Quantitative electron microscopic description of heart muscle cells: Application to normal, hypertrophied and thyroxin-stimulated hearts

Quantitative Electron Heart Muscle Cells

Microscopic Description of

Application to Normal, Hypertrophied and Thyroxin-Stimulated Hearts

ERNEST PAGE, MD L P McCALLISTER, PhD

Chicago, IlImos

Mathematical techniques can be used to extract quantitative information from tissue electron micrographs of heart muscle. With these methods it has been possible to measure the fractions of myocardial cell volume made up of myofibrils, mitochondria, sarcotubules, T system and sarcoplasm, as well as the membrane areas per unit of cell volume of plasma membrane, sarcotubular membrane and mitochondrial cristae. The techniques have been used to quantitate the changes in the ultrastructure of myocardial cells in experimental left ventricular hypertrophy and in experimentally induced hypothyroidism before and after treatment with thyroxin. These measurements have demonstrated striking and physiologically significant changes in the ultrastructural composition of the cells. The pattern of ultrastructural change in ventricular hypertrophy differs in characteristic ways from the pattern during thyroxin-stimulated cardiac cellular growth. These observations have suggested certain generalizations about the constraints to which growing myocardial cells are subject.

New and simple microchemical methods have been developed to determine the cardiac contents of myofibrils and mitochondrial cristae. The methods are applicable to samples of heart muscle obtained at autopsy, surgery or biopsy, as well as to experimental material. Application of these techniques to the study of experimental left ventricular hypertrophy and thyroxin-stimulated growth of myocardial cells has confirmed and extended the results of quantitahve measurements on electron micrographs. It is suggested that the ap- proaches described can form the basis of a quantitahve electron microscopic and microchemical pathology of heart muscle.

From the Departments of Medlclne and Physiology, the Prltzker School of Medlclne, the Unlverslty of Chlcago, Chlcago, III This work was supported by U S Public Health Service Grant HL-10503, Myocardlal In- farction Research Unit Contract PH 43-68- 1334, and grants from the American Heart Assoclatlon and the Chlcago and llllnols Heart Associations During the course of this work Dr Page was a recipient of a U S Public Health Service Career Development Award and Dr McCalllster of a postdoctoral tratneeship under the U S Public Health Service Training Grant HL-05673-05

Address for reprints Ernest Page, MD, Department of Medicme, Unlverslty of Chl- cage, Hospital P 0 Box 137. 950 E 59th St , Chicago, III 60637

The application of electron microscopy to heart muscle promised to provide physlologlsts mterested m the normal and abnormal func- tioning of heart muscle with two sorts of mformatlon On the one hand, the electron microscope was expected to reveal early changes that are produced m cellular ultrastructures by physlologlc or pathologic processes On the other hand, it was antlclpated that electron mlcroscoplc observations might explain the mechamsms of cellular phenomena that depend on the mtrlcate details of cellular fme structure and especially on the geometric relation between the various types of cellular membranes. Techmcally satisfactory electron mlcrographs of nearly all types of mammalian heart muscle have been available for more than 10 years l Nevertheless, the expected mformatlon about ultrastructural changes and ultrastructural geometry has not been forthcommg To obtain such mformatlon, it 1s neces- sary to apply quantltatlve methods of analysis to tissue electron mlcrographs of heart muscle, and such methods have not been avall- able until now

Recently, it has become possible to apply to electron mlcrographs of mammalian heart muscle mathematical techmques for measurmg

172 February 1973 The American Journal of CARDIOLOGY Volume 31

the amounts of the various ultrastructural components m the myocardial ce11.2-4 In our paper we de- scribe the use of these techniques to study several problems of Interest to cardiologists: (1) the ultrastructural composition of normal left ventricular muscle and the changes m that composition during experimental left ventricular hypertrophy; and (2) the ultrastructural composition of left ventricular muscle in experimental hypothyroidism and the changes in this composition that result from treatment with thyroxin.

Principles

The nature of ultrastructural changes in heart muscle during physiologic and pathologic processes: Prenatal and postnatal growth or pathologic processes like myocardial hypertrophy produce graded changes in the ultrastructural composition of myocardial cells. These graded changes are quantitative; that is, as a result of the change, there is more or less of a particular component in the cell. Such changes are not usually of the all or none type, hence, they are not necessarily detectable by inspection of the electron micrograph. Instead, the detection of changes requires the application of sensitive mathematical methods to the electron micrographs.

