the journal of bioloqical chemistry vol. …the journal of bioloqical chemistry vol. 250, no. 3,...

THE JOURNAL OF BIOLOQICAL CHEMISTRY Vol. 250, No. 3, Issue of February 10, pp. 952-960, 1975

Printed in U.S.A.

Conformation and Spin State in Methemoglobin*

(Received for publication, May 17, 1974)

PRESTON HENSLEY, STUART J. EDELSTEIN, DAVID C. WHARTON, AND QUENTIN H. GIBSON

From the Section of Biochemistry, Molecular and Cell Biology, Cornell University, Ithaca, New York 14850

SUMMARY

The properties of human methemoglabin have been investigated under a wide variety of conditions to determine its conformation and to test for evidence of the T state conformation which has been proposed by Perutz to exist in the presence of high spin ligands and inositol hexaphosphate (IHP). Subunit dissociation was measured as a criterion for the T state since marked differences in the tetramer-dimer equilibrium exist for oxyhemoglobin (R state) and deoxyhemoglobin (T state). In the absence of IHP, complexes of methemoglobin with both high spin ligands (water, fluoride) or low spin ligands (azide, cyanide) show extensive dissociation in 2,2-bis(hydroxymethyl)-2,2’,2”-nitriloethanol buffers, pH 6, 0.1 M NaCl, with values of the tetramer-dimer dissociation constant (K4,2) near 10W5 M. The addition of IHP lowers K4,2 to a value near 1OW M for all forms of methemoglobin. Combination of IHP with methemoglobin promotes a conformational change, but the change is apparently independent of spin state. The conformation acquired in the presence of IHP is not identical with the T state (K4,* N lo-l2 M) and can also occur with hemoglobin in the ferrous form, as revealed by a substantial reduction in K4,2 for CO-hemoglobin upon addition of IHP. Subunit dissociation has also been measured using the haptoglobin reaction, since haptoglobin binds only to hemoglobin dimers. The haptoglobin experiments give results that are qualitatively in agreement with the conclu- sions reached by ultracentrifuge measurements. Similar results are also obtained by estimating the degree of dissociation on the basis of the material which aggregates following mixing with dithionite. The effect of IHP on azide-binding kinetics with methemoglobin has also been examined. Change in reactivity is observed upon addition of IHP, but the principal effect is an enhancement of the rate of reaction of the fi chains. Changes in the reactivity of the /393 sulfhy- dry1 group of methemoglobin also accompany addition of IHP, but in a manner which is largely independent of the spin state of the iron. Similar changes are again found with CO-hemoglobin upon addition of IHP. The rate of binding of bromthymol blue also shows some changes upon addition of IHP, but the changes are more pronounced for deoxyhemoglobin than for methemoglobin. Since the results obtained did not appear to indicate a significant role for spin state in the changes observed, additional studies were undertaken Lsing

* The work was supported by the National Institutes of Health Grants HL-13591.04 (S. J. E.), HL-10633-08 (U. C. W.), and GM- 14276.08 (q. H. G.) ; National Science Foundation Grant GB-41448 (S. J. E.) ; and by an Alfred P. Sloan Foundation Research Fellow- ship (S.J.E.).

EPR spectroscopy. The principal effect of IHP on spin- related parameters was found to be a slight shift in the apparent pK of the transition from aqua- to hydroxymethemoglobin. However, on the acid side of the pK no significant enhancement of spin was caused by addition of IHP. An additional test was performed with methemoglobin S. Since gelation of sickle cell hemoglobin is associated with the T state, methemoglobin S was examined for gelation in the presence of IHP. However, with a 31% solution at 37” for 1 hour, no gelation could be detected. Thus, while conformational changes in methemoglobin are clearly promoted by IHP, the changes appear to involve neither a correlation with spin state nor a substantial population of a normal T state structure. Quantitative estimates of the population in various states are presented in the accompanying paper.

In recent years many features of the cooperative binding of oxygen by hemoglobin have been defined in precise structural terms. The three-dimensional st,ructures of the liganded and unliganded forms as well as of numerous variants have been resolved (l-7). Kinetic aspects of binding of oxygen and other ligands have been described, including the recognition of important differences in the contributions of the individual cz and /3 chains (8-11). The suggestion that many disparate functional properties of hemoglobin could be represented by a simple model based on two conformational states T, and R (12, 13), has been supported by several recent experimental findings, including kinetic (14-16) and nuclear magnetic resonance (17-19) measurements. Studies on the dissociation of hemoglobin into half- molecules, o$3 dimers, have revealed an absence of cooperative ligand binding (20-24) and permitted assignment of the value of L as the ratio of the tetramer-dimer dissociation constants for the T and R states (25, 26). The equilibrium between the states is sensitive to pH, structural changes in the hemoglobin and small molecules, such as organic phosphates which bind predominantly to the T state (27, 28). The stabilization of the T state by organophosphates, particularly IHl’,’ is sufficient to switch several forms of dcoxyhemoglobin, including the valence hybrids (17)) carboxypeptidase-treated variants (15)) and mu- tants such as Bethesda (14), from predominance of the R state to predominance of the T state.

1 The abbreviations used are: IHP, inositol hexaphosphate; bis- tris, 2,2-bis(hydroxymethyl)-2,2’,2”-nitriloethanol; p-MB, p- mercuribenzoate; Pipes, piperazine-N ,N’-bis(S-ethanesulfonic acid).

