Medical Physiology, 3rd Edition

From Contractile Filaments to a Regulated Pump

In Chapter 9, we examined the general features of muscle contraction and compared the properties of skeletal, cardiac, and smooth muscle. In this chapter, we examine how some of the features of cardiac muscle underlie cardiac performance.

The entry of Ca2+ from the outside triggers Ca2+-induced Ca2+ release from the sarcoplasmic reticulum

Excitation-contraction (EC) coupling in cardiac ventricular myocytes (see pp. 242 -243) is similar to EC coupling in skeletal muscle (see pp. 229–230). One major difference is that, in the case of skeletal muscle, the initiating event is the arrival of an action potential at the neuromuscular junction, the release of acetylcholine, and the initiation of an end-plate potential. In the ventricular myocyte, action potentials in adjacent myocytes depolarize the target cell through gap junctions (see p. 483) and thereby generate an action potential.

As in a skeletal muscle fiber (see pp. 229–230), the depolarization of the plasma membrane in the ventricular myocyte invades T tubules that run radially to the long axis of the myocyte. Unlike skeletal muscle cells, cardiac myocytes also have axial T tubules that run parallel to the long axis of the cell and interconnect adjacent radial T tubules.

Another major difference in EC coupling between cardiac and skeletal muscle is the way that the L-type Ca2+ channels (Cav1.2, dihydropyridine receptors) in the T-tubule membrane activate the Ca2+-release channels made up of four RYR2 molecules in the sarcoplasmic reticulum (SR) membrane. In skeletal muscle, the linkage is mechanical and does not require Ca2+ entry per se. If you place skeletal muscle in a Ca2+-free solution, the muscle can continue contracting until its intracellular Ca2+ stores become depleted. In contrast, cardiac muscle quickly stops beating. Why?

In cardiac muscle, Ca2+ entry through the L-type Ca2+ channel Cav1.2 (Fig. 22-11, red arrow No. 1) is essential for raising [Ca2+]i in the vicinity of the RYR2 on the SR. A subset of Cav1.2 channels may be part of caveolae. This trigger Ca2+ activates an adjacent cluster of RYRs in concert, causing them to release Ca2+ locally into the cytoplasm by Ca2+-induced Ca2+ release (CICR; see Fig. 22-11, red arrow No. 2). In the CICR coupling mechanism, the action of this Ca2+ is analogous to that of a neurotransmitter or chemical messenger that diffuses across a synapse to activate an agonist-gated channel, but in this case the synapse is the intracellular diffusion gap of ~15 nm between plasma-membrane Cav channels and RYR channels on the SR membrane. The CICR mechanism is a robust amplification system whereby the local influx of Ca2+from small clusters of L-type Cav channels in the plasma membrane triggers the coordinated release of Ca2+ from the high-capacity Ca2+ stores of the SR. Such single CICR events can raise [Ca2+]i to as high as 10 µM in microdomains of ~1 µm in diameter. These localized increases in [Ca2+]i appear as calcium sparksimageN9-3 when they are monitored with a Ca2+-sensitive dye by confocal microscopy. If many L-type Ca2+channels open simultaneously in ventricular myocytes, the spatial and temporal summation of many elementary Ca2+ sparks leads to a global increase in [Ca2+]i. The time course of this global [Ca2+]i increase in ventricular myocytes lasts longer than that of the action potential (compare blue and black curves in inset in Fig. 22-11) because the RYR Ca2+-release channels remain open for a longer time than L-type Ca2+channels.


FIGURE 22-11 Role of Ca2+ in cardiac contraction. The inset pertains to ventricular myocytes. Vm, membrane potential.

Atrial myocytes have a poorly developed T-tubule system. In atrial cells, depolarization-induced activation of L-type Cav channels on the plasma membrane triggers Ca2+ release from RYR channels in the peripheral SR (i.e., closest to the plasma membrane), eliciting subsurface Ca2+ sparks. These peripheral Ca2+ sparks then activate a wave of CICR that propagates inwardly throughout the central SR network of the atrial myocyte.

A global rise in [Ca2+]i initiates contraction of cardiac myocytes

The basic structure of the thin and thick filaments in cardiac muscle is the same as in skeletal muscle (see Fig. 9-5). After [Ca2+]i increases, Ca2+ binds to the cardiac isoform of troponin C (TNNC1; see Table 9-1), and the Ca2+-TNNC1 complex releases the inhibition of the cardiac isoform of troponin I (TNNI3) on actin. As a result, the tropomyosin (TPM1) filaments bound to cardiac troponin T (TNNT2) on the thin filament shift out of the way (see Fig. 9-6), allowing myosin to interact with active sites on the actin. ATP fuels the subsequent cross-bridge cycling (see Fig. 9-7). Because the heart can never rest, cardiac myocytes have a very high density of mitochondria and thus are capable of sustaining very high rates of oxidative phosphorylation (i.e., ATP synthesis).

The cross-bridge cycling causes thick filaments to slide past thin filaments, generating tension. The time course of cardiac tension development is delayed relative to the time course of the global surge in [Ca2+]i(compare red and blue curves in inset of Fig. 22-11).

