| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Cardiovascular Genome Unit, Department of Medicine
2 Department of Anesthesiology, Brigham and Women’s Hospital, Harvard Medical School, Boston 02115
3 Division of Cardiovascular Medicine, Department of Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan
4 Department of Biostatistics, Harvard School of Public Health, Boston, Massachusetts
5 Cardiology Branch, National Heart, Lung, and Blood Institute, Bethesda, Maryland
6 Department of Laboratory Medicine and Pathobiology, University of Toronto, Ontario, Canada
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B-crystallin, antagonizer of myc transcriptional activity, ß-dystrobrevin, calsequestrin, lipocortin, and lumican). Microarray technology provides us with a genomic approach to explore the genetic markers and molecular mechanisms leading to heart failure. cDNA microarray; normalization; real-time reverse transcription-polymerase chain reaction
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although the genetic defect is unknown in most cases of familial DCM and determined in only half of cases of familial HCM, it is clear that both diseases exhibit nonallelic and allelic genetic heterogeneity (6, 14, 16, 34). The HCM phenotype is caused in most patients by distinct mutations in one of several sarcomeric genes, whereas cytoskeletal mutations are more closely related to DCM (6, 21). However, recent evidence provides clues that the same cytoskeletal or sarcomeric gene mutation may be associated with either DCM or HCM in the same family, possibly operating through two different series of events that remodel the heart (38, 44). Furthermore, the link between genetic mutation and contractile dysfunction remains an enigma. Genomic technology enables us to look at the differential expression of tens of thousands of genes simultaneously, and to compare patterns of gene expression during disease development and progression (10). However, approaches to the analysis of the large database generated from microarray platforms are still evolving. In this study, we constructed a spotted cDNA microarray using clones from various cardiovascular cDNA libraries sequenced and annotated in our laboratory to test the hypothesis that DCM and HCM end-stage heart failure developed via different molecular pathways and that, therefore, the two diseases presented different gene expression profiles. We also proposed methods for slide normalization, which was slide-dependent and nonlinear, and for gene-specific expression levels assessment using a hierarchical model. The feasibility of our approach was validated by real-time reverse transcription-polymerase chain reaction (RT-PCR). Using cDNA microarray, we obtained preliminary molecular portraits of DCM- and HCM-related end-stage heart failure.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Building a cDNA library and construction of spotted cDNA microarray.
The cDNA library was constructed using various cardiovascular tissues, including HCM in end-stage heart failure, fetal hearts (8–12 wk), normal adult hearts and aorta, with the lambda ZAP Express vector system (Stratagene, La Jolla, CA) as previously described (19, 20). Expressed sequence tags (ESTs) were generated using a well-established PCR and cycle-sequencing-based approach. All ESTs were searched against the nonredundant GenBank/EMBL/DDBJ and dbEST databases using the BLAST algorithm on a Unix platform (Sun Microsystems). Assignment of putative identities for ESTs required an expected value of 10-10 or less and a minimum of 95% nucleotide identity (9). ESTs matching to known genes were classified into seven different functional groups described previously (20). Individual nonredundant cDNA clones were isolated into unique pools and amplified by PCR in 96-well microplates. PCR products with a final volume of 50 µl were precipitated with 5 µl of 3 M ammonium acetate and 125 µl of 95% ethanol at -20°C overnight. The plates were centrifuged at 4,000 rpm for 30 min at 4°C, and the supernatant was decanted. Following two washes with 50 µl of 70% ethanol, the resulting DNA pellets were air-dried and resuspended in 20 µl of 3x SSC (sodium chloride/ sodium citrate buffer). A total of 10,368 nonredundant PCR products (including 96 bacteria clones as negative controls) were spotted onto Corning CMT-GAPS amino-Silane-coated glass microarray slides (Corning, Corning, NY) using the model GMS 417 arrayer (Affymetrix, Santa Clara, CA) and postprocessed using succinic anhydride blocking according to the manufacturer’s manual (4). The unique clones include 2,496 (24.3%) known genes, 3,296 (32.1%) matched ESTs, and 4,480 (43.6%) novel genes. The possibility of contaminated cDNA clones was reduced by prescreening with agarose gel electrophoresis.
Making cDNA probes and hybridization.
