Cardiac Inflammation Contributes to Changes in the Extracellular Matrix in Patients With Heart Failure and Normal Ejection Fraction

Transcription

1 Cardiac Inflammation Contributes to Changes in the Extracellular Matrix in Patients With Heart Failure and Normal Ejection Fraction Dirk Westermann, MD*; Diana Lindner, PhD*; Mario Kasner, MD; Christine Zietsch, PhD; K. Savvatis, MD; F. Escher, MD; J. von Schlippenbach, MD; C. Skurk, MD; Paul Steendijk, PhD; Alexander Riad, MD; Wolfgang Poller, MD; Heinz-Peter Schultheiss, MD; Carsten Tschöpe, MD Downloaded from by guest on June 3, 218 Background The pathophysiology of heart failure with normal ejection fraction (HFNEF) is still under discussion. Here we report the influence of cardiac inflammation on extracellular matrix (ECM) remodeling in patients with HFNEF. Methods and Results We investigated left ventricular systolic and diastolic function in 2 patients with HFNEF and 8 control patients by conductance catheter methods and echocardiography. Endomyocardial biopsy samples were also obtained, and ECM proteins as well as cardiac inflammatory cells were investigated. Primary human cardiac fibroblasts were outgrown from the endomyocardial biopsy samples to investigate the gene expression of ECM proteins after stimulation with transforming growth factor-. Diastolic dysfunction was present in the HFNEF patients compared with the control patients. In endomyocardial biopsy samples from HFNEF patients, we found an accumulation of cardiac collagen, which was accompanied by a decrease in the major collagenase system (matrix metalloproteinase-1) in the heart. Moreover, a subset of inflammatory cells, which expressed the profibrotic growth factor transforming growth factor-, could be documented in the HFNEF patients. Stimulation of primary human cardiac fibroblasts from HFNEF patients with transforming growth factor- resulted in transdifferentiation of fibroblasts to myofibroblasts, which produced more collagen and decreased the amount of matrix metalloproteinase-1, the major collagenase in the human heart. A positive correlation between cardiac collagen, as well as the amount of inflammatory cells, and diastolic dysfunction was evident and suggests a direct influence of inflammation on fibrosis triggering diastolic dysfunction. Conclusions Cardiac inflammation contributes to diastolic dysfunction in HFNEF by triggering the accumulation of ECM. (Circ Heart Fail. 211;4:44-52.) Key Words: collagen inflammation remodeling diastolic dysfunction Patients with heart failure with normal ejection fraction (HFNEF) have increased mortality, 1 4 and morbidity has also been found to be higher compared with patients with HF and reduced EF. 5 One of the underlying mechanisms leading to the clinical symptomatology in patients with HFNEF is diastolic function abnormality with increased diastolic stiffness, but nondiastolic function abnormalities with exerciseinduced changes in systolic velocity, chronotropic incompetence, and ventricular-vascular uncoupling have also been demonstrated to contribute to this disease Despite the growing prevalence of HFNEF in the past 15 years, a disease affecting about half of the HF population, knowledge about its molecular mechanisms is still limited, especially because the pathways leading to HFNEF are not confined to 1 pathology. Intracellular changes with elevated cardiomyocyte resting tension, as well as a shift in titin isoforms, are important in patients with severe HFNEF. 19,2 Furthermore, accumulation of cardiac collagen was shown to be present in this disease and to contribute to the aggravation of diastolic function Enhanced endothelial migration of inflammatory cells into the myocardium might influence the development of these changes, especially in regard to changes in the extracellular matrix (ECM). 24,25 Editorial see p 5 Clinical Perspective on p 52 We hypothesized that inflammation is 1 important trigger of cardiac fibrosis and therefore it plays an important role in HFNEF. We show herein that increased inflammation triggers collagen accumulation in HFNEF patients and that both Received December 15, 29; accepted September 29, 21. From the Department of Cardiology and Pneumology (D.W., D.L., M.K., C.Z., K.S., F.E., J.v.S., C.S., A.R., W.P., H.-P.S., C.T.), Charité, Universititäts-Medizin Berlin, Campus Benjamin Franklin, Berlin, Germany, and Department of Cardiology (P.S.), Leiden University Medical Center, Leiden, The Netherlands. *Drs Westermann and Lindner contributed equally to this work. Correspondence to Dirk Westermann, MD, Department of Cardiology, Campus Benjamin Franklin, Hindenburgdamm 3, 122 Berlin, Germany. dirk.westermann@web.de 211 American Heart Association, Inc. Circ Heart Fail is available at DOI: /CIRCHEARTFAILURE

