The Mesa Trail and the Interacting Heads Motif of Myosin II

Abstract

Myosin II molecules in the thick filaments of striated muscle form a structure in which the heads interact with each other and fold back onto the tail. This structure, the “interacting heads motif” (IHM), provides a mechanistic basis for the auto-inhibition of myosin in relaxed thick filaments. Similar IHM interactions occur in single myosin molecules of smooth and nonmuscle cells in the switched-off state. In addition to the interaction between the two heads, which inhibits their activity, the IHM also contains an interaction between the motor domain of one head and the initial part (subfragment 2, S2) of the tail. This is thought to be a crucial anchoring interaction that holds the IHM in place on the thick filament. S2 appears to cross the head at a specific location within a broader region of the motor domain known as the myosin mesa. Here, we show that the positive and negative charge distribution in this part of the mesa is complementary to the charge distribution on S2. We have designated this the “mesa trail” owing to its linear path across the mesa. We studied the structural sequence alignment, the location of charged residues on the surface of myosin head atomic models, and the distribution of surface charge potential along the mesa trail in different types of myosin II and in different species. The charge distribution in both the mesa trail and the adjacent S2 is relatively conserved. This suggests a common basis for IHM formation across different myosin IIs, dependent on attraction between complementary charged patches on S2 and the myosin head. Conservation from mammals to insects suggests that the mesa trail/S2 interaction plays a key role in the inhibitory function of the IHM.

Introduction

Myosin II is a motor protein that, in conjunction with actin, produces movement in both muscle and nonmuscle cells. The molecule is a hexamer consisting of two identical heavy chains and four light chains (Fig. 1A). The N-terminal half of each heavy chain (HC) is folded into a globular head (subfragment 1 or S1), consisting of a motor domain (MD) and a regulatory domain (RD), while the C-terminal half consists of a long α-helical region that dimerizes with the other HC to form an extended coiled-coil tail. An essential light chain (ELC) and regulatory light chain (RLC) is associated with the regulatory domain of each head. Under physiological conditions, the tail regions self-associate to form polymeric “thick” filaments—the functional form of myosin II in muscle and nonmuscle contractile systems (Fig. 1C). The motor domains, lying on the surface of the thick filament, contain both actin-binding and ATPase sites, and are responsible for cell motility and muscle contraction via actin-myosin sliding interactions powered by ATP.

Figure 1. The structure of myosin molecules and filaments.

A. Schematic representation of the myosin molecule (based on [40]). Heavy chains are green and blue (MDs shown as ellipsoids); ELCs (yellow) and RLCs (red) are in the neck region (regulatory domain). LMM shows proteolytic tail fragment (light meromyosin). Other labels defined in text. B. Schematic representation of myosin molecule folded into the 10S conformation, with blocked and free heads forming an IHM; yellow asterisk marks region of interaction between BH and FH. The first segment of the tail (S2) interacts with the mesa region on the back surface of the BH. BH consists of heavy chain (green), ELC (orange), and RLC (yellow). FH shows heavy chain (blue), ELC (magenta), and RLC (red). Based on [11, 19]. C. Density map of 3D reconstruction of tarantula thick filament showing IHM with fitted atomic model of BH (green HC, orange ELC, yellow RLC), FH (cyan HC, pink ELC and beige RLC), and proximal portion of S2 (red). Yellow asterisk as in (B). Unpublished data from [1].

The interacting heads motif (IHM)

In relaxed muscle, the myosin heads on the thick filament surface fold back on to the tail and interact with each other in a conformation known as the “interacting heads motif” (IHM) [1] (Fig. 1C). In the IHM, the actin-binding region of one head (the “blocked” head or BH) is blocked by an interaction with the motor domain of the other (“free”) head (FH), preventing binding of the BH to actin. The FH, in turn, has its converter domain covered by interaction with the BH [5] (Fig. 1C). These asymmetric interactions suggest a structural basis for the auto-inhibition of myosin in relaxed muscle, by inhibition of actin binding in the BH and ATP hydrolysis in the FH [5]. The interactions also sequester the heads close to the thick filament surface, minimizing contact with actin. It has been suggested that the IHM may be the structural basis for the super-relaxed state (SRX) of myosin [6–8].