When we discussed the mechanics of skeletal muscle in Chapter 9, we introduced the concept of a length-tension diagram (see Fig. 9-9C), which is a plot of muscle tension as a function of muscle length. The length parameter in such a plot can be either the length of the whole skeletal muscle or the length of a single sarcomere. For heart muscle, which wraps around the ventricle, the length parameter can be either the ventricular volume, which is analogous to whole-muscle length, or the sarcomere length. The sarcomere, stretching from one Z line to another, is the functional unit in both skeletal and cardiac muscle.

Phosphorylation of phospholamban and of troponin I speeds cardiac muscle relaxation

With the waning of the phase 2 plateau of the cardiac action potential (see Fig. 21-4B, top panel), the influx of Ca2+ through L-type Ca2+ channels decreases, which lessens the release of Ca2+ by the SR. By itself, halting of Ca2+ entry and release can only prevent a further increase in [Ca2+]i. The actual relaxation of the contractile proteins depends on four processes: (1) extrusion of Ca2+ into the extracellular fluid (ECF), (2) reuptake of Ca2+ from the cytosol by the SR, (3) uptake of Ca2+ from the cytosol into the mitochondria, and (4) dissociation of Ca2+ from troponin C. Processes 2 and 4 are highly regulated.

Extrusion of Ca2+ into the ECF

Even during the plateau of the action potential, the myocyte extrudes some Ca2+. After the membrane potential returns to more negative values, Ca2+ extrusion (see Fig. 22-11, green arrow No. 1) gains the upper hand and [Ca2+]i falls. In the steady state (i.e., during the course of several action potentials), the cell must extrude all the Ca2+ that enters the cytosol from the ECF through L-type Ca2+ channels. As in most other cells (see p. 126), this extrusion of Ca2+ into the ECF occurs by two pathways: (1) a sarcolemmal Na-Ca exchanger (NCX1), which operates at relatively high levels of [Ca2+]i; and (2) a sarcolemmal Ca pump (cardiac subtypes 1, 2, and 4 of plasma-membrane Ca ATPase, or PMCA), which may function at even low levels of [Ca2+]i. However, PMCA contributes only modestly to relaxation. Because PMCA is concentrated in caveolae, which contain receptors for various ligands, its role may be to modulate signal transduction.

Reuptake of Ca2+ by the SR

Even during the plateau of the action potential, some of the Ca2+ accumulating in the cytoplasm is sequestered into the SR (see Fig. 22-11, green arrow No. 2) by the cardiac subtype of the sarcoplasmic and endoplasmic reticulum Ca pump SERCA2a (see p. 118). Phospholamban (PLN),imageN22-7 an integral SR membrane protein with a single transmembrane segment, is an important regulator of SERCA2a. In SR membranes of cardiac, smooth, and slow-twitch skeletal muscle, unphosphorylated PLN can exist as a homopentamer that may function in the SR as an ion channel or as a regulator of Cl channels. The dissociation of the pentamer allows the hydrophilic cytoplasmic domain of PLN monomers to inhibit SERCA2a. However, phosphorylation of PLN by any of several kinases relieves PLN's inhibition of SERCA2a, allowing Ca2+ resequestration to accelerate. The net effect of phosphorylation is an increase in the rate of cardiac muscle relaxation. PLN-knockout mice have uninhibited SERCA2a Ca pumps and thus an increased velocity of muscle relaxation.



Contributed by Emile Boulpaep

Phospholamban (PLN or PLB) is a 6-kDa integral membrane protein with 52 amino acids and a single transmembrane domain. The protein kinases that can phosphorylate PLN include PKA, sarcoplasmic reticulum calmodulin kinase (SRCaM kinase—a distinct Ca2+-CaM–dependent protein kinase), and a cGMP-dependent kinase (see pp. 66–67).

As noted in the text, the phosphorylated PLN can exist as a homopentamer that may function in the SR as an ion channel. Structural biology studies of PLN indicate that, at its narrowest point, the pore radius is 1.8 Å.

Phosphorylation of PLN by protein kinase A (PKA) explains why β1-adrenergic agonists (e.g., epinephrine), which act through the PKA pathway (see p. 57), speed up the relaxation of cardiac muscle. Phosphoprotein phosphatase 1 (PP1) dephosphorylates PLN, thereby terminating Ca2+ reuptake.

Uptake of Ca2+ by Mitochondria

The mitochondria take up a minor fraction of the Ca2+ accumulating in the cytoplasm (see Fig. 22-11, green arrow No. 3). The inner mitochondrial membrane contains large-conductance, highly selective Ca2+channels (MiCas) that are inwardly rectifying. At potentials of –160 mV, which are typical for energized mitochondria, the MiCa channels carry a substantial Ca2+ current. Unlike many other Ca2+ channels, MiCa does not inactivate as intramitochondrial [Ca2+] rises to micromolar concentrations.

Dissociation of Ca2+ from Troponin C

As [Ca2+]i falls, Ca2+ dissociates from troponin C (see Fig. 22-11, green arrow No. 4; see p. 233), blocking actin-myosin interactions and causing relaxation. β1-adrenergic agonists accelerate relaxation by promoting phosphorylation of troponin I, which in turn enhances the dissociation of Ca2+ from troponin C.

