Myorod and myosin were prepared from the posterior adductor of Crenomytilus grayanus (mussel) as described in our previous works [7, 18]. In the case of myorod, the final purification stage was modified. The 33% to 43% ammonium sulfate fraction [18] was dissolved in solution containing (in mM): 500 KCl, 2 EDTA, 2 NaN₃, 0.5 PMSF, 0.05 leupeptin, 1 DTT, 50 Tris-HCl, pH 8.0. The fraction was dialyzed against the following solution (in mM): 75 KCl, 2 EDTA, 2 NaN₃, 5 β-Mercaptoethanol, 50 Tris-HCl, pH 8.0 and precipitated at 15000×g for 30 min. The pellet was dissolved in urea buffer containing 6 M urea and (in mM): 600 KCl, 1 NaN₃, 2 DTT, 0.5 EGTA, 50 Tris-HCl, pH 8.0, and clarified at 25,000×g for 30 min. The supernatant (50 mg of total protein) was loaded onto a gel-filtration chromatography column (Sepharose CL-6B, Sigma, 60 cm) equilibrated by the same urea buffer. Fractions from the first peak (CF1, tubes 12–21), from the second peak (CF2, tubes 23–33) and from the third peak (CF3, tubes 35–41) (Fig 1A) were selected by SDS–PAGE (Fig 1B), combined and exhaustively dialyzed at 4°C against a solution containing (in mM): 30 KCl, 2 NaN₃, 2 DTT, 1 MgCl₂, 20 imidazole-HCl, pH 6.5.

Rabbit skeletal tissues were donated from the vivarium of the G.B. Elyakov Pacific Institute of Bioorganic Chemistry (Vladivostok, Russia). All procedures were approved by the Animal Care Committee of A.V. Zhirmunsky Institute of Marine Biology, Far East Branch of the Russian Academy of Sciences (Protocol N 21 from 08.09.2014). Rabbit skeletal actin was prepared as described previously [12] and was freeze-dried with 10% sucrose as F-actin. Rabbit skeletal muscle myosin was prepared according to Margossian and Lowey [19].

Protein concentrations were determined either by microbiuret method or by Bradford method calibrated with Crenomytilus grayanus myosin and myorod as described [7].

The actomyosin reconstituted from rabbit F-actin and rabbit or mussel myosin in presence or absence of unphosphorylated (UnPhos) or phosphorylated (Phos) myorod was prepared in two modes. In the first case, filaments of myosin and myorod were preformed separately under the same conditions by quick dilution of proteins from solution A (in mM): 500 KCl, 2 MgCl₂, 0.2 EGTA, 1 DTT to 100 mM KCl by solution B containing (in mM): 2 MgCl₂, 2 NaN₃, 0.3 CaCl₂, 20 imidazole-HCl, pH 7.0. The preformed myosin filaments were then mixed with F-actin and preformed UnPhos or Phos myorod filaments and incubated for 20 min at 4°C.

In the second mode AM was prepared by mixing of myosin and F-actin in high ionic strength (solution A) in the presence or absence of UnPhosMR or PhosMR. The mixtures then were diluted to 100 mM KCl by solution B and were incubated for 20 min at 4°C. It was shown that in this particular condition myosin / myorod co-filaments form [7]. The weight ratio of myosin:myorod:F-actin was 1:1:2 in both types of actomyosins.

Myorod was phosphorylated by smMLCK as described previously [16]. Briefly, 3 μM of smMLCK, 7 μM of calmodulin (Sigma), 0.1 mM Ca²⁺ and 0.4–1 mM ATP were added to the protein in solution containing (in mM): 150 KCl, 0.5 MgCl₂, 2 NaN₃, 0.5 DTT, 20 imidazole-HCl, pH 7.0, and incubated for 1 hour at 25°C. Phosphorylated protein was diluted to 30 mM KCl by modified solution B with 0.5 mM MgCl₂ and sedimented at 5000×g for 15 min. The pellet was re-suspended and washed three times by the same solution.

The same procedures, except the addition of smMLCK, were carried out for UnPhosMR, which was used as a control in the Mg²⁺-ATPase assay.

The Mg²⁺-ATPase activity of AM reconstituted from rabbit F-actin, rabbit or mussel myosin and various concentrations of UnPhosMR or PhosMR was determined in a medium containing (in mM): 100 KCl, 0.5 MgCl₂, 0.5 DTT, 0.1 CaCl₂, 10 imidazole-HCl, pH 7.2. In some cases, the co-filaments of myosin/UnPhosMR or myosin/PhosMR were formed prior to the measurement of ATPase activity by mixing the proteins in high ionic strength solutions followed by a decrease of the ionic strength as described above. The ATPase reaction was started after 10 min of incubation at 25°C by adding Mg²⁺-ATP to 0.4 mM and was terminated after 10 min by adding trichloroacetic acid to 5 mM. The amount of P_i liberated was evaluated colorimetrically at 625 nm by the slightly modified method of Fiske and Subbarow as described [12]. Results of ATPase assay are reported as means ± SEM with 3 to 5 different measurements.