Myosin II can also exist in a monomeric (non-filamentous) form, in one of two conformations, known as 6S and 10S, based on their sedimentation rates in the analytical ultracentrifuge [9, 10]. At high ionic strength (0.5 M), myosin is in the 6S form, in which the tail is extended, with flexibly attached heads [11]. At physiological ionic strength in the presence of ATP, smooth and nonmuscle myosin II adopt the 10S conformation, in which the tail folds back on itself at two hinge points, and the heads interact with each other (as in the IHM) and with specific regions of the folded tail (Fig. 1B) [3, 5, 10, 12–14]. The ionic strength dependence of the conformation suggests that the main interactions causing folding are electrostatic. The interactions between the two heads in the folded molecule [5] are essentially the same as those in the IHM of relaxed thick filaments [1] (Fig. 1B cf. 1C). Filaments [1, 15–17] and molecules [3–5, 18–20] from a wide range of species and tissues show a similar IHM structure, implying that the IHM is fundamental to myosin II function.

The mesa trail

In addition to interaction between the two heads in the IHM, there is an interaction, caused by their backward folding, between the blocked head and the proximal region of the tail (subfragment 2, or S2) (Figs. 1B, 2A) [1]. This appears to be a crucial anchoring interaction [21] that stabilizes the IHM structure, as myosin lacking this region of tail is not inhibited [22]. The blocked head/S2 interaction is common to the IHM of both thick filaments and folded myosin II molecules [1, 3, 14], and is also seen in the myosin proteolytic fragment, heavy meromyosin (HMM; Fig. 1A), which has S2 but lacks the distal two thirds of the tail (LMM, Fig. 1A) [3]. Structural observations suggest that S2 lying across the blocked head would sterically inhibit binding of the free head to actin as well as stabilizing the blocked head in its folded-back configuration [14]. In filaments, it may additionally serve to hold the myosin heads close to the filament backbone [1].

Figure 2. Atomic models showing the course of the mesa trail.

A. Ribbon representation of 3JBH atomic model of tarantula myosin IHM showing FH and BH HCs (cyan and green respectively), associated ELCs and RLC, and S2. Based on [1]. B. Ribbon representation of the blocked head motor domain of 3JBH showing the structural regions involved in the mesa trail and the adjacent overlying region of S2 (transparent solid). Beta strands 6 and 7 (cyan), Helix W (green), Helix-Loop-Helix motif (red), loop 2 (magenta), Helix O/β-strand 5 linker (pink), Helix E/β-strand 4 linker (orange). C. Sphere representation of blocked head motor domain of 3JBH showing charged residues on surface of mesa trail and of adjacent overlying region of S2. Negative residues are red and positive residues are blue. D. Same as (C) but showing coulombic surface charge representation. The path of the mesa trail is delineated by yellow dashed lines. E. Surface charge distribution in atomic model of tarantula myosin showing putative complementary patches of opposite charge on S2 and the adjacent MD on the BH (ellipses). Surface on S2 would be flipped around the yellow vertical axis to create the S2/mesa trail interface.

