Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y V O L . 6 7 , N O . 9 , 2 0 1 6

ª 2 0 1 6 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N I S S N 0 7 3 5 - 1 0 9 7 / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j a c c . 2 0 1 5 . 1 2 . 0 3 5

Evidence for Mechanisms Underlyingthe Functional Benefits of a MyocardialMatrix Hydrogel for Post-MI Treatment

Jean W. Wassenaar, BS,a,b Roberto Gaetani, PHD,a,b Julian J. Garcia,a,b Rebecca L. Braden, MS,a,b Colin G. Luo, MS,c

Diane Huang, BS,c Anthony N. DeMaria, MD,c Jeffrey H. Omens, PHD,a,c Karen L. Christman, PHDa,b

ABSTRACT

Fro

an

Na

do

Un

bo

All

Ma

BACKGROUND There is increasing need for better therapies to prevent the development of heart failure after

myocardial infarction (MI). An injectable hydrogel derived from decellularized porcine ventricular myocardium has been

shown to halt the post-infarction progression of negative left ventricular remodeling and decline in cardiac function in

both small and large animal models.

OBJECTIVES This study sought to elucidate the tissue-level mechanisms underlying the therapeutic benefits of

myocardial matrix injection.

METHODS Myocardial matrix or saline was injected into infarcted myocardium 1 week after ischemia–reperfusion in

Sprague–Dawley rats. Cardiac function was evaluated by magnetic resonance imaging and hemodynamic measurements

at 5 weeks after injection. Whole transcriptome microarrays were performed on RNA isolated from the infarct at 3 days

and 1 week after injection. Quantitative polymerase chain reaction and histologic quantification confirmed expression of

key genes and their activation in altered pathways.

RESULTS Principal component analysis of the transcriptomes showed that samples collected from myocardial matrix-

injected infarcts are distinct and cluster separately from saline-injected control subjects. Pathway analysis indicated that

these differences are due to changes in several tissue processes that may contribute to improved cardiac healing after MI.

Matrix-injected infarcted myocardium exhibits an altered inflammatory response, reduced cardiomyocyte apoptosis,

enhanced infarct neovascularization, diminished cardiac hypertrophy and fibrosis, altered metabolic enzyme expression,

increased cardiac transcription factor expression, and progenitor cell recruitment, along with improvements in global

cardiac function and hemodynamics.

CONCLUSIONS These results indicate that the myocardial matrix alters several key pathways after MI creating a

pro-regenerative environment, further demonstrating its promise as a potential post-MI therapy. (J Am Coll Cardiol

2016;67:1074–86) © 2016 by the American College of Cardiology Foundation.

P rogression from acute myocardial infarction(MI) to chronic heart failure (HF) begins withan initial ischemic injury, resulting in progres-

sive myocyte loss through both necrotic andapoptotic mechanisms, and migration of inflamma-tory cells into the injured myocardium. An increase

m the aDepartment of Bioengineering, University of California, San Dieg

d the cDepartment of Medicine, University of California, San Diego, La Joll

tional Institutes of Health National Heart, Lung, and Blood Institute (1R

ctoral fellowships from the California Institute for Regenerative Medicine

iversity of California, San Diego Medical Scientist Training Program T32

ard member of Ventrix, Inc. Dr. Christman is a co-founder, board member,

other authors have reported that they have no relationships relevant to

nuscript received November 11, 2015; accepted December 14, 2015.

in matrix metalloproteinases from the inflammatoryinfiltrate further exacerbates the decline in heartfunction by digesting the extracellular matrix (ECM),followed by subsequent deposition of fibrillar cross-linked collagen. The heart has recently beenrecognized as an organ capable of some degree of

o; bSanford Consortium for Regenerative Medicine;

a, California. This research was funded in part by the

01HL113468). Ms. Wassenaar was supported by pre-

and the American Heart Association, as well as the

GM007198-40. Dr. DeMaria is a scientific advisory

consultant, and holds equity interest in Ventrix, Inc.

the contents of this paper disclose.

AB BR E V I A T I O N S

AND ACRONYM S

CPC = cardiac progenitor cell

ECM = extracellular matrix

HF = heart failure

IHC = immunohistochemistry

LV = left ventricle

J A C C V O L . 6 7 , N O . 9 , 2 0 1 6 Wassenaar et al.M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6 Effects of Myocardial Matrix Hydrogel After MI

1075

self-regeneration (1). However, this is insufficient tocompensate for the billions of cardiomyocytes lost af-ter MI (2). Additionally, function of surviving cardio-myocytes is also altered after infarction. The hearthas a high energy demand and recent studies haveshown that dysregulation in cardiac metabolismafter MI contributes notably to cardiac dysfunctionin HF (3).

SEE PAGE 1087MI = myocardial infarction

PPAR = peroxisome

proliferator-activated receptor

qPCR = quantitative real-time

erase chain reaction

Interest in developing alternative treatments for MIhas been expanding. Such therapies include variouscells, biological molecules, acellular biomaterials, orcombinations thereof. Meta-analyses of initial celltherapy trials suggest only a modest effect on cardiacfunction (4), and given low cell survival rates and theirlargely paracrine mechanism of action, there has beenincreasing interest in the use of injectable acellularscaffolds (5). If designed appropriately, these bio-materials can be delivered throughminimally invasiveapproaches and stimulate cardiac repair, while avoid-ing many of the complications associated with a livingproduct (6). Our group previously developed aninjectable myocardial matrix hydrogel, derived fromdecellularized porcine ventricular ECM (7), which canbe delivered with a transendocardial catheter. Thishydrogel was shown to reduce negative left ventricular(LV) remodeling and the decline in cardiac function inboth rat (8) and pig (9) models when delivered 2 weeksafter MI. Herein, we examined whether the materialcould improve global cardiac function and hemody-namics when delivered 1 week after MI in a rat model,and utilized a transcriptomics-directed approach toidentify the underlying mechanisms by which thematrix improves post-MI repair.

