Cardiomyopathy-associated and basic residue mutations in myopalladin alter actin binding, bundling, and structural stability

Abstract

Myopalladin (MYPN) is a striated muscle-specific protein essential for sarcomere integrity and actin filament organization. The Ig3 domain of MYPN mediates direct binding and bundling of filamentous actin (F-actin), and several cardiomyopathy (CM)-associated mutations cluster within this domain. To elucidate the molecular basis of MYPN-actin interactions, we performed alanine-scanning mutagenesis of conserved basic residues and evaluated CM-linked variants of the Ig3 domain using co-sedimentation assays. These mutations impaired actin binding and abolished bundling activity, despite preserving secondary structure in most cases. Circular dichroism revealed that the P961L variant adopts a more open, partially unfolded β-sandwich conformation, distinct from the wild type (WT), which correlates with its aggregation and mislocalization in Drosophila cardiomyocytes. In vivo expression of full-length WT and mutant MYPN in Drosophila muscle confirmed Z-disc localization for WT, while R955W and P961L mutants formed aberrant clusters. Biochemical assays demonstrated that MYPN Ig3 promotes actin polymerization and bundling even under non-polymerizing conditions, and its interaction with actin is electrostatically driven but calciumindependent. Sedimentation equilibrium analysis confirmed that MYPN Ig3 functions as a monomer, suggesting that bundling arises from dual actin-binding surfaces rather than from self-oligomerization. These findings establish the Ig3 domain as a structurally robust, actin-regulatory module and identify disruption of MYPN–actin interactions as a pathogenic mechanism in CM. Our study provides the first direct evidence linking Ig3 domain mutations to impaired actin dynamics and sarcomere disorganization, offering mechanistic insight into MYPN-associated cardiomyopathies.

1 |. INTRODUCTION

Cardiomyopathies (CM) are a heterogeneous group of myocardial disorders that impair cardiac function and represent a major global public health concern (Elliott et al., 2008; Maron et al., 2006; Towbin & Bowles, 2002). These disorders can be classified as hypertrophic (HCM), dilated (DCM), restrictive (RCM), arrhythmogenic, or left ventricular noncompaction (LVNC) CM and can be further classified as genetic or acquired forms.

Mutations in the MYPN gene, encoding the sarcomeric protein myopalladin (MYPN, NM_032578.3), are relatively common among genetic CM (Purevjav et al., 2012). MYPN is a 145 kDa protein expressed specifically in striated muscle, where it acts as a scaffold linking regulatory proteins in the I-band with structural proteins in the Z-disc assembly while also shuttling to the nucleus to regulate gene expression. A member of the palladin/MYPN/myotilin family, MYPN shares 68% sequence identity with palladin (PALLD) (Bang et al., 2001; Filomena et al., 2020; Filomena et al., 2021). Like PALLD, MYPN contains five immunoglobulin (Ig) domains and a proline-rich region; however, MYPN and PALLD produce opposing effects on actin dynamics. PALLD promotes polymerization, while MYPN inhibits it, though MYPN more strongly stabilizes filamentous (F)-actin against depolymerization (Filomena et al., 2020).

Among the 66 known MYPN missense variants associated with CM, six (F954L, R955Q, R955W, P961L, C1002W, and R1042C) are localized within the actin-binding Ig3 domain (Noureddine & Gehmlich, 2023). Some variants, like P961L, disrupt sarcomere integrity, while others such as R955W preserved structure (Meyer et al., 2013). Despite the clinical significance of MYPN mutations in CM, the role of MYPN in cardiac structure and function, as well as the molecular pathophysiology underlying MYPN-associated phenotypic heterogeneity remains poorly understood (Noureddine & Gehmlich, 2023; Purevjav et al., 2012). Understanding how MYPN maintains Z-disc integrity in muscle tissue may provide insight into therapeutic strategies for MYPN-linked CM.

In this study, we investigate the actin-binding specificity of the MYPN Ig3 domain using charge-neutralizing mutagenesis and actin co-sedimentation assays. Our data reveal that F-actin binding is mediated by conserved basic charge clusters within the Ig3 domain. Given that several cardiomyopathy-associated MYPN mutations localize to this region, we hypothesized that these pathogenic variants disrupt actin interactions and consequently impair sarcomere integrity, contributing to disease pathogenesis. While the effects of these mutations on MYPN-actin interactions had not been previously examined, our findings show that pathogenic variants impair F-actin binding. Furthermore, the DCM-associated P961L mutation prevented stable Ig fold adoption, while other mutations maintained the overall structure but altered the domain’s stability. To assess the functional consequences in vivo, we expressed full-length wild-type (WT) and two DCM mutants, R955W and P961L, in Drosophila cardiomyocytes and body wall muscles, revealing distinct subcellular mislocalization patterns relative to Z-disc markers.

Additionally, we found that electrostatic interactions contribute to MYPN Ig3-F-actin binding and that the Ig3 domain functions as an actin polymerizing and bundling protein, even under non-polymerizing conditions. Together, these data establish a mechanistic link between MYPN mutations, defective actin binding, and potential sarcomere disorganization in CM.

2 |. RESULTS

2.1 |. F-actin binding and bundling are impaired by neutral substitutions at conserved basic residues

Previous studies identified two surface-exposed basic patches in the PALLD Ig3 domain (K₁₃-LKHYK₁₈ and K₅₁), along with key lysine residues K36 and K46 (Beck et al., 2013; Sargent et al., 2025). Lysine 38, positioned adjacent to K51 in the tertiary structure, also contributes to the second basic patch and has been shown to regulate actin polymerization and bundling (Yadav et al., 2016). Sequence alignment of PALLD Ig3 with MYPN Ig3 reveals equivalent basic residues in MYPN (Figure 1a). In the AlphaFold3-predicted structure of MYPN Ig3, residues K949, R950, K952, and R955 cluster into a basic patch on the surface, while K987 and R988 form a second patch on the opposite face (Figure 1b) (Abramson et al., 2024). This suggests that basic surfaces may represent a shared, evolutionarily conserved actin-binding mechanism. Therefore, we hypothesized that MYPN Ig3 also binds F-actin directly via these basic residues. To test this, we generated alanine substitution mutations at conserved basic positions in the isolated Ig3 domain of MYPN. Co-sedimentation assays with F-actin revealed that alanine substitutions at K949, R950, K952, R955, K987, and R988 resulted in a slight increase in the apparent dissociation constant (K_d) (Figures 2a and S1A–G); however, these differences were not statistically significant (p > 0.05). Conversely, the K987A (p < 0.05), R988A (p < 0.05), and K952A (p < 0.01) mutants showed significantly reduced B_max values compared to WT (Figure 2a), indicating a lower overall binding capacity. This reduction suggests that charge-neutralizing mutations in the Ig3 domain disrupt key residues required for actin interaction, thereby decreasing the number of effective binding sites per molecule. Although the apparent affinity (K_d) remains largely unchanged, the diminished number of functional sites would limit the ability of MYPN Ig3 to engage actin filaments effectively.

FIGURE 1.

Sequence alignment of MYPN and PALLD Ig3 domains reveals conserved basic patches. (a) Alignment shows conserved basic residues (blue). MYPN K952, R955, and R988 correspond to PALLD K15 (952), K18 (955), and K51 (988), respectively. Additionally, PALLD includes other critical residues involved in F-actin interaction (K36 (973), K38 (975), and K46 (983)). In MYPN, blue denotes charge-neutralizing mutations; orange denotes a cardiomyopathic mutation (R955 is shared by both mutation types). MYPN Ig3 numbering is based on the full-length human protein, whereas PALLD numbering refers to the isolated mouse Ig3 domain. (b) AlphaFold3 predicted structure of the MYPN Ig3. Two basic patches are indicated: Patch 1 (K949, R950, K952, and R955) and Patch 2 (K987 and R988). Basic side chains are shown as blue sticks.

