The myosin lever arm was first recognized almost 30 years ago as the first structure of myosin subfragment 1 (S1) was described (Rayment et al., 1993a; Rayment et al., 1993b), and the lever arm’s function in amplifying the motion of the converter domain was subsequently confirmed (Uyeda et al., 1996; Geeves and Holmes, 1999). Much work since has focused on understanding the functional significance of different lever arm features across myosin classes (Spudich and Sivaramakrishnan, 2010). However, less attention has been paid to the functional consequences of point mutations in the myosin lever arm. Such mutations in the human β-cardiac myosin lever arm are a frequent cause of hypertrophic cardiomyopathy (HCM), a disease characterized by hypercontractility and eventual hypertrophy of the left ventricle (Ho et al., 2009). While many HCM-causing mutations in the myosin motor domain have been described and characterized (Nag et al., 2015; Adhikari et al., 2016; Kawana et al., 2017; Nag et al., 2017; Adhikari et al., 2019; Sarkar et al., 2020; Vander Roest et al., 2021; Spudich, 2019; Liu et al., 2018), HCM-causing mutations in the lever arm remain understudied, despite the fact that the lever arm has a relatively high rate of these mutations and plays a key role in transducing the chemical energy of ATP hydrolysis into physical motion (Uyeda et al., 1996; Geeves and Holmes, 1999).

HCM is a leading cause of genetic heart disease, affecting up to 1 in 200 in the US population (Semsarian et al., 2015). Mutations leading to HCM have been found in genes encoding a variety of sarcomeric proteins; however, the vast majority occur in either the β-cardiac myosin heavy chain or cardiac myosin-binding protein-C (Konno et al., 2010). Because left ventricular hypercontractility typically precedes hypertrophy in HCM patients (Ho et al., 2009), it has been hypothesized that sarcomeric mutations lead to hypercontractility at the molecular scale. Early work in this area suggested that mutations might increase myosin’s motor activity by increasing its actin-activated ATPase rate, motility velocity, or force production (Tyska et al., 2000; Debold et al., 2007). While some mutations may function by this mechanism (Adhikari et al., 2016; Adhikari et al., 2019; Liu et al., 2018), lately, a more compelling hypothesis has gained traction: many HCM-causing mutations appear to reduce myosin’s ability to form an autoinhibited state (Nag et al., 2017; Adhikari et al., 2019; Sarkar et al., 2020; Vander Roest et al., 2021; Spudich, 2019; Moore et al., 2012; Spudich, 2015; Alamo et al., 2017; Robert-Paganin et al., 2018; Anderson et al., 2018). Loss of an autoinhibited state could lead to additional myosin heads acting to generate force during contraction, thus leading to hypercontractility.

An autoinhibited state of myosin was first described in smooth muscle myosin over 20 years ago (Trybus, 1991), but its relevance to β-cardiac myosin and HCM was only more recently recognized. In the structural view of this smooth muscle autoinhibited state, the myosin heads fold back onto their own subfragment 2 (S2) tail in a conformation known as the interacting heads motif (IHM; Wendt et al., 2001; Woodhead et al., 2005). One of the two heads in the dimer has its actin-binding interface buried in the folded structure; this head is referred to as the ‘blocked head’, while the other is called the ‘free head’, since its actin-binding interface is not hidden structurally. Many myosin types have been shown to assume this folded back IHM structure (Woodhead et al., 2005; Jung et al., 2008; Al-Khayat et al., 2013), and it is now thought to be a common feature across the myosin II class (Lee et al., 2018).

The IHM structure has been correlated to an ultra-low basal ATPase rate (three orders of magnitude below the actin-activated ATPase rate) in the absence of actin called the ‘super relaxed state’ (SRX; Anderson et al., 2018; Al-Khayat et al., 2013; Stewart et al., 2010; McNamara et al., 2015; Rohde et al., 2018). Heads lacking the S2 tail mostly have a faster basal ATPase rate (two orders of magnitude below the actin-activated ATPase rate) referred to as the ‘disordered relaxed state’ (DRX). Additionally, actin-activated ATPases comparing myosin with and without its S2 tail have shown that when the S2 tail is present, the apparent ATPase rate decreases, suggesting that the S2 tail promotes autoinhibition (Nag et al., 2017; Adhikari et al., 2019; Sarkar et al., 2020; Vander Roest et al., 2021; Trybus et al., 1997). Therefore, myosin containing the S2 tail is thought to be in equilibrium between an open state available for actin binding and the closed IHM conformation. While these structural states are correlated to specific basal and actin-activated ATPase rates, there may be circumstances in which the structural states are uncoupled from their respective rates (Chu et al., 2021; Nag and Trivedi, 2021). Thus, while these functional assays measure autoinhibition by the myosin S2 tail, they are only correlative measures for structural states (i.e., the IHM).

HCM-causing mutations in the myosin lever arm could lead to hypercontractility by (1) disrupting S2 tail-based autoinhibition, (2) increasing intrinsic motor activity without affecting autoinhibition, or (3) affecting both intrinsic motor activity and autoinhibition. The lever arm’s role in the formation of the autoinhibited state has previously been explored in mollusk myosin, which is primarily regulated via molecular switching from the autoinhibited off state to an open state in the presence of Ca²⁺. A number of structural studies collectively concluded that three joints in the molluscan myosin lever arm appear to be potential sources of flexibility which may be necessary for the formation of the IHM (Houdusse et al., 2000; Houdusse and Cohen, 1996; Himmel et al., 2009; Pylypenko and Houdusse, 2011; Brown et al., 2011): a region at the extreme N-terminus of the lever arm, termed as the ‘pliant’ region, a typical bent region between the essential light chain (ELC) and the regulatory light chain (RLC) binding regions, and a region at the C-terminus of the lever arm in the RLC-binding area termed the ‘hook’ or ‘ankle’ joint. Recently, high-resolution cryo-electron microscopy (cryo-EM) structures of the folded-back IHM state in smooth muscle myosin demonstrated that the lever arm must take on a different conformation in each of the asymmetric folded heads in the IHM, further confirming the importance of lever arm positioning in the folded state (Scarff et al., 2020; Yang et al., 2020; Heissler et al., 2021). Mutations in the lever arm could reduce myosin’s S2 tail-based autoinhibition by negatively impacting any of these structural requirements for forming the IHM. Alternatively, lever arm mutations could increase intrinsic motor activity, leading to hypercontractility. For example, mutations could increase the rigidity of the lever arm and/or alter its light chain-binding properties, which could in turn influence power output. Indeed, several previous studies investigating RLC mutations have suggested that mutations can modulate lever arm compliance (Greenberg et al., 2010; Sherwood et al., 2004; Karabina et al., 2015; Farman et al., 2014). Lever arm mutations could also allosterically affect the motor domain, potentially leading to increased motor activity.

In the present study, we examined five adult-onset HCM-causing mutations in the myosin lever arm to determine how they affect myosin function: D778V, L781P, S782N, A797T, and F834L (Figure 1). We selected these mutations based on their confirmed pathogenicity and to allow investigation of mutations across the lever arm. We also aimed to investigate mutations in or near regions thought to be important in lever arm function, including the pliant region (D778V, L781P, and S782N), the bent region between the light chains (A797T), and the hook joint (F834L). We hypothesized that mutations in the lever arm could affect myosin either by (1) altering its intrinsic motor activity, and/or (2) reducing its ability to form the autoinhibited state. We found that the three mutations in the pliant region of the lever arm, D778V, L781P, and S782N, led to changes in both myosin’s motor activity and the formation of the autoinhibited state. The two mutations in the light chain-binding regions, A797T and F834L, clearly impacted myosin’s ability to form the autoinhibited state, likely explaining their contributions to hypercontractility. Thus, mutations in the lever arm have varying impacts on myosin function depending on whether they are in the pliant region or the light chain-binding regions.

Figure 1

Download asset Open asset

The myosin lever arm performs one of the most important functions of the myosin chemomechanical cycle: it amplifies the transduction of the chemical energy of ATP hydrolysis into physical motion, allowing for myosin’s motor function. This crucial role explains why the lever arm is both highly conserved and a hotspot for HCM-causing mutations. Here, we have characterized five such mutations spanning the pliant region and the light chain-binding regions, giving insights into the role of the lever arm in generating force and power output, as well as its role in myosin S2 tail-based autoinhibition. The impacts of the mutations on myosin function segregate them naturally into two categories: the light chain-binding region mutations, A797T and F834L, behave very differently from the pliant region mutations, D778V, L781P, and S782N.

The functional changes in myosin’s motor activity caused by the pliant region mutations were dramatic and ran the full spectrum of potential outcomes: D778V increased actin-activated ATPase of the 2-hep construct along with in vitro motility velocity and k₀, L781P decreased motility velocity, k₀, and step size, and S782N did not result in changes in any parameter except the force sensitivity of the detachment rate $\delta$. That these mutations can have such varied and substantial effects on myosin activity speaks to the importance of the pliant region in coupling the activity of the lever arm to the rotation of the converter domain.