952

by guest on June 30, 2020http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

953

Following suggestions by Hoard that in low spin heme com- pounds the iron is planar with the heme and outside the plane of the heme in high spin complexes (29, 30), I’erutz has re-examined the positions of the iron atoms in hemoglobin; a displacement of about 0.75 A from the plane of the heme ring in deoxyhemoglobin was revealed (31). The differences in position of the iron dependent on spin state arise from the fact that the radius of the iron in the high spin form present in deoxyhemoglobin is too large to be accommodated by the heme ring without some displacement. In oxyhemoglobin the iron is low spin and the radius is diminished so that an in-plane position of the iron can be accommodated. In t,he ferric form (methemoglobin), its position is intermediate, with a displacement of about 0.3 A from the heme plane (31). This observation led I’erutz to propose that the position of the iron slightly out of the plane in methemoglobin would facilitate its transition to the T state upon addition of 1HP (32). In support of this view he advanced spectral studies on mixed valence hemoglobins and recently a series of papers with more extensive studies by a variety of methods (33-35). On the basis of these result,s F’erutz and co- workers propose a coupling of spin state and conformation, such that the T conformation enhances the high spin character of the

iron and thereby contributes to the diminished affinity for oxygen

in the T state. Transition to the R state releases the tension on the iron lowering the spin state and contributing to the enhanced affinity for oxygen of the R state. This proposal has major im- plications both in terms of the role of metals in biochemical systems and the detailed mechanism of actiou of hemoglobin

and therefore warrants close examination. In this paper the question of coupling of spin state and conformation is examined with a variety of kinetic and spectroscopic methods in conjunc- tion with subunit dissociation measurements. Studies on the oxidation-reduction reaction of hemoglobin and a quantitative

analysis of the conformational equilibria of methemoglobin will be presented in the follov+ing paper.

EXPERIMENTAL PROCEDURES

Materials-Pooled samples of human hemoglobin were obtained from the Tompkins County Hospital. Hemolysates were prepared as described earlier (36). Methemoglobin was prepared by reacting hemolysates with a slight excess of ferricyanide. The mixture was then freed from excess ferricyanide and organic phosphates by passage through a column of Sephadex G-25. Solutions were used within 1 week of preparation. The sample of sickle cell hemoglobin used in these studies was kindly provided by Dr. 8. Charache.

UZtracentrijugution-Sedimentation equilibrium experiments were performed with a Beckman model E ultracentrifuge equipped with a monochromator, photoelectric scanner, and an on-line computer system for automated data collection and analysis (37). Up to 15 solutions were examined simultaneously using five multichannel cells with a six-hole titanium rotor. All experiments were performed at a rotor temperature of 20”. The dat,a stored in the computer were fitted to a two-term exponential equation (21) with terms corresponding to hemoglobin dimers (mol wt 32,250) and tetramers (mol wt 64,500). From the coefl- cients of the two terms (corresponding to the concentrations of dimers and tetramers), the equilibrium constant for the tetramer- dimer dissociation reaction, Kd,z, was calculated directly.

Kinetic Measurements-Stopped flow measurements were made by following absorbance or fluorescence changes with a Durrum- Gibson apparatus interfaced to an on-line computer. Details of the system have recently been described (10). Data reported here represent average values obtained from three successive mixing events. All kinetic measurements were performed at 20”.

EPR Measureme,lts-Spectra were recorded with a Varian model E-3 spectrometer equipped for measurements at the temperature of liquid NS (77 K). The spectra shown in this paper represent the first derivatives of the absorption curves.

3oL 0 6 12 I.8 24 30 36 42 48 54

Concenhatlon ( p M heme)

FIG. 1. Sedimentation eauilibrium studies on various forms of hemoglobin. Data is presented in terms of weight average molecular weight aersua concentration. The buffer was 0.05 M bis-tris, pH 6.0, 0.1 IM NaCl. All experiments were conducted at 20”.

RESULTS

Ultracentrijugation-In order to characterize the quaternary stability of methemoglobin in the presence and absence of ligands and IHP, sedimentation equilibrium experiments were performed under conditions which reveal the tetramer-dimer equilibrium. Data collected uith the aid of the online computer were examined by fitting the points to the sum of exponentials and expressing the data as 111 versus c. Fig. 1 presents the results in this form and emphasizes the differences between methemoglobin alone, methemoglobin plus IHP, and deoxyhemoglobin. The results can be expressed in terms of dissociation constants (K& calculated from the coefficients of the exponential fitting equa- tions, and values of K4,2 = 3 x 1O-5 M and K~J = lop9 M

are obtained for methemoglobin alone and with added IHE’.* The value of K4,2 calculated for deoxyhemoglobin is less than 10~I1 M, in agreement with the values in the IO-‘* M range de- termined under similar conditions with the CO binding method (25, 26).

