Drosophila myosin function and muscle structure are modulated by isoform-specific amino acid residues of the relay domain

Abstract

The relay domain of the myosin molecular motor communicates with the nucleotide binding site of the molecule, the actin-binding region, and the converter domain. Interactions between the relay and converter domains are critical for the converter to elicit movement of the lever arm, which yields the molecule’s power stroke. An invariant tryptophan relay residue at the converter interface is flanked by two poorly conserved amino acids. Here we utilize mutant versions of the Drosophila melanogaster muscle myosin heavy chain to probe the function of these variable relay residues, residues 509 and 511. We replaced either or both poorly-conserved residues within an embryonic myosin isoform with those of the indirect flight muscle isoform. All such replacements eliminate in vitro actin motility, suggesting abnormal communication between the relay and converter domains that disables the power stroke. Interestingly, we found that the variations at residues 509 and 511 differentially disrupt both myosin ATPase activity and indirect flight muscle ultrastructure, with the double mutant showing the most severe effect upon ATPase activity and the least severe effect on emergent adult ultrastructure. Isoform-specific interactions may be implicated in these differential effects, as our molecular modeling at two stages of the mechanochemical cycle defined differences in contacts between the embryonic, flight muscle and chimeric 509 and 511 relay residues with a key residue in the converter domain. Overall, our studies show that two poorly conserved amino acid residues at the relay-converter interface are important to isoform-specific function of myosin and to the structural integrity of muscle.

1. Introduction

The myosin motor protein of muscle thick filaments interacts with actin-containing thin filaments in an ATP-dependent fashion to yield muscle contraction. This process is dependent upon a series of internal rearrangements of myosin domains that ultimately result in production of a power stroke, wherein the myosin lever arm rotates while the myosin motor domain is attached to actin [1]. The cocking of the lever arm prior to its rotation is dependent upon signals arising from the nucleotide binding pocket as to the state of ATP hydrolysis. These signals are transmitted through the relay domain and the SH1 helix to the converter domain, which is adjacent to the lever arm [1]. At the relay-converter interface, an invariant tryptophan relay residue has been shown to change its orientation during this process [2]. Interestingly, the residues that flank the invariant tryptophan are not well conserved among myosins of different organisms and can vary within muscle myosin isoforms of an individual organism (Fig. 1A).

Fig. 1.

Comparison of myosin relay domain sequences of various phyla; locations of alternative domains of Drosophila myosin. A. Amino acid sequences of relay domains of striated muscle myosins, muscle-like myosins or non-muscle myosins of various phyla compared to the Drosophila melanogaster EMB isoform (yellow, non-conserved residue; grey, well conserved residue; invariant tryptophan 510 is shown in bold, with variable residues 509 and 511 flanking it). The indirect flight muscle isoform (IFI) differs from EMB at five positions. Phyla: Arthropoda (D. melanogaster: fruit fly, C. vittatus: scorpion); Nematoda (C. elegans); Annelida (Chaetopterus sp. THS-2011: parchment tube worm); Mollusca (M. yessoensis: scallop, A. californica: sea hare); Platyhelminthes (S. mediterranea: planaria); Echinodermata (S. purpuratus: sea urchin); Chordata (G. gallus: chicken, H. sapiens: human); Cnidaria (H. vulgaris: fresh-water polyp); Sponges (D. avara: marine sponge); D. melanogaster zipper: non-muscle myosin II; Amoebozoa (D. discoideum: slime mold). NCBI accession identifiers are listed in the right-most column. B. Location of alternatively-encoded regions mapped onto the crystal structure of Drosophila EMB in the post-rigor configuration (PDB:5W1A; [3]). Exon 3 encodes alternative N-terminal regions (magenta); exon 7 encodes alternative nucleotide binding pocket lip regions (cyan); exon 9 encodes alternative relay domains (blue); exon 11 encodes alternative converter domains (green).

