cells Article Opposite Effects of Moderate and Extreme Cx43 Deficiency in Conditional Cx43-Deficient Mice on Angiotensin II-Induced Cardiac Fibrosis Laura Valls-Lacalle 1,2,† , Corall Negre-Pujol 1,2,† , Cristina Rodríguez 2,3 , Saray Varona 2,4 , Antoni Valera-Cañellas 1,2 , Marta Consegal 1,2 , Jose Martínez-González 2,4 and Antonio Rodríguez-Sinovas 1,2, * 1 Cardiovascular Diseases Research Group, Department of Cardiology, Vall d’Hebron University Hospital and Research Institute, Universitat Autònoma de Barcelona, Departament de Medicina, Pg. Vall d’Hebron 119-129, 08035 Barcelona, Spain; laura.valls@vhir.org (L.V.-L.); corall_np@hotmail.com (C.N.-P.); tonivalerac@hotmail.com (A.V.-C.); marta.consegal@vhir.org (M.C.) 2 Centro de Investigación Biomédica en Red sobre Enfermedades Cardiovasculares (CIBERCV), 28029 Madrid, Spain; CRodriguezS@santpau.cat (C.R.); SVarona@santpau.cat (S.V.); JMartinezG@santpau.cat (J.M.-G.) 3 Institut de Recerca del Hospital de la Santa Creu i Sant Pau, IIB-Sant Pau, 08041 Barcelona, Spain 4 Instituto de Investigaciones Biomédicas de Barcelona (IIBB-CSIC), IIB-Sant Pau, 08041 Barcelona, Spain * Correspondence: antonio.rodriguez.sinovas@vhir.org; Tel.: +34-934894038; Fax: +34-934894032 † Both authors contributed equally.  Received: 30 September 2019; Accepted: 21 October 2019; Published: 22 October 2019  Abstract: Connexin 43 (Cx43) is essential for cardiac electrical coupling, but its effects on myocardial fibrosis is controversial. Here, we analyzed the role of Cx43 in myocardial fibrosis caused by angiotensin II (AngII) using Cx43fl/fl and Cx43Cre-ER(T)/fl inducible knock-out (Cx43 content: 50%) mice treated with vehicle or 4-hydroxytamoxifen (4-OHT) to induce a Cre-ER(T)-mediated global deletion of the Cx43 floxed allele. Myocardial collagen content was enhanced by AngII in all groups (n = 8–10/group, p < 0.05). However, animals with partial Cx43 deficiency (vehicle-treated Cx43Cre-ER(T)/fl ) had a significantly higher AngII-induced collagen accumulation that reverted when treated with 4-OHT, which abolished Cx43 expression. The exaggerated fibrotic response to AngII in partially deficient Cx43Cre-ER(T)/fl mice was associated with enhanced p38 MAPK activation and was not evident in Cx43 heterozygous (Cx43+/- ) mice. In contrast, normalization of interstitial collagen in 4-OHT-treated Cx43Cre-ER(T)/fl animals correlated with enhanced MMP-9 activity, IL-6 and NOX2 mRNA expression, and macrophage content, and with reduced α-SMA and SM22α in isolated fibroblasts. In conclusion, our data demonstrates an exaggerated, p38 MAPK-dependent, fibrotic response to AngII in partially deficient Cx43Cre-ER(T)/fl mice, and a paradoxical normalization of collagen deposition in animals with an almost complete Cx43 ablation, an effect associated with increased MMP-9 activity and inflammatory response and reduced fibroblasts differentiation. Keywords: angiotensin II; fibrosis; collagen; connexin 43; hypertrophy 1. Introduction Connexins are a family of membrane proteins with a characteristic structure consisting of four transmembrane domains, cytoplasmic amino- and carboxi-terminal domains, and an extracellular loop [1]. Hexameric connexin assemblies, known as connexons or hemichannels, dock with connexons from adjacent cells, to form intercellular channels, that put into contact the cytoplasms of neighboring cells. In the heart, these intercellular channels gather at cardiomyocyte poles [2], forming plaques Cells 2019, 8, 1299; doi:10.3390/cells8101299 www.mdpi.com/journal/cells Cells 2019, 8, 1299 2 of 17 called gap junctions. Gap junctions constitute low resistance pathways that are essential to allow electrical current flow between connected cells, and thus, coordinated cardiac contraction [3]. A total of 21 different connexin genes have been described in the human genome, whereas 20 have been found in mice [4]. Each of