Gene expression analysis of ischemic and nonischemic cardiomyopathy: shared and distinct genes in the development of heart failure
Michelle M. Kittleson1,
Khalid M. Minhas1,
Rafael A. Irizarry2,
Shui Q. Ye3,
Gina Edness1,
Elayne Breton1,
John V. Conte4,
Gordon Tomaselli1,
Joe G. N. Garcia3 and
Joshua M. Hare1
1 Division of Cardiology, Department of Medicine, Johns Hopkins Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland
3 Division of Pulmonary and Critical Care Medicine, Department of Medicine, Johns Hopkins Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland
4 Cardiothoracic Division, Department of Surgery, Johns Hopkins School of Medicine, Johns Hopkins University, Baltimore, Maryland
2 Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health, Johns Hopkins University, Baltimore, Maryland
 |
ABSTRACT
|
---|
Cardiomyopathy can be initiated by many factors, but the pathways from unique inciting mechanisms to the common end point of ventricular dilation and reduced cardiac output are unclear. We previously described a microarray-based prediction algorithm differentiating nonischemic (NICM) from ischemic cardiomyopathy (ICM) using nearest shrunken centroids. Accordingly, we tested the hypothesis that NICM and ICM would have both shared and distinct differentially expressed genes relative to normal hearts and compared gene expression of 21 NICM and 10 ICM samples with that of 6 nonfailing (NF) hearts using Affymetrix U133A GeneChips and significance analysis of microarrays. Compared with NF, 257 genes were differentially expressed in NICM and 72 genes in ICM. Only 41 genes were shared between the two comparisons, mainly involved in cell growth and signal transduction. Those uniquely expressed in NICM were frequently involved in metabolism, and those in ICM more often had catalytic activity. Novel genes included angiotensin-converting enzyme-2 (ACE2), which was upregulated in NICM but not ICM, suggesting that ACE2 may offer differential therapeutic efficacy in NICM and ICM. In addition, a tumor necrosis factor receptor was downregulated in both NICM and ICM, demonstrating the different signaling pathways involved in heart failure pathophysiology. These results offer novel insight into unique disease-specific gene expression that exists between end-stage cardiomyopathy of different etiologies. This analysis demonstrates that transcriptome analysis offers insight into pathogenesis-based therapies in heart failure management and complements studies using expression-based profiling to diagnose heart failure of different etiologies.
microarray; heart failure
 |
INTRODUCTION
|
---|
DILATED CARDIOMYOPATHY is a common cause of congestive heart failure and the leading cause of cardiovascular morbidity and mortality in the United States (27). Dilated cardiomyopathy can be initiated by a variety of factors, such as ischemia, pressure or volume overload, myocardial inflammation or infiltration, and inherited mutations (14). A prevailing hypothesis is that, despite the varied inciting mechanisms that initiate the heart failure syndrome, there is a final common pathway that drives heart failure progression (47). Because of this, there is limited research into specific molecular events that are unique to the underlying process. This issue is especially relevant in the two major forms of dilated cardiomyopathy, nonischemic (NICM) and ischemic (ICM). Although NICM and ICM have similar presentations (26), they are characterized by different pathophysiology, prognosis, and response to therapy (19, 21, 23, 24, 32, 42), and understanding their different pathophysiological mechanisms is essential in guiding future therapies.
The emergence of microarray technology to simultaneously assess mRNA levels of tens of thousands of genes offers a novel approach to compare and contrast the myocardial transcriptome of NICM and ICM. Although previous studies have examined changes in gene expression in failing vs. nonfailing (NF) hearts (1, 5, 44, 45, 51), they have focused only on NICM. The goal of this study was to simultaneously examine the differences in transcriptomes between either NICM or ICM and normal hearts to establish a set of shared and unique genes differentially expressed in the two major causes of heart failure. The present approach is distinct, but complementary, to our previous study (33) in which we used the method of nearest shrunken centroids (46) to determine a clinical prediction algorithm (i.e., a gene expression-based biomarker) that differentiated between NICM and ICM. The current analysis offers insight into both disease-specific pathogenesis and therapeutics. Furthermore, an understanding of the distinctions with potential pathophysiological underpinnings between these two conditions supports and complements ongoing biomarker development efforts to differentiate heart failure of different etiologies (33).
 |
METHODS
|
---|
Patient population.
The study sample was composed of 31 end-stage cardiomyopathy and 6 NF hearts. Myocardial tissue from end-stage cardiomyopathy patients was obtained at the time of left ventricular assist device (LVAD) placement or cardiac transplantation from two institutions: 1) Johns Hopkins Hospital in Baltimore, MD (n = 24 NICM and ICM samples and 6 NF samples), and 2) University of Minnesota in Minneapolis, MN (n = 7 NICM samples). Samples from the latter institution were collected and prepared independently (11), and gene expression data files were kindly provided.
Discarded myocardial tissue from the left ventricular free wall or apex obtained during surgery was immediately frozen in liquid nitrogen and stored at 80°C. There is no evidence that differences in left ventricular sampling sites contribute to sample variability, and, in our previous experience, the sampling of tissue from these two sites did not contribute to variability in gene expression (33). The dissecting pathologist selectively excluded areas of visible fibrosis from the portion stored for analysis. Because myocardial tissue obtained at LVAD placement and unused donor hearts are considered discarded tissue, we obtained an exemption from the Johns Hopkins Institutional Review Board for sample collection and medical chart abstraction without written informed consent.
Differential gene expression.