Why would it not be simpler to isolate the components of the cardiac cell from homogenates of heart muscle by biochemical techniques and to measure the content of each component thus isolated by chemical analysis? One reason is that such techniques are tedious, and usually require a large amount of tissue An even more important consider- ation is, that for the ultrastructures that make up myocardial cells, such techniques do not usually give reliable estimates of how much of a given component was present m the intact cell Fractionation procedures may yield a small amount of a highly purified component, but they usually cause significant de- struction or loss of that component. Conversely, fractionation procedures designed to yield most of the cellular component originally present m the intact myocardmm almost invariably yield that component in a form contaminated to a sigmficant degree with other components.

Quantitative analysis (“stereology”) of tissue electron micrographs: The mathematical techniques of point counting and the determination of membrane areas by counting intersections with lines of a test grid are powerful new tools for the extrac- tion of quantitative information from electron micrographs of tissues.s-7 These techniques, which have become known under the rubric “stereology, ” are based on the analysis of photographic prints of electron micrographs like that shown in Figure 1. The choice of the mathematical treatment appropriate for a particular problem m cardiac ultrastructure de- pends on the orientation of the section and on the distribution within the tissue of the particular structure being investigated. Moreover, the application of mathematical techniques to analysis of electron ml-

ELECTRON MICROSCOPIC DESCRIPTION OF HEART MUSCLE CELLS-PAGE AND McCALLlSTER

FIGURE 1 Electron mrcrograph of normal rat left ventrtcular myocardtal cell In cross sectron, showmg techmque used for quanhtattve analysts The prtnt IS overlald with a thin sheet of transparent plastic mscrtbed with a rectangular grid The grid IS

used for pomt counhng and Me mtegratton as descrtbed In Ref- erences 2 and 3 (X5.184, reduced by 34 percent )

crographs of tissue makes sense only if the tissue is well preserved, artifacts arising from inadequate fixation, shrinkage or swelling due to the use of non- isotonic fixatives, and other preparative artifacts in- validate the analysis.

Quantitative Description of Normal Ventricular Myocardium

Table I presents a quantitative description of the ultrastructural composition of myocardial cells from the normal rat left ventricle; it shows that the normal cell volume is composed of 47.6 percent myofibrils, 35.8 percent mitochondria, 1 percent T system, 3.5 percent sarcotubules and 12 percent sarcoplasm (cytoplasm), nuclei and other structures. Table II gives the surface areas per unit of cell volume of 4 types of membranes. It is apparent that the mitochondrial cristae are by far the most extensive mem- branous component of the cell. This finding is not surprising, smce the cristae are the sites of oxidative phosphorylation and thus the predominant source of energy the heart muscle utilizes for contraction. The plasma membrane or sarcolemma is composed of 2 regions: (1) the external cellular envelope (external

February 1973 The American Journal of CARDIOLOGY Volume 31 173


TABLE I

Compositfon of Myocardlal Cells from Rat Left Ventncles*

Organelle

Percent of

Cell Volume Princrpal Functions

Myofrbrrls 47 6 Contrachon and relaxatron

107

Mrtochondna 35.8 Respiratron, oxrdatrve

f06 phosphorylatron

T system 10 Transmrssron of electrrcal

*o 1 srgnal at ceil surface into

Interior of cell

Sarcotubulest 35 Control of contrachon and

f02 relaxahon by regulatron

of ronrzed calcium

concentration

Terminal crsternae 0 35 Store calcium and release

zto.02 rt into the sarcoplasm

(cytoplasm)

Rest of sarcotubu- 3 15 Collect calcium from

lar network zto 02 sarcoplasm and return it

to terminal cisternae

Sarcoplasm (cyto- 12 1

plasm), nucler and

other structures

mg of calcmm to the sarcotubular membrane, and possibly also by transporting calcium mto the lumen of the tubules, thereby causing the myofibrils to relax. In the final step, the calcium bound to or within the noncisternal portion of the sarcotubular network is translocated back to the terminal cisternae for temporary storage

Ultrastructural Changes in Experimental Left Ventricular Hypertrophy

Changes in contents of mitochondria and myofibrils: It 1s particularly mstructive to examme how the proportions and absolute amounts of the various ultrastructures that make up a myocardial cell change when the cell is stimulated to grow. One experimental example of the response of myocardial ultrastructures to a growth stimulus is the hypertrophy that results when the ascending aorta is partially constricted with a silver band 4.1e The consequent increase m the pressure against which the left ventricle has to elect blood produces a progressive increase m the size of myocardlal cells without at the same time mcreasmg the number of these cells

