Cardiovascular Research 66 (2005) 410 419 www.elsevier.

com/locate/cardiores

Age-dependent changes in myocardial matrix metalloproteinase/tissue inhibitor of metalloproteinase profiles and fibroblast function
Merry L. Lindsey a,*, Danielle K. Goshorna, Christina E. Squiresa, G. Patricia Escobar a, Jennifer W. Hendricka, Joseph T. Mingoiaa, Sarah E. Sweterlitscha, Francis G. Spinalea,b
a

Division of Cardiothoracic Surgery Research, Room 629, Strom Thurmond Research Building, 770 MUSC Complex, Medical University of South Carolina, 114 Doughty Street, P.O. Box 250778, Charleston, SC 29425, United States b Ralph A. Johnson Veterans Administration Medical Center, Charleston, SC 29425, United States Received 30 August 2004; received in revised form 3 November 2004; accepted 24 November 2004 Avaialble online 13 December 2004 Time for primary review 28 days

Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

Abstract Objective: To evaluate the effects of aging on left ventricular (LV) geometry, collagen levels, matrix metalloproteinase (MMP) and tissue inhibitor of metalloproteinase (TIMP) abundance, and myocardial fibroblast function. Methods: Young (3-month-old; n =28), middle-aged (MA; 15-month-old; n =17), and old (23-month-old; n =16) CB6F1 mice of both sexes were used in this study. Echocardiographic parameters were measured; collagen, MMP, and TIMP levels were determined for both the soluble and insoluble protein fractions; and fibroblast function was evaluated. Results: LV end-diastolic dimensions and wall thickness increased in both MA and old mice, accompanied by increased soluble protein and decreased insoluble collagen. Immunoblotting revealed differential MMP/TIMP profiles. Compared to MA levels, MMP-3, MMP-8, MMP9, MMP-12, and MMP-14 increased, and TIMP-3 and TIMP-4 decreased in the insoluble fraction of old mice, suggesting increased extracellular matrix (ECM) degradative capacity. Fibroblast proliferation was blunted with age. Conclusion: This study, for the first time, identified specific differences in cellular and extracellular processes that likely contribute to agedependent ECM remodeling. D 2004 European Society of Cardiology. Published by Elsevier B.V. All rights reserved.
Keywords: Matrix metalloproteinase; Tissue inhibitor of metalloproteinase; Extracellular matrix; Matrix remodeling

1. Introduction Cardiovascular disease is a leading cause of morbidity and mortality that occurs with increasing incidence in the elderly [1,2]. With aging, the myocardium undergoes structural remodeling and hypertrophy. An important component of structural remodeling is remodeling of the extracellular matrix (ECM). The ECM integrates cell and tissue function by providing a scaffold on which cells migrate, grow, and differentiate [3]. A major constituent of the ECM is fibrillar collagen. Under basal conditions in the
* Corresponding author. Tel.: +1 843 876 5186; fax: +1 843 876 5187. E-mail address: lindseml@musc.edu (M.L. Lindsey).

young adult, collagen turnover occurs at a rate that allows for normal replacement and the maintenance of left ventricular (LV) structure and function. The determinants that regulate collagen turnover, particularly as a function of aging, remain poorly understood. Matrix metalloproteinases (MMPs) are key enzymes involved in ECM turnover. The MMP family is comprised of more than 25 individual members divided into specific classes based on in vitro substrate specificity for various ECM components. MMP activity is inhibited by tissue inhibitors of metalloproteinases (TIMPs), a family currently composed of four members [4]. MMPs are involved in several cardiovascular disease processes, including plaque rupture [5,6]; aneurysm formation and rupture [7]; LV

0008-6363/$ - see front matter D 2004 European Society of Cardiology. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.cardiores.2004.11.029

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419

411

remodeling following pressure and/or volume overload [8 12]; and during all stages of congestive heart failure progression [1315]. Whether age-related changes in MMPs and TIMPs provide a molecular mechanism for changes in ECM structure has not been demonstrated. The myocardial fibroblast is a predominant cell regulating ECM turnover by altering (1) ECM synthesis and deposition, (2) ECM degradation and turnover through the production and release of MMPs, and/or (3) mechanical tension on the collagen network [16]. Based on previous observations, age-related changes in fibroblast function could provide a cellular mechanism for changes in ECM structure. Therefore, the goal of this study was to evaluate aging mice for changes in LV structure; altered levels of collagen, MMPs, and TIMPs; and differences in fibroblast function.

