367

Atrophy Reversal and Cardiocyte Redifferentiation inReloaded Cat Myocardium

Ed W. Thompson, Thomas A. Marino, Cornelius E. Uboh, Robert L. Kent, andGeorge Cooper, FV

From the Departments of Anatomy, Pharmacology, and Physiology, and the Cardiology Section of the Department of Medicine, TempleUniversity School of Medicine, Philadelphia, Pennsylvania

SUMMARY. We have recently described rapid cardiac atrophy in response to decreased load. Thepresent study was designed to determine whether this atrophy is solely a degenerative responseof damaged myocardium or is, instead, an adaptive response of viable myocardium. A discreteportion of cat myocardium was unloaded by severing the chordae tendinae of a single rightventricular papillary muscle. One week later, the muscle was reloaded by attachment of its apexto the ventricular free wall. This allowed the load to be removed and restored without alteringthe blood supply, innervation, or frequency of contraction of the tissue. In myocardium unloadedfor 1 week, the cardiocyte cross-sectional area and the volume densities of mitochondria andmyofibrils decreased significantly. Large areas of cytoplasm were devoid of organelles, and thefew remaining myofilaments were oriented in a variety of directions rather than longitudinallywithin the cell. Upon reloading for 1 week, the cardiocyte cross-sectional area, volume density ofmitochondria, and ultrastructural organization all returned to normal. The volume density of themyofibrils increased toward control, and they reoriented with respect to the long axis of thecardiocyte. The contractile function of the papillary muscles, which was depressed as early as 1day after unloading and almost absent at times later than 3 days after unloading, returned tonormal after 2 weeks of reloading. This study demonstrates that adult mammalian myocardiumresponds to unloading with a marked loss of cellular differentiation, organization, and functionwhich is fully reversible with reloading. This plasticity in response to load may well be the basicmechanism responsible for the development and maintenance of normal cardiac structure andfunction. (CircRes 54: 367-377, 1984)

THE hemodynamic loading environment of adultmammalian myocardium is an important determi-nant of its biological properties; increases or de-creases in the demands placed upon the heart areaccompanied by changes in cardiac size, composi-tion, and function. Within certain limits of load andtime, these changes represent the normal physiolog-ical adaptations required to maintain optimal pumpfunction. However, when the load falls outside theselimits for longer times, progressive structural andfunctional changes occur.

The factors determining a physiological vs. path-ological myocardial response to increased load havebeen the subject of very extensive investigations,recently reviewed in detail by a number of investi-gators (Alpert, 1983). However, cardiac hypertrophyappears to be, in simplest terms, an unusual exten-sion of the normal postnatal response of the heartto increasing circulatory demands; adult cardiocytesenlarge in response to hemodynamic overloads bythe addition of structurally normal contractile unitsin series and in parallel with those already in place(Bishop, 1971). Our recent initial investigations (To-manek and Cooper, 1981; Cooper and Tomanek,1982) of the other end of the potential spectrum ofmyocardial stress and strain have shown that un-loaded adult myocardium undergoes profound andrapid atrophy. The most striking features of thisarrophic myocardium are the loss of ultrastructural

differentiation and the associated loss of effectivecontractile function.

This finding would be essentially a biological cur-iosity if this arrophic response were found to repre-sent fixed degeneration. If structural and functionalatrophy were, instead, found to be reversible with arestoration of load, it would provide a tool by whichthe dynamic regulation of adult myocardial structureand function by changing load conditions could bedefined. In addition, it would provide insight intothe primary importance of load, first, to the reali-zation of normal cardiac properties during devel-opment, and, second, to the maintenance of theseproperties in adult life.

Therefore, in the present investigation we reim-posed a load upon a previously unloaded myocardialsegment to determine whether atrophied myocar-dium undergoes fixed degeneration or, instead, boththe atrophy and the associated abnormalities couldbe reversed; that is, whether unloaded and atrophicmyocardium retains structural and functional re-sponsiveness to a restored load.

Methods

Experimental Model Preparation

Right ventricular papillary muscles of adult cats wereused for these studies. This cylindrical myocardial segmenthas a uniform orientation of the longitudinal cardiocytes

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

368 Circulation Research/ Vol. 54, No. 4, April 1984

in parallel with the tissue long axis. In addition, thispreparation can be unloaded and reloaded without alter-ing the blood supply, innervation, or frequency of con-traction (Cooper and Tomanek, 1982). Adjacent, intactpapillary muscles from the same right ventricles were usedas paired controls. Thus, both muscles were exposed toidentical conditions, with the single exception of load,throughout the experimental period. The right ventricularfree wall weight for each animal was compared with thebody weight and the left ventricular weight to verify thatright ventricular hypertrophy was not present. Each catalso was examined for pleural effusion, ascites, and he-patic enlargement; the absence of these conditions wasused as evidence that the animal was not in right ventric-ular failure.

Surgical Procedures

Adult cats (2-3 kg) were fully anesthetized with keta-mine hydrochloride (25 mg/kg, im) and then paralyzedwith succinylcholine (2 mg/kg, iv) prior to tracheal intu-bation and mechanical ventilation. A right thoracotomywas performed between the 4th and 5th ribs under sterileconditions. The pericardium was opened widely, and elas-tic bands were placed posterior to the azygous vein andthe superior and inferior venae cavae. A 1-cm-long portionof the right ventricular free wall was clamped, avoidingmajor epicardial vessels, and this excluded portion wasthen incised. After producing temporary occlusion of ve-nous return to the heart by tightening the three elasticbands, we removed the clamp to open the ventricularcavity. A single thin, cylindrical papillary muscle wasidentified, and its chordae tendinae were severed com-pletely. The ventricular incision then was quickly re-clamped while the ventricle was superfused with warmnormal saline, and venous inflow to the heart waspromptly restored. In all cases, the flow of venous bloodto the heart was stopped for less than 1 minute. Afterclosure of the ventriculotomy and thoracotomy, the ani-mal was allowed to recover.

