Synergistic modulation by chloride and organic phosphates of hemoglobin from bear (Ursus arctos)

Synergistic modulation by chloride and organic phosphates of hemoglobin from bear (Ursus arctos)

J. Mol. Biol. (1994) 236, 1401-1406 Synergistic Modulation by Chloride and Organic Phosphates of Hemoglobin from Bear (Ursus arctos) Massimo Colettal...

611KB Sizes 1 Downloads 40 Views

J. Mol. Biol. (1994) 236, 1401-1406

Synergistic Modulation by Chloride and Organic Phosphates of Hemoglobin from Bear (Ursus arctos) Massimo Colettal, Saverio G. Condo2, Roberto Scatena2, Maria E. Clementi Silvia Baroni3, S&i N. Sletten4, Ole Brix4 and Bruno Giardina3t ‘Department of Molecular, Cellular and Animal Biology University of Camerino, Via F. Camerini, 2-62032 Camerino (MC),

Italy

2Department of Experimental Medicine and Biochemical Sciences University of Rome “Tor Vergata” Via 0. Raimondo-00173 Rome, Italy ‘CNR Center for the Chemistry of Receptors and Institute of Chemistry and Clinical Chemistry Faculty of Medicine, Catholic University “Sacro Cuore” Largo F. Vito, l-00195 Rome, Italy 4Zoological Laboratory, University of Bergen Allegaten 41, N-5007 Bergen, Norway The oxygen binding properties of hemoglobin (Hb) from brown bear (Ursu.s arctos) have been studied focussing on the effect of heterotropic ligands, and the behaviour has been compared with that of human HbA, taken as a prototype of mammalian Hbs. It has been observed that in bear Hb chloride ions and 2,3-diphosphoglyceric acid (Gri(2,3)P,) can modulate the oxygen affinity in a synergistic way such that their individual effect is enhanced whenever they are both present in saturating amounts. The thermodynamic analysis of such a feature indicates that in bear Hb there are two classes of chloride binding sites, one acting synergistically with Gri(2,3)P, and another one, which likely overlaps with the organic phosphate interaction cleft, and is therefore fully operative only in the absence of Gri(2,3)P,. The behaviour of the last site is similar to that observed in human HbA, where the effect of Cl- and Gri(2,3)P, is mutually exclusive. The interaction energy between chloride and Gri(2,3)P, synergistic binding sites appears to be O,-linked so that the interplay may have a relevant physiological role in modulating the oxygen transport in brown bear. This behaviour is associated with a marked pH-dependence of the oxygenation enthalpy in bear Hb, such that under acidotic and hypercloruremic conditions, oxygen supply to peripheral tissues could be maintained essentially unaltered even under low temperature conditions. Keywords:

oxygen binding; heterotropic effects; enthalpy of oxygenation; mammalian hemoglobins; acidosis and hypercloruremia

mational changes (Brunori et al., 1985). In this respect, it is worth pointing out that the product of Fulfilment of O2 transport function has pushed the central exon is enough for accomplishing the hemoglobins (Hb$) to develop in the course of main function of O2 binding, whereas the products evolution a common molecular mechanism based on of the two external exons add the potentiality for an the principles of ligand-linked quaternary conforexternal modulation of the functional properties (Eaton, 1982). Hence, O2 binding at the hemes is t Author to whom all correspondence should be modulated in different species to a different extent addressed. by functional interactions with groups reacting with $ Abbreviations used: Hb, hemoglobin; Gri(2,3)P,, solvent components. 2,3-diphosphoglyceric acid; bis-Tris, Within this overall scheme, hemoglobins exhibit a 2-[bis(2-hydroxyethyI)amino]-2-hydroxymethylgreat deal of variation in terms of absolute affinities 1,3-propanediol. 1401 0 1994 AcademicPressLimited OO22-2S36/94/101401-06 $08.00/O 1. Introduction

