at 37o C, the dissociation constant for MgATP used were 38 ttmol/l [13]. This study was conducted at 25°C and a dissociation constant for MgATP of 50/~mol/1 was used [24]. Oxygen saturation Oxygenated Deoxygenated
[Mg2+ ]free (#mol/l) 193 97
Reference 27 27
Oxygenated Deoxygenated
258 163
13 13
Oxygenated Ox. sat. = 38%
225 196
this study this study
spectra, deoxygenation tends to decrease the measured [Mg2+]rr~. This unexpected finding is probably due to the equation for q~ used (see Materials and Methods) where q , = [ A T P f ] / [ A T P t ] . Here ATP t denotes all species of ATP not complexed to Mg 2÷ and ATP t the total concentration of ATP [13]. Included in the free species of ATP is hemoglobin bound ATP, and as deoxygenation increases the hemoglobin binding of ATP relative to MgATP [13,31,32] the above fraction is increased. [Mg2+]free is thereby decreased when using Eqn. 3 to calculate the level, as it is assumed that all ATP is available for binding to Mg 2÷. Using a more complete analysis, where the differential binding of MgATP and ATP to hemoglobin is included, one may obtain an increase in [MgE+]fre e by deoxygenation using the spectral data and appropriate dissociation constants for ATP-hemoglobin complexes [13]. These results indicate that [Mg2+]free measurements in human erythrocytes using Eqns. 2 and 3 are biased by ATP binding to hemoglobin. A number of factors, such as oxygen saturation of hemoglobin, 2,3-DPG concentration and intracellular pH, are likely to influence the binding of ATP to hemoglobin [13,31,32]. A change of these factors may alter q~ and thereby the N M R measured [Mg2+]fr~ without a corresponding change in true [Mg2+]free or more likely a change that is in the opposite direction of the N M R measured. This may explain the contradictory results regarding an association between essential hypertension and [Mg2+]fr~ in human erythrocytes [14,18] and may even explain the apparently lowered [Mg2+]free in stored erythrocytes which so far lacks explanation [15-17]. In addition we have manipulated the concentrations of the major Mg2+-binding metabolites ATP and 2,3-
174 D P G to investigate the sensitivity of the measured [Mg2÷]free on these metabolites. As shown in Table II, increasing the 2,3-DPG concentration 5-fold induces a decrease in [Mg2+]fre e from 225 to 104/~mol/1, while a simultaneous increase of A T P 2-fold and 2,3-DPG 5-fold induces the [Mg2+]fre e to go down as low as 61/~mol/1. Although the perturbations of 2,3-DPG and A T P in this study are highly unphysiological, one may expect the [Mg2+]rree tO be dependent on the concentrations of ATP and 2,3-DPG at more physiological levels at any given total Mg 2÷ level.
Acknowledgements Anne Marie Jacobsen is thanked for technical assistance, while Soren Wamberg is thanked for help with equipment and advice. This research was partly supported by a junior research fellowship to A.P. from the Danish Research Academy.
References 1 Henderson, T.O., Costello, A.J.R. and Omachi, A. (1974) Proc. Natl. Acad. Sci. USA 71, 2487-2490. 2 Labotka, R.J., Glonek, T., Hruby, M.A. and Honig, G.R. (1976) Biochem. Med. 15, 311-329. 3 Kagimoto, T., Hayashi, F., Yamasaki, M., Morino, Y., Akasaka, K. and Kishimoto, S. (1978) Experimentia 34, 1092-1093. 4 Lam, Y.-F., Lin, A.K.-L.C. and Ho, C. (1979) Blood 54, 196-209. 5 Labotka, R.J. and Honig, G.R. (1980) Am. J. Hematol. 9, 55-65. 6 Sarpel, G., Lubansky, H.J., Danon, M.J. and Omachi, A. (1981) Arch. Neurol. 38, 271-274. 7 Swanson, M.S., Angle, C.R., Stohs, S.J., Wu, S.T., Salhany, J.M., Eliot, R.S. and Markin, R.S. (1983) Proc. Natl. Acad. Sci. USA 80, 169-172. 8 Zwerling, H.K., Diamond, J.N., Levy, G.C. and Clark, D.A. (1986) Magn. Res. Med. 4, 10-14. 9 Ambruso, D.R., Hawkins, B., Johnson, D.L., Fritzberg, A.R., Klingensmith III, W.C. and McCabe, E.R.B. (1986) Biochem. Med. Metabol. Biol. 35, 376-383. 10 Petersen, A., Jacobsen, J.P. and Horder, M. (1987) Scand. J. Clin. Lab. Invest. 47, 233-237.