If a mechanical load was to be restored to the papillarymuscle, the animal was again anesthetized 1 week afterpapillary muscle unloading and placed on artificial venti-lation, as described above. The original thoracotomy wasreopened, the ventricle was clamped, and the ventricularincision was reopened. After venous inflow occlusion andremoval of the clamp as before, the papillary muscle wassecured to the right ventricular free wall near the tricuspidannulus by means of a thin suture placed through thejunction of the chordae with the apex of the muscle. Themuscle was reloaded at a length at which it was straight,but not taut, during diastole. The ventriculotomy then wasreclamped, and flow to the heart was restored. Again,flow of venous blood to the heart was interrupted for lessthan 1 minute. After closure of the ventricular and thoracicincisions, the cat was allowed to recover spontaneously.

Anatomical Evaluation

Tissue Preparation

Samples were taken for analysis of unloading at either1 or 2 weeks after that procedure. For analysis of reload-ing, 1 week of unloading was followed by either 1 or 2weeks of reloading prior to sampling. One- and 2-weekreloading data were combined for analysis, since therewas no significant difference between these two groups.For each of the terminal procedures, the cat was anesthe-tized and artificially ventilated as described above, and

the chest was opened through a midstemal incision. Thedescending aorta was cannulated retrograde from a gravityperusion apparatus. After the injection of sodium heparin(1,000 U, iv), the heart was arrested in diastole and clearedof blood by perfusion for 3 minutes at 120 mm Hg pressurewith oxygenated Locke's solution containing 2% procaine.This was followed by perfusion for 10 minutes withisotonic 1.25% glutaraldehyde in sodium cacodylatebuffer. The heart was removed, and separate weights weredetermined for the right ventricular free wall and the leftventricle plus septum. The unloaded or reloaded papillarymuscle and an adjacent control muscle in each case thenwere excised and placed in cold fixative for 3-8 hours.These specimens were postfixed in 1% osmium tetroxidefor 60 minutes, dehydrated through a series of gradedalcohols, and embedded in vinyl cydohexene dioxide resin(Spurr, 1969). Three types of sections were prepared fromeach specimen: 1-fim cross-sections from the mid-portionof each muscle were cut and stained with toluidine blue(Trump et al., 1961) for the light microscopic determina-tion of cardiocyte cross-sectional area; l-/im cross- andlongitudinal sections were cut and used for determiningtissue composition; cross- and longitudinal 60- to 90-nmsections were cut and stained with uranyl acetate (Watson,1958) and lead citrate (Reynolds, 1963) for ultrastructuralstudy.

Tissue Examination

Cardiocyte cross-sectional area was determined at amagnification of 1600X for both control and experimentalmuscles by planimetry with a digitizer interfaced to asmall computer. To ensure that measurements were takennear the midpoint of each cell, we included only profilescontaining a nucleus. Fifty cells from each of two blockscut in cross-sections were counted for each muscle, andthese values were averaged. We determined the relativevolume densities of the myocardial tissue components(muscle, endothelium, blood vessel lumen, and connectivetissue) at the same magnification by stereological analysis(Elias et al., 1971; Page et al., 1971), using a 609-pointgrid projected onto the image of the tissue. The volumedensity of each component of the myocardium was deter-mined by the following relationship: Vx/V, = P»/P,, inwhich *PX" represents the number of intersection pointsfalling on the component, and "P," represents the totalnumber of points over the tissue. To ensure representativesampling, two blocks each of transversely and longitudi-nally sectioned tissue from each animal were examined,and quantitative measurements were made on three ran-dom 5,469 jim2 areas from each block. Thus, a total of65,625 fim2 of tissue was randomly sampled for eachpapillary muscle. From these data, mean values werecalculated for each muscle.

Myocardial ultrastructure was characterized by exam-ining thin sections in a transmission electron microscope.Random areas of each grid were photographed andprinted at a final magnification of 10,000x. The relativecardiocyte volume densities of mitochondria and myo-fibrils, as well as the tissue volume density of extracellularcollagen, were determined as described above, using a165-point grid printed onto a thin transparent sheet andplaced over the electron micrographs. The ratio of theexternal sarcolemmal surface area of the cardiocyte to itsvolume was calculated by the following formula: Area/volume = (ir/2)C/a-Pc in which "C represents one-halfof the total number of intersections of the external surfacemembrane with vertical or horizontal lines of the grid, 'a '

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

Thompson et a/./Myocardial Atrophy Reversal

represents the length of one side of a small square on thegrid divided by the final magnification of the photographicprint, and *PC' represents the total number of points overthe cell (Page et al., 1971). Two blocks each of transverselyand longitudinally sectioned tissue were used for eachmuscle, and quantitative determinations were made onfive random photographs from each block. An area of 143nm2 was analyzed from each photograph. A total of 2,856/im2 of tissue was randomly sampled for each papillarymuscle, and mean values for each muscle then werecalculated.

Functional EvaluationSeven days after the unloading operation, the reloading

operation described above was performed. Fourteen daysafter this second operation, the cats were anesthetizedwith sodium pentobarbital (25 mg/kg, ip), and rapid car-diectomy was performed. A reloaded or a control musclewas excised and mounted in a myograph (Cooper et al.,1981). The chordal end of each muscle was fixed to thelever of a photoelectric displacement transducer mountedabove the muscle by a tie at the chorda-muscle junction.The ventricular end of the muscle was enclosed rigidly ina sharp clip sintered to a metal rod; the other end of thisrod was screwed directly onto a semiconductor straingauge. Tension generated by the muscle was measuredwith very little stray compliance (<0.7 /im/mN) over therange of force studied, and the enclosed clip producedonly discrete end-segment damage of similar extent incontrol or reloaded muscles. Details of the transducersystem and associated equipment are reported elsewhere(Cooper, 1976).