free wall. One half was snap frozen for immunoblotting and the other half was used immediately for myocardial fibroblast isolation. 2.3. Protein isolation and collagen content A highly reproducible sample fractionation protocol based on differential protein solubility was used to extract proteins from LV samples. To extract easily soluble proteins (e.g., cytoplasmic proteins), the LV half was homogenized in soluble lysis buffer (0.25 M sucrose and 1 mM EDTA) and centrifuged at 13,000 g for 10 min [19]. To extract insoluble proteins (e.g., membrane proteins), the insoluble pellet was resuspended in chaotropic membrane extraction reagent (7 M urea, 2 M thiourea, and the detergent amidosulfobetaine-14; Sigma). Fractionation purity was confirmed by immunoblotting for thioredoxin (N90% in the soluble fraction) and SERCA2 (N95% in the insoluble fraction). Protein concentrations were determined using the Bradford assay. Because of the high urea content, insoluble protein extracts were diluted 1:40 to ensure compatibility with the Bradford assay. All samples (5 Ag) were run on gels to confirm the accuracy of protein concentrations. Collagen volume fractions were determined in protein extracts using the microplate picrosirius red assay [20,21]. Soluble collagen was measured in the soluble fraction; insoluble collagen was measured in the insoluble fraction. Equal amounts of myocardial extracts (10 Ag total protein) were added to triplicate wells of a 96-well microtiter plate. The samples were dried in the incubator and stained for 1 h with 100 Al of 0.1% picrosirius red in saturated picric acid (w/v). The dye was solubilized in 100 Al of 0.1 M NaOH, and the plates were read by spectrophotometry (Multiskan MCC/340) at an absorbance of 540 nm. Vitrogen 100 purified collagen (Collagen Biomaterials) was used as a positive control and to generate a standard curve. The amount of collagen per 10 Ag total protein was obtained from the standard curve and multiplied by the total protein to give total collagen levels. Total collagen levels were divided by the initial LV wet weight to obtain microgram collagen per milligram LV wet weight. 2.4. Immunoblotting Immunoblotting was performed as described previously [22] using antibodies for MMP-2, MMP-3, MMP-9, MMP12, MMP-13, MMP-14, TIMP-1, and TIMP-4 (Chemicon), MMP-7, MMP-8, and TIMP-2 (Oncogene), TIMP-3 and Mac-3 (Cedarlane), and the discoidin domain receptor (DDR2; Genex Bioscience). With the exception of MMP2, the MMP antibodies recognized both pro and active forms (Table 1). A sample of breast cancer homogenate was run on each gel as a positive control and to verify the correct molecular weight size band. Recombinant proteins were also used as positive controls when further confirmation was required. All samples for each set were run on one of two

Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

2. Methods 2.1. Mice CB6F1 mice are the F1 hybrid of C57BL/6 and BALB/c mice and were selected because they are genetically heterogeneous and display hybrid vigor. Young (3 months old; n =11 female and 17 male), middle-aged (MA; 15 months old; n =6 female and 12 male), and old (23 months old; n =6 female and 11 male) CB6F1 mice were purchased from the Aged Rodent Colonies maintained by the National Institute of Aging. The age groups (3, 15, and 23 months old) correspond to young adult, middle-aged adult, and old but not senescent adult, respectively. All animal procedures were conducted in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 8523, revised 1996). All studies were approved by the Institutional Animal Care and Use Committee at the Medical University of South Carolina. 2.2. Echocardiography and tissue collection For echocardiographic acquisition, the mice were anesthetized with 12% isoflurane (depending on the individual mouse), with heart rates maintained at z400 beats per minute [17]. Echocardiographic acquisition and analysis were performed using the Sonos 5500 (Agilent Technologies) equipped with a high-band linear 15.6 MHz transducer to obtain m-mode images. LV dimensions and wall thickness were measured as previously described [18]. Echocardiography was not performed on one MA and one old mouse because the mice died before echocardiograms were acquired. For tissue collection, the mice were anesthetized with 5% isoflurane, the coronary vasculature was flushed with saline, and the heart was excised. The LV and RV were separated and weighed individually. The LV was divided longitudinally into two halves so that each half contained septum and

412

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419

Table 1 MMP and TIMP antibodies used for immunoblotting Antibody MMP-2 MMP-3 MMP-7 MMP-8 MMP-9 MMP-12 MMP-13 MMP-14 TIMP-1 TIMP-2 TIMP-3 TIMP-4 Company Chemicon Chemicon Oncogene Oncogene Chemicon Chemicon Chemicon Chemicon Chemicon Oncogene Cedarlane Chemicon Catalog # AB19015 AB810 PC492 PC493 AB804 AB19051 AB8114 AB815 AB8116 IM11L CL2T3 AB816 MMP forms recognized pro pro, pro, pro, pro, pro, pro, pro, active active active active active active active

26-well 420% Criterion TrisHCl gels (Bio-Rad). The two gels and two blots were handled identically and simultaneously to minimize intergel variability. Ponceau staining of the blots confirmed the successful transfer, and the positive control sample, which was run on both gels, served to confirm consistency between the two blots. 2.5. Myocardial fibroblast isolation Myocardial fibroblast isolation from mice is technically challenging, and yields vary tremendously depending on which strain of mice is used (unpublished observation). Several mouse myocardial fibroblast isolation protocols have been published, from which we developed the following protocol [2325]. The LV half was placed into cold Dulbeccos modified eagle media (DMEM; Gibco-BRL/ Invitrogen; Grand Island, NY) with 10% fetal bovine serum (FBS) and 1% antibioticantimycotic solution (penicillin, streptomycin, amphotericin B; Cellgro) and used immediately for fibroblast isolation using the outgrowth technique. Each LV was used to establish separate primary cultures. The cells displayed characteristic fibroblast morphology, and cell cultures were confirmed to be pure fibroblasts by immunocytochemistry using a composite of antibodies. The antibodies used were against vimentin (Sigma; positive), desmin (Sigma; negative), factor VIII (Sigma; negative), the discoidin domain receptor 2 (DDR2; Genex Bioscience; positive), a smooth muscle actin (Sigma; positive), and proly-4-hydroxylase (Chemicon; positive) [26,27]. Passage 14 cells were used for the functional assays. 2.6. Fibroblast functional assays and protein levels Proliferation, migration, and adhesion indexes were evaluated using protocols published for mice, rat, and/or dog and modified as described below. For proliferation assays, myocardial fibroblasts were plated at 1104 per well in quadruplicate on 0.2% gelatin-coated 24-well plates and incubated for 24 h. The wells were washed with phosphatebuffered saline, and the cells were fixed for 20 min with

