Increased afterload induces pathological cardiac hypertrophy: a new in vitro model
- First Online:
- Cite this article as:
- Hirt, M.N., Sörensen, N.A., Bartholdt, L.M. et al. Basic Res Cardiol (2012) 107: 307. doi:10.1007/s00395-012-0307-z
- 1.9k Downloads
Increased afterload results in ‘pathological’ cardiac hypertrophy, the most important risk factor for the development of heart failure. Current in vitro models fall short in deciphering the mechanisms of hypertrophy induced by afterload enhancement. The aim of this study was to develop an experimental model that allows investigating the impact of afterload enhancement (AE) on work-performing heart muscles in vitro. Fibrin-based engineered heart tissue (EHT) was cast between two hollow elastic silicone posts in a 24-well cell culture format. After 2 weeks, the posts were reinforced with metal braces, which markedly increased afterload of the spontaneously beating EHTs. Serum-free, triiodothyronine-, and hydrocortisone-supplemented medium conditions were established to prevent undefined serum effects. Control EHTs were handled identically without reinforcement. Endothelin-1 (ET-1)- or phenylephrine (PE)-stimulated EHTs served as positive control for hypertrophy. Cardiomyocytes in EHTs enlarged by 28.4 % under AE and to a similar extent by ET-1- or PE-stimulation (40.6 or 23.6 %), as determined by dystrophin staining. Cardiomyocyte hypertrophy was accompanied by activation of the fetal gene program, increased glucose consumption, and increased mRNA levels and extracellular deposition of collagen-1. Importantly, afterload-enhanced EHTs exhibited reduced contractile force and impaired diastolic relaxation directly after release of the metal braces. These deleterious effects of afterload enhancement were preventable by endothelin-A, but not endothelin-B receptor blockade. Sustained afterload enhancement of EHTs alone is sufficient to induce pathological cardiac remodeling with reduced contractile function and increased glucose consumption. The model will be useful to investigate novel therapeutic approaches in a simple and fast manner.
KeywordsAfterload enhancementCardiac hypertrophyCardiac metabolismCardiac tissue engineeringEndothelin receptor antagonistFibrosis
Afterload enhanced or enhancement
Engineered heart tissue
Nonselective endothelin receptor antagonist
- ETA-RA (ETB-RA)
Selective endothelin-A receptor antagonist (endothelin-B)
Transverse aortic constriction
Aging of the population and better survival of patients affected by coronary heart disease have led to an increasing number of patients with chronic heart failure, the common end stage of virtually all cardiac diseases. Its prevalence in the general population is 1–2 %, but reaches more than 10 % in octogenarians . Left ventricular hypertrophy is the most important antecedent risk factor for the development of heart failure . Cardiac hypertrophy is clearly double-edged. Under exercise or pregnancy it is beneficial and referred to as physiologic, under sustained hypertension or aortic stenosis cardiac hypertrophy is pathologic. Both types of hypertrophy are characterized by an increase in cardiomyocyte size, but only pathological hypertrophy is associated with increased apoptosis and fibrosis as well as functional changes such as altered cellular Ca2+ homeostasis, ion channel remodeling, reduced contractile force and relaxation velocity, predisposing factors for malignant arrhythmia and heart failure . Despite considerable efforts, the mechanisms underlying the differences between physiological and pathological hypertrophy remain incompletely understood.
The classical model to study cardiac hypertrophy has been transverse aortic constriction (TAC), in which the afterload of hearts is increased by banding of either the thoracic or the abdominal aorta . TAC reliably results in cardiac hypertrophy, but functional outcomes such as heart failure can be variable. Further disadvantages of this method are its costs in terms of labor and money, the impossibility of investigating human myocyte biology, and limitations inherent to in vivo models. Specifically, it is difficult to differentiate direct load-induced alterations from systemic and humoral mechanisms. In vitro models of cardiac hypertrophy have been developed to overcome some of these limitations, e.g. neonatal rat cardiomyocytes stimulated by α-adrenergic agonists or endothelin-1 , or, when cultured on silicone membranes, by phasic or tonic stretch . Cardiomyocytes in cell culture are isotropically oriented, which may limit their informative value, as cardiomyocytes in heart tissue are highly organized. Moreover, passive stretch mimics an increase in preload rather than afterload, the more frequent cause of cardiac hypertrophy. Finally, standard 2D monolayer cell cultures do not allow measurement of contractile function, an important parameter differentiating “physiological” from “pathological” hypertrophy.
