Alternatively spliced exon regulates context-dependent MEF2D higher-order assembly during myogenesis

Gönczi, Mónika; Teixeira, João M. C.; Barrera-Vilarmau, Susana; Mediani, Laura; Antoniani, Francesco; Nagy, Tamás Milán; Fehér, Krisztina; Ráduly, Zsolt; Ambrus, Viktor; Tőzsér, József; Barta, Endre; Kövér, Katalin E.; Csernoch, László; Carra, Serena; Fuxreiter, Monika

doi:10.1038/s41467-023-37017-7

Download PDF

Article
Open access
Published: 10 March 2023

Alternatively spliced exon regulates context-dependent MEF2D higher-order assembly during myogenesis

Nature Communications volume 14, Article number: 1329 (2023) Cite this article

1702 Accesses
2 Citations
3 Altmetric
Metrics details

Subjects

Abstract

During muscle cell differentiation, the alternatively spliced, acidic β-domain potentiates transcription of Myocyte-specific Enhancer Factor 2 (Mef2D). Sequence analysis by the FuzDrop method indicates that the β-domain can serve as an interaction element for Mef2D higher-order assembly. In accord, we observed Mef2D mobile nuclear condensates in C2C12 cells, similar to those formed through liquid-liquid phase separation. In addition, we found Mef2D solid-like aggregates in the cytosol, the presence of which correlated with higher transcriptional activity. In parallel, we observed a progress in the early phase of myotube development, and higher MyoD and desmin expression. In accord with our predictions, the formation of aggregates was promoted by rigid β-domain variants, as well as by a disordered β-domain variant, capable of switching between liquid-like and solid-like higher-order states. Along these lines, NMR and molecular dynamics simulations corroborated that the β-domain can sample both ordered and disordered interactions leading to compact and extended conformations. These results suggest that β-domain fine-tunes Mef2D higher-order assembly to the cellular context, which provides a platform for myogenic regulatory factors and the transcriptional apparatus during the developmental process.

Human skeletal muscle aging atlas

Article Open access 15 April 2024

Multimodal cell atlas of the ageing human skeletal muscle

Article Open access 22 April 2024

Nuclear mRNA decay: regulatory networks that control gene expression

Article 18 April 2024

Introduction

Myogenic lineage establishment and maintenance of terminal myogenic phenotype is controlled by precise dynamics of temporal and spatial expression, and hierarchical relationship between four myogenic regulatory factors (MRFs)¹. Myocyte-specific Enhancer Factor 2 (Mef2) protein belongs to the MADS (MCM1-Agamous-Deficiens-serum response factor) box family of transcription factors, which potentiate the transcriptional activity of myogenic proteins². Mef2 activity is precisely controlled during muscle development; in addition, the four isoforms encoded by separated genes in vertebrates further increase its functional variety³. The four MEF2 genes (MEF2A, -B, -C and -D) have spatially and temporally distinct expression during development⁴. Loss-of-function studies showed that mice lacking MEF2A and MEF2C were not viable owing to cardiac dysfunctions, however, MEF2D null mice were viable with cardiac defects⁵.

Mef2 is activated by myogenic regulatory factors^6,7 and the synergy with basic helix-loop-helix (bHLH) proteins is critical for regulating the expression of muscle-specific genes⁸. The interactions between Mef2 and myogenic proteins appears to be rather complex, involving a dimerisation of Mef2 as well as its bHLH/E12 partners⁹. A physical interaction between these dimers through the DNA-binding and dimerisation motifs has been proposed⁸. While bound to DNA, the MyoD basic region was shown to mediate the protein-protein interactions with Mef2¹⁰. Mutations of two basic residues of MyoD bHLH region, for example, impeded interactions with Mef2 and resulted in myogenic inactivity¹¹. The proximity of these proteins on DNA and their higher local concentrations contribute to transcriptional synergy during myogenesis¹⁰.

Tissue-specific alternative splicing further regulates the activity of Mef2 proteins during muscle differentiation^3,12. For example, the inclusion of a short acidic region, termed as the β-domain was in particular observed to potentiate transcription in all MEF2 gene products as compared to the splicing variants lacking this segment. The β-domain function was position-independent suggesting an autonomous transcriptional activity, not related to DNA binding or dimerisation¹³. Tissue-specific alternative splicing of short, disordered motifs often rewires protein interaction networks^14,15. In accord, the alternative inclusion of the β-domain was proposed to influence the coordination with myogenic proteins during differentiation. The characteristic feature of this seven-residue segment is the abundance of acidic residues, the mutation of which abolishes transcription¹³. Therefore, the β-domain was proposed to serve as an acidic blob¹⁶ within the Mef2 disordered transactivation domain (TAD, Fig. 1a), which may directly interact with the basal machinery. These acidic blobs function without strong structure and sequence constraints¹⁷ through a multitude of alternative interactions formed between the transcription factor and its partner¹⁸. Growing structural data on transcriptional assemblies^19,20 support that interaction plasticity is beneficial for fine-tuned regulation of the transcription machinery²¹. In addition, transcription regulation is also enabled by higher-order organisation of proteins^22,23, including a wide range of non-stoichiometric clusters²⁴.

**Fig. 1: Predicted structure and dynamics characteristics of the wild-type Mef2D and variant sequences.**

Based on these insights, we reasoned that the β-domain may influence the higher-order organisation of Mef2 proteins, enabling their coordination with both the myogenic regulatory factors and the basal machinery. To investigate how alternative inclusion of the β-domain contributes to myogenic transcription, we designed eight Mef2D variants with distinct β-domain properties using a combination of FuzPred method of predicting disordered interactions²⁵ and the FuzDrop method of predicting the probability of forming protein condensates²⁶. The variants with similar (var1, var2), mobile (var3, var4) and rigid (var5 – var8) β-domains as compared to the wild-type Mef2D (Figs. 1b, 1c) were also predicted to have different cellular behaviors and higher-order assembly (Fig. 1d). We in particular anticipated that the Mef2D variants with different β-domain dynamics will respond differently to changes in the cellular conditions during myogenesis. Then we have assessed the transcriptional activity of the Mef2D variants in reference to the wild-type protein and the isoform lacking the β-domain, termed as β-. We found a negative relationship between the β-domain dynamics and the transcriptional activity; more rigid variants were more potent to increase transcriptional activity in C2C12 cells. In parallel, we observed that more rigid variants, in particular var8 progressed the early phase of differentiation. Then we have characterised how β-domain dynamics affects Mef2D higher-order assembly. In accord with the predictions, we found that Mef2D formed nuclear puncta in C2C12 cells, the mobility of which resembled liquid-liquid phase separated condensates. In addition, we also observed solid-like aggregates in the cytosol, in particular promoted by a rigid β-domain variant (var8). Strikingly, solid aggregates were also formed by the mobile var4, which was predicted to undergo liquid-liquid phase separation as well as to readily convert between liquid-like and solid-like higher-order states (Fig. 1d). NMR and molecular dynamics simulations confirmed that depending on its flanking regions, the β-domain samples both ordered interactions, which can lead to solid-like aggregates as well as disordered interactions, which can result in liquid-like condensates. Based on these results we reason that the alternatively spliced β-domain serves as an interaction element for Mef2D higher-order assembly, which is required for transcription potentiation and differentiation. The characteristics of the assembly are fine-tuned by β-domain dynamics so that Mef2D can efficiently communicate with both the myogenic regulatory proteins and the basal machinery.