Modulation

1402

of Heterotropic Effects in Bear Hb

for Oz and in their susceptibility to metabolic effectom, being all finalized to fully meet, the physiological requirements of a given species. This type of functional tuning is developped to a high degree in all mammalian hemoglobins and is primarily based on the ability of the effector to preferentially bind one of the two quaternary conformations. In partibinding of anions, such as cular, preferential chloride and Gri(2,3)P,, to the low O2 affinity T state brings about an easier delivery of oxygen to the respiring tissues. Tn this respect, it seems important to analyse in different species the effect of chloride ions, which are known to be present in large amounts in the blood (~100 mM), and the possible interplay with the main erythrocytic effector, i.e. Gri(2,3)P, (De Bruin & Janssen, 1973; Perutz et al., 1980). Moreover, it may be worthwhile to recall that recent results (Giardina et al., 1990; Coletta et al., 1992) have outlined the role of temperature and its interplay with heterotropic ligands in the regulation of Hb function in several mammals. In view of these considerations, the functional cha.ra.cteriza.tion of hemoglobin from bear (Ursus nrctos) here reported has been performed focussing on the possible interplay between chloride ions? Gri(2,3)P, and temperature. Further,

a parallel

investigation

aspects

of this regulatory

I

I

7.0

8.0

PH

has been carried

out on human HbA, allowing consideration tional

log Pm

of addi-

mechanism.

The results appear of particular evolutionary interest and outline new aspects of the interactions between relevant

heterotropic physiological

ligands that importance.

2. Materials

may

be

of

and Methods

Blood samples were drawn. under anaesthesia. from t,he cephalic vein of a brown bear (CTr.uu.sawlos) aged 22 months, in the Zoological Garden of the LJniver&y of Oulu (Finland). Red blood cells were washed 3 times wit,h isotonic KaCI solution by rentrifugation at 3000 revs/min and the packed cells lysed by adding 2 volumes of cold hypotonic phosphate buffer. Stroma were removed by centrifugat,ion at 10.000 revs/mm for 30 min. Removal of organic phosphate and inorganic ions \vas obtained by passing the hemolysate first through a Sephadex C-25 column. equilibrated with 0.01 M Tris. HCI (pH 8.0). 91 M NaCI, and then by recycling the protein solut,ion through a column of mixed-bed ionexchange resin (Bio-Rad AC 501 X8). Solutions of Gri(2.3)P, were prepared hy dissolving the, sodium salt of 2%diphosphoglyceric acid (Sigma) in water or in the appropriate buffer. Oxygen binding isotherms have been determined by the tonometric method (Antonini & Brunori. 1971: Giardina & Amiconi. 1981): in the absence and in the presence of allosteric effecters, between 20°C and 40°C. The overall oxygenation enthalpy (6H, kcal/mol; 1 kcal = 4184 ,I). corrected for the solubilization heat of 0, (= -3.0 kcal/moJ; see Antonini & Brunori, 1971), has been calculated from the integrated van’t Hoff equation: 6H = ((T,T,)/(T,-T,))6Inp,/1000, where p, is the median pressure of the ligand (Wyman. 1964). Over the temperature range explored between 20°C

$5

i0

8il

PH Figure 1. Effect of pH on the oxygen affinity (in trrms of log 11~) of brown hear Hh (A) and human HI) (13) at 20°C’. Buffer conditions: 0.1 $1 bis-Tris or Tris. H(‘I l)lus tbl M Sac’1 in the absence (open circles) and in thr Jupsrnrr (optn t,rianglrs) of 3 mM Cri(2.3)Pz. Filled circalrs rcbprrsrnt rsptriments performed in 0.1 >I chloride-frc,r Hrpes buffer plus different concentrations of (:ri(2.3)1’2 which are saturating for HhO, at rash pH as it has bren calculated from sets of experiments as those reJ)ortetl in Figure 2. For further experimental details. see the text.

and WY’. van’t Hoff plots were linrar within the rxperimental error. An average standard deviation of +89, fn~ values of p, and of f l5”/e for 6H values \vcre calculated.