METHODS

All procedures in this study were approved by theCommittee on Animal Research at the University ofCalifornia, San Diego, and the Association for theAssessment and Accreditation of Laboratory AnimalCare. Myocardial matrix or saline was injected into thearea of ischemia 1 week after 25 min of ischemia–reperfusion in female Sprague-Dawley rats. Rat Gene2.0 ST arrays (Affymetrix, Inc., Santa Clara, California)were used for whole transcriptome analysis of infarctand border zone at 3 days and 1 week after injection(see Online Figure 1 for a representative H&E slidefrom infarcted hearts that was included for analysis),followed by validation of the expression of key genesby quantitative real-time polymerase chain reaction(qPCR, primer sequences in Online Table 1). Cardiacmagnetic resonance imaging and hemodynamics

recordings were performed at 5 weeks afterinjection (6 weeks after MI). Histology andimmunohistochemistry (IHC) were used toquantify phenotypic changes. For further de-tails, refer to the Online Appendix.

RESULTS

CARDIAC FUNCTION. Myocardial matrixinjection (n ¼ 8) significantly reduced thepercent change in ejection fraction (p ¼ 0.028)and end-systolic volume (p ¼ 0.004) com-pared with saline (n ¼ 7) (Figure 1A) from

6 days after MI (1 day before injection) to 6 weeks afterMI (5 weeks after injection). There was a similar butnonsignificant trend for end-diastolic volume (p ¼ 0.11). Cardiac magnetic resonance data are provided inOnline Table 2. LV hemodynamics were measuredusing a microtipped manometer pressure catheter at 6weeks after MI (Figure 1B). Compared with saline (n ¼5), myocardial matrix-injected hearts (n ¼ 5) hadsignificantly higher LV peak systolic pressure (p ¼ 0.002), myocardial relaxation (–dP/dtmax; p ¼ 0.003),and myocardial contractility (þdP/dtmax; p ¼ 0.002).

TRANSCRIPTOMICS. Differences in transcriptomesbetween saline- and matrix-treated infarcts wereglobally examined using both principal componentanalysis and hierarchical clustering. Saline- andmyocardial matrix-injected samples did not clusterseparately at 3 days after injection. However, by 1weekafter injection, both principal component analysis(Figure 2A) and hierarchical clustering (Figures 2Band 2C) showed separation of the transcriptomes(p ¼ 0.06), indicating a shift in global gene expression.A false discovery rate of q < 0.05 was used to deter-mine the differentially expressed genes for biologicalinterpretation (Online Table 3). A Panther over-representation test showed that differences in tran-scripts from ECM (Gene Ontology [GO]: 0031012) andthose responsible for muscle contraction (GO:006936)for the 1-week time point were nonsignificant (p¼ 0.37and p¼ 0.88, respectively), indicating that differencesin gene expression were not due to sampling vari-ability. Using Ingenuity Pathway Analysis (IPA)(Qiagen, Redwood City, California), we identified themain effects of the myocardial matrix at 3 daysafter injection as downregulation of apoptosis, upre-gulation of blood vessel development, and increasein cell movement (Online Table 4). By 1 week, severalpathways were significantly activated, includingdownregulation of cell death and hypertrophy, andupregulation of many metabolic processes and genetranslation/transcription (Online Table 5). Further de-tails, as well as the complete list of differentially

polym

FIGURE 1 Cardiac Magnetic Resonance and Hemodynamics Analysis

0

-2

-4

-6

-8EF A

bsol

ute

Chan

ge (%

)

*

**

**** **

60

40

20

0ESV

Abso

lute

Cha

nge

(μμL)

EDV

Abso

lute

Cha

nge

(μL) 100

80

60

40

20

0

SalineMatrix

6

4

2

0

LVED

P (m

m H

g)

150

100

50

0

LVPS

P (m

m H

g)

5000

4000

3000

2000

1000

0

-dP/

dtm

ax (m

m H

g/s)

5000

4000

3000

2000

1000

0

+dP/

dtm

ax (m

m H

g/s)

A

B

(A) Cardiac magnetic resonance (CMR) was performed to compare percent changes in ejection fraction (EF), end-systolic volume (ESV), and end-diastolic volume (EDV)

from 6 days after myocardial infarction (MI) (1 day before injection) to 6 weeks after MI (5 weeks after injection) at study termination. (B) Left ventricular end-diastolic

pressure (LVEDP), left ventricular peak systolic pressure (LVPSP), myocardial relaxation (-dP/dtmax), and myocardial contractility (þdP/dtmax) were assessed by cath-

eterization before humane killing. *p < 0.05. **p < 0.01.

Wassenaar et al. J A C C V O L . 6 7 , N O . 9 , 2 0 1 6

Effects of Myocardial Matrix Hydrogel After MI M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6

1076

expressed genes, can be found in Online Tables 6through 13. Based on the differences elucidated bythis transcriptome analysis,we performedqPCRonkeygenes in the identified pathways (Online Figure 2) andfurther assessed differences at the tissue level usingIHC as described later.

FIGURE 2 Transcriptomes Cluster Separately at 1 Week After Inject

Component 1 (32.27%)

Com

pone

nt 2

(11.2

3%)

A B

Principal component analysis (A) and hierarchical clustering (B) of infarc

after myocardial matrix injection is distinct from control saline injection b

expressed at 1 week after injection. RNA from 2 infarcts were combined

(n ¼ 3 arrays per group, per time point). Blue ¼ saline; Red ¼ matrix; t

INFLAMMATORY RESPONSE. Pathways involved inthe immune response including migration and infil-tration of various cell types were predicted at both the3-day and 1-week time points (Online Tables 4 and 5,Figure 3A, Online Figure 3). Similarly, a substantialnumber of genes at both time points —26.6% at day

ion

C

t transcriptomes of all samples indicated that global gene expression

y 1 week. (C) Hierarchical clustering of the 2,144 genes differentially

for analysis on 1 microarray chip to reduce biological variability

riangles ¼ 3 days; squares ¼ 1 week.