Details of sample preparation, microarray hybridization, data normalization, and validation with quantitative PCR are available as Supplemental Material (see the Physiological Genomics web site).1
We identified differentially expressed genes in two comparisons: 1) NICM vs. NF hearts and 2) ICM vs. NF hearts. Statistically significant changes in gene expression were identified using significance analysis of microarrays (SAM) (49). SAM identifies genes with statistically significant changes in expression by identifying a set of gene-specific statistics (similar to the t-test) and a corresponding false discovery rate (FDR; similar to a P value adjusted for multiple comparisons). Using the "one class" option, we identified genes with an FDR of <5% (corresponding to a P value adjusted for multiple comparisons <0.05) and an absolute fold change of
2.0. This threshold has been used in other similar studies (44) and may maximize specificity (20). These differentially expressed genes were visualized by hierarchical clustering (2) and heat mapping (22), using Euclidean distance with complete linkage.
 |
RESULTS
|
---|
Clinical specimens.
Subjects with ischemic (n = 10) or nonischemic (n = 21) end-stage cardiomyopathy exhibited severely reduced ejection fraction, left ventricular dilation, elevated pulmonary arterial and wedge pressures, and reduced cardiac index (Table 1). ICM subjects were older, all male, more often on angiotension-converting enzyme inhibitors, and less often on intravenous inotropic therapy. Compared with no-LVAD patients, pre-LVAD patients had lower ejection fraction, higher pulmonary capillary wedge pressure, and lower cardiac index. The NF hearts (n = 6) were from unused cardiac transplant donors. The unused donor subjects were younger (median age 42 yr with interquartile range 2450 yr) and predominantly male, and echocardiographic and hemodynamic information and medications were not available.
Differential gene expression: NICM vs. NF and ICM vs. NF.
There were 257 genes differentially expressed between NICM and NF samples and 72 genes differentially expressed between ICM and NF samples, with a FDR of <5% and an absolute fold change of
2.0. Of the differentially expressed genes, only 41 were common to both NICM and NF and ICM and NF comparisons. As a measure of variability of gene expression, the coefficient of variation for these differentially expressed genes is depicted in Supplementary Fig. S1. The coefficient of variation is low and comparable for both NICM and ICM.
Differentially expressed genes common to both NICM-NF and ICM-NF comparisons.
The majority of the 41 shared genes fell into functional classes of cell growth, maintenance, and signal transduction (Fig. 1). Genes implicated in the fetal gene program induction were among those differentially expressed, including downregulation of
-myosin heavy chain polypeptide-6 (36) and upregulation of atrionatriuretic peptide receptor C (18). In the cell growth and maintenance class, there were multiple probes corresponding to hemoglobin
- and ß-chains. There were also genes involved in signal transduction, including endothelin receptor type A and monocyte chemotactic protein-1. In addition, there were genes encoding components of the sarcomere (
-myosin heavy chain noted above), the cytoskeleton (collagen type-21
and ficolin), and the extracellular matrix (asporin). The majority of the genes were upregulated in NICM and ICM hearts compared with NF hearts, and for all 41 shared genes, fold changes were remarkably similar in direction and magnitude between NICM-NF and ICM-NF comparisons (Table 2).

View larger version (15K):
[in this window]
[in a new window]
|
Fig. 1. Percentage of known genes in each functional category that were significantly regulated in both nonischemic (NICM) and ischemic cardiomyopathy (ICM) compared with nonfailing (NF) hearts (black bars). Light gray bars: unique to NICM hearts. White bars: unique to ICM hearts. Dark gray bars: the representation of these functional categories on the array. There is no correlation between the representation of genes on the array and distribution of genes in the comparisons. APO, apoptosis; BIN, binding; CAT, catalytic activity; CEL, cell adhesion; CGM, cell growth/maintenance; CYT, cytoskeleton; DEV, development; INF, inflammatory response; MET, metabolism; NUC, nucleus; SIG, signal transduction; TRA, transcription.
|
|
Differentially expressed genes unique to the NICM-NF comparison.
Of the 216 genes that were uniquely differentially expressed in NICM hearts, the majority fell into metabolism, cell growth, maintenance, signal transduction, and binding (Fig. 1 and Supplementary Table S1). The genes involved in metabolism included angiotensin I-converting enzyme-2 (ACE2) and genes involved in fatty acid and cholesterol metabolism (acyl-CoA synthetase long chain family member-3 and oxysterol binding protein-like-8). In cell growth and maintenance, upregulated genes included cyclin-dependent kinase inhibitor-1B and delta sleep-inducing peptide, a vagal-potentiating peptide with influences on cardiac rhythm (39). Genes involved in signaling pathways were upregulated, including signal transducer and activator of transcription-1 and -4 (members of the JAK/STAT signaling pathway), as well as receptors for leptin, growth hormone, transforming growth factor-ß, and platelet-derived growth factor. Several genes implicated in inflammation and the immune response showed increased expression in NICM hearts, including interleukin-27, a major histocompatibility complex molecule, and a component of the complement pathway, H factor-1. There were also several genes related to cell adhesion, apoptosis, and development. All genes were upregulated in NICM hearts except one: a zinc transporter that was downregulated twofold.
Differentially expressed genes unique to the ICM-NF comparison.
The 31 genes uniquely differentially expressed between ICM and NF hearts were predominantly in functional classes of cell growth, maintenance, catalytic activity, and signal transduction (Fig. 1 and Supplementary Table S2). They also included genes implicated in the fetal gene program induction, including upregulation of natriuretic peptide precursor B, atrial natriuretic factor, and an embryonic atrial myosin light chain polypeptide (14).
Differentially expressed genes and functional categories.
As shown in Fig. 1, the majority of genes on the array (>50%) belonged to functional classes of binding and metabolism; a moderate number of genes (1540%) were in the classes of catalytic activity, cell growth/maintenance, development, nucleus, signal transduction, and transcription; and few genes (<10%) belonged to classes of apoptosis, cell adhesion, cytoskeleton, and inflammatory response (the combined percentages total over 100%, since genes can belong to more than one functional category). This pattern does not match that of our data (P < 0.001 in a
2-test). This suggests that the differences in functional categories identified were not solely a function of their representation on the microarray.
