Hemoglobin Philly (p35 Tyrosine Phenylalanine):
Studies in the Molecular Pathology of Hemoglobin
RONALD F. RIEDER, FRANK A. OsKI, and J. B. CLEGG
From the Department of Medicine, State University of New York, DownstateMedical Center, Brooklyn, New York 11203; Department of Pediatrics, Schoolof Medicine, University of Pennsylvania, Philadelphia, Pennsylvania 19104;Nuffield Unit of Medical Genetics, Department of Medicine, University ofLiverpool, Liverpool, Great Britain
A B S T R A C T An abnormal unstable hemoglobin, he-moglobin Philly, was found in three members of a fam-ily, each of whom had evidence of a chronic hemolyticstate. The presence of the mutant protein was sug-gested by the rapid appearance of inclusion bodies uponincubation of erythrocytes with brilliant cresyl blue andby the increased heat precipitability of the hemoglobin.However, no abnormal hemoglobin could be demon-strated by electrophoresis or column chromatography.Sulfhydryl titration of the hemolysates with p-mercuri-benzoate indicated that there was an average of fourreactive sulfhydryl groups per hemoglobin molecule in-stead of the usual two. The total number of hemoglobinsulfhydryl groups was normal; six groups were mea-sured when denatured globin was reacted with 5,5'-dithi-obis [2-nitrobenzoic acid]. This indicated that the in-creased sulfhydryl reactivity was due to an increasedavailability to p-mercuribenzoate of the usually unreac-tive hemoglobin cysteines at u112 and a104. After treat-ment for i hr with 4-5 moles of p-mercuribenzoate permole of hemoglobin, electrophoresis revealed that 30-35%of the hemoglobin had been dissociated into a- andf-chains. Normal hemolysates revealed negligible split-ting after 72 hr of similar treatment. The a- and f-chainsof hemoglobin Philly were separated from the unsplithemoglobin A by carboxymethyl cellulose chromatog-raphy. Fingerprint and amino acid analyses revealedthat tyrosine #35 was replaced by phenylalanine. In he-moglobin Philly there is loss of the normal hydrogenbond between the tyrosine hydroxyl group and the car-boxyl group of aspartic acid a126 at the aifli contact.This shifts the equilibrium from hemoglobin tetramerstoward monomers, exposing the f112 and a104 cysteines.In the cell, precipitation of the unstable monomers maycontribute to erythrocyte destruction.
Received for publication 24 January 1969.
INTRODUCTIONThe severity of the clinical manifestations of the dif-ferent abnormal hemoglobins is varied. Many of thesemutant proteins produce no discernible effect on thepatient, while some cause a mild compensated hemolysis,and others are associated with severe hemolytic disease(1). The detection of an abnormal hemoglobin in manypatients presenting with clinical states within thisspectrum has frequently been straightforward. Since thedemonstration of hemoglobin S (2), the widespread useof electrophoresis has resulted in the discovery of over100 mutant forms of human hemoglobin (3).
Recently it has become evident that a proportion ofpatients with hereditary nonspherocytic hemolytic diseaseof the Heinz body type possess an abnormal unstablehemoglobin (4). In several instances the electrophoreticor chromatographic separation of a variant protein fromhemoglobin A has been difficult or only partially suc-cessful (3). Where complete chemical analyses havebeen performed, several of the mutations have beenshown to consist of amino acid substitutions involvingno change in charge or occurring in the interior of thehemoglobin molecule, so that surface charge and elec-trophoretic mobility are unaffected (3).The present report describes a hemoglobinopathy
which was discovered in three members of a family, eachof whom had evidence of a chronic hemolytic state.The demonstration of a mutant hemoglobin was ham-pered by the inability to separate a variant protein fromhemoglobin A by any of the usual methods of electro-phoresis or chromatography. Proof of the presence ofhemoglobin Philly, and its isolation and purification, de-pended upon a unique property of this new abnormal he-moglobin.The properties of hemoglobin Philly have provided an
explanation for the pathophysiologic state of the affected
The Journal of Clinical Investigation Volume 48 1969 1627
patients. The structural defect of the abnormal hemoglo-bin has provided information on the importance of oneinterchain bond in the maintenance of the structural in-tegrity of the normal hemoglobin molecule.
METHODSHandling of blood samples. Routine hematologic exami-
nations by standard methods (5) and enzyme assays wereperformed on heparinized freshly drawn venous blood speci-mens. Other studies were done on samples anticoagulated andmailed packed in ice. Transit time was approximately 1 day.Preparation of hemolysates. Erythrocytes were washed
three to five times with 0.9% NaCl. For studies of hemo-globin, hemolysis was effected by shaking the cells with anequal volume of distilled water and i volume of carbontetrachloride or by stirring with 4 volumes of distilled waterfollowed by the addition of adequate 4 N NaCl to restore thesodium concentration to 0.15 N. Stroma was removed bycentrifugation.Erythrocyte enzyme assays. Hemolysates were prepared
by diluting the washed red cell suspension 1:10 in distilledwater. Hexokinase, phosphoglucose isomerase, phosphofruc-tokinase, aldolase, triose phosphate isomerase, phospho-glycerate mutase, enolase, and lactic dehydrogenase wereassayed by previously described methods (6) with tri-ethanolamine buffer, pH 8.0, at a final concentration of 0.1moles/liter. Phosphoglycerate kinase was assayed as de-scribed by Loder and De Gruchy (7), glyceraldehyde-3-phosphate dehydrogenase as described by Schrier (8), andglucose-6-phosphate dehydrogenase by the method of Zink-ham and Lenhard (9). Pyruvate kinase was measured bythe procedure of Rose and Warms (10) with phosphoenol-pyruvate in a final concentration of 2.0 mmoles/liter and aden-osine diphosphate (ADP) at 1.5 mmoles/liter. Glutathionereductase was assayed according to the method of Horn (11).
Effect of redox dye ont erythrocytes and hemoglobin. Onepart of a 1% solution of brilliant cresyl blue, in 0.9% sodiumchloride was added to two parts of heparinized blood. Thismixture was incubated at 370C, and blood films were pre-pared at intervals over the next 24 hr.
Other studies. Methemoglobin was measured by themethod of Evelyn and Malloy (12) on fresh hemolysatesand after 24 hr of incubation at 370C. The heat stability ofhemoglobin was tested, with and without the addition ofNaCN, according to the procedure of Grimes, Meisler, andDacie (13). The technique of Betke, Marti, and Schlicht wasused to assay alkali-resistant hemoglobin (14). The ferro-hemoglobin solubility test was performed according toItano (15).