FIGURE 2.

Alanine mutations in conserved basic residues of MYPN Ig3 domain reduce F-actin binding and impair cross-linking. (a) Co-sedimentation binding curves where a constant Ig3 concentration (10 μM) was titrated with varying F-actin (0–30 μM). The data were fitted with a hyperbolic curve, assuming specific binding only, to obtain apparent dissociation constants (K_d) and maximum binding B_max which are shown to the right of the graph. Values represent the mean ± SD from three independent experiments (n = 3). Statistical analysis by Welch’s and Brown-Forsythe ANOVA revealed no significant differences between the K_d values obtained for any of the Ig3 variants and WT (p > 0.05). However, significant differences in B_max were found for the K987A (*), R988A (*), and K952A (**) variants compared to WT. Asterisks denote statistical significance (*p < 0.05, **p < 0.01). (b,c) Low-speed co-sedimentation bundling assay where constant F-actin concentration (10 μM) was incubated with Ig3 at 1:1 (b) and 1:2 (c) molar ratios. The percentage of pelleted F-actin after low-speed centrifugation represents bundles (black bars). The pellet from high-speed centrifugation represents polymerized F-actin (gray bars). The supernatant contains G-actin (white bars). Data are presented as mean ± SD (n = 3). Asterisks indicate statistical significance from Welch’s and Brown-Forsythe ANOVA comparing WT to mutants: (b) 1:1 ratio and (c) 1:2 ratio (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001).

Although no single mutation completely abolished F-actin binding, every mutant exhibited a significant loss of crosslinking or bundling activity (p < 0.05). This demonstrates that while the mutants retain affinity for a single filament, they are specifically impaired in the ability to bridge two filaments (Figures 2b,c and Figure S1n–t), a trend consistent with observations for analogous PALLD Ig3 mutations (Beck et al., 2013). This striking loss of bundling underscores that actin crosslinking requires multiple simultaneous filament interactions mediated by MYPN. When B_max is reduced, as observed for most mutants, fewer filaments can be engaged at once, severely compromising bundle formation even when some binding persists.

Furthermore, for mutants with K_d and B_max values comparable to WT, the observed bundling defects may indicate an alternative mechanism such as actin-induced dimerization of the Ig3 domain, similar to that reported for PALLD (Vattepu et al., 2015). These mutants may disrupt an actin-induced dimerization interface required for bundling, while still preserving actin-binding affinity. This explains why bundling defects are more pronounced than binding defects and underscores the critical role of two conserved actin-binding surfaces in the MYPN Ig3 domain. Together, these findings position bundling as a highly sensitive readout of MYPN Ig3 integrity and reinforce its critical role in cytoskeletal organization.

2.2 |. Cardiomyopathic mutations impair actin binding and bundling by MYPN Ig3

The essential contribution of basic residues to F-actin binding and cross-linking, combined with the presence of pathogenic mutations in the Ig3 domain of MYPN (Figure 3), suggested that these variants might impair MYPN-F-actin interactions and thereby compromise sarcomere integrity in cardiac muscle (Noureddine & Gehmlich, 2023). To directly test this hypothesis, we generated site-directed mutants of disease-associated variants, expressed and purified the recombinant proteins, and assessed their actin-binding properties.

FIGURE 3.

Domain architecture of MYPN and location of CM variants. The domain organization of MYPN is shown along with related proteins (200 kDa isoform 1 of PALLD and myotilin), highlighting Ig domains and polyproline-rich (PR) regions. The locations of a subset of CM variants from the 66 identified CMs for MYPN are indicated. Specific mutations within the Ig3 domain and their associated diseases are detailed in the boxes above MYPN.

There were no significant differences among the K_d values of the CM-associated variants compared to WT (p > 0.05) (Figure 4a and Figure S1h–m). Similarly, B_max values were not significant different (p > 0.05) from WT for most CM variants, except R955Q, which showed a significant reduction (p < 0.05). The K_d and B_max values for P961L could not be determined (ND) because nonlinear least-squares fitting of the P961L data failed to converge due to a high background signal in the pellet even in the absence of F-actin. This suggests that the observed sedimentation was not due to specific actin binding but rather to non-specific interactions, such as protein precipitation or aggregation during the assay. Increasing F-actin concentrations produced only a minimal increase in pelleted P961L, and its markedly reduced F-actin bundling activity further confirmed impaired, non-specific interactions (Figure 4b,c). Furthermore, circular dichroism (CD) analysis described later indicated that P961L is partially unfolded, consistent with the difficulty in purifying soluble protein and suggesting that misfolding underlies its aggregation propensity.

FIGURE 4.

Cardiomyopathy-associated MYPN Ig3 domain mutants show reduced F-actin binding and dramatic decrease in bundling. (a) Binding curves generated from co-sedimentation assays where constant Ig3 concentration (10 μM) was titrated with varying F-actin (0–30 μM). The data were fitted with a hyperbolic curve, assuming specific binding only, to obtain apparent dissociation constant (K_d) and maximum binding B_max which are shown to the right of the graph. Values represent the mean ± SD from three independent experiments (n = 3). Statistical comparisons using Welch’s and Brown-Forsythe ANOVA indicated no significant differences among the K_d values of the variants compared to the WT (p > 0.05). The B_max values for all variants were not significantly different (p > 0.05) from the WT, except for R955Q, which showed a significant difference (*p < 0.05). The K_d and B_max value of P961L was not determined (ND) because nonlinear least-squares fitting could not converge as the mutant protein exhibited a high background signal in the pellet even in the absence of actin. (b,c) Low-speed co-sedimentation bundling assay with constant F-actin concentration (10 μM) mixed with Ig3 at 1:1 (b) and 1:2 (c) molar ratios. The percentage of pelleted F-actin after low-speed centrifugation represents bundles (black bars). The pellet from high-speed centrifugation represents polymerized F-actin (gray bars). The supernatant contains G-actin (white bars). Data are presented as mean ± SD (n = 3). Asterisks indicate statistical significance from Welch’s and Brown-Forsythe ANOVA comparing WT to mutants: (b) 1:1 ratio and (c) 1:2 ratio (**p < 0.01 and ***p < 0.001).

Consistent with the charge-neutralization mutants described in the previous section, all CM-associated variants exhibited a pronounced loss of actin bundling activity despite largely retaining binding affinity (Figure 4b,c and Figure S1u–z). This parallel trend reinforces that bundling is more sensitive to structural perturbations than binding alone and highlights the critical role of multiple actin-binding surfaces in maintaining filament crosslinking. Together, these findings underscore that the MYPN Ig3 domain is a functionally critical actin-binding module, and that both charge-neutralizing and disease-associated mutations compromise its ability to regulate filament organization. In particular, the DCM variants F954L, R955Q, and R955W cluster near the basic patch 1 (Figure 1), reinforcing the importance of this region as a key actin-interaction interface. Disruption of F-actin engagement at or adjacent to this site likely contributes to Z-line disorganization and sarcomere instability observed in cardiomyopathic hearts.