One feature that was shared among the pliant region mutations was a reduction in the force sensitivity δ. Reduced force sensitivity suggests a mechanism where mutations partially uncouple the force perceived at the ATP-binding site from the force applied at the anchor point at the C-terminus of the lever arm. We previously observed a decrease in δ for the P710R mutation, which is located in between the SH helices and the converter domain approximately 20 Å from the pliant region (Vander Roest et al., 2021). Pliant region mutations may likewise reduce rigidity, thus uncoupling the force transduction and reducing δ. The term ‘pliant region’ itself comes from the varied positions of this region observed in different crystal structures (Houdusse et al., 2000; Gourinath et al., 2003), suggesting that its compliance or ability to assume multiple conformations is an important feature of its function; thus, any alteration to the rigidity of this region would likely have strong impacts on myosin function in sarcomeres. It is possible that reductions in δ as a result of pliant region mutations could impact contraction and/or relaxation kinetics in patients with these specific mutations.

Additionally, divergent changes resulting from the pliant region mutations may be a function of the identities of the mutations themselves. For example, the L781P mutation introduces a helix-breaking proline into a continuous α-helix. If the lever arm α-helix is destabilized, this could explain the observed decrease in step size for the L781P mutation: a more bent lever arm would generally produce a lower step size. Additionally, the D778V mutation replaces a highly polar amino acid with a more hydrophobic one. This mutation dramatically increases the detachment rate, suggesting that amino acids in the pliant region, which are very distant from the actin-binding domain, can play a role in the highly allosteric process of actin detachment. Conversely, the S782N mutation has comparatively few impacts on the kinetics of the myosin motor, suggesting that some mutations in the pliant region may be well tolerated in the context of motor domain function.

The sum total of the effects of pliant region mutations on motor activity (outside of potential impacts on autoinhibition) is reflected in the changes in calculated average power output (Figure 4C), which takes into account actin-activated ATPase rate k_cat, force-dependent detachment rate k_det(F), force-independent detachment rate k₀, and step size d. For the pliant region mutations, D778V appears to be hypercontractile (particularly at high resistive forces), L781P appears to be hypocontractile, and S782N is within error of WT. A prominent model in the field suggests that HCM mutations increase myosin activity in ensemble, resulting in hypercontractility at the cellular level (Spudich, 2019). The magnitude of the increases in D778V’s actin-activated ATPase activity and motility velocity was comparable to the early-onset mutations D239N and H251N (Adhikari et al., 2016) and not far from the changes observed for I457T, which had the largest changes we’ve observed to date (Adhikari et al., 2019). Thus, it is possible that these increases alone could lead to hypercontractility in a cellular context. However, the same cannot be said of L781P and S782N, where a different rationale would have to apply, given that these mutations resulted in either decreased or no change in contractility, respectively. Thus, we additionally considered the possibility that these mutations might impact myosin S2 tail-based autoinhibition.

The pliant region mutations showed a unique phenotype where they had a smaller reduction between the actin-activated ATPase rate of the 2-hep vs the 25-hep constructs as compared to WT, suggesting reduced autoinhibition, but did not show any statistically significant changes in the SRX/DRX ratio in the single turnover ATPase assay. We have used these two assays to study a number of mutations to date (Adhikari et al., 2019; Sarkar et al., 2020; Vander Roest et al., 2021), and all previously studied mutations had shown concordant behavior in these two assays. One way to interpret this data is that the actin-activated ATPase of 25-hep suggests that autoinhibition is disrupted by the mutations, but only in the presence of actin, such that the single turnover ATPase is unchanged. Previous data supports a model where the presence of actin can influence and disrupt myosin autoinhibition (Rohde et al., 2018), so it stands to reason that mutations could specifically affect this activity. This model would suggest that in the in vivo context, myosin activity would be increased specifically during contraction, when actin is available for myosin binding, thus increasing force production, but during relaxation, when actin-binding is blocked by tropomyosin, there would be the same proportion of SRX compared to WT. It is clear from structures of smooth muscle myosin in the folded state that the conformation of the pliant region is different in the two heads (Scarff et al., 2020; Yang et al., 2020; Heissler et al., 2021): the blocked head pliant region takes on a bent conformation, while in the free head pliant region is much straighter. Thus, logically, mutations could destabilize the IHM by changing the available positions of the pliant region, leading to the reduced autoinhibition we observed and thus hypercontractility. Alternatively, it is possible that this study of the pliant region mutations may be limited by isolating the myosin outside of the context of sarcomeres, where structural constraints or additional proteins (such as myosin binding protein-C) could impact myosin activity and/or the formation of the autoinhibited state. Regardless, future experiments introducing pliant region mutations into a cardiomyocyte model would be useful for helping to test whether these mutations indeed result in net hypercontractility at the cellular level, and if so, how that would arise and lead to cellular hypertrophy.

In contrast to the pliant region mutations, the light chain-binding region mutations both had very little or no impact on myosin’s in vitro motility velocity or actin-activated ATPase activity of the 2-hep construct. However, both mutations had clear impacts on myosin’s S2 tail-based autoinhibition, reflected by both an increase in the S2 tail-containing 25-hep actin-activated ATPase rate and a decrease in the proportion of SRX. Thus, it is likely that the light chain-binding regions of the lever arm play an important role in myosin S2 tail-based autoinhibition. Most myosin mutations that have been investigated in the context of myosin autoinhibition have been in regions thought to be directly involved in forming structurally necessary contacts for the IHM state, including mutations in both the myosin ‘mesa’ region and the S2 tail (Nag et al., 2017; Adhikari et al., 2019), which likely interact with each other in the IHM, as well as mutations at the putative head-head interface (Sarkar et al., 2020). In contrast, A797T and F834L do not appear to be involved in such contacts in modeled structures of β-cardiac myosin in the IHM state (Nag et al., 2017; Alamo et al., 2017; Robert-Paganin et al., 2018). Thus, if the defects in autoinhibition that we observed for A797T and F834L indeed indicate reductions in the IHM structural state, these mutations are likely to affect the IHM structure by an allosteric mechanism. A simple model could be that the A797T and F834L mutations affect the position or rotation of the light chains about the lever arm. While our data shows that the light chains were still able to load onto the lever arm at the proper ratios, it is possible that specific light chain positioning is required to access the folded state. Indeed, in structures of smooth muscle myosin in the folded state (Scarff et al., 2020), it was shown that the RLC in particular took a position such that its phosphorylation sites were very close together in the IHM state, which the authors hypothesized created energetically unfavorable charge-charge interactions when the sites were phosphorylated (phosphorylation has been shown to move heads out of the autoinhibited state). Additionally, it was shown that a specific interaction of the C-termini of the RLCs with the hook joint (near F834L) on both the free and blocked heads may be important for formation of the IHM (Heissler et al., 2021). Thus, positioning of the light chains appears to be important for forming the folded state.

Alternatively, it is possible that the light chain-binding region mutations do not affect the positioning of light chains; instead, they could affect the structure of the lever arm itself in ways that prevent autoinhibition. It is apparent from the modeled structures of β-cardiac myosin in the IHM state that the lever arm may form an unusual conformation to allow the heads to fold back. In all three structures, the myosin heads and S2 tails overlay remarkably well; however, the lever arms vary dramatically (Nag et al., 2017; Alamo et al., 2017; Robert-Paganin et al., 2018; Figure 8A-C). Notably, in all three modeled structures, the lever arm’s α-helix is broken; this differs from known structures of the folded state in other myosins, where the α-helix is largely unbroken (Scarff et al., 2020; Yang et al., 2020; Heissler et al., 2021; Figure 8D–F). This suggests that the homology modeled structures of human β-cardiac myosin may differ from the real structure (especially in the lever arm region) and emphasizes the need for an atomic-resolution structure of human β-cardiac myosin in the ‘off’ state. Thus, it is possible that a specific or unusual conformation of the lever arm might be required to allow the IHM structure, and mutations might preclude specifically those conformations. For example, in molluscan myosin, where the folded back state has been extensively studied, the lever arm plays a key role in the formation of the folded back state, and a few regions in particular have been identified as important. The first is the slightly bent region of the lever arm directly in between the two light chains; in molluscan myosin, this is where Ca²⁺ directly binds to myosin and activates it from the folded back autoinhibited state (Houdusse and Cohen, 1996). Coincidentally, the A797T mutation is very near to that slight bend (Fig. S7G). Another region that has been identified as important in molluscan myosin is the ‘hook’ or ‘ankle’ joint near the C-terminus of the lever arm (Pylypenko and Houdusse, 2011; Brown et al., 2011). This feature is common across the myosin II class and can bend to very different angles, which is thought to help allow for the folded back conformation. The F834L mutation is very close to that ankle joint, and thus could affect the conformation about that joint (Fig. S7H). In any case, a high-resolution structure of human β-cardiac myosin in the IHM state will be useful to further understand the structural requirements of the lever arm for autoinhibition.