FIGURE 3 Inflammatory Response

3days–Saline 3days–Matrix

1week–Saline 1week–Matrix

CXCL10CCL3L3IL1RNCD68ITGB2SPP1CCR2MMP12G6PDLBPTIMP1HMOX1LOXWISP2

C6ANGPT2IGFBP4AGTR1CFHCTGFANXA1UNC50NFKBIAS1PR1TEKCAV1CASP3GNAO1CXCL10MAPK14Ccr2AKT2P2RY1MAPK8IARAFTHRBVEGFAGJA1PDK1PRKCEABCC8ABCC8CCL2CCL11CD36 Saline Matrix

Saline Matrix

Saline Matrix

Three Day

One Week*

#

800

600

400

200

0

Tryp

tase

+ Ce

lls

80

60

40

20

0

Tryp

tase

+ Ce

lls

2.0x1006

1.5x1006

1.0x1006

5.0x1005

0.0

CD68

+ (P

ixel

s)

A B C

D

E

(A) Expression of genes (red ¼ upregulated; green ¼ downregulated) involved in inflammation at 3 days and 1 week. (B) CD68þ staining for macrophages,

visualized by 3,3’ diaminobenzidine (brown) in a myocardial matrix-injected infarct 3 days after injection. (C) Quantification of CD68 staining in the

infarct wall from 3 slides at 3 days after injection. (D) Tryptaseþ (red) cells in myocardial matrix injected heart at 1 week after injection; nuclei are stained

blue with Hoescht 33342. (E) Quantification of all tryptaseþ mast cells in the infarct wall from 3 slides at 3 days and 1 week after injection.

Scale bar ¼ 50 mm. #p ¼ 0.052. *p < 0.05.

J A C C V O L . 6 7 , N O . 9 , 2 0 1 6 Wassenaar et al.M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6 Effects of Myocardial Matrix Hydrogel After MI

1077

3 and 9.8% at 1 week—were characterized aspart of immune response process (GO:0002376).Increased expression of CD68 (p ¼ 0.045) and matrixmetalloproteinase 12 (p ¼ 0.043) (Online Figure 2A)were confirmed by qPCR, indicative of increasedmacrophage infiltration as a result of matrix injection,yet IHC analysis using a CD68 antibody (Figure 3B) didnot show a difference in macrophage infiltrationbetween saline- and matrix-injected infarcts at 3 days(Figure 3C). There was, however, a trend toward anincrease in infiltrating tryptaseþmast cells (Figure 3D)at 3 days (p ¼ 0.052), which reached significance by1 week (p ¼ 0.032) (Figure 3E).

BLOOD VESSEL FORMATION. Activation of bloodvessel development was predicted by IPA at 3 daysafter injection (Online Table 4, Figure 4A). Although

increased vessel development was not directly pre-dicted at 1 week after injection, many growth factorsassociated with angiogenesis and neovascularizationwere identified (Figure 4A). These included a decreasein angiopoietin-2 (p ¼ 0.012) and increases in acidicfibroblast growth factor (p ¼ 0.053) and vascularendothelial growth factors A and B (p ¼ 0.034 andp ¼ 0.009, respectively), confirmed by qPCR (OnlineFigure 2B). Infarct vascularization was then exam-ined by IHC (Figure 4B). Although not different at3 days after injection, capillary density was signifi-cantly greater in the matrix-injected group at 1 week(p ¼ 0.038) (Figure 4C). Arteriole density showed asimilar result, with no difference at 3 days, but therewas a trend toward an increase at 1 week in the matrixgroup (p ¼ 0.056) (Figure 4D). When subdivided intolarge (Online Figure 4A), medium (Online Figure 4B),

FIGURE 4 Blood Vessel Formation

3days–Saline 3days–Matrix

1week-Saline 1week-Matrix

CXCL10ITGB2SPP1G6PDVCAM1PGFHMOX1

CTGFTGFBR1IGF2PDGFDFGF1FGF9TBRG4VEGFBFGF16FGF13VEGFA

15

10

5

0Saline Matrix

% v

WF+

(Pix

els)

Saline Matrix

15

10

5

0

Three Day One Week

*

#

Saline Matrix Saline Matrix

Three Day One Week100

80

60

40

20

0

Tota

l Art

erio

le D

ensit

y (p

er m

m2 )

100

80

60

40

20

0

A

B

C

D

(A) Expression of vessel development genes at 3 days and growth factors at 1 week. (B) Representative images of vessel staining with endothelial cells labeled by

von Willebrand factor (vWF) (green) and smooth muscle cells labeled with a-smooth muscle actin (red). Endothelial cells (C) and arteriole density (D) are quantified

within the infarct. Scale bar ¼ 50 mm. #p < 0.1. *p < 0.05.

Wassenaar et al. J A C C V O L . 6 7 , N O . 9 , 2 0 1 6

Effects of Myocardial Matrix Hydrogel After MI M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6

1078

and small (Online Figure 4C) arterioles, there was atrend at 3 days for increased small arterioles withinmatrix-injected infarcts (p ¼ 0.082), which shifted at1 week to a significant increase in medium (p ¼ 0.021)and a trend for an increase in large diameter vessels(p ¼ 0.065).

CARDIOMYOCYTE APOPTOSIS. Decreased apoptosiswas predicted consistently by IPA at both 3 days and1 week after injection (Online Tables 4 and 5,Figure 5A, Online Figures 5 and 6). Increased expres-sion of antioxidative enzymes heme oxygenase 1 (p ¼0.015) at 3 days and catalase (p ¼ 0.002) at 1 week, aswell as the antiapoptosis regulator Bcl-2 (p ¼ 0.006)at 1 week, were confirmed using qPCR (OnlineFigure 2C). To determine the effect that myocardialmatrix injection may have on apoptosis of car-diomyocytes specifically, anti–cleaved-caspase 3staining with colabeling of cardiomyocytes usinga-actinin was performed (Figure 5B). Quantificationof the number of caspase-3–expressing car-diomyocytes within the infarct wall showed a trendtoward decreased apoptotic cardiomyocytes within

the infarct wall (p ¼ 0.085) at 3 days after injection(Figure 5C).

