Abstract
Background & aims Visceral smooth muscle cells (SMCs) are an integral component of the gastrointestinal (GI) tract and are critical for regulating motility. SMC contraction is regulated by changes in post-translational signaling and the state of differentiation. Impaired SMC contraction is associated with significant morbidity and mortality but the mechanisms regulating the expression levels of SMC-specific contractile proteins, including the role of long non-coding RNAs (lncRNAs), remains largely unexplored. Herein, we have uncovered an important role of Carmn (Cardiac mesoderm enhancer-associated noncoding RNA), a SMC-specific lncRNA, in regulating the phenotype of visceral SMCs of the GI tract.
Methods Analysis of GTEx and publicly available single-cell RNA sequencing (scRNA-seq) datasets from embryonic, adult human and mouse GI tissues were used to identify SMC-specific lncRNAs. The functional role of Carmn was investigated using a novel GFP knock-in (KI) reporter/knockout (KO) mouse model. Bulk RNA sequencing (RNA-seq) and single nuclei RNA sequencing (snRNA-seq) of colonic muscularis were used to investigate underlying mechanisms.
Results Unbiased in silico analyses and GFP expression patterns in Carmn GFP KI mice revealed that Carmn is specifically expressed in SMCs in human and mouse GI tract. Premature lethality was observed in global Carmn KO (gKO) and inducible SMC-specific KO (iKO) mice due to colonic pseudo-obstruction, severe distension of the GI tract with blockages in cecum and colon segments. Histology, whole-gut GI transit time and muscle myography analysis revealed severe dilation, significantly delayed GI transit and impaired GI contractility in Carmn KO mice versus control mice. Bulk RNA-seq of colonic muscularis revealed that Carmn deficiency promotes SMC de-differentiation as evidenced by up-regulation of extracellular matrix genes and down-regulation of SMC contractile genes including Mylk, a key regulator of SMC contraction. SnRNA-seq further revealed SMC Carmn deficiency not only compromised myogenic motility by reducing expression of contractile genes but also impaired neurogenic motility by disrupting cell-cell connectivity in the colonic muscularis. These findings may have translational significance as silencing CARMN in human colonic SMCs significantly attenuated contractile gene expression including MYLK and decreased SMC contractility. Luciferase reporter assays showed that CARMN enhances the transactivation activity of the master regulator of SMC contractile phenotype, myocardin, thereby maintaining the GI SMC myogenic program.
Conclusion Our data suggest that Carmn is indispensable for maintaining GI SMC contractile function in mice, and that loss of function of CARMN may contribute to human visceral myopathy. To our knowledge this is the first study showing an essential role of lncRNA in the regulation of visceral SMC phenotype.
Introduction
The major functions of the gastrointestinal (GI) tract are the digestion of food, absorption of nutrients and removal of waste. To accomplish these critical functions, smooth muscle cells (SMCs) populating the walls of the GI tracts generate the peristaltic forces that efficiently move food from one segment to the next or to retain food in one region until digestion and absorption are completed 1. Loss of GI motility is a characteristic of several smooth muscle-motility diseases including chronic intestinal pseudo-obstruction (CIPO) which is characterized by impaired contractility and functional intestinal obstruction 2, 3. Smooth muscle-motility disorders have been shown to be caused by genetic mutations resulting in loss of function of SMC-contractile genes such as MYH11 4, MYLK 5, LMOD1 6, ACTG2 7, ACTA2 8 and MYL9 9 or the inactivation of their upstream regulatory transcription factors, such as Srf 10, 11 and Myocd 12, two master regulators that form a complex binding to CArG elements within SMC-contractile gene loci 13. The etiology of CIPO is multifactorial, and the importance of epigenetic regulators such as long non-coding RNAs (lncRNAs) in this disease remains largely unexplored.
LncRNAs are a new class of RNA transcripts that have lengths exceeding 200 nucleotides but no apparent protein-coding potential. Accumulating evidence suggests that lncRNAs act as non-coding regulatory molecules that play critical roles in a variety of physiological and pathological conditions 14, 15. Although several lncRNAs have been shown to contribute to SMC biology 16-23, these studies have exclusively focused on vascular SMCs. To our knowledge, thus far there has no reports on the functional role of lncRNAs in visceral SMCs.
In the current study, data mining of multiple independent publicly available omics datasets, revealed a novel lncRNA, CARMN, as the most abundantly expressed lncRNA in visceral SMCs of both human and mouse. CARMN was initially identified as an important regulator of cardiac differentiation in vitro 24, 25. Recently, we showed that CARMN is a highly abundant and conserved SMC-specific lncRNA, which plays a critical role in maintaining vascular SMC contractile phenotype 26-28. Unexpectedly, we found that both germline and SMC-specific inducible deletion of Carmn in mice resulted in premature lethality due to colonic pseudo-obstruction. These data suggest that Carmn is indispensable for GI function and to the best of our knowledge, it is the first lncRNA shown to have an important role in regulating GI motility. Our study further suggests that down-regulation or loss of function mutations of CARMN may contribute to CIPO in humans.
