Proc. Natl. Acad. Sci. USAVol. 73, No. 9, pp. 3288-3292, September 1976Medical Sciences

Inhibition of erythrocyte sickling by cystamine, a thiol reagent(hemoglobin S polymerization/oxygen affinity/intracellular hemoglobin concentration)

WAFFA HASSAN, YVES BEUZARD, AND JEAN ROSAUnit6 de Recherches sur les Anemies, I.N.S.E.R.M. U. 91, Hopital Henri Mondor, 94010 Creteil, France

Communicated by Helen M. Ranney, June 28, 1976

ABSTRACT Incubation of sickle cells with cystamine, athiol reagent, resulted in the formation of an intracellular S-ethylamine derivative. The rate of the reaction was dependentupon the cystamine concentration, the temperature, and theduration of the incubation. The cystamine-treated cells dem-onstrated a marked inhibition of sickling under hypoxic con-ditions, a decrease in their mean corpuscular hemoglobin con-centration, and a significant increase in their oxygen affinity.The oxygen affinity of these cells was less dependent on theirmean corpuscular hemoglobin concentration than that of un-treated sickle cells. The minimum gelling concentration of S-ethylamine deoxyhemoglobin S was slightly increased. Cys-tamine did not affect the intracellular pH nor the 2,3-di hos-phoglycerate level. The exact contribution of the interrelatedfactors in cystamine inhibition of sickling (changes in oxygenaffinity, mean corpuscular hemoglobin concentration, andminimum gelling concentration) has yet to be determined.

Under hypoxic conditions, hemoglobin S (Hb S) moleculespolymerize either in concentrated solutions (1, 2) or withinerythrocytes (3, 4), giving rise to sickled cells. The polymer-ization of Hb S within the erythrocytes results in an abnormallylow oxygen affinity (5, 6), which is dependent on Hb concen-tration within the cells containing Hb S (S cells) (7).

Several agents have been reported to inhibit sickling, suchas cyanate (8), nitrogen mustard (9), alkylurea (10), dimethyladipimidate (11), and, more recently, bis(N-maleimidomethyl)ether (12). Cystamine, a thiol reagent (NH2-CH2-CH2-S-S-CH2-CH2-NH2), has been shown to bind to Hb (13). The S-ethylamine Hb derivative migrated as a single electrophoreticband distinct from that of the unreacted Hb. This propertyprovided a simple means of quantitation of the cystamine-reacted fraction. The S-ethylamine Hb exhibited a high oxygenaffinity and a reduced Bohr effect (14). In view of these prop-erties, it was reasonable to suspect that cystamine could exertan antisickling effect.

In the present report, the effects of cystamine on the sicklingof S cells and on some related phenomena are presented.

MATERIALSFresh blood samples from individuals homozygous (SS), het-erozygous (AS), and doubly heterozygous (SC) for sickle hem-oglobinopathy and from normal adults were drawn into EDTAas anticoagulant. The composition of this anticoagulant was 1%EDTA, 0.6% NaCI, and 6% glucose (wt/vol); it maintained aconstant level of 2,3-diphosphoglycerate (P2-G) in the eryth-rocytes during a minimum period of 7 days at 4°. Cystaminedichloride (2,2-dithio-bisethylamine dichloride) was provided

Abbreviations: Hb, hemoglobin; MetHb, methemoglobin (ferrihem-oglobin); Hb S, sickle cell hemoglobin; S cells, cells containing morethan 94% of Hb S; SA cells, cells containing Hb S and Hb A; SC cells,cells containing Hb S and Hb C; MCHc, mean corpuscular Hb con-

centration; cyst-Hb, Hb reacted with cystamine (S-ethylamine de-rivative of Hb); P2-G, 2,3-diphosphoglycerate; P50, partial pressureof oxygen at which Hb is half saturated with oxygen; Ins-P6, inositolhexaphosphate.

by Merck, purified glutaraldehyde (25% aqueous solution) wasobtained from TAAB laboratories, and enzymes and substrateswere obtained from Boehringer. Cellulose acetate was suppliQdby Chemetron (Milano), and Sephadex by Pharmacia. Allchemicals were of analytical grade.

