Gene expression profiles in human cardiac cells subjected to hypoxia or expressing a hybrid form of HIF-1{alpha}

CANWEN JIANG, HSIENWIE LU, KAREN A. VINCENT, SRINIVAS SHANKARA, ADAM J. BELANGER, SENG H. CHENG, GEOFFREY Y. AKITA, RALPH A. KELLY, MARK A. GOLDBERG and RICHARD J. GREGORY

Genzyme Corporation, Framingham, Massachusetts 01701


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS AND MATERIALS
 RESULTS
 DISCUSSION
 REFERENCES
 
The cellular response to hypoxia depends on rapid posttranslational modifications of proteins as well as regulation of gene expression. We performed serial analysis of gene expression (SAGE) on human cardiac cells under normoxia, subjected to hypoxia, or infected with Ad2/HIF-1{alpha}/VP16 (an adenoviral vector expressing a stable hybrid form of hypoxia-inducible factor 1{alpha}) or Ad2/CMVEV (an empty vector). Of the 97,646 SAGE tags that were sequenced, 27% matched GenBank entries, while an additional 32% matched expressed sequence tags (ESTs) in UniGene. We analyzed 161 characterized genes or ESTs with a putative identification. Expression of 35, 11, and 46 genes was increased by hypoxia, infection with Ad2/EVCMV, or infection with Ad2/HIF-1{alpha}/VP16, respectively, compared with normoxia; conversely, 20, 11, 38 genes, respectively, were expressed at lower levels. Genes regulated by hypoxia were associated with transcription, biosynthesis, extracellular matrix formation, glycolysis, energy production, cell survival, and cell stress. Changes following infection with Ad2/HIF-1{alpha}/VP16 mimicked the hypoxic response to a certain extent. Infection with Ad2/CMVEV affected expression of genes that were associated with extracellular matrix formation and membrane trafficking. Differential expression of select genes was confirmed using TaqMan in additional human cardiac cells and rat neonatal ventricular myocytes. These data provide insight into gene expression underlying the diverse and complex cellular response to hypoxia, expression of HIF-1{alpha}/VP16, or adenoviral infection.

serial analysis of gene expression; adenoviral vector


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS AND MATERIALS
 RESULTS
 DISCUSSION
 REFERENCES
 
THE MAINTENANCE of an adequate supply of oxygen to aerobic organisms is vital for survival. Cellular hypoxia induces a wide range of profound physiological and pathophysiological responses in an organism at both the systemic and cellular levels, including angiogenesis, erythropoiesis, and glycolysis (13). The hypoxic response is dependent on rapid posttranslational modifications of proteins as well as regulation of gene expression. The paradigm for gene regulation in response to hypoxia involves oxygen sensing, followed by a series of signal transduction mechanisms that lead to induction or repression of transcription and changes in mRNA stability or the rate of translation (8). Hypoxia-inducible factor 1 (HIF-1), a heterodimeric basic helix-loop-helix-PAS (bHLH-PAS) domain transcription factor, plays a key role in transcriptional activation of hypoxia inducible genes (34). HIF-1 is composed of two constitutively expressed subunits, HIF-1{alpha} and HIF-1ß (aryl hydrocarbon nuclear translocator). Although HIF-1ß heterodimerizes with several other bHLH-PAS proteins, HIF-1{alpha} is unique to HIF-1. Although the precise mechanism is not understood, HIF-1 is controlled by cellular oxygen concentration such that the level of HIF-1{alpha} protein, via regulation of protein stability, and HIF-1 DNA binding activity both increase exponentially as oxygen concentrations decrease. We have generated an adenoviral vector (Ad2/HIF-1{alpha}/VP16) expressing a hybrid form of HIF-1{alpha} that is stable under normoxic conditions. The hybrid transcription factor is composed of the DNA-binding and dimerization domains from HIF-1{alpha} and the transactivation domain from herpes simplex virus VP16 protein (33).

Recently, progress has been made in understanding the molecular mechanisms responsible for altered gene expression in response to hypoxia and of transcriptional activation of hypoxia-responsive genes by HIF-1 (26, 28). To understand genome-wide gene regulation by hypoxia and HIF-1, we performed serial analysis of gene expression (SAGE) on human cardiac cells. SAGE is a tool for rapid and quantitative analysis of gene expression of a given cell population under specific conditions (19, 32). One of the defining characteristics of genome-scale expression profiling is that the examination of so many diverse genes opens a window on all the processes that actually occur and not merely the single process one intended to observe. In addition, we sought to determine whether viral proteins are expressed in cardiac cells following adenovirus-mediated gene transfer and, if so, what is the impact of the expressed viral proteins on cardiac cells.


    METHODS AND MATERIALS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS AND MATERIALS
 RESULTS
 DISCUSSION
 REFERENCES
 

Adenovirus vectors.
The adenovirus vector expressing HIF-1{alpha}/VP16 (Ad2/HIF-1{alpha}/VP16) was generated as described previously (2, 33). Ad2/CMVEV was constructed in a fashion similar to that of Ad2/HIF-1{alpha}/VP16 except that the vector lacked a transgene.

