Journal of Clinical InvestigationVol. 42, No. 2, 1963

STUDIES OF ACUTE-PHASEPROTEIN. II. LOCALIZATION OFCx-REACTIVE PROTEIN IN HEARTIN INDUCED MYO-

CARDIAL INFARCTION IN RABBITS *

By IRVING KUSHNER,t LOUIS RAKITAt AND MELVIN H. KAPLAN§

(From the Department of Medicine, Metropolitan General Hospital and Western ReserveUniversity School of Medicine, Cleveland, Ohio)

(Submitted for publication September 28, 1962; accepted November 1, 1962)

Cx-reactive protein (CxRP), which appears inthe blood of rabbits during inflammation (1), isanalogous and immunologically related to C-reac-tive protein (CRP) of man (2). The site of ori-gin of these acute-phase proteins and the mecha-nism of their production are problems as yet un-resolved.

Previous work from this laboratory has de-scribed an immunofluorescent technique for thedetection of CxRP in inflamed rabbit tissue (3).After the induction of an inflammatory lesion inthe rabbit by the intramuscular injection of typhoidvaccine, CxRP could be found in necrotic myo-fibers within the inflammatory lesion, and not inany other rabbit organ or tissue studied. At notime was CxRP detected in the cells of the inflam-matory exudate. This same localization of CxRPto necrotic myofibers was observed in rabbitsmade markedly neutropenic by the injection of ni-trogen mustard. These observations rendered un-likely the derivation of acute-phase protein frominflammatory cells or from a product of the ac-tivity of such cells. Rather, these observationshave supported the concept that CxRP has itsorigin in the tissues undergoing inflammatory ornecrotic change.

The present work represents an extension ofthese findings to inflammatory and necrotic lesionsof myocardium. The clinical association of CRPin man with myocardial infarction and rheumaticmyocarditis prompted consideration that thisacute-phase reactant might be derived from in-jured myocardial myofibers. Accordingly, the

*Work performed under grants-in-aid from the Na-tional Heart Institute, U. S. Public Health Service(H-3726) and (H-6403), and the Cleveland Foundation.

t Research Fellow, Helen Hay Whitney Foundation.t Senior Research Fellow, U. S. Public Health Service.§ Established Investigator, American Heart Association.

immunohistochemical localization of CxRP wasinvestigated in rabbits in which infarction of themyocardium was produced by coronary artery oc-clusion. The results of these studies indicatedthat in such lesions, CxRP was localized to, andprobably had its origin in, cardiac myofibers un-dergoing necrotic changes.

MATERIALS AND METHODS

Preparation of antiserum. CxRP was prepared as apurified, delipidated, CxRP-Cx carbohydrate precipitateas previously described (3), and was solubilized in 1 percent NaCl containing 0.0001 M sodium dihydrogenethylene diamine tetraacetate (EDTA). This solubilizedprecipitate will henceforth be referred to as CxRP. Agoat antiserum to CxRP was prepared by an immuniza-tion procedure that included several courses of alum-precipitated CxRP given as previously described (3),followed by a single booster of 90 Acg of protein N ad-ministered subcutaneously and intramuscularly in Freund'sadjuvant. A highly reactive antiserum was obtained thatyielded a minimal reaction with normal rabbit serum inprecipitin tests in capillary tubes. After absorption ofantiserum with lyophilized normal rabbit serum, the anti-body content of the serum (CxRPA) as determined byquantitative precipitation was 910 lAg antibody N per ml.Agar gel diffusion studies employing this CxRPA re-vealed a single strong band appearing within 1 day anda second very weak band appearing after 8 to 10 dayswhen CxRPA reacted against acute-phase rabbit serumor against CxRP, the same pattern seen with previousantisera prepared without the use of Freund's adjuvant(3).

Inmmunofluorescentt techniques. Preparation of fluores-cent conjugates, absorption procedures, fluorescence mi-croscopy, fixation and staining techniques, and criteriafor immunohistochemical specificity have all been de-scribed previously (3).