Material and Methods
Generation of Carmn global KO mice
The Carmn KO/GFP KI reporter mice were generated by insertion with a promoterless, reversed splicing acceptor/membrane-bound GFP gene trap cassette, which is flanked by 2 pairs of oppositely orientated Lox2272 sites and LoxP sites into intron 2 of the Carmn gene (referred to as conditional PFG allele) as we recently reported 26. Upon Cre-mediated deletion of exon 2 and inversion of the reversed GFP cassette (PFG to GFP), the inserted splicing acceptor will prematurely terminate Carmn expression by splicing exon 1 to the GFP cassette while turning on GFP expression under control of the endogenous Carmn promoter (referred to as GFP, or KO allele). GFP expression therefore faithfully displays endogenous Carmn expression in vivo while disrupting Carmn expression. To generate Carmn global KO (gKO) mice, we first crossed CarmnPFG/PFG female mice with male mice ubiquitously expressing Cre (CMV-Cre, JAX, cat. #: 006054) 29 under the control of a human cytomegalovirus minimal promoter. The resultant heterozygous offspring mice were then intercrossed to obtain CarmnGFP/GFP (gKO) mice, littermate WT (control) and CarmnGFP/WT (Het) mice. Both male and female mice of the cohort of gKO and control mice are used.
Generation of SMC-specific inducible Carmn KO mice
To generate SMC-specific Carmn inducible KO (iKO) mice, we first generated Myh11-CreERT2+; CarmnPFG/WT mouse line, by crossing CarmnPFG/PFG female mice with male mice expressing tamoxifen-inducible Cre driven by the SM-specific Myh11 gene promoter (Myh11-CreERT2+) 30. Subsequently Myh11-CreERT2+; CarmnPFG/WT male mice were bred with CarmnPFG/PFG female mice to generate Myh11-CreERT2+; CarmnPFG/PFG and Myh11-CreERT2+; CarmnPFG/WT mice. The dual fluorescence reporter mTmG (membrane-targeted tandem dimer Tomato, mTomato or mT; membrane-targeted green fluorescent protein, mGFP or mG) mice (Stock No: 007676) 31 were purchased from the Jackson Laboratory. To generate the control mice, Myh11-CreERT2+ male mice were bred with homozygous mTmG+/+ female mice. The resulted Myh11-CreERT2+; mTmG+/- SMC-lineage tracing mice were used as control to exclude potential cytotoxicity caused by ectopic expression of GFP and Cre in SMCs of Carmn iKO mice. To delete Carmn specifically in SMCs, 8-10 weeks old male Myh11-CreERT2+; mTmG+/- (Control), Myh11-CreERT2+; CarmnPFG/PFG (iKO) and Myh11-CreERT2+; CarmnPFG/WT (iHet) mice were intraperitoneally injected with tamoxifen (1mg/mouse/IP) for 2 rounds of 5 days each, with a 2 days’ break in between. Only male mice are used in this study because Myh11-CreERT2+ transgene is only located in the Y chromosome 30. All mice used in this study are maintained on a C57BL/6J background. The primers for mouse genotyping are listed in Supplemental Table 1. The use of experimental animals has been approved by the Institutional Animal Care and Use Committee and Biosafety committee at Augusta University in accordance with NIH guidelines.
Statistical Analysis
GraphPad Prism (version 9.2.0) was used for the statistical analysis. All data are expressed as mean ± SEM of at least 3 independent experiments. Tests used for statistical significance evaluations are specified in figure legends. An unpaired 2-tailed t test was used for data involving 2 groups only. Two-way analysis of variance was used for data involving >2 groups. Values of p < 0.05 were considered statistically significant, except for analysis of omics data, which used false discovery rate (FDR)-adjusted p < 0.05 as the threshold for statistical significance.
A Detailed description of methods is provided in the Online Supplemental Material.
Results
Identification of CARMN as a SMC-specific lncRNA in the human and mouse gut
To begin to screen for potentially important lncRNAs in GI tract, we first identified the most abundant lncRNAs that are expressed in human colon and small intestine tissues by analyzing the publicly available GTEx database 32. This analysis identified the top 10 highly enriched lncRNAs in adult human gut tissues (Figure 1A). To further examine the expression of these lncRNAs at a single cell level, we re-analyzed the single-cell transcriptome dataset of 62,849 cells isolated from duodenum, jejunum, ileum and colon of 6-10 weeks’ post-conception embryonic human gut (https://www.gutcellatlas.org) 33. Based on the gene expression signature of each cell, cells were divided into 45 clusters from 24 cell types (Figure 1B and Supplemental Figure 1A). UMAP (Uniform Manifold Approximation and Projection) visualization of gene expression revealed that among the top 10 most abundant lncRNAs, CARMN is the most restrictedly expressed lncRNA in SMCs, whose expression pattern is almost identical to that of the canonical SMC markers, such as MYH11, MYLK, ACTA2, LMOD1, TAGLN and CNN1 and the SMC-specific transcription cofactor MYOCD 27, 28, 34 (Figure 1C-D and Supplemental Figure1B). In addition to SMCs, CARMN is also expressed in myofibroblasts and pericytes, both of which exhibit some extent of SMC features (Figure 1D). To examine the expression of CARMN in adult human colon tissues, we next de novo analyzed the scRNA-seq dataset that was generated from 16 adult human colons (GSE156905) 35. UMAP visualization revealed that, the expression of CARMN is specific in SMCs, paralleling with the expression of well-known SM-specific markers MYH11 and MYLK (Figure 1E and Supplemental Figure 1C). Although bulk RNA-seq showed the lncRNAs MALAT1, NEAT1 and PGM5-AS1 are highly expressed in adult human gut (Figure 1A), none of them display a SMC-specific expression pattern as CARMN as revealed by the scRNA-seq analysis (Figure 1C & E). Furthermore, re-analyzing scRNA-seq dataset of adult mouse ileum and colon revealed that Carmn is restricted to SMCs, similar to the expression pattern of the well-known SMC-markers Myh11 and Mylk and the SMC-specific transcription co-factor Myocd (Supplemental Figure 1D).