After the muscles had been mounted in the myograph,they were superfused at 29°C by a solution of the follow-ing composition (mM): CaCl^ 2.5; KC1, 4.7; MgSO4, 1.2;KH2PO4, 1.1; NaHCO3, 24.0; Na acetate, 20.0; NaCl, 98.0;and glucose, 10.0 with 10 units of zinc insulin added perliter. This solution was equilibrated with 95% O2-5% CO2,with a resultant pH of 7.4, and circulated past the musclefrom a 1-liter reservoir. Each muscle was preloaded lightlyand stimulated at 0.2 Hz until a stable mechanical re-sponse was obtained. Field stimuli 5-10% above thresholdof alternating polarity with no DC offset between stimuli,were employed to minimize electrolytic contamination.

Force-velocity and force-shortening curves were con-structed during 0.5-Hz contractions of each muscle, asfollows: first, a preload of about 5 mN/mm2 muscle cross-sectional area was used to define the first point on thesecurves; successive afterload increments of 5 mN/mm2

were then added to define further points on these curvesuntil a maximum isotonic force was reached. The maxi-mum velocity and extent of shortening at each load weremeasured. After this, isometric length-tension curves wereconstructed by beginning at a relatively short musclelength at which active tension generation was first noted,and then proceeding in 0.2-mm increments until thelength producing maximum isometric tension, L™^ wasexceeded slightly. Further details of these techniques havebeen described before (Cooper et al., 1973).

After each experiment, muscle length was measured bya micrometer with a known preload attached to the mus-cle. This length, along with the passive tension portion ofthe length-tension relationship, allowed calculation ofmuscle length at L™,. Muscle cross-sectional area wascalculated from this length and from the dry weightobtained as the constant weight reached at 100°C. A wet-to-dry weight ratio of 4 and a specific gravity of 1 were

369

assumed. That is, area (mm2) = dry weight X 4 (mm3)/length (mm). Results were normalized in terms of musclelength at Lma, and cross-sectional area.

Statistical AnalysisTo ensure that adequate sampling was used for the

anatomical experiments (Eisenberg and Cohen, 1983), anested sampling analysis of variance was done (Shay,1975) on normal cat right ventricular papillary muscles.To check our sampling size for the parameters obtainedwith light microscopy, the volume density of myocyteswas considered. It was determined that the minimumadequate sample was two regions per block, two blocksper papillary muscle, and two animals per group. In thisstudy, we used three regions per block, four blocks permuscle, and four animals per group. For the data to beobtained from electron micrographs, the same analysiswas performed on the relatively invariant volume densityof myofibrils from the cardiocytes of normal papillarymuscles, and it was determined that the minimum ade-quate sample consisted of one picture per block, threeblocks per animal, and four animals per group. For therelatively more variable external saicolemmal surfacearea:volume ratio, this analysis showed that the minimumadequate sample was four pictures per block, three blocksper animal, and four animals per group. In this study, weused five pictures per block, four blocks per animal, andfour animals per group. Therefore, the sampling tech-niques we used met or exceeded the sampling require-ments.

For the results of both the anatomical and the functionalstudies, each value is expressed as mean ± SE. Compari-sons of ventricular weights among the treatment groupswere done by one-way analysis of variance. The contrac-tile function and the tissue or cellular composition ofexperimental papillary muscles were compared with thoseof control muscles by a two-tailed paired Student's f-test.Significant differences were said to exist when P was lessthan 0.05.

Results

Characteristics of the Experimental ModelNeither ascites, pleural fluid, nor hepatic engorge-

ment was found in any cat from any of the experi-mental groups. The ratio of right ventricular weightto left ventricular weight was 0.34 ± 0.04 gAg forthe 1-week unloaded animals, 0.34 ± 0.04 g/kg forthe 2-week unloaded animals, and 0.42 ± 0.08 gAgfor the reloaded animals. These values were notsignificantly different by one-way analysis of vari-ance. It has been established previously (Cooper andTomanek, 1982) that selective transection of thechordae tendinae of a single papillary muscle doesnot result either in substantial tricuspid regurgitationor in subsequent volume overloading and hypertro-phy of the right ventricle; the mean right atrialpressure and the ratios of right ventricular weightto body weight and of right ventricular weight toleft ventricular weight do not change during thisunloading procedure, and the cross-sectional area ofcardiocytes in control papillary muscles is the sameas that in muscles from sham-operated cats. Thus,neither right ventricular hypertrophy nor failure

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

370 Circulation Research/Vol. 54, No. 4, April 1984

results from these procedures, and any changes canbe attributed solely to changes in the loading con-ditions of the muscle.

Cardiac Anatomy

Gross and Histological Appearance

Atrophy of the unloaded right ventricular papil-lary muscle was obvious by gross examination at1 week after unloading, and was much more pro-nounced at 2 weeks; unloaded muscles were boththinner and shorter than adjacent intact papillarymuscles. In all cats studied, the distal end of theunloaded muscle remained unattached to surround-ing structures, eliminating the possibility of onlypartial unloading during the experimental period.Upon light microscopic examination, cardiocytesstained less intensely than normal after the firstweek of unloading, but otherwise appeared normal.Necrosis of myocytes in response to unloading wasnot observed. The further cellular atrophy that oc-curs at later times after unloading has been describedelsewhere (Tomanek and Cooper, 1981; Cooper andTomanek, 1982).

When the papillary muscle was reloaded for 1 or2 weeks after 1 week of unloading, gross atrophywas reversed, and the diameter and length of thereloaded muscle returned to normal. The stainingintensity of cardiocytes in this tissue returned to thatof the normal myocardium, and the reloaded tissuecould not be distinguished from control tissue bylight microscopic examination.

Cardiocyte Cross-Sectional Area

Table 1 illustrates the decrease in cross-sectionalarea of myocardial cells in unloaded tissue and theirreturn to normal size 1 week after the load wasrestored. The atrophy which occurred during thefirst week of unloading was much more pronouncedafter 2 weeks. When the myocardium was reloadedfor 1 or 2 weeks after 1 week of unloading, cardi-ocyte cross-sectional area returned to the controlvalue. Together with the reversal of gross atrophyof the muscle, these changes in cardiocyte cross-sectional area demonstrated that the myocardiumresponds to the return of mechanical load by anincrease in volume of the myocardial cells, restoringthem to normal size.