Results

Designing Mef2D variants with altered β-domain dynamics and higher-order assembly

Proteins sample a wide range of states from ordered to disordered conformations both in free and bound forms^27,28 and variations of structure and interaction patterns lead to functional plasticity²⁹. Most of the transactivation region (TAD) of Mef2D is predicted to be disordered by the Espritz method³⁰ in the free form, whereas the β-domain and its flanking regions appear to be more structured (Supplementary Fig. 1a). We used the FuzPred method to estimate the interaction behaviors, as this approach does not require the knowledge of the specific partner²⁵. These predictions indicate that the majority of the TAD region forms heterogeneous, disordered interactions³¹, whereas the β-domain serves as a more ordered recognition element (Supplementary Fig. 1b), which can establish well-defined contacts with the partner. In addition, a few motifs, (239–247, 331–339, 355–363, 427–434 residues, Supplementary Fig. 1b) are predicted to form ordered interactions, out of which K245 was identified as an acetylation site³². Regions, which form disordered interactions through heterogeneous and variable contacts, often promote liquid-liquid phase separation²⁶. Based on FuzDrop method²⁶, the majority of the Mef2D TAD is predicted to promote formation of liquid-liquid phase separated condensates (Fig. 1a). The β-domain, however, cannot spontaneously form condensates due to structural order (Fig. 1a). On the other hand, the β-domain is estimated to have a multiplicity of binding modes (MBM, Fig. 1a), which suggests a change between ordered and disordered interactions, the latter facilitating Mef2D partitioning in condensates³³.

We used three residue-specific dynamics characteristics to design β-domain variants: i) probability of structural disorder in the free state (p_D, computed by the Espritz method³⁰); ii) probability of disordered interactions (p_DD, computed by the FuzPred method²⁵); iii) droplet-promoting probability as predicted by the FuzDrop method (p_DP, computed by the FuzDrop method²⁶). Then we averaged these values for the β-domain (286-292 residues) and assessed the impact of mutations on three dynamics parameters (p_D, p_DD, p_DP) using structural disorder as the primary ranking (Methods). Then we combined mutations to increase the impact on β-domain dynamics and higher-order assembly. Two variants with similar β-domain dynamics to the wild-type (var1, var2) had additional negatively charged residues introduced at the C-terminal flanking region (Figs. 1b, 1c). Two variants with mobile β-domains (var3, var4) contained positively charged and polar residues (Figs. 1b, 1c). The four variants with rigid β-domains (var5 – var8) contained more aromatic and hydrophobic residues (Figs. 1b, 1c).

In addition, we also predicted how the variants affect Mef2D higher-order assembly based on the droplet-promoting probability (p_DP), which characterises the ability to spontaneously undergo liquid-liquid phase separation, and the multiplicity of binding modes (MBM), which indicates the likelihood of changing between liquid-like and solid-like states³¹. The complex cellular behavior is represented on the droplet landscape^34,35, which displays the droplet-probability as a function of the MBM³³ (Fig. 1d). For example, var3 and var4 both have disordered β-domain (Fig. 1c, Supplementary Table 1), yet they are predicted to exhibit different sensitivity to the cellular context. var3 has a positively charged lysine at position 292, which preferably forms disordered interactions and liquid-like condensates, almost independently on other cellular factors (low MBM, Fig. 1d). In contrast, var4 has an asparagine at position 292, which forms a NNN motif with the neighboring residues. This asparagine motif can sample both disordered and ordered interactions and thus is predicted to form liquid-like condensate or solid-like state depending on the cellular conditions (high MBM, Fig. 1d)³³. Similarly, introducing an asparagine into the C-terminal flanking region of var1 or adding aromatic groups to the β-domain of var5 and var6 also increases the sensitivity to the cellular conditions (Fig. 1d). Introducing both a negative and aromatic residue in var8 results in comparable cellular sensitivity to the wild-type protein (Fig. 1d). Interestingly, var8 is located at the diagonal of the droplet landscape, suggesting that the transition from droplet to amyloid states can be achieved through a combination of deposition and condensation pathways^33,34. The dynamic characteristics of the wild-type protein and the variants are detailed in Supplementary Table 1.

Taken together, we designed eight Mef2D variants containing two to three mutations with distinct impact on β-domain dynamics and higher-order assembly. These mutations have not been identified before and their combinations served to probe the role of the β-domain in muscle cell differentiation.

β-domain rigidity potentiates transcription activity

First, we measured the transcriptional activity of Mef2D variants in cycling and differentiated C2C12 cells. The variants were cloned into pCMV eukaryotic expression vectors and were transfected together with two reporter plasmids: MCK-Luc and pSV-βgal (Methods). Transcription efficiency was characterised by the luciferase activity of the variants, which was normalised to the galactosidase signal. In non-differentiating cells, all variants with rigid β-domains (var5 – var8) exhibited increased transcriptional activity as compared to the wild-type protein (Fig. 2a). Interestingly, var3 and var4 with mobile β-domains also exhibited slightly elevated transcriptional activity (Fig. 2a). In differentiated C2C12 cells, however, only the var8 exhibited significantly higher transcription activity as compared to the wild-type Mef2D (Fig. 2b). The β- variant was less potent to activate transcription in accord with previous results¹³ (Figs. 2a, 2b).

**Fig. 2: Transcriptional activity of MEF2D variants in C2C12 cells.**

Then we measured the transcriptional activity without the endogenous Mef2D background using knock-out (KO) cultures generated by CRISPR/Cas9 technique³⁶ (Methods). The C2C12 KO cell line was not capable to generate multinucleated myotubes in the absence of Mef2D (Fig. 2c). The transcriptional activity of the variants without the endogenous wild-type background exhibited similar trends as observed in non-differentiated C2C12 cells. The variants with rigid β-domain (var5 – var8) potentiated transcription as compared to the wild-type Mef2D, while the variants with similar β-domain dynamics as the wild-type (var1, var2) showed comparable activity (Fig. 2D). The variants with mobile β-domain did not exhibit a uniform behavior, while the transcriptional activity of var3 was slightly decreased, that of var4 considerably increased as compared to the wild-type, indicating a more complex mechanism.

Transcriptional activity negatively correlated with the structural disorder of the β-domain (Supplementary Table 2), suggesting that its rigidity promotes interactions with the transcriptional apparatus.

β-domain rigidity promotes early myotube development

In the next step, we assessed the impact of the different variants on skeletal muscle differentiation followed by desmin immunocytochemistry and transmission light microscopy (Methods, Fig. 3). From day 2, we observed multinucleated myotubes with wild-type Mef2D and all variants, and to lesser extent in case of the β- variant. A marked difference between different variants appeared in particular on day 3, when more rigid variants promoted myotube formation as compared to the wild-type (Fig. 3a). At the end of the differentiation process a similar amount of myotubes were formed with all variants, except the one lacking the β-domain (Fig. 3a). The progress of differentiation was quantified by the fusion index, defined as the ratio of the nuclei number in multinucleated myocytes versus the total number of nuclei within the visual fields (Methods, Supplementary Fig. 3). Quantitative analysis corroborated that variants with rigid β-domain, in particular var8, promoted the early phase of differentiation (Supplementary Fig. 3). In addition, var4 with mobile β-domain, also showed a faster myotube development, which was not observed in the presence of var3 (Supplementary Fig. 3).

**Fig. 3: Differentiation progress in the presence of Mef2D variants.**

Mef2D physically interacts with myogenic regulatory factors^8,9, thus we monitored their protein expression levels by Western blot analysis at each day of the differentiation process. We found increased levels of both the early differentiation regulator MyoD (Fig. 3b) and the late differentiation marker desmin (Fig. 3c) in the presence of rigid β-domain variants at the early phase of differentiation (days 1 and 2). The impact on desmin was prolonged also on day 3 of differentiation (Fig. 3c). This suggests that more ordered β-domains stabilise Mef2D interactions with the myogenic regulatory factors contributing to potentiation of transcription¹⁰. Along this line, we did not find higher desmin or MyoD levels in the presence of var4 with mobile β-domain (Figs. 3b, 3c), although this variant also exhibited progress in myotube development.

Taken together variants with rigid β-domain promote the early phase of differentiation and stabilise the myogenic factors.

