Temperature independence of the alkaline Bohr effect in pig red cells and pig haemoglobin solutions

Biochimica et Biophysica Acta, 708 (1982) 105-111 Elsevier Biomedical Press

105

BBA 31347

TEMPERATURE INDEPENDENCE OF THE ALKALINE BOHR EFFECT IN PIG RED CELLS AND PIG HAEMOGLOBIN SOLUTIONS MARTINE SINET a, BRIGITTE BOHN b, PATRICK GUESNON b and CLAUDE POYART b,,

,7 Unitb 13 I N S E R M , H@ital Claude Bernard, 75019 Paris and b Unitb 27 I N S E R M , 42, rue Desbassayns de Richemont, 92150 Suresnes (France) (Received April 5th, 1982)

Key words: Hemoglobin; Oxygen affinity; CI

binding; Bohr effect," Temperature dependence," (Pig)

The influence of temperature on the oxygen affinity and the alkaline Bohr effect of pig red cells and pig heamoglobin solutions has been compared to that of human adult red cells and human adult hemoglobin. Pig red cells and pig llb evidence a lower affinity for oxygen in various conditions of pH, temperature and salt concentration, in the presence as well as in the absence of organic phosphates. It has been observed that the alkaline Bohr effect of pig Hb was reduced by 20-25% compared to Hb A0 and independent of changes in temperature, contrary to human Hb A o. Titrations of pig Hb with CI- indicate a lower heterotropic effect of this anion at low concentration of the salt. It is concluded that this may be the origin of the temperature independence of the alkaline Bohr effect in pig Hb. Conversely, the temperature dependence of the alkaline Bohr effect of Hb A o should be related to the oxygen-linked binding of CI- at the al-ct 2 interface.

Thermodynamic studies of ligand binding to the haems in human haemoglobin have been extensively performed [1-7]. These studies have brought out valuable information on the functional properties of haemoglobin and more specifically on the modulating effects of hetrotropic cofactors on the haem affinity for the ligand [3,8,9]. Numerous recent studies have underlined the exothermic nature of any electrostatic interactions within the tetramer, whether due to salt-bridge formation during the deoxygenation step or to oxygen-linked binding of anions such as chloride or organic phosphates [10,11]. These characteristics may be extented as well to proton binding. Because these heterotropic cofactors bind preferentially to deoxyheamoglobin [10-12], they compensate at least partially for the endothermic release of oxygen and vice-versa. The change in the

* To whom correspondence should be addressed. 0167-4838/82/0000-0000/$02.75 © 1982 Elsevier Biomedical Press

apparent heat of oxygenation (A Happ) of the haems in most mammalian haemoglobins is larger at pH 7 than at pH 8, larger in the presence of anions than in their absence and larger in the T than in the R quaternary structures [13]. Studies of haemoglobin function at varying temperature therefore constitute an important tool for understanding the precise tuning by heterotropic effectors, of oxygen binding or release. Physiologically this thermodynamic approach is also of interest not only in the study of the respiratory properties of cold-blood species such as fish [9,14,15], but also in mammalian species where temperature gradients of 10 to 15°C commonly occur between the core of the organism and its periphery [16]. A good example of this fine tuning is illustrated by the fact that the increase in oxygen affinity of blood with decreasing temperature is accompanied by an increase in the alkaline Bohr effect [17]. Are these thermodynamic effects of similar magnitude in all species, given the wide

106

variation in the protein structures, or are there differences which could illustrate adaptive mechanisms to various metabolic rates or to the environmental conditions? In the present work we report the results obtained in studying the effects of temperature on the functional properties of large white pig blood isolated haemoglobin solutions compared to those of the human adult haemoglobin (A0). The former attracted our interest, as we observed that the alkaline Bohr effect dit not vary with temperature, contrasting with what can be observed in most mammalian species, including man [2,5,17]. Detailed studies directed at the mechanism of this surprising behaviour have been interpreted according to the known primary structure of this Hb [18] and in the line of the debated mechanism at the origin of the alkaline Bohr effect in Hb A 0 [19]. Material and Methods