The effects of ligands of methemoglobin on the dissociation equilibrium were also examined both in the presence and absence of 1HP. Data were analyzed to obtain dissociation constants, as described for Fig. 1, and indicate that regardless of whether the ligand is high spin (fluoride) or low spin (azide), the dissociation behavior is similar. The results for fluoride methemoglobin were obtained by performing measurements at several fluoride concentrations to eliminate the effects of ionic strength since fluoride binding is so weak (38) that high concentrations of NaF are required to approach saturation. Experiments w-ere also performed to study the effects of IHP on CO-hemoglobin and these

2 In these calculations, data were fit on the basis of 1 molecule of IHP bound per tetramer, with the corresponding adjustments in molecular weight and in to take into account the contributions of IHP. Bindine of 2 molecules of IHP oer dimer was also assumed, since the expezments with p-MB indicate a direct effect of IHP on the sulfhydryl reactivity of the dimers and since this stoichiometry gave the best fit on the basis of a statistical analysis of the data (P. Hensley, J. K. Moffat and S. Edelstein, in preparation). Spectral studies indicated that no change in the extinction coefh- cient of hemoglobin was required since spectral changes in the Soret region caused by binding of IHP were too small to cause a significant change in the estimates of the dissociation constants when compared to the standard deviations of the measurements themselves (53).


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

954

60 HbCO+ IHP

1 2 4 6 8 IO 12 14 16 16 20

Concentration (IL M heme)

FIG. 2. Dependence of molecular weight on concentration for CO-hemoglobin in the presence and absence of IHP. Other details were as in Fig. 1.

results are summarized in Fig. 2. A clear reduction in the extent of dissociation occurs in the presence of IHP, suggesting that the effect of IHP is not limited to ferric hemoglobins where the iron may be slightly out of the heme plane (29-31).

Haptoglobin Experiments-Since the question of whether methemoglobin can assume the T state is so important in evaluat- ing the recent hypothesis regarding spin state and conformation, the general conclusion from ultracentrifugation that methemoglobin even in the presence of IHP dissociates somewhat more than deoxyhemoglobin was also investigated using the haptoglobin reaction. With normal deoxyhemoglobin,. haptoglobin is essentially unreactive, since its binding appears to require free cu/3 dimers (39). In contrast, haptoglobin reacts readily with oxy- or CO-hemoglobin and the time course of the binding can be analyzed to give values of the tetramer-dimer dissociation constant generally in good agreement with ultracentrifuge results (39). When methemoglobin was examined in the haptoglobin reaction no indications of a T state such as is found for deoxyhemoglobin were obtained. Even in the presence of IHP with aquomethemoglobin (high spin), the haptoglobin reaction readily occurred. With a rapid phase attributable to the presence of dimers (Fig. 3A), the slow phase would then be a reflection of dissociation of tetramers into dimers which can then react with haptoglobin. From the proportion of the material reacting rapidly, an estimate of the extent of dissociation can be obtained. The results confirmed the ultracentrifuge observations of a decrease in dissociation when IHI’ was added, but there was more dissociation than that observed with deoxghemoglobin in the T state. Similar results were found with azide methemoglobin, a low spin derivative (Fig. 3B), and oxyhemoglobin (Fig. 3C). Thus influence of IHP on the extent of dissociation is to a first approximation independent of the spin state of the 3+ iron and independent of valence.

Dithionite Experiments-A third independent method was employed to measure the effects of IHP on subunit. dissociation. If dilute solutions of osyhemoglobin or methemoglobin are re- acted with dithionite, deoxghemoglobin is formed. The reduction of methemoglobin is considerably slower than the removal of oxygen from oxyhemoglobin. However, for both species, reaction with dithionite will convert dimers initially present to deoxydimers. Since deoxyhemoglobin is effectively undissoci-

ated at all concentrations above about 10-n M, the deoxydimers then aggregate with a spectral change and reaction rate characteristic of dimer association (40). Thus, the fraction of the molecules which participates in this reaction is a reflection of the initial concentration of dimers present. Results for methemoglobin in the presence and absence of IHP are presented in Fig. 4. The initial phase of the reaction reflects changes due to reaction with dithionite per se. However, the slower changes are due to association of deoxydimers and the AA at the zero time obtained by extrapolation to the ordinate is a measure of the concentration of dimers initially present. Some correction is necessary for the finite time of the dithionite reaction so that the extrapolation should not be to zero time, but to about halfway through the rapid phase of the reaction. For methemoglobin (6 PM heme) alone the slow phase corresponds to an absorbance change equivalent to about 3 PM (heme), roughly in agreement with the extent of dissociate measured by other methods. Upon addition of IHP the amount of heme is substantially reduced, to about 0.3 PM, but is still present in measurable abundance. The increased scatter in the points for the determination with IHP reflects the much smaller changes in absorbance present under these conditions and the decreased rate reflects the diminished amount of reactive material.

Azide Binding to dlethemoglobin-In order to test possible effects of IHP on the electronic structure of the iron, the kinetics of azide binding were examined since changes in the iron might reasonably be expected to influence the rate of azide binding. As seen in Fig. 5, an effect of II-II’ on the time course of azide binding is observed. The relatively homogeneous reaction in the absence of IHP becomes more obviously biphasic in the presence of IHP. Analysis of a series of curves recorded at different wavelengths indicates that the more rapid phase corresponds to the reaction of the /3 chains. Thus the effect of IHP on the azide reaction is to cause a preferential increase in the rate of binding to the /3 chains, while leaving the behavior of the (Y chains effectively unaltered.