Mef2D forms higher-order assemblies in C2C12 cells with both liquid-like and solid-like properties

Then we analyzed the cellular organisation of representative Mef2D variants and overexpressed the wild-type, var3, var4 and var8 in C2C12 cells grown under cycling conditions and in presence of differentiation medium. Rapid turnover and accumulation upon inhibition of the proteasome indicates that Mef2D and its variants are disordered³⁷ (Supplementary Fig. 4a, Methods). Confocal microscopy revealed that wild-type Mef2D and var3, var4 and var8 formed higher-order assemblies, which are mainly located inside the nucleus (Fig. 4a). Fluorescence recovery after photobleaching (FRAP) analysis indicated high mobility of the wild-type protein as well as var3, var4 and var8 inside the nucleoplasm (Supplementary Fig. 4b) and nuclear foci (Fig. 4c), resembling liquid-liquid phase separated condensates³⁸. FRAP analysis in U2OS cells corroborated the presence of mobile nuclear condensates (Supplementary Fig. 5), in case of all variants, irrespective of the β-domain dynamics. These results are in line with the FuzDrop predictions²⁶ that Mef2D and all the variants are capable of spontaneously undergoing liquid-liquid phase separation (p_LLPS ~ 0.99).

**Fig. 4: Mef2D forms liquid-like and solid-like higher order assembly.**

In addition, we observed cytoplasmic foci reminiscent of higher-order structures²⁴ (Fig. 4a). In contrast to the nuclear condensates, the cytoplasmic foci exhibited low mobility (Fig. 4d), indicating solid-like states. The percentage of cells displaying cytoplasmic higher-order assemblies was significantly higher for var8 with rigid β-domain, both in cycling and differentiating myoblasts compared to wild-type Mef2D (Fig. 4b). Surprisingly, we also observed that var4 with mobile β-domain tends to promote the formation of cytoplasmic aggregates in cycling cells (Fig. 4b). The increased propensity of var8 and var4 to form higher-order structures was not due to higher expression levels, since we found similar protein levels in C2C12 cells (Supplementary Fig. 4a).

As both var8 and var4 promoted the early phase of myotube development, we reasoned that ordered higher-order assemblies facilitate myogenesis. In accord, we found a negative correlation between the transcription activity in non-differentiating cells and the probability of β-domain to form liquid-like condensates (Supplementary Table 2).

Structure and dynamics of Mef2D model peptides

To provide molecular insights how the β-domain modulates the Mef2D higher-order assembly, we have employed different NMR methods and molecular dynamics simulations (Methods) using a 37-residue peptide model of the wild-type Mef2D and representative variants (var3, var4 and var8) containing the β-domain and its flanking regions (UniProt code: Q14814, 265-301 residues, Fig. 1b).

Order parameters (S²) derived from the backbone chemical shifts based on the Random Coil Index (RCI)³⁹ indicated the disordered nature of all peptides (Supplementary Fig. 6a). var8 exhibited an increased ordering in particular at D29 and W31 mutant residues. Higher R2/R1 values of var8, in particular at the C-terminal region suggested conformational exchange between compact and extended conformers corroborating partial ordering (Supplementary Fig. 6b). Diffusion ordered spectroscopy (DOSY) indicated a faster diffusion of var8 (D = 2.95*10⁻¹⁰ ± 1.15*10⁻¹⁰ m²/s) than the wild-type peptide (D = 1.54*10⁻¹⁰ ± 2.61*10⁻¹² m²/s) suggesting more compact conformations (Supplementary Fig. 6c). In contrast, lower diffusion coefficient of var3 (Supplementary Fig. 6c) suggested more extended structures, consistently with the decreased order parameters (Supplementary Fig. 6a).

In accordance with the NMR data, molecular dynamics simulations indicated that the studied Mef2D model peptides (wild-type, var3, var4 and var8) sample a large regime of their conformational space (Supplementary Fig. 7a). In particular, the disordered Mef2D peptides adopt both compact and extended conformations (Supplementary Fig. 7b, Methods) and exhibit distinct contact patterns depending on their β-domain properties (Figs. 1c, 1d, Methods). The characteristic contacts in var8 model peptide are formed dominantly within the β-domain (56 % of all contacts), through a consecutive set of residues (Fig. 5a). The β-domain less frequently contributes to intra-molecular contacts in var4 and the wild-type peptide (34 % and 36% of the contacts, respectively) and is rarely involved in structure formation of var3 (Fig. 5a). In line with the NMR data, structure analysis indicates that var8 forms more compact structures (Fig. 5b, Supplementary Fig. 7b) mediated by hydrophobic contacts (Leu26 and Leu28 residues) and partly aromatic interactions (His25 and Trp31, Fig. 5b, Supplementary Fig. 7c). var4 can sample both compact conformations mediated by polar and aromatic contacts (Thr22, His25 and Asn30 residues), as well as extended conformations mediated by charge-charge interactions mostly with the C-terminal region (Glu23 and Arg33, Fig. 5b, Supplementary Fig. 7b and c). The compact conformations of the wild-type peptide were also mediated by hydrophobic interactions, which are less frequent than those in var8 (Fig. 5b, Supplementary Fig. 7b). In addition, the wild-type peptide also forms charge-charge interactions with the N-terminal region (eg. Asp27-Arg2, Fig. 5b). In contrast, var3 rarely samples compact conformations and forms some local, transient contacts at both the C- and N-terminal regions (Fig. 5b, Supplementary Fig. 7b).

**Fig. 5: Representative structural features of Mef2D variant peptides from molecular dynamics simulations.**

Structure analysis of model peptides suggests that different biophysical forces shape the conformational landscape of the Mef2D variants, and mutations in the β-domain remarkably change the structural variability and compactness of the peptide models.

Discussion

Elaborating the relationship between the protein state and its function presents a new paradigm in understanding of cellular processes⁴⁰. Transcriptional programmes controlling cell-line identity are frequently regulated by high density of transcriptional apparatus^41,42, often generating distinct compartments⁴³. These transcriptional clusters, also termed as super-enhancers²³, are stabilised by dynamic and heterogeneous interactions²². In general, the material states of such higher-order assemblies must be precisely controlled^24,35. Solidifying the liquid-like state of a-kinase anchoring protein AKAP95, for example, has been shown to interfere with alternative splicing and initiate tumorigenesis⁴⁴, while ASC interactions are conditioned to the hardening of NLRP6 condensates⁴⁵.

Skeletal muscle development may also involve transcriptional clusters⁴⁶ formed by myogenic proteins and Mef2^2,8. Proximity of their DNA-binding sites (Supplementary Fig. 8) promotes physical interactions between these factors^8,10, which is required to activate transcription of muscle-specific genes⁴⁶. We reasoned that the alternative spliced, acidic β-domain contributes to the orchestration of myogenic regulatory factors. Based on sequence analysis and structure predictions (Fig. 1), we anticipated that the β-domain serves as an ordered interaction element within the disordered transactivation region (Supplementary Fig. 1). Thus, by introducing mutations into the β-domain and its flanking region based on FuzDrop predictions, we aimed at modulating Mef2D higher-order assembly.

We have found a negative correlation between the transcriptional activity of the Mef2D variants and β-domain dynamics (Fig. 2, Supplementary Table 2); in particular rigid β-domain variants (var5 –var8) were more potent to activate transcription in C2C12 cells before differentiation (Fig. 2). Along these lines, the increased transcription activity of the rigid β-domain variants also promoted the early phase of myotube development (Fig. 3, Supplementary Fig. 3), also reflected by altered expression of MyoD and desmin (Fig. 3). Surprisingly, var4 with mobile β-domain also promoted transcription and early phase differentiation (Figs. 2, 3, Supplementary Fig. 3), without increasing the expression of myogenic regulatory proteins (Fig. 3).