2.3 |. MYPN Ig3 promotes actin polymerization and crosslinking during filament assembly

Previous studies showed that PALLD Ig3 can crosslink both pre-formed actin filaments and filaments generated during polymerization, with stronger bundling observed when Ig3 is present during filament assembly (Gurung et al., 2016). To determine whether MYPN Ig3 exhibits similar behavior, we compared its bundling activity under three conditions: (i) with G-actin (globular actin) in non-polymerizing buffer (G-buffer), (ii) with G-actin in polymerizing conditions (F-buffer), and (iii) with pre-polymerized filaments (i.e., regular bundling in F-buffer). These distinct conditions allowed us to distinguish Ig3’s role in nucleating polymerization from its ability to cross-link pre-formed filaments. Figure 5 shows that MYPN Ig3 promotes bundling in all three conditions, and the extent of bundling increases with Ig3 concentration (See also Figure S1aa–ac). Notably, actin polymerized in the presence of MYPN (i.e., co-polymerization) produced significantly more bundled filaments than when MYPN was added after polymerization at 10 μM Ig3, indicating that Ig3 enhances filament assembly when present during polymerization. This difference diminished at higher Ig3 concentrations (20 μM), where nearly 100% of actin was bundled in both co-polymerization and postpolymerization conditions, suggesting a saturation effect. These results demonstrate that MYPN Ig3 not only bundles pre-formed filaments but also actively promotes filament assembly, highlighting its dual role in actin network organization.

FIGURE 5.

MYPN Ig3 domain promotes actin polymerization and bundling under multiple conditions. The MYPN Ig3 domain (5, 10 or 20 μM) was incubated with G-actin under either non-polymerizing G-buffer (white bars) or polymerizing, F-buffer (gray bars) conditions. For comparison, MYPN Ig3 was also added to pre-polymerized F-actin (10 μM) in F-buffer conditions (black bars). The percentage of actin recovered in the pellet after low-speed centrifugation represents bundled filaments. Data are presented as the mean ± SD (n = 3). Statistical analysis using Welch’s and Brown-Forsythe ANOVA compared bundling efficiency of WT Ig3 at three concentrations (5, 10, and 20 μM) across conditions. Controls without WT Ig3 are shown at 0 μM. Significant differences are indicated by asterisks: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, and ns denotes not significant.

2.4 |. Electrostatic interactions mediate MYPN Ig3-F-actin binding

Most actin-binding proteins (ABP) depend, to some degree, on electrostatic interactions to engage the negatively charged surface of F-actin. Many of these proteins contain clusters of basic residues within their actin-binding regions. Increasing the ionic strength, by raising the salt concentration, often reduces the affinity for actin due to the disruption of these electrostatic interactions (Amann et al., 1998; Dixon et al., 2008; Heier et al., 2017; Hüttelmaier et al., 1999; Lee et al., 2004; Li et al., 1998; Tang, Szymanski, et al., 1997). To determine whether MYPN Ig3-F-actin binding is influenced by ionic strength, we performed co-sedimentation assays across a range of KCl concentrations (25–200 mM).

As shown in Figure 6a,f-actin binding was maximal at 100 mM KCl, the standard concentration used in prior assays, and significantly reduced at 50 and 200 mM (p < 0.05 and <0.001, respectively) (see also Figure S1ad–af). Binding at 25 mM KCl was comparable to that at 100 mM (not significantly different at p > 0.05), indicating that both low and physiological ionic strengths support interaction, whereas high salt disrupts it. These results suggest that electrostatic interactions contribute to MYPN Ig3 F-actin binding and that elevated ionic strength partially weakens this interaction, similar to PALLD Ig3 (Dixon et al., 2008). Notably, MYPN Ig3 exhibited maximal binding at physiological ionic strength (100 mM KCl), unlike PALLD Ig3, which binds most strongly at low salt. This difference implies that MYPN may have evolved distinct electrostatic tuning to maintain stable interactions in the cardiac cytoskeletal environment.

FIGURE 6.

MYPN Ig3-F-Actin interaction is electrostatically driven but not modulated by Ca²⁺. (a) Co-sedimentation of 20 μM Ig3 with 5 μMF-actin at increasing KCl concentrations (25, 50, 100, and 200 mM). The percentage of Ig3 bound to F-actin (pellet fraction) is shown in black, while the unbound Ig3 (supernatant fraction) is shown in white. Data are presented as the mean ± SD (n = 3). Statistical analysis using Welch’s and Brown-Forsythe ANOVA compared the binding efficiency of WT Ig3 at 100 mM physiological KCl to that at other KCl concentrations. Significant differences are indicated by asterisks: *p < 0.05, ***p < 0.001, and ns denotes not significant. (b) Binding curves of Ig3 (10 μM) titrated with F-actin (0–30 μM) were obtained in the presence of Ca²⁺ or EGTA. Apparent K_d and B_max values are shown for conditions of +Ca²⁺ (pCa = 4), −a²⁺ (pCa = 10), and regular binding. Data are presented as the mean ± SD (n = 3). No significant differences in K_d or B_max between conditions, as assessed by Welch’s and Brown-Forsythe ANOVA (p > 0.05).

2.5 |. MYPN Ig3 binding to actin is independent of calcium

Upon muscle cell activation, sarcoplasmic Ca²⁺ concentrations rise from a resting concentration of ~100 nM to ~1 μM (Bootman, 2012). Previous studies have shown that the N2A domain of titin, a sarcomeric ABP, exhibits Ca²⁺-dependent binding to F-actin (Dutta et al., 2018). To determine whether the MYPN Ig3 domain is similarly regulated by Ca²⁺, we performed co-sedimentation assays under Ca²⁺-rich (pCa = 4) and Ca²⁺-free (EGTA, pCa = 10) conditions. As shown in Figure 6b, neither the apparent K_d nor the B_max values of MYPN Ig3 were significantly affected by Ca²⁺ (see also Figure S1ag–ah). The K_d was 2.86 ± 1.15 μM in the presence of Ca²⁺ and 1.61 ± 0.60 μM in the absence of Ca²⁺, both comparable to the baseline K_d under standard assay conditions (1.84 ± 0.63; p > 0.05). Similarly, B_max values showed no significant differences across conditions (p > 0.05), indicating that MYPN Ig3–actin binding is Ca²⁺-independent. This stable interaction may help preserve Z-line integrity during rapid Ca²⁺ fluctuations that occur with muscle contraction. Such stability could also protect binding interfaces for other regulatory proteins, such as titin, that respond dynamically to calcium signaling.

2.6 |. Most cardiomyopathy-linked and charge-neutralizing mutations preserve the Ig fold of MYPN

To determine whether any of the point mutations affect the secondary structure of the MYPN Ig3 domain, we performed far-UV CD spectroscopy. The CD spectrum of WT Ig3 exhibited a maximum near 205 nm and a minimum around 215 nm, consistent with the signature of antiparallel β-sheets, which typically exhibit a peak at 195 nm and a trough at 218 nm (Figure S2) (Greenfield, 2006). These results confirm the β-sheet-rich architecture of the Ig3 domain in MYPN. Notably, the spectral features of MYPN Ig3 are slightly shifted from canonical β-sheet signatures, likely due to the β-sandwich topology of this domain, in which two opposing β-sheets are stacked together (Bork et al., 1994; Otey et al., 2009). This geometry, along with strand twisting, can influence CD signal positions (Whitmore & Wallace, 2008).