Human heart failure samples were obtained from the left ventricular free wall of explanted hearts from three idiopathic DCM patients without identifiable etiologies or antecedent myocarditic episode and from two patients with HCM (disease caused by Arg719Gln ß-myosin heavy chain mutation in both) during cardiac transplantation. Normal adult heart tissues of three donors were obtained from the left ventricular free wall of hearts not used for cardiac transplantation and pooled as reference samples. Thirty-five micrograms of total RNA from pooled normal adult heart, DCM, or HCM samples was oligo-dT primed, and probe synthesis was performed in the presence of either Cy3-dUTP (pooled normal) or Cy5-dUTP (pooled DCM or HCM) (Amersham Biosciences, Piscataway, NJ). Briefly, RNA was prepared in 8 µl of DEPC water, to which 1 µl of oligo-dT primer (0.5 µg/µl, GIBCO-BRL) was added. The mixture was incubated at 70°C for 5 min, then ice-chilled immediately. Four microliters of 5x first-strand buffer, 2 µl of 10x low-T dNTP (5 mM dATP, dCTP, and dGTP, and 2 mM dTTP), 2 µl of Cy3- or Cy5-dUTP, 2 µl of 0.1 M DTT, and 1 µl of RNaseOUT RNase inhibitor (40 U/µl, Invitrogen) were added, mixed well, and heated to 65°C for 5 min. One microliter of Superscript II (200 U/µl, GIBCO-BRL) was then added and incubated for 30 min at 42°C, followed by the addition of another 1 µl of Superscript II for 40 min at 42°C. The reverse transcription reaction was terminated with 2.5 µl of 500 mM EDTA, and the mixture was heated to 65°C for 1 min. Then, 5 µl of 1 M NaOH was added for 10 min at 65°C to hydrolyze RNA, and the pH was neutralized with 12.5 µl of 1 M Tris buffer (pH 7.5). Following purification of the labeled probe by gel exclusion chromatography (ProbeQuant G-50, Amersham), two cDNA probes of interest were mixed, reduced to a volume of about 5 µl, and combined with 30 µl of hybridization solution [stock solution containing 100 µl of DIG EasyHyb hybridization solution (Roche), 5 µl of yeast tRNA (10 mg/ml), and 5 µl of salmon sperm DNA (10 mg/ml) as blocking agents]. The probe solution was then hybridized to the arrayed slide at 37°C overnight. The next day, slides were washed first with 1x SSC to remove the coverslip; next, the slides were given three successive washes of 0.1% SDS and 1x SSC at 50°C for 15 min each, followed by a rinse with 1x SSC at room temperature. The slides were dried by a 5-min spin in a conical tube at 500–1,000 rpm to remove excess fluid. Scanning of the slide was performed using the model GMS 418 scanner (Affymetrix) at 532 nm (Cy3) and 635 nm (Cy5). The DCM experiments were performed on four slides in replicate experiments performed separately (slide number N = 2 for each experiment); HCM experiments were done on three slides with one replicate experiment (N = 2) and another one separate single slide experiment. All diseased heart samples were labeled with Cy5, and normal heart samples were labeled with Cy3. Pooled normal adult heart samples were labeled with Cy3 or Cy5, and then hybridized on four slides [2 replicate experiments (N = 2 for each experiment)] that were used as calibration experiments. Replication of hybridization experiments will greatly reduce the misclassification rate (25, 27); thus we performed replicate experiments in our study to exclude the nonconsistent data.
Image acquisition and data processing.
Raw scanned images were processed using ScanAlyze 2.44 microarray image analysis software (Michael Eisen, Stanford University, CA, http://rana.lbl.gov/EisenSoftware.htm). Cy3 and Cy5 scans for each slide were superimposed onto each other, and values corresponding to the fluorescence intensity for each spot were obtained and exported to an Excel spreadsheet. Local background was subtracted from the fluorescence value of each spot to obtain a "net" value. To avoid false-positive results generated from poor quality spots with weaker fluorescence signals, we filtered the weak hybridization spots first to account for incomplete hybridization. The CH1GTB2 and CH2GTB2 background criteria, which mean the fraction of pixels with greater than 1.5 times the background intensity, were used with a cutoff values of 0.5 in both channels (Michael Eisen, http://rana.lbl.gov/EisenSoftware.htm). We also included one tray of bacterial clones as negative controls. Spots with net signal intensities on the Cy3 or Cy5 channels less than the median intensities of bacterial clones were also filtered out to exclude signals due to nonspecific binding.
Bioinformatics.