We observed that Mef2D forms both nuclear and cytoplasmic foci (Fig. 4). The mobility of the nuclear foci in both C2C12 and U2OS cells resembled mobile nuclear condensates formed through liquid-liquid phase separation (Fig. 4c, Supplementary Figs. 4 and 6). In contrast, we observed the lack of mobility in the cytoplasmic foci (Fig. 4d), indicating the presence of solid-like aggregates. Formation of solid-like assemblies were promoted by var8 with rigid β-domain, as well as by var4 with mobile β-domain (Fig. 4b). var4 was predicted to spontaneously undergo liquid-liquid phase separation as well as to convert between liquid-like and solid-like states (Fig. 1d, Supplementary Table 1). The presence of higher-order structures with both liquid-like and solid-like characters may be explained by two molecular mechanisms. In the first scenario, these assemblies are formed independently, the liquid-like condensates through phase separation and the solid-like aggregates through ordered assembly or deposition^33,34. In the second scenario, the maturation of the liquid-like condensates leads to the formation of solid-like aggregates. The distinct behaviors of the mobile β-domain variants var3 and var4 seem to be consistent with the second scenario. These two variants exhibit remarkably different cellular context-sensitivity as indicated by the droplet landscape (Fig. 1d). The context-dependent var4 is more potent to activate transcription and promotes differentiation as compared to the less context-dependent var3 (Figs. 2, 3), indicating that a transition of Mef2D between liquid-like and solid-like states may be required during myogenesis. We cannot exclude the possibility that the liquid-like and solid-like states are formed independently due to the different cellular milieu in the nucleus and cytosol. In this case, however, formation of these higher-order assemblies must be regulated through independent pathways during muscle cell differentiation. Instead, we favor the simpler model, in which the transition between the liquid-like and solid like states is required for potentiation of transcriptional programmes during myogenesis.

Based on these results, we propose that transcription potentiation of Mef2D is linked with its higher-order assembly, which enables simultaneous interactions with both the basal machinery and the myogenic factors to activate distinct gene expression programs for muscle lineage development. Our results support the model that Mef2D assemblies are formed by disordered interactions of the transactivation domain. The alternatively spliced β-domain serves as an ordered interaction element to stabilise the disordered assembly. The sensitivity of the β-domain and its flanking regions to the cellular context may facilitate switching between ordered and disordered interactions and, thereby, the adaptation of this segment to the cellular conditions during different stages of the muscle development. Along these lines, we observe both liquid-like and solid-like Mef2D assemblies in C2C12 cells, raising the possibility of their interconversion. The transcription activation and differentiation progress by the droplet-promoting and context-dependent var4, in contrast to the context-insensitive var3, conforms to this scenario. A similar mechanism likely operates during myofibrillogenesis to organise Z-disc formation by FATZ condensates⁴⁷.

In summary, our results shed light on how dynamic properties of an alternatively spliced segment fine-tune Mef2D higher-order assembly to the cellular context, providing a complex regulatory mechanism for gene-expression during skeletal muscle development.

Methods

Computational design

We considered the human Mef2D sequence (UniProt code: Q14814; https://legacy.uniprot.org/uniprot/Q14814) and computed three residue-specific dynamics characteristics: i) structural disorder in the unbound state using the Espritz method (p_D)³⁰ as implemented into the FuzPred algorithm (https://fuzpred.bio.unipd.it) ii) disordered interactions as predicted by the FuzPred method (p_DD)²⁵ iii) droplet-promoting probability as predicted by the FuzDrop method (p_DP)²⁶ and we averaged the values of dynamics descriptors (p_D, p_DD, p_DP) for the β-domain (residues 286-292). We introduced mutations in 286-301 residues and computed the differences between the mutant and wild-type β-domain dynamics: i) structural disorder Δp_D = p_D(mut) - p_D(wt); ii) disordered interactions Δp_DD = p_DD(mut) - p_DD(wt); iii) droplet-promoting probability Δp_DP = p_DP(mut) - p_DP(wt), where each probability value is the average of the respective probabilities of the β-domain residues. Mutations with largest impact on the β-domain Δp_D and with most diverse chemical nature were selected (Fig. 1B). We also considered the droplet-promoting probability (Δp_DP) and the multiplicity of binding modes (MBM) for ranking, to generate variants with distinct higher-order assembly³¹. The droplet landscape was used to inform on the cellular behavior of the higher-order assembly, in particular the likelihood of changing between liquid-like and solid-like higher-order assemblies^33,34 (Fig. 1D). The variants included two sequences with similar β-domain dynamics to the wild-type (var1, var2), two variants with more mobile β-domain (var3, var4), and four variants with more rigid β-domain (var5 - var8) as compared to the wild-type (Figs. 1B, 1C, 1D). The dynamics characteristics of the variants are summarised in Supplementary Table 1.

Cell culturing and differentiation

HEK293 (human embryonic kidney) cells and mouse immortalized C2C12 myoblast cell lines (both from ATCC) were cultured in High Glucose DMEM (Biosera, Nuaille, France) supplemented with 10% fetal bovine serum (Gibco by Life Technologies, Carlsbad, California, USA), 1% Penicillin-Streptomycin (Gibco), and 1% L-glutamine (Biosera). The medium was changed every other day, and cells were sub-cultured at 80-90% confluence. The culturing media was exchanged to DMEM containing 2% horse serum (Gibco) to differentiate C2C12 myoblasts into multinucleated myotubes. Differentiation was followed by transmission microscopy for 6 days.

CRISPR-Cas9 experiments

C2C12 cells were transfected with MEF2D specific CRISPR/Cas9 knockout and HDR plasmid constructs, targeting 3 different places in coding sequence (in exon# 3, 4 and 5; Santa Cruz Technology, Dallas, TX, USA). Forty-eight hours after the lipid-mediated transfection puromycin selection (1.5 µg/ml) was applied for 5 days than single cells were sorted by the presence of GFP and PRF expressed from the KO and HDR vectors, respectively, using FACS Aria III flow cytometer (BD Biosciences). MEF2D-KO colonies were grown from individual single cells.

Determination of transcription activity, reporter plasmids

MEF2D variants with modified dynamical properties were cloned into pCMV-HA eukaryotic expression vector (Genscript, Piscataway, NJ, USA). To determine the transcriptional activity of Mef2D variants, MCK-Luc reporter plasmid (Addgene, Watertone, MA, USA) and pSV-β-galactosidase control vector (Promega) were used together with the variant plasmids. Control and MEF2D-KO cells (description for the generation of KO cultures is in the section above) were transfected with the vectors of the variants using Lipofectamine 2000 transfection reagent (Life Technologies, Carlsbad, CA, USA). 48 hours later luciferase and galactosidase activity were determined by Luciferase and β-galactosidase enzyme assays (both from Promega) following the manufacturer´s instructions. Luminescent and absorbance data were collected by Synergy H1 microplate reader to quantify the transcriptional activity of the samples. Microsoft Excel 2016 was used to process transcription and differentiation data.

Immunocytochemistry and calculation of fusion index

Cultured cells were fixed with 4% PFA followed by washing with 0.1 M glycine in PBS to neutralize excess formaldehyde. Cells were permeabilized with 0.25% Triton-X (Sigma-Aldrich) and blocked with PBS containing 1% BSA. The samples were incubated with primary antibodies (anti-desmin, 1:1000) diluted in blocking solution overnight at 4 °C. Following the application of fluorophore-conjugated secondary antibodies (Biotinylated anti-mouse IgG (1:300), Alexa Fluor 488 conjugated Streptavidin (1:100); both from Thermo Fisher Scientific) mounting medium containing DAPI (Vector Laboratories, Burlingame, USA) was added to each coverslip. Immunolabelled samples were investigated with a Zeiss LSM510 laser scanning confocal microscope (Zeiss, Oberkochen, Germany). Fusion index was quantified as the ratio of the nuclei number in myocytes with two or more nuclei versus the total number of nuclei within a visual field.

Immunofluorescence on cultured cells

C2C12 cells were grown on polylysine-coated glass coverslip. C2C12 cells were transfected using Lipofectamine 3000 (Life Technologies) following manufacturer instructions. 24 hrs post-transfection, cells were washed with cold PBS and fixed with 3.7% formaldehyde in PBS for 9 minutes at room temperature, followed by permeabilization with ice-cold acetone for 5 minutes at −20 °C. PBS containing 3% BSA and 0.1% Triton X-100 was used for blocking and incubation with primary (MEF2D, 14353-1-AP, Proteintech) and secondary antibodies (Alexa Fluor® 488, A‐21202, Thermo Scientific). Images were obtained using a Leica TCS-SP8 (Leica Microsystems) and a 63x oil immersion lens.