CD analysis revealed that nearly all MYPN Ig3 domain mutants maintain a β-sheet-rich architecture similar to the WT, as assessed by DichroWeb secondary structure analysis (Table S1 and Figure S2) (Miles et al., 2022; Whitmore & Wallace, 2004). The CD spectra of most variants closely overlapped with WT (Figure S2a,b), indicating preservation of the Ig fold. Minor deviations were observed for R950A and C1002W, but these did not reflect major structural disruption. Quantitative estimates using DichroWeb’s K2D method confirmed that WT and all mutants with the exception of P961L contain 10%–12% α-helix, 39%–47% β-sheet, and 44%–49% random coil, consistent with the β-sandwich topology of immunoglobulin domains. Collectively, these results indicate that, aside from the P961L case, neither cardiomyopathy-associated nor charge-neutralizing mutations significantly alter the secondary structure or hydrophobic core packing of the Ig3 domain.

The P961L mutant exhibited a distinct CD profile (Figure S2b); yet, the minimum near 215 nm indicates residual β-sheet structure. Secondary structure analysis revealed reduced β-sheet (~36%), increased random coil (~51%), and a slight rise in α-helix (~13%) content compared to the WT (Table S1). This combination suggests a perturbed β-sheet signature and partial unfolding. Consequently, while P961L retains some β-sheet elements, its higher random coil content likely allows a more open β-sandwich conformation, promoting aggregation.

2.7 |. Mutations modulate thermal stability without disrupting folding

Although CD spectroscopy confirmed that most disease-associated and charge-neutralizing mutations do not alter the secondary structure of MYPN Ig3, subtle differences in domain stability could still affect function or increase susceptibility to misfolding under physiological stress. To assess these effects, we measured the thermal stability of WT and mutant Ig3 domains using CD thermal denaturation. WT MYPN Ig3 unfolded cooperatively with a thermal denaturation midpoint (T_M) of 48.50 ± 0.85°C (Figure 7). Variants K987A and F954L exhibited modest but significant (***p < 0.001 and **p < 0.01, respectively) and reproducible increases in T_M (50.48 ± 0.20 and 50.11 ± 0.21°C, respectively), consistent with enhanced stability through modified intramolecular interactions. In contrast, C1002W and R950A displayed significantly reduced thermal stability (****p < 0.0001; T_M = 45.10 ± 0.33°C and 42.38 ± 0.34°C) and less cooperative unfolding (significant only for C1002W, ****p < 0.0001) transitions, indicative of local destabilization or increased conformational flexibility. Interestingly, two different mutations at the same position (R955W and R955Q) had opposing effects on unfolding cooperativity. The variant R955W exhibited an intermediate state during thermal denaturation and significantly reduced cooperativity (****p < 0.0001). Conversely, the R955Q mutation showed significantly greater cooperativity than the WT (p < 0.05). Despite these differences in unfolding mechanism, the T_M values for both variants (47.85 ± 0.18°C for R955Q and 48.05 ± 0.45°C for R955W) did not differ significantly (p > 0.05) from that of the WT.

FIGURE 7.

Mutations alter MYPN Ig3 thermal stability. Thermal denaturation of WT and mutant Ig3 was monitored by CD at 205 nm. Data were globally fit to a two-state unfolding model to determine the T_M and Hill coefficient. Melting temperature values and Hill coefficients shown in each graph represent the mean from three or more independent experiments with the associated error reflecting the standard deviation of the nonlinear curve-fitting results. Values for mutants marked with a dragger (†) were obtained from a single measurement. (a) Unfolding curves for alanine mutations and (b) CM mutations, compared with WT. Statistical analysis by Welch’s and Brown-Forsythe ANOVA revealed significant differences in T_m value for the mutants R950A (****), K987A (***), F954L (**), and C1002W (****). Hill coefficient was significantly different compared with WT for R955Q (*), R955W, and C1002W (****). Asterisks denote statistical significance (*p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001). No significant differences between the T_m values and Hill coefficient obtained for other Ig3 variants and WT (p > 0.05).

The melting curve for P961L lacked a typical sigmoidal transition between folded and unfolded states (Figure S3). This aligns with our previous observation that P961L may already exist in a partially unfolded state with a more open β-sandwich conformation. Therefore, P961L likely fails to undergo a cooperative unfolding transition upon heating because of its hydrophobic patches are already more solvent-exposed, promoting aggregation rather than unfolding.

No significant changes in thermal stability or cooperativity were observed for the remaining mutants (K949A, K952A, R955A, R988A, and R1042C) (Figure 7a,b). Overall, these results indicate that the MYPN Ig3 domain is structurally robust, tolerating many substitutions without major disruption, although P961L is a clear exception. Similar effects have been observed in other Ig-like domains, such as titin I21, where single-residue substitutions (e.g., C3575S) decreased stability without altering the overall fold (Martinez-Martin et al., 2023). By integrating structural and thermodynamic analyses, we demonstrate that the MYPN Ig3 domain tolerates many mutations without major structural disruption, yet select variants fine-tune its stability in ways that may contribute to cardiomyopathy pathogenesis.

2.8 |. MYPN Ig3 domain exists as a monomer in solution

To determine the oligomeric state of the MYPN Ig3 domain in solution, we performed sedimentation equilibrium analytical ultracentrifugation (SE-AUC), a quantitative method for measuring native molecular mass in solution. Nonlinear least-squares fitting of the concentration gradients to a self-association model revealed no evidence of higher-order species (Figure S4). Across multiple protein concentrations and rotor speeds, the measured molecular mass was 11.7 kDa, closely matching the predicted monomeric mass of 12 kDa and showed no concentration-dependent increase in molecular weight. These results indicate that MYPN Ig3 domain exists as a stable monomer in solution under our experimental conditions. This is consistent with previous findings for the homologous PALLD Ig3 domain, which also functions as a monomer despite its actin-bundling capability (Beck et al., 2008; Dixon et al., 2008; Mykkanen et al., 2001). This suggests that multivalent F-actin interactions are achieved through the structural topology or surface chemistry of individual Ig3 domains, rather than via domain organization.

In contrast, other Ig-like domains, particularly in structural proteins such as filamin and titin, can mediate dimerization or form elongated multimers. For example, certain filamin Ig domains participate in homodimeric or heterodimeric associations that are essential for actin cross-linking and mechanical resistance (Nakamura et al., 2007). Similarly, titin’s Ig and FnIII domains can form supramolecular assemblies that contribute to sarcomeric elasticity (Tskhovrebova & Trinick, 2003). The fact that MYPN Ig3 and its PALLD homolog retain actin-bundling activity as monomers highlights a distinct molecular strategy involving the use of highly specialized binding interfaces rather than multimerization to achieve filament crosslinking. This may reflect unique functional constraints at the Z-line where tight spatial packing of sarcomeric proteins requires precise, monomeric interactions.

2.9 |. MYPN function in drosophila muscle tissues

To investigate the impact of MYPN mutations in vivo, we generated humanized Drosophila models to analyze full-length MYPN protein localization in larval heart and muscle tissues. The Drosophila genome does not encode for MYPN, so we expressed wild-type and mutant versions of the 145 kDa full-length protein in transgenic fly lines utilizing the GAL4-UAS binary expression system (Brand & Perrimon, 1993). The Mef2 promoter drives expression of the transactivator GAL4 protein in heart and skeletal muscle throughout the Drosophila life cycle (Ranganayakulu et al., 1998). GAL4 binds to upstream activating sequences (UAS) to induce expression of MYPN proteins.