these genes code for a different protein, all sharing a similar structure and high homology [4]. Connexin 43 (Cx43) is the most ubiquitous connexin isoform, being widely distributed in most tissues, including cardiac cells (cardiomyocytes, fibroblasts, and endothelial and smooth muscle cells) [5,6]. The best known function of Cx43 in the heart is to allow electrical coupling between neighboring cardiomyocytes [3]. The importance of gap junction channels in electrical coupling is demonstrated by evidence showing that altered gap junction coupling through Cx43 intercellular channels increases arrhythmogenesis [7]. However, Cx43 has additional gap junction-dependent and independent functions [8]. Thus, Cx43 gap junction channels allow the existence of chemical coupling between the cytoplasms of connected cells, permitting transfer of most ions and second messengers [9]. Chemical coupling through Cx43 has been suggested to be involved in propagation of cell death during myocardial ischemia reperfusion, as demonstrated by the fact that gap junction uncouplers reduce infarct size after transient ischemia in different animal models [10]. In the same direction, isolated hearts from transgenic mice devoid of Cx43 had reduced infarctions after transient global ischemia [11,12]. Gap junction-independent functions of Cx43 in the heart include those ascribed to free sarcolemmal hemichannels and mitochondrial Cx43. Free hemichannels at the sarcolemma have been suggested to play a prominent role in paracrine communication and may affect cell death and survival [8,13]. In addition, some data suggest that mitochondrial Cx43 plays a key role in ischemic preconditioning, as protection is abolished when Cx43 translocation to the mitochondria is reduced [11,12]. Less known are the effects of Cx43 on scar formation and collagen deposition, especially in the heart. Whereas in most tissues reduction of Cx43 expression or functionality is associated with beneficial effects on scar formation [14–16], data on the role of Cx43 in cardiac remodeling and collagen deposition are not conclusive. Thus, some authors have reported that reduced Cx43 expression in heterozygous Cx43+/- mice attenuates collagen content in infarcted and non-infarcted regions and reduces ventricular remodeling and scar area (i.e., infarct size) several weeks after permanent coronary ligature [17,18]. In addition, application of the cell-permeable Cx43 carboxiterminal mimetic peptide αCT1, that competitively inhibits the interaction of endogenous Cx43 with zonula occludens-1, to cryoinjured mouse hearts was associated with reduced gap junction remodeling and lower inducibility of ventricular arrhythmias 7 to 9 days after injury [19], and decreased left ventricular volumes 8 weeks later [20]. In contrast to these findings, it has been reported that reduced cellular coupling in Cx43Cre-ER(T)/fl mice, expressing 50% of normal Cx43 content, associates with increased collagen deposition after pressure overload induced by transverse aortic constriction [21]. However, most of these studies have used mice models with partial Cx43 deficiency (about 50%) present from birth. This fact might have resulted in compensatory changes that may account for some of those findings. In addition, data on scar size might be biased by the acute effects of Cx43 deficiency on cell injury [10]. In this context, the aims of this work were to analyze the effects of an inducible, and almost complete, Cx43 deficiency on cardiac fibrosis and left ventricular remodeling after angiotensin II (AngII)-induced pressure overload in Cx43Cre-ER(T)/fl mice, and to analyze the mechanisms involved. 2. Materials and Methods An expanded description of the methods used in this work can be obtained at the Online Supplement. 