EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m
w w w. e l s e v i e r. c o m / l o c a t e / y e x c r
Research Article
Hypoxia regulates inflammatory gene expression in endothelial cells Lionel Flamant, Sébastien Toffoli, Martine Raes, Carine Michiels ⁎ Laboratory of Biochemistry and Cellular Biology (URBC), FUNDP-University of Namur, 61 rue de Bruxelles, 5000 Namur, Belgium
A R T I C L E I N F O R M AT I O N
A BST RAC T
Article Chronology:
Hypoxia can activate the endothelium toward a pro-inflammatory phenotype and enhance
Received 30 May 2008
leukocyte adhesion. This process is involved in pathological conditions such as vascular
Revised version received
remodeling or ischemia–reperfusion injury. This study was aimed to obtain a global picture of
29 October 2008
the response of the endothelial cells to hypoxia with respect to inflammatory genes. To this
Accepted 23 November 2008
purpose, we used a low density DNA microarray specifically designed to quantitate the expression
Available online 7 December 2008
of genes involved in the inflammatory pathways and a customized real-time PCR array. The expression of several pro-inflammatory genes known to be NF-kB target genes was decreased after
Keywords:
the incubation of endothelial cells under hypoxia. In parallel, a decrease in the DNA binding activity
Endothelial cell
of this transcription factor was observed. On the other hand, HIF-1 DNA binding activity was
Inflammatory genes
increased as well as the expression of several genes known to be regulated by this factor. Among
Leukocyte adhesion
them are several pro-inflammatory genes whose overexpression could account for the increase in
Hypoxia
leukocyte adhesion to the hypoxic endothelial cells. We concluded that hypoxia does not shift the
NF-kB
endothelial cell phenotype to a more pro-inflammatory one probably because of a decrease in the
HIF-1
expression of several cytokines. On the other hand, a clear response to hypoxia was observed with HIF-1 probably playing an important role in this process. © 2008 Elsevier Inc. All rights reserved.
Introduction Hypoxia is a characteristic feature of many human pathologies: it drives angiogenesis in tumors, takes part in ischemic disorders and participates in several inflammatory syndromes including atherosclerosis and chronic inflammation. One major cell type involved in these pathologies is the endothelial cell: being at the interface between blood and tissue, it is at the first line to sense hypoxia. Moreover, since the endothelium regulates leukocyte, platelets and smooth muscle cell functions [1], changes in the endothelial cell phenotype can result in vasoconstriction, platelet activation and aggregation and leukocyte recruitment, adhesion and activation. Leukocytic invasion into ischemic tissues is now well documented [2]. Studies that aimed to understand the mechanisms through
⁎ Corresponding author. Fax: +32 81 72 41 35. E-mail address:
[email protected] (C. Michiels). 0014-4827/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.yexcr.2008.11.020
which the hypoxia-induced recruitment of leukocytes in vivo might occur show that endothelial cells are important players notably via the secretion of several cytokines. Short term hypoxia can directly activate endothelial cells toward a pro-inflammatory phenotype and enhance leukocyte adhesion to these activated endothelial cells (for a review, [3]). This activation mainly occurs through the synthesis of lipid mediators. Long term response is mediated by regulating gene expression. The major regulator of transcription under hypoxia is HIF-1 (hypoxiainducible factor-1). This factor is composed of two subunits: HIF1β/ARNT which is constitutively expressed and HIF-1α whose protein level is regulated by oxygen tension. Under normoxia, the protein is hydroxylated on two prolines and targeted for proteasomal degradation. When oxygen is no longer present, the
734
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
Table 1 – Effects of increasing incubation time under hypoxia on the expression of inflammatory genes
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
735
Table 1 (continued)
EAhy926 cells were incubated during 4, 8, 16 or 24 h under normoxia or hypoxia and total RNA was extracted and retrotranscribed before being hybridized onto the Inflammation Dual Chips. Results are presented in induction fold (hypoxia vs normoxia) for two independent experiments. Results highlighted in yellow are quantitative and statistically significant. Results highlighted in green are qualitative and statistically significant. Results highlighted in purple are the positive controls for hypoxia.
hydroxylation no longer occurs, HIF-1α is stabilized, translocates into the nucleus and dimerizes with ARNT to form the transcriptionally active HIF-1 [4], HIF-1 target genes are aimed to adapt cells to the low oxygen tension conditions: they include genes encoding erythropoietin, VEGF and glucose metabolism regulating proteins [4]. Among them are also some pro-inflammatory genes like IL-8 [5] and COX-2 [6]. Studies on the overall transcriptomic response of endothelial cells to hypoxia have already been published [7–9]. However, they did not focus on the inflammatory response of the endothelial cells under hypoxia and it is difficult from these studies to obtain a global picture about the inflammatory phenotype of the endothelial cells exposed to hypoxia. The present study was aimed to obtain such a picture. To this purpose, we used a low density DNA microarray specially designed to quantitate the expression of genes involved in the inflammatory pathways. Results were
completed by using a customized real-time PCR array and with functional studies. We concluded that hypoxia does not shift the endothelial cell phenotype to a more pro-inflammatory one despite a slight enhanced adhesiveness for leukocytes.
Materials and methods Cell culture and hypoxia incubation EAhy926 endothelial cells [10] were maintained in Dulbecco's Modified Eagle Medium (DMEM) (4.5 g/l D-glucose, Invitrogen) containing 10% (v/v) fetal calf serum (Invitrogen). The human acute monocytic leukemia cell line THP-1 was purchased from the ATCC and cultured in RPMI-1640 (Invitrogen) supplemented with 10 mM HEPES, 1 mM sodium pyruvate, 4.5 g/l glucose, 1.5 g/l
736
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
sodium bicarbonate, 0.05 mM 2-mercaptoethanol and 10% fetal calf serum (Invitrogen). HUVEC were purchased from Lonza and cultivated as instructed, using the Clonetics® Endothelial Cell Systems. For hypoxia experiments (1% O2), cells were incubated in serumfree CO2-independent medium (Invitrogen) supplemented with 1 mM L-glutamine (Sigma). Normoxic control cells were incubated in the same conditions but in normal atmosphere (20% O2). Positive controls were performed by stimulating the cells in the presence of 10 ng/ml TNF-α (human recombinant, R&D Systems).