2 Westermann et al Inflammation and Fibrosis in HFNEF 45 inflammation and fibrosis are correlated to diastolic dysfunction. We show that inflammatory cells express transforming growth factor (TGF)-, which directly induces changes in cardiac fibroblasts as increasing collagen accumulation and decreasing the degradation system (matrix metalloproteinase [MMP]-1 and their tissue inhibitor [TIMP]). This was associated with a transdifferentiation of cardiac fibroblasts to myofibroblasts, known to be a potent contributor of pathological remodelling in diseased tissues. These findings suggest that cardiac inflammation by increased transendothelial migration is 1 potent trigger of diastolic dysfunction in the HFNEF population. Collagen type I P=.1 P<.1 Collagen type III I P=.2 P=.3 Downloaded from by guest on June 3, 218 Methods Study Population Patients presenting with New York Heart Association class II and III HF, an EF 5%, and diastolic dysfunction on echocardiography were evaluated for participation in this study. Persistent atrial fibrillation and pulmonary diseases were excluded. Significant coronary artery disease or heart valve diseases were ruled out by angiography. Clinically relevant parvovirus B19 infection with a genome equivalent 5 was ruled out in all patients included in the study. 26 Twenty patients met the inclusion criteria, were enrolled, and are referred to as the HFNEF group, with respect to recent guidelines. 27 In addition, 8 patients without signs of congestive HF who underwent coronary angiography for evaluation of chest pain were also enrolled in this study and served as controls. In all patients, endomyocardial biopsy samples were obtained the day after hemodynamic function was analyzed. Cardiac conditions were stable before catheterization in all patients. All patients gave informed, written consent. The research protocol was approved by the local institutional review board. Pressure-Volume Measurements and Echocardiography The conductance catheter was used to assess pressure-volume measurements in all patients, as recently described in more detail. 6 In brief, left ventricular (LV) diastolic function was characterized by LV end-diastolic pressure and isovolumetric relaxation ( ). Furthermore, we calculated the exponential curve fit to the diastolic LV pressure-volume points during a transient preload reduction to determine the load-independent diastolic stiffness constant (LV stiffness constant ). We analyzed LV end-systolic pressure and EF as a parameter of systolic function. The slope of the end-systolic pressure-volume relation was calculated as a load-independent parameter for cardiac contractility. Mitral and pulmonary venous Doppler flow velocities were recorded in the apical 4-chamber view with a VingMed System FiVe (GE Healthcare, Chalfont St. Giles, UK) as well as the LV filling index by the ratio of transmitral flow velocity to annular velocity (E/E lateral), as previously described. 7 Right ventricular systolic pressure (in mm Hg) was measured before the endomyocardial biopsy samples were taken. Endomyocardial Biopsy Endomyocardial biopsy samples were obtained from the right side of the ventricular septum of each patient with use of a flexible bioptome (Westmed) via the femoral vein approach. Right ventricular systolic pressure was measured. The tissue pieces were frozen in LN 2 and stored at 7 C for subsequent analysis. 28 From all patients, 1 endomyocardial biopsy sample was used for immunohistological staining, and 1 was used for measuring gene expression. (All measurements were performed once, n 2 in HFNEF and n 8 in controls.) Collag gen type I gene expression relative e to control 2 1 Collag gen type III gene expression relative e to control Cell Culture Primary cardiac fibroblasts were obtained by outgrowth from biopsy specimens from 5 HFNEF patients and were incubated in Iscove s basal medium (Biochrom AG, Berlin, Germany) containing 1% human serum, 1% fetal calf serum (FCS), 1 U/mL penicillin (Biochrom AG), and 1 g/ml streptomycin (Biochrom). The cardiac fibroblasts were seeded in 24-well plates, and when the culture was 8% confluent, they were serum-starved in Iscove s basal medium containing.5% FCS (PAA, Cölbe, Germany) for 16 hours. They were then stimulated with 5 ng/ml TGF- 1 (PeproTech, Hamburg, Germany) for 6 or 24 hours compared with no stimulation for 6 or 24 hours. All experiments were done with cells between the second and fourth passage with n 6 wells per experiment for all time points and all patients. For each patient, a mean was calculated from the 6 cell culture wells for each experiment (basal conditions and stimulated for all time points), and the means of all patients are presented in the figures. Native fibroblasts were stained with antibodies against CD31, desmin, vimentin, P4HB, and -smooth muscle actin and compared with HL1 (experimental atrial cardiomyocytes) as well as with endothelial cells. The monocyte cell line THP-1 was cultured in RPMI 164 medium (PAA) containing 1% FCS (PAA), 1 U/mL penicillin, and 1 g/ml streptomycin (PAA). Type I Carboxy-Terminal Telopeptide of Collagen and Propeptide of Procollagen Type I in Serum Serum was collected 1 hour before invasive measurements, and type I carboxy-terminal telopeptide as a breakdown product of collagen type 1, as well as procollagen type I as a marker of type I collagen synthesis, was measured as described previously. 29 RNA Isolation Total RNA was extracted from the myocardial biopsy samples or cell culture wells by the Trizol method (GIBCO BRL, Carlsbad, Calif) Figure 1. Box-and-whisker plots showing increased levels of collagen types I and III proteins in endomyocardial biopsy samples from patients with HFNEF compared with controls. Increases in the collagen I to III ratio as well of collagen type I C-telopeptide were measured in serum. The mrna abundance of collagen types I and type III was increased in HFNEF compared with control subjects. AF indicates area fractions. *P.5 vs controls.