* Data rn Tables I and II are based on prevrously unpublrshed

results for left ventncles from 5 normal female rats weighing

200 to 220 g

t Values for sarcotubular volume are mlnlmal estimates

because some of these structures will be sectioned tangentially

and thus not measured The true values may be up to 25

percent greater

sarcolemma), which is the locus of the action potential and hence of electrical excitation and conduc- tion; and (2) the transverse tubular, or T system, which consists of tubular mfoldmgs of the external sarcolemma and serves to transmit the message that an action potential is passing across the cell surface to the interior of the cell The data m Table II mdi- cate that the plasma membrane lmmg the T system mcreases the total surface area of the cell by about one third. The membrane luring the sarcotubules, or sarcoplasmic reticulum, is also divided mto 2 por- tions 1 portion, the terminal cisternae or diads, is located m the sarcoplasm beneath the plasma membrane lmmg both the external and T tubular surface 8.9 The terminal cisternae function as storage depots for calcium, they translate the electrical sig- nal at the cell surface, as well as that transmitted down the T tubules, as a command to release calcium mto the sarcoplasm The calcium released mter- acts with a calcium-receptor protein on the myofibril, causing the myofibrils to contract The second or noncisternal component of the sarcotubules is a network that courses over the surface of and to some extent through the myofibrillar contractile material This portion of the sarcotubular system functions to lower the cytoplasmic calcmm level after the myofibrils have contracted It does so by a selective bmd-

Figure 2 shows the fractions of cell volume made up of myoflbrils and mitochondria at intervals of 1 to 49 days after aortic constriction The upper graph indicates that 10 or more days after constriction of the aorta the fraction of cell volume made up of myoflbrils is uniformly increased. At the same time, as shown m the lower graph, the fraction of cell volume made up of mitochondria shows a significant decrease The increase m myofibrillar volume occurs partially at the expense of mltochondrlal volume and partially by encroachment on the 12 percent of cell volume normally made up predominantly of sarcoplasm This observation indicates that the cumulative volume of myofibrils. the structures whose con- tractions are responsible for most of the energy con-

TABLE II

Membrane Areas in Myocardial Cells from Rat Left Ventricles

Membrane Type

Membrane

Area

Plasma membrane

External sarcolemma

External sarcolemma

+ T system

Mrtochondnal cnstae

Sarcotubules*

0 30 It 0 02

0 39 f 0 02

111 1

1 22 i 0 05

2 51 t 0 08

*Values for sarcotubules are mrnimal estimates subject to a 25 percent correction as described In Table I


Units

pm2 membrane area

pm? cell volume

yrnL membrane area

Mm3 cell volume

pm2 membrane area

Mm3 cell volume

pm* membrane area

pm3 cell volume

pm2 membrane area

pm3 myofibnllar

volume


sumption of the myocardial cell, 1s mcreasmg; at the same time the cumulative volume of mitochondria, the structures that produce the adenosine tnphos- phate (ATP) used as a “fuel” by the myofibrils, is diminishing. A progressron of this disproportion may well be lmphcated in the ultimate development of heart failure by the hypertrophied myocardium Slmllar observations have been made in other animal models. 11-1s

Since the quantitative analysis of electron mlcrographs 1s a new method, we have sought independent techniques to correlate with this analysis. We have developed new microchemical methods for assaying the amount of myofibril and the amount of mitochondrial cristal membrane per unit of dry weight of heart. For the mlcrochemlcal measurement of myoflbrillar content it has proved convenient to determine a fraction of magnesium that IS sequestered within the thin (actm) filaments of the myofibrils at the time the myoflbril 1s assembled.14 For the mlcro- chemical measurement of the amount of mltochondrlal cristal membrane per unit of dry weight of heart, a useful assay has been the cellular copper content15; this assay reflects the copper in the cytochrome a and cytochrome a oxldase of the mitochon- drlal crlstae which account for more than 90 percent

FIGURE 3 Effect of hypertrophy on ventrlc- ular content of sequestered myoflbrlllar mag- neslum (an assay for myoflbrlllar mass) and cell copper (an assay for mltochondrlal cristae) A and C show ratios (r) of magnesium content and copper content of hypertrophled ventricles to the respective contents of ventricles from sham-operated littermates Values lymg above the horizontal line for r = 1 00 indicate a net accumulation of myoflbrillar magnesium or of copper by the myocardial cells of the hypertrophied ventricles B and D show the data of A and C replotted to Illustrate the relation of the increases in the contents of myoflbrlllar magnesium and copper to the increase in ventricular dry weight which occurs m hypertrophy After 5 days of aortic constrlction the points for myofibrlllar magnesium (6) lie above the lme r’ = 1 00, mdicatmg that the myoflbrlllar contribution to the dry weight has increased proportionately more than ventricular dry weight as a whole, 2 or more days after aortic constriction the points for cellular copper content (D) lie below the line r’ = 1 00, m- dicatmg that the mitochondrial contribution to the dry weight has increased proportlon- ately less than ventricular dry weight as a whole When myofibrlls and mltochondrla were Isolated from left ventricular homogenates, the sequestered magnesium content per gram of myofibrillar protein and the copper content per gram of mltochondrial protein were found to be the same in hypertrophied and control hearts 4.15