METHODSCystamine dichloride (0.22 M) was dissolved in 0.15 M diso-dium phosphate solution. The solution was adjusted to pH 7.4with monosodium phosphate (0.15 M). Aliquots of this cys-tamine solution were added to the cell suspensions to give finalcystamine concentrations ranging from 0.5 to 5 mM. Aliquotsof 5% cell suspensions were incubated for 60 or 120 min at 370or 40 with or without cystamine. After incubation the cells werewashed and resuspended in phosphate buffers. The final vol-ume of the packed cells was 40% of the total. The mean cor-puscular Hb concentration (MCHC) of control and cystam-ine-treated cells was determined from Hb and hematocritmeasurexsents, except when stated otherwise. Intracellular pHwas determined on packed frozen and thawed cells, at 370, witha Radiometer pH meter fitted with a microelectrode unit (typeE 5021a). Erythrocyte P2-G level was determined accordingto Rose and Leibowitz (15). Hemolysates were prepared ac-cording to Drabkin (16). Hb electrophoresis was performed oncellulose acetate strips at pH 8.6 with Tris-EDTA-borate buffer(17); the proportions of the various fractions was determinedby densitometry. Solutions of S-ethylamine Hbs (cyst-Hbs) wereprepared from cells that had been incubated with 2-5 mMcystamine for 1 hr at 37°. Organic phosphates and free cys-tamine were removed by chromatography on a column (2.5 X40 cm) of Sephadex G-25.

Sickling Experiments. Aliquots (0.1 ml) of washed andpacked cells were suspended in 10 ml of 0.15 M phosphatebuffer at pH 7.45. The suspensions were subsequently deoxy-genated by evacuation in a 250 ml tonometer at partial oxygenpressures of 0, 10, 20, 30, 40, or 50mm Hg (0, 1.3, 2.6, 3.9, 5.2,or 6.5 kPa) for 10 min and incubated at 370C for a further 10min. Deoxygenated cells were then anaerobically transferredinto a deoxygenated, phosphate-buffered glutaraldehyde so-lution (5% vol/vol) for fixation. The percentage of sickle cellswas determined on at least 500 cells by phase contrast micros-copy. We have designated sickle cells as cells that had sharppointed projections, as well as deformed cells that were notbiconcave discs.

Preparation of Fresh Reconstituted Cells with DifferentMean Corpuscular Hb Concentrations. Fresh packed andwashed erythrocytes (0.5 ml) were lysed in volumes of 0.05 MNaCl solution ranging from 0.4 to 1.5 ml. Careful mixing fol-lowed by incubation at 0° for 5 min was essential to ensurecomplete cell lysis. In some experiments, cystamine was addedto the lysed cells for the preparation of cells containing equalproportions of Hb and cyst-Hb. Isotonicity was restored byaddition of 1.3 M KCl solution. Cell preparations were subse-

3288

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

021

Proc. Natl. Acad. Sci. USA 73 (1976) 3289

quently incubated at 370 for 1 hr in order to allow resealing;the cells were washed and finally suspended in the appropriatebuffer. Control electrophoresis, oxygen affinity measurements,cellular P2-G and MCHC determinations were performedwithout delay, using a Coulter S or manual technique.Oxygen Dissociation Studies. Oxygen dissociation curves

were determined on erythrocytes and on Hb solutions at 370by the spectrophotometric method of Benesch et al. (18), asmodified by Bellingham and Huehns (19), with a Unicam SP800. Oxygen dissociation curves of the erythrocytes were de-termined in 0.15 M phosphate buffers at pH 7.15 or 7.45.Oxygen equilibrium of 0.2% solutions of stripped Hlbs and ofcyst-Hbs was determined in 0.05M Tris-HCl buffer in the pHrange 7.15-7.45 or in 0.05 M [bis-(2-hydroxyethyl)amino]-tris(hydroxymethyl)methane (Bis-Tris) at pH 6.45. The Bohreffect was determined from graphs relating the logarithm ofthe partial pressure of oxygen at which Hb is half saturated withoxygen (P50) to pH. The interaction with P2-G was studied bydetermination of the P50 of stripped S-ethylaiiine and controlHb solutions, to which different concentrations of P2-G hadbeen added.

