Spatial expression of Myocardin protein in normal and disease states

Abstract

Vascular smooth muscle cell (VSMC) dedifferentiation, a phenomenon found in virtually all vascular diseases, is characterized by a transcriptional switch from a contractile to a phenotypically modulated state. Myocardin (MYOCD) is a smooth muscle cell-restricted co-activator that is necessary and sufficient for the differentiation of VSMC through the transcriptional activation of SMC-restricted cytoskeletal and contractile genes. Despite 25 years of research on MYOCD, the reliable expression of this protein continues to be poorly represented and understood. Accordingly, we generated a novel rat model carrying HA-tagged MYOCD to address pervasive disparities in the literature and elucidate MYOCD protein expression in vivo. Western blotting studies documented highest MYOCD protein expression in aorta, bladder and uterus with low or undetectable expression in all other tissues of the rat, including heart. Predictive modeling supports the C-terminus of MYOCD to be most immunoreactive where the HA tag and one other commercial immunogen reside. Of note, the latter immunogen was used to generate what appears to be the most trustworthy commercial antibody against MYOCD. In vitro transcription/translation and phosphatase treatment of ectopic and endogenous MYOCD protein reveal an intrinsically high molecular weight of MYOCD that has eluded prior reports. Importantly, we present the very first spatial expression profile of MYOCD protein in several mouse and rat tissues under baseline and vascular injury conditions. The results offer the SMC community new resources and insight into the reliable detection of MYOCD protein.

Graphical Abstract

New & Noteworthy:

Endogenous MYOCD protein has eluded reliable detection due to pervasive commercial antibody failures, likely due to protein disorder. A CRISPR-generated 3×HA knock-in rat enabled the first in vivo spatial characterization of MYOCD, revealing its presence in SMC-rich tissues and, notably, within the microvasculature of SMC-poor organs. Phosphorylation largely accounts for the high molecular weight of MYOCD. Strikingly, MYOCD is detected in neointimal cells after vascular insult, suggesting persistence or re-expression during vascular remodeling.

Introduction

Vascular smooth muscle cells (VSMCs) constitute the predominant cell type within conducting and muscular blood vessels, providing support through the organization of an intricate extracellular matrix and, in many instances, contractile force to coordinate normal blood flow. Under vascular disease conditions, VSMCs undergo phenotypic modulation to a transitional state, which serves as an intermediate for several subsequent VSMC subtypes^[1–3]. Notably, this process of dedifferentiation is not universally detrimental to vascular health; for example, SMC phenotypic modulation contributes to the formation of the protective fibrous cap in atherosclerosis^[4–6]. Similarly, adaptive remodeling of the vessel wall in response to altered flow and hemodynamic stress is beneficial during the onset of vascular disease, and loss of this adaptability has been linked to increased susceptibility to vessel rupture and dissection^[7].

Myocardin (MYOCD) was initially identified as a cofactor of serum response factor (SRF), and together the SRF-MYOCD complex bound CArG elements to powerfully activate cardiac contractile genes ^[8]. Subsequent reports found SRF-MYOCD to be a critical molecular switch for VSMC differentiation and the acquisition of a quiescent, contractile cell state^[9–16]. The current paradigm posits that low MYOCD levels set in motion the initial switching of contractile SMCs to functionally diverse VSMC subclasses, positioning MYOCD as a critical nodal point for the phenotypic state of VSMCs. While the expression profile of Myocd mRNA is well known, MYOCD protein expression in vivo has eluded rigorous analysis, leaving several pertinent questions unresolved.

Indeed, understanding MYOCD protein expression has been a persistent challenge as commercially available antibodies do not reliably detect endogenous MYOCD. For example, in vitro studies have consistently reported MYOCD at an apparent molecular weight of ~100 kDa, corresponding to its predicted size based on amino acid sequence ^[17]. However, the authentic MYOCD molecular weight of protein in mice migrates at 130 - 140 kDa^[18]. The basis for this discrepancy remains unsettled and has been largely overlooked in the MYOCD literature. Given this enduring problem, we developed a new endogenously tagged rat MYOCD model to extend previous work in the mouse^[18] and begin evaluating MYOCD spatial expression in tissues. The latter is significant as no study has yet to report on the spatial expression profile of MYOCD protein. Here, we provide strong support of a higher-than-predicted molecular weight of MYOCD protein in the rat and some understanding as to the basis for this unexpected finding. We also offer some insight into why so many antibodies may have failed. Finally, we present the first extensive spatial profile of MYOCD protein across adult mouse and rat tissue types, in embryonic heart, and under disease modeling conditions.

Methods

Generation of Myocd^3xHA MYOCD rat model

In spring 2021, a rat line carrying a C-terminal 3xHA tag at the endogenous Myocd locus was generated. All design and microinjection steps were performed by Cyagen Biosciences (Suzhou, Jiangsu, China). Briefly, fertilized rat zygotes underwent three-component CRISPR ^[19] with Cas9 protein, a sgRNA targeting the Myocd locus, and a single-strand donor DNA containing three copies of the HA epitope (Table 1). Founder pups were identified by PCR and confirmed by Sanger sequencing. Two correctly targeted founders were transferred to Augusta University for colony establishment. Each founder was backcrossed once to outbred, wild-type Sprague–Dawley (SD) rats, and litters were genotyped by PCR and sequencing. To minimize the risk of deleterious phenotypes arising from inbreeding, no line was intercrossed beyond two generations before back-crossing to fresh breeder rats. For experimental studies, animals were randomized into two groups by genotype. Male and female rats were studied throughout. Experiments were carried out without blinding. The 3xHA-tagged Myocd (Myocd^3xHA) strain has been cryopreserved and is available upon request from the Rat Resource & Research Center (RRRC; Columbia, Missouri; https://www.rrrc.us). All animal procedures were approved by the Institutional Animal Care and Use Committee at Augusta University (protocol 2019-1000).

Table 1: Materials used.

Table of CRISPR components, oligos, and antibodies used

	Sequence			Usage
sgRNA 1	GTGTTTACCACTGCTGTAAG			Myocd3xHA Rat Model
Donor Oligo (3xHA in bold)	CAACCTCACCTCCAGCGGGCCCAGCATTTTCAACATCG ATTTTCTGGATGTCACGGATCTCAATTTGAATTCCCCTA TGGACCTCCACTTACAGCAGTGGGGCTCCTACCCATA CGATGTTCCTGACTATGCGGGCTATCCCTATGACGTC CCGGACTATGCAGGATCCTATCCATATGACGTTCCAG ATTACGCTTAAACACCTGAGGTACAAGTGTTATGAGAG CTCAGTAGGAATTCCATGGGGGAGAGCACACAACCAG AAATGTGTGATCCAAAAGATAAAGGGGAGAAAA			Myocd3xHA Rat Model
Name	Sequence			Usage
HAMyocd_Rat_primerF	CGGAATCTCCTTGGGAAACA			Genotyping
HAMyocd_Rat_primerR	TTGTCACAGAAACCTCCATCC			Genotyping
Myocd_qPCR_primerF	ATGGATGAACTCCTGGATGT			Fig1B
Myocd_qPCR_primerR	AAGACTTCCAGGTGTTCCTC			Fig1B
HA_qPCR_primerF	CCCATACGATGTTCCTGACTATG			Fig1C
HA_qPCR_primerR	CTCTTGTACCTCGGGTGTTT			Fig1C
Antibody	Company	Lot	Catalog	Usage
anti-HA	CST	Lot 8	3724S	WB: Fig1D–E, Fig2B, Supp. Fig2B, Fig3A–D, Supp. Fig3 IFM: Fig4A–F, Supp. Fig4A-K, Fig5E–F, Supp. Fig5A
anti-ACTA2	Sigma	058M4761V	C6198	IFM: Fig4A–F, Supp. Fig4A-K, Supp. Fig5A-B
anti-MYOCD	Invitrogen	WI3368968A	PA5-95700	WB: Fig2B
anti-MYOCD	Sigma	118M4871V	SAB4200539	WB: Fig2C
anti-MYOCD	Proteintech	157771	51132-1-ap	WB: Supp. Fig2A
anti-ACTB	Affinity	54o2802	T0022	WB loading control: Fig1D, Supp Fig3
anti-HSPA8 (Hsp70)	Proteintech	NA	10995	Supp Fig3
anti-HA-HRP	Sigma	NA	12013819001	Immuno-GoldEM: Supp. Fig4L
anti-GAPDH	Millipore	2036696	07-1580	All WB loading controls, excluding Fig1D
anti-IgG	Santa Cruz	K1017	sc-2-25	Glycosidase positive control

abbreviations: Myocd^3xHA, 3x HA tagged Myocardin; WB, Western blot; EM, electron microscopy; IFM, immunofluorescence microscopy; NA, not available.