Western blot analysis

For Western blot analysis cells were harvested in lysis buffer, the protein content of samples was measured by a BCA protein assay (Pierce, Rockford, IL, USA). Protein samples were separated in 7.5% SDS–polyacrylamide gels, then transferred to nitrocellulose membranes (Whatman, Maidstone, England). After blocking non-specific binding sites, membranes were exposed to primary antibodies overnight at 4 °C as follows: mouse monoclonal anti-MEF2D (1:500; BD Biosciences, San Jose, CA, USA), monoclonal mouse myoD (Novus Biologicals, Abington, UK) antibody in 1:200 dilution, mouse monoclonal anti-desmin in 1:1000 dilution (BD Biosciences) and goat polyclonal anti-actin (1:250; Santa Cruz Biotechnology, Dallas, USA). Following incubation with horseradish peroxidase-conjugated, respective anti-goat or anti-mouse IgG antibodies (Bio-Rad, Hercules, CA, USA; 1:1000, 1 hour) the immunoreactive bands were visualized by enhanced chemiluminescence (Pierce or Millipore, Billerica, MA, USA). The optical density of signals was measured using a Kodak Gel Logic 1500 system and further analyzed by using ImageJ 1.40g freeware.

Confocal microscopy and Fluorescence recovery after photobleaching (FRAP) analysis

FRAP measurements on C2C12 and U2OS cells transfected with GFP-tagged Mef2D WT, var3, var4 and var8 were performed using a confocal microscope Leica TCS SP8 (Leica Systems). For FRAP analysis we used a 63x oil immersion objective. Bleaching was obtained by using a laser intensity of 100% at 405 nm for 1 s. Recovery was recorded for 120 time points after bleaching (120 s). Analysis of the recovery curves were carried out with the FIJI/ImageJ. Fluorescent density analysis was performed using FIJI/ImageJ and selecting specific region of interest (ROI). The flow of the protein was measured by quantifying the recovery of the bleached area at the cost of the unbleached region and using a custom written FIJI/ImageJ routine. The bleached region was corrected for general bleaching during image acquisition. We quantified the molecules that move from the unbleached region to the bleached region, leading to recovery of the bleached region. Prior to FRAP analysis, the images were corrected for drift using the StackReg plug‐in function of the FIJI software suite. The equation used for FRAP analysis is: ((Ibleach − Ibackground)/(Ibleach(t0) − Ibackground(to)))/((Itotal‐Ibackground)/(Itotal(t0) − Ibackground(to))), where Itotal is the fluorescence intensity of the entire cellular structure, Ibleach represents the fluorescence intensity in the bleach area, and Ibackground the background of the camera offset. FRAP curves were averaged to obtain the mean and standard error mean.

Graphics and statistics

Pooled data were expressed as mean ± standard error of the mean (SEM). The mean and SEM were calculated as weighted averages and weighted standard errors with weight corresponding to the number of samples and number of experiments. Box plots and line graphs were constructed using Origin 8.6 graphic software. In all box-plot type presentations the rectangles in the box plots present the median and the 25 and 75 percentile values, while the error bars point to 1 and 99%. The differences between control cultures and cells overexpressing Mef2D variants were assessed using one-way analysis of variance (ANOVA) and student’s t-test function of Microsoft Excel 2016, the significance and a p value of less than 0.05 was considered statistically significant. Vector graphics were prepared by Inkscape 0.92.5 (2060ec1f9f, 2020-04-08), https://www.inkscape.org.

Molecular Dynamics (MD) simulations

Model system

The starting models of the 37 residue peptides of the wild-type protein, var4 and var8 terminated by acetyl (ACE) and N-methyl (NME) groups were generated by the LEaP module of AMBERTools17⁴⁸. The solute was neutralised by adding Na⁺ and Cl^- ions randomly and then solvated by 12 Å of TIP3 water⁴⁹. Periodic boundary conditions using a truncated octahedron geometry with cell edge of 151 Å was used with particle-mesh Ewald method⁵⁰ for long-range electrostatic interactions and 8 Å non-bonded interaction cut-off. All simulations were performed using a Charmm36m force field, which was developed for intrinsically disordered proteins⁵¹.

Computational protocol

Extended conformations of the models were subjected to minimization by steepest descent and conjugate gradient methods. The system was heated from 0 K to 50 K in 50 ps using Langevin thermostat, then from 50 K to 310 K in 1 ns using Berendsen thermostat while applying a harmonic restraint of 4 kcal mol⁻¹/Å² for all atoms. Water density was adjusted to ~1 g/cm³ in 500 ps NPT simulation using Berendsen barostat, applying 1 kcal mol⁻¹/Å² restraint on the backbone atoms, which was then released for 1 ns NPT simulation. Then three independent 100 ns NPT simulations at 310 K and 1 bar were carried out for each model, with 2 fs timestep applying the SHAKE algorithm on the hydrogen geometry. Coordinates were saved in each 10 ps. The translational center of mass motions was removed in a wrapping procedure every 1000 steps.

MD analysis

The MD trajectories were analysed using the MDTraj Python package v1.97⁵². We considered the 70 to 100 ns trajectory of each replica based on the RMSD values in reference to the average structure (Supplementary Fig. 7). Radius of gyration of was computed based on the method developed for disordered proteins⁵², using MDTraj Python package. Clustering of the conformations of the disordered peptides was based on the contact map similarity³¹. Contacts were computed using GetContacts (https://getcontacts.github.io, commit: b0777f7148a327387133d9632f8ed1e34bfaa282). We defined the similarity between two contact maps as: \({f}_{S}=\frac{{C}_{i}\cap {C}_{j}}{{(C_{i} \cup {C}_{j})}-(C_{i}\cap {C}_{j})}\), where C_i and C_j are the set of contacts in snapshot i and j, the numerator is the common contacts of the two sets and the denominator is the total number of different contacts of the two sets. We clustered two snapshots if f_S > 0.5, and the clusters were joined based on the frequencies of the common contacts. For graphical clarity, we visualised contacts only in the four most populated clusters. The calculations were performed with Python (https://pypi.org) and Numpy (https://numpy.org) libraries, plots were prepared using the matplotlib python package (https://matplotlib.org/). Structure visualisation was done using PyMol (The PyMOL Molecular Graphics System, Version 2.0 Schrödinger, LLC).

NMR analysis

37-residue peptide sequences of the wild-type, var3, var4 and var8 containing the β-domain and its flanking residues (UniProt code: Q14814, 265-301 residues) were ¹⁵N labelled on leucines (as underlined SRKPDLRVITSQAGKGLMHHLTEDHLDLNNAQRLGVS). All NMR experiments were carried out on a Bruker Avance Neo (700 MHz ¹H frequency) spectrometer equipped with a Prodigy TCI cryoprobe, and the temperature was set to 298 K. Chemical shifts were referenced to DSS. Peptides were dissolved in 5% D₂O and 95% H₂O 20 mM PBS buffer solution, with a final peptide concentration of 1 mM. For NMR data acquisition and evaluation, Bruker’s Topspin 4.0 software package was utilized.

NMR resonance assignment

2D correlation spectra, including ¹H-¹H TOCSY and ¹H-¹H ROESY with ¹⁵N decoupling, ¹³C-¹H HSQC, and ¹³C-HSQC-TOCSY, ¹⁵N-¹H HSQC, ¹⁵N-¹H HSQC-TOCSY, and -ROESY.

Determination of order parameters (S²) from chemical shifts

Chemical shifts (C_α, C_β, and H_α) were used to extract Random Coil Indices and NMR order parameters (S²). First, the offset-corrected data were converted to secondary shifts, then site-specific corrections were applied to calculate Random Coil Indices⁵³. Residue order parameters (S²) were derived from the chemical shifts of Cα, Cβ, and Hα backbone atoms³⁹.