Cell culture and experimental protocols.
The investigation conforms to principles outlined in the Declaration of Helsinki (Cardiovasc Res 35: 2–3, 1997). Human cardiac cells from three fetuses, passaged twice, were obtained from Clonetics (San Diego, CA). Cells from one fetal heart were used to generate total RNA for SAGE analysis, whereas cells from two additional fetal hearts were used in the validation experiments. The cells were cultured in DMEM supplemented with 20% FBS and 1% chick embryo extract. Approximately 90% of the cells stained positive with antibodies against both smooth muscle actin and sarcomeric actin, confirming their origin as fetal cardiac cells (data not shown). A pilot study showed that subjecting the cells to hypoxia (1% O2) for 8 h induced vascular endothelial growth factor (VEGF) at the mRNA and protein level. Extending the hypoxic conditions to 24 h did not further increase VEGF expression.

The normoxic group was maintained under control conditions (21% O2, 5% CO2, 37°C). Two additional groups were infected with Ad2/HIF-1{alpha}/VP16 or Ad2/CMVEV at a multiplicity of infection of 200 for 6 h and then maintained for an additional 42 h. The hypoxic group was placed in a hypoxia chamber (1% O2, 5% CO2, 37°C) for 8 h. The cells from these four groups were harvested with RNAzol B (Tel-test) for isolation of total RNA.

Neonatal rat ventricular cardiomyocytes were isolated as described previously (20). The cells were seeded at a density of 2.0 x 105 cells/cm2 on rat collagen type I coated dishes or chamber slides and maintained in DMEM-F12 medium containing 10% FBS and ITS (a mixture of bovine insulin, human transferrin, and sodium selenite) supplement for 72 h. The cells grown on dishes were then subjected to experimental protocols identical to that of their counterparts from human fetal hearts, and total RNAs were harvested for TaqMan analysis. The parallel cell samples grown on chamber slides were stained with a mouse antibody against sarcomeric tropomyosin-{alpha} plus Alexa-fluor-568-labeled goat-anti-mouse secondary antibody.

SAGE.
SAGE was performed as described previously (32). The SAGE tags were recorded as four tag libraries and analyzed using the modified SAGE Software Package and SAGEmap (19, 32). The tags that were differentially expressed threefold or greater between each pair of libraries were selected for further analysis. The characterized genes and the ESTs with a putative identification whose expression in one of the experimental SAGE libraries (hypoxia, Ad2/HIF-1{alpha}/VP16, or Ad2/CMVEV) combined with the normoxia library was at least four SAGE tags were statistically evaluated. The method described by Audic and Claverie (3) was used to compare the difference of the proportions between two libraries for each tag of interest.

TaqMan analysis.
TaqMan 5' nuclease fluorogenic quantitative PCR was performed as described previously (23). mRNA was quantified using the appropriate standard curves and was then normalized to 18S rRNA or ß-actin.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS AND MATERIALS
 RESULTS
 DISCUSSION
 REFERENCES
 

Summary of SAGE analysis.
The number of sequenced tags is summarized in Table 1. These tags represented from 7,013 to 9,943 unique transcripts. Twenty-seven percent of the tags matched characterized GenBank entries while an additional 32% matched expressed sequence tag (EST) entries in UniGene. The remaining tags have not been described.


View this table:
[in this window]
[in a new window]
 
Table 1. Overall summary of SAGE analysis

 
We analyzed the abundance of select housekeeping genes in the four libraries. GAPDH, a known HIF-1 target gene, was elevated in the Ad2/HIF-1{alpha}/VP16 library as expected, whereas ß-actin was relatively unchanged. We also compared the cardiac libraries to several SAGE libraries generated from noncardiac cells. Each of the noncardiac libraries was sequenced to ~40,000 tags. The abundance of GAPDH and ß-actin was comparable in all but the Ad2/HIF-1{alpha}/VP16 library.

Statistical analysis of SAGE data.
We assumed that the number of copies of a specific mRNA per cell was a fraction or a proportion of the total number of mRNA molecules in that cell. The same proportion of specific tags should be present in the SAGE library of all sequenced tags. Thus we did not convert the number of tags detected in a given library to the number of transcripts per cell. Comparison of the different SAGE libraries showed that the majority of the genes in fetal cardiac cells were expressed at similar levels. However, 944, 760, and 1,505 tags in cardiac cells subjected to hypoxia, infected with Ad2/HIF-1{alpha}/VP16 or Ad2/ CMVEV were differentially expressed at least threefold compared with that of normoxia, respectively. Of the differentially expressed tags, 744 characterized genes or ESTs with a putative identification expressed at least four tags in one of the experimental SAGE libraries (hypoxia, Ad2/HIF-1{alpha}/VP16, or Ad2/CMVEV) combined with the normoxia library (Table 1). The method described by Audic and Claverie (3) was utilized to calculate the probability values for comparing the abundance of these transcripts in the hypoxia, Ad2/HIF-1{alpha}/VP16, or Ad2/CMVEV library to that of the normoxia library. The probability values for the 161 genes that exhibited the most differential expression are shown in Tables 2–7GoGoGoGoGo`Go. Of these 161 genes, 35, 11, and 46 were increased in hypoxia, Ad2/CMVEV, and Ad2/HIF-1{alpha}/VP16, respectively, compared with normoxia. Conversely, 20, 11, and 38 tags were expressed at lower levels in hypoxia, Ad2/CMVEV, and Ad2/HIF-1{alpha}/VP16, respectively, compared with normoxia. These differentially expressed genes were categorized into functional groups for further analysis (Tables 27).