Blood from large white pigs was collected on heparin at the slaughter-house and kept in icewater until studied on the same day. Human adult blood was obtained from healthy non-smoking members of the laboratory. Pig Hb and Hb A o were isolated from minor components by DEAE-Sephadex chromatography using a linear gradient from pH 7.9 to 6.9 in 0.05 M Tris-HCl buffer. The isolated components were further stripped of organic and inorganic anions by ion-exchange chromatography (Dowex 70, AG50 1 X 8), concentrated under pressure and stored in liquid nitrogen until used. Much care was taken to remove completely the diphosphoglycerate from pig Hb solutions. This was best achieved by incubating for 24-36 hours the fresh pig red cells at 37°C prior to chromatography. This resulted in a decrease of the diphosphoglycerate content from 1.8 m o l / m o l tetramer (its normal value in the pig) to 0.2 to 0.3 m o l / m o l Hb 4 which could then be more easily stripped off the Hb during the ion-exchange. The purity of the isolated component was checked by isolectric focusing, which showed one single band migrating at p I 6.83 and 6.955 for pig and human Hb, respectively. The metHb content of the final stock solution was less than 2% as determined by usual spectrophotometric criteria.

Oxygen-binding curves in red cells suspensions were determined with the Hemox-Analyzer (TCS Medical Products, Southampton, PA, U.S.A.), according to methods already described [20,21]. This double-wavelength spectrophotometer allows precise and simultaneous recording of the changes in absorbance of the red cell suspensions or haemoglobin solutions with changes in Po2 on an x-y recorder. 50-60 ~1 of whole blood are added initially to 6 ml of 0.05 M Bistris-HC1 buffer or Tris-HC1 buffer (for pH values~ > 7.8) in 0.14 M NaC1 adjusted to the desired pH at the temperature of the measurements. The red cell suspension was first equilibrated under pure oxygen during 2-3 min and then deoxygenated slowly with pure humidified argon until zero Po2 is attained. Oxygen-binding curves in purified haemoglobin solutions were carried out with the same apparatus. To suit the light-path of the optical cell (1.2 cm) solutions of Hb were 60-70 ~tM on a haem basis. The compositions of the various buffers used is given in the legends of the figures and tables. The optical system and the humidified gases (either argon or oxygen in argon mixture) were thermostatted with a Haake K F3 thermostat (Haake, F.R.G). The temperature inside the optical cells was controlled with a thermoprobe with a accuracy of ±0.1°C. At the end of each experiment a sample of the solutions was withdrawn from the cell and its pH was measured with a Radiometer microassembly system connected to a P H M 62 meter (Radiometer, Copenhagen, Denmark) at the same temperature as that of the experiment. It was checked repeatedly that the final pH was identical to the pH of the initial buffer solutions. To ascertain our results in Hb solutions obtained with the Hemox-Analyzer two other checks were performed. Firstly a series of oxygen-binding curves in pig and human Hb were performed at a concentration of 0.8-0.9 mM in haem using the step-by-step equilibrating method with a Cary 219 spectrophotometer (Varian, U.S.A.). Similar values of Ps0 and cooperativity were obtained with the two methods. Secondly we checked for metHb formation in Hb solutions at the end of each run by measuring, in the oxygenated sample, the ratio A576.5/Aso O. No more than 5% of metHb was formed during the run in conditions where oxidation was maximum (pH

107

6.8, 30°C). All experiments showing a ratio A576.5/As00 lower than 2.85 were discarded. The following experiments were performed: in red cell suspensions, measurements of the change in logp50 with pH from 6.8 to 7.8 at 25, 30, 37 and 42°C, in isotonic saline plus 0.05 M Tris or BistrisHC1 buffer. In Hb solutions the same measurements were performed in various conditions of NaC1 concentration and/or diphosphoglycerate from pH 6.8 to 8.0, in Bistris or Tris-HC1 at 15, 20, 25 and 30°C. Logp50 and ns0, the index of haem-haem interaction at Ps0, were calculated from the Hill's equation. The value of the Bohr effect Alogps0/ApH was estimated from the slope of the regression line, in the pH range 7.0 to 7.8, where linearity may be assumed as a good approximation. Changes in AHapp, in various conditions of pH and salt concentration, were calculated from the slopes of the Van t' Hoff plots. Results and Discussion