Immunofluorescence staining and confocal microscopy Endothelial cells were seeded at 400,000 cells/well on glass coverslips in 24-well plates. After 24 h incubation in standard conditions, cells were incubated 16 h under normoxia or hypoxia or in the presence of TNF-α, thereafter medium was removed and cells were fixed 10 min with PBS containing 4% paraformaldehyde (Merck). Fixed cells were then washed three times with PBS-BSA 2%, cells were incubated at 4 °C overnight with the primary antibody (anti-ICAM-1, # BBA3, R&D Systems, dil. 1/100). Cells were washed three times as described above and the secondary antibody conjugated to Alexa fluorochrome (488 nm, dil. 1/500)
was added. After 1 h incubation, cells were washed three times with PBS. For nuclei labeling, cells were incubated 30 min with TOPRO-3 (Molecular Probes, dil. 1/80) in the presence of 2 mg/ml RNAse then washed three times with PBS. Finally, glass coverslips were mounted in Mowiol for observation in confocal microscopy (Leica). Semi-quantitative observations were performed with a constant PMT value.
Cytokine microarray SignalChip Human Cytokine microarrays (Eppendorf Array Technologies) were used to analyze the protein level of 20 human cytokines, namely eotaxin, GM-CSF, IFNγ, IL-1α, IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, IL-12p40, IL-12p70, IL-17, IP-10, MIP-1α, MIP-1β, RANTES, TNFα, TNF-RI and TNF-RII. The assay is based on a sandwich ELISA, where capture antibodies are spotted on glass slides in an array format. Samples are contacted with the array, and detection is performed using labeled detection antibodies. Positive and negative controls are included to normalize the data, and calibration curves constructed from known amounts of purified cytokines are used to generate quantitative data. Arrays were contacted with 8.7 μl of sample diluted 10 times and detection was performed in fluorescence, according to the manufacturer's
Fig. 1 – Effect of hypoxia on the expression of inflammatory genes. EAhy926 cells were incubated during 4, 8, 16 or 24 h under normoxia or hypoxia and total RNA was extracted and retrotranscribed before being hybridized onto the Inflammation Dual Chips. Results are presented in induction fold (hypoxia vs normoxia) as means for two independent experiments.
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
Table 2 – Effect of hypoxia (16 hours) on the expression of inflammatory genes
737
Table 2 (continued) protocol. Scanning was performed using a ScanArray scanner, and
EAhy926 cells were incubated during 16h under normoxia or hypoxia and total RNA was extracted and retrotranscribed before being hybridized onto the Inflammation Dual Chips or being amplified in realtime using the Taqman Low Density Arrays. Results are presented in induction fold (hypoxia vs normoxia) for three independent experiments (means ± 1 SD). Results highlighted in pink are N 1 + 1 SD, results highlighted in blue are < 1–1 SD.
738
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
protocol. Scanning was performed using a ScanArray scanner, and signals were quantified using the Imagene software (PerkinElmer). Data analysis was performed using the data analysis software provided with the arrays.
IL-8 ELISA IL-8 secreted in the incubation medium was assayed using an ELISA (Quantikine from R&D Systems) according to the procedure
provided by the supplier. Results are expressed in pg of IL-8 reported to μg of proteins assayed by the Folin method.
Assay for prostacyclin After the incubation, the medium was collected for measurement of PGI2. PGI2 was determined by EIA according to the manufacturer's recommendations (Cayman). Results are expressed in pg of PGI2 reported to μg of proteins assayed by the Folin method.
Fig. 2 – Effect of hypoxia on the expression of CD31, MIF, CCL2, THBD and MMP2. EAhy926 cells and two different cultures of HUVEC were incubated during 16 h under normoxia or hypoxia and total RNA was extracted and retrotranscribed before quantitative PCR using appropriate primers. Results are presented in induction fold (hypoxia vs normoxia) as means ± 1 S.D. for three independent experiments.
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
Leukocyte adhesion assay Endothelial cells were seeded at 400,000 cells/well in 96-well plates. After 24 h incubation in standard conditions, cells were incubated 16 h under normoxia or hypoxia or in the presence of TNF-α. They were then incubated 1 h at 37 °C in the presence or not of neutralizing anti-ICAM-1 antibodies (anti-ICAM-1, # BBA3, R&D Systems, diluted 1/10 in PBS). Thereafter, the solution was removed and calcein-labeled THP-1 monocytes (150,000 monocytes/100 μl/well) were added to the wells and incubated at 37 °C. After 30 min incubation, the wells were rinsed three times with PBS. 200 μl of PBS were added and the fluorescence associated to the adherent leukocytes was determined (excitation at 485 nm — emission at 520 nm, Fluoroskan Ascent).
Nuclear protein extraction Nuclear protein extractions in high salt buffer were prepared as previously described [11]. Briefly, cells seeded in 75 cm2 flasks (Corning) were incubated under normoxic or hypoxic conditions for 16 h. At the end of the incubation, cells were rinsed with PBS containing 1 mM Na2MoO4 and 5 mM NaF. They were then incubated on ice for 3 min with 10 ml cold Hypotonic Buffer (HB 1×, 2 mM HEPES, 0.5 mM NaF, 0.1 mM Na2MoO4, 0.01 mM EDTA) and harvested in 500 μl HB containing 0.5% NP-40 (Sigma). Cell lysates were centrifuged 30 s at 13,000 rpm and sedimented nuclei were resuspended in 50 μl HB 10× containing 20% glycerol and a protease inhibitor cocktail (Complete, Roche) and phosphatase inhibitors (1 mM Na3VO4, 5 mM NaF, 10 mM p-nitrophenylphosphate, 10 mM β-glycerophosphate). Extraction was performed for 30 min at 4 °C by the addition of 100 μl HB 10× containing 20% glycerol, 0.8 M NaCl and protease/phosphatase inhibitors.