zincformalin (Anatech). The plates were stained with 1% methylene blue. After eluting stain with acid alcohol (0.05 M HCl in 50% ethanol), the plates were read at an absorbance of 620 nm [2830]. To compare migration properties, fibroblasts were maintained in serum-free media (DMEM supplemented with a 1 solution of insulin, transferrin, and selenium, 0.1% bovine serum albumin, 10 Ag/mL ascorbate, and 1 antibiotic/antimycotic solution) for 24 h and then passaged. Serum starved cells were seeded (3104 per well in duplicate wells) into the upper compartment of a transwell tissue-culture-treated insert (Costar Brand, Corning, Acton, MA) of 24-well tissue culture plates. Preliminary studies were performed to determine optimal cell densities and incubation times. FBS (10%) served as a chemoattractant and was placed in the lower chamber of each well. Following 4 h incubation, fibroblasts that had migrated to the underside of the membrane were fixed and stained with 1% methylene blue as described above. The migration index was calculated as a percentage of the average optical density for the young fibroblast group. For adhesion rates, myocardial fibroblasts were plated at 3104 per well in duplicate on ECM-coated 24-well plates (BD BioCoat Cellware; Becton Dickinson, Franklin Lakes, NJ). Substrates analyzed for fibroblast adhesion characteristics were used at concentrations recommended [31] and included plastic, gelatin (0.2 mg/mL), laminin (2 Ag/mL), fibronectin (2.5 Ag/mL), collagen I (5 Ag/mL), and collagen IV (5 Ag/mL). Plates were incubated for 10 min and then washed with PBS to remove nonadherent cells. Initial experiments using 30- and 60-min incubations demonstrated that saturation was already reached by 10 min. Fibroblasts were fixed and stained with 1% methylene blue as described above. To analyze protein levels in fibroblast cell pellets, confluent fibroblasts were incubated in serum-free media for 48 h, and the cell pellets were homogenized in chaotropic membrane extraction reagent to obtain total protein. Immunoblotting was performed using antibodies for a smooth muscle actin (Sigma) and MMP-9. 2.7. Data analysis Data are presented as meanFstandard error of the mean (S.E.M.). Statistical analyses were performed using Intercooled Stata 8.0 for Windows (Stata, College Station, TX). Echocardiographic, proliferation, and adhesion data were analyzed by repeated measures analysis of variance, with comparisons between individual groups made using the Bonferroni-corrected t -test. A two-tailed value of p b0.05 was considered statistically significant. The migration data were normalized as a percentage of the average value for the young or MA group and analyzed using a one-sample t -test to compare MA and old values to young or MA levels set at 100%. Immunoblotting data were analyzed as total MMP data (which included all forms of the enzyme in the pooled

Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419 Table 2 LV geometry, necropsy data, and total protein levels in aging mice Young Sample size (n ) Heart rate (bpm) End diastolic dimensions (mm) Fractional shortening (%) Posterior wall thickness in diastole (mm) Left ventricle mass (mg) Body weight (g) Left ventricle mass to tibia length (mg/mm) Soluble protein (Ag/mg LV wet wt) Insoluble protein (Ag/mg LV wet wt) Soluble collagen (Ag/mg LV wet wt) Insoluble collagen (Ag/mg LV wet wt) 28 446F9 4.32F0.09 33F1 0.75F0.01 96F2 26.0F0.4 5.6F0.2 28.2F1.4 109.4F6.2 4.3F0.4 17.2F1.4 Middle-aged 17 470F10 4.92F0.09* 29F1 0.82F0.02* 152F7* 39.5F1.3* 8.8F0.5* 35.7F2.2* 103.0F6.3 4.9F0.4 13.7F1.3 Old 16 475F14 4.87F0.12* 30F1 0.84F0.02*

413

progressed in the same direction from the young to MA to old groups; (2) changes that occurred from the young to MA group and then remained stable in the old group; (3) changes that occurred only in the old group; and (4) changes that occurred in one direction in MA and then reversed direction in the old group [32].

3. Results
155F8* 41.7F1.2* 8.3F0.5* 35.1F1.8* 98.9F5.3 4.8F0.4 12.6F0.9*

3.1. Effect of aging on LV morphometrics LV mass, end-diastolic dimensions, and posterior wall thickness increased in the MA and old groups, while fractional shortening was preserved (Table 2). The LV mass calculated from echocardiographic measurements correlated significantly with LV mass taken at necropsy (r =0.85; p b0.001; n =61). 3.2. Effect of aging on total protein and collagen levels Protein from young (n =24), MA (n =15), and old LV samples (n =14) were differentially extracted into soluble and insoluble fractions. The percentage of total protein obtained for the combined fractions per initial wet weight was 14.7F0.5%, 14.2F0.7%, and 14.0F0.7% for young, MA, and old groups ( p =0.68), indicating that equal yields were recovered. Concordant with increases in LV mass, total protein and collagen levels increased in the MA and old groups. Protein

Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

* p b0.05 young vs. middle-aged or old mice.

soluble and insoluble fractions) and for each MMP form individually for each of the two fractions. A one-sample t test was used to compare MA and old values to the average young value. Linear regressions were performed to determine relationships between groups, using age as the independent variable. Changes that occurred with aging were divided into four types: (1) changes that continually

Fig. 1. Aging differentially regulates MMP-8 and TIMP-4 expression. Soluble MMP-8 levels were unchanged (A), while insoluble MMP-8 levels increased in the old age group (B)+breast cancer homogenate positive control sample. Soluble TIMP-4 levels were unchanged (C), while insoluble TIMP-4 levels decreased in the old age group (D). All immunoblots are representative. All samples were run on one of two gels. For each set, two lanes from each of the two gels are shown for each age group.