Relaxation measurements

¹⁵N-relaxation data (T₁, T₂) were obtained by Bruker pseudo 3D ¹⁵N-¹H correlation experiments with double INEPT-transfer and decoupling during acquisition. The relaxation delay times were set to 10, 20, 40, 80, 120, 160, 200, 300, 400, 600, 1200, and 2400 ms for T₁ and the Carl-Purcell-Meiboom-Gill (CPMG) pulse trains of 17, 34, 51, 68, 85, 102, 119, 153, 204, 254, and 305 ms were used for T₂-measurements. 64 increments were collected, and the number of transients was 8 for T₁/T₂-measurements and 80 for the heteronuclear (¹⁵N-¹H) NOE experiment to obtain a proper S/N ratio. Reduced Spectral Densities⁵⁴ and R2/R1 ratios of ¹⁵N nuclei of isotopically labeled leucine residues were derived from relaxation data (¹⁵N T₁, T₂ and NOE).

Diffusion ordered spectroscopy

(DOSY) Diffusion coefficients were derived from a measurement with a stimulated spin echo LED (Longitudinal-Eddy-current Delay) sequence with bipolar gradients and two spoil gradients⁵⁵. The length of the diffusion delay (75 ms) and gradient pulses (2 ms) were optimized by short, one-dimensional (1D) experiments. The gradient strength was split to 32 increments, linearly varying between 5 % and 95 % of its maximum value. The maximum gradient strength of the probe was 57.7 G/cm. The number of scans was set to 16. Data processing was performed and 2D DOSY plots were generated by the dosy module of Topspin. Diffusion coefficients were calculated by an exponential fit of the signal intensity decays obtained with increasing gradient strength.

ChIP-Seq Protocol

C2C12 myoblasts and myotubes from late differentiation stage were cross-linked using disuccinimidyl glutarate (DSG, Sigma) followed by fixation with 1% methanol-free ultrapure formaldehyde (Thermo Scientific). After washing, cells were harvested in ChIP lysis buffer and centrifuged at 12,000g for 1 min at 4 °C. Resuspended chromatin samples were sonicated to get fragments between 100 and 2000bp. Chromatin IP was performed using anti-MEF2D antibody (NBP-1-80567, Novus Biologicals) with overnight incubation at 4 °C. Antibody-chromatin complex was centrifuged (3,500 rpm, 20 min, 4 °C), then samples were incubated with paramagnetic beads (Dynabeads Protein A, Thermo Scientific) for 6h at 4 °C. Antibody-chromatin-bead complexes were washed with different buffers, then DNA-protein complexes were eluted and de-cross-linked by adding NaCl (0.2 M). ChIPed DNA was purified using Qiagen MinElute columns and prepared for sequencing according to the manufacturer’s instructions (Illumina).

ChIP-seq data analysis

The library was sequenced at the Genomic Medicine and Bioinformatics Core Facility at the University of Debrecen, Hungary using single end 75 bp Illumina technology. The raw reads were aligned to the mm10 reference genome sequence using Burrows-Wheeler Aligner. The resulted file was further analyzed using the Homer approach⁵⁶, involving the peak calling, the denovo motif finding and the annotation of the peaks and motifs under the peaks. The top 1000 peaks based on the Homer’s peak scores were analyzed for denovo motifs. The most enriched motifs, the hlf motif (https://jaspar.genereg.net/matrix/MA0043.2/), and the Mef2A motif were used to annotate the peaks in the Mef2D enhancer region. In addition to Mef2D, transcription factors binding to the creatine kinase gene (Mck) enhancer region were derived from the ChIPSummitDB database (https://summit.med.unideb.hu/summitdb/).

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

Source data are provided with this paper. The data that support the findings are available in the data repository of University of Debrecen at https://doi.org/10.48428/ADATTAR/3RFQRC. ChIP-seq sequencing data can be accessed in the SRA database of NCBI project number PRJNA887931, and GEO accession code GSE224053. Transcription factors binding to the creatine kinase gene (Mck) enhancer region used for ChIP-seq data analysis were derived from the ChIPSummitDB database (https://summit.med.unideb.hu/summitdb/). Mef2D sequence was derived from the UniProt database (https://www.uniprot.org/) Q14814 (main isoform). NMR assignments have been deposited in BMRB database with the codes BMR51840 [https://doi.org/10.13018/BMR51840] (wt), BMR51841 [https://doi.org/10.13018/BMR51841] (var3), BMR51842 [https://doi.org/10.13018/BMR51842] (var4), BMR51843 [https://doi.org/10.13018/BMR51843] (var8). Molecular dynamics trajectories were deposited as https://doi.org/10.5281/zenodo.7657119. Source data are provided with this paper.

Code availability

The scripts used to process the trajectories and the results generated by MDTraj Software version 1.9.7 and GetContacts (https://getcontacts.github.io/commit:b0777f7) as well as to produce the plots used in the article are available: https://drive.google.com/drive/folders/1IDFZwTUmoCPy4T_B_AGRKCL2OQq_rZOC. The link also contains the processed MD trajectories used for analysis. The underlying libraries are: Python, Numpy (https://numpy.org) and matplotlib (https://matplotlib.org/).