Tissue Composition

The relative volume densities of the tissue com-ponents of unloaded and reloaded myocardium arepresented in Figure 1. The decrease in size of themyocardial cells during the first week of atrophywas accompanied by proportional changes in theother components of the myocardium, since thevolume densities of blood vessel lumen, endothe-lium, and connective tissue did not change. As atro-phy progressed, a modest relative increase in con-nective tissue was noted, but the other elements ofthe myocardium remained constant.

Reversal of atrophy also involved proportionalchanges in the muscular, vascular, and connectivetissue components of the myocardium. Figure 1

TABLE 1

Myocardial Morphometrics

Control Experimental

Cardiocyte cross-sectional area (JM\2)Mitochondnal volume density (%)Myofibrillar volume density (%)Mitochondria/myofibril ratioSurface area/volume ratioCollagen volume density (%)

Cardiocyte cross-sectional area

Cardiocyte cross-sectional area (jtmMitochondnal volume density (%)Myofibrillar volume density (%)Mitochondria/myofibril ratioSurface area/volume ratioCollagen volume density (%)

1 week of unloading

166.5 ± 14.6 150.6 ± 18.0*20.4 ±0.9 15.1 ±1.0*46.6 ± 2.7 27.4 ± 3.1*0.45 ± 0.05 0.60 ± 0.07*0.24 ± 0.03 0.24 ± 0.050.9 ±0.3 1.0 ±0.3

2 weeks of unloading

191.5 ±3.3 118.6 ±8.4*

Reloading following unloading

175.7 ±20.1 160.2 ±9.416.1 ±1.8 13.4 ±0.846.2 ±1.5 43.4 ±1.7*

0.36 ± 0 05 0.32 ± 0.020.20 ±0.03 0.21 ±0.050.8 ±0.3 0.9 ±0.4

Results for the two groups of papillary muscles are expressed as mean ± SE. Mitochondrial andmyofibrillar volume density are expressed as percent of cardiocyte volume; collagen volume density isexpressed as percent of myocardial tissue volume. There were five cats in the 1-week unloaded groupand four cats in the 2-week unloaded group. The reloaded group consisted of cats unloaded for 1 weekfollowed by 1 or 2 weeks of reloading; in this group, five cats were examined for cross-sectional area,and four cats were analyzed for the other parameters measured.

' Significant difference between the control and experimental muscles.

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

Thompson et a(./Myocardial Atrophy Reversal 371

IOO-,

1 8<Hoc

IWK UNLOADED

m CONTROLE 2 UNLOADED

0-3

UIDOTHEUUM COmtTIMUC

2WK UNLOADED

E 3 CONTROLE 2 UNLOADED

iRELOADED

CONTROLRELOADED

n-4

CNDOTHCUUH C0MN.T1I9UC UKEN ENDOTHELIUM CONH.TI33UE

FIGURE 1. Tissue composition of myocardium in response to removal and restoration of load. Volume densities of cardiocytes, blood vessel lumen,endothelium, and connective tissue in myocardium unloaded for 1 or 2 weeks or reloaded for 1 week are compared with values from adjacentcontrol muscles. The contributions of muscular and vascular elements remain unchanged in all groups. Connective tissue increased after 2 weeksof unloading, but was unchanged in the 1-week-unloaded and 1-week-reloaded tissues. The volume density of each component is represented asa percentage of the total composition of the myocardium and is expressed as mean ± SE. * P < 0.05 by Student's t-test.

shows that volume densities of these elements werethe same as the values that we have shown beforein control myocardium (Marino et al., 1983a, b) andin myocardium in which cardiocytes were enlargingin response to reloading. Thus, during both the earlystages of atrophy, and during its reversal, changesin the various components of the myocardium werecoordinated to maintain a constant compositionwithin the myocardial tissue.

Myocardial infrastructure

Qualitative changes within the myocardial cells 1week after either the removal or the reimposition ofload are shown in transverse section in Figure 2 andin longitudinal section in Figure 3. In the unloadedtissue shown in Figures 2B and 3B, cardiocytes typ-ically showed a marked loss of myofibrils and manyareas of cytoplasm free of organelles. Persistingbundles of contractile filaments were seen primarilyat the periphery of the cell and were disorganized,lying at many angles to the long axis of the cell. Thedensity of Z-bands was much less uniform fromsarcomere to sarcomere, and the typical bandingpattern was lost. Mitochondria continued to lie nearcontractile elements but appeared less numerousthan normal. The sarcolemma, intercalated discs,and nuclei of unloaded cardiocytes appeared un-changed, as did the endothelial cells and the inter-stitial components of the tissue.

In contrast, 1 week of reloading, following 1 weekof unloading, resulted in substantially complete re-versal of the degenerative changes of atrophy. Fig-ures 2C and 3C show that, once again, myofibrilsand mitochondria fill the cytoplasm, and that theextensive organelle-free regions characteristic ofatrophic tissue are no longer present. As demonstra-ted by the longitudinally sectioned tissue shown inFigure 3C, the contractile elements are realignedwith the long axis of the cell, the typical bandingpattern is present, and adjacent sarcomeres are inregister. Cardiocytes showing only a partial response

to the change in loading conditions were seen oc-casionally, but none showed the loss and disorgan-ization of myofibrils which was characteristic ofunloaded myocardium. Myofibrils were uniformlyaligned with the long axis of the cell and were thepredominant constituent of the cytoplasm.