The specific side chains involved in the S2/BH interaction are not yet known, as neither myosin nor HMM has been crystallized or resolved at atomic resolution by cryo-EM. However, with the current level of information, it appears that S2 crosses the blocked head at a specific location [1, 23, 24] within a broader region of the motor domain known as the myosin mesa – a flat region suggested to be involved in multiple protein-protein interactions [25]. We have designated this region of interaction with S2 the “mesa trail” owing to its narrow, linear path across the wider mesa. Structural and sequence analysis has previously shown conservation of the mesa in vertebrate skeletal and cardiac myosin IIs, but not in smooth and nonmuscle myosins [25]. This has led to the suggestion that the mesa may have evolved specifically for interaction with myosin binding protein C (MyBP-C), a thick filament component present in skeletal and cardiac muscle, but absent from smooth and nonmuscle cells. Here, we make a detailed study specifically of the trail region of the mesa, comparing structure and sequence across multiple myosin IIs and species. We show that there is good conservation of positive and negative charge distribution along the mesa trail, which appears to be complementary to the charge distribution (also conserved) on the corresponding region of S2. This conservation occurs not only across skeletal and cardiac, but also smooth, nonmuscle, and invertebrate myosin IIs. We conclude that a common structural mechanism, based on electrostatic attraction, underlies mesa trail-S2 interaction across all the species we have studied. Conservation suggests that this interaction plays a critical role in the functioning of myosin II.

Results

Our goal was to determine whether there is a conserved interaction path between S2 and the myosin mesa in different types of muscle and different species. To accomplish this, we examined the three-dimensional distribution of surface residues and surface charge potential in available myosin head atomic structures (Table S1) using UCSF Chimera [26], and compared equivalent residues in the same regions.

Surface charge analysis

We first examined the path of the S2/mesa interaction in the only structure where this has been explicitly observed – the IHM in the cryo-EM structure of tarantula thick filaments [1]. PDB 3JBH is a pseudo-atomic structure derived from fitting a tarantula myosin homology model to the cryo-EM density map of the tarantula filament [24]. The location and distribution of charged residues exposed on the surface of S2 and its putative areas of contact with the BH MD were analyzed in Chimera (Fig. 2). S2 exhibits a strong band of negative charge known as Ring 1 [27] where it crosses the BH (Figs. 2C, E). Analysis of charged amino acid residues on the surface of the BH shows several that could putatively interact with this part of S2 (Fig. 2C) [1, 25, 27]. The distribution of the associated surface charge potential indicates that there is a high concentration of positive charge in the central portion of this region (Fig. 2D), which could interact with the overall negative Ring 1. Although positively charged side chains predominate in this region of the BH, there are also some smaller groups of negative residues. Detailed comparison suggests that these could interact with small complementary positive charge patches on S2 (Fig. 2E; Movie 1).

We designate the region of putative interaction between S2 and the BH mesa [25] the mesa trail because of its narrow, extended path across the mesa. We refer to the region of the mesa trail adjacent to Ring 1 of S2 as the central trail, the N-terminal region as the proximal trail and the C-terminal region as the distal trail (Fig. 2C). The proximal trail includes a portion of the C-terminal of Helix O, the Helix O/β-strand 5 linker, Helix E/β-strand 4 linker and β-strands 6 and 7 on the surface of the upper 50K domain (Fig. 2B; Movies 2, 3). The central trail includes the surface of the C-terminal region of Loop 2, Helix W, and the N-terminal region of the Helix-Loop-Helix motif in the Lower 50K domain. The distal trail includes the remaining surface of the Helix-Loop-Helix including the Activation Loop and the H-Loop.

We compared the mesa trail of the BH in 3D crystallographic structures of myosin heads from a variety of species and muscle types, all of which exhibit an IHM. Our goal was to see whether there was a consistent pattern of charged surface residues and associated surface charge potential that could provide a common basis for the interaction of S2 with the BH. Nine S1 models were analyzed, covering a range of myosin II molecules from vertebrate cardiac, skeletal, smooth and nonmuscle cells to invertebrate striated muscle and Dictyostelium myosin II (Table S1). Myosin-5, which does not form an IHM [28], was included as a negative control.

Charged surface residues populated the mesa trail for all species studied, including myosin-5 (Fig. 3A). The charge distribution indicated the presence of patches of positive charge in the central region of the trail for all the myosin IIs that exhibit an IHM (Fig. 3B). In contrast, Dictyostelium, which forms only a partial IHM [4], and especially myosin-5, which does not form an IHM at all ([28], had predominately negative surface charge distributions over most of the trail. This suggests that IHM formation may be dependent on complementary charge interactions between S2 (mostly negative) and the BH mesa trail (mostly positive) in all species that form an IHM.