The aim of the present study was to generate a novel in vitro model that exhibits a tissue-like arrangement of cardiomyocytes and faithfully monitors the effects of increased afterload on cardiac tissue. One of the key factors was to switch from conventional 2-dimensional cardiomyocyte culture to a 3-dimensional format, called fibrin-based engineered heart tissue (EHT) . The effect of afterload enhancement was compared to pharmacological stimulation of EHTs.
A detailed description of materials and methods can be found at Supplemental Material online.
Generation of EHTs
EHTs were generated as previously described with minor modifications . Namely, we increased cell concentration from 4.1 to 5.0 × 106 cells/mL (+22 %), we omitted Matrigel in EHTs, and reduced the initial volume of a single EHT from 145 to 100 μL (−31 %). Briefly, ventricular heart cells (the atria were carefully excised) from neonatal Wistar and Lewis rats (postnatal day 0 to 3) were isolated by a fractionated DNase/Trypsin digestion protocol. This procedure was reviewed and approved by the Ethics Commission of the Medical Association of Hamburg. Rat ventricular heart cells, fibrinogen, thrombin and DMEM (2×, to match the volumes of fibrinogen and thrombin and thus ensuring isotonic conditions) were mixed and pipetted into molds, which were obtained by casting 2 % agarose (in PBS) around Teflon® (polytetrafluoroethylene) spacers in a 24-well culture dish (Online Table I). After polymerization of fibrin (1–2 h), EHTs were transferable to new cell culture dishes filled with medium (Online Fig. I).
Manufacturing silicone racks with adjustable post resistance
EHTs were generated in casting molds, each of them containing a pair of small silicone tubes. They were part of custom-made silicone racks (Sylgard® 184, Dow Corning), each carrying four pairs of silicone tubes (Fisher Scientific 3100504; Online Fig. I). The lower apertures of the tubes were closed by silicone discs, which ameliorated adherence of EHTs and prevented intrusion of medium. The silicone tubes were elastic and obeyed Hooke’s law for springs. They formed a defined resistance for beating EHTs, which could be expressed as force or spring constant k, its value under baseline conditions being 0.95 N/m (Online Fig. II). Inserting a metal brace into the two silicone tubes increased the resistance opposed by the silicone posts by a factor of 12 (k = 11.5 N/m). Metal braces were handmade from stainless steel Nubryte®Wire 016 (GAC, size 016, length: 14′′). The conditions after insertion of metal braces are referred to as afterload enhanced (AE) in this study. The preload of EHTs was 0.8 mN (tension generated by the silicone posts in diastole; i.e. resting state).
Cell culture conditions
EHTs were maintained in 37 °C, 7 % CO2, and 40 % O2 humidified cell culture incubators throughout experiments. EHT medium for the first 8 days of culture consisted of DMEM (Biochrom F0415), 10 % horse serum inactivated (Gibco 26050), 2 % chick embryo extract, 1 % penicillin/streptomycin (Gibco 15140), insulin (10 μg/mL, Sigma-Aldrich 857653), and aprotinin (33 μg/mL, Sigma-Aldrich A1153). Between day 8 and 13 medium was used with the same composition except for reduced horse serum content (4 %). Up to day 13 medium was changed three times per week, afterwards daily. From day 13 on EHTs were kept in serum-free medium, i.e. the above medium without horse serum plus triiodothyronine (T3, 0.5 ng/mL, European Commission—Joint Research Centre IRMM-469). Supplementation of low concentrations of hydrocortisone (50 ng/mL, Sigma-Aldrich H0888) from day 13 on increased spontaneous beating rate and thereby reduced experimental drop-out rates (due to lack of beating) to less than 5 %. EHTs started to beat coherently one week after casting. Thereafter EHTs were filmed every other day. The movies were analyzed by a customized software (CTMV) which automatically calculated force, frequency, fractional shortening, contraction time T1 and relaxation time T2 of each single EHT, as recently published (Fig. 7a, b; Online Movie I) .
