Detection and quantification of myocardial cell death: Application of monoclonal antibodies specific for cardiac myosin

Journal of Molecular and Cellular Cardiology

(1982)

14, Suppl. 3, 139-l

46

Detection and Quantification of Myocardial Cell Death: Application of Monoclonal Antibodies Specific for Cardiac Myosin* Edgar Haher, Hugo A. Katus, John G. Hurrell, Gary R. Matsueda, Paul Ehrlich, Vincent R. Zurawski, Jr. and Ban-An Khaw Cardiac Unit, Massachusetts General Hospital, and Harvard Medical School, Boston, Massachusetts, USA E. HABER, H. A. KATUS, J. G. HURKELL, G. R. MATSLEDA, P. EHRLICH, V. R. Zuto~ws~! AND B.-A. KHZILY. Detection and Quantification of Myocardial Cell Death: Application of Monoclonal Antibodies Specific for Cardiac Myosin. Journal of Molecular and Cellular Cardiolou (1982) 14, Suppl. 3, 139-146. The development of cell-membrane permeability to macromolecules is likely to be an unequivocal criterion of cell death. In order to utilize this phenomenon as an organ-specific probe in the diagnosis of myocardial infarction, the leakage of cardiac-myosin light chains from damaged myocardium as well as the entry and fixation of radioactively labeled cardiac myosin-specific antibody into cells was examined. In 25 rabbits immunized with cardiac-myosin light chains, none yielded antisera that discriminated fully between cardiac- and skeletal-muscle light chains, thus failing to yield an organ-specific reagent. While immunospecitically purified rabbit antisera raised in response to immunization with cardiac myosin resulted in useful in uiuo imaging reagents for myocardial infarcts, the quantity of antibody produced by these methods was severely limiting. The application of somatic-cell fusion of mouse myeloma cells with mouse plasma cells secreting either antibody specific for cardiac myosin or cardiac-myosin light chains allowed for the cloning of hybrid cell lines that continuously secreted homogeneous antibodies. The cardiac-(light chain) antibodies, utilized either singly in a conventional radioimmunoassay or in pairs in a sandwich immunoradiometric assay, markedly enhanced the resolution between cardiac- and skeletal-muscle light chains. Either the myosin light chain or myosinspecific cell lines could be amplified by growth in viva as plasmacytomas, permitting the harvesting of gram quantities of antibodies. Thus, the problem of supply was readily solved. These monoclonal antibodies provide a powerful tool for the localization and quantification of myocardial necrosis in uiuo. KEY WoRDs:

iVlyocardia1

infarct;

Myosin;

Monoclonal

antibodies;

Myosin

light chains.

Introduction Vascular occlusive disease is the major cause of death and disability in the more economically advanced countries of the world. Consequently, it is uot at all surprising that a major research eabrt has been expended in devising methods to salvage ischemic tissues. The evaluation of these experimental approaches requires an objective definition of cell death, preferably one that can be discerned reliably either in the living animal or patient. The physical destruction of the cell membrane is certainly an irreversible phenomenon. By this we mean not simply the failure of membrane function in the regulation of the ionic environment of the cell, but actual loss of the barrier between intra- and extracellular space, so that the macromolecular consents of the cell are dispersed. The leakage of macromolecules from the cell has long been assumed to be a marker of loss of cell-

* Supported

by funds

made

available

by the trustees

membrane integrity and as such has become a frequently utilized experimental and clinical tool [15, 21, ZZ]. More recently radiolabeled antibodies specific for intracellular constituents have been used as markers for cells that expose their cytoplasmic antigens to extracellular fluid [4]. The wide application of these methods as sha.rp experimental tools has been impaired either by lack of organ specificity or by the lack of availability of reagents (antibodies) reproducible in large quantities so that the same measurements might be made in different laboratories. ‘I‘he application of the method of somatic-cell fusion to the in vitro and in viuo propagation ofhomogeneous clones of antibody-secreting cells had revolutionized the production of antibodies. While the conventional antiserum contains many antibodies to a single antigen that differ from one