The role of the poorly conserved relay residues is of particular interest in Drosophila melanogaster, where three versions of the relay domain are encoded by alternative versions of exon 9 [4]. Each isoform has different residues flanking the invariant tryptophan, despite the fact that the 57 amino acids encoded by the three alternative exons differ at only 6 other residues. In addition to alternative isoform variation in the relay domain of Drosophila myosin, the interacting converter domain displays isoform-specific variation as a result of alternative RNA splicing of exon 11. Alternative exon-encoded regions of Drosophila myosin are depicted on the embryonic isoform (EMB) crystal structure in Fig. 1B.

We previously examined the roles of alternative relay domains in Drosophila by substituting the relay from the indirect flight muscle isoform (IFI) into EMB and vice-versa [5]. Expressing the chimeric myosins in Drosophila indirect flight muscles showed that substitution of the EMB relay into the IFI molecule permitted normal myosin and muscle function. In contrast, EMB myosin that contains the IFI relay domain was unable to transport actin in vitro, had reduced ATPase activity, and yielded more defects in indirect flight muscle myofibril assembly and stability compared to muscles expressing EMB.

Here we examine how the poorly conserved myosin residues flanking the invariant tryptophan disable EMB function, in order to shed light on their isoform-specific properties. We inserted either or both of the IFI relay residues into the EMB construct and tested effects upon in vitro actin motility, ATPase activity, and myofibril assembly and stability. In addition, we modeled the interaction of these residues with a key converter domain residue in endogenous and chimeric molecules. We find that either or both IFI residues disable in vitro motility, suggesting abnormal relay-converter interactions that are borne out by molecular modeling. Further, actin-activated ATPase activity is reduced most severely when both amino acids are replaced. Interestingly, however, the double substitution yields the best myofibril assembly and stability in young adults. Overall, despite the poor conservation of these relay residues, they clearly play an important role in tuning Drosophila myosins, suggesting that other organisms utilize this approach for producing myosins with differing functional properties.

2. Materials and methods

2.1. DNA constructs, mutagenesis and germline transformation

Transgenes EMB-D509N, EMB-A511D and EMB-D509N-A511D were constructed using the backbone of the EMB Drosophila myosin heavy chain (Mhc) plasmid, which includes the Mhc transcriptional promoter [6]. A 4.7-kb Xba I/Apa I fragment containing the 5′ end of EMB was subcloned into the pBluescriptKS vector (Agilent, Santa Clara, CA, USA) to yield Emb5’. A 5.1-kb Apa I/Kpn I fragment containing the 3′ end of EMB was likewise subcloned to yield Emb3’. A 1.27-kb Nru I-Apa I fragment from Emb5′, which encodes a portion of exon 8 through exon 12, was cloned into Apa I and Eco RV sites of pBluescriptKS to yield Emb9b. Site-directed mutagenesis (QuickChange II kit, Agilent) was performed to yield EMB-D509N, EMB-A511D and EMB-D509N-A511D. For EMB-D509 N, the mutagenesis primer was 5′-GGCAT-TAACTGGGCCTTCATC-3’; for EMB-A511D, the primer was 5′-GGCATTGACTGGGACTTCATC-3’; for EMB-D509-A511D the primer was 5′-GGCATTAACTGGGACTTCATC-3’ (nucleotide coding changes in bold). Mutagenesis was confirmed by sequencing. Each subclone was digested with Bgl II and Apa I to produce a 979-bp fragment that was individually cloned back into Emb5’. Each clone was then digested with Xba I and Apa I to produce a 4.7-kb fragment that was ligated with the 5.1-kb fragment of Apa I/Kpn I-digested Emb3′ into pCaSpeR 4 [7], which had been digested with Xba I and Kpn I. All sites used in the cloning process were sequenced to ensure accuracy. Note that to be consistent with our previous publications, the chicken skeletal muscle myosin numbering system is used for all residue numbers [8].