Clustering.
The heat maps with clustering algorithms for the two comparisons, ICM-NF and NICM-NF, are shown in Fig. 2. The NF samples formed a distinct cluster from the ICM samples. For the NICM-NF comparison, there were two dominant clusters. One dominant cluster contained only NICM samples obtained from patients at the time of LVAD implantation (NICM/pre-LVAD). The other dominant cluster contained two subgroups: 1) predominantly NF samples and 2) the remaining portion of NICM samples, which were all obtained from patients who did not have an LVAD before cardiac transplantation (NICM/no-LVAD). Thus there was a clear discrimination among the NICM samples obtained from 1) patients who required LVADs before cardiac transplantation and 2) patients who survived to cardiac transplantation without LVAD support.

View larger version (61K):
[in this window]
[in a new window]
|
Fig. 2. Hierarchical clustering of genes based on similarity in gene expression and relatedness of samples. Each row represents a gene, and each column represents a sample. Sample prefix "T" denotes end samples from patients at the time of cardiac transplantation without left ventricular assist devices (no-LVAD), "LC" denotes samples obtained from patients at the time of LVAD placement (pre-LVAD), and "N" denotes NF samples. The suffix "i" denotes ICM samples, and the suffix "ni" denotes NICM samples. The color in each cell reflects the level of expression of the corresponding gene in the corresponding sample, relative to its mean level of expression in the entire set of samples. Expression levels greater than the mean are shaded in blue, and those below the mean are shaded in red. Circled samples denote the predominant etiology clusters, and samples labeled with an arrow fall outside of their appropriate cluster. A: ICM vs. NF. The no- and pre-LVAD samples do not form distinct clusters. B: NICM vs. NF. The no- and pre-LVAD samples form distinct clusters, as indicated.
|
|
To determine the specificity of the profiles, we also created a heat map with clustering algorithm for all 288 genes that were identified as differentially expressed in at least one of the two comparisons (Supplementary Fig. S2). Samples formed three distinct etiology clusters, NF, ICM, and NICM, but this was likely due to the presence of shared differentially expressed genes. To confirm the specificity of the differentially expressed genes, we performed two additional heat maps with clustering (Supplementary Fig. S3, A and B): first, NF and ICM samples using only those genes identified as differentially expressed between NF and NICM samples, and second, NF and NICM samples using only those genes identified as differentially expressed between NF and ICM samples. If, as we assumed, the genes uniquely identified as differentially expressed in ICM relative to NF hearts were truly unique to the ICM-NF comparison, then a heat map of these genes in NICM and NF hearts should demonstrate no clustering by etiology, and vice versa for NICM genes in ICM hearts. This was the case: as expected, in both heat maps, the samples did not cluster by etiology, indicating that the unique differentially expressed genes were specific to the given comparison.
Validation.
We selected 16 genes of potential biological interest and validated the microarray findings in NICM, ICM, and NF hearts using quantitative PCR (qPCR). As shown in Fig. 3, qPCR confirmed 27 of the 32 microarray predictions with regard to fold change; 11 of these agreed completely in fold change and significance. Of the five that did not agree on fold change, three were nonsignificantly changed in both comparisons (the leptin receptor in ICM, serine proteinase inhibitor, clade E, member-1 in NICM, and the acyl-CoA synthetase long chain family member-3 in ICM), leaving only two clear disagreements: S100 calcium binding protein-A8 was significantly downregulated by qPCR but nonsignificantly upregulated by microarray, and lumican was significantly upregulated in ICM by microarray and nonsignificantly downregulated by qPCR. Notably, of the 10 genes significantly expressed only in one comparison, NICM or ICM, relative to NF hearts, 17 of the 20 comparisons were confirmed by fold change and/or significance, again confirming the specificity of the uniquely identified genes.

View larger version (19K):
[in this window]
[in a new window]
|
Fig. 3. Independent assessment of gene expression levels. To validate selected microarray findings using a complementary methodology, we quantified transcript abundance of 16 genes using quantitative PCR (qPCR). Fold change is expression in NICM and ICM hearts compared with NF hearts according to qPCR (black bars) and microarrays (gray bars). ACE2, angiotensin-converting enzyme-2; ATP1B3, ATPase, Na+/K+ transporting, beta 3 polypeptide; FACL3, acyl-CoA synthetase long chain family member-3; HBA2, hemoglobin A2; LEPR, leptin receptor; LUM, lumican; MYH6, myosin heavy chain 6; NAP1L3, nucleosome assembly protein 1-like-3; NPR3, atrionatriuretic peptide receptor C; PHLDA1, pleckstrin homology-like domain family A member-1; RPS4Y, ribosomal protein S4, Y-linked; S100A8, S100 calcium binding protein A8; SERPINE1, serine (or cysteine) proteinase inhibitor, clade E, member-1; SLC39A8, solute carrier family-39, member-8; TNFRSF11B, tumor necrosis factor receptor superfamily member-11b; TXNIP, thioredoxin interaction protein. *P < 0.05 compared with NF hearts by Wilcoxon rank sum test. P < 0.05 by significance analysis of microarrays.
|
|
 |
DISCUSSION
|
---|
The principal finding of this investigation is that cardiomyopathies of different etiologies exhibit both shared and distinct changes in gene expression compared with NF hearts. Remarkably, of the >22,000 transcripts present on the Affymetrix microarray platform, only a total of 288 genes (12%) are differentially expressed in NICM and ICM relative to NF hearts, and 41 of these genes are common to both comparisons with comparable fold changes. This suggests that there are both shared and distinct mechanisms that contribute to the development of heart failure of different etiologies, which supports the recent identification of a gene expression-based diagnostic biomarker that differentiates between ICM and NICM cardiomyopathy (33). In addition, a better understanding of these distinctions encourages ongoing efforts to develop cause-specific therapies specifically targeted at NICM and ICM (7).