Reaction with p-JIB--As an aid in establishing the conformational state of hemoglobin under different conditions, the time course of the reaction of methemoglobin with p-M13 has been examined. Since the rates of reaction with p-MIS for oxyhemoglobin (R state) and deoxyhemoglobin differ widely, by up to SO-fold (41, 42), the reaction should have important diagnostic value in determining conformational changes. In initial experiments relatively small changes in reactivity toward p-lln with methemoglobin were observed upon addition of ligand and these changes were very similar regardless of whether the methemoglobin was saturated with high spin or low spin ligands. At that time Dr. Sanford Simon communicated a preliminary version of his work (34) and indicated that large IHl’ effects were obtained in Pipes buffer at slightly acid pH values. Under these conditions we too find somewhat larger IHP effects. Under the normal buffer conditions used for the bulk of the other studies on methemoglobin however, it is found that IHPinduced changes are considerably smaller than the difference between oxyhemoglobin and deoxyhemoglobin and show no consistent dependence on spin state. For example, the time course of the reaction of p-Ml% for both aquomethemoglobin and azide methemoglobin was studied in the presence and absence of IHP. As seen in Fig. 6A, some reduction in reaction rate accompanies addition of IHI’ and the reduction is somewhat greater for the predominantly high spin aquomethemoglobin than for the predominantly low spin azide methemoglobin. However, this difference is due to a slower rate for azide methemoglobin in the absence of IHP. In


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

955

TIME (set)

\- NO ADDITIONS

\

.

1 I

,\; 3

TIME bed

the presence of IHP very similar rates are observed for the two forms; therefore, to the extent that reactivity of p93 -SK reflects conformation, there is no difference in the presence of 1HP for the high spin and low spin forms. A similar pattern of re- activities is observed at pH 7.4 (Fig. 6B). For both pH values some heterogeneity in the reactions is apparent., as shown by the nonlinearity of the first order plots (Fig. 6, A and B). It is difficult therefore to compare these results directly with those reported by l’erutz et al. (34) since their results are presented as single rate constants. Thus, for the reaction at pH 7.4 the differences in the latter stages of the reaction. The small differences in the initial phases would agree with the results for very similar conditions presented by Perutz et al. (34).

Since the degree of dissociation of the various forms of methemoglobin varies with added IHP, studies u-ere conducted to determine the extent to which changes in the rate of reaction with

TIME (set)

FIG. 3. Reaction of hemoglobin with haptoglobin. Fraction fluorescence change remaining versus time. Hemoglobin concentration: 36~~, bis-tris buffer, pH 6.0,2Op~ eq of haptoglobin. A, methemoglobin; B, methemoglobin azide; C, oxyhemoglobin.

p-MB reflect changes in the tetramer-dimer dissociation constant. The react,ion of p-MB with aquomethemoglobin was examined at two different concentrations, 20 PM (heme) and 5 PM (heme), where the proportion dimer should be appreciably different according to results cited earlier. (Ko independent dissociation measurements were made on these solutions but results are obtained in phosphate buffer under conditions for which the tetramer-dimer equilibrium has been more thoroughly characterized.) No significant difference in time course was observed for the two concentrations, either in the presence or in the absence of IHP (Fig. 7). Moreover when IHP was placed in the syringe containing the p-MU solution instead of in the methemoglobin solution (Fig. 7), the results were effectively unchanged from the results with IHP in the syringe containing the methemoglobin solution. Therefore, the IHP would appear to be acting on dimeric molecules (c$?) as m-e11 as tetrameric molecules ((~$3~).


ww

.jbc.org/D

ownloaded from

oprn manager datamatics

http://www.jbc.org/

956

\ .

.OOll I 0 .6 1.2 I.8 2.4 3.0 36

ITISSC

FIG. 4. Reaction of hemoglobin with dithionite. Absorbance change on a logarithmic scale versus time for 6 PM methemoglobin in the presence and absence of IHP (100 PM) in bis-tris buffer, pH 6.

0.1 -

.x3-

O-.. no IHP -.

0.. “\ 01

200

TIME (msec)

FIG. 5. Reaction of azide with methemoglobin. Conditions: 0.05 M bis-tris, pH 6, 50 rnM azide. Observation at 430 nm with 4-mm cell in the presence and absence of 25 I.~M IHP.

Although 1HP pulls the tetramer-dimer equilibrium in favor of the tetramer, aggregation could not progress substantially in the time of the p-M13 reaction. Since the T state is accessible only to tetrameric forms of hemoglobin, it follows that the effects of IHP must be due, at least in part, to interactions with R-type structures.

The general effect of a depression in the reactivity of /393 -SW toward p-MB caused by IHP is not limited to methemoglobin. Effects of similar magnitude are also observed with CO-hemoglobin as summarized in Fig. 8.