Ultracentrifuge studies. Sedimentation analyses weremade with a Spinco model E ultracentrifuge at 52,000 rpmand 4VC. Measurements were made on Schlieren photographsobtained using a No. 29 Wratten filter and 1 N spectro-scopic plates.
Electrophoresis. Standard methods were employed forhemoglobin electrophoresis on starch gel (16), agar gel (17),and starch block (18). The method of Chernoff and Pettitwas utilized for the electrophoresis of globin polypeptidechains on starch gel in urea-containing buffers (19).Column chromatography. Column chromatography of
hemoglobin solutions on carboxymethyl cellulose (CMC) andon Amberlite CG-50 was performed according to previouslydescribed methods (20, 21).
Sulfhydryl determinations. The reactive sulfhydryl groupsof hemoglobin were measured spectrophotometrically by
titration with p-mercuribenzoate (PMB) according to themethod of Boyer (22) as modified by Benesch and Benesch(23). Total sulfhydryl groups in denatured globin prepara-tions were measured by the method of Ellman (24) using5,5'-dithiobis[2-nitrobenzoic acid] (DTNB), as describedby Guidotti (25).
Preparation of a- and 8-chains of hemoglobin. The heme-bearing PMB derivatives of the chains of hemoglobin A wereprepared by reacting the protein with a twelvefold molarexcess of PMB at pH 6.2, according to the method of Bucciand Fronticelli (26) as modified by Rosemeyer and Huehns(27). Isolation of the a-PMB and 8-PMB chains was thenachieved by CMC chromatography or starch-block electro-phoresis.
Fingerprint analysis. Heme was removed from the PMBderivatives of hemoglobin a- and ,8-chains by the acid-acetonemethod (28). PMB was removed from the globin by dis-solving the lyophilized protein in nitrogen-saturated 8 Murea containing 0.3 M 2-mercaptoethanol and allowing themixture to stand at room temperature for 3 hr under ni-trogen. The protein was subsequently separated from thelow molecular weight materials by passage through aSephadex G25 column equilibrated with 0.5% formic acid.Regeneration of sulfhydryl groups was checked using DTNB.Aminoethylation and trypsin digestion were carried out aspreviously described (29). High voltage paper electrophoresisof the tryptic digests was done with pyridine-acetic acid-water buffers at pH 4.7 and 6.4. The chromatography stepof the two-dimensional fingerprints was performed usingsolvent systems composed of pyridine-isoamyl alcohol-water, 35: 35: 37, or butanol-acetic acid-pyridine-water,15: 3:10:12. Automatic amino acid analyses were done onmaterial eluted from the fingerprints and hydrolysed in 6 NHCl containing 1 mg of phenol per ml in order to preventloss of tyrosine.
Materials. Sigma Chemical Co., St. Louis, Mo., providedPMB sodium salt, cystamine (2,2'-dithiobis[ethylamine dihy-drochloride], DTNB, and iodoacetamide. Iodoacetanmide-1-J'C,0.05 mc in 4.2 mg, was obtained from New England NuclearCorp., Boston, Mass.
RESULTSClinical data and case report. The propositus (II-1)
(Table I), a 9 yr old girl of Italian and German-Frenchancestry, was originally evaluated at 8 yr of age be-cause of the finding of persistant reticulocytosis andsplenomegaly. There is no history of jaundice or darkurine. At 6 yr of age the child was treated for severalmonths with iron because of anemia, but no data aboutthis episode are available. During the present investiga-tion the patient has had intermittent splenomegaly, thespleen edge being palpable at 1-4 cm below the left costalmargin. Hepatomegaly has not been observed. There isno family history of anemia, jaundice, or dark urine. Noother member of the immediate family has splenomegaly.
Table I shows the clinical hematologic data of thepatient and her family. The propositus, the father (I-1),and a brother (II-2) have elevated reticulocyte counts.Blood samples obtained from all three persons showed theformation of multiple intraerythrocytic inclusion bodieswhen incubated with the redox dye, brilliant cresyl blue
1628 R. F. Rieder, F. A. Oski, and J. B. Clegg
TABLE IClinical Hematologic Data
Methemoglobin Hemoglobin heatstability Autohemolysis RBC
Unin- Incu- inclu-Subjects Hemoglobin Reticulocytes cubated* bated* Control* NaCN: Controll Glucosel sions
g/100 ml % % % precipitated %I-1 14.6 2.2-3.6 36.4 3.9 +1-2 11.0 0.7 0II-1 12.6-13.6 5.1-8.1 3.0 15.4 26.3 8.5 38.4 2.3 +11-2 14.0 2.9 +I1-3 11.0 1.1 0
BCB, brilliant cresyl blue.* Normai control 1.5% before incubation and 1.2% after incubation.Normal control 6.4% precipitated, with NaCN 5.9% precipitated.
Normal control less than 1.2% hemolysis.
(BCB). Precipitates were visible in the cells at 1 hr andwere prominent by 2 hr. Inclusions were present in 99%of the cells. The mother (I-2) and a sister (II-3) havenormal reticulocyte counts, and their red cells did notform inclusion bodies upon exposure to BCB. Red cellmorphology in all family members is normal. Upon in-cubation, at 370C for 24 hr, the blood of the patient II-1showed increased formation of methemoglobin (Table I).Autohemolysis was increased and was corrected by theaddition of glucose to a concentration of 30 mmoles/liter. As shown in Table I, hemolysates from both thefather and the propositius exhibited increased hemo-globin instability when incubated at 50'C. However,12-24 hr of heating was necessary to cause hemoglobinprecipitation, and this was prevented by the previousaddition of NaCN.The red blood cell survival of patient II-1 was mea-
sured by use of 51Cr. The erythrocyte half-life was re-duced to 12 days. The osmotic fragility of the patient'sred cells was normal. Upon incubation at 370C for 24 hrthe osmotic fragility increased abnormally.
Determinations of hemoglobin F (1.1%), hemoglobinAs (3.7%), and the ferrohemoglobin solubility (4.7 g/liter) gave normal results.
