Summary
Background
The mechanism involved in the onset of aortic valve (AoV) disease remains unclear despite its poor prognosis and frequency. Recently, we reported that Krox20 ( EGR2 in humans) is involved in AoV development and dysfunction.
Aim
Analyze Krox20 heterozygous mice ( Krox20 +/− ) to discover whether incomplete expression of Krox20 can cause valvular diseases.
Methods
Transcriptional levels of Col1a2/COL1A2 and Krox20/EGR2 in AoVs from Krox20 +/− mice and human patients operated on for severe aortic regurgitation were evaluated by quantitative reverse transcription-polymerase chain reaction (qRT-PCR). Human control valves were obtained from three transplanted patients without AoV disease. Twenty-one heterozygous Krox20 +/− mice were compared with 35 controls at different ages. Three independent measurements of valve thickness were performed on magnified tissue sections using Image J software. In vivo valve structure and function were evaluated using the high-frequency Vevo ® 2100 echocardiogram.
Results
qRT-PCR analysis using AoVs from patients with severe aortic regurgitation showed a decrease in EGR2 expression associated with significant downregulation of COL1A2 expression ( P < 0.05). Similar results were observed in the AoVs of Krox20 +/− mice. Anatomical examination revealed that incomplete invalidation of Krox20 caused significant thickening of the aortic leaflet compared with controls (145 ± 22 vs. 75 ± 24 μm; P = 0.01). Within the mutant group, this thickening worsened significantly over time ( Krox20 +/− mice aged > 7 vs. < 7 months: 136 ± 48 vs. 102 ± 41 μm; P < 0.001). Moreover, the aortic leaflets of embryonic day 18.5 Krox20 +/− embryos were significantly more thickened than those from controls, suggesting that this disease begins during embryonic development. Echo-Doppler analysis showed a significant increase in AoV dysfunction in heterozygous versus control mice (53% vs. 17%; P < 0.001), suggesting a tight relationship between valve architecture and function. Morphometric analysis revealed that the most severe AoV dysfunction was always associated with the most thickened valves. Classic histological analysis revealed that mutant AoVs had extracellular matrix disorganization, with features of human myxomatous degeneration, including excess of proteoglycan deposition in spongiosa and reduction of collagen fibre in fibrosa, but no calcification.
Conclusion
Decreased expression of Krox20 in mice causes degeneration of the aortic leaflets and disorganization of the extracellular matrix, causing valvular dysfunction.
Résumé
Contexte
Les mécanismes impliqués dans la survenue des valvulopathies aortiques restent à ce jour non élucidés malgré leur fréquence et leur mauvais pronostic. Récemment, nous avons montré que Krox20 ( EGR2 chez l’Homme) joue un rôle important dans la valvulogenèse aortique et la survenue de dysfonction valvulaire.
Objectif
Notre objectif était de savoir si la perte d’un allèle de Krox20 pouvait conduire à un défaut valvulaire.
Méthodes
Nous avons mesuré par PCR quantitative en temps réel (QPCR) le niveau d’expression du gène Col1a2/COL1A2 et de Krox20/EGR2 au sein de valves aortiques provenant de souris hétérozygotes Krox20 +/− et de patients opérés d’insuffisance aortique sévère et sur des valves témoins issues de patients transplantés indemnes de valvulopathie aortique. Nous avons comparé un groupe de souris hétérozygotes Krox20 +/− ( n = 21) à un groupe témoin ( n = 35) à différents âges de vie. Les échocardiographies ont été réalisées à l’aide d’un échocardiogramme Vevo ® 2100. Les coupes histologiques ont été obtenues après inclusion des échantillons de cœurs murins en paraffine et section au microtome. L’épaisseur valvulaire a été mesurée à trois reprises grâce au logiciel Image J software.
Résultats
L’analyse par QPCR révèle une réduction de l’expression du gène EGR2 associée à une diminution significative de COL1A2 ( p < 0,05) dans les valves aortiques de patients opérés de valvulopathie aortique concordant avec une diminution de Col1a2 au sein de valves aortiques de souris mutantes. L’examen anatomique montre que l’invalidation incomplète de Krox20 conduit à un épaississement valvulaire significatif par rapport au groupe témoin (145 ± 22 vs 75 ± 24 μm ; p = 0,01). L’épaississement de la valve aortique s’aggrave au cours du temps (âge > 7 mois versus < 7 mois, 136 ± 48 vs 102 ± 41 μm ; p < 0,001). De plus, les feuillets valvulaires d’embryons mutants sont significativement plus épais que les feuillets des embryons témoins à E18,5, suggérant une apparition de la dégénérescence valvulaire dès la vie embryonnaire. L’analyse échocardiographique montre une augmentation significative de valvulopathies aortiques chez les souris Krox20 +/− (53 % vs 17 % ; p < 0,001), suggérant une relation étroite entre architecture valvulaire et dysfonctionnement. L’analyse morphométrique des échantillons démontre que les valvulopathies les plus sévères sont associées aux épaississements valvulaires les plus importants. Enfin, les feuillets aortiques des souris Krox20 +/− présentent une désorganisation de la matrice extracellulaire évocatrice de dégénérescence myxoïde incluant un excès de protéoglycans dans la spongiosa , une raréfaction des fibres de collagène dans la fibrosa et une absence de calcification.