Figure 3. Atomic models of the blocked head motor domains of the nine myosins studied.

A. Shows the charged surface residues lying under the path of S2 (as in Fig. 2C). B. As in (A) but showing the coulombic surface potential (as in Fig. 2D).

Conservation analysis

In addition to studying the location of charged residues and the distribution of surface charge patches on S2 and the mesa trail in atomic models for different myosin IIs, we also compared the amino acids in the same structural locations. This allowed us to quantify the degree of charge conservation along the mesa trail and S2. The atomic models were first structurally aligned in 3D using UCSF Chimera (Fig. S1) and the results used as a guide for the alignment of the sequences. We then compared the aligned sequences of the models (Fig. 4). The sequences of myosins where 3D structural information was unavailable were not included as there is not always a direct relationship between sequence conservation and structural conservation (including surface exposure), which is crucial to understanding surface interactions.

Figure 4. Alignment of amino acid surface residues on the mesa trail and associated S2 region for the nine myosins analyzed in this study.

The residues were aligned according to the equivalent locations in 3D space in the respective atomic models. The charged surface residues are highlighted in blue for positive (lysine and arginine), light blue for partial positive (histidine) and red for negative (aspartic and glutamic acid). Charged amino acids in the alignment that were structurally not accessible for potential interaction with S2 are highlighted in very light blue or red. The residues are grouped according to location on the mesa trail and S2.

Sequence analysis showed significant conservation of charged residues exposed on the surface of the mesa trail, consistent with the structural analysis described above. The degree of conservation was categorized according to the number of myosins having conserved charge at each location (Table 1; Fig. 5). The seven species exhibiting full IHM conformations were used as a reference so that scoring was not biased towards the partial IHM species (Dictyostelium) or the species that has no IHM (myosin 5a). High conservation was defined as conservation of surface charge in at least six of the seven myosins at any given location, moderate conservation as 4 or 5, and low conservation as 2 or 3 out of the 7 myosins. Lack of conservation (i.e. only one species had a charged residue at a specific location) had two categories: a unique residue of either charge if none of the other species had charged residues, or a charge reversal if charge conservation was exhibited among the other species. High conservation was the largest category, with similar degrees of conservation among all species, except for Dictyostelium and myosin 5a, where there was a progressive decrease (Table 1, Fig. 5). A similar pattern was observed for moderate conservation, except that smooth and nonmuscle myosin were similar to Dictyostelium and myosin 5a. There appeared to be no definitive pattern for low conservation. There were almost no residues that were unique or exhibited charge reversal for any species, except Dictyostelium and myosin 5a, where there were significant numbers in both categories (Fig. 5, red rectangles). The highest proportion of conserved charges appeared to be located in the central trail and the corresponding Ring 1 of S2 (Fig. 5), with positive residues predominating on the central trail and negative residues on Ring 1.

Table 1.

Frequency of charge conservation in the mesa trail and S2

	Conservation
	High			Moderate			Low			Unique			Reversal
Species	MT	S2	All	MT	S2	All	MT	S2	All	MT	S2	All	MT	S2	All
Cardiac	14	10	24	7	3	10	6	4	10	0	0	0	0	0	0
Skeletal	14	10	24	7	3	10	5	4	9	0	0	0	0	0	0
Tarantula	12	10	22	8	4	12	6	3	9	1	1	2	0	0	0
Scallop	14	8	22	7	5	12	6	4	10	1	0	1	0	0	0
Insect	12	9	21	7	3	10	2	3	5	0	0	0	0	0	0
Smooth	14	10	24	3	3	6	4	6	10	0	0	0	0	0	0
Nonmuscle	14	10	24	3	3	6	6	7	13	1	0	1	0	0	0
Dictyostelium	10	6	16	4	3	7	3	7	10	1	4	5	2	5	7
Myosin 5a	9	4	13	3	2	5	5	6	11	8	4	12	2	6	8

Table shows frequency of charged residue conservation on the surface of the mesa trail (MT) and associated S2 based on the results of Fig. 4. High, moderate, low, unique and reversal are defined in the text.