Quantitative RT-PCR

Snap-frozen tissue samples were homogenized using Takara BioMashers. The resulting lysate was further disrupted in QIAzol Lysis Reagent (QIAGEN, 79306) at room temperature. Total RNA was extracted using the miRNeasy Mini Kit (QIAGEN, 217004) per the manufacturer’s instructions. One microgram of RNA was reverse-transcribed into complementary DNA (cDNA) using the iScript^™ cDNA Synthesis Kit (Bio-Rad, 1708890). For quantitative real-time PCR (qPCR), cDNA was mixed with SsoAdvanced^™ Universal SYBR^® Green Supermix (Bio-Rad, 1725271), gene-specific primers (Table 1), and nuclease-free water. Reactions were run on a CFX Connect Real-Time PCR Detection System (Bio-Rad). Ct values were exported to Excel, normalized to the housekeeping gene Gapdh to calculate ΔΔCt, and plotted in GraphPad Prism Version 10.

Western blot

Rats were anesthetized and perfused via the left ventricle with 0.9% NaCl to remove blood. The aorta was excised, rinsed briefly in PBS, then transferred to ice-cold PBS. Adventitial fat was removed under a stereomicroscope using fine forceps, and the medial smooth muscle layer was peeled away from the adventitia. Samples were snap-frozen in liquid nitrogen and stored at −80 °C. For protein extraction, frozen tissues were crushed using Takara BioMashers (9790A). The tissues were then further homogenized in fresh RIPA buffer. Homogenates were cleared by centrifugation at 14,000 × g for 10 minutes at 4 °C, and supernatants were collected. HEK293 cells were washed twice with ice-cold PBS to remove residual media and trypsin, then lysed in ice-cold RIPA buffer (Sigma, R0278) with protease inhibitors (Sigma, 04693159001) by vigorous vortexing and pipetting. Cell lysates were cleared as above. Protein concentrations were determined by BCA assay (Bio-Rad, 5000112). Aliquots (20 - 40 μg) were mixed with Laemmli sample buffer and heated at 95 °C for 10 min. Proteins were resolved in 4–20% SDS–PAGE gradient gels (Bio-Rad, 5678094) in 1× Tris–Glycine–SDS running buffer (Bio Basic, A0030) for 2.5 h at constant voltage. Gels were imaged for total protein before transfer to PVDF membranes (Bio-Rad, 1620177) using the Trans-Blot Turbo system (Bio-Rad, 1704273). Membranes were blocked for 15 minutes in blocking buffer (Bio-Rad, 12010020), then incubated overnight at 4 °C with primary antibodies (Table 1) at a dilution of 1:1000. After three 5 minutes washes in PBST (PBS + 0.1 % Tween-20), membranes were incubated for 1 h at room temperature with HRP-conjugated secondary antibodies (Table 1) at a dilution of 1:5000, washed again, and developed with WesternBright Sirius HRP substrate (Advansta, K-12043D10). Blots were imaged on a ChemiDoc system (Bio-Rad, 120031). Critically, the banding location of the Myocd^3xHA band is dependent on the percentage of polyacrylamide. Both 4-20% gels and 7% gels were utilized throughout this study. Gel percentages used are indicated in figure legends. The anti-HA antibody used here (anti-HA, CST (3724S)) is stated by the provider to produce a nonspecific band at 100 kDa.

Antibody epitope alignment and AIUPred prediction

MYOCD antibody epitope sequences were obtained from vendor websites and aligned to the primary amino acid sequence of rat MYOCD using the ggMSA package in RStudio. The resulting alignment files were exported as graphical outputs and imported into Adobe Illustrator for integration with annotated MYOCD functional domains^{[12, 20]}. Intrinsic disorder and binding regions were predicted using the AIUPred web interface and the ANCHOR2 associated software^[21]. The full-length amino acid sequence of the longest MYOCD isoform was submitted to generate both disorder/binding propensities. These outputs were then overlaid with the epitope alignments and functional domain map in Illustrator for final figure assembly.

Cell culture and plasmid transfection

HEK293 cells were maintained in 60 mm dishes in Dulbecco’s Modified Eagle Medium (DMEM; Gibco, 11965-092) supplemented with 10% fetal bovine serum (FBS; Thermo, 10082147). Culture medium was replaced every 48 h, and cells were passaged with trypsin–EDTA (Worthington Biochemical, LS003570) once they reached ~80 % confluency. Cell lines were not assessed for myoplasm or other contaminants. For plasmid transfection, cells at ~80 % confluency were prepared in antibiotic-free, DMEM containing 10% serum and transfected with 250 ng of plasmid DNA using PolyJet^™ In Vitro DNA Transfection Reagent (Sigma, SL100688) following the manufacturer’s protocol. After 24 h incubation at 37 °C, the transfection medium was replaced with complete media and cells were harvested for protein analysis. Plasmids used were v1, v2, and v3, corresponding to different isoforms of Myocardin^[22]. The length of each isoform is as follows: v1, 983; v2, 935; v3, 904 amino acids. Plasmids are available on request.

Post-translational modification and in vitro transcription/translation

To assess post-translational modifications (PTMs), 1 μg of total protein (tissue, HEK293, or in vitro translated lysate) was incubated overnight with either PNGase F (NEB, P0704S) for N-glycan removal or Antarctic Phosphatase (NEB, M0289S) or Calf Intestinal Alkaline Phosphatase (Thermo Scientific 18009019) for dephosphorylation at 37°C. Treated samples were then processed for Western blotting as described above. In vitro–transcription/translation lysates were generated using the TNT T7 Coupled Wheat Germ Extract System (Promega, L4140) according to the supplier’s protocol.

Myocd^fl8/fl8 mouse model

The exon 8 floxed Myocd mouse model^[23] was graciously provided as cryopreserved embryos by Dr. Michael Parmacek (University of Pennsylvania) and a colony re-established here at Augusta University. Animals were crossed with Itga8-CreER^T2 mice ^[24] and injected with 50 mg tamoxifen/kg body weight (Millipore, T5648-5G) dissolved in sunflower oil (Sigma, S5007) for five days. After a 10-day washout, aorta and liver were isolated for protein extraction. Lysates were then used in Western blots to test MYOCD antibody efficacy.

Immunoprecipitation and Mass Spectrometry

After procuring aortae, the medial SMC layers were isolated. Endothelial layers were removed by brushing the inside of the aorta and adventitial layers were removed physically by peeling the medial layer off using fine forceps. Aortic SMC nuclei were isolated using the Chromium nuclei isolation kit (10x Genomics, Cat#PN-1000493) as per the manufacturer’s protocol. Medial aortic lysate was prepared by homogenizing isolated nuclei in Co-IP lysis buffer (50 mM Tris-HCl pH 8.0, 150 mM NaCl, 1mM EDTA, 1% triton x-100, 5 μL/mL DTT, protease inhibitors) or IP lysis buffer (50 mM Tris-HCl pH 7.4, 150 mM NaCl, 1 mM EDTA, 1% NP-40, 0.05% SDS, protease inhibitors) with a glass Dounce homogenizer on ice (≤ 10 min). A probe sonicator was used with two pulses at 5 seconds each on ice to rupture any remaining membranes when using the Co-IP buffer. Lysates were cleared by centrifugation at maximum speed for 10 minutes at 4 °C. A 1% aliquot of the supernatant was reserved as input. The remaining lysate was precleared using A/G conjugated beads (Pierce, 88802) for two hours at 4 °C. Lysates were then incubated overnight at 4 °C, end-over-end, with Pierce^™ Anti-HA Magnetic Beads (Pierce, 88836). Beads were collected magnetically and washed three times each in lysis buffer + protease inhibitors and PBS plus protease inhibitors (5 minutes per wash at 4 °C). Bound proteins were eluted by boiling in Laemmli buffer for 10 min, beads were removed, and eluates were analyzed by Western blotting as above. This was done for HA and confirmatory Co-IP blots. Samples were also sent to the Emory Proteomics Core for Mass Spectrometry analysis. Results were analyzed in Rstudio with peptide counts for each protein averaged across the three wild type replicates and the three Myocd HA-tagged replicates to obtain WT_avg and HA_avg. Enrichment was quantified using a stabilized log₂ fold-change: log ₂[(HA_avg + 0.5)/(WT_avg + 0.5)]. To weight the score by peptide support, a total abundance term (HA_avg + WT_avg) was multiplied by the log₂ fold-change to generate a final enrichment score. Proteins were ranked by this score, and low-support proteins were filtered out, except for MYOCD, which was retained and shown with its true rank.

Histopathology and immunofluorescence microscopy

Rat tissues were first immersed in 4% paraformaldehyde at 4 °C for 24–48 hours, then processed and embedded in paraffin for histological examination. Blocks were sectioned at 5 μm on a Microm HM 355S microtome and incubated at 58 °C overnight to secure the sections onto slides. After deparaffinization, slides were stained with hematoxylin and eosin using a Leica Autostainer XL (ST5010, Leica Biosystems) following the manufacturer’s instructions, and images were captured on an Olympus Bx43 microscope. For immunofluorescence, deparaffinized sections underwent heat-mediated antigen retrieval in Dako Target Retrieval Solution (Agilent Technologies, S236784-2). Slides were then incubated overnight at 4 °C with primary antibodies at a dilution of 1:200 (Table 1). After washing, fluorescent secondary antibodies were applied for 30 min at a dilution of 1:200. Finally, coverslips were mounted with ProLong Glass antifade mounting medium containing NucBlue (ThermoFisher, Cat. #P36981) to stain nuclei. Fluorescence images were acquired on a Zeiss LSM 900 confocal microscope using ZEN BLUE software. HA-positive nuclei were manually quantified using the Cell Counter plugin in Fiji (ImageJ). Regions positive for ACTA2 immunofluorescence were first identified, and only nuclei within these ACTA2-positive areas were considered for analysis. HA-positive spots within these regions were counted manually, and the resulting counts were normalized to the total number of DAPI-stained nuclei within the same ACTA2-positive regions to determine the proportion of MYOCD-HA-positive cells.