View this table:
[in this window]
[in a new window]
 
Table 2. Upregulation of gene expression by hypoxia

 

View this table:
[in this window]
[in a new window]
 
Table 3. Downregulation of gene expression by hypoxia

 

View this table:
[in this window]
[in a new window]
 
Table 4. Upregulation of gene expression following infection with Ad2/CMVEV

 

View this table:
[in this window]
[in a new window]
 
Table 5. Downregulation of gene expression following infection with Ad2/CMVEV

 

View this table:
[in this window]
[in a new window]
 
Table 6. Upregulation of gene expression following infection with Ad2/HIF-1{alpha}/VP16

 

View this table:
[in this window]
[in a new window]
 
Table 7. Downregulation of gene expression following infection with Ad2/HIF-1{alpha}/VP16

 
Gene expression in response to hypoxia.
Several transcription factors including HIF-1, activation protein 1, and nuclear factor {kappa}B are known to mediate hypoxia-induced gene expression (26, 28). In this study, only c-fos was induced by hypoxia. We also observed induction of cytoplasmic phospholipase A2 (cPLA2) interaction protein. This gene was not previously associated with hypoxia. In addition, negative cofactor 2{alpha}, a transcription repressor, was downregulated.

Hypoxia induces upregulation of specific proteins despite a general downregulation of protein synthesis. We observed induction of several genes that were associated with protein synthesis. Increased de novo synthesis of specific proteins may help repair or replace damaged proteins (29). Several genes that were associated with protein turnover were also upregulated. The products of these genes, namely proteasome activator, SUMO-1-activating enzyme E1N subunit, and ARD-1 N-acetyltransferase, may accelerate the degradation of damaged proteins. It appears that there is a coordinated hypoxic response to repair damaged proteins by increasing protein synthesis and by accelerating the degradation of damaged proteins.

Hypoxia is known to induce cytokines and growth factors (21, 26). At the current depth of sequencing, only a modest increase of VEGF tags was detected, whereas TaqMan analysis revealed a fourfold increase in the SAGE samples. In addition, lysyl hydroxylase and metalloproteinase 14 were also upregulated by hypoxia. Collagen cross-linking by lysyl hydroxylase allows collagen fibril formation and stabilization. Metalloproteinase 14 is capable of digesting select extracellular matrix (ECM) components (16). Furthermore, cell division cycle protein 42, p27, and dynamitin were upregulated, suggesting a complex hypoxic effect on cell growth.

Apoptosis occurs in hypoxic cardiomyocytes in vitro and in ischemic myocardium in vivo (12, 31). The induction of apoptosis-associated proteins by hypoxia, especially the ratio of the anti-apoptotic Bcl-2 to the pro-apoptotic Bax, determines the extent of apoptosis (1). We did not observe upregulation of Bcl-2 and Bax. However, survivin and gp130 were induced (1, 17). These genes may be part of a stress-activated survival pathway. In independent experiments, no increase in apoptosis was detected using the terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) assay in fetal cardiac cells exposed to hypoxia (O2 < 1%) for 8 h (data not shown).

Gene expression following infection with Ad2/CMVEV or Ad2/HIF-1{alpha}/VP16.
Transcripts corresponding to the adenoviral genes E3A and E3B were detected in cardiac cells infected with Ad2/HIF-1{alpha}/VP16. By contrast, we only detected pIX (two tags) in cells infected with Ad2/CMVEV. Expression of the E3 genes following infection with Ad2/HIF-1{alpha}/VP16 but not Ad2/EVCMV was confirmed in a separate experiment using TaqMan analysis. Adenoviral genes have been shown to be expressed in vascular endothelial cells infected with a wild-type E4 but not E4-deleted adenoviral vector (27). However, it is unlikely that E4 contributed to the expression of viral genes, since both Ad2/HIF-1{alpha}/VP16 and Ad2/EVCMV were based on a backbone that retained the ORF6 of E4 only. Preliminary studies suggest that VP16 is partially responsible for this differential viral gene expression, although the precise mechanism remains to be determined (K. A. Vincent, unpublished data).