.58-

% .38,

I=

:: 34- a L

30-

. AORTIC CONSTRICTION o SHAM -OPERATED CONTROLS

t 0 10 20 30 40 50

DAYS AFTER OPERATION

FIGURE 2. Effect of left ventricular hypertrophy In rats on the fraction of myocardial cell volume made up of myoflbrlls and mitochondria, respectively Standard errors of the mean are smaller than the circles used to present the data In this figure and Figures 3, 4 and 8, hypertrophy was produced by constnct- mg the ascending aorta of rats with a silver band for periods of 1 to 50 days Data replotted from Page et al 4

l

l

;I m

: 0 4 12 20 26 60

l

k DAYS AFTER AORTIC CONSTRICTION

l

MYOFIERILS 0

0 0

0

l

a 0

MITOCHONDRIA

l

6

1.20 - 0

l.10-

l l

EO- , 1 .* 1 0 4 12 20 28 60

DAYS AFTER AORTIC CONSTRICTION

l

I l

’ : l l ! 8

l

l l

I .* ,

1 2 3 49

DAYS AFTER AORTIC CONSTRICTION -;!‘I DAYS AFTER AORTIC CONSTRICTION L

February 1973 The Amerlcan Journal of CARDIOLOGY Volume 31 175


0 CONTROL •a HYPERTROPHIED

-A- -B-

z 0.0 -I L EXTERNAL EXTERNAL EXTERNAL EXTERNAL

SARCOLEMMA SARCOLEMMA SARCOLEMMA SARCOLEMMA +T-SYSTEM +T-SYSTEM

FIGURE 4. Effect of ventrtcular hypertrophy on surface to volume raho of cell (A) and on the absolute area of plasma membrane In a segment of cell 1 sarcomere (2 pm) In length (B) A shows that a compensatory growth of the plasma membrane Iln- mg the T system keeps the surface to volume ratlo of the hypertrophied cell at normal values Data In B (and Fig 6A, 7B, 10A. 12B and 13A) were calculated from the experlmentally deter- mmed surface to volume ratios assuming that the 2~-long segment IS approximately a right circular cylinder B shows that the absolute areas of the membranes lmmg the external sarcolemmal envelope and T system, respectively, have slgnlflcantly In- creased in response to the hypertrophic stimulus

of the copper content m myocardial cells These 2 microchemical assays are m prmciple applicable to specimens of heart muscle obtained at postmortem study and at surgery, they should therefore prove useful for clmical mvestigations on pathologic material

Figure 3, A and C, shows that with progressively longer durations of aortic constriction both the sequestered myofibrillar magnesium content (and thus myofibrillar mass) and the cellular copper content (and thus the amount of mitochondrial cristal membrane) increase The magnesium content increases proportionately more than the dry weight of the ventricle as a whole (Fig 3B), conversely, as shown m Figure 3D, the ventricular copper content increases less than the ventricular dry weight The 2 newly developed microchemical assays therefore confirm the results obtained by quantitative analysis of electron micrographs

Effects on external sarcolemma and T system: Figure 4 summarizes the results of measurements on the external sarcolemma and T system m hypertrophied left ventricles after 10 days of aortic constric-

tion. When myocardial cells hypertrophy m response to the increased work load, the cross-sectional diameter and area (and thus the volume) of the cells increase The shape of a myocardial cell may be approximated by a right circular cylinder that has a curved surface made up of the external envelope of plasma membrane, the external sarcolemma By virtue of then larger diameter and volume, the hypertrophied myocardial cells may be expected to have a smaller surface to volume ratio, that is, the ratio of external sarcolemmal membrane area per unit of cell volume may be expected to fall as the cell increases m size Figure 4A shows that there is m fact a sigmficant decrease m this ratio, which falls from a value of 0 36 m the control cells to 0 32 m the hypertrophied cells Such a decrease m surface to volume ratio would present the cell with a serious problem The problem arises because the plasma membrane that makes up the external sarcolemmal envelope performs certain vital functions. These functions mclude, among others, uptake of metabohc nutrients from the blood and interstitial spaces, out- ward transport mto the mterstitial spaces and blood of waste products and metabolic mtermediates like lactic acid, which might otherwise accumulate m the cell, and movements of ions associated with electrical excitation