We used a rank-invariant method to identify nondifferentially expressed genes on each slide across various signal intensities (46). After selecting nondifferentially expressed genes and fitting them into a normalization curve (using the Lowess smoothing procedure in S-plus), we extrapolated the normalization curve to normalize genes with extremely high or low intensities. After obtaining the normalized data for each filtered spot, we use a hierarchical linear model to assess gene expression levels and incorporated the calibration experiments as prior knowledge of variance components in the analysis of comparative experiments. The 95% posterior interval for each gene was displayed as upper and lower quartiles. Genes with a score lower than 0.025 (which indicates 2.5% probability of logarithmic Cy5/Cy3 expression level below 0) were selected as upregulated (e.g., increased gene expression), whereas genes with score greater than 0.975 (which indicates 97.5% probability of logarithmic Cy5/Cy3 expression level below 0) were selected as downregulated (e.g., decreased gene expression). This model allows us to select differentially expressed genes from replicate experiments, eliminating those spots with high variation in their expression ratios across slides.
Real-time RT-PCR.
To confirm the expression patterns of upregulated or downregulated genes, we chose several genes for further analysis using quantitative real-time RT-PCR in a 96-well format. For each gene of interest, real-time RT-PCR was performed for pooled DCM (n = 3) or HCM (n = 2) heart RNA samples which were used in microarray experiments, and pooled normal adult heart RNA samples were used as reference. Triplicate aliquots of each pooled RNA sample were used in the same reactions. As an internal control, primers for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were designed and amplified in parallel with the genes of interest. One-step real-time RT-PCR was performed using 50 ng of total RNA per reaction. Primers were designed using the Primer3 program (http://www.genome.wi.mit.edu), verified for complementarity (http://www.basic.nwu.edu/biotools/oligocalc.html), and searched against the public database to confirm unique amplification products (http://www.ncbi.nlm.nih.gov). Primers were generally 20 base pairs long and were chosen to generate PCR products of 100 to 150 base pairs. The melting temperature range was between 59 and 61°C. The PCR products were checked by 2% agarose gel electrophoresis for each reaction. Water control and no reverse transcriptase control were also performed on all DCM, HCM, and normal heart samples to eliminate the possibility of significant genomic DNA contamination. The primer sequences are listed in Table 1. All reactions were carried in 50-µl volumes containing 25 µl of SYBR Green PCR Master Mix (Perkin-Elmer Applied Biosystems, Foster City, CA), 0.25 µl of Multiscribe reverse transcriptase (50 U/µl, Perkin-Elmer), 0.5 µl of RNaseOUT RNase inhibitor, 50 ng of sample RNA, and 10 pmol of each forward and reverse primer. Reactions in 96-well format were performed in the Perkin-Elmer ABI Prism 7700 sequence detection system. The cycling parameters were 30 min at 48°C (reverse transcription), heated to 95°C for 10 min, and followed by 40 cycles of PCR (15 s at 94°C and 1 min at 60°C). The threshold cycle value (CT) represents the cycle at which a statistically significant increase in the normalized reporter signal (Rn) above a chosen threshold can first be detected, according to the manufacturer’s manual. Threshold is defined as the average standard deviation of Rn for the early cycles, multiplied by an adjustable factor. To determine relative expression levels in each RNA population, a standard curve was plotted on the basis of expressions of GAPDH in serial dilutions of pooled normal adult heart RNA (200 ng, 100 ng, 50 ng, 25 ng, 12.5 ng, 6.25 ng, and 3.12 ng) and a no-template control. For all experimental samples, the relative RNA quantity of each sample was determined from the standard curve, divided by the corresponding amount of GAPDH control to achieve a normalized value. Fold differences were calculated by dividing the mean of DCM or HCM samples by the averaged amount of normalized mRNA generated in the normal adult samples, and standard error (SE) was calculated for each category. All reaction results for the same samples with a coefficient of variation greater than 10% were retested. Statistical significance was defined by P < 0.05 using Student’s t-test.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
B-crystallin, antagonizer of myc transcriptional activity, ß-dystrobrevin, calsequestrin, lipocortin, and lumican) were found differentially expressed in either DCM or HCM (Fig. 3B). One gene (copper/zinc superoxide dismutase) did not display differential expression on HCM microarray data, while increased expression (1.6-fold increase in HCM, P = 0.047) was demonstrated by real-time RT-PCR (Table 5).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Molecular portraits of DCM- and HCM-related end-stage heart failure.