For study of the histological appearance of sites re-vealing immunofluorescent staining, a double-stainingtechnique was used. After sites of staining specific forCxRP were photographed, the coverslips were floated offthe slide, the section was stained with hematoxylin andeosin, and the site of CxRP localization was identified.

286

Cx-REACTIVE PROTEIN IN MYOCARDIALINFARCTION

Production of myocardial lesions. Adult, New Zea-land, white, male rabbits weighing 3.3 to 4.3 kg wereanesthetized with 90 mg of intraperitoneal Nembutalland maintained with intravenous Pentothal 2 (1 per cent)as needed. After endotracheal intubation via a tracheos-tomy incision, the animals were maintained on artificialrespiration. A thoracotomy was performed through aleft lateral chest incision in the fourth or fifth inter-costal space. The anterior descending branch of the leftcoronary artery was identified. Direct epicardial elec-trocardiographic recordings were obtained with a cot-ton-tipped, wick electrode of copper wire from the wholeanterior and lateral wall of the epicardial surface throughthe intact pericardium, and were recorded on a 4-channelCambridge Simpliscribe electrocardiograph. A smallportion of the pericardium immediately over the arterywas excised and a suture was placed around the arterywith an atraumatic needle. The ligature was tightenedeither directly, completely occluding the artery, or overa probe to allow for quick release if ventricular fibrilla-tion occurred. If the latter did not occur, complete oc-clusion was effected. Repeat epicardial electrocardio-graphic leads were then recorded as before the tie. Thechest was closed and the animal permitted to come outof anesthesia. After varying intervals, up to 48 hoursafter ligation, the animals were reanesthetized and main-tained on artificial respiration while the thoracotomyincision was reopened and epicardial electrocardiographicrecordings were made to ascertain the site of infarction.In some instances, the animals died before a terminalelectrocardiogram could be obtained. After sacrifice,tissue was obtained from the heart both distal and proxi-mal to the ligation, from intercostal muscle at the thora-cotomy site, and from lung, liver, spleen, kidney, andmesenteric lymph node. The tissue blocks were quick-frozen in a mixture of dry ice and alcohol and stored at-25° C. Tissues were sectioned in the cryostat at athickness of 4 ,u, and the slides were air dried before fixa-tion and staining.

Electrocardiographic interpretation. In the analysis ofthe electrocardiograms, development of Q waves or en-largement of pre-existing Q waves was considered to beevidence of infarction and necrosis. Elevation of RS-Tsegments and inversion of T waves were considered evi-dence of injury and ischemia, respectively.

Determination of serum CxRP concentration. Bloodsamples were obtained from the marginal ear vein priorto any surgical incision and from the inferior vena cavaat the time of sacrifice.

After collection, blood samples were permitted to standovernight at 40 C, and the serum was separated by cen-trifugation. Estimation of serum CxRP concentrationwas performed by the capillary precipitin technique andwas graded by the height of precipitin in the column inmillimeters (4).

1 Pentobarbital sodium, Abbott Laboratories, NorthChicago, Ill.

2Thiopental sodium, Abbott Laboratories, North Chi-cago, Ill.

RESULTS

Fourteen rabbits were subjected to coronary ar-tery ligation, and all showed electrocardiographicevidence of myocardial ischemia and injury im-mediately after ligation (Figure 1). At intervalsof 4 to 48 hours after the ligation, prior to sacri-fice, terminal electrocardiograms revealed electri-cal evidence of necrosis in localized portions ofmyocardium distal to the ligature in all animalssurviving this procedure.

Histologic studies of heart tissue taken from thezone of necrosis, demarcated by the electrode re-cordings, were stained with hematoxylin and eosin.These sections revealed the classical changes as-sociated with myocardial infarction (5-7): loss ofcross striations, hyalinization, increased eosino-philic staining of involved myofibers, and loss ofnuclei (Figure 29). Homogenization and other evi-dences of disturbance of fibrillar structure wereobserved in phosphotungstic acid- and hematoxy-lin-stained preparations. Vacuolization of myo-fibers was observed occasionally. The infarctedarea generally showed variable infiltration withleukocytes, with interstitial edema.