The overlap of thick and thin filaments cannot explain the unusual shape of the cardiac length-tension diagram

We discussed passive and active length-tension diagrams for skeletal muscle in conjunction with Figure 9-9C and D. We obtain a passive length-tension diagram by holding a piece of resting skeletal or cardiac muscle at several predefined lengths and measuring the tension at each length (Fig. 22-12A, green and violet curves). We obtain the active length-tension diagram by stimulating the muscle at each predefined length (i.e., isometric conditions) and measuring the increment in tension from its resting or passive value (see Fig. 22-12A, turquoise and brown curves).


FIGURE 22-12 Length-tension diagram. Compared with A, in B EDV on the x-axis is used as an index of sarcomere length. (Because EDV was difficult to measure before the days of echocardiography, Starling actually used left atrial pressure as an index of the degree of filling.) Starling measured pressure on the y-axis as an index of tension. Thus, systolic pressure replaces active tension, and diastolic pressure replaces passive tension. In C, left atrial pressure on the x-axis is used as an index of sarcomere length, and stroke work (systolic pressure × ejected volume) on the y-axis is used instead of tension.

The passive length-tension diagrams for skeletal and cardiac muscle are quite different. The passive tension of a skeletal muscle (see Fig. 22-12A, green curve) is practically nil until the length of the sarcomere exceeds 2.6 µm. Beyond this length, passive tension rises slowly. On the other hand, the passive tension of cardiac muscle (see Fig. 22-12A, violet curve) begins to rise at much lower sarcomere lengths and rises much more steeply. As a result, cardiac muscle will break if it is stretched beyond a sarcomere length of 2.6 µm, whereas it is possible to stretch skeletal muscle to a sarcomere length of 3.6 µm.

The reason for the higher passive tension is that the noncontractile (i.e., elastic) components of cardiac muscle are less distensible. The most important elastic component is the giant protein titin (see p. 234), which acts as a spring that provides the opposing force during stretch and the restoring force during shortening (see Fig. 9-4B).

The active length-tension diagrams also differ between skeletal and cardiac muscle. The active tension of skeletal muscle (see Fig. 22-12A, turquoise curve) is high and varies only modestly between sarcomere lengths of 1.8 and 2.6 µm. In Chapter 9, we attributed the shape of this curve to the degree of myofilament overlap (see pp. 238–240). In cardiac muscle (see Fig. 22-12A, brown curve), active tension has a relatively sharp peak when the muscle is prestretched to an initial sarcomere length of ~2.4 µm. As the prestretched sarcomere length increases from 1.8 to 2.4 µm, active tension rises steeply. We cannot account for this rise by an increase in the overlap of thick and thin filaments because the filament dimensions of cardiac and skeletal muscle are similar. Rather, the rise in tension at longer sarcomere lengths in cardiac muscle probably has two general causes: (1) Raising the sarcomere length above 1.8 µm increases the Ca2+ sensitivity of the myofilaments. One mechanism controlling the Ca2+ sensitivity may be interfilament spacing between thick and thin filaments, because fiber diameter varies inversely with fiber length. As we stretch the muscle to greater sarcomere lengths, the lateral filament lattice spacing is less than in an unstretched fiber so that the probability of cross-bridge interaction increases. Increased cross-bridge formation in turn increases the Ca2+ affinity of TNNC1, thereby recruiting more cross-bridges and therefore producing greater force. Another mechanism could be that, as the muscle elongates, increased strain on titin either alters lattice spacing or alters the packing of myosin molecules within the thick filament. (2) Raising the sarcomere length above 1.8 µm increases tension on stretch-activated Ca2+ channels, thereby increasing Ca2+ entry from the ECF and thus enhancing Ca2+-induced Ca2+ release.

As cardiac sarcomere length increases above 2.4 µm, active tension declines precipitously, compared with the gradual fall in skeletal muscle. Once again, this fall-off does not reflect a problem in the overlap of thin and thick filaments. Instead, titin increases the passive stiffness of cardiac muscle and may also impede development of active tension at high sarcomere lengths.

Starling's law states that a greater fiber length (i.e., greater ventricular volume) causes the heart to deliver more mechanical energy

Long before the development of the sliding-filament hypothesis and our understanding that active tension should depend on sarcomere length, Ernest Starling in 1914 anticipated the results of Figure 22-12A using an isolated heart-lung preparation. imageN22-10Starling's law states that “the mechanical energy set free on passage from the resting to the contracted state depends on the area of ‘chemically active surfaces,’ i.e., on the length of the fibres.” Therefore, the initial length of myocardial fibers determines the work done during the cardiac cycle. Figure 22-12B shows the results of experiments that Starling performed on the intact heart. Starling assumed that the initial length of the myocardial fibers is proportional to the end-diastolic volume. Further, he assumed that tension in the myocardial fibers is proportional to the systolic pressure. Therefore, starting from a volume-pressure diagram, Starling was able to reconstruct an equivalent length-tension diagram (Table 22-4).

TABLE 22-4

Equivalent Units for Converting Between a Three-Dimensional Heart and a Linear Muscle Fiber











Extent of shortening


Stroke volume


Velocity of shortening


Velocity of ejection





mm Hg

or force



or tension



This table shows the equivalence between dimensions when converting from the contraction of a hollow organ, such as a cardiac ventricle, to a linear model of a single muscle contraction.


Starling’s Law

For the original description of Starling’s law, see the landmark paper: Patterson SW, Piper H, Starling EH: The regulation of the heart beat. J Physiol 48:465-513, 1914.