Erythrocyte - enzymes and intermediary metabolism.In an attempt to determine the cause of the increasedhemolysis, a variety of red cell enzymes and metabolicintermediates was measured in the blood of patientII-1. The following determinations gave normal values:red cell glutathione, glutathione stability, glutathioneperoxidase, triphosphopyridine nucleotide (TPNH) anddiphosphopyridine nucleotide (DPNH) glutathione re-ductases, catalase, glucose-6-phosphate dehydrogenase,6-phosphogluconate dehydrogenase, hexokinase, glucose-phosphate isomerase, phosphofructokinase, aldolase,
glyceraldehyde phosphate dehydrogenase, triose phos-phate isomerase, phosphoglycerate kinase, phosphoglyc-erate mutase, enolase, pyruvate kinase, lactic dehydro-genase, adenosine triphosphate (ATP), ATPase, acetyl-cholinesterase, erythrocyte glucose consumption, andlactate production. The hexose monophosphate shuntpathway, as reflected in UCOs production, was normaland responded normally to methylene blue stimulation.Hemoglobin electrophoresis. Because of the produc-
tion of intraerythrocytic inclusion bodies upon BCBtreatment, and because of the demonstration of decreasedheat stability of the hemoglobin in the presence of anormal erythrocyte enzyme complement, an attempt wasmade to demonstrate the existence of an abnormal hemo-globin. Conventional vertical starch-gel electrophoresisat pH 8.6 and pH 7.0 of hemolysates obtained from thepropositus and her family revealed no abnormal hemo-globin band. When the gel was overloaded with hemo-globin from patient II-1, a benzidine-positive band wasnoted in the position of free a-chains (Fig. 1), suggest-ing the presence, in the hemolysate, of an unstablef-chain mutant (30). On agar-gel electrophoresis at pH6.0, a minor abnormal band was noted in a positionanodal to hemoglobin A (Fig. 2). This appeared to beprecipitated, denatured hemoglobin rather than a trueelectrophoretic band. No further migration of this bandoccurred when the current was passed at right angles tothe original direction of flow. However, another streakof denatured hemoglobin was subsequently noted to re-main behind the then sideways-moving hemoglobin A, afinding suggesting continuing denaturation of an un-stable component having a true mobility similar to hemo-globin A. Starch-gel electrophoresis in 8 M urea, of glo-bin prepared from the propositus' hemolysate, did not
Hemoglobin Philly 1629
a Hb A Hb A2
FIGURE 1 Hemoglobin electrophoresis on starch gel at pH 8.6. The gel was overloaded andstained with benzidine to show the presence of free a-chains in the hemolysate from patient II-1(lower sample). The anode is to the right of the figure.
reveal an abnormally migrating a- or P-globin compo-nent.Column chromatography. Because of suggestive evi-
dence of the presence of an unstable hemoglobin with anelectrophoretic mobility identical with hemoglobin A, anattempt was made to demonstrate an abnormal hemo-globin by column chromatography. No new band waspresent when hemolysates from patient II-1 and subjectII-2 were chromatographed on CMC, Amberlite CG-50in developers 2,4, and 6, or on Sephadex G200.Absorption spectra. Complete visible and ultraviolet
absorption spectra were obtained on deoxy- and carbon-monoxyhemoglobin solutions obtained from the proposi-tus, with a Cary model 14 recording spectrophotometer.These spectra were identical with a normal control andthe OD 280: OD 540 ratio was normal, a fact indicatingthe presence of the usual heme: globin ratio.
Titration of reactive sulfhydryl groups in hemoglobin.Because of a recent suggestion (31) that a blockade ofthe sulfhydryl (- SH) groups of globin plays a role inthe precipitation of unstable hemoglobins and in the in-duction of hemolysis, the reactive sulfhydryl groupswere measured in hemolysates obtained from the proposi-tus, the brother, and the father. Hemoglobin A has threepairs of cysteine residues. Each a-chain has one cysteineat position 104, while the f-chains have residues atpositions 93 and 112 (32). The P93 cysteines are situ-ated at the surface of the hemoglobin tetramer, whilethe P112 and a104 cysteines are in the interior of themolecule at the ap, contact (33). Ordinarily, under mildconditions, only the cysteines at P93 in the native hemo-globin molecule are available for reaction with sulfhy-dryl reagents (34).The addition of PMB to hemoglobin solutions results
in the attachment of this reagent to the reactive - SHgroups on the protein. When PMB reacts with sulfhy-
dryl residues there is a specific increase in the opticalabsorption at 250 my (22). Stepwise addition of PMBresults in a continued increase in optical density until thefree cysteinyl residues in the mixture are exhausted.Knowing the quantity of PMB added, one can calculatethe number of -SH groups that have reacted (22, 23).This technique was used to compare the number of reac-tive sulfhydryl residues in the hemolysates from patientsII-1 and I-1 with a normal hemolysate. A hemoglobinA solution with sulfhydryl groups blocked by iodoaceta-mide was used as the blank cuvette solution.
In the normal hemolysate 2 mmoles of PMB reactedwith each millimole of hemoglobin, a finding indicatingthat there was an average of two reactive cysteinyl resi-dues for each hemoglobin molecule (Fig. 3). When he-molysates from the propositus and the father weretested, the optical density increased until 4 mmoles ofPMB had been added per mmole of hemoglobin. Thisresult indicated that on the average there were fourreactive cysteinyl groups per hemoglobin molecule. Asimilar result was obtained when the hemoglobin ofsubject II-2 was tested.
Determination of total sulfhydryl residues. In orderto decide whether the above results indicated the pres-ence of additional globin cysteine residues or simply mir-rored an increased reactivity of those residues normallypresent at P112 and a104, the total number of - SHgroups in the hemolysate was measured. Globin wasprepared from the hemolysate of subject II-1 and froma normal hemolysate. The protein was dissolved in so-dium phosphate buffer at pH 7.15 containing 11% so-dium dodecyl sulfate (SDS). SDS denatures the poly-peptide chains making all cysteinyl residues reactive.The total sulfhydryl groups were determined by reac-tion with DTNB and measurement of the optical densityat 415 myA (24, 25). Both hemolysates gave results close
1630 R. F. Rieder, F. A. Oski, and J. B. Clegg
to the expected six sulfhydryl groups per tetramer. Thevalue obtained for normal globin was 6.8; globin fromthe propositus gave a value of 6.3. These results sug-gested that the additional reactive sulfhydryl groupsfound in the hemolysate from I-1, II-1, and II-2 werederived from the normally unreactive cysteine residuesat 8112 and a104. The mutation in the abnormal hemo-globin did not appear to involve a new cysteine insertion.