414

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419

Table 3 Total myocardial MMP and TIMP levels in middle-aged and old mice, as a percentage of young or middle-aged levels Middle-aged (% young) Sample size, n MMP-2 MMP-3 MMP-7 MMP-8 MMP-9 MMP-12 MMP-13 MMP-14 TIMP-1 TIMP-2 TIMP-3 TIMP-4 712 169F40 70F10* 82F18 85F9 68F10* 72F12* 80F10 50F8* 24F3* 79F19 128F12* 102F11 Old (% young) 1011 154F34 113F12 93F10 108F11 117F11 84F10 72F4* 77F7* 60F25 93F13 97F17 81F7* Old (% middle-aged) 1011 91F20 162F17* 114F12 127F12 173F16* 117F14 89F5 154F15* 252F103 119F16 76F13 80F7*

increase in LV mass (Table 2). The increase in soluble protein contributed to an increased soluble to insoluble protein ratio. The ratios were 0.27F0.02, 0.36F0.03, and 0.36F0.02 for young, MA, and old groups, respectively ( p b0.05 for young vs. MA and young vs. old). In contrast, collagen levels decreased in the insoluble fraction, indicating that insoluble collagen levels were not maintained for the given increase in LV mass (Table 2). The decrease in insoluble collagen contributed to an increased soluble to insoluble collagen ratio. The ratios were 0.28F0.03, 0.38F0.03, and 0.41F0.04 for young, MA, and old groups, respectively ( p b0.05 for young vs. old). 3.3. Effect of aging on MMP and TIMP profiles MMP and TIMP levels were determined by immunoblotting. Representative immunoblots are shown in Fig. 1. Total MMP and TIMP levels were analyzed as a percent change from young and MA levels (Table 3). MMP-2, MMP-7, MMP-8, and TIMP-2 levels were not changed between the young and MA or old groups or between MA and old groups. Compared to young levels, MMP-3, MMP9, MMP-12, MMP-14, and TIMP-1 were decreased and TIMP-3 increased in the MA group. MMP-13, MMP-14, and TIMP-4 were decreased in the old group. Compared to MA levels, MMP-3, MMP-9, and MMP-14 were increased and TIMP-4 decreased in the old group.
Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

The densitometric units of all immunoreactive bands in both fractions were combined to obtain a total MMP or TIMP value. Data are presented as a percentage (meanFS.E.M.%) of the average density for the young or middle-aged groups, which were set at 100%. * p b0.05 young vs. middle-aged or old or middle-aged vs. old.

levels were increased in the soluble and insoluble fractions, while collagen levels increased in the soluble fraction only. When normalized to the initial LV wet weight, protein levels remained elevated in the soluble fraction, indicating that the increase in soluble protein was greater in proportion to the

Table 4 Myocardial MMP and TIMP levels in young, middle-aged, and old LV Sample size MW (kD) Soluble Young 1821 MMP-2 MMP-3 MMP-7 MMP-8 MMP-9 MMP-12 MMP-13 MMP-14 72 57 45 28 19 64 58 92 88 54 60 48 65 54 45 40 29 28 24 28 24 989F135 1188F139 648F102 1409+138 ND ND 603F51 381F44 ND 298F45 194F34 141F24 173F24 440F74 189F39 184F27 217F31 1073F203 731+167 1786F350 896F47 ND Middle-aged 1012 1644F402 514F73* 679F226 1153F260 ND ND 486F81 301F36* ND 171F55* 227F53 118F18 120F22* 138F27* 95F13* 120F25* 134F23* 258F35* 575F140 2271F430 856F83 ND Old 1012 1368F373 479F38* 314F61*,y 1315F139 ND ND 557F50 461F79 ND 43F5*,y 96F10*y 90F6*,y 125F13* 314F73y 100F14* 150F14* 68F4*,y 649F266 683F93 2753F443 725F84 ND Insoluble Young 1422 70F11 166F22 143F20 ND ND ND 143F22 833F70 ND 230F34 589F80 ND ND 109F25 ND ND ND ND ND 1331F352 202F35 23F4 Middle-aged 912 141F40 690F135* 267F58 ND ND ND 140F47 504F124* ND 170F45 363F106 ND ND 132F59 ND ND ND ND ND 1591F583 269F57 15F3 Old 1012 265F92 583F99* 693F167*,y ND ND ND 264F51*,y 992F78y ND 262F58 627F70y ND ND 285F68*,y ND ND ND ND ND 113F32*,y 157F29y 20F2y

TIMP-1 TIMP-2 TIMP-3 TIMP-4

Data are presented as densitometric (arbitrary) units (meanFSEM). NDnot detected in that fraction. * p b0.05 young vs. middle-aged or old. y p b0.05 middle-aged vs. old.

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419

415

Fig. 2. Summary of matrix metalloproteinase (MMP) and tissue inhibitor of matrix metalloproteinase (TIMP) changes with aging. Four types of changes occurred: (1) progressive change from young to middle-aged to old groups; (2) change from young to middle-aged and old groups but not between middle-aged and old groups; (3) change only in old group; or (4) change in one direction between young and middle-aged groups then a reverse in direction between middleaged and old groups. z increased; A decreased; NCno change; NDnot detected.

Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

MMP and TIMP levels from each fraction were compared to young levels (Table 4). MMP-7, MMP-13, TIMP-1, and TIMP-2 were detected in the soluble fraction only. MMP-2, MMP-7, TIMP-2, and TIMP-4 levels were not changed between the young and MA or old groups. In the soluble fraction, MMP-3, MMP-9, MMP-12, MMP-13, MMP-14, and TIMP-1 levels decreased in the MA group. Soluble MMP-3, MMP-12, MMP-13, and MMP-14 levels remained decreased in the old group. In the insoluble fraction, MMP-3 levels increased and MMP-9 levels decreased in the MA group. MMP-3, MMP-8, and MMP-14 levels were increased, while TIMP-3 levels were decreased in the insoluble fraction of the old group. MMP and TIMP levels in the old group were also compared to MA levels (Table 4). MMP-2, MMP-7, TIMP1, and TIMP-2 were not different between the MA and old groups. In the soluble fraction of the old group, MMP-3, MMP-12, and MMP-13 levels decreased, and MMP-14 levels increased. In the insoluble fraction of the old group, MMP-3, MMP-8, MMP-9, MMP-12, and MMP-14 levels increased, and the levels of TIMP-3 and TIMP-4 decreased. Changes between the groups are summarized and characterized in Fig. 2.

3.4. Effect of aging on cell numbers in vivo Levels of Mac-3, a macrophage marker, were not different between the groups, suggesting that increased MMP levels were not due to increased macrophage numbers. Levels of DDR2, a fibroblast marker, decreased in the MA and old groups to 64F13% and 74F8% of young levels, respectively ( p b0.05 for both). 3.5. Effect of aging on fibroblast function and MMP-9 expression Fibroblasts are the principal cell regulators of ECM turnover. Given the changes in collagen, MMP, and TIMP levels, we characterized the age-related fibroblast phenotype. We successfully established primary cultures from 58 of the 63 LV samples used for fibroblast isolation. Of the five samples that did not establish, all were from young LV

Fig. 3. Fibroblasts isolated from old (n =13) mice displayed lower proliferative capacity compared to fibroblasts isolated from young (n =19) and middle-aged (n =14) mice (*p b0.05 for young vs. old and yp b0.05 for middle-aged vs. old).

Fig. 4. MMP-9 levels were decreased in fibroblasts isolated from the MA (n =11) and old (n =10) groups compared with young (n =14). (Top) Representative immunoblots for MMP-9 and a smooth muscle actin. (Bottom) The ratio of MMP-9 to a smooth muscle actin relative to young levels (*p b0.05 for young vs. middle-aged and old).

416

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419

samples (four female and one male). The cell counts at first passage were 39 F5 10 4 cells for young ( n =14), 29F6104 cells for MA (n =11), and 33F8104 cells for old mice (n =10; p =0.48), indicating the equal and adequate yields were obtained for all three age groups. For these mice, the time from isolation to first passage was 19F1 days for the young, 20F3 days for MA, and 21F1 days for the old mice ( p =0.72), indicating that the rate of establishing culture was similar between the three groups. All fibroblasts were positive for a smooth muscle actin, indicating that some differentiation into a myofibroblast phenotype occurred in culture. Fibroblasts from old mice displayed lower proliferative capacity than fibroblasts from young and MA mice (Fig. 3). There was no correlation between proliferation and passage number(r =0.02; p =0.88), indicating that proliferation rates did not change with passaging. The migration index, normalized to the average of young values (n =13), was 61F6% for MA (n =12; p b0.001) and 81F10% for old myocardial fibroblasts (n =9; p =0.11). Adhesion to plastic (n =812), gelatin (n =1319), fibronectin (n =1319), laminin (n =814), collagen I (n =1319), and collagen IV (n =1319) were similar for all age groups ( p =n.s.). Changes in adhesive characteristics therefore did not explain the decrease in migration in the MA group. Immunoblotting of cell pellets demonstrated equal levels of a smooth muscle actin between the three groups ( p =0.48 and 0.41 for young vs. MA and old, respectively). MMP-9 levels, normalized to a smooth muscle actin levels, were lower in fibroblasts isolated from the MA and old groups (Figs. 3 and 4). 3.6. Regression analyses To determine which parameters correlated with age, regression analyses were performed using age as the independent variable (Table 5). Insoluble MMP-3, insoluble MMP-14, and total collagen correlated positively with age. Soluble MMP-12, soluble MMP-14, insoluble collagen, and fibroblast proliferation correlated negatively with age.
Table 5 Regression analyses using age as the independent variable Dependent variable LV mass MMP-3 45 kD insoluble Soluble protein Soluble to insoluble protein ratio Soluble to insoluble collagen ratio Posterior wall thickness (diastole) End-diastolic dimension MMP-14 65 kD insoluble Total collagen MMP-12 54 kD soluble MMP-14 40 kD soluble Fibroblast Proliferation Insoluble collagen TIMP-1 soluble r 0.65 0.57 0.43 0.42 0.41 0.41 0.39 0.39 0.28 0.53 0.51 0.48 0.35 0.31 p b0.001 0.001 0.001 0.002 0.003 0.007 0.010 0.022 0.041 b0.001 b0.001 b0.001 0.011 0.046