Figure 5. Sequence alignment of mesa trail and S2 surface residues.

The data are the same as in Fig. 4, but grouped according to the degree of sequence conservation (cf. Table 1). This serves to highlight the high degree of charge conservation across all animal species in the mesa trail and S2 (long bracket) and low conservation in Dictyostelium and myosin 5a (short bracket), which show partial or no IHM formation respectively. The latter myosins showed a much larger number of charge reversals or unique residues (red rectangles).

In summary, the species that exhibited a complete IHM appeared to have the most conservation of charged residues on the surface of the mesa trail. Myosin 5a (which does not form an IHM [28]) had the least conservation whereas Dictyostelium (which forms a partial IHM [4]) was intermediate between the two groups.

Discussion

IHM interactions

The IHM is the result of interaction of the blocked head MD with the free head MD and ELC, and of the blocked head MD with S2. The head-head interface involves BH actin-binding loops that interact with the FH catalytic domain, converter, and ELC [1, 5, 19, 21, 24, 29]. There is also evidence that the RLCs from the BH and FH may form an “interaction complex” with the initial portion of S2 [30]. Structural and mutational data suggest that the S2/BH MD interaction involves the distal half of the initial 11 heptads of S2 interacting with a specific region of the BH MD, which we have designated the mesa trail, and that this may be a major factor contributing to IHM stability [1, 22, 27]. The IHM interactions are likely to be electrostatic, as they are abolished at high ionic strength [11]. Structural data show that they are also relatively weak [1, 4, 15], consistent with the known properties of ionic bonds in proteins. No single side chain interaction alone is likely to account for the overall stability of the IHM and it is likely that the IHM interactions involve broader patches of complementary charge rather than specific individual side chains. Isolated S2 fragments bind weakly to myosin S1, indicating that there are significant attractive forces between these regions even in the absence of the parent molecule, and this binding is stronger at lower ionic strength [31]. These observations are consistent with the mesa trail concept and the involvement of ionic bonds in its interaction with S2.

The mesa trail

The BH/S2 interaction, which we propose occurs along the mesa trail, appears to be essential to the IHM conformation. This is suggested by the observation that it can be observed in both myosin and HMM even when the BH/FH interaction is broken [3, 32]. Thus BH/S2 interaction is apparently stronger than the head-head interaction. It has been suggested that the BH/S2 interaction may be crucial for initial IHM formation while BH/FH interaction contributes to IHM stability [21, 24]. Any perturbation of the BH/S2 interaction is therefore likely to disrupt the whole IHM conformation.

The mesa trail lies on a broader area of the MD surface (the mesa) that has been proposed as a candidate for specific protein-protein interactions [25]. The mesa contains parts of several subdomains, helices and loops that exhibit a predominance of basic residues exposed on the surface. It has been shown that a large area of the mesa is highly conserved in vertebrate cardiac and skeletal myosins but not in smooth and nonmuscle myosins [25]. This has led to the proposal that the mesa is a binding site for MyBP-C, which is present in cardiac and skeletal muscles but not in smooth and nonmuscle cells [25, 31]. Our structural analysis shows that when only the trail region of the mesa is considered, there is good conservation of charge not only across vertebrate cardiac and skeletal muscles, but also vertebrate smooth, nonmuscle, and invertebrate muscle myosins (Fig. 3). Many regions within both the trail and S2 also exhibit a high degree of sequence conservation among the species studied (Figs. 4, 5), supporting our structural observations. We conclude that the trail is a specialized region of the mesa that is conserved across all species and tissue types in which myosin II folds into an IHM, and is likely to be the region of the BH that interacts with S2 to stabilize this conformation. Conservation of charge distribution along the trail, and its complementarity to S2 charge distribution, suggests that S2-trail interaction plays a fundamental role in myosin II function in all animals.

