Purity of Glycera dibranchiata monomer hemoglobin components III and IV determined by isoelectric focusing

Comp. Biochem. Physiol. Vol. 92B, No. 4, pp. 619-622, 1989

0305-0491/89 $3.00 + 0.00 © 1989 Pergamon Press pie

Printed in Great Britain

PURITY OF GLYCERA DIBRANCHIATA M O N O M E R HEMOGLOBIN COMPONENTS III A N D IV D E T E R M I N E D BY ISOELECTRIC FOCUSING IOANNIS CONSTANTINIDIS,* RICHARD L. KANDLERt and JAMES D. SATTERLEE~ Department of Chemistry, University of New Mexico, Albuquerque, NM 87131, USA (Tel: (505) 277 6655) (Received 18 May 1988) Abstract--1. The Glycera dibranchiata monomer hemoglobin components III and IV display behavior upon high voltage isoelectric focusing which is similar, but not identical to the behavior demonstrated by monomer hemoglobin component II (Constantinidis and Satterlee (1987). Biochemistry 26, 7779-7786). 2. Both components III and IV show multiple line behavior and formation of significant amounts of apoprotein when solutions of each holoprotein are focused on polyacrylamide gels. 3. The apoprotein of each component focuses as a single line, indicating that this is the most unambiguous estimate of purity for these proteins. 4. The purity of the component III and IV preparations can be estimated to be at least 95%.

INTRODUCTION

The proteins of the monomer hemoglobin fraction from Glycera dibranchiata (Kandler and Satterlee, 1983; Kandler et al., 1984; Cooke and Wright, 1985; Parkhurst et al., 1980) are interesting because one hemoglobin has been crystallized from this mixture that displays an exceptional amino acid substitution for the distal histidine (E-7 H i s ~ L e u ) (Padlan and Love, 1974; Imamura et al., 1972). Despite this substitution, which imparts unusual kinetic properties (Parkhurst et al., 1980; Mintorovitch and Satterlee, 1988), the overall structural similarity to sperm whale myoglobin is striking (Padlan and Love, 1974, Satterlee, 1984). Based upon the unusual kinetic properties now becoming apparent, these monomer hemoglobins are likely to attract increased attention. Establishing a uniform purity assay for all components, from the different preparation methods currently in use is clearly essential if future work from different laboratories is to be judged comparable. Although Parkhurst et al. (1980) were the first to suggest that the monomer fraction contained three distinct hemoglobins, they failed to experimentally substantiate the homogeneity of their individual isolated hemoglobin components. In fact, evidence has been presented that calls into question the homogeneity of their "component 2" preparation (Kandler et al., 1984). This ambiguity in determining the analytical purity of preparations of the Glycera dibranchiata monomer hemoglobin is quite general (Cooke and Wright, 1985; Imamura et al., 1972; Padlan and Love, 1974; Seamonds et al., 1971; Weber et al., 1977) and originates from the unavailability (until recently) of a well defined purity criterion. We have used different techniques from those workers

mentioned above in isolating three major individual hemoglobins from the monomer fraction, moreover, we have also been the first to subject these individual hemoglobin preparations to high resolution analytical procedures (Kandler and Satterlee, 1983; Kandler et al., 1984; Constantinidis and Satterlee, 1987; Constantinidis et al,, 1988). Most of that work, including the identification of an isoelectric focusing purity criterion, was devoted solely to monomer hemoglobin component II. Recently, we have partially sequenced all three major monomeric hemoglobin components obtained by our isolation methods and successfully cloned and sequenced the eDNA for one of them (P. Simons and J. Satterlee, unpublished results). Despite this work, which implies a high level of giobin purity for preparations of all of our components, it is only the Glycera dibranchiata monomer component II hemoglobin component whose purity has been established by biochemical means. Here we wish to present results that establish the purity and isoelectric focusing behavior of the other two major monomeric hemoglobins obtained in our preparations (components III and IV) and contrast the isolation procedures. MATERIALS AND METHODS