Effect of Inositol Hexaphosphate (Ins-F6). The effect ofIns-P6 was studied by comparing the spectral changes under-gone by the ferri form (MetHb) of Hb S and cyst-Hb S upon theaddition of an excess of Ins-P6 (20). Hb S and cyst-Hb S werefirst oxidized to the corresponding MetHb in the presence ofpotassium ferricyanide and then chromatographed on a Se-phadex G25 column equilibrated with 0.1 M sodium phosphatebuffer (pH 6.5). Difference spectra were recorded in the ab-sence and in the presence of a 5 molar excess of Ins-P6 0.1 Mphosphate buffer at pH 6.5 with a Cary 118 C spectropho-tometer.

Gelation Studies. Gelation experiments were performed bythe method of Singer and Singer (21), as modified by Bookchinand Nagel (22). Solutions of Hbs and of cyst-Hbs were subjectedto simultaneous vacuum concentration and dialysis against 0.15M potassium phosphate buffer (pH 7.35) at 4°. Hb S, cyst-Hb,and mixtures of Hbs S/cyst-S, S/A, cyst-S/A, S/cyst-A, andcyst-S/cyst-A were studied. All samples were subjected tocontrol electrophoresis, MetHb measurements, and pH deter-minations before and after gelation. Samples containing morethan 2% of MetHb (23) before, or more than 6% after, gelationwere discarded. On account of the high oxygen affinity of thecyst-Hbs, nitrogen was bubbled for 90 min; this period wassufficient to assure a complete deoxygenation of all the samples,as proven by recording the spectra.

Effect of Cystamine on Some Enzyme Activities andSubstrates in Erythrocytes. The activities of glucose-6-phos-phate dehydrogenase (EC 1.1.1.49), 6-phosphogluconate de-hydrogenase (EC 1.1.1.44), hexokinase (EC 2.7.1.1), pyruvatekinase (EC 2.7.1.40), phosphofructokinase (EC 2.7.1.11), glu-tathione reductase (EC 1.6.4.2), and phosphoglyceraldehydedehydrogenase (EC 1.2.1.12) and the concentration of ATPwere determined in the cells treated with 1 mM cystamine andin control cells by the methods of Beutler (24).

RESULTSThe electrophoretic patterns obtained from hemolysates of A,SA, SC, and S cells that had been incubated for 1 hr with 1 mMcystamine are compared to those of controls (Fig. 1). A homo-geneous, slowly migrating band was present in addition to theunreacted Hb in each incubated sample. This result was similarto those obtained with Hb solutions as previously described (13,25, 26). It indicated that cystamine can penetrate the cellmembrane and react with intracellular Hb. The reaction pro-

® a aI

a

b b' c c' d dl '

& HbA

-cyst-HbAHbS W -wW

3tcyst-HbS- -

HbA2 d *- HbC

-cyst-HbA2- ^- cyst-HbCCarbonicanhydrase

0)

FIG. 1. Electrophoretic patterns of hemolysates from control cellsand cells treated with 1 mM cystamine. (a, a') A cells; (b, b') SA cells;(c, c') SC cells; (d, d') S cells. Letters with primes indicate reacted cells.Cell suspensions, at 5% hematocrit value, in 0.15M phosphate buffer(pH 7.45), were incubated with cystamine at a final concentration of1 mM for 1 hr at 37°. Electrophoresis of the cell lysates was performedon cellulose acetate at pH 8.6.

ceeded at identical rates for intracellular Hbs S, C, and A (Fig.1). However, it was dependent on the cystamine concentration,the temperature, and the duration of incubation of the cells(Fig. 2). The incubation of the cells with 1 mM cystamine for1 hr at 370 resulted, therefore, in the reaction of 55 ± 5% of theintracellular hemoglobin.The mean corpuscular Hb concentration (MCHC) of A or

S erythrocytes was constantly decreased when cells were in-cubated with cystamine in isotonic medium. The decrease inMCHC reached 5 + 0.5% and 10 ± 1% at 1 mM and 2 mMcystamine concentrations, respectively, and was associated withan equivalent increase in the mean corpuscular volume.

