Oxygen Equilibrium Characteristics of Abnormal

Hemoglobins: Hirose (a2 0 37Ser) L Ferrara ( 47Gly 2

Broussais (a2 90Asfli2), and Dhofar (a2258Arg)

SHIGERUFUJrrA

From the First Department of Medicine, Faculty of Medicine, KqJushuUniversity, Fukuoka, Japan

A BS T R A C T The oxygen equilibrium characteristicsof four structural variants of hemoglobin A were cor-related with their amino acid substitutions.

Hemoglobin Dhofar, in which the proline at E2(58)9is replaced by arginine, had normiial oxygen equilibriumcharacteristics.

Hemoglobin L Ferrara, in which the aspartic acidat CD5(47)a is replaced by glycine, and hemoglobinBroussais, in which the lvsine at FG2(90)a is replacedby asparagine, both showed a slightly elevated oxygenaffinity; nevertheless both demonstrated a normal heme-heme interaction and a normal Bohr effect.

Hemoglobin Hirose, in which the tryptophan at C3(37),3 is replaced by serine, showed abnormalities of alloxygen equilibrium characteristics; i.e., increased oxy-gen affinity, diminished heme-heme interaction, andreduced Bohr effect.

These results suggest that aspartic acid at CD5(47)aand lysine at FG2(90)a are involved in the functionof the hemoglobin molecule, despite the fact that thesepositions are not located directlv in the heme or thea-,8-contact regions.

Tryptophan at C3 (37)13 is located at contact be-twveen al- and /32-subunits. It is suggested that the sub-stitution by serine might disturb the quarternary struc-

ture of the mutant hemoglobin molecule during transi-tion from oxy-form to deoxy-form resulting in an alter-ation of the heme function.

This work was presented in part at the 33rd Annual Meet-ing of the Japan Hematological Society.

Dr. Fujita's present address is Division of HumanGenetics, Department of Medicine, Cornell University Medi-cal College, New York.

Received for publicationt 6 Auguist 1971 and in revised formn13 June 1972.

INTRODUCTION

Since Pauliig. Itano, Singer, and Wells (1) denmon-strated sickle cell hemoglobin in 1949, over a hundredstructural variants of human hemoglobin have been re-ported. In some cases, a substitution of a single amiinoacid residue alters the functional properties of hemo-globin and the mutant hemoglobin is associated withclinical manifestations.

The introduction of X-ray crystallography analysis fa-cilitated the description of the detailed architecture andthe construction of atomic model of hemoglobin molecule(2-6).

Correlated studies of structure and function of heimio-globins have contributed not only to our understandingof the disordered mechaniisimis resulting from molecularalterations, but also to knowledge of the interrelationbetween structure and function of the norimial hemo-globin nolecule (7).

Since 1957, the Biochemical Laboratory of the FirstDepartment of 'Medicine, Faculty of Medicine, KvuslhuUniversitv has examined electrophoretically over 50.000blood specimens from successive clinic patients with theaim of detecting hemoglobin variants. As a consequeince,11 kindreds with inherited structural variants have beendiscovered and in 9 of them the amino acid sequenceswere identified (8-15). The present paper deals wN-ithoxygen equilibrium characteristics of four of these-hemoglobin L Ferrara (a247GIYp2) (8, 16), hemoglobinBroussais (a0asnp2) ( 13, 17), hemoglobin Dhofar (a2p58Ar2 )(12, 18), and hemoglobin Hirose(a,#23ser) (14). Thefunctional properties of these hemoglobins were discussedon the molecular level based on the atomic model pro-posed by Perutz and his colleagues, and comments were

made on the pathophysiological mechanisms involved inthe functional aberrations.

.2520 The Journal of Clinical Investigation Volume 51 October 1972

., ~~~1,~~ ~ ~~~.I

0 LO 2D

log P02

FIGURE 1 Oxygen equilibria of unfractionated hemolysatesfrom a normal adult (O----O) and from a heterozygotefor hemoglobin Hirose ( ,and those of hemo-globins A (0- O) and Hirose (-*4) isolated fromthe same hemolysate. The data were obtained at pH 7.01,20°C. Hemoglobin concentrations were 0.117o in 0.1 M phos-phate buffer. The curves were plotted according to Hill'sequation.

METHODS

Preparation of hemoglobin compontentts. Blood sampleswere taken from the cubital veins of normal and affectedindividuals using heparin sodium as the anticoagulant. Thered blood cells were separated from sera and washed fourtimes with cold physiological saline solution. Hemolysateswere prepared according to the method of Drabkin (19).Erythrocytes were lysed by mixing 1 vol of washed packedcells, 1 vol of cold distilled water, and 0.5 vol of toluene.From the hemolysates which had been refrigerated over-

night, the hemoglobin layer was separated by centrifugation.Purification of hemoglobin A and abnormal hemoglobins

was carried out on starch block electrophoresis in 0.05 M

sodium-barbital buffer at pH 8.6 according to Kunkel andWallenius (20), or on column chromatography of DEAESephadex A-50 (Pharmacia Fine Chemicals, Inc., Pisca-taway, N. J.) using Tris-HCl buffer according to Huismanand Dozy (21). The purity of the separated hemoglobinswas confirmed by thin starch gel electrophoresis with Tris-EDTA-borate buffer at pH 8.6 (10).