The binding of the adenoviral vector to the receptor alone is sufficient to induce endogenous genes in HeLa cells (27). By contrast, low-level expression of E4 proteins is implicated in induction of endogenous genes in vascular endothelial cells (6). In cardiac cells infected with adenoviral vectors, we did not detect the induction of the genes that were previously observed to be upregulated in vascular endothelial cells or HeLa cells. However, both Ad2/CMVEV and Ad2/ HIF-1{alpha}/VP16 downregulated damage-specific DNA binding protein 1 and E2-EPF, possibly reflecting a stress response. Since with the exception of pIX there was no viral gene expression detected following infection with Ad2/CMVEV, regulation of endogenous genes by Ad2/CMVEV and Ad2/HIF-1{alpha}/VP16 is likely attributable to the binding of the adenoviral vector to the cells rather than viral gene expression.

Despite the lack of expression of viral genes, infection with Ad2/CMVEV regulated genes in several categories. Negative cofactor 2{alpha}, which may suppress gene expression (18), was induced. Ad2/CMVEV also upregulated Smt3p and N-acetyltransferase. Conjugation with Smt3p and acetylation by N-acetyltransferase may be protein modifications that are critical to cell cycle regulation (22, 35).

Upregulation of endothelin-1 converting enzyme, Smt3p, N-acetyltransferase, and collagen appears to be a response to cell stress such as viral infection. Induction of lysophospholipase I and copine I and downregulation of interleukin-8 (IL-8) and damage-specific DNA binding protein 1 may reflect host defense mechanisms following viral infection. Ad2/HIF-1{alpha}/VP16 also downregulated Thy-1 cell surface antigen, small inducible cytokine A2, and the pentaxin-related protein, suggesting a similar response to viral infection. However, although hypoxia and Ad2/HIF-1{alpha}/VP16 caused a complex expression pattern of genes associated with cell growth, infection with Ad2/CMVEV downregulated several genes in this category, including cyclin D1.

As expected, Ad2/HIF-1{alpha}/VP16 upregulated several known HIF-1 inducible genes, including VEGF, Glut-1, Glut-3, LDH-A, aldolase, enolase, insulin-like growth factor binding protein 3 (IGFBP3), and other glycolytic enzymes. However, some differentially expressed genes were not previously known to be regulated by hypoxia or HIF-1. Plasminogen activator inhibitor 1 (PAI1) and endothelial differentiation factor 1 (EDF1) were induced by Ad2/HIF-1{alpha}/VP16. IGFBPs modulate the biological activity of IGF that may regulate VEGF-dependent neovascularization (30). Expression of PAI1 is correlated with endothelial cell migration and tube formation in vitro, and cancer cell invasion and angiogenesis in vivo (4). In addition, Ad2/HIF-1{alpha}/VP16 induced genes associated with ECM. Thus HIF-1{alpha}/VP16 may induce an angiogenic response involving growth factors, ECM proteins, proteinase regulators, and EDF1.

HIF-1 mediates hypoxia-induced apoptosis through increasing p53 and reducing Bcl-2 protein levels (10). Ad2/HIF-1{alpha}/VP16 induced Nip3 that impairs the anti-apoptotic function of Bcl-2 (11). However, gp130 was upregulated. Several genes that might directly or indirectly promote cell survival by various mechanisms were also upregulated by Ad2/HIF-1{alpha}/VP16. These genes included tumor necrosis factor-{alpha} (TNF-{alpha}) inducible protein and IGFBP3 (7, 25). IGFBPs, which sequester the IGF from the IGF receptor, may protect cells from a variety of apoptotic stimuli, although IGFBP3 may have an IGF-independent pro-apoptotic effect (8). An independent experiment to assess the net outcome of the transcriptional activation of these genes showed no increase in apoptosis, as measured using the TUNEL assay.

Independent confirmation of SAGE data.
VEGF is upregulated by either hypoxia or HIF-1{alpha}. No VEGF-A transcript was present in normoxia, whereas there were two and four transcripts in hypoxia and Ad2/HIF-1{alpha}/VP16, respectively. TaqMan detected a 4- and 102-fold increase by hypoxia and Ad2/HIF-1{alpha}/VP16, respectively. The reason underlying this discrepancy remains to be understood. To further validate select genes, independent tests were performed to measure the expression levels of an additional 15 genes using TaqMan in the total RNA samples used in the SAGE analysis (23). The expression profiles of 13 genes, as measured by SAGE and TaqMan, were very similar (Fig. 1A). The combined number of transcripts that each of these genes expressed under control (normoxia) and one of the experimental conditions (hypoxia or infection with Ad2/HIF-1{alpha}/VP16) was 4 to 27. The average abundance was 7.4 ± 1.4 transcripts (mean ± SE), similar to the average abundance of all the differentially expressed genes analyzed (8.2 ± 1.8). Of these 13 genes, IGFBP3, Nip3, Glut-1, lysyl hydroxylase, enolase 1{alpha}, p27, and aldolase A were known hypoxia-responsive genes. The fact that these known hypoxia-responsive genes were selected for confirmation might have caused an overestimate of the successful confirmation rate. The remaining genes were not previously shown to be associated with HIF-1 or hypoxia. These genes included myosin regulatory light chain (MRLC), prolyl 4-hydroxylase {alpha} II, rhodanese, pentaxin-related gene, EDF1, and gp130. However, induction of two genes, namely integrin-ß and NADH-ubiquinone oxidoreductase, was not confirmed. These two genes expressed four and three transcripts in the Ad2/HIF-1{alpha}/VP16 library, and one and zero tags in the normoxia library, respectively. They were relatively less abundant compared with the majority of the 13 genes that were confirmed by TaqMan. The TaqMan assay was optimized using the appropriate standard curves as described previously (23). mRNA levels of the gene of interest were normalized to that of 18S rRNA or ß-actin. In this study, levels of 18S rRNA were particularly stable under different experimental conditions. The results obtained using the TaqMan method were reproducible. Comparison of SAGE and TaqMan analysis in select genes suggests that the current cutoff is adequate to ensure differential expression. However, for genes that were not abundant in the SAGE libraries, the SAGE observations deserve further confirmation by independent tests.



