(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .
Filamin C is Essential for mammalian myocardial integrity [1]
['Tongbin Wu', 'Department Of Medicine', 'University Of California San Diego', 'La Jolla', 'California', 'United States Of America', 'Yujun Xu', 'Lunfeng Zhang', 'Department Of Pharmacology', 'Skaggs School Of Pharmacy']
Date: 2023-02
FLNC, encoding filamin C, is one of the most mutated genes in dilated and hypertrophic cardiomyopathy. However, the precise role of filamin C in mammalian heart remains unclear. In this study, we demonstrated Flnc global (Flnc gKO ) and cardiomyocyte-specific knockout (Flnc cKO ) mice died in utero from severely ruptured ventricular myocardium, indicating filamin C is required to maintain the structural integrity of myocardium in the mammalian heart. Contrary to the common belief that filamin C acts as an integrin inactivator, we observed attenuated activation of β1 integrin specifically in the myocardium of Flnc gKO mice. Although deleting β1 integrin from cardiomyocytes did not recapitulate the heart rupture phenotype in Flnc knockout mice, deleting both β1 integrin and filamin C from cardiomyocytes resulted in much more severe heart ruptures than deleting filamin C alone. Our results demonstrated that filamin C works in concert with β1 integrin to maintain the structural integrity of myocardium during mammalian heart development.
The precise role of filamin C in mammalian heart development had not been determined, in part due to the lack of cardiac phenotypes in previously described Flnc knockout mice, which still had truncated filamin C expressed in the heart. In this study, we analyzed a true Flnc knockout mouse line, in which filamin C protein was completely ablated. Flnc knockout mice developed massive ruptures in their myocardium but not in the endocardium, suggesting filamin C is essential for the structural integrity of myocardium. On the other hand, we did not find overt abnormalities of sarcomeric structure in cardiomyocytes of Flnc knockout mice, indicating that filamin C is likely not required for sarcomeric assembly as previously observed in FLNC null iPSC-CMs. Moreover, contrary to the dogma that filamins are integrin inactivators, we found that filamin C plays an unexpected role in integrin activation and works in concert with β1 integrin to ensure the structural integrity of the myocardium.
To this end, we generated a floxed Flnc mouse line [ 19 ] and analyzed Flnc global knockout (Flnc gKO ) and Flnc cardiomyocyte-specific knockout (Flnc cKO ) mice. Both Flnc gKO and Flnc cKO mice died before embryonic day (E) 11.5 from severely ruptured ventricular myocardium, indicating filamin C is required to maintain the structural integrity of myocardium in mammalian heart. By immunofluorescence analyses, we found downregulation of key extracellular matrix (ECM) proteins which might partially explain the heart rupture phenotype. Surprisingly, we did not observe obvious sarcomere disarray in cardiomyocytes of Flnc gKO mice, suggesting that filamin C is not required for sarcomere assembly in vivo. Interestingly, we observed attenuated activation of β1 integrin specifically in myocardium of Flnc gKO mice. However, deleting β1 integrin from cardiomyocytes did not recapitulate the heart rupture phenotype in Flnc knockout mice, whereas deleting both β1 integrin and filamin C from cardiomyocytes resulted in larger heart ruptures. Our results demonstrated that filamin C works in concert with β1 integrin to maintain the structural integrity of myocardium during mammalian heart development.
Several studies sought to elucidate the function of filamin C in heart. A nonsense mutation identified in the teleost fish medaka causes myocardial rupture in heart ventricles, suggesting that filamin C is involved in maintenance of structural integrity of cardiac muscle [ 16 ]. Ablating filamin C in human induced pluripotent stem cell–derived cardiomyocytes (hiPSC-CMs) led to sarcomere disarray [ 17 ]. Surprisingly, mice with homozygous deletion of the last 8 exons of Flnc did not show overt cardiac phenotypes [ 18 ]. However, these Flnc knockout mice still expressed a truncated form of filamin C protein in the heart [ 18 ]. Thus, a bona fide Flnc knockout mouse model is required to study the precise role of filamin C in mammalian heart.