DNA-binding assay DNA-binding assays using TransAM ELISA kit (Active Motif) for detecting transcription factor DNA binding activity was performed according to the manufacturer's recommendations. Briefly, 5 μg of nuclear proteins were incubated for 1 h in a 96-well plate coated with a double-stranded oligonucleotide containing the consensus sequence recognized by the transcription factor to be assayed. The transcription factor bound to DNA was detected using a specific primary antibody (rabbit anti-p65 (SC-372 Santa Cruz), rabbit antiIRF-1 (SC-497 Santa Cruz) or mouse anti-HIF-1α (# 610958 Becton Dickinson)). Colorimetric reaction was then performed with a HRPconjugated anti-rabbit or anti-mouse IgG antibody and absorbance was measured at 450 nm in a spectrophotometer.
739
extraction kit (Promega), quality was checked with a bioanalyzer (Agilent Technologies) and 20 μg were used for retrotranscription in the presence of biotin-11-dCTP (Perkin-Elmer) and Superscript II Reverse Transcriptase (InVitrogen), as described previously [12]. Hybridizations on the arrays were carried out as described by the manufacturer and reported previously [12]. Detection was performed with a cyanin 3-conjugated IgG anti-biotin (Jackson Immuno Research Laboratories). Fluorescence of hybridized arrays was scanned using a Packard ScanArray (Perkin-Elmer) at a resolution of 10 μm.
Real time RT-PCR After the incubation, total RNA was extracted using the Total RNAgent extraction kit (Promega). mRNA contained in 2 μg total RNA was reverse transcribed using SuperScript II Reverse Transcriptase (InVitrogen) and oligodT primers according to the manufacturer's instructions. Forward and reverse primers for aldolase, TGF-b1, CTGF, MIF, MMP-2, PDGF-B, C/EBPb, ICAM-1, CD31, THBD, CCL2 and EPHB-2 were designed using the Primer Express 1.5 software (Applied Biosystem). Amplification reaction assays contained 1× SYBR Green PCR Mastermix (Applied Biosystem) and primers (Eurogentec) at the optimal concentrations. A hot start at 95 °C for 5 min was followed by 40 cycles at 95 °C for 15 s and 65 °C for 1 min using an ABI PRISM 7000 SDS thermal cycler (Applied Biosystem). RPL13A was used as the reference gene for normalization and mRNA expression level was quantified using the threshold cycle method.
Taqman low density array After the incubation, total RNA was extracted using the Total RNAgent extraction kit (Promega). mRNA contained in 2 μg total RNA was reverse transcribed using the “High Capacity cDNA Archive” kit from Applied Biosystems according to the manufacturer's instructions. 100 ng of retrotranscribed total RNA in 50 μl are then mixed with 50 μl of the “Taqman Universal PCR master Mix” (Applied Biosystems) and loaded into one of the 8 fill ports of the microfluidic array. Customized arrays allowing the detection of 48 genes were used. mRNA expression level was quantified using the threshold cycle method with β-actin as the reference gene.
Statistical analysis Data are reported as means ± 1 S.D. One way analysis of variance and pairwise multiple comparison procedures (Holm-Sidak method) were performed with the program SigmaStat V.3.11 (Systat Software, Inc.) and used where appropriate.
Gene expression analysis on DNA microarray We used a low-density DNA array allowing the gene expression analysis for 310 genes related to inflammation (DualChip® human inflammation, Eppendorf). Results using these reliable and validated arrays developed by Eppendorf were reported elsewhere [12–14]. The method is based on a system with two identical arrays on a glass slide and three identical sub-arrays (triplicate spots) per array. Endothelial cells cultured in 75 cm2 flasks (Corning) were incubated for different incubation times under normoxic and hypoxic conditions. At the end of the incubation, total RNA was extracted with the Total RNAgents
Results Effects of hypoxia on the expression of inflammatory genes In order to study the effects of hypoxia on the expression of a high number of inflammatory genes, we used a low density DNA array harboring 310 probes for inflammatory genes. We first studied the influence of the duration of hypoxia: cells were incubated 4, 8, 16 or 24 h under normoxia or hypoxia and the expression was quantified compared to the corresponding control. Table 1 shows
740
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
the complete list of the genes that were detected (107 genes). Out of these 107 genes, 7 were significantly upregulated for the two duplicates at least at one endpoint while 39 were significantly downregulated for the two duplicates at least at one endpoint. Fig. 1 shows the results for some of them: of note is the upregulation of CTGF, IGFBP2, MIF and NRG1. On the other hand, ICAM-2 AND ICAM-3 were downregulated (ICAM-1 mRNA was not detected) as well as THBD and PGF. These results were validated by real-time RT-PCR for 8 genes and very similar results were obtained (data not shown). Since the increase in expression was higher at 16 h than at 24 h for 6 genes out of the 7 upregulated genes, we chose this endpoint for the further experiments. In order to gain insight for the low expressed genes that were not detected by the DNA microarray, we repeated the experiments at 16 h in order to quantify the mRNA level of 48 inflammatory genes using a customized array for real-time PCR reactions. The same samples were hybridized in parallel on the DNA microarray to compare the results. 20 μg were hybridized instead of 10 μg in the previous experiments, in order to increase the number of detected mRNAs. All the data are presented in Table 2. For the genes that were detected by both methods, similar down- or upregulation levels were obtained: such examples are CCL2, CSF3 and THBD for downregulated genes and F2R for upregulated genes. Of note were the downregulation of several well known proinflammatory genes: CCL2 (MCP-1), CCL5 (RANTES), CSF1, CSF2, CSF3, CXCL2 (MIP2a), LTA (lymphotoxin A), LTB (lymphotoxin B), PLA2G4C (cPLA2) and the upregulation of other pro-inflammatory genes: CTGF, CXCR4 (receptor for SDF-1), ICAM-1, MIF, PTGS2 (COX2). Changes in anti- and pro-angiogenic gene expression are also observed: upregulation of ENG (endogline), EPHB2 (ephrin B2), MMP2 (matrix metalloproteinase 2), VEGF and downregulation of FGF1 (acidic FGF), FGF2 (basic FGF), PGF (PlGF), THBD (thrombomodulin), and TIMP1/2. EAhy926 cells were often used as representative endothelial cells. Many works from our laboratory [15–17] and others [18,19] have demonstrated that these cells behave like primary HUVEC regarding different pathways. However, in order to confirm that the changes in gene expression induced by hypoxia in EAhy926 cells are relevant to endothelial cells, two different cultures of primary HUVEC have been tested. Fig. 2 shows that hypoxia induced similar changes in gene expression in HUVEC than in EAhy926 cells, except for THBD for which no change was observed in HUVEC.