CARDIAC METABOLISM. At 1 week after injection,several genes associated with oxidative metabolismand mitochondrial biogenesis were upregulated andpredicted to be activated by IPA (Online Table 5,Figure 6A, Online Figures 7 and 8). Similarly, GOanalysis classified 53% of the differentially expressedtranscripts to be involved in metabolic processes(GO:008152). Within these genes, we identifiedseveral upregulated nuclear receptors involved incardiac metabolism and mitochondrial biogenesis,including increased expression of the transcriptionco-activator peroxisome proliferator-activated re-ceptor (PPAR) gamma coactivator 1-alpha (PGC-1a)(p ¼ 0.006) and several of its target receptors,including estrogen-related receptor-g (p ¼ 0.001) andPPARa (p ¼ 0.032) and b/d (p ¼ 0.018), which wereconfirmed by qPCR (Online Figure 2D). To determinewhether changes in metabolism were associatedspecifically with cardiomyocytes, IHC was per-formed to identify expression of PGC-1a within

FIGURE 5 Apoptosis

3days–Saline 3days–Matrix

1week–Saline 1week–Matrix

1week–Saline 1week–Matrix

ALOX15PLAC8G6PDCD53Gsta1SPP1IL1RNAGTR1TIMP1HMOX1MAP2K6REG3A

RPS3Atg5DDIT4HDAC1S100A10PtmaPPIAADAM17TNFRSF11BPTGES3C6OLR1SPP1ANGPT2PAWRPDE5ASERPINE2IGFBP3ANXA1NFKBIAAGTR1CFHPTNCTGFGFRA1DPP4HLA–ATNFSF10IGF2JUNBCASP3NTRK3BIRC2TOP2AITGA1KLF6SNRKCAV1DDIT3REG3A

MAPK14TSC1MARK2PDE4AKLK1GNAO1BCL2Cyb5r3SLC25A10GSK3ACD59CATPPARDPARK7MAP2K3ELK1KCNK3INPPL1PTGDSPPP2R1AGHRAKT2INSRMAP2K4NAMPTEHD4UCP2PRKAA2PRKACAPPIFMAPK8IP1GATA4PGC1aCDNFAKAP1MFN2GJA1MAPTPHLDA1VEGFAPARK2MEF2DNCS1ERBB4MMP7

Saline Matrix

200

150

100

50

0Casp

3+ C

ardi

omyo

cyte

s

Saline Matrix

20

15

10

5

0

Three Days One Week

#

A

B

C

(A) Expression of genes involved in apoptosis at 3 days and 1 week. (B) Examples of positive cleaved-caspase 3-expression (red) in a-actininþ cardiomyocytes

(green); nuclei are stained blue with Hoescht 33342 (merged image [top] and red-channel only [bottom]). (C) Quantification of all cleaved-caspase 3

expressing cardiomyocytes within the border zone of 3 slides. Scale bar ¼ 50 mm. #p ¼ 0.085.

J A C C V O L . 6 7 , N O . 9 , 2 0 1 6 Wassenaar et al.M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6 Effects of Myocardial Matrix Hydrogel After MI

1079

cardiomyocytes (Figure 6B). Quantification of PGC-1aþcardiomyocytes adjacent to the infarct scar revealedthat myocardial matrix-injected hearts exhibited ahigher percentage of PGC-1a expression comparedwith saline-injected hearts (p ¼ 0.009) (Figure 6C).

CARDIAC DEVELOPMENT AND PROGENITORS. Re-view of significant GO terms showed several associ-ated with muscle development, including heartdevelopment (GO:0007507) (Figure 7A), andqPCR-confirmed consistent elevated expression ofGATA4 (p ¼ 0.022), Nkx2.5 (p ¼ 0.009), MEF2d

(p ¼ 0.004), myocardin (p ¼ 0.004), Tbx5 (p ¼ 0.012),and Tbx20 (p ¼ 0.043) (Online Figure 2E). We alsoassessed infiltration of cells expressing cKit, acommonly used marker for cardiac progenitor cells(CPCs). Because cKit is also expressed by mast cells,tissue sections were costained with antimast celltryptase (Figure 7B). Quantification of cKitþ/tryptase-cells in the infarct wall showed a trend at 3 days(p ¼ 0.067), which attained significance at 1 week(p ¼ 0.031) in the matrix-injected group (Figure 7C).Costaining showed some of these cells expressed thecardiac transcription factor Nkx2.5 (Online Figure 9).

FIGURE 6 Myocardial Metabolic Gene Expression

1week–Saline 1week–MatrixAQP1ACSL3ACSL5DPP4NUCB2CD36DNM1LERRgAKT2ACOX1GHRPGC1aGPAMACSL6AQP7ECI2ACAA2ACSL1PRKAA2MLYCDHADHBRAB4AACADVLPHYHEEF1A2CPT2PPARaACACBHADHHADHASLC2A4INSCATINPPL1INSRPGC1bSLC27A1PPARdERBB4KAT2BRPS6KB1UCP3STAT5A

Saline Matrix

80

70

60

50

40

30PGC1α

+ Ca

rdio

myo

cyte

s (%

)

**

A B

C

(A) Expression of metabolic genes at 1 week. (B) Example of peroxisome proliferator-activated receptor gamma coactivator 1-alpha (PGC-1a)

expression (red) in a-actininþ cardiomyocytes (green) adjacent to the infarct; nuclei are stained blue with Hoescht 33342 (merged image [top]

and red-channel only [bottom]). Thick arrows point to positive PGC-1a stained cardiomyocyte nuclei and thin arrows point to negative

nuclei. (C) Quantification of PGC-1 a þ nuclei expression percentage; >200 cardiomyocytes were quantified per heart. Scale bar ¼ 50 mm.