Filamins (FLNA, FLNB, FLNC) are large actin-binding and -crosslinking dimeric proteins, with each subunit ranging from 240 to 280 kDa [ 5 ]. Filamin C (FLNC) is predominantly expressed in striated muscle tissues [ 6 ], and is localized to the Z-disc [ 7 ], intercalated disc (ICD) [ 8 ], and costamere [ 6 ]. Filamin C contains an N-terminal actin-binding domain (ABD) and 24 C-terminal immunoglobulin (Ig)-like domains [ 5 ], which are responsible for protein dimerization and interacting with myotilin and FATZ-1 at Z-discs [ 9 , 10 ]. C-terminal Ig-like domains also interact with β1 integrin [ 11 ] and sarcoglycans [ 6 ] at the costamere, a structural and functional component that bridges and strengthens the connection of the Z-discs to the sarcolemma [ 12 ]. Thus, filamin C is proposed to serve as a link between myofibrils and sarcolemma [ 7 , 13 , 14 ]. In vitro studies have demonstrated filamins inactivate integrin by competing with talin for binding to the cytoplasmic domain of the integrin β subunit [ 15 ]. However, the functional consequences of loss of filamins, especially filamin C, on integrin activation and its potential role in filamin C-related cardiomyopathy has not been explored in vivo.
To further explore the genetic interaction of Flnc and Itgb1, and its contribution to myocardial wall integrity, we generated cardiomyocyte-specific Flnc/Itgb1 double knockout mice (Flnc/Itgb1 dcKO ) ( Fig 5B ). Strikingly, Flnc/Itgb1 dcKO mice had much more severe heart rupturing which led to myocardial disintegration ( Fig 5C ), compared with Flnc cKO mice. Consequently, the endocardium of Flnc/Itgb1 dcKO mice became inflated, presumably due to the lack of mechanical support from the myocardium ( Fig 5C ). Flnc/Itgb1 dcKO mice also had larger pericardial effusions and more pronounced overall growth retardation ( Fig 5C ). Taken together, our findings suggest that filamin C works in concert with β1 integrin to maintain the structural integrity of myocardium during mammalian heart development.
To assess whether decreased activation of β1 integrin in cardiomyocytes contributed to the heart rupture phenotype observed in Flnc gKO mice, we generated cardiomyocyte-specific Itgb1 (encoding β1 integrin) knockout mice using Xmlc2 Cre (Itgb1 cKO ) and compared them with cardiomyocyte-specific Flnc knockout mice (Flnc cKO ) ( Fig 5B ). Consistent with observations in Flnc gKO mice, Flnc cKO mice had obvious ruptures in their myocardium at E10.5 ( Fig 5C ). Although we previously reported that Itgb1 cKO develop heart ruptures at E14.5 [ 31 ]. Itgb1 cKO hearts did not show rupture at E10.5 ( Fig 5C ). These findings indicated that decreased β1 integrin activation in Flnc gKO mice may only partially account for the myocardial rupture phenotype.
Filamins are known integrin inactivators and abnormal activation of β1 integrin can lead to impaired cell proliferation, differentiation and migration [ 15 ]. To determine whether β1 integrin, a dominant integrin β isoform in cardiomyocytes [ 31 ], was ectopically activated in Flnc gKO mice, we performed IF using an antibody (9EG7) [ 32 ] against the activated ligand-bound conformation of β1 integrin and an antibody against total β1 integrin. Surprisingly, we found that activated β1 integrin was reduced in the myocardium of Flnc gKO hearts without changes in the total β1 integrin expression ( Fig 5A ). In contrast, both activated and total β1 integrin were upregulated in regions of endocardium proximal to rupture sites of Flnc gKO hearts ( Fig 5A ).
To investigate whether the heart rupturing was caused by cardiomyocyte hypoplasia resulting from decreased cardiomyocyte proliferation and/or increased cardiomyocyte apoptosis, we measured cardiomyocyte proliferation and apoptosis rates in E8.5 to E10.5 Flnc gKO hearts and littermate controls. Although cardiomyocyte proliferation was markedly reduced and cardiomyocyte apoptosis was increased in E10.5 Flnc gKO hearts compared with controls, both parameters were indistinguishable between Flnc gKO and controls from E8.5 to E9.5 ( S2B–S2C Fig ). Because the heart rupture phenotype was already evident in E9.5 Flnc gKO hearts, these findings indicated that heart rupturing was not caused by cardiomyocyte hypoplasia.