Effects of hypoxia on the DNA binding activity of HIF-1, IRF-1 and NF-kB Changes in gene expression are the result of changes in gene transcription or in mRNA stability. In order to investigate whether the activity of transcription factors known to be involved in regulating gene expression either under hypoxia (HIF-1) or during inflammation (NF-kB and IRFs), we assessed the DNA binding activity of three of these transcription factors: HIF-1, NF-kB and IRF1. The DNA binding activity of HIF-1 was increased by 6-fold under hypoxia (Fig. 3A). Such an increase has already been described for this cell line [20]. The DNA binding activity of IRF-1 was decreased by 40% after 16 h incubation under hypoxic conditions (Fig. 3B). This result is coherent with the 60% decrease in IRF-1 mRNA expression that we detected using the low density DNA microarray (Table 2). Similarly, the DNA binding activity of NF-kB was lower under
Fig. 3 – Effect of hypoxia on the DNA binding activity of HIF-1, IRF-1 and NF-kB. EAhy926 cells were incubated during 16 h under normoxia or hypoxia and nuclear extracts were recovered. The DNA binding activity was then measured using specific TransAM assays. Results are given as means ± 1 SD (n = 3). ⁎⁎ and ⁎⁎⁎, p < 0.01 and p < 0.001 vs normoxia.
hypoxic conditions than in normoxic conditions (Fig. 3C). Again, this observation is parallel to the decrease in p65 (NFKB1) mRNA expression under hypoxia (see Table 2).
Effects of hypoxia on pro-inflammatory mediator secretion Since numerous changes (up- and downregulation) in proinflammatory and anti-inflammatory genes were observed and
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
741
Fig. 4 – Effect of hypoxia on the secretion of pro-inflammatory cytokines. (A, B) EAhy926 cells were incubated during 4, 8, 16 or 24 h under normoxia or hypoxia and the medium was recovered before being hybridized onto the Cytokine Chip. An example of such an experiment is given in A and the quantified results are given for IL-8 and soluble TNF-R1 in B. Results are presented in optical density (OD) as means for two independent experiments. (C) EAhy926 cells were incubated during 2, 4 or 16 h under normoxia or hypoxia and the medium was recovered before being assayed for IL-8 using an ELISA. Results are given in pg of IL-8/μg proteins, as means ± 1 SD (n = 3). ⁎⁎⁎, p < 0.001 vs normoxia.
since changes in mRNA expression are not always reflected at the protein level, we wanted to assess the actual secretion of proinflammatory cytokines under hypoxic conditions. For that, we used a protein array able to detect 20 cytokines (Fig. 4A). Of that 20 cytokines, only two of them varied significantly under hypoxic conditions in comparison to normoxic conditions: IL-8 and sTNFR1. Both of them were constitutively secreted by the endothelial cells under normoxia. Hypoxia markedly decreased IL-8 secretion already after 16 h incubation while sTNF-R1 release was decreased only after 24 h incubation (Fig. 4B). These results were confirmed
using a specific ELISA for IL-8 (Fig. 4C). TNF-R1 mRNA level was also significantly decreased as detected by the low density DNA array (Table 2). IL-8 mRNA was not detected by the low density DNA array probably due to a too low expression level. It was detected using the real-time PCR array but no significant change was observed, due to a very high variability of the results probably because it was very near the detection limit of the method. Among the genes whose expression varied under hypoxia, are PLA2G4C (cPLA2, downregulated) and PTGS2 (COX2, upregulated) which are involved in prostaglandin synthesis. In order to assess
742
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
Discussion
Fig. 5 – Effect of hypoxia on the secretion of PGI2. EAhy926 cells were incubated during 2, 4 or 16 h under normoxia or hypoxia and the medium was recovered before being assayed for PGI2 using an EIA. Results are given in pg of PGI 2/μg proteins, as means ± 1 SD (n = 3).
the overall consequence of these variations on prostaglandin synthesis, we quantified PGI2 secretion using a specific EIA assay. Fig. 5 shows that low amounts of PGI2 are secreted constitutively by the endothelial cells but hypoxia did not significantly affected PGI2 secretion.
Effects of hypoxia on leukocyte adhesion onto endothelial cells An increase in the expression of ICAM-1 mRNA was observed after 16 h incubation under hypoxia (Table 2). We thus wanted to confirm this result by quantitative RT-PCR in both endothelial cell types: a 70% increase was indeed detected in EAhy926 cells. However, hypoxia induced a decrease in ICAM-1 mRNA level in both HUVEC cultures (Fig. 6). We also checked whether ICAM-1 protein expression was also changed, by immunofluorescence staining. Fig. 6 shows that hypoxia did not induce ICAM-1 expression on the surface of the endothelial cells, EAhy926 and HUVEC, while TNF-α did markedly increase its expression. One the most important consequences of endothelial cell activation during inflammation is the increase in leukocyte adhesion to these endothelial cells. In order to investigate whether hypoxia affects endothelial cell adhesiveness, unstimulated THP-1 cells (monocytic cell line) were added to endothelial cells after 16 h incubation under normoxia or hypoxia or in the presence of TNF-α used as a positive control. Fig. 7A shows that hypoxia significantly increased the adhesiveness of endothelial cells for leukocytes however to a lower extent than TNF-α. Neutralizing anti-ICAM-1 antibodies did not modify this increase (Fig. 7B). It has to be noted that the fact that hypoxia-increase in EAhy926 cell adhesiveness for leukocytes was lower in this case may be due to the fact that the cells were incubated 1 h in the presence of the antibodies after the hypoxia incubation before THP-1 were added. These antibodies were functional since they did inhibit the increase in THP-1 adhesion induced by TNF-alpha. Another mechanism than ICAM-1 overexpression is thus probably involved. On the other hand, hypoxia did not increase the adhesiveness of HUVEC for leukocytes while TNF-α had a strong effect (Fig. 7C).