Effect of Cystamine on Oxygen Dissociation Properties.The P50 values of S and A cells were studied before and afterreaction with cystamine. The cystamine-treated S and A cellsshowed various degrees of increased oxygen affinity whichdepended on the extent of reaction of the intracellular Hb withcystamine. Reaction of about 55% of the intracellular Hb re-sulted in a 32% decrease in the P50 values of both A and S cellsat pH 7.15 and 7.45 (Table 1). The P2-G levels and pH values,in controls and cystamine-treated cells, were identical. Theincrease in oxygen affinity of cystamine-treated cells could nottherefore be explained by variations in P2-G concentrations orin intracellular pH.

Results of the oxygen dissociation properties of Cyst-Hbs Aand S are shown in Table 2. The oxygen affinity of cyst-Hb S(or A) was considerably greater than that of unreacted Hbswithout modification of the cooperativity. The Bohr effect over

100

80

60

I 40

20

00.5 1

Cystamine,mM

2 5

FIG. 2. Effect of cystamine concentration on the rate of formationof cyst-Hb S in S erythrocytes. (0) Cells incubated for 2 hr at 370; (o)cells incubated for 1 hr at 37°; (0) cells incubated for 2 hr at 40.Ranges include data from experiments on cells from 10 subjects.

Medical Sciences: Hassan et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

021

3290 Medical Sciences: Hassan et al.

Table 1. Effect of cystamine on the oxygen affinityof S and A cells (P.,)

Ps0 (mm Hg)

A cells S cells

pHof +1mM +1mMbuffers Controls cystamine* Controls cystaminet

7.15 31 ± 1.5 23-25 .62 ± 5 41 ± 27.45 24 ± 1 16-17.5 40 ± 2 28 ± 1.5

Fifty-five plus or minus five percent of Hb was reacted with cysta-mine. Experiments were performed in 0.15M phosphate buffer at370.* Three normal subjects.t Seven patients.

the pH range 7.15-7.45 was reduced to -0.36 as compared to-0.50 for unreacted Hb S. These results obtained on solutionsof cyst-Hb S are similar to those obtained by Taylor et al. (14)on solutions of cyst-Hb A. A lower interaction with P2-G forcyst-Hb S (or A) was also found. The data in Table 2 indicatethat the average effect of different concentrations of P2-G onthe shift towards the right of the oxygen dissociation curve ofcyst-Hb S is 30% less than its effect on that of unreacted HbS.

Effect of Cystamine on Sickling. At a partial oxygen pres-sure of 10 mm Hg, the cystamine-treated S cells showed a re-duced tendency to sickle, which correlated with the cystamineconcentration and hence with the proportion of the cystam-ine-reacted Hb (Fig. 3a). The percentage of sickle cells wasreduced from 80% in untreated cells to 22% in cells treated with1 mM cystamine, i.e., cells in which about 55% of the Hb wasreacted. In cells treated with 2 mM cystamine, in which thereaction of intracellular Hb with cystamine was complete,erythrocyte sickling did not occur (Fig. 3a). However, evenhigher concentrations of cystamine (3 mM) did not apparentlyaffect the irreversibly sickled cells, since their number in eachsample (3-12%) remained unchanged.Some inhibition of sickling by 1 mM cystamine occurred in

completely deoxygenated cells. Only 60% of these cells werepresent in sickled form, as compared to 90% in preparations ofuntreated cells. The effect on the inhibition of sickling by 1 mMcystamine was markedly enhanced when the oxygen pressurewas raised from 0 to 20 mm Hg (Fig. 3b). This was expected

Table 2. Effect of reaction of cystamine with Hb on theoxygen affinity (PO) of hemoglobin S or A and

on the P2-G effect on P

Pso (mm Hg)

Molar ratio cyst-HbBuffer P2-G/Hb Hb S or A S or A

0.05 M Bis-Tris, 0 32 19pH 6.45, 370

0.05M Tris HCI 0 13.5 11pH 7.45, 370

0.05MTris HCl, 0 18 14pH7.15, 37° 3 28 20

4.5 32 239 40 26.5

20 42 28

0 o.s 1 2 3

Cystamine, mM

.S10 20 30 40 50

P02 (mm Hg)20 40 0o 50 1o0Percent OxyHb

FIG. 3. (a) Effect of increasing concentrations of cystamine onsickling of S erythrocytes at a P02 of 10 mm Hg. (b) Influence ofoxygen pressure on sickling of control erythrocytes (solid line) anderythrocytes treated with 1 mM cystamine (dashed line). Sicklingexperiments were done in 0.15 M phosphate buffer, pH 7.45 at 37°.Different symbols represent results obtained with cells from differentpatients. (c) Sickling is related to the hemoglobin oxygen saturation(same data as shown in b).

from the high oxygen affinity of cyst-Hbs. The same data areshown in Fig. Sc, in which the abscissa represents the meanoxygen saturation of S cells from three patients. It clearly ap-pears that the proportion of sickle cells is higher at a givenoxygen saturation in the untreated samples than in the cys-tamine-reacted cells. These results indicate that a part of theantisickling effect of cystamine is independent of its influenceon oxygen affinity.

