Biochem. J. (1964), 93, 401

Changes in the Haemoglobin Types of Sheep as a Response to Anaemia

BY G. VAN VLIET*Baarn, The Netherland8

AND T. H. J. HUISMANDepartment of Biochemistry, Medical College of Georgia, Augusta, Ga., U.S.A.

(Received 2 March 1964)

The existence of two genetic variants of haemo-globin in sheep of different breeds has been demon-strated (e.g. Harris & Warren, 1955; Cabannes &Serain, 1955; Evans, King, Cohen, Harris &Warren, 1956; Evans, Harris & Warren, 1957; vander Helm, van Vliet & Huisman, 1957; Huisman,van Vliet & Sebens, 1958 a). The haemoglobin typemoving more rapidly towards the anode in paperelectrophoresis at pH 8 6 was called Hb-A or Hb-II,and the slow-moving haemoglobin has been desig-nated as Hb-B or Hb-I. Further experiments haveshown that the two haemoglobin types differ inchromatographic behaviour on the cation ex-changers CM-cellulose and Amberlite IRC-50, inrate of denaturation by alkaline reagents, and insolubility in concentrated salt solutions; analyses ofthe amino acid compositions of the two variantshave shown that Hb-A contains more threonine,serine and glutamic acid and Hb-B more glycine,alanine and aspartic acid (van der Helm et al. 1957;Huisman, van de Brande & Meyering, 1960). Thehybridization behaviour of the haemoglobins wasstudied by Shreffler & Vinograd (1962), who demon-strated that Hb-B more readily dissociates into itssub-units than does Hb-A at both acid and alkalinepH. Studies of the equilibria of the haemoglobinvariants with oxygen revealed a higher affinity ofthe Hb-A type as compared with the Hb-B type(Huisman et al. 1958a).

It has been shown (Huisman et al. 1958a;Huisman, van der Helm, Visser & van Vliet, 1958 b;van Vliet, 1960) that quantitative changes in theamounts of Hb-A and Hb-B may occur in a sheepheterozygous for both variants (AB sheep) whensuch an animal is submitted to the stress of severeanaemia. According to the observations of Blunt &Evans (1963) this response to anaemia may be dueto either an increased production of foetal haemo-globin or the formation of a new haemoglobin,electrophoretically similar to Hb-B and differentfrom foetal haemoglobin (Hb-F). The presentpaper describes additional studies of the changes inthe haemoglobin in AB sheep and in sheep homo-zygous for Hb-B (B sheep) before, during and

* Present address: Research Laboratories, Philips-Duphar Inc., Weesp, The Netherlands.

26

after experimental anaemia. Results suggesting theformation of a new haemoglobin type (Hb-C) in theAB sheep, but not in the B sheep, are presentedand some characteristic properties of the Hb-C aredescribed.

METHODS

Material. Adult sheep of the Texel breed were studied.Two AB ewes (no. 2357 and no. 9087, weighing 93 and81 kg. respectively) and one male (no. 289, body wt. 70 kg.)were made anaemic by repeated bleeding according to thescheme presented below. Blood samples from these animalswere collected at regular intervals and analysed. Inaddition, blood samples from two lambs heterozygous forHb-A and Hb-B and two lambs homozygous for Hb-B werestudied at periodic intervals after birth. All animals, whowere studied while in the animal quarters of the Labora-tory, were kept free of parasites by the administration ofantihelminthics. It may therefore be assumed that no addi-tional blood loss due to intestinal parasites had occurred.

Haematological studies and the preparation of erythrocytehaemoly8ates. The blood samples were collected in heparinand centrifuged to remove the plasma. The erythrocyteswere washed three times with 0 9% NaCl solution to whichwere added 50 units of penicillin and 250jug. of strepto-mycin/ml. The packed erythrocytes were mailed immedi-ately from The Netherlands to the U.S.A. by air. Nodeterioration of the samples was noted under these condi-tions. Haematological studies were performed in thehaematological laboratory of the Department of InternalMedicine of the State University of Utrecht, The Nether-lands (Dr K. Punt), by the conventional laboratory pro-cedures (Wintrobe, 1962). The erythrocytes of sheep 2357were also separated by centrifugation into young and olderythrocytes by the method of Borum, Figueroa & Perry(1957).

