Are the class 18 myosins Myo18A and Myo18B specialist sarcomeric proteins?

Abstract

Initially, the two members of class 18 myosins, Myo18A and Myo18B, appeared to exhibit highly divergent functions, complicating the assignment of class-specific functions. However, the identification of a striated muscle-specific isoform of Myo18A, Myo18Aγ, suggests that class 18 myosins may have evolved to complement the functions of conventional class 2 myosins in sarcomeres. Indeed, both genes, Myo18a and Myo18b, are predominantly expressed in the heart and somites, precursors of skeletal muscle, of developing mouse embryos. Genetic deletion of either gene in mice is embryonic lethal and is associated with the disorganization of cardiac sarcomeres. Moreover, Myo18Aγ and Myo18B localize to sarcomeric A-bands, albeit the motor (head) domains of these unconventional myosins have been both deduced and biochemically demonstrated to exhibit negligible ATPase activity, a hallmark of motor proteins. Instead, Myo18Aγ and Myo18B presumably coassemble with thick filaments and provide structural integrity and/or internal resistance through interactions with F-actin and/or other proteins. In addition, Myo18Aγ and Myo18B may play distinct roles in the assembly of myofibrils, which may arise from actin stress fibers containing the α-isoform of Myo18A, Myo18Aα. The β-isoform of Myo18A, Myo18Aβ, is similar to Myo18Aα, except that it lacks the N-terminal extension, and may serve as a negative regulator through heterodimerization with either Myo18Aα or Myo18Aγ. In this review, we contend that Myo18Aγ and Myo18B are essential for myofibril structure and function in striated muscle cells, while α- and β-isoforms of Myo18A play diverse roles in nonmuscle cells.

Keywords: MYO18A; MYO18B; knockout (KO) mice; sarcomere; stress fibers; unconventional myosins.

FIGURE 1

Domain structures and sequence alignments of Myo18A and Myo18B. (A), Domain structures of mouse (Mus musculus) Myo18Aα, Myo18Aβ, Myo18Aγ, and Myo18B. Sequence identity (Id) and similarity (Sim) are indicated for various regions. Abbreviations: KE-rich, lysine (K) and glutamic acid (E)-rich; PDZ, post-synaptic density (PSD) protein, Drosophila discs-large (Dlg) tumor suppressor protein, and zonula occludens (ZO) protein; P-loop, phosphate-binding loop; SW1, switch-1; SW2, switch-2; IQ, IQ (isoleucine (I) and glutamine (Q)) calmodulin-binding motif; aa, amino acids; P-rich, proline (P)-rich; S-rich, serine (S)-rich; NLS, nuclear localization signal. (B), Alignment of the extreme N-terminal ends of Myo18Aγ and Myo18B (marked by red boxes in panel (A), highlighting high identity and similarity. In the aligned sequences, vertical lines (“|”) indicate positions of identity (amino acids are identical), colons (“:”) indicate positions of similarity (amino acids have similar side chain properties), single dots (“.”) indicate mismatch of amino acids, and dashes (“–”) indicate gaps in a sequence relative to its counterpart. The gray filled boxes show alignment of Mus musculus (Ms) Myo18A and Ms Myo18B P-loop, switch-1, and switch-2 sequences with those of Ms Myo6, Ms Myh9 (nonmuscle myosin IIA), Ms Myo5a, and Ms Myo1e. Conserved sequence motifs are highlighted in bold black, while deviations in Myo18A and Myo18B are shown in red. Inserts specific for Myo18A and Myo18B are shown in orange.

FIGURE 2

Protein structure prediction of (Mus musculus) Myo18Aγ. (A), AlphaFold v2.3.2 was used to perform structure predictions via a Jupyter Notebook hosted on GitHub. Due to computational power constraints, predictions were restricted to amino acid residues 1 to 1400, encompassing the N-terminal extension, as well as the head domain. Visual representation of the structure was generated using ChimeraX. For clarity, only the head domain is represented (residues 281–1200). The ribbon representation of the head domain is shaded in gray, except for features forming and surrounding the nucleotide-binding pocket: the P-loop (G473–T480) is colored green, SW-1 (S520–Q533) is blue, and SW-2 (D758–Q763) is cyan. In addition, a small loop extension following SW-2 (G769–A774) is highlighted pink and the moderately large Myo18A-loop (E1024–L1052) is yellow. (B), Ribbon representation of the head domain color-coded for the confidence score (pLDDT, predicted local distance difference test) of the model generated by AlphaFold. (C), Close-up view of the nucleotide-binding pocket, albeit including the synthetic substrate ADP-AlF₄ (transparent sand colored surface), incorporated by overlaying the structure 1MND from the PDB. (D), Side chains of residues which deviate from censensus sequences are shown as sticks (see also Figure 1B). (E), Rotated view (90° rotation relative to panels (C, D) highlighting the proximity of the moderately long Myo18A-loop (yellow) to the nucleotide-binding site. (F), View optimized to show the side chains of residues within SW-1 and SW-2 which deviate (colored red) from the consensus sequences.