On day 14 hypertrophic interventions started. For most experiments we studied four different groups in parallel (n = 4–6 per group, one 24-well culture dish). Afterload of EHTs was increased by inserting metal braces (AE)—a mere mechanical intervention—in one group. In two other groups hypertrophic growth was triggered by either phenylephrine (PE, (R)–(–)–Phenylephrine hydrochloride, Sigma-Aldrich P6126) at 20 μmol/L or endothelin-1 (ET-1, Sigma-Aldrich E7764) at 5 nmol/L. Most interventions were carried out for 7 days (except cell size measurements and whole genome gene expression analysis, which both were analyzed after 5 days). EHTs without metal braces and without pharmacological agents served as controls (Online Fig. III). Endothelin receptors were blocked for selected experiments in the AE group. Endothelin receptor antagonists (ET-RA) in a concentration well above IC50 were added to the medium 2 h before metal braces were inserted and medium for all subsequent medium changes also contained ET-RAs (Online Table II).
Establishment of serum-free cell culture conditions for EHTs
Activation of the hypertrophic gene program
To corroborate our mRNA findings on the protein level, Western Blot analysis of β-MHC and immunohistochemistry of ANP were performed. β-MHC levels were low in control EHTs and robustly increased by AE and pharmacological stimulation (Fig. 2b). Immunohistochemistry staining conditions for ANP were optimized on adult rat heart tissue (Fig. 2c). The majority of atrial cardiomyocytes were ANP positive with a distinct perinuclear staining, whereas hardly any ANP signal was detectable in rat ventricles. Under the same staining conditions control EHTs showed weak and diffuse ANP signals only in scattered cells at the surface of the EHT. In AE-EHTs many cardiomyocytes throughout the tissue construct were ANP positive, also with the characteristic perinuclear staining. Distinct ANP positive cardiomyocytes were also observed in ET-1- and PE-treated EHTs, but preferentially at the free edges (Fig. 2d).
Cardiomyocytes in EHTs enlarge with increased afterload or pharmacological stimulation
Unbiased analysis of alterations in gene expression and its comparison to TAC in mice
Compared to control EHTs between 572 and 1,405 genes were differentially expressed in the three hypertrophy groups. The fold change threshold for upregulation was set to 1.5, for downregulation to 0.66, accordingly. In detail, 648 genes were upregulated and 388 downregulated by AE. ET-1-stimulation upregulated 385 genes and downregulated 187 genes, PE induced 776 and repressed 629 genes. All typically upregulated genes in the hypertrophic gene program (see above) were represented in the list of the most strongly upregulated genes (Online Table IV). Notably, in the AE group 3 out of 4 typically upregulated hypertrophy-associated genes (Myh7, Acta1, Nppa) were found among the 5 most strongly upregulated genes (of overall 22,000 investigated genes). The fourth hypertrophic gene Nppb was ranked on position 60.
After conversion of up- or downregulated rat gene symbols in the AE group to mouse gene symbols they were compared to two mouse TAC data sets (Fig. 4 b–e): TAC I from Toischer et al.  was based on an Affymetrix® expression array analysis, TAC II from Lee et al.  was obtained by RNA-Seq and provided data for short-term TAC (hypertrophic state) and long-term TAC (failing state). Seven genes (amongst others Myh7, Acta1, Nppa) were upregulated irrespective of species and analysis system in all three afterload enhancement procedures, i.e. AE-EHT, TAC I and short-term TAC II (Fig. 4b). The same analysis for long-term TAC II revealed more shared upregulated genes, e.g. extracellular matrix proteins and profibrotic or fibroblast markers (Fig. 4c). Genes collectively downregulated in AE, TAC I and long-term TAC II included Adra1b (alpha 1B adrenoceptor) and Atp2a2 (Serca2a) (Fig. 4e).
An enrichment analysis revealed three molecular KEGG pathways in which genes upregulated in all 3 EHT hypertrophy groups (AE, ET-1, PE) were overrepresented: rno 04512 extracellular matrix, rno 03010 ribosome, and rno 00010 glycolysis. As transcriptional activation of ribosomal proteins seemed evident in a model of cardiac hypertrophy we subsequently focused on investigating glycolysis and extracellular matrix of EHTs under mechanical or pharmacological stimulation.