Superprecipitation and clearing of actomyosin (AM) were recorded for changing optical density of actomyosin suspension by a laser diffraction particle sizer 3600E (Malvern) in small-angle spectrophotometer mode at 633 nm. The reaction was run by adding 0.5 mM ATP to AM suspension containing: F-actin, 0.2 mg/ml; myosin, 0.1 mg/ml; myorod, 0.1 mg/ml, in solution containing (in mM): 100 KCl, 0.2 EGTA, 0.5 MgCl₂, 1 NaN₃, 20 Tris-HCl, pH 7.2. These conditions simulate the relaxation medium which reduced affinity of myosin to actin [20]. Dissociation of actomyosin bridges in this condition leads to a clearing, meaning a decrease in optical density of actomyosin suspension.

Crude myorod fraction obtained at 33–43% ammonium saturation contained admixtures of myosin and twitchin (Fig 1). To obtain a pure protein this fraction was applied to a gel filtration chromatography, where three main peaks were collected (Fig 1A), pooled accordingly to SDS-PAGE (Fig 1B) and extensively dialyzed. Combined fraction 1 (CF1, enriched in myosin and twitchin), combined fraction 2 (CF2, enriched in myorod with some amount of twitchin and myosin) and combined fraction 3 (CF3, contained pure myorod) were then used in standard Mg²⁺-ATPase assay.

CF1 strongly increased Mg²⁺-ATPase activity of AM formed from skeletal myosin and F-actin, whereas CF2 activated myosin’s ATPase significantly less than CF1 (Fig 2). Finally, CF3 containing pure myorod did not activate myosin’s ATPase activity at all (Fig 2). Thus, we confirmed that pure unphosphorylated myorod did not potentiate Mg²⁺-ATPase activity of skeletal or molluscan [7] myosins.

We recently found that presumably myorod N-terminal phosphorylation increases the actin-activated Mg²⁺-ATPase activity of myosin in the natural complex of molluscan thick filament proteins consisting of twitchin, myosin and myorod [17]. These data suggest that there are conditions under which myorod modulates the actin–myosin interaction in a phosphorylation-dependent manner.

To better understand the mechanism of such a modulation, we tried to reproduce this effect using the purified proteins: UnPhosMR and PhosMR (CF3), rabbit or molluscan myosins and rabbit F-actin. As it is shown in Fig 3A, addition of separately preformed filaments of unPhosMR or PhosMR to the reconstituted AM with preformed myosin filaments had no effect on the myosin’s Mg²⁺-ATPase activity (Fig 3A).

Next, the mode of actomyosin formation was changed. Myosin, UnPhosMR or PhosMR and F-actin were mixed in a high ionic strength solution (500 mM KCl) where myosin and myorod were in monomeric form and subsequently assembled into the co-filaments by quickly decreasing the ionic strength. The resulting complex was examined using the Mg²⁺-ATPase assay. Again, UnPhosMR did not activate Mg²⁺-ATPase activity of myosin, whereas PhosMR significantly increased the Mg²⁺-ATPase activity of myosin (Fig 3B).

Although purified PhosMR modulates myosin’s ATPase activity, it did not affect the mechanical properties of the AM either by itself or in the state of co-assembled into filaments with myosin at “clearing” condition, i.e. under conditions of reduced affinity of myosin to actin (Fig 4B, AM-PhosMR and AM-UnPhosMR). On the other hand, the addition of CF1 fraction (inactivated by urea myosin and twitchin) stimulates superprecipitation of actomyosin (Fig 4A, curve AM + CF1) because these proteins form non-dissociated cross-links with actin.

The hypothesis that co-polymerization of myosin and PhosMR is required for the alteration of myosin’s ATPase activity was verified in a series of experiments in which myosin / myorod co-filaments were formed at different molar ratios of myorod to myosin for a constant myosin concentration.

As shown in Fig 5A phosphorylated myorod caused two-fold increase of actin-activated Mg²⁺-ATPase of myosin for a 1:1 PhosMR / myosin ratio. A higher ratio of PhosMR to myosin leads to a reduction in activity. In contrast, the ATPase of myosin assembled into co-filaments with UnPhosMR remained unchanged up to a 2:1 myorod: myosin ratio; the activity was inhibited by a further increase of the ratio. PhosMR also affected the ATPase activity of rabbit myosin in a similar way as of molluscan myosin (Fig 5B).

In molluscan catch muscle myorod is located on the surface of the paramyosin core of thick filaments [1, 2, 4], most likely forming a complex with myosin and twitchin in vivo. Myosin and myorod have identical C-terminal parts [8], with the same ACD domain responsible for the filament assembly [10], and can interact in vitro to form co-filaments [7]. It was also shown that trace admixtures of myosin in myorod preparations as well as addition of twitchin to myorod drastically alters the structure of resulting filaments [7, 11].