References

Hernández-Hernández, J. M., García-González, E. G., Brun, C. E. & Rudnicki, M. A. The myogenic regulatory factors, determinants of muscle development, cell identity and regeneration. Semin Cell Dev. Biol. 72, 10–18 (2017).
Article PubMed PubMed Central Google Scholar
Molkentin, J. D. & Olson, E. N. Combinatorial control of muscle development by basic helix-loop-helix and MADS-box transcription factors. Proc. Natl Acad. Sci. USA 93, 9366–9373 (1996).
Article ADS CAS PubMed PubMed Central Google Scholar
Taylor, M. V. & Hughes, S. M. Mef2 and the skeletal muscle differentiation program. Semin Cell Dev. Biol. 72, 33–44 (2017).
Article CAS PubMed Google Scholar
Martin, J. F. et al. A Mef2 gene that generates a muscle-specific isoform via alternative mRNA splicing. Mol. Cell Biol. 14, 1647–1656 (1994).
CAS PubMed PubMed Central Google Scholar
Kim, Y. et al. The MEF2D transcription factor mediates stress-dependent cardiac remodeling in mice. J. Clin. Invest 118, 124–132 (2008).
Article CAS PubMed Google Scholar
Dedieu, S., Mazères, G., Cottin, P. & Brustis, J. J. Involvement of myogenic regulator factors during fusion in the cell line C2C12. Int J. Dev. Biol. 46, 235–241 (2002).
CAS PubMed Google Scholar
Dodou, E., Xu, S. M. & Black, B. L. mef2c is activated directly by myogenic basic helix-loop-helix proteins during skeletal muscle development in vivo. Mech. Dev. 120, 1021–1032 (2003).
Article CAS PubMed Google Scholar
Molkentin, J. D., Black, B. L., Martin, J. F. & Olson, E. N. Cooperative activation of muscle gene expression by MEF2 and myogenic bHLH proteins. Cell 83, 1125–1136 (1995).
Article CAS PubMed Google Scholar
Kaushal, S., Schneider, J. W., Nadal-Ginard, B. & Mahdavi, V. Activation of the myogenic lineage by MEF2A, a factor that induces and cooperates with MyoD. Science 266, 1236–1240 (1994).
Article ADS CAS PubMed Google Scholar
Black, B. L., Molkentin, J. D. & Olson, E. N. Multiple roles for the MyoD basic region in transmission of transcriptional activation signals and interaction with MEF2. Mol. Cell Biol. 18, 69–77 (1998).
Article CAS PubMed PubMed Central Google Scholar
Davis, R. L. & Weintraub, H. Acquisition of myogenic specificity by replacement of three amino acid residues from MyoD into E12. Science 256, 1027–1030 (1992).
Article ADS CAS PubMed Google Scholar
Sebastian, S. et al. Tissue-specific splicing of a ubiquitously expressed transcription factor is essential for muscle differentiation. Genes Dev. 27, 1247–1259 (2013).
Article CAS PubMed PubMed Central Google Scholar
Zhu, B., Ramachandran, B. & Gulick, T. Alternative pre-mRNA splicing governs expression of a conserved acidic transactivation domain in myocyte enhancer factor 2 factors of striated muscle and brain. J. Biol. Chem. 280, 28749–28760 (2005).
Article CAS PubMed Google Scholar
Ellis, J. D. et al. Tissue-specific alternative splicing remodels protein-protein interaction networks. Mol. Cell 46, 884–892 (2012).
Article CAS PubMed Google Scholar
Buljan, M. et al. Tissue-Specific Splicing of Disordered Segments that Embed Binding Motifs Rewires Protein Interaction Networks. Mol. Cell 46, 871–883 (2012).
Article CAS PubMed PubMed Central Google Scholar
Sigler, P. B. Transcriptional activation. Acid blobs and negative noodles. Nature 333, 210–212 (1988).
Article ADS CAS PubMed Google Scholar
Tompa, P. & Fuxreiter, M. Fuzzy complexes: polymorphism and structural disorder in protein-protein interactions. Trends Biochem Sci. 33, 2–8 (2008).
Article CAS PubMed Google Scholar
Graham, T. A., Ferkey, D. M., Mao, F., Kimelman, D. & Xu, W. Tcf4 can specifically recognize beta-catenin using alternative conformations. Nat. Struct. Biol. 8, 1048–1052 (2001).
Article CAS PubMed Google Scholar
Brzovic, P. S. et al. The acidic transcription activator Gcn4 binds the mediator subunit Gal11/Med15 using a simple protein interface forming a fuzzy complex. Mol. Cell 44, 942–953 (2011).
Article CAS PubMed PubMed Central Google Scholar
Hatos A., Monzon A. M., Tosatto S. C. E., Piovesan D., Fuxreiter M. FuzDB: a new phase in understanding fuzzy interactions. Nucleic Acids Res 50, D509–D517 (2021).
Article PubMed Central Google Scholar
Fuxreiter, M. et al. Malleable machines take shape in eukaryotic transcriptional regulation. Nat. Chem. Biol. 4, 728–737 (2008).
Article CAS PubMed PubMed Central Google Scholar
Chong, S. et al. Imaging dynamic and selective low-complexity domain interactions that control gene transcription. Science 361, eaar2555 (2018).
Article PubMed PubMed Central Google Scholar
Sabari, B. R. et al. Coactivator condensation at super-enhancers links phase separation and gene control. Science 361, eaar3958 (2018).
Article PubMed PubMed Central Google Scholar
Wu, H. & Fuxreiter, M. The Structure and Dynamics of Higher-Order Assemblies: Amyloids, Signalosomes, and Granules. Cell 165, 1055–1066 (2016).
Article CAS PubMed PubMed Central Google Scholar
Miskei, M., Horvath, A., Vendruscolo, M. & Fuxreiter, M. Sequence-Based Prediction of Fuzzy Protein Interactions. J. Mol. Biol. 432, 2289–2303 (2020).
Article CAS PubMed Google Scholar
Hardenberg, M., Horvath, A., Ambrus, V., Fuxreiter, M. & Vendruscolo, M. Widespread occurrence of the droplet state of proteins in the human proteome. Proc. Natl Acad. Sci. USA 117, 33254–33262 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Sormanni, P. et al. Simultaneous quantification of protein order and disorder. Nat. Chem. Biol. 13, 339–342 (2017).
Article CAS PubMed Google Scholar
Fuxreiter, M. Fuzziness in Protein Interactions-A Historical Perspective. J. Mol. Biol. 430, 2278–2287 (2018).
Article CAS PubMed Google Scholar
Gianni, S. et al. Fuzziness and Frustration in the Energy Landscape of Protein Folding, Function, and Assembly. Acc. Chem. Res 54, 1251–1259 (2021).
Article CAS PubMed PubMed Central Google Scholar
Walsh, I., Martin, A. J., Di Domenico, T. & Tosatto, S. C. ESpritz: accurate and fast prediction of protein disorder. Bioinformatics 28, 503–509 (2012).
Article CAS PubMed Google Scholar
Barrera-Vilarmau, S., Teixeira, J. M. C. & Fuxreiter, M. Protein interactions: anything new? Essays Biochem 66, 821–830 (2022).
Article CAS PubMed PubMed Central Google Scholar
Choudhary, C. et al. Lysine acetylation targets protein complexes and co-regulates major cellular functions. Science 325, 834–840 (2009).
Article ADS CAS PubMed Google Scholar
Horvath, A., Vendruscolo, M. & Fuxreiter, M. Sequence-based Prediction of the Cellular Toxicity Associated with Amyloid Aggregation within Protein Condensates. Biochemistry 61, 2461–2469 (2022).
Article CAS PubMed Google Scholar
Vendruscolo, M. & Fuxreiter, M. Sequence determinants of the aggregation of proteins within condensates generated by liquid-liquid phase separation. J. Mol. Biol. 434, 167201 (2021).
Article PubMed Google Scholar
Vendruscolo, M. & Fuxreiter, M. Protein condensation diseases: therapeutic opportunities. Nat. Commun. 13, 5550 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Ran, F. A. et al. Genome engineering using the CRISPR-Cas9 system. Nat. Protoc. 8, 2281–2308 (2013).
Article CAS PubMed PubMed Central Google Scholar
Asher, G., Reuven, N. & Shaul, Y. 20S proteasomes and protein degradation “by default”. Bioessays 28, 844–849 (2006).
Article CAS PubMed Google Scholar
Zhu, L. & Brangwynne, C. P. Nuclear bodies: the emerging biophysics of nucleoplasmic phases. Curr. Opin. Cell Biol. 34, 23–30 (2015).
Article CAS PubMed PubMed Central Google Scholar
Cilia, E., Pancsa, R., Tompa, P., Lenaerts, T. & Vranken, W. F. From protein sequence to dynamics and disorder with DynaMine. Nat. Commun. 4, 2741 (2013).
Article ADS PubMed Google Scholar
Fuxreiter, M. & Vendruscolo, M. Generic nature of the condensed states of proteins. Nat. cell Biol. 23, 587–594 (2021).
Article CAS PubMed Google Scholar
Whyte, W. A. et al. Master transcription factors and mediator establish super-enhancers at key cell identity genes. Cell 153, 307–319 (2013).
Article CAS PubMed PubMed Central Google Scholar
Hnisz, D. et al. Super-enhancers in the control of cell identity and disease. Cell 155, 934–947 (2013).
Article CAS PubMed Google Scholar
Cho, W. K. et al. Mediator and RNA polymerase II clusters associate in transcription-dependent condensates. Science 361, 412–415 (2018).
Article ADS CAS PubMed PubMed Central Google Scholar
Li, W. et al. Biophysical properties of AKAP95 protein condensates regulate splicing and tumorigenesis. Nat. cell Biol. 22, 960–972 (2020).
Article CAS PubMed PubMed Central Google Scholar
Shen, C. et al. Phase separation drives RNA virus-induced activation of the NLRP6 inflammasome. Cell 184, 5759–5774.e20 (2021).
Article CAS PubMed PubMed Central Google Scholar
Sandmann, T. et al. A temporal map of transcription factor activity: mef2 directly regulates target genes at all stages of muscle development. Dev. cell 10, 797–807 (2006).
Article CAS PubMed Google Scholar
Sponga, A. et al. Order from disorder in the sarcomere: FATZ forms a fuzzy but tight complex and phase-separated condensates with alpha-actinin. Sci. Adv. 7, eabg7653 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
D. A. Case DSC et al. AMBER 2017 (University of California, San Francisco 2017).
Jorgensen, W. L., Chandrasekhar, J., Madura, J. D., Impey, R. W. & Klein, M. L. COMPARISON OF SIMPLE POTENTIAL FUNCTIONS FOR SIMULATING LIQUID WATER. J. Chem. Phys. 79, 926–935 (1983).
Article ADS CAS Google Scholar
Darden, T., York, D. & Pedersen, L. PARTICLE MESH EWALD - AN N.LOG(N) METHOD FOR EWALD SUMS IN LARGE SYSTEMS. J. Chem. Phys. 98, 10089–10092 (1993).
Article ADS CAS Google Scholar
Huang, J. et al. CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nat. Methods 14, 71–73 (2017).
Article CAS PubMed Google Scholar
Bernado, P. & Blackledge, M. A self-consistent description of the conformational behavior of chemically denatured proteins from NMR and small angle scattering. Biophys. J. 97, 2839–2845 (2009).
Article ADS CAS PubMed PubMed Central Google Scholar
Berjanskii, M. V. & Wishart, D. S. A simple method to predict protein flexibility using secondary chemical shifts. J. Am. Chem. Soc. 127, 14970–14971 (2005).
Article CAS PubMed Google Scholar
Kaderavek, P. et al. Spectral density mapping protocols for analysis of molecular motions in disordered proteins. J. Biomol. NMR 58, 193–207 (2014).
Article CAS PubMed Google Scholar
Wu, D. H., Chen, A. D. & Johnson, C. S. An Improved Diffusion-Ordered Spectroscopy Experiment Incorporating Bipolar-Gradient Pulses. J. Magn. Reson. 115, 260–264 (1995).
Article ADS CAS Google Scholar
Heinz, S. et al. Simple combinations of lineage-determining transcription factors prime cis-regulatory elements required for macrophage and B cell identities. Mol. Cell 38, 576–589 (2010).
Article CAS PubMed PubMed Central Google Scholar
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Hatos, A., Tosatto, S. C. E., Vendruscolo, M. & Fuxreiter, M. FuzDrop on AlphaFold: visualizing the sequence-dependent propensity of liquid-liquid phase separation and aggregation of proteins. Nucleic Acids Res 50, W337–344 (2022).
Article PubMed PubMed Central Google Scholar