Increased glycolysis in AE-EHTs
Fibrosis in hypertrophied EHTs
Fibrosis is a hallmark of pathological cardiac hypertrophy and the fibrosis signals in the enrichment analysis and the comparison to mouse TACs prompted us to further investigate remodeling of EHTs after hypertrophic stimulation. Transcript concentrations of important structural and regulatory proteins of cardiac connective tissue were extracted from the whole-transcriptome analysis and verified by reverse transcription followed by quantitative PCR (Online Table V). Collagen-1 and collagen-3 are the major ECM components of the healthy heart and collagen-1 in particular is augmented in fibrotic hearts. In EHTs, transcript concentration of the collagen-1 pro-alpha chain Col1a1 was increased by AE (1.55×), ET-1 (1.81×) and PE (1.33×). Fibronectin 1, a collagen linker protein, was also upregulated by AE (1.48×), ET-1 (2.16×) and PE (2.08×). In contrast, elastin and fibrillins, both elements of elastic fibers, were downregulated in all 3 hypertrophy groups: elastin (AE 0.40×; ET-1 0.51×; PE 0.12×) and fibrillin-1 (AE 0.55×; ET-1 0.64×; PE 0.51×). No changes were observed in the transcript concentrations of TGFβ1, in contrast to its downstream target connective tissue growth factor (CTGF), which was induced in all hypertrophy groups (AE 1.92×; ET-1 1.86×; PE 1.70×).
Functional consequences of afterload enhancement
To distinguish whether impaired contractility reflected real effects of the hypertrophic interventions or simple injury of the tissue by mechanical manipulation, we inserted metal braces for only short periods of time. Even after repeating this procedure on three consecutive days it did not impair contractility, but rather had a slight stimulatory effect (Online Fig. VI). Markers of cell death or tissue injury remained unchanged compared to control EHTs. Interestingly, a comparison of EHTs cast on silicone posts with low spring constants (k = 0.29) to our standard spring constant without metal braces (k = 0.95) revealed higher forces and bigger cardiomyocytes indicating that at the low range of spring constants EHTs might also be suitable for investigating physiological hypertrophy (Online Fig. VI).
Prevention of the deleterious afterload enhancement effects by endothelin receptor blockade
In this study we developed a new model of cardiac hypertrophy based on our EHT technology. The main findings were as follows: (1) Low concentrations of triiodothyronine (T3) in cell culture medium allowed the cultivation of beating EHTs without serum for more than a week. (2) A sole increase of afterload of EHTs triggered the activation of the hypertrophic gene program and (3) lead to an increase in cardiomyocyte size, which was measurable by myocyte specific dystrophin staining. (4) Cardiomyocyte hypertrophy could also be induced by the α-agonist phenylephrine or endothelin-1. (5) Gene expression patterns in the mechanical and pharmacological groups showed large overlap, but were greatly different from control EHTs. (6) Glucose consumption per beat was increased in afterload enhanced EHTs. (7) All 3 hypertrophic interventions lead to fibrotic activation as well as sustained impairment of contractile force and relaxation both of which were preventable by endothelin receptor blockade. The data show, as we believe for the first time, that an increase in afterload, independent of systemic neurohumoral activation or a mismatch between blood supply and demand, is sufficient to induce the whole spectrum of pathological hypertrophy in cardiac muscle tissue.
Cardiac preload is defined as end-diastolic wall tension. In this study, we refer to preload as the tension exerted by the silicone posts on EHTs in the resting state (metered value 0.8 mN, calculated from the observed deflection of silicone posts in the resting state). The load opposing shortening of the ventricular muscle fibers is termed ventricular afterload . This load in EHTs is the sum of the preload and the tension produced by the contraction of the EHT. The latter is considerably increased when EHTs have to beat against stiff metal braces. Besides neurohumoral activation chronically increased hemodynamic load is the main trigger of cardiac hypertrophy in patients . We evaluated both in EHTs. On the one hand phenylephrine (α-adrenergic agonist) , and endothelin-1 (ETA and ETB receptor agonist) [26, 34] were chosen as well-established humoral inductors of cardiomyocyte hypertrophy. Both agonists bind to receptors which are coupled to Gq/G11-proteins and subsequently activate phospholipase C. The consequence of mechanical load imposed on rat cardiomyocytes in vitro was evaluated in numerous studies by others  and us , or even on human cardiomyocytes derived from embryonic or induced pluripotent stem cells . In these studies cardiomyocytes were stretched passively, i.e. they were subjected to preload enhancement . Other in vitro models of cardiac hypertrophy on the basis of a short-term culture of excised rabbit trabeculae also investigated preload increases [5, 18]. However, data from large epidemiological studies such as the Framingham Heart Study revealed that afterload enhancement (and especially sustained hypertension) is far more important than preload enhancement for the development of heart failure . This finding was also supported in a comparison of TAC mice (afterload) and shunt mice (preload) .