Electrophoretic and chromatographic studies. Starch-gelelectrophoresis was carried out by a slight modification(Huisman, 1963) of the procedure developed by Smithies(1955). Chromatographic separation of Hb-A, Hb-B, Hb-Cand Hb-F was obtained with DEAE-cellulose (Huisman &Dozy, 1962). The same procedure, but on a larger scale,was used for the isolation of greater quantities of thedifferent haemoglobin types (Huisman & Dozy, 1962).

Characterization of the isolated haemoglobin component8.Absorption-spectral measurements were performed with aBeckman DK-2 spectrophotometer and quartz cuvettes of1-0 cm. light-path. The solubility of the three isolatedhaemoglobin variants A, B and C in concentrated saltsolutions was determined by the method of Derrien (1952):

Bioch. 1964, 93

401

G. vA VLIET AND T. H. J. HUISMANthe final haemoglobin concentration was 1-0 g./100 ml.Gross structural differences between the polypeptide chainsof the haemoglobins were studied by hybridization experi-ments and by starch-gel electrophoresis in sodium formate-formic acid buffer, pH 19 (Muller, 1960; Huisman, 1963).The hybridization procedure was similar to that describedfor human haemoglobin types (Gammack, Huehns, Leh-mann & Shooter, 1961; Huisman, 1963). A small volume(0 5 ml.) of a particular haemoglobin solution (5 g./100 ml.)was mixed with an equal quantity of canine haemoglobinsolution of approximately similar concentration and themixture separated into two equal portions. One portion wasdialysed for 7 hr. against 0.1 M-sodium acetate-acetic acidbuffer, pH 4 7, at 40 followed by dialysis overnight againstthe tris-EDTA-borate buffer used in starch-gel electro-phoresis (Huisman, 1963). The control mixture was treatedsimilarly except that the dialysis was against water ratherthan sodium acetate-acetic acid buffer. The formation ofhybrid haemoglobins was studied by the starch-gel-electrophoretic procedure mentioned above.

Oxygen equilibrium experiment8. Construction of oxygendissociation curves at 370 was performed by the proceduredescribed by Brinkman & Dirken (1940). Total erythro-cyte haemolysates containing 5 g. of haemoglobin/100 ml.and isolated haemoglobin components in solutions of 0-8-1-0 g. of haemoglobin/100 ml. were studied. The solutions,which had been dialysed for 24 hr. against a large volume ofa 0-1 M-potassium phosphate buffer, pH 7-4, at 40 were eachequilibrated with oxygen at a different pressure (10-80 mm.Hg) in 150 ml. tonometers at 40 mm. pCOa and 37°.The determination of the percentage oxygen saturation wasmade with a Zeiss spectrophotometer (model PMQ-II), bythe procedure described by Jonxis & Boeve (1956) for theconcentrated haemoglobin solutions, and by that describedby Meyering, Israels, Sebens & Huisman (1960) for the moredilute haemoglobin solutions. The pH measurements werecarried out on a Radiometer model 4 pH-meter at 250; eachvalue obtained was corrected to 370 by using the factorgiven by Rosenthal (1948). The logP50 value, which is thevalue of the P02 at which 50% saturation of the haemo-globin with oxygen was observed, and the factor n, ameasure of the haem-haem interaction, were calculatedfrom the linear relationships attained by plotting thelogarithms of the P02 against the logarithms of y/(100 - y),y representing the percentage oxygen saturation of thehaemoglobin.

RESULTS

Electrophoretic and chromatographic character-istics of sheep haemoglobins. An example of theseparation ofHb-A and Hb-B by starch-gel electro-phoresis is presented in Fig. 1. The haemoglobinsample obtained from the AB sheep 2357 undernormal conditions consisted of two major fractions,the fast-moving Hb-A and the slower-movingHb-B, which possesses an electrophoretic mobilityidentical with that of human Hb-A. A minuteamount of a slower-moving fraction was also de-tectable. This component, designated Hb-C, waspresent in much larger quantities during the experi-mental anaemia and ultimately replaced Hb-Aentirely (Fig. 1).