11 Monti, J.P., Martinez, S., Braguer, D., Courtina, C., Francois, G. and Crevat, A. (1987) Clin. Chem. 33, 2284-2285. 12 Petersen, A., Pedersen, E.J. and Quistorff, B. (1989) Biochim. Biophys. Acta 1012, 267-271. 13 Gupta, R.J., Benovic, J.L. and Rose, Z.B. (1978) J. Biol. Chem. 253, 6172-6176. 14 Resnick, L.M., Gupta, R.J. and Laragh, J.H. (1984) Proc. Natl. Acad. Sci. USA 81, 6511-6515. 15 Bock, J.L., Wenz, B. and Gupta, R.J. (1985) Blood 65, 1526-1530. 16 Bock, J.L., Wenz, B. and Gupta, R.J. (1987) Biochim. Biophys. Acta 928, 8-12. 17 Bock, J.L. and Yusuf, Y. (1988) Biochim. Biophys. Acta 941, 225-231. 18 Woods, K.L., Walmsley, D., Haegerty, A.M., Turner, D.L. and Lian, L.-Y. (1988) Clin. Sci. 74, 513-517. 19 Matuura, T., Kohno, M., Kanayama, Y. et al. (1987) Biochem. Biophys. Res. Commun. 143, 1012 1017. 20 Petersen, A., Jacobsen, J.P. and Horder, M. (1987) Magn. Res. Med. 4, 341-350. 21 Deuticke, B., Duhm, J. and Dierkesmann, R. (1971) Pfliigers Arch. 326, 15-34. 22 Zachara, B. (1977) Transfusion 17, 628-634. 23 Lundin, A., Hasenson, M., Persson, J. and Pousette, ,~. (1986) in Bioluminescence and Chemiluminescence (Deluca, M. and McElroy, W., eds.), part B, pp. 27-42, Academic Press, New York. 24 Gupta, R.J., Gupta, P., Yushok, W.D. and Rose, Z.B. (1983) Biochem. Biophys. Res. Commun. 117, 210-216. 25 Bonnay, M.M., Legac, E.M., Brunier, A.P. and Bohuon, C.J. (1988) Clin. Chem. 34, 2154-2155. 26 F6ray, J.-C. and Garay, R. (1986) Biochim. Biophys. Acta 856, 76-84. 27 Labotka, R.J. (1984) Biochemistry 23, 5549-5555. 28 Flatman, P.W. and Lew, V.L. (1980) J. Physiol. 305, 13-30. 29 Hamasaki, N., Asakura, T. and Minamaki, S. (1970) J. Biochem. 68, 157-161. 30 Rapoport, I., Berger, H., Rapoport, S.M., Eisner, R. and Gerber, G. (1976) Biochim. Biophys. Acta 428, 193-204. 31 Bunn, H.F., Ransil, B.J. and Chao, A. (1972) J. Biol. Chem. 246, 5273-5279. 32 Gupta, R.J., Benovic, J.L. and Rose, Z.B. (1978) J. Biol. Chem. 253, 6165-6171. 33 Murphy, E., Gabel, S.A., Funk, A. and London, R.E. (1988) Biochemistry 27, 526-528. 34 Sontheimer, G.M., Kuhn, W. and Kalbitzer, H.R. (1986) Biochem. Biophys. Res. Commun. 134, 1379-1386.