The purified hemoglobins, the unfractionated hemolysatescontaining abnormal hemoglobins (abnormal hemolysates),and the unfractionated hemolysates from normal individuals

(normal hemolysates) were dialyzed against 0.1 M phos-phate buffer (Na2HPO4-KH2PO4) at 40C for 20 hr. Afterdialysis, each sample was diluted to a 0.1% hemoglobin con-centration and the oxygen equilibrium was measured at200C at different pHs. The hemoglobin concentration wasdetermined spectrophotometrically after conversion to pyri-dine hemochromogen (22).

Measurement of oxygen eqilibrium of hemoglobins. Theoxygen equilibrium of hemoglobin was recorded automati-cally as a successive deoxygenation curve according to themethod of Imai, Morimoto, Kotani, Watari, Hirata, andKuroda (23). The oxygen partial pressure (P02) in thesample was measured with a Beckman polarographic oxygensensor, model 39065 (Beckman Instruments, Inc., Fullerton,Calif.), and the percentage of oxyhemoglobin was estimatedspectrophotometrically using a monochromatic light beam at560 m,A. The values were recorded continuously on an X-Yrecording chart. The temperature of the samples was mea-sured by a thermistor and maintained at 20°C to within±0.10°C by thermodules throughout the measurement of oxy-gen equilibria. The curves were readily reproducible; themaximum standard error was approximately 3% near thehalf saturation point.

Before and after the oxygen equilibrium measurement,the visible absorption spectra of samples were recorded by

1.2

1.0

to

Q80

-i

0.4

3.0

c 2.0

1.0

6b0 6.5 7.0pH

7.5 80

FIGuRE 2 pH dependence of oxygen affinity and values ofHill constant n for hemoglobin A (0) and hemoglobinHirose ( 0 ).

Oxygen Equilibrium Characteristics of Abnormal Hemoglobins

o a

o 00

0~~~~~~~~

2521

I','

50

01 10

OXYGEN PRESSURE(mmHg)100

FIGLRE 3 Oxygen equilibrium curves of lhenlolvsate from a heterozygotefor hemoglobini Hirose (- - ---), purified hemoglobin A (- - - -) and puri-fied hemoglobin Hirose ( ) under the same conditionis as in Fig. 1. Thedotted line is the curve calculated from the data of hemoglobin A andlhemoglobini Hirose oIn the assumption that the lhemolysate contains 40%lhemoglobini Hirose.

a Beckman DK-2 self-recording spectrophotometer (Beck-man Instruments, Inc.) in order to assess the amount ofmethemoglobin wlhich was formed (during the measurementof oxygen equilibrium. Methemoglobin in the samples was

calculated from extinction coefficients (24). The quantity ofmethemoglobin formed during the oxygeni equilibrium study

was less than 7%.The fractional oxygen saturation of hemoglobin, Y, w-as

calculatedl by following formula:

Y= (ODMeoxy-01)) ( OD(ieoxy-ODoxy)

where ODdeoxy and ODoxy are the optical denisities of de-oxygenated and oxygenated lhemoglobin respectively and

D)D are the optical densities coniverted from the transmit-tances measured during the oxygen equilibrium study.

The oxygeni affinity of hemoglobiin wvas exp)ressed by P50w-hich is the oxygen partial pressure at half saturation of

hemoglobin with oxygen. pH (lepindeince of oxygen affinity

of hemoglobin, i.e. the Bohr effect (25), was estimated fromthe values of P50 at pH values ranging from 7 to 8. Theformula y = Alog P-,/ApH was usedl.

Values of n in Hill's equation (26), which is the numeri-cal expression for heme-heme interaction, were calculatedfrom the most linear part of the slope in curves of thelog(Y/(1 - Y) ) plotted againist log Po2.

The oxygen equilibrium studies were carried out within7 days after the blood had beeni taken.

RESULTS

Oxygen equilibriumn characteristics of hciieo globinHirose (aprSer). The oxygen equilibrium curves ofpurified hemoglobin Hirose and hemoglobin A are shownin Fig. 1. It can be noted that the oxygen equilibriumcurve of hemoglobin Hirose is shifted markedly to the

left of heimioglobin A and that the slope of the curve is

decreased. These findinigs indicate a increased oxygen

affinity and reduced hemiie-lheme interaction of hemloglobiniHirose. The P- at plH 7.01, 20°C of hemiioglobin Hirosewas oinly 2.6 mmHg compared Nith 11 mmHg for he-

moglobin A. The average value of n for hemoglobinHirose N-as 1.48±0.24 (SD, N = 5) compared witlh 2.97±0.21 (SD, N = 5) for purified hemoglobin A (Fig. 2).A pH increase from 6.5 to /7.9 raises the n value of hemo-globin Hirose froml 1.26 to 1.88. The Bohr effect of hemo-globin Hirose is markedly decreased: the a-value was

- 0.13 for hemiioglobiln Hirose and -0.53 for hemoglobinA.

The unfractionated hemolysate of an individual heter-ozygous for hemoglobin Hirose contained approximately40% hemiioglobin Hirose. The oxygen equilibrium curve

of the abnormal unfractionated hemolvsate is shiftedto a position approximately intermediate between thoseof purified lhemoglobin Hirose and purified hemoglobillA (Fig. 1). The curve show-s a biphalsic configuration,more analogous to that of hemiioglobini Hirose at the lower

part and more analogous to that of hemoglobin A at theupper part of the curve. The oxygen dissociation curves

shlowni in Fig. 3 illustrate the close agreement betweenthe observed dissociation curve in a hemolvsate from a

heterozygous carrier and the expected curve calculatedfrom the purified mutant and normal hemoglobins.