Figures 4B and 5 show the mechanism by which the myocardial cell compensates for the decrease m surface to volume ratio associated with an increase m cell diameter The compensation is the result of an increase m the area of plasma membrane lmmg the T system The plasma membrane lming the T system may be considered an internal sarcolemma, that is, an extension of the external sarcolemmal envelope by an mfoldmg of the plasma membrane mto the interior of the cell Figure 4 indicates that, as the cell expands m diameter, additional T system membrane is made to such an extent as to keep constant the ratio [total (external + internal) sarcolemmal membrane area]/[cell volume] The cell thus be- comes progressively honeycombed with T system channels contammg a fluid having the composition of extracellular fluid (Fig 51, these channels run both transversely and parallel to the long axis of the cell * 9 l6 It seems possible that m extreme degrees of myocardial hypertrophy this compensatory mechanism might break down This possibility merits further study A failure to mamtam the total surface to volume ratio of the cell at a value sufficient to sus- tam the essential functions of the plasma membrane might contribute to the transition from myocardial hypertrophy to heart failure

Growth of the sarcotubular system: Figure 6 d- lustrates the behavior of the sarcotubular system m the hypertrophymg heart Although not shown, the fraction of myocardial cell volume made up of sarcotubules was found to remam approximately the same m the hvnertronhied heart. even though the myocardial cell”s*had enlarged To ‘maintain thk fraction constant. an additional sarcotubular volume


ELECTRON MICROSCOPIC DESCRIPTION OF HEART MUSCLE CELLS-PAGE AND MCALLISTER

CELL WITH LARGE DIAMETER CELL WITH SMALL DIAMETER 0 0 0 0 4 0 0 0 0 0

0 0 0 0 0 0 a c) 0 0 0 0 ,” D ,” ,” _ _

1. rt f”t rt Fl _n.- L 3Dc_31_IL__iL1 tt. .kl u- IlC_3[__31IIDC

t-l H r--M ‘External

i&l ‘L

Sarcolemma T-System

FIGURE 5. Effect of cell growth on relative contrlbutlons of external sarcolemma and T system to total area of plasma membrane (schematic and not to scale) As the cell enlarges, progressively more of Its total surface of plasma membrane (that IS, external sarcolemma plus T system) IS made up of plasma membrane lining the T system For slmpllclty the T system IS shown as a rectangular array, although other geometric ar- rangements have not been ruled out

must therefore have been synthesized in the same proportion as the cell volume as a whole. A more re- vealing insight mto the physiology of hypertrophy is provided by considering the area of sarcotubular membrane. Figure 6 shows that the total sarcotubular membrane area per unit of cell volume increases significantly during hypertrophy, that is, the cell produces proportionately more sarcotubular membrane than it produces cell volume as a whole. An- other cellular component that increases dispropor- tionately is myofibrdlar volume and mass (Fig. 2 and 3A). The sarcotubular membrane is the crucial control system for myofibrillar function. This membrane controls both contraction and relaxation by regulat- mg the concentration of ionized calcium to which the myofibrils are exposed, and thereby the amount of calcium they bind. It is therefore of interest to examme the relation between sarcotubular membrane area and myofibrillar volume during hypertrophy (Fig 6C). This examination leads to a striking conclusion. As the cell grows during hypertrophy, sarcotubular membrane area and myofibrillar volume increase proportionately and their ratio thus remains constant Looked at in another way, the organelle responsible for contraction and relaxation and the membrane system responsible for the control of that organelle are made m a constant proportion to each other, and both grow proportionately more than the cell as a whole.

Myocardial Ultrastructure in Experimental Hypothyroidism and Its Response to Thyroxin

When immature (200 g) female rats undergo thy- roidectomy and are treated with lsliodine, the fol- lowing occur: Their growth is arrested; their body weight fails to increase or decreases slightly, the weight of the heart stabilizes at a lower than normal value; and the ratio of heart weight to body weight falls significantly below that of control rats with intact thyroid function, thus indicating not only that body weight has diminished but also that the weight of the heart has decreased proportionately more than body weight as a whole. If, 30 days after thyroidecto-

my, such animals receive an injection of thyroxm at a dose sufficient to replace the secretion of the ab- sent thyroid glands, the myocardial cells respond by growing rapidly in volume but not m number We have recently examined the way the various cardiac ultrastructures participate in this growth.l?