Electrophoretic demonstration of hemoglobin Philly.The finding of additional reactive sulfhydryl groups inthe hemolysates from affected family members stronglyindicated the presence of an abnormal hemoglobin andsuggested a method of separating the mutant protein(hemoglobin Philly) from hemoglobin A. Cystaminecombines with the two reactive cysteines of hemoglobinto place one additional positive charge on each a-chain
(35). This increased positive charge retards the migra-tion of hemoglobin A towards the anode during elec-trophoresis at pH 8.6. It was postulated that hemoglobinPhilly might react with more than two cystamine mole-cules and thus be retarded in its migration even morethan hemoglobin A, thereby affording a separation ofthe two proteins. However, the electrophoretic migra-tions of the hemoglobin in a normal hemolysate and thehemoglobin from patient II-1 were equally decreased af-ter exposure to excess cystamine, and no separation ofhemoglobin Philly from hemoglobin A occurred. Theseresults suggested that the additional reactive sulfhydrylgroups in hemoglobin Philly, while available to PMB,were not available to cystimine. An attempt was nextmade to use the extra negative charges placed on hemo-
Hb A and Philly
Hb C and F
FIGuRE 2 Agar-gel electrophoresis at pH 6.2. A band of denatured hemoglobinremains behind the main hemoglobin band in the hemolysate from patient II-1(second from the top). Bromthymol blue stain. The anode is to the left of thefigure.
Hemoglobin Philly 1631
N 0.6- -
0l I-/ FAT ER
. ,/*.n2/DAUGHTER-~~~~X NORMAL
0.40.5 1. 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 55
mM PMB/mM Hb
FIGuRE 3 Sulfhydryl titration with p-mercuribenzoate (PMB) of hemolysatesfrom patients I-1 and II-1 compared with a normal. The optical density isgraphed against the ratio of the concentrations of PMB and hemoglobin in thereaction mixtures. The break in the curve occurred at a ratio of approxi-mately 2 in the normal and at a ratio of approximately 4 in the samples fromthe patients.
globin Philly by PMB to effect a separation of the mu-tant protein from hemoglobin A.The hemolysate from the propositus was treated for 2
hr with 4 mmoles of PMB per mmole of hemoglobinin 0.05 M potassium phosphate, pH 7.0, the buffer usedin the spectrophotometric titration of the -SH groups.After exposure to PMB (Fig. 4) the hemolysate frompatient II-1 showed three new benzidine-positive elec-trophoretic bands in the positions of f-PMB, a-PMB,and free a-chains (26, 27). This finding indicated thatPMB had dissociated hemoglobin Philly into a- andfl-chain subunits. No splitting of the hemoglobin in anormal hemolysate was produced by PMB under theseconditions, even after 72 hr of exposure. Similar split-ting was produced by PMB treatment of hemolysates ob-tained from subjects I-1 and II-2, but no splitting ofhemoglobin occurred in the hemolysate obtained fromsubject I-2 (Fig. 5).
Electrophoretic mobility of the subunits of hemoglobinPhilly. Hemoglobin A can be split into a- and f-chainsby exposure for 12 hr to 12 mmoles of PMB per mmoleof hemoglobin in 0.2 M NaCl, 0.075 M sodium-potassiumphosphate, at pH 6.0 (26, 27). A normal hemolysate andone containing hemoglobin Philly were exposed to thismore strenuous treatment. Electrophoresis (Fig. 6)showed that both hemolysates were resolved into mono-mers, and that the migrations of corresponding sub-units were identical. The mobilities of the separatedchains were the same as those of the subunits formed
upon the mild treatment of hemoglobin Philly withPMB (Fig. 4). This finding confirmed the observationthat there were no charge differences between the cor-responding subunits of hemoglobin A and hemoglobinPhilly. In addition, globin chains prepared from the iso-lated subunits of hemoglobin Philly had mobilities onurea gel identical with the subunits of hemoglobin A.(The preparation of hemoglobin Philly subunits is de-scribed below in the section on fingerprint analysis.)
Quantity of PMB required to dissociate hemoglobinPhilly. The minimum molar ratio of PMB to hemoglo-bin required to obtain splitting of hemoglobin Phillywas determined. Varying amounts of PMB were mixedwith constant volumes of hemolysate from patient II-1in 0.05 M potassium phosphate, pH 7.0. The mixtureswere allowed to stand in the refrigerator for 1 hr andthen subjected to starch-gel electrophoresis. There wasa definite appearance of monomers at the ratio of 3mmoles of PMB per mmole of hemoglobin (Fig. 7). Byvisual estimation maximum production of monomers oc-curred when the ratio reached 4 to 5. No splitting of anormal hemolysate occurred under these conditions(Fig. 8). These data indicate that the titration of he-moglobin Philly with more than 2 mmoles of PMB permmole of hemoglobin resulted in splitting of the hemo-globin into monomers.
Ultracentrifuge studies. Sedimentation velocity ex-periments were done on the hemolysate from patientII-1. The hemoglobin concentration was varied from 2
1632 R. F. Rieder, F. A. Oski, and J. B. Clegg
to 12 mg/ml in 0.01 M potassium phosphate buffer,pH 7.0. A single. monodisperse boundary was ob-served with an sOaw value of 4.55S. This is the expectedvalue for the hemoglobin tetramer. Thus evidence for thepresence of a significant amount of free monomers inthe untreated hemolysate could not be established by thistechnique.
Reaction of hemoglobin Philly with iodoacetamide.The observation that cystamine was unable to combinewith all the PMB-reactive - SH groups of hemoglobinPhilly prompted an investigation of the reactivity of theabnormal hemoglobin to another sulfhydryl reagent, io-doacetamide. When hemoglobin A is treated with amolar excess of iodoacetamide the alkylating reagentattaches only to the cysteines at 893 (34). All the reac-tive cysteines are blocked, and subsequent combinationwith PMB is prevented. This was not the case with theabnormal hemolysate. The PMB titration of a hemo-globin solution from patient II-1, previously treatedwith 10 mmoles of iodoacetamide per mmole of hemo-globin, indicated that approximately 2 mmoles of free
B --PMB A -PMB2
-SH group remained to react with PMB after iodo-acetamide treatment (Fig. 9).