Immunogold electron microscopy

Immunogold electron microscopy was performed as in a previous publication^[24]. Briefly, homozygous (Myocd^HA) mice were perfused with heparinized PBS followed by fixative comprising 4% paraformaldehyde/0.2% glutaraldehyde in 0.1 M sodium cacodylate (NaCac) buffer, pH 7.4. The aorta was isolated, immersion fixed in the same 4% paraformaldehyde/0.2% glutaraldehyde and dehydrated through a graded series of ethanol (25%, 50%, 70%, 80%, 95%). Samples were then given to the Augusta University Electron Microscopy Core and embedded in LR White Resin (Electron Microscopy Sciences, Ft Washington, PA). Semi-thin sections of 750 nm were cut and stained with Toluidine Blue to map areas of interest. Thin sections (50 nm) were cut with a diamond knife (Diatome/Electron Microscopy Sciences) on a Leica EM UC7 ultramicrotome (Leica Microsystems, Inc, Bannockburn, IL) and collected on Pioloform coated nickel slot grids. Grids, section-side down, were floated on drops of etching solution (5% Sodium Metaperiodate) in PBS for 5 minutes, then washed with PBS. Aldehydes were quenched 30 minutes with 1M ammonium chloride in PBS. Grids were then blocked in Aurion Blocking Solution (Electron Microscopy Sciences, Ft. Washington, PA), in PBS for 2-4 hours at RT. Grids were floated on drops of anti-HA antibody (Sigma, 12013819001), diluted 1:100 in Aurion BSA-C incubation solution (Electron Microscopy Sciences) overnight at 4 degrees and washed in PBS. Grids were then floated on drops of anti-primary species-specific Aurion Ultrasmall gold reagent (Electron Microscopy Sciences) diluted 1:200 in Aurion BSA-C for 2 hours at RT, then washed in PBS, followed by deionized water. Ultrasmall particles were silver enhanced for 19 minutes using HQ Silver^™ (Nanoprobes, Yaphank, NY), then washed in ice cold deionized water to halt enhancement. Grids were then stained with 4% uranyl acetate and lead citrate to increase contrast. Sections were observed in a JEOL 1400 Flash transmission electron microscope (JEOL USA Inc., Peabody, MA) at 120 kV and imaged with One View Digital Camera Controller (Gatan Inc., Pleasanton, CA).

Balloon injury of rat common carotid artery

Rats aged 14–16 weeks were used for carotid artery balloon injury experiments as described^[25]. Rats weighing below 350 grams were excluded from balloon injury studies. Significant dropout was observed when animals below these cutoffs were utilized. Each rat was anesthetized with isoflurane using a Kent SomnoSuite^® system, calibrated for the animal’s weight and set to a flow rate of 3-4%. Adequate depth of anesthesia was confirmed by the absence of a toe-pinch reflex. Analgesia was provided via subcutaneous injection of buprenorphine (0.03 mg/mL, 0.05 mg/kg), and ophthalmic ointment was applied to both eyes. The rat was placed on a heating pad to maintain core body temperature throughout the procedure. The surgical site was prepared by removing hair with Nair, then disinfecting with povidone iodine followed by 70% ethanol. A 1.5 cm longitudinal incision was made over the left carotid region using a scalpel. Blunt dissection of the overlying muscle layers exposed the common carotid artery (CCA) and its bifurcation into the internal (ICA) and external carotid arteries (ECA). During this step, care was taken to avoid injury to the vagus nerve. Permanent ligatures were placed on the ECA about 5 mm distal to the bifurcation, leaving sufficient length for the ensuing arteriotomy; the occipital branch of the ECA was also ligated. Vascular clamps were then applied to the CCA and ICA to isolate blood flow. Lidocaine (Vedco, 50989) was applied topically at the bifurcation to induce vasodilation. An arteriotomy was made in the ligated segment of the ECA, and a 2F Fogarty balloon catheter (Edward LifeSciences, 120602FP) was inserted and advanced into the CCA. The balloon was inflated to approximately 1.5 atm, distending the artery by about 1.5-fold, then gently withdrawn and rotated three times before removal. A permanent ligature was tied proximal to the arteriotomy and clamps were released to re-establish circulation. The incision was closed with skin sutures (esutures, J494G), the wound was swabbed again with povidone iodine, and the area was irrigated with sterile saline. Each rat was then housed individually under a heat lamp or on a heating pad until full recovery from anesthesia, with close postoperative monitoring. Carotid arteries were harvested two weeks after surgery and either fixed or flash-frozen for downstream analyses.

Ligation injury of mouse carotid artery

To further assess changes in MYOCD expression following acute vascular injury, 10 week old homozygous Myocd^HA mice^[16] underwent complete ligation of the left carotid artery, as previously described^[26]. Following injury, mice were maintained for four weeks before euthanasia. The injured left carotid artery and the uninjured right carotid artery (used as internal control), were then harvested, fixed in 4% paraformaldehyde and subjected to histological analysis.

Statistics

Normality of the data was assessed using the Shapiro–Wilk test. Depending on the comparison, paired or unpaired one- or two-tailed t-tests were applied to evaluate differences between experimental and control groups. For analyses involving multiple groups, one-way or two-way ANOVA followed by Tukey’s post hoc test was conducted. No formal power analysis was performed; sample sizes were determined based on prior experience with similar experimental models and established conventions in the field. All statistical analyses were carried out using GraphPad Prism 9 (GraphPad Software). Data are presented as mean ± standard deviation, and statistical significance was defined as p < 0.05.

Results

Generation and characterization of the Myocd^3xHA rat

To assess the true nature of MYOCD protein, we generated a knock-in rat line harboring a 3×HA epitope tag at the 3’ end of the Myocd locus (Myocd^3xHA) to unambiguously define endogenous MYOCD protein expression (Fig. 1A). To confirm the integrity of the targeted locus, Sanger sequencing of genomic DNA at the 3′ end was carried out on homozygous tagged (Myocd^HA) and wild-type (Myocd^WT) rats. The results revealed no nucleotide differences outside the 96bp 3x HA tag insert, verifying faithful integration of the knock-in allele without disruption of the surrounding genomic sequence (Supp. Fig 1). Quantitative RT-PCR using Myocd-specific primers revealed comparable Myocd mRNA in heart, with little detectable expression in liver (Fig. 1B). Of note, the heart was chosen for evaluation as it has a high basal level of Myocd mRNA^{[8, 9, 27, 28]}. We further validated the expression of Myocd mRNA in multiple tissues with HA-specific qPCR primers (Fig. 1C). Myocd mRNA was most abundant in SMC-rich tissues with very low levels in most other tissues, except for heart. As expected, wild type rats showed no product validating the specificity of the HA primers (Fig. 1C).

Figure 1: Schematic and characterization of the Myocd^3xHA rat model.

A, Schematic of the Myocd locus showing insertion of the 3×HA tag at the 3′ end of the coding sequence. B, RT–qPCR analysis showing comparable Myocd mRNA expression in Myocd^HA and wild-type animals. Myocd^WT and homozygous samples are shown in gray and orange, respectively. n=3 independent animals. C, RT–qPCR analysis of Myocd expression across multiple tissues from Myocd^HA rats. n=3 independent animals. D, Western blot and quantitation of tissue lysates from Myocd^HA rat. The black arrow indicates the Myocd^3xHA band migrating just below ~150 kDa with 7% gel used. anti-HA, CST (3724S). n=4 independent animals. E, Western blot and quantitation of aortic lysates from Myocd^WT, Myocd^WT/HA, and Myocd^HA rats. 4-20% gel was used. The black arrow indicates the true Myocd^3xHA band. anti-HA, CST (3724S). n=6 independent animals.

abbreviations: Myocd^3xHA , 3x HA tagged Myocardin; Myocd^WT, wild type; Myocd^WT/HA, heterozygous; Myocd^HA, homozygous (3xHA tag); Ao, aorta; Bl, bladder; Br, brain; Co, colon; He, heart; Ki, kidney; Li, liver; Lu, lung; Sk.M, skeletal muscle; Sm.I, small intestine; St, stomach; Ut, uterus; WT Ao, wild-type aorta;

Leveraging the HA epitope tag, we successfully detected endogenous MYOCD protein in vivo across a panel of rat tissues (Fig. 1D). The observed distribution showed robust protein levels in SMC-rich tissues such as aorta, bladder, and uterus. Notably, the Myocd^3xHA protein migrated just below 150 kDa, substantially larger than the predicted molecular weight derived from the full-length human and rat amino acid sequence (Supp. Fig. 1A). We next compared Myocd^3xHA protein levels in wild-type, heterozygous, and homozygous aortic tissue. This analysis suggested bi-allelic expression of Myocd mRNA (Fig. 1E).