His diagram for diastole (see Fig. 22-12B, purple curve), which shows a rising pressure (tension) with increased EDV (fiber length), is very similar to the early part of the passive length-tension diagram for cardiac muscle (see Fig. 22-12A, violet curve). His diagram for systole (see Fig. 22-12B, red curve) is more or less equivalent to the ascending phase of the active length-tension diagram for cardiac muscle (see Fig. 22-12A, brown curve). Therefore, Starling's systole curve shows that the heart is able to generate more pressure (i.e., deliver more blood) when more is presented to it.

ventricular performance curve (see Fig. 22-12C) is another representation of Starling's length-tension diagram, but it is one a clinician can obtain for a patient. A ventricular performance curve shows stroke work (P · ΔV, see Equation 22-4), which includes Starling's systolic pressure (itself an estimate of muscle tension) on the y-axis, plotted against left atrial pressure, which corresponds to Starling's end-diastolic volume (itself an estimate of muscle length), on the x-axis. What we learn from performance curves obtained for living subjects is that Starling's law is not a fixed relationship. For instance, the norepinephrine released during sympathetic stimulation—which increases myocardial contractility (as we will see below in this chapter)—steepens the performance curve and shifts it upward and to the left (see brown arrow in Fig. 22-12C). Similar shifts occur with other positive inotropic agents (e.g., cardiac glycosides), that is, drugs that increase myocardial contractility. Note also that ventricular performance curves show no descending component because sarcomere length does not increase beyond 2.2 to 2.4 µm in healthy hearts.

The velocity of cardiac muscle shortening falls when the contraction occurs against a greater opposing force (or pressure) or at a shorter muscle length (or lower volume)

The functional properties of cardiac muscle—how much tension it can develop, how rapidly it can contract—depend on many factors, but especially on two properties intrinsic to the cardiac myocyte.

1. Initial sarcomere length. For the beating heart, a convenient index of initial sarcomere length is EDV. Both initial sarcomere length and EDV are measures of the preload imposed on the cardiac muscle just before it ejects blood from the ventricle during systole. Starling's law, in which the independent variable is EDV, focuses on preload.

2. Force that the contracting myocytes must overcome. In the beating heart, a convenient index of opposing force is the arterial pressure that opposes the outflow of blood from the ventricle. Both opposing force and arterial pressure are measures of the afterload the ventricular muscle must overcome as it ejects blood during systole. Experiments on isotonic contractions focus on the afterload, factors that the ventricle can sense only after the contraction has begun.

Figure 22-13 shows how one might measure the velocity of shortening in a way that is relevant for a cardiac muscle facing both a preload and an afterload. In Figure 22-13A, the muscle starts off at rest, stretched between a fixed support (bottom of the muscle) and the left end of a lever (top of the muscle). A weight attached to the other end of the lever, but resting on a table, applies stretch to the muscle to the extent allowed by the screw, which adjusts the “stop” of the lever's left end. Thus, the combination of the weight and the screw determines initial sarcomere length (i.e., preload). At this time, the muscle cannot sense the full extent of the weight. The more we stretch the muscle by retracting the screw, the greater the preload. When we begin to stimulate the muscle, it develops a gradually increasing tension (see Fig. 22-13C, lower blue curve), but the length between the fixed point and the left end of the lever (see Fig. 22-13A) remains constant. That is, the muscle cannot shorten. Therefore, in the first phase of the experiment, the muscle exerts increasing isometric tension.

When the muscle has built up enough tension, it can now begin lifting the weight off the table (see Fig. 22-13B). This phase of the contraction is termed afterloaded shortening. The tension now remains at a fixed afterload value (flat portion of lower curve in Fig. 22-13C) but the muscle gradually shortens (rising portion of upper blue curve). Therefore, in the second phase of the experiment, the muscle exerts isotonic contraction. From the slope of the upper curve in Figure 22-13C, we can compute the velocity of shortening at a particular afterload.

This experiment roughly mimics the actions of ventricular muscle during systole. Initially, during its isometric contraction, our hypothetical muscle increases its tension at constant length, as during the isovolumetric contraction of the cardiac cycle shown in segment CD of Figure 22-9. The initial length corresponds to EDV, the preload. Later, the muscle shortens while overcoming a constant force (i.e., generating a constant tension), as during the ejection phase of the cardiac cycle shown in segment DEF of Figure 22-9. The tension corresponds to arterial pressure, the afterload.

What happens if we vary the afterload (i.e., change the weight)? As we already observed in our discussion of skeletal muscle, it is easier to lift a feather than a barbell. Thus, with a heavier weight, the muscle develops a lot of tension, but shortens slowly (see Fig. 22-13D, red tracings). Conversely, with a lighter weight, the muscle develops only a little bit of tension, but shortens rapidly (purple curves).

If we plot the velocities of shortening in Figure 22-13D as a function of the three different afterloads being lifted, we obtain the purple, blue, and red points on the load-velocity curve in Figure 22-13E. The velocity of muscle shortening corresponds to the outflow velocity of the ventricle (see Fig. 22-8D, E). Thus, at higher opposing arterial pressures the outflow velocity should decrease. The black curve in Figure 22-13E applies to a muscle that we stretched only slightly in the preload phase (i.e., low preload in Fig. 22-13A). The red curve in Figure 22-13E shows a similar load-velocity relationship for a muscle that we stretched greatly in the preload phase (i.e., high preload). In both cases, the velocity of shortening increases as the tension (i.e., afterload) falls.