of the Massachusetts

General

Hospital

E. Haber

140

another not only in affinity but in the specific part of the antigen molecule recognized, the monoclonal antibodies that are the product of the cell-fusion technique are a homogeneous molecular species that is uniform in affinity and capable of binding only to a single epitope on the surface of the antigen molecule. Once a desired clone of cells has been isolated, it is for all practical purposes immortal and capable of producing indefinite quantities of antibody. In this paper we will demonstrate how the application ofmonoclonal antibodies for cardiac myosin and for cardiac-myosin light chains provides increased discrimination in the diagnosis of myocardial infarction and afIords an opportunity for the widespread application of these tools. In an effort to enhance sensitivity and specificity in the detection of myocardial-cell necrosis, the radioimmunoassay for the MB-isozyme of creatine kinase was developed [18, 271. However, normal plasma contains some of this enzyme, and the antiserum that is raised to the BB-isozyme also reacts with enzyme from the brain [IL?]. Another intramyocardial-cell protein that provides an opportunity for sensitive measurement is myoglobin. Plasma concentrations of myoglobin, as measured by radioimmunoassay, have been demonstrated

Materials Preparation

of monoclonal

Either DEAE

to rise subsequent to myocardial infarction [23]. However, cardiacand skeletal-muscle myoglobin are identical [13]. Since a variety of injuries to skeletal muscle, as well as strenuous exercise, release measureable amounts ofmyoglobin into the circulation [z+], the source ofthe myoglobin may, at times, be unclear. Cardiac-myosin light chains are structurally [16, ZO] and immunologically [x] different from skeletaland smooth-muscle myosin light chains and thus may provide a unique cardiac-specific antigen. Although differences exist, there is sullicient immunologic similaritybetween myosin light chains from various tissues that antibodies that uniquely recognize one and not the others have not been obtained [8,25]. The sera of 25 rabbits immunized with human cardiac-myosin light chains were examined by us at varying times after immunization. Cross-reactivity between cardiacand skeletal-muscle light chains varied between 10 and 100%. An antiserum could not be found that would provide the requisite specificity useful in clinical assay. We report here a solution of this problem and a method that employs monoclonal antibodies that recognize two different epitopes on the same antigen and thus markedly enhances the resolution possible in immunoassay.

and Methods

antibodies

Human cardiac and skeletal myosin were isolated from ventricular myocardium or psoas muscle obtained at necropsy. Myosin light chains were separated from myosin heavy chains [K] and their homogeneity assessed by SDS gel electrophoresis. Somatic-cell fusion was performed according to the general method ofliijhler and Milstein [IO]. Balb/c mice were hyperimmunized with several injections of either unmodified or gluteraldehyde-(crosslinked) human cardiac-myosin light chains in complete Freund’s adjuvant injected intraperitoneally, followed two months later by an intravenous booster injection. Three days thereafter the spleens were excised, and the spleen cells were fused with 10’ PS-NSl/l-Ag4-1 (NSl) cells [u] in 30% polyethyleneglycol. The cells were seeded in four culture plates (Costar, Cambridge, MA). Ofa total of 229 wells, 89.5% showed growth in selective medium [I.?]. To screen for antibody in the culture

Development

et al.

of myosin-light

the mixed proteins from ascites fluid or a cellulose-purified IgG fraction was coupled

specificfor

myosin light chains

medium, a solid-phase assay was employed [19] utilizing myosin light chains immobilized on plastic and iodinated goat-(anti-mouse) F(ab’), as second antibody. In 35% of the wells showing cell growth, antimyosin light chain antibody could be detected. The experiments depicted in Figure 1 were performed directly on culture supernatants without an attempt at cloning. Three cell lines, selected for further characterization, were subcloned three times to assure monoclonality, utilizing limiting dilution computed to yield one cell per well. The experiments depicted in Figures 2 and 3 and Tables 1 and 2 represent results obtained with the products of these monoclonal cell lines. The cell cultures were propagated in vivo utilizing Balb/c mice primed with Pristane. Antibody-enriched ascites could be collected 7-14 days after tumor-cell injection by paracentesis. Isotypic analysis showed IgG, antibodies in all four cell lines.

chain radioimmunoassay to cyanogen bromide-activated Sepharose using 2-3 mg protein per ml of Sepharose.