4. Discussion 4.1. Major findings LV remodeling resulting in LV hypertrophy is a common occurrence in the aging myocardium [2]. The present study hypothesized that age-related LV hypertrophy was a result of changes in ECM remodeling. Accordingly, the goal of this study was to examine the effects of aging on LV structure, collagen, MMP, and TIMP levels, and myocardial fibroblast function in mice. The unique findings of this study were as follows. First, the age-related increase in LV mass was due to increases in both LV dimensions and wall thickness. Second, soluble protein levels increased and insoluble collagen levels decreased with age. Soluble MMPs decreased, insoluble MMPs increased, and TIMPs decreased in both fractions. Third, fibroblasts have decreased functional capacity and produce less MMP-9 with age. This study provides mechanistic insight into ECM remodeling events that occur as a function of aging. The shift in MMP localization from soluble to insoluble fractions may indicate increased recruitment of the MMP to an insoluble substrate. Increased MMP activity correlated with decreased insoluble collagen levels, suggesting that ECM degradation was stimulated in aging mice. The primary difference between MA and old mice was a further shift in MMPs from soluble to insoluble fractions, suggesting that increased MMP degradative capacity serves an important function in regulating collagen levels in old mice. The continued increase in MMP levels in old mice suggests that increased MMP activity is required to maintain increased LV size. This is the first study to correlate age-dependent changes in LV structure with changes in specific MMPs and TIMPs. 4.2. Aging effects on LV size Age-related LV hypertrophy occurred in CB6F1 mice. Body weight and LV mass increased significantly with aging, similar to previous studies in other mice strains, rats, and humans [23,33]. None of these parameters were different between the MA and old groups, indicating that growth occurred primarily during the transition from young to MA and was maintained thereafter (a type 2 change). Increases in both LV dimension and wall thickness contributed to the increase in LV mass. 4.3. Aging effects on collagen, MMP, and TIMP levels With age, decreased soluble MMP and TIMP levels accompanied the increase in soluble protein, and increased insoluble MMP and decreased insoluble TIMP levels accompanied the decrease in insoluble collagen. Together, the results suggest two interpretations. First, the decrease in insoluble collagen levels is likely due to increased MMP and decreased TIMP levels in the insoluble fraction.

Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419

417

Changes in collagen solubility reflect changes in the biochemical characteristics of the collagen. Coupled with an overall increase in total protein, collagen synthesis may be increased but matched with a greater increase in collagen degradation. Alternatively, the increased soluble to insoluble ratio may indicate a change in collagen type or degree of cross-linking [34]. Second, changes in the soluble fraction (decreased MMP-12 and MMP-14 and increased protein) correlated with the increase in wall thickness and provided a mechanism for increased LV size with aging. A major finding of this study was that changes to specific MMPs occur during aging rather than a global change in all MMP levels. In the soluble fraction, the predominant types of changes were type 1 (progressive changes with aging) and type 4 changes (directional change in MA that reversed in the old group), indicating that changes in soluble MMP levels occurred throughout the aging process. In contrast, type 3 changes (change only in the old group) characterized the insoluble fraction. A primary difference between the MA and old groups, therefore, was the increased levels of MMPs and decreased levels of TIMPs in the insoluble fraction. The changes in MMP levels between MA and old groups indicate a change in myocardial ECM composition, while LV mass is still maintained. The use of three time points allowed nonlinear changes to be assessed and the rate of change to be determined [32]. Interestingly, changes in total MMP levels from the pooled fractions mirrored changes in the soluble but not insoluble fraction. The increases in total MMP-3, MMP-9, and MMP-14 suggest increased synthesis of these particular MMP types. The differential MMP-3 expression between soluble and insoluble fractions highlights the need to use fractionated samples to obtain information on age-related distribution patterns as shifts between fractions cannot be discerned by monitoring total myocardial levels. The decrease in MMP-3, MMP-9, MMP-12, and MMP-14 in the soluble fraction was accompanied by increases of the same MMPs in the insoluble fraction, suggesting that increased binding to insoluble substrates occurred with aging. Proteins likely to be found in the soluble fraction include cytoplasmic and easily soluble ECM proteins. Proteins likely to be found in the insoluble fraction include membrane and insoluble ECM proteins. The fact that MMP-7 and TIMP-1, known soluble proteins, were only seen in the soluble fraction confirms this idea. MMP-14, a membrane-bound MMP type, is known to recycle intracellularly and may be found in both fractions [35]. The age-related change in MMP-14 localization suggests that degradation and signaling pathways through this MMP were altered. MMP-3 is a well-established activator of other MMPs [36], and the increase of MMP-3 in the insoluble fraction supports the idea of increased collagen degradation with aging. The localization of MMP-12 around the vasculature supports a role for MMP-12 in regulating changes in arterial compliance and stiffening that occur with aging [37]. MMP14 is a membrane-type MMP with putative roles in