Figure 8

Download asset Open asset

In summary, this study allowed for a detailed analysis of lever arm function, specifically focusing on how HCM-causing mutations in both the pliant and light chain-binding regions of the lever arm alter myosin activity. While the light chain-binding region mutants clearly showed a phenotype of reducing myosin autoinhibition, potentially by disrupting the folded back IHM conformation, the pliant region mutants had a more complicated phenotype that begs further investigation. With a high density of mutations across a very small region of myosin, the pliant region seems to play an intriguing role in both transducing force from the motor domain to the lever arm and potentially forming the IHM structure. Future research into the lever arm region, particularly using in vivo models, could further clarify the role of the lever arm in both force transduction and formation of the folded state.

All data generated or analysed during this study are included in the manuscript and supporting file; source data files have been provided for all data figures.

1. 1. Adhikari AS
   2. Kooiker KB
   3. Sarkar SS
   4. Liu C
   5. Bernstein D
   6. Spudich JA
   7. Ruppel KM

   (2016) Early-Onset Hypertrophic Cardiomyopathy Mutations Significantly Increase the Velocity, Force, and Actin-Activated ATPase Activity of Human β-Cardiac Myosin

   Cell Reports 17:2857–2864.

   https://doi.org/10.1016/j.celrep.2016.11.040

   • PubMed
   • Google Scholar
2. 1. Adhikari AS
   2. Trivedi DV
   3. Sarkar SS
   4. Song D
   5. Kooiker KB
   6. Bernstein D
   7. Spudich JA
   8. Ruppel KM

   (2019) β-Cardiac myosin hypertrophic cardiomyopathy mutations release sequestered heads and increase enzymatic activity

   Nature Communications 10:2685.

   https://doi.org/10.1038/s41467-019-10555-9

   • PubMed
   • Google Scholar
3. 1. Aksel T
   2. Choe Yu E
   3. Sutton S
   4. Ruppel KM
   5. Spudich JA

   (2015) Ensemble force changes that result from human cardiac myosin mutations and a small-molecule effector

   Cell Reports 11:910–920.

   https://doi.org/10.1016/j.celrep.2015.04.006

   • PubMed
   • Google Scholar
4. 1. Al-Khayat HA
   2. Kensler RW
   3. Squire JM
   4. Marston SB
   5. Morris EP

   (2013) Atomic model of the human cardiac muscle myosin filament

   PNAS 110:318–323.

   https://doi.org/10.1073/pnas.1212708110

   • PubMed
   • Google Scholar
5. 1. Alamo L
   2. Ware JS
   3. Pinto A
   4. Gillilan RE
   5. Seidman JG
   6. Seidman CE
   7. Padrón R

   (2017) Effects of myosin variants on interacting-heads motif explain distinct hypertrophic and dilated cardiomyopathy phenotypes

   eLife 6:e24634.

   https://doi.org/10.7554/eLife.24634

   • PubMed
   • Google Scholar
6. 1. Anderson RL
   2. Trivedi DV
   3. Sarkar SS
   4. Henze M
   5. Ma W
   6. Gong H
   7. Rogers CS
   8. Gorham JM
   9. Wong FL
   10. Morck MM
   11. Seidman JG
   12. Ruppel KM
   13. Irving TC
   14. Cooke R
   15. Green EM
   16. Spudich JA

   (2018) Deciphering the super relaxed state of human β-cardiac myosin and the mode of action of mavacamten from myosin molecules to muscle fibers

   PNAS 115:E8143–E8152.

   https://doi.org/10.1073/pnas.1809540115

   • PubMed
   • Google Scholar
7. 1. Brown JH
   2. Kumar VSS
   3. O’Neall-Hennessey E
   4. Reshetnikova L
   5. Robinson H
   6. Nguyen-McCarty M
   7. Szent-Györgyi AG
   8. Cohen C

   (2011) Visualizing key hinges and a potential major source of compliance in the lever arm of myosin

   PNAS 108:114–119.

   https://doi.org/10.1073/pnas.1016288107

   • PubMed
   • Google Scholar
8. 1. Chu S
   2. Muretta JM
   3. Thomas DD

   (2021) Direct detection of the myosin super-relaxed state and interacting-heads motif in solution

   The Journal of Biological Chemistry 297:101157.

   https://doi.org/10.1016/j.jbc.2021.101157

   • PubMed
   • Google Scholar
9. 1. Dawson JF
   2. Sablin EP
   3. Spudich JA
   4. Fletterick RJ

   (2003) Structure of an F-actin trimer disrupted by gelsolin and implications for the mechanism of severing

   The Journal of Biological Chemistry 278:1229–1238.

   https://doi.org/10.1074/jbc.M209160200

   • PubMed
   • Google Scholar
10. 1. De La Cruz EM
    2. Ostap EM

    (2009) Kinetic and equilibrium analysis of the myosin ATPase

    Methods in Enzymology 455:157–192.

    https://doi.org/10.1016/S0076-6879(08)04206-7

    • PubMed
    • Google Scholar
11. 1. Debold EP
    2. Schmitt JP
    3. Patlak JB
    4. Beck SE
    5. Moore JR
    6. Seidman JG
    7. Seidman C
    8. Warshaw DM

    (2007) Hypertrophic and dilated cardiomyopathy mutations differentially affect the molecular force generation of mouse alpha-cardiac myosin in the laser trap assay

    American Journal of Physiology. Heart and Circulatory Physiology 293:H284–H291.

    https://doi.org/10.1152/ajpheart.00128.2007

    • PubMed
    • Google Scholar
12. 1. Farman GP
    2. Muthu P
    3. Kazmierczak K
    4. Szczesna-Cordary D
    5. Moore JR

    (2014) Impact of familial hypertrophic cardiomyopathy-linked mutations in the NH2 terminus of the RLC on β-myosin cross-bridge mechanics

    Journal of Applied Physiology (Bethesda, Md 117:1471–1477.

    https://doi.org/10.1152/japplphysiol.00798.2014

    • PubMed
    • Google Scholar
13. 1. Geeves MA
    2. Holmes KC

    (1999) Structural mechanism of muscle contraction

    Annual Review of Biochemistry 68:687–728.

    https://doi.org/10.1146/annurev.biochem.68.1.687

    • PubMed
    • Google Scholar
14. 1. Gourinath S
    2. Himmel DM
    3. Brown JH
    4. Reshetnikova L
    5. Szent-Györgyi AG
    6. Cohen C

    (2003) Crystal structure of scallop Myosin S1 in the pre-power stroke state to 2.6 a resolution: flexibility and function in the head

    Structure (London, England) 11:1621–1627.

    https://doi.org/10.1016/j.str.2003.10.013

    • PubMed
    • Google Scholar
15. 1. Greenberg MJ
    2. Kazmierczak K
    3. Szczesna-Cordary D
    4. Moore JR

    (2010) Cardiomyopathy-linked myosin regulatory light chain mutations disrupt myosin strain-dependent biochemistry

    PNAS 107:17403–17408.

    https://doi.org/10.1073/pnas.1009619107

    • PubMed
    • Google Scholar
16. 1. Greenberg MJ
    2. Shuman H
    3. Ostap EM

    (2014) Inherent force-dependent properties of β-cardiac myosin contribute to the force-velocity relationship of cardiac muscle

    Biophysical Journal 107:L41–L44.

    https://doi.org/10.1016/j.bpj.2014.11.005

    • PubMed
    • Google Scholar
17. 1. Heissler SM
    2. Arora AS
    3. Billington N
    4. Sellers JR
    5. Chinthalapudi K

    (2021) Cryo-EM structure of the autoinhibited state of myosin-2

    Science Advances 7:eabk3273.

    https://doi.org/10.1126/sciadv.abk3273

    • PubMed
    • Google Scholar
18. 1. Himmel DM
    2. Mui S
    3. O’Neall-Hennessey E
    4. Szent-Györgyi AG
    5. Cohen C

    (2009) The on-off switch in regulated myosins: different triggers but related mechanisms

    Journal of Molecular Biology 394:496–505.

    https://doi.org/10.1016/j.jmb.2009.09.035

    • PubMed
    • Google Scholar
19. 1. Ho CY
    2. Carlsen C
    3. Thune JJ
    4. Havndrup O
    5. Bundgaard H
    6. Farrohi F
    7. Rivero J
    8. Cirino AL
    9. Andersen PS
    10. Christiansen M
    11. Maron BJ
    12. Orav EJ
    13. Køber L

    (2009) Echocardiographic strain imaging to assess early and late consequences of sarcomere mutations in hypertrophic cardiomyopathy

    Circulation. Cardiovascular Genetics 2:314–321.

    https://doi.org/10.1161/CIRCGENETICS.109.862128

    • PubMed
    • Google Scholar
20. 1. Homburger JR
    2. Green EM
    3. Caleshu C
    4. Sunitha MS
    5. Taylor RE
    6. Ruppel KM
    7. Metpally RPR
    8. Colan SD
    9. Michels M
    10. Day SM
    11. Olivotto I
    12. Bustamante CD
    13. Dewey FE
    14. Ho CY
    15. Spudich JA
    16. Ashley EA

    (2016) Multidimensional structure-function relationships in human β-cardiac myosin from population-scale genetic variation

    PNAS 113:6701–6706.