Effect of Ins-P6 on cyst-MetHb S. The addition of a 5 molarexcess of Ins-P6 to MetHb S in 0.1 M phosphate buffer at pH6.5 produced spectral changes identical to those observed withsolutions of MetHb A (20), and indicated conversion to thedeoxy T conformation. The addition of the same concentrationof Ins-P6 to cyst-MetHb S resulted in much smaller spectralchanges than those obtained with MetHb S (Fig. 4), indicatinga stabilization of the oxy R conformation.Dependence of the Oxygen Affinity of Cystamine-Treated

Cells on MCHC. The dependence of the P5o on the MCHC ofcontrol and cystamine-treated A and S cells is shown in Fig. 5.The cystamine-treated cells contained 55 ± 5% of their re-spective intracellular Hbs in a reacted form, as deduced fromcontrol electrophoresis. Unlike the A cells, the S cells displayeda great dependence of the P50 upon MCHC. Their regressioncoefficient (0.95 mm Hg/g dl) was higher than that of A cells

E4

460 500 540 580 620 660 700nm

FIG. 4. Difference spectra between the original MetHb solutions(20 gM) in 0.1 M phosphate buffer (pH 6.5) and solutions at the samepH to which a 5 molar excess of Ins-P6 had been added. (Continuousline) MetHb S; (dashed line) cyst-MetHb S; (dotted line) base line.Values of the extinction coefficient £ are in moles of heme.

Proc. Natl. Acad. Sci. USA 73 (1976)

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

021

Proc. Nati. Acad. Sci. USA 73 (1976) 3291

0

S cellso A cellsA C*btUuiscdelA qs. tACIN

50

40

30

l0 20 30MC HC (/dl)

FIG. 5. Dependence of the oxygen affmnity ofS and A cells treatedwith 1 mM cystamine on MCHC. Oxygen affinity measurements were

performed at 370 in 0.15 M phosphate buffer and at pH 7.15.

(0.165 mm Hg/g-dl) over the concentration range 35-5 g/dl.Similar results have previously been reported (7). The regression

line of S cells converged with that of A cells at an MCHC of 5g/dl. By contrast, the regression line of cystamine-treated Scells, with a regression coefficient of 0.89 mm Hg/g dl, con-

verged with that of A cells at an MCHC of 20 g/dl. At MCHCsbelow 20 g/dl the scatter of the data points of the cystamine-treated S cells did not permit definition of a regression line.

Effect of Cystamine on Gelation Behavior of Hb S. Theresults of gelation studies are in Table 3. The valuesfor the minimum gelling concentration of deoxy Hb S (23.9g/dl) and for a deoxygenated mixture of 60% Hb A and 40%Hb S (31.5 g/dl) are similar to those previously reported (22).The minimum concentration of Hb required for gelation ofsolutions containing cyst-Hb S was higher than that for solutionscontaining unreacted Hb S (Table 3). In contrast cystamine-reacted Hb A behaved similarly upon gelation to unreacted HbA.

Effect of Cystamine on Some Erythrocyte Enzymes andSubstrates. Most of the enzyme activities and substrates assayedin sickle erythrocytes were not modified by incubation with 1mM cystamine. However, pyruvate kinase and phosphofruc-tokinase activities were only 75 and 60% of the controls, re-

spectively. Conversely, the activity of 3-phosphoglyceratemutase (EC 2.7.5.3) was significantly increased.

0

DISCUSSIONThe studies presented in this report provide evidence thatcystamine, a thiol reagent, is a strong inhibitor of sickling. Thisagent appears to modify three interrelated processes that un-

derlie the sickling of S erythrocytes: the Hb concentrationwithin the S cells, their oxygen affinity, and the gelation of HbS.