3 46 Circ Heart Fail January P=.4 P=.2 P=.1 Downloaded from by guest on June 3, 218 MMP-1 1 VCAM P=.4 TIMP-1 % DHE positv ve nuclei MMP-2 Additional purification was performed with the ChargeSwitch total RNA cell kit (Invitrogen, Karlsruhe, Germany). The yield of purified total RNA was analyzed by checking the UV absorbance at 26 nm on a NanoDrop ND-1 (Agilent Technologies, Boeblingen, Germany) spectrophotometer. Real-Time Reverse Transcription Polymerase Chain Reaction The ABI-inventoried TaqMan gene expression assays (each included forward and reverse primers as well as a fluorescently marked probe) used for preamplification (from the biopsy samples) and for the real-time reverse transcription polymerase chain reaction were obtained from Applied Biosystems. Prior real-time reverse transcription polymerase chain reaction cdna samples (1 to 25 ng) of the biopsy samples were preamplified with pooled gene expression assays with the TaqMan PreAmp Master Mix (early access) in a final volume of 25 L. Quantification of housekeeping CDKN1B transcripts as an internal control for the amount and quality of cdna was performed for all samples. Gene expression for collagen types I and III was analyzed from cardiac biopsy samples (primers from Applied Biosystems). Data for cell culture were also normalized to human CDKN1B mrna levels as an endogenous control (unaffected by TGF- treatment) and are expressed relative to untreated controls according to the formula 2 CT. TGF- ELISA The THP-1 cells were starved in Iscove s medium containing.5% FCS (PAA), 1 U/mL penicillin, and 1 g/ml streptomycin (PAA) 16 hours before the experiment. THP-1 cells (3 millions [ml]) were treated with 1 ng/ml phorbol 12-myristate 13-acetate P= Figure 2. Decreases in protein levels of MMP-1 and increases in levels of MMP-2 and TIMP-1 were found in endomyocardial biopsy samples from patients with HFNEF compared with controls. Increased oxidative stress by DHE staining was documented in HFNEF compared with control subjects. The adhesion molecule VCAM-1 was also increased in HFNEF compared with control subjects. AF indicates area fractions. *P.5 vs controls. (Sigma-Aldrich) for 12, 24, or 48 hours. The cell culture supernatant was then used for measuring the total TGF- 1 concentration with the TGF 1 E max immunoassay system (Promega). The cell culture supernatant was first diluted 1:5 in Dulbecco s phosphate-buffered saline (PAA) and then acidified with HCl to ph 2.6 for 15 minutes, followed by neutralization with NaOH before using the sample for the TGF 1 E max immunoassay system. Histology Immunohistological staining was performed by standard techniques. The antibodies were purchased from Chemicon (anti-mmp1, CD3, CD11a, CD45, vascular cell adhesion molecule [VCAM]-1, and collagen types I and III) and Calbiochem (anti-timp-1 and anti MMP-2). Immunohistochemical staining images were quantified by digital image analysis. For double immunofluorescence, we used Tissue Tec embedded cryosections. The following primary antibodies were used: mouse anti -sarcomere actin (1:5, Santa Cruz) and mouse anti TGF- (1:5, Serotec). The secondary antibodies were labeled with fluorescein isothiocyanate and tetramethylrhodamine isothiocyanate (Sigma, St. Louis, Mo). Nuclei were visualized with 4,6-diamidino-2-phenylindole (Sigma). Slides were embedded in fluorescent mounting medium (DakoCytomation, Glostrup, Denmark), coverslipped, and analyzed by fluorescence microscopy (Zeiss, Jena, Germany). Sections were incubated with 1 mol/l dihydroethidium to detect O 2 and visualized by fluorescence microscopy, whereby DHE-positive nuclei were counted with respect to the number of total nuclei. Statistical Analysis Data are shown in box-and-whisker plots (Figures 1 through 3) and as mean SEM (Figures 4 and 5). For comparison of the HFNEF 3 8 P<.1 P=.2 P< D3 + cells / mm C 2 1 CD D11a + cells / mm m CD D45 + cells / mm m Figure 3. Increased numbers of CD3, CD11, and CD45 cells were found in endomyocardial biopsy samples from patients with HFNEF compared with controls. *P.5 vs controls.

4 Westermann et al Inflammation and Fibrosis in HFNEF 47 Downloaded from by guest on June 3, 218 Figure 4. A, Representative histologic image with double staining of CD11 cells with the profibrotic growth factor TGF-. B, In vitro experiments with phorbol 12-myristate 13-acetate activated monocytes (THP-1 cells) showed increased production of TGF- mrna and protein levels of TGF- in a time-dependent manner. *P.5 vs individual controls. group with the control group, the nonparametric Mann Whitney U test was used for data that were not normally distributed. Fisher s exact test was used to analyze categorical variables. A probability value.5 was considered statistically significant. Data were analyzed with Graphpad 5.1 (PRISM, San Diego, Calif) and SPSS (version 15.; SPSS Inc, Chicago, Ill). Results Patient Demographics Patient characteristics are summarized in Table 1. There were no significant differences between HFNEF patients and controls with respect to age, sex, or body surface area. The prevalence of hypertension and diabetes mellitus was not statistically different between the groups (Table 1). Hemodynamic Data Systolic function was not changed in HFNEF compared with controls. In contrast, diastolic dysfunction was evident in the HFNEF group, with an increased LV stiffness constant, increased LV end-diastolic pressure, and prolonged isovolumetric relaxation. Echocardiography revealed that the cardiac dimensions of both groups were not significantly different. Diastolic dysfunction with increased E/E lateral was documented in the HFNEF group. Moreover, significant cardiac hypertrophy was documented in HFNEF compared with control subjects (Table 2). Right ventricular systolic pressure, as 1 parameter of pulmonary hypertension, was not significantly different in HFNEF and controls (29.7; range, 24 to 33 mm Hg, vs 25; range, 21 to 28.5 mm Hg; P.4). Figure 5. In vitro cell culture experiments of primary cardiac fibroblasts stimulated with TGF- for 6 or 24 hours. Increased gene expression of -smooth muscle actin (a-sma), connective tissue growth factor (CTGF), and collagen type I (Col I) were found in stimulated cells compared with unstimulated cells. MMP-1 was decreased after stimulation with TGF-. MMP-2 was increased after stimulation. *P.5 compared to the relative controls.