    https://doi.org/10.1073/pnas.1606950113

    • PubMed
    • Google Scholar
21. 1. Houdusse A
    2. Cohen C

    (1996) Structure of the regulatory domain of scallop myosin at 2 A resolution: implications for regulation

    Structure (London, England 4:21–32.

    https://doi.org/10.1016/s0969-2126(96)00006-8

    • PubMed
    • Google Scholar
22. 1. Houdusse A
    2. Szent-Gyorgyi AG
    3. Cohen C

    (2000) Three conformational states of scallop myosin S1

    PNAS 97:11238–11243.

    https://doi.org/10.1073/pnas.200376897

    • PubMed
    • Google Scholar
23. 1. Jung HS
    2. Burgess SA
    3. Billington N
    4. Colegrave M
    5. Patel H
    6. Chalovich JM
    7. Chantler PD
    8. Knight PJ

    (2008) Conservation of the regulated structure of folded myosin 2 in species separated by at least 600 million years of independent evolution

    PNAS 105:6022–6026.

    https://doi.org/10.1073/pnas.0707846105

    • PubMed
    • Google Scholar
24. 1. Karabina A
    2. Kazmierczak K
    3. Szczesna-Cordary D
    4. Moore JR

    (2015) Myosin regulatory light chain phosphorylation enhances cardiac β-myosin in vitro motility under load

    Archives of Biochemistry and Biophysics 580:14–21.

    https://doi.org/10.1016/j.abb.2015.06.014

    • PubMed
    • Google Scholar
25. 1. Kawana M
    2. Sarkar SS
    3. Sutton S
    4. Ruppel KM
    5. Spudich JA

    (2017) Biophysical properties of human β-cardiac myosin with converter mutations that cause hypertrophic cardiomyopathy

    Science Advances 3:e1601959.

    https://doi.org/10.1126/sciadv.1601959

    • PubMed
    • Google Scholar
26. 1. Konno T
    2. Chang S
    3. Seidman JG
    4. Seidman CE

    (2010) Genetics of hypertrophic cardiomyopathy

    Current Opinion in Cardiology 25:205–209.

    https://doi.org/10.1097/HCO.0b013e3283375698

    • PubMed
    • Google Scholar
27. 1. Lee KH
    2. Sulbarán G
    3. Yang S
    4. Mun JY
    5. Alamo L
    6. Pinto A
    7. Sato O
    8. Ikebe M
    9. Liu X
    10. Korn ED
    11. Sarsoza F
    12. Bernstein SI
    13. Padrón R
    14. Craig R

    (2018) Interacting-heads motif has been conserved as a mechanism of myosin II inhibition since before the origin of animals

    PNAS 115:E1991–E2000.

    https://doi.org/10.1073/pnas.1715247115

    • PubMed
    • Google Scholar
28. 1. Liu C
    2. Kawana M
    3. Song D
    4. Ruppel KM
    5. Spudich JA

    (2018) Controlling load-dependent kinetics of β-cardiac myosin at the single-molecule level

    Nature Structural & Molecular Biology 25:505–514.

    https://doi.org/10.1038/s41594-018-0069-x

    • PubMed
    • Google Scholar
29. 1. McNamara JW
    2. Li A
    3. Dos Remedios CG
    4. Cooke R

    (2015) The role of super-relaxed myosin in skeletal and cardiac muscle

    Biophysical Reviews 7:5–14.

    https://doi.org/10.1007/s12551-014-0151-5

    • PubMed
    • Google Scholar
30. 1. Moore JR
    2. Leinwand L
    3. Warshaw DM

    (2012) Understanding cardiomyopathy phenotypes based on the functional impact of mutations in the myosin motor

    Circulation Research 111:375–385.

    https://doi.org/10.1161/CIRCRESAHA.110.223842

    • PubMed
    • Google Scholar
31. 1. Nag S
    2. Sommese RF
    3. Ujfalusi Z
    4. Combs A
    5. Langer S
    6. Sutton S
    7. Leinwand LA
    8. Geeves MA
    9. Ruppel KM
    10. Spudich JA

    (2015) Contractility parameters of human β-cardiac myosin with the hypertrophic cardiomyopathy mutation R403Q show loss of motor function

    Science Advances 1:e1500511.

    https://doi.org/10.1126/sciadv.1500511

    • PubMed
    • Google Scholar
32. 1. Nag S
    2. Trivedi DV
    3. Sarkar SS
    4. Adhikari AS
    5. Sunitha MS
    6. Sutton S
    7. Ruppel KM
    8. Spudich JA

    (2017) The myosin mesa and the basis of hypercontractility caused by hypertrophic cardiomyopathy mutations

    Nature Structural & Molecular Biology 24:525–533.

    https://doi.org/10.1038/nsmb.3408

    • PubMed
    • Google Scholar
33. 1. Nag S
    2. Trivedi DV

    (2021) To lie or not to lie: Super-relaxing with myosins

    eLife 10:e63703.

    https://doi.org/10.7554/eLife.63703

    • PubMed
    • Google Scholar
34. 1. Pylypenko O
    2. Houdusse AM

    (2011) Essential “ankle” in the myosin lever arm

    PNAS 108:5–6.

    https://doi.org/10.1073/pnas.1017676108

    • PubMed
    • Google Scholar
35. 1. Rayment I
    2. Holden HM
    3. Whittaker M
    4. Yohn CB
    5. Lorenz M
    6. Holmes KC
    7. Milligan RA

    (1993a) Structure of the actin-myosin complex and its implications for muscle contraction

    Science (New York, N.Y.) 261:58–65.

    https://doi.org/10.1126/science.8316858

    • PubMed
    • Google Scholar
36. 1. Rayment I
    2. Rypniewski WR
    3. Schmidt-Bäse K
    4. Smith R
    5. Tomchick DR
    6. Benning MM
    7. Winkelmann DA
    8. Wesenberg G
    9. Holden HM

    (1993b) Three-dimensional structure of myosin subfragment-1: A molecular motor

    Science (New York, N.Y.) 261:50–58.

    https://doi.org/10.1126/science.8316857

    • PubMed
    • Google Scholar
37. 1. Robert-Paganin J
    2. Auguin D
    3. Houdusse A

    (2018) Hypertrophic cardiomyopathy disease results from disparate impairments of cardiac myosin function and auto-inhibition

    Nature Communications 9:4019.

    https://doi.org/10.1038/s41467-018-06191-4

    • PubMed
    • Google Scholar
38. 1. Rohde JA
    2. Roopnarine O
    3. Thomas DD
    4. Muretta JM

    (2018) Mavacamten stabilizes an autoinhibited state of two-headed cardiac myosin

    PNAS 115:E7486–E7494.

    https://doi.org/10.1073/pnas.1720342115

    • PubMed
    • Google Scholar
39. 1. Sarkar SS
    2. Trivedi DV
    3. Morck MM
    4. Adhikari AS
    5. Pasha SN
    6. Ruppel KM
    7. Spudich JA

    (2020) The hypertrophic cardiomyopathy mutations R403Q and R663H increase the number of myosin heads available to interact with actin

    Science Advances 6:eaax0069.

    https://doi.org/10.1126/sciadv.aax0069

    • PubMed
    • Google Scholar
40. 1. Scarff CA
    2. Carrington G
    3. Casas-Mao D
    4. Chalovich JM
    5. Knight PJ
    6. Ranson NA
    7. Peckham M

    (2020) Structure of the shutdown state of myosin-2

    Nature 588:515–520.

    https://doi.org/10.1038/s41586-020-2990-5

    • PubMed
    • Google Scholar
41. 1. Schindelin J
    2. Arganda-Carreras I
    3. Frise E
    4. Kaynig V
    5. Longair M
    6. Pietzsch T
    7. Preibisch S
    8. Rueden C
    9. Saalfeld S
    10. Schmid B
    11. Tinevez J-Y
    12. White DJ
    13. Hartenstein V
    14. Eliceiri K
    15. Tomancak P
    16. Cardona A

    (2012) Fiji: an open-source platform for biological-image analysis

    Nature Methods 9:676–682.

    https://doi.org/10.1038/nmeth.2019

    • PubMed
    • Google Scholar
42. 1. Semsarian C
    2. Ingles J
    3. Maron MS
    4. Maron BJ

    (2015) New perspectives on the prevalence of hypertrophic cardiomyopathy

    Journal of the American College of Cardiology 65:1249–1254.

    https://doi.org/10.1016/j.jacc.2015.01.019

    • PubMed
    • Google Scholar
43. 1. Sherwood JJ
    2. Waller GS
    3. Warshaw DM
    4. Lowey S

    (2004) A point mutation in the regulatory light chain reduces the step size of skeletal muscle myosin

    PNAS 101:10973–10978.

    https://doi.org/10.1073/pnas.0401699101

    • PubMed
    • Google Scholar
44. 1. Sommese RF
    2. Sung J
    3. Nag S
    4. Sutton S
    5. Deacon JC
    6. Choe E
    7. Leinwand LA
    8. Ruppel K
    9. Spudich JA

    (2013) Molecular consequences of the R453C hypertrophic cardiomyopathy mutation on human β-cardiac myosin motor function

    PNAS 110:12607–12612.