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 1. Correlation between fold increase measured by SAGE and TaqMan in the same total RNA samples that were used for SAGE analysis (A), between fold increase in the total RNA samples for SAGE analysis and that generated from an additional two fetal hearts, as measured by TaqMan (B), or between the fold increase in the fetal human cardiac cells and neonatal rat ventricular myocytes, as measured by TaqMan (C). With use of a linear curve fit, the r values are 0.82, 0.75, and 0.95, respectively.

 
Because of the relatively large quantity of mRNA required from each sample, SAGE was only performed on one sample per condition. To partially validate the SAGE observations, cardiac cells from two additional fetuses were subjected to identical experimental conditions. Total RNA samples were then analyzed using TaqMan. Differential expression of eight genes, including IGFBP3, Nip3, VEGF, gp130, MRLC, rhodanese, EDF1, and aldolase A was confirmed (Fig. 1B). These results suggest that the observations obtained from this SAGE analysis are readily confirmed in cells from multiple hearts. Again, for genes that were not abundant in the SAGE libraries, one must be cautious in interpreting the results, because of the fact that cells from one heart only were used in the analysis.

Validation in neonatal rat ventricular myocytes.
Differential gene expression constitutes the molecular basis for the control and manifestation of cardiac growth and development at different stages. For example, the fetal heart consistently exhibits higher levels of expression of genes involved in transcription and translation (protein synthesis) and lower levels of expression of cell structure/motility genes than the adult heart. In addition, parallels are revealed between cardiovascular developmental and pathophysiological progress. Thus differential response to hypoxia and HIF-1 between fetal and adult cardiac cells is anticipated. This requires experimental validation of SAGE findings in adult human tissues. However, it is difficult to maintain viable adult human cardiac cells that can be subjected to hypoxia or infection with Ad2/HIF-1{alpha}/VP16. As an alternative strategy, neonatal rat ventricular myocytes were isolated and cultured as previously described (20). Three days after seeding at high density (5 x 105 cells/cm2), the majority of the neonatal rat ventricular myocytes were beating in normal medium and stained positively with antibodies against sarcomeric tropomyosin-{alpha} (data not shown). The cells were then subjected to hypoxia or infected with Ad2/HIF-1{alpha}/VP16 or Ad2/CMVEV. The total mRNA samples were harvested for TaqMan analysis. Six of seven genes and ESTs that were differentially expressed in fetal human cardiac cells by SAGE analysis were confirmed by TaqMan analysis in neonatal rat ventricular cells. These genes included Nip3, IGFBP3, lysyl OX, EDF1, MRLC, and VEGF. However, brain natriuretic peptide (BNP) was not induced in rat ventricular cells. These results suggest that the majority of the observations in fetal human cardiac cells were readily confirmed in a well-characterized ventricular cell model. It should be noted that selection of five known hypoxia-responsive genes for confirmation might have caused an overestimate of the successful confirmation rate.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS AND MATERIALS
 RESULTS
 DISCUSSION
 REFERENCES
 
When cardiomyocytes are exposed to hypoxia, there is an increased expression of growth factors, glucose transporters, glycolytic enzymes, vasomotor regulators, and stress proteins (26, 28). HIF-1 mediates transcriptional activation of hypoxia-responsive genes in several categories (28). Our data provide a global assessment of the hypoxic response at the molecular level and underscore the utility of large-scale, parallel gene expression analysis for the study of complex biologic responses. Since the genes analyzed represent 20–30% of the human genome, they also provide a necessary framework for interpreting the significance of individually expressed genes. In addition, some differentially expressed genes were not previously associated with hypoxia or HIF-1.