**p < 0.01.

Wassenaar et al. J A C C V O L . 6 7 , N O . 9 , 2 0 1 6

Effects of Myocardial Matrix Hydrogel After MI M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6

1080

HYPERTROPHY AND FIBROSIS. At 1 week after in-jection, cardiac hypertrophy was highly representedin the pathways predicted to be downregulated byIPA (Online Table 5, Figure 8A, Online Figure 10).Increased expression of negative regulators of hy-pertrophy DUSP5 (p ¼ 0.003) and DYRK1a (p ¼ 0.004)as well as a decrease in the positive regulator NUPR1(p ¼ 0.048) were all confirmed by qPCR (OnlineFigure 2F). The effects on hypertrophic LV remodel-ing were therefore assessed at 5 weeks after injection

by measuring cardiomyocyte diameter to evaluatecardiomyocyte hypertrophy (Figure 8B) and stainingwith Mason’s trichrome to evaluate interstitialfibrosis in the remote myocardium. Quantificationof cross-sectional areas showed a trend in thematrix-injected hearts (n ¼ 7) toward less hypertro-phic cardiomyocytes (p ¼ 0.061) (Figure 8C) comparedwith saline (n ¼ 8). Moreover, myocardial matrix in-jection was associated with significantly reducedinterstitial fibrosis (p < 0.001) (Figure 8D).

FIGURE 7 Cardiac Development

1week–Saline 1week–MatrixPcolceCpSema3dEltd1Nrp2Cdh5Sema5aTbx20Tbx5Sema4dEmr1Adam19CntnapSgcdSgcgSctrCdh2Gata4Popdc2MyocdGata5Nkx2.5Ldb3Mef2dCdh20

Saline Matrix

80

60

40

20

0cK

it+/T

rypt

ase-

Cel

ls

Three Day

#

Saline Matrix

80

60

40

20

0

cKit+

/Try

ptas

e- C

ells

One Week*

A

B

C

(A) Expression of genes involved in heart development. (B) Example of cKitþ(green)/tryptase- (red) cells with Hoescht-labeled nuclei (blue)

in the myocardial matrix-injected infarct after 1 week. (C) Quantification of cKitþ/tryptase- cells throughout border zone of 3 slides at 3 days

and 1 week after injection. Scale bar ¼ 50 mm. #p ¼ 0.067. *p < 0.05.

J A C C V O L . 6 7 , N O . 9 , 2 0 1 6 Wassenaar et al.M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6 Effects of Myocardial Matrix Hydrogel After MI

1081

DISCUSSION

Previous small and large animal studies havedemonstrated that injection of acellular myocardialmatrix improved cardiac function and reducednegative LV remodeling when delivered 2 weeks afterMI. Importantly, biocompatibility, hemocompati-bility, and lack of arrhythmias were demonstrated(8,9). In the current study, we analyzed infarct geneexpression using whole-transcriptome microarrays togain a comprehensive understanding of the tissue-level mechanism of action of the myocardial matrixhydrogel. Gene expression was analyzed at both3 days and 1 week after injection based on previousstudies indicating that cellular infiltration into theinjected hydrogel was most pronounced during thefirst week (9). By 1 week after injection, the tran-scriptomes of saline and matrix-injected infarctsclustered separately. Similar analyses have also beenapplied to other experimental therapies for MI

including cell transplantation (10–12) and injection ofcell-derived products (13); however, to our knowl-edge, no other biologic-based therapy has been re-ported to induce a distinct transcription signature ata global level. In this study, the key modulatedpathways included inflammation, reduction ofapoptosis and cardiac hypertrophy, metabolism, andblood vessel and cardiac development (CentralIllustration).

Gene expression differences within the infarctsuggested an increase in macrophage migration inresponse to the matrix injection with increased tran-scription of CD68, a macrophage marker, and matrixmetalloproteinase 12, a macrophage-specific protease(14). However, differences in transcription patternscould also be attributed to changes in immune cellbehavior, because an increase in macrophages wasnot demonstrated. Collective analysis of all differen-tially expressed transcripts related to inflammationwas not conclusive regarding whether there was a

FIGURE 8 Hypertrophic Remodeling

1week–Saline 1week–MatrixAGTR1ADAM17CTGFNUPR1DUSP5DYRK1AMAPK14MMP7KCND2GATA4AKAP1FBXO32RRADVEGFB Saline Matrix

Saline Matrix

1000

800

600

400

200

0

4

3

2

1

0

Inte

rstit

ial F

ibro

sis (%

)

**

#

Card

iom

yocy

te A

rea

(μm

2 )

A

D

C

E

B

(A) Expression of genes involved in hypertrophic response. (B) Representative image of the remote myocardium stained with laminin antibody (green) to outline

cardiomyocytes and Hoescht 33342 to visualize nuclei (blue). (C) Quantification of cardiomyocyte cross-sectional area, averaged $300 cells per heart. (D) Represen-

tative Masson’s trichrome staining of interstitial fibrosis (left, saline; right, matrix). (E) Quantification of interstitial fibrosis from 5 slides per heart. Scale bar ¼ 50 mm.

#p ¼ 0.061. **p < 0.01.

Wassenaar et al. J A C C V O L . 6 7 , N O . 9 , 2 0 1 6

Effects of Myocardial Matrix Hydrogel After MI M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6

1082

predominance of either M1 or M2 macrophage acti-vation, which has been attributed to proinflam-matory and pro-remodeling responses, respectively(15). Both phenotypes are likely necessary for post-infarct repair as depletion of either M1 or M2 macro-phages inhibited the ability of neonatal hearts toregenerate after MI (16). In the process of identifyingc-Kitþ progenitors, which stain negative for tryptase,we detected a notable increase in the number oftryptaseþ mast cells in the matrix-injected groups.Although mast cells are traditionally associated withan allergic response, they are also involved in neo-vascularization and regulation of the immuneresponse (17). After ischemia, rapid mast celldegranulation occurs, triggering recruitment of otherleukocytes and preventing cardiomyocyte apoptosis(18). Additionally, mast cell products are known to beinherently angiogenic and stimulate endothelialsecretion of angiogenic chemokines (19). Although itis known that mast cell activation may be thefirst step in the acute inflammatory response toimplanted biomaterials (20), implication on the

reparative response of decellularized materials hasnot been reported.