When the afterload is so large that no shortening ever occurs, that afterload is the isometric tension, shown as the point of zero velocity on the x-axis of Figure 22-13E. As expected from Starling's law, the greater the initial stretch (i.e., preload), the greater the isometric tension. In fact, at any velocity (see Fig. 22-13E, dashed horizontal line), the tension is greater in the muscle that was stretched more in the preload phase (red curve)—a restatement of Starling's law.

In summary, at a given preload (i.e., walking up the black curve in Fig. 22-13E), the velocity of shortening for cardiac muscle becomes greater with lower afterloads (i.e., opposing pressure). Conversely, at a given afterload—that is, comparing the black and red curves for any common x value (see Fig. 22-13E, dashed vertical line)—the velocity of shortening for cardiac muscle becomes greater with a greater preload (i.e., sarcomere length).

Finally, the curves in Figure 22-13E do not represent a fixed set of relationships. Positive inotropic agents shift all curves up and to the right. Thus, a positive inotropic agent allows the heart to achieve a given velocity against a greater load, or to push a given load with a greater velocity.

Another way of representing how velocity of shortening depends on the initial muscle length (i.e., preload) is to monitor velocity of shortening during a single isotonic contraction. If we first apply a large preload to stretch a piece of muscle to an initial length of 9.0 mm (see Fig. 22-13F) and then stimulate it, the velocity instantly rises to a peak value of ~8.5 mm/s; it then gradually falls to zero as the muscle shortens to 7.5 mm. If we start by applying a smaller preload, thereby stretching the muscle to an initial length of 8.5 or 8.0 mm, the peak velocity falls. Thus, initial length determines not only the tension that cardiac muscle can generate, but also the speed with which the muscle can shorten.

Increases in heart rate enhance myocardial tension

Heart muscle tension has a special dependence on the frequency of contraction. If we stimulate isolated heart muscle only a few times per minute, the tension developed is much smaller than if we stimulate it at a physiological rate of 70 times per minute. The progressive rise of tension after an increase in rate—the positive staircase phenomenon—was first observed by Henry Bowditch in 1871. Underlying the staircase phenomenon is an increase in SR Ca2+ content and release. The larger SR Ca2+ content has three causes. First, during each action potential plateau, more Ca2+ enters the cell through Cav1.2 L-type Ca2+ channels, and the larger number of action potentials per minute provides a longer aggregate period of Ca2+ entry through these channels. Second, the depolarization during the plateau of an action potential causes the Na-Ca exchanger NCX1 to operate in the reverse mode, imageN21-2 allowing Ca2+ to enter the cell. At higher heart rates, these depolarizations occur more frequently and are accompanied by an increase in [Na+]i, which accentuates the reversal of NCX1, both of which enhance Ca2+ uptake. Third, the increased heart rate stimulates SERCA2a, thereby sequestering in the SR the Ca2+ that entered the cell because of the first two mechanisms. The mechanism of this stimulation is that the rising [Ca2+]i, through calmodulin (CaM), activates CaM kinase II, which leads to phosphorylation of PLN; phosphorylation of PLN in turn enhances SERCA2a.

Contractility is an intrinsic measure of cardiac performance

Now that we know that the performance of the heart depends on such factors as degree of filling (i.e., preload), arterial pressure (i.e., afterload), and heart rate, it would be useful to have a measure of the heart's intrinsic contractile performance, independent of these extrinsic factors. Contractility is such a measure.

Contractility is a somewhat vague but clinically useful term that distinguishes a better-performing heart from a poorly performing one. In a patient, imageN22-8 it is difficult to assess cardiac performance by use of the approaches illustrated in Figures 22-12 and 22-13. One clinically useful measure of contractility is the ejection fraction (see p. 519). However, according to Starling's law, ejection depends on EDV (i.e., preload), which is external to the heart. Two somewhat better gauges of contractility are the rate of pressure development during ejection (ΔPt) and the velocity of ejection. Both correlate well with the velocity of shortening in Figure 22-13E and F, and they are very sensitive guides to the effect of inotropic interventions.


Assessment of Contractility in Patients

Contributed by Emile Boulpaep

For two reasons, it is not practical to assess cardiac performance in a patient by using the approaches outlined in Figures 22-12 and 22-13. First, with patients, we do not deal with isolated muscles in vitro. Second, the aforementioned figures require that we study the muscle under the artificial conditions of only isometric (i.e., preloaded) or only isotonic (i.e., afterloaded) contractions. During a full cardiac cycle, of course, these conditions alternate.


FIGURE 22-13 Effect of preload and afterload on velocity of shortening. In A, the developed tension is not yet sufficient to lift the weight (i.e., afterload). In B, the muscle, which has now developed sufficient tension to lift the weight, shortens against a constant afterload. In C, the slope of the blue curve (ΔLt) is the velocity of shortening. The velocities of shortening for three different afterloads (tensions) in D are plotted as the three colored points of the lower curve in E. In F, the x-axis has the longest lengths on the left, so that “time” runs from left to right (arrows). Note that the family of curves is enclosed by the envelope created by the curve for the greatest initial length.