Cystamine reduced the MCHC slightly, but in a consistentmanner. Its effect on membrane permeability may be due tothe reaction of membrane SH groups with cystamine. Themodifications in MCHC were not accompanied by any changein the intracellular pH nor in the intracellular concentrationof organic phosphates which could affect oxygen affinity andsickling (27). The reduction in MCHC produces a delay in thesickling time, as demonstrated by kinetic studies of the gelationof deoxy Hb S (28) and of the sickling time of erythrocytes atzero oxygen pressure (29). The variation in MCHC seen incystamine-treated S cells could be associated in a similar mannerwith a delay in sickling. The cystamine-treated cells exhibiteda marked increase in their oxygen affinity. The part played inthis increase by changes in the MCHC is probably small com-pared to the direct effect of cystamine on the oxygen affinityof Hb.

These modifications in the oxygen affinity of cystamine-reacted cells have to be compared to those observed withbis(N-maleimidomethyl) ether, an antisickling agent thatcrosslinks the cysteine ,693 to the histidine fy97 (12). This sub-stance increased the oxygen affinity but abolished the coop-erativity of oxygen binding in the S cells. This result contrastswith the normal cooperativity of the cystamine-treated Hb.Such a discrepancy indicates that these two SH reagents mayplay a different role in the inhibition of sickling.The slight but constantly increased minimum gelling con-

centration of reacted Hb S was in contrast to the absence of an.effect by p-hydroxy mercuribenzoate. This reagent, whichreacts with cysteine p393, raised the 02 affinity for Hb but didnot alter the minimum gelling concentration of Hb S or of a

mixture of Hbs S and A (30). The modifications of minimumgelling concentration produced by cystamine suggest alterationsin some bonds that form the polymer or gel. Moreover, sincethe mixtures of Hb S and cyst-Hb A have the same minimumgelling concentration as mixtures of Hb S and unreacted Hb A,it seems that Hb S itself must be altered. This supports a

mechanism of conformational change that results from thestabilization of the quaternary oxy state and is suggested by thedecreased interaction of cyst-Hb with P2-G and by the atten-uated spectral changes that occurred in the presence of excess

Ins-P6. This hypothesis is consistent with the conclusions ofBookchin and Nagel in studies of the minimum gelling con-

centration of half-liganded Hb S (31).

Table 3. Effect of cystamine on minimum gelling concentration in g/dl of Hb S and mixtures of Hbs A and S

MGC,*% Hb S % Cyst-Hb S % Hb A % Cyst-Hb A mean value No. of(a2Ap2S) (a2A S-cYSt) (a2A2A) (A2A -y2ACYst) (g/dl) Range specimens

100 23.9 23.2-24.0 710 90 26.6 25.0-28.8 740 60 31.5 30.4-32.4 740 60 31.2 30.4-32.0 5

40 60 32.7 32.4-33.2 540 60 32.8 32.4-33.6 5

* Minimum gelling concentration.

Medical Sciences: Hassan et al.

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

021

3292 Medical Sciences: Hassan et al.

Preliminary studies of the action of cystamine on erythrocyteenzyme activities revealed no major modification. However,extensive studies are required to determine the extent of thecellular alterations induced by cystamine. Preliminary findingsin animals (32) and in healthy human subjects (33) have beenreported, but careful toxicological studies have yet to be per-

formed.

We acknowledge Dr. John Chapman for careful assistance in re-

viewing manuscript, and are indebted to Dr. Cassius de Linval fromFort de France for providing blood samples from patients with sicklecell anemia. This research was supported by grants from l'InstitutNational de la Sante et de la Recherche M6dicale, la Fondation de laRecherche Medicale Frangaise, and from l'Universit6 de Paris Val deMarne.

1. Allen, D. W. & Wyman, J., Jr. (1954) Rev. Hematol. 9,155-157.

2. Murayama, M. (1957) J. Biol. Chem. 228,231-240.3. Bessis, M., Nomarski, G., Thiery, J. P. & Breton Gorius, J. (1958)

Rev. Hematol. 13,249-270.4. Finch, J. T., Perutz, M. F., Bertles, J. F. & Dobler, J. (1973) Proc.