In the current study we confirmed our previous conclusion that pure UnPhosMR does not activate myosin’s actin-activated Mg²⁺-ATPase (Figs 2, 3 and 5). In addition, we showed that PhosMR, if co- assembled into filaments with myosin, is able to potentiate the Mg²⁺-ATPase activity of myosin (Fig 5). This suggests that the properties of the co-filaments were determined not only by the identical C-terminal domains of myosin and myorod, but also by the N-terminus of myorod and, moreover, by its state of phosphorylation. It seems that the potentiation of the myosin activity requires an additional interaction of the phosphorylated N-terminal domain of myorod with myosin heads or the S-2 fragment. This finding is consistent with the higher affinity of a synthetic phosphorylated N-terminal peptide of myorod for myosin filaments compared to that of the unphosphorylated peptide [13].

With increasing concentration of PhosMR, the Mg²⁺-ATPase activity of myosin enhanced rapidly and achieved the maximum at equal molar ratio of PhosMR-to-myosin (Fig 5). We suppose that this elevated activity is determined by the maximum number of additional contacts between myorod and myosin that is established at an equimolar ratio between the proteins assembled into the co-filaments. It could alternatively be achieved through conformational changes within the myosin head or its S-2 fragment, since data obtained from experiments on mechanical properties of AM do not support the increasing of actin-myosin interaction in the presence of PhosMR (Fig 4) as it was shown for twitchin [12].

The inhibition of actin-activated Mg²⁺-ATPase of myosin resulting from a further increase of myorod content (2:1 myorod:myosin and higher) might also occur due to adjacent myosin molecules being more ordered with higher ratio of myorod in co-filaments. In this scenario, the binding and cycling of one myosin head on actin might exert a force on adjacent myosin-actin cross-bridges, which could inhibit their ATPase activity. A similar inhibitory effect on the Mg²⁺-ATPase activity of myosin with increasing myosin rod content in co-filaments was observed by Stepkowski et al. [21].

It should be noted that the potentiation of myosin’s ATPase in the presence of PhosMR does not depend on the source of myosin during the formation of AM. It is likely that molluscan myosin does not specialize for an interaction with myorod. However, the activating effect was more pronounced in co-filaments assembled from rabbit myosin and PhosMR (Fig 5).

The current understanding of the catch mechanism is based on the phosphorylation-dependent interaction of twitchin with actin without the participation of additional proteins [3, 12, 18, 15, 22, 23]. However, twitchin is a minor protein of the catch muscle [15] and its content may not be sufficient for the development of catch state. Moreover, twitchin is widely spread in different muscles that are not related to the catch contraction [5], while myorod is expressed exclusively in molluscan catch muscle [4, 5].

Also, N-terminal synthetic peptides as well as full-length myorod interact with F-actin and thin-filaments in phosphorylation- and Ca²⁺-dependent modes, respectively [13]. Myorod is phosphorylated by smMLCK at its N-terminus at the Thr141 position [16] and to a lesser extent by twitchin [17] which contains a MLCK-domain [14]. It was also shown that myorod as well as myosin are phosphorylated by Ca^{2 +}-dependent myosin-associated kinase at their C-terminal parts [17, 24] in the non-helical C-terminal tailpiece [25], most likely at the Ser1693 residue [8] which corresponds to Ser973 in myorod [24]. Myosin-associated kinase is active only in the absence of calcium ions [17, 24, 25] and therefore might be important for catch development [25].

The phosphorylation of myorod at the two different sites of the molecule indicates the possibility of regulation of its unknown function. For example, twitchin is seems to be a force-sensitive kinase, which is thought to be activated by stretch that results in its autophosphorylation [26]. The idea that twitchin is a mechanical link interconnecting thin and thick filaments [12] though its interaction with actin [18] and myosin [12; 27] or myorod and paramyosin [12] supports this possibility. Stretching of muscle might trigger a chain of events responsible for N-terminal phosphorylation of myorod and potentiating Mg²⁺-ATPase of myosin during the contraction. The mechanism how phosphorylation activates myorod might include changing of the conformation of the N-domain or rupture of links between N-terminal domains of myorod similar to the dissociation of the two heads of myosin during phosphorylation [28].

In this study, we found a functional link between phosphorylation of the N-terminus of myorod and the level of actin-activated ATPase activity of myosin. This is an example of signal transmission along myorod-myosin complex affecting mechano-chemical features of myosin. However, the function of myorod is unlikely explicable solely based on properties of the myorod—myosin complex. Rather, the myorod—myosin complex is an integral part of the thick filament structures consisting of paramyosin, myosin, myorod and twitchin. Nevertheless, the results of this study clearly indicate that myorod is not simply a structural component of thick filaments but may act as an active player of catch muscle contraction.