In the separation of the Hb-A, Hb-B and Hb-C,starch-gel electrophoresis was found to be superiorto any other electrophoretic procedure, paperelectrophoresis particularly. The degree to whichthe three haemoglobins were separated by DEAE-cellulose chromatography is shown in Fig. 2. Asexpected from the electrophoretic mobilities, theHb-C was eluted first, followed successively byHb-B and Hb-A. Since the separation of thehaemoglobin variants was almost complete, calcu-lation of the proportion of each haemoglobin typein mixtures was possible. Fig. 2 also shows achromatogram of the haemoglobin of a newbornlamb (no. 11) from a blood sample collected 15 daysafter birth. Four haemoglobin fractions weredetectable. The major haemoglobin fraction, whichwas identified as Hb-F, was eluted between theHb-B and the Hb-A. A small but distinct pro-portion of Hb-C was also present. The greater partof each haemoglobin component, eluted from thesecolumns, was combined and concentrated by CM-cellulose chromatography (Huisman & Meyering,1960). Some idea of the degree of purity of theseisolated components can be obtained by referenceto their mobilities in starch-gel electrophoresis(Fig. 3). Mutual contamination of the haemoglobinvariants was minimal or absent. Fig. 3 also showsthat the small proportion of Hb-C, which wasdemonstrated in the blood of the newborn lamb byDEAE-cellulose chromatography, was detectable

(a) (b) (c) (d)

Origin

Hb-C

Hb-B

Hb-A

Fig. 1. Separation of the haemoglobin types A, B and C bystarch-gel electrophoresis. Experimental details are givenin the text. Total Hb concentrations: (a) 11-2 g./100 ml.;(b) 7-8 g./100 ml.; (c) 6-1 g./100 ml.; (d) 4-3 g./100 ml. Allblood samples were taken from sheep 2357.

402 1964

HAEMOGLOBIN TYPES OF SHEEP DURING ANAEMIA

by starch-gel electrophoresis. The percentages ofthe different haemoglobin components in bloodsamples from newborn lambs, which were collectedat different time-intervals after birth and wereanalysed by the DEAE-cellulose chromatographicprocedure, are shown in Table 1. Two lambs, no. 10and no. 11, were heterozygous for Hb-A and Hb-B,and two lambs, no. 12 and no. 13, homozygous forHb-B. The results demonstrate the fast disap-pearance of Hb-F after birth and its replacementby either Hb-A and Hb-B or Hb-B alone, and alsosuggest the presence of small but distinct amountsof Hb-C in newborn AB sheep and its absence innewborn B sheep.

R&sponse of an AB sheep to extreme blood 1088.Sheep 2357 was submitted to experimental anaemia.The quantities of blood removed two or three timeseach week varied from 250 to 750 ml. and areplotted in the upper section of Fig. 4. As a result ofthis treatment a very severe anaemia developed(Table 2 and Fig. 4), which was microcytic andslightly hypochromic with a moderate reticulo-cytosis. The total haemoglobin concentration de-creased from 12-2 g./100 ml. to a final value of2-8 g./100 ml. Electrophoretic and chromato-

graphic analyses of the haemoglobin in the bloodsamples collected at various intervals revealed acontinuous decrease in the percentage of Hb-Aand a corresponding increase in the percentage ofHb-C, whereas no significant change in the per-centage of Hb-B was observed. No Hb-A wasdetectable during the final 2 weeks of the anaemicperiod, whereas the total proportion of Hb-C in-creased to such an extent that it was only slightlyless than that of Hb-B (Fig. 4). Significant differ-ences in the proportions of Hb-A and of Hb-C werefound for young and old erythrocytes. There wasa marked rise in the percentage of Hb-C in theyoung erythrocytes as compared with that of olderythrocytes during the course of the anaemia, andthe concentration of Hb-A was lower in the youngerythrocytes than in the older cells. No significantdifference in the percentage of Hb-B in the two celllayers was observed.

Response of an AB sheep and a B sheep to moder-ate blood loss and the recovery from the anaemia. Asimilar experiment was carried out with the AB

(a) (b) (c) (. ) (e)

0 60 120 180 240 300 360Vol. of effluent (ml.)

Fig. 2. Separation of the different haemoglobin types byDEAE-cellulose chromatography. Experimental detailsare given in the text. (a) Blood sample of sheep 2357(total Hb concn.: 7-8 g./100 ml.); (b) blood sample of lamb11, collected 15 days after birth.

Origin

Hb-A

H b-A .: 0

Fig. 3. Starch-gel-electrophoretic patterns of haemoglobintypes, isolated by DEAE-cellulose chromatography.Experimental details are given in the text. (b) Hb-B; (c)Hb-F; (d) Hb-A. For the purpose ofcomparison the separa-tions of the haemoglobins present in a blood sample ofsheep 2357 (total Hb conen.: 4-3 g./100 ml.) and in a bloodsample of lamb 11 (15 days after birth) are also shown;these samples are labelled (a) and (e) respectively.