Mapping MYOCD immunogens to functional domains and predicted structure

Examination of publicly available data on MYOCD antibodies from commercial sources indicated a major discrepancy in quality control studies, as several antibodies were tested in cell types or tissues which do not express MYOCD (Table 2). Many such antibodies have been used to report a spurious band of ~100 kDa. To begin to elucidate the basis for such faulty detection of MYOCD protein, we aligned the immunogens of commercially available antibodies with annotated MYOCD protein domains (Fig. 2A, upper). Many immunogens are shared across companies and target similar regions such as the RPEL or basic domains of MYOCD (Fig. 2A, upper). These immunogens were further aligned against predictions of intrinsically disordered regions and protein binding sites generated by AIUPred (Fig. 2A, lower). Strikingly, nearly the entire MYOCD protein is predicted to be intrinsically disordered, with all commercial immunogens mapping partially or entirely within disordered domains, except for one near the end of the C-terminus (Fig. 2A). Of note, this immunogen is closely juxtaposed to the placement of the 3xHA tag (Fig. 2A). Interestingly the antibody raised to this C-terminal immunogen (Table 2) consistently detected endogenous MYOCD at its true molecular weight (Supp. Fig. 2A).

Table 2: MYOCD commercial antibody validation and general characteristics.

WB¹ with A549, HeLa, HepG2, MCF7 cells and brain; IHC² with brain, heart, ovary, spleen, thymus; WB³ of spleen and pancreas; WB⁴ with HCT-116, HEK-293 (overexpressed MYOCD), SW-480 cells; IHC⁵ with rat and chicken heart, IFM with A10 SMC.

Company	Type	Host	Immunogen	Quality control
ABclonal	Polyclonal	Rabbit	Synthetic peptide, hMYOCD	IHC, brain
abinScience	Polyclonal	Rabbit	Synthetic peptide, aa2-139, hMYOCD	30 kDa rhMYOCD
Affinity	Polyclonal	Rabbit	Synthetic peptide, internal hMYOCD	HepG2/A549 cells
Aviva Systems Bio	Polyclonal	Rabbit	Synthetic peptide, aa241-290, mMYOCD	WB, pancreas
Biorbyt	Polyclonal	Rabbit	Synthetic peptide, aa671-770, hMYOCD	IHC, stomach
BosterBio†	Polyclonal	Rabbit	Synthetic peptide, aa200-280, hMYOCD	WB, HeLa cells
CosmoBio	Polyclonal	Rabbit	Synthetic peptide, aa675-704, hMYOCD	IFM, U2OS cells
Creative Diagnostics	Polyclonal	Rabbit	Synthetic peptide, aa250-350, hMYOCD	NA
CUSABIO	Polyclonal	Rabbit	hMYOCD	NA
G-Biosciences	Polyclonal	Rabbit	Custom synthesized peptide	NA
GeneTex‡	Polyclonal	Rabbit	Synthetic peptide, N-term, mMYOCD	WB, pancreas
Invitrogen	Polyclonal	Rabbit	Synthetic peptide, aa221-271, hMYOCD	WB1, IHC2
Leading Biology	Polyclonal	Rabbit	Synthetic peptide, aa743-777, mMYOCD	WB, C2C12 cells
MyBioSource	Polyclonal	Rabbit	Human MYOCD peptide-KLH	WB, brain
Novus	Polyclonal	Rabbit	Recombinant aa34-97, hMYOCD	IFM, BJ cells
Origene	Polyclonal	Rabbit	Synthetic peptide, aa241-290, hMYOCD	NA
ProteoGenix	Polyclonal	Rabbit	Recombinant aa2-139, hMYOCD	NA
Proteintech	Polyclonal	Rabbit	Synthetic peptide, aa918-938, hMYOCD	WB, 3T3 cells
R&D Systems#	Monoclonal	Mouse	Recombinant aa97-290, hMYOCD	WB, MCF-7 cells
RayBiotech	Polyclonal	Rabbit	Synthetic peptide, aa650-750, hMYOCD	WB, heart
RayBiotech±	Polyclonal	Rabbit	Mouse MYOCD, C-terminus	WB, C2C12 cells
Santa Cruz	Monoclonal	Mouse	Epitope at aa391-410, hMYOCD	WB, SW480 cells
Sigma	Polyclonal	Rabbit	Epitope within aa241-290, hMYOCD	WB3
Sigma	Polyclonal	Rabbit	hMYOCD	WB4, IHC5
Sigma	Monoclonal	Mouse	Synthetic peptide, hMYOCD	WB, rhMYOCD

abbreviations: aa, amino acid; hMYOCD, human MYOCD; mMYOCD, mouse MYOCD; IHC, immunohistochemistry; rh, recombinant human MYOCD; WB, Western blot; IFM, immunofluorescence microscopy; NA, not available.

^†Also available at St. John’s Lab

^‡Also available at LSBio

^#Also available at ThermoFisher and MyBioSource

^±Also available at abcepta

Figure 2: MYOCD antibodies do not reliably detect endogenous protein.

A, (upper): MYOCD primary amino acid sequence with alignment of commercial antibody immunogens. Red immunogen (Proteintech) has successfully detected endogenous MYOCD by WB. Functional domains are annotated beneath immunogens with the yellow arrow and rectangle indicating the 3xHA tag. (lower): AIUpred energy-based prediction of (red) disordered regions and (blue) binding affinity. These predictions are aligned to the MYOCD sequence above. B, HEK293 cells transfected with Myocd isoforms (v1, v2, v3) or an empty plasmid and immunoblotted with either anti-MYOCD or anti-HA, n=4 immunoblots testing MYOCD antibodies on independent transfected lysates. anti-MYOCD, Invitrogen PA5-95799; anti-HA, CST (3724S). 7% gel was used. C, Immunoblot using a commercial anti-MYOCD antibody with WT rat lysates and HEK293 cells with (+) or without (−) transfection of the v1 Myocd plasmid, n=3 independent animals where endogenous MYOCD was not recognized by a commericial antibody. The black arrow indicates the true MYOCD band. anti-MYOCD, Sigma SAB4200539. 7% gel was used.

abbreviations: WB, western blot; RPEL, Arg–Pro–Glu–Leu domain; ++, basic domain; Q, glutamine-rich domain; SAP, SAF-A/Acinus/PIAS domain; LZ, leucine-zipper domain; TAD, transactivating domain; WT, wild type; v1, 983 aa; v2, 935 aa; v3, 904 aa; aa, amino acids.

As an additional control, Myocd^HA rat tissue showed MYOCD protein migrating at the same molecular weight as exogenous MYOCD expression in transfected HEK293 cells, with comparable detection seen with a commercial MYOCD antibody. (Fig. 2B). Multiple Myocd isoforms showed consistent detection by immunoblotting using either HA or a commercial antibody. Notably, although a commercial antibody recognized MYOCD in transfected cells expressing supraphysiologic levels of the protein, a separate commercial antibody raised to a nearly identical immunogen failed to detect endogenous MYOCD protein in SMC-rich tissues (Fig. 2C), underscoring its limited efficacy in detecting in vivo MYOCD protein.

Phosphorylation of MYOCD in vivo

To date, analyses of MYOCD post-translational modifications (PTMs) have been restricted to in vitro studies of exogenous protein bearing an epitope tag^[29–35]. Two such studies defined putative phosphorylation sites for MYOCD^{[30, 36]}, which we aligned to several species for conservation in an effort to narrow the window of potential sites of MYOCD modification (Supp. Fig. 1B). Here, we performed a series of in vitro and in vivo studies to begin elucidating the basis for a higher-than-predicted molecular weight of MYOCD. HEK293 lysates overexpressing MYOCD isoforms revealed a reduction in the molecular weight of Myocd^3xHA following phosphatase treatment (Fig. 3A). Consistent with these findings, treatment of in vivo lysates from bladder and aorta with two distinct phosphatases similarly reduced the MW of MYOCD, mirroring our in vitro assay (Fig. 3B). In contrast, digestion of SMC-rich tissue lysates with a glycosidase to remove N-linked glycans did not produce a detectable MW shift relative to sham controls suggesting the absence of glycosylation on MYOCD protein (Fig. 3B & Supp. Fig. 2B). Critically, a conditional knockout of Myocd with minimal MYOCD in aorta, served as a control to rule out spurious bands such as the frequently misidentified 95-100 kDa band (Supp Fig. 3B).

Figure 3: Constitutive phosphorylation of Myocardin in vivo and IP-MS.

A, Western blot of HEK293 cell lysates ± Myocd plasmid of varying isoforms treated with or without Antarctic Phosphatase. n=3 independently treated cell lysates expressing exogenous MYOCD. anti-HA, CST (3724S). 7% gel was used. B, In vivo assessment of MYOCD phosphorylation and glycosylation by Western blot. Bladder and aorta lysates were treated with either PBS (negative), PNGase-F, CIAP, or AnP. IgG was used as a positive control for glycosidase treatment. n=3 independent animals. anti-HA, CST (3724S). 4-20% gel was used. C, In vitro transcription–translation assay. IVT lysates were run alongside HEK293 lysates with and without MYOCD overexpression. Samples were treated ± glycosidase or phosphatase enzymes and immunoblotted with an anti-HA antibody. n=4 independent IVT experiments. anti-HA, CST (3724S). 7% gel was used. D, Immunoprecipitation followed by immunoblot using anti-HA beads on Myocd^WT or Myocd^HA rat aorta protein lysate. The black arrow indicates the true Myocd^3xHA band. n=3 independent animals. anti-HA, CST (3724S). 4-20% gel was used. E, Venn diagram showing the overlapping number of peptides across species and tissue from Myocd^HA-mass spectrometry results. Red arrows denote phosphatase shifted MYOCD^3xHA band, black arrows on the same blot indicate untreated MYOCD^3xHA.

abbreviations: Myocd^3xHA, 3xHA tagged Myocardin; Myocd^WT, wild type; Myocd^HA, homozygous (3xHA tag); v1, 983 aa; v2, 935 aa; v3, 904 aa; aa, amino acids; CIAP, Calf Intestinal Alkaline Phosphatase; AnP, Antarctic Phosphatase; IVT, In vitro transcription-translation; WB, Western blot; IB, immunoblot; IP, immunoprecipitation.