The changes in cardiocyte ultrastructure observedin the early stages of myocardial unloading andreloading were quantified by stereological analysisand are presented in Table 1. One week of unloadingresulted in a 26% decrease in the volume density ofmitochondria and a 41% decrease in the volumedensity of myofibrils. Because the mitochondrialresponse lagged behind that of the myofibrils, thesetwo changes produced a 31% increase in the ratioof these two components. No significant change wasobserved on the ultrastructural level in the volumedensity of tissue collagen during the first week ofunloading. This agrees with the light microscopicobservation that no change occurred in the connec-tive tissue component of the myocardium duringthis time. The ratio of the external sarcolemmalsurface area of the cardiocyte to its volume was alsomaintained as the cell underwent atrophy.

Table 1 also shows that when atrophy was re-versed by reloading, the cardiocyte volume densitiesof both mitochondria and myofibrils approachedcontrol values. This process appeared to be completeafter 1 week of reloading, since further changes werenot seen with two weeks of reloading; thus the datain the bottom third of Table 1 are pooled as indicatedthere. The cellular fraction represented by mito-chondria in the cardiocytes from reloaded musclewas no longer significantly different from that inthe intact muscle. Whereas the volume density ofmyofibrils was still significantly lower in cardiocytesfrom reloaded muscle than in control cardiocytes,this value represents a return to 94% of the controlvalue, and it is apparent that a return toward thenormal condition is well under way. The ratio ofmitochondria to myofibrils was again within normal

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

372 Circulation Research/Vol. 54, No. 4, April 1984

FIGURE 2. Electron micrographs of control (panel A), 1-week-unloaded (panel B), and 1-week-unloaded, I -week-reloaded (panel O papillarymuscles cut in cross-section. Extensive loss of contractile elements in the unloaded tissue resulted in large areas of organelle-free cytoplasm, andthe remaining myofibrils are present predominantly near the periphery of the cell (panel B). Upon reloading for 1 week after one week of unloading(panel O, myofibrils again fill the cytoplasm and have a uniform orientation. Cap, capillary; ID, intercalated disc; Mf, myofibrils; Mt, mitochondria;N, nucleus.

limits, and, as in the unloaded tissue, the ratio ofexternal sarcolemmal surface area to volume re-mained constant as these changes were occurring.

Cardiac FunctionThe dimensions of the right ventricular papillary

muscles selected from cats for the study of contrac-

tile behavior were not significantly different in thecontrol and reloaded groups: muscle length was 6.40± 0.86 mm control and 7.75 ± 0.64 mm reloaded;muscle cross-sectional area was 0.63 ±0.16 mm2

control and 0.68 ± 0.14 mm2 reloaded. The upperlimits of papillary muscle cross-sectional area formetabolic support by diffusion at 29°C has been

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

Thompson et a/./Myocardial Atrophy Reversal 373

FIGURE 3. Electron micrographs of control (panel A), 1-week-unloaded (panel B), and 1-week-unloaded, 1-week-reloaded (panel Q papillarymuscles cut longitudinally. Loss and disorientation of contractile elements is the predominant feature of the unloaded myocardium (panel B);remaining myofibrils lie at various orientations both within the same cell (arrows) and in adjacent cells, and their typical banding pattern hasbeen lost. In tissue reloaded for I week after 3 week of unloading (panel Q, the myofibrils again appear normal with respect to quantity,orientation, and banding. Sarcomeres in adjacent myofibrils are again in register. Abbreviations are the same as in Figure 1

(Cooper, 1979). Noused in the present

found to be about 1.10 mmmuscle larger than this wasstudy.

The mechanical data for isotonic contractions,shown in Figure 4, demonstrate that the decrease inboth the velocity and extent of shortening at allloads, shown before (Cooper and Tomanek, 1982)

for the unloaded muscles, returned to normal in thereloaded muscles.

The mechanical data for isometric contractions arepresented in Figure 5 and in Table 2. We had pre-viously found (Cooper and Tomanek, 1982) a re-duction in the active tension developed during con-traction at all muscle lengths studied in the unloaded

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

374 Circulation Research/Vo/. 54, No. 4, April 1984

0 8

0 6

0 4

Q2

1

-

-

-

V

10

•Control, N-5• Relooded, N-5

\

\

20 30 40

02r

OJ

00,

TABLE 2Papillary Muscle Mechanics At

0 a 20 30 40

FIGURE 4. Left panel: average force-velocity values ± se for thecontrol and reloaded muscles. Right panel: extent of shortening vs.force for the same muscles.

muscles, and an increase, more pronounced atlonger muscle lengths, in the tension required tostretch the resting muscle. Figure 5 shows that boththe active and resting length-tension relationshipsreturned to normal with reloading. Table 2 providesa summary of isometric contractile function at L^x,that muscle length at which developed tension isgreatest. No decrement with respect to any of thesemeasures of isometric contractile performance wasfound for the reloaded muscles.

Discussion

This study demonstrates for the first time thatadult mammalian myocardium, despite the fact thatit cannot regenerate by hyperplasia, retains a re-markable plasticity in response to load. In the initiallaboratory investigation of unloaded myocardium(Tomanek and Cooper, 1981; Cooper and Tomanek,1982), we found progressive cardiocyte atrophy:there was a marked loss of contractile elements, andthe residual sarcomeres were no longer oriented inparallel with the long axis of either the muscle or itsconstituent cardiocytes. This cardiocyte atrophy was

70

60

_ 50

I 40

5 30

I20

10

0

•Control, N - 5

• ReJoodod, N-5

-22 -18 -M -10 -6 -2 ^ T _ *ZLength (% donga from L _ , )

FIGURE 5. Average length-tension values ± S£ for the control andreloaded muscles. The convex pair of lines, beginning on the left withthe upper pair of points nearest the ordinate, represents active tension.The concave pair of lines, beginning on the left with the lower pairof points nearest the ordinate, represents resting tension.