A crucial point in developing the new in vitro model of hypertrophy was to reduce hypertrophic culture medium supplements. Our previously published EHT protocol relied on culture medium containing 10 % horse serum . As serum is a strong inducer of cardiomyocyte hypertrophy , we reduced horse serum concentration during EHT development and completely omitted it one day before starting hypertrophic interventions. This was possible by supplementing the medium with triiodothyronine (T3), well known to be essential for cardiac development [7, 31]. T3 itself is capable of provoking cardiomyocyte hypertrophy, although this form is usually referred to as physiological. Taking this into consideration, a T3 concentration in the lower physiological range was chosen, approximately 100-fold lower than that reported to induce hypertrophy .
The stimulation of EHTs with phenylephrine or endothelin-1 induced the canonical hypertrophic (or fetal) gene program with upregulation of ANP, BNP, β-MHC and α-skeletal actin as well as downregulation of SERCA2a [9, 38]. Remarkably, sole afterload enhancement showed almost an identical pattern not only regarding the direction but also the extent of changes. Importantly, altered transcript concentrations were accompanied by similar changes in the protein concentrations as shown for two representative examples (ANP and β-MHC). Of note, we chose the β-MHC antibody clone NOQ 7.5.4D, which has recently been reported by the Simpson group as the only specific available β-MHC antibody . As nicely illustrated in that study, the fetal gene program does not necessarily indicate hypertrophy or even pathological hypertrophy. We therefore concentrated on precisely determining cardiomyocyte size. Cell sizes were measured in situ, i.e. in fixed native EHTs, to prevent any changes in cell morphology caused by enzymatic isolation from their extracellular matrix. A specific sarcolemma staining with a dystrophin antibody ensured that only myocytes were included in the analysis and facilitated automated and blinded measurement of cardiomyocyte sizes. The increase in cross-sectional area of cardiomyocytes by approximately 30 % was similar in all hypertrophy groups and thus within a reasonable and frequently reported range [1, 15].
The microarray analysis displayed that control and hypertrophied EHTs differed distinctly in their expression patterns. The upregulated genes in AE-EHTs displayed modest overlap with upregulated genes in mouse TAC. However, even gene expression between two identical procedures in the same species (Mouse TAC I and II) was widely different. Strikingly, the total overlap of upregulated genes in all three groups (with short-term TAC II) was very consistent with the “hypertrophic gene program” and the overlap increased considerably (particularly by fibrotic markers) when including long-term TAC II, which the authors referred to as failing phenotype .
Two KEGG pathways, Glycolysis and Extracellular Matrix, were identified in the enrichment analysis of the transcriptome to be overrepresented in the panel of upregulated genes of all three hypertrophy groups. Alterations in cardiomyocyte energetics in pathological cardiac hypertrophy with a shift from fatty acids (normally 60–90 %) to glucose as the primary myocardial energy substrate are described [3, 8, 27, 36]. However, these experiments relied on hypertrophied myocardium (in vivo) which after explantation was analyzed in vitro. We observed an upregulation of glycolytic genes and an increase in glucose consumption per beat (+48 %) in EHTs hypertrophied by AE—despite lower contractile forces. To our knowledge, this is the first time this metabolic shift has been observed in vitro in its entirety. Apparently, cardiac tissue has a native metabolic plasticity without the need for regulation by the whole organism.