A third assessment of contractility focuses on the physiological relationship between pressure and volume during the cardiac cycle. In the era of echocardiography, these volume data are now reasonably easy to obtain. We return to the ventricular pressure-volume loop that we introduced in Figure 22-9 and redraw it as the purple loop in Figure 22-14A. In this example, the EDV is 120 mL. Point D′ on the loop represents the relationship between pressure and volume at the end of the isovolumetric contraction, when the aortic valve opens. If we had prevented the aortic valve from opening, ventricular pressure would have continued to rise until the ventricle could generate no additional tension. In this case, the pressure would rise to point G′, the theoretical maximum isovolumetric pressure. We could repeat the measurement at very different EDVs by decreasing or increasing the venous return. Point G represents the maximum isovolumetric pressure for an EDV below 120 mL (orange loop), and point G″ represents this pressure for an EDV above 120 mL (green loop). The gold dashed line through points G, G′, and G″ in Figure 22-14A would describe the relationship between pressure and EDV under isometric conditions (i.e., aortic valve closed)—the equivalent of an isometric Starling curve (e.g., brown curve in Fig. 22-12A). The steeper this line, the greater the contractility.


FIGURE 22-14 Assessment of contractility by the use of a ventricular pressure-volume loop. The purple pressure-volume loop is the normal curve in Figure 22-9. In A, at the same normal state of cardiac contractility, the red loop is generated by decreasing EDV, and the green loop is generated by increasing EDV. The slope of the line through the points at the end of systole (F, F′, and F″) represents the ESPVR.

It is impossible to measure maximum isovolumetric pressures in a patient because it is hardly advisable to prevent the aortic valve from opening. However, we can use the end-systolic pressure at point F′ on the normal pressure-volume loop with an EDV of 120 mL (purple loop in Fig. 22-14A). For an EDV below 120 mL (orange loop), the corner point would slide down and to the left (point F). Conversely, for an EDV above 120 mL (green loop), the corner point would slide upward and to the right (point F″). The corner points of many such pressure-volume loops fall along a line—the end-systolic pressure-volume relation (ESPVR)imageN22-9—that is very similar to that generated by the points G, G′, and G″.


Using the End-Systolic Pressure-Volume Relation in Lieu of an Isometric Starling Curve

Contributed by Emile Boulpaep

The ESPVR is a load-insensitive index of left ventricular contractility. This relation has been measured in small animal species using a conductance-catheter technique—whereby several electrodes along a catheter are used to compute the electrical conductance, which can be converted to absolute ventricular volume—and indwelling pressure gauges. Sato and colleagues (1998) tested the ESPVR under different contractility states: baseline condition, after sympathectomy, and after β blockade in rats. The general slope of the curve decreased with decreasing contractility; that is, modestly after sympathectomy and strongly after β blockade. The curve was not always perfectly straight. The ESPVR was slightly convex toward the pressure axis under baseline conditions, linear under sympathectomy conditions, and slightly concave toward the pressure axis under β-blockade conditions.


Feldman MD, Mao Y, Valvano JW, et al. Development of a multifrequency conductance catheter-based system to determine LV function in mice. Am J Physiol Heart Circ Physiol. 2000;279:H1411–H1420.

Georgakopoulos D, Kass DA. Estimation of parallel conductance by dual-frequency conductance catheter in mice. Am J Physiol Heart Circ Physiol. 2000;279:H443–H450.

Ito H, Takaki M, Yamaguchi H, et al. Left ventricular volumetric conductance catheter for rats. Am J Physiol Heart Circ Physiol. 1996;270:H1509–H1514.

Kubota T, Mahler CM, McTiernan CF, et al. End-systolic pressure-dimension relationship of in situ mouse left ventricle. J Mol Cell Cardiol. 1998;30:357–363.

Sato T, Shishido T, Kawada T, et al. ESPVR of in situ rat left ventricle shows contractility-dependent curvilinearity. Am J Physiol Heart Circ Physiol. 1998;274:H1429–H1434.

Uemura K, Kawada T, Sugimachi M, et al. A self-calibrating telemetry system for measurement of ventricular pressure-volume relations in conscious, freely moving rats. Am J Physiol Heart Circ Physiol. 2004;287:H2906–H2913.

Effect of Changes in Contractility

The ESPVR is a clinically useful measure of contractility. Enhancing the contractility increases the slope of the ESPVR line, just as it increases the steepness of the ventricular performance curves (see brown arrow in Fig. 22-12C). For example, imagine that—with the same EDV and aortic pressure as in the control situation (purple area and gold ESPVR line in Fig. 22-14B)—we increase contractility. We represent increased contractility by steepening the ESPVR line (from gold to red dashed line in Fig. 22-14B). The result is that ejection continues from point D′ to a new point F (red loop in Fig. 22-14B) until the left ventricular volume reaches a much lower value than normal. In other words, enhanced contractility increases stroke volume. Decreasing contractility would flatten the slope of the ESPVR and decrease stroke volume.

Effect of Changes in Preload (i.e., Initial Sarcomere Length)

A pressure-volume loop nicely illustrates the effect of increasing preload (i.e., increasing filling or EDV) without changing contractility. Starting from the control situation (see Fig. 22-14C, purple area), increase of the EDV shifts the isovolumetric segment to the right (segment CD on the red loop). Because the volume change along segment DEF is larger than for the control situation, stroke volume increases—as predicted by Starling's law.