Control Reloaded

Active force (rriN/mm )Resting force (rnN/mm2)Resting/total force+dF/dt (mN/mm2/sec)Time-to-peak force (msec)-dF/dt (mN/mm2/sec)Relaxation time (msec)

64.20 ±17.14 ±0.21 ±278 ±356 ±179 ±590 ±

1.573.270.04713515

62.6116.80.21272362185576

±±±±±±±

1.283.200.0311111221

Results are expressed as mean ± SE. There were five catsstudied, with both a control and a reloaded papillary muscleremoved from the right ventricle of each. Resting/total force isresting force divided by the sum of the active and resting force.The terms +dF/dt and —dF/dt refer to the maximum rates of riseand fall of active force, respectively. Time-to-peak force is thetime from the end of the latent period to peak active force;relaxation time is the time from peak active force to completerelaxation.

accompanied by a depression of both isometric andisotonic contractile function. These changes oc-curred very rapidly: atrophic changes were observedwithin 1-3 days of unloading, and contractile func-tion was virtually absent at times greater than 3 daysafter unloading.

The present study demonstrates that myocardialatrophy, at least in its early stages, is completelyreversible. The loss or restoration of normal cardi-ocyte intracellular organization with absent or re-stored load suggests that the degree and direction ofstress placed on the myocardium is possibly theprimary determinant of both the cellular content ofcontractile elements, and their orientation with re-spect to the cellular axis along which stress is trans-mitted during contraction and/or strain is transmit-ted during relaxation. The importance of this struc-tural plasticity is emphasized by the accompanyingfunctional changes.

Experimental ModelThe right ventricular papillary muscle provides an

ideal model for studying cardiac unloading and re-loading, since the transection of chordae tendinaeunloads the muscle, and reattachment of the chor-dae tendinae reloads it. In this cylindrical muscle,both the cardiocytes and their myofilaments arearranged in parallel with the direction of tensiongeneration, allowing the relation between contractilestructure and contractile function to be studied in ageometrically defined tissue. In addition, vascularand neural constituents enter the papillary muscleat its base; we have found that severing the chordaetendinae does not interfere with the blood supply,innervation, or frequency of contraction in this prep-aration (Cooper and Tomanek, 1982). In this study,an intact adjacent papillary muscle from the sameright ventricle was used as the control in each case;thus, experimental variables other than load werefully accounted for, since they affected the control,unloaded, and reloaded muscles equally.

In the present study, load was reimposed after

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

Thompson et a/./Myocardial Atrophy Reversal

1 week of unloading. This time for reloading waschosen because by 1 week after unloading, there aresignificant decreases in cardiocyte cross-sectionalarea as well as in the volume densities of mitochon-dria and myofibrils, and mechanically effective con-tractile function has already disappeared (Cooperand Tomanek, 1982). However, the loss of myocar-dial cells which might occur at later times afterunloading (Tomanek and Cooper, 1981) is not seenat this early time.

Tissue CompositionThe major finding of this study with respect to

the quantity and proportion of cardiac tissue com-ponents is that, both during atrophy and duringatrophy reversal, when cardiocyte size in terms ofcross-sectional area decreases and then increasesback to the control value, the cardiocyte contributionto the myocardium remains relatively constant andin proportion to the amount of the other tissueelements in which it is embedded. Specifically, Fig-ure 1 shows that the percent of each tissue elementwe examined remains about the same, whereas thetotal amount of cardiac tissue falls with atrophy andthen rises again with atrophy reversal.

The most striking difference with respect to tissuecomposition between a reversible load reduction, inthe present study of atrophy, and the reversible loadincreases, in other studies of hypertrophy, is that—in the present study of atrophy reversal—the pro-portion of connective tissue does not change,whereas, with hypertrophy, the proportion of con-nective tissue goes up fairly uniformly and may ormay not return to normal with hyertrophy reversal.In a recent pair of studies examining a chronicprogressive pressure overload of the non-failing catright ventricle, we found that collagen increasesalong with the duration of hypertrophy (Cooper etal., 1981), but that this increase returns to the controlvalue when the pressure overload is relieved(Cooper, 1982). In contrast, in view of the demon-stration that an acute pressure overload severeenough to cause congestive heart failure and sub-stantial short-term mortality (Spann et al., 1967)produces fairly gross histological fibrosis (Bishopand Melsen, 1976), it is unlikely that a full reversalof increased collagen content would be observedfollowing a prolonged, severe pressure overload.Similarly, if full cardiac unloading were allowed topersist until actual cardiocyte loss became promi-nent, the resultant relative increase in connectivetissue would be expected to persist even after thereimposition of a normal load.

An intriguing aspect of the interrelationship ofthe various cardiac tissue components during un-loading and reloading is that both the muscle cellsand their supportive and nutritive matrix change insynchrony. Whereas this may or may not be thecase during normal neonatal cardiac development(Legato, 1979a; Olivetti et al., 1980), the retentionof the very complex and specific interrelationships

375

of the various tissue elements in adult myocardium(Borg and Caulfield, 1981; Borg et al., 1983), wherecardiocyte division is not possible (Zak, 1974;Cooper et al., 1981), would probably be a necessarycondition for the return of normal structure and,thus, normal function, after either atrophy or hy-pertrophy had taken place. Thus, our finding withrespect to the synchronous changes in tissue com-position in atrophy and hypertrophy, respectively,would provide a structural foundation for the returnto normal function in each case.

Ultrastructural CompositionThe major finding of this study with respect to

the intracellular organization of the cardiac musclecells is that these cardiocytes undergo a remarkablededifferentiation during unloading and then redif-ferentiation during reloading, especially with respectto the contractile filaments. In view of the srikingloss and axial disorientation of the myofilamentsduring atrophy (Figs. 2B and 3B), it would not beexpected that these cells could generate substantialforce or shortening in parallel with the original longaxis. Indeed, our previous study of the contractileproperties of these papillary muscles during earlyatrophy demonstrates that effective contractile func-tion is essentially absent (Cooper and Tomanek,1982). This concordance of ultrastructure with func-tion is paralleled by the present data showing, first,a regeneration and axial realignment of the myo-filaments (Figs. 2C and 3C), and, second, a returnof normal contractile function (Figs. 4 and 5).