5 48 Circ Heart Fail January 211 Downloaded from by guest on June 3, 218 Table 1. Characteristics of the HFNEF and Control Groups Characteristics Control (n 8) HFNEF (n 2) P Value Age, y 56 (5 69) 6 (43 68).21* NYHA class II 12 NYHA class III 8 NT-pro-BNP, pg/ml 51 (28 67) 573 ( ).1* Medications -blockers ACE inhibitors/arbs Ca 2 channel blockers Diuretics Concomitant diseases Hypertension Diabetes mellitus 6.14 NYHA indicates New York Heart Association; NT-pro-BNP, NT-pro-brain natriuretic peptide; ACE, angiotensin-converting enzyme; and ARB, angiotensin receptor blocker. Patient characteristics and data are shown as median (25% 75% percentile). *t test. Fisher s exact test. Extracellular Matrix Collagen type I and III expression in the cardiac biopsy samples was significantly higher in patients with HFNEF than in controls. Moreover, mrna abundance of both collagen subtypes was increased in HFNEF patients (Figure 1). The ratio of collagen type I to type III was increased in HFNEF patients. TGF- mrna expression in the biopsy samples was increased in HFNEF patients compared with controls (59% compared with controls, P.4). Histologic expression levels of cardiac MMP-1, the major human collagenase, were decreased in HFNEF, whereas TIMP-1 was increased in HFNEF individuals compared with controls (Figure 2). Table 2. Echocardiographic and Hemodynamic Results Control HFNEF P Values Chamber dimensions, mm LVEDD 47 (42 49) 46 (4 5).573* Septum 9.8 ( ) 12.3 ( ).1* Posterior wall 9.2 ( ) 11.4 ( ).1* Tissue Doppler E/E 6.1 ( ) 13.9 ( ).11* (lateral) Pressure-volume loops LVP, mm Hg 131 ( ) 136 ( ).412 dp/dt max, mm Hg/s 1783 ( ) 1882 ( ).535 ESPVR, mm Hg/mL 1.1 (.8 1.3) 1.2 (1 1.5).386 LVEDP, mm Hg 6 (4 9) 16 (14 24).1*, ms 48 (4 49) 64 (63 78).1* Diastolic stiffness.1 (.6.4).6 (.3.12).1* LVEDD, LV end-diastolic diameter; LVP, LV pressure; ESPVR, end-systolic pressure volume relation; LVEDP, LV end-diastolic pressure; and, isovolumetric relaxation time. Data are shown as median (25% 75% percentile). *t test. Type I carboxy-terminal telopeptide, as 1 serum marker for collagen degradation, was increased in the sera of HFNEF patients compared with controls (mean SEM, vs g/l; P.25). Procollagen type I, a serum marker for collagen production, was also increased in the HFNEF group compared with controls (mean SEM, vs g/l; P.18). Inflammation The inflammatory cells marked by CD3, CD11a, and CD45 were all increased in the cardiac tissues of the HFNEF patients compared with controls (Figure 3). Moreover, the adhesion molecule VCAM-1 was increased in HFNEF compared with control patients (Figure 2). In double staining, inflammatory cells revealed a secretion of the profibrotic growth factor TGF- (Figure 4). More cells (mostly cardiomyocytes and endothelial cells) in the HFNEF group were positive after DHE staining as a marker of oxygen radical production in comparison with controls (Figure 2). No change in systemic inflammation as measured by C-reactive protein was found between the groups (1.3.5 vs mg/dl; P NS). Fibroblast Cell Culture After stimulation with TGF- for 6 or 24 hours, connective tissue growth factor and collagen type 1A1 mrnas were upregulated compared with controls. The transdifferentiation to myofibroblasts, as suggested by -smooth muscle actin mrna expression, was observed after 24 hours of stimulation with TGF-. MMP-1 was not increased after 6 or 24 hours, but MMP-2 was upregulated after 24 hours. With an increased TIMP-1 expression, the MMP-1 to TIMP-1 ratio as 1 indicator of MMP-1 activity was downregulated after stimulation with TGF- for 6 or 24 hours (Figure 5). THP-1 Cell Culture Activated THP-1 cells produced TGF- mrna compared with their nonactivated controls. This was followed by a time-dependent production of TGF- protein, which showed significantly increased protein levels 24 hours after stimulation of the THP-1 monocytes (Figure 4B). Discussion The salient finding of the current study is that cardiac inflammatory cells documented in endomyocardial biopsy samples from patients with HFNEF induce extracellular remodelling with increased accumulation of collagen. TGF-, excreted by inflammatory cells, is 1 potent stimulus for the regulatory changes of the ECM, as shown in cell culture experiments with primary cardiac fibroblasts derived from endomyocardial biopsy samples of patients with HFNEF. These changes contribute to diastolic dysfunction, which is 1 of the underlying pathologies of HFNEF, which was documented by a correlation between collagen (as well as inflammatory cells) and diastolic dysfunction.

6 Westermann et al Inflammation and Fibrosis in HFNEF 49 Downloaded from by guest on June 3, 218 The molecular changes leading to diastolic dysfunction are still debated, and different pathways can affect the pathology of HFNEF. Altered isoform expression of the giant muscle protein titin can determine the elastic properties of the heart. 19 Moreover, increased myocyte tension, which can be prevented by experimental phosphorylation of titin, contributes to diastolic dysfunction in patients with severe HFNEF. 2,3 Additional modulatory effects on titin stiffness may arise from disulfide bonding under oxidant stress also affecting LV compliance. 31 Nevertheless, the steep part of the diastolic pressure-volume relation is mainly modified by the ECM, as can be observed when comparing stretch lengths between whole-muscle stripes and single myocytes, 32 suggesting that excess collagen might further aggravate diastolic dysfunction. Cardiac collagen is a stable protein with a low turnover (8 to 12 days), 33 but its balance can be disrupted in pathological conditions. Next to ischemia, which may lead to replacing fibrosis, for example, increased wall stress, angiotensin II, and TGF- may induce profibrotic processes leading to pathological tissue fibrosis. 34 The tensile strength of collagen type 1 approximates that of steel. Therefore, LV chamber stiffness in vivo will be affected by the amount of cardiac ECM. 35 In this study, we showed that total cardiac collagen content was increased in patients with HFNEF, in agreement with other work that showed increased fibrosis in this disease. 