To validate and visualize the expression of Carmn in mice, we generated a Carmn KO/GFP KI mouse model with insertion of a promoterless, reversed splicing acceptor/membrane-bound GFP gene trap cassette, which is flanked by two pairs of oppositely orientated Lox2272 sites and LoxP sites into the intron between exon 2 and 3 in mouse Carmn gene locus (PFG allele) 26 (Figure 1F). Cre-mediated recombination leads to inversion of the reversed GFP cassette (PFG to GFP, KO allele) and deletion of Carmn exon 2. Hence, Carmn transcription will be prematurely terminated by splicing of exon 1 to the splice acceptor of the GFP cassette, resulting in GFP expression driven by the endogenous Carmn gene promoter. To trace Carmn expression in vivo, we crossed the female CarmnPFG/WT (CarmnP/W) mice with male mice expressing global Cre to invert the GFP cassette (Figure 1G). Although ubiquitous Cre recombines the PFG allele globally, data from the direct visualization of GFP fluorescence and the immune-fluorescence staining of the SM-marker MYH11 in K/W mouse colon tissue revealed that Carmn specifically expresses in MYH11 positive SMCs (Figure 1H). Similar findings were also observed in the section of jejunum from the K/W mouse (Supplemental Figure 1E). Taken together, data from these unbiased in silico analysis and from the novel Carmn KI GFP reporter mouse model demonstrate that CARMN is a lncRNA abundantly and specifically expressed in visceral SMCs in both human and mouse gut.
Global deletion of Carmn in mice results in premature death due to the lethal colonic pseudo-obstruction
We next sought to explore functional role of Carmn in vivo by generating Carmn global KO (gKO) mice via intercrossing the female K/W and male K/W mice. The obtained WT and heterozygotes (Het) littermates serve as controls (Figure 2A). During the course of intercrossing Carmn Het mice, we unexpectedly observed approximate 60% of pregnant dams develop dystocia phenotype, suggesting haploinsufficiency of Carmn causes an impotency of uterine contraction to ensure successful parturition (Figure 2B-C). Carmn gKO mice were born at expected Mendelian ratio without gross abnormalities. However, on the weaning day of postnatal day 21, the gKO mice showed unhealthy appearance and smaller body size (Figure 2D) with significantly lower body weight compared to control mice (Figure 2E). Unexpectedly, after weaning and switching to the normal chow diet, approximate 60% of gKO mice exhibited an acute lethality within 1 week with all remaining gKO mice dying within 15 months. The gKO mice displayed an enlarged abdomen caused by dilated duodenum, jejunum and colon tracts due to infilled air and accumulation of stool (Figure 2G). Photograph of the isolated GI tracts further revealed that caecum and colon segments are the major distended parts with accumulation of feces in gKO mice (Figure 2H), suggesting a lethal colonic pseudo-obstruction. No apparent phenotype was observed in bladder, despite the high expression of Carmn in this tissue (Figure 2G). Histological analysis by HE staining further revealed severe dilation of colon and thinning muscle layers of gKO mice compared to control mice (Figure 2I-J). Similar results were also observed in the jejunum (Supplemental Figure 2A-B). Moreover, ultrastructural examination by transmission electron microscopy revealed that Carmn KO SMCs in both colon and jejunum display markedly abnormal structures, including damaged endoplasmic reticulum with dramatically dilated lumen and massive membrane laminated vacuoles with embedded membranes and organelles such as mitochondria, suggesting a stressed and degenerative phenotype of SMCs (Figure 2I and Supplemental Figure 2C). Taken together, these data demonstrate that global deletion of Carmn causes lethal colonic pseudo-obstruction and dystocia, and suggest that Carmn is indispensable for the GI and uterine function.
Inducible SMC-specific deletion of Carmn in adult mice recapitulates the lethal phenotype of Carmn gKO mice
To uncover the functional role of Carmn specifically in adult SMCs, we generated Carmn inducible SMC-specific KO mice by crossing tamoxifen (TAM)-inducible transgenic mice carrying SMC-specific gene Myh11 promoter driven CreERT2 (Myh11-CreERT2) with CarmnPFG/PFG mice (referred to as iKO mice hereafter, Figure 3A). To exclude potential cytotoxicity caused by ectopic expression of GFP and Cre in SMCs of SM-specific Carmn iKO mice, we used the SM-lineage tracing mice (referred to as WT control mice hereafter) as control in which Cre and membrane-tagged GFP specifically express in SMCs as iKO mice by crossing Myh11-CreERT2 mice with Rosa26-mTmG dual fluorescence reporter mice (Figure 3B). We also included a cohort of mice in which a single Carmn allele is specifically deleted in adult SMCs as an additional control (referred to as iHet hereafter). TAM was peritoneally injected into these 3 cohorts of 8-10 weeks old male mice for 10 times and then the mouse body weight was monitored (Figure 3C). We found at day 62 post the first TAM injection, the body weight of iKO mice is significantly reduced compared to the WT and iHet mice. We observed that iKO mice start to display mortality at day 33, and none of them can survival beyond 130 days after the first TAM injection (Figure 3D). Dissecting the iKO mice at day 30 post the 1st TAM injection revealed almost identical the pathological events observed in gKO mice, including drastic distention of GI tract that is infilled with abundant air and feces, especially in the cecum and colon segments (Figure 3E-F). Histological analysis of colon and jejunum of iKO mice showed dilation and thinner muscular layer compared to the WT and iHet control mice, similar findings as gKO mice (Figure 3G-H and Supplemental Figure 3A-B). Ultrastructural images revealed a similar severe degenerative phenotype of SMCs in both colon and jejunum of iKO mice as observed in gKO mice (Figure 3I and Supplemental Figure 3C). Taken together, these data demonstrate that deletion of Carmn specifically in postnatal SMCs of adult mice recapitulates the lethal colonic pseudo-obstruction phenotype occurred in Carmn gKO mice, leading to premature lethality.