Effect of Changes in Afterload

A pressure-volume loop also illustrates the effect of an increased afterload (i.e., increase in aortic pressure). Starting from the control situation (see Fig. 22-14D, purple area), an increase of aortic pressure shifts the upper right corner of the loop from point D′ (purple loop) to D (red loop) because the ventricle cannot open the aortic valve until ventricular pressure reaches the higher aortic pressure. During the ejection phase—assuming that contractility (i.e., slope of the ESPVR) does not change—the ventricle necessarily ejects less blood until segment DEF intersects the ESPVR line. Therefore, an increase in afterload (at constant contractility) causes the loop to be taller and narrower, so that stroke volume and ejection fraction both decrease. However, if we were to increase contractility (i.e., increase the slope of the ESPVR), we could return the stroke volume to normal.

Positive inotropic agents increase myocardial contractility by raising [Ca2+]i

Modifiers of contractility can affect the dynamics of cardiac muscle contraction independent of preload or afterload. These factors have in common their ability to change [Ca2+]i. When these factors increase myocardial contractility, they are called positive inotropic agents. When they decrease myocardial contractility, they are called negative inotropic agents (Boxes 22-3 and 22-4).

Box 22-3

Cardiac Hypertrophy

Either volume overload or pressure overload can mechanically compromise the heart. A volume overload is an excessive EDV (i.e., preload). For example, a large AV shunt would volume overload both the left and right sides of the heart. The increased EDV leads to an increase in stroke volume (see Fig. 22-14C), which elevates cardiac output. Systemic arterial pressure usually remains normal. A pressure overload is an excessive pressure in the ventricle's outflow tract (i.e., afterload). For the left heart, the problem would be an increase in systemic arterial pressure (i.e., hypertension). The increased aortic pressure leads to a decrease in stroke volume (see Fig. 22-14D). However, because of a compensatory increase in heart rate, cardiac output usually remains normal. When, over time, the adaptive process of hypertrophy becomes inadequate to cope with demand, the result is mechanical dysfunction and, ultimately, heart failure (see Box 22-4).

Because cells of the adult heart are terminally differentiated, stimuli that might be mitogenic in other cells cannot elicit cell division in the heart, but rather cause the cardiac myocytes to hypertrophy and increase muscle mass. Elite athletes develop physiological hypertrophy, in which the cardiac cells increase proportionally both in length and in width. Volume overload leads to eccentric hypertrophy characterized by increases in myocyte length out of proportion to width. Pressure overload causes concentric hypertrophy with a relatively greater increase in myocyte width.

A host of events may trigger hypertrophy, including various hypertrophic factors, increases in [Ca2+]i, and mechanical forces.

Hypertrophic Factors

Agents implicated in cardiac hypertrophy include the cardiac cytosolic protein myotrophin (Myo/V1) and the cytokine cardiotrophin 1 (CT-1), as well as catecholamines, angiotensin II (ANG II), endothelin 1, insulin-like growth factor 2, transforming growth factor-β, and interleukin-1. Catecholamines and ANG II both activate the mitogen-activated protein kinase (MAPK) cascade (see pp. 68–69). Farther downstream in the signal-transduction pathway, the transcriptional response to hypertrophic stimuli includes the zinc-finger transcription factor, GATA-4 (see pp. 80–81), and perhaps also the transcription factors SRF and Sp1, as well as the TEF-1 family.


Elevated [Ca2+]i may be both a trigger for hypertrophy and part of signal-transduction pathways that lead to hypertrophy. [Ca2+]i in heart cells is probably elevated initially during chronic volume or pressure overloads, just as [Ca2+]i would be elevated in a normal heart that is working hard. Elevated [Ca2+]i may activate calcineurin, a Ca2+-dependent phosphatase (see p. 58). After being dephosphorylated by calcineurin, the transcription factor NFAT3 can enter the nucleus and bind to GATA4 (see above), which transcriptionally activates genes responsible for hypertrophy. Mice that express constitutively activated forms of calcineurin develop cardiac hypertrophy and heart failure.

Mechanical Factors

Mechanical stretch induces the expression of specific genes. The mechanical sensor that triggers cardiac hypertrophy may be MLP (muscle LIM protein), part of the myocardial cytoskeleton. Stretch activates a phosphorylation cascade of protein kinases: Raf-1 kinase, extracellular signal–regulated kinase (ERK), and a separate subfamily of the MAPKs called SAPKs (for stretch-activated protein kinases). These various kinases regulate gene expression by activating the transcription factor AP-1 (see pp. 81–82).

The pathways we have just discussed lead to several changes in gene expression within cardiac myocytes during hypertrophy. In addition to synthesizing many housekeeping proteins, hypertrophic cardiac myocytes undergo other changes that are more specific for contraction. Some of the most striking changes include reduced levels of the mRNA encoding three critical proteins in the membrane of the SR: (1) the Ca2+-release channel, (2) PLN, and (3) SERCA2. In addition, cardiac hypertrophy is associated with increased levels of mRNA for the skeletal α-actin, which is normally expressed in fetal heart, but not in adult heart. Hypertrophic hearts also have increased expression of the angiogenic factor vascular endothelial growth factor (VEGF; see p. 481).