Oxygen equtilibriumn characteristics of hemnoglobin LFerrara.(asIGIY Po). As shown in Fig. 4, unfractionatedhemolysate which contained approximately 16% hemo-

2522 S. Fujita

z0

z0

x

0

zwi

u

/1//i,/

//

/ ~ ~~~I//~

/ /

globin L Ferrara, showed normal functional propertieswhen compared with an unfractionated normal hemoly-sate.

The oxygen affinity of purified hemoglobin L Ferrarawas increased when compared with purified hemoglobinA (Fig. 4). P.O at pH 7.02, 20°C was 7.5 mmHg forhemoglobin L Ferrara and 10.9 mmHg for hemoglobinA. The average n value in Hill's equation was 2.67±0.24(SD, N =5) for hemoglobin L Ferrara and 2.91±0.02(SD, N = 5) for hemoglobin A. The Bohr effect was sim-ilar in both hemoglobins (Fig. 5). The 'a-value was

0.51 for hemoglobin L Ferrara and - 0.56 for hemo-globin A.

Oxygen equilibrium characteristics of hemoglobinBroussais(as 90sAP,i). Unfractionated hemolysate from an

individual heterozygous for hemoglobin Broussais showedoxygen equilibrium characteristics similar to that of un-fractionated hemolysate from a normal individual (Fig.

0

0 A02.

log '02FIGURE 4 Oxygen equilibria of unfractionated hemolysatesfrom a normal adult (O----O) and from a heterozygotefor hemoglobin L Ferrara ( ---*), and those of hemo-globins A (0O O) and L Ferrara (* *)isolatedfrom the same hemolysate. The'data were obtained at pH7.02, 20°C. Hemoglobin concentrations were 0.19'o in 0.1 Mphosphate buffer. The curves were plotted according toHill's equation.

1.2

1.o [

0

-j

0.8[

0.6

0.4.

3.0

6.0 6.5 7.0

pH7.5 &0

FIGuRE 5 pH dependence of oxygen affinity and values ofHill constant n for hemoglobin A (0) and hemoglobin LFerrara (*).

6). The abnormal hemolysate contained, however, only17% hemoglobin Broussais.

Oxygen equilibrium curves of purified hemoglobinBroussais and purified hemoglobin A are shown in Fig.6. The curve for hemoglobin Broussais is shifted slightlyleft as compared with that of hemoglobin A under thesame experimental contditions. P5o at pH 7.04, 20°C was9.6 mmHg for hemoglobin Broussais 'and 10.6 mmHgfor hemoglobin A. The average n value in Hill's equationwas 2.41±0.16 (SD, N = 5) for hemoglobin Broussaisand 2.56+0.13 (SD, N = 5) for hemoglobin A. The Bohreffect was similar in both hemoglobins (Fig. 7).

The oxygen affinity of purified hemoglobin A was

slightly higher than that of the unfractionated normalhemolysate. This may imply that the functional proper-ties of the hemoglobin are artificially modified duringpurification. If the purification procedure affected hemo-globin Broussais to a greater degree than hemoglobinA, it might result in the slightly increased oxygen affinityin the former. Before and after the measurement of oxy-gen equilibrium, the visible spectrum (450-700 m/A) ofpurified hemoglobin Broussais was compared with puri-fied hemoglobin A. No differences were discovered.

Oxygen Equilibrium Characteristics of Abnormal Hemoglobins

Ia

o 6 0 0 0

0~~~~~~6~~~~~~~~

@~

'.2 5. 26J3

0

log P02

FIGURE 6 Oxygen equilibria of unfractionated hemolysatesfrom a normal adult (0----O) and from a heterozygotefor hemoglobin Broussais (o--), and those of hemo-globins A (0 O) and Broussais (* *) isolatedfrom the same hemolysate. The data were obtained at pH7.04, 20°C. Hemoglobin concentrations were 0.1% in 0.1 m

phosphate buffer. The curves were plotted according toHill's equation.

Differential alteration of hemiioglobin A and hemoglobinBroussais by the isolation procedure is not excluded;nevertheless the slight increase in oxygen affinity appears

more likely to be attributable to the properties of hemo-

globin Broussais itself.Oxygen, cquilibriu st cliaractcristics of 1lic, oglobinl

Dhofar(aej825.-r"). The hemolysate from an individualheterozygous for hemoglobin Dhofar contained approxi-mately 50% hemoglobin Dhofar. The abnormal hemolv-sate showed similar oxygen equilibrium characteristicsto that of an unfractionated normal hemolysate (Fig. 8).

The oxygen equilibrium curves of purified hemoglobinDhofar and purified hemoglobin A are shown in Fig. 8.P5o at pH 6.95, 20°C was 8.6 mmHg for hemoglobinDhofar and 8.5 mmHg for hemoglobin A. The average

n value in Hill's equation was 2.69+0.04 (SD, N = 5) for

hemoglobin Dhofar and 2.68±0.03 (SD, N = 5) for he-

moglobin A. The Bohr effect was similar in both hemo-

globins (Fig. 9). The value of P.0 for hemoglobin A in

this experiment is smaller than those obtained in theother experiments. It was assumed, therefore, that thepurification procedure altered slightly the normal hemo-

globin A and possibly also hemoglobin Dhofar. Never-theless, oxygen equilibrium characteristics-oxygen affi-nity, n values and the Bohr effects of purified hemo-

globin Dlhofar and hemoglobin A were indistinguishable.Unfractionated hemolvsate containing approximately50% henmoglobin Dhofar also showed oxygen equilibriumcharacteristics identical with unfractionated normal he-

molNsate. The functional properties of henmoglobinDhofar appeared to be similar to that of hemoglobin A.