To further assess the contribution of phosphorylation relative to other modifications, we performed in vitro transcription–translation (IVT) of full-length Myocd^3xHA. The IVT product migrated at a position comparable to phosphatase-treated HEK293-derived MYOCD (Fig. 3C), suggesting constitutive phosphorylation represents the predominant PTM of MYOCD in vivo. Notably, the IVT MYOCD band did not shift to the predicted molecular weight of 95-100 kDa, suggesting the presence of additional factors or conformational effects.

Conserved MYOCD-interacting proteins identified by cross-species mass spectrometry

No studies have assessed the MYOCD interactome in vivo. Here, we sought to identify potential MYOCD-binding partners through co-immunoprecipitation (co-IP) coupled with tandem mass spectrometry (MS). Immunoprecipitation of endogenous MYOCD from nuclear lysates of aortic VSMCs derived from the medial layer yielded several-fold enrichment relative to input, confirming that the HA tag permits capture of endogenous MYOCD (Fig. 3D). We next employed co-IP-MS analysis to investigate potential SRF-independent mechanisms of MYOCD function. Despite multiple replicates, only a small amount of MYOCD was recovered relative to the total protein content (Table 3). This limited recovery is consistent with previous studies using a Myocd^3xHA mouse (Table 6). Furthermore, the immunoprecipitation of HA produced several non-specific bands below the expected ~150 kDa band possibly confounding MYOCD enrichment and increasing background noise (Fig. 3D). To identify conserved MYOCD-interacting partners, we cross-referenced the rat Myocd^3xHA MS results with data obtained from Myocd^3xHA mice, using both aorta and bladder for comparison (Fig. 3E). Complete lists of candidate MYOCD-associated proteins with unions between rat aorta, mouse aorta, and mouse bladder are provided in Tables 4–6. Notably, three proteins identified in our cross-species comparison: ACTB^[37], HSP90AB1 (Hsp90)^[31], and HSPA8 (Hsp70)^[31], have been previously reported to interact with tagged MYOCD in vitro. To further substantiate these findings, we performed confirmatory co-immunoprecipitation followed by Western blotting and were able to validate binding interactions with ACTB and HSPA8 (Supp. Fig 3), whereas HSP90AB1 could not be confirmed under our experimental conditions. The identification of previously reported interactors by mass spectrometry and their initial confirmation by co-IP support the reliability of our interactome findings, despite limited MYOCD recovery.

Table 3: Immunoprecipitation and tandem mass spectrometry results for Myocd^HA rat aorta.

Table of enriched peptides from immunoprecipitation and mass spectrometry using HA-magnetic beads on protein lysates from rat aorta. Average enrichment across samples for Myocd^HA or Myocd^WT is listed. MYOCD is pinned to the top.

Rank	Accession No.	Protein Name	Avg #Peptides in HA	Avg #Peptides in WT
259	A0A0G2K4C5	Myocardin OS=Rattus norvegicus OX=10116 GN=Myocd PE=4 SV=1	3.00	0
1	F1LTJ5	Uncharacterized protein OS=Rattus norvegicus OX=10116 PE=1 SV=3	61.33	45.33
2	Q99P77	GTP-binding protein 4 OS=Rattus norvegicus OX=10116 GN=Gtpbp4 PE=2 SV=3	9.00	1.00
3	G3V6P7	Myosin, heavy polypeptide 9, non-muscle OS=Rattus norvegicus OX=10116 GN=Myh9 PE=1 SV=1	45.33	36.33
4	P15800	Laminin subunit beta-2 OS=Rattus norvegicus OX=10116 GN=Lamb2 PE=1 SV=1	20.33	11.67
5	G3V9Y1	Myosin, heavy polypeptide 10, non-muscle, isoform CRA_b OS=Rattus norvegicus OX=10116 GN=Myh10 PE=1 SV=1	35.33	27.00
6	F1MAN8	Laminin subunit alpha 5 OS=Rattus norvegicus OX=10116 GN=Lama5 PE=1 SV=2	17.67	9.67
7	F1M853	Ribosome-binding protein 1 OS=Rattus norvegicus OX=10116 GN=Rrbp1 PE=1 SV=3	9.00	2.00
8	G3V9M6	Fibrillin 1 OS=Rattus norvegicus OX=10116 GN=Fbn1 PE=1 SV=1	69.33	62.00
9	F1MAA7	Laminin subunit gamma 1 OS=Rattus norvegicus OX=10116 GN=Lamc1 PE=1 SV=1	16.33	9.33
10	E9PTU4	Myosin-11 OS=Rattus norvegicus OX=10116 GN=Myh11 PE=1 SV=2	102.67	96.00
11	A0A0G2K6S9	Myosin-11 OS=Rattus norvegicus OX=10116 GN=Myh11 PE=1 SV=1	102.00	95.33
12	Q6T487	Alpha-actinin-1 OS=Rattus norvegicus OX=10116 GN=Actn1 PE=1 SV=1	18.67	13.00
13	F1LM84	Nidogen-1 OS=Rattus norvegicus OX=10116 GN=Nid1 PE=1 SV=1	14.00	8.67
14	G3V6S0	Spectrin beta chain OS=Rattus norvegicus OX=10116 GN=Sptbn1 PE=1 SV=3	6.00	1.00
15	F1LST1	Fibronectin OS=Rattus norvegicus OX=10116 GN=Fn1 PE=1 SV=3	25.00	20.00
16	A0A0G2K013	Alpha-actinin-4 OS=Rattus norvegicus OX=10116 GN=Actn4 PE=1 SV=1	13.67	8.67
17	D3ZZT9	Collagen type XIV alpha 1 chain OS=Rattus norvegicus OX=10116 GN=Col14a1 PE=1 SV=3	45.00	40.33
18	P06761	Endoplasmic reticulum chaperone BiP OS=Rattus norvegicus OX=10116 GN=Hspa5 PE=1 SV=1	9.00	4.33
19	B5DFC9	Nidogen-2 OS=Rattus norvegicus OX=10116 GN=Nid2 PE=2 SV=1	7.00	2.50
20	G3V6Y6	Alpha-1,4 glucan phosphorylase OS=Rattus norvegicus OX=10116 GN=Pygb PE=1 SV=2	12.00	7.67
21	D3ZAF5	Periostin OS=Rattus norvegicus OX=10116 GN=Postn PE=1 SV=1	11.33	7.00
22	G3V7U4	Lamin-B1 OS=Rattus norvegicus OX=10116 GN=Lmnb1 PE=1 SV=1	5.00	1.00
23	G3V7L1	Utrophin OS=Rattus norvegicus OX=10116 GN=Utrn PE=1 SV=3	5.00	1.00
24	F1MAF2	A-kinase-anchoring protein 17A OS=Rattus norvegicus OX=10116 GN=Akap17a PE=1 SV=1	5.00	1.00
25	Q63356	Unconventional myosin-Ie OS=Rattus norvegicus OX=10116 GN=Myo1e PE=1 SV=1	5.00	1.00
26	F1M987	Signal peptide, CUB domain and EGF-like domain-containing 1 OS=Rattus norvegicus OX=10116 GN=Scube1 PE=4 SV=1	10.33	6.33
27	P15205	Microtubule-associated protein 1B OS=Rattus norvegicus OX=10116 GN=Map1b PE=1 SV=3	6.33	2.33
28	D3ZUL3	Collagen type VI alpha 1 chain OS=Rattus norvegicus OX=10116 GN=Col6a1 PE=1 SV=1	18.00	14.33
29	Q3B7D6	F-spondin OS=Rattus norvegicus OX=10116 GN=Spon1 PE=1 SV=1	21.00	17.67
30	Q811S9	Guanine nucleotide-binding protein-like 3 OS=Rattus norvegicus OX=10116 GN=Gnl3 PE=1 SV=1	15.67	12.67
31	Q923Z2	Tropomyosin 1, alpha, isoform CRA_a OS=Rattus norvegicus OX=10116 GN=Tpm1 PE=1 SV=1	10.67	7.67
32	Q5RJR9	Collagen-binding protein OS=Rattus norvegicus OX=10116 GN=Serpinh1 PE=1 SV=1	8.67	5.67
33	P04633	Mitochondrial brown fat uncoupling protein 1 OS=Rattus norvegicus OX=10116 GN=Ucp1 PE=1 SV=2	4.00	1.00
34	Q63355	Unconventional myosin-Ic OS=Rattus norvegicus OX=10116 GN=Myo1c PE=1 SV=2	26.67	24.00
35	P11442	Clathrin heavy chain 1 OS=Rattus norvegicus OX=10116 GN=Cltc PE=1 SV=3	11.00	8.50
36	G3V8L3	Lamin A, isoform CRA_b OS=Rattus norvegicus OX=10116 GN=Lmna PE=1 SV=1	24.00	21.67
37	G3V7C6	Tubulin beta chain OS=Rattus norvegicus OX=10116 GN=Tubb4b PE=1 SV=2	17.33	15.00
38	F1LNH3	Collagen type VI alpha 2 chain OS=Rattus norvegicus OX=10116 GN=Col6a2 PE=1 SV=2	16.00	13.67
39	F1LXS3	Matrilin 2 OS=Rattus norvegicus OX=10116 GN=Matn2 PE=4 SV=3	4.50	2.00
40	A0A0G2JX64	Tropomyosin 1, alpha, isoform CRA_i OS=Rattus norvegicus OX=10116 GN=Tpm1 PE=1 SV=1	10.67	8.33
41	B1WBZ1	Embryonal Fyn-associated substrate OS=Rattus norvegicus OX=10116 GN=Efs PE=2 SV=1	9.67	7.33
42	P32089	Tricarboxylate transport protein, mitochondrial OS=Rattus norvegicus OX=10116 GN=Slc25a1 PE=1 SV=1	3.50	1.00
43	D3ZIF0	Zinc finger protein 512 OS=Rattus norvegicus OX=10116 GN=Zfp512 PE=1 SV=1	3.50	1.00
44	Q9WUL0	DNA topoisomerase 1 OS=Rattus norvegicus OX=10116 GN=Top1 PE=1 SV=1	3.50	1.00
45	A0A0G2K0Q7	Myosin light chain kinase, smooth muscle OS=Rattus norvegicus OX=10116 GN=Mylk PE=1 SV=1	7.33	5.00
46	Q5BJP4	RNA binding motif protein 39 OS=Rattus norvegicus OX=10116 GN=Rbm39 PE=1 SV=1	4.33	2.00
47	C0JPT7	Filamin A OS=Rattus norvegicus OX=10116 GN=Flna PE=1 SV=1	61.67	59.67
48	D4A917	Latent-transforming growth factor beta-binding protein 4 OS=Rattus norvegicus OX=10116 GN=Ltbp4 PE=1 SV=2	16.67	14.67
49	P62909	40S ribosomal protein S3 OS=Rattus norvegicus OX=10116 GN=Rps3 PE=1 SV=1	15.00	13.00