These results complement our recent identification of a gene expression profile that differentiates between ICM and NICM cardiomyopathy (33). In that analysis, we used prediction analysis of microarrays (PAM) (46) to identify and validate a 90-gene profile that could differentiate between NICM and ICM. Unlike the current analysis, PAM identifies the smallest number of genes that succinctly characterizes a class. These genes do not necessarily have biologic significance, since they are chosen based on the stability of their expression rather than a combination of magnitude and stability (46). This prior study demonstrated that gene expression profiles correlated with clinical parameters in heart failure patients and supported ongoing efforts to incorporate expression profiling-based biomarkers in determining prognosis and response to therapy in heart failure.
The current study has a distinctly different purpose and uses different samples and statistical methods. Instead of identifying and validating a gene expression profile as a diagnostic biomarker, the current study focuses on novel gene discovery: identifying differentially expressed genes to better understand the similarities and differences between the two major forms of cardiomyopathy, ICM and NICM. In addition, because we were interested in the genesis of cardiomyopathy, we compared both ICM and NICM with NF hearts (the prior study did not involve NF hearts). Finally, in the current study, we used SAM (49) to identify differentially expressed genes and validated our findings with qPCR, as opposed to using PAM (46) and validating our findings by testing the gene expression prediction profile in an independent set of samples.
Thus the two studies target two different goals of microarray analysis, using a pattern of gene expression as a biomarker vs. examining gene expression for novel gene discovery (7, 15). These findings of the unique and shared genes expressed in NICM and ICM relative to NF hearts complement those of the prior study. Both demonstrate that unique gene expression exists in the two major forms of cardiomyopathy. On one hand, this allows a pattern of gene expression to function as a diagnostic biomarker. On the other hand, the unique patterns of gene expression can be further investigated to better define cause-specific therapies for heart failure. These two analyses are clearly not redundant, since they used different sets of samples and different statistical methods and, most importantly, had different purposes. Furthermore, given the complementary nature of the two analyses, it is not surprising that only four of the genes in the current study were observed in our prior identification of a gene expression profile that differentiated between ICM and NICM (33). The current analysis also focused on differential gene expression and thus targeted different genes than the one investigating prediction (46).
The current study is unique for a number of reasons. First, we have studied 37 samples, which is a large number relative to gene expression studies in cardiomyopathy to date (1, 3, 5, 10, 11, 25, 28, 44, 45, 51). There are no accepted means of calculating sample size and power in microarray experiments, but because our study examines a larger number of samples than prior studies, we have increased power to detect significant changes in gene expression. Furthermore, we have the added advantage of uniformity among samples: all NICM hearts were from individuals with idiopathic cardiomyopathy, and the clinical characteristics were reasonably similar within groups.
The second unique feature of this study is that we have not compared only failing and NF hearts, as in many previous studies (1, 5, 45, 51), but extended this analysis to compare the differential gene expression of NICM and ICM relative to NF hearts. This offers further insight into the mechanisms involved in the development of heart failure of varying etiologies. Many genes are shared between NICM and ICM relative to NF hearts, and this is consistent with clinical experience: the presentation and standard treatment for cardiomyopathy of both etiologies are similar (27). However, despite similar presentations and therapies, NICM and ICM are distinct diseases; patients with ICM have decreased survival compared with their NICM counterparts (21, 24) and respond differently to therapies (19, 23, 32, 42). Thus an understanding of the distinctions between the two conditions at the level of gene expression may guide future efforts to design etiology-based therapies.
The predominance of metabolism genes in NICM hearts suggests that the derangements involved in the genesis and maintenance of NICM may be metabolic in nature. This is supported by an early trial of beta-blockers in heart failure that demonstrated a greater mortality benefit in NICM than ICM (13). Beta-blockers improve myocardial efficiency by shifting myocardial metabolism from free fatty acids to glucose. The increase in fatty acid metabolism genes specifically in NICM in our analysis would explain why beta-blockers may be particularly beneficial in NICM. Furthermore, our results suggest that future etiology-specific therapies in NICM could target metabolic pathways, including those of fatty acid or cholesterol synthesis. One particularly relevant example is ranolazine. This investigational compound shifts myocardial cells from fatty acid to glucose metabolism and is currently being investigated as a treatment for myocardial ischemia (9). On the basis of our results, this drug could also be helpful in patients with NICM.
In ICM, on the other hand, our results suggest that abnormalities in catalytic activity may predominate, and an anti-ischemic protective effect of the specific catalytic enzymes identified, serine proteinase inhibitors, has been previously observed in pigs subject to experimentally induced myocardial ischemia (31). Given our results, it may be possible that such enzymes could also be beneficial in patients with ICM.
Our work agrees to an extent with the findings of a similar analysis of differential gene expression by Steenman et al. (44), in which pooled samples of NICM and ICM were compared with one NF sample, and 95 differentially expressed genes were identified between failing and NF hearts. Compared with our list of 288 genes, we found 8 genes in common (Supplementary Table S3). There are a number of reasons why our results differed from those of the prior study. The prior study had only one NF heart, and it was from a patient with cystic fibrosis. This heart is likely very different, not only in age but also in hemodynamic parameters, from a heart from an unused cardiac transplant donor. In addition, we used different statistical algorithms for normalization and identification of differentially expressed genes. We normalized with robust multiarray average, which has been shown to offer better detection of differentially expressed genes than Affymetrixs default preprocessing algorithm (29). We identified differentially expressed genes with SAM, which has been validated in a number of studies (6, 41, 49, 50) and may be more accurate than other commonly used methods for identifying differentially expressed genes, such as t-tests (43). In addition, our analysis may have more external validity because we studied more samples (37 vs. 7 patients) with individually hybridized, as opposed to pooled, data. Individual hybridization may be more accurate than pooling because it allows the estimation of the within-group variance for each gene (38).
Some of the genes shown to be differentially expressed in our study have been previously identified as differentially expressed in studies of NICM vs. NF hearts, with remarkably similar fold changes between studies (Supplementary Table S3). Commonly identified genes include those involved in the fetal gene program (14), including natriuretic peptide precursor B, atrial natriuretic factor, cardiac muscle myosin heavy chain, and atrial alkali myosin light chain. The majority of genes are upregulated in NICM and ICM hearts vs. NF hearts, and this has also been noted in prior studies (1, 5, 44, 45, 51). This is likely due to biological differences, since prior studies all used different methods to normalize data and identify differentially expressed genes. Furthermore, because the expression of many of these genes was confirmed with qPCR in these prior studies, this offers indirect further confirmation of the validity of our differentially expressed genes. This highlights the critical point in microarray analysis used for gene discovery: the results should be considered hypothesis generating, and the gene expression should be confirmed with other quantitative techniques, such as qPCR (15).
Through qPCR, we confirmed the expression of 27 of the 32 comparisons with 16 genes of interest in heart failure. Of greatest interest are the novel genes from our analysis, including ACE2 and a member of the tumor necrosis factor (TNF) receptor superfamily (TNFRSF11B, also known as osteoprotegerin). We show that, in subjects with end-stage cardiomyopathy, ACE2 is significantly upregulated in NICM but not ICM. ACE2 is expressed predominantly in vascular endothelial cells of the heart and kidney, and ACE and ACE2 have different biochemical activities. Angiotensin I is converted to angiotensin I-9 (with 9 amino acids) by ACE2 but is converted to angiotensin II, which has eight amino acids, by ACE. Whereas angiotensin II is a potent blood vessel constrictor, angiotensin I-9 has no known effect on blood vessels but can be converted by ACE to a shorter peptide, angiotensin I-7, which is a blood vessel dilator (4). Interestingly, loss of ACE2 was associated with upregulation of hypoxia-inducible genes, suggesting a role for ACE2 in mediating the response to cardiac ischemia, although ACE2 was not significantly changed in ICM cardiomyopathy in our study (17). The upregulation of ACE2 in NICM but not ICM cardiomyopathy cannot be ascribed to the increased prescription of ACE inhibitors in ICM cardiomyopathy subjects, because, unlike ACE, ACE2 is insensitive to inhibition by ACE inhibitors (48). Thus our results offer insight into a possible new etiology-specific therapeutic target in heart failure.
Another novel finding of interest is the significant downregulation of a member of the TNF receptor subfamily, TNFRSF11B, in both NICM and ICM. Levels of TNF have been shown to be upregulated in chronic heart failure (34), and increasing levels of TNF have been correlated with disease severity (40). However, in clinical trials, soluble TNF-
antagonists did not reduce mortality or heart failure hospitalizations (12, 37). One might speculate that this lack of benefit may relate somehow to the downregulation of the TNF receptor in chronic heart failure.
The results of the unsupervised hierarchical clustering algorithm suggest that hearts of patients with NICM who do not undergo LVAD implantation resemble NF hearts more than those of NICM patients who require an LVAD before cardiac transplantation. An examination of their baseline characteristics confirms this: NICM-LVAD patients are a sicker subset, with higher pulmonary capillary wedge pressure and increased need for intravenous inotropes, two known markers of poor prognosis in chronic heart failure patients (8, 16). Although there are documented changes in gene expression between hearts before and after LVAD support (3, 10, 11, 25), there is no evidence that differential gene expression exists between end-stage cardiomyopathy samples obtained before LVAD placement and at the time of cardiac transplantation in patients without LVADs or between patients with different clinical presentations. Because this result was obtained with an unsupervised clustering algorithm, it is free of bias of predefined categories (35). Although is it possible that the differences were due, in part, to the use of seven NICM-LVAD samples from an outside institution, this is less likely based on our prior results with these samples, which indicated that the institution of origin did not contribute to variability in gene expression (33), and because the outside institution samples themselves did not form a distinct cluster. This unanticipated difference between end-stage NICM patients could offer insight into the differential gene expression of different stages of heart failure. This requires further study, and lends credence to the notion that gene expression can be correlated with clinically relevant parameters in heart failure patients to aid in determining prognosis and response to therapy.
Although the analysis of gene expression using oligonucleotide microarrays is a powerful technique, limitations warrant mention. Not all genes are represented on the Affymetrix U133A arrays used in this study, and therefore the knowledge that can be acquired from these experiments remains incomplete. In addition, an NF unused donor heart is not the same as a normal heart, because circumstances causing a donor heart to be ineligible for cardiac transplantation, such as infection or prolonged hypotension, can also affect gene expression. One study suggested that the differential gene expression identified between failing and NF hearts may have been due to age and gender differences rather than differences in ventricular function (5). However, normal age- and sex-matched hearts are impossible to obtain, and other researchers have used comparable unused donor hearts in their experiments (1, 5, 45, 51).
Another limitation of this study is that microarray analysis is essentially hypothesis generating. However, in the tradition of such studies in the microarray literature (1, 3, 5, 10, 11, 25, 30, 44, 45, 51), this is a hypothesis-generating analysis with biological validation of select genes confirmed by qPCR. We have followed the practice of other studies in the field and extended the analysis to include more samples with different etiologies of heart failure and a careful comparison with the results of prior studies (in Supplementary Table S4), which is unprecedented in the literature thus far. For this reason, we believe that these analyses, while mainly hypothesis generating, do have significant value and should be made available to other individuals interested in microarray analysis of ICM and NICM.
In conclusion, we offer a novel addition to the analysis of differential gene expression between failing and NF hearts by providing new insight into the genetic pathways involved in the genesis of cardiomyopathy. By comparing differential gene expression in NICM and ICM relative to NF hearts, we have shown that there are a number of common and unique genes involved in the development of heart failure of differing etiologies. This analysis will provide valuable insight into the pathophysiology of heart failure and offers a basis for future studies of cause-specific therapies in the complex management of heart failure patients.
 |
GRANTS
|
---|
Research was supported by National Heart, Lung, and Blood Institute Grants 1U01-HL-066583 (J. G. N. Garcia) and 5RO1-HL-065455 (J. M. Hare). J. M. Hare is a recipient of the Paul Beeson Physician Faculty Scholar in Aging Research Award and is supported by the Donald W. Reynolds Cardiovascular Clinical Research Center.
 |
ACKNOWLEDGMENTS
|
---|
We are indebted to Maryann Albaugh, Thomas Brown, Tera Lavoie, Stacey Nash, and Kristy Swigert for valuable assistance in sample collection and microarray processing.
 |
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, Ross 1059, 720 Rutland Ave, Baltimore, MD 21205 (e-mail: jhare{at}mail.jhmi.edu).
doi:10.1152/physiolgenomics.00255.2004.
1 The Supplemental Material for this article (Supplemental Figs. S1S3, Supplemental Tables S1S3) is available online at http://physiolgenomics.physiology.org/cgi/content/full/00255.2004/DC1. 
 |
REFERENCES
|
---|
- Barrans JD, Allen PD, Stamatiou D, Dzau VJ, and Liew CC. Global gene expression profiling of end-stage dilated cardiomyopathy using a human cardiovascular-based cDNA microarray. Am J Pathol 160: 20352043, 2002.[Abstract/Free Full Text]
- BioConductor. Welcome to Bioconductor, http://www.bioconductor.org/ (accessed 30 June 2004).
- Blaxall BC, Tschannen-Moran BM, Milano CA, and Koch WJ. Differential gene expression and genomic patient stratification following left ventricular assist device support. J Am Coll Cardiol 41: 10961106, 2003.[CrossRef][ISI][Medline]
- Boehm M and Nabel EG. Angiotensin-converting enzyme 2a new cardiac regulator. N Engl J Med 347: 17951797, 2002.[Free Full Text]
- Boheler KR, Volkova M, Morrell C, Garg R, Zhu Y, Margulies K, Seymour AM, and Lakatta EG. Sex- and age-dependent human transcriptome variability: implications for chronic heart failure. Proc Natl Acad Sci USA 100: 27542759, 2003.[Abstract/Free Full Text]
- Bullinger L, Dohner K, Bair E, Frohling S, Schlenk RF, Tibshirani R, Dohner H, and Pollack JR. Use of gene-expression profiling to identify prognostic subclasses in adult acute myeloid leukemia. N Engl J Med 350: 16051616, 2004.[Abstract/Free Full Text]
- Butte A. The use and analysis of microarray data. Nat Rev Drug Discov 1: 951960, 2002.[CrossRef][ISI][Medline]
- Califf RM, Adams KF, McKenna WJ, Gheorghiade M, Uretsky BF, McNulty SE, Darius H, Schulman K, Zannad F, Handberg-Thurmond E, Harrell FE Jr, Wheeler W, Soler-Soler J, and Swedberg K. A randomized controlled trial of epoprostenol therapy for severe congestive heart failure: the Flolan International Randomized Survival Trial (FIRST). Am Heart J 134: 4454, 1997.[ISI][Medline]
- Chaitman BR, Pepine CJ, Parker JO, Skopal J, Chumakova G, Kuch J, Wang W, Skettino SL, and Wolff AA. Effects of ranolazine with atenolol, amlodipine, or diltiazem on exercise tolerance and angina frequency in patients with severe chronic angina: a randomized controlled trial. JAMA 291: 309316, 2004.[Abstract/Free Full Text]
- Chen MM, Ashley EA, Deng DX, Tsalenko A, Deng A, Tabibiazar R, Ben Dor A, Fenster B, Yang E, King JY, Fowler M, Robbins R, Johnson FL, Bruhn L, McDonagh T, Dargie H, Yakhini Z, Tsao PS, and Quertermous T. Novel role for the potent endogenous inotrope apelin in human cardiac dysfunction. Circulation 108: 14321439, 2003.[Abstract/Free Full Text]
- Chen Y, Park S, Li Y, Missov E, Hou M, Han X, Hall JL, Miller LW, and Bache RJ. Alterations of gene expression in failing myocardium following left ventricular assist device support. Physiol Genomics 14: 251260, 2003.[Abstract/Free Full Text]
- Chung ES, Packer M, Lo KH, Fasanmade AA, and Willerson JT. Randomized, double-blind, placebo-controlled, pilot trial of infliximab, a chimeric monoclonal antibody to tumor necrosis factor-alpha, in patients with moderate-to-severe heart failure: results of the anti-TNF Therapy Against Congestive Heart Failure (ATTACH) trial. Circulation 107: 31333140, 2003.[Abstract/Free Full Text]
- CIBIS-II Investigators and Committees. The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II): a randomised trial. Lancet 353: 913, 1999.[CrossRef][ISI][Medline]
- Colucci WS and Braunwald E. Pathophysiology of heart failure. In: Heart Disease: A Textbook of Cardiovascular Medicine, edited by Braunwald E, Zipes D, and Libby P. Philadelphia, PA: WB Saunders, 2004, p. 503533.
- Cook SA and Rosenzweig A. DNA microarrays: implications for cardiovascular medicine. Circ Res 91: 559564, 2002.[Abstract/Free Full Text]
- Costanzo-Nordin MR, OConnell JB, Engelmeier RS, Moran JF, and Scanlon PJ. Dilated cardiomyopathy: functional status, hemodynamics, arrhythmias, and prognosis. Cathet Cardiovasc Diagn 11: 445453, 1985.[ISI][Medline]
- Crackower MA, Sarao R, Oudit GY, Yagil C, Kozieradzki I, Scanga SE, Oliveira-dos-Santos AJ, da Costa J, Zhang L, Pei Y, Scholey J, Ferrario CM, Manoukian AS, Chappell MC, Backx PH, Yagil Y, and Penninger JM. Angiotensin-converting enzyme 2 is an essential regulator of heart function. Nature 417: 822828, 2002.[CrossRef][ISI][Medline]
- Day ML, Schwartz D, Wiegand RC, Stockman PT, Brunnert SR, Tolunay HE, Currie MG, Standaert DG, and Needleman P. Ventricular atriopeptin. Unmasking of messenger RNA and peptide synthesis by hypertrophy or dexamethasone. Hypertension 9: 485491, 1987.[Abstract]
- Doval HC, Nul DR, Grancelli HO, Perrone SV, Bortman GR, and Curiel R. Randomised trial of low-dose amiodarone in severe congestive heart failure. Grupo de Estudio de la Sobrevida en la Insuficiencia Cardiaca en Argentina (GESICA). Lancet 344: 493498, 1994.[CrossRef][ISI][Medline]
- Draghici S. Statistical intelligence: effective analysis of high-density microarray data. Drug Discovery Today 7: S55S63, 2002.[CrossRef][ISI][Medline]
- Dries DL, Sweitzer NK, Drazner MH, Stevenson LW, and Gersh BJ. Prognostic impact of diabetes mellitus in patients with heart failure according to the etiology of left ventricular systolic dysfunction. J Am Coll Cardiol 38: 421428, 2001.[CrossRef][ISI][Medline]
- Eisen MB, Spellman PT, Brown PO, and Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA 95: 1486314868, 1998.[Abstract/Free Full Text]
- Felker GM, Benza RL, Chandler AB, Leimberger JD, Cuffe MS, Califf RM, Gheorghiade M, and OConnor CM. Heart failure etiology and response to milrinone in decompensated heart failure: results from the OPTIME-CHF study. J Am Coll Cardiol 41: 9971003, 2003.[CrossRef][ISI][Medline]
- Felker GM, Thompson RE, Hare JM, Hruban RH, Clemetson DE, Howard DL, Baughman KL, and Kasper EK. Underlying causes and long-term survival in patients with initially unexplained cardiomyopathy. N Engl J Med 342: 10771084, 2000.[Abstract/Free Full Text]
- Hall JL, Grindle S, Han X, Fermin D, Park S, Chen Y, Bache RJ, Mariash A, Guan Z, Ormaza S, Thompson J, Graziano J, Sam Lazaro SE, Pan S, Simari RD, and Miller LW. Genomic profiling of the human heart before and after mechanical support with a ventricular assist device reveals alterations in vascular signaling networks. Physiol Genomics 17: 283291, 2004.[Abstract/Free Full Text]
- Hare JM, Walford GD, Hruban RH, Hutchins GM, Deckers JW, and Baughman KL. Ischemic cardiomyopathy: endomyocardial biopsy and ventriculographic evaluation of patients with congestive heart failure, dilated cardiomyopathy and coronary artery disease. J Am Coll Cardiol 20: 13181325, 1992.[ISI][Medline]
- Hunt SA, Baker DW, Chin MH, Cinquegrani MP, Feldmanmd AM, Francis GS, Ganiats TG, Goldstein S, Gregoratos G, Jessup ML, Noble RJ, Packer M, Silver MA, Stevenson LW, Gibbons RJ, Antman EM, Alpert JS, Faxon DP, Fuster V, Gregoratos G, Jacobs AK, Hiratzka LF, Russell RO, and Smith SC Jr. ACC/AHA Guidelines for the Evaluation and Management of Chronic Heart Failure in the Adult: Executive Summary A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Revise the 1995 Guidelines for the Evaluation and Management of Heart Failure): Developed in Collaboration With the International Society for Heart and Lung Transplantation; Endorsed by the Heart Failure Society of America. Circulation 104: 29963007, 2001.[Free Full Text]
- Hwang JJ, Allen PD, Tseng GC, Lam CW, Fananapazir L, Dzau VJ, and Liew CC. Microarray gene expression profiles in dilated and hypertrophic cardiomyopathic end-stage heart failure. Physiol Genomics 10: 3144, 2002.[Abstract/Free Full Text]
- Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, and Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res 31: e15, 2003.[Abstract/Free Full Text]
- Kaab S, Barth AS, Margerie D, Dugas M, Gebauer M, Zwermann L, Merk S, Pfeufer A, Steinmeyer K, Bleich M, Kreuzer E, Steinbeck G, and Nabauer M. Global gene expression in human myocardium-oligonucleotide microarray analysis of regional diversity and transcriptional regulation in heart failure. J Mol Med 82: 308316, 2004.[CrossRef][ISI][Medline]
- Khan TA, Bianchi C, Voisine P, Feng J, Baker J, Hart M, Takahashi M, Stahl G, and Sellke FW. Reduction of myocardial reperfusion injury by aprotinin after regional ischemia and cardioplegic arrest. J Thorac Cardiovasc Surg 128: 602608, 2004.[Abstract/Free Full Text]