Conclusion
Nos résultats montrent que la diminution de l’expression de Krox20 chez la souris est à l’origine d’une dégénérescence des feuillets aortiques et d’une désorganisation de la matrice extracellulaire responsable de dysfonction valvulaire.
Background
Valvular heart disease is the second major cardiac concern after coronary heart disease. Aortic valve (AoV) disease is the most frequent, occurring in 2.5% of the population in industrialized countries . Despite its high incidence and negative prognostic effect, the underlying mechanism of AoV disease remains unknown. AoV architecture changes throughout the life, suggesting environmentally mediated dysregulation in the pathological condition . However, it is now recognized that adult valve disease may originate in anomalies of valve development .
Valvulogenesis occurs after the initial stages of cardiogenesis, as a result of endocardial cushion formation and extensive remodelling of the extracellular matrix (ECM). The atrioventricular cushions give rise to the mitral and tricuspid valves, while the aortic and pulmonary valves arise from outflow tract cushions. Cushion formation initiates with expansion of the cardiac jelly, the ECM residing between the myocardium and the endocardium. The essential first step is an endothelial-to-mesenchymal transformation of endocardial cells, induced by signalling factors coming from the myocardium . Semilunar valve development is distinguished from atrioventricular valve formation by the invasion of neural crest-derived cells, which participate in outflow tract septation . The mature valve structure is composed of a highly organized ECM, with three layers rich in elastin (ventricularis), proteoglycans (spongiosa) and fibrillar collagen (fibrosa). The fibrosa is situated on the aortic aspect of the semilunar valves ; this layer is composed predominantly of fibrillar collagens types I and III, which provide tensile stiffness. There is a tight relationship between valve architecture and function. Misexpression and disorganization of fibrillar collagen in the ECM are associated with AoV dysfunction, as well as aortic stenosis or regurgitation, caused by impaired valvulogenesis and altered architecture of the aortic leaflet. Collagen accumulation in fibrotic degeneration promotes aortic stenosis by stiffening of the valve during systole. Conversely, collagen degradation in myxomatous degeneration causes leaflet prolapse and aortic regurgitation during diastole. To date, the mechanisms involved in transcriptional regulation of collagen genes remain unknown.
Recently, we demonstrated that the transcription factor Krox20 (called EGR2 in human) is expressed in the semilunar and atrioventricular valve primordia, and that its function is required for cardiac valve development. Indeed, in vivo imaging and histological analysis revealed that complete loss of Krox20 in the mouse results in AoV dysfunction, with a significant thickening and disorganization of the ECM, including excess of proteoglycan and reduction of fibrillar collagens .
In this study, we used in vivo imaging to show that Krox20 heterozygous mice also have AoV dysfunction. In addition, histological and transcriptional analyses reveal that AoV anomalies are associated with disorganization of the collagen layer and impaired Col1a2 transcriptional levels.
Methods
Animal experiments
This work was approved by the “comité d’éthique pour l’expérimentation animale” (Marseille Animal Care Committee; Protocol No. 38-09102012) and conformed to Directive 2010/63/EU of the European Parliament. Krox20 lacZ ( Krox20 − ) mice were intercrossed with the C57BL/6 mice to generate Krox20 +/− heterozygous offspring that were obtained at expected Mendelian ratios. After echocardiography analysis, adult Krox20 +/− mice were sacrificed via intraperitoneal injection of pentobarbital sodium (0.5 mL). Heart samples were harvested, fixed in 4% paraformaldehyde and stored in 1X phosphate buffered saline.
Histological analysis
Standard histological procedures were used . Heart tissues from Krox20 +/− and littermate controls were paraffin embedded and cut at 8 μm per tissue section. Sections were stained with Masson’s trichrome (Sigma, St. Louis, MO, USA), orcein (Sigma), alcian blue (Sigma), Von Kossa (Sigma) and alizarin red (Sigma), according to the manufacturer’s instructions.
Quantification of valve anomalies
Regarding quantification of valve thickness, the distal region of the leaflets or leaflets of valves were used over a minimum depth of 100 μm using a DM5000 microscope with LAS software (Leica Microsystems, Wetzlar, Germany). Image J software was used to measure the surface and the thickness of the valve. Three independent measurements were taken per leaflet, from 10 different sections, and the values were averaged. At least six animals were used per genotype for statistical analysis.