Carmn deficiency impairs GI motility and colonic contractility in mice
Previous studies have shown that the impairment of visceral SMC contraction in the GI tract is responsible for colonic pseudo-obstruction associated with CIPO in humans 5, 6, 8, 9. To directly assess the contractile function of GI tract of gKO and iKO mice in vivo, the whole-gut transit time from food intake to excretion were measured as a functional readout of GI motility. We found that the whole-gut transit time was significantly increased in both Carmn KO mouse models (gKO and iKO) compared to their respective control mice (Figure 4A-B), indicating that Carmn is indispensable for the GI motility. The impairment of gut motility in iKO mice resulted in larger stool size in diameter and darker color stool (Figure 4C-D). As the phenotype mainly developed in the colon segment, next we measured the colonic contractile activity ex vivo using myography. We found that the spontaneous colonic contractile activity in both Carmn gKO and iKO mice is significantly attenuated compared to that of control mice (Figure 4E-F and Supplemental 4A-B). Moreover, the amplitude of colonic contraction elicited by stimulation of KCl (60 mM) or the muscarinic agonist Carbachol (1 µM Cch), both of which can initiate smooth muscle contraction through MYLK-dependent phosphorylation of myosin regulatory light chain (RLC) 36, 37, was significantly impaired in Carmn gKO and iKO mice compared to their respective control mice (Figure 4G-N and Supplemental Figure 4C-F). In summary, these data indicate that Carmn is indispensable for GI motility and colonic contractility.
Carmn deficiency induces visceral SMC de-differentiation
To understand the molecular mechanism that underlies Carmn deficiency-induced colonic pseudo-obstruction phenotype, bulk RNA-sequencing (RNA-seq) was performed on the muscularis isolated from colon and jejunum tissues of Carmn iKO and control mice (Figure 5A). Data from the RNA-seq analysis demonstrated that Carmn iKO resulted in 318 significantly down-regulated genes and 243 up-regulated genes in colon muscularis (Figure 5B and Supplemental Table 2), while 561 significantly down-regulated genes and 635 up-regulated genes in jejunum muscularis, respectively (Figure 5C and Supplemental Table 2). Gene ontology (GO) analysis of down-regulated genes in colon muscularis revealed that Carmn deficiency significantly attenuates the genes involved in maintaining the muscle homeostasis, such as muscle contraction, muscle tissue development and muscle cell differentiation (Figure 5D & F). In contrast, up-regulated genes in colon muscularis are significantly over-represented in pathways responding to tissue repair such as matrix remodeling, wound healing, and inflammatory process including response to lipopolysaccharide (Figure 5E-F). In Carmn KO jejunum muscularis, the down-regulated genes are involved in metabolic processes (Supplemental 5A), while the up-regulated genes are associated with immune cell infiltration and inflammatory reaction (Supplemental 5B), a similar finding observed in iKO colon tissues. To better understand the potential genes directly regulated by Carmn, we further compared the overlapping differentially expressed genes in both colon and jejunum muscularis. The results showed by the Venn diagram demonstrated that there are 136 overlapping downregulated and 128 overlapping upregulated genes between Carmn iKO colon and jejunum muscularis, respectively (Figure 5G). Strikingly, GO analysis of the overlapping down-regulated genes revealed that these over-represented genes belonging to muscle contraction and development, almost identical to the results observed in colon. These data collectively suggest that the down-regulation of these contractile genes is a primary cause for the hypomotility of GI tracts observed in Carmn deficient mice (Figure 5H and Supplemental Figure C). Consistently, GO analysis of the up-regulated genes in common revealed the shared pathways are involved in inflammatory response and matrix remodeling (Supplemental Figure 5D-E), implicating a secondary effect to ileus paralytics induced by deficiency of Carmn. To further pinpoint the key targets of Camrn with critical roles in maintaining colonic SMC homeostasis, we analyzed the common genes involved in multiple smooth muscle homeostasis signaling pathways. This analysis revealed that Mylk is the only overlapping gene, which has been shown to be the principal regulator of the myosin II molecular motor in the initiation of smooth muscle contraction essential for normal gastrointestinal motility (Figure 5I) 38. Taken together, the integrative analysis of the bulk RNA-seq datasets on both colon and jejunum muscular layers suggest that Carmn deficiency disrupts the homeostatic expression of muscle contractile genes such as Mylk while promoting expression of genes involved in response to tissue injury and inflammation.