DISCUSSION

It is assumed that most amino acid residues at the exter-

nal surface of the hemoglobin molecule do not influencethe function of hemoglobin (7) and consequently theirreplacenment would not affect the functional propertiesof the molecule. AMany mutant hemoglobiins with aminoacid substitutions occurring at the external surface ofthe molecule exhibit nornmal oxvgen equilibrium char-

acteristics, wNith the exception of a fewx variants (27-30).Both hemiioglobins Broussais and L Ferrara have an

1.2

1.0 [0

fll

X 0.80

-J

0.6[

0.4.~

o0

c

2.0

6.0 6.5 7.0 7.5 &0pH

FIGURE 7 pH dependence of oxygen affinity and values of

Hill constant n for hemoglobin A (0) and hemoglobinBroussais ( O ).

2524 S. Fujita

O 0

0

CD~~~~~~~~~~~~~~~~~~~I

O L0

log '02FIGURE 8 Oxygen equilibria of unfractionated hemolysatesfrom a normal adult (O --O) and from a heterozygotefor hemoglobin Dhofar (*-- ,and those of hemo-globins A (0O O) and Dhofar (* *) isolated fromthe same hemolysate. The data were obtained at pH 6.95,20'C. Hemoglobin concentrations were 0.1% in 0.1 M phos-phate buffer. The curves were plotted according to Hill'sequation.

amino acid substitution at the external surface of thehemoglobin molecule. They showed slightly elevated oxy-gen affinity with normal herne-heme interaction and nor-mal Bohr effect. X-ray crystallography analysis at 5.5 Aresolution (3) suggested that in hemoglobin A the sidechain of lysine at FG2(90)a may form a salt bridge withthe propionic acid side chain of heme, although this hv-pothesis could not be confirmed by analysis at 2.8 A reso-lution (6). The lysine at FG2(90)a is replaced bv as-

paragine in hemoglobin Broussais. The mechanism bywhich the amino acid substitution at this site in hemo-globin Broussais alters the oxygen affinity of the moleculeis not clear. However, lysine at FG2(90) a is an invariantresidue in all mammalianl hemoglobins whose primarystructures have been determined (7). This residue mavhave some special role on the functional properties ofhemoglobin.

Nagel, Ranney, Bradley, Jacobs, and Udem (31) re-ported that hemoglobin L Ferrara had almost normal

functional properties, and had a P5o value smaller thanthat of hemoglobin A by 21% at pH 7.4, 10°C. In thepresent study, values of P5o in hemoglobin L Ferrara were

23-31% smaller than those of hemoglobin A. This dif-ference between hemoglobin A and hemoglobin L Ferrarais significant. We concluded, therefore, that hemoglobinL Ferrara has a higher oxygen affinity than hemoglobinA.

In hemoglobin L Ferrara the amino acid substitutioninvolves the site of CD5(47) of the a-chain. On the CDcorner of the a-chain in hemoglobin A amino acid resi-dues at CD1(43), CD3(45) and CD4(46) have contactwith heme but aspartic acid at CD5(47) is located at theexternal suface of the hemoglobin molecule and has no

contact with heme or other subunits. The oxygen affinityof hemoglobin increases when this aspartic acid is re-

placed by glycine in hemoglobin L Ferrara. Nagel et al.(31) also reported that hemoglobin L Ferrara showedheat unstability at 55°C or higher temperature. In a poly-peptide chain, glycine is permitted conformations whichare forbidden to aspartic acid or other amino acids (32).

1.2j

1.0

0

In

Ji

0.61.

0.41.

3.0

c

2.0

6.0 6.5 7.0 7.5 8.0pH

FIGURE 9 pH dependence of oxygen affinity and values ofHill constant n for hemoglobin A (0) and hemoglobinDhofar ( 0 ).

Oxygen Equilibrium Characteristics of Abnormal Hemoglobins

* ffi i

a * a

0.8[

2525

TABLE IFunctional Properties of Abnormal Hemoglobins with A4mino Acid Substitution at ail-02-Contact

Oxygen Heme-heme BohrDesignation Amino acid substitution affinity interaction effect Reference

J Cape town FG4(92)a Arg - Glin High Diminished Normal 38Chesapeake FG4(92)a Arg - Leu High Diminished Normal 36G Georgia G2 (95)a Pro Leu High Dliminished Reduced 40Rampa G2 (95)a Pro Ser High Diminished Red tced 40Yakima Gl (99)# Asp His High Diminished Normnal 39Kempsey GI (99)0 Asp Asoi High Diminished Normal 37Kansas G4(102)o Asni Thr Low Diminished Normal 41Hirose C3 (37)0 Try -Ser High Diminished Reduced This paiper

In hemoglobin L Ferrara, therefore, the conform-lationi ofthe CD corner of the a-chain may be different froml thatof hemoglobin A, resulting in a clhange of hemiie funictionanld unstability of structure.