pericellular proteolysis and signaling [35,38]. The decreases in soluble MMP-12 and MMP-14 would support the accumulation of soluble ECM fragments. Matricryptins, biologically active enzymatic ECM degradation products, have been shown to stimulate signaling pathways and affect cell growth [39]. The attenuation of MMPs in the soluble fraction potentially promotes the accumulation of ECM fragments and affects downstream signaling cascades. The fact that fibroblast proliferation and migration were altered with aging suggests a role for these ECM fragments in regulating fibroblast function. The results from this study provide the foundation for future studies that evaluate the effects of ECM fragments on LV structure and ECM remodeling, particularly during aging. 4.4. Aging effects on fibroblast function Myocardial fibroblasts exhibited decreased proliferation and migration in the MA group. Proliferation decreased further in the old age group. In the old mice, the decrease in fibroblast proliferation rates corresponded with the decrease in myocardial DDR2 levels, which together suggest a decreased number of fibroblasts in the aging myocardium. In the MA mice, the fact that DDR2 levels were decreased in vivo but fibroblast proliferation was unaltered in vitro suggests that apoptotic rates may be increased. Apoptosis was not evaluated in this study. Concordant with our results, migration and proliferation are both attenuated in dermal fibroblasts isolated from mice of increasing age [40]. The decrease in fibroblast-derived MMP-9 was consistent with the drop in myocardial MMP-9 levels in the MA group. The rebound in myocardial MMP-9 in the old group is likely due to other cell contributors. We did not observe a change in Mac-3 levels, suggesting that macrophage levels were unaltered with age. The amount of MMP-9 produced per macrophage could be higher, however. Alternatively, smooth muscle, endothelial, and mast cells also potentially contribute MMP-9 [15]. The decreased proliferative and migratory capacities in aging fibroblasts suggest that global fibroblast function may be impaired, which would provide an age-related cellular mechanism for altered ECM levels. Studies examining the response of aging myocardial fibroblasts to growth factor and cytokine stimulation would support this concept. 4.5. Clinical relevance MMPs have demonstrated roles in several components of LV remodeling, including the inflammatory, angiogenic, and hypertrophic responses. MMP inhibition has been used to successfully attenuate LV remodeling following myocardial infarction [41,42] and during the progression to heart failure [43]. The data presented here establish age-related baseline differences in LV echocardiographic parameters, ECM and MMP levels, and fibroblast functions that may influence the effects of MMP inhibition during disease pathologies such
Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

418

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419 [18] Creemers EEJM, Davis JN, Parkhurst AM, Leenders P, Dowdy KB, Hapke E, et al. Deficiency of TIMP-1 exacerbates LV remodeling after myocardial infarction in mice. Am J Physiol Heart Circ Physiol 2003; 284:H364 71. [19] Hanson BJ, Schulenberg B, Patton WF, Capaldi RA. A novel subfractionation approach for mitochondrial proteins: a three-dimensional mitochondrial proteome map. Electrophoresis 2001;22:950 9. [20] Walsh BJ, Thornton SC, Penny R, Breit SN. Microplate reader-based quantitation of collagens. Anal Biochem 1992;203:187 90. [21] Marotta M, Martino G. Sensitive spectrophotometric method for the quantitative estimation of collagen. Anal Biochem 1985;150:86 90. [22] Wilson EM, Moainie SL, Baskin JM, Lowry AS, Deschamps AM, Mukherjee R, et al. Region- and type-specific induction of matrix metalloproteinases in post-myocardial infarction remodeling. Circulation 2003;107:2857 63. [23] Burgess ML, McCrea JC, Hedrick HL. Age-associated changes in cardiac matrix and integrins. Mech Ageing Dev 2001;122:1739 56. [24] Jaffre F, Callebert J, Sarre A, Etienne N, Nebigil CG, Launay J-M, et al. Involvement of the serotonin 5-HT2B receptor in cardiac hypertrophy linked to sympathetic stimulation: control of interleukin-6, interleukin1{beta}, and tumor necrosis factor-{alpha} cytokine production by ventricular fibroblasts. Circulation 2004;110:969 74. [25] Zohar R, Zhu B, Liu P, Sodek J, McCulloch CA. Increased cell death in osteopontin-deficient cardiac fibroblasts occurs by a caspase 3-independent pathway. Am J Physiol Heart Circ Physiol 2004;287:H730 9. [26] Olaso E, Labrador JP, Wang L, Ikeda K, Eng FJ, Klein R, et al. Discoidin domain receptor 2 regulates fibroblast proliferation and migration through the extracellular matrix in association with transcriptional activation of matrix metalloproteinase-2. J Biol Chem 2002;277:3606 13. [27] Wang J, Lukse E, Seth A, McCulloch CA. Use of conditionally immortalized mouse cardiac fibroblasts to examine the effect of mechanical stretch on alpha-smooth muscle actin. Tissue Cell 2001; 33:86 96. [28] Pelletier B, Dhainaut F, Pauly A, Zahnd J. Evaluation of growth rate in adhering cell cultures using a simple colorimetric method. J Biochem Biophys Methods 1988;16:63 74. [29] Harrison NK, Dawes KE, Kwon OJ, Barnes PJ, Laurent GJ, Chung KF. Effects of neuropeptides on human lung fibroblast proliferation and chemotaxis. Am J Physiol 1995;268:L278 83. [30] Scragg MA, Ferreira LR. Evaluation of different staining procedures for the quantification of fibroblasts cultured in 96-well plates. Anal Biochem 1991;198:80 5. [31] Davis JM. Basic cell culture. Second edition. Oxford7 Oxford University Press; 2002. p. 381. [32] Coleman P, Finch C, Joseph J. The need for multiple time points in aging studies. Neurobiol Aging 2004;25:3 4. [33] Slama M, Ahn J, Varagic J, Susic D, Frohlich ED. Long-term left ventricular echocardiographic follow-up of SHR and WKY rats: effects of hypertension and age. Am J Physiol Heart Circ Physiol 2004;286:H181 5. [34] Norton GR, Tsotetsi J, Trifunovic B, Hartford C, Candy GP, Woodiwiss AJ. Myocardial stiffness is attributed to alterations in cross-linked collagen rather than total collagen or phenotypes in spontaneously hypertensive rats. Circulation 1997;96:1991 8. [35] Osenkowski P, Toth M, Fridman R. Processing, shedding, and endocytosis of membrane type 1-matrix metalloproteinase (MT1MMP). J Cell Physiol 2004;200:2 10. [36] Lijnen HR. Matrix metalloproteinases and cellular fibrinolytic activity. Biochemistry (Mosc) 2002;67:92 8. [37] Lakatta EG, Levy D. Arterial and cardiac aging: major shareholders in cardiovascular disease enterprises: Part I. aging arteries: a bset upQ for vascular disease. Circulation 2003;107:139 46. [38] Tam EM, Morrison CJ, Wu YI, Stack MS, Overall CM. Membrane protease proteomics: isotope-coded affinity tag MS identification of

as pressure overload and myocardial infarction. These results reveal unique targets for future interventions to modify age-dependent matrix remodeling.