3. Results and Discussion Figure 1 shows the 0, Bohr effect of bear Hh and human HbA under diiferent experimental conditions. The occurrence of a synergistic effect between Gri(2,3)P, and Cl- in the modulation of 0, affinity of bear Hb shows up very clearly. Thus. in this Hb

Modulation

of Heterotropic Effects in Bear Hb

-log M

m

.. 4

i

6

--4

i

b

-log M Figure 2. Effect of ~hloritle ions (open virdes) and (01~~~triangles) concentration on the oxygen affinity of Lyman hrar HI, (A) and of human HbA (B) in (PI M Hrl~s buffer I)H ‘7.Z at ZO’Y’. Fillet1 symI~ols represent: (circlrs) the effWt of chloride ions on log pm in the= Imwnw of saturating amounts of 0.1 $1 Cri(“.S)l’,: and (triangkls) thr effect of (:ri(2.3)P, on logp, in the preswxv of saturating amounts of (‘I- (2 M XaC’I). The continuous lines are ol,tainrrl by fitting experimental data to N(Il (I ). (:ri(2.9)P2

the Gri(2,3)P,-linked increase of p,,, is further enhanced, t,hough t)o a different extent for various pH values, by the presence of chloride ions (see Figure 1A). This appears more evident from the comparison with the behaviour observed in the case of human HbA (Figure 113). As a matter of fact, in t.he case of human HbA t.he effect observed for Gri(2.3)P, alone (at saturating concentrations) on the oxygen affinity (i.e. log p,) is not further enhanced by the addition of 01 $1 Cl-. A similar behaviour is observed essent,ially for all other mammalian hemoglobins and it differs strikingly from that found in the case of bear Hb (Figure 1A), which stands alone up to now in displaying a mutually synergistic effect between Cl- and Gri(2,3)P,. The maximum amplitude of the synergistic effect is attained on the acidic limb of the alkaline Bohr effect (Figure 1A). However, it must be pointed out that the dependence of log p, reported in Figure 1 refers to a constant, concent,ration of Gri(2,3)P, (=3 mM), a value that might be insufficient to saturate also the oxygenated form at all pH values. Therefore, the apparent decrease of the synergistic effect on the alkaline side of the Bohr effect curve does not necessarily imply that the phenomenon is

1403

directly proton-linked, but it may be simply related to the pH-dependence of the Gri(2,3)P, effect, as pointed out also for Gri(2,3)P, binding to human HbA (Antonini et al., 1982). On the other hand, at pH I 7.5 the extent of the synergistic effect appears constant throughout the pH range investigated with a dlog p, z 030 (see Figure l), suggesting that the eventual groups involved in the phenomenon should have a pK, >> 7.5. In order to shed more light on this phenomenon, especially in relation to its potential physiological relevance, we have analysed in detail such a synergistic effect of Cl- and Gri(2,3)P, on the oxygen binding properties of bear Hb and of human HbA at pH 7.3, a pH value close to the physiological one and at which the synergistic effect is already fully operative. Figure 2A displays the dependence of the oxygen affinity for bear Hb on the concentration of either Cl- and Gri(2,3)P, alone, or when the other is already present in saturating amounts (i.e. 2 M Clor 0.1 M Gri(2,3)P,). The synergistic effect appears evident in the right portion of Figure 2 where addition of either Cm or Gri(2,3)P, when the partner ligand is already present in saturating amounts brings about a further decrease of oxygen affinity tha.t is dependent on the concentration of the added heterotropic effector. This feature indicates the existence of binding sites different for Cl- and Gri(2,3)P, allowing simultaneous interaction with the protein without negative interference. Furthermore, it suggests that in this Hb the two classes of heterotropic interacting sites have direct and separat.e communication pathways with the heme and the observed phenomenon is not a pseudo-linkage (Wyman, 1968). In the case of human HbA the same experiment gives a completely different result (Figure 2B). Thus, in this case the addition of Gri(2,3) in the presence of saturating amounts of C1- (i.e. 2 M) does not show any increase of log p,, and the same is true also for the addition of Cl- in the presence of saturating amounts of Gri(2,3)P, (i.e. 30 mM: Figure 21~). This behaviour displayed by human HbA suggests that Cl- and Gri(2,3)P, share the same O,-linked binding site such that their effect is mutually exclusive. It is interesting to outline, as previously reported, tha,t, unlike that of other animals living in cold environments (Giardina et al., 1990; Coletta et al., 1992), bear Hb displays a pH-dependent exothermic oxygenation enthalpy (Figure 3), with values at pH 7.3 (6H = -293 kJ/mol, see Table 2) that are intermediate between those of human HbA (6H N -42 kJ/mol) and of ruminant Hbs (6H N - 12 kJ/mol: Coletta et al., 1992). Data reported in Figure 2 allow a further analysis (both fol deoxyHb and oxytlb) of Cl- and Gri(2,3)P, effects in human HbA and bear Hb. In the case of bear Hb it is possible to calculate the same parameters for either heterotropic effector in the presence of saturating amounts of the partner ligand (see the left portion of both panels in Figure

1404

Modulation

of Heterotropic

Effects

in Bear Hb

.4

Table 1 Values of the parameters for binding of Gri(2,3)P, and Cl- to bear Hb and human HbA, 01 M Hepes (pH 7.3) at 20°C according to scheme II

.