Further evidence that the mesa trail is an essential component of the IHM is provided by myosin 5a and Dictyostelium myosin II. In myosin 5a, which does not form an IHM [28], there is significantly less sequence conservation of both the trail and the corresponding region of S2 (Fig. 5; Table 1). This is reflected in the lack of a distinct mesa trail in the surface charge distribution, which is predominately negative in this region (Fig. 3B). Dictyostelium, which only forms a partial IHM [4], supports this view, with lower conservation of both surface charge and sequence than the animal myosin IIs (Figs. 3, 5).

The mesa trail contains a relatively high proportion of basic residues, giving it a predominately positive charge (Fig. 2D). The results show broadly similar distribution patterns of charged side chains in the mesa trails from different species, but with significant variation in the surface potential distribution (Fig. 3). Only two atomic models for the N-terminal region of S2 are available (cardiac S2 crystal [27] and tarantula S2 homology [24]), but all of the charged residues are known to be on the surface. Therefore, the linear sequences of S2 can be used directly for comparison with the mesa trails. However, the rotational orientation of S2 about its long axis where it runs over the trail cannot be accurately determined for the IHM at the resolutions currently available, so only the relative axial location of charge patches along S2 in relation to the mesa trail charge patches can be considered.

Inspection of the locations of the charged surface residues along the mesa trails indicates that they tend to form two distinct lines (most noticeable in skeletal and cardiac), one on each side of the trail (Fig. 3A). If S2 is positioned over the space between the lines, there may be possible pairs of side chain interactions with the pairs of residues located in lateral positions on each of the two chains of the S2 coiled coil. However, the charged residues along the trail are not arranged symmetrically (Figs. 2C, 3A), and any corresponding pairs of side chains on S2 may therefore need also to be asymmetric to maximize the interactions. S2 has a relatively flexible coiled-coil structure with a tendency for disorder towards the N-terminus (where S2 connects to the heads), and the N-terminal 9 heptads have a uniquely asymmetric coiled-coil structure [27]. These 60 residues cover the distance from the S1/S2 junction to the distal edge of the BH (Fig. 2A), and the asymmetry between the two chains of S2 may facilitate electrostatic interactions between S2 and the S1 surface. A caveat to the significance of the two lines of charged residues along the trail is that when surface charge potential is calculated, charged patches occupy the center of the trail as well (Fig. 3B), suggesting electrostatic interaction of the face of S2 with the trail. A possible complementary match of the positive and negative surface charge patches on the BH MD and S2 for tarantula HMM is shown in Fig. 2E and Movie 1.

The mesa trails in the BH and FH are in different environments in the IHM. While the BH trail is involved in the interaction with S2, probably providing the initial foundation for IHM formation, the trail in the FH is completely exposed and does not take part in head-head or head-tail interactions (Fig. S2). Much of the mesa trail/S2 interaction interface is highly conserved, but some parts are not. The regions of moderate and low conservation may be involved in interactions only related to the FH or relate to differences in the strength of the BH/S2 interaction between species. For example, the IHM (and therefore presumably the BH/S2 interaction) appears to be weaker in vertebrate skeletal and cardiac myosins than in regulated myosins [19]. This could be reflected in differences in the respective sequences and charge distributions of the mesa trails. Also, in addition to the highly conserved regions shared by all species that exhibit an IHM, there are subgroups of highly conserved regions that are only shared between certain species (Fig. 6). Striated myosins share a small subgroup with high conservation predominately in the mesa trail, which may possibly be related to FH interactions in helical thick filaments. The vertebrate striated myosins contain a subgroup (also mostly in the trail) which may be related to interactions with MyBP-C. The invertebrate striated myosins share subgroups of mesa trail residues that may be related to their relatively higher IHM stability and also a small subgroup of S2 residues with smooth and nonmuscle myosin that may impart additional IHM stability. Smooth and nonmuscle myosins share a relatively large, exclusive subgroup that is evenly distributed between the mesa trail and S2, which may contribute to the high IHM stability and super-relaxed state of these regulated myosins.