Immunostaining of the larvae heart tube revealed localization of human MYPN to the Z-disc of cardiac tissue using our GAL4-mediated expression system (Mef2-GAL4 > UAS-MYPN WT; Figure 8a). This result nicely validates the Drosophila musculature as an evolutionary conserved model to study mammalian muscle structure and function as human MYPN is also located at the Z-disc (Bang et al., 2001). The Drosophila larval body wall muscles are larger than the heart tube and the organization of sarcomeres in series within each muscle cell is ideal for studying sarcomeric protein localization. As expected due to the lack of MYPN in Drosophila, there was no MYPN signal in w¹¹¹⁸ control muscles (Figure 8b). Expression of human full-length WT MYPN is located at the Z-disc (Figure 8b, open white triangle) in larval muscles. To verify our immunostaining results, we also generated transgenic flies containing GFP-tagged versions of MYPN and found that MYPN-GFP showed the same Z-disc localization as the untagged protein (Figure S5b). Importantly, the overall pattern of larval muscles was not altered upon expression of MYPN or MYPN mutant proteins appended with GFP (Figure S5a).

FIGURE 8.

MYPN is located at the Z-disc and Ig domain mutations cause aberrant MYPN accumulation. (a) Immunostaining of Drosophila larval heart tube expressing Mef2 > MYPN WT. MYPN protein is localized to the Z-disc (magenta, open white triangle). F-actin is shown in gray. Scale bar: 10 Mm. (b) L3 larval muscles stained with phalloidin to visualize F-actin or with an antibody against MYPN. F-actin and MYPN colocalize at the Z-disc (open white triangle), in Mef2 > MYPN WT muscles. MYPN R955W and MYPN P961L Ig domain mutations show abnormal accumulation of MYPN (white asterisk). Scale bar: 10 Mm. (c) Western blot showing the expression levels of MYPN (Yamamoto et al., 2013) in the indicated genotypes. Total protein staining serves as a loading control. One-way ANOVA. n = 3. (d) Quantification of MYPN expression levels.

In MYPN R955W and P961L mutants, we discovered that MYPN was largely present at its expected location (Figure 8b, open white triangle) but also accumulated abnormally in clusters away from the Z-disc or appeared as widened Z-discs (Figure 8b, asterisks). Similar phenotypes were observed for GFP-tagged MYPN mutants, where R955W-GFP signal extended into the thin filament (Figure S5b, open white arrow) and P961L-GFP resulted in larger aggregates (Figure S5B, asterisks). Overexpression of all MYPN-GFP proteins resulted in substantial accumulation near the nuclei (Figure S5c). Western blotting was used to confirm expression levels of all MYPN proteins and confirmed that the differences in subcellular localization were not due to differences in expression levels (Figure 8c,d and Figure S5d,e).

3 |. DISCUSSION

The identification of 66 MYPN missense variants associated with DCM, HCM, RCM, and LVNC CM (Gu et al., 2017; Noureddine & Gehmlich, 2023), combined with observed myofibrillar abnormalities in patients (Meyer et al., 2013) and MYPN’s interactions with key sarcomeric structural components, collectively underscores its role in maintaining sarcomere integrity and normal cardiac function (Filomena et al., 2020; Filomena et al., 2021). Since the underlying disease mechanisms remain poorly understood and the functional role of MYPN in the heart has remained elusive, the development of new therapies for CM has been hampered. Previous studies have primarily focused on MYPN’s role in signaling pathways and nuclear shuttling disruption (Huby et al., 2014; Purevjav et al., 2012). These investigations examined specific MYPN mutations (Y20C and Q529X) located in the region that interacts with cardiac ankyrin repeat protein (CARP/Ankrd1), a stress-inducible transcriptional co-factor that suppresses genes involved in heart failure and hypertrophy development (Huby et al., 2014; Purevjav et al., 2012). Our study focuses on disease-associated mutations within the MYPN Ig3 domain, the minimal actin-binding region responsible for thin filament interaction and assembly. To better understand MYPN’s role in cardiac function, we characterized how the Ig3 domain modulates actin dynamics and determined the functional alterations caused by CM mutations.

Ig domains are evolutionarily conserved protein–protein interaction modules present in several cytoskeleton-associated proteins (Otey et al., 2009), including MYPN, myomesin (Auerbach et al., 1999), titin (Linke, 2000), MyBP-C (Okagaki et al., 1993), MyBP-H (Okagaki et al., 1993), filamin C (Gonzalez-Morales et al., 2017), obscurin (Benian & Mayans, 2015), PALLD (Beck et al., 2013), and myotilin (Kostan et al., 2021). Although these domains share a similar overall structure, they engage with a wide range of binding partners, from actin and myosin to nuclear components, prompting the question of what structural features underlie their specificity. Several known actin-binding motifs, such as calponin homology domain, Wiskott-Aldrich syndrome protein homology domain-2 motifs, gelsolin homology, actindepolymerizing factor homology, and the lysine-rich actin-binding domain 3 (ABD3) motif rely on clusters of positively charged residues for actin interaction (Beck et al., 2013; Gimona et al., 2002; Paunola et al., 2002; Poukkula et al., 2011; Way et al., 1991). PALLD, MYPN, and myotilin constitute a closely related subfamily of Ig-domain containing proteins, and in both PALLD and myotilin, F-actin binding is mediated by conserved basic clusters. Specifically, myotilin relies on residues K354, K358, and K359 (Kostan et al., 2021); while PALLD contains lysine-rich sequences (K13/950, K18/955, K36/973, K46/983, and K51/988), where substitution of these residues abolishes F-actin binding (Beck et al., 2013; Sargent et al., 2025).

To elucidate the molecular basis of MYPN’s regulation of actin dynamics, we examined the role of positively charged residues in the Ig3 domain’s actin-binding region. Charge-neutralizing mutational analysis revealed that two surface-exposed basic patches on opposing faces of the Ig3 domain are essential for actin binding. Mutations in either basic patch did not significantly reduce F-actin binding affinity and no single mutation completely abolished the interaction, suggesting that multiple or additional residues also contribute to binding. Thus, MYPN, like its related family members, relies on clusters of surface-exposed basic residues to mediate F-actin binding (Beck et al., 2013; Kostan et al., 2021). Although some mutants retain partial affinity for actin, their reduced B_max limits the extent of filament engagement, which is critical for crosslinking and bundle formation. Actin bundling requires multiple simultaneous interactions to stabilize filament networks; therefore, a lower binding capacity translates into markedly impaired bundling activity, as observed in our assays.

In cells, actin filaments assemble into diverse higher-order architectures, and actin cross-linking proteins can regulate this structural complexity (Karp, 2009). Crosslinking generally requires either two actin-binding sites within a single protein (e.g., PALLD, fascin, fimbrin, and espin) or dimerization of monomeric proteins that contain only one binding site (e.g., myotilin) (Karp, 2009; Salmikangas et al., 2003). Our mutational data show that neutralizing either basic patch significantly impaired actin cross-linking, mirroring effects observed with analogous mutations in PALLD’s Ig3 domain (Beck et al., 2013). Our data highlight the indispensable role of these residues in actin bundle formation. Sedimentation equilibrium analysis confirmed that MYPN Ig3 is monomeric in solution, indicating that bundling likely arises from dual actin-binding surfaces rather than from self-oligomerization. We therefore propose that the two basic clusters on opposite faces of the Ig3 domain permit a single MYPN molecule to simultaneously engage and bundle two actin filaments. However, bundling may involve a more complex mechanism, such as actin-induced dimerization of the Ig3 domain which was observed in PALLD Ig3 (Vattepu et al., 2015). Mutations in the basic patch could disrupt the dimerization interface necessary for bundling without affecting actin binding.