Endothelial cell activation is an important feature of inflammatory response. Since hypoxia is involved in chronic inflammatory diseases, it is important to understand how hypoxia influences endothelial cell inflammatory phenotype. This study was aimed to obtain a global picture of the endothelial cell response to hypoxia by investigating changes in inflammatory gene expression. In parallel the activity of three transcription factors was assessed in order to understand how the changes in gene expression could be regulated. A summary of these results is presented in Fig. 8. The major regulator of cellular responses to hypoxia is HIF-1. Unsurprisingly, the DNA binding activity of HIF-1 was observed to be increased in our experimental conditions. In parallel, an increased expression of genes already known to be transcriptionally regulated by HIF-1 has been evidenced. This is the case for aldolase, VEGF and endoglin (ENG) [4], PDGF-β [21], CXCR4 [22], CTGF [23], MIF [24], and MMP2 [25]. Two other genes were downregulated, despite being known to be regulated by HIF: FGF2 (bFGF, [26]) and PlGF [9]; probably because the expression of these genes also depends on the activity of other transcription factors. Indeed, NF-kB is known to also regulate PDGF-β expression and its activity is decreased under hypoxia in these experiment conditions (see below). Similarly, COX2 has been described to be upregulated by HIF-1 under hypoxic conditions [6,27], as observed here for both endothelial cell types (HUVEC, data not shown), but in other stimulating conditions, NF-kB [28] and IRF-1 [29] whose activity is decreased here can also regulate its expression. Most of our data are thus in accordance with the data from the literature and thus validate our approach. On the other hand, the major regulator of inflammatory responses is NF-kB. The results shown here evidenced a decrease in the mRNA level of NFKB1 (p65) which is parallel to a decrease in the DNA binding activity of NF-kB. Our results indicated that in our experimental conditions, (1% O2), hypoxia does not lead to NF-kB activation in endothelial cells. There are reports in the literature reporting NF-kB activation under hypoxia in different cell types: macrophages [30], neutrophils [31], endothelial cells [32] as well as cancer cells [33]. The exact mechanism responsible for NF-kB activation under hypoxia is not clear but it may involve ROS [34]. However, there are also other reports that show that NF-kB is activated during the reoxygenation phase following a previous hypoxia incubation period and in these cases, ROS are clearly involved in NF-kB activation but there is no activation of NF-kB during the hypoxia phase [35,36]. In conclusion, according to the experimental conditions and probably the cell types, hypoxia does or does not activate NF-kB but the exact mechanisms for this fine tuning are not yet known. Among the known NF-kB target genes [28] whose mRNA are detectable by the two arrays used here, a few of them were not changed. This is the case for IL-6, CSF2 (GM-CSF) or IL-8. On the other hand, most of them displayed a decreased expression: IL-1α and IL-1β, CCL2 (MCP-1), CCL5 (RANTES), CXCL2 (Groβ), CSF1, CSF3, LTA/B (lymphotoxins A and B). This is consistent with the decreased DNA binding activity of NF-kB. Another NF-kB target gene whose expression was decreased in our experimental conditions is IRF-1. Again, in parallel, the DNA binding activity of this transcription factor was observed to be decreased as well as the expression of one of its target genes, TAP1 [37]. The only NF-kB
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
743
Fig. 6 – Effect of hypoxia on ICAM-1 expression. EAhy926 cells and two cultures of HUVEC were incubated during 16 h under normoxia or hypoxia or in the presence of TNF-α at 10 ng/ml, used a positive control. (A, B) After the incubation, total RNA was extracted and retrotranscribed before quantitative PCR using appropriate primers. Results are presented in induction fold (hypoxia vs normoxia) as means ± 1 SD for three independent experiments. (C, D) After the incubation, cells were fixed and staining for ICAM-1 (green). Nuclei are stained in blue using TOPRO-3. Cells were observed using a confocal microscope with a constant photomultiplier.
target gene whose expression was increased under hypoxia was ICAM-1, in EAhy926 cells but not in HUVEC. We compared our results to the ones published by Manalo et al. [9]. Some of the hypoxia-inducible genes they described are also found in our study: this is the case for VEGF, PDGF-B, CXCR4, MMP2, TGF-b1, while others are different. This can be explained by the fact that they incubated the cells for 24 h under hypoxia while we did for 16 h and they used artery endothelial cells while we used umbilical vein-derived cells. It is well recognized that the endothelium is highly sensitive to ischemia–reperfusion (I/R), increasing leukocyte recruitment and adhesion. This hyperadhesiveness of leukocytes to endothelial cells contributes to I/R induced tissue injury [3,38]. Most of the studies aimed to understand the underlying mechanisms responsible for this increase in leukocyte adhesion have been designed so that only the effect of the reoxygenation phase was investigated (ex. [39–41]). In addition, some studies investigated the effects of hypoxia on leukocytes and showed that hypoxia can modulate leukocyte adhesion to EC in that case [42–44]. In our study, the increase in leukocyte adhesion can only be due to endothelial cell stimulation since we used unstimulated, normoxic THP-1. This study was aimed to dissect the mechanisms that may be involved during the first hypoxia phase and on endothelial cells. A slight increase in leukocyte adhesion to the endothelial cells was indeed observed directly after the incubation of the
endothelial cells under hypoxia for 16 h for EAhy926 cells but not for HUVEC. This increase in leukocyte adhesiveness is probably not due to a binding to ICAM-1 since no increase in the expression of the protein was observed despite the increase of the mRNA level and there was no effect of neutralizing antiICAM-1 antibodies. This result is in accordance with other results from the literature that showed no effect of hypoxia on ICAM-1 expression on HUVEC or HAEC-1 [45,46] or even a decreased expression [47] while these studies demonstrated an increase in ICAM-1 expression during the reoxygenation phase. VCAM-1 mRNA was not detected so that it is probably not either via this molecule that the adhesion is mediated. Finally, CD31 mRNA was downregulated by hypoxia in both endothelial cell types. Recruitment of leukocytes may prime these cells for enhanced adhesion: CCL2 (MCP-1) and CCL5 (RANTES) have been described to be involved in the recruitment of monocytes into hypoxic tissues [48]. However, in our studies, the mRNA level of both cytokines was markedly decreased under hypoxia. IL-8 is also known to display chemotactic properties. Our results showed that hypoxia did not change the mRNA level of IL-8 and even decreased the IL-8 secretion by the endothelial cells. The discrepancy between these observations and the ones of Kim et al. [5] that showed an enhanced IL-8 and MCP-1 secretion by endothelial cells under hypoxia may be due to the different endothelial cell types (EAhy929 vs human pulmonary artery endothelial cells). Other chemokines may also be involved in
744
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
Fig. 7 – Effect of hypoxia on leukocyte adhesion. EAhy926 cells and HUVEC were incubated during 16 h under normoxia or hypoxia or in the presence of TNF-α at 10 ng/ml, used a positive control. (A) After the incubation, calcein-labeled leukocytes were added for 60 min and endothelial cells were rinsed. Results are given in RFU (relative fluorescence units) as means ± 1 SD (n = 6). ⁎ and ⁎⁎⁎, p < 0.05 and p < 0.001 vs normoxia. (B, C) After the incubation, cells were incubated 1 h at 37 °C in the presence or not of neutralizing anti-ICAM-1 antibodies. Thereafter, the solution was removed and calcein-labeled leukocytes were added for 60 min and endothelial cells were rinsed. Results are given in RFU (relative fluorescence units) as means± 1 SD (n = 6). ⁎ and ⁎⁎, p < 0.05 and p < 0.01 vs normoxia; (⁎⁎), p < 0.01 vs TNF-α.