- Kittleson M, Hurwitz S, Shah MR, Nohria A, Lewis E, Givertz M, Fang J, Jarcho J, Mudge G, and Stevenson LW. Development of circulatory-renal limitations to angiotensin-converting enzyme inhibitors identifies patients with severe heart failure and early mortality. J Am Coll Cardiol 41: 20292035, 2003.[CrossRef][ISI][Medline]
- Kittleson MM, Ye SQ, Irizarry RA, Minhas KM, Edness G, Conte JV, Parmigiani G, Miller LW, Chen Y, Hall JL, Garcia JGN, and Hare JM. Identification of a gene expression profile that differentiates between ischemic and nonischemic cardiomyopathy. Circulation 110: 34443451, 2004.[Abstract/Free Full Text]
- Levine B, Kalman J, Mayer L, Fillit HM, and Packer M. Elevated circulating levels of tumor necrosis factor in severe chronic heart failure. N Engl J Med 323: 236241, 1990.[Abstract]
- Liu ET and Karuturi KR. Microarrays and clinical investigations. N Engl J Med 350: 15951597, 2004.[Free Full Text]
- Lowes BD, Minobe W, Abraham WT, Rizeq MN, Bohlmeyer TJ, Quaife RA, Roden RL, Dutcher DL, Robertson AD, Voelkel NF, Badesch DB, Groves BM, Gilbert EM, and Bristow MR. Changes in gene expression in the intact human heart. Downregulation of alpha-myosin heavy chain in hypertrophied, failing ventricular myocardium. J Clin Invest 100: 23152324, 1997.[Abstract/Free Full Text]
- Mann DL, McMurray JJ, Packer M, Swedberg K, Borer JS, Colucci WS, Djian J, Drexler H, Feldman A, Kober L, Krum H, Liu P, Nieminen M, Tavazzi L, van Veldhuisen DJ, Waldenstrom A, Warren M, Westheim A, Zannad F, and Fleming T. Targeted anticytokine therapy in patients with chronic heart failure: results of the Randomized Etanercept Worldwide Evaluation (RENEWAL). Circulation 109: 15941602, 2004.[Abstract/Free Full Text]
- Peng X, Wood CL, Blalock EM, Chen KC, Landfield PW, and Stromberg AJ. Statistical implications of pooling RNA samples for microarray experiments. BMC Bioinformatics 4: 26, 2003.[CrossRef][Medline]
- Pokrovsky VM and Osadchiy OE. Regulatory peptides as modulators of vagal influence on cardiac rhythm. Can J Physiol Pharmacol 73: 12351245, 1995.[ISI][Medline]
- Rauchhaus M, Doehner W, Francis DP, Davos C, Kemp M, Liebenthal C, Niebauer J, Hooper J, Volk HD, Coats AJ, and Anker SD. Plasma cytokine parameters and mortality in patients with chronic heart failure. Circulation 102: 30603067, 2000.[Abstract/Free Full Text]
- Sarwal M, Chua MS, Kambham N, Hsieh SC, Satterwhite T, Masek M, and Salvatierra O Jr. Molecular heterogeneity in acute renal allograft rejection identified by DNA microarray profiling. N Engl J Med 349: 125138, 2003.[Abstract/Free Full Text]
- Singh SN, Fletcher RD, Fisher SG, Singh BN, Lewis HD, Deedwania PC, Massie BM, Colling C, and Lazzeri D. Amiodarone in patients with congestive heart failure and asymptomatic ventricular arrhythmia. Survival Trial of Antiarrhythmic Therapy in Congestive Heart Failure. N Engl J Med 333: 7782, 1995.[Abstract/Free Full Text]
- Singhal S, Kyvernitis CG, Johnson SW, Kaiser LR, Liebman MN, and Albelda SM. Microarray data simulator for improved selection of differentially expressed genes. Cancer Biol Ther 2: 383391, 2003.[ISI][Medline]
- Steenman M, Chen YW, Le Cunff M, Lamirault G, Varro A, Hoffman E, and Leger JJ. Transcriptomal analysis of failing and nonfailing human hearts. Physiol Genomics 12: 97112, 2003.[Abstract/Free Full Text]
- Tan FL, Moravec CS, Li J, Apperson-Hansen C, McCarthy PM, Young JB, and Bond M. The gene expression fingerprint of human heart failure. Proc Natl Acad Sci USA 99: 1138711392, 2002.[Abstract/Free Full Text]
- Tibshirani R, Hastie T, Narasimhan B, and Chu G. Diagnosis of multiple cancer types by shrunken centroids of gene expression. Proc Natl Acad Sci USA 99: 65676572, 2002.[Abstract/Free Full Text]
- Towbin JA and Bowles NE. Molecular genetics of left ventricular dysfunction. Curr Mol Med 1: 8190, 2001.[CrossRef][Medline]
- Turner AJ, Tipnis SR, Guy JL, Rice G, and Hooper NM. ACEH/ACE2 is a novel mammalian metallocarboxypeptidase and a homologue of angiotensin-converting enzyme insensitive to ACE inhibitors. Can J Physiol Pharmacol 80: 346353, 2002.[CrossRef][ISI][Medline]
- Tusher VG, Tibshirani R, and Chu G. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci USA 98: 51165121, 2001.[Abstract/Free Full Text]
- Valk PJ, Verhaak RG, Beijen MA, Erpelinck CA, Barjesteh van Waalwijk van Doorn-Khosrovani S, Boer JM, Beverloo HB, Moorhouse MJ, van der Spek PJ, Lowenberg B, and Delwel R. Prognostically useful gene-expression profiles in acute myeloid leukemia. N Engl J Med 350: 16171628, 2004.[Abstract/Free Full Text]
- Yung CK, Halperin VL, Tomaselli GF, and Winslow RL. Gene expression profiles in end-stage human idiopathic dilated cardiomyopathy: altered expression of apoptotic and cytoskeletal genes. Genomics 83: 281297, 2004.[CrossRef][ISI][Medline]