2.1. Mice and Experimental Protocol Studies were carried out in adult conditional, Cx43-deficient, Cx43Cre-ER(T)/fl animals and in its Cx43fl/fl littermates (4–6 months old, genetic background: C57BL/6J) of both sexes (sex ratio 1:1). Cx43Cre-ER(T)/fl animals were developed by Eckardt et al [22]. In them, the coding region of one Cells 2019, 8, 1299 3 of 17 of the Cx43 alleles was replaced by Cre-ER(T), a fusion construct of the Cre recombinase and a specifically mutated domain of the human estrogen receptor. Treatment with 4-hydroxytamoxifen (4-OHT) leads to binding of the drug to the ER(T) domain, inducing Cre activity, whereas ER(T) is insensitive to the natural ligand β-estradiol. Cre activation causes a global deletion of Cx43, after recognition of loxP sites flanking the second Cx43 allele, not modifying expression of other connexin isoforms [7,12,22,23]. Osmotic minipumps were subcutaneously implanted, at the interscapular space, and under 2% isofluorane anesthesia (model 1002, Alzet, DURECT Corporation, Cupertino, CA, USA) to continuously deliver saline or AngII (1000 ng/kg/min, Sigma-Aldrich, St Louis, MO, USA) at a constant flow of 0.25 µl/h, for 14 days [24]. One day after pump implantation, Cx43Cre-ER(T)/fl and Cx43fl/fl mice were intraperitoneally injected with vehicle (oil) or 4-OHT to induce Cx43 ablation in the former, as described before [12,22]. This protocol was chosen to ensure that the reduction in Cx43 content coincided with the time of maximal collagen deposition, and to avoid excessive mortality associated with Cx43 deficiency [7,12,22]. In addition, Cx43+/- mice and their wild-type Cx43+/+ littermates (genetic background: C57BL/6J) were used to characterize some of the findings. These animals were a kind gift from Dr. S. Poelzing (Virginia, USA). At the end of the experimental procedure, mice were euthanized with a sodium pentothal overdose (1.5 g/kg, intraperitoneal). Hearts were quickly excised and weighted. The basal half was immediately snap-frozen in liquid N2 , whereas the apical area was fixed overnight with 4% paraformaldehyde and embedded in paraffin. Cardiac hypertrophy was calculated as the ventricular weight/body weight ratio. Animals were distributed in the following experimental groups: Cx43fl/fl + oil, Cx43fl/fl +4-OHT, Cx43Cre-ER(T)/fl + oil, and Cx43Cre-ER(T)/fl + 4-OHT (n = 8–11/group), together with Cx43+/+ and Cx43+/- (n = 4–7/group), either treated with saline or AngII. To check whether changes in collagen deposition observed in Cx43Cre-ER(T)/fl mice injected with oil after AngII treatment were secondary to increased p38 MAPK activity, 8 additional animals from this group were intraperitoneally treated with 10 mg/Kg/day SB203580, a p38 MAPK inhibitor (prepared at 6.67 mmol/L in 33% DMSO), beginning the first day after pump implantation to the end of the experimental protocol. 2.2. Systolic Cardiac Function by Transthoracic Echocardiography Echocardiographic measurements were performed at baseline and at the end of the experimental protocol, with a Vivid q portable ultrasound system, using an ILS 12 MHz transducer (GE Healthcare, WI, USA), as described [25]. 2.3. Collagen Content and Cardiomyocyte Size Cardiac fibrosis was analyzed in paraffin-embedded histological sections (4 µm) stained with Picrosirius Red (Sigma-Aldrich, MO, USA) as described [24]. Mean cardiomyocyte cross-sectional area was measured in sections stained with hematoxylin and eosin [25]. 2.4. Immunohistochemistry Immunostaining was performed in paraffin-embedded cardiac sections (4 µm) by overnight incubation (4 ◦ C) with antibodies against LAMP-2/Mac-3 (#sc-19991, 1:200, Santa Cruz Biotechnology Inc., Dallas, TX, USA) or MMP9 (#ab38898, 1:500, Abcam, Cambridge, UK) as previously described [24]. Each staining was performed by duplicate. 