    https://doi.org/10.1073/pnas.1309493110

    • PubMed
    • Google Scholar
45. 1. Spudich JA
    2. Watt S

    (1971) The Regulation of Rabbit Skeletal Muscle Contraction

    Journal OF Biological Chemistry 246:4866–4871.

    https://doi.org/10.1016/S0021-9258(18)62016-2

    • PubMed
    • Google Scholar
46. 1. Spudich JA
    2. Sivaramakrishnan S

    (2010) Myosin VI: an innovative motor that challenged the swinging lever arm hypothesis

    Nature Reviews. Molecular Cell Biology 11:128–137.

    https://doi.org/10.1038/nrm2833

    • PubMed
    • Google Scholar
47. 1. Spudich JA

    (2015) The myosin mesa and a possible unifying hypothesis for the molecular basis of human hypertrophic cardiomyopathy

    Biochemical Society Transactions 43:64–72.

    https://doi.org/10.1042/BST20140324

    • PubMed
    • Google Scholar
48. 1. Spudich JA

    (2019) Three perspectives on the molecular basis of hypercontractility caused by hypertrophic cardiomyopathy mutations

    Pflugers Archiv 471:701–717.

    https://doi.org/10.1007/s00424-019-02259-2

    • PubMed
    • Google Scholar
49. 1. Stewart MA
    2. Franks-Skiba K
    3. Chen S
    4. Cooke R

    (2010) Myosin ATP turnover rate is a mechanism involved in thermogenesis in resting skeletal muscle fibers

    PNAS 107:430–435.

    https://doi.org/10.1073/pnas.0909468107

    • PubMed
    • Google Scholar
50. 1. Sung J
    2. Nag S
    3. Mortensen KI
    4. Vestergaard CL
    5. Sutton S
    6. Ruppel K
    7. Flyvbjerg H
    8. Spudich JA

    (2015) Harmonic force spectroscopy measures load-dependent kinetics of individual human β-cardiac myosin molecules

    Nature Communications 6:7931.

    https://doi.org/10.1038/ncomms8931

    • PubMed
    • Google Scholar
51. Book

    1. Sung J
    2. Mortensen KI
    3. Spudich JA
    4. Flyvbjerg H

    (2017) How to Measure Load-Dependent Kinetics of Individual Motor Molecules Without a Force-Clamp n Single-Molecule Enzymology

    In: Spies M, Chemla YRB, editors. Nanomechanical Manipulation and Hybrid Methods. Academic Press. pp. 1–29.

    https://doi.org/10.1016/bs.mie.2016.08.002

    • Google Scholar
52. 1. Trybus KM

    (1991) Assembly of cytoplasmic and smooth muscle myosins

    Current Opinion in Cell Biology 3:105–111.

    https://doi.org/10.1016/0955-0674(91)90172-u

    • PubMed
    • Google Scholar
53. 1. Trybus KM
    2. Freyzon Y
    3. Faust LZ
    4. Sweeney HL

    (1997) Spare the rod, spoil the regulation: necessity for a myosin rod

    PNAS 94:48–52.

    https://doi.org/10.1073/pnas.94.1.48

    • PubMed
    • Google Scholar
54. 1. Tyska MJ
    2. Hayes E
    3. Giewat M
    4. Seidman CE
    5. Seidman JG
    6. Warshaw DM

    (2000) Single-molecule mechanics of R403Q cardiac myosin isolated from the mouse model of familial hypertrophic cardiomyopathy

    Circulation Research 86:737–744.

    https://doi.org/10.1161/01.res.86.7.737

    • PubMed
    • Google Scholar
55. 1. Uyeda TQ
    2. Abramson PD
    3. Spudich JA

    (1996) The neck region of the myosin motor domain acts as a lever arm to generate movement

    PNAS 93:4459–4464.

    https://doi.org/10.1073/pnas.93.9.4459

    • PubMed
    • Google Scholar
56. 1. Vander Roest AS
    2. Liu C
    3. Morck MM
    4. Kooiker KB
    5. Jung G
    6. Song D
    7. Dawood A
    8. Jhingran A
    9. Pardon G
    10. Ranjbarvaziri S
    11. Fajardo G
    12. Zhao M
    13. Campbell KS
    14. Pruitt BL
    15. Spudich JA
    16. Ruppel KM
    17. Bernstein D

    (2021) Hypertrophic cardiomyopathy β-cardiac myosin mutation (P710R) leads to hypercontractility by disrupting super relaxed state

    PNAS 118:e2025030118.

    https://doi.org/10.1073/pnas.2025030118

    • PubMed
    • Google Scholar
57. 1. Wendt T
    2. Taylor D
    3. Trybus KM
    4. Taylor K

    (2001) Three-dimensional image reconstruction of dephosphorylated smooth muscle heavy meromyosin reveals asymmetry in the interaction between myosin heads and placement of subfragment 2

    PNAS 98:4361–4366.

    https://doi.org/10.1073/pnas.071051098

    • PubMed
    • Google Scholar
58. 1. Woodhead JL
    2. Zhao F-Q
    3. Craig R
    4. Egelman EH
    5. Alamo L
    6. Padrón R

    (2005) Atomic model of a myosin filament in the relaxed state

    Nature 436:1195–1199.

    https://doi.org/10.1038/nature03920

    • PubMed
    • Google Scholar
59. 1. Yang S
    2. Tiwari P
    3. Lee KH
    4. Sato O
    5. Ikebe M
    6. Padrón R
    7. Craig R

    (2020) Cryo-EM structure of the inhibited (10S) form of myosin II

    Nature 588:521–525.

    https://doi.org/10.1038/s41586-020-3007-0

    • PubMed
    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "HCM mutations in the pliant and light chain-binding regions of the lever arm of human β-cardiac myosin have divergent effects on myosin function" for consideration by eLife. Your article has been reviewed by 4 peer reviewers, including James R Sellers as Reviewing Editor and Reviewer #2, and the evaluation has been overseen by Anna Akhmanova as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: E Michael Ostap (Reviewer #3).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

Thank you for submitting your manuscript to eLife. It was examined by three expert reviews and myself. The general conclusion is that the work is thorough, significant and, for the most part, very well documented. All of the reviewers felt that if their concerns can be adequately addressed, the work merits publication.

There were three major concerns. The first two deals with whether you have appropriate statistical power to back up your claims. One reviewer was concerned that the number of biological replicates is only two and this is magnified by the second concern which is the significant variability in the values reported in Table S1 for the WT sample. For example, the kcat varies from 4.64 to 6.04 /S in the two WT biological replicates and their Km values show even more variability for that replicate. Since all the data are normalized to the WT control, this raises concerns. In the manuscript you attribute this to "assay variation", but this degree of difference is surprising and reinforces the need for a third replicate. All of the reviewers appreciate the work involved in prepping and assaying six proteins and so we did not ask this lightly. Can you distinguish between assay variation and the case where the first WT replicate simply had less active motors? The data reported for the mutants are much tighter, but even their variation makes us wonder whether the small differences reported as deviations from WT are valid. Please either provide data from another replicate or give us strong reasons why we should not be concerned over this.

The third major concern deals with the optical trapping results. One reviewer commented that you should show some representative primary data showing single interactions and that a plot of displacement distributions would be important to see, given that these mutations are in the lever arm. More detail on the assumptions that go into the fittings of the harmonic force-spectroscopy data is also asked for.

In addition, each reviewer had specific items that should be addressed.

Reviewer #2 (Recommendations for the authors):

P.22 Figure S3 Legend. Authors in discussing A797 stated "This region has previously been implicated in the formation of the folded state" and later on that F834 which is in the ankle joint that "this joint has been previously shown to be important in autoinhibition". I'm assuming that the folded state and autoinhibition are both referring to the IHM state. Also, no reference is given for these statements.

Table S1 and P.7, line 5. The actin titrations of the myosin's ATPase activity aretypically fit to a rectangular hyperbola such as that used for analysis of the substrate dependence of enzyme kinetics with the Michaelis-Menten formalism. However, actin is not a substrate for myosin and so the term describing the actin concentration at half maximal Vmax should not be called a Km. Many authors use Kactin or Katpase to describe this value.

Reviewer #3 (Recommendations for the authors):

The authors need to better outline the assumptions that go into the fitting of the harmonic force-spectroscopy data. Would the analysis find non-linear mechanical behaviors of the lever caused by the mutation? What if the force-dependence didn't follow Bell's-Law like kinetics? I think the technique is totally appropriate for the study, but its limitations should be outlined.

What are the uncertainties on the plots in Figure 3A and Figure 4.

The Plots in Figures 2 and 5 are plotted as normalized values. Why not plot the ATPase values? It would be concerning if this is because there is too much variability from the assays.

It would be useful for Table S1 to report the mutant25-Hep/WT25-Hep kcat ratio.

Reviewer #4 (Recommendations for the authors):

The critique below list specific questions for the authors to consider, ordered as they occur in the manuscript. Quotes indicate text in the manuscript on the indicated line.

Page 2 line 8. Regarding "Frequent cause of hypertrophic…" can the authors state how frequent is frequent?