Oxygen-binding isotherms in intact red cell suspensions Fig. 1 (part A) illustrates the temperature dependence of logp50 in pig and human fresh erythrocytes. The affinity for oxygen of pig red cells is lower than that of human red cells at each temperature level and with high and similar values of co-operativity (ns0 ;~ 2.7). These figures are in accordance with known values for pig blood reported by others [22-24]. It may be seen also from Fig. 1 that the slopes of the two lines are much different. (Alogpso/At) being 0.017 and 0.027 for pig and human red cells, respectively, indicating a 40% increase in the heat of oxygenation of pig erythrocytes compared to human red cells. This was not unexpected given that the concentration of diphosphoglycerate in adult pig red cells amounted to 1.83- 0.06 mol/mol Hb4, i.e., twice that of the human. The surprise came when we observed the same phenomenon in diphosphoglycerate-depleted red cells (~< 0.05 mol/mol Hb4) as shown in Table I. Surprising also was the reduction of the alkaline Bohr effect in pig red cells at each level of temperature and the marked reduction of its temperature dependence (Fig. l, part B). The slope of

A log PSO

1.5

1.0

30~ n SO

20

is

l

~

[]

o

I

I

20

3~0

/,0 t°C I

3;

,2

o

c B

ApH -0.6

-0.5

2's

3'o

3'7

oc

Fig. 1. A. Temperature dependence of logps0 in adult pig (D) and adult human (m) fresh red cells suspensions. Conditions were: 0.14 M NaCI/0.05 M Bistris, pH 7.4. Oxygen binding isotherms were measured with the Hemox-Analyzer. Temperature was controlled within 0.1°C. The inset shows the values of the apparent cooperativity for the two series of experiments. Bars indicate ± 1 S.E. of the mean value. B. Variations of the alkaline Bohr effect calculated from the changes in logp5 o with pH at varying temperatures in human and pig fresh red cells suspensions, pH was varied at each temperature level from 6.8 to 7.8. Conditions as in panel A.

AlogPs0/ApH over t (°C) was approximately halved in pig blood compared to human blood. Because the red cells constitute a complex environment where changes in intracellular pH with

108

,o~~

TABLE 1 VARIATIONS OF THE APPARENT HEAT OF O X Y G E N A T I O N IN PIG E R Y T H R O C Y T E S C O M P A R E D TO H U M A N E R Y T H R O C Y T E S

/ /¢ /

05

Values shown are A H , pp in k c a l / m o l Hb 4. 2,3-DPG, 2,3-diphosphoglycerate. Pig

//o

/1

Human

pH

Fresh

2,3-DPGdepleted

Fresh

2,3-DPGdepleted

7.0 7.2 7.4 7.6 7.8

-6.8 -%0 -7.2 -7.3 -7.5

--7.5 -7.9 -8.3 -8.7 -9.1

-10.8 --11.3 --11.7 --12.1 -12.6

-11.6 --12.5 --13.3 14.2 -15.1

elevated concentration of diphosphoglycerate may affect the significance of the Bohr factor calculated from extracellular pH values [25] we carried out similar experiments in purified Hb solutions where the milieu could be accurately controlled.