2.5. Real Time-PCR Total RNA from frozen mouse myocardium was isolated using the TriPure Isolation Reagent (Roche Diagnostics, Indianapolis, IN) following manufacturer’s instructions [24]. Quantification of mRNA levels was performed by real-time PCR using an ABI PRISM 7900HT sequence detection system (Applied Biosystems, Foster City, CA, USA) and specific primers and probes provided by Applied Cells 2019, 8, 1299 4 of 17 Biosystems (Assay-on-Demand system) or Integrated DNA technologies (Coralville, IA, USA). Relative mRNA levels were determined using the 2−∆∆Ct method. 2.6. Total Myocardial Homogenates Snap-frozen myocardium was homogenized (Diax 600 homogenizer, Heidolph, Germany) in homogenization ice-cold buffer (in mmol/L: Tris-HCl 20, NaCl 140, EDTA 0.8 (pH 7.8), Tween 20 0.1%, sodium fluoride 1, sodium orthovanadate 1, and a protease cocktail inhibitor (1%)). Protein lysates were obtained from the supernatant after centrifugation at 750 g for 10 minutes (4 ◦ C). 2.7. Gelatin Zymography Enzymatic activities of metalloproteinase (MMP)-9 and MMP-2 in murine cardiac homogenates were measured by zymography, as previously described [26]. Gels were scanned in a CanonScan 9000F scanner and densitometrically analyzed (Image Studio Lite v5.0, LI-COR Corporate, Lincoln, NE, USA). 2.8. Cardiac Fibroblasts Isolation Cardiac fibroblasts were isolated from a pool of 2–4 hearts of adult mice by differential centrifugation of cardiac cells after digestion with a mix of collagenase-trypsin as previously described [24]. 2.9. Western Blot Analysis Protein extracts from mouse hearts, and from fibroblasts in culture, were electrophoretically separated on 10% polyacrylamide gels and analyzed by Western blot according to standard procedures [12,27]. 2.10. Statistics Data are expressed as mean ±SEM. Statistical analysis was performed using SPSS 15.0 (IBM, Armonk, NY, USA). Differences were assessed by one-way or two-way ANOVA and Tukey post hoc tests. Repeated measures ANOVA (MANOVA) was used for echocardiographic data. Differences were considered significant when p < 0.05. 3. Results Two Cx43Cre-ER(T)/fl mice injected with 4-OHT and infused with saline, and three AngII-infused, died during the 14 days interval and were excluded from further analysis. This mortality corresponds to that previously associated with Cx43 deficiency in this model [7,12,22]. 3.1. AngII Treatment Induces a Similar Hypertrophic Response in Hearts from Cx43fl/fl and Cx43Cre-ER(T)/fl Mice Independently of Cx43 Levels Treatment with AngII for 14 days induced an increase in cardiac weight/body weight (CW/BW) ratio, indicative of cardiac hypertrophy, of similar magnitude in all groups, independently of Cx43 levels (Figure 1a). The lack of influence of Cx43 on the hypertrophic response to AngII was confirmed by assessment of cardiomyocyte cross-sectional area, which was similarly enhanced in all groups (Figure 1b), and by echocardiography, showing increased LVPW and IVS (Figure 1c–d) at day 14, with no changes in ejection fraction, LVEDD, or LVESD (Figure 1e–g). Two-way ANOVA analysis demonstrated a significant effect of AngII infusion (p < 0.001) on CW/BW and cardiomyocyte cross-sectional area, with no significant differences between experimental groups and lack of interactions between infusion treatment and group. Consequently, with these data, AngII induced a significant induction of the classic hypertrophic marker ANP in all groups (two-way ANOVA, p < 0.001), with no effect of group allocation and lack of interaction between both factors (Figure 1h). Cells 2019, 8, 1299 5 of 17 Cells 2019, 8, x FOR PEER REVIEW 5 of 17 Figure 1. AngII treatment induces a similar cardiac hypertrophic response in hearts from Cx43fl/flfl/fland Figure 1. AngII treatment induces a similar cardiac hypertrophic response in hearts from Cx43 and Cx43Cre-ER(T)/fl mice, independently of Cx43 expression levels. Changes in cardiac weight/body weight Cx43Cre-ER(T)/fl mice, independently of Cx43 expression levels. Changes in cardiac weight/body weight (CW/BW) ratio (a) and cardiomyocyte cross-sectional area (b) in Cx43fl/fl (fl/fl) and Cx43Cre-ER(T)/fl (Cre/fl) (CW/BW) ratio (a) and cardiomyocyte cross-sectional area (b) in Cx43fl/fl (fl/fl) and Cx43Cre-ER(T)/fl mice, treated with saline or angiotensin II for 14 days. * (p < 0.05) indicates significant differences vs. the (Cre/fl) mice, treated with saline or angiotensin II for 14 days. * (p < 0.05) indicates significant corresponding saline-treated group. (c–g) Changes in left ventricular posterior wall thickness (LVPW), differences vs. the corresponding saline-treated group. (c–g) Changes in left ventricular posterior interventricular septum thickness (IVS), left ventricular end-diastolic internal diameter (LVEDD), left wall thickness (LVPW), interventricular septum thickness (IVS), left ventricular end-diastolic ventricular end-systolic internal diameter (LVESD) (all expressed vs. body weight), and ejection fraction internal diameter (LVEDD), left ventricular end-systolic internal diameter (LVESD) (all expressed vs. (EF) in the same animals. Data from saline-treated animals are shown pooled as a single group. * body weight), and ejection fraction (EF) in the same animals. Data from saline-treated animals are (p < 0.05) indicates significant differences vs. all groups except Cx43fl/fl +4-OHT + AngII and Cx43fl/fl + shown pooled as a single group. * (p < 0.05) indicates significant differences vs. all groups except oil + AngII in C and D, respectively. (h) Changes in myocardial ANP mRNA. * (p < 0.05) indicates Cx43fl/fl +4-OHT + AngII and Cx43fl/fl + oil + AngII in C and D, respectively. (h) Changes in myocardial significant differences vs. the corresponding saline-treated group. ANP mRNA. * (p < 0.05) indicates significant differences vs. the corresponding saline-treated group. As happened in Cx43fl/fl and Cx43Cre-ER(T)/fl mice, chronic exposure to AngII induced a similar As in increase happened CW/BW in Cx43 ratio fl/fl and Cx43Cre-ER(T)/fl mice, chronic exposure to AngII induced a similar in both Cx43+/- animals (3.80 ± 0.13 in saline-treated (n = 4) to 4.8 ± 0.20 increase in CW/BW ratio in both Cx43 mg/g in AngII-infused mice (n = 6), p = 0.004) +/- animals (3.80 ± 0.13 in saline-treated (n = 4) to 4.8 ± 0.20 mg/g and in their wild-type littermates (from 4.20 ± 0.09 in in AngII-infused saline-treated mice (n = 7) (n =± 6), to 4.60 0.20p mg/g = 0.004) and in their mice in AngII-treated wild-type (n = 6),littermates p = 0.04). (from 4.20 ± 0.09 in saline-treated (n = 7) to 4.60 ± 0.20 mg/g in AngII-treated mice (n = 6), p = 0.04). 3.2. AngII Treatment Differentially Modulates Cardiac Fibrosis in Cx43Cre-ER(T)/fl Mice 3.2. AngII Treatment Differentially Modulates Cardiac Fibrosis in Cx43Cre-ER(T)/fl Mice Myocardial interstitial collagen was low in all saline-infused experimental groups, with a collagen volume Myocardial interstitial fraction ranging collagen between 2% and was 3%low in all (Figure saline-infused 2). In experimental contrast, treatment groups, with AngII for 14with daysa collagen volume fraction ranging between 2% and 3% (Figure 2). In contrast, treatment increased collagen content in Cx43 mice injected with oil from 2.67 ± 0.32% to 7.30 ± 1.19% (p < 0.05) fl/fl with AngII for 14 days (Figure increased 2a). The enhancedcollagen content collagen in Cx43in deposition fl/fl mice injected with oil from 2.67 ± 0.32% to 7.30 ± response to AngII was even higher, in oil-treated 1.19% Cx43 (p < 0.05) Cre-ER(T)/fl (Figure 2a). The enhanced collagen animals, expressing 50% of normal Cx43 deposition content in response (Figure 2b).to AngII was even higher, in oil-treated Cx43Cre-ER(T)/fl animals, expressing 50% of normal Cx43 content (Figure 2b). To assess whether the marked enhancement in collagen content in response to AngII observed in Cx43Cre-ER(T)/fl + oil mice could be explained by their mild Cx43 deficiency, this approach was also conducted in a mouse model having a similar partial reduction of Cx43 (Cx43+/- mice). Unexpectedly, AngII did not increase collagen deposition in Cx43+/- mice over levels found in AngII-treated Cx43+/+ animals (Figure 2c), which may suggest that findings obtained in Cx43Cre-ER(T)/fl + oil mice were, in fact, not directly related to the level of Cx43 expression. Cells 2019, 8, 1299 6 of 17 Cells 2019, 8, x FOR PEER REVIEW 6 of 17 Figure Figure 2. 