In the genes with altered expression in DCM and HCM human heart failure samples, we demonstrated the commonly upregulated or downregulated and differentially expressed genes of DCM and HCM, despite similar clinical situations of end-stage heart failure, in our patients. In our study, the number of transcripts involved in cell/organism defense, especially the immune response subcategory, was more in the DCM than in the HCM samples, possibly reflecting the diverse etiology of DCM. In contrast, HCM is most commonly caused by sarcomeric gene mutations. Although myocardial remodeling characterizes both conditions to some extent, hypertrophic processes might be more prominent in HCM, as evidenced in this and previous studies by greater expression of ribosomal genes in HCM (19). DCM and HCM also presented different portraits of downregulated genes. As the percentage of known genes was derived from our cDNA microarray database, not dealing with the entire population of known genes, the difference in the functional classification subcategory between DCM and HCM should be further confirmed and interpreted with caution. Moreover, the expression levels, not only the numbers, of genes in a certain subcategory are more related to biological significance. Despite similar clinical features of end-stage heart failure, the gene defects leading to DCM or HCM and secondary consequences of those defects differ in these two diseases (31, 38, 39), and indeed this is what we found in this study. Our molecular portraits of DCM and HCM are the first example of predictive markers for these two diseases (17).
Commonly upregulated or downregulated genes in DCM and HCM.
Our results also show several genes commonly upregulated or downregulated in both DCM and HCM samples. Atrial natriuretic peptide and SERCA displayed increased and decreased expression levels consistently in all the slides, respectively, and therefore served as positive control in this study (45). Copper/zinc superoxide dismutase and heat shock protein 90 were upregulated, reflecting cardiac response to oxidative stress in end-stage heart failure (40). Increased elongation factor 2 expression levels contributed to the activation of protein synthesis in myocardial hypertrophy and decompensated heart failure (50), and dephosphorylation of elongation factor 2 resulted in its activation and subsequently increased protein synthesis (12). Elevated calcium-activated neutral protease activity was reported in an isoproterenol-induced cardiac hypertrophy rat model, but its role in cardiomyopathy-related heart failure has not yet been corroborated (1). Decreased elastin/collagen ratio was one of the causes of adverse extracellular matrix remodeling in heart failure (32). Decorin is an extracellular matrix proteoglycan and is a member of leucine-rich protein family. Decorin can neutralize the activity of transforming growth factor-ß, which increases in hypertrophic heart failure through collagen accumulation. Decreased protein levels of decorin were shown in a spontaneously hypertensive heart failure rat model (32). Our RT-PCR results demonstrated increased decorin mRNA expression in human end-stage heart failure, suggesting a translational regulation defect, species variation, or the difference between the disease models used. CD59 is a complement regulatory protein (33), and the implications of its increased expression in heart failure are still unknown. Decreased phosphofructokinase expression in both DCM and HCM reflected low activity of glycolysis in heart failure (2).
Differentially expressed genes in DCM and HCM.
Several genes were found to be differentially expressed in DCM and HCM. In DCM, expression levels of atrial myosin alkali light chain, calsequestrin, lipocortin, and lumican were increased; whereas thioredoxin reductase was decreased. In HCM, expression levels of B-crystallin and desmin increased, whereas mRNA levels of the antagonizer of myc transcriptional activity and ß-dystrobrevin decreased. Calsequestrin, a sarcoplasmic reticulum Ca2+ storage protein, was highly expressed in our DCM samples. Calsequestrin may play a role in the Ca2+ regulatory pathway of heart failure, as was indicated in one transgenic mouse model with overexpression of calsequestrin developing cardiac hypertrophy and heart failure (35). Reprogramming of gene expression in the failing myocardium is denoted by the change from -myosin heavy chain (MHC) to ß-MHC gene expression. Myosin light chain isoform changes were described in the mRNA and protein levels of failing myocardium (7, 36), but the implication of increased expression of atrial myosin alkali light chain in DCM is still not clear. Recently, increased lumican expression was demonstrated in ischemic and reperfused rat heart, suggesting its contribution to myocardial fibrosis and regulation of collagen fiber assembly in DCM (3). Lipocortin belongs to an annexin family of calcium-dependent phospholipid-binding proteins, and several annexins were elevated in the failing heart (5), although they differ in cytosolic phospholipase A2 activity (24). Lipocortin also has anti-inflammatory actions and has been shown to reduce myocardial ischemia-reperfusion injury through leukocyte recruitment (8). Thioredoxin reductase has antioxidant activity (18), but its role in DCM warrants further investigation.