Although a hypertrophied myocardial cell may be able to do more work than a nonhypertrophied cell, it has a lower “contractility” when normalized to its cross-sectional area. Why should hypertrophied cardiac muscle not be as good as normal muscle? Possibilities include alterations in the transient increases in [Ca2+]i during the cardiac action potential and alterations in the expression of the contractile filaments, particularly the myosin isoenzymes.

Box 22-4

Cellular Basis of Heart Failure

Heart failure is among the most common causes of hospitalization in developed countries for people aged 65 years or older and is a leading cause of death. People whose hearts cannot sustain an adequate cardiac output become breathless (because blood backs up from the left side of the heart into the lungs) and have swollen feet and ankles (because blood backs up from the right side of the heart and promotes net filtration in systemic capillaries; see Box 20-1). On the cellular level, decreased contractility in heart failure could be a result of cardiac hypertrophy (see Box 22-3), reflecting alterations in the transient increases of [Ca2+]i, the expression of the contractile filaments, or both.

Changes in [Ca2+]i physiology could reflect altered properties of the Cav1.2 L-type Ca2+ channel in the plasma membrane or the Ca2+-release channel RYR2 in the SR membrane. In an animal model of hypertension-induced cardiac hypertrophy that leads to heart failure, the Cav1.2 channels exhibit an impaired ability to activate RYR2 through Ca2+-induced Ca2+ release. A distortion of the microarchitecture in hypertrophic cells, and thus a distortion of the spacing between Cav1.2 channels and RYR2, could be responsible for impaired coupling. Each of the four RYR2 molecules in the Ca2+-release channel associates with a molecule of calstabin 2 (also known as the FK506-binding protein, FKBP12.6) that, together with other proteins, forms a macromolecular complex regulating the Ca2+-release channel. Depletion of calstabin 2 in heart failure results in leaky RYR2 channels that continually release Ca2+ into the cytosol. High [Ca2+]i makes the heart prone to delayed afterdepolarizations (see p. 506), ventricular arrhythmias, and sudden death. imageN21-16

Changes in the expression of contractile proteins can reduce contractility. Two isoforms of myosin heavy chain, αMHC and βMHC, are present in the heart (see Table 9-1). The speed of muscle shortening increases with the relative expression of αMHC. In human heart failure, the amount of αMHC mRNA, relative to total MHC mRNA, falls from ~35% to ~2%.

An interesting animal model of heart failure is the knockout mouse that lacks the gene encoding MLP, the muscle LIM protein (see pp. 530–531). MLP-deficient mice have the same disrupted cytoskeletal architecture seen in failing hearts. In addition, these mice have a dilated cardiomyopathy. Although humans with failing hearts are generally not deficient in MLP, the evidence from these knockout mice suggests that the MLP system could play a role in certain forms of cardiomyopathy.

Positive Inotropic Agents

Factors that increase myocardial contractility increase [Ca2+]i, either by opening Ca2+ channels, inhibiting Na-Ca exchange, or inhibiting the Ca pump—all at the plasma membrane.

1. Adrenergic agonists. Catecholamines (e.g., epinephrine, norepinephrine) act on β1 adrenoceptors to activate the α subunit of Gs-type heterotrimeric G proteins. The activated αs subunits produce effects by two pathways. First, αs raises intracellular levels of cAMP and stimulates PKA (see p. 57), which can then act by the mechanisms summarized in Table 23-2 to increase contractility and speed relaxation. Second, αscan directly open L-type Ca2+ channels in the plasma membrane, which leads to an increased Ca2+ influx during action potentials, increased [Ca2+]i, and enhanced contractility.

2. Cardiac glycosides. Digitalis derivatives inhibit the Na-K pump on the plasma membrane (see p. 117) and therefore raise [Na+]i. We would expect the increased [Na+]i to slow down the Na-Ca exchanger NCX1, to raise steady-state [Ca2+]i, and to enhance contractility. Recent evidence suggests that cardiac glycosides may also increase [Ca2+]i by a novel pathway—increasing the Ca2+ permeability of Na+ channels in the plasma membrane.

3. High extracellular [Ca2+]. Acting in two ways, elevated [Ca2+]o increases [Ca2+]i and thereby enhances contractility. First, it decreases the exchange of external Na+ for internal Ca2+. Second, more Ca2+ enters the myocardial cell through L-type Ca2+ channels during the action potential.

4. Low extracellular [Na+]. Reducing the Na+ gradient decreases Ca2+ extrusion through NCX1, raising [Ca2+]i and enhancing contractility.

5. Increased heart rate. As we noted in introducing the staircase phenomenon (see p. 528), an increased heart rate increases SR stores of Ca2+ and also increases Ca2+ influx during the action potential.

Negative Inotropic Agents

Factors that decrease myocardial contractility all decrease [Ca2+]i.

1. Ca2+-channel blockers. Inhibitors of L-type Ca2+ channels (see pp. 189–192)—such as verapamil, diltiazem, and nifedipine—reduce Ca2+ entry during the plateau of the cardiac action potential. By reducing [Ca2+]i, they decrease contractility.

2. Low extracellular [Ca2+]. Depressed [Ca2+]o lowers [Ca2+]i, both by increasing Ca2+ extrusion through NCX1 and by reducing Ca2+ entry through L-type Ca2+ channels during the plateau of the cardiac action potential.

3. High extracellular [Na+]. Elevated [Na+]o increases Ca2+ extrusion through NCX1, thereby decreasing [Ca2+]i.