Reaction with Bromthymol Blue-The reaction of hemoglobin with bromthymol blue can be used to distinguish the R and T states along with parameters already described such as p93 -SH reactivity and subunit dissociation. While differences exist in the extent of binding of bromthymol blue to deoxyhemoglobin and oxyhemoglobin, differences in the rate of reaction of hrom- thymol blue with the two states of hemoglobin are particularly marked; the velocity of combination of bromthymol blue with deoxyhemoglobin is considerably higher than for the liganded forms of hemoglobin (43). Since both subunit dissociation and @93 reactivity for methemoglobin indicate some change upon

I I 0 6 I2 18 24 30 36

.0021 9 6 I2 I6 24 30 36

mS*C

1 FIG. 6. Reaction of methemoglobin and azide methemoglobin

with p-MB. Data obtained in 0.05 M Pipes and 25 PM p-MB with 20 PM methemoglobin (0) or azide methemoglobin (0) in the presence and absence of IHP (200 PM). A, data for pH 6.8; B, data for pH 7.4.

addition of IHI’, but give a conformation which is not identical with the normal T state, e‘xperiments have been performed on the reaction of methemoglobin with bromthymol blue to determine the degree to which bromthymol blue binding resembles that of one or the other conformational states. As seen in Fig. 9, the characteristic rapid rate for deoxyhemoglobin and slow rate for methemoglobin at pH 7 are observed. However, addition of IHPresults in a drastic decrease in the bromthymol blue binding rate for deoxyhemoglobin and a much smaller decrease in bromthymol blue binding rate for methemoglobin. The IHP-dependent decrease in bromthymol blue binding rate is somewhat larger at pH 6.6 (Fig. 10) and effectively absent at higher pH (above 7.4). Thus, while bromthymol blue binding emerges as an indicator sensitive to the presence of IHl’ (in addition to a marked sensitivity to pH and ionic strength), the major effects are found with deosyhemoglobin. Some change with methemoglobin is observed but the rate of binding is still considerably faster than observed with deoxyhemoglobin.

EPR Neasurements-The studies already described are con- cerned with the effects of changes in the iron resulting from IHP- stabilized conformational changes as revealed by sedimentation and kinetic measurements. If 1HPstabilized conformational changes are in fact spin state-dependent, reciprocal effects


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

957

.06 .05’ - .04 1 .”

i‘ \ ~“-%~o~ +--n-

m 0, A-----A--.- A .”

1 ‘. -=--l-o-,

L

,001 1 0 6 12 10 24 30 36

meet

A _ --=--0 . . ,+IHP

‘---., --0. l --0,

dewy Hb

G b / 9 006 \ Q B

04 -e.. r o\-“c~.....~;~l H P “\ O\ ‘01

.02 O\ O---.9 \o,;Alone

01 , y----By,-

40 00 120 I60 200 240 - meet . . . FIG. 7. Reaction of methemoglobin with p-MB. Condltlons:

0.05 M Pipes, pH 6.0, 0.25 j&M p-MB with methemoglobin at 20 MM FIG. 9. Reaction of hemoglobin with bromthymol blue. A, or 5 pM (where indicated as 1:4 dilution) with IHP (0) or without reaction of 40 &M bromthymol blue with 40 PM methemoglobin or

IHP (0). In one case, IHP was added with the p-MB (A). deoxyhemoglobin in 0.05 M phosphate buffer, pH 7.0, in the presence and absence of IHP (500~~). 6(, reaction of methemoglobin

.06

01 -

108 -

106-

\ no IHP

0

I I

16 32 48 msec

FIG. 8. Reaction of CO-hemoglobin with p-MB. 50 pM p-MB, 25 pM CO-hemoglobin, 0.05 M phosphate buffer, pH 7.0.

Conditions :

in 0.1 M phosphate buffer, pH 6.6. Other conditions were as in

1

5.0 s2.59 g=2.18 9.1.88

I

-A- /-- DT-

tively. The conditions of EPR spectroscopy were: microwave

FIG. 10. EPR spectra of methemoglobin (15 mM in heme) in 0.01 M potassium phosphate buffer, pH 6.0. A, without IHP;

frequency, 9.125 GHz; microwave power, 100 milliwatts; modula-

B, with 20 mM IHP.

tion amplitude, 10 G; scanning time, 16 min; temperature, 77 K.

Both samples contain 20 mM NaCl. C and D represent a lo-fold amplification of spectra A and B, respec-

should be detectable by studying the EPR signal of the iron under different conditions. Therefore, El’R spectra were re-

corded on methemoglobin to explore this point. As seen in Fig. 10, spectra of aquomethemoglobin are very similar in the presence and absence of IHP at, pH 6.0 and indicate a predomi-

nance of high spin material in both cases. Small changes can be detected at higher pH where the transition to hydroxymethemoglobin begins to appear, as seen in Fig. 11. However, when these changes are normalized and plotted versus the ratio of IHP

to hemoglobin tetramers (Fig. la), the changes are not saturated at a ratio of one. These results appear to indicate that (a) aquomethemoglobin at pH 6.0 is almost entirely high spin in the absence of IHP and 1HP causes little further increase in high

spin character, and (b) IHP may influence the transition between aquo- and hydroxymethemoglobin at slightly alkaline pH, but this effect is at least partially nonspecific (in terms of the single organic phosphate site between the p chains) since the effect


ww

.jbc.org/D

ownloaded from

oprn manager datamatics

http://www.jbc.org/

958

q=6.0 9.2.59 92.18 i gz1.88

I I I

FIG. 11. EPR spectra of methemoglobin (2.68 mM in heme) in 0.01 M potassium phosphate, pH 8.0, and 0.02 M NaCl. A, no IHP; B, 0.67 mM IHP; C, 1.34 mM IHP; D, 2.01 mM IHP; E, 2.68 mM IHP; F, 3.35 rnM IHP; G, 4.02 mM IHP. The conditions of EPR spectroscopy were: microwave frequency, 9.124 GHz; microwave power, 100 milliwatts; modulation amplitude, 20 G; scanning time, 16 min; temperature, 77 K.

persists at ratios of 1HP to hemoglobin of greater than unity under conditions where binding is stoichiometric.