Our procedures for isolating and purifying the three major monomeric hemoglobins from Glycera dibranchiata have previously been described in detail (Kandler and Satterlee, 1983; Kandler et al., 1984; Constantinidis and Satterlee, 1987). As previously reported, separations with the uniformly oxidized (ferric) proteins have advantages over separations involving the reduced (ferrous; CO ligated) proteins (Kandler and Satterlee, 1983; Kandler et al., 1984). Both techniques have been used in this work. Isoelectric focusing methods employed here, using precast LKB poly*Present address: NMR Laboratory, Johns Hopkins Uni- acrylamide gels that were not prefocused, were exactly those versity Medical School, Baltimore, MD, USA. previously described in detail (Constantinidis and Satterlee, ]'Present address: Travenol Laboratories, Roundlake, IL, 1987). Gels were either stained with Coomassie Blue in USA. order to detect protein bands, or were simply fixed and left ~Author to whom correspondence should be addressed. unstained in order to selectively detect the heine containing 619

620

IOANNIS CONS'IANT1NII)IS ~'{ ~1/.

bands (red color) (Kandler and Satterlee, 1983; Kandler et al., 1984). Gels were scanned and integrated using instruments previously described (Constantinidis and Satterlee. 1987). RESULTS AND DISCUSSION

Purity and components II1 and I V Three major monomeric hemoglobins (components II, III and IV) can be isolated from Glycera dibranehiata erythrocytes (Kandler and Satterlee, 1983; Kandler et al., 1984; Cooke and Wright, 1985). In previous work on component II it has been demonstrated that the gel electrophoresis and IEF behavior, although unusual, can be understood in detail (Constantinidis and Satterlee, 1987). Those results indicated identical electrophoresis behavior for all three components, thereby showing that they could not be discriminated on the basis of size. Furthermore, for component II, multiple lines were observed on IEF gels of the holoprotein, regardless of ligation or oxidation state, whereas the apoprotein focused as a single line (Constantinidis and Satterlee, 1987). The results of identical IEF experiments for the unseparated monomer fraction and the individual monomer hemoglobin components III and IV, isolated by our method, are presented in Fig. 1. The multiple line behavior of the holoprotein lanes of each (unseparated: lane A; component II: lane C: component III: lane E, component IV: lane G) is obvious. Component II is included in this figure for comparison. In each holoprotein lane stained protein bands occur in three general pl areas: 6.8, 7.3 and 7.7. However, the single band displayed by the apoprotein of each component (lanes D, F, H) indicates that only a single globin is present in these preparations. It is obvious from the comparisons in Fig. 1 that the pI 7.7 band corresponds to apoprotein and that it is present in every individual monomer component holoprotein lane. The conditions that contribute to formation of apoprotein were illustrated in detail for component II (Constantinidis and Satterlee, 1987). Note, also, that in the apoprotein lane of the unseparated monomer fraction (lane B) two major bands occur at pI 7.7. This doublet reflects the slight difference in isoelectric point for apo-component II (lanes C, D) compared to the apoprotein of components III and IV, which display essentially identical isoelectric points (lanes E-H). It confirms the constitution of the unseparated fraction. The remaining two areas demonstrating protein bands for the holohemoglobin lanes (A, C, E, G) are at pl 6.8 and 7.3. Both regions contain multiple lines. This work and that reported before, for component II, illustrate that these bands occur only for the holoproteins. This multiple band effect demonstrated on LKB precast polyacrylamide gels occurs for all three monomer hemoglobin components and is associated with the presence of the heme group (Constantinidis and Satterlee, 1987). Myoglobin also shows this multiple band behavior on our precast LKB gels, whereas apomyoglobin does not (Constantinidis and Satterlee, 1987). We have shown (Kandler and Satterlee, 1983; Constantinidis and Satterlee, 1987) that the relative intensities of the