Figure 6. Charged residue conservation along the mesa trail and associated S2.

Conservation is categorized according to species having a complete IHM and within those, groups specific to different types of myosin, shown with black outlines. From top, these groups are 1. striated, smooth, nonmuscle; 2. Striated only; 3. Vertebrate striated only; 4. Invertebrate striated only; 5. Invertebrate, smooth and nonmuscle; 6. Smooth and nonmuscle only.

A recent cryo-EM reconstruction of insect flight muscle thick filaments provides a valuable test of the significance of the BH/S2 interaction. The IHM in this specialized muscle has a quite different orientation from that observed in all other muscles to date [33], and, most significantly, there appears to be no interaction between S2 and the blocked head. Interestingly, the IHM formed by isolated myosin molecules from insect flight muscle is much less stable than for insect skeletal muscle (the type included in our analysis here) [4]. This might be due to the absence of the S2/BH interaction, as in the filaments. This correlates with weaker negative charge of flight muscle myosin S2 in its region of contact with the blocked head when compared with other species [34], supporting the importance of charge attraction in creating the S2/mesa trail interface. Sequence comparison also suggests a lower level of charge attraction in the IHM interfaces of flight compared with skeletal myosin in insects [24, 34]. However, an X-ray crystal model is not available to allow a direct 3D comparison.

S2 interactions in regulation and the SRX

The IHM is an important conformation for regulating myosin activity. “Regulated” myosins [35] appear to have more stable IHMs than “unregulated” myosins, in both single molecules and filaments [1, 4, 15, 18, 19]. The importance of the BH/S2 interaction in regulating myosin is suggested by the loss of regulation if S2 is shortened to 7 heptads or fewer [22], consistent with the observation that an S2 of at least 9 heptads is required to bridge the distance from the S1/S2 junction, over the mesa trail, to the distal boundary of the BH (Fig. 2A). In unregulated myosin systems, the principal regulatory switch resides on the thin filaments [35], resulting in a reduced necessity for a strong IHM, as is found experimentally [15, 19]. However, the IHM may function as more than a simple regulatory switch. There is evidence that it may be the structural basis for the highly inhibited, energy-conserving super-relaxed state (SRX) of both vertebrate and invertebrate muscles [6–8, 36]. The BH/S2 interaction is also a component of the disordered-relaxed state (DRX), where the FH is mobile [37].

Mesa Trail Mutations

Several mutations in β-cardiac myosin (MYH7 gene) leading to hypertrophic cardiomyopathy (HCM) are in highly conserved locations in the mesa trail [21, 25, 38, 39]. Each of these mutations results in a loss of positive charge (Table 2), which would be expected to weaken S2 binding to the blocked head and thus weaken the IHM as a whole. This is supported by R453C, R249Q and H251N HCM mutations, which appear to lower S1/S2 binding affinity [31, 38]. These residues are located in the proximal region of the mesa trail and are conserved across the striated muscle myosins (Fig. 5). Reduction in the affinity of S2 for the BH may make more heads available for interaction with actin, thus leading to the hypercontractility characteristic of HCM [21, 31]. While loss of positive charge in the mesa trail leads to HCM and hypercontractility, the replacement of negative charge by positive charge (E525K) leads to the only case of dilated cardiomyopathy (DCM) related to mutations in the mesa trail (Table 2). Here it is possible that increased positive charge in the trail increases attraction of the negatively charged S2, making the IHM more stable than the wild type. This could make fewer heads available for interaction with actin, thus contributing to the reduced contractility found in DCM [40]. The impact of these single residue mutations at conserved sites in the mesa trail and S2 serves to emphasize the critical role played by interactions between the mesa trail and S2 in normal muscle function, and suggests why these regions have been so well conserved.

Table 2.