Carmn deficiency disrupts homeostatic SMC contractile gene program and compromises the cell-cell communication in the colonic muscularis
To obtain a comprehensive understanding of changes in transcriptome of Carmn deficient SMCs at the single cell level, we isolated cell nuclei from colonic muscularis of Carmn iKO and control mice and performed snRNA-seq (Supplemental Figure 6A). UMAP analysis of 5,072 nuclei from both control and Carmn iKO samples revealed 17 distinct clusters with 13 cell types as defined by the cell type-specific gene expression profile (Figure 6A and Supplemental Figure 6B-E). As expected, Carmn was specifically detected in SMCs positive for the well-known SMC marker genes Myh11 and Mylk and the SMC-specifically expressed transcription co-factor Myocd (Figure 6A and Supplemental Figure 6E). Cell composition analysis revealed that compared to control group, the percentage of SMCs, glial cells, Ctfr+ cells and Goblet cells is decreased, while the percentage of fibroblast, fibroblast-like, macrophage and T cells is increased in Carmn iKO samples (Figure 6B). The decreased SMCs, Ctfr+ cells and Goblet cells may contribute to the imbalance of colon muscularis homeostasis 39, while the increased fibroblast or fibroblast-like, macrophage and T cells may be responsible for the tissue repair and inflammatory response as secondary to the severe impairment of GI motility observed in Carmn iKO mice 40, 41. To unravel the transcriptome changes of SMCs following Carmn deletion, we next conducted a differential analysis specifically in SMC clusters. This analysis identified 123 down-regulated and 265 up-regulated genes, respectively, in the SMCs of Carmn iKO compared to control (Figure 6C and Supplemental Table 3). GO analysis of the down-regulated genes in iKO SMCs revealed that Carmn deficiency significantly attenuates the genes involved in regulating muscle contraction and synapse organization (Supplemental Figure 6F), while the top 10 enriched pathways of the up-regulated genes are mainly responsible for extracellular matrix organization and adhesion (Supplemental Figure 6G), pathways tightly related with tissue repair 42. Consistent with the bulk RNA-seq data, genes related to muscle contraction including Mylk and Chrm2 were among the significantly down-regulated genes while the matrix genes such as Ccn2, Fn1 and Eln were significantly increased in the iKO SMCs, indicating a de-differentiation status of iKO SMCs (Supplemental Figure 6H). Analysis of the differentially expressed genes (DEGs) in colon muscularis revealed by bulk RNA-seq and in SMC clusters revealed by snRNA-seq uncovered 29 co-downregulated genes and 34 co-upregulated genes, respectively (Figure 6D). Intriguingly, GO analysis of these co-downregulated and co-upregulated genes showed that these genes are involved in the pathways regulating muscle contraction and extracellular matrix organization, respectively (Figure 6E-F). Taken together, results from the colon muscularis bulk RNA-seq and snRNA-seq unambiguously demonstrate that deletion of Carmn in SMCs leads to a de-differentiation phenotype of SMCs by down-regulating contractile genes while up-regulating genes involved in extracellular matrix remodeling.
To define the sub-populations of SMCs in mouse colon in greater detail, we closely examined the transcriptome of the 1,529 SMCs identified in both iKO and control samples. Unsupervised Seurat-based clustering of those SMCs revealed 7 distinct sub-populations defined by cluster-specific marker genes (Figure 6G-H). Interestingly, we found that, compared to control WT SMCs, Carmn iKO SMCs contain a reduced percentage of sub-cluster 4 and a higher percentage of sub-cluster 6 (Figure 6I). Although all 7 SMC subpopulations are Myh11 positive (the 1st panel in the top row in Figure 6J), expression of Myh11 gene is lower whereas expression Mylk gene (the 2nd panel in the top row of Figure 6J) is nearly completely absent in the SMC sub-cluster 4 in Carmn iKO SMCs, compared to control. In contrast, the most significantly enriched 2 genes in the sub-cluster 4 SMCs but depleted in iKO SMCs are Nrg3 (Neuregulin 3, a ligand binding to ERBB-family receptors) and Rbfox1 (RNA binding protein, fox-1 homolog) (Figure 6J), both are neuron-enriched genes (Supplemental Figure 6E). Notably, NRG3 gene mutations in enteric neuron system cause Hirschsprung’s disease (HSCR) in humans 43, a birth defect that characterized with bowl hypoperistalsis. In contrast, SMCs in sub-cluster 6 are the only subset of SMCs positive of the fibroblast-specific gene Fbn2 (the 1st panel in the bottom row of Figure 6J). Furthermore, the most significantly upregulated genes in iKO SMC sub-cluster 6 are Ccn1, Ccn2 and Thbs1, all of which are fibroblast-specific genes and are key mediators for smooth muscle extracellular matrix homeostasis (Figure 6J) 44. To better understand the transcriptome characteristics of the SMC sub-cluster 4 and 6, GO analysis of the enriched genes in both sub-clusters showed that the enriched genes are involved in the pathways regulating axonogenesis and matrix remodeling, respectively (Supplemental Figure 7A-B). Therefore, we refer to the sub-cluster 4 as neuron-like SMCs and sub-cluster 6 as fibroblast-like SMCs. To investigate the transition of SMC phenotypes, we applied the Monocle 2 algorithm to perform the single-cell pseudotime trajectory analysis 45-47. This analysis revealed SMCs form a continuous transition and progressively branched into two trajectories, ending with SMC sub-cluster 4 and 6, respectively (Figure 6K and Supplemental Figure 7C). Collectively, this comprehensive snRNA-seq analysis revealed that the SMCs in mouse colon muscularis can be divided into three groups: the normal contractile SMCs (clusters 1, 2, 3, 5, 7) on which Carmn deficiency has little effects, the novel neuron-like SMCs (sub-cluster 4) that is depleted in the Carmn iKO mice, and the novel fibroblast-like SMCs (sub-cluster 6) which is increased in the Carmn iKO SMCs. Thus, we hypothesize that the higher percentage of the fibroblast-like de-differentiated SMCs are responsible for tissue repair, while the reduced percentage of the neuron-like SMCs in Carmn iKO mice may impair the neurogenic motility that synergistically communicates with other cell types 48, 49. To test this hypothesis, we performed cell-cell communication among these cells using CellChat 50. This analysis revealed that, compared with WT mice, neuron-like SMCs in Carmn iKO cells lost the cell-cell communication with neuron2, other contractile SMCs, glial cells, ICC and myenteric neurons (Figure 6L and Supplemental Figure 7D). The overall signaling patterns showed that compared to WT cells, Carmn iKO SMCs gain cellular signaling involved in SMC de-differentiated, such as IGF, THBS and TGFb, which closely correlated with the increased percentage of fibroblast-like SMCs in Carmn iKO mice (Figure 6M). In contrast, neuron-like SMCs lost the NRXN, SEMA3 and NRG signals while gaining FN1 ligand-mediated signal (Figure 6M and Supplemental Figure 7E). Taken together, the data from the snRNA-seq analysis demonstrate that Carmn deficiency in SMCs not only compromises the myogenic motility by reducing expression of contractile genes essential for SMC contraction but also impairs neurogenic motility by disrupting the cell-cell connectivity in the colonic muscularis, collectively contributing to the colonic pseudo-obstruction phenotype of Carmn iKO mice.