Response of mitochondria and myofibrils to thyroxin: Figure 7 compares the results of quantitative analysis on electron micrographs of hearts from hypothyroid control rats with those from hypothyroid animals in the same batch given thyroxin injections (180 pg/kg body weight) every 4 days and sacrificed after 24 days. Treatment with thyroxin caused a

q CONTROL EB HYPERTROPHY

Y 0

I-

A

14-

%

a%

12s

WB

!q lo-

OJ

C

FIGURE 6. Effect of ventricular hypertrophy on sarcotubular membrane area As myocardial cells enlarge during hypertrophy, both the absolute area of sarcotubular surface (A) and the sarcotubular surface area per unit of cell volume (B) increase, by contrast, the area of sarcotubular membrane per unit of myofibrlllar volume (C) remains constant In this figure and In Figure 13 sarcotubular membrane areas are mmlmal values, and the true value may be as much as 25 percent greater because some membranes, vlsuallzed in tangential section, are not measured Smce the values for control and hypertrophied hearts are subject to the same percentage error, the compari- son between the 2 states remains valid

February 1973 The Amencan Journal of CARDIOLOGY Volume 31 177


[7 HYPOTHYROID CONTROL

q THYROXIN -TREATED

60-

40-

20-

O- MYOFIBRILS MITOCHON- CELL

FIGURE 7 Response of myoftbrlls and mtto- chondrla In left ventrrcular myocardlal cells from hypothyroid rats to treatment with thyroxln A, cross-sectlonal area of cell, B and C, absolute content of myoflbrlls and mltochondrla In a 1 sarcomere-long cylmdrlcal cell segment, D and E, percent of cell volume occupied by myofibrlls and mltochondrla, respectively As cells grow their con-

MYOFIBRILS MITOCHON- tents of myoflbrlls and mitochondrta in-

CROSS-SECTIONAL DRIA AREA

A B C D

DRIA

E

crease, the Increase in myoflbrtls IS propor- tlonal to the Increase in cell volume, whereas the Increase m mitochondrla IS proportionately greater than that in cell volume l7

nearly 1 4-fold increase m the volume of the myocardial cells (Fig 7A) Concomitantly the total amount of myofibrils, as measured by the volume in a 1 sarcomere-long (2 pm) segment of myocardial cell, m- creased sigmficantly (Fig 7B). Nevertheless, the fraction of cell volume made up of myofibrils re- mained very nearly constant (Fig 7D) The volume of myoflbrils had therefore grown m the same proportion as cell volume The total amount of mlto- chondria m a 1 sarcomere-long segment also m- creased (Fig 7C) However, m the case of mitochon-

drla, unhke that of myoflbrlls, a larger fraction of cell volume was made up of mitochondria after treatment with thyroxm (Fig 7E), the volume of ml- tochondria had therefore increased proportionately more than cell volume In these respects the pattern of ultrastructural growth m response to thyroxm (Fig 7) is entirely different from that m response to constriction of the ascending aorta (Fig 2)

To confirm this conclusion, we have also examined the content of sequestered myofibrillar magnesium (a microchemical assay of myofibrillar mass) and the

0 tiyP0TtiyR010 CONTROL •JTHYR~XIN -TREATED 0 HYPOTHYROIO CONTROL BB THYROXIN-TREATED

;; 028.

?

p 024.

K

z 020

FIGURE 8 Response of sequestered myofibrillar magnesium FIGURE 9 Response of cellular copper content (an Index of the (an index of myofibrlllar content) to thyroxin treatment In left content of mltochondrial crlstae) to thyroxln treatment In left ventricle of hypothyroid rats Both left ventricular dry weight (A) ventricles of hypothyroid rats Left ventricular dry weight (A) and the total amount of sequestered myoflbrillar magnesium In and copper content (B) increase. but the increase in copper the ventricle (B) Increase proportionately, the magnesium con- content IS proportionately greater than that In dry weight the tent per unit of dry weight (C) therefore remains constant I7 copper content per umt dry weight (C) therefore Increases l5

178 February 1973 The Amencan Journal of CARDIOLOGY Volume 31


= I; 2000

aif? +f ‘0 0 q$ 1500

%I U-J Fg 1000

Ya

9

500

? 0

A

0 HYPOTHYROID CONTROL q THYROXIN-TREATED

FIGURE 10. Response of membrane area of mltochondrlal crls-

tae to thyroxm In left ventricular myocardlal cells from hypothyroid rats Total membrane area (A), membrane area per unit of mitochondrial volume (6) and membrane area per unit of myoflbrillar volume (C) all Increase slgnlflcantly l7

cellular copper content (which reflects the content of mitochondrial cristae). Figure 8B shows that the total content of sequestered myofibrillar magnesium increased, but the content per umt of dry weight re- mained approximately the same before and after treatment with thyroxin (Fig. 8C). The total cellular copper content also increased (Fig. 9B); however, the copper content per unit of dry weight increases in response to thyroxin treatment (Fig 9C). These results suggest a disproportionate increase m the mitochondrial content of the tissue and m particular a disproportionate increase m the amount of mitrochondrial cristae. Such increases would be consistent with biochemical studies on the rat heart previously reported by Gross.18