26-2

Vol. 93 =403

G. v" VLIET AND T. H. J. HUISMAN

Table 1. Percentages of the different haemoglobin fractions in the blood of lambs at various ages

Experimental details are given in the text.

Lamb Ageno. (days)10 15

3511 15

35210

12 1232

13 325

Hb-A(%)27-351-424-647-857-20000

Hb-B(%)19-334-819-036-042-831-676-67-5

72-6

Hb-C(%)1-60-92-01-400000

Hb-F(%)51-812-954.414-80

68-423-492-527-4

10 - the AB sheep 9087 to this treatment were com-parable with those seen in sheep 2357; the Hb-A

)[5R m r.......f was partially replaced by the Hb-C, but the concen-0 ll l ll .0 .. .S L. .1L. . .. |tration ofHb-B was not affected. For example, the

proportions of Hb-A, Hb-B and Hb-C in the bloodsamples collected 165 days after the start of theexperiments were 24-9, 45-5 and 29-6 % respec-

6L \ tively, whereas 57 % of Hb-A, 43 % of Hb-B and0 % of Hb-C were found in the blood sample col-

4 F lected before the induction of the anaemia. A con-siderable formation of Hb-C was maintained in the

2t- t post-anaemic period. Although the percentage ofHb-C decreased rapidly with a corresponding rise in

° ~ ° the percentage of Hb-A, the total concentration ofHb-C in the blood remained fairly constant

15[; (Fig. 5). The haematological changes observed for

10 the B sheep 289 submitted to a similar experi-mental anaemia were comparable with those found

5p , for the AB sheep 9087. No differences, however,were observed in the type of haemoglobin produced.

0 l5 | 2 | The electrophoretic and chromatographic mobilities0 5 10 15 20 of the single haemoglobin type present in thisTime (weeks) animal before, during and after the anaemic state

hanges in the concentrations of Hb-A (O) Hb-B were identical and no formation of an additionalHb-C (0) in the blood of sheep 2357 before and haemoglobin component was observed.vere experimental anaemia. Experimental details Physicochemical properties of the haemoglobinsin the text. A, B, (7 and F. The visible-light-absorption spectra

of the oxy, deoxy, carbonmonoxy, met and cyan-met derivatives of the four haemoglobin types,

087 and with the B sheep 289. The quanti- isolated by DEAE-cellulose chromatography, werelood removed each week were slightly less identical. Also, no striking differences were ob-ose removed from sheep 2357. This treat- served between the ultraviolet-absorption spectra3sulted in a rather severe anaemia with of the four haemoglobin types. The solubilities ofhypochromic and microcytic erythrocytes Hb-A, Hb-B and Hb-C in phosphate solutions ofoderate reticulocytosis (Table 3 and Fig. 5). different molarities, as determined by the tech-imals were maintained at a total haemo- nique of Derrien (1952), were markedly different.concentration of approx. 5 g./100 ml. for The relatively low solubility of Rb-A comparedwhereupon the bleeding was discontinued. with that of Rb-B (van der Relm et at. 1957) wasrecovery from the anaemia resulted with a confirmed; the solubility of Hb-C was intermediateace of the microcytosis and hypochromia as between the solubilities of Hb-A and Hb-B.from the values of the mean corpuscular The existence of possible gross structural differ-and mean corpuscular haemoglobin. The ences between Hb-A, Hb-B and Hb-C was studiedin the haemoglobin types as a response of by starch-gel electrophoresis of the corresponding

e-̂ -~ 10

00C0 0

0-0_

C,;v:.

0

0 6.

0

Fig. 4. C(A) and ]during seare given

sheep 9(ties of bthan th(ment reseverelyand a m

The aniglobin c

3 weeks,A rapidpersisteijudged:volumechanges

404 1964

I

I

HAEMOGLOBIN TYPES OF SHEEP DURING ANAEMIA

t m

Q.,, 8..2 r-2 coclqtwmw:zP r--l P-4 dq mQ 0 ;4 $4

. .4 C) C.) 4-4.'.) P4-6(1) -

9 (1)

COd CO Col CO CO CO C-

C'- CO 010fI 0N

- CO -cOO COCO CO CO C-I 0) COS-O CO L 14 C-C~ C

110 COCO CO COeseas ) Xo

co_ _cb .*i^ ms

00C' CO _4Co o _: _: cq Ca _-:

oo Qr =t_- _- _-

$ 0-0)

1010 1 10109 10 10.d O(m-4 14 o li 004in -4

- COCOCOaCOCOCO -CO' COCOCOCOCO~

fr

.0~w

E'-

CO

¢CC

o

,.D

0>

EH

0)

x 0 0 0110;.O2 AC4Cf4~oC E.d

C)

10o 10

101

0 O C> &Ac0o0

-.