Echocardiography
In vivo valve structure and function were evaluated using an ultra high-frequency, high resolution ultrasound (Vevo ® 2100; VisualSonics, Inc., Toronto, ON, Canada). A total of 35 controls and 21 mutant mice were analysed for our study. The chests of the mice were treated with a chemical hair remover to reduce ultrasound attenuation. Heart rate and core body temperature were continuously monitored. Mice were anaesthetized with 1–2% isofluorane inhalation, and placed on a heated platform to maintain their temperature during the analysis. Two-dimensional imaging was recorded with a 40 MHz transducer to capture long- and short-axis projections with guided M-Mode, B-Mode and colour and pulsed-wave Doppler. Doppler interrogation was performed on the semilunar valve outflow in the parasternal long-axis view to assess stenosis and regurgitation using a sample volume toggle to optimize the angle of interrogation. A modified right parasternal long-axis view was required in some cases to ensure ascertainment of the maximum velocity. All measurements were obtained using an angle of interrogation < 50°. Aortic regurgitation was defined as valve incompetence with reversal of flow in diastole. Statistical significance was determined using Student’s t -test and defined as P < 0.05.
Human AoV samples and processing
This study was performed in accordance with institutional guidelines and was approved by the local research ethics committee at the University Hospital of Marseille (No. 2013-A01020-45). Written consent was obtained from all patients involved. Control valves were collected from three transplanted hearts with no valvulopathy. Four abnormal AoVs were collected from patients operated on for severe aortic dysfunction. Ribonucleic acid (RNA) was extracted as described below.
Real-time quantitative reverse transcription-polymerase chain reaction (qRT-PCR)
The AoVs and surrounding tissue containing the leaflets, annulus and aortic root were manually dissected from 3-month-old Krox20 mutant mice. After genotyping, samples from six mice of the same genotype were pooled, and RNA was isolated using a NucleoSpin ® RNA/Protein Kit (Macherey-Nagel GmbH & Co. KG, Düren, Germany) in accordance with the manufacturer’s instructions. Human AoV tissues were lysed on TRIzol ® (Life Technologies, Carlsbad, CA, USA), and RNA was extracted using an RNeasy ® Mini Kit (Qiagen, Hilden, Germany). Reverse transcriptions were performed by using a first strand complementary deoxyribonucleic acid synthesis kit (Roche, Basel, Switzerland) in accordance with the manufacturer’s instructions. A LightCycler ® 480 SYBR Green I Master mix (Roche) was used for real-time qRT-PCR analysis with a LightCycler ® 480 (Roche), following the manufacturer’s instructions. The gene-specific primers used in this study have been described previously . Each experiment was performed in triplicate for each genotype. Samples were normalized to endogenous housekeeping gene ( TBP or GAPDH genes for mouse or human samples, respectively). Level changes were calculated by the comparative cycle threshold (ΔΔCT) method. Normalized expression levels in the control ( Krox20 +/+ ) were set to 1.0 for each gene.
Statistical analysis
All parametric data are expressed as mean ± standard error of the mean. Statistics were carried out using Student’s t -test to compare variances and the Chi 2 test for two independent samples. A P value < 0.05 was considered significant.
Results
Analysis of COL1A2/Col1a2 and EGR2/Krox20 expression in human and mouse AoV disease
Molecular mechanisms leading to extracellular matrix (ECM) disturbance, such as collagen fibres in the AoV, are still unclear. We collected explants of diseased human AoV, with no obvious fibrocalcific remodelling, from patients operated on for severe AoV regurgitation (mean age 43.25 ± 12.75 years), to determine transcriptional levels of fibrillar collagen genes by real-time qRT-PCR. Control valve tissues were obtained during cardiac transplantation from patients with end-stage heart failure and no AoV disease. Among all the fibrillar collagen genes tested, only COL1A2 was significantly downregulated in diseased AoV compared with controls ( P < 0.05) ( Fig. 1 A). Interestingly, qRT-PCR showed 50% reduced expression of KROX20/EGR2 , known as a direct regulator of fibrillar collagen , in all diseased AoVs compared with controls ( Fig. 1 B). This finding emphasizes that downregulation of KROX20/EGR2 expression leads to severe AoV dysfunction. In order to test whether the haploin sufficiency of Krox20 could be involved in the reduction of collagen fibres, we performed qRT-PCR analysis in Krox20 heterozygous mice at 3 months of age. Our data revealed a significant reduction in Krox20 expression in the AoV leaflets of heterozygous mice ( P < 0.05) ( Fig. 1 C). In addition, the level of Col1a2 transcripts was downregulated in the AoVs of Krox20 +/− mice ( Fig. 1 D), confirming the requirement of Krox20 for the activation of that collagen.