Page 2 line 12. Please avoid ambiguous and qualitative terms such as "relatively high" and "frequent" (see previous comment) as they are… ambiguous.

Page 3 line 14. Regarding "there may be conditions" please state what these conditions may be.

Page 4 line 6. Why were these mutations chosen to study in comparison to other mutations in this region of the protein?

Page 6 line 1. General comment on the Results in this manuscript:

In general, the quality of the work is high. However, I have reservations about the publication of manuscripts that are based solely on two biological replicates of recombinantly expressed protein. Most of the line-by-line critiques below reflect that concern. The paper uses four experiments (ATPase, single-turnover, motility, and single-molecule force spectroscopy) to draw conclusions. How many independent biological replicates of protein expression and purification were used for these experiment? Where there only two in total that were tested across all experiments? This should be stated more clearly in the methods and the number of biological replicates increased as stated in Public Review.

Page 7 line 7. How is the 16+/-9% being computed here? Is this from a single replicate or from the normalized data compared across the 2 biological replicates. Please state more clearly in the text how the effect size and error is computed. The reported effect sizes should be from multiple biological replicates.

Page 7 line 15. In Figure 2, why are the ATPase data displayed as representative ATPase curves? The data have been normalized and so the biological replicates should be able to be plotted on the same curve. Doing so increases transparency and quality of presentation. The normalized biological replicates should be able to be displayed on the same plot. Each replicate could be fit independently, or a single fit to the N = 2 (preferably N > 2) set of data fit to a single function.

Page 7 line 21. Please increase biological replicates used in these studies. In the N = 2 experiments noted here what is the +/- error being presented?

Page 8 line 3. Regarding "mean +/- SD." State what mean is being plotted and what SD is being plotted. Is this the mean of the three technical replicates from a single biological replicate or the mean of all data across multiple biological replicates. If it is biological replicates, SD is a rather limited error estimate as there are only two replicates.

Page 10 line 9. Are the single molecule experiments performed with biological replicates? Please do this if it wasn't done. The data do not make it clear which data sets in the single molecule experiments were from independent biological replicates. Please state how the reported mean values and SEM are computed in relationship to biological replicates. They should be computed across independent biological replicates as should statistical significance.

Page 11 Figure 3A. How many molecules are these curves determined from? Please include confidence intervals for these curves. Presumably, these plots should be determined from data across all biological replicates.

Page 11 Figure 3. As asked above, please explain which data are from a single biological replicate and which are from multiple replicates. The methods states that the single molecule experiments were performed from multiple biological replicates. Can the authors graphically indicate which molecules were from which prep? A table similar to S1, summarizing the statistics of the single molecule measurements would be highly valuable.

Page 13 Figure 4. Please provide confidence intervals with these curves and propagate the uncertainty from experimental measurements that they are calculated from, including uncertainty from biological replicates into the determination of these fundamental biophysical properties.

Page 14 Figure 6. As with Figure 2 ATPase data, please depict data of the control normalized biological replicates, not a single "representative" data set.

Page 15 Figure S2. Please provide additional detail on the experiment such as what replicate the data points come from. Do the data come from a single biological replicate or multiple biological replicates. Ideally, the data should be from more than 2 biological replicates.

Page 17 line 3. Is the N listed in this line and in Table S2, technical replicates or biological replicates? How many different preparations of protein contributed to this data?

Page 29 line 22-25. The authors state, "we have previously observed that additional biological replicates do not typically differ when the two biological replicates match" as justification for not providing additional replicates. Given the intrinsic uncertainty in all experiments performed in this study, please perform more biological replicates and ensure that all data presented in the paper are from more than two biological replicates of recombinant protein expression.

Page 30 line 17-21. Were the single turnover experiments performed using material obtained from more than one biological replicate? Please do so as indicated elsewhere in the review.

Essential revisions:

Thank you for submitting your manuscript to eLife. It was examined by three expert reviews and myself. The general conclusion is that the work is thorough, significant and, for the most part, very well documented. All of the reviewers felt that if their concerns can be adequately addressed, the work merits publication.

There were three major concerns. The first two deals with whether you have appropriate statistical power to back up your claims. One reviewer was concerned that the number of biological replicates is only two and this is magnified by the second concern which is the significant variability in the values reported in Table S1 for the WT sample. For example, the kcat varies from 4.64 to 6.04 /S in the two WT biological replicates and their Km values show even more variability for that replicate. Since all the data are normalized to the WT control, this raises concerns. In the manuscript you attribute this to "assay variation", but this degree of difference is surprising and reinforces the need for a third replicate. All of the reviewers appreciate the work involved in prepping and assaying six proteins and so we did not ask this lightly. Can you distinguish between assay variation and the case where the first WT replicate simply had less active motors? The data reported for the mutants are much tighter, but even their variation makes us wonder whether the small differences reported as deviations from WT are valid. Please either provide data from another replicate or give us strong reasons why we should not be concerned over this.

We appreciate the reviewers’ concerns with variability in the ATPase assays—we ourselves have previously noted this variability and understand why it may give pause. Our expression system allows us to express and highly purify human β-cardiac myosin fragments using mouse C2C12 myoblasts. Over the past ~11 years of using this system, we have observed variability in our measured k_cat values for both single- and double-headed WT myosin fragments (sS1, S1, 2hep, and 25hep) that we have sought to understand and minimize. We believe the single largest contribution to this variability is likely differences in post-translational modifications (PTMs) to myosin that occur during their production in the myoblast cell line. Indeed, we have performed mass spectrometry on selected purified samples and have seen that the list of modifications is extensive (>>100). While this finding is interesting, we have not pursued it both because of prohibitive cost and more importantly because such modifications, occurring in a heterologous expression system, are of unclear physiological relevance to the PTMs that occur in human cardiomyocytes. Please note that the most extensively studied PTM of the myosin motor, phosphorylation of the RLC, is controlled for in all our studies as we exchange on a bacterially expressed human RLC that is of course unphosphorylated. This stands in contrast to other groups who use this expression system but leave the endogenous mouse skeletal RLCs in place, with no control for phosphorylation status.

We hypothesize that alterations in PTMs may be due to subtle differences in cell density or differentiation state despite standardized protocols for cell growth and differentiation. These changes in turn may be secondary to slight differences in growth factors, etc. This issue has been exacerbated during the pandemic because of interruptions to lab access and supply chains. For example, the experiments reported in this manuscript were carried out over a period of 2-2.5 years, during which different batches of C2C12 cells (both from same ATCC stock), DMEM, fetal bovine serum and horse serum were necessarily used for myoblast cell culture.

Despite these issues, we have found over the course of studying over 50 mutations in human βcardiac myosin – that comparing proteins grown and harvested from the same batch of cells at the same time under the same conditions minimizes the effect of this variability and allows reproducible comparisons of the biochemical and biomechanical effects of any given mutant to its simultaneously prepared and assayed WT control. Indeed, we have seen remarkably consistent % changes from WT for a given mutation despite variations in absolute activity. Therefore, all of the data in this paper compares proteins that were expressed in parallel as described above. Please note that this means that the WT 2-hep to which the editor alludes was prepared independently on 12 different occasions (12 biological replicates) rather than two. We have also repeated the WT 2hep and WT 25hep protein preparations so that the updated data includes 5 biological replicates of this pair of proteins.

To address concerns around variability and number of replicates, we have done the following:

1. We have now provided three additional replicates of WT 2- vs. 25-hep to further validate the difference between WT 2- and 25-hep. These are reported in table S1 as replicates 3 – 5. When only two replicates were shown, the WT 25-hep/WT 2-hep ratio was 0.63 ± 0.10; with the addition of three biological replicates, the average ratio is now 0.61 ± 0.10. This additional data can be found in Figure 3 —figure supplement 1, and Supplementary File 1.

2. We include a new supplemental figure (Figure 3 —figure supplement 1) that shows all the nonnormalized ATPase replicates. Our goal in including the normalized data in the main text is to allow the reader to focus on the relevant differences between WT and mutants. Note that this normalization is to simultaneously prepared and assayed WT protein—we do not normalize to protein that was prepared on a different day or from a different batch of cells. The normalization eliminates slight background variability due to myoblast growth conditions, temperature, or protein age (as described above), allowing comparison of relevant differences between mutant and WT proteins. However, we recognize the value in including the raw data and have thus added Figure 3 —figure supplement 1.

We hope that these revisions help to allay the reviewers’ concerns about the variability in the ATPase results. The results that are used to make conclusions are statistically significant.

The third major concern deals with the optical trapping results. One reviewer commented that you should show some representative primary data showing single interactions and that a plot of displacement distributions would be important to see, given that these mutations are in the lever arm. More detail on the assumptions that go into the fittings of the harmonic force-spectroscopy data is also asked for.

We thank the reviewers for this valuable suggestion. We have now added a figure (Figure 4 —figure supplement 1) in the supporting information with representative primary data for each myosin assayed. The figure also includes plots of displacement distributions. We have also modified the text (page 22 line 23 to line 27) to better describe the assumptions that go into the fittings of harmonic force-spectroscopy data.