O04

0 ND

04) PCO4-4E-0C

ce O~~~-0-0iCC

04

00

10 1010bb e

011001_cq oo I: 0)

CO 10CO CO CO CO C

_CO CO 01-OO

_C01 CO CO C4

Vol. 93 405

0' .-I

o to

A o

Ca

Ca

ti0 I)040-14 _

,d

0

fri PO0 0 ~

ODol000

00

PO.

cOf

0

104

CO

0)

0)

00

.0

0-

H1-

-0'.,4 0-

CaE

10

.E 5 C ;.¢)

G. VAN VLIET AND T. H. J. HUISMANglobins in sodium formate-formic acid buffer,pH 19 (Muller, 1960; Huisman, 1963), since withthis technique a complete separation of the twobasic polypeptide chains of these proteins can beobtained. Examples of the results obtained areshown in Fig. 6. The three globin types share onepolypeptide chain, which has been identified as thea-polypeptide chain (Muller, 1961). The non-oc-polypeptide chains possessed different electro-phoretic mobilities, indicating differences in struc-tural composition. The non-cc-chain (i.e. the y-chain) of the foetal haemoglobin showed an electro-phoretic mobility that was slightly greater thanthat of the corresponding chain of Hb-C underthese experimental conditions.The presence of identical a-polypeptide chains

and different non-cc-polypeptide chains in Hb-A,Hb-B and Hb-C was confirmed by the formation ofhybrid haemoglobins from mixtures of thesehaemoglobin types with canine haemoglobin.Results of such hybridization experiments areshown in Fig. 7. The starch-gel-electrophoreticpatterns of the hybrid mixtures each show the

o_0z~-0_

0 -

S

0

0*-4

0 4

v)

presence of two newly formed haemoglobinfractions. The fraction that showed the slowestelectrophoretic mobility and that is composed ofthe a-chains of sheep haemoglobin and the s-chains of canine haemoglobin, ocxheePpan (Ruis-man, 1963), was present in each of the three hybridmixtures, indicating the presence of identical cc-polypeptide chains in the sheep haemoglobin typesA, B and C. The other hybrid haemoglobins pro-duced by recombining the sub-units of caninehaemoglobin and the three sheep haemoglobinsbehaved differently in starch-gel electrophoresis.The mobility of that produced from mixtures ofcdnine haemoglobin and Hb-A was greater thanthat ofHb-A. The component formed in mixtures ofcanine haemoglobin and Hb-B moved ahead of theHb-B and was slightly slower than Hb-A. Thehybrid haemoglobin formed in mixtures of caninehaemoglobin and Hb-C possessed an electro-phoretic mobility that was slightly greater thanthat of Hb-B.

(a) (b) (c)Origin

25s.20 boc._ 5

7 7060ft

30-315

0 1

5e __ _

P 0 50100 150 200

non-

J250

Time (days)

Fig. 5. Changes in the concentrations of Hb-A (0), Hb-B(A) and Hb-C (-) in the blood of sheep 9087 before, duringand after moderate experimental anaemia. Experimentaldetails are given in the text.

Fig. 6. Separation of the polypeptide chains of Hb-C (a),Hb-B (b) and Hb-A (c) by starch-gel electrophoresis insodium formate-formic acid buffer, pH 1-9. Experimentaldetails are given in the text.

406 1964

Vol. 93 HAEMOGLOBIN TYPES OF SHEEP DURING ANAEMIA 407

Physiological propertie8 of haemoglobin8 A, B saturation values were plotted versus the correctedand C. Fig. 8 shows examples of the oxygen dis- logarithms of the P02 at pH 7-0 (at 37°). Thesociation curves of Hb-A, Hb-B and Hb-C, which difference in oxygen affinity between the Hb-A andhad been isolated by DEAE-cellulose chromato- Hb-B, which has been described by Huisman et al.graphy. The curves were constructed from data (1958a), was confirmed. The affinity of Hb-C forobtained for solutions containing approx. 1 g. of molecular oxygen was identical with that of Hb-A;haemoglobin/100 ml., which were dialysed against the logP50 values for both haemoglobin types wereOI-M-potassiumphosphatebuffer,pH7-4.Theoxygen 1-240, whereas the logP50 value for Hb-B was 1-440.