Our in vitro data on CM causing mutations demonstrate that all variants impair F-actin binding and have a significant effect on actin bundling. This aligns with data from the muscle-specific α-actinin-2 ABD, in which HCM-associated mutations (A119T and G111V) also reduce F-actin affinity while preserving secondary structure (Haywood et al., 2016). Among the DCM-causing mutants, P961L has the most dramatic effect, abrogating specific actin interactions. This trend of impaired actin-binding extends to other DCM-associated mutants (F954L, R955Q, and R955W), an HCM-associated mutant (C1002W), and an LVNC-associated mutant (R1042C), suggesting a common pathological mechanism for CM.

Our copolymerization assays revealed that MYPN Ig3 not only bundles pre-formed filaments but also promotes actin polymerization from G-actin under non-polymerizing conditions (Figure 5), suggesting that Ig3 can nucleate or stabilize early actin oligomers, facilitating filament elongation. This activity likely arises from the presence of two basic actin-binding surfaces on opposite faces of the Ig3 domain, enabling simultaneous engagement of multiple actin monomers and reducing the kinetic barrier for filament assembly. At first glance, these results appear to contrast with prior work reporting that MYPN Ig3 inhibits actin polymerization in a dose-dependent manner using pyrene-actin fluorescence assays under polymerizing conditions (Filomena et al., 2020). We propose that these observations reflect context-dependent roles of MYPN Ig3. Under conditions that normally prevent polymerization (G-buffer) Ig3 may act as a nucleator by bridging actin monomers, whereas under polymerizing conditions, high concentrations of Ig3 could inhibit elongation by sequestering actin or capping filament ends. Thus, MYPN Ig3 likely exhibits dual functionality, promoting filament assembly at low to moderate concentrations or under unfavorable conditions, while limiting elongation and favoring crosslinking at high concentrations.

This nuanced behavior suggests that MYPN Ig3 fine-tunes actin dynamics rather than serving as a simple activator or inhibitor, consistent with its role in maintaining sarcomere structure. In vivo, this polymerizationpromoting function may be critical for sarcomere maintenance and remodeling. At the Z-disc, where actin filaments anchor and undergo dynamic turnover, MYPN could help reinforce filament networks during mechanical stress or repair processes. By accelerating filament assembly and bundling, Ig3 may stabilize thin filament arrays, ensuring proper alignment and tension transmission across the sarcomere. This constitutive polymerization activity, combined with bundling, positions MYPN as a structural organizer rather than a passive scaffold, highlighting its potential role in preserving Z-disc integrity during cycles of contraction and relaxation.

The ability of MYPN Ig3 to promote actin polymerization and bundling may also have broader implications for transcriptional regulation. By reducing the cytoplasmic G-actin pool, Ig3 could facilitate MRTF-A nuclear translocation and activate SRF-dependent gene expression, linking MYPN’s structural role at the Z-disc to signaling pathways that govern muscle growth and stress adaptation (Filomena et al., 2020). Unlike PALLD, which also stimulated actin polymerization but primarily functions in cytoskeletal remodeling, MYPN’s polymerization activity may serve a dual role by reinforcing sarcomere architecture while modulating gene expression programs important for cardiac stress responses.

Expression of full-length human WT MYPN in Drosophila larval muscles demonstrated proper Z-line localization, despite the absence of endogenous MYPN in the fly genome. Immunostaining and GFP-tagged constructs revealed that the R955W and P961L mutant proteins accumulated into clusters, incorporating improperly into Z-discs. Notably, GFP-P961L exhibited prominent protein aggregates. Anti-MYPN staining further demonstrated widened Z-discs and, in the case of GFP-R955W, partial diffusion of MYPN into thin filaments. Consistent with these findings, the biopsies from DCM patients, the R955W showed normal staining patterns, however P961L exhibited profound pathological effects disrupting MYPN’s interactions with α-actinin and leading to abnormal protein distribution and altered cytoskeletal organization with severely compromised sarcomere structure (Meyer et al., 2013). These findings demonstrate how CM mutations disrupt MYPN’s interactions with both actin and other sarcomeric proteins, leading to Z-disc structural defects representing a likely mechanism underlying their pathological effects.

CD analysis revealed that all mutants, except P961L, retained secondary structure content and displayed wavelength scans similar to WT, preserving the Ig-fold β-sandwich. However, melting temperature measurements indicated that some mutations altered thermal stability without disrupting the overall fold. In contrast, the CD spectrum of the P961L mutant reflected a partially unfolded conformation, indicating a deviation from the characteristic Ig-fold. This loss of native architecture demonstrates that P961 is important for maintaining the structural integrity of the domain. Notably, despite the severe misfolding and loss of bundling activity observed for the P961L Ig3 domain in vitro, the full-length P961L mutant in Drosophila still localized correctly to the Z-disc but did not form larger aggregates, suggesting that other domains of MYPN may compensate for impaired Ig3 function in vivo.

The electrostatic model of F-actin binding has been proposed for many ABPs, as these typically contain clusters of positively charged residues on their solvent-exposed surfaces, while F-actin itself possesses strongly anionic subdomains (Angelini et al., 2006; Beck et al., 2013). Furthermore, the polyelectrolyte nature of F-actin is known to be stabilized by long-range electrostatic interactions (Tang et al., 1997a; Tang & Janmey, 1996). To determine whether MYPN–F-actin binding is electrostatically driven, we examined the salt sensitivity of the MYPN Ig3–F-actin interaction. We found that MYPN-actin interactions are strongest at physiological salt concentrations (100 mM KCl) and lower salt (25 mM KCl) but are significantly reduced at higher salt concentrations (200 mM KCl). This salt-dependent reduction is consistent with electrostatic contributions. Interestingly, MYPN Ig3 behaves differently from closely related proteins PALLD, myotilin, and coronin 1B. Its binding at 100 mM KCl remains just as strong as binding at 25 mM KCl, whereas F-actin binding by these other ABPs consistently decreases with increasing salt concentration (Cai et al., 2007; Dixon et al., 2008; Kostan et al., 2021). Nevertheless, the reduced co-sedimentation of MYPN with F-actin at elevated salt concentrations supports a role for electrostatic interactions in MYPN–F-actin binding.

The interaction between actin and certain ABPs, such as including filamin A, titin (T2 fragment and N2A region), and non-muscle α-actinins, is known to be regulated by calcium (Drmota Prebil et al., 2016; Dutta et al., 2018; Nakamura et al., 2007). Given that MYPN ablation does not alter myofilament calcium sensitivity (Filomena et al., 2020), we examined whether the presence of MYPN influences F-actin binding in a calciumdependent manner, as seen with other ABPs. Our experiments demonstrate that calcium does not modulate the binding between MYPN’s Ig3 domain and F-actin. Thus, the calcium-independent interaction between the Ig3 domain and F-actin could further explain the lack of effect on myofilament calcium sensitivity upon its ablation.

In summary, this study provides a potential mechanistic link between MYPN mutations and cardiomyopathy, revealing how disruption of MYPN’s actin-binding function drives aberrant accumulation of mutant proteins at the Z-disc and compromises sarcomere integrity. Importantly, we uncover that the Ig3 domain harbors two strategically positioned, solvent-exposed actin-binding sites, underscoring its pivotal role in orchestrating actin dynamics. Together, these findings position the Ig3 domain as a critical nexus in cardiac cytoskeletal regulation and a potential focal point for understanding and targeting MYPN-associated CMs.