DeoxyHb

3

A. Human HbA

K, 0-l) K, (M-*1

E n 5

13. Bear Hb K, (M-l) K, W-‘1 K, Of-‘) K, Of-‘)

2

K, (M-l) t Calculated and K,,K,.

3.2

3.3

3.4

Figure 3. van’t Hoff isochores of brown bear Hb in @I M bis-Tris or TrisHCl buffer plus O-1 M NaCl and 3 mM Gri(2,3)P, at pH 68 (open circles), pH 7.0 (filled circles), pH 7.3 (open triangles) and pH 7.6 (filled triangles).

5

P+Gri(2,3)P, ci-

+Gri(2,3)P,

minimum

P-Gri(2,3)P, + Cl-

I K; P-Gri(2,3)P,-Cl-

scheme:

(Scheme I)

KG where P can be either deoxyHb or oxyHb, and Kc (r=d or o; y=l, 2, 3 or 4) are the association equilibrium constants for the binding of heterotropic ligand to deoxyHb (= K$ or to oxyHb (= KF). The behaviour of data reported in Figure 2 can be described by the following equation: logp,

336fO.79 1.02( +063) x IO2

l@(f0.16) x 10’ 152( kW3.5) x IO5 1.25( kO.55) x IO3 1.16 x 1O’t 1.64x0.16)x 10’

6+30( * 1.20) x 1.40( kO.54) x 7.70( * 4.20) x 1.21 x 104t 880( & 1.20) x

according

to the crossed relat.ionship

(K;K;)/(K;K;)

l/T4?

2), according to the following

207f33 444( f 1.22) x IO5

conservation principle requires between the four independently meters, such that:

1

IP-Cl-6

OxyHb

= logpi+

Rlog((l +K; [LI)/(l +K;[W,

(1)

where logp, is the observed oxygen affinity, logpz is the oxygen affinity of human HbA and of bear Hb in the absence of heterotropic ligands under consideration, and R is the ratio between the number of binding sites for the ligand and the number of subunits. In the case of human HbA, the behaviour observed (see Figure 2B) allows only the determination of two equilibrium constants for scheme I, namely K; and K; (see Table 1). Conversely, in the case of bear Hb, the possibility of measuring all four overall binding parameters is a unique experimental opportunity and it represents a stringent test for the applicability of the minimum mechanism reported in scheme I, since in this case the energy

100 102 IO2 loo

K, K,/K2

a relationship measured para-

= M = 1.

On the basis of these considerations, the analysis indicated that scheme I is unsatisfactory and a mechanism more complex must be taken into consideration. The simplest extension of scheme 1 may consider the possible existence in bear Hb of two classes of binding sites for Cl-, one of which is overlapping with the binding cleft of Gri(2.3)Pz (thus competing with it), whereas the other one is positively interacting with the polyphosphate binding site. Hence, the scheme could be rewritten as follows: P+Gri(2,3)P, + Cll

fi;.

P-Gri(2,3)Pz + Cl-

P-Cl- +Gri(2,3)P, + Cl-

‘K;

P-Cl--Gri(2,3)P,

I K;

I K;

P-(cl-),

I K;

4

(Scheme II)

However, it must be remarked that the dependence alone does not of 1% Pnl on Cl- concentration display a significant heterogeneity for the two chloride ion binding sites (see Figure 2). Therefore, the chloride binding sites are represented as two classes only from a topological standpoint (see Bonaventura et al., 1976; Nigen et al., 1980) but are not functionally heterogeneous, and we can reasonably attribute them the same equilibrium binding constant in the absence of the other effector (i.e. KT = K’;). In such a case, applying equation (1) in the analysis of functional data a value of R = O-25 has been employed for Gri(2,3)P, binding and of R = 95 has been used for Cl-. On the basis of these considerations, the data reported in the right portion of Figure 2 under saturating amounts of either one of the effecters can be rationalized in the