In another experiment aliquots of an iodoacetamide-treated hemolysate containing hemoglobin Philly wereexposed to varying amounts of PMB and the reactionmixtures subjected to starch-gel electrophoresis. Fig. 10shows that the addition of PMB to iodoacetamide-treated hemoglobin Philly resulted in enhanced dissoci-ation of the abnormal hemoglobin into a and P subunits.The appearance of monomers was apparent at a PMB:hemoglobin ratio of only 0.5 and appeared to reach amaximum at a ratio of 2. Iodoacetamide alone did notproduce the appearance of dissociated subunits. No split-ting occurred when iodoacetamide was added afterthe hemolysate had been treated with 2 mmoles of PMBper mmole of hemoglobin.The hemolysate from patient II-1 was reacted with
a 10 mM excess of iodoacetamide to which had beenadded a tracer amount of iodoacetamide-1-'4C. Afterstanding 5 hr at room temperature the mixture was dia-lysed to remove unreacted reagent. PMB, 4 mmoles per
NORMAL + PMB
II-1 + PMB
FIGURE 4 Starch-gel electrophoresis at pH 8.6 of hemolysates treated for 2 hr with p-mercuri-benzoate, 4 mmoles/mmole of hemoglobin, in 0.05 M potassium phosphate at pH 7.0. The treatedsample from patient II-1 has bands in the positions of 8-PMB, a-PMB, and free a-chains.There was a slight increase in the mobility of hemoglobin A and A2 upon addition of the nega-tively charged PMB. Ba refers to a mutant hemoglobin A2 in the normal sample. This ac-counts for the decreased intensity of staining of the minor hemoglobin band in the controlspecimen. Benzidine stain.
Hemoglobin Philly 1633
FiGuRE 5 Starch-gel electrophoresis of PMB-treated hemolysates obtained from membersof the propositus' family. The conditions were the same as in Fig. 4. The specimens from thepropositus, II-I, the father, I-1, and the brother, II-2, showed the presence of a and P sub-units. No dissociation was present in the samples from the mother, I-2, and from an unrelatednormal individual.
mmole of hemoglobin, was then added and the mixture The specific activity of P-Philly was 4270 cpm/mg, he-placed on a CMC column. The fi-Philly, hemoglobin A, moglobin A, 2450 cpm/mg, and a-Philly, 180 cpm/mg.and a-Philly fractions were collected and further puri- The absence of a-chain counts indicated that the cysteinefied by starch-block electrophoresis. Radioactivity was at a104 was not available to iodoacetamide. The cysteinemeasured with a low background gas-flow counter. at fl112 did not appear to have reacted either. The ratio
NORMAL + PMB
NORMAL HEMOLYSATE* | | | 6OX DC : ~~I I-1 + PMBII-1 HEMOLYSATE
FIGURE 6 Starch-gel electrophoresis of PMB-treated hemoglobin. The hemolysates from anormal person and patient II-1 were almost completely dissociated into subunits by exposurefor 12 hr to 12 mmoles of PMB per mmole of hemoglobin, in 0.2 N NaCl, 0.075 M potassium-sodium phosphate, pH 6.0. The electrophoretic mobilities of the a- and ,B-chains of the patient'shemoglobin were the same as the corresponding subunits from the normal individual.
1634 R. F. Rieder, F. A. Oski, and J. B. Clegg
of the specific activity of P-Philly to the specific activityof hemoglobin A was 2: 1. A ratio of 4: 1 would havebeen expected if both P-Philly -SH groups had re-acted with the radioactive iodoacetamide. These resultsare consistent with a limitation, in hemoglobin Phillyas in hemoglobin A, of the attachment of iodoacetamideto only the surface -SH groups at f93.
Determination of the percentage of hemoglobin Phillyin hemolysates. Hemolysates containing hemoglobinPhilly were treated with various amounts of PMB forJ-24 hr at 4C in the titration buffer. The reactionproducts were separated using starch-block electropho-resis or a CMC column equilibrated with 0.01 M sodiumphosphate, pH 6.2, and employing a gradient formedby mixing equal volumes of pH 6.7 and pH 8.4 0.01 Msodium phosphate (Fig. 11). The percentage of hemo-
globin split by PMB varied but was most often in therange of 30-35% (Table II). The time of exposure toPMB did not appear to greatly affect the proportionof the hemoglobin dissociated into monomers. The he-moglobin A peak from the CMC chromatography of thetreated hemolysate from patient II-1 was treated againwith varying amounts of PMB. No further splittingwas noted when the mixtures were subjected to starch-gel electrophoresis. This finding indicated that the phe-nomenon of easy dissociation by PMB was restricted to adefinite fraction of the hemoglobin in the hemolysate.
Fingerprint analysis of hemoglobin Phily. The a- andfl-chains of hemoglobin Philly were prepared free ofhemoglobin A. Hemolysates from patients I-1 and IH-were treated with 5 mmoles of PMB per mmole of he-moglobin and allowed to react for 1 hr. This split he-
Hb A and Philly
6 5 4 3 2 1 o.5 0
mM PMB / mM Hb
FIGURE 7 Treatment of patient's (II-1) hemolysate at pH 7.0 with varying amounts of PMB.There was evidence of subunit dissociation at concentration ratios of PMB to hemoglobingreater than 2. A slight increase in the mobilities of Hb A and A2 is also noted. This is prob-ably due to the negative charge of PMB.
Hemoglobin Phily 1635
moglobin Philly but left hemoglobin A intact. The mix-tures were then chromatographed on CMC and thefl-PMB, a-PMB, and free a-chain fractions collected.On CMC the unsplit hemoglobin A is eluted between thefl-PMB and a-PMB fractions (Fig. 11). Heme was re-moved by the acid-acetone method (28) and the globinlyophilized. The - SH groups were regenerated bytreatment of the protein with 2-mercaptoethanol in 8 Murea. Completeness of removal of PMB was ascertainedby sulfhydryl titration using the Ellman reagent (24, 25).The free chains were then treated with ethylenimine andsubjected to trypsin digestion (29). Fingerprint analysisof the a-chain of hemoglobin Philly gave patterns identicalwith that of a-A. The fingerprint of f-Philly showedthat peptide trypsin IV (TpIV) had a slightly faster
chromatographic mobility than f-A TpLV (Fig. 12).In addition, the ninhydrin stain of the fl-Philly peptidewas much weaker than that for the corresponding pep-tide in f-A. Specific stains for tryptophan and argininewere positive on the f-Philly TpIV spot, a finding indi-cating that these amino acids were present. The additionof urea to the tryptic digest before fingerprinting in-creased the yield of P-Philly TpIV, as judged by theincreased intensity of the ninhydrin and arginine stains.This result suggested that the low yield was due to ad-sorption of the abnormal peptide to other peptides, per-haps to those in the neutral band. The abnormal spot wascut from a fingerprint stained with 0.02% ninhydrin,eluted, and hydrolysed with 6 N HC1 containing phenol.The addition of phenol prevents the loss of any tyro-
.:.:.: . . :.::: :. :: .... ..... .... :* :: . . :: : .... : . .... .:*.: ::.
.... ..... . .... .. ...