Deletion of Carmn attenuates expression of genes essential for SMC contraction
Next we validated the bulk RNA-seq and snRNA-seq data by assessing the expression of contractile genes regulated by Carmn. Results from qRT-PCR and Western blotting assays demonstrated that MYLK is consistently down-regulated in both jejunum and colon muscularis of both gKO and iKO mice, compared to their respective control mice, suggesting that Mylk is a major target gene regulated by Carmn (Figure 7A-F and Supplemental Figure 8A-B). Carmn deficiency did not significantly affect the expression of the key transcription factor Srf and its cofactor Myocd (Figure 7A and D). IF staining further revealed that the MYLK immuno-fluorescence intensity in both colonic and intestinal muscular layers of both gKO and iKO mice are markedly weaker than that of controls (Figure 7G-H and Supplemental Figure 7C-D). Taken together, these data demonstrate that Carmn deficiency leads to decreased expression of the key SM-contractile gene Mylk that is essential for SM contraction, indicating that the lethal colonic pseudo-obstruction phenotype observed in Carmn KO mice is attributable to the decreased expression of Mylk gene in Carmn deficient SMCs.
CARMN is critical for contraction of human colonic SMCs by regulating MYLK gene expression
We next sought to explore the translational relevance of CARMN-mediated contractile function in humans by using human colonic smooth muscle cells (HuCoSMCs). To examine whether CARMN regulates contractile gene expression in human GI SMCs, HuCoSMCs were transfected with control or CARMN antisense oligonucleotide (ASO) for 48 hours. qRT-PCR and Western blot analysis revealed that silencing of CARMN significantly increases expression of the proliferative gene PCNA and the matrix gene CCN2 while attenuating the expression of almost all contractile genes examined except MYH11 at both the mRNA and protein levels, compared to silencing control cells (Figure 8A-C). Since MYLK-mediated phosphorylation of RLC is prerequisite for the initiation of SMC contraction, we examined KCl and Cch-mediated MYLK-pMLC signaling pathway by Western blotting in HuCoSMCs with or without CARMN. The results from these experiments consistently showed that knocking down CARMN effectively attenuates the MYLK expression, thereby blocking MYLK-mediated phosphorylation of RLC (pMLC) (Figure 8D-E). To determine whether depletion of CARMN causes a functional loss of contractile competence of SMCs, we performed collagen gel contraction assays using HuCoSMCs in the presence or absence of CARMN. Results from this assay showed an approximate 50% decrease in contractile activity in CARMN-depleted HuCoSMCs (Figure 8F-G). It has been shown that the binding of the transcriptional complex of MYOCD and SRF to CArG elements residing in the promoter regions of SMC contractile genes is an important mechanism for regulating SM-specific gene expression, thereby maintaining the SMC contractile phenotype 51. Moreover, our recent study showed that Carmn directly binds with MYOCD, but not SRF to potentiate MYOCD-mediated myogenic function 26. Since MYLK has been known as a SRF target gene 52 and its SMC-specific expression overlaps with CARMN (Figure 1D), we hypothesize that MYLK is a novel target of CARMN/MYOCD/SRF complex in visceral SMCs. To test this hypothesis, we performed luciferase assays using WT or CArG mutant of Mylk gene reporter with or without co-transfection of expression plasmids of Carmn and Myocd. We also performed a similar luciferase assay in Lmod1 gene reporter because we found depletion of CARMN leads to down-regulation of LMOD1 in HuCoSMCs (Figure 8A-C) and LMOD1 has been reported as an SRF/MYOCD-dependent gene and its mutation causes megacystis microcolon intestinal hypoperistalsis syndrome (MMIHS) in mouse and human 6. Data from these luciferase reporter assays in HuCoSMCs demonstrated that Carmn alone has no effect on the basal activity of Mylk and Lmod1 gene promoters but significantly enhances MYOCD-mediated transactivation activity in a CArG-dependent manner (Figure 8H). Taken together, these results suggest that Carmn is indispensable for visceral SMC contraction by maintaining cell-cell connectivity and regulating SRF/MYOCD complex-dependent MYLK and LMOD1 gene expression and MYLK-dependent contractile signaling in trans. Global deletion or SMC-specific deletion of Carmn in mice results in attenuated expression of MYLK and comprised cell-cell connectivity, leading to premature lethality due to the gut dysmotility-induced pseudo-obstruction (Figure 8I).