P element-mediated transformation [9] was used by BestGene, Inc. (Chino Hills, CA, USA) to generate transgenic lines expressing the w⁺ eye color marker. Each transgenic insert was mapped to its chromosomal location using balancer chromosomes and standard genetic crosses. Three independent lines for each transgenic construct were crossed into the Mhc¹⁰ background, which is null for myosin in indirect flight and jump muscles [10].

2.2. Verification of transgenic line identity, myosin expression level and flight ability

RNA from transgenics in the Mhc¹⁰ background was reverse transcribed and sequenced to confirm expression of transgenes in the absence of endogenous Mhc transcripts. To this end, RNA from upper thoraces of two-day-old female adult flies was extracted using LiCl₂ [11]. cDNA was synthesized using a New England Biolabs Protoscript cDNA synthesis kit (Ipswich, MA, USA), with 0.5 μg RNA and 3 μmol of reverse primer specific to Mhc exon 12 (5′- GAGCTTCTTGAAGCCCTTACGG -3′). PCR was performed with 1 μl of cDNA along with 3 μmol of the aforementioned reverse primer and an exon 8 forward primer (5′-CGATACCGCCGAGCTGTACAG-3′). Initial incubation for 60 s at 94 °C was followed by 30 cycles of: 30 s at 94 °C, 30 s at 55 °C and 2 min at 68 °C. PCR-generated fragments were subcloned using the CloneJet PCR Cloning Kit (Thermo Scientific, Waltham, MA, USA) and sequenced. At least ten independent cDNA subclones were sequenced for each transgene, all showing the expected mutation.

To ensure that myosin heavy chain expression was at normal levels for each transgene in the Mhc¹⁰ background, SDS polyacrylamide gel electrophoresis was employed [12]. Six dissected upper thoraces from two-day-old female flies were homogenized in 60 μl SDS gel buffer. Eight μl of sample were loaded on a 9% polyacrylamide gel. This was repeated with five freshly prepared samples. After electrophoresis, protein accumulation was determined using Coomassie blue stained gels that were digitally scanned and analyzed on NIH image software. All lines showed levels of myosin heavy chain relative to actin that were at or above 92% of wild-type ratios.

Flight ability assessment was performed on at least 100 two-day-old adult female flies from each transgenic line at 22 °C. Flight was determined by the ability to fly up (U), horizontal (H), down (D) or not at all (N), as previously described [13]. All transgenics showed no flight ability.

2.3. Myosin isolation

Myosin was isolated from dissected dorso-longitudinal indirect flight muscles as previously described [5,14]. Briefly, fibers dissected from 100 to 200 young flies were collected in York Modified Glycerol (YMG: 20 mM potassium phosphate, pH 7.0, 2 mM MgCl₂,1 mM EGTA, 10 mM DTT, and a protease inhibitor mixture). Material was pelleted and resuspended in YMG with 2% Triton X-100, stored on ice for 30 min, centrifuged, rinsed in YMG and re-centrifuged. The pellet was resuspended in 1 M KCl, 5 mM MgCl₂, 0.5 mM EGTA, 10 mM sodium pyrophosphate, 10 mM DTT, 50 mM potassium phosphate, pH 6.8, and protease inhibitor for 15 min on ice. Following centrifugation, the supernatant was diluted to an ionic strength of 40 mM and stored overnight. After ultracentrifugation, the pellet was dissolved in 2.4 M KCl, 100 mM histidine, pH 6.8, 0.5 mM EGTA, 10 mM DTT, incubated on ice and then slowly diluted to 0.3 M KCl. Following centrifugation, the supernatant was diluted 10-fold, placed on ice for 1 h and then centrifuged. The pellet was dissolved in myosin storage buffer (0.5 M KCl, 20 mM MOPS, pH 7.0, 2 mM MgCl₂, 10 mM DTT) and concentration was determined by absorbance at 280 nm.