Studies with Hemoglobin S-Since the various physical studies reported here indicate some conformational change in methemoglobin upon addition of lHP, further experiments were conducted with methemoglobin S to determine the extent to which the new conformation shares properties with the T state. He- moglobin S has the advantage for these studies that gelation readily occurs only in the T form. Other liganded forms may participate in gelling but they diminish the efficacy of the process, requiring higher concentrations of hemoglobin to achieve gelation (44). Published reports on methemoglobin S deal exclu- sively with cyanomethemoglobin (44, 45) ; therefore, experiments were performed with aquomethemoglobin. A cyclic series of tests were conducted on a sample of hemoglobin S (26%). Gelation was observed in the deoxy state which was reversed upon cooling. A gel was formed upon warming with 1HP added and broke down upon cooling. The sample was then treated with a small excess of ferricyanide to convert it to methemoglobin and failed to gel upon warming. Addition of 1Hl’ also failed to cause gelation but upon re-forming deoxyhemoglobin by addition of a small excess of dithionite, gelation occurred which was reversed upon cooling. Therefore, it is not likely that methemoglobin in the presence of IHP assumes a normal T conformation. Repet,itions of the experiment with concentrations of hemoglobin S up to 3Ioj, also failed to reveal gelation with

IO 20 30 40 5.0 60

IHP/TETRAMER

FIG. 12. Effect of IHP on the amplitude of the high spin ferric EPR signal (g = 6.0) of methemoglobin. l , pH 7.5; 0, pH 8.0; 0, pH 8.5. All samples were buffered with 0.01 M potassium phosphate and contained 0.02 M NaCl. The conditions of EPR spectroscopy were identical with those for Fig. 2.

methemoglobin. Brieh13 has reported gelation of methemoglobin in the presence of IHP at 35%. However, this concentration is roughly twice that required for gelation of deoxyhemoglobin under similar conditions and argues for the nonidentity of the conformation of deoxyhemoglobin and methemoglobin in the presence of IHP.

DISCUSSION

The experiments reported here provide a clear test of the hypothesis of I’erutz linking spin state and conformation (32-35). According to the hypothesis, formation of the T state places tension on the hemes which enhances the high spin character of the iron and is responsible for the diminished affinity of the T state toward ligands. The conclusion that spin state and conformation are tightly coupled is based on the following line of argument: (a) spectroscopic indices can be established which reflect an R --) T transition upon addition of IHP for deoxyhemoglobin (33) ; (b) spectroscopic changes (as well as changes in @93 -SH reactivity) upon addition of IHP to high spin forms of methemoglobin indicate an R + T transition (34); and (c) the nature of the spectroscopic changes plus magnetic susceptibility measurements indicate that a heightening of the spin state of the iron accompanies the R --) T transition (35). In testing the hypothesis linking conformation and spin state for methemoglobin, two principal questions may be raised: first, does the conformation of methemoglobin acquired in the presence of IHI’ have the same properties as the T state and, second, does the formation of this conformation require high spin iron in the hemes?

Concerning the first question, the results presented here clearly demonstrate a conformational change for methemoglobin upon addition of IHP, but the conformation formed is not identical with the T state. Its subunit dissociation and SH reactivity are intermediate between the properties of the R and ‘I’ states; gelation experiments on methemoglobin S indicate that the T state cannot be the predominant conformation present. Studies on azide reactivity indicate that the major effect of addition of IHP to methemoglobin is a change in the reactivity of the P chains. Similar effects are observed with liganded ferrous he-

3 R. W. Briehl, personal communication.


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

959

moglobin (9, 10) where no T 4 R transition appears to be in- volved. Therefore, it would be premature to conclude that addition of IHP to methemoglobin results in acquisition of the T state conformation. Some fraction of the population may acquire a conformation which resembles the T state, but this form is not likely to predominate. Actual limits on the proportion of molecules which may be in the T state can be set by an analysis of the oxidation-reduction reaction as described in the following paper. Concerning the second question, it appears that formation of the conformation of methemoglobin promoted by IHP is not dependent on the spin state of the iron. Similar effects on subunit dissociation and p-MB reactivity are observed for both high spin and low spin forms.

In contrast to the predictions of the Perutz hypothesis, the results presented here are interpreted most directly as indicating that the conformational change in methemoglobin promoted by IHP is not linked to spin state. The failure to verify the hypothesis of Perutz linking spin state and conformation had led us to re-examine the evidence upon which the hypothesis is based. While certain observations in the recent papers by Perutz et al. (33-35) are consistent with the linkage of spin state and conformation in which population of the T state enhances the high spin character of the heme iron, other observations are at variance with this conclusion. For example:

1. Noncooperative forms of deoxyhemoglobin which are fixed

found the semihemoglobin (heme only on (Y chains) which is largely fast reacting with CO, with a rate characteristic of the R state, is rendered largely slow reacting with CO, with a rate characteristic of the T state, upon addition of protoporphyrin. Since protoporphyrin contains no iron, yet stabilizes a T state, this result argues against a requirement of iron in a high spin form for formation of the T state. Similarly, Noble et al. (48) observed that protoporphyrin globin reacts with bromthymol blue, p-MB, and N-ethylmaleimide at rates which are similar to those observed for deoxyhemoglobin. Thus in this case as well, properties of the T state are observed in the absence of a metal, so that the suggestion of a special requirement for high spin iron seems unnecessarily limiting. The cooperative oxygenation of cobalt hemoglobin in which the cobalt is low spin in the T state provides further arguments against any structural significance to the high spin to low spin transition of the iron in oxygenation of ferrous hemoglobin (49, 50).

in the R state display the same magnetic susceptibility as normal T state forms (33) ; thus, any coupling of conformation and spin state is apparently below the level of detectability.