Mutations in cardiac myosin heavy chain mesa trail and S2 (MYH7) that lead to cardiac disease

Mutation	Disease	Location	Conservation	Reference
R169G	HCM	MT Proximal	High	[25]
R204H	HCM	MT Proximal	Low	[41]
R249Q	HCM	MT Proximal	High	[41]
H251N	HCM	MT Proximal	Moderate	[38]
K450E	HCM	MT Proximal	High	[41]
R453C	HCM	MT Proximal	Moderate	[31]
E525K	DCM	MT Central	High	[25]
R652G	HCM	MT Central	High	[21]
K657Q	HCM	MT Central	High	[21]
R663H	HCM	MT Central	High	[41]
E894G	HCM	S2 Ring 1	High	[41]
E903K	HCM	S2 Ring 1	High	[41]
D906G	HCM	S2 Ring 1	Low	[31]

MT, mesa trail; HCM, hypertrophic cardiomyopathy; DCM, dilated cardiomyopathy.

Conclusion

Our analysis shows that there is good conservation of surface charge on the mesa trail and S2 across all the species and types of myosin II we have studied that exhibit an IHM. This suggests that mesa trail/S2 interaction is a critical requirement for the normal functioning of myosin in striated, smooth and nonmuscle cells across the animal kingdom. The interaction appears to be essential to the stability of the IHM, which sequesters myosin heads on the thick filament surface in the relaxed or super-relaxed state. The importance of the mesa trail/S2 interaction is highlighted by diseases in cardiac muscle that result from mutations likely to weaken the interaction, leading to hypercontractility and HCM, and those that might strengthen it, resulting in hypocontractility and DCM. The one known exception to the existence of this interaction, insect flight muscle myosin, shows no interaction between S2 and the blocked head, and this correlates with a reduced negative charge in the part of S2 that would cross the blocked head. Insect flight muscle can contract more than one hundred times per second, and may need to ensure rapid availability of heads to actin by avoiding their binding to S2.

Methods

To determine whether there is a conserved interaction path between S2 and the myosin mesa, we examined the three-dimensional distribution of surface residues and of surface charge potential in different myosin head atomic structures using UCSF Chimera [26], and compared residues in the same locations.

Although the resolution of atomic structures is sufficient to determine the precise location of specific residues in S1 and S2 crystals, the IHM models produced from docking these structures into EM density maps have lower resolution when the putative intra- and intermolecular interactions are considered. Therefore, only potential interaction sites can be suggested at the currently available resolutions. In addition, the atomic structures were derived from crystals of single heads (S1) and not from two-headed fragments or myosin filaments, which could affect the specific conformation of local regions within the IHM. Thus, our analysis will produce only an approximation of the interactions that may be involved in the stability of the IHM. Because there is not always a direct correspondence between the conservation of surface residues and the distribution of surface charge patches (Figs. 3A, B), we also used the electrostatic potential utility in Chimera to create a map of the distribution of charge patches on the surface, revealing the regional effect of the combination of surface side chains. It should be noted that most of Loop 2 is missing in the atomic models derived from crystals. However, the complete Loop 2 is present in the tarantula homology model, where it gives an apparently higher positive charge than the X-ray models.

Importantly, linear sequence information alone does not reveal whether or not a conserved residue is located on the molecular surface. Even if a sequence appears to be conserved, the 3D structure in that region may not be identical. The local 3D structure can also be affected by residues in adjacent, non-conserved regions. For example, identical residues may be located on the surface in some cases but not accessible to the surface in others. Our 3D analysis of structurally equivalent residues (Figs. 4–6) allowed a degree of quantification of conservation levels, which supported visual inspection of the different structures. We only studied myosins for which a corresponding atomic structure was available by aligning the atomic models in 3D using Chimera (Fig. S1) and comparing the residues at equivalent structural locations.

Highlights.

• The mesa is a region of the myosin II head involved in the interacting-heads motif

• Charge distribution is studied in a region of the mesa that binds myosin S2

• Surface charge and amino acids are well-conserved along this “mesa trail”

• The mesa trail/S2 interaction is a key feature of the interacting-heads motif

• The importance of the mesa trail is revealed by disease mutations in this region