Discussion
CIPO is a rare but life-threatening disease characterized by severe intestinal dysmotility. Functional and histopathologic studies have shown that GI dysmotility in CIPO patients results from multiple mechanisms, including defects in neurons, SMCs or interstitial cells of Cajal (ICC) 53. Mutations affecting the expression and function of proteins of the contractile apparatus and cytoskeleton are the most common cause of myopathic CIPO 3. While lncRNAs have important epigenetic roles in SMC, their role in CIPO is poorly understood. Our study revealed a surprisingly critical role of a SMC-specific lncRNA, CARMN, in regulating visceral SMC contractility. Our findings raise the possibility that polymorphisms in the CARMN gene could lead to loss of expression and/or function, which possibly could contribute to some cases of myopathic CIPO.
Herein, we demonstrated that loss of Carmn expression in mice critically impairs colonic motility, disrupts colonic homeostasis and eventually results in lethal colonic pseudo-obstruction. To the best of our knowledge, our study is the first to show that a lncRNA is essential for maintaining visceral SMC contractility. The impact of loss of Carmn is seen with global deletion prior to development and SMC-specific inducible ablation in adulthood, with both mouse models developing a severe GI motility disorder with degenerative visceral myopathy, resembling SM dysfunction-mediated CIPO in humans. Future studies are warranted to examine whether there is a reduction in expression level or loss-of-function of mutations of CARMN in patients with CIPO. Furthermore, our comprehensive snRNA-seq analysis revealed that Carmn deficiency affects neurexins (NRXNs) and Neuregulins (NRGs)-mediated neuronal signaling. NRXNs and NRGs are synaptic cell-adhesion molecules that connect pre- and postsynaptic neurons at synapses, mediate trans-synaptic signaling, and shape neural network properties by specifying synaptic functions 54. How Carmn expression in SMC affects NRXN and NRG signaling remains undefined. Moreover, SNP mutations within NRXN are associated with susceptibility to Hirschsprung’s disease 55. We propose that CARMN may play a broader role, not only in SMCs per se, but also in regulating cell-cell communications that are crucial for coordinating GI motility.
We also found that pregnant Carmn heterozygous mice develop a dystocia phenotype, likely due to impaired uterine contractions. Notably, starting from day 21 when the pups were weaned to regular chow diet from milk feeding, 60% of Carmn global knockout mice succumbed to the diet switch within 2 weeks, suggesting that Carmn is required for force overload response in both uterine and GI tract SMCs. A limitation of our study is that we were unable to assess the specific functional role of Carmn in postnatal uterine SMCs because the inducible Myh11-CreERT2 transgene we used is only active in male mice 30. A new SMC-specific Cre mouse model in which Cre is active in both sexes of mice is required to address this question.
During the preparation of this study, Vacante et al reported on the use of CRISPR/Cas9-mediated genome editing to constitutively knockout Carmn in mice 27. Using this mouse model, they found that loss of Carmn reduced the expression of SMC contractile genes in the mouse aorta, similar to our findings in the mouse GI tract. However, this study did not report a gastrointestinal phenotype in Carmn KO mice. Recently, Ni et al used antisense GapmeRs specific to Carmn to knockdown Carmn in mice 28. They showed that despite the reduced efficacy of antisense GapmeRs, achieving approximately 50% efficiency in knocking down Carmn, they significantly decreased expression of SM-specific contractile genes in the mouse aorta. However, this study also did not report any GI phenotype in mice treated with Carmn antisense GapmeRs. The reasons for these apparent discrepancies are not clear although we can speculate that they may relate to a focus on the vasculature rather than the GI tract, residual or altered function of the Carmn in the CRISPR knockout, differences in the genetic background of mice and/or the low efficiency of knockdown in vivo by GapmeRs.
Previous studies have shown that both the SRF transcription factor and its cofactor, MYOCD are necessary for the development and maintenance of the visceral SMC contractile phenotype. Targeted deletion of either gene in SMCs of adult mice resulted in severe gut dysmotility, characterized by weak peristalsis and dilation of the digestive tract 11, 12. Notably, the SMC-specific inducible Carmn KO mice in our study exhibited remarkable similarities in the gross, microscopic and molecular findings of the SM-specific inducible Myocd and Srf deletion 10, 11 ,12, 56. This is not completely surprising given that Carmn was found to bind to and potentiate MYOCD/SRF complex function as we previously reported in vascular SMCs 26, but the extent of regulation by Carmn is remarkable. Our findings support a role for CARMN as a critical binding partner of the MYOCD/SRF complex that has a vitally important role in maintaining the function of visceral SMCs. However, the lethal GI myopathy phenotype resulting from Myh11 promoter-driven Cre has precluded us from assessing a potential role of Carmn in vascular SMCs in vivo. To address this important question, future studies will need to be conducted by using a vascular SMC-specific Cre mouse model.