19,21,36 Furthermore, not only was collagen type I found to be increased in this study population, but also there was a change in the collagen type I to III ratio in favor of the stiffer collagen type I, similar to that in patients with systolic HF. 37 In addition to increased protein levels of collagen types I and III, their mrna abundance was also increased in endomyocardial biopsy samples from patients with HFNEF compared with controls. Another marker of excessive collagen production is type I carboxy-terminal telopeptide, a degradation product of collagen with increased collagen turnover together with propeptide of procollagen type I, a serum marker of collagen production, both of which were increased in the HFNEF group. Oxidative stress was increased in HFNEF patients, as DHE staining revealed, which might be the result of increased LV stiffness. Multiple studies have shown that 1 of the best-known inducers of collagen production is the profibrotic growth factor TGF-. It also has profound effects on ECM homeostasis, in part through its ability to alter the balance between MMPs and their TIMP inhibitors. The endogenous collagen degradation system is regulated by increased activity of MMPs overcoming their tissue inhibitors. 38 Nevertheless, there are different MMPs, and the substrate affinity of those proteases is different. 39 MMP-1 (interstitial collagenase) is known to degrade collagen fibers, and therefore it likely favors collagen degradation. Through the activity of the activator protein-1 transcription factor, TGF- can, on one hand, repress MMP-1 gene expression 4 and, on the other hand, increase TIMP-1 expression. Congruent with these ex vivo data, we have show an upregulation of TIMP-1 protein and a downregulation of MMP-1 protein levels in the biopsy samples from HFNEF patients, which leads to a significant decrease in the MMP-1 TIMP-1 ratio. This inhibition of the collagen degradation system could be 1 mechanism contributing to the accumulation of ECM in HFNEF patients and the initiation of diastolic dysfunction over a longer time period. 41 Interestingly, Lopez and colleagues 22 showed that this ratio was increased in patients with systolic HF, whereas it was unchanged in their hypertensive HF group. Whether this process is dynamic or changes with time, or whether this mechanism represents a distinct difference between both HF subtypes remains to be clarified. Furthermore, and in contrast to the activator protein-1 mediated downregulation of MMP-1, we found increased levels of MMP-2 (controlled by activator protein-2), which is a known gelatinase and has substrate affinity for denatured fibrillar collagen as well as for the basement membrane. 42 Some studies have described MMP-2 serum levels in patients with hypertensive and diastolic HF, but the results are inconsistent, with some studies showing an increase, 43 and others showing no difference or even a decrease in MMP Recently, increased levels of MMP-2 were shown to predict HF in patients with diastolic dysfunction and hypertension. In that study, MMP-2 was a better prognostic marker than the well-known HF biomarker brain natriuretic peptide. 29 Several experimental studies in MMP-2 knockout animals have helped to advance understanding of the molecular function of MMP-2. Matsumura and colleagues 44 showed in MMP-2 knockout mice a decrease in invading inflammatory cells and a decrease in LV rupture after myocardial infarction. Those authors demonstrated that destruction of basement membrane proteins facilitates the transendothelial migration of immunocompetent cells, thereby triggering cardiac inflammation. In light of such findings, we investigated the number of inflammatory cells in our patients and showed that HFNEF is associated with increased cardiac inflammation, with high numbers of CD3, CD11a, and CD45 cells. Moreover, the adhesion molecule VCAM-1, which attracts immunocompetent cells to the endothelium and initiates transendothelial migration, was also increased in the HFNEF group. This result is especially interesting, because VCAM-1 is upregulated by angiotensin II, which may be increased in regard to known risk factors like hypertension and diabetes mellitus in HFNEF. Recently, it was shown that immunocompetent cells like T-cells (CD3 ) can indeed alter tissue remodelling in vitro, 45 and we have shown that cardiac inflammation is associated with excessive collagen accumulation in experimental diabetic cardiomyopathy in 1 animal model of HFNEF. 25 There is experimental and clinical evidence that inflammatory cells might modulate cardiac function in HF with reduced 46,47 and normal 48,49 EF. The direct effects of these cells are still under debate, but it has been suggested that increased inflammation is associated with the development of systolic HF by distinct changes in the ratio of MMP to TIMP. 5,51 In line with these data, we have demonstrated herein that these inflammatory cells express TGF- in cardiac tissue, as shown by immunohistochemical double staining. Moreover, THP-1 cells, when they become acti-