We attempted to assess differential gene regulation by hypoxia and expression of HIF-1{alpha}/VP16. However, two factors compromised the interpretation of the data. First, the hypoxic condition (length, severity, etc.) and the multiplicity of infection of Ad2/HIF-1{alpha}/VP16 that would yield comparable levels of HIF-1{alpha} mRNA and protein were not optimally matched. Instead, the hypoxic condition was chosen, as it resulted in adequate expression of known hypoxia-inducible genes. Second, infection with adenovirus and expression of viral genes might have impacted gene expression in the host cells. Cells infected with Ad2/CMVEV were used as a control for the viral infection. However, substantial differences in expression of viral transcripts between the cells infected with Ad2/HIF-1{alpha}/VP16 and Ad2/CMVEV compromised the value of the control and complicated the analysis. The contribution of viral infection to differential gene expression in cells subjected to hypoxia or infected with Ad2/HIF-1{alpha}/VP16 could not be excluded using the Ad2/CMVEV control. There were noticeable differences in altered gene expression induced by hypoxia and infection with Ad2/HIF-1{alpha}/VP16. For example, hypoxia upregulated several transcription factors or regulators. Some of these transcription-related genes were not previously shown to mediate gene expression in response to hypoxia. Upregulation of these genes might lead to secondary activation of other genes. By contrast, infection with Ad2/HIF-1{alpha}/VP16 induced only activating transcription factor 4 (ATF4). These results suggest that hypoxia is capable of inducing a broader response compared with HIF-1{alpha}/VP16. Second, Ad2/HIF-1{alpha}/VP16 dramatically induced BNP, whereas hypoxia only moderately increased BNP expression. Hypoxia is known to increase BNP release from isolated hearts, possibly via increased stability of the BNP mRNA (15). It is unlikely that viral infection per se contributed to this differential regulation, since Ad2/CMVEV did not induce BNP. This result may be attributable to different levels of HIF-1 protein in cells subjected to hypoxia and infected with Ad2/HIF-1{alpha}/VP16. It is also possible that induction of BNP by hypoxia is rapidly downregulated, whereas overexpression of HIF-1{alpha}/VP16 may overcome the downregulation. The measurement of BNP mRNA levels at 8 h missed the peak of the hypoxic induction. Alternatively, hypoxia triggered an inhibitory mechanism that attenuated the transcriptional activation of BNP by HIF-1. It remains to be determined whether the activation domain of VP16 confers much stronger activity to HIF-1 than the endogenous activation domain of HIF-1{alpha}.

Gradual occlusion of atherosclerotic coronary arteries is frequently associated with development of collateral circulation. However, the collateral development is rarely adequate to compensate fully for the flow lost to occlusion of native coronary arteries. In several animal models, administration of a single growth factor promotes myocardial angiogenesis. Safety and efficacy of therapeutic angiogenesis have also been evaluated in patients with ischemic heart disease (14). We reasoned that administration of HIF-1{alpha}, the regulated subunit of a transcription factor (HIF-1) upstream of angiogenic factors, might result in a coordinated angiogenic response. In a rabbit model of hind limb ischemia, administration of a plasmid encoding HIF-1{alpha}/VP16 increases angiogenesis and blood supply (33). The SAGE data suggest that HIF-1{alpha}/VP16 induced growth factors, matrix proteins, proteinases, and EDF1. Coordinated transcriptional activation of these genes may provide an environment for migration, adhesion, proliferation, and differentiation of endothelial cells, which are essential steps of angiogenesis.

Adenoviral vectors are being widely investigated for use in gene therapy for heart disease. We analyzed gene expression in human cardiac cells following infection with Ad2/HIF-1{alpha}/VP16 or Ad2/CMVEV. Both vectors utilized the same adenoviral backbone (deletion of E1, wild-type E2 and E3, deletion of E4 except ORF6). Only expression of pIX (two tags) was detected in cells infected with Ad2/CMVEV, whereas adenoviral E3A and E3B transcripts were observed in cells infected with Ad2/HIF-1{alpha}/VP16. However, despite little detectable expression of viral protein, infection with Ad2/CMVEV regulated expression of host genes that were associated with transcription, ECM formation, membrane trafficking, and cell growth and mitosis. The consequence of upregulation of these genes remains to be determined.

The fact that SAGE was performed on one sample only per condition prevented advanced statistical analysis. Thus the scope of differential expression might have been overestimated or underestimated, as the cutoff might have been arbitrary. For example, there were only two VEGF tags in the hypoxia SAGE library, whereas TaqMan analysis detected a fourfold induction of VEGF by hypoxia. It remains to be understood why SAGE did not detect the fourfold increase in VEGF expression induced by hypoxia. However, validation of differential expression of 13 out of 15 genes examined by TaqMan analysis suggests that the current scope of differential expression is adequate. Furthermore, we confirmed differential expression of eight genes in cells from an additional two hearts by TaqMan analysis. This independent confirmation validated the observations of differential gene expression in multiple tissue samples.

To generate the relatively large quantity of mRNA required for SAGE, passaged fetal cardiomyocytes were used. Approximately 90% of the cells stained positive with antibodies against both smooth muscle actin and sarcomeric actin, indicating their origin as fetal cardiomyocytes. However, it is difficult to precisely define to what extent the passaged cells resembled the phenotype of the native cardiomyocytes. Furthermore, differential gene expression constitutes the molecular basis for the control and manifestation of cardiac growth and development at different stages. To address these concerns, we deployed a well-characterized rat ventricular cell model to partially validate the observations in fetal human cardiac cells. These cells were beating and stained positively with tropomyosin-{alpha}. Indeed, six out of seven genes were confirmed.