2.4. In vitro motility

In vitro actin motility was assessed using nitrocellulose-coated coverslips in a flow cell to which Drosophila myosin was affixed [5,14]. Unlabeled phalloidin-stabilized F-actin filaments were flowed into the cell to react with anomalous myosin molecules that irreversibly bound to actin in the presence of ATP. Following a wash, actin filaments containing TRIT-C-labeled phalloidin were added along with ATP to view actin filament motility on active myosin via fluorescence microscopy.

2.5. ATPase activity

ATPase activities of myosin were measured as previously described in detail [5,14]. Briefly, CaATPase activity was determined by adding 2 μg of myosin to CaATPase buffer (10 mM imidazole pH 6.0, 0.1 M KCl, 10 mM CaCl₂). ATPase activity was initiated at 25 °C by addition of [γ-³²P] ATP. After 15 min, the reaction was quenched with 1.8 M HClO₄. Extraction and scintillation counting were carried out as described previously [14]. Basal MgATPase was determined using MgATPase buffer (10 mM imidazole pH 6.0, 20 mM KCl, 2 mM MgCl₂, 0.1 mM CaCl₂). Actin-stimulated MgATPase was assessed by the addition of 0.1 to 2 μM chicken breast muscle F-actin. After 25 min, the reaction was quenched. Basal Mg-ATPase activities were subtracted from actin-activated data points. Actin-activated ATPase activity (V_max) and the Michaelis constant (K_m) were obtained by fitting all data points from preparations of each myosin with the Michaelis-Menten equation using SigmaPlot software. Statistical comparisons of ATPase data were carried out using one-way ANOVA and Tukey’s post hoc test.

2.6. Indirect flight muscle ultrastructure

Transmission electron microscopy was used to determine the effects of transgene expression on indirect flight muscle structure in the myosin-null (Mhc¹⁰) background at four different stages of development: late-stage pupae, 2-h-old adults, 2-d-old adults and 1-week-old adults. Samples were prepared by fixation of dissected tissue in 3% paraformaldehyde, 2% glutaraldehyde, 100 mM sucrose, 2 mM EGTA, 100 mM sodium phosphate buffer, pH 7.2. After overnight treatment, samples were post-fixed in 1.0% osmium, 100 mM sodium phosphate, pH 7.2, dehydrated, and then embedded in Spurr’s resin. Details of sample preparation were published previously [15]. Cross-sections and longitudinal sections of indirect flight muscles were obtained from females for each transgenic line, with at least three different samples examined for each line at each stage of development.

2.7. Protein structure analysis

Scallop muscle myosin II crystal structures in the pre-power stroke state (PDB: 1QVI) [16] and post-power stroke (actin-detached) state (PDB: 1KK8) [17] were used as templates for the S-1 amino acid sequences of the Drosophila embryonic (EMB) and indirect flight muscle (IFI) myosin isoforms. The ESyPred3D homology modeling server [18] was employed to produce the homology coordinates and the output was viewed using PyMOL (http://www.pymol.org/).

3. Results

3.1. Effects of relay domain mutations on in vitro motility and ATPase activity

We produced transgenic Drosophila lines expressing embryonic myosin (EMB) containing substituted residues surrounding invariant tryptophan 510 of the relay domain, mutating either or both flanking residues to match those of the indirect flight muscle myosin isoform (IFI). The mutant proteins are designated EMB-D509N, EMB-A511D and EMB-D509N-A511D. Previously published work utilizing wild-type EMB myosin [6] or EMB myosin containing the entire IFI relay domain (designated EMB-9a) [5] yielded flies lacking flight ability. In the current work, absence of flight was observed for EMB-D509N, EMB-A511D and EMB-D509N-A511D adults, despite normal levels of myosin accumulation (see Materials and methods).