Using bulk RNA-seq and snRNA-seq approaches to investigate mechanisms, we identified that Carmn deficiency disrupts the homeostasis of the colon and impairs colonic motility mainly through down-regulation of an array of SMC contractile genes especially the Mylk gene. Consistent with previous studies showing that MYLK is an SRF/MYOCD target gene 52, we found that Carmn-induced transcriptional activation of the MYLK gene promoter is CArG-dependent. It is well accepted that SMC regulatory light chain (RLC) phosphorylation by calcium/calmodulin-dependent MLCK is a prerequisite for SM contraction and critical for maintaining the physiological movement of hollow organs 38, 57-59. We found that deletion of Carmn in both murine visceral SMCs in vivo and human colonic SMCs in vitro not only significantly down-regulated Mylk gene expression, but also attenuated KCl and Cch-induced RLC phosphorylation. Together with the previous study showing that inducible SMC-specific deletion of Mylk in mouse results in markedly reduced RLC phosphorylation, gut motility and premature lethality, our data suggests MYLK, as a CARMN target gene, at least in part, is responsible for mediating CARMN function in visceral SMCs. However, we cannot rule out the possibility that other down-regulated genes such as Dmpk and Chrm2 may play additional functional roles since these genes are also SRF/MYOCD targets genes and have been shown to be important for smooth muscle contraction 56, 60. Interestingly, a recent computational study suggested that Carmn could regulate the expression of downstream genes including MYLK via formation of RNA-DNA-DNA triplexes 61. Although this idea is provocative, confirmation on whether Carmn regulates the expression of SMC-contractile genes via RNA-DNA interactions needs to be tested in wet lab experiments. Based on data showing a reduction in a novel population of neuron-like SMCs in Carmn KO mice, our snRNA-seq data further suggest that Carmn regulates cellular connectivity between SMCs and other cell types. We propose that cell-cell decoupling may partly account for the decreased contractility observed in the GI tracts of Carmn KO mice. Collectively, our data show that SM-specific lncRNA Carmn plays a critical role in maintaining visceral SMC homeostasis that is likely to occur through multiple mechanisms.
CARMN is an evolutionarily conserved smooth muscle cell-enriched lncRNA that was initially identified as the host gene for MIR143/145 cluster 26, which is the best-characterized microRNA pair regulating SMC differentiation and phenotypic modulation 62-65. Using bulk RNA-seq data we found that Carmn deletion robustly decreases the expression of miR143/145 cluster in colon tissues (Supplemental Table 2). However, previous studies demonstrated that germline deletion of the miR143/145 cluster in mice resulted in no overt developmental defects at baseline, although it did contribute to a defect in epithelial regeneration in response to intestinal injury 66. In the current study, we found that both global deletion and SMC-specific deletion of Carmn resulted in a lethal colonic pseudo-obstruction phenotype, which is very different from the negligible phenotype observed in mice absent the miR143/145 cluster. Together, these results indicate that the ability of Carmn to maintain colonic SMC homeostasis is independent of the miR143/145 cluster although Carmn deletion can affect miR143/145 expression. This finding is also consistent with recent studies showing that CARMN can regulate the plasticity of vascular SMC phenotypes independent of the MIR143/145 cluster in vitro 26-28. However, miR145 was reported to strongly express in zebrafish gut smooth muscle and loss of miR145 in zebrafish resulted in deficits of smooth muscle maturation and contraction 67, 68, a similar finding to our mouse study. Given that the putative mouse Carmn ortholog was not annotated and identified in zebrafish until in our recent report 26 and that the Carmn gene is localized within the miR143/145 gene locus, we speculate that the GI phenotype observed in zebrafish with knockout of miR145 is likely due to an accidental deletion of Carmn. To confirm this, a Carmn-specific knockout/knockdown zebrafish model should be generated to directly address the distinct roles of Carmn versus miR145.
In summary, our study has uncovered a novel role of the lncRNA, CARMN, as a major regulator of gastrointestinal motility and viability through actions in visceral SMCs that serve to maintain contractile function and intercellular connectivity. A greater understanding of the GI roles of lncRNAs such as CARMN adds to our tool kit in helping to understand and diagnose human visceral myopathy diseases such as CIPO.
What You Need to Know
Background and Context
Visceral smooth muscle cells (SMCs) are critical for regulating motility of gastrointestinal (GI) tract. Loss of GI motility causes several motility diseases including chronic intestinal pseudo-obstruction (CIPO) in humans. The functional role of long non-coding RNAs (lncRNAs) in regulating GI motility is largely unexplored.
New Findings
Carmn is a SMC-specific lncRNA that is conserved in humans and mice. Carmn deficiency in mice results in premature lethality due to colonic pseudo-obstruction by compromising intestinal motility through reduced expression of contractile genes and impairing cell-cell connectivity in the colonic muscularis.
Limitations
Study limitations include having not yet validated the sub-population of neuron-like SMCs in mice and lack of evidence to show the decreased expression of CARMN in human CIPO patients’ samples.
Impact
Carmn is indispensable for maintaining GI SMC contractile function in mice, but also it suggests that loss of function of CARMN may contribute to human visceral myopathy such as CIPO.
Footnotes
Conflict of interest statement: The authors have no conflict of interest to disclose.
Data availability statement: The original data, analytic methods, and materials will be made available to other researchers upon reasonable requests.
Source of funding: The work at the J.Z. laboratory is supported by grants from the National Heart, Lung, and Blood Institute, NIH (HL157568 and HL149995). J.Z. is a recipient of Established Investigator Award (17EIA33460468) and Transformational Project Award (19TPA34910181) from American Heart Association. X.H. and K.D. are supported by a postdoctoral fellowship (836341) and a Career Development Award (938570), respectively, from American Heart Association.