It is of interest that in hemoglol)in Hasharon the re-placement of aspartic acid at CD5(47)a by histi(lilnemlakes the structure of the mutaint hemoglobini moleculetinstable (33, 34). The present data in regardls to thehiglh oxygen affinity of hemiioglobini L Ferrara and struc-tural unstability of hemoglobiins L Ferrara and Hasliaronindicate that aspartic acid at CD5(47)a in hemoglobinA plays an important role in maintenance of normalstructure and function of the hemoglobin molecule, al-thouglh the atomic model of this position does not indi-cate this.

In lhemloglobin Dhofar proline at E2 (58) P is replaced1w) argininie. Proline residues play anl important role inthe conformation of the helical region of the molecule(2, 32). In hemoglobin Dhofar the amliino acid substitui-tion occurs at the helix E w-hiclh participates in the for-nmation of the heme pocket anid wlhichi containis the distalhistidine. It could be expected, therefore, that substantialfunctional changes would be induced by the substitutionof arginine for proline. Contrary to our expectation,hemoglobin Dhofar does not differ significalntly fromhemoglobin A in oxygen affinity, heimie-heme interaction,or the Bohr effect. During oxygenatioin of hemoglobin,the distance between the porphyrin alnd helix E is widenedto make room for the oxygen molecule (35). It can beargued that since the E2(58)P position is near the cornerof helix E at the outer surface of the molecule and argi-nine can be thus permitted to extend its long side chainexternally without disturbing neighboring residues orsubunits, the amino acid substitution does nIot effect thenmovemelnt of the helix E anid, thus., the functional proper-ties of the molecule remzaini uniialteredl.

Hemoglobin Hi rose involves an amillno acid substitutionat the al-P2-contact: trvptoplian at (3 (37) P is replacedby serine. Tryptophan at C3(37)tP in hemiioglobini A hascontacts witlh five amiiino acid residues of the a-chain in-

cluding FG4(92) aArg, FG5(93) aVal, Gl (94) aAsp,G2(95)aPro, and HC2(140)aTyr (6). These contactscomprise 28 atoms which are miore than one-third of theatoms making up the ai-P2-contact (6), and conceivablysuclh contacts canl not be adequately maintained in hemo-globin Hirose.

During tranisitioln fr-oim oxy-formi to deoxy-formi of thelhemoglobiin mlolecule, the relative displacement of the,P-chain to the a-chain is greater at the as-,P2-contact thaniat the as-Pi-contact. The relative displacemenit of atomsat the former conitact can be as much as 5.7 A (4, 6).Oxygen equilibrium characteristics of seveni abnormalhemoglobins, in wthich amino acid substitutions occur atthe al-,82-contact, have been reported (36-41). Six of themhave high oxygen affinity and one has low oxygen affinity.All of them show reduced heme-lheme interaction, anidtwo of them show reduced Bohr effect (Table I).Hemoglobin Hirose also showed a reduced Bohr effect'in addition to high oxygen affinity and diminished hemle-heme interaction.

In hemiioglobin Hirose, breakdown of conitact betweentrvptophan at C3(37)P and tyrosine at HC2(140)a seenmschiefly to account for the functional disturbance.

The importance of the al-P2-contact in the transmissionof heme-heme interaction has been pointed out by severalinvestigators (6, 7). Briehl and Hobbs (42) suggestedon the basis of ultraviolet spectrum studies of hemoglo-bin that tryptophan at C3 (37),P played an importantrole on interchain interaction. Perutz (35) also showedlthat contact between tryptophan at C3(37) P and tyro-sine at HC2(140)a took part in conformational changeof P-chain from oxy-form to deoxy-form. The breakdow-nof this contact may disturb the conformational change ofthe P-chain during the oxygenation-deoxygenation reac-tion, resulting in decreased heme-heme interaction andhigh oxygen affinity.

The C-terminal histidine of the P-chain conitribtutesto 50'% of the Bohr effect (43) anid a-amino grousof the a-chaini contributes to another 25'%, of theBohr effect (44). The sites or amino acid residutes

2.526 S. Fujita

which may contribute to the remaining one-quarterof the Bohr effect have still not been identified precisely.Tryptophan at C3(37)P has no direct contact withl thoseresidues responsible for the Bohr effect. Since tyrosineat HC2(140) a contributes to the liberation of Bohr pro-ton bound to valine at NA1(1)a (35), breakdown ofcontact between tryptophan at C3(37)1 and tyrosine atHC2(140)a may influence the release of Bohr proton bychanges in quarternary conformation in the transitionfrom deoxy-form to oxy-form. This can not explain allof the decreased Bohr effect in hemoglobin Hirose. He-moglobin Hirose moves nmore slowly than hemoglobin Aon starch gel electrophoresis at pH 8.6, although trypto-phani and serine are both neutral amino acids. It islikely that the amino acid replacement in hemoglobinHirose affects not only the al-A%-contact but also the stericconformation of the molecule.

It is the general impression that high oxygen affinityof structural variants other than unstable hemoglobinsusually leads to erythrocytosis (36-40) and low oxygenaffinity results in reduced hemoglobin concentration inthe peripheral blood (15, 29). In spite of the obvious ab-normality in oxygen equilibrium characteristics, nospecific hematological and clinical findings due to theabnormal hemoglobins were observed in the cases heter-ozygous for hemoglobin L Ferrara or hemoglobinHirose.