The second ultrastructural feature of interest isthe decline and rise of the mitochondrial comple-ment of the dedifferentiating and redifferentiatingcardiocytes. The volume fraction of the mitochon-dria does not decline as much as that of the myo-fibrils, and the ratio of these two organelles returnsto control with reloading. While one or severalmechanisms by which changing load causes parallelalterations in diverse elements on the level of bothtissue and cell are unclear, it is interesting that theamounts of these structurally and functionally di-verse components change together during both un-loading and reloading.

This interrelationship of ultrastructure and func-tion has at least one other possible parallel in adultmyocardium. The cellular morphology of the un-loaded cardiocytes strongly resembles that of thecells of the cardiac conduction system. Purkinje cellsand atrioventricular bundle cells have large areas ofsarcoplasm devoid of organelles, with myofibrilslocated peripherally and in various orientations withrespect to the cellular long axis (Thornell and Eriks-son, 1981; Marino, 1979). This may well be relatedto the fact that these cells are embedded in relativelylarge amounts of stiff connective tissue (Bencosmeet al., 1969), and thus are structurally insulated fromthe normal systolic and diastolic stress and strain. Ithas also been noted that, in the Purkinje cells ofventricular false tendons, the orientation of the myo-

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

376 Circulation Research/Vol. 54, No. 4, April 1984

fibrils can be correlated with the direction of passivetension to which they are subjected during ventric-ular filling (Thomell et al., 1976; Thornell and Er-icksson, 1981). This correlation of the morphologyof contractile proteins in unloaded and reloadedmyocardium with that of conduction system cardi-ocytes raises the distinct possibility that the mor-phology of cardiocytes with respect to their contrac-tile organelles is a simple function of the extent anddirection of the load placed upon these cells.

Contractile FunctionMyocardial unloading produces a severe, progres-

sive loss of contractile function that is characterizedby a decrease in the active properties of shorteningand tension generation and an increase in passivestiffness (Cooper and Tomanek, 1982). Straightfor-ward structural bases for both the active and passivechanges are available in the loss and disorientationof the sarcomeres and in the increase in collagencontent. The full return of both active and passivemechanical properties to normal with reloading,coincident with the structural reorganization, lendsstrong support to the concept that the functionalchanges are structurally based.

Relation to Cardiac Myogenesis and GrowthThe unloaded cardiocytes in Figures 2B and 3B

closely resemble neonatal cardiocytes (Fig. 3A inLegato, 1979a, Fig. 1 in Legato, 1979b) in two im-portant respects: (1) the sarcoplasm is only partiallyoccupied by myofilaments and mitochondria, and(2) the uniform axial orientation of the existingsarcomeres is not yet established. During both nor-mal postnatal myogenesis and reloading of atrophicmyocardium, four changes result in the cardiocyteattaining normal adult morphology: (1) the sarcom-eres both develop and increase greatly in volumedensity (Olivetti et al., 1980; and Table 1 of thisstudy), (2) the volume density of the mitochondriaincreases (Legato, 1979; and Table 1 of this study),(3) the ratio of mitochondria to myofibrils attains anormal adult value (Legato, 1979b; Table 1 of thisstudy) and (4) the sarcomere axis aligns with theload axis appropriate to each particular region ofworking ventricular myocardium (Fig. 3, B vs. C). Inaddition, when load along a previously establishedaxis of myocardial stress increases in the adult, car-diac hypertrophy is accomplished by the serial andparallel addition of new sarcomeres along the orig-inal load axis (Bishop, 1971; Marino et al., 1983b).

Thus, a morphological parallel exists between car-diac myogenesis with its extension in the adult intohypertrophy and the redifferentiation of atrophied,dedifferentiated myocardium in response to loadreimposition. In normal cardiac myogenesis, a com-plex interplay of humoral, neural, and loading con-ditions contributes to the eventual realization of theadult cardiocyte (Zak, 1974; Rakusan et al., 1978).In cardiac hypertrophy, a number of humoral,

neural, and loading conditions probably contributeto the adaptive process. It has been suggested thatsympathetic nerves (Ostman-Smith, 1979, 1981)and circulating catecholamines (Simpson et al.,1982) may play the most important roles. In contrast,in this model of reversible cardiac atrophy, wherehumoral and neural changes affect experimental andcontrol tissues equally, variable load alone seems tobe of primary importance to the changes observed.

ConclusionThe major finding of the present study is that

when load is varied without independent changesin other factors, the removal of load results in aprompt dedifferentiation and atrophy of the affectedcardiocytes, and the reimposition of load leads toequally prompt cardiocyte growth and redifferentia-tion, with a return of normal contractile function.This would suggest that load may be the primarydeterminant of myocardial structure and ultimatelyof myocardial function.

We wish to thank Thomas Vinagucrra, Steven Vincigucrra, BngtdKane, and Joseph Severdia for their technical assistance with thesestudies, and Maxine Blob for the preparation of the manuscript. Thesupport of the Lancaster, York-Adams, Northwestern, Keystone,Southwestern, and Southeastern Pennsylvania, and the DelawareChapters of the American Heart Association is also appreciated.

Supported by Grants HL 29146, HL 29718, and HL 17631 fromthe National Institutes of Health, and by Grant 82-800 from theAmerican Heart Association.

Address for reprints: George Cooper, IV, M.D., Section of Cardiol-ogy, Temple University Hospital, 3401 N. Broad Street, Philadelphia,Pennsylvania 19140.

Received September 19, 1983; accepted for publication January 20,1984.