Download references

Acknowledgements

This research was supported by AIRC IG 26229 (M.F.), AFM-Telethon project 23542 (S.C.) MDA project 952598 (S.C.), National Research, Development and Innovation Office of Hungary grants NKFI/ OTKA NN 128368 (K.E.K.), 2019-1.1.1-PIACI-KFI-2019-00109 (J.T., V.A.), Thematic Excellence Program grant TKP2021-NKTA-34 (B.E.), Hungarian Academy of Sciences HAS-11015 (M.F.), European Regional Development Fund projects GINOP-2.3.3-15-2016-00004 and GINOP-2.3.2–15-2016–00044), NKFI OTKA: K 137600 (L.C.) and K 132814 (B.E.). L.C. acknowledges the support by UD Scientific Excellence. S.C. acknowledges Dr. Jonathan Vinet and the microscopy facility CIGS for technical support.

Author information

Authors and Affiliations

Department of Physiology, Faculty of Medicine, University of Debrecen, Egyetem tér 1, H-4032, Debrecen, Hungary
Mónika Gönczi, Zsolt Ráduly & László Csernoch
ELKH Cell Physiology Research Group, Department of Physiology, Faculty of Medicine, University of Debrecen, Egyetem tér 1, H-4032, Debrecen, Hungary
Mónika Gönczi, Zsolt Ráduly & László Csernoch
Department of Biomedical Sciences, University of Padova, Via Ugo Bassi 58/B, 35131, Padova, Italy
João M. C. Teixeira, Susana Barrera-Vilarmau & Monika Fuxreiter
Department of Biomedical, Metabolic and Neural Sciences, University of Modena and Reggio Emilia G. Campi 287, 41125, Modena, Italy
Laura Mediani, Francesco Antoniani & Serena Carra
Department of Inorganic and Analytical Chemistry, University of Debrecen, Egyetem tér 1, H-4032, Debrecen, Hungary
Tamás Milán Nagy, Krisztina Fehér & Katalin E. Kövér
MTA-DE Molecular Recognition and Interaction Research Group, University of Debrecen Egyetem tér 1, H-4032, Debrecen, Hungary
Tamás Milán Nagy, Krisztina Fehér & Katalin E. Kövér
Department of Biochemistry and Molecular Biology, Faculty of Medicine, University of Debrecen, Egyetem tér 1, H-4032, Debrecen, Hungary
Viktor Ambrus, József Tőzsér & Endre Barta
Department of Genetics and Genomics, Institute of Genetics and Biotechnology, Hungarian University of Agriculture and Life Sciences, Szent-Györgyi A. út 4, H-2100, Gödöllő, Hungary
Endre Barta

Authors

Mónika Gönczi
View author publications
You can also search for this author in PubMed Google Scholar
João M. C. Teixeira
View author publications
You can also search for this author in PubMed Google Scholar
Susana Barrera-Vilarmau
View author publications
You can also search for this author in PubMed Google Scholar
Laura Mediani
View author publications
You can also search for this author in PubMed Google Scholar
Francesco Antoniani
View author publications
You can also search for this author in PubMed Google Scholar
Tamás Milán Nagy
View author publications
You can also search for this author in PubMed Google Scholar
Krisztina Fehér
View author publications
You can also search for this author in PubMed Google Scholar
Zsolt Ráduly
View author publications
You can also search for this author in PubMed Google Scholar
Viktor Ambrus
View author publications
You can also search for this author in PubMed Google Scholar
József Tőzsér
View author publications
You can also search for this author in PubMed Google Scholar
Endre Barta
View author publications
You can also search for this author in PubMed Google Scholar
Katalin E. Kövér
View author publications
You can also search for this author in PubMed Google Scholar
László Csernoch
View author publications
You can also search for this author in PubMed Google Scholar
Serena Carra
View author publications
You can also search for this author in PubMed Google Scholar
Monika Fuxreiter
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

Conceptualisation: M.F., Experiment design: S.C., M.G., K.E.K.; Data collection: M.G., L.M., F.A., Z.R., K.F., T.M.N., V.A., E.B., J.T.; Analysis of the results: M.G., J.M.C.T., S.B.V., K.F., E.B., L.C., S.C., M.F.; Visualisation: M.G., J.M.C.T., S.B.V., T.M.N., V.A., E.B., S.C.; Supervision: S.C., K.E.K., L.C., J.T., M.F.; Writing the original manuscript: M.F., M.G., K.E.K; Revising the manuscript: S.C., M.F.

Corresponding authors

Correspondence to Serena Carra or Monika Fuxreiter.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

: Nature Communications thanks the other anonymous reviewer(s) for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Reporting Summary

Source data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Gönczi, M., Teixeira, J.M.C., Barrera-Vilarmau, S. et al. Alternatively spliced exon regulates context-dependent MEF2D higher-order assembly during myogenesis. Nat Commun 14, 1329 (2023). https://doi.org/10.1038/s41467-023-37017-7

Download citation

Received: 10 March 2022
Accepted: 24 February 2023
Published: 10 March 2023
DOI: https://doi.org/10.1038/s41467-023-37017-7

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Subjects

Abstract

Similar content being viewed by others

Human skeletal muscle aging atlas

Multimodal cell atlas of the ageing human skeletal muscle

Nuclear mRNA decay: regulatory networks that control gene expression

Introduction

Results

Designing Mef2D variants with altered β-domain dynamics and higher-order assembly

β-domain rigidity potentiates transcription activity

β-domain rigidity promotes early myotube development

Mef2D forms higher-order assemblies in C2C12 cells with both liquid-like and solid-like properties

Structure and dynamics of Mef2D model peptides

Discussion

Methods

Computational design

Cell culturing and differentiation

CRISPR-Cas9 experiments

Determination of transcription activity, reporter plasmids

Immunocytochemistry and calculation of fusion index

Immunofluorescence on cultured cells

Western blot analysis

Confocal microscopy and Fluorescence recovery after photobleaching (FRAP) analysis

Graphics and statistics

Molecular Dynamics (MD) simulations

Model system

Computational protocol

MD analysis

NMR analysis

NMR resonance assignment

Determination of order parameters (S2) from chemical shifts

Relaxation measurements

Diffusion ordered spectroscopy

ChIP-Seq Protocol

ChIP-seq data analysis

Reporting summary

Data availability

Code availability

References

Acknowledgements

Author information

Authors and Affiliations

Contributions

Corresponding authors

Ethics declarations

Competing interests

Peer review

Peer review information

Additional information

Supplementary information

Supplementary information

Reporting Summary

Source data

Source data

Rights and permissions

About this article

Cite this article

Share this article

Comments

Search

Quick links

Determination of order parameters (S²) from chemical shifts