To investigate more directly the effect of thyroxm on the cristae of cardiac mitochondria we compared the areas of cristal membrane per unit of mitochondrial volume in hearts from hypothyroid and thyroxin-treated rats. Figure 10 shows that treatment with thyroxm, in addition to mcreasing the absolute amount of mitochondria in the ventricle (Fig. 7C and 9B) and the fraction of cell volume made up of mitochondria (Fig. 7E), causes an increase in the total area of cristal membrane (Fig. 10A) and m the area of cristal membrane per unit of mitochondrial volume (Fig. 10B). Stated differently, thyroxin produces a modest increase in the density with which cristal membranes (Fig. 11) are packed. Even in the hypothyroid heart cristae are already rather densely packed, a reflection of the large and constant de- mand by the heart on the functions of respiration and oxidative phosphorylation carried out by the cristae. The possibilities for packing additional cristae mto existing mitochondria are therefore limited in mitochondria from ventricular myocardium, pre- sumably, further increases m the total area of mito-

FIGURE 11. High magnlflcatlon electron micrograph of cardiac

mitochondria from animal Inlected with thyroxin, showq dense packing of crlstae l7 (X55,330, reduced by 34 percent )

chondrial cristal membranes available to the cell must be a result primarily of the production of additional mitochondria or the enlargement of the existing mitochondria, or both

From a different point of view, these observations on mitochondrial cristae provide interesting insights into the abnormalities present in the hearts of rats deprived of thyroid hormone. As already noted, these animals have a reduced heart weight per unit of body weight and their myocardial cells are smaller. In proportion to their reduced size, these myocardial cells have a normal concentration of myofibrils. However, the membrane area of mitochondrial cristae per unit of myofibrillar volume is sharply reduced m the hypothyroid state (Fig. 1OC). Even if these remaining cristae function normally, their abil- ity to supply ATP to the contractile elements would be less per unit of myofibrillar volume than in the normal cell. It is possible that the reduced rate of ATP hydrolysis observed m myofibrlls isolated from hearts of hypothyroid ammalsls may be an adjust- ment to this mitochondrial defect.

Growth of plasma membrane and sarcotubules: Figures 12 and 13 present the effects of thyroxin on the membrane areas per unit of cell volume for the external sarcolemma and T system and for the



0 HYPOTHYROID CONTROL •j THYROXM-TREATED 0 HYPOTHYROID CONTROL q THYROXIN-TREATED

-A- -B-

EXTERNAL EXTERNAL EXTERNAL EXTERNAL

SARCOLEMMA SARCOLEW SARCOLEMMA SARCOLEMMA

+ T-SYSTEM +T-SYSTEM

FIGURE 12 Effect of treatment wrth thyroxm on the surface to volume rabo of the cell (A) and on the absolute area of plasma membrane In a 1 sarcomere-long segment of cell (6) in left ventrrcular myocardral cells from hypothyroid rats The total surface to volume ratro IS kept constant by a compensatory growth of plasma membrane lrnmg the T system, as shown schematl- tally In Figure 5 I7

sarcotubules The results lead to the same conclu- sions previously obtained for myocardial cells enlarg- ing in response to constriction of the ascending aorta. During growth of myocardial cells m response to thyroxm, the total surface to volume ratio (that is, the area of external sarcolemmal plus T tubular membrane per unit of cell volume) is again found to be constant, whereas the conventional surface to volume ratio (that is, the area of external sarcolemma per unit of cell volume) decreases (Fig 121. at the same time the area of sarcotubular membrane per unit myofibrillar volume remains constant This result indicates that new sarcotubular membrane and new myofibrillar volume are made m a constant proportion to each other (Fig 13)

1

2

3

180

FIGURE 13 Effect of treatment with thyroxln on sarcotubular membrane area In left ventricular myocardlal cells from hypothyroid rats The total sarcotubular membrane area In a segment of cell 1 sarcomere long increases (A), but membrane area per unit of cell volume (6) and membrane area per unit of myoflbrlllar volume (C) both remaln constant l7