Similar experiments were carried out with moreconcentrated total haemolysates of the red blood

(a) (b) (c) (d) e) cells of the AB sheep 9087 and the B sheep 289,which were collected at various times before, during

C H C H C Hd and after the experimental anaemia. The totalhaemoglobin concentrations were adjusted to

Or gin approx. 5 g./100 ml. before the construction of theoxygen affinity curves. The results of these studiesare expressed as the logP50 values at pH 7-0 (at

.!.' ..'' ' i'37°) and are shown in Fig. 9. The proportions of the

Hb-C ~~~~~~~~~~threehaemoglobin types of the different red-cellH b-C threehaemolysates of sheep 9087 and the total haemo-Hb-C n :X; globin concentrations at the time of collection are

also presented in Fig. 9. No significant change inH b-A the logP50 values was observed; any difference

noted fell within the limit of accuracy of themethod, namely 0-040, which is two standard devi-ations of a mean value of 1-400. Similar resultswere obtained on analysing the various haemo-

+ lysates of the erythrocytes of the B sheep 289. The

Fig. 7. Formation of hybrid haemoglobins from mixturesof sheep haemoglobins with canine haemoglobin as demon-strated by starch-gel electrophoresis. Experimental detailsare given in the text. (b) Mixture of Hb-C and caninehaemoglobin; (c) mixture of Hb-B and canine haemoglobin;(d) mixture of Hb-A and canine haemoglobin. C, Control;H, hybridization experiment. The haemoglobins in the totalerythrocyte haemolysates of sheep 2357 (total Hb concn.:6-1 g./100 ml.) and of the dog are also shown; these samplesare labelled (a) and (e) respectively.

100 r

0

00-4-

bo

80 H

60 F

40 -

20

A

1-0 1-2 1.4 1-6 1-8

log P02

Fig. 8. Oxygen dissociation curves of isolated Hb-A (Q),Hb-B (A) and Hb-C (0). Experimental details are givenin the text.

'0

o-

bo

OQ

co)

C-

0

0 -

0P

4 C-

&.C)0

1-6

1 5

1-4

1-3

15

10

5

-, AA Au AA

0

0 50

A A A AAA A ~AAA L

a3 a

o =,n

O

100 150 200Time (days)

Fig. 9. Changes in the percentages of Hb-A (0), Hb-B (A)and Hb-C (-), and the logarithms of the P50 values of thetotal haemoglobins (El) of sheep 9087 before, during andafter moderate experimental anaemia. The logarithms ofthe P50 values of the total haemolysates (A) of the B sheep289, being collected under identical conditions, are alsogiven. Experimental details are given in the text.

c

E

I

I

I

I I I

G. vN VLIET AND T. H. J. HUISMAN

mean value observed for logP50 was 1-495, with asimilar value of 0-040 as two standard deviations ofthe mean. The mean values of the exponent n inHill's equation, representing the physiologicalinteraction of the haem groups, were 2-0 + 0-25 forthe haemolysates of sheep 9087 and 1-95 + 0*2 forthose of sheep 289.

DISCUSSION

The simultaneous use of more advanced electro-phoretic and chromatographic techniques hasprovided additional information on the changes inthe quantities of the haemoglobin types of sheepduring the stress of severe anaemia. As a result ofthis the interpretation of the observations of thechanges in the proportions of the two haemoglobintypes in AB sheep during severe blood loss, madeby Huisman et al. (1958 a), is no longer tenable. Thepresent findings strongly suggest that under suchcircumstances a new haemoglobin type withdistinct physicochemical properties is produced.This haemoglobin component, designated Hb-C,replaces one of the two adult haemoglobin types,namely Hb-A, whereas the formation of the secondhaemoglobin, Hb-B, seems not to be affected. Theoxygen affinity of Hb-C was found to be identicalwith that of Hb-A. The new haemoglobin type wasoften detectable in small amounts in the blood ofheterozygous sheep that were suffering from a mildhypochromic microcytic anaemia.The results of the hybridization experiments and