Natl. Acad. Sci. USA 70,718-722.5. May, A. & Huehns, E. R. (1972) Haematol. Bluttransfus. 10,

279.6. Seakins, M., Gibbs, W. N., Milner, P. F. & Bertles, J. F. (1973) J.

Clin. Invest. 52,422-432.7. May, A. & Huehns, E. R. (1975) Br. J. Haematol. 30, 317-

335.8. Cerami, A. & Manning, J. (1971) Proc. Natl. Acad. Sci. USA 68,

1180-1183.9. Roth, E. F., Jr., Neuman, G., Vanderhoff, G., Nagel, R. C., Ka-

plan, B. & Jaffee, E. R. (1973) Proceedings of the AmericanSociety of Hematology, 16th Meeting, p. 110.

10. Elbaum, D., Nagel, R. L., Bookchin, R. M. & Herskowits, T.(1974) Proc. Natl. Acad. Sci. USA 71, 4718-4722.

11. Lubin, B. H., Pena, V., Mentzer, W. C., Bymun, E., Bradley, T.B. & Packer, L. (1975) Proc. Natl. Acad. Sci. USA 72, 43-46.

12. Zak, S. J., Geller, G. R., Finkel, B., Tukey, D. P., McCormack,

M. K. & Krivit, W. (1975) Proc. Natl. Acad. Sci. USA 72,4153-4156.

13. Smithies, 0. (1965) Science 150, 1595-1598.14. Taylor, J. F., Antonini, E., Brunori, M. & Wyman, J. (1966) J.

Biol. Chem. 241,241-248.15. Rose, Z. B. & Leibowitz, J. (1970) Anal. Biochem. 35, 177-

180.16. Drabkin, D. L. (1949) Arch. Biochem. 21,244-231.17. Garel, M. C., Hassan, W., Coquelet, M. T., Goossens, M. & Rosa,

J. (1976) Biochim. Biophys. Acta 24,97-104.18. Benesch, R., MacDuff, G. & Benesch, R. E. (1965) Anal. Biochem.

11,81-87.19. Bellingham, A. J. & Huehns, E. R. (1968) Nature 218, 924-

926.20. Perutz, M. F., Fersht, A. R., Simon, S. R. & Roberts, G. C. K.

(1974) Biochemistry 13, 2174-2186.21. Singer, K. & Singer, L. (1953) Blood 8, 1008-1023.22. Bookchin, R. M. & Nagel, R. L. (1971) J. Mol. Biol. 60, 263-

270.23. Benesch, E. R., Benesch, R. & Yung, S. (1973) Anal. Biochem.

55,245-248.24. Beutler, E. (1971) Red Cell Metabolism (Grune & Stratton, New

York).25. Garel, M. C., Cohen Solal, M., Blouquit, Y., & Rosa, J. (1974)

FEBS Lett. 43,93-96.26. Rosa, J., Beuzard, Y., Thillet, J., Garel, M. C. & Caburi, J. (1975)

in Abnormal Haemogloblns and Thalassemia, ed. Schmidt, R.M. (Academic Press, New York), pp. 79-116.

27. Jensen, M., Bunn, H. F., Halikas, G., Kan, Y. W. & Nathan, D.G. (1973) J. Clin. Invest. 52,2542-2547.

28. Hofrichter, J., Ross, P. D. & Eaton, W. A. (1974) Proc. Natl. Acad.Sci. USA 71, 4864-4868.

29. Zarkowsky, H. S. & Huchmuth, R. M. (1975) J. Clin. Invest. 56,1023-1034.

30. Nagel, R. L. & Bookchin, R. M. (1975) in Sickle Cell Anemia andother Hemoglobinopathies, ed. Levere, R. D. (Academic Press,New York), pp. 51-66.

31. Bookchin, R. M. & Nagel, R. L. (1973) J. Mol. Biol. 76, 233-239.

32. Mordukhovich, V. V. (1975) Radiobiologiya 15, 132-136.33. Vasin, M. V. & Levedeva, N. N. (1975) Kosm. Biol. Aviakosmi-

cheskaya Med. 9,54-57. (U.S.S.R.).

Proc. Natl. Acad. Sci. USA 73 (1976)

Dow

nloa

ded

by g

uest

on

Aug

ust 2

8, 2

021