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. ... ..* :~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~....: .. 4.
mM PMB / mM Hb
FIGURE 8 Treatment of a normal hemolysate at pH 7.0 with varying amounts of PMB.No subunit dissociation occurred. PMB imparts a slight mobility change to Hb A andHb A2.
1636 R. F. Rieder, F. A. Oski, and J. B. Clegg
....::. . : ... ...
sine present, and this finding was verified by analysis ofP-TpIV prepared from hemoglobin A. The amino acidanalysis of fi-Philly TpIV revealed that the tyrosineresidue was missing, and a phenylalanine residue waspresent (Table III). The tyrosine in hemoglobin A isat position 35 of the p-chain, counting from the aminoterminal end, and occupies position number one in theC helix. (32, 33). Hemoglobin Philly can therefore bedesignated as #35 (Cl) tyrosine -* phenylalanine.
DISCUSSIONA number of patients have now been found with con-genital nonspherocytic hemolytic anemia due to unstablemutant hemoglobins (4). These hemoglobins charac-teristically cause hemolysis even when present in theheterozygous state. Large numbers of Heinz bodies haveappeared in the peripheral blood of patients with un-stable hemoglobins who have undergone splenectomy(13, 36). In nonsplenectomized patients, inclusion bod-ies have been readily produced, in vitro, by exposureof erythrocytes to redox dyes (4, 37). Precipitation ofhemoglobin upon heating to 50'C has been a character-istic of some of these mutant proteins (4, 13, 30).
All the amino acid substitutions found in the unstablehemoglobins have involved residues in the interior ofthe molecule (3). Substitutions causing increased easeof denaturation have been observed at positions in con-tact with the heme groups, at certain general internalpositions, and at some of the contacts between subunits.
I 2 3mM PMB/mMHb
4 5 6
FIGURE 9 Sulfhydryl titration with PMB of an iodoaceta-mide-treated hemolysate from patient II-1 compared with anuntreated hemolysate from patient II-1 and an untreatednormal hemolysate. An iodoacetamide-blocked normal he-molysate was the solution in the blank cuvette. Iodoacetamidedid not completely block the sulfhydryl groups in the ab-normal hemoglobin; two titratable groups remained.
TABLE I IEstimation of Amount of Hemoglobin Philly in Hemolysates
Amount of each fraction foundAmount of Reaction Method of separation
Subject PMB added time of products Hb ,B-PMB Hb A Hbc-PMB Hba
millimoks PMB Hr %
mihimoles HbII-1 5 starch block 23.0 64.0* 13.0?II-1 10 a starch block 29.7 52.6* 17.7tI-1 5 1 starch block 13.6 69.6 6.9 9.9I-1 5 1 CMC chromatography 16.4 64.9 10.1 8.6II-I 5 1 starch block 13.0 76.3* 10.7tII-1 5 4 starch block 14.1 73.5* 12.4tII-1 5 8 starch block 13.6 74.3* 12.1tII-1 5 24 starch block 14.4 71.8* 13.8tII-1 6 1 starch block 16.2 68.5* 15.3tII-1 10 1 starch block 16.6 65.9* 17.5t
PMB, p-mercuribenzoate; CMC, carboxymethyl cellulose.* A+A2.?a-PMB + a. A2 + a-PM B.
Hemoglobin Philly 1637
H-I Hemolysote- IA Treoted
Kx H-I Hemolysate
For the most part, the positions involved are occupiedby nonpolar residues.As in the present case, a number of these hemoglobin
variants have not been separated easily from hemoglobinA by electrophoresis (38, 39). All attempts to separateintact hemoglobin Philly from hemoglobin A have beenunsuccessful. The increased tendency of hemoglobinPhilly to dissociate in the presence of PMB allowed elec-trophoretic and chromatographic purification of the sub-units of this unstable hemoglobin.
Initially the addition of PMB to hemoglobin A re-sults in the attachment of the reagent to the f93 cysteinesat the surface of the hemoglobin molecule and is associ-ated with a reduction in the molecular weight of theprotein. The hemoglobin tetramer dissociates at the alP2
TABLE II IMole Ratios of Amino Acids Found in fl-Philly Tp IV
compared with Values Expected in ,8-A Tp IV
Amino acid P-Philly/Tp IV P-A/Tp IV Difference
Arg 0.9(1) 1Glu 1.2(1) 1Leu 1.7(2) 2Phe 0.9 (1) +1Pro 0.8 (1) 1Thr 1.0(1) 1Try* (1) 1Tyr 1 -1Val 1.7(2) 2
* By staining reaction on paper.
IODOACETAMIDE-TREATED II-l HIEMOLYSATE
543 A 05 4 3) 2 1 0 '. 5 0 . 25 0
mM PM.B / mM HZb
FIGURE 10 Starch-gel electrophoresis of patient's iodoacetamide-reacted hemolysate after treat-nient with varying amounts of PMB. The enhanced subunit dissociation began at a concen-tration ratio of PMB to hemoglobin greater than 0.25.
1638 R. F. Rieder, F. A. Oski, and J. B. Clegg
kE 0.8-ie 0.7a0 0.6
4 * * ~~~~~~pH
t *~~~~~- OD. 540 m
10 30 50 70 90 1io 170TUBE NUMBER
FIGURE 11 Carboxymethyl cellulose column chromatography of a PMB-treated hemolysate from patient I-1. Peak 1 consisted of Philly B-PMB, peakII was hemoglobin A, peak III was predominantly hemoglobin As, and a-PMB,and peak IV was made up of free a-chains. The components of the variouspeaks were verified by starch-gel electrophoresis, fingerprint analyses, andsulfhydryl titration.
contact into aijl dimers (27). The sulfhydryls at P112and al04 remain unavailable to PMB because they lieburied in the ai#% contact (3, 27, 33). Under conditionsof low pH and higher ionic strength, which increase thetendency of dimers to dissociate into monomers, thea104 and P112 cysteines become available for substitutionwith PMB. This results in complete irreversible dis-sociation of the tetramer into subunits (27). The presentreport indicates that a similar mechanism was responsiblefor the PMB dissociation of hemoglobin Philly. Thechief difference would appear to be the mildness of theconditions under which the variant hemoglobin wasdissociated.The initial two equivalents of PMB probably attached
to the surface 693 cysteines of hemoglobin Philly, asdissociation into a and P subunits was not detected onstarch-gel electrophoresis. Further addition of PMBthen resulted in the production of monomers. The spec-trophotometric titration of the abnormal hemolysatesindicated that approximately 4 mmoles of cysteine re-acted per mmole of hemoglobin. Using a figure of 3.6cysteines reacted per molecule of total hemoglobin in thehemolysate, if 60% of the protein consisted of hemo-globin A, and only two residues of cysteine reacted oneach hemoglobin A molecule, it can be calculated thateach molecule of hemoglobin Philly had all six sulfhy-dryls titrated.