In hemoglobin L Ferrara and hemoglobin Broussais,the change in oxygen affinity is slight and the proportionof the abnormal hemoglobins in hemolysates is quitesmall. Unfractionated hemolysates from these individualshave, therefore, oxygen equilibrium characteristics simi-lar to that of unfractionated normal hemolysates. Thecarriers of these abnormal hemoglobins do not showclinical or hematological manifestations resulting fromthe abnormal hemoglobins.

Hemoglobin Hirose is very aberrant in all its oxygenequilibrium characteristics, despite the fact that no ab-normal hematologic findings were demonstrable. Theoxygen equilibrium curve of an unfractionated hemolv-sate, containing approximately 40% hemoglobin Hirose,was biphasic showing a very high oxygen affinity at lowoxygen tension, whereas the upper part of the oxygenequilibrium curve resembled that of hemoglobin A inshape and position.

In the absence of an interaction between hemoglobinA and a high oxygen affinity mutant, the oxygen equi-librium curve of the hemolysate containing these twocomponents should reflect the presence of the abnormalhemoglobin at low oxygen tension and hemoglobin Aat high oxygen tension. If the two hemoglobins are notindependent in reacting with oxygen, the entire oxygena-tion curve should be shifted to the left of hemoglobin A(45). Hemoglobin Hirose corresponds to the former case

(Fig. 3), and under physiological conditions, that is,above 75% oxygen saturation of blood (46), hemoglobinA chiefly contribute to exchange of oxygen. Becausethe carriers of hemoglobin Hirose show no hematologicalabnormality, there must exist some way of compensatingfor the disturbed heme function other than increase inerythropoiesis.

As seen in the present cases, structural and functionalaberrations resulting from molecular alteration do notalways seem to correspond with the quality and the quan-tity of clinical manifestations. The following factors areall involved with the manifestation of clinical symptomsresulting from disturbed heme function: (a) magnitudeof change in oxygen affinity of henmoglobin, (b) propor-tion of abnormal hemioglobin in the hemolysate, (c) in-teraction between normal and abnormal hemoglobins inthe heterozygous state during interaction with oxygen,and (d) reactivity of abnormal hemoglobin with organicphosphates prescent in erythrocytes, such as 2,3-diphos-phoglycerate (DPG). The latter factor has been shown toaffect the oxygen affinity of hemoglobin (47, 48). Hemo-globin A has higher oxygen affinity than hemoglobin F,but in the presence of DPG, the comparative oxygen affini-ties are reversed (49). If an amino acid substitution al-ters the affinity of hemoglobin for DPG, then the hemo-globin will behave in a different way "in vivo" as com-pared with "in vitro" where DPGor other organic phos-phates are not present. The effect of DPGor other or-ganic phosphates on function of hemoglobin Hirose hasnot yet been clarified. Further study is necessary to in-terpret the reasons for the normal hematological findingsin the presence of the disturbed heme function in indi-viduals heterozygous for hemoglobin Hirose.

ACKNOWLEDGMENTSThe author wishes to express his thanks to Professor T.Yanase, Dr. M. Hanada, and Dr. Y. Ohta of Kyushu Uni-versity for their continuous support and advice; to Dr. K.Imai of Osaka University for arrangement of the instrumentassembly for oxygen equilibrium studies; to Dr. H. B.Hamilton of Atomic Bomb Casualty Commission in Hiro-shima for his valuable suggestions; and to Dr. H. Cleveand Dr. S. D. Litwin of Cornell University for their carefulreading of this manuscript. The author is indebted to Dr.T. Kumamoto of Moji Hospital of Japanese National Rail-way Corporation, Dr. T. Fujimura, Dr. K. Kawasaki, andDr. K. Yamaoka of Kyushu University for help in obtain-ing the blood samples.

REFERENCES1. Pauling, L., H. A. Itano, S. J. Singer, and I. C. Wells.

1949. Sickle cell anemia, a molecular disease. Science(Wash. D. C.). 110: 543.

2. Perutz, M. F. 1965. Structure and function of hemo-globin. I. A tentative atomic model of horse oxyhemo-globin. J. Mol. Biol. 13: 646.

Oxygen Equilibrium Characteristics of Abnormal Hemoglobins 25a27

3. Perutz, M. F., J. C. Kendrew, and H. C. Watson. 1965.Structure and function of hemoglobin. II. Some relationsbetween polypeptide chain configuration and amino acidsequence. J. lIol. Biol. 13: 669.

4. Muirhead, H., J. MX1. Cox, L. Mazzarella, and A. F.Perutz. 1967. Structure and function of hemoglobin.III. A three-dimensional Fourier synthesis of humandeoxyhemoglobin at 5.5 A resolution. J. Mllol. Biol. 28:117.

5. Bolton, WV., J. MI. Cox, and NI. F. Perutz. 1968. Struc-ture and functioni of hemoglobin. IN. A three-dimen-sional Fourier synthesis of horse dleoxyhemoglobin at5.5 A resolution. J. MIol. Biol. 33: 283.

6. Perutz, M. F., H. Muirhead, J. 'M. Cox, and L. C. G.Goaman. 1968. Three-dimensionial Fourier synithesis ofhorse oxyhemoglobin at 2.8 A reso'lutioni: the a+omicmodel. Natuire (Louid.). 219: 131.