leukocyte recruitment and activation to the hypoxic endothelium: monocytes and macrophages respond to the chemotactic effects of CXCL12 (SDF-1) via the CXCR4 receptor and hypoxia has been shown to significantly upregulate the expression of both proteins [22,49]. Our results showed that CXCL12 mRNA level may be induced under hypoxia but it did not reach statistical significance while CXCR4 mRNA was indeed markedly increased in the endothelial cells. Another possible pathway that would need further investigation is the role of lipid molecules such as PAF or prostaglandins; indeed an increase in COX2 expression was observed in both cell types. Whether these pathways are
involved in the increased adhesion of leukocytes to hypoxic endothelial cells remains to be determined. Further investigation will thus be needed to identify the exact mechanism(s) responsible for the leukocyte adhesion to the hypoxic EAhy926 endothelial cells in our experimental conditions. Chronic hypoxia influences the release of vasoconstrictors, growth factors and pro-inflammatory molecules in endothelial cells. Hypoxic exposure of the endothelium induces an inflammatory response within the vessel wall responsible for vascular remodeling [2] but also I/R tissue damages. Our data show that hypoxia did lead to a slight increased adhesiveness of leukocytes to
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
745
Fig. 8 – Schematic representation of the effects of hypoxia on the activity of several transcription factors and on gene expression. Corresponding proteins are located at their known subcellular localization. Proteins encoded by HIF-1 target genes are in red, proteins encoded by IRF-1 target genes are in yellow and proteins encoded by NF-kB target genes are in blue.
the EAhy926 endothelial cells but probably not to a proinflammatory phenotype. The global gene expression analyses preformed in this work suggest that HIF-1 is one main regulator of gene expression in endothelial cells in the experimental conditions used here while NF-kB activity is decreased. The expression of several cytokines is decreased while others are upregulated (Fig. 8), indicating a highly complex process where numerous interactive events are taking place. All of them must be taken into account if a better picture is to be determined of the effects of hypoxia on inflammatory reactions.
Acknowledgments Sébastien Toffoli is a recipient of a FNRS-Télévie grant. Carine Michiels is research director of FNRS (Fonds National de la Recherche Scientifique, Belgium). This article presents results of the Belgian Program on Interuniversity Poles of Attraction initiated by the Belgian State, Prime Minister's Office, Science Policy Programming. The responsibility is assumed by its authors. We are grateful to Prof. C. Edgell (Pathology Department, University of North Carolina) for kindly donating the EAhy926 cells. We thank the Fonds de la Recherche Fondamentale Collective for financial support. We also acknowledge the financial support of les Laboratoires Pierre Fabre.
REFERENCES
[1] C. Michiels, Endothelial cell functions, J. Cell Physiol. 196 (2003) 430–443. [2] K.R. Stenmark, K.A. Fagan, M.G. Frid, Hypoxia-induced pulmonary vascular remodeling: cellular and molecular mechanisms, Circ. Res. 99 (2006) 675–691. [3] C. Michiels, T. Arnould, J. Remacle, Endothelial cell responses to hypoxia: initiation of a cascade of cellular interactions, Biochim. Biophys. Acta 1497 (2000) 1–10. [4] G.L. Semenza, Targeting HIF-1 for cancer therapy, Nat. Rev. Cancer 3 (2003) 721–732. [5] K.S. Kim, V. Rajagopal, C. Gonsalves, C. Johnson, V.K. Kalra, A novel role of hypoxia-inducible factor in cobalt chloride- and hypoxia-mediated expression of IL-8 chemokine in human endothelial cells, J. Immunol. 177 (2006) 7211–7224. [6] A. Kaidi, D. Qualtrough, A.C. Williams, C. Paraskeva, Direct transcriptional up-regulation of cyclooxygenase-2 by hypoxia-inducible factor (HIF)-1 promotes colorectal tumor cell survival and enhances HIF-1 transcriptional activity during hypoxia, Cancer Res. 66 (2006) 6683–6691. [7] S.B. Scheurer, J.N. Rybak, C. Rosli, D. Neri, G. Elia, Modulation of gene expression by hypoxia in human umbilical cord vein endothelial cells: a transcriptomic and proteomic study, Proteomics 4 (2004) 1737–1760. [8] W. Ning, T.J. Chu, C.J. Li, A.M. Choi, D.G. Peters, Genome-wide analysis of the endothelial transcriptome under short-term chronic hypoxia, Physiol. Genomics 18 (2004) 70–78.