4 |. MATERIALS AND METHODS

4.1 |. Mutagenesis and cloning

Human WT MYPN Ig3 domain cDNA (res 944-1048; NM_032578.3) was previously cloned into the pTBSG vector containing an N-terminal His₆-tag and a tobacco etch virus (TEV) protease cleavage site (Filomena et al., 2020; Filomena et al., 2021). MYPN Ig3 domain mutants were prepared using the QuickChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA) and the Q5 Site-Directed Mutagenesis Kit (New England Biolabs, Ipswich, MA). All the primers were obtained from Integrated DNA Technologies Inc. (Coralville, IA, USA). Human MYPN Ig3 WT was used as the template in the PCR reactions for all mutants except K952A, for which the K952A/R955A template was used. All constructs were confirmed by sequencing (Sequetech Corporation, Mountain View, CA). All primer sequences are provided in Supplemental Table S2.

The pTBSG Ig3 vector carrying the P961L mutation was digested with SpeI and KpnI restriction enzymes to linearize the plasmid and create a fusion site for MBP tag insertion. The MBP tag DNA fragment was obtained by digesting the pTBMalE Ig34 vector with the same enzymes (SpeI and KpnI). The vector and insert were ligated using T4 DNA ligase. Successful cloning of pTBMalE P961L was confirmed by agarose gel electrophoresis and DNA sequencing.

4.2 |. Protein expression and purification

All Ig3 domains were expressed and purified from E. coli BL21(DE3) cells as previously described (Filomena et al., 2020), with slight modifications. Namely, the Ni-NTA column elution buffer here contained 500 mM imidazole instead of 250 mM and all Ni-NTA buffers were adjusted to pH 8.0 (previously pH 7.4). The pTBMalE P961L construct was purified following the same protocol as described above up to the TEV protease digestion step. TEV cleavage was performed in the presence of 50 mM TCEP and BAL protease inhibitor. Following digestion, the eluate was diluted 1.6-fold with Ni-NTA lysis buffer. All buffers used were identical to those for other proteins, except the NaCl concentration was maintained at 500 mM. After confirming TEV digestion, the protein sample was subjected to additional purification using an amylose resin column (NEB) to remove the MBP tag. The flow-through fractions and lysis washes from the first amylose column were passed through a second amylose column to ensure complete removal of MBP. The combined flow-through fractions from both columns were then concentrated and further purified by size-exclusion chromatography on HiLoadR 16/600 SuperdexR 200 pg. (GE Healthcare) using the previously described storage buffer (Filomena et al., 2020), while maintaining 500 mM NaCl throughout the process. Final protein purity was verified by SDS-PAGE analysis.

4.3 |. Actin co-sedimentation assays

Actin was purified from rabbit muscle acetone powder (Pel-Freez Biologicals) following the method described by (Spudich & Watt, 1971). Binding assays were used to determine the apparent dissociation constant (K_d) of MYPN Ig3 domain variants binding to F-actin. The assays were conducted as previously described (Filomena et al., 2020), with details that were not included in the original method. The binding assays were performed using a fixed concentration (10 μM) of Ig3 constructs, which were incubated with pre-polymerized F-actin (0–30 μM) in 1X actin polymerization buffer (10 mM Tris [pH 8.0], 100 mM KCl, 2 mM MgCl₂, and 2 mM DTT), with all subsequent steps performed according to the referenced method. The K_d (μM) was determined by calculating the molar fraction of bound MYPN construct relative to the total MYPN in the assay, multiplying this fraction by the total MYPN concentration (10 μM) to obtain the concentration of bound protein, and plotting these values against the corresponding F-actin concentration (Kostan et al., 2021). Data fitting was performed in GraphPad Prism (v 10) using the one-site specific binding equation (Albraiki et al., 2023). To assess how basic residue and CM mutations affected MYPN Ig3’s ability to bundle or cross-link F-actin, bundling assays were conducted. Pre-polymerized F-actin (10 μM) was incubated with MYPN Ig3 variants (0–20 μM; data are shown for 10 and 20 μM) in 1X actin polymerization buffer for 1 h. The samples were then subjected to low-speed centrifugation (5000g) to sediment cross-linked F-actin bundles (B fraction). Subsequently, the supernatant was subjected to high-speed centrifugation (150,000g) to pellet non-cross-linked F-actin bound to Ig3 (P fraction). The final supernatant (S fraction) contained soluble G-actin monomers and unbound Ig3 (Zimmermann et al., 2016). Each of the fractions was analyzed by SDS-PAGE as previously described (Filomena et al., 2020). The actin bands in all the fractions were quantified using ImageJ software (Kostan et al., 2021) and the percentage of total actin in each fraction was calculated.

4.4 |. Actin copolymerization assays

Copolymerization assays were performed as previously described (Albraiki et al., 2023). Assays under polymerization conditions were conducted in 1X actin polymerization buffer (F-buffer), while assays under non-polymerization conditions were conducted in G-buffer (2 mM Tris–HCl [pH 8.0], 0.2 mM CaCl₂, 0.5 mM DTT, 0.2 mM ATP, and 0.01% NaN₃). MYPN Ig3 was stored in a buffer that does not induce actin polymerization. The assay procedure was identical to the actin bundling assay mentioned above.

4.5 |. Salt dependence of F-actin-MYPN Ig3 binding

MYPN Ig3 WT (20 μM) was incubated with 5 μM F-actin in 1X actin polymerization buffer with varying KCl concentrations (25, 50, 100, and 200 mM) for 1 h at room temperature in a total volume of 100 μL (Dixon et al., 2008). After incubation, the samples were centrifuged at 150,000g for 30 min. The salt dependence of the F-actin-Ig3 interaction was assessed as described in the actin binding assays section. Negative controls without actin were prepared in parallel to confirm that MYPN Ig3 remained soluble under low salt conditions (Dixon et al., 2008).

4.6 |. Calcium (Ca²⁺) dependence in F-actin-MYPN Ig3 interactions

The WT Ig3 domain (10 μM) was incubated with varying concentrations of polymerized actin (0–30 μM) in the presence of Ca²⁺ ions (pCa = 4) in 1X AP buffer at room temperature for 1 h in a total volume of 100 μL. The Ig3 bound to F-actin was sedimented as a pellet by centrifugation at 150,000g for 30 minutes. The effect of Ca²⁺ on the interaction between Ig3 and actin filaments was evaluated by constructing binding curves as previously described in the actin binding assays section. A negative control for Ig3-F-actin binding was performed under identical conditions but in the absence of Ca²⁺, achieved by adding EGTA to a final concentration of 1 mM (pCa = 10) (Dutta et al., 2018).