CxRP was first detected in the serum as earlyas 41 to 5 hours after arterial ligation (Table I).It was found in the blood in progressively increas-ing amounts in the remaining interval of the 48-hour observation period after ligation. In im-munohistochemical studies, CxRP was first de-tected in the heart in the zone distal to the liga-tion at 4j to 5 hours after ligation and was demon-strable only in this zone of infarcted heart tissuefor the remainder of the 48 hours of the study. Itcould not be demonstrated in heart at 2 and 4 hoursafter ligation. Thus, the localization in tissue andappearance of CxRP in the blood showed a simi-lar time-course.

In the zone of affected myocardium, CxRP waspresent at the outer edge of myofibers and in fine,round or linear inclusions within the substance ofnecrotic myofibers (Figures 3-8). Microprecipi-tate was often present in these sites (Figure 8) aswell as occasionally in the interstitium betweennecrotic myofibers. The presence of CxRP inmyofibers was not apparently related to the pres-ence of inflammatory cells. An identical patternof specific staining was observed in single, scat-tered myofibers at the periphery of the infarcted

287

IRVING KUSHNER,LOUIS RAKITA, AND MELVIN H. KAPLAN

A. CONTROL

~fI ~ff-1I-I

B. 10 MINS.F' 1i dtL ±LI 4

T -i- iij K!1 L S111,

C. 4 HOURS

FIG. 1. RABBIT 1065, EPICARDIAL LEAD. A. CONTROL,BEFORE LIGATION. B. 10 MINUTES AFTER CORONARYAR-

TERY LIGATION. Marked ST segment elevation indicativeof injury. C. 4 HOURSAFTER CORONARYARTERYLIGATION.

QS indicative of loss of electrical potentials, consideredas evidence of necrosis.

C-)

V4

Cle

El)

-_

m~..;¢ s

c-)

00

.0

0a

l;0

c-)

FIG. 2. RABBIT 738, SACRIFICED 48 HOURSAFTER LIGA-TION OF CORONARYARTERY. SITE OF INFARCTED MYO-CARDIUM. Necrotic myofibers show eosinophilia, hyalini-zation, and a homogeneous appearance. Frozen section.Hematoxylin and eosin stain. (X 312)

U

0

-I0

$3

Cd0

Cd

0

0

E.E

cn

E "C1. a

z ..

c- =) Q

V

o0

,o co2t

N- d

0d o

Cd

.u0

E

0

0.

w.. W0

CU._&jC'

COb

0 o0 c o0 oo

o0

0 0

00 000

0 00 0o

00 0 00 0000

00 0 0 0 0 00

00 0 c o0 oo

o 0 + + 0+ +

* *o++*+++*+++++

CddCd Cd

Cd+ + +000 &0 >*1- 4C-

C- C-

0000 00000 -

eq U) Ul) 00 00~ )9t, tI 00

288

.v

C-

Cd

E0

.

0

~0C-

'0CUE

*

Cx-REACTIVE PROTEIN IN MYOCARDIALINFARCTION

FIG. 3. RABBIT 738, SACRIFICED 48 HOURS AFTER LIGATION. SECTION OF INFARCTED

MYOCARDIUMSTAINED WITH CxRPA CONJUGATE. CxRP is localized to edges of myo-fibers in focal droplets and granules, and within substance of myofibers. (X312)

FIG. 4. SAME SECTION AS FIGURE 3. AFTER IMMUNOFLUORESCENTSTAINING, COVERSLIP

WAS FLOATED OFF AND TISSUE STAINED WITH HEMATOXYLIN AND EOSIN. Necrotic myo-

fibers, eosinophilic and hyaline in appearance, with loss of nuclei. (X312)

area remote from infiltrating cells (Figure 7).The extent and intensity of the staining reactionfor CxRP in the involved tissue increased pro-gressively during the 48 hours of observation.