7 5 Circ Heart Fail January 211 Downloaded from by guest on June 3, 218 vated, which is 1 hallmark of transendothelial migration, produce TGF- on the mrna and protein level, as evidenced by in vitro studies of phorbol 12-myristate 13-acetate activated THP-1 cells. In addition, TGF- gene expression was increased in the HFNEF biopsy samples. To test the direct effect of TGF- expressed by inflammatory cells in this study, we performed experiments in a cell culture system with primary human cardiac fibroblasts derived from endomyocardial biopsy samples of HFNEF patients. After stimulation with TGF-, fibroblasts expressed increased amounts of -smooth muscle actin, 1 marker for their transdifferentiation into pathologically activated myofibroblasts. Connective tissue growth factor was concomitantly increased, leading to higher collagen gene expression after 6 and 24 hours, which explains the collagen accumulation found in the endomyocardial biopsy samples. On the other hand, the degradation system of collagen, especially the MMP-1 to TIMP-1 ratio, was downregulated after stimulation with TGF- in the cell culture system, which is another explanation for the extensive fibrosis seen in HFNEF hearts. MMP-2, in line with its known function in degrading the basement membrane to allow for easier transendothelial migration of inflammatory cells into cardiac tissue, was upregulated after TGF- stimulation. Interestingly, TGF- gene expression was increased after stimulation, which suggests that the already transdifferentiated myofibroblasts might have induced further activation of fibroblasts and therefore enhanced the profibrotic process in a paracrine fashion. Viral agents are 1 possible inducer of cardiac inflammation. The parvovirus B19, which is associated with endothelial and diastolic dysfunction, 28 is known to cause myocarditis. Nevertheless, it was recently shown that in patients with dilated cardiomyopathy, a viral load of only 5 genome equivalents is a clinically relevant threshold for the maintenance of myocardial inflammation. 26 Because this extent of viral load was ruled out in all patients, the exact mechanism of cardiac inflammation has still to be determined, although animal models of HF are often associated with cardiac inflammation. 24 Recently, it was shown 52 that activated myofibroblasts also produce chemokines to fuel the inflammatory process, which might induce a vicious circle of inflammation triggering fibrosis and diastolic dysfunction. If hemodynamic changes in HFNEF induce proinflammatory changes to start cardiac inflammation, these have to be determined in detail in further studies. We suggest that small numbers of invading inflammatory cells stimulate cardiac fibrosis by expressing TGF- and inducing a pathological transdifferentiation from fibroblasts to myofibroblasts. This will not only stimulate gene expression of collagen but also induce a decline in MMP-1 activity and an increase in MMP-2 mrna abundance, as well as paracrine TGF- production. Increased stiffness due to ECM accumulation will increase oxidative stress, suggested to be associated with increased endothelial activation. Together with increased MMP-2, which has been suggested to disrupt the basal membrane, this might induce a vicious circle fuelling inflammation leading to fibrosis and ultimately, to progression of the disease. Conclusions In this study, HFNEF was characterized by an increase in cardiac inflammation. This inflammatory process triggers cardiac collagen accumulation by inducing collagen gene expression and inhibiting the cardiac degradation system. Inhibiting the transendothelial migration of inflammatory cells into cardiac tissue might be a future therapeutic concept in HFNEF. Sources of Funding This study was supported by the Deutsche Forschungsgemeinschaft (SFB-TR-19, A2, and B5) and by the European Commission, Research Directorate General, FP7-Health-21, MEDIA (26149), Brussels, Belgium. None. Disclosures References 1. Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, Redfield M. Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med. 26;355: Bhatia RS, Tu JV, Lee DS, Austin PC, Fang J, Haouzi A, Gong Y, Liu P. Outcome of heart failure with preserved ejection fraction in a population-based study. N Engl J Med. 26;355: Tribouilloy C, Rusinaru D, Mahjoub H, Souliere V, Levy F, Peltier M, Slama M, Massy Z. Prognosis of heart failure with preserved ejection fraction: a 5 year prospective population-based study. Eur Heart J. 28;29: Gaasch WH, Delorey DE, Kueffer FJ, Zile R. Distribution of left ventricular ejection fraction in patients with ischemic and hypertensive heart disease and chronic heart failure. Am J Cardiol. 29; 14: Ahmed A, Zile MR, Rich MW, Fleg JL, Adams KF Jr, Love TE, Young JB, Aronow WS, Kitzman DW, Gheorghiade M, Dell Italia LJ. Hospitalizations due to unstable angina pectoris in diastolic and systolic heart failure. Am J Cardiol. 27;99: Westermann D, Kasner M, Steendijk P, Spillmann F, Riad A, Weitmann K, Hoffmann W, Poller W, Pauschinger M, Schultheiss HP, Tschope C. Role of left ventricular stiffness in heart failure with normal ejection fraction. Circulation. 28;117: Kasner M, Westermann D, Steendijk P, Gaub R, Wilkenshoff U, Weitmann K, Hoffmann W, Poller W, Schultheiss HP, Pauschinger, M. Tschope C. Utility of Doppler echocardiography and tissue Doppler imaging in the estimation of diastolic function in heart failure with normal ejection fraction: a comparative Doppler-conductance catheterization study. Circulation. 27;116: Zile MR, Baicu CF, Gaasch H. Diastolic heart failure abnormalities in active relaxation and passive stiffness of the left ventricle. N Engl J Med. 24;35: Lam CS, Roger VL, Rodeheffer RJ, Bursi F, Borlaug BA, Ommen SR, Kass DA, Redfield M. Cardiac structure and ventricular-vascular function in persons with heart failure and preserved ejection fraction from Olmsted County, Minnesota. Circulation. 27;115: Phan TT, Abozguia K, Nallur Shivu G, Mahadevan G, Ahmed I, Williams L, Dwivedi G, Patel K, Steendijk P, Ashrafian H, Henning A, Frenneaux M. Heart failure with preserved ejection fraction is characterized by dynamic impairment of active relaxation and contraction of the left ventricle on exercise and associated with myocardial energy deficiency. J Am Coll Cardiol. 29;54: Wachter R, Schmidt-Schweda S, Westermann D, Post H, Edelmann F, Kasner M, Luers C, Steendijk P, Hasenfuss G, Tschope C, Pieske B. Blunted frequency-dependent upregulation of cardiac output is related to