By 1 week after injection, myocardial matrix hadalso significantly increased infarct neovasculature.Interestingly, the changes over time seemed to bedue to a decrease in endothelial cells and arteriolesin the saline-injected control subjects. Immediatelyafter MI, hypoxia-inducible factor expression triggerstranscriptional activation of many angiogenic factors(21); however, a decrease in vascular density withtime has been previously reported (22), possibly dueto vessel regression. Notably, 1 factor known toinduce this process is angiopoietin 2, which wasmore highly expressed in the saline group comparedwith the matrix. Preservation of the infarct vascula-ture may be a result of the pro-angiogenic milieuinduced by the myocardial matrix, whether indi-rectly, through its effects on other cell types such asimmune cells, or directly, by creating a new physicalscaffold for vessel infiltration or releasing bioactivematricryptic peptides from partial proteolysis of theECM (23).

CENTRAL ILLUSTRATION Effects of Myocardial Matrix Hydrogel After Myocardial Infarction: Mechanisms Underlying theFunctional Benefits

Hydrogelinjected into infarct

Myocardial Matrix Hydrogel

Injection of Myocardial Matrix Hydrogel: •Modulates inflammatory response •Reduces cardiomyocyte apoptosis •Enhances blood vessel development •Alters myocardial metabolism •Induces cardiac transcription factor expression •Recruits progenitor cells •Diminishes cardiac hypertrophy and fibrosis

Wassenaar, J.W. et al. J Am Coll Cardiol. 2016; 67(9):1074–86.

Injection of myocardial matrix 1 week after myocardial infarction (MI) into the infarcted area induced various tissue level changes that reduced negative left ventricular

remodeling and improved hemodynamics. Altering these key pathways created a pro-regenerative environment, potentially preventing or slowing development of

heart failure.

J A C C V O L . 6 7 , N O . 9 , 2 0 1 6 Wassenaar et al.M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6 Effects of Myocardial Matrix Hydrogel After MI

1083

Pathway analysis from the microarray data pre-dicted consistent downregulation of apoptosis andcell death. Specifically, matrix-injected infarctsexpressed higher levels of heme oxygenase 1 at 3 daysand catalase at 1 week, both of which are stress-induced enzymes that reduce reactive oxygen spe-cies. In the histologic analysis, we demonstrated atrend toward a reduction in cardiomyocyte apoptosisin the infarct wall. After an infarct, cardiomyocytedeath peaks 24 h after the injury then decreases, butcontinues to be increased above baseline levels for atleast 12 weeks (24). Therefore, delivery of themyocardial matrix could play an important role insalvaging cardiomyocytes that are pre-apoptotic.Pathway analysis also predicted upregulation ofgenes involved in mitochondrial metabolism. It iswell-demonstrated that myocardial ischemia reduces

substrate oxidation, resulting in an increased relianceon glycolysis (25). After matrix injection, we foundincreased expression of PGC-1a, PPARa, PPARb/d, andestrogen-related receptor-g. The PGC-1a activates thetranscription of these nuclear receptors to increasefatty acid uptake, oxidative phosphorylation, andmitochondrial biogenesis (26). Expression of PGC-1atargets is known to be downregulated in both rodentmodels and patients with HF (27). We showed an in-crease in PGC-1a in cardiomyocytes in matrix-injectedanimals. In comparison, injections of various bonemarrow cells into the infarct myocardium have beenassociated with decreased expression of genes relatedto mitochondrial function (10,12). Angiotensin-converting enzyme inhibition using captopril, acommon treatment for HF, similarly did not rescuechanges in fatty acid metabolism (28).

Wassenaar et al. J A C C V O L . 6 7 , N O . 9 , 2 0 1 6

Effects of Myocardial Matrix Hydrogel After MI M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6

1084

The adult mammalian heart has a limited ability toregenerate, and many recent efforts have attemptedto enhance this ability after MI (2). GO analysis sug-gested activation of terms related to heart and car-diovascular system development, which led us toinvestigate whether there was an upregulation ofcardiac transcription factors known to play a role incardiac regeneration. We found increased expressionof 6 cardiac transcription factors. Of these, GATA4,myocardin, Nkx2.5, Tbx5, and Tbx20 are all expressedthroughout various stages of embryonic cardio-genesis (29), whereas GATA4, Nkx2.5, and Tbx5 arefrequently used to identify various CPC subsets (30).Increased expression of MEF and GATA4 were re-ported after injection of bone marrow cells andmononuclear cell secretomes (10,13); however, to ourknowledge, concomitant increased transcription ofseveral cardiac transcription factors has not beenreported previously. A lineage tracing study will,however, be necessary to determine whether themyocardial matrix can induce cardiac regeneration,because increased expressions of these factors canoccur during other processes. For example, GATA4,myocardin, Nkx2.5, and Tbx20 are expressed by adultcardiomyocytes and are required for their survivaland function (31–34); cardiac hypertrophy is associ-ated with elevated GATA4, MEF2d, and Nkx2.5(35,36); and cardiac fibroblasts can express high levelsof GATA4 and Tbx20 (37). We, however, show an in-crease of 6 transcription factors for cardiac develop-ment, along with a decrease in cardiac hypertrophyand fibrosis. We also found a significant increase inckitþ/tryptase-CPCs in matrix-injected hearts. Previ-ously, the myocardial matrix was shown to promotecardiac differentiation of cKitþ CPCs in vitro (38). Inthis study, we found some cKitþ cells coexpressingNkx2.5, which has been used to identify cardiaclineage differentiation in vivo (39). The importanceof c-kitþ cells in post-infarct regeneration, however,has been a controversial subject and 2 recent studiesarrived at opposite conclusions (40,41). Additionally,cKitþ cells may contribute to other effects ofthe myocardial matrix injection, such as neo-vascularization (41).