Modulation

of Heterotropic Effects in Bear Hb

following way: (1) the dependence of log p, on Clconcentration in the presence of saturating amounts of Gri(2,3)P, refers only to the reaction 3 in scheme II (since only one class of site is available for Clunder these conditions); (2) the dependence of log Pill on Gri(2,3)Pz in the presence of saturating amounts of Cl- concerns reactions 4 and 5 (since binding of Gri(2,3)P, requires the displacement of Cl- from the second class of binding sites). The calculated equilibrium constants are reported in Table 1. An internal check of the applicability of scheme IT to the behaviour of bear Hb can be performed calculating the value of the parameter M (see above), which now turns out to be 1, as expected according to the energy conservation principle. The parameters calculated according to scheme IT deserve further comments. First of all, it must be remarked as the affinity of chloride ions is approximately tenfold higher in deoxy bear Hb than in deoxy human Hb, whereas for the oxygenated derivat,ive the affinity difference decreases significantly (see Table 1). Such a difference is not observed in the case of Gri(2,3)P,, which displays in both proteins similar values within the limit,s of uncertainty. The final result is that both in the absence and in the presence of saturating amount of either one of t,wo effecters, the 0, affinity of bear Hb is somewhat higher than for human HbA. Moreover, addition of one heterotropic effector in the presence of the other one brings about in bear Hb a further decrease of oxygen affinity, which then becomes closely similar to that displayed by human HbA (see Figure 2). Therefore, the same functional effect is obtained in human HbA by the addition of only one effector, and in bear Hb after t,he addition of both effecters. The O,-linked binding of inorganic anions to human Hb is well known (Tmai, 1982). In particular, there is considerable evidence for the existence of three main binding sites per a/? dimer: (1) an intrasubunit site between the a-amino group of Vallp and the side-chain of Lys82p (Arnone, 1972); (2) an int,rasubunit site between the a-amino group of Valla and the hydroxyl group of Serl3la (Perutz, 1970; Nigen et al., 1980); (3) an intersubunit site between the guanidium group of Argl4la and the a-amino group of Valla of the opposite subunit (Peru&, 1970). However, recent crystallographic observations on the mutant Hb Rothschild (p37 see Kavanaugh et al., 1992) have Trp+Arg, indicated that a single point mutation may trigger the appearance of a new O,-linked chloride binding cleft at the aIPz interface. However, such a situation does not apply to our case, since Hb from bear shows the same amino acid residues of human HbA in the a,Pz contact region (Hofmann et al., 1986). Therefore, the location of the new synergistic chloride binding site in bear Hb remains an open question. The physiological importance of the synergistic effect reported in this paper can be fully appreciated considering the relevance of the “chloride shift” in

1405

Table 2 Overall heat of oxygenation (611) in kJlmo1 of oxygen displayed by brown bear Hb in 0.1 M bis-Tris or Tris. HC1 buffers plus 0.1 M NaCl and 3 mM Gri(2,3) P, PH

6H (kJ/mol of 02)

68 7.0 7.3 7.6

-9-6 -142 -293 - 50.2

The values have been calculated from the van’t Hoff equation using data reported in Figure 3 and are corrected for the heat contribution from oxygen solubilization (z - 12.5 kJ/mol).

the regulation of oxygen unloading and loading at the level of capillaries (Brix et al., 1990). Moreover, it must be pointed out that the 6H for oxygen binding displays a marked pH-dependence (see Figure 3 and Table 2) such that under conditions of acidosis, likely to be realized during the winter sleep, the exothermicity of oxygenation is drastically reduced. Therefore, under these conditions, even a lowering of temperature at the level of peripheral tissues would have essentially no consequence for their 0, supply. Further, although no direct information is available in the case of brown bear, it must be outlined that acidosis is usually associated in other mammals to hypercloruremia; hence, the peculiar synergistic effect of Cl- and Gri(2,3)P, on the oxygen binding properties of bear Hb could have a physiological relevance in rendering possible a normal oxygen supply at low and temperatures even under hypercloruremic acidotic conditions. Therefore, as a final conclusion bear Hb seems to represent an additional variation on the theme of t,he relationships between environmental* cofiditions and functional modulation, since in this case the appropriate oxygen supply is realized through both the interaction between different anionic cofactors and through temperature effects, in a complex but very efficient interplay among all these effecters. The financial support Universita’ e della Ricerca