Oxygen-binding isotherms to &olated pig and human Hb with varying temperatures and pH Fig. 2 shows results comparable to those obtained in red cells suspensions. Pig Hb has a lower affinity for oxygen than Hb A 0, while keeping normal cooperativity in the absence of organic phosphates. Similarly the influence of temperature on Ps0 was lower in pig Hb than in Hb A 0. This is shown in Table II from the values of AH, pp for the two Hb and in conditions of varying chloride and diphosphoglycerate concentration. The low value of A Happ in pig Hb at low chloride concentration (first column in Table II) indicates that the difference in the apparent heat of oxygenation between the two Hb is not exclusively related to differences in anion binding, but also to additional electrostatic bonds in the pig tetramer. This was confirmed by the lower value of K l, the equilibrium constant of oxygen-binding in the T structure, of pig Hb at pH 9 and in 10 mM NaC1/TrisHC1 buffer, which was 0.148 mmHg ~ and 0.232 mmHg t for pig and human Hb, respectively, at 25°C. Fig. 3 demonstrates that the value of the alkaline Bohr effect of purified Hb solutions is lower in pig Hb than in Hb A 0 and independent

/a

/

15

20

25

30 t'C

-05

0

05

10 log P02

Fig. 2. Left: Hill plots of oxygen-binding curves for purified stripped haemoglobin solutions, showing the low affinity of pig haemoglobin ([3) compared to h u m a n Hb A 0 ( l l L The slopes of the Hill plots at Ps0 were 2.64 and 2.58, respectively. Conditions: 0.01 M NaCI/0.02 M Bistris buffer, pH 7.0, at 25°C, 70 /~M haem, 12-mm cell light-path. Inset: the temperature dependence of the oxygen affinity of purified pig and h u m a n Hb, to be compared with panel A in Fig. 1. Conditions as described for the left panel.

of the changes in temperature. Because the reduction of the alkaline Bohr effect in pig Hb amounted to approx. 20% at 25°C, 0.1 M NaC1 when compared to that of Hb A 0, we postulated that, according to the theory proposed by Perutz [12,19], this might be due to the loss of one Bohr group and more precisely to the disappearance of the Val-1-a-oxygen-linked Bohr group. However, upon examination of the primary structure of pig Hb reported by Braunitzer et al. [18] it is seen that all

T A B L E II C H A N G E S IN AHapp IN PIG A N D H U M A N A D U L T H A E M O G L O B I N IN T H E PRESENCE OF C I - A N D DIPHOSPHOGLYCERATE AH,pp values are expressed in k c a l / m o l Hb 4. DPG, diphosphoglycerate.

Hb A o PigHb

0.01 M C1-

0.1 M C1

0.01 M C1- + 0.5 m M D P G

-- 17.6 -12.6

-- 13.2 -11.2

- 10.9 -8.8

109 i

w

!

!

log PS0

A log P50 &pH 1.5

_--

-0.6 j

j/

--.-a

~

1.0

-0.5 [ ] []

l's

'1

[]

[]

25

2's

3'0 toc

Fig. 3. Temperature dependence of the alkaline Bohr effect in pig (r-q) and human (11) Hb. Conditions were: 0.1 M NaCI, 0.05 M Bistris-HCl or Tris-HCl buffer, 70 #M haem. The Bohr effect was calculated in the pH interval 7.0 to 7.8 where changes in logp50 vary linearly with pH to a good approximation.

the amino acid residues responsible for the acid and alkaline Bohr effects in Hb A 0 are present in pig Hb. Moreover, all the residues known from crystallographic data to be in electrostatic interactions with His-146-fl, the major alkaline Bohr group, or responsible for the oxygen-linked binding of chloride or diphosphoglycerate in Hb A 0 are present as well in pig Hb. Upon more careful examination of our results we found, however, some indication that chloride-binding to pig Hb might be different from to Hb A 0. This is indicated in column 2 of Table II which shows that the change in AH, pp upon addition of 0.1 M chloride at pH 7.0 was much lower ( + 1 . 4 kcal/mol) in pig Hb than in Hb A 0 ( + 4 . 4 kcal/mol). By contrast, the change in AHap p upon addition of a 30-fold molar excess of diphosphoglycerate was approximately the same for the two Hb (columns 2 and 3). That the oxygen-linked binding of diphosphoglycerate was similar in pig and human Hb is illustrated in Fig. 4. This shows also that the Alogp5 o p i g - human Hb was maintained over the whole range of diphosphoglycerate concentration, thus supporting the view of a greater