4B [I] Purified

Monoclonal

Antibodies

Specific

for Cardiac

Myosin

111

3c7

Dilution

FIGURE I. Binding of hybridoma antibodies to cardiac- or skeletal-muscle light chains immobilized on plastic microtiter plates. Dilution ofthe cell-culture supernatant is plotted against ct/min of ‘251-labeledgoat-(anti-mouse) Fa b antibody bound to the plate that recognizes the myosin light chain-specific antibody.

anti-myosin-(light chain) antibody and human heart-myosin light chains were iodinated with I mu “‘4 using the lactoperoxidase method [14] to a specific activity of 10 @/pg. Goat-(anti-mouse) F(ab’)Z was iodinated with 1 mCi “‘1 using chloramine T as oxidant [Z]. The antibody-binding capacity of Sepharose-substituted antibody was examined by measuring the binding of “‘I-myosin light chains. The quantity of antibody substituted

onto Sepharose was determined by the binding (of ““I-goat-(anti-mouse) F(ab’),. The capacity of ““I-substituted antibodies to bind to myosin light chains was examined in a similar manner utilizing Sepharose 4B to which myosin light chains had been bound. The sandwich assay was carried om utilizing Sepharose-antibody, myosin light chains, and rZ51-antibody.

Results and Discussion Examples of the varieties of antigenic specificities found are depicted in Figure 1. Antibody 3C7 typifies the most common type. There is complete identity of reactivity between cardiac- and skeletalmyosin light chains, indicating recognition of common antigenic determinants. Antibody 2B2 is representative of a less frequent type. These monoclonal antibodies demonstrate a partial crossreactivity between skeletal- and smooth-muscle light chains. Perhaps only a part of a common antigenie determinant is recognized. The least common monoclonal antigenic determinant is typified by lE6. This antibody shows no evidence of crossreactivity between skeletal- and cardiac-muscle light chains, and thus recognizes the unique determinants of the cardiac light chain. It can provide the basis for the development of an entirely specific immunoassay for cardiac cell damage. Three monoclonal antibodies that exhibited varying degrees of cross-reactivity between cardiacand skeletal-muscle light chains were selected for study. When immobilized on Sepharose, each

bound ‘““I-(cardiac myosin) light chains effectively allowing ready determination of cross-reactivity with skeletal-muscle light chains utilizing a solidphase competitive radioimmunoassay (Figure 2). Increasing concentration of either unlabeled cardiac or skeletal light chains resulted in decreased binding of ‘““I-cardiac light chains. Antibody 1C5 proved to be fully cross-reactive (Figure 2a). 2B9 was 17.5% cross-reactive (Figure 2b), and 4F10 was 25% cross-reactive (Figure 2~). The antibodies appear to possess varying affinities for myosin light chains, the most cross-reactive antibody, lC5, provided the greatest sensitivity in measurement. Figure 3 demonstrates the enhanced specificity that is inherent in an assay where two epitopes ajre independently measured. One monoclonal amibody is immobilized by covalent linkage of Sepharose. When exposed to an antigen mixture, only those molecules recognized by that antibody will adhere. A second antibody labeled with tz51 is then added. It will bind only to those antibody molecules that are both immobilized on the column

E. Haber

142

et al.