ReferencesAlpert NR (ed) (1983) Perspectives in Cardiovascular Research,

vol 7, Myocardial Hypertrophy and Failure. New York, RavenPress

Bencosme SA, Tnllo A, Alanis J, Benitez D (1969) Correlativeultrasrructural and electrophysiological study of the Purkinjesystem of the heart J Electrocardiol 2: 27-38

Bishop SP (1971) Ulrrasrructural alterations in canine myocardialhypertrophy. In Cardiac Hypertrophy, edited by NR AlpertNew York, Academic Press, pp 107-124

Bishop SP, Melsen LR (1976) Myocardial necrosis, fibrosis, andDNA synthesis in experimental cardiac hypertrophy inducedby sudden pressure overload. Circ Res 39: 238-245

Borg TK, Caulfield JB (1981) The collagen matrix of the heart.Fed Proc 40: 2037-2041

Borg TK, Johnson LD, Lill PH (1983) Specific attachment ofcollagen to cardiac myocytes in vivo and in vitro. Dev Biol 97:417-423

Cooper G (1976) The myocardial energetic active state: I. Oxygenconsumption during tetanus of cat papillary muscle. Circ Res39: 695-704

Cooper G (1979) Myocardial energetics during isometric twitchcontractions of cat papillary muscle. Am J Physiol 236: H244-H253

Cooper G (1982) Reversibility of chronic progressive pressureoverload. Circulation (abstr). 66 (suppl. II): 252

Cooper G, Tomanek RJ (1982) Load regulation of the structure,composition and function of mammalian myocardium. Circ Res50: 788-798

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

Thompson et a/./Myocardial Atrophy Reversal 377

Cooper G, Puga FJ, Zujko KJ, Harrison CE, Coleman HN (1973)Normal myocardial function and energetics in volume-overloadhypertrophy in the cat. Circ Res 32: 140-148

Cooper G, Tomanek RJ, Ehrhardt JC, Marcus ML (1981) Chronicprogressive pressure overload of the cat right ventricle. CircRes 48: 488-497

Eisenberg BR, Cohen IS (1983) The ultrastructure of the cardiacPurkin)e strand in the dog: a morphometnc analysis. Proc RSoc Lond [Biol] 217: 191-213

Elias H, Hennig A, Schwartz DE (1971) Stereology: Applicationsto biomedical research. Physiol Rev 51: 158-200

Legato MJ (1979a) Cellular mechanisms of normal growth in themammalian heart. I. Qualitative and quantitative features ofventricular architecture in the dog from birth to five months ofage. Cir Res 44: 250-262

Legato MJ (1979b) Cellular mechanisms of normal growth in themammalian heart. II. A quantitative and qualitative comparisonbetween the right and left ventricular myocytes in the dog frombirth to five months of age. Circ Res 44: 263-279

Marino TA (1979) The atrioventricular node and bundle in theferret heart: a light and quantitative electron microscopic study.Am J Anat 154: 365-392

Marino TA, Houser SR, Martin FG, Freeman AR (1983a) Anultrastructural morphometnc study of the papillary musde ofthe right ventricle of the cat. Cell Tissue Res 230: 543-552

Marino TA, Houser SR, Cooper G (1983b) Early morphologicalalterations of pressure overloaded cat right ventricular myocar-dium. Anat Rec 207: 417-426

Olivetti G, Anversa P, Loud AV (1980) Morphometric study ofearly postnatal development in the left and right ventricularmyocardium of the rat. II. Tissue composition, capillary growth,and sarcoplasrruc alterations. Circ Res 46: 503-512

Ostman-Smith I (1979) Adaptive changes in the sympatheticnervous system and some effector organs of the rat followinglong term exercise or cold acclimation and the role of cardiacsympathetic nerves in the genesis of compensatory cardiachypertrophy. Acta Physiol Scand [Suppl] 108: 1-118

Ostman-Smith I (1981) Cardiac sympathetic nerves as the finalcommon pathway in the induction of adaptive cardiac hyper-trophy. Clin Sci 61: 265-272

Page E, McCallister LP, Power B (1971) Stereological measure-ments of cardiac ulrrastructures implicated in excitation-con-traction coupling Proc Natl Acad Sci USA 68: 1465-1466

Rakusan K, Raman S, Layberry R, Korecky B (1978) The influenceof aging and growth on the postnatal development of cardiacmuscle in rats. Circ Res 42: 212-218

Reynolds ES (1963) The use of lead citrate at high pH as anelectron opaque stain in electron microscopy. J Cell Biol 17:208-213

Shay J (1975) Economy of effort in electron microscope mor-phometry. Am J Pathol 81: 503-512

Simpson P, McGrath A, Savion S (1982) Myocyte hypertrophyin neonatal rat heart cultures and its regulation by serum andby catecholamines. Circ Res 51: 787-801

Spann JF, Buccino RA, Sonnenblick EH, Braunwald E (1967)Contractile state of cardiac muscle obtained from cats withexperimentally produced ventricular hypertrophy and heartfailure. Circ Res 21: 341-354

Spurr AR (1969) A low-viscosity epoxy resin embedding mediumfor electron microscopy. J Ultrastruct Res 26: 31-43

Thomell LE, Encksson A (1981) Filament systems in the Purkinjefibers of the heart. Am J Physiol 241: H291-H305

Thomell LE, Sjostrom M, Andersson KE (1976) The relationshipbetween mechanical stress and myofibrillar organization inheart Purkinje fibers. J Mol Cell Cardiol 8: 689-695

Tomanek R], Cooper G (1981) Morphological changes in themechanically unloaded myocardial cell. Anat Rec 200: 271-280

Trump BF, Smuckler A, Benditt EP (1961) A method for stainingepoxy sections for light microscopy. J Ultrastruct Res 5: 343-348

Watson ML (1958) Staining of tissue sections for electron micros-copy with heavy metals. J Biophys Biochem Cytol 4: 475—478

Zak R (1974) Development and proliferative capacity of cardiacmuscle cells. Circ Res 35: (suppl II): 17-26

INDEX TERMS: Cardiac dedifferentiation • Cardiac differentia-tion • Cardiac atrophy • Cardiac development

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from

E W Thompson, T A Marino, C E Uboh, R L Kent and G Cooper, 4thAtrophy reversal and cardiocyte redifferentiation in reloaded cat myocardium.

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1984 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.54.4.3671984;54:367-377Circ Res. 

http://circres.ahajournals.org/content/54/4/367World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Research Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on May 25, 2018

http://circres.ahajournals.org/D

ownloaded from