B-crystallin and desmin were shown to be highly expressed in HCM samples in our study. Altered expression in their mRNA in HCM has been reported previously (19, 47). In a study on the genetic dissection of left ventricular noncompaction or Barth syndrome, Ichida et al. (22) reported a novel mutation of the gene for -dystrobrevin, a cytoskeletal protein, in a family with this disease. Whether the decreased ß-dystrobrevin in our HCM sample has genetic implications needs further study. Proto-oncogene expression might mediate the hypertrophic mechanism in heart failure (23). Enforced expression of myc proto-oncogene invokes a proliferation stimulus, and its activity might be suppressed by the Mad (antagonizer of myc transcriptional activity) family of proteins (53). It is intriguing that we showed decreased expression of Mad in HCM, partly contributing to the sustained increased myc proto-oncogene expression in HCM.
Pitfalls in two-color system cDNA microarray study.
Microarray experiments using two-color comparisons present potential pitfalls for data analysis. We do not measure gene expression level directly, but rather fluorescence intensity recorded by a scanner. Many factors influence the observed intensity levels including the following: differences of the amount of overall mRNA between two samples, concentration, brightness, dye labeling efficiency, exposure time, hybridization, washing stringencies, and scanning camera sensitivity (49). These factors may produce a multiplicative effect, creating a need for bias correction, or normalization, between the two color systems. We demonstrated that the normalization in two-color cDNA microarray experiments was slide-dependent and nonlinear, especially among those genes with lower fluorescence intensities. For selecting differentially expressed genes, many reports arbitrarily use 1.5-, 2.0-, or 3.0-fold changes as cutoff criteria, which may overlook differentially expressed genes with lesser, but statistically significant, fold changes (27).
Our real-time RT-PCR (41, 43, 52) results supported the hierarchical model we used to assess expression level and revealed that the accuracy rate of differentially expressed genes obtained from microarray data depended on fold change and the level of statistical significance. The chance of false-negative selection could not be assessed from this study, as we did not randomly pick up the nondifferentially expressed genes for RT-PCR confirmation. However, the expression of copper/zinc superoxide dismutase (which was increased in RT-PCR, but not on microarray analysis) in HCM samples may exemplify this caveat.
Study limitations.
In this study, we were limited by small DCM and HCM sample sizes, and the molecular mechanisms involved in cardiac hypertrophy and decompensated heart failure are complex. Furthermore, gene expression may be confounded by drug treatment in these end-stage heart failure patients. Therefore, we used pooled DCM, HCM, and adult normal heart samples for hybridization to minimize the effect of biological variation, and thus we tested the feasibility of our data mining algorithm. We also used the concept of confidence interval and statistical strength to minimize the possibility of false-positive results and to overcome the variability in expression levels across experiments (29, 49). Our study included only end-stage heart failure samples, and this was an obstacle to the observation of gene expression changes as heart failure developed.
In conclusion, an approach to select differentially expressed genes in cDNA microarray experiments was proposed and validated in this study. This offers the basis for advanced microarray data mining. Based on our results, we suggest that investigators could narrow down the gene lists for further study by choosing genes with more stringent levels of statistical significance and adding fold change criteria based on standard deviation of preselected reference genes (11). Several candidate genes with differential expression in DCM or HCM open the avenue for further diagnostic or therapeutic targets. The DNA microarray technology greatly facilitates the molecular characterization of DCM- and HCM-related heart failure on a genomic scale and provides us a preliminary molecular portrait of these two diseases in end-stage heart failure.
|
ACKNOWLEDGMENTS |
|---|
This work was supported in part by grants from the Heart and Stroke Foundation of Ontario, the Medical Research Council of Canada, the Canadian Genome Analysis and Technology Program (CGAT), and the National Institutes of Health (Grants 5RO1-HL-5851603, 5P5O-HL-5931603, and 5RO1-HL-6166102). J. J. Hwang was supported by the National Taiwan University Hospital, Taipei, Taiwan. C.-W. Lam was supported by the Croucher Foundation, Hong Kong and The Chinese University of Hong Kong. V. J. Dzau is the recipient of National Institutes of Health Merit Award 5R37-HL-3561016.
Part of the content was presented at the 5th Annual Scientific Meeting of the Heart Failure Society of America, Washington DC, September 9–12, 2001.