(b)

8060-

3

IOO-

0

SC

8060-

0’

0.1

1

I

IO

100

Myosin FIGURE presence (* -a), added.

of

2. Binding increasing unlabeled

(a), Antibody

of””

1c5;

I cardiac-myosin concentrations skeletal-muscle (b) antibody2B9;

1

I

/

1000

10000

00000

light

chains

light chains to monoclonal of unlabeled cardiaclight chains added (0 -O),

I

(ng)

or

antibodies immobilized on Sepharose light skeletal-muscle myosin unlabeled cardiac-muscle light

in the chains chains

(c) antibody4FlO.

and possess the epitope for which it is specific. Thus immobilized radioactivity represents the recognition of two different epitopes on the antigen molecule. In Figure 3a a fully cross-reactive antibody is combined with a partially cross-reactive one, resulting in apparent cross-reactivity that reflects only the discrimination of the partially cross-reactive antibody (Table 1). In Figure 3b and 3c enhanced resolution is afforded by combining two partially cross-reactive antibodies (Table 2). A logical consequence of the last argument is that the use of the same antibody, both immobilized to Sepharose and labeled, should result in no apparent binding of the label. This is demonstrated in Figure 3a. When antibody 2B9 is both immobilized on the support and labeled, no radioactivity remains on the solid support after washing. Monoclonal antibodies recognize a single antigenie determinant, which should allow for enhanced specificity of recognition; and thus one should be able to select an antibody specific for the

TABLE antigen

1. Competitive

Antibody

assay

utilizing

labeled

Measured fractional cross-reactivity

IC5

1 .oo

2B9 4FlO

0.17 0.25

epitope that defines the difference between two very similar molecules (Figure 1). Yet it appears that cross-reactivity is often seen. With monoclonal antibodies this is probably a manifestation of the sharing of parts of an epitope by two antigens. A method that entailed the combination of two antibodies that recognize different epitopes on the same molecule should enhance specificity markedly (Figure 4). The differentiation between cardiacand skeletal-muscle myosin appears to offer a good test case. The molecules are different structurally,

Monoclonal

Antibodies

Specific

I 13

for Cardiac Myosin

(a)

(b)

F iooc 0 8047 8 ;=I 60D

+

IO

100

1000

o’--p/p

10000

Myosin

0.1

light chains

I

IO

100

1000

10000

(ng)

FIGURE 3. Binding of ““I-labeled second antibody to antigen that had first been bound to Sepharose-immobilized antibody. Percent maximal labeled-antibody binding in antigen ezcess is plotted against the amount of cardiacor skeletal-muscle light chains added. (a) skeletal-muscle light chains (Ol ), cardiac-muscle light chains (O-- 0) measured with antibody 2B9 immobilized on Sepharose and antibody lC5 labeled. Cardiac-muscle light chains A) measured with antibody 2B9 immobilized on Sepharose and antibody 2B9 labeled. (b) Cardiac-muscle (Alight chains (O0), skeletal-muscle light chains (O0) measured with antibody 4F10 immobilized on Sepharose and antibody lC5 labeled. (c) Cardiac-muscle light chains (O0), skeletal-muscle light chains 4) measured with antibody 4F 10 immobilized on Sepharose and antibody 2B9 labeled. (unpublished data). P------

TABLE

2. Bideterminant

assay utilizing

antibody

representation

Measured cross-reactivity

0.17 0.25 0.037

lC5 and 2B9 lC5 and 4FlO 2B9 and 4FlO

4. Schematic

second

Calculated cross-reactivity

Antibodies

FIGURE

labeled

ofthe

bideterminant

0.17 0.25 0.043

immunoassay

utilizing

two monoclonal

antibodies.