4.7 |. Circular dichroism (CD)

Purified MYPN Ig3 WT and mutant domains were dialyzed into CD buffer (10 mM sodium phosphate [pH 7.5], 50 mM NaF, and 1 mM TCEP). For the P961L mutant, the NaF concentration in the CD buffer was increased to 150 mM to stabilize this mutant. CD spectra were collected in the far-UV region (180–260 nm) at 20°C using a Jasco J-810 spectropolarimeter. Thermal denaturation was monitored at 205 nm with Ig3 WT samples at various protein concentrations (5.5–34 μM) to confirm consistent spectral changes. For comparison of all thermal denaturation curves, the protein concentrations for all Ig3 variants were maintained at approximately 30 μM (30–34 μM), except for R955W and C1002W, which were at 21 and 18 μM, respectively. All CD spectra were collected using a 1.00 mm pathlength quartz cuvette, ensuring consistency across all domains. To determine protein concentrations, theoretical extinction coefficients (ε₂₈₀) were calculated using Expasy-ProtParam (Gasteiger et al., 2003), with ε₂₈₀ of 8480 M⁻¹ cm⁻¹ for the WT and all mutants except for tryptophan (W) mutants which had an ε₂₈₀ of 13,980 M⁻¹ cm⁻¹. Thermal denaturation was studied by monitoring the CD signal at 205 nm as the temperature increased from 20 to 90°C, with a temperature ramp rate of 1°C/min and a 16 s time constant. Temperature control was maintained using a Julabo F25-ME refrigerated and heating circulator. Thermal denaturation midpoints (T_M) were estimated by fitting the changes in CD signal at 205 nm to a sigmoidal function, considering a two-state unfolding process, using GraphPad Prism. Molecular ellipticity (ME) signals were normalized to represent the percentage of folded protein across a temperature range of 20–90°C. Raw ellipticity data were converted to ME ([θ], degrees·m²·dmol⁻¹) using the standard formula as given below within the Spectra Manager software, which incorporates the path length (cm), protein concentration (μM), and the number of amino acid residues (108) in the Ig3 domain.

$$\text{ME}\left[\theta \right]=\frac{100\times \text{raw}\text{ellipcticity}\left(\theta \right)}{\text{concentration}\left(\frac{\text{mol}}{\text{L}}\right)\times \text{pathlength}\left(\text{cm}\right)\times \text{number}\text{of}\text{amino}\text{acids}}$$

Protein secondary structure content was quantified by analyzing CD spectra using the Dichroweb server (Miles et al., 2022; Whitmore & Wallace, 2004; Whitmore & Wallace, 2008). Spectra (260–180 nm) were processed with the K2D method.

4.8 |. Sedimentation equilibrium by analytical ultracentrifugation

Sedimentation equilibrium was performed with gelfiltered MYPN WT Ig3 in HEPES buffer (20 mM HEPES [pH 7.5], 150 mM NaCl, 2 mM DTT). A Beckman Coulter Optima XL-I analytical ultracentrifuge equipped with a four position An-60 Ti analytical rotor and a UV absorption detector was used for these experiments. The epon charcoal-filled centerpiece with six channels was filled with 110 μL of protein samples of three different concentrations (34, 58, and 314 μM) in the three channels of the top row while the three channels in the bottom row were filled with 120 μL of dialysis buffer before sealing the cell with a torque of 120 inchpounds. Samples were centrifuged at 27,000 rpm followed by 37,000 rpm until sedimentation and diffusion reached equilibrium. The state of equilibrium was recognized by monitoring raw absorbance spectra along the radial position over time. Further, the absorption wavelength was selected depending on the concentration of sample protein as the protein molecules tend to concentrate at the bottom of the cell due to centrifugal force. For this assay, the absorption of samples at concentrations 34 and 58 μM was collected at 280 nm while the absorption of the sample at 314 μM was measured at 305 nm. The radial absorption scans of protein concentration profiles were collected using XL-I Windows control software. The resulting data were analyzed using nonlinear least squares analysis according to the equation below using the Origin software provided by the instrument manufacturer (Demeler, 2010).

$$c\left(r\right)=c_O\exp \left[\frac{M_b{\omega }^2}{RT}\left(\frac{r^2-r_O^2}{2}\right)\right]=c_O\exp \left[\sigma \left(\frac{r^2-r_O^2}{2}\right)\right]$$

4.9 |. Expression of MYPN in Drosophila

4.9.1 |. Drosophila stocks and husbandry

All stocks were reared at 31°C on standard cornmealmolasses-yeast media. w1118 was used as the WT control in Figure 8b. The Mef2-GAL4 (Bloomington Drosophila Stock Center; RRID:BDSC_27390) driver was used to direct expression in muscle tissue.

4.10 |. Creation of transgenic flies

All UAS-MYPN lines (hMYPN WT, hMYPN R955W, hMYPN P961L, hMYPN WT-GFP, hMYPN R955W-GFP, and hMYPN P961L-GFP) were created using gene synthesis of full-length, 145 kDa human MYPN (NM_032578.3) by Genscript and subcloned into the pUAST vector using EcoRI and HpaI restriction sites. Plasmid DNA from all constructs was purified with a Qiagen Maxi Kit (Hilden, Germany), sequence verified, and sent to Rainbow Trangenics (Camarillo, CA) for the creation of transgenic flies. The insertions were balanced on their respective chromosomes. All lines were balanced over the Cyo, Tb (BL36335), or TM3, Sb/TM6, Tb balancers.

4.11 |. Larval muscle staining

Wandering L3 larvae were placed onto a Sylgard plate, pinned near the mouth hooks at the anterior end and the internal organs were removed. The muscle carcasses were fixed in 4% formaldehyde and immunostained using standard immunostaining protocols, with normal goat serum used as a blocking agent in PBT. Tissues were stained with the following primary antibodies: rabbit anti-MYPN (1:200; kindly provided by Marie-Louise Bang) (Bang et al., 2001) or secondary antibodies: Alexa Fluor anti-Rabbit 488 or 594 (1:400, Molecular Probes, Eugene, OR). Phalloidin 488 or 594 was used to label F-actin (1:400, Molecular Probes, Eugene, OR). A Zeiss 700 confocal microscope was used to capture the images. Image processing and analysis was performed using a combination of Zen Black (Zeiss), ImageJ (NIH), and Adobe Photoshop. Images taken at 5x are displayed as maximum intensity projections. Data acquisition at increased magnifications (63×) is presented as single plane confocal images.

4.12 |. Western blotting

Three whole L3 larvae were placed into 3× SDS sample buffer (188 mM Tris–HCl [pH 6.8], 3% [w/v] SDS, 30% [v/v] glycerol, 0.01% [w/v] bromophenol-blue, and 15% [v/v] β-mercaptoethanol), boiled at 95°C for 3 min, homogenized, boiled for an additional 10 min at 95°C, and centrifuged at 20,000g for 1 min to pellet debris. The resulting protein samples were separated by SDS-PAGE and transferred to the nitrocellulose blotting membrane (pore size 0.45 μm, Cytiva, Marlborough, MA) using the Trans-Blot^® TurboTM Transfer System (Bio-Rad, Hercules, CA). Membranes were stained with the Revert 700 Total Protein Stain (LI-COR Biosciences, Lincoln, NE) as a loading control. Membranes were probed with the rabbit anti-MYPN (1:1000; kindly provided by Marie-Louise Bang) (Bang et al., 2001). IRDye 800CW secondary antibodies (LI-COR Biosciences, Lincoln, NE) were used at 1:10,000. Membranes were developed using the LI-COR Odyssey XF and quantitation of protein levels was performed in Empiria Studio Software (LI-COR Biosciences, Lincoln, NE).

4.13 |. Statistical analysis

Statistical significance for in vitro and in vivo experiments was determined using one-way ANOVA Welch’s and Brown-Forsythe for comparisons between groups. Data are presented as individual data points or mean ± SD. Significance was defined as p < 0.05. All analyses were performed using GraphPad Prism version (10).

Supplementary Material

Additional supporting information can be found online in the Supporting Information section at the end of this article.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.