2. The various spectral and kinetic indices used to distinguish the response to IHP of high and low spin forms of methemoglobin show qualitatively similar effects for both forms. The effects with high spin forms are argued to be quantitatively greater (34) ; these arguments are considered in the following paper.

hlarked spectral and paramagnetic changes in nitrousoxide- hemoglobin, a low spin material, up011 addition of IHP (51, 52) provide further indications that such changes are not restricted to high spin forms. Thus we conclude that the conformational change in methemoglobin promoted by IHP is largely independent of spin state. In addition the transition to a T-like state upon addition of IHP can be only partial, as revealed by an analysis of the oxidation-reduction reaction by which methemoglobin is formed from deoxyhemoglobin. Such an analysis is presented in the following paper which includes a quantitative evaluation of the possible R t--t T equilibria for methemoglobin and a further discussion of spin state and conformation.

Acknowledgment-We thank Dr. Jack Peisach for helping us with the preliminary EPR experiments.

3. Azide binding affinity is virtually independent of 1HP (34) so that evidence is lacking that the conformational state favored by IHI’ has any functional properties distinct from lHP-free methemoglobin; indeed the kinetic results present here suggest that the effects of 1HP are limited to /3 chains and therefore argue against an R -+ T type transition that might be expected to alter the behavior of both chains.

REFERENCES

1. XUIRHEAD, H., Cox, J. M., MAZZARELLA, L. & PERUTZ, M. F. (1967) J. Mol. Uiol. 28, 117-156

2. MUIRHEAD, H. & GREER, J. (1970) Nature 228,516-519 3. GREER, J. (1971) J. Mol. Biol. 69, 99-105 4. GREER, J. (1971) J. Mol. BioZ. 69,107-126 5. GREER, J. (1971) J. Mol. Biol. 62,241-249 6. PERUTZ. M. F. & TENEYCK, L. F. (1971) Cold Spring Harbor

4. The changes in magnetic susceptibility reported for aquomethemoglobin upon addition of IHI’ are very small (35), on the same order as the changes reported for deoxyhemoglobin (33) between pH 7 and pH 8.6 (where no T -+ R transition is believed to occur) and even smaller than changes observed in normal deoxyhemoglobin itself upon addition of organic phosphates (46). Little alteration in spin state (but some decrease in subunit dissociation constant) for methemoglobin upon addition of IHP has also been reported from relaxation measurements.4 In fact, the only major change in magnetic susceptibility upon addition of IHP is found for the low spin form azide methemoglobin (35) for which the R 3 T transition is inacces-

Symp: Qua&. Biol. 36, 295-310 7. PERUTZ. M. F.. PULSINELLI. P. D. & RANNEY, H. M. (1972)

Nat&e New bol. 237,259-i63 8. GRAY, R. D. & GIBSON, Q. H. (1971) J. Biol. Chem. 246,5176-

5178 9. GRAY. R. D. & GIBSON. &. H. (1971) J. Biol. Chem. 246,7168-

7174 -

10. OLSON, J. S. & GIBSON, Q. H. (1971) J. Biol. Chem. 247, 1713- 1726

11. GIBSON, Q. H. (1973) Proc. Natl. Acad. Sci. U. S. A. 70, l-4 12. EDELSTEIN, S. J. (1971) Nature 230,224-227 13. HOPFIELD. J. J.. SHULMAN. R. G. & OGAWA, S. (1971) J. Mol.

Biol. 61; 425-443 14. OLSON, J. S. & GIBSON, Q. H. (1972) J. Biol. Chem. 247,3662-

3670 sible according to the Perutz hypothesis.

It seems, therefore, that the spin state of the iron is unrelated to conformation. Differences in reactivity of the R and T states would be due to differences in accessibility generated by structural differences in the globin. This view is consistent with the allosteric model in which the conformational transition results from the differential affinities of t,he R and T states which pull a pre-existing equilibrium and is also supported by a number of

independent experiments. For example, Parkhurst et al. (47)

4 A. M. Schwartz and P. R. Schimmel, personal communication.

15. MOFFAT, K., OLSON, J. S., GIBSON, Q. H. & KILMARTIN, J. V. (1973)‘J. &ol. Chem. 248,6387-6393

16. CASSOLY. R.. GIBSON. Q. H.. OGAWA. S. & SHULMAN, R. G. (1971) koChem. Biophys. Rk. Corn&n. 44, 1015-1021

L7. OGAWA, S. & SHULMAN, R. G. (1972) J. Mol. Biol. 70,315-336 18. OGAWA, S., MAYER, A. & SHULMAN, R. G. (1972) Biochem.