~

1'o

2 3 DPG

~,1/M Hb/,

3'o

Fig. 4. The dependence of the oxygen affinity of pig (rq) and human (11) Hb upon diphosphoglycerate [DPG] Na salt (mol/mol Hb4). 70 /~M haem, 0.1 M NaC1, 0.05 M Bistris buffer, 25°C, pH 7.0. The abcissa is on a log scale.

intrinsic stability of pig Hb over that of Hb A 0 and independent of heterotropic cofactors. Titrations of pig Hb and Hb A 0 with chloride at constant pH were then performed at different temperatures. The results are shown in Table III. Whereas the amount of oxygen-linked chloride (Alogps0/A[C1-]) decreases as expected with increasing temperature in Hb A 0 solutions, it was independent of temperature in pig Hb. The complete chloride titration curve at 15°C is given in Fig. 5 together with its first derivative (inset of the figure). This indicates that the amount of oxygen-linked chloride in pig Hb is mostly suppressed at low concentration of the anion, which suggests that the abnormality lies at the chloride high-affinity binding site [26,27]. This site has been attributed to the a-amino group of the Nterminal valine of both a-chains, interacting with the guanidinium group of the C-terminal arginine TABLE III OXYGEN-LINKED BINDING OF CIValues shown are ~ logp50/A [C1- ], expressed as mmol/mmol haem oxygenated.

Hb A o Pig Hb

15°C

25°C

30°C

0.52 0.44

0.47 0.48

0.36 0.44

110

Io9 P50

,az//

05

i

/ ,/~{

o/

. t~ Leg PSO

..,,,,,

7,

fect in human Hb A 0 is rehited to the oxygen-linked binding of C1- at the Otl-a2 site. Support for this conclusion may be found also in the similarity of alkaline Bohr effect measured by differential titration of protons at 4°C and 25°C in Hb A 0 solutions at low concentration of chloride (5 raM) [27,30]: (3) our results confirm the important role of C1- in regulating haemoglobin function in mammals, as abundantly demonstrated by de Bruin et al. [31].

Acknowledgments Fig. 5. Titration of purified pig ([]) and human (11) Hb with CI- at 15°C, pH 7.0, 70 /~M haem. The inset represents the first derivative of the two curves, illustrating the shift towards the right of the chloride-binding curve in pig Hb compared to that of Hb A 0.

We are pleased to acknowledge the helpful comments of M.F. Perutz. We thank also A.M. Aubertin and J. Grellier for secretarial help. This work has been supported by funds from INSERM.

References of the opposite sister chain [28]. How can we reconcile our results with the known structure of pig Hb where both Val-l-a~ and Arg-141-e~ 2 are present as in Hb A0? There is in the sequence of the residues of the pig a-chains one substitution which may be the clue to the peculiar behaviour of pig Hb with changes in temperature, that is, the replacement of Serine-131-a (human) by asparagine in pig a-chains. Arnone and Williams [28] have described an electrostatic interaction between the hydroxyl group of Ser-131-a and the N-terminal Val-1-a in the same chain and through a small anion such as C1-. This non-oxygen-linked bridge was labelled X~- on their original scheme [28,29]. Because the presence of an amide group in Asn-131-a in place of the hydroxyl group of serine may lower the pK value of Val-l-a in the deoxy form (or increase it in the oxy form), we postulate that this can lead to an inhibition of the oxygen-linked binding of C1 at the Otl-Ot2 interface. We conclude (1) that the alkaline Bohr effect in pig blood and pig hemoglobin is inhibited by 20-25% due to the decreased participation of the Val-1-a Bohr group and related to an inhibition of the amount of oxygen-linked chloride binding: (2) due to the exothermic nature of CI- binding to deoxyhemoglobin this would explain the temperature independence of the alkaline Bohr effect in pig haemoglobin. Conversely it follows that the temperature dependence of the alkaline Bohr ef-