Page E, Fozzard HA Capacltlve, resrstrve and syncytral 4 propertres of heart muscle ultrastructural and physrologlcal consrderabons Appendrx comprlation of articles on cardiac ultrastructure In, Structure and Functron of Muscle, 2nd edition (Bourne GH, ed) New York, Academic Press. In 5 press Page E, McCalhster LP, Power B Stereologrcal measure- 6 ments of cardiac ultrastructures implicated in excitatton- contractron coupling Proc Nat Acad SCI USA 68 1465- 1466.1971 7 Mobley BA, Page E The surface area of sheep cardiac Purklnje flbres J Physiol 220 547-563, 1972 8

References

Page E, Poltment PI, Zak R, et al Myofrbrrllar mass In rat and rabbit heart muscle correlation of microchemical with stereological measurements in normal and hypertrophrc hearts Clrc Res 30 430-439. 1972 Underwood E* Quantitative Stereology Reading Massachu- setts, Addison Wesley Publishing Co, 1970, p l-274 Sitte H Morphometrische Untersuchungen an Zellen In, Quantitative Methods In Morphology (Weibel ER, ElIas H. ed) Berlin, Sprmger Verlag, 1967, p 167-197 Ehas H, Henmg A, Schwartz DE Stereology appltcatlon to biomedical research Physiol Rev 51 158-200. 1971 Page E* Tubular systems in Purkinje cells of the cat heart J

February 1973 The Amertcan Journal of CARDIOLOGY Volume 31

The observations that, as the cell grows, the total surface to volume ratio and the ratio of sarcotubular membrane area per unit of myofibrillar volume remam constant thus appear to be valid for 2 very different growth processes It is tempting to speculate that the phenomena underlying the constancy of these ratios under different conditions may represent fundamental laws by which myocardial cells add new units of ultrastructure For example, it is conceivable that a fixed volume of new myofibrils and the sarcotubular membranes that control myofibrillar function for that volume are coded for as a unit by the genetic material m the nucleus of the myocardial cell, according to this view, the cell would increase its volume by the systematic addition of such units At the same time, the total sarcolemmal surface, which supplies nutrients to the cell and also provides the membrane mechanisms for excitation and excitation-contraction couplmg, is constrained to grow m such a way as not to impair these essential functions It is further conceivable that the accumulation of new plasma membrane lmmg the T system is coded for by the nuclear deoxyribonucleic acid (DNA) so as to regulate cell growth m accordance with these constraints


Ultrastruct Res 17 72-83, 1967 9 Page E; Correlabons between electromrcroscoprc and physr-

ologrcal observations In heart muscle J Gen Physrol 51 211-220.1968

10 Narr KG, Cutrietta AF, Zak R, et al: Brochemrcal correlates of cardiac hypertrophy 1 Experimental model changes In heart weight, RNA content, and nuclear RNA polymerase actrvrty Crrc Res 23 251-262, 1968

11 Anversa P, Vitoii-Mazra L, Visioii 0, et al. Experimental cardiac hypertrophy a quantitative ultrastructural study in the compensatory stage J Molec Cell Cardiol 3 213-228, 1971

12 Woiienberger A, Schutze W: Ueber das Volumenverhaeltnrs von Mrtochondrren zu Myofrbrrllen rm chromsch ueberlast- eten, hypertrophierten Herzen Naturwrssenschaften 49 161-162.1962

13 Poche R, DeMeiios-Mattos CM, Rembarz HW, et al* Ueber das Verhaeltnis Mrtochondrren Myofrbrrllen in den Herzmu-

skelzellen der Ratte ber Druckhypertrophre des Herzens Vrrchow Arch [Path Anat] 344 100-110, 1968

14 Weber A, Herz E, Reiss I: The role of magnesium in the relaxation of myofrbrris Brochemrstry 8 2266-2270, 1969

15 Page E, McCaiiister LP, Eariey J. Manuscript In preparation 16 Meddoff DA, Page E: Extensions of the cardiac plasma

membrane from the intercalated disk of one cell into the transverse tubules of the adjacent cell J Ultrastruct Res 24 508~521,1968

17 McCaiiister LP, Page E. Effects of thyroxrn on ultrastructure of rat myocardral cells a stereological study J Uitrastruct Res, In press

18 Gross NJ: Control of mrtochondrral turnover under the rnflu- ence of thyroid hormone J Ceil BIOI 48 29-50, 1971

19 Lifschitz MD, Kayne HL Cardiac myofrbrrllar ATPase actrvi- ty In hypophysectomrzed thyrordectomized rats Biochem Pharmacoll5 405-407, 1966


Documents

Quantitative electron microscopic description of heart muscle cells: Application to normal, hypertrophied and thyroxin-stimulated hearts