8 Westermann et al Inflammation and Fibrosis in HFNEF 51 Downloaded from by guest on June 3, 218 impaired relaxation in diastolic heart failure. Eur Heart J. 29;3: Borlaug BA, Melenovsky V, Russell SD, Kessler K, Pacak K, Becker LC, Kass A. Impaired chronotropic and vasodilator reserves limit exercise capacity in patients with heart failure and a preserved ejection fraction. Circulation. 26;114: Tschope C, Westermann D. Heart failure with normal ejection fraction: pathophysiology, diagnosis, and treatment. Herz. 29;34: Borlaug BA, Lam CS, Roger VL, Rodeheffer RJ, Redfield M. Contractility and ventricular systolic stiffening in hypertensive heart disease: insights into the pathogenesis of heart failure with preserved ejection fraction. J Am Coll Cardiol. 29;54: Tan YT, Wenzelburger F, Lee E, Heatlie G, Leyva F, Patel K, Frenneaux M, Sanderson E. The pathophysiology of heart failure with normal ejection fraction: exercise echocardiography reveals complex abnormalities of both systolic and diastolic ventricular function involving torsion, untwist, and longitudinal motion. J Am Coll Cardiol. 29;54: Tschope C, Paulus WJ. Is echocardiographic evaluation of diastolic function useful in determining clinical care? Doppler echocardiography yields dubious estimates of left ventricular diastolic pressures. Circulation. 29;12: 81 82; discussion Penicka M, Bartunek J, Trakalova H, Hrabakova H, Maruskova M, Karasek J, Kocka V. Heart failure with preserved ejection fraction in outpatients with unexplained dyspnea: a pressure-volume loop analysis. J Am Coll Cardiol. 21;55: Deleted in proof. 19. van Heerebeek L, Borbely A, Niessen HW, Bronzwaer JG, van der Velden J, Stienen GJ, Linke WA, Laarman GJ, Paulus J. Myocardial structure and function differ in systolic and diastolic heart failure. Circulation. 26;113: Borbely A, van der Velden J, Papp Z, Bronzwaer JG, Edes I, Stienen GJ, Paulus J. Cardiomyocyte stiffness in diastolic heart failure. Circulation. 25;111: van Heerebeek L, Hamdani N, Handoko ML, Falcao-Pires I, Musters RJ, Kupreishvili K, Ijsselmuiden AJ, Schalkwijk CG, Bronzwaer JG, Diamant M, Borbely A, van der Velden J, Stienen GJ, Laarman GJ, Niessen HW, Paulus J. Diastolic stiffness of the failing diabetic heart: importance of fibrosis, advanced glycation end products, and myocyte resting tension. Circulation. 28;117: Lopez B, Gonzalez A, Querejeta R, Larman M, Diez J. Alterations in the pattern of collagen deposition may contribute to the deterioration of systolic function in hypertensive patients with heart failure. J Am Coll Cardiol. 26;48: Muller-Brunotte R, Kahan T, Lopez B, Edner M, Gonzalez A, Diez J, Malmqvist K. Myocardial fibrosis and diastolic dysfunction in patients with hypertension: results from the Swedish Irbesartan Left Ventricular Hypertrophy Investigation versus Atenolol (SILVHIA). J Hypertens. 27;25: Heymans S, Hirsch E, Anker SD, Aukrust P, Balligand JL, Cohen- Tervaert JW, Drexler H, Filippatos G, Felix SB, Gullestad L, Hilfiker- Kleiner D, Janssens S, Latini R, Neubauer G, Paulus WJ, Pieske B, Ponikowski P, Schroen B, Schultheiss HP, Tschope C, Van Bilsen M, Zannad F, McMurray J, Shah M. Inflammation as a therapeutic target in heart failure? a scientific statement from the Translational Research Committee of the Heart Failure Association of the European Society of Cardiology. Eur J Heart Fail. 29;11: Westermann D, Rutschow S, Jager S, Linderer A, Anker S, Riad A, Unger T, Schultheiss HP, Pauschinger M, Tschope C. Contributions of inflammation and cardiac matrix metalloproteinase activity to cardiac failure in diabetic cardiomyopathy: the role of angiotensin type 1 receptor antagonism. Diabetes. 27;56: Bock CT, Klingel K, Kandolf R. Human parvovirus B19-associated myocarditis. N Engl J Med. 21;362: Paulus WJ, Tschope C, Sanderson JE, Rusconi C, Flachskampf FA, Rademakers FE, Marino P, Smiseth OA, De Keulenaer G, Leite-Moreira AF, Borbely A, Edes I, Handoko ML, Heymans S, Pezzali N, Pieske B, Dickstein K, Fraser AG, Brutsaert L. How to diagnose diastolic heart failure: a consensus statement on the diagnosis of heart failure with normal left ventricular ejection fraction by the Heart Failure and Echocardiography Associations of the European Society of Cardiology. Eur Heart J. 27;28: Tschope C, Bock CT, Kasner M, Noutsias M, Westermann D, Schwimmbeck PL, Pauschinger M, Poller WC, Kuhl U, Kandolf R, Schultheiss P. High prevalence of cardiac parvovirus B19 infection in patients with isolated left ventricular diastolic dysfunction. Circulation. 25;111: Martos R, Baugh J, Ledwidge M, O Loughlin C, Murphy NF, Conlon C, Patle A, Donnelly SC, McDonald K. Diagnosis of heart failure with preserved ejection fraction: improved accuracy with the use of markers of collagen turnover. Eur J Heart Fail. 29;11: Jaber WA, Maniu C, Krysiak J, Shapiro BP, Meyer DM, Linke WA, Redfield M. Titin isoforms, extracellular matrix, and global chamber remodeling in experimental dilated cardiomyopathy: functional implications and mechanistic insight. Circ Heart Fail. 28;1: Grutzner A, Garcia-Manyes S, Kotter S, Badilla CL, Fernandez JM, Linke A. Modulation of titin-based stiffness by disulfide bonding in the cardiac titin N2-B unique sequence. Biophys J. 29;97: Kruger M, Linke WA. Titin-based mechanical signalling in normal and failing myocardium. J Mol Cell Cardiol. 29;46: Laurent GJ. Dynamic state of collagen: pathways of collagen degradation in vivo and their possible role in regulation of collagen mass. Am J Physiol. 1987;252:C1 C Weber KT. Extracellular matrix remodeling in heart failure: a role for de novo angiotensin II generation. Circulation. 1997;96: Zile MR, Brutsaert DL. New concepts in diastolic dysfunction and diastolic heart failure, part I: diagnosis, prognosis, and measurements of diastolic function. Circulation. 22;15: Querejeta R, Lopez B, Gonzalez A, Sanchez E, Larman M, Martinez Ubago JL, Diez J. Increased collagen type I synthesis in patients with heart failure of hypertensive origin: relation to myocardial fibrosis. Circulation. 24;11: Pauschinger M, Knopf D, Petschauer S, Doerner A, Poller W, Schwimmbeck PL, Kuhl U, Schultheiss P. Dilated cardiomyopathy is associated with significant changes in collagen type I/III ratio. Circulation. 1999;99: Spinale FG. Matrix metalloproteinases: regulation and dysregulation in the failing heart. Circ Res. 22;9: Westermann D, Savvatis K, Schultheiss HP, Tschope C. Immunomodulation and matrix metalloproteinases in viral myocarditis. J Mol Cell Cardiol. 