Isolated myosin containing EMB-D509N, EMB-A511D or EMB-D509N-A511D mutations failed to promote movement of fluorescent actin filaments in the in vitro motility assay (Table 1), suggesting that relay-converter interaction is critical for driving actin motility. In these semi-automated assays, we were unable to identify gliding movement of any actin filaments in flow cells containing mutant myosin despite employing multiple biological replicates for each mutant. Previously, such lack of motility was found with EMB-9a myosin, where the entire IFI relay domain encoded by exon 9a was substituted into EMB [5].

Table 1.

Actin-sliding velocities and Ca⁺², Mg⁺², actin-stimulated Mg⁺² ATPase kinetics of EMB-D509N, EMB-A511D and EMB-D509N-A511D myosins.

Myosin	Actin sliding velocity ± SD (μms−1)	Basal Ca-ATPase ± SD (s−1)	Basal Mg-ATPase ± SD (s−1)	Actin-stimulated Vmax ± SD (s−1)	Km (actin) ± SD (μM)
EMB-D509N (n = 4)	0	5.83 ± 0.87**	0.10 ± 0.01*,†	0.62 ± 0.07††	0.20 ± 0.08
EMB-A511D (n = 5)	0	5.94 ± 0.82**	0.13 ± 0.02***	0.96 ± 0.15***	0.10 ± 0.08
EMB-D509N-A511D (n = 4)	0	3.76 ± 0.47	0.06 ± 0.01	0.42 ± 0.06	0.06 ± 0.03
EMB§	0.7 ± 0.2 (n = 3)	4.47 ± 2.20 (n = 6)	0.08 ± 0.04 (n = 7)	1.10 ± 0.38 (n = 7)	0.29 ± 0.18 (n = 7)

Statistically significant difference relative to EMB-D509N-A511D:

^*= p < 0.05;

^**= p < 0.01;

^***= p < 0.001.

Statistically significant difference of EMB-D509 N relative to EMB-A511D:

^†= p < 0.05;

^††= p < 0.01.

Values calculated by one-way ANOVA with Tukey’s multiple comparison test (GraphPad Prism version 10.4.2). Primary data are found in Appendix A. Supplementary data (ATPase data and statistics).

^§Data for EMB isoform [5] provided for comparison purposes.

We next performed ATPase activity assays to assess effects of relay domain substitutions on enzymatic activity. The double mutation was more effective at reducing ATPase activities than either single mutation, as Ca-ATPase, Mg-ATPase and actin-stimulated Mg-ATPase activities for EMB-D509N-A511D were less than those of EMB-D509N or EMB-A511D (Table 1). The double substitution reduced these activities below those previously seen for intact EMB ([5,19]; Table 1), yielding values akin to EMB-9a [5]. The double mutation also enhanced actin affinity relative to Mg-ATPase activity (K_m), compared to each single mutation, although these differences were not statistically significant (Table 1). Further, the EMB-D509N mutation produced statistically significant reductions in both basal Mg-ATPase and actin-stimulated Mg-ATPase activities compared to EMB-A511D (Table 1), suggesting residue 509 is more critical to myosin function than residue 511.

3.2. Effects of relay domain mutations on indirect flight myofibril assembly and stability

Transmission electron microscopy was employed to study the effects of the relay domain mutations on indirect flight muscle ultrastructure. We examined cross-sections and longitudinal sections of late-stage pupae, 2-h-old adults, 2-d-old adults and 1-week-old adults. While EMB-D509N late-stage pupae show assembly defects, with abnormal myofibril shapes and gaps in sarcomeres (Fig. 2A,D), both EMB-A511D (Fig. 2B,E) and EMB-D509N-A511D pupae (Fig. 2C,F) display normal hexagonal packing of thick and thin filaments and intact sarcomeres. Normal structures are largely retained in the double mutants in 2-h-old adults (Fig. 2I,L), but more severe degeneration occurs by this age in both single mutants (Fig. 2G,H,J,K). By 2 d, all of the mutants show disarray (Fig. 2M–R) and this becomes more severe by 1 week for EMB-A511D (Fig. 2T,W) and EMB-D509N-A511D (Fig. 2U,X). EMB-D509N one-week-old adults (Fig. 2S,V) appear to retain their structure compared to 2-d-old adults. Interestingly, despite having the poorest ATPase activity, the double mutant assembles myofibrils normally (late-stage pupae) and largely retains their structure into very young adulthood (d 2).