746
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
[9] D.J. Manalo, A. Rowan, T. Lavoie, L. Natarajan, B.D. Kelly, S.Q. Ye, J.G. Garcia, G.L. Semenza, Transcriptional regulation of vascular endothelial cell responses to hypoxia by HIF-1, Blood 105 (2005) 659–669. [10] C.J. Edgell, C.C. McDonald, J.B. Graham, Permanent cell line expressing human factor VIII-related antigen established by hybridization, Proc. Natl. Acad. Sci. U. S. A. 80 (1983) 3734–3737. [11] Y. Chen, A. Takeshita, K. Ozaki, S. Kitano, S. Hanazawa, Transcriptional regulation by transforming growth factor beta of the expression of retinoic acid and retinoid X receptor genes in osteoblastic cells is mediated through AP-1, J. Biol. Chem. 271 (1996) 31602–31606. [12] F. de Longueville, D. Surry, G. Meneses-Lorente, V. Bertholet, V. Talbot, S. Evrard, N. Chandelier, A. Pike, P. Worboys, J.P. Rasson, B. Le Bourdelles, J. Remacle, Gene expression profiling of drug metabolism and toxicology markers using a low-density DNA microarray, Biochem. Pharmacol. 64 (2002) 137–149. [13] F. de Longueville, F.A. Atienzar, L. Marcq, S. Dufrane, S. Evrard, L. Wouters, F. Leroux, V. Bertholet, B. Gerin, R. Whomsley, T. Arnould, J. Remacle, M. Canning, Use of a low-density microarray for studying gene expression patterns induced by hepatotoxicants on primary cultures of rat hepatocytes, Toxicol. Sci. 75 (2003) 378–392. [14] F. Debacq-Chainiaux, C. Borlon, T. Pascal, V. Royer, F. Eliaers, N. Ninane, G. Carrard, B. Friguet, F. de Longueville, S. Boffe, J. Remacle, O. Toussaint, Repeated exposure of human skin fibroblasts to UVB at subcytotoxic level triggers premature senescence through the TGF-beta1 signaling pathway, J. Cell Sci. 118 (2005) 743–758. [15] S. Toffoli, O. Feron, M. Raes, C. Michiels, Intermittent hypoxia changes HIF-1alpha phosphorylation pattern in endothelial cells: unravelling of a new PKA-dependent regulation of HIF-1alpha, Biochim. Biophys. Acta 1773 (2007) 1558–1571. [16] C. Gustin, E. Delaive, M. Dieu, D. Calay, M. Raes, Upregulation of pentraxin-3 in human endothelial cells after lysophosphatidic acid exposure, Arterioscler. Thromb. Vasc. Biol. 28 (2008) 491–497. [17] C. Gustin, M. Van Steenbrugge, M. Raes, LPA modulates monocyte migration directly and via LPA-stimulated endothelial cells, Am. J. Physiol. Cell Physiol. 295 (2008) C905–914. [18] J.E. Suggs, M.C. Madden, M. Friedman, C.J. Edgell, Prostacyclin expression by a continuous human cell line derived from vascular endothelium, Blood 68 (1986) 825–829. [19] D. Bouis, G.A. Hospers, C. Meijer, G. Molema, N.H. Mulder, Endothelium in vitro: a review of human vascular endothelial cell lines for blood vessel-related research, Angiogenesis 4 (2001) 91–102. [20] S. Paris, H. Denis, E. Delaive, M. Dieu, V. Dumont, N. Ninane, M. Raes, C. Michiels, Up-regulation of 94-kDa glucose-regulated protein by hypoxia-inducible factor-1 in human endothelial cells in response to hypoxia, FEBS Lett. 579 (2005) 105–114. [21] M. Yamakawa, L.X. Liu, T. Date, A.J. Belanger, K.A. Vincent, G.Y. Akita, T. Kuriyama, S.H. Cheng, R.J. Gregory, C. Jiang, Hypoxia-inducible factor-1 mediates activation of cultured vascular endothelial cells by inducing multiple angiogenic factors, Circ. Res. 93 (2003) 664–673. [22] T. Schioppa, B. Uranchimeg, A. Saccani, S.K. Biswas, A. Doni, A. Rapisarda, S. Bernasconi, S. Saccani, M. Nebuloni, L. Vago, A. Mantovani, G. Melillo, A. Sica, Regulation of the chemokine receptor CXCR4 by hypoxia, J. Exp. Med. 198 (2003) 1391–1402. [23] D.F. Higgins, M.P. Biju, Y. Akai, A. Wutz, R.S. Johnson, V.H. Haase, Hypoxic induction of Ctgf is directly mediated by Hif-1, Am. J. Physiol. Renal. Physiol. 287 (2004) F1223–1232. [24] J.A. Baugh, M. Gantier, L. Li, A. Byrne, A. Buckley, S.C. Donnelly, Dual regulation of macrophage migration inhibitory factor (MIF) expression in hypoxia by CREB and HIF-1, Biochem. Biophys. Res. Commun. 347 (2006) 895–903. [25] K.G. Shyu, F.L. Hsu, M.J. Wang, B.W. Wang, S. Lin, Hypoxia-inducible factor 1alpha regulates lung adenocarcinoma cell invasion, Exp. Cell Res. 313 (2007) 1181–1191.