After treatment with iodoacetamide the splitting ofhemoglobin Philly was achieved upon the addition of
less than 2 mmoles of PMB per mmole of hemoglobin.Assuming a 60: 40 ratio of hemoglobin A to hemoglo-bin Philly, if iodoacetamide blocked the P93 cysteines onall the hemoglobin molecules and rendered hemoglobin Aunreactive to PMB, it can be calculated that a subse-quent titration value of 1.6 PMB residues reflected anattack on all four remaining cysteines in the hemoglobinPhilly fraction. This titration was associated with theappearance of a and P monomers on starch-gel electro-phoresis. Thus after iodoacetamide treatment, the al104and 8112 cysteines remained reactive to PMB. The re-sults of the experiments using radioactive iodoacetamideand unlabeled PMB support this interpretation. Almostno radioactivity was found on the dissociated a subunits.The P-Philly fraction had a specific activity only twicethat of hemoglobin A, a finding suggesting that in he-moglobin Philly as in hemoglobin A, the attachment ofthe alkylating agent occurred at P93 but not P112.The unavailability of the alc04 and P112 cysteines to
iodoacetamide was further indicated by the inability ofan excess of the alkylating agent alone to split hemoglo-bin Philly even after the #93 cysteines had reacted withPMB.The increased availability of the alr04 and P112 cys-
teines of hemoglobin Philly to PMB can be explained bythe amino acid substitution in the mutant protein. Thenormal dissociation of hemoglobin is thought to occur atthe acif contact with formation of all dimers (27). Theequilibrium of the dimer-monomer conversion is strongly
Hemoglobin Philly 1639
. 4 II
I+FIGURE 12 Tracings of ninhydrin-stained fingerprints of ,8-A globin and 8-Philly globin.Peptide p-Philly TpIV was invariably very faint and had a slightly faster chromatographicmobility than 8-A TpIV. The sample was applied as a thin vertical line (indicated) on a stripof filter paper, and electrophoresis was performed (horizontal direction) at pH 4.7 in pyridine-acetic acid-water buffer. The strip was then sewn to a sheet of filter paper followed bydescending chromatography in butanol-acetic acid-water-pyridine (vertical direction).
1640 R. F. Rieder, F. A. Oski, and J. B. Clegg
in favor of dimers. In hemoglobin Philly there is a re-placement of tyrosine by phenylalanine at position 35 ofthe j3-chain. Normally this tyrosine is situated at the a,#,contact, and its hydroxyl group forms a hydrogen bondwith the carboxyl group of an aspartic acid residue inthe a-chain at position 126 (33). The side chain ofphenylalanine cannot form this hydrogen bond. Thissubstitution probably weakens the a,#, contact and shiftsthe hemoglobin tetramer-dimer-monomer equilibriumtowards formation of monomers.' This effect would leadto exposure of the internal sulfhydryls, making themavailable to PMB. As phenylalanine is an almost per-fect isomorphous replacement of tyrosine, it is unlikelythat any conformational changes are involved.'
There is no explanation at present for the inabilityof excess iodoacetamide or cystamine to attach to thea104 and P112 cysteines and cause monomer formation.The shifting of the equilibrium towards monomers is
also probably responsible for the decreased life-span ofthe patient's erythrocytes. Free a- and ,-chains are muchless stable than hemoglobin A and tend to precipitate outof solution, and this presumably leads to formation ofintraerythrocytic inclusions, or Heinz bodies (40). Se-questration by the spleen of red cells with inclusionsthen occurs (41). In addition, the spleen may remove in-clusions from erythrocytes with subsequent circulationof the "pitted" cells (42). If some loss of membrane re-sulted from this splenic activity these cells would beshorter lived. Attachment of Heinz body inclusions tothe red cell membrane by means of mixed disulfide bondshas also been postulated (31), resulting in damage tothe cell membrane and ensuing hemolysis (43).
ACKNOWLEDGMENTSWe thank Dr. Robert Kaye for referring the propositus forstudy.This work was supported by U. S. Public Health Service
Grants AM-12401 and HD-01919 and also received financialsupport from the World Health Organization. Some of theclinical studies were performed on the Clinical ResearchCenter of the University of Pennsylvania supported by Grant3MOlFR4006. Dr. Oski is a recipient of U. S. Public HealthService Career Development Award 1-KO3-HD38600.
REFERENCES1. Heller, P. 1966. Hemoglobinopathic dysfunction of the
red cell. Amer. J. Med. 41: 799.2. Pauling, L., H. A. Itano, S. J. Singer, and I. C. Wells.
1949. Sickle cell anemia, a molecular disease. Science(Washington). 110: 543.
3. Perutz, M. F., and H. Lehmann. 1968. Molecular pathol-ogy of human hemoglobin. Nature (London). 219: 902.
4. Rieder, R. F., and T. B. Bradley, Jr. 1968. HemoglobinGun Hill: an unstable protein associated with chronichemolysis. Blood. 32: 355.
Perutz, M. F. Personal communication.
5. Cartwright, G. E. 1968. Diagnostic Laboratory Hema-tology. Grune & Stratton, Inc., New York. 4th edition.
6. Chapman, R. G., M. A. Hennessey, A. M. Waltersdorph,F. M. Huennekens, and B. W. Gabrio. 1962. Erythro-cyte metabolism. V. Levels of glycolytic enzymes andregulation of glycolysis. J. Clin. Invest. 41: 1249.
7. Loder, P. B., and G. C. de Gruchy. 1965. Red-cell en-zymes and co-enzymes in non-spherocytic congenitalhaemolytic anaemias. Brit. J. Haematol. 11: 21.
8. Schrier, S. L. 1966. Organization of enzymes in humanerythrocyte membranes. Amer. J. Physiol. 210: 139.
9. Zinkham, W. H., and R. E. Lenhard, Jr. 1959. Meta-bolic abnormalities of erythrocytes from patients withcongenital nonspherocytic hemolytic anemia. J. Pediat.55: 319.
10. Rose, I. A., and J. V. B. Warms. 1966. Control of gly-colysis in the human red blood cell. J. Biol. Chem. 241:4848.
11. Horn, H. D. 1963. Glutathion reductase. In Methods ofEnzymatic Analysis. M. V. Bergmeyer, editor. AcademicPress Inc., New York. 875.