( A ) Targeting strategy for generating Flnc knockout mice. Global Flnc knockout mice were generated by crossing Flnc floxed mice (Flnc fl/fl ) with Sox2 Cre mice, while cardiomyocyte-specific Flnc knockout mice were generated by crossing Flnc fl/fl mice with Xmlc2 Cre mice or cTnT Cre mice. After the deletion of exon 9–13 of the Flnc gene by Cre recombinase, a premature termination codon (PTC) will be introduced to exon 14 and trigger non-sense mediated mRNA decay (NMD) of Flnc mRNA. ( B ) Western blot confirms the complete depletion of filamin C protein in Flnc global knockout mice (Flnc -/- or Flnc gKO ). GAPDH is used as a loading control. ( C ) Percentage of live wild-type (Flnc +/+ ), heterozygous (Flnc +/- ) and knockout (Flnc -/- ) embryos from E8.5 to E12.5. Exact number of each genotype and each developmental stage: E8.5, 4:9:4 (Flnc +/+ : Flnc +/- : Flnc -/- ); E9.5, 16:32:15; E10.5, 13:25:11; E11.5, 4:7:4*; E12.5, 4:6:0. *: dead/under resorption. ( D ) Wholemount images of control and Flnc global knockout (Flnc gKO ) embryos at E9.5, E10.5 and E11.5. Yellow arrows indicate pericardial effusion; Red arrow indicates heart rupture; Scale bar, 1 mm.
By in situ hybridization, we demonstrated that Flnc was specifically expressed in heart and somites from embryonic day (E) 9.5 to 11.5 ( S1A Fig ). The expression pattern of Flnc is consistent with the observation that heart and skeletal muscle are most affected in patients with mutations in FLNC [ 4 , 20 ]. However, previously described Flnc knockout mice only had defects in skeletal muscles but not in heart, probably owing to the hypomorphic nature of the mutant Flnc allele in that study [ 18 ]. Thus, a bona fide Flnc knockout mouse model is required to fully understand roles of filamin C in heart. To this end, we generated global Flnc knockout mice (Flnc -/- or Flnc gKO ) by crossing Flnc fl/fl [ 19 ] mice with Sox2 Cre mice [ 21 ] ( Fig 1A ). Western blot and immunofluorescence analyses confirmed that filamin C protein was completely absent in Flnc gKO mice (Figs 1B and S1B ).
Discussion
In this study, we demonstrated that filamin C played an essential role in maintaining the structural integrity of myocardium, as FlncgKO mice had severely ruptured ventricular myocardium but intact endocardium. Interestingly, CD31-positive thrombi and chest wall overgrowth were observed at the rupture sites, and β1 integrin and ECM proteins were upregulated in the endocardium. These phenomena are likely compensatory mechanisms to prevent complete heart rupturing. However, FlncgKO mice did not survive past E11.5, indicating that filamin C is essential for heart development and embryonic survival. Mechanistically, although several cell junction and dystrophin-associated glycoprotein complex (DGC) proteins were unchanged, key extracellular matrix (ECM) proteins were downregulated in myocardium of FlncgKO mice which may partially explain the heart rupture phenotype, reminiscent of our findings in kindlin-2 knockout mice [31]. Contrary to the belief that filamin C functions as an integrin inactivator, we observed attenuated activation of β1 integrin specifically in the myocardium of FlncgKO mice. To further investigate whether downregulation of activated β1 integrin was key to cardiac phenotypes in FlncgKO mice, we generated β1 integrin cardiomyocyte-specific knockout mice (Itgb1cKO). However, Itgb1cKO mice did not recapitulate the early heart rupture phenotype observed in Flnc knockout mice, whereas deleting β1 integrin and filamin C simultaneously from cardiomyocytes resulted in much more severe heart ruptures. Our results suggest filamin C works in concert with β1 integrin to maintain the structural integrity of myocardium during mammalian heart development.
FLNC is among the most mutated genes in dilated cardiomyopathy (DCM) and hypertrophic cardiomyopathy (HCM) patients [4], underscoring the essential role of filamin C in cardiac development and function. However, previously described Flnc knockout mice with homozygous deletion of the last 8 exons of Flnc only had defects in skeletal muscle but not in cardiac muscle [18]. Further studies revealed that these mice still expressed a truncated form of filamin C protein lacking the last four immunoglobulin (Ig)-like repeats and the hinge 2 region [18]. While the truncated filamin C protein was expressed at a lower level than wildtype, the reduction in Flnc mRNA levels was less pronounced in heart than in limb muscle [18], which may explain why there are phenotypes in skeletal muscles but not in heart of Flnc knockout mice. These observations also suggested that the N-terminal actin-binding domain and 19 Ig-like repeats (~82% of wild-type protein), even at much lower levels than that of wild-type FLNC proteins, are sufficient for filamin C to function normally in heart. Thus, the hypomorphic nature of the mutant Flnc allele renders it unsuitable for studying the function of filamin C in the heart. To address this problem, we generated Flnc knockout mice by deleting exon 9–13 of the Flnc gene, which introduced a premature termination codon (PTC) within exon 14 and subjected Flnc mRNA to nonsense-mediated mRNA decay (NMD). In line with this, Flnc mRNA levels were drastically downregulated in Flnc global knockout mice according to our RNA-seq data (Fig 3B and S1 Table). Although a small amount of N-terminal truncated protein (not recognizable by our filamin C antibodies that were raised against C-terminal regions of filamin C protein) could be generated, the truncated protein is unlikely to be functional as it only includes the N-terminal actin-binding domain and two Ig-like domains (486 amino acids, ~17% of wild-type protein).