144

E. Haber et al.

yet elicited antibodies (mixtures) and most monoclonal antibodies we have tested utilizing a conventional solid-phase radioimmunoassay (Figure 2 and Table 1) exhibit varying degrees of cross-reactivity. When the assay is constructed so that two different monoclonal antibodies must bind to the same antigen, the resultant cross-reactivity observed corresponds as might be expected from theoretical considerations to approximately the

Monoclonal,

myosin-specific

antibc Idies in myocardial-infarct

We have previously demonstrated the feasibility of using radiolabeled, affinity column-purified anticanine cardiac myosin antibody fragments administered intravenously for localization and visualization of necrotic regions of infarcted myocardium in VZUO[3, 41. When cell death occurs, the intracellular protein, myosin, is exposed to extracellular fluid. It is then available to react with labeled antibodies or antibody fragments. Myosin is very insoluble in, physiological fluids so that membrane breakdown does not result in antigen loss. Conventionally elicited antibodies labeled with various radioisotopes such as ‘“51-Ab(Fab’)2 [17], r3’I-Ab(Fab’), [s], “‘In-DTPA-Ab Fab [7], 67Ga-DTPA Fab [6], and synTc-DTPA-Ab Fab [9] have been used successfully in infarct tissue. While elicited antibodies have been very useful in pilot studies, they present major impediments in wide-scale clinical application of myocardialinfarct imaging. Elicited antibodies are intrinsically heterogeneous. As a consequence, different batches of serum even from the same donor animal are likely to present a different spectrum of antigenic specificities. Antibodies must be affinity purified from serum since a monospecilic antibody preparation is required in imaging studies. An admixture of radiolabeled proteins that are not antibodies to the antigen to be imaged results not only in an

Preparation

product of the two fractional cross-reactivities (Table 2). It is likely that the two epitopes selected must be sufficiently far from one another on the molecule’s surface so that steric hindrance between the two monoclonal antibodies does not occur. We have not as yet seen an example of this problem, but Figure 3a1 clearly demonstrates that it is impossible to bind two antibody molecules to the same epitope.

increased radiation dose to the patient but also unnecessary background radioactivity. Both affinity purification and repeated immunization require large amounts of scarce antigens. In myocardial-infarct imaging, the antigen in question is human cardiac myosin, a material obviously in limited supply. The hybridoma method of antibody production [II] provides a solution to all these problems. Antibodies from monoclonal cell cultures are homogeneous and reproducible in industrial quantities. Since the major product of the cell line is a specific antibody (especially when non-secreting myeloma cell lines are used in the initial cell fusion), affinity chromatography is often unnecessary, and the antibody may be isolated by simple ionexchange chromatography. Finally, antigen requirements are minimal. Initial immunization is generally carried out in mice, each animal requiring only a few micrograms of antigen. Once a desired cell line is obtained, further immunization is not needed. An additional dividend inherent in the selection of desirable clones is that highly purified antigen is not needed. If a mixture of antigens is used in the initial immunization, only those clones of hybrid cells that secrete antibody to the desired antigen need to be propagated.

of monoclonal ai ntibodies speciicfor

Balb/c mice were immunized with human cardiac myosin purified from left ventricular myocardium obtained at necropsy utilizing the same method as previously described for canine myosin [4]. A similar schedule of immunization was utilized as described above for cardiac light chain; the cellfusion procedure was also carried out in an identical manner; and screening for antibody in the culture medium was performed utilizing a solid-phase assay as above. Antimyosin antibody activity could

imaging

myosin

be detected in 75% of wells showing cell growth in selective media. The cell line, WM-2, the product of which was utilized in the imaging experiments described here, was subcloned three times utilizing limiting dilution to one cell per well to assure monoclonality. This antibody reacted with both canine and human cardiac myosin and thus could be used for imaging experimental canine myocardial infarcts. It was labeled with g9ntechnetium using the technique described above.