No CxRP could be demonstrated in normalmyocardial tissue proximal to the ligation, nor inthe normal-appearing myofibers observed distal tothe ligation, even in the presence of infiltrationwith inflammatory cells. In other organs studied,including liver, kidney, spleen, lung, and mesen-teric lymph node, CxRP was observed only invascular lumens and interstitial spaces. CxRPwas also demonstrated in the intercostal skeletalmuscle adjacent to the line of thoracotomy in-cision. The localization of CxRP in this tissuetraumatized by surgery was limited to the outeredges of and inclusions within necrotic myofibers.This distribution was similar to that observedpreviously in inflammatory lesions of muscle in-duced by injection of typhoid vaccine (3).

DISCUSSION

The localization of CxRP in inflammatory le-sions of skeletal muscle produced by an exogenousinflammatory agent, typhoid vaccine, had sug-gested previously the probable origin of CxRP inthis lesion from necrotic muscle fibers. In thepresent work, coronary artery ligation was selectedas a method of producing myocardial injury. Inthe heart, specific staining of necrotic myocardialcells in the infarcted area was limited to the outeredge, possibly including sarcolemmal or subsar-colemmal segments of the cells, and to inclusionsand vacuoles within the sarcoplasm of these cells.This association of CxRP with necrotic but notwith normal myofibers is consistent with the hy-pothesis that CxRP is produced as a result of in-flammatory or necrotic changes in cardiac myo-fibers. The alternative possibility, that the locali-zation described represents secondary deposition of

289

IRVING KUSHNER, LOUIS RAKITA, AND MELVIN H. KAPLAN

FIG. 5. RABBIT 738, SACRIFICED 48 HOURSAFTER CORONARYARTERY LIGATION. SECTIONOF INFARCTED MYOCARDIUM. CxRP is distributed along the edge and within necrotic fibers.Myofibers with morphologic evidence of necrosis also exhibit a faint diffuse autofluorescence.(X312)

FIG. 6. SAME SECTION AS FIGURE 5, STAINED WITH HEMATOXYLIN AND EOSIN AFTERCOVERSLIP FLOATED OFF. Myofibers containing CxRP show necrotic changes. (X 312)

290

Cx-REACTIVE PROTEIN IN MYOCARDIALINFARCTION

CxRP from the blood into myofibers with alteredpermeability, cannot be excluded, although it isconsidered less likely in view of the highly char-acteristic cytologic distribution of CxRP in seg-ments of myofibers at the periphery of the cell andin inclusions and vacuoles. Were this localiza-tion due to penetration of necrotic myofibers byCxRP from the blood or tissue fluids, a diffuseor random distribution in the myofiber might havebeen expected rather than this delicate, consistentpattern. Further, possibility of the origin ofCxRP from other tissues was opposed by the fail-ure of CxRPto be detected in noninfarcted portionsof myocardium, or in cells of other organs, includ-ing liver, kidney, spleen, lung, and lymph nodes.The observation of CxRP in necrotic skeletal mus-cle at the site of thoracotomy was consistent withorigin of CxRP from this site of tissue injury aswell.

The inflammatory stimulus in the present workwas endogenous in that it consisted of the tissueresponse to a local ischemic state. The associa-tion of CxRP with necrotic cardiac myofibers inthis lesion paralleled observations of CxRP inskeletal muscle necrosis induced with an exoge-nous agent, typhoid vaccine. Thus, it does notseem necessary to ascribe a special function toexogenous agents in eliciting CxRP, except thatnecrosis be induced. Other structures of heartand skeletal muscle, including connective tissueand blood vessel walls, did not show localizationof CxRP. It would be of interest to determinethe distribution of CxRP in other types of inflam-mation, such as that associated with the hyper-sensitivity state, in which blood vessels and inter-stitial tissue are perhaps more directly affected.