This first genome-wide expression profiling will provide insight into the patterns of gene expression following hypoxia, expression of HIF-1{alpha}/VP16, or infection with adenoviral vectors in human cardiac cells. For example, preconditioning (brief episodes of hypoxia or ischemia in vivo) is known to protect cardiac cells from subsequent lethal hypoxia insults. Remarkable progress has been made in the last few years in our understanding of the cellular and molecular mechanisms underlying this cardioprotective adaptation (5). In this study, we reported the first time that protective genes such as gp130 and survivin were upregulated by hypoxia. We also observed that the pentaxin-related gene that was associated with innate immunity was downregulated by hypoxia or viral infection. It will be interesting to determine whether there is any relationship between innate immunity and cell stress induced by nonpathogen insults such as hypoxia. These findings will facilitate further studies to understand the cellular response to hypoxia and preconditioning.


    ACKNOWLEDGMENTS
 
We thank Yuxia Luo, Steve Madden, Clarence Wang, and the Virus Production Facilities at Genzyme Corporation for support.


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

Address for reprint requests and other correspondence: C. Jiang, Genzyme Corporation, 31 New York Ave., Framingham, MA 01701-9322 (E-mail: canwen.jiang{at}genzyme.com).

10.1152/physiolgenomics.00058.2001.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 METHODS AND MATERIALS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Ambrosini G, Adida C, and Altieri DC. A novel anti-apoptosis gene, survivin, expressed in cancer and lymphoma. Nat Med 3: 917–921, 1997.[ISI][Medline]
  2. Armentano D, Sookdeo CC, Hehir KM, Gregory RJ, St George JA, Prince GA, Wadsworth SC, and Smith AE. Characterization of an adenovirus gene transfer vector containing an E4 deletion. Hum Gene Ther 6: 1343–1353, 1995.[ISI][Medline]
  3. Audic S and Claverie JM. The significance of digital gene expression profiles. Genome Res 7: 986–995, 1997.[Abstract/Free Full Text]
  4. Bajou K, Noel A, Gerard RD, Masson V, Brunner N, Holst-Hansen C, Skobe M, Fusenig NE, Carmeliet P, Collen D, and Foidart JM. Absence of host plasminogen activator inhibitor 1 prevents cancer invasion and vascularization. Nat Med 4: 923–928, 1998.[ISI][Medline]
  5. Bolli R. Later phase of preconditioning. Circ Res 87: 972–983, 2000.[Abstract/Free Full Text]
  6. Bruder JT and Kovesdi I. Adenovirus infection stimulates the Raf/MAPK signaling pathway and induces interleukin-8 expression. J Virol 71: 398–404, 1997.[Abstract]
  7. Buerke M, Murohara T, Skurk C, Nuss C, Tomaselli K, and Lefer AM. Cardioprotective effect of insulin-like growth factor I in myocardial ischemia followed by reperfusion. Proc Natl Acad Sci USA 92: 8031–8035, 1995.[Abstract]
  8. Bunn HF and Poyton RO. Oxygen sensing and molecular adaptation to hypoxia. Physiol Rev 76: 839–885, 1996.[Abstract/Free Full Text]
  9. Butt AJ, Firth SM, and Baxter RC. The IGF axis and programmed cell death. Immunol Cell Biol 77: 256–262, 1999.[ISI][Medline]
  10. Carmeliet P, Dor Y, Herbert JM, Fukumura D, Brusselmans K, Dewerchin M, Neeman M, Bono F, Abramovitch R, Maxwell P, Koch CJ, Ratcliffe P, Moons L, Jain RK, Collen D, Keshert E, and Keshet E. Role of HIF-1{alpha} in hypoxia-mediated apoptosis, cell proliferation and tumour angiogenesis. Nature 94: 485–490, 1998.
  11. Chen G, Ray R, Dubik D, Shi L, Cizeau J, Bleackley RC, Saxena S, Gietz RD, and Greenberg AH. The E1B 19K/Bcl-2-binding protein Nip3 is a dimeric mitochondrial protein that activates apoptosis. J Exp Med 186: 1975–1983, 1997.[Abstract/Free Full Text]
  12. Fliss H and Gattinger D. Apoptosis in ischemic and reperfused rat myocardium. Circ Res 79: 949–956, 1996.[Abstract/Free Full Text]
  13. Guillemin K and Krasnow MA. The hypoxic response: huffing and HIFing. Cell 89: 9–12, 1997.[ISI][Medline]
  14. Hamawy AH, Lee LY, Crystal RG, and Rosengart TK. Cardiac angiogenesis and gene therapy: a strategy for myocardial revascularization. Curr Opin Cardiol 14: 515–522, 1999.[ISI][Medline]
  15. Hanford DS and Glembotski CC. Stabilization of the B-type natriuretic peptide mRNA in cardiac myocytes by {alpha}-adrenergic receptor activation: potential roles for protein kinase C and mitogen-activated protein kinase. Mol Endocrinol 10: 1719–1727, 1996.[Abstract]
  16. Hiraoka N, Allen E, Apel IJ, Gyetko MR, and Weiss SJ. Matrix metalloproteinases regulate neovascularization by acting as pericellular fibrinolysins. Cell 95: 365–377, 1998.[ISI][Medline]
  17. Hirota H, Chen J, Betz UA, Rajewsky K, Gu Y, Ross J, Muller W, and Chien KR. Loss of a gp130 cardiac muscle cell survival pathway is a critical event in the onset of heart failure during biomechanical stress. Cell 97: 189–198, 1999.[ISI][Medline]
  18. Kim S, Na JG, Hampsey M, and Reinberg D. The Dr1/DRAP1 heterodimer is a global repressor of transcription in vivo. Proc Natl Acad Sci USA 94: 820–825, 1997.[Abstract/Free Full Text]
  19. Lash AE, Tolstoshev CM, Wagner L, Schuler GD, Strausberg RL, Riggins GJ, and Altschul SF. SAGEmap: a public gene expression resource. Genome Res 10: 1051–1060, 2000.[Abstract/Free Full Text]
  20. Lee HR, Henderson SA, Reynolds R, Dunnmon P, Yuan D, and Chien KR. Alpha 1-adrenergic stimulation of cardiac gene transcription in neonatal rat myocardial cells: effects on myosin light chain-2 gene expression. J Biol Chem 263: 7352–7358, 1988.[Abstract/Free Full Text]
  21. Levy AP, Levy NS, Loscalzo J, Calderone A, Takahashi N, Yeo KT, Koren G, Colucci WS, and Goldberg MA. Regulation of VEGF in cardiac myocytes. Circ Res 76: 758–766, 1995.[Abstract/Free Full Text]
  22. Li SJ and Hochstrasser M. A new protease required for cell-cycle progression in yeast. Nature 398: 246–251, 1999.[ISI][Medline]
  23. Livak KJ, Flood SJ, Marmaro J, Giusti W, and Deetz K. Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. PCR Methods Appl 4: 357–362, 1995.[ISI][Medline]
  24. Oltvai ZN, Milliman CL, and Korsmeyer SJ. Bcl-2 heterodimerizes in vivo with a conserved homolog, Bax, that accelerates programmed cell death. Cell 74: 609–619, 1993.[ISI][Medline]
  25. Opipari AW Jr, Hu HM, Yabkowitz R, and Dixit VM. The A20 zinc finger protein protects cells from tumor necrosis factor cytotoxicity. J Biol Chem 267: 12424–12427, 1992.[Abstract/Free Full Text]
  26. Piacentini L and Karliner JS. Altered gene expression during hypoxia and reoxygenation of the heart. Pharmacol Ther 83: 21–37, 1999.[ISI][Medline]
  27. Ramalingam R, Rafii S, Worgall S, Hackett NR, and Crystal RG. Induction of endogenous genes following infection of human endothelial cells with an E1(-)/E4(+) adenovirus gene transfer vector. J Virol 73: 10183–10190, 1999.[Abstract/Free Full Text]
  28. Semenza GL. Regulation of mammalian O2 homeostasis by hypoxia-inducible factor 1. Annu Rev Cell Dev Biol 15: 551–578, 1999.[ISI][Medline]
  29. Sharma HS, Snoeckx LH, Sassen LM, Knoll R, Andres J, Verdouw PD, and Schaper W. Expression and immunohistochemical localization of Hsp-70 in preconditioned porcine myocardium. Ann NY Acad Sci 723: 491–494, 1994.[ISI][Medline]
  30. Smith LE, Shen W, Perruzzi C, Soker S, Kinose F, Xu X, Robinson G, Driver S, Bischoff J, Zhang B, Schaeffer JM, and Senger DR. Regulation of VEGF-dependent retinal neovascularization by insulin-like growth factor-1 receptor. Nat Med 5: 1390–1395, 1999.[ISI][Medline]
  31. Tanaka M, Ito H, Adachi S, Akimoto H, Nishikawa T, Kasajima T, Marumo F, and Hiroe M. Hypoxia induces apoptosis with enhanced expression of Fas antigen mRNA in cultured neonatal rat cardiomyocytes. Circ Res 75: 426–433, 1994.[Abstract]
  32. Velculescu VE, Zhang L, Vogelstein B, and Kinzler KW. Serial analysis of gene expression. Science 270: 484–487, 1995.[Abstract]
  33. Vincent KA, Shyu K, Luo Y, Magner M, Tio RA, Jiang C, Goldberg MA, Akita GY, Gregory RJ, and Isner JM. Angiogenesis is induced in a rabbit model of hindlimb ischemia by naked DNA encoding a HIF-1{alpha}/VP16 hybrid transcriptional factor. Circulation 102: 2255–2261, 2000.[Abstract/Free Full Text]
  34. Wang GL and Semenza GL. Purification and characterization of hypoxia-inducible factor 1. J Biol Chem 270: 1230–1237, 1995.[Abstract/Free Full Text]
  35. Whiteway M and Szostak JW. The ARD1 gene of yeast functions in the switch between the mitotic cell cycle and alternative developmental pathways. Cell 43: 483–492, 1985.[ISI][Medline]