abbreviations: Myocd^WT, wild type; Myocd^HA, homozygous (3xHA).

Table 6:

Mass spectrometry results for Myocd^HA rat aorta compared against Myocd^HA mouse aorta and bladder

Gene	RatAo	MusAo	MusBl
ACAA2	7	9	2
ACTA1	19	16	16
ACTB *	21	15	18
ACTN1	37	6	5
ACTN4	27	3	5
ACTR3	1	1	1
ALB	6	11	15
ANXA1	4	2	3
ANXA2	14	10	8
ANXA5	1	6	7
ANXA6	6	9	13
AOC3	2	4	2
AQP1	1	1	1
ARF3	6	4	4
ARG1	1	2	5
ATP5F1A	15	12	12
ATP5F1C	11	1	2
ATP5MF	2	1	1
ATP5PO	8	1	3
BCAR1	4	30	29
BGN	13	7	5
CALM1	6	5	3
CAP1	1	3	6
CAPZA2	8	1	1
CAVIN1	13	2	8
CDC42	5	2	1
CFL1	4	3	5
CLTC	21	8	8
CNN1	8	6	11
COL14A1	53	1	7
COL1A1	11	1	1
COL6A1	22	11	8
COL6A2	21	10	3
CSRP1	9	3	7
CYB5R3	1	6	6
DCN	11	5	6
DECR1	1	2	1
DPYSL3	2	5	4
DSP	10	41	34
DSTN	6	7	7
EEF2	2	6	10
EFS	14	1	3
EHD2	20	6	2
EIF5A	2	1	2
FBN1	96	8	2
FHL1	14	7	6
FLNA	97	30	52
GAPDH	10	12	14
GDI2	1	1	2
GSN	12	4	9
H1F4	9	3	3
H1F5	6	1	1
HADHA	12	14	6
HADHB	3	8	1
HBA-A1	3	4	6
HNRNPA2B1	2	3	5
HNRNPK	7	2	4
HSP90AB1 *	7	4	16
HSPA5	14	10	9
HSPA8 *	15	15	11
HSPB1	5	1	5
IQGAP1	50	1	3
JUP	9	16	18
KRT1	9	10	10
KRT14	14	19	14
KRT16	11	12	8
KRT17	16	18	12
KRT19	8	11	15
KRT2	9	10	9
KRT42	13	18	15
KRT5	20	21	21
KRT75	19	16	11
KRT77	7	13	10
KRT79	9	10	8
KRT8	8	8	15
LGALS1	4	1	3
LMCD1	3	2	1
LMNA	32	4	7
LMNB1	8	1	1
LMOD1	8	1	1
LPP	8	7	12
LUM	10	8	7
MYL12B	9	6	7
MYL9	10	7	9
MYOCD **	3	9	12
NACA	3	1	1
NDUFA4	2	3	1
PARVA	3	1	2
PDLIM3	2	2	5
PDLIM7	10	3	3
PGM5	19	4	4
PKM	3	12	17
POSTN	24	13	4
PPIA	2	5	5
PPP1R12A	1	1	6
PPP2R2A	2	1	1
PRELP	16	5	3
RACK1	1	1	2
RAN	7	1	2
RPL11	7	1	1
RPL12	4	1	2
RPL18	5	1	3
RPL19	4	1	1
RPL22	4	1	2
RPL27	6	1	2
RPL35	5	1	2
RPL6	11	1	2
RPLP2	2	1	1
RPS13	7	1	2
RPS16	9	2	1
RPS27A	5	3	2
RPS3	17	2	2
RPS6	13	1	3
RPS7	12	1	2
RPS8	14	2	3
RPS9	15	1	1
RPSA	15	1	2
RRAS	6	2	2
S100A11	1	1	1
SERPINH1	11	1	1
SLC25A4	13	7	8
SLC25A5	12	8	7
SMTN	7	2	3
SPTAN1	10	3	10
SRSF1	3	2	1
SRSF7	3	2	2
TAGLN	9	12	15
TGM2	29	6	5
TLN1	20	4	11
TPM1	31	8	5
TPM2	15	6	5
TRA2B	3	2	2
TUBB2A	17	13	11
TUBB5	19	16	14
TUFM	4	2	1
TXN1	1	1	1
UQCRC2	6	5	3
UQCRFS1	1	3	1
VCL	14	15	11
VDAC2	2	2	1
VIM	28	13	21
YWHAZ	1	8	6

Full table of overlapping peptides in cross-species HA–mass spectrometry. Proteins with one asterisk (*) have supporting publications for interacting with Myocardin. MYOCD is denoted with two asterisks (**). Columns indicate peptide count.

abbreviations: Myocd^HA, homozygous (3xHA); RatAo, homozygous rat aorta; MusAo, homozygous mouse aorta; MusBl, homozygous mouse bladder.

Table 4: Mass spectrometry results for Myocd^HA rat aorta compared against Myocd^HA mouse aorta.

Full table of overlapping peptides in cross-species HA–mass spectrometry. Columns indicate peptide count.

Gene	RatAo	MusAo
AEBP1	16	1
CALML3	3	2
CAPN2	4	1
CAPZB	6	1
CAT	2	2
COL15A1	12	2
COL18A1	18	7
COL1A2	5	1
COL4A1	4	1
COL4A2	5	3
COL6A6	17	2
CRIP2	3	1
CSRP2	5	4
DBN1	10	1
EFEMP1	3	1
FBLIM1	6	1
FBLN5	10	7
FN1	44	6
GNB1	1	2
HNRNPA3	2	1
HTRA1	6	1
ITGA8	1	1
KRT12	2	1
KRT28	3	3
LAMA4	22	1
LAMB2	37	6
LAMC1	23	5
LTBP4	26	7
NEDD4	6	1
NID1	26	1
NID2	11	1
OGN	9	1
PTGIS	1	6
PXDN	31	1
RAB5C	2	2
SFXN3	7	1
SLC25A1	4	2
SOD2	1	1
TGFB1I1	9	2
TGFB1	4	1
TINAGL1	12	2
UCP1	4	7
UQCR10	1	1
VCAN	4	1
YBX1	8	1

abbreviations: Myocd^HA, homozygous (3xHA); RatAo, homozygous rat aorta; MusAo, homozygous mouse aorta.