21;48: Hall MC, Young DA, Waters JG, Rowan AD, Chantry A, Edwards DR, Clark M. The comparative role of activator protein 1 and Smad factors in the regulation of TIMP-1 and MMP-1 gene expression by transforming growth factor- 1. J Biol Chem. 23;278: Heymans S, Schroen B, Vermeersch P, Milting H, Gao F, Kassner A, Gillijns H, Herijgers P, Flameng W, Carmeliet P, Van de Werf F, Pinto YM, Janssens S. Increased cardiac expression of tissue inhibitor of metalloproteinase-1 and tissue inhibitor of metalloproteinase-2 is related to cardiac fibrosis and dysfunction in the chronic pressure-overloaded human heart. Circulation. 25;112: Spinale FG. Myocardial matrix remodeling and the matrix metalloproteinases: influence on cardiac form and function. Physiol Rev. 27;87: Martos R, Baugh J, Ledwidge M, O Loughlin C, Conlon C, Patle A, Donnelly SC, McDonald K. Diastolic heart failure: evidence of increased myocardial collagen turnover linked to diastolic dysfunction. Circulation. 27;115: Matsumura S, Iwanaga S, Mochizuki S, Okamoto H, Ogawa S, Okada Y. Targeted deletion or pharmacological inhibition of MMP-2 prevents cardiac rupture after myocardial infarction in mice. J Clin Invest. 25; 115: Mikko M, Fredriksson K, Wahlstrom J, Eriksson P, Grunewald J, Skold M. Human T cells stimulate fibroblast-mediated degradation of extracellular matrix in vitro. Clin Exp Immunol. 28;151: Abbate A, Bonanno E, Mauriello A, Bussani R, Biondi-Zoccai GG, Liuzzo G, Leone AM, Silvestri F, Dobrina A, Baldi F, Pandolfi F, Biasucci LM, Baldi A, Spagnoli LG, Crea F. Widespread myocardial inflammation and infarct-related artery patency. Circulation. 24;11: Lisman KA, Stetson SJ, Koerner MM, Farmer JA, Torre-Amione G. The role of inflammation in the pathogenesis of heart failure. Curr Cardiol Rep. 22;4: Sciarretta S, Ferrucci A, Ciavarella GM, De Paolis P, Venturelli V, Tocci G, De Biase L, Rubattu S, Volpe M. Markers of inflammation and fibrosis are related to cardiovascular damage in hypertensive patients with metabolic syndrome. Am J Hypertens. 27;2: Westermann D, Van Linthout S, Dhayat S, Dhayat N, Schmidt A, Noutsias M, Song XY, Spillmann F, Riad A, Schultheiss HP, Tschope C.

9 52 Circ Heart Fail January 211 Tumor necrosis factor- antagonism protects from myocardial inflammation and fibrosis in experimental diabetic cardiomyopathy. Basic Res Cardiol. 27;12: Heymans S, Pauschinger M, De Palma A, Kallwellis-Opara A, Rutschow S, Swinnen M, Vanhoutte D, Gao F, Torpai R, Baker AH, Padalko E, Neyts J, Schultheiss HP, Van de Werf F, Carmeliet P, Pinto M. Inhibition of urokinase-type plasminogen activator or matrix metalloproteinases prevents cardiac injury and dysfunction during viral myocarditis. Circulation. 26; 114: Towbin JA. Inflammatory cardiomyopathy: there is a specific matrix destruction in the course of the disease. Ernst Schering Research Foundation Workshop. Springer, Heidelberg, Germany. 26: Souders CA, Bowers SL, Baudino A. Cardiac fibroblast: the renaissance cell. Circ Res. 29;15: Downloaded from by guest on June 3, 218 CLINICAL PERSPECTIVE Heart failure with normal ejection fraction (HFNEF) can be diagnosed in 5% of patients with the clinical syndrome of HF. Nevertheless, the HFNEF syndrome seems to be heterogeneous and therefore will present with different pathologies leading to the disease. It is assumed that an accumulation of cardiac collagen is 1 important mediator in the development of HFNEF. We show here that increased transendothelial migration of inflammatory cells triggers a profibrotic phenotype by activating fibroblasts into myofibroblasts in cardiac tissue. Activated myofibroblasts, characterized by -smooth muscle actin, are well known for excessive collagen production in other diseased tissues and express significantly more collagen compared with normal cardiac fibroblasts. Moreover, inflammatory cells reduced the expression of the collagen degradation system, the matrix metalloproteinases, again promoting cardiac fibrosis as 1 trigger for a stiff and noncompliant ventricle. This transdifferentiation was induced by transforming growth factor-, which was expressed by inflammatory monocytes in the heart. Interestingly, we document here that these inflammatory cells can be found more frequently in cardiac tissue of the HFNEF group than in the control group. Therefore, continuous low-grade inflammation might be 1 key player in the development and/or progression of HFNEF. This would make anti-inflammatory strategies inhibiting transendothelial migration of cells into the cardiac tissue an interesting approach in treating HFNEF, a disease plagued by limited data.

10 Downloaded from by guest on June 3, 218 Cardiac Inflammation Contributes to Changes in the Extracellular Matrix in Patients With Heart Failure and Normal Ejection Fraction Dirk Westermann, Diana Lindner, Mario Kasner, Christine Zietsch, K. Savvatis, F. Escher, J. von Schlippenbach, C. Skurk, Paul Steendijk, Alexander Riad, Wolfgang Poller, Heinz-Peter Schultheiss and Carsten Tschöpe Circ Heart Fail. 211;4:44-52; originally published online November 12, 21; doi: /CIRCHEARTFAILURE Circulation: Heart Failure is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX Copyright 21 American Heart Association, Inc. All rights reserved. Print ISSN: Online ISSN: The online version of this article, along with updated information and services, is located on the World Wide Web at: Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in Circulation: Heart Failure can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. Once the online version of the published article for which permission is being requested is located, click Request Permissions in the middle column of the Web page under Services. Further information about this process is available in the Permissions and Rights Question and Answer document. Reprints: Information about reprints can be found online at: Subscriptions: Information about subscribing to Circulation: Heart Failure is online at: