Deficiency of different nitric oxide synthase isoforms activates divergent transcriptional programs in cardiac hypertrophy

Thomas P. Cappola1, Leslie Cope3, Amy Cernetich2, Lili A. Barouch2, Khalid Minhas2, Rafael A. Irizarry3, Giovanni Parmigiani3, Sarfraz Durrani4, Tera Lavoie2, Eric P. Hoffman4, Shui Q. Ye2, Joe G. N. Garcia2 and Joshua M. Hare2

National Heart, Lung, and Blood Institute Program in Genomic Applications-HopGene, Philadelphia, Pennsylvania 19104
1 Department of Medicine, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania 19104
2 Department of Medicine, Johns Hopkins Medical Institutions, Baltimore, Maryland 21287
3 Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland 21287
4 Center for Genetic Medicine, Children’s National Medical Center, Washington, District of Columbia 20010


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 DISCLOSURES
 REFERENCES
 
Decreased nitric oxide synthase (NOS) activity induces left ventricular hypertrophy (LVH), but the transcriptional pathways mediating this effect are unknown. We hypothesized that specific NOS isoform deletion (NOS3 or NOS1) would activate different transcriptional programs in LVH. We analyzed cardiac expression profiles (Affymetrix MG-U74A) from NOS-/- mice using robust multi-array average (RMA). Of 12,422 genes analyzed, 47 genes were differentially expressed in NOS3-/- and 67 in NOS1-/- hearts compared with wild type (WT). Only 16 showed similar changes in both NOS-/- strains, most notably decreased heat-shock proteins (HSP10, 40, 70, 86, 105). Hypertrophied NOS1-/- hearts had unique features, including decreased myocyte-enriched calcineurin interacting protein and paradoxical downregulation of fetal isoforms ({alpha}-skeletal actin and brain natriuretic peptide). Cluster analyses demonstrated that NOS1 deletion caused more pronounced changes in the myocardial transcriptome than did NOS3 deletion, despite similar cardiac phenotypes. These findings suggest that the transcriptional basis for LVH varies depending on the inciting biochemical stimulus. In addition, NOS isoforms appear to play distinct roles in modulating cardiac structure.

ventricular remodeling; mice; knockout; gene expression profiling; genomics


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 DISCLOSURES
 REFERENCES
 
LEFT VENTRICULAR HYPERTROPHY (LVH) may be defined as an increase in cardiac myocyte size accompanied by remodeling of the interstitial matrix and gross cardiac enlargement. LVH is a characteristic pathological feature of common heart diseases such as congestive heart failure and hypertension and is an independent predictor of cardiovascular death (15). As a result, the pathogenesis of LVH has been a major focus of cardiovascular research. A number of insults have been shown to induce LVH, including mechanical overload, oxidative stress, neurohormonal activation, and inherited mutations in sarcomeric proteins (reviewed in 17, 26, 27, 30, 32). This stereotypic response to diverse injuries suggests that a final common transcriptional pathway may ultimately be responsible. To date, a final common "hypertrophy gene program" has not been identified. However, alterations in many genes have been reported, including fetal cardiac isoforms, immediate-early genes, growth factors, redox genes, and members of the MAP kinase, ß-adrenergic, and Ca2+-signaling pathways among others (17, 26, 27, 30, 32). Activation of some of these pathways [particularly fetal isoforms such as or ß-myosin heavy chain (ß-MHC), {alpha}-skeletal actin, and brain natriuretic peptide (BNP)] is now so widely accepted that it is often included as part of the definition of cardiac hypertrophy.

We and others have demonstrated that physiological cardiac nitric oxide (NO) signaling opposes LVH (3, 5, 35). In mice, deletion of nitric oxide synthases (NOS) leads to age-associated LVH as manifested by an increase in both ventricular wall thickness and mass. Notably, deletion of NOS3 (endothelial NOS) causes hypertrophy associated with systemic hypertension, whereas deletion of NOS1 (neuronal NOS) causes LVH with normal systemic blood pressure (3, 11). These findings suggest that NOS1 deficiency induces LVH via load-independent signals and, therefore, may activate a distinct transcriptional pathway.

We hypothesized that deletion of NOS3 or NOS1 would activate different cardiac hypertrophy transcriptional programs. Because a large number of genes have been implicated in LVH pathogenesis, we employed a genomic approach using microarrays and quantitative polymerase chain reaction (qPCR) to analyze multiple genes simultaneously. We tested our hypothesis using three approaches: 1) exploratory analysis of differentially expressed genes in NOS-/- mice, 2) candidate-gene analysis using sets of genes implicated in LVH pathogenesis, and 3) cluster analysis comparing global patterns of gene expression. Our findings demonstrate that hypertrophied NOS3-/- and NOS1-/- hearts have distinct molecular phenotypes, suggesting unique roles of NOS isoforms in the pathogenesis of LVH.


    METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 DISCLOSURES
 REFERENCES
 
Animals.
We studied transgenic mice with homozygous deletions of NOS3 or NOS1, as well as C57BL/6 wild-type (WT) controls (Jackson Laboratories, Bar Harbor, ME). NOS3-/- and NOS1-/- mice initially had a mixed background of C57BL/6 and Sv129 and were then backcrossed at least 10 generations onto C57BL/6 (11, 12). All experimental protocols were approved by the Animal Care and Use Committee of the Johns Hopkins University.

Microarray hybridization.
Myocardial RNA was isolated from individual NOS3-/-, NOS1-/-, and WT mice (n = 3 in each group; age 20 ± 3 mo) using the Trizol reagent and Qiagen RNeasy columns. Seven of the mice were male, and two were female (one NOS3-/- and one NOS1-/-). Individual cDNAs were prepared from each RNA isolate using reverse transcriptase (GIBCO-BRL SuperScript). Each cDNA was subsequently used as a template to make biotin-labeled cRNA using an in vitro transcription reaction (Enzo), resulting in a single cRNA for each heart. Each cRNA was hybridized with an individual Affymetrix MG-U74A oligonucleotide array, which was subsequently processed and scanned according to the manufacturer’s instructions. All arrays were hybridized in the same batch to avoid variability in hybridization conditions. Each array quantifies the expression of 12,422 transcripts (including full-length mRNA sequences and expressed sequence tag clusters) derived from build 74 of the UniGene database (http://www.affymetrix.com). Data were saved as raw image files and converted into probe set data (as "* .cel" files) using Microarray Suite (MAS 5.0). All data are available for download at http://www.hopkins-genomics.org.

Analysis of microarray data.
We used robust multi-array average (RMA) to analyze Affymetrix probe set data (13). Software for RMA is available for download (http://www.bioconductor.org) for use in the R-package for statistical computing (http://www.r-project.org). There are four stages to RMA. First, probe set data ("* .cel" files) from all arrays are simultaneously normalized using quantile normalization, which eliminates systematic differences between GeneChips, without significantly altering the relative intensity of probes within a GeneChip. Second, mean optical background level for each array is estimated, and the intensity for each probe is adjusted to remove this. Third, the normalized, background-corrected data is transformed to the log2 scale. Finally, a median-polish procedure is used to combine multiple probes into a single measure of expression for each gene on each array. Three types of analysis were subsequently performed on this normalized expression data as described below.

Differentially expressed genes in NOS-/- hearts compared with WT controls.
To determine regulated genes in NOS3-/- and NOS1-/- myocardium, we applied the following criteria. First, data were filtered to include genes present above background on at least one array according to MAS 5.0. Next, we required at least a 1.8-fold change in expression between WT and NOS-/- in order for a gene to be considered of biological interest. Finally, changes in gene expression were filtered for statistical significance using the following test statistic

where SErma are the gene-specific standard errors derived from RMA. Genes were considered as differentially expressed for the 2.5% most extreme high and low values of {tau}. Assuming that {tau} follows a t-distribution, this threshold corresponds to a two-sided P < 0.0009 in our data. The identity of each differentially expressed gene was determined using annotation databases (http://www.affymetrix.com/index.affx and http://unchip.org:8080/bio/) or via BLAST searches of the corresponding expressed sequence tag.

Candidate-gene analysis.
We assembled a list of 71 candidate genes previously implicated in the pathogenesis of LVH (see the Supplemental Tables A and B, available as Supplementary Material at the Physiological Genomics web site)1 and identified the corresponding probe sets on the MG-U74A array. In effect, this collection of probe sets defines a hypertrophy-specific subset. Expression levels for each of the candidates were obtained from the normalized expression profiles of NOS3-/-, NOS1-/-, and WT mice (n = 3 for each genotype). For each candidate gene, two-sided unpaired Student’s t-tests were performed comparing expression levels between NOS-/- mice and their WT controls, and genes that showed a change compared with control at the P < 0.05 level were selected. Expression of these genes in each of the NOS-/- samples was normalized to the mean expression level in WT mice, and the results were displayed in an Eisen plot (7), in which red indicates increased expression and green indicates decreased expression compared with WT.

Cluster analysis.
We compared global patterns of gene expression among the mouse strains using hierarchical clustering. All arrays (NOS3-/-, NOS1-/-, WT) were normalized and converted to measures of absolute gene expression using RMA. Data were filtered to include genes that were present above background on at least one array according to MAS 5.0 and that had a max/min expression ratio across all arrays of at least 1.8. Clusters were constructed using Cluster software (http://rana.lbl.gov) using average linkage clustering and three different measures of similarity: Pearson correlation coefficients, Spearman rank correlation coefficients, and Kendall’s tau (7, 8).

Validation.
qPCR was performed to validate changes in seven selected genes using the same myocardial RNA samples analyzed with microarrays. Each RNA sample was treated with DNase I to remove any contaminating genomic DNA and was subsequently used to synthesize cDNA using an oligo(dT)12–18 primer and a first-strand synthesis kit (Invitrogen). Primers for qPCR were designed using Primer Express 2.0 software. Each sample was run in duplicate (6 PCR assays for each mouse strain) on a GeneAmp 7900 Sequence Detection System (PE Applied Biosystems) and analyzed using SDS 2.0 software (Applied Biosystems). For each gene of interest, a standard curve was generated using serial dilutions of WT cDNA. The quantity of gene transcript in unknown samples was estimated using this standard curve, with GAPDH as a normalizer. SYBR green reagent (Applied Biosystems) served as a reporter throughout all experiments. Levels of transcript normalized to GAPDH were compared between experimental and control samples using a Wilcoxon rank-sum test, with P < 0.05 as the criterion for statistical significance.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 DISCLOSURES
 REFERENCES
 
Hypertrophy phenotype.
We have previously described the phenotypic characteristics of NOS-/- mice as assessed by histology, echocardiography, and cardiac catheterization with pressure-volume analysis (3). Both NOS3-/- and NOS1-/- mice develop age-associated LVH characterized by an increase in left ventricular wall thickness and cardiac mass. We confirmed LVH in our experimental animals by direct measurement of cardiac mass normalized to body weight. Compared with WT controls (5.0 ± 0.4 mg/g), NOS3-/- (6.4 ± 0.9 mg/g) and NOS1-/- (6.2 ± 0.8 mg/g) mice exhibited similar increases in cardiac mass at age 20 mo (28% and 25%, respectively).

Differentially expressed genes in NOS-/- hearts.
Of the 12,422 genes and expressed sequence tags present on the microarrays, 6,409 were detected at levels higher than background in at least one of the strains (NOS3-/-, NOS1-/- or WT). As shown in Fig. 1, 47 genes were differentially expressed in NOS3-/- hearts and 65 were differentially expressed in NOS1-/- hearts compared with WT controls. A complete annotated list of these genes is available in the Supplemental Tables A and B. We categorized each of the differentially expressed genes by known or proposed gene function (Table 1). These genes participate in cellular responses that are universally important for cell proliferation (growth and differentiation, transcription, protein turnover, apoptosis), as well as pathways that may be related specifically to cardiac hypertrophy (structure, metabolism, stress response, redox control). A number of genes related to immunoglobulin synthesis were also differentially expressed, perhaps reflecting different degrees of immune activation in NOS3-/- and NOS1-/- mice compared with WT. Alternatively, they could result from different antigenic exposures in the three mouse strains. Remarkably, the direction of change in expression (increase or decrease) was similar for only 16 genes (Table 2).



View larger version (21K):
[in this window]
[in a new window]
 
Fig. 1. Visual display of differential gene expression in NOS-/- hearts compared with wild-type (WT) controls. RNA from NOS3-/- (left), NOS1-/- (right), and WT hearts (n = 3 for each group) was analyzed using individual Affymetrix MG-U74 arrays. The robust multi-array average (RMA) algorithm was used to normalize all nine arrays and to compute average gene expression values for each strain. For each of 12,422 genes, expression ratios comparing NOS-/- and WT hearts were computed and are displayed on the ordinate. The abscissa indicates the basal level of gene expression in WT controls. Genes that show high degrees of differential expression lie far above or far below the dashed lines. We selected genes that showed at least a 1.8-fold change from WT expression and that met criteria for statistical significance. As indicated by the crosses, 47 genes were differentially expressed in NOS3-/- hearts and 65 were differentially expressed in NOS1-/- hearts compared with WT. NOS, nitric oxide synthase.

 

View this table:
[in this window]
[in a new window]
 
Table 1. Differentially expressed genes in NOS-/- mice, grouped by function

 

View this table:
[in this window]
[in a new window]
 
Table 2. Genes with similar changes in expression in NOS3-/- and NOS1-/- relative to WT

 
Candidate-gene analysis.
Many genes previously implicated in cardiac hypertrophy were not differentially expressed in our exploratory analyses. Therefore, we performed a focused analysis using a candidate-gene approach to determine the changes in known hypertrophy genes, which may have been masked by large changes in other, perhaps less relevant, genes.

We performed a literature search to identify candidate hypertrophy genes, 71 of which were present on the MG-U74A microarray. A complete list of candidates is available in the Supplemental Tables A and B. Of these candidates, 32 (52%) were changed in at least one of the experimental strains compared with WT. These results are displayed in Fig. 2. Unlike classic models of load-induced hypertrophy (24), NOS-/- hearts did not show an isoform switch from {alpha}- to ß-MHC or upregulation of {alpha}-skeletal actin. (These genes did not meet criteria for differential expression and are not displayed in Fig. 2.) Likewise, there was no upregulation of natriuretic peptides in NOS-/- hearts. In fact, {alpha}-skeletal actin and BNP were paradoxically underexpressed in the NOS1-/- strain despite the presence of substantial LVH. We also found a potent downregulation of myocyte-enriched calcineurin interacting protein (MCIP-1), an inhibitor of calcineurin-mediated hypertrophy (25), in the NOS1-/- strain. Transcripts for the Ca2+ handling proteins (L-type Ca2+ channel, SERCA2a, ryanodine receptor) were downregulated to a greater extent in NOS3-/- than in NOS1-/-.



View larger version (51K):
[in this window]
[in a new window]
 
Fig. 2. Candidate-gene analysis. We performed a literature search to identify candidate hypertrophy genes, 71 of which were present on the MG-U74A microarray. Of these candidates, 32 genes showed a significant change compared with control at the P < 0.05 level and are displayed. Data are normalized to average level of expression in WT controls. Red indicates increased expression vs. control; green indicates decreased expression; E, NOS3-/-; N, NOS1-/-. Genes are grouped by functional categories. Unlike classic models of load-induced hypertrophy, brain natriuretic peptide (BNP) and {alpha}-skeletal actin were paradoxically decreased in hypertrophied NOS1-/- hearts. MCIP-1, an inhibitor of calcineurin-mediated hypertrophy, was selectively decreased in NOS1-/-, and may promote left ventricular hypertrophy (LVH) in this strain.

 
Cluster analysis.
Our results to this point suggested distinct transcriptional profiles in NOS3-/- and NOS1-/- hearts. To test this formally, we used hierarchical clustering. This computational method groups experimental samples into clusters according to similarity in patterns of gene expression across a large number of genes. The results are displayed in a dendrogram (Fig. 3) with short branches indicating a high degree of similarity, providing an objective way to compare relationships among samples (13). Of the 12,422 genes and expressed sequence tags present on the microarrays, 253 were detected at levels higher than background and had a max/min expression ratio greater than 1.8 and were included in the clustering algorithm. Using three different similarity functions, we found that the NOS1-/- hearts clustered as a distinct group. In contrast, the expression profiles among NOS3-/- and WT mice showed a higher degree of similarity, providing further evidence that the transcriptional responses to NOS3 and NOS1 deletion in the heart are distinct.



View larger version (11K):
[in this window]
[in a new window]
 
Fig. 3. Cluster analysis of gene expression in NOS-/- hearts. Expression profiles from each strain were normalized using RMA and filtered to remove genes that showed minimal change in expression. The remaining 229 genes were used as input for hierarchical clustering using three different similarity functions: Pearson correlation coefficients (A), Spearman correlation coefficients (B), and Kendall’s tau (C). In all three analyses, expression profiles showed a high degree of similarity in NOS3-/- and WT mice, whereas NOS1-/- mice were a distinct group. E, NOS3-/-; N, NOS1-/-.

 
Quantitative PCR.
We selected seven genes of biological interest in LVH and validated the microarray findings in NOS3-/-, NOS1-/-, and WT hearts using qPCR (Table 3). As shown in Fig. 4B, qPCR confirmed 12 of 14 microarray predictions, including decreased heat-shock proteins (HSP) in NOS3-/- and NOS1-/- and decreased MCIP-1, BNP, and {alpha}-skeletal actin in NOS1-/-. Neither qPCR nor microarray detected significant changes in {alpha}- or ß-MHC in either strain. In two cases ({alpha}-skeletal actin and MCIP-1), the microarrays failed to detect changes that were detected by the more sensitive qPCR technique in the NOS3-/- hearts. There were no disagreements among microarray and qPCR findings in the direction of change in expression compared with control.


View this table:
[in this window]
[in a new window]
 
Table 3. Primers used for quantitative PCR

 


View larger version (14K):
[in this window]
[in a new window]
 
Fig. 4. Quantitative PCR (qPCR). To validate selected microarray findings using an independent methodology, we quantified transcript abundance of seven genes using qPCR. A: transcript abundance of seven genes normalized to GADPH copy number in WT, NOS3-/-, and NOS1-/- mice. B: fold change in expression in NOS3-/- and NOS1-/- mice compared with WT according to qPCR (black bars) and microarrays (gray bars). *P < 0.05 compared with WT by Wilcoxon rank-sum test. {dagger}P < 0.05 by RMA. qPCR confirmed 12 of 14 microarray predictions, including decreased HSP70 in NOS3-/- and NOS1-/-, and decreased MCIP-1, BNP, and {alpha}-skeletal actin in NOS1-/-. Neither qPCR nor microarrays detected significant changes in {alpha}- or ß-MHC in either strain. In two cases ({alpha}-skeletal actin and MCIP-1), microarrays failed to detect changes that were detected by the more sensitive qPCR technique in the NOS3-/- hearts.

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 DISCLOSURES
 REFERENCES
 
The principal finding of this investigation is that cardiac hypertrophy in NOS3-/- and NOS1-/- mice is associated with distinct transcriptional changes despite similar cardiac phenotypes. We demonstrated this using three analytic approaches: exploratory analysis of a large number of genes, candidate-gene analysis of relevant hypertrophy genes, and hierarchical clustering based on global patterns of gene expression. In addition, we validated selected microarray results using qPCR as an independent methodology. Our findings support the notion that the transcriptional basis of cardiac hypertrophy can vary, with specific changes dependent upon the inciting biochemical stimulus.

Some of our findings are counterintuitive, such as the absence of changes in myosin isoforms and paradoxical decreases in BNP and {alpha}-skeletal actin. However, they add to a growing literature revealing the diversity of transcriptional changes associated with hypertrophy. Aronow et al. (2) compared expression profiles from mice with hypertrophy induced by four distinct transgenic alterations and found that no single hypertrophy program was responsible. A number of functional and transgenic studies suggest that calcium/calcineurin signaling pathways play a central role in mediating hypertrophy (9, 19). However, pharmacological inhibition of calcineurin as a strategy for preventing LVH has had mixed results (31, 36), indicating that calcineurin signaling may not mediate all forms of hypertrophy. Studies in humans indicate that increased fetal isoform expression, which is regarded as a defining characteristic of hypertrophy in animal models, may not always accompany human heart failure. Rather, fetal isoform expression may remain unchanged, while adult isoform expression decreases (23).

The divergent transcriptional responses in NOS3-/- and NOS1-/- reflect the functional specificity of NOS isoforms in the cardiovascular system (3). NOS1 is localized to the sarcoplasmic reticulum (SR) and modulates the function of the SR calcium release channel (ryanodine receptor). We have shown that deletion of NOS1 impairs basal SR calcium cycling (3). Long-term dysregulation of calcium cycling has been shown to alter cardiac gene transcription and stimulate hypertrophy (9). In contrast, NOS3, which is localized to the sarcolemma, has no effect on basal calcium cycling in the heart, but is a potent cause of systemic hypertension due to effects on the peripheral vasculature (12). These different functional effects, i.e., altered calcium cycling and altered vascular load, may mediate the different downstream changes in cardiac gene transcription that we observed in NOS1-/- and NOS3-/-. The selective downregulation of MCIP-1 in NOS1-/- hearts, which we demonstrated both on microarray and on qPCR (Figs. 2 and 4), offers another link between NOS1 deletion and calcium signaling. MCIP-1 is an endogenous inhibitor of the calcineurin signaling pathway. When transgenically overexpressed, MCIP-1 prevents hypertrophy induced by calcineurin overexpression, ß-adrenergic stimulation, or exercise training (25). Decreased MCIP-1 in NOS1-/- may therefore promote calcineurin signaling and contribute to hypertrophy.

Although we have emphasized the divergent transcriptional changes in NOS-/- mice, some genes showed similar changes in expression in both strains (Table 2). Most notably, a number of HSPs are markedly reduced in both NOS-/- strains. In particular, HSP70 is downregulated, both on microarray and on qPCR (Fig. 4A). HSP70 protects against adverse ventricular remodeling and prevents myocardial infarction in the setting of ischemia/reperfusion injury (4, 20). HSP downregulation in NOS-/- hearts is consistent with previous findings that NO induces HSP70 expression via activation of the transcription factor HSF1 (34). Impaired HSP function in NOS deficiency may contribute to worsening hypertrophy and cardiac dysfunction. Another noteworthy finding is that Bag3, an inhibitor of apoptosis, is underexpressed in both NOS-/- strains (Table 2), consistent with the well-established role of apoptosis in cardiac hypertrophy and remodeling (6).

Among the transgenic causes of hypertrophy explored in model systems, hypertrophy caused by NOS deletion may be especially relevant to human congestive heart failure. Heart failure is characterized by an imbalance between the formation and degradation of reactive oxygen species within the myocardium (14, 18), and the resultant oxidative stress is sufficient to stimulate cardiac hypertrophy (21, 28). Inactivation of NO signaling by reactive oxygen species may be one of the major mechanisms whereby oxidative stress promotes hypertrophy (10). Moreover, treatments that have been shown to attenuate cardiac hypertrophy in heart failure (angiotensin converting enzyme inhibitors, beta-blockers, and hydralazine/nitrates), increase NO bioavailability and promote NO signaling (1, 33). Our findings identify specific transcriptional targets of NO signaling in cardiac hypertrophy, some of which may be targets for pharmacological intervention.

Expression profiling is a powerful technique that offers tremendous promise, but it creates substantial analytic challenges resulting from the analysis of many genes in a small number of samples (22). We addressed these at multiple levels. Of the methods to convert Affymetrix probe set data into measures of gene expression, which include software supplied by the manufacturer, model-based methods (16), and RMA, we chose RMA based on its superiority in the analysis of small data sets (13). For the exploratory analyses, we used strict selection criteria to define differential gene expression to select biologically significant genes and to minimize spurious findings that result from multiple comparisons. Furthermore, we independently validated findings of interest using qPCR. We also used a candidate-gene approach to detect changes in highly relevant genes that could be overlooked in exploratory analyses.

Several limitations warrant mention. Microarray technologies only assess the abundance of transcripts, which may not correlate with the abundance, localization, or function of corresponding proteins. However, many genes (e.g., HSPs) are principally regulated at the level of transcription and therefore may be accurately characterized by microarrays. We chose to use the entire heart to limit the impact of tissue heterogeneity on quantification of transcript abundance; however, this prevented us from discerning chamber-specific alterations in gene transcription. This may be important in light of the tendency of NOS3-/- mice to develop pulmonary hypertension and right ventricular hypertrophy when they are subjected to chronic hypoxemia (29). Although we analyzed 71 candidate hypertrophy genes, our list of candidates was limited to genes present on the microarray, and not all relevant hypertrophy genes could be analyzed. Our data were obtained in murine models of age-associated LVH. For this reason, our findings may not directly parallel the pathogenesis of LVH in humans and may not reflect the early transcriptional changes caused by hypertension or myocardial infarction. However, prior investigations have demonstrated altered NOS signaling in aged hearts, suggesting that age-associated hypertrophy in NOS knockouts is a pathophysiologically relevant phenotype (37).

In summary, our analyses show that deficiency of different NOS isoforms activates divergent transcriptional programs in cardiac hypertrophy. In addition, deficiency of HSPs may play a central role in initiating or potentiating LVH, and downregulation of MCIP-1 may promote LVH in the absence of NOS1. These findings add to a growing literature suggesting that the transcriptional basis of LVH varies depending on the inciting biochemical stimulus. Moreover, we provide new insights into the role of NOS isoforms in maintaining normal cardiac architecture.


    DISCLOSURES
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 DISCLOSURES
 REFERENCES
 
This research was supported by National Institutes of Health Grants 1U01-HL-066583 (to J. G. N. Garcia) and 5R01-HL-065455 (to J. M. Hare).

T. P. Cappola was a recipient of a Pfizer Postdoctoral Fellowship in Cardiovascular Medicine. J. M. Hare is a recipient of a Paul Beeson Physician Faculty Scholars in Aging Research Award.


    FOOTNOTES
 
Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

Address for reprint requests and other correspondence: J. M. Hare, Johns Hopkins Hospital, Division of Cardiology, 600 N. Wolfe St., Carnegie 568, Baltimore, MD 21287-6568 (E-mail: jhare{at}mail.jhmi.edu).

10.1152/physiolgenomics.00156.2002.

1 The Supplementary Material for this article (Tables A and B) is available online at http://physiolgenomics.physiology.org/cgi/content/full/00156.2002/DC1. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 DISCLOSURES
 REFERENCES
 

