In this review, we discuss the three major mechanisms (Frank-Starling mechanism, heart rate/frequency-dependent contraction, and beta-adrenergic stimulation) that govern cardiac output as well as affect calcium sensitivity, mainly at the level of troponin C, compare calcium sensitivity measurements in skinned/permeabilized and intact muscle preparations, in order to shed light on myofilament protein mutations that have the potential to be translated to further our understanding of cardiac physiology in vivo. We recognize that drugs, metabolites, pH, etc., also critically impact on calcium sensitivity (and potentially cardiac output), but that these are typically not primarily resulting from sarcomeric mutations, and were deemed beyond the scope of this review.

At the beginning of the cardiac contraction cycle, calcium ions enter cardiomyocytes via voltage-activated L-type calcium channels, leading to calcium-induced calcium release (CICR) from the sarcoplasmic reticulum (SR). The release of calcium from the SR increases free calcium ion concentration from approximately 100 nM to 1 μM, making more calcium available for binding to troponin C (TnC), a subunit in the troponin complex. Calcium ions binding to TnC initiates a cascade of events leading to force generation via interaction between thin and thick filaments, i.e., by the cycling of cross-bridges. In order to relax, calcium must come off TnC to cease activation, and to allow dissociation of thin and thick filaments to occur and relax the muscle. Calcium ions are recycled back into the SR via SR Ca²⁺ ATPase (SERCA) or extruded out of the cell via Na⁺/Ca²⁺ exchanger (NCX).

Myofilament calcium sensitivity is a concept used by researchers to simplify the complex, dynamic process of cardiac contraction, and relaxation into a relationship between the concentration of free calcium ions available for binding to TnC and the amount of force generated by the muscle. In failing myocardium, calcium sensitivity has been reported to either increase or decrease depending on the etiology of the disease (Willott et al., 2010). As the calcium sensitivity increases, the contractility of the muscle typically increases, but this also means that relaxation may be often impaired if calcium dissociates from TnC more slowly. The vast majority of previous studies have utilized mechanically and chemically skinned muscle preparations for measuring calcium sensitivity because they are able to reduce a complex system into one that only contains two variables: free calcium ions and force of contraction by the myofilaments. This reductionist approach has revealed important mechanical properties of the thin and thick filaments but does not sufficiently translate into the contracting heart. The force-generating capacity of cardiac muscle in vivo takes into account not only the simple association and dissociation rate of calcium from TnC but the entire intracellular environment that includes various kinases and phosphatases, for example. From previous studies, it is clear that the myofilaments play an integral role in cardiac muscle contraction and relaxation. Therefore, the myofilaments are an important target in treatment of heart failure, which continues to afflict millions of lives today with limited treatment options. It is imperative that we utilize the invaluable knowledge the cardiac muscle physiology field has already generated regarding calcium sensitivity and produce new data to not only further our understanding of the physiology of a dynamically contracting heart in vivo but also more effectively translate our findings to the clinic.

A typical approach to assess myofilament calcium sensitivity is via construction of a force-pCa curve and determining a potential left- or right-ward shift of the curve (Figure 1). A left-ward shift indicates an increased calcium sensitivity, as a given steady-state force can be attained using a lower concentration of free calcium. On the other hand, a right-ward shift indicates a decreased calcium sensitivity, as a muscle requires a higher concentration of free calcium to generate a given steady-state force. A deeper insight into this steady-state model reveals that, while a change in myofilament calcium sensitivity can reflect altered dynamic behavior, one must also know at least one additional parameter to do so. The equilibrium dissociation constant, K_d, of TnC is a ratio between the calcium association rate constant to TnC (k_on) and the calcium dissociation rate constant from TnC (k_off) (Figure 1). TnC however does not work in isolation (Davis and Tikunova, 2008; Biesiadecki et al., 2014). There are many factors that collaboratively change the sensitivity of the myofilament activation and deactivation by calcium. No current models fully explain the complex integration of all components on the governing of thin-filament calcium binding (see Siddiqui et al., 2016). Thus, for the remainder of this review, we will discuss on on-rate (k_on) and off rate (k_off), as the apparent on- and off-rates of the myofilament system, reflecting the effective on- and off-rates of myofilament activation and deactivation not necessarily reflecting solely Ca²⁺ binding to TnC.

Myofilament calcium sensitivity increases when the k_on increases relative to the k_off, resulting in an overall decrease in K_d. In other words, the k_on does not necessarily have to increase to increase TnC's calcium sensitivity. As long as the k_off decreases by a larger percentage compared to the k_on, one would observe an increase in calcium sensitivity. This is an important distinction because having an absolute increase in the k_on would lead to increased activation of the myofilament and thus increased force generation in our model of cardiac muscle twitch (Figure 2). Our model is written in Labview (National Instruments) and uses a simple mathematical calcium transient: [Ca²⁺]_i = Amplitude^*time^*e^∧(-Downamplitude^*time/τ). This calcium transient (light blue trace in both Figures 2, 3) with kinetic parameters that reflect literature values drives, via on and off rate, the thin filament activation level (reflecting TnC-Ca²⁺ binding). This thin filament activation allows cross-bridge formation using the simple 2 state model, governed by an on-rate (f), and an off-rate (g). The model incorporates cross-bridge attachment and detachment rates and thin filament activation levels to generate twitches in real time. Our program allows us to change various parameters such as temperature, calcium transient relaxation constant, cross-bridge attachment rate, and cross-bridge detachment rate. In all the used simulations in this review (Figures 2, 3), all parameters other than the calcium-TnC k_on and k_off rates were kept constant. Our model reports various twitch kinetic parameters such as time-to-peak (TTP), and relaxation to 50% (RT50) in real time. We generated our cardiac twitches by initially changing these parameters to best mimic the typical cardiac muscle twitch kinetics we have observed in intact human trabeculae (Milani-Nejad et al., 2015). It is believed that the contraction kinetics of a muscle are much slower than the k_on, which has traditionally been believed to be diffusion-limited, and that changes in the k_on do not affect the contraction kinetics (Bers, 2001; Davis and Tikunova, 2008). Therefore, an increase in the k_on would result in an increase in developed tension but not necessarily in faster contraction kinetics. On the other hand, having an absolute decrease in the k_off would result in slowed relaxation (Figure 3). The rate-limiting steps of relaxation kinetics are complex and involve several distinct processes that at least partially overlap in time. As reviewed in previous literature, the main processes involved are thought to be the decline of the intracellular calcium concentration, transient calcium coming off TnC (k_off), and cross-bridge cycling kinetics (Biesiadecki et al., 2014). Therefore, a decrease in the k_off of TnC could slow down relaxation kinetics. An increase in the k_on or a decrease in the k_off would culminate in an increased calcium sensitivity but would have drastically different effects on dynamic twitch kinetics.

Despite the unchanged calcium sensitivity, significant consequences for dynamic contraction can occur when the k_on and k_off increase or decrease by the same factor (Figure 3). This would result in no apparent change in steady-state calcium sensitivity, however, if the k_on and k_off both increase, the muscle would develop a higher force and relax faster. If the k_on and k_off both decrease, the muscle would develop a lower force and relax slower. Again, this type of hypothetical analysis reveals that a second important reason that simply measuring the myofilament calcium sensitivity as a stand-alone measurement is not sufficient for translation into dynamic contraction and relaxation.

Another way to determine the dynamic contraction and relaxation behavior of a muscle is by measuring the twitch kinetic parameters of an intact muscle. This information, combined with assessments of force-pCa (intact or skinned), will provide supplemental information on the developed tension as well as the contraction and relaxation kinetics, which can also allow, albeit indirect, more insight into the relative contribution of k_on and k_off, as changes in k_on would more heavily influence the developed tension while k_off the relaxation kinetics. In summary, the assessment of a force-pCa curve of a muscle at steady-state does not contain sufficient information to make an unambiguous translation into dynamic behavior of a muscle. When dynamic twitch kinetic parameters are not available, but k_on and/or k_off are assessed in addition to assessment of steady-state myofilament sensitivity, the translation of skinned fiber data toward potential dynamic behavior is greatly enhanced.

To our knowledge, it is virtually impossible to assess the k_on and k_off in intact muscles, as researchers have typically used in vitro fluorimetry to measure changes in fluorescence in isolated TnC at steady-state, but the knowledge of the two rate constants can yield important insights into the contraction and relaxation kinetics of a working muscle (Tikunova and Davis, 2004). Since the dynamic behavior is critically important in various disease states, assessment of kinetics, as well as kinetic reserve is critically important to further understand cardiac malfunction in disease (Janssen et al., 2016).

Cardiac output is heavily regulated, and this regulation is mainly governed by three mechanisms: length-dependent activation, frequency-dependent activation, and beta-adrenergic activation (Janssen, 2010). These three factors all encompass modulation of myofilament calcium sensitivity, which has important implications for the dynamic behavior of cardiac muscle, and we will discuss each of these factors below in more detail.

Most previous studies that assessed myofilament calcium sensitivity have utilized skinned muscle preparations, reporting approximate half-max force at pCa of 6 (EC₅₀). However, intact muscles have been reported to exhibit higher calcium sensitivity compared to skinned muscles (Gao et al., 1994; Varian et al., 2006; Monasky et al., 2010). Later studies, in intact muscle at physiological temperature (Varian et al., 2006; Monasky et al., 2010) confirmed a high sensitivity for calcium in intact muscle compared to published values in skinned/permeabilized muscle. After the muscle skinning process, many intracellular components excluding the myofilament are lost. This naturally leads one to wonder what sensitizing intracellular components are lost during the process of muscle skinning. One possibility might be various kinases that increase the calcium sensitivity of myofilaments. For example, protein kinase D (PKD) has been reported to increase calcium sensitivity via phosphorylation of Ser³¹⁵ cardiac myosin binding protein C (cMyBP-C) (Dirkx et al., 2012). However, PKD can also reduce calcium sensitivity via phosphorylation of troponin I (TnI) Ser^23/24 (Cuello et al., 2007). It may be possible that there are “natural” calcium sensitizers other than kinases that are (partially) lost upon skinning. Carnosine-like compounds and taurine are examples of cytosolic compounds that have been shown to alter myofilament calcium sensitivity that may be lost during permeabilization (Steele et al., 1990; Lamont and Miller, 1992). Recently, S-glutathionylation of cMyBP-C as well as phosphorylation of TnI by adenosine monophosphate (AMP) kinase have been shown to increase calcium sensitivity (Nixon et al., 2012; Patel et al., 2013). Phosphatase 2A (PP2A), which is associated with various calcium handling and myofilament proteins such as the L-type calcium channel and myosin light chain 2 (MLC-2), also increases calcium sensitivity (Wijnker et al., 2011).

However, it is important to note that there are many desensitizing cytosolic components as well. For example, other kinases such as protein kinase A (PKA) and protein kinase C (PKC) have both been reported to decrease calcium sensitivity, not increase it (Herzig and Rüegg, 1980; Strang et al., 1994; de Tombe and Stienen, 1995; van der Velden et al., 2006). It has been well-documented that protein kinase A (PKA) phosphorylation of TnI 23/24 results in desensitization of myofilament (Layland et al., 2005). Protein kinase C (PKC) and protein kinase G (PKG) can phosphorylate myofilament proteins such as cMyBP-C to reduce calcium sensitivity (Pfitzer et al., 1982; van der Velden et al., 2006). O-linked N-acetyl-D-glucosaminylation of cardiac myofilament also decreases calcium sensitivity (Ramirez-Correa et al., 2008).

Another aspect to consider with the skinning procedure is that sarcomeric lattice spacing increases due to the procedure (Irving et al., 2000). A number of studies have reported that decreased lattice spacing (analogous to increased muscle length) leads to increased calcium sensitivity (McDonald and Moss, 1995; Wang and Fuchs, 1995). The increase in sarcomeric lattice spacing in skinned muscle preparations may lead to decreased calcium sensitivity. However, Konhilas et al. (2002) have reported that osmotic compression of the lattice spacing does not affect the length-calcium sensitivity relationship.

It is not clear at the moment whether desensitizing or sensitizing cytosolic components play a bigger role in intact muscle preparations. In addition, most of the investigations on these cytosolic components were performed on skinned muscle in vitro, which makes it difficult for one to predict their actual roles in vivo. However, one must carefully consider the implication of the loss of cytosolic signaling molecules and the changes in the myofilament geometry on their direct or indirect effect on calcium sensitivity during the permeabilization process.

Myofilament proteins work in conjunction to allow the cardiac muscle to contract and relax in response to changes in intracellular free calcium ion concentration. Therefore, it is not surprising that genetic mutations in many of the myofilament proteins impact calcium sensitivity. However, only a few mutations have been characterized to show their translation into dynamic behavior of a muscle. For the purpose of characterization of dynamic contraction of a muscle in vivo, it is necessary but not sufficient to show changes in calcium sensitivity. One must also either report biochemical changes in the k_on or k_off or report twitch force development and kinetics in an intact muscle to show what exact contributing factors changed calcium sensitivity. In this review, we highlight a few mutations that have originally been found in patients with dilated cardiomyopathy (DCM) or hypertrophic cardiomyopathy (HCM) and discuss how these mutations affect myofilament calcium sensitivity and dynamic behavior.

The myofilament is crucial in the regulation of contractile and relaxation behavior of the cardiac muscle, especially in the pathophysiology of heart failure. Force-pCa curves generated from skinned muscle preparations are able to reduce the complex environment of a muscle into a much simpler relationship between isometric force and free calcium ions. While such data are necessary and important, as a stand-alone assessment however, the knowledge of myofilament calcium sensitivity alone is not sufficient for the extrapolation to dynamic behavior of a muscle representative of that in vivo. Heart failure continues to be one of the leading causes of death in the U.S., and standard-of-care treatment of the disease has been largely limited to beta blockers, diuretics, angiotensin converting enzyme inhibitors, and calcium channel blockers, which have been in use for decades. More studies on calcium sensitivity that incorporate either assessment of k_on or k_off, or assess dynamic behavior are needed in the field of cardiac physiology to improve interpretation of the impact of myofilament mutations, and for strategizing of novel treatment for the patients who continue to suffer from the disease.

PJ: concept of review, concept of illustrations, edited final draft. JC: wrote initial draft, made illustrations. MZ, BB, JD: helped discuss concept, edited final draft.