Previous studies with myocardial matrix testeddelivery at 2 weeks after MI (8,9). In the currentstudy, we injected the material 1 week after MI,demonstrating that the matrix can also be effectiveat an earlier time point. Hemodynamics, which hadnot been studied previously, further demonstratedimprovements in LV peak systolic pressure, myo-cardial contractility, and myocardial relaxation.As further evidence that the matrix attenuates nega-tive LV remodeling, decreased hypertrophy was

suggested by transcriptional analysis as early as 1week after injection, and there was a trend toward areduction in cardiomyocyte area and a significantdecrease in interstitial fibrosis by 5 weeks.

STUDY LIMITATIONS. Limitations of this studyinclude the use of a small animal model, relativelyshort time points of assessment after injection, andthe more mild ischemia–reperfusion model, which islikely more representative of the acute or subacute MIpopulation rather than severe remodeling in HF pa-tients. Although changes in cardiac function weresignificant in this and a previous rat MI study, withsimilar injections and time points (8), greater in-creases in function were observed in a large animalmodel where the infarct was more severe, multipleinjections were performed across the infarct, andcardiac function was examined out to 3 months afterinjection (9). Unlike cells or growth factors, whichhave had diminished efficacy in moving from small tolarge animals, biomaterials may have the capacity forgreater improvements in larger animals (42), which ispromising for future clinical translation. This studyalso tested a porcine-derived ECM hydrogel in a ratmodel, which has potential for xenograft-elicitedinflammation. However, we observed similar im-provements in a porcine MI model with theporcine-derived material (9), suggesting that theinterspecies effects are not a major factor. Also, thisstudy mimicked the xenogeneic porcine materialsource being used in an ongoing clinical trial (NCT02305602).

CONCLUSIONS

We demonstrated decreased negative LV remodelingand improved hemodynamics following deliveryof myocardial matrix 1 week after MI. We providedboth transcriptional and histologic evidence thatthe myocardial matrix mediated this by inducingvarious tissue level changes. These results providefurther evidence for the promise of the myocardialmatrix as a therapy to prevent development of HFafter MI.

ACKNOWLEDGMENTS The authors thankDr. NicholasWebster of the San Diego VA/VAMF Microarray andNGS Core for assistance with microarray analysis aswell as Dr. Christian Metallo for helpful discussionsregarding metabolism.

REPRINT REQUESTS AND CORRESPONDENCE: Dr.Karen L. Christman, Department of Bioengineering,University of California, 2880 Torrey Pines ScenicDrive, La Jolla, California 92037. E-mail: christman@eng.ucsd.edu.

PERSPECTIVES

COMPETENCY IN MEDICAL KNOWLEDGE: In animal

models, injection of a hydrogel derived from decellular-

ized porcine ventricular myocardium attenuated the

decline in cardiac function after MI.

TRANSLATIONAL OUTLOOK: Clinical trials are

needed to evaluate the safety and efficacy of injectable

hydrogels prepared from myocardial extracellular matrix

in patients with MI.

J A C C V O L . 6 7 , N O . 9 , 2 0 1 6 Wassenaar et al.M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6 Effects of Myocardial Matrix Hydrogel After MI

1085

RE F E RENCE S

1. Leri A, Kajstura J, Anversa P. Role of cardiacstem cells in cardiac pathophysiology: a paradigmshift in human myocardial biology. Circ Res 2011;109:941–61.

2. Xin M, Olson EN, Bassel-Duby R. Mendingbroken hearts: cardiac development as a basis foradult heart regeneration and repair. Nat Rev MolCell Biol 2013;14:529–41.

3. Wang ZV, Li DL, Hill JA. Heart failure and loss ofmetabolic control. J Cardiovasc Pharm 2014;63:302–13.

4. Delewi R, Andriessen A, Tijssen JGP, Zijlstra F,Piek JJ, Hirsch A. Impact of intracoronary celltherapy on left ventricular function in the settingof acute myocardial infarction: a meta-analysis ofrandomised controlled clinical trials. Heart 2013;99:225–32.

5. Rane AA, Christman KL. Biomaterials for thetreatment of myocardial infarction a 5-yearupdate. J Am Coll Cardiol 2011;58:2615–29.

6. Ungerleider JL, Christman KL. Concise review:injectable biomaterials for the treatment ofmyocardial infarction and peripheral artery dis-ease: translational challenges and progress. StemCells Transl Med 2014;3:1090–9.

7. Singelyn JM, DeQuach JA, Seif-Naraghi SB,Littlefield RB, Schup-Magoffin PJ, Christman KL.Naturally derived myocardial matrix as an inject-able scaffold for cardiac tissue engineering. Bio-materials 2009;30:5409–16.

8. Singelyn JM, Sundaramurthy P, Johnson TD,et al. Catheter-deliverable hydrogel derived fromdecellularized ventricular extracellular matrix in-creases endogenous cardiomyocytes and pre-serves cardiac function post-myocardial infarction.J Am Coll Cardiol 2012;59:751–63.

9. Seif-Naraghi SB, Singelyn JM, Salvatore MA,et al. Safety and efficacy of an injectable extra-cellular matrix hydrogel for treating myocardialinfarction in pre-clinical animal studies. Sci TranslMed 2013;5:173ra25.

10. Jameel MN, Li Q, Mansoor A, et al. Long-termfunctional improvement and gene expressionchanges after bone marrow-derived multipotentprogenitor cell transplantation in myocardialinfarction. Am J Physiol Heart Circ Physiol 2010;298:H1348–56.

11. Burt RK, Chen Y-h, Verda L, et al. Mitoticallyinactivated embryonic stem cells can beutilized as an in vivo feeder layer to nurse

damaged myocardium following acute myocardialinfarction: a pre-clinical study. Circ Res 2012;111:1286–96.

12. Lachtermacher S, Esporcatte BB, da Silva deAzevedo Fortes F, et al. Functional and tran-scriptomic recovery of infarcted mouse myocar-dium treated with bone marrow mononuclearcells. Stem Cell Rev and Rep 2012;8:251–61.

13. Pavo N, Zimmermann M, Pils D, et al. Long-acting beneficial effect of percutaneously intra-myocardially delivered secretome of apoptoticperipheral blood cells on porcine chronic ischemicleft ventricular dysfunction. Biomaterials 2014;35:3541–50.