from the Minister0 dell’ Scientifica e Tecnologica of

Italy (MURST 40%) and from National Research Council is gratefully Dr. Matti

acknowledged. The authors Nuutinen for stimulating

want to thank and helpful

discussion. References Antonini,

E. & Brunori, IM. (1971). Hemoglobin and

myoglobin

in

their

reactions

with

ligands.

Frontiers of Biology (Neuberger, A. t Tatum,

In

E. L.,

eds), North-Holland Publ. Co., Amsterdam. Antonini, E., Condo’, S. G., Giardina, B., Ioppolo, C. 6 Bertollini, A. (1982). The Effect of pH and D-glycerate 2,3-biphosphate on the 0, equilibrium of normal and SH(b93)-modified human hemoglobin. Eur. J. Biochena. 121, 325328. Arnone (1972). X-ray diffraction study of binding of

1406

Modulation

of Heterotropic

i,3-diphosphoglycerate to human deoxyhaemoglobin. Nature (London), 237, 146*149. Bonaventura, J., Bonaventura, C., Sullivan, B., Ferruzzi, G., McCurdy, P. R., Fox, J. & Moo-Penn, W. F. Hemoglobin providence. Functional (1976). consequences of two alterations of the 2,3-diphosphoglycerate binding site at position p82. J. Biol. Chem. 251, 7563-7571.

Brix.0.. Thomsen, B.. Nuutinen. M., Hakala, A., Pudas. J. & Giardina, B. (1990). The chloride shift may facilitate oxygen loading and unloading to/from the hemoglobip from the brown bear (lJr.slls nrclos). Comp. Biochem. Physiol. 95B, 865468. Brunori, M., Coletta. M. & Giardina, B. (1985). Oxygen carrier proteins. In MetalZoProteins (Harrison, P.. ed.). vol. 2, pp. 263-331, MacMillan, London. Coletta, M., Clementi, M. E., Ascenzi, P.. Petruzzelli, R,., Condo’. S. G. $ Giardina, B. (1992). A comparative study of the temperature dependence of the oxygenbinding properties of mammalian hemoglobins. Eur. J. Biochem. 204. 1155-l 157. De Bruin, S. H. & Janssen, L. H. M. (1973). The interaction of 2,3-diphosphoglycerate with human hemoglobin. Effects on the alkaline and acid Bohr effect. J. Biol. Chem. 248, 2774-2777. Eaton, W. A. (1982). The relationship between coding sequences and function in haemoglobin. Nat,ure (London), 284. 183-185. Giardina, B. 6 Amiconi, G. (1981): Measurement of binding of gaseous and nongaseous ligands to hemoglobins by conventional spectrophotometrir procedures. Methods Enzymol. 76, 4 17427.

Effects in Bear Hb

Giardina. B., C’OIIC~O’S. . G., Petruzzelli. It.. Bardgard. A. t I3rix, 0. (1990). Thermodynamics of oxygen bincling to arctic hemoglobins. The case of reindeer. Biophys. Chem. 37. 28 l-286. Hofmann. 0.. Schreitmiiller. T. Br. Braunitzer. G. (1986). The primary structure of polar bear and asiatir black bear hemoglobin. Biol. (‘hem. Hoppe-Seyler. 367. 53-59. Imai. K. (1982). Atlosteric Effecls in Hnemoglobin. Cambridge University Press, Cambridge. LTK. Kavanaugh. ,I. S., Rogers, P. H.. Case. 1). A. BE Arnone. A. (1992). High-resolution X-ray study of deosyhemoglobin Rothschild 37/l Trp+Arg: a mutation that. creat,es an intersubunit chloride-binding site. Riochrmistry. 31, 411 IL-k1 21. K’igen. A. M.. Manning.
Edited by A. R. Fersht

(Received 13 July 1993; a.ccepted 24 November 1993)