starch-gel-electrophoretic studies at low pH leavelittle doubt that one of the two polypeptide chainsof Hb-C differs from the corresponding chains ofthe other haemoglobin types, whereas the samesecond polypeptide chain, the cx-chain, is present inall four haemoglobin variants. Preliminary resultsof experiments, which are not described in detail inthe present paper, indicate the existence of morethan one difference in the amino acid compositionof Hb-C as compared with those of Hb-A, Hb-Band Hb-F. Particularly striking were the differ-ences in the number of methionine and isoleucineresidues. No isoleucine was present in either Hb-Aor Hb-B, whereas 2 and 4 residues of this aminoacid were found in Hb-C and Hb-F respectively.The number of methionine residues were 6, 8, 2 and4 for Hb-A, Hb-B, Hb-C and Hb-F respectively.Because of these differences and of other deviationsin the amino acid compositions of Hb-A and Hb-B(van der Helm et al. 1957), it seems of basic im-portance to investigate the primary structure of thedifferent sheep haemoglobin variants. Such studiesare at present in progress.

There is apparently a specific replacement ofHb-A and not of Hb-B by another haemoglobintype in experimental anaemia. Although the

present study failed to demonstrate any change inthe Hb-B component, the possibility of the forma-tion of an altered haemoglobin type with identicalelectrophoretic and chromatographic propertiescannot be excluded. Application of additionaltechniques, including structural studies of the'Hb-B' types isolated before and during severeanaemia, may prove of further value. The physio-logical characteristics of Hb-B, as determined byits oxygen equilibrium, did not change during theperiod of severe anaemia.The continuous presence of rather large amounts

of Hb-C in the post-anaemic period is also note-worthy. During the 70 days after the last bleedingof the AB sheep 9087 the total concentration ofHb-C present in the blood had remained constantat approx. 1 5 g./100 ml. (Fig. 5), whereas duringthe recovery phase a continuous rise in the concen-trations of both Hb-A and Hb-B was observed.One might therefore postulate that during thisperiod distinct erythrocyte types containingspecific (mixtures of) haemoglobin variants areproduced. The rate of production of erythrocytescontaining mainly Hb-C and probably Hb-B (or anunknown variant with similar properties) seems tobalance their destruction rate, whereas the forma-tion of erythrocytes with mainly Hb-A andprobably Hb-B greatly exceeds the rate of break-down of these cells.The nature of the genetic control of the sheep

haemoglobins A and B is of fundamental interestbecause of the fact that the non-oc-polypeptidechains of these haemoglobin types appear to differby perhaps 10-20 amino acid residues, though theevidence both from pedigrees and population datafor single-gene control of these polypeptide chainsseems to be convincing (van der Helm et al. 1957;Huisman et al. 1958a). Different mechanisms ofmode of inheritance, including one postulatingsingle-gene control of the two different polypeptidechains by a pair of alleles at a regular locus, havebeen discussed (Shreffier & Vinograd, 1962). Theproblem has become more complicated by thepresent evidence for the existence of a regulatorymechanism that is active in anaemic states andresults in the increased production of a third (andpossibly a fourth) type of haemoglobin in hetero-zygous AB sheep. It seems, at present, not possibleto develop a genetic scheme of mechanisms thatwill adequately explain all the observed facts. Anyfinal conclusion must be based on the results ofdetailed analyses of the primary structures of thehaemoglobin types.

SUMMARY

1. The response of sheep heterozygous for thetwo haemoglobin variants Rb-A and Rb-B and

1964408

Vol. 93 HAEMOGLOBIN TYPES OF SHEEP DURING ANAEMIA 409

sheep homozygous for Hb-B to the stress of severeexperimental anaemia has been reinvestigated.

2. In animals subjected to extreme blood lossthe Hb-A was replaced entirely by a new haemo-globin variant (Rb-C), whereas the production ofHb-B apparently was not affected. Under condi-tions of moderate blood loss, the replacement ofHb-A by Hb-C was partial. Hb-C was detectable infairly constant concentrations in the blood ofheterozygous sheep during the recovery phase ofthe anaemia.

3. Hb-C differs from Hb-A and from Hb-B inelectrophoretic mobility and in its behaviour inDEAE-cellulose chromatography. HaemoglobinsA, B, C and foetal haemoglobin seem to share thesame oc-polypeptide chain, but the non-cc-chains ofeach of the four haemoglobin types are distinctlydifferent.

4. The affinity of Hb-C for molecular oxygenwas identical with that of Hb-A and thereforedifferent from that of Hb-B. No changes in oxygenaffinities of total erythrocyte haemolysates of aheterozygous AB sheep and a homozygous B sheepwere observed during the experimental anaemia.