Biophys. Res. Commun. 49, 1485-1491 19. LINDSTROM, T. R., Ho, C. & PISCIOTTA, A. V. (1972) Nature

New Biol. 237, 263-264 20. EDELSTEIN, S. J. & GIBSON, Q. H. (1971) in Probes of Struc-

ture and Function of Macromolecules and Membranes, (CHANCE;, B., YONETANI, T. & MILDVAN, A. S., eds) pp. 417-429, Vol. II, Academic Press, New York


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

960

21. EDELSTEIN, S. J., RF,HMAR, M. J., OLSON, J. S. & GIBSON, Q. H. globin in Their Reactions with Ligands, North Holland Pub- (1970) J. Biol. Chem. 246.4372-4381 lishing Co., Amsterdam

22. A&~&N, M. E., MOFF~T, J. K. & GIBSON, Q. H. (1971) J. Biol. Chem. 246, 2796-2807

23. KELLETT, G. L. (1971) Nature New 6?iol. 234, 189-191 24. HEWITT, J. A., KILM~IRTIN, J. V., TE:NEYCK, L. F. & PKRUTZ.

25. M. F. (1972) PTOC. Natl. Acad. Sci. CT. S. A: 69,203-207

THOMAS. J. 0. & EDELSTEIN. S. J. (1972) J. Biol. Chem. 247.

26. 7870-7874

. I

THOMAS, J. 0. & EDELSTEIN, S. J. (1973) J. Biol. Chem. 248, 2901-2905

27. BENF,SCH, R. E., BENESCH, R. & RENTHAL, R. (1971) Nature New Biol. 234, 174-176

28. ARNONE, A. (1972) Nature 237, 146-149 29. HOARD, J. L., HAMOR, M. J., HAMOR, T. A. & CIIUGHEY,

30. 31. 32. 33.

W. S. (1965) J. Am. Chem. Sot. 87, 2312-2319 HOARD, J. L. (1971) Science 174, 1295-1302 PERUTZ, M. F. (1970) n’ature 228,726-739 PE:RUTZ, M. F. (1972) Nature 237, 495-499 PI~RUTZ, M. F., LADNER, J. E., SIMON, S. 12. & Ho, C. (1974)

Biochemistry 13, 2163-2173 34.

35.

PERUTZ. M. F., FERSHT, A. R., SIMON, S. R. & ROBERTS, G. C. F. (1974) &&hemistr& 13, 2174-2186

PERUTZ. M. F.. HEIDNER. F. J.. LADNIXR. J. E.. BEETLESTONE. J. G.,‘Ho, C’. & SL~\DE’, E. P. (1974) i+ioche&try 13, 21871 2200

39.

40. 41. 42.

43.

44.

45

46.

47.

48.

49.

50.

51.

36. GIBSON, Q. H. (1970) J. Biol. Chem. 246,3285-3288 52. 37. CREPEAU, R. H., EDXLSTEIN, S. J. & REHM‘ZR, M. J. (1972)

Anal. Biochem. 60,213-233 53. 38. ANTONINI, E. & BRUNORI, M. (1971) Hemoglobin and Myo-

NAGKL, R. L. & GIBSON, Q. H. (1971) J. Biol. Chem. 246, 69-73

KELLETT, G. L. & GUTFRF,UND, H. (1970) Nature 227, 921-926 GIBSON, Q. H. (1973) J. Biol. Chem. 248,1281-1284 ANTONINI, E. & BRUNORI, M. (1969) J. Biol. Chem. 244,3909-

3912 ANTONINI, E., WYMAN, J., MORETTI, R. & ROSSI-FANELLI, A.

(1963) Biochim. Biophys. Acta 71, 124-138 BOOKCHIN, It. M. & NAGEL, R. L. (1971) J. Mol. Biol. 60,

263-270 BOOKCHIN, R. M. & NAGXL, R. L. (1973) J. Mol. Biol. 76,

233-239 ALPERT, Y., BANERJEE, R. & DENIS, J. (1973) Nature New

Biol. 243, 80-81 PARKHURST, L. J., GERACI, G. & GIBSON, Q. H. (1970) J. Biol.

Chem. 246, 4131-4135 NOBLE, R.. W., Rossr, G. & BERNI, R. (1972) J. Mol. Biol. 70,

689-696 HOFFMAN, B. M. & PETERING, D. H. (1970) PTOC. Natl. Acad.

Sci. U. S. A. 67,637-643 DICKINSON, L. C. & CHIEN, J. C. W. (1973) J. Biol. Chem. 248,

5005-5011 CASSOLY, R. (1974) C. R. Acad. Sci., in press RSIN, H., RIST.\U, 0. & SCHELER, W. (1972) Fed. Eur.

Biochem. Sot. Lett. 24,24 CREPEAU, R. II., HF,NSLEY, C. P. & EDELSTEIN, S. J. (1974)

Biochemistry, l&4860-4865


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

P Hensley, S J Edelstein, D C Wharton and Q H GibsonConformation and spin state in methemoglobin.

1975, 250:952-960.J. Biol. Chem.

http://www.jbc.org/content/250/3/952Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here

http://www.jbc.org/content/250/3/952.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


ww

.jbc.org/D

ownloaded from

http://www.jbc.org/content/250/3/952

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;250/3/952&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/250/3/952

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=250/3/952&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/250/3/952

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/content/250/3/952.full.html#ref-list-1

http://www.jbc.org/

the journal of bioloqical chemistry vol. …the journal of bioloqical chemistry vol. 250, no. 3,...

Documents