In addition, each reviewer had specific items that should be addressed.

Reviewer #2 (Recommendations for the authors):

P.22 Figure S3 Legend. Authors in discussing A797 stated "This region has previously been implicated in the formation of the folded state" and later on that F834 which is in the ankle joint that "this joint has been previously shown to be important in autoinhibition". I'm assuming that the folded state and autoinhibition are both referring to the IHM state. Also, no reference is given for these statements.

We have adjusted the text in the figure legend to refer more explicitly to the IHM state and have added references for these statements. We thank the reviewer for this comment.

Table S1 and P.7, line 5. The actin titrations of the myosin's ATPase activity aretypically fit to a rectangular hyperbola such as that used for analysis of the substrate dependence of enzyme kinetics with the Michaelis-Menten formalism. However, actin is not a substrate for myosin and so the term describing the actin concentration at half maximal Vmax should not be called a Km. Many authors use Kactin or Katpase to describe this value.

We have changed instances of “KM” to “Kapp” (as in, apparent K). We thank the reviewer for this comment.

Reviewer #3 (Recommendations for the authors):

The authors need to better outline the assumptions that go into the fitting of the harmonic force-spectroscopy data. Would the analysis find non-linear mechanical behaviors of the lever caused by the mutation? What if the force-dependence didn't follow Bell's-Law like kinetics? I think the technique is totally appropriate for the study, but its limitations should be outlined.

We thank the reviewer for this valuable suggestion. Various aspects of the strengths and limitations associated with the harmonic force spectroscopy technique and assumption involved in the fitting of the harmonic force spectroscopy data has been described in detail previously by our research group (Sung, J. et al., Methods in Enzymology, 2010, 475, 321-375; Sun, J. et al., Nature Communications, 2015, 6, Article number 7931; Liu, C. et al., Nature Structural Molecular Biology, 2018, 25, 505-514; and Vander Roest, A. S. et al., Proceedings of the National Academy of the National Academy of Sciences U.S.A., 2021, 118, e2025030118.). However, we agree that a detailed explanation of certain strengths and weaknesses of the HFS method and corresponding data analysis would certainly improve the current manuscript and help the readers to perceive our results with greater insight. We have now modified the method section accordingly, to address this concern (page 22 line 23 to line 27).

What are the uncertainties on the plots in Figure 3A and Figure 4.

The uncertainty in Figure 3A (now Figure 4A) is now shown as dotted lines. We have also included the individual fits for each molecule, which is now shown in Figure 4 —figure supplement 1 M, N, O, P. The uncertainties for Figure 4 (now Figure 5) are also now shown as dotted lines. We thank the reviewer for this comment.

The Plots in Figures 2 and 5 are plotted as normalized values. Why not plot the ATPase values? It would be concerning if this is because there is too much variability from the assays.

We have now included full non-normalized values as Figure 3 —figure supplement 1. The purpose of showing normalized values in the main text is not to conceal any variability (as evidenced by the fact that we had already transparently included all of the individual fitted parameters for each experiment in table S1 – now Supplementary File 1), but to allow readers to focus on the relevant differences between mutant and WT.

It would be useful for Table S1 to report the mutant 25-Hep/WT 25-Hep kcat ratio.

Table S1 (now Supplementary File 1) is designed for readers to be able to see the real ATPase values for each independent biological replicate. As described in the methods, biological replicates of each mutant protein were only prepared with a paired WT 2-hep control, not a WT 25-hep control. Data for WT 25-hep was collected separately with its own WT 2-hep controls, as shown in the bottom row of the table. It would be misleading and inappropriate to normalize the mutant 25-hep values to the WT 25-hep values from separate, unpaired protein preparations. The rightmost column of Supplementary File 1 (average mutant 25-hep/mutant 2-hep ratios) is designed to allow the reader to make the relevant comparison.

Reviewer #4 (Recommendations for the authors):

The critique below list specific questions for the authors to consider, ordered as they occur in the manuscript. Quotes indicate text in the manuscript on the indicated line.

Page 2 line 8. Regarding "Frequent cause of hypertrophic…" can the authors state how frequent is frequent?

Based on a simple ClinVar search of HCM mutations, we found that 73 unique positions in the myosin head are associated with likely pathogenic or confirmed pathogenic HCM mutations out of a total 777 amino acids (~9%), while in the lever arm, 11 unique positions out of a total 61 amino acids (~18%) in the lever arm are associated with likely pathogenic or confirmed pathogenic HCM mutations, suggesting ~2x enrichment for mutations in the lever arm. However, we know that the ClinVar database does not include all known HCM mutations, and that the levels of evidence vary somewhat from submitter to submitter, so this is a relatively weak analysis that we do not feel is adequate for publication. Thus, we have relied on the word “frequent” to convey this sentiment. We believe a separate bioinformatic analysis would be warranted to determine how much more frequent these mutations are, but we believe that analysis is outside the scope of the current work.

Page 2 line 12. Please avoid ambiguous and qualitative terms such as "relatively high" and "frequent" (see previous comment) as they are… ambiguous.

See comment above. While we appreciate the importance of avoiding ambiguity, an inherent challenge of understanding the frequency of mutations in a human disease is the lack of available high-quality data that exhaustively annotates every known mutation and whether the mutation is benign, potentially pathogenic, or definitively pathogenic. However, our rudimentary analysis suggests that “relatively high” and “frequent” are warranted. We are not aware of any publication that rigorously assesses the frequency of HCM-causing lever arm mutations.

Page 3 line 14. Regarding "there may be conditions" please state what these conditions may be.

We have added the parenthetical “for example, in the presence of drugs,” which is supported by reference 35, where they showed a difference in the relative increase in IHM vs. SRX upon the addition of mavacamten. Anecdotally, we have also noted that omecamtiv mecarbil may uncouple the relationship between the structural state and SRX-like rates (unpublished data).

The phrase is intentionally ambiguous because the goal is to point out that an individual rate constant can arise from multiple protein states; an individual rate is, by rule, not necessarily definitive proof of a specific structural state. In this case, there’s no reason to believe that an SRXlike rate couldn’t arise from any number of different structural states myosin may take. This point is discussed at-length in our prior publication (see ref. 24).

Page 4 line 6. Why were these mutations chosen to study in comparison to other mutations in this region of the protein?

We have added justification for selecting these mutations (page 4 lines 7-10), and we thank the reviewer for this comment.

Page 6 line 1. General comment on the Results in this manuscript:

In general, the quality of the work is high. However, I have reservations about the publication of manuscripts that are based solely on two biological replicates of recombinantly expressed protein. Most of the line-by-line critiques below reflect that concern. The paper uses four experiments (ATPase, single-turnover, motility, and single-molecule force spectroscopy) to draw conclusions. How many independent biological replicates of protein expression and purification were used for these experiment? Where there only two in total that were tested across all experiments? This should be stated more clearly in the methods and the number of biological replicates increased as stated in Public Review.

We hope that the analysis and adjustments reported in the “essential revisions” section will help allay the reviewer’s concern in this regard. The results reported in this paper represent a great number of individual biological replicates across all samples and mutants.

Page 7 line 7. How is the 16+/-9% being computed here? Is this from a single replicate or from the normalized data compared across the 2 biological replicates. Please state more clearly in the text how the effect size and error is computed. The reported effect sizes should be from multiple biological replicates.

16 ± 9% is a composite of the two biological replicates. To compute the error, the errors of the fit for each biological replicate (17 ± 11% and 15 ± 7%) were propagated using the standard formula for error propagation in a mean:${\displaystyle {\mathrm{\Delta }\chi }_{\text{est}}=\frac{\sqrt{{\mathrm{\Delta }\chi }^{2}1+\mathrm{\Delta }{x}^{2}2}}{\sqrt{}N}}$ Error propagation was used instead of SD due to the limitation of only having two biological replicates, which makes SD inappropriate. Note that error propagation resulted in larger errors as compared to computed standard deviations for the biological replicates (which is appropriate in this context). We have clarified our use of error propagation in the corresponding methods section.

Page 7 line 15. In Figure 2, why are the ATPase data displayed as representative ATPase curves? The data have been normalized and so the biological replicates should be able to be plotted on the same curve. Doing so increases transparency and quality of presentation. The normalized biological replicates should be able to be displayed on the same plot. Each replicate could be fit independently, or a single fit to the N = 2 (preferably N > 2) set of data fit to a single function.

Due to edits suggested in the “essential revisions” section, we have now included Figure 3 —figure supplement 1, which shows all of the non-normalized replicates for all of the reported actin-activated ATPase results. We hope this will adequately address the reviewer’s concern around transparency.

From a technical perspective, plotting the individual replicates on the same chart results in a different fitting result as compared to plotting the replicates separately and averaging the resulting fitting parameters (even when the experiments are both normalized). While there are numerous mathematical reasons for this difference, a key technical reason in the context of these experiments is that the KM’s vary across biological replicates. Additionally, increasing the number of datapoints in a fitted function artificially reduces the standard error of the fit, making composite data appear misleadingly certain. For these reasons, we do not think it would be appropriate to fit data from separate biological replicates to a single function.