Spatial localization of Myocd^3xHA across diverse tissues

To date, the reliable spatial localization of endogenous MYOCD protein in vivo has not been reported. We addressed this gap by sampling SMC enriched Myocd^HA tagged rat tissues such as aorta, bladder, and uterus. Each of these tissues displayed discrete nuclear localization, consistent with the first reporting of MYOCD in cells^[8] (Fig. 4A–C, Supp. Fig. 4). More than 85% of ACTA2-positive cells in different SMC tissues showed nuclear MYOCD protein (Fig. 4A–C, Supp. Fig. 4A-C). To further substantiate these findings across species, we examined Myocd^3xHA Myocd mouse tissues from a previously reported model^[18]. Here, discrete nuclear localization of MYOCD was observed in adult mouse aorta, bladder, and developing E10.5 embryonic mouse heart (Fig. 4D–F). Next, an analysis of MYOCD protein expression was performed in several SMC-poor tissues. Despite the variable background staining of such tissues as cortex of the brain, epithelial lining of the colon, cardiomyocytes, tubules of the kidney, and alveoli of the lungs, we noted nuclear MYOCD protein in vessels of these tissues and in some visceral smooth muscle (eg, intestine, colon) (Supp. Fig. 4D-J). To further interrogate the subcellular localization of MYOCD in SMC, we employed immunogold electron microscopy (IEM). HA-IEM revealed gold particle localization within the nucleus of aortic SMC of Myocd^HA mice. Interestingly, most gold particles were found to localize in heterochromatic regions of SMC nuclei (Supp. Fig. 4L). Collectively, these findings demonstrate MYOCD protein expression in the nucleus of vascular and visceral SMCs across a broad range of postnatal tissues.

Figure 4: Spatial localization of MYOCD protein in rat and mouse tissues.

Representative immunostaining of Myocd^3xHA rat A, abdominal aorta; B, bladder; and C, uterus. All quantitation represents independent biological replicates. Immunostaining results from Myocd^3xHA mouse D, descending aorta; E, bladder and F, e10.5 embryonic heart. Scale bars represent 50 μm. Fluors are pseudo-colored: red, ACTA2; white, HA (MYOCD); blue, DAPI. Yellow arrows indicate HA-positive nuclei within ACTA2-positive regions. Dotted lines denote boundaries around vessels. anti-HA, CST (3724S) was used for all HA immunostaining.

abbreviations: Myocd^3xHA, 3x HA tagged Myocardin; Myocd^WT, wild type; Myocd^HA, homozygous (3xHA tag); A, adventitia; M, media; L, lumen.

Endogenous MYOCD protein localization following vascular injury

Myocd mRNA is well known to decrease under pathological conditions in vitro and in vivo ^{[38, 39]}. However, accurate in vivo characterization of endogenous MYOCD protein following vascular injury has not been reported. We performed balloon de-endothelialization of the left carotid artery (LCA) in 16-week-old Myocd^HA rats and allowed lesions to develop for two weeks (Fig. 5A, 5B). Injured LCAs exhibited robust neointimal lesion formation, evident by gross morphology and confirmed by H&E staining at levels distal to the suture site (Fig. 5C, 5D). We next examined Myocd^3xHA localization in tissue sections from both Myocd^HA and Myocd^WT balloon-injured animals. MYOCD protein staining was readily detected in the medial layer of the carotid artery. Surprisingly, we also observed some MYOCD protein in the neointimal region of injured LCAs (Fig. 5E, Supp. Fig. 5A). No appreciable nonspecific staining was observed in primary-omitted control vessels (Supp. Fig. 6B). We supplemented these findings with a cross-species comparison to the Myocd^3xHA mouse model using ligation injury of the carotid artery (Fig. 5F). As in the rat injury model, the mouse injury model showed similar neointimal Myocd^3xHA positivity in the LCA and expected medial SMC staining of both the LCA and RCA. Collectively, these findings reveal a previously unrecognized spatial distribution of MYOCD protein within the neointima after acute injury.

Figure 5: MYOCD is spatially localized in neointimal cells in rat and mouse injury models.

A, Cartoon schematic of the balloon de-endothelialization procedure performed on the rat left carotid artery. Created using BioRender. B, Balloon injury timeline from week 16 to staining of harvested vessels. C, Whole-mount image of ballooned rat carotid artery with red lines indicating the suture site and descending levels of sectioning. D, Representative H&E staining of ballooned LCA and control RCA. Scale bars represent 200 μm. E, Immunostaining of rat LCA and RCA after balloon injury. n = 6 independent rats. White squares indicate the zoomed region shown in the split panels. Blue: DAPI; red: ACTA2; white: HA (MYOCD). F, Myocd^3xHA mouse LCA and RCA after ligation injury with H&E and immunostaining of adjacent sections. Medial border is demarked by the dashed line in panels E, D, and F. n = 5 independent mice. Blue: DAPI; white: HA (MYOCD). Scale bars for E and F images represent 50 μm. anti-HA, CST (3724S) was used for all HA immunostaining.

abbreviations: Myocd^3xHA, 3xHA tagged Myocardin; ECA, external carotid artery; ICA, internal carotid artery; LCCA, left common carotid artery; LCA, left carotid artery; RCA, right carotid artery; M, media; A, adventitia; NI, neointima.

Discussion

Reliable detection of endogenous MYOCD protein has remained an unresolved challenge in SMC biology. Here, we developed and characterized a rat model harboring a C-terminal 3×HA epitope tag at the endogenous Myocd locus. Characterization of this rat model at both the mRNA and protein levels confirmed the preferential expression of MYOCD in smooth muscle-enriched tissues^[14]. Further, consistent with previous reports of Myocd mRNA^{[8, 10, 18, 22, 27]}, we observed MYOCD protein expression in embryonic heart and adult aorta, bladder, and uterus. Surprisingly, but consistent with a prior report in mice^[18], Myocd mRNA was readily detected in adult rat heart, but its protein expression was barely detected, a discordance that warrants further investigation.

MYOCD is frequently reported at an apparent molecular weight of 95-100 kDa, corresponding to its primary amino acid sequence prediction. However, as demonstrated above and in a previous mouse study^[18], authentic MYOCD migrates at ~130-150 kDa, depending on the percentage gel used. Multiple Western blots presented in this study reveal a nonspecific band of ~100 kDa that persists even in conditional Myocd knockout tissue, whereas the true higher molecular weight band is absent. These findings provide compelling evidence that the higher molecular weight species represents bona fide MYOCD. Notably, MYOCD is not unique in this regard; MRTFA, with a primary amino acid sequence of 1,029 amino acids in mouse, has a predicted molecular weight of 109 kDa, but has been shown to also migrate at ~150 kDa. ^[40]. The underlying basis for this higher-than-predicted molecular weight remains undefined.

Notably, a commercial MYOCD antibody and the HA antibody yielded comparable detection of overexpressed protein; however, this concordance did not extend to the endogenous context where commercial antibodies have failed to detect authentic MYOCD in SMC-rich tissue lysates^[14]. Here, however, we discovered a single commercial antibody that reliably detected endogenous MYOCD. Despite its efficacy by immunoblot, this antibody failed to detect MYOCD by immunofluorescence on either paraffin or frozen tissue sections (data not shown). Interestingly, this antibody was raised against twenty amino acids at the C-terminus of MYOCD, a region closely adjacent to the site targeted by the 3×HA tag. The success of both this antibody and the HA tag may be attributable to the low predicted disorder of the MYOCD C-terminus, which would permit greater accessibility of the underlying amino acid sequence for antibody binding. While C-termini are generally disordered across the proteome^[41], MYOCD’s C-terminus appears to be more structured, which may contribute to its favorable recognition by antibodies. Indeed, the C-terminus represents an advantageous site for protein tagging. In a systematic comparison of N- and C-terminal GFP fusion proteins, all C-terminal fusions localized correctly to their predicted cellular compartments, whereas fewer than half of N-terminal fusions did so, with many exhibiting either no expression or aberrant localization^[42]. Similarly, systematic studies in yeast have demonstrated that C-terminal tagging better supports both mitochondrial and nuclear localization compared to N-terminal tagging^[43]. In contrast, N-terminal tagging of transcription factors can compromise protein stability, as demonstrated for SMAD5, where N-terminally tagged protein was markedly less abundant than C-terminal or internal tagged variants^[44]. Taken together, these findings support the C-terminus of MYOCD as an optimal site for epitope tagging and antibody-based detection.

To elucidate the basis for the discrepancy in MYOCD molecular weight, we evaluated two PTMs known to alter electrophoretic mobility: phosphorylation and glycosylation^{[45, 46]}. Several prior studies have examined MYOCD PTMs^{[29–36, 47, 48]}, but all such analyses were conducted using ectopically expressed protein in cultured cells. Here, we directly assessed PTMs on endogenous MYOCD in vivo. Treatment of tissue lysates with glycosidase produced no discernible shift in MYOCD migration, whereas IgG, included as a positive control^[49], exhibited an expected subtle mobility change characteristic of glycosidase treatment. In contrast, phosphatase treatment induced a pronounced downward shift in MYOCD mobility to a position just above the 100 kDa nonspecific band. Surprisingly, neither phosphatase-treated nor IVT-generated MYOCD collapses to the predicted 100 kDa suggesting that intrinsic properties of the protein, potentially including regions of disorder^[50] or unusual amino acid composition^[51], contribute to its anomalous migration. Unexpected band migration independent of PTMs has been documented for other transcription factors. For example, CTCF migrates at 130 kDa despite a predicted molecular weight of 82 kDa, a discrepancy confirmed by mass spectrometry to result from intrinsic protein properties rather than covalent modifications^[52]. Similarly, IVT MYOCD rules out PTMs as a driving factor for MYOCD MW shift. On the other hand, PTMs have been shown in other transcription factors to generate substantial molecular weight shifts, as observed with MYC, where polyubiquitination produces high molecular weight species above the unmodified protein^[53], and heat shock factors, where SUMOylation produces a characteristic upward shift^{[54, 55]} .