Fig. 2.

Effects of relay domain mutations on myofibril assembly and stability in the dorsal longitudinal indirect flight muscle. A. Transverse section from EMB-D509N late-stage pupa. Myofibrils show disruption in assembly of thick and thin filaments, with gaps in the myofibril structure. B. Transverse sections from EMB-A511D and C. EMB-D509N-A511D late-stage pupae. Myofibrils show normal morphology with double-hexagonal packing of thick and thin filaments. D. Longitudinal section from EMB-D509N late-stage pupa. Sarcomeres contain gaps and frayed edges. E. Longitudinal section from EMB-A511D and F. EMB-D509N-A511D late-stage pupae. Sarcomeres appear normal with regular M – and Z-lines. G. Transverse section from EMB-D509N 2-h-old adult shows continued disruption in myofibril morphology and packing of thick and thin filaments. H. Transverse section from EMB-A511D 2-h-old adult displays disrupted myofibril morphology. I. Transverse section from EMB-D509N-A511D 2-h-old adult shows normal filament packing. J. Longitudinal sections from EMB-D509N and K. EMB-A511D 2-h-old adults show sarcomere gaps with disrupted M – and Z-lines. L. Longitudinal section from EMB-D509N-A511D 2-h-old adult shows more subtle disruption of M – and Z-lines, with fewer gaps than single mutants. M. Transverse sections from EMB-D509N and N. EMB-A511D 2-d-old adults show similar levels of disruption as 2-h-old adults. O. Transverse section from EMB-D509N-A511D 2-d-old adult shows somewhat more disruption of myofibril morphology than 2-h-old adults. P. Longitudinal sections from EMB-D509N and Q. EMB-A511D 2-d-old adults have enhanced degradation of M – and Z-lines compared to 2-h-old adults. R. Longitudinal section from EMB-D509N-A511D 2-d-old adult. Sarcomeres show disruption of filament packing, with retention of Z-lines. S. Transverse section from EMB-D509N 1-week-old adult retains structure similar to 2-d-old adult. T. Transverse sections from EMB-A511D and U. EMB-D509N-A511D 1-week-old adults show continued myofibril morphology degradation, with scattered thick and thin filaments throughout the myofibril. V. Longitudinal section from EMB-D509N 1-week-old adult shows gaps in sarcomeres with poorly defined M – and Z-lines. W. Longitudinal section from EMB-A511D 1-week-old adult displays sarcomeres that appear over contracted. X. Longitudinal section from EMB-D509N-A511D 1-week-old adult shows continued gapping and fraying, with some retention of M – and Z-lines. The myofibril shown in each panel is representative of the population at that given stage of development. Scale bars, 0.5 μm. M, M-line. Z, Z-line. m, mitochondrion.

3.3. Modeling the effects of relay domain mutations on molecular interactions

The crystal structures of scallop muscle myosin were used as templates to model Drosophila myosin in the pre- and post-power stroke states (Fig. 3). Structures from this organism serve as useful models since they originate from invertebrate striated muscle and are solved in multiple stages of the power stroke [16,17]. We assessed relay interaction with residue R759 in the converter domain, as we previously found that this residue is critical to myosin interaction with the relay loop [20, 21].

Fig. 3.