[26] M. Calvani, A. Rapisarda, B. Uranchimeg, R.H. Shoemaker, G. Melillo, Hypoxic induction of an HIF-1alpha-dependent bFGF autocrine loop drives angiogenesis in human endothelial cells, Blood 107 (2006) 2705–2712. [27] R.J. Cook-Johnson, M. Demasi, L.G. Cleland, J.R. Gamble, D.A. Saint, M.J. James, Endothelial cell COX-2 expression and activity in hypoxia, Biochim. Biophys. Acta 1761 (2006) 1443–1449. [28] H.L. Pahl, Activators and target genes of Rel/NF-kappaB transcription factors, Oncogene 18 (1999) 6853–6866. [29] T. Taniguchi, K. Ogasawara, A. Takaoka, N. Tanaka, IRF family of transcription factors as regulators of host defense, Annu. Rev. Immunol. 19 (2001) 623–655. [30] C.C. Fong, Q. Zhang, Y.F. Shi, R.S. Wu, W.F. Fong, M. Yang, Effect of hypoxia on RAW264.7 macrophages apoptosis and signaling, Toxicology 235 (2007) 52–61. [31] S.R. Walmsley, C. Print, N. Farahi, C. Peyssonnaux, R.S. Johnson, T. Cramer, A. Sobolewski, A.M. Condliffe, A.S. Cowburn, N. Johnson, E.R. Chilvers, Hypoxia-induced neutrophil survival is mediated by HIF-1alpha-dependent NF-kappaB activity, J. Exp. Med. 201 (2005) 105–115. [32] H. Matsushita, R. Morishita, T. Nata, M. Aoki, H. Nakagami, Y. Taniyama, K. Yamamoto, J. Higaki, K. Yasufumi, T. Ogihara, Hypoxia-induced endothelial apoptosis through nuclear factor-kappaB (NF-kappaB)-mediated bcl-2 suppression: in vivo evidence of the importance of NF-kappaB in endothelial cell regulation, Circ. Res. 86 (2000) 974–981. [33] Q. Shi, X. Le, J.L. Abbruzzese, B. Wang, N. Mujaida, K. Matsushima, S. Huang, Q. Xiong, K. Xie, Cooperation between transcription factor AP-1 and NF-kappaB in the induction of interleukin-8 in human pancreatic adenocarcinoma cells by hypoxia, J. Interferon Cytokine Res. 19 (1999) 1363–1371. [34] J.M. Lluis, F. Buricchi, P. Chiarugi, A. Morales, J.C. Fernandez-Checa, Dual role of mitochondrial reactive oxygen species in hypoxia signaling: activation of nuclear factor-{kappa}B via c-SRC and oxidant-dependent cell death, Cancer Res. 67 (2007) 7368–7377. [35] R. Natarajan, B.J. Fisher, D.G. Jones, A.A. Fowler, Atypical mechanism of NF-kappaB activation during reoxygenation stress in microvascular endothelium: a role for tyrosine kinases, Free Radic. Biol. Med. 33 (2002) 962. [36] M. Tamatani, N. Mitsuda, H. Matsuzaki, H. Okado, S. Miyake, M.P. Vitek, A. Yamaguchi, M. Tohyama, A pathway of neuronal apoptosis induced by hypoxia/reoxygenation: roles of nuclear factor-kappaB and Bcl-2, J. Neurochem. 75 (2000) 683–693. [37] H. Nguyen, J. Hiscott, P.M. Pitha, The growing family of interferon regulatory factors, Cytokine Growth Factor Rev. 8 (1997) 293–312. [38] D.N. Granger, P.R. Kvietys, M.A. Perry, Leukocyte–endothelial cell adhesion induced by ischemia and reperfusion, Can. J. Physiol. Pharmacol. 71 (1993) 67–75. [39] H. Ichikawa, S. Flores, P.R. Kvietys, R.E. Wolf, T. Yoshikawa, D.N. Granger, T.Y. Aw, Molecular mechanisms of anoxia/reoxygenation-induced neutrophil adherence to cultured endothelial cells, Circ. Res. 81 (1997) 922–931. [40] S. Kokura, N. Yoshida, T. Yoshikawa, Anoxia/reoxygenation-induced leukocyte–endothelial cell interactions, Free Radic. Biol. Med. 33 (2002) 427–432. [41] S. Cassie, M.F. Masterson, A. Polukoshko, M.M. Viskovic, L.A. Tibbles, Ischemia/reperfusion induces the recruitment of leukocytes from whole blood under flow conditions, Free Radic. Biol. Med. 36 (2004) 1102–1111. [42] T. Kong, H.K. Eltzschig, J. Karhausen, S.P. Colgan, C.S. Shelley, Leukocyte adhesion during hypoxia is mediated by HIF-1dependent induction of beta2 integrin gene expression, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 10440–10445. [43] M.C. Bosco, M. Puppo, C. Santangelo, L. Anfosso, U. Pfeffer, P. Fardin, F. Battaglia, L. Varesio, Hypoxia modifies the transcriptome of primary human monocytes: modulation of novel immunerelated genes and identification of CC-chemokine ligand 20 as a new hypoxia-inducible gene, J. Immunol. 177 (2006) 1941–1955. [44] C. Murdoch, M. Muthana, C.E. Lewis, Hypoxia regulates
EX PE R IM E NTAL CE L L RE S E ARCH 315 ( 2 0 0 9 ) 733 – 747
macrophage functions in inflammation, J. Immunol. 175 (2005) 6257–6263. [45] T.B. Corcoran, A. Engel, G.D. Shorten, The influence of propofol on the expression of intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) in reoxygenated human umbilical vein endothelial cells, Eur. J. Anaesthesiol. 23 (2006) 942–947. [46] R.C. Ziegelstein, C. He, Q. Hu, Hypoxia/reoxygenation stimulates Ca2+-dependent ICAM-1 mRNA expression in human aortic endothelial cells, Biochem. Biophys. Res. Commun. 322 (2004) 68–73.
747
[47] C. Willam, R. Schindler, U. Frei, K.U. Eckardt, Increases in oxygen tension stimulate expression of ICAM-1 and VCAM-1 on human endothelial cells, Am. J. Physiol. 276 (1999) H2044–2052. [48] K.R. Stenmark, N.J. Davie, J.T. Reeves, M.G. Frid, Hypoxia, leukocytes, and the pulmonary circulation, J. Appl. Physiol. 98 (2005) 715–721. [49] D.J. Ceradini, A.R. Kulkarni, M.J. Callaghan, O.M. Tepper, N. Bastidas, M.E. Kleinman, J.M. Capla, R.D. Galiano, J.P. Levine, G.C. Gurtner, Progenitor cell trafficking is regulated by hypoxic gradients through HIF-1 induction of SDF-1, Nat. Med. 10 (2004) 858–864.