1 Astrup, P., Engel, J., Severinghaus, J.W. and Munson, E. (1965) Scandinav. J. Clin. Lab. Invest. 17, 515-523 2 Rossi-Bernardi, L. and Roughton, F.J.W. (1967) J. Biol. Chem. 242, 784-792 3 Imai, K. and Tyuma, I. (1973) Biochem. Biophys. Res. Commun. 51, 52-58 4 Atha, D.H. and Ackers, G.K. (1974) Biochemistry 13, 2376-2382 5 Hlastala, M.P., Woodson, R.D. and Wranne, B. (1977) J. Appl. Physiol. 43, 545-550 6 Gill, S.J., Gaud, H.T. and Barisas, B.G. (1980) J. Biol. Chem. 255, 7855-7857 7 Reeves, R.B. (1980) Respir. Physiol. 42, 317-328 8 Hedlund, B.E., and Lovrien, R. (1974) Biochem. Biophys. Res. Commun. 61,859-867 9 Barisas, B.G. and Gill, S.J. (1979) Biophys. Chem. 9, 235244 10 Benesch, R.E., Benesch, R. and Yu, C.I. (1969) Biochemistry 8, 2567-2571 11 Ima'i, K. and Yonetani, T. (1975) J. Biol. Chem. 250, 7093-7098 12 Perutz, M.F. (1970) Nature (London) 228, 726-739 13 Imai, K. (1979) J. Mol. Biol., 133, 233-247 14 Saffran, W.A. and Gibson, Q.H. (1979) J. Biol. Chem. 254, 1666-1670 15 Mayo, K.H. and Chien, J.C.W. (1980) J. Mol. Biol. 142, 63-73 16 Castaing, M., Bursaux, E. and Poyart, C. (1982) Eur. J. Biochem. 121, 573-578 17 Antonini, E., Wyman, J., Brunori, M., Fronticelli, C., Bucci, E. and Rossi-Fanelli, A. (1965) J. Biol. Chem. 240, 10961103 18 Braunitzer, G., Schrank, B., Stangl, A. and Scheithauer, U. (1978) Hoppe Seylers' Z. Physiol. Chem., 359, 142-146 19 Perutz, M.F., Kilmartin, J.V., Nishikura, K., Fogg, J.H.,

ill

20 21 22 23 24 25 26

Butler, P.J.G. and Rollema, H.S. (1980) J. Mol. Biol. 138, 649-670 Azakura, T. (1979) Crit. Care Med. 7, 391-395 Wajcman, H., Aguilar i Bascompte, J., Labie, D., Bohn, B. and Poyart, C. (1982) J. Mol. Biol. 156, 185-202 Baumann, R., Teischel, F., Zoch, R. And Bartels, H. (1973) Respir. Physiol. 19, 153-161 Tweedale, P.M. (1973) Respir. Physiol. 19, 145-152 Kim, H.D. and Duhm, J. (1974) Am. J. Physiol. 226, 1001-1006 Duhm, J. (1971) Pfliigers Arch. 326, 341-356 Chiancone, E., Norne, J.E., Forsen, J., Bonaventura, J., Brunori, M., Antonini, E. and Wymart, J. (1975) Eur. J. Biochem. 55, 385-390

27 Rollema, H.S., de Bruin, S.H., Janssen, L.H.M. and Van Os, G.A.J. (1975) J. Biol. Chem. 250, 1333-1339 28 Arnone, A. and Williams, D.J. (1977) in Molecular Interactions of Hemoglobin (Labie, D., Rosa, J. and Poyart, C., eds.), Vol. 70, pp. 15-22, Institut National de la Sant6 et de la Recherche M~dicale, Paris 29 O'Donnell, S., Mandaro, R., Schuster, T.M. and Arnone, A. (1979) J. Biol. Chem. 254, 12204-12208 30 Poyart, C.F., Guesnon, P. and Bohn, B. (1981) Biochem. J. 195, 493-501 31 De Bruin, S.H., Rollema, H.S., Janssen, L.H.M. and Van Os, G.A.J. (1974) Biochem. Biophys. Res. Commun. 58, 210-215