Location of potentially-interacting residues in the relay and converter domains modeled using the crystal structures of scallop myosin S-1 in the pre- (PDB 1QVI; [15]) and post (actin-detached)- (PDB 1KK8; [16]) power stroke states. A. Indirect flight muscle myosin isoform (IFI) in pre- (left) and post-power stroke (right) states. Relay domain polar residue N509 (red) potentially interacts with the converter domain basic residue R759 (green) best in the post-power stroke state. The relay domain acidic residue D511 (grey) appears to interact with the converter domain basic residue R759 in the post-power stroke state, but not in the pre-power stroke state. B. EMB in the pre- (left) and post-power stroke (right) states. Relay domain acidic residue D509 (red) interacts best with the converter domain basic residue R759 (green) in the post-power stroke state. The relay domain hydrophobic residue A511 (grey) appears unlikely to interact with the converter domain basic residue R759 in either state. C. IFI relay domain mutations inserted into the EMB isoform in the pre- (left) and post-power stroke (right) states. Relay domain polar residue D509 N (red) interacts best with converter domain basic residue R759 (green) in the post-power stroke (right) state. Relay domain acidic residue A511D (grey) may weakly interact with the converter domain basic residue R759 in the post-power stroke state. The chicken skeletal muscle myosin numbering system was used for all residue numbers [7].

In the pre-power stroke state of the indirect flight muscle isoform (IFI), relay domain polar residue N509 and negatively-charged D511 are fairly distant from positively-charged R759 of the converter domain (Fig. 3A, left). This configuration is significantly changed during the mechanochemical cycle, as both relay residues interact with the converter residue in the post-power stroke state (Fig. 3A, right).

For EMB, relay domain residue D509 distantly interacts with converter residue R759 in the pre-power stroke state (Fig. 3B, left), but forms a strong interaction in the post-power stroke configuration (Fig. 3B, right). The negatively charged EMB D509 residue permits stronger interaction than for polar N509 of IFI, forming a salt bridge. This stabilization may account for the reduced fluctuation of the EMB relay and converter domains compared to IFI, which had been observed by molecular dynamics studies [3]. In EMB, relay domain hydrophobic residue A511, with its nonpolar methyl side chain, is not expected to interact with the positively-charged R759 relay residue (Fig. 3B).

The two IFI relay domain substitutions into the EMB molecule (EMB-D509N-A511D) are modeled in Fig. 3C. The differences in interactions of the relay domain with the converter domain are influenced by both the relay domain mutations and by their interface with the EMB converter domain, which is encoded by a different version of exon 11 than the IFI converter [22]. The D509 N mutation in the pre-power stroke state replaces a charged relay-converter interaction in EMB (Fig. 3B, left) with a weaker polar interaction, yielding a more distant interface with R759 (Fig. 3C, left). However, the interaction of this residue in the post-power stroke state (Fig. 3C, right) is more proximal than for the salt bridge in EMB (Fig. 3B, right) or the polar interaction in IFI (Fig. 3A, right). The potential charged-based interaction with A511D remains distant in the pre-power stroke state (Fig. 3C, left) as in IFI (Fig. 3A, left) and is weaker in the post power-stroke state (Fig. 3A–C, right panels). Overall, it is clear that the interaction between the relay and converter domains of the EMB-D509N-A511D molecule match neither those of EMB or IFI as a result of insertion of the IFI residues at positions 509 and 511. This abnormal relay-converter configuration could affect the communication pathway between the relay and converter domains, influencing converter-lever arm movement.

4. Discussion

Drosophila melanogaster muscle myosin heavy chain serves as an important tool for probing protein domains that are key to myosin isoform-specific function, as alternative splicing of RNA transcripts highlight areas that must drive functional differences. The alternative domains identified in the globular head (Fig. 1B) point to the interacting relay (blue) and converter (green) as two of four motor domain regions encoded by alternative exons that yield isoform-specific functionality. Interestingly, despite the presence of three alternative exons encoding the Drosophila relay, the encoded domains are highly conserved, as is the case throughout phylogeny (Fig. 1A). Our current study indicates that two amino acid residue variations, those adjacent to the invariant tryptophan, are important for both biochemical (Table 1) and ultrastructural (Fig. 2) functions of myosin isoforms. Molecular modeling suggests that isoform-specific versions of these amino acid residues affect the relative orientation and interaction of the relay and converter domains (Fig. 3).