Monoclonal

Antibodies

Specific

for Cardiac

Myosin

145

The canine myocardial-infarct model Each dog was anesthetized by intravenous pentobarbitol; a left thoracotomy was performed under sterile conditions, and the left anterior descending (LAD) coronary artery was occluded with a silk suture approximately two-thirds the distance from the apex to the base in order to produce approximately 30% left ventricular-wall cyanosis [5, 171. Four hours following LAD coronary artery occlusion, the ligature was removed and 5 mCi of 08m technetium-labeled antimyosin hybridoma b’M-2 &as injected into the LAD coronary artery. A scintigraphic image (left lateral and anteroposterior views) was obtained 18 h later (Figure 5) that demonstrated the localization of the labeled antibody in the anteroapical region of the heart.

The power of specific antibody as a vehicle for imaging specific antigens in uivo has been amply demonstrated. The availability of monoclonal antibodies as well as better radioactive labeling techniques will allow wider application of this approach to the detection of antigens that are rare and diff;cult to purify. Candidates for imaging include specific tumor antigen, intracellular antigens for detection of necrotic tissue, as well as organor tissue-specific antigens. Thus two major properties inherent to monoclonal antibodies, selectivity and availability ‘in essentially limitless quantities, have effected a potential solution to significant problems in myocardial-infarct diagnosis.

FIGURE 5. Left lateral (IMAGE 1) and anteroposterior (IMAGE 2) scintigrams showing localization ggmTc-labeled monoclonal DTPA-antimyosin (WM-2) Fab fragments 18 h post-intracoronary administration canine experimental myocardial infarction. The lower central activity is also due to liver activity.

of in a

REFERENCES i. 2. 3. 4. 5. 6.

In Methoak in Enpnology, Vol. 22, Enzyme CUATRECASAS, P. & ANFINSEN, C. B. Affinity chromatography. purification and related techniques, W. B. Jakoby, Ed., Chap. 31. New York: Academic Press (1971). GREENWOOD, F. C., HUNTER, W. M. & GLOVER, J. S. The preparation of ‘3’I-labelled human growth hormone of high specific radioactivity. BiochemicalJournal 89, 114-123 ( 1963). HOFFER, P. B., LATHROP, K., BEKERMAN, C., FANG, V. S. & REFETOFF, S. Use of 131-I-CEAantibodyas a tumolr scanning agent. Journal of Nu&ar Medicim~l5, 323-327 (1974). KHAW, B. A., BELLER, G. A., HABER, E. & SMITH, T. W. Localization of cardiac myosin-specific antibody in experimental myocardial infarction. Journal of Clinical Investigation58,439-446 (1976). KHAW, B. A., BELLER, G. A. & HABER, E. Experimental myocardial infarct imaging following administration of Iodine- 13 1 labeled antibody (Fab’):! fragments specific for cardiac myosin. Circulation 57,743-750 ( 1978). KHAW, B.A., FALLON, J. T., KATUS, H., ELMALEH, D., STRAUSS, H. W., LOCKE, E., POHOST, G. M. & HABER, E. Positron imaging of experimental myocardial infarction with 68Ga-DTPA-antimyosin antibody. Cirnrktion 59-69 (II), 135 (1979).

146 7. 8.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21.

22. 23.

24. 25. 26. 27.