12. Evelyn, K. A., and H. T. Malloy. 1938. Microdetermina-tion of oxyhemoglobin, methemoglobin, and sulfhemo-globin in a single sample of blood. J. Biol. Chem. 126:655.
13. Grimes, A. J., A. Meisler, and J. V. Dacie. 1964. Con-genital Heinz-body anaemia. Further evidence on thecause of Heinz-body production in red cells. Brit. J.Haematol. 10: 281.
14. Betke, K., H. R. Marti, and I. Schlicht. 1959. Estimationof small percentages of fetal hemoglobin. Nature (Lon-don). 184: 1877.
15. Itano, H. A. 1953. Solubilities of naturally occurringmixtures of human hemoglobin. Arch. Biochem. Biophys.47: 148.
16. Smithies, 0. 1965. Characterization of genetic variantsof blood proteins. Vox Sang. 10: 359.
17. Marder, V. J., and C. L. Conley. 1959. Electrophoresisof hemoglobin on agar gels. Frequency of hemoglobinD in a Negro population. Johns Hopkins Med. J. 105:77.
18. Kunkel, H. G., R. Ceppellini, V. Muiller-Eberhard, andJ. Wolf. 1957. Observations on the minor basic hemo-globin component in the blood of normal individuals andpatients with thalassemia. J. Clin. Invest. 36: 1615.
19. Chernoff, A. I., and N. M. Pettit, Jr. 1965. The aminoacid composition of hemoglobin. III. A qualitativemethod for identifying abnormalities of the polypeptidechains of hemoglobin. Blood. 24: 750.
20. Chernoff, A. I., N. Pettit, and J. Northrop. 1965. Theamino acid composition of hemoglobin. V. The prepara-tion of purified hemoglobin fractions by chromatographyon cellulose exchangers and their identification by starchgel electrophoresis using tris-borate-EDTA buffer.Blood. 25: 646.
21. Allen, D. W., W. A. Schroeder, and J. Balog. 1958.Observations on the chromatographic heterogeneity ofadult and fetal human hemoglobin: a study of the effectsof crystallization and chromatography on the hetero-geneity and isoleucine content. J. Amer. Chem. Soc. 80:1628.
22. Boyer, P. D. 1954. Spectrophotometric study of the reac-tion of protein sulfhydryl groups with organic mercurials.J. Amer. Chem. Soc. 76: 4331.
23. Benesch, R., and R. E. Benesch. 1962. Determination of-SH groups in proteins. Methods Biochem. Anal. 10: 43.
Hemoglobin Philly 1641
24. Ellman, G. L. 1959. Tissue sulfhydryl groups. Arch.Biochem. Biophys. 82: 70.
25. Guidotti, G. 1967. Studies on the chemistry of hemo-globin. 1. The reactive sulfhydryl groups. J. Biol. Chem.242: 3673.
26. Bucci, E., and C. Fronticelli. 1965. A new method forthe preparation of a and j8 subunits of human hemo-globin. J. Biol. Chem. 240: PC551.
27. Rosemeyer, M. A., and E. R. Huehns. 1967. On themechanism of the dissociation of haemoglobin. J. Mol.Biol. 25: 253.
28. Rossi-Fanelli, A., E. Antonini, and A. Caputo. 1958.Pure native globin from human haemoglobin: prepara-tion and some physico-chemical properties. Biochem.Biophys. Acta. 28: 221.
29. Clegg, J. B., M. A. Naughton, and D. J. Weatherall.1966. Abnormal human hemoglobins. Separation andcharacterization of the a and 6 chains by chromatography,and the determination of two new variants, Hb Chesa-peake and Hb J (Bangkok). J. Mol. Biol. 19: 91.
30. Huehns, E. R., and E. M. Shooter. 1965. Human haemo-globins. J. Med. Genet. 2: 48.
31. Jacob, H. S., M. C. Brain, J. V. Dacie, R. W. Carrell,and H. Lehmann. 1968. Abnormal haem binding andglobin SH group blockade in unstable haemoglobins.Nature (London). 218: 1214.
32. Braunitzer, G., K. Hilse, V. Rudloff, and N. Hilschmann.1964. The hemoglobins. Advan. Protein Chem. 19: 1.
33. Perutz, M. F., H. Muirhead, J. M. Cox, and L. C. G.Goaman. 1968.. Three-dimensional Fourier synthesisof horse oxyhaemoglobin at 2.8 A resolution: the atomicmodel. Nature (London). 219: 131.
34. Guidotti, G., and W. Konigsburg. 1964. The character-ization of modified human hemoglobin. I. Reaction withiodoacetamide and N-ethylmaleimide. J. Biol. Chem. 239:1474.
35. Smithies, 0. 1965. Disulfide-bond cleavage and formationin proteins. Science (Washington). 150: 1595.
36. Hutchison, H. E., P. H. Pinkerton, P. Waters, A. S.Douglas, H. Lehmann, and D. Beale. 1964. HereditaryHeinz-body anaemia, thrombocytopenia, and haemoglo-binopathy (Hb. K6ln) in a Glasgow family. Brit. Med.J. 2: 1099.
37. Rieder, R. F., W. H. Zinkham, and N. A. Holtzman.1965. Hemoglobin Zurich. Clinical, chemical and kineticstudies. Amer. J. Med. 39: 4.
38. Dacie, J. V., N. K. Shinton, P. J. Gaffney, Jr., R. W.Carrell, and H. Lehmann. 1967. Haemoglobin Hammer-smith (p42(CD1) Phe-- Ser). Nature (London). 216:663.
39. Beretta, A., V., Prato, E. Gallo, and H. Lehmann. 1968.Haemoglobin Torino-a43 (CD1) Phenylalanine --Va-line. Nature (London). 217: 1016.
40. Fessas, P., D. Loukopoulos, and A. Kaltsoya. 1966.Peptide analysis of the inclusions of erythroid cells in,8-thalassemia. Biochim. Biophys. Acta. 124: 430.
41. Rifkind, R. A. 1966. Destruction of injured red cellsin vivo. Amer. J. Med. 41: 711.
42. Crosby, W. H. 1959. Normal functions of the spleenrelative to red blood cells: a review. Blood. 14: 399.
43. Jacob, H. S., and J. H. Jandl. 1962. Effects of sulfhydrylinhibition on red blood cells. II. Studies in vivo. J. Clin.Invest. 41: 1514.
1642 R. F. Rieder, F. A. Oski, and J. B. Clegg
Click here to load reader