  1. Ambrosioni E, Bacchelli S, Esposti DD, and Borghi C. Beta-blockade in hypertension and congestive heart failure. J Cardiovasc Pharmacol 38, Suppl 3: S25–S31, 2001.[ISI][Medline]
  2. Aronow BJ, Toyokawa T, Canning A, Haghighi K, Delling U, Kranias E, Molkentin JD, andDorn GW II. Divergent transcriptional responses to independent genetic causes of cardiac hypertrophy. Physiol Genomics 6: 19–28, 2001.[Abstract/Free Full Text]
  3. Barouch LA, Harrison RW, Skaf MW, Rosas GO, Cappola TP, Kobeissi ZA, Hobai IA, Lemmon CA, Burnett AL, O’Rourke B, Rodriguez ER, Huang PL, Lima JA, Berkowitz DE, and Hare JM. Nitric oxide regulates the heart by spatial confinement of nitric oxide synthase isoforms. Nature 416: 337–339, 2002.[ISI][Medline]
  4. Benjamin IJ and McMillan DR. Stress (heat shock) proteins: molecular chaperones in cardiovascular biology and disease. Circ Res 83: 117–132, 1998.[Abstract/Free Full Text]
  5. Calderone A, Thaik CM, Takahashi N, Chang DL, and Colucci WS. Nitric oxide, atrial natriuretic peptide, and cyclic GMP inhibit the growth-promoting effects of norepinephrine in cardiac myocytes and fibroblasts. J Clin Invest 101: 812–818, 1998.[Abstract/Free Full Text]
  6. Condorelli G, Morisco C, Stassi G, Notte A, Farina F, Sgaramella G,de Rienzo A, Roncarati R, Trimarco B, and Lembo G. Increased cardiomyocyte apoptosis and changes in proapoptotic and antiapoptotic genes bax and bcl-2 during left ventricular adaptations to chronic pressure overload in the rat. Circulation 99: 3071–3078, 1999.[Abstract/Free Full Text]
  7. Eisen MB, Spellman PT, Brown PO, and Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95: 14863–14868, 1998.[Abstract/Free Full Text]
  8. Fisher L andVan Belle G. Biostatistics: a Methodology for the Health Sciences. New York: Wiley, 1993.
  9. Frey N, McKinsey TA, and Olson EN. Decoding calcium signals involved in cardiac growth and function. Nat Med 6: 1221–1227, 2000.[ISI][Medline]
  10. Hare JM. Oxidative stress and apoptosis in heart failure progression. Circ Res 89: 198–200, 2001.[Free Full Text]
  11. Huang PL, Dawson TM, Bredt DS, Snyder SH, and Fishman MC. Targeted disruption of the neuronal nitric oxide synthase gene. Cell 75: 1273–1286, 1993.[ISI][Medline]
  12. Huang PL, Huang Z, Mashimo H, Bloch KD, Moskowitz MA, Bevan JA, and Fishman MC. Hypertension in mice lacking the gene for endothelial nitric oxide synthase. Nature 377: 239–242, 1995.[ISI][Medline]
  13. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U, and Speed TP. Exploration, normalization, and summaries of high density oligonucleotide array probe level data. Biostatistics 4: 249–264, 2003.[Abstract/Free Full Text]
  14. Keith M, Geranmayegan A, Sole MJ, Kurian R, Robinson A, Omran AS, and Jeejeebhoy KN. Increased oxidative stress in patients with congestive heart failure. J Am Coll Cardiol 31: 1352–1356, 1998.[ISI][Medline]
  15. Levy D, Larson MG, Vasan RS, Kannel WB, and Ho KK. The progression from hypertension to congestive heart failure. JAMA 275: 1557–1562, 1996.[Abstract]
  16. Li C and Wong WH. Model-based analysis of oligonucleotide arrays: expression index computation and outlier detection. Proc Natl Acad Sci USA 98: 31–36, 2001.[Abstract/Free Full Text]
  17. MacLellan WR and Schneider MD. Genetic dissection of cardiac growth control pathways. Annu Rev Physiol 62: 289–319, 2000.[ISI][Medline]
  18. Mallat Z, Philip I, Lebret M, Chatel D, Maclouf J, and Tedgui A. Elevated levels of 8-iso-prostaglandin F2{alpha} in pericardial fluid of patients with heart failure: a potential role for in vivo oxidant stress in ventricular dilatation and progression to heart failure. Circulation 97: 1536–1539, 1998.[Abstract/Free Full Text]
  19. Molkentin JD, Lu JR, Antos CL, Markham B, Richardson J, Robbins J, Grant SR, and Olson EN. A calcineurin-dependent transcriptional pathway for cardiac hypertrophy. Cell 93: 215–228, 1998.[ISI][Medline]
  20. Okubo S, Wildner O, Shah MR, Chelliah JC, Hess ML, and Kukreja RC. Gene transfer of heat-shock protein 70 reduces infarct size in vivo after ischemia/reperfusion in the rabbit heart. Circulation 103: 877–881, 2001.[Abstract/Free Full Text]