We acknowledge our gratitude to Dr K. Punt and Miss J.Bos (Utrecht, The Netherlands) for their help in the hae-matological studies, and to Miss A. M. Dozy, Mrs C. A.Reynolds, Mrs M. Sheley and Miss J. M. Schillhorn vanVeen for skilful technical assistance. The study was inpart supported by U.S. Public Health Service Grantsno. H-5168 and no. H-6982.

REFERENCES

Blunt, M. H. & Evans, J. V. (1963). Nature, Lond., 200,1215.

Borum, E. R., Figueroa, W. G. & Perry, S. M. (1957).J. clin. Invest. 36, 676.

Brinkman, R. & Dirken, M. N. J. (1940). Acta brev. neerl.Physiol. 10, 228.

Cabannes, R. & Serain, C. (1955). C.R. Soc. Biol., Paris,149, 1193.

Derrien, Y. (1952). Biochim. biophys. Acta, 8, 631.Evans, J. V., Harris, H. & Warren, F. L. (1957). Biochem.

J. 65, 42P.Evans, J. V., King, J. W. B., Cohen, B. L., Harris, H. &

Warren, F. L. (1956). Nature, Lond., 178, 849.Gammack, D. B., Huehns, E. R., Lehmann, H. & Shooter,

E. M. (1961). Acta genet. 11, 1.Harris, H. & Warren, F. L. (1955). Biochem. J. 60, xxix.Huisman, T. H. J. (1963). Advanc. clin. Chem. 6, 231.Huisman, T. H. J. & Dozy, A. M. (1962). J. Chromat. 7,

180.Huisman, T. H. J. & Meyering, C. A. (1960). Clin. chim.

Acta, 5, 103.Huisman, T. H. J., van de Brande, J. & Meyering, C. A.

(1960). Clin. chim. Acta, 5, 375.Huisman, T. H. J., van der Helm, H. J., Visser, H. K. G.& van Vliet, G. (1958b). In Abnormal Haemoglobins,p. 181. Ed. by Jonxis, J. H. P. & Delafresnaye, J. F.Oxford: Blackwell Scientific Publications Ltd.

Huisman, T. H. J., van Vliet, G. & Sebens, T. (1958a).Nature, Lond., 182, 171.

Jonxis, J. H. P. & Boeve, H. W. (1956). Acta med. scand.115, 157.

Meyering, C. A., Israels, A. L. M., Sebens, T. & Huisman,T. H. J. (1960). Clin. chim. Acta, 5, 208.

Muller, C. J. (1960). Nature, Lond., 186, 643.Muller, C. J. (1961). Doctoral Thesis. Assen, The Nether-

lands: Van Gorcum and Co.Rosenthal, T. B. (1948). J. biol. Chem. 187, 25.Shreffler, D. C. & Vinograd, J. (1962). Biochim. biophys.

Acta, 65, 101.Smithies, 0. (1955). Biochem. J. 61, 629.van der Helm, H. T., van Vliet, G. & Huisman, T. H. J.

(1957). Arch. Biochem. Biophys. 72, 331.van Vliet, G. (1960). Doctoral Thesis. Groningen, The

Netherlands: van Genderen.Wintrobe, M. M. (1962). Clinical Hematology. Philadel-

phia: Lea and Febiger.

Biochem. J. (1964), 93, 409

Effect of Cofactors, Oestrogens and Magnesium Ions on the Activityand Stability of Human Glutamate Dehydrogenase

BY J. C. WARREN, D. 0. CARR AND S. GRISOLIADepartment of Biochemi8try, Univer8ity of Kan8a8 School of Medicine, Kan8as City 3, Kan8., U.S.A.

(Received 27 February 1964)

It is well known that glutamate dehydrogenase[L-glutamate-NAD(P) oxidoreductase, EC 1.4.1.3]undergoes aggregation and deaggregation under avariety of conditions, including changes in concen-tration (Olson & Anfinsen, 1952). It has beenclaimed (Tomkins, Yielding & Curran, 1961) that

with aggregation there are reciprocal changes in therates of oxidation of glutamate and other aminoacids (such as alanine) that are substrates for thisenzyme. Other investigators believe that in therange of 0.1-1.0 mg./ml. (e.g. completely deaggre-gated to essentially fully aggregated enzyme)