Our previous work [5] showed that, compared to intact EMB, substitution of the entire IFI relay domain into the EMB myosin backbone resulted in lack of in vitro motility, reduced ATPase activity, as well as enhanced degeneration of myofibrillar structure compared to EMB. The communication pathway between the relay domain and the nucleotide binding pocket [23] is presumably affected by the domain exchange, which leads to increased ADP affinity and hence a reduction in ATPase activity [24]. Here we found that the effects of only substituting one, the other, or both IFI amino acid residues flanking the invariant tryptophan yielded similar results to substituting the entire alternative region containing five differences (residues 494, 498, 505, 509, 511: Fig. 1A). This suggests that the tryptophan-flanking residues are key to maintaining the mechanochemical cycle and myofibril assembly/stability. Interestingly, while the substitution of both flanking amino acids most drastically reduced ATPase activity (Table 1), the double substitution yielded the best myofibrillar structure in newly emergent adults (Fig. 2I). This result is akin to our analysis of some other Mhc mutations, where under-functional myosin does not severely affect sarcomere structure [25,26].

In contrast to expectations based on absence of in vitro motility and reduced ATPase activity [5], mechanics assays on fibers expressing EMB containing the entire IFI relay domain yielded increased stiffness, higher maximum power output and increased fiber kinetics compared to EMB [27]. This suggests that the IFI relay is capable of producing IFI-like mechanical properties under the oscillatory load induced by a mechanical rig during sinusoidal analysis. Such load is not present in in vitro motility or ATPase assays. Thus, it will be interesting to determine whether either or both of the IFI residues flanking the invariant tryptophan are key to defining mechanical properties in fiber experiments.

We previously studied the effects of mutating IFI relay residues 509 and 511 in the context of an IFI myosin that contains the mutation R759E in its converter domain [21]. This converter mutation in the relay domain interaction region substantially reduces in vitro motility, IFI ATPase activity, flight ability and results in myofibrillar degeneration [20]. To examine the interaction of IFI relay residues 509 and 511 with IFI R759E, we individually mutated residues 509 or 511 to lysine in an attempt to establish a salt bridge between the converter and relay. Interestingly, the N509K mutation suppressed all of the R759E mutant phenotypes, whereas D511K did not, suggesting that residue 509 plays a major role in the inter-domain interaction. This conclusion is borne out for EMB myosin by the current study, where mutating EMB residue 509 had significantly greater effects on both basal and actin-activated Mg-ATPase activity than a residue 511 mutation (Table 1). Each of our molecular models shows a closer proximity for residue 509 to converter residue 759 (Fig. 3), further supporting this functional interaction.

A recent study examined the role of five alternative relay domains of Locusta migratoria muscle myosins expressed in vitro using the Sf9 baculovirus system [28]. The authors found that the alternative relay domains affected actin-activated Mg-ATPase activity, with residues 474 and 509 serving isoform-specific functions. As residue 474 does not differ among the Drosophila muscle myosin isoforms, the locust study corroborates our finding that residue 509 is a key element in myosin isoform function.

5. Conclusion

Our integrative analysis demonstrates that relay domain residues 509 and 511 of Drosophila EMB myosin are key elements in the myosin mechanochemical cycle, as mutation of either or both residues eliminates in vitro actin filament motility and negatively impacts myosin ATPase activity and myofibril ultrastructure. Molecular modeling and previous genetic suppression analysis indicate that interaction with the converter domain is disrupted by these mutations. The 509 and 511 relay domain residues are among the few that show isoform-specific or phylogenetic variation. This observation, in conjunction with the aforementioned biochemical and ultrastructural results, indicate their importance in fine-tuning myosin isoform function.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.bbrc.2026.153439.

Availability of data and materials

All data are included in the main text or in the supplementary appendix. Constructed plasmids and transgenic Drosophila stocks are available from the corresponding author upon request.