Approximately 5 hours elapsed between the in-flammatory stimulus and the appearance of CxRPin the blood in some of the rabbits studied. Thisis more rapid than has been described previously(8), and is three hours earlier than was seen inprevious studies after the intramuscular injectionof typhoid vaccine. The report by Kroop and

Shachman (9) that the strong stimulus of majorsurgery evokes a more rapid and more markedCRPresponse than that seen after minor surgery,might suggest that both the speed and magnitudeof the acute-phase protein response are propor-tional to the severity of the inflammatory stimulus.Other variables to be considered, however, are thekind of inflammatory stimulus and the tissue af-fected. In the present work, the combined inflam-matory stimulation induced by surgical thoraco-tomy and myocardial infarction may possibly ac-count for the particularly rapid appearance ofCxRP in the rabbits studied.

SUMMARY

The histologic localization of Cx-reactive pro-tein (CxRP) was studied in inflammatory lesionsof heart tissue induced by infarction secondary tocoronary artery ligation. CxRP was found withinand around necrotic myofibers, beginning approxi-mately 5 hours after ligation, and could not befound in normal myocardial sites or in other or-gans studied, except for the intercostal skeletalmuscle site traumatized during thoracotomy. Theseobservations are consistent with the hypothesisthat acute-phase protein is produced at the in-flammatory site as a result of inflammatory or ne-crotic tissue change. These experimental resultsdirect attention to the possible origin of C-reactiveprotein (CRP) from injured cardiac myofibersin infarctive and rheumatic inflammation of hearttissue in man.

ACKNOWLEDGMENT

The technical assistance of Catherine Rezou, MomoyeKansaki, and Hayes Brooks is acknowledged.

REFERENCES

1. Anderson, H. C., and McCarty, M. The occurrencein the rabbit of an acute phase protein analogousto human C-reactive protein. J. exp. Med. 1951,93, 25.

FIG. 7. RABBIT 738, 48 HOURSAFTER LIGATION. CxRP is detected in an isolated myo-fiber distant from the mass of necrotic myofibers elsewhere in the section. CxRP is dis-tributed in focal sites at edge and in inclusions within myofiber. (X250)

'FIG. 8. RABBIT 1082, SACRIFICED 8 HOURSAFTER LIGATION. Section of infarcted myo-cardium. CxRP is demonstrated as a fine microprecipitate on the edges of myofibers show-ing early necrosis. (X250)

291

IRVING KUSHNER, LOUIS RAKITA, AND MELVIN H. KAPLAN

2. Gotschlich, E., and Stetson, C. A., Jr. Immunologiccross-reactions among mammalian acute phaseproteins. J. exp. Med. 1960, 111, 441.

3. Kushner, I., and Kaplan, M. H. Studies of acutephase protein. I. An immunohistochemical methodfor the localization of Cx-reactive protein in rab-bits. Association with necrosis in local inflam-matory lesions. J. exp. Med. 1961, 114, 961.

4. Anderson, H. C., and McCarty, M. Determination ofC-reactive protein in the blood as a measure ofthe activity of the disease process in acute rheu-matic fever. Amer. J. Med. 1950, 8, 445.

5. Karsner, H. T., and Dwyer, J. E. Studies in infarc-tion. IV. Experimental bland infarction of the

myocardium, myocardial regeneration and cicatri-zation. J. med. Res. 1916, 34, 21.

6. Mallory, G. K., White, P. D., and Salcedo-Salgar, J.The speed of healing of myocardial infarction.Amer. Heart J. 1939, 18, 647.

7. Saphir, 0. A Text on Systemic Pathology. NewYork, Grune and Stratton, 1958, p. 87.

8. Hedlund, P. Clinical and experimental studies on

C-reactive protein (acute phase protein). Actamed. scand. 1961, suppl. 361.

9. Kroop, I. G., and Shachman, N. H. The effect ofsurgical trauma and the rheumatic state on C-re-active protein formation. Clin. Res. Pror. 1955, 3,119.

292