7. Perutz, M. F., and H. Lehmanan. 1968. Miolecular l)ath-ology of human hemoglobiin. ANatuore (Louid.). 219: 902.

8. Hanada, MI., NM. Seita, I. Ohya, and K. Yamaoka. 1963.Studies on abnormal hemoglobins. (II) The structuralabnormalities of hemoglobin Shimonoseki, hemoglobinKokura and hemoglobin Umi. Proc. Svi, ip. Cheni.Physiol. Path ol. 3: 136.

9. Hanada, M., Y. Ohta, T. Imamura, T. Fujimura, K.Kawasaki, K. Yamaoka, and 'I. Seita. 1964. Studies onabnormal hemoglobinis in w-estern Japran. Jfp. J. Hunti.Geniet. 9: 253.

10. Imamura, T. 1966. Hemoglobin Kagoshima: an ex-ample of hemoglobin Norfolk in a Japanese family. Amii.J. Hul111. Gec;ct. 18: 584.

11. Ohta, Y., T. Imamura, T. Fujimura, K. Kawasaki, K.Yamaoka, and M. Hanada. 1967. A chemical abnor-mality in hemoglobin MIiyada from Japanese individuals.Jfp. J. Hum1tl. Gcuct. 12: 127.

12. Kawasaki, K., -Y. Ohta, T. Imamura, and( AI. Hanada.1967. Characterization of hemiioglobini Yukuhashi. Jfp.J. C/in. Hemi(atol. 8: 175.

13. Fujimura, T., Y. Ohta, T. Imaiimura, K. Kawasaki, K.Yamaoka, and M. Hanada. 1967. Characterization ofhemoglobin Tagawa-I. .4cta Haciiiotol. Jfp. 30: 639.

14. Yamaoka, K. 1971. Hemoglobin Hirose a2/3237 (C3)tryptophan yielding serine. Blood J. Hcnioatol. 38: 730.

15. Imamura, T., S. Fujita, Y. Ohta, M. Hanada, and T.Yanase. 1969. Hemoglobin Yoshizuka (G1O(108)/3 as-paragine -> aspartic acid) a newx- varianit wsith a re-duced oxygen affinity from a Japaanese family. .1. Clin.Invest. 48: 2341.

16. Lehmanin, H., and R. XV. Carrell. 1969. Variations in thestructure of human hemoglobin. With particular refer-ence to the unstable hemoglobins. Br. .Mled. Bull. 25: 14.

17. de Traverse, P. M1., H. Lehmanin, M. L. Coquelet, D.Beale, and W. A. Isaacs. 1966. Etude d'une hemoglobineJa non encore decrite, dans une familie francaise. C. R.Soc. Biol. 160: 2270.

18. Marengo-Rowe, A. J., P. A. Lorkin, E. Gallo, andH. Lehmann. 1968. Hemoglobin Dhofar-a new variantfrom Southern Arabia. Biochinm. Biophvs. Acta. 168: 58.

19. Drabkin, D. L. 1946. Spectrophotometric studies. XIV.The crystallographic and optical properties of the hemo-globin of man in comparison with those of other species.J. Biol. Cheiwi. 164: 703.

20. Kunkel, H. G., and G. WVallenius. 1955. New hemoglobinin normal adult blood. Scienlce (Wash. D. C.). 122: 288.

21. Huisman, T. H. J., and A. M. Dozy. 1965. Studies onithe heterogeneity of hemoglobin. IX. The use of tris(hy-droxymethyl) aminomethane-HCI buffers in the anion-exchange chromatography of hemoglobins. J. Chrom11a-togr. 19: 160.

22. Paul. K. G., H. Theorell, and A. Akeson. 1953. Themolar light absorption of pyridine ferroprotoporl,hyrill(pyridine hemochromogeni). .Act(a Clheiwi. Scand. 7: 1284.

23. Imai, K.. H. MIorimoto, NI. Kotani, H. XVatari, WV.Hirata, and M. Kuroda. 1970. Studies oni the funictionof abnormal hemoglobins I. AnI improved method forautomatic measuremelnt of the oxygen equilibrium curveof hemog'obitn. Biochlin. Bioplhvs. Acta. 200: 189.

24. Beniesclh, R., G. Maccluff, and( R. E. Benesch. 1965. De-termination of oxy)gen equilibria wvith a versatile newtonometer. Anial. BIio-lchn. 11: 81.

25. Bolr, C.. K. Hasselbal, andl A. Krogh. 1904. Uebereinieni in biologischer Beziehung wichtigen Einfluss, (lendie Kohlen-saurespaninunig des Blutes auf desseni Sauer-stoffbindung uebt. Skauid. 4rcli. Phlvsiol. 16: 402.

26. Hill, A. V. 1910. The possible effects of the aggregationof the molecules of hemoglobin on its dissociationcurves. J. Phv,siol. (Loiid.). 40: P4.

27. Bookchiin, R. \f., R. L. Nagel, and H. 'M. Raniney.1967. Structure and I)roperties of lhemoglobin C Har-lem, a human hemoglobin variant with amino acid sub-stitutions in 2 residues of the g-polypeptide chain. J.Biol. C/1lte l. 242: 248.

28. Huisman, T. H. J., J. Still, and(l C. -M. Nechtman. 1963.The oxygen eqjuilibria of some slow-moving humanhemiolobini types. Bioc/timi. Biophl vs. Acta. 74: 69.