In a recent report, filamin C was ablated in in vitro cultured human induced pluripotent stem cell–derived cardiomyocytes (FLNC−/− hiPSC-CMs), which exhibited defects in sarcomere assembly and decreased thin filament gene expression, suggesting that filamin C plays a role in sarcomere assembly and thin filament gene expression [17]. To determine whether filamin C possesses similar functions in vivo, we examined overall sarcomere structure in E9.5 FlncgKO cardiomyocytes by immunofluorescence analyses. However, we did not observe any sarcomere disarray as seen in FLNC−/− hiPSC-CMs. In addition, our RNA-seq analysis revealed very modest downregulation of thin filament genes including Lmod2 (Log 2 FC = -0.33), Tnni3 (Log 2 FC = -0.30) and Synpo2 (Log 2 FC = -0.31) (S1 Table), which is in stark contrast to the dramatic downregulation of thin filament genes in FLNC−/− hiPSC-CMs [17]. Our findings suggest that filamin C is dispensable in sarcomere assembly and has minimal impact on the expression of thin filament genes in vivo. It is worth noting that filamin A and filamin B were not upregulated in FLNC−/− hiPSC-CMs according to our examination of the transcriptomics and proteomics data from that study [17]. Thus, upregulation of filamin A and filamin B, or lack thereof, could explain why there are no defects of sarcomere assembly in cardiomyocytes of FlncgKO mice but sarcomere disarray in FLNC−/− hiPSC-CMs. Future studies, i.e., ablating all three filamins from cardiomyocytes in vivo, might be necessary to elucidate roles of filamins in sarcomere assembly.
Filamin C interacts with β1 integrin [11] and sarcoglycans [6] at the costamere to serve as a link between myofibrils and sarcolemma. Our discovery of myocardial wall ruptures in Flnc knockout mice provided strong support for filamin C’s essential structural role in myocardium integrity. On the other hand, filamins are well-known integrin inactivators that function by competing with talin for binding to the cytoplasmic domain of the integrin β subunit [15], and abnormal activation of β1 integrin can lead to impaired cell proliferation, differentiation and migration [15]. To determine whether β1 integrin is ectopically activated in filamin C-ablated cardiomyocytes which could account for the observed cardiac phenotypes, we examined expression and localization of activated and total β1 integrin by immunofluorescence. Surprisingly, while total β1 integrin expression and localization were unchanged, activated β1 integrin was reduced in myocardium of FlncgKO hearts. To further investigate whether the attenuated activation of β1 integrin was key to cardiac phenotypes in FlncgKO mice, we generated cardiomyocyte-specific Itgb1 knockout mice (Itgb1cKO) and compare them with cardiomyocyte-specific Flnc knockout mice (FlnccKO). However, ablating β1 integrin in cardiomyocytes did not cause myocardial rupture at E10.5, a stage when FlnccKO mice already had severe rupturing in their myocardium. Considering that Itgb1cKO indeed develop heart ruptures later at E14.5 [31], these findings indicated that attenuated β1 integrin activation alone may only partially account for myocardial ruptures in FlncgKO mice. Another possibility is that some residual β1 integrin proteins may still remain in E10.5 Itgb1cKO mice due to their remarkably long half-life [33], and these remaining β1 integrin proteins can be normally activated in the presence of filamin C. If this is the case, simultaneously ablating filamin C and β1 integrin should recapitulate the phenotypes of FlnccKO mice. However, the Flnc/Itgb1dcKO mice we generated had even more severe heart rupturing phenotypes than FlnccKO mice, suggesting that filamin C maintains the integrity of myocardium through both integrin-dependent and integrin-independent pathways. Future studies are needed to delineate detailed molecular mechanisms by which filamin C facilitates β1 integrin activation in cardiomyocytes.
[END]
---
[1] Url:
https://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1010630
Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.
via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/