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: C. C. Liew, Cardiovascular Genome Unit, Dept. of Medicine, Brigham and Women’s Hospital, Harvard Medical School, 75 Francis St., Thorn 1326, Boston, MA 02115 (E-mail: cliew{at}rics.bwh.harvard.edu; URL, http://tcgu.bwh.harvard.edu).
10.1152/physiolgenomics.00122. 2001.
1 Supplementary Material to this article (APPENDIX Tables A and B) is available online at http://physiolgenomics.physiology.org/cgi/content/full/10/1/31/DC1. 
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
B-crystallin causes aberrant desmin and B-crystallin aggregation and cardiomyopathy in mice.
Circ Res89
:3
–5,2001
.This article has been cited by other articles:
|
|
 |
|
  T. Koyanagi, L. Y. Wong, K. Inagaki, O. V. Petrauskene, and D. Mochly-Rosen Alteration of gene expression during progression of hypertension-induced cardiac dysfunction in rats Am J Physiol Heart Circ Physiol, July 1, 2008; 295(1): H220 - H226. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Kuner, A. S. Barth, M. Ruschhaupt, A. Buness, L. Zwermann, E. Kreuzer, G. Steinbeck, A. Poustka, H. Sultmann, and M. Nabauer Genomic analysis reveals poor separation of human cardiomyopathies of ischemic and nonischemic etiologies Physiol Genomics, June 10, 2008; 34(1): 88 - 94. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. E. Haddad, L. J. Saunders, S. D. Crosby, M. Carles, F. del Monte, K. King, M. R. Bristow, F. G. Spinale, T. E. Macgillivray, M. J. Semigran, et al. Human cardiac-specific cDNA array for idiopathic dilated cardiomyopathy: sex-related differences Physiol Genomics, April 21, 2008; 33(2): 267 - 277. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. O. Gramolini, T. Kislinger, R. Alikhani-Koopaei, V. Fong, N. J. Thompson, R. Isserlin, P. Sharma, G. Y. Oudit, M. G. Trivieri, A. Fagan, et al. Comparative Proteomics Profiling of a Phospholamban Mutant Mouse Model of Dilated Cardiomyopathy Reveals Progressive Intracellular Stress Responses Mol. Cell. Proteomics, March 1, 2008; 7(3): 519 - 533. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Prante, H. Milting, A. Kassner, M. Farr, M. Ambrosius, S. Schon, D. G. Seidler, A. E. Banayosy, R. Korfer, J. Kuhn, et al. Transforming Growth Factor beta1-regulated Xylosyltransferase I Activity in Human Cardiac Fibroblasts and Its Impact for Myocardial Remodeling J. Biol. Chem., September 7, 2007; 282(36): 26441 - 26449. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Ashrafian and H. Watkins Reviews of Translational Medicine and Genomics in Cardiovascular Disease: New Disease Taxonomy and Therapeutic Implications: Cardiomyopathies: Therapeutics Based on Molecular Phenotype J. Am. Coll. Cardiol., March 27, 2007; 49(12): 1251 - 1264. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Ojaimi, K. Qanud, T. H. Hintze, and F. A. Recchia Altered expression of a limited number of genes contributes to cardiac decompensation during chronic ventricular tachypacing in dogs Physiol Genomics, March 14, 2007; 29(1): 76 - 83. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. S. Ginsburg, D. Seo, and C. Frazier Microarrays Coming of Age in Cardiovascular Medicine: Standards, Predictions, and Biology J. Am. Coll. Cardiol., October 17, 2006; 48(8): 1618 - 1620. [Full Text] [PDF]
|
|
|
|
|
|
A. S. Barth, R. Kuner, A. Buness, M. Ruschhaupt, S. Merk, L. Zwermann, S. Kaab, E. Kreuzer, G. Steinbeck, U. Mansmann, et al. Identification of a Common Gene Expression Signature in Dilated Cardiomyopathy Across Independent Microarray Studies J. Am. Coll. Cardiol., October 17, 2006; 48(8): 1610 - 1617. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Schiekofer, I. Shiojima, K. Sato, G. Galasso, Y. Oshima, and K. Walsh Microarray analysis of Akt1 activation in transgenic mouse hearts reveals transcript expression profiles associated with compensatory hypertrophy and failure Physiol Genomics, October 11, 2006; 27(2): 156 - 170. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Donahue, D. A. Marchuk, and H. A. Rockman Redefining Heart Failure: The Utility of Genomics J. Am. Coll. Cardiol., October 3, 2006; 48(7): 1289 - 1298. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 |