Cardiac fibrosis is an undisputed criterion of pathological hypertrophy . Increased collagen-1 deposition has been associated with increased myocardial stiffness [2, 28] and a decrease in the elastin to collagen ratio has been found in decompensated cardiac hypertrophy . In our model, the increased gene expression of several pro-fibrotic genes, the downregulation of fibrillin and elastin and the immunohistochemical detection of increased deposition of freshly synthesized collagen-1 indicate that afterload enhancement as well as the pharmacological interventions induced a fibrotic response as part of a pathological remodeling process. These findings highlight two advantages of EHTs. On the one hand, fibrin as ECM for tissue engineering allows unambiguous detection of freshly synthesized collagen-1 whereas collagen-1 as used by us previously  does not. On the other hand, EHTs likely enable “more physiological” crosstalk between cell types than 2-D cultures, because all native ventricular heart cell types are present and organized in a 3-D tissue-like structure. The present findings indeed show that EHTs allow studying processes as complex as cardiac fibrosis, known to involve activated fibroblasts (myofibroblasts), endothelial cells, cardiomyocytes and monocytes [2, 24]. Whether the remodeling involved (cell-type-specific) changes in cell proliferation was not evaluated and requires further studies.
A unique advantage of EHTs is the possibility to measure contractile function over a long period of time. The observed 35 % loss of contractile force after AE was initially surprising because it is to be viewed on the background of increased myocyte size (+28 %). However, the sum of myocyte cross-sectional area in AE was almost identical to that in control, indicating loss of some myocytes on the one hand and reduced force generation per muscle mass on the other. Moreover, relaxation of AE-EHTs was clearly impaired as it is known from hypertrophied myocardium . These parameters suggest that the cardiomyocyte hypertrophy after AE was pathological and not the consequence of a physiological training effect (which we observed when we increased spring constants of the elastic posts from k = 0.29 [solid] to 0.95 [hollow]) or reversal of an unphysiological “starving process” due to lack of serum. An unambiguous classification of cardiac hypertrophy into a physiological or pathological form is usually not possible due to intermediate conditions . However, most phenomena we observed are generally attributed to pathological hypertrophy: increase of α-skeletal actin and β-MHC with concomitant decrease of thyroid hormone receptors α and β , decrease of SERCA2a, a metabolic shift towards glycolysis, Gq-coupled signaling, cardiac fibrosis, loss of contractility, and impaired relaxation [8, 14]. We believe that these observations are important for several reasons. First, as far as we are aware of, it is the first time that all of these hallmarks of pathological hypertrophy can be measured and are reproduced together in an in vitro model. Second, it shows that afterload enhancement as such induces pathological hypertrophy, independent of other well known factors such as systemic neurohumoral activation  and the mismatch between the growth of blood vessels and myocytes . Intriguingly, the deleterious effects of afterload enhancement on contractile parameters and fibrosis were preventable by endothelin blockers, nicely supporting either the early concept of autocrine or paracrine release of ET-1 and Ang II [30, 42], or a direct mechanically induced activation of ETA-receptors as proposed for angiotensin II-receptors , or both. The fact that we did not find altered endothelin gene expression (data not shown) argues for the predominant importance of the ligand-independent mechanism. Third, the model is simple, robust and, in contrast to other cardiomyocyte cell culture techniques, stable for weeks. It is highly standardized, reproducible and can be easily established in any standard laboratory with animal facility. Thus, it will be useful in investigating molecular processes involved in pathological cardiac hypertrophy and in evaluating potential therapeutic approaches.
We thank Melanie Neumann of the HEXT Mouse Pathology Facility for assistance in histological analyses, Malik Alawi of the HEXT Bioinformatics Service Facility for assistance with expression array data, Bülent Aksehirlioglu for helping in manufacturing the silicone racks, June Uebeler for technical assistance, and Dagmar Claussen for taking photographs. Furthermore, we thank the Core Facility for High-throughput Genetics and Genomics of the Leibniz Institute for Arteriosclerosis Research (LIFA) for performing the microarray experiments. This study was supported by funds from the Deutsche Forschungsgemeinschaft (HA 3423/3-1, Es 88/12-1), Deutsche Herzstiftung (F/13/10), European Union (FP6 EUGene Heart, FP7 Angioscaff). HEXT is supported by funds from the Freie und Hansestadt Hamburg.
Conflict of interest
This article is distributed under the terms of the Creative Commons Attribution License which permits any use, distribution, and reproduction in any medium, provided the original author(s) and the source are credited.