29. Stamatoyannopoulos, G., J. T. Parer, and C. A. Finch.1969. Physiologic implications of a lenmoloblin with de-creased oxygen affinity (lhemoglobiin Seattle). N. EuglI.J. .l/cd. 281: 915.

30. Imai, K.. H. Mlorimoto, M. Kotani, S. Shibata, T. 'Mi-vaji, and K. Matsutomo. 1970. Studies on the functioniof abnormal hemoglobinis. II. Oxygen equilibrium ofabnormal hemoglobins: Shimonoseki, Ube II. Hikari,Gifu and Agenogi. Biocthiri. Biop/hys. Acta. 200: 197.

31. Nagel, R. L., H. M1. Raniiey, T. B. Bradley, A. Jacobs,and L. Udem. 1969. Hemoglobin L Ferrara in a Jewishfamily associated with a hemolytic state in the propo-situs. Blood J. Hcuitotol. 34: 157.

32. Ramakrislman, C., anid G. N. Ramachandrani. 1965.Stereoclhemical criteria for polypeptide and protein chainiconifornmations. II. Allowed conformationis for a pair ofpe)tide unlits. Bil p/vs. J. 5: 909.

33. Halbrecht, I., W. A. Isaacs, H. Lelhmannii, and F.Ben-Porat. 1967. Hemoglobin Hasharon (a47 Asparticaci(I -> Histidine). Isr. J. JId. Sci. 3: 827.

34. Charache, S., A. M. Mondzac, U. Gessner, and E. E.Gayle. 1969. Hemoglobin Hasharon(a,47 i(CD5)g2) a

hemoglobini found in low conceintratioln. J. Clin. Invest.48: 834.

35. Perutz, M. F. 1970. Stereochemistry of cooperativeeffects in hemoglobin. Nature (Louid.). 228: 726.

36. Charache, S., D. J. XVeatherall, and J. B. Clegg. 1966.Polycythemia associated with a hemoglobinopathy. J.Cl/ii. Invest. 45: 813.

37. Reed, C. S., R. Hampson, S. Gordon, R. T. Jones,M. J. Novy, B. Brimhall, M. J. Edwards, and R. D.Koler. 1968. Erythrocy-tosis secondary to inicrease(doxygen affinity of a mutant hemoglobin, HemoglobinKempsey. Blood J. Hemtatol. 31: 623.

2528 S. Fujita

38. Lines, J. G., and R. McIntosh. 1967. Oxygen bindingby hemoglobin J-Cape town(a292 Arg-->Gln). Nature(Lond.). 215: 297.

39. Novy, M. J., M. J. Edwards, and J. Metcalfe. 1967.Hemoglobin Yakima. II. High blood oxygen affinityassociated with compensatory erythrocytosis and normalhemodynamics. J. Clin. Invest. 46: 1848.

40. Smith, L. L., C. F. Plese, B. P. Barton, S. Charache,J. B. Wilson, and T. H. J. Huisman. 1972. Subunitdissociation of the abnormal hemoglobins G Georgia(a95LeU (G2)p2) and Rampa( 2,1er (G2)p2). J. Biol.Chem. 247: 1433.

41. Bonaventura, J., and A. Riggs. 1968. Hemoglobin Kan-sas, a human hemoglobin with a neutral amino acidsubstitution and an abnormal oxygen equilibrium. J.Biol. Chem. 243: 980.

42. Briehl, R. W., and J. V. Hobbs. 1970. Ultraviolet differ-ence spectra in human hemoglobin. I. Difference spectrain hemoglobin A and their relation to the function ofhemoglobin. J. Biol. Chem. 245: 544.

43. Perutz, M. F., H. Muirhead, L. Mazzarella, R. A.Crowther, J. Greer, and J. V. Kilmartin. 1969. Identifi-cation of residues responsible for the alkaline Bohr effectin hemoglobin. Nature (Lond.). 222: 1240.

44. Kilmartin, J. V., and L. Rossi-Bernardi. 1969. Inhibitionof C02 combination and reduction of the Bohr effect inhemoglobin chemically modified at its a-amino groups.Nature (Lond.). 222: 1243.

45. Bellingham, A. J., and E. R. Huehns. 1968. Compensa-tion in hemolytic anemias caused by abnormal hemo-globins. Nature (Lond.). 218: 924.

46. Bartels, H., R. Beer, E. Fleischer, H. J. Hoffheinz, J.Krall, G. Rodewald, J. Wenner, and I. Witt. 1955. Bes-timmung von Kurzschlussdurchblutung und Diffusions-kapazitaet der Lunge bei Gesunden und Lungen Kran-ken. Pfluegers Arch. Gesamte Physiol. Menschen Tiere.261: 99.

47. Benesch, R., and R. E. Benesch. 1967. The effect oforganic phosphates from the human erythrocyte on theallosteric properties of hemoglobin. Biochem. Biophys.Res. Communi. 26: 162.

48. Chanutin, A., and R. R. Curnish. 1967. Effect of organicand inorganic phosphates on the oxygen equilibrium ofhuman erythrocytes. Arch. Biochem. Biophys. 121: 96.

49. Tyuma, I., and K. Shimizu. 1970. Effect of organicphosphates on the difference in oxygen affinity betweenfetal and adult human hemoglobin. Fed. Proc. 29: 1112.

Oxygen Equilibrium Characteristics of Abnormal Hemoglobins 2529