E. Haber et al. KHAW, B. A., FALLON, J. T., STRAUSS, H. W. & HABER, E. Myocardial infarct imaging with Indium-11 ldiethylene triamine pentaacetic acid-anticanine cardiac my&n antibodies. Science 209, 295-297 (1980). KHAM’, B. A., GOLD, H. K., FALLON, J. T. & HABER, E. Detection of serum cardiac myosin light chains in acute experimental myocardial infarction: radioimmunoassay of cardiac myosin light chains. Circulation, 58, 1130-l 136 (197%). KHAW, B. A., STRAUSS, H. W., CARVALHO, A., GOLD, H., LOCKE, E. & HABER, E. Labeling anti-cardiac myosin Fab and human fibrinogen with technetium-99m. Submitted to Journal of Nuclear Medicine. KBHLER, G. & MILSTEIN, C. Continuous cultures offused cells secreting antibody ofpredelined specificity. Nature 256,495-497 (1975). KOHLER, G. & MILLSTEIN, C. Derivation of specific antibody-producing tissue culture and tumor lines by cell fusion. European Journal of Immunology 6,5 1 l-5 19 ( 1976). LITTLEFIELD, J. W. Selection of hybrids from matings of Iibroblasts in vitro and their presumed recombinants. Science 145,709-710 (1964). LUGINBUHL, W. H. Immunologic investigation of preparations of human myoglobin. American Journal of Clinical Pathology 38,4871193 (1962). MARCHALONIS, J. J. An enzymatic method for the trace iodination of immunoglobulins and other proteins. Biochemical Journal 113,299-305 ( 1969). NYDICK, I., WROBLEWSKI, F. & LADWE, J. S. Evidence for increased serum glutamic oxalacetic transaminase (SGO-T) activity following graded myocardial infarcts in dogs. Circulation 12, 161-168 (1955). PERRIE, W. T. & PERRY, S. V. An electrophoretic study of the low-molecular-weight components of myosin. BiochemicalJournal 119,31-38 (1970). PRESSMAN, D. & KEIGHLEY, G. Zone of activity ofantibodies as determined by use of radioactive tracers; zone of activity of nephritoxic antikidney serum. Journal of Immunology 59, 141-146 (1948). ROBERTS, R., SOBEL, B. E. & PARKER, C. W. Radioimmunoassay for creatine kinase isoenzymes. Science 194, 855-857 (1976). ROSENTHAL, J. D., HAYASHI, K. & NOTKINS, A. L. Comparison of direct and indirect solid-phase microradioimmunoassays for the detection ofviral antigens and antiviral antibody. AHlied Microbiolog,v 25,567~573 (1973). (1973). SARKAR, S., SRETER, F. A. & GERGELY, J. Light chains ofmyosin from white, red, and cardiac muscles. Proceedings of the National Academy of Science, USA 68,946-950 ( 197 1) SHELL, W. E., KJEKSHUS, J. K. & SOBEL, B. E. Quantitative assessment ofthe extent ofmyocardial infarction in the conscious dog by means ofanalysis ofserial changes in serum creatine phosphokinase activity. Journal of Clinical Investigation 50, 2614-2625 (1971). SOBEL, B. E. & SHELL, W. E. Serum enzyme determinations in the diagnosis and assessment of myocardial infarction. Circulation 45,47 1-482 (1972). STONE, M. J., WILLERSON, J. T., GOMEZ-SANCHEZ, C. E. &WATERMAN, M. R. Radioimmunoassay ofmyoglobin in human serum: results in patients with acute myocardial infarction. Journnl of Clinical Investigation 56, 1334-1339 (1975). THORN, G. W., ADAMS, R. D., BRAUNWALD, E., ISSELBACHER, K. J. & PETERSDORF, R. G. Harri.son’s Principles of Znteml Medicine, 8th Edn, pp. 1975, 1990 and 1991. New York: McGraw-Hill (1977). TRAHERN, C. A., GERE, J. B., KRAUTH, G., II & BIGHAM, D. A. Clinical assessment ofserum myosin light chains in the diagnosis ofacute myocardial infarction. American Journal of Cardiolog)l41(4), 641-645 (1978). WIKMAN-COFFELT, J., ZELIS, R., FENNER, C. & MASON, D. T. Comparative purification ofmyocardial myosin and antigenic specificity of the two light chains. Pre@rative Biochemisty 3,439-449 (1973). WILLERSON, J. T., STONE, M. J., TING, R., MUKHERJEE, A., GOMEZ-SANCHEZ, C. E., LEWIS, P. & HERSH, L. B. Radioimmunoassay of creatine kinase-B isoenzyme in human sera: results in patients with acute myocardial infarction. Proceedings of the National Academy of Science, USA 74, 17 1 I-1 7 15 ( 197 7).