GEL pH 7.7.* 7.3.,

A B C D E F G H

6.8 ~

Fig. 1. Stained potyacrylamide isoelectric focusing gel. Lane A: unseparated monomer fraction holohemoglobin: t,ane B: unseparated monomer fraction apohemoglobin; Lane C: component II holohemoglobin: Lane D: component II apohemoglobin; Lane E: component lII holohemoglobin; Lane F: component 1II apohemoglobin; Lane G: component IV holohemoglobin: Lane H: component IV apohemoglobin. pI 6.8 and 7.3 bands reflect changes in oxidation and ligation state of the heme in holoprotein component II and similar work for components III and IV (data not shown) produced identical results. The possible sources of this effect were discussed in the component II work (Constantinidis and Satterlee, 1987). Figure 1 shows that the globins (apoprotein) of monomer components III and IV each focus as a single line, so that, as demonstrated for component II, a discriminating purity criterion is high voltage isoelectric focusing of the apoprotein. The purity of our monomer component IIl and IV preparations can be established by scanning and integrating gels like that shown in Fig. 1. Given the range of gel loading conditions during multiple determinations, we estimate that our individual monomer component III and IV preparations consist of a globin with a single pI to the extent of 95%. This high degree of homogeneity is supported by high resolution N M R results which indicate no detectable heme containing impurities are present in these preparations (Constantinidis et al., 1988). Apoprotein is present immediately q/ter lysis A question has arisen whether the detection of apoprotein in holoprotein gel lanes is a consequence of the handling procedures since isolation and purification takes approximately three weeks. Figure 2A represents a gel which was over-loaded with erythrocyte lysate solution immediately following isotonic lysis of washed cells. Lane A has been stained, lane B has not and the dark areas (in B) reflect visually detectable heme-containing bands. The pI 7.7 band can clearly be seen on the stained lane, but not on the unstained lane. By comparing its pl to the data presented in Fig. 1, and considering its invisibility on the nonstained gel, it is concluded that this band is due to the apoprotein. This indicates that apoprotein formation is not a consequence of prolonged handling, since it is in evidence on IEF gels immediately

Glycera hemoglobin

621

A GEL pH

B A e - "--

Gel pH

C

O

7.6,,~ 7.3,,

7.6-I,

6.8 7.3,*

Fig. 2. Polyacrylamide isoelectric focusing gel. (A) Lane A: stained, overloaded gel of a lysate solution loaded immediately after isotonic lysis; Lane B: unstained gel of the same lysate solution showing the heme containing components as dark bands. (B) Stained isoelectric focusing gel of the band representing "component 2" from reduced, CO-hemoglobin isocratic isolation (Lane C) and the component 3 from the same isolation (Lane D). after erythrocyte lysis. Rather, this phenomenon seems to be characteristic of these monomer hemoglobins, or to their interactions with the gel. This situation implies that the heme association equilibrium constant with globin is lower than that displayed by myoglobin, for which no similar apoprotein formation on gels is observed (Constantinidis and Satterlee, 1987). This conclusion provides experimental support for a previous suggestion, based upon u.v.-visible spectroscopy of the Glycera dibranchiata monomer hemoglobins (Seamonds et al., 1971), that, " . . . the globin moiety may be less tightly bound to the heme than in other hemoglobins and myoglobins." Comparing isolation procedures

The results presented here, in conjunction with the IEF behavior of component II (Kandler and Satterlee, 1983; Constantinidis and Satterlee, 1987) and previous N M R and chromatography results (Kandler et al., 1984; Constantinidis and Satterlee, 1987; Constantinidis et al., 1988) constitute the most detailed analytical examination of the Glycera dibranchiata monomer hemoglobins yet published. It was noted above that our isolations involve using the oxidized (ferric state) proteins, whereas other isolations have employed the reduced (ferrous state) proteins (Cooke and Wright, 1985; Parkhurst et al., 1980; Weber et al., 1977). The work of Parkhurst et al. (1980) is particularly relevant to these findings because it produced often cited oxygen and carbon monoxide binding kinetics results. In repeating the procedure described in Parkhurst et al. (1980) our N M R results revealed that the protein band corresponding to "component II" (Parkhurst et al., 1980) actually contained a combination of two hemoglobins (components II and III) that are produced by our isolation method (Kandler et al.,

1984). The reason for this is apparently the change in pI of the hemoglobins in the reduced, CO ligated state (compared to the oxidized form) and concomitant lower resolution that results from this on the ion exchange column. In Figure 2B, Lane C, we present an IEF gel of the so-called "component 2" isolated by our repeat experiments with the ferrous COhemoglobins (Kandler et aL, 1984) employing as accurately as possible the previous conditions (Parkhurst et al., 1980). Near pI 7.3 this gel displays more complexity than shown by any of our isolated component hemoglobins (Fig. 1) or to the "component 3" protein from this same reduced protein isolation (Fig. 2, Lane D). One line (indicated by arrow) near 7.3 corresponds to the pI of our component II, whereas other bands (marked by asterisks) correspond to the pI of our component III. Further confirmation of this is shown by the doublet of bands near pI 7.7, indicating that apoprotein bands characteristic of our component II and III (or IV) simultaneously occur in this lane. The results of Figure 2B, in conjunction with previous N M R results (Kandler et al., 1984) confirm the lack of resolution in single column ion exchange separations (at pH 6.8) of the ferrous CO-ligated monomer hemoglobins and suggest that the previous "component 2" actually consists of a mixture of the components that are isolable using our procedure. In view of this we suggest that it may be necessary to exercise caution in interpreting the published kinetics results (Parkhmst et al., 1980), until the purity and homogeneity of t':e individual components from that isolation are adequately established.

Acknowledgements--We are grateful for support from the National Institutes of Health (DK 30912; HL01758, a Research Career Development Award) to J.D.S. This work

IOANNIS CONSTANTINIDIS Ct al.

622

was carried out during the period that J.D.S. was a Fellow of the Alfred P. Sloan Foundation.

REFERENCES Constantinidis I. and Satterlee J. D. (1987) Isoelectric focusing purity criteria and ~H nmr detectable spectroscopic heterogeneity in the major isolated monomer hemoglobins from Glycera dibranchiam. Biochemistry 26, 7779-7786. Constantinidis I , Satterlee J. D., Pandey R. K., Leung H-K. and Smith K. M. (1988) Assignment of selected hyperfine proton nmr resonances in the met forms of Glycera dibranchiam monomer hemoglobins and comparisons with sperm whale metmyoglobin. Biochemistry 27, 3069-3076. Cooke R. M. and Wright P. E. (1985) Differences in amino acid composition and heme electronic structure of the multiple monomeric hemoglobin components of Glycera dibranchiata. Biochim. biophys. Acta 832, 357-364. Imamura T., Baldwin T. O. and Riggs A. (1972) The amino acid sequence of the monomeric hemoglobin component from the bloodworm Glycera dibranchiata. J. biol. Chem. 247, 2785 2797. Kandler R. L. and Satterlee J. D. (1983) Significant heterogeneity in the monomer fraction of Glycera dibranchiata hemoglobins. Detection, partial isolation and characterization of several protein components. Comp. Biochem. Physiol. 75B, 499 503.

Kandler R. L., Constantinidis I. and Satterlce J. D. (1984l Evaluation of the extent of heterogeneity in the Glvcera dihranchiata monomer haemoglobin fraction by the use of n.m.r, and ion-exchange chromatography. Biochem. J. 226, 131 138. Mintorovitch J. and Satterlee J. D. (1988) Anomalously slow cyanide binding to Glycera dihranchiata monomer methemoglobin component II: implication for the equilibrium constant. Biochemt~,tO, 27, 8045 8050. Padlan E. A. and Love W. E. (1974) Three dimensional structure of hemoglobin from the polychaete annelid GO,cera dihranchiata at 2.5 A resolution. J. biol. Chem. 249, 4067 4078. Parkhurst L. J., Sima P. and Goss D. J. (1980) Kinetics of oxygen and carbon monoxide binding to the hemoglobins of GO'cera dibranchiata. Biochemistry 19, 26882692. Satterlee J. D. (1984) Anomalous pH dependence of the heme-bound carbon monoxide spectroscopic properties in the Glycera dibranchiata monomer hemoglobin fraction compared to vertebrate hemoglobins. Biochem. biophys. Acta 791, 384 394. Seamonds B., Forster R. E. and George P. (1971) Physicochemical properties of the hemoglobins from the common bloodworm G(vcera dihranchiata. J. biol. Chem. 246, 5391 5397. Weber R. E., Sullivan 8 , Bonaventura J. and Bonaventura C. (1977) The haemoglobin systems of the bloodworms Glvcera dibranchiata and G. americana. Oxygen binding properties of haemolysates and component haemoglobins. Comp. Biochem. Physiol. 58B, 183 187.