2. Cardiacfibrosis Cardiac fibrosisinduced inducedbyby AngII AngII in in Cx43 Cre-ER(T)/fl Cx43Cre-ER(T)/fl andand Cx43Cx43 +/- mice. +/- mice. Representative images images Representative (upper panels) and mean quantification (lower graphs) of interstitial (upper panels) and mean quantification (lower graphs) of interstitial collagen collagen deposition, expressed deposition, expressed as percentage as percentage ofoftotal totalmyocardial myocardial area, area, in in Cx43 Cx43fl/fl((a) fl/fl fl/fl) ((a) and fl/fl) Cx43 and Cre-ER(T)/fl ((b) Cre/fl) mice, after Cx43Cre-ER(T)/fl ((b) Cre/fl) mice, after treatment, for 14 days, with saline or AngII. Bar represents 50 m. (c) shows changes in fibrosis in treatment, for 14 days, with saline or AngII. Bar represents 50 µm. (c) shows changes in fibrosis in wild-type and Cx43 +/- animals. The approximate amount of Cx43 expression is indicated, in wild-type and Cx43+/- animals. The approximate amount of Cx43 expression is indicated, in parenthesis, parenthesis, below the name of each group. * (p < 0.05) indicates significant differences vs. the below the name of each group. * (p < 0.05) indicates significant differences vs. the corresponding corresponding saline-treated group.  (p < 0.05) indicates significant differences vs. all Cx43fl/fl or saline-treated group. τ (p < 0.05) indicates significant differences vs. all Cx43fl/fl or Cx43Cre-ER(T)/fl Cx43Cre-ER(T)/fl AngII-treated animals. AngII-treated animals. The effects of AngII on collagen deposition observed in Cx43fl/fl mice injected with oil were not To assess modified whether when theof animals marked enhancement this genotype in collagen were treated with content 4-OHT (Figure in response 2a). Into contrast, AngII observed the in Cx43 Cre-ER(T)/fl + oil mice could be explained by their mild Cx43 deficiency, this approach was also marked enhancement in collagen content observed in mild Cx43-deficient Cx43 Cre-ER(T)/fl mice was conducted reverted in inaanimals mouse model with anhaving almosta complete similar partial reduction Cx43 deletion of Cx43 (Cx43 (Cx43 Cre-ER(T)/fl +/- mice). treated Unexpectedly, with 4-OHT) (Figure AngII did not2b), increase with collagen values collagen approaching deposition those+/-found in Cx43 miceinover Cx43 found in AngII-treated Cx43+/+ fl/fl mice after AngII (Figure 2a). levels Two-way animals (FigureANOVA analysismay 2c), which demonstrated suggest that significant findings effects obtainedof both in Cx43 allocation+and groupCre-ER(T)/fl oil infused mice were, in treatment, and a significant interaction between fact, not directly related to the level of Cx43 expression. both variables in Cx43 fl/fl and Cx43Cre-ER(T)/fl animals, but only for infused treatment in Cx43+/- mice. The effects of AngII on collagen deposition observed in Cx43fl/fl mice injected with oil were not Analysis of slices using polarized light microscopy showed a significant increase in mature modified when animals of this genotype were treated with 4-OHT (Figure 2a). In contrast, the marked collagen, after AngII, in all groups except in Cx43Cre-ER(T)/fl animals treated with 4-OHT (Supplemental enhancement Figure S1). in collagen content observed in mild Cx43-deficient Cx43Cre-ER(T)/fl mice was reverted in animals with an almost complete Cx43 deletion (Cx43Cre-ER(T)/fl treated with 4-OHT) (Figure 2b), with3.3.collagen Paradoxical values approaching Overexpression those of mRNAs found Coding for in Proteins fl/fl mice Cx43Involved inafter Collagen AngII (Figure Turnover in 2a). Two-way Cx43-Deficient ANOVA analysisMice demonstrated significant effects of both group allocation and infused treatment, and a significant fl/fl and Cx43Cre-ER(T)/fl animals, but only for Next, interaction we analyzedbetween expression both variables of mRNAs in Cx43 coding for proteins involved in collagen synthesis infused treatment inP4HA1), +/- Cx43 maturation mice. (COL1A1, TGFβ1, (LOX) and degradation (TIMP1, TIMP2) in both wild-type Analysis (Cx43 fl/fl) andof slices using mice, Cx43-deficient polarized infused light withmicroscopy saline or AngII. showed As can a be significant seen in Figureincrease in mature 3, when treated with saline, hearts from Cx43 fl/fl mice andCre-ER(T)/fl from Cx43 Cre-ER(T)/fl animals injected with oil (50% collagen, after AngII, in all groups except in Cx43 animals treated with 4-OHT (Supplemental Cx43S1). Figure expression) showed similar levels of all studied mRNAs. In contrast, saline-treated Cx43Cre-ER(T)/fl mice injected with 4-OHT depicted a marked induction of COL1A1, TGFβ1, LOX, and Cells 2019, 8, 1299 7 of 17 Cells 2019, 8, x FOR PEER REVIEW 7 of 17 3.3.TIMP1. ParadoxicalSuchOverexpression induction occurs of mRNAs despiteCoding for Proteins Involved these saline-infused in Collagen animals did not Turnover in interstitial deposit more Cx43-Deficient Mice collagen that remaining groups (Figure 2b). Furthermore, Next, we analyzed treatment withof expression AngIImRNAsincreased coding expression of these for proteins four in involved mRNAs collagenin Cx43 fl/fl mice synthesis and in Cx43 (COL1A1, Cre-ER(T)/fl TGFβ1, P4HA1), animals injected with maturation (LOX)oil,and as compared degradationwith(TIMP1, corresponding TIMP2) mice implanted in both wild-type with saline-containing (Cx43 osmotic pumps, fl/fl ) and Cx43-deficient but not mice, infused in saline with mice with a marked or AngII. As canCx43 deficiency be seen in Figure(Cx43 Cre-ER(T)/fl 3, when +4-OHT) treated with(Figure 3). In this saline, hearts last from Cx43 fl/fl mice group, the mRNA and from levels Cx43ofCre-ER(T)/fl COL1A1,animals TGFβ1,injected and LOX, although with oil (50% still elevated, Cx43 expression) wereshowed significantly lowerofinallAngII-infused similar levels studied mRNAs. than in saline-infused In contrast, saline-treated animals.Cre-ER(T)/fl Cx43 Two-way ANOVA analysis for COL1A1, TGFβ1, and LOX showed a significant mice injected with 4-OHT depicted a marked induction of COL1A1, TGFβ1, LOX, and TIMP1. effect of group allocation Such(p < 0.001) and significant interactions with infused treatment (p < 0.05), but no induction occurs despite these saline-infused animals did not deposit more interstitial collagen thateffects of AngII treatment. No significant remaining groupschanges (Figure were2b). observed for P4HA1 and TIMP2 in any case. Figure 3. Expression of mRNAs coding for proteins involved in collagen turnover in AngII-treated Cx43 Cre-ER(T)/fl Figure 3. Expression of mRNAs mice. Levels of mRNAscoding for proteins coding involved for proteins in collagen involved in collagen turnover in AngII-treated synthesis (COL1A1, Cx43 Cre-ER(T)/fl mice. Levels of mRNAs coding for proteins involved in collagen synthesis (COL1A1, P4HA1) (a–c), maturation (LOX) (d) and degradation (TIMP1, TIMP2) (e–f), expressed as percentage Cx43fl/fl vs. Table + oil mice 1. P4HA1) (a–c), treatedmaturation (LOX) with saline, (d) and in both degradation wild-type fl/fl , fl/fl) and (Cx43(TIMP1, TIMP2) (e–f), expressed Cx43-deficient mice as percentage (Cre/fl), implanted vs. with Cx43osmotic fl/fl + oilpumps mice containing treated with saline, saline in both or AngII. < 0.05) indicates * (p wild-type (Cx43fl/fl , fl/fl) and significant Cx43-deficient differences vs. themice (Cre/fl), implanted corresponding with osmotic saline-treated (p < 0.05) group. #pumps containing saline or AngII. shows significant * (p