Acknowledgements The authors acknowledge the following sources of support: HL-10337 (MLL), HL-75360 (MLL), HL-45024 (FGS), HL-97012 (FGS), P01-48788 (FGS), and a VA Career Development Award (FGS).

References
[1] Cheitlin MD. Cardiovascular physiology-changes with aging. Am J Geriatr Cardiol 2003;12:9 13. [2] Lakatta EG. Arterial and cardiac aging: major shareholders in cardiovascular disease enterprises: Part III. Cellular and molecular clues to heart and arterial aging. Circulation 2003;107:490 7. [3] Scott-Burden T. Extracellular matrix: the cellular environment. NiPS 1994;9:110 5. [4] Brew K, Dinakarpandian D, Nagase H. Tissue inhibitors of metalloproteinases: evolution, structure and function. Biochim Biophys Acta 2000;1477:267 83. [5] Galis ZS, Sukhova GK, Lark MW, Libby P. Increased expression of matrix metalloproteinases and matrix degrading activity in vulnerable regions of human atherosclerotic plaques. J Clin Invest 1994;94: 2493 503. [6] Galis ZS, Khatri JJ. Matrix metalloproteinases in vascular remodeling and atherogenesis: the good, the bad, and the ugly. Circ Res 2002;90: 251 62. [7] Carmeliet P. Proteinases in cardiovascular aneurysms and rupture: targets for therapy? J Clin Invest 2000;105:1519 20. [8] Nagatomo Y, Carabello BA, Coker ML, McDermott PJ, Nemoto S, Hamawaki M, et al. Differential effects of pressure or volume overload on myocardial MMP levels and inhibitory control. Am J Physiol Heart Circ Physiol 2000;278:H151 61. [9] Chancey AL, Brower GL, Peterson JT, Janicki JS. Effects of matrix metalloproteinase inhibition on ventricular remodeling due to volume overload. Circulation 2002;105:1983 8. [10] Cox MJ, Sood HS, Hunt MJ, Chandler D, Henegar JR, Aru GM, et al. Apoptosis in the left ventricle of chronic volume overload causes endocardial endothelial dysfunction in rats. Am J Physiol Heart Circ Physiol 2002;282:H1197 205. [11] Kim HE, Dalal SS, Young E, Legato MJ, Weisfeldt ML, DArmiento J. Disruption of the myocardial extracellular matrix leads to cardiac dysfunction. J Clin Invest 2000;106:857 66. [12] Stroud JD, Baicu CF, Barnes MA, Spinale FG, Zile MR. Viscoelastic properties of pressure overload hypertrophied myocardium: effect of serine protease treatment. Am J Physiol Heart Circ Physiol 2002;282: H2324 35. [13] Li YY, Feldman AM. Matrix metalloproteinases in the progression of heart failure: potential therapeutic implications. Drugs 2001;61:1239 52. [14] Lindsey M, Lee RT. MMP inhibition as a potential therapeutic strategy for CHF. Drug News Perspect 2000;13:350 4. [15] Spinale FG. Matrix metalloproteinases: regulation and dysregulation in the failing heart. Circ Res 2002;90:520 30. [16] Kanekar S, Hirozanne T, Terracio L, Borg TK. Cardiac fibroblasts: form and function. Cardiovasc Pathol 1998;7:127 33. [17] Roth DM, Swaney JS, Dalton ND, Gilpin EA, Ross Jr J. Impact of anesthesia on cardiac function during echocardiography in mice. Am J Physiol Heart Circ Physiol 2002;282:H2134 40.

Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014

M.L. Lindsey et al. / Cardiovascular Research 66 (2005) 410419 undescribed MT1-matrix metalloproteinase substrates. PNAS 2004; 101:6917 22. [39] Labat-Robert J. Age-dependent remodeling of connective tissue: role of fibronectin and laminin. Pathol Biol 2003;51:563 8. [40] Ashcroft GS, Mills SJ, Ashworth JJ. Ageing and wound healing. Biogerontology 2002;3:337 45. [41] Lindsey ML, Gannon J, Aikawa M, Schoen FJ, Rabkin E, LoprestiMorrow L, et al. Selective matrix metalloproteinase inhibition reduces left ventricular remodeling but does not inhibit angiogenesis after myocardial infarction. Circulation 2002;105:753 8.

419

[42] Yarbrough WM, Mukherjee R, Escobar GP, Mingoia JT, Sample JA, Hendrick JW, et al. Selective targeting and timing of matrix metalloproteinase inhibition in post-myocardial infarction remodeling. Circulation 2003;108:1753 9. [43] Spinale FG, Coker ML, Kromback SR, Mukherjee R, Hallak H, Houck WV, et al. Matrix metalloproteinase inhibition during the development of congestive heart failure. Circ Res 1999;85:364 76.

Downloaded from http://cardiovascres.oxfordjournals.org/ by guest on April 4, 2014