Why Is there a Limit to the Changes in Myofilament Ca2+-Sensitivity Associated with Myopathy Causing Mutations?

Mutations in striated muscle contractile proteins have been found to be the cause of a number of inherited muscle diseases; in most cases the mechanism proposed for causing the disease is derangement of the thin filament-based Ca2+-regulatory system of the muscle. When considering the results of experiments reported over the last 15 years, one feature has been frequently noted, but rarely discussed: the magnitude of changes in myofilament Ca2+-sensitivity due to myopathy-causing mutations in skeletal or heart muscle seems to be always in the range 1.5–3x EC50. Such consistency suggests it may be related to a fundamental property of muscle regulation; in this article we will investigate whether this observation is true and consider why this should be so. A literature search found 71 independent measurements of HCM mutation-induced change of EC50 ranging from 1.15 to 3.8-fold with a mean of 1.87 ± 0.07 (sem). We also found 11 independent measurements of increased Ca2+-sensitivity due to mutations in skeletal muscle proteins ranging from 1.19 to 2.7-fold with a mean of 2.00 ± 0.16. Investigation of dilated cardiomyopathy-related mutations found 42 independent determinations with a range of EC50 wt/mutant from 0.3 to 2.3. In addition we found 14 measurements of Ca2+-sensitivity changes due skeletal muscle myopathy mutations ranging from 0.39 to 0.63. Thus, our extensive literature search, although not necessarily complete, found that, indeed, the changes in myofilament Ca2+-sensitivity due to disease-causing mutations have a bimodal distribution and that the overall changes in Ca2+-sensitivity are quite small and do not extend beyond a three-fold increase or decrease in Ca2+-sensitivity. We discuss two mechanism that are not necessarily mutually exclusive. Firstly, it could be that the limit is set by the capabilities of the excitation-contraction machinery that supplies activating Ca2+ and that striated muscle cannot work in a way compatible with life outside these limits; or it may be due to a fundamental property of the troponin system and the permitted conformational transitions compatible with efficient regulation.

Mutations in striated muscle contractile proteins have been found to be the cause of a number of inherited muscle diseases; in most cases the mechanism proposed for causing the disease is derangement of the thin filament-based Ca 2+ -regulatory system of the muscle. When considering the results of experiments reported over the last 15 years, one feature has been frequently noted, but rarely discussed: the magnitude of changes in myofilament Ca 2+ -sensitivity due to myopathy-causing mutations in skeletal or heart muscle seems to be always in the range 1.5-3x EC 50 . Such consistency suggests it may be related to a fundamental property of muscle regulation; in this article we will investigate whether this observation is true and consider why this should be so. A literature search found 71 independent measurements of HCM mutation-induced change of EC 50 ranging from 1.15 to 3.8-fold with a mean of 1.87 ± 0.07 (sem). We also found 11 independent measurements of increased Ca 2+ -sensitivity due to mutations in skeletal muscle proteins ranging from 1.19 to 2.7-fold with a mean of 2.00 ± 0.16. Investigation of dilated cardiomyopathy-related mutations found 42 independent determinations with a range of EC 50 wt/mutant from 0.3 to 2.3. In addition we found 14 measurements of Ca 2+ -sensitivity changes due skeletal muscle myopathy mutations ranging from 0.39 to 0.63. Thus, our extensive literature search, although not necessarily complete, found that, indeed, the changes in myofilament Ca 2+ -sensitivity due to disease-causing mutations have a bimodal distribution and that the overall changes in Ca 2+ -sensitivity are quite small and do not extend beyond a three-fold increase or decrease in Ca 2+ -sensitivity. We discuss two mechanism that are not necessarily mutually exclusive. Firstly, it could be that the limit is set by the capabilities of the excitation-contraction machinery that supplies activating Ca 2+ and that striated muscle cannot work in a way compatible with life outside these limits; or it may be due to a fundamental property of the troponin system and the permitted conformational transitions compatible with efficient regulation.
Mutations in striated muscle contractile proteins have been found to be the cause of a number of inherited muscle diseases; in most cases the mechanism proposed for causing the disease is derangement of the thin filament-based Ca 2+regulatory system of the muscle. Hypertrophic cardiomyopathy and hypercontractile diseases of skeletal muscle, such as distal arthrogryposis and "stiff child syndrome, " have been linked to a higher myofilament Ca 2+ -sensitivity (Marston, 2011;Donkervoort et al., 2015). In contrast dilated cardiomyopathy mutations are commonly, but not exclusively, linked to decreased Ca 2+ -sensitivity. Mutations in contractile proteins that are linked to nemaline myopathy and related skeletal muscle myopathies have also been found to be associated with reduced Ca 2+ sensitivity (Marttila et al., 2012(Marttila et al., , 2014. The causative connection between myofilament Ca 2+ -sensitivity and muscle dysfunction is a field of intensive research that is too complex to consider in this account. However, when considering the results of such experiments reported over the last 15 years, one feature has been frequently noted, but rarely discussed. The magnitude of changes in myofilament Ca 2+ -sensitivity due to myopathycausing mutations in skeletal or heart muscle seems to be always in the range 1.5-3x EC 50 . Such consistency suggests it may be related to a fundamental property of muscle regulation; in this article we will investigate whether this observation is true and consider why this should be so. Most investigations have found increased Ca 2+ -sensitivity in muscle with hypertrophic cardiomyopathy (HCM) and restrictive cardiomyopathy (RCM)-causing mutations. Our literature search found 71 independent measurements of the mutation-induced change of EC 50 ranging from 1.15 to 3.8-fold with a mean of 1.87 ± 0.07 (sem) ( Table 1). We also found 11 independent measurements of increased Ca 2+ -sensitivity due to mutations in skeletal muscle proteins ranging from 1.19 to 2.7-fold with a mean of 2.00 ± 0.16 ( Table 2).
Dilated cardiomyopathy-causing mutations were initially found to decrease Ca 2+ -sensitivity but more recent studies have indicated the situation is more complex. DCM-linked mutations can both increase and decrease Ca 2+ -sensitivity depending on the individual mutations, moreover the direction of change can be different with a single mutation measured in different systems (Marston, 2011;Memo et al., 2013). This is illustrated in Table 3 where 42 independent determinations show a range of EC 50 wt/mutant from 0.3 to 2.3. In addition we found 14 measurements of Ca 2+ -sensitivity changes due skeletal muscle myopathy mutations ranging from 0.39 to 0.63 (Table 4).
Thus, our extensive literature search, although not necessarily complete, found that, indeed, the changes in myofilament Ca 2+sensitivity due to disease-causing mutations have a bimodal distribution and that the overall changes in Ca 2+ -sensitivity are quite small and do not extend beyond a 3-4-fold increase or decrease in Ca 2+ -sensitivity. Indeed when all the findings are plotted as a histogram one finds that increases in Ca 2+ -sensitivity on a log scale have an approximately normal distribution with mean increase in Ca 2+ -sensitivity (EC 50 wt/mutant) of 1.86-fold (corresponding to pCa 50 = 0.255 ± 0.015), whilst the decreases in Ca 2+ sensitivity have a mean EC 50 wt/mutant of 0.54-fold (corresponding to pCa 50 of -0.286 ± 0.01; Figure 1A). It  is also worth noting that this small Ca 2+ -sensitivity shift is observed independent of the measurement method Figure 1B compares the pCa 50 distribution measured by unloaded assays (actomyosin ATPase or in vitro motility) and by loaded assays (force measurements in skinned muscles, cell, and isolated myofibrils). The mean magnitude of the Ca 2+ -sensitivity change is about 20% less when measured in loaded assays. What could be the underlying reason for this consistent and small effect of mutations on EC 50 ? We will consider two possible mechanisms that are not necessarily mutually exclusive. Firstly, it could be that the limit is set by the capacity of the EC coupling system that supplies activating Ca 2+ and that striated muscle cannot work in a way compatible with life outside these limits; alternatively it may be due to a fundamental property of the troponin system and the permitted conformational transitions compatible with efficient regulation.
Before attempting to discuss these mechanisms it is worthwhile considering some additional evidence on Ca 2+sensitivity shifts. Perhaps the most puzzling observation is that there appears to be no correlation between the Ca 2+ -sensitivity shift and disease severity. Skeletal myopathy mutations that cause life-threating muscle weakness from birth and often require mechanical assistance in breathing (Ravenscroft et al., 2015), have the same Ca 2+ -sensitivity shifts as dilated cardiomyopathy mutations which are considerably less lethal (Hershberger et al., 2013). Whilst heart muscle has compensatory strategies not available in skeletal muscle to account for this difference, the small change in Ca 2+ -sensitivity even in the most severe skeletal muscle disease might be indicative of a fundamental structure-based limit on changes in EC 50 .
Consideration of the Ca 2+ -sensitivity shifts in cardiomyopathies (Tables 1, 3) do not indicate any correlation with disease severity. Any relationship that may exist is masked by the extreme variability of Ca 2+ -sensitivity shift measurements. For instance, the "severe" TNNI3 R145G HCM/RCM-linked mutation features at both extremes of the Ca 2+ -sensitivity range (1.15x and 3.65x); for the 6 assays in the table the mean is 1.84, close to the mean of all 71 HCM measurements (1.87). The same variability can be seen with other mutations where multiple values are available: ACTC E99K, n = 5, 1.24-2.45 mean 1.85; TPM1 E180G, n = 4, 1.30-2.75, mean 1.78. The second relevant observation is that the physiological modulation of cardiac muscle myofilament Ca 2+ -sensitivity due to phosphorylation of troponin I by protein kinase A has been known to be a 2-3-fold shift for many years . Table 5 lists a number of recent determinations of this Ca 2+ -sensitivity shift in several species and measured by both loaded and unloaded assays illustrating its small range. Figure 1C shows how the magnitude and distribution of measured changes is similar to the changes induced by disease-causing mutations. It would be logical to conclude that this represents the range of achievable Ca 2+ sensitivity shifts in cardiac muscle due to the limitations of the EC coupling system. In principle, it should be possible to go beyond the Ca 2+sensitivity limits set by EC coupling in an in vitro system where Ca 2+ binding affinity can be much greater or much less than the native troponin. Cardiac troponin C presents extreme examples in a single molecule. Only site II binds Ca 2+ in the physiologically relevant range (2.5 × 10 5 M −1 ) and so is solely responsible for Ca 2+ -regulation (Holroyde et al., 1980). A few amino acid changes in the EF-hand motifs results in sites that do not bind Ca 2+ (Site I) or sites that bind Ca 2+ 200x tighter (sites III and IV) and are permanently occupied by Ca 2+ or Mg 2+ (Li and Hwang, 2015). Thus, it would seem that neither a very high Ca 2+ sensitivity nor a very low one are able to participate in regulation. How much deviation of Ca 2+ affinity from the norm is compatible with muscle regulation?
It is known that for mutations, the small Ca 2+ -sensitivity changes correlate with Ca 2+ binding affinity to thin filaments (Robinson et al., 2007). In a study of mutations induced in skeletal muscle troponin C, Davis et al. achieved a 243-fold range of Ca 2+ binding affinities for troponin C. However, this did not translate into such a great range when Ca 2+ -binding was measured in the presence of TnI (96-148) and caused a still smaller shift in the Ca 2+ -sensitivity of force production (Davis et al., 2004). Thus, the most extreme Ca 2+ -sensitizing mutation, V45Q increased TnC Ca 2+ binding affinity 19-fold, but the increase was only 3.1-fold when measured in the presence of the TnI peptide and Ca 2+ -sensitivity in skinned fibers was just 2.3-fold more than wild-type. This is within the same  (data from Tables 1, 3). Unloaded assays are IVMA and ATPase, loaded assays are Fibers TG, Myofibrils TG, Fibers X, Human fibers, For decreased Ca 2+ sensitivity mean unloaded pCa50 is -0.27 ± 0.02 and mean loaded is -0.21 ± 0.03, p = 0.05. For increased Ca 2+ -sensitivity mean unloaded pCa50 is 0.26 ± 0.02 and mean loaded is 0.021 ± 0.02, p = 0.04. (C) Distribution of change in Ca 2+ -sensitivity due to troponin I phosphorylation (EC 50 unphosphorylated/EC 50 phosphorylated). Data from Table 5. The mean change is 0.50 ± 0.06-fold (n = 9), pCa50 = −0.30. range of many HCM-causing mutations ( Table 1). A similar picture emerges from Cardiac troponin C where the single regulatory Ca 2+ -binding site simplifies the argument: V44Q increases Ca 2+ -binding affinity to TnC 6.5-fold but increases myocyte Ca 2+ -sensitivity by just 3.4-fold (Parvatiyar et al., 2010). Thus, it seems that the structure of troponin and its interactions with the rest of the thin filament does limit the consequences of a modification that increases Ca 2+ binding affinity. A slightly different situation arises when Ca 2+ binding affinity is less than wild-type. Davis et al., noted that the mutations that decreased Ca 2+ binding affinity the most (F26Q, 63-fold, I37Q, 24-fold and I62Q, 10-fold) could not properly regulate force in skinned fibers since they only produced about 13% of the maximal force of wild-type muscle at saturating Ca 2+ concentrations. On the other hand, two less extreme mutations, M81Q and F78Q decreased Ca 2+ -sensitivity whilst retaining the same maximum force production as wild type. In these cases, again, the increased Ca 2+ binding affinity for TnC was substantially greater than the increased Ca 2+ -sensitivity of skinned fibers (5.9x vs. 1.8x for M81Q and 8.4x vs. 4.2x for F78Q). Thus, thin filament structure seems to limit the possible effects of changes in Ca 2+ -binding affinity.
It is self-evident that changing myofilament Ca 2+ sensitivity will affect contractile output in muscle. It is well-established that EC 50 for skinned muscle fibers is about 1 µM and FIGURE 2 | The effects of changing Ca 2+ -sensitivity on contractility. Ca 2+ -activation curves for mouse myofibrils with EC 50 of 0.8 µM, a Hill coefficient of 4 and a [Ca 2+ ] i range from 500 nM at peak to 100 nM when relaxed (pink box). The curves with two-fold higher Ca 2+ sensitivity, as found with HCM mutations, four-fold higher Ca 2+ sensitivity and 0.5-fold Ca 2+ -sensitivity, as may be found in some DCM mutations, is plotted for comparison.
that Ca 2+ -activation of contraction is highly cooperative. Most measurements suggest a five-fold range in free Ca 2+ concentration during a cardiac muscle contraction. Peak Ca 2+ concentration is about 600 nM at rest and can be substantially higher during adrenergic stimulation, thus normally muscle is only partially activated (Negretti et al., 1995;Dibb et al., 2007). Figure 2 shows a real life example: in a mouse model of HCM (ACTC E99K) we measured both the Ca 2+ -activation curve for myofibrils and the contractility of intact papillary muscle as well as the Ca 2+ -transient (Song et al., 2013). Under the conditions of this experiment the Ca 2+ transient was the same in Wild-type and ACTC E99K muscle, Ca 2+ sensitivity was 0.8 µM for wildtype and 0.34 µM for ACTC E99K with a Hill coefficient of about 4. The increase in Ca 2+ -sensitivity due to the ACTC E99K HCM mutation corresponds to an approximately four-fold increase in twitch force in the absence of a change in the Ca 2+ -transient that was actually observed.
We can use this model to consider what would happen if Ca 2+ -sensitivity changed beyond the normal range. If myofilament Ca 2+ -sensitivity was 4 times normal, maximum force would reach close to 100%, leaving no range for it to be modulated by adrenergic agents. Moreover, it is likely that the muscle would not fully relax, since, based on the five-fold range of the Ca 2+ transient even at the lowest Ca 2+ level force would be 5-10%, a substantial fraction of the peak force of wild-type muscle, thus the hypercontractile phenotype would impose a major defect in relaxation, much more severe than the diastolic dysfunction associated with HCM mutations with only a 1.8-fold average Ca 2+ sensitivity increase.
If myofilament Ca 2+ -sensitivity were decreased to half the normal, contractility would be very low indeed. The fact that mutations that decrease Ca 2+ -sensitivity are not lethal and indeed in transgenic mice, may exhibit little phenotype, is probably due to a compensatory increase in the Ca 2+ -transient (Du et al., 2007). However, this compensation may not be enough to support normal contraction in the long term, leading to DCM, the phenotype commonly associated with reduced Ca 2+sensitivity.

CONCLUSION
The objective of this article was to confirm that Ca 2+sensitivity of contractility only varies within an narrow range of three-fold above and below the normal EC 50 at rest and to investigate why this should be. The high cooperativity of muscle activation by Ca 2+ means there is a narrow [Ca 2+ ] range between relaxed and active muscle. It would appear that the excitation-contraction coupling machinery of the cell has limited ability to change the amplitude of the Ca 2+ -transient or baseline [Ca 2+ ] to compensate for changes in EC 50 ; thus increased Ca 2+ -sensitivity would be limited by inability to relax and reduced Ca 2+ -sensitivity would be limited by inability to contract. It is intriguing that the Ca 2+ -sensitivity range of the thin filament itself is independently limited. Mutations that change Ca 2+ -binding affinity to TnC by a large amount nevertheless only produce a small change in EC 50 for activation of loaded or unloaded contractility in vitro. Whether this property is an evolutionary adaptation that limits the deleterious effects of mutations in thin filaments or simply fortuitous in unknown.

AUTHOR CONTRIBUTIONS
The author confirms being the sole contributor of this work and approved it for publication.

FUNDING
SM's research is funded by British Heart Foundation programme grant RG/11/20/29266.