14. Dean RA, Cox JH, Bellac CL, et al. Macrophage-specific metalloelastase (MMP-12) truncates andinactivates ELRþ CXC chemokines and generatesCCL2, -7, -8, and -13 antagonists: potential roleof the macrophage in terminating poly-morphonuclear leukocyte influx. Blood 2008;112:3455–64.

15. Martinez FO, Sica A, Mantovani A, Locati M.Macrophage activation and polarization. FrontBiosci 2008;13:453–61.

16. Aurora AB, Porrello ER, Tan W, et al. Macro-phages are required for neonatal heart regenera-tion. J Clin Invest 2014;124:1382–92.

17. Bischoff SC. Role of mast cells in allergic andnon-allergic immune responses: comparison ofhuman and murine data. Nat Rev Immunol 2007;7:93–104.

18. Frangogiannis NG, Smith CW, Entman ML. Theinflammatory response in myocardial infarction.Cardiovasc Res 2002;53:31–47.

19. Somasundaram P, Ren G, Nagar H, et al. Mastcell tryptase may modulate endothelial cellphenotype in healing myocardial infarcts. J Pathol2005;205:102–11.

20. TangL,JenningsT,EatonJW.Mastcellsmediateacute inflammatory responses to implanted bio-materials. Proc Natl Acad Sci U S A 1998;95:8841–6.

21. Lee SH, Wolf PL, Escudero R, Deutsch R,Jamieson SW, Thistlethwaite PA. Early expressionof angiogenesis factors in acute myocardialischemia and infarction. New Engl J Med 2000;342:626–33.

22. Virag JI, Murry CE. Myofibroblast and endo-thelial cell proliferation during murine myocardialinfarct repair. Am J Pathol 2003;163:2433–40.

23. Davis GE, Bayless KJ, Davis MJ, Meininger GA.Regulation of tissue injury responses by the expo-sure ofmatricryptic siteswithin extracellularmatrixmolecules. Am J Pathol 2000;156:1489–98.

24. Palojoki E, Saraste A, Eriksson A, et al. Car-diomyocyte apoptosis and ventricular remodelingafter myocardial infarction in rats. Am J PhysiolHeart Circ Physiol 2001;280:H2726–31.

25. Neely JR, Morgan HE. Relationship betweencarbohydrate and lipid metabolism and the energybalance of heart muscle. Annu Rev Physiol 1974;36:413–59.

26. Duncan J, Finck B. PPAR/PGC-1 regulation ofmetabolism in cardiac disease. In: Patterson C,Willis MS, editors. Translational Cardiology. NewYork: Humana Press, 2012:83–111.

27. Rowe GC, Jiang A, Arany Z. PGC-1 Coactivatorsin cardiac development and disease. Circ Res2010;107:825–38.

28. Jin H, Yang R, Awad TA, et al. Effects of earlyangiotensin-converting enzyme inhibition on car-diac gene expression after acute myocardialinfarction. Circulation 2001;103:736–42.

29. Marín-García J. Signaling Pathways in Cardio-vascular Development. In: Signaling in the Heart.New York, NY: Springer, 2011:155–96.

30. Akhmedov AT, Marín-García J. Myocardialregeneration of the failing heart. Heart Fail Rev2012;18:815–33.

31. Oka T, Maillet M, Watt AJ, et al. Cardiac-specific deletion of Gata4 reveals its requirementfor hypertrophy, compensation, and myocyteviability. Circ Res 2006;98:837–45.

32. Shen T, Aneas I, Sakabe N, et al. Tbx20 regu-lates a genetic program essential to adult mousecardiomyocyte function. J Clin Invest 2011;121:4640–54.

33. Toko H, Zhu W, Takimoto E, et al. Csx/Nkx2-5is required for homeostasis and survival of cardiacmyocytes in the adult heart. J Biol Chem 2002;277:24735–43.

34. Huang J, Min Lu M, Cheng L, et al. Myocardinis required for cardiomyocyte survival and main-tenance of heart function. Proc Natl Acad Sci U S A2009;106:18734–9.

35. Kim Y, Phan D, van Rooij E, et al. The MEF2Dtranscription factor mediates stress-dependentcardiac remodeling in mice. J Clin Invest 2008;118:124–32.

Wassenaar et al. J A C C V O L . 6 7 , N O . 9 , 2 0 1 6

Effects of Myocardial Matrix Hydrogel After MI M A R C H 8 , 2 0 1 6 : 1 0 7 4 – 8 6

1086

36. Azakie A, Fineman JR, He Y. Myocardial tran-scription factors are modulated during pathologiccardiac hypertrophy in vivo. J Thorac CardiovascSurg 2006;132:1262–71.e4.

37. Furtado MB, Costa MW, Adi Pranoto EM, et al.Cardiogenic genes expressed in cardiac fibroblastscontribute to heart development and repair. CircRes 2014;114:1422–34.

38. French KM, Boopathy AV, DeQuach JA, et al.A naturally derived cardiac extracellular matrixenhances cardiac progenitor cell behavior in vitro.Acta Biomater 2012;8:4357–64.

39. Aminzadeh MA, Tseliou E, Sun B, et al.Therapeutic efficacy of cardiosphere-derivedcells in a transgenic mouse model of non-ischaemic dilated cardiomyopathy. Eur Heart J2015;36:751–62.

40. Ellison Georgina M, Vicinanza C, SmithAndrew J, et al. Adult c-kitpos cardiac stem cellsare necessary and sufficient for functional cardiacregeneration and repair. Cell 2013;154:827–42.

41. van Berlo JH, Kanisicak O, Maillet M, et al.c-kitþ cells minimally contribute cardiomyocytesto the heart. Nature 2014;509:337–41.

42. Laurencin CT, Khan Y. Regenerative engi-neering. Sci Transl Med 2012;4:160ed9.

KEY WORDS biomaterial, extracellularmatrix, heart failure, infarction, microarray

APPENDIX For an expanded Methods sec-tion, and supplemental tables, figures, andreferences, please see the online version of thisarticle.