An important finding in this report is the first spatial characterization of endogenous MYOCD protein expression in vivo. Immunofluorescence microscopy revealed discrete nuclear localization of MYOCD within SMC-enriched tissues, consistent with the original description of MYOCD as a nuclear cofactor^[8]. The HA tag also permitted identification of MYOCD-positive nuclei within the microvasculature of non-SMC tissues, bringing to light an overlooked spatial context of MYOCD protein. Perhaps the most unexpected finding of this study was the presence of MYOCD protein within neointimal cells following vascular injury. Both balloon-injured rat and ligated mouse carotid arteries exhibited MYOCD immunoreactivity in the neointimal region after insult, a result observed consistently across multiple independent animals. This finding stands in contrast to the well-documented decrease in Myocd mRNA that accompanies SMC dedifferentiation and phenotypic modulation^[38]. The persistence or, more likely, reemergence of MYOCD protein in intimal cells suggests potential functions outside of SMC gene regulation and opens new avenues of investigation into the role of MYOCD in vascular remodeling and disease states. Expanding this finding to vascular diseases with non-uniform medial expansion such as abdominal aortic aneurysm or atherosclerosis, would be of interest. Whether these neointimal MYOCD-positive cells derive from medial SMCs or alternative sources remains to be determined.

To assess MYOCD subnuclear localization, we used immunogold EM labeling of Myocd^3xHA mice and observed a striking pattern of MYOCD over heterochromatic regions of the nucleus. This finding was unexpected as MYOCD would be predicted to localize to CArG-rich promoters within active euchromatin^{[9, 11, 12]}. One possible explanation for this localization is that MYOCD belongs to the SAP domain family of nuclear proteins. SAP domains have been shown to bind A/T-rich scaffold attachment regions and satellite DNA sequences that are concentrated in heterochromatin ^[56]. HNRNPU, the founding member of this protein family^[57], specifically binds mouse DNA within heterochromatin^[58]. Notably, transcription factor sequestration at heterochromatin has emerged as a regulatory mechanism. For example, C/EBPα localizes to heterochromatin through binding to repetitive DNA, which functionally sequesters C/EBPα and reduces its transcriptional capacity; preventing this heterochromatic localization resulted in elevated promoter binding and increased target gene activation^[59]. More recently, TEAD1 was shown to form puncta at heterochromatin that serve as transcriptionally inactive storage depots^[60]. Given that MYOCD can form condensates^{[61, 62]}, it is reasonable to speculate that the heterochromatic localization observed here represents a similar sequestration phenomenon.

A related observation concerns the discordance between Myocardin mRNA and protein in the heart. Myocd mRNA is highly expressed in both developing and adult myocardium^{[8–10, 27]}, as well as the failing myocardium^[63], yet protein expression diminishes markedly in the normal adult heart. Here, we demonstrate the presence and localization of MYOCD protein in the developing E10.5 mouse myocardium. The functional importance of MYOCD protein during cardiogenesis is underscored by direct reprogramming studies, wherein MYOCD is an essential component of transcription factor cocktails that convert fibroblasts into cardiomyocyte-like cells^[64–66]. Notably, while the core reprogramming factors GATA4, MEF2C, and TBX5 (also written as GMT) are sufficient for cardiac reprogramming, enhanced reprogramming was observed when MYOCD was included ^{[64, 67]}. Furthermore, studies in cardiac stem cells revealed that MYOCD addition to GMT triggered expression of cardiac structural proteins that were not induced by GMT alone, suggesting MYOCD plays a critical role in cardiac transcriptional programs ^[68]. This observation parallels our finding of robust MYOCD protein in developing myocardium and may reflect a developmental requirement for MYOCD protein to establish the cardiomyocyte contractile apparatus. Interestingly, conditional deletion of Myocd in the postnatal heart leads to dilated cardiomyopathy with disrupted cardiomyocyte structural organization and lethality^[28], suggesting that even low-level MYOCD protein expression imparts important function in maintaining cardiomyocyte integrity in the postnatal myocardium.

The rat model described here offers several practical advantages over our previously reported mouse model^[18]. The rat is substantially larger, permitting a single aorta to yield protein quantities equivalent to three or four mouse aortas. Furthermore, Sprague-Dawley rats represent an outbred strain, more closely approximating the genetic heterogeneity of human populations than inbred mouse lines^[69]. These attributes, combined with the expanded methodological applications demonstrated here, including immunostaining, immunoprecipitation-mass spectrometry, PTM analysis, and spatial characterization following vascular injury, position this model as a valuable resource for the SMC research community.

Limitations of the study

Several limitations of the present study warrant consideration. First, lineage tracing was not incorporated into our vascular injury experiments. Consequently, assertions regarding the SMC origin of neointimal MYOCD-positive cells remain inferential and will require fate-mapping approaches for definitive resolution. In this context, Cre-mediated recombination of loxP sites in rats has proven inefficient, imposing significant constraints on genetic manipulation strategies in this species (unpublished). Second, immunoprecipitation of MYOCD yielded several nonspecific bands which may contribute to increased background noise in the proteomics results. Finally, mass spectrometry analysis revealed limited MYOCD peptide recovery, a finding consistent with prior work in a mouse model^[18]. Despite substantially greater tissue mass and protein input for immunoprecipitation, MYOCD did not constitute a significant proportion of the recovered peptides, suggesting that the HA epitope may not be optimal for this application. Notably, this contrasts with Western blot analysis of immunoprecipitated samples, which demonstrated robust enrichment of Myocd^3xHA relative to input, indicating that the limitation lies specifically in the mass spectrometry workflow rather than in immunoprecipitation efficiency. It is critical to note that none of our attempted runs produced SRF peptides, the canonical MYOCD cofactor^[8]. We speculate this is due to a transient interaction between the two binding partners. Experiments where MYOCD/SRF are ectopically expressed generate robust Co-IP results suggesting a supraphysiologic level of the protein is needed to observe binding^{[8, 12, 70]}.

In summary, the Myocd^3xHA rat model enables spatial profiling of endogenous MYOCD protein in normal and disease tissues and provides new insight into its unusually high molecular weight in protein gels, a finding suggestive of a constitutively phosphorylated state in vivo. The detection of MYOCD protein in neointimal cells following vascular injuries across species challenges prevailing models of MYOCD regulation during SMC phenotypic modulation. Finally, observations here underscore the necessity for rigorous antibody validation in MYOCD studies and set the stage for deeper investigation into the functional consequences of MYOCD phosphorylation and the role of this coactivator in phenotypically modulated SMCs during vascular disease.

Supplementary Material

Supplemental Figs. S1–S5: [DOI:10.5281/zenodo.20165506]

Table 5: Mass spectrometry results for Myocd^HA rat aorta compared against Myocd^HA mouse bladder.

Full table of overlapping peptides in cross-species HA–mass spectrometry. Columns indicate peptide count.

Gene	RatAo	MusBl
ACTR1A	2	2
ARPC1B	5	1
ATP5PD	2	3
ATP5PF	1	1
BTF3	2	1
CAVIN2	2	4
COL12A1	5	2
COPB2	1	1
CORO1C	10	1
DDX1	2	1
DDX17	10	1
DES	18	16
DYNC1H1	1	1
EIF4A1	1	1
EWSR1	1	1
FBLN1	2	3
FERMT2	4	3
FGA	8	2
FGB	12	2
FLNC	8	4
GLUD1	2	3
GPX1	1	1
H2AX	5	4
HNRNPL	1	1
HNRNPU	11	6
IDH2	3	2
JPH2	13	1
MACROH2A1	13	2
NAP1L1	2	2
NCL	1	4
NIBAN1	1	1
P4HB	1	5
PCBP2	1	2
PDCD4	6	1
PHB2	2	1
PPP1R12B	3	3
PZP	3	2
RBMX	1	1
RPL10	11	2
RPL10A	9	1
RPL13A	6	1
RPL17	10	3
RPL18A	5	1
RPL21	6	1
RPL24	8	1
RPL27A	8	1
RPL3	17	2
RPL34	6	1
RPL37A	5	1
RPL4	14	3
RPL7	11	3
RPL7A	12	2
RPLP0	4	2
RPN1	11	1
RPN2	3	1
RPS10	4	1
RPS11	11	1
RPS15	6	1
RPS15A	6	2
RPS19	6	4
RPS2	11	4
RPS20	4	1
RPS23	8	2
RPS28	3	1
S100A6	1	1
SEPTIN7	1	1
SFPQ	6	1
SLK	13	1
SND1	5	1
SNRPD3	2	1
SPTBN1	7	7
SRSF10	1	1
SYNPO2	10	1
TCP1	1	5
TOP1	5	1
UGP2	1	1

abbreviations: Myocd^HA, homozygous (3xHA); RatAo, homozygous rat aorta; MusBl, homozygous mouse bladder.