  21. Pimentel DR, Amin JK, Xiao L, Miller T, Viereck J, Oliver-Krasinski J, Baliga R, Wang J, Siwik DA, Singh K, Pagano P, Colucci WS, and Sawyer DB. Reactive oxygen species mediate amplitude-dependent hypertrophic and apoptotic responses to mechanical stretch in cardiac myocytes. Circ Res 89: 453–460, 2001.[Abstract/Free Full Text]
  22. Rao JS and Bond M. Microarrays: managing the data deluge. Circ Res 88: 1226–1227, 2001.[Free Full Text]
  23. Razeghi P, Young ME, Alcorn JL, Moravec CS, Frazier OH, and Taegtmeyer H. Metabolic gene expression in fetal and failing human heart. Circulation 104: 2923–2931, 2001.[Abstract/Free Full Text]
  24. Rockman HA, Ross RS, Harris AN, Knowlton KU, Steinhelper ME, Field LJ,Ross J Jr, and Chien KR. Segregation of atrial-specific and inducible expression of an atrial natriuretic factor transgene in an in vivo murine model of cardiac hypertrophy. Proc Natl Acad Sci USA 88: 8277–8281, 1991.[Abstract]
  25. Rothermel BA, McKinsey TA, Vega RB, Nicol RL, Mammen P, Yang J, Antos CL, Shelton JM, Bassel-Duby R, Olson EN, and Williams RS. Myocyte-enriched calcineurin-interacting protein, MCIP1, inhibits cardiac hypertrophy in vivo. Proc Natl Acad Sci USA 98: 3328–3333, 2001.[Abstract/Free Full Text]
  26. Sadoshima J and Izumo S. The cellular and molecular response of cardiac myocytes to mechanical stress. Annu Rev Physiol 59: 551–571, 1997.[ISI][Medline]
  27. Sawyer DB, Siwik DA, Xiao L, Pimentel DR, Singh K, and Colucci WS. Role of oxidative stress in myocardial hypertrophy and failure. J Mol Cell Cardiol 34: 379–388, 2002.[ISI][Medline]
  28. Siwik DA, Tzortzis JD, Pimental DR, Chang DL, Pagano PJ, Singh K, Sawyer DB, and Colucci WS. Inhibition of copper-zinc superoxide dismutase induces cell growth, hypertrophic phenotype, and apoptosis in neonatal rat cardiac myocytes in vitro. Circ Res 85: 147–153, 1999.[Abstract/Free Full Text]
  29. Steudel W, Scherrer-Crosbie M, Bloch KD, Weimann J, Huang PL, Jones RC, Picard MH, and Zapol WM. Sustained pulmonary hypertension and right ventricular hypertrophy after chronic hypoxia in mice with congenital deficiency of nitric oxide synthase 3. J Clin Invest 101: 2468–2477, 1998.[Abstract/Free Full Text]
  30. Sugden PH. Signalling pathways in cardiac myocyte hypertrophy. Ann Med 33: 611–622, 2001.[ISI][Medline]
  31. Sussman MA, Lim HW, Gude N, Taigen T, Olson EN, Robbins J, Colbert MC, Gualberto A, Wieczorek DF, and Molkentin JD. Prevention of cardiac hypertrophy in mice by calcineurin inhibition. Science 281: 1690–1693, 1998.[Abstract/Free Full Text]
  32. Watkins H, Seidman JG, and Seidman CE. Familial hypertrophic cardiomyopathy: a genetic model of cardiac hypertrophy. Hum Mol Genet 4: Spec No 1721–1727, 1995.
  33. Wittstein IS, Kass DA, Pak PH, Maughan WL, Fetics B, and Hare JM. Cardiac nitric oxide production due to angiotensin-converting enzyme inhibition decreases beta-adrenergic myocardial contractility in patients with dilated cardiomyopathy. J Am Coll Cardiol 38: 429–435, 2001.[ISI][Medline]
  34. Xu Q, Hu Y, Kleindienst R, and Wick G. Nitric oxide induces heat-shock protein 70 expression in vascular smooth muscle cells via activation of heat shock factor 1. J Clin Invest 100: 1089–1097, 1997.[Abstract/Free Full Text]
  35. Yang XP, Liu YH, Shesely EG, Bulagannawar M, Liu F, and Carretero OA. Endothelial nitric oxide gene knockout mice: cardiac phenotypes and the effect of angiotensin-converting enzyme inhibitor on myocardial ischemia/reperfusion injury. Hypertension 34: 24–30, 1999.[Abstract/Free Full Text]
  36. Zhang W, Kowal RC, Rusnak F, Sikkink RA, Olson EN, and Victor RG. Failure of calcineurin inhibitors to prevent pressure-overload left ventricular hypertrophy in rats. Circ Res 84: 722–728, 1999.[Abstract/Free Full Text]
  37. Zieman SJ, Gerstenblith G, Lakatta EG, Rosas GO, Vandegaer K, Ricker KM, and Hare JM. Upregulation of the nitric oxide-cGMP pathway in aged myocardium: physiological response to L-arginine. Circ Res 88: 97–102, 2001.[Abstract/Free Full Text]