Page 7 line 21. Please increase biological replicates used in these studies. In the N = 2 experiments noted here what is the +/- error being presented?

This is the same data described in page 7 line 7 above, which we address in the response to the comment above. We recognize that the increase of 16% is marginal, but it is statistically significant and therefore relevant to report (though statistical significance is not necessarily indicative of biological significance). In fact, we would have preferred if this increase was not significant, as it’s surprising on its own that a mutation in the pliant region, fairly distal from the active site, could have an effect on the ATPase rate. Nevertheless, when differences are statistically significant, we have an obligation to report them.

The conclusions in the paper on the function of D778V do not hinge on this individual increase of 16%, but instead take into account all of the observed effects of the D778V mutation on motor function, including findings from the optical trapping and in vitro motility. For example, in figure 4C (overall power output), if the 16% increase in ATPase is not used to compute power output and instead we use an equivalent ATPase to the WT, the power output for D778V is still greater than that of WT. The increased power output is primarily due to the increased velocity of D778V which is, in turn, due to its increased detachment rate observed in single molecule optical trapping. The increased velocity of D778V as measured by the function of detachment rate and step size (equation 5) is corroborated by the observed increase in in vitro motility velocity.

Page 8 line 3. Regarding "mean +/- SD." State what mean is being plotted and what SD is being plotted. Is this the mean of the three technical replicates from a single biological replicate or the mean of all data across multiple biological replicates. If it is biological replicates, SD is a rather limited error estimate as there are only two replicates.

Standard deviations here represent the standard deviation of the data points shown in the chart, which collectively come from at least 3 separate protein preparations for each protein (as noted above in Author response table 1 in the response to the question on page 6 line 1). This has been clarified in the figure legend.

In the context of the in vitro motility experiments (and optical trapping experiments), the definitions of technical and biological replicates are not necessarily immediately evident. For example, for each slide prepared, three separate imaging frames are averaged together to collectively determine the MVEL of a given slide channel. In that sense, each data point shown is already a summary of triplicate measurements. Additionally, some people would consider aliquots of the same protein prep but thawed and assayed on separate days using new stocks of actin and other motility proteins as separate biological replicates. In light of this complexity, it makes most sense to treat each slide as an individual replicate, which is what we have done here. As mentioned above, each slide is measured in triplicate.

Page 10 line 9. Are the single molecule experiments performed with biological replicates? Please do this if it wasn't done. The data do not make it clear which data sets in the single molecule experiments were from independent biological replicates. Please state how the reported mean values and SEM are computed in relationship to biological replicates. They should be computed across independent biological replicates as should statistical significance.

Single molecule experiments were performed across 2-3 independent protein preparations, as noted in the table in the response to the reviewer’s question above (page 6 line 1). We have also added Figure 4 —figure supplement 2, which shows the data points colored to indicate which independent protein preparation the molecules originated from. Molecules from unique protein preparations do not appear to systematically differ from molecules from separate protein preparations.

In many ways, the optical trapping experiments present a similar complication in the definition of technical and biological replicates as the in vitro motility experiments, discussed above in response to the question for page 8 line 3. For each molecule, multiple independent binding events are collected (see newly included figures Figure 4 —figure supplement 1 E-H, where each data point represents an individual binding event from an individual molecule). These measurements are summarized into a force-velocity fit for each molecule (Figure 4 —figure supplement 1 I-L), which represents summary data from many individual measurements. Therefore, it makes most sense to treat each molecule as an independent replicate, which is what we have done here. While error bars for each molecule are not shown to preserve visual clarity, each molecule does have its own error measurement for k₀ and δ. (Step size d is calculated as an averaged summary of all of the events from an individual molecule; thus each molecule does not have a measured error in step size.)

Page 11 Figure 3A. How many molecules are these curves determined from? Please include confidence intervals for these curves. Presumably, these plots should be determined from data across all biological replicates.

The curves in Figure 3A show the average fit across all molecules. The number of molecules for each protein has been added to the figure legend, and we have now shown the error of the averages as dotted lines in Figure 3A. The fits for each individual molecule are now also shown in Figure 4 —figure supplement 1 M-P.

Page 11 Figure 3. As asked above, please explain which data are from a single biological replicate and which are from multiple replicates. The methods states that the single molecule experiments were performed from multiple biological replicates. Can the authors graphically indicate which molecules were from which prep? A table similar to S1, summarizing the statistics of the single molecule measurements would be highly valuable.

We have added Figure 4 —figure supplement 2, which shows the data points colored to indicate which independent protein preparation the molecules originated from. We thank the reviewer for this comment.

Page 13 Figure 4. Please provide confidence intervals with these curves and propagate the uncertainty from experimental measurements that they are calculated from, including uncertainty from biological replicates into the determination of these fundamental biophysical properties.

Figure 4 (now Figure 5) now shows confidence intervals for each curve, which are propagated from the uncertainty of experimental measurements of each protein. We thank the reviewer for this comment.

Page 14 Figure 6. As with Figure 2 ATPase data, please depict data of the control normalized biological replicates, not a single "representative" data set.

We have replaced the main text versions of these figures with plots showing the fitted curves for each kinetic assay, along with an average fitted curve for each WT and mutant 25-hep. We have also added the same plots for the supplemental figure showing WT and mutant 2-hep single turnover results (now Figure 7 —figure supplement 2), and moved the representative fitted curves to a new supplemental figure (now Figure 7 —figure supplement 1). We thank the reviewer for this comment.

Page 15 Figure S2. Please provide additional detail on the experiment such as what replicate the data points come from. Do the data come from a single biological replicate or multiple biological replicates. Ideally, the data should be from more than 2 biological replicates.

We hope the table shown in our response to the reviewer’s earlier point (page 6 line 1) helps clarify biological and technical replicates for these experiments. We have also now noted the number of unique protein preparations for each protein construct in the corresponding methods section. We thank the reviewer for this comment.

Page 17 line 3. Is the N listed in this line and in Table S2, technical replicates or biological replicates? How many different preparations of protein contributed to this data?

As above, we hope Author response table 1 in our response to the reviewer’s earlier point (page 6 line 1) helps clarify, along with the adjustment to the methods section.

Page 29 line 22-25. The authors state, "we have previously observed that additional biological replicates do not typically differ when the two biological replicates match" as justification for not providing additional replicates. Given the intrinsic uncertainty in all experiments performed in this study, please perform more biological replicates and ensure that all data presented in the paper are from more than two biological replicates of recombinant protein expression.

We hope that our response to the “essential revisions” has helped allay the reviewer’s concern in this regard.

Page 30 line 17-21. Were the single turnover experiments performed using material obtained from more than one biological replicate? Please do so as indicated elsewhere in the review.

As above, we hope Author response table 1 in response to the reviewer’s earlier point (page 6 line 1) helps clarify, along with the adjustment to the methods section.

Author details

1. Makenna M Morck

   1. Stanford Cardiovascular Institute, Stanford University School of Medicine, Stanford, United States
   2. Department of Biochemistry, Stanford University School of Medicine, Stanford, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Validation, Writing - original draft

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-7561-892X

2. Debanjan Bhowmik

   1. Stanford Cardiovascular Institute, Stanford University School of Medicine, Stanford, United States
   2. Department of Biochemistry, Stanford University School of Medicine, Stanford, United States

   Present address

   Transdisciplinary Research Program, Rajiv Gandhi Centre for Biotechnology, Thiruvananthapuram, India

   Contribution

   Data curation, Formal analysis, Investigation, Methodology, Writing – review and editing

   Competing interests

   No competing interests declared

3. Divya Pathak

   1. Stanford Cardiovascular Institute, Stanford University School of Medicine, Stanford, United States
   2. Department of Biochemistry, Stanford University School of Medicine, Stanford, United States

   Contribution

   Investigation, Writing – review and editing

   Competing interests

   No competing interests declared

4. Aminah Dawood

   1. Stanford Cardiovascular Institute, Stanford University School of Medicine, Stanford, United States
   2. Department of Biochemistry, Stanford University School of Medicine, Stanford, United States

   Contribution

   Resources

   Competing interests

   No competing interests declared

5. James Spudich

   1. Stanford Cardiovascular Institute, Stanford University School of Medicine, Stanford, United States
   2. Department of Biochemistry, Stanford University School of Medicine, Stanford, United States

   Contribution

   Conceptualization, Funding acquisition, Writing – review and editing

   Competing interests

   JAS is cofounder and on the Scientific Advisory Board of Cytokinetics, Inc, a company developing small molecule therapeutics for treatment of hypertrophic cardiomyopathy

6. Kathleen M Ruppel

   1. Stanford Cardiovascular Institute, Stanford University School of Medicine, Stanford, United States
   2. Department of Biochemistry, Stanford University School of Medicine, Stanford, United States

   Contribution

   Conceptualization, Formal analysis, Investigation, Project administration, Resources, Supervision, Writing – review and editing

   For correspondence

   kruppel@stanford.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-0971-5331

• 748

  Page views

• 129

  Downloads

• 3

  Citations

Article citation count generated by polling the highest count across the following sources: PubMed Central, Crossref, Scopus.