October 2003
Volume 44, Issue 10
Free
Biochemistry and Molecular Biology  |   October 2003
Coordinate Activation of HIF-1 and NF-κB DNA Binding and COX-2 and VEGF Expression in Retinal Cells by Hypoxia
Author Affiliations
  • Walter J. Lukiw
    From the Neuroscience Center and
    Department of Ophthalmology, Louisiana State University Health Sciences Center, New Orleans, Louisiana; and the
  • Anna Ottlecz
    Novartis Institute for Biomedical Research, Basel, Switzerland.
  • George Lambrou
    Novartis Institute for Biomedical Research, Basel, Switzerland.
  • Monika Grueninger
    Novartis Institute for Biomedical Research, Basel, Switzerland.
  • Joelle Finley
    From the Neuroscience Center and
  • Hilary W. Thompson
    From the Neuroscience Center and
    Department of Ophthalmology, Louisiana State University Health Sciences Center, New Orleans, Louisiana; and the
  • Nicolas G. Bazan
    From the Neuroscience Center and
    Department of Ophthalmology, Louisiana State University Health Sciences Center, New Orleans, Louisiana; and the
Investigative Ophthalmology & Visual Science October 2003, Vol.44, 4163-4170. doi:10.1167/iovs.02-0655
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Walter J. Lukiw, Anna Ottlecz, George Lambrou, Monika Grueninger, Joelle Finley, Hilary W. Thompson, Nicolas G. Bazan; Coordinate Activation of HIF-1 and NF-κB DNA Binding and COX-2 and VEGF Expression in Retinal Cells by Hypoxia. Invest. Ophthalmol. Vis. Sci. 2003;44(10):4163-4170. doi: 10.1167/iovs.02-0655.

      Download citation file:


      © 2016 Association for Research in Vision and Ophthalmology.

      ×
  • Supplements
Abstract

purpose. Proinflammatory signaling mechanisms are implicated in the induction of retinal neovascularization (NV) during ischemic retinopathies. This study examined transcription factor (TF) AP-1, HIF-1, and NF-κB DNA-binding in relation to cyclooxygenase (COX)-2 and VEGF RNA and protein levels in hypoxia-triggered monkey choroidal retinal (RF/6A) endothelial cells. Effects of the carboxamide CGP43182 were tested on COX-2 and VEGF activation and prostaglandin (PG)E2 release.

methods. RF/6A cells were subjected to hypoxia for 1 and 3 hours, at which times RNA and proteins were isolated. Potential AP-1, hypoxia-inducible factor (HIF)-1 and NF-κB DNA-binding sites were identified using DNA sequence search algorithms and were analyzed using gel-shift assay. COX-2 and VEGF RNA, protein, and PGE2 levels were quantified by RT-PCR, Western analysis, and enzyme immunoassay, respectively. Tubular morphogenesis was analyzed with phase-contrast imaging microscopy.

results. Nuclear AP-1, HIF-1 and NF-κB promoter DNA binding increased 1.5-, 4-, and 3-fold, respectively, after 1 hour of hypoxia. COX-2 RNA was elevated five- and fourfold after 1 and 3 hours of hypoxia, respectively. VEGF RNA and protein abundance lagged behind COX-2 induction but were each increased two- to threefold 3 hours after hypoxia. CGP43182 was found to inhibit NF-κB DNA binding, COX-2 and VEGF gene expression, PGE2 release, and hypoxia-induced tubular morphogenesis.

conclusions. Maximum HIF-1 and NF-κB DNA binding immediately before COX-2 expression suggests that these TFs are important regulators of COX-2 induction in hypoxic RF/6A cells. IL-1β emulated AP-1, HIF-1, and NF-κB DNA binding during hypoxia and may be a novel cytokine trigger for NV. CGP43182 appears to be an effective inhibitor of NV. VEGF expression appears to be regulated through dual interdependent mechanisms involving HIF-1 directly and indirectly through NF-κB-mediated COX-2 expression and PGE2 production.

Neovascularization (NV) is orchestrated by the coordinate induction of a family of growth factor genes and most prominently by vascular endothelial growth factor (VEGF). VEGF, one of the most potent angiogenic factors known, also possesses endothelial cell-specific mitogenic effects that closely correlate with NV during embryonic development and normal systemic physiology. However, VEGF also contributes to pathophysiologic conditions in diverse tissue types. For example, VEGF gene transcription spatially and quantitatively correlates with NV in chronic fetal anemia, in malignant tumors, in retinal NV, and in models of hypoxic ischemia, thus suggesting VEGF’s key role in growth, health, and disease. 1 2 3 Hypoxia is thought to be a crucial physiologic stimulus for VEGF upregulation that precedes NV, and low cellular oxygen tension rapidly induces a number of transient genetic signals through which this is accomplished. Primarily, hypoxia induces the activation of hypoxia-inducible factor (HIF)-1, a transcription factor (TF) that provides the means for changes in gene expression in response to alterations in cellular oxygenation. 4 5 6 7 HIF-1 recognizes the minimal cis-acting hypoxia-responsive element (HRE) core DNA sequence, 5′-RCGTG-3′ (R is any purine) that is enriched in the inducible prostaglandin synthase cyclooxygenase (COX)-2, in VEGF, and other oxygen-sensitive gene promoters. 4 5 6 7 HIF-1 plays a role in the upregulation of VEGF in the ischemic retina, where increased levels of HIF-1 DNA binding show temporal and spatial correlation with enhanced VEGF gene transcription. 4 5 7 NF-κB, another oxygen-sensitive TF that recognizes multiple DNA sequences of 5′-GGACTT-3′ in the COX-2 but not in the VEGF gene promoter, 6 8 9 plays an important role in autoimmune, chronic inflammatory, and vascular diseases. NF-κB also contributes to altered hypoxia-induced signal-transduction events through the expression of target genes that encode cytokines and chemokines, such as interleukin (IL)-1, -2, -6, -8, TNFα, inflammatory enzymes such as COX-2, inducible nitric oxide synthase (iNOS), cPLA2 and cell-adhesion molecules. 9 10 11  
Hypoxia also potently triggers COX-2 transcription in the rat lung, 12 in human vascular endothelium, 13 and in human neural cells in primary culture. 14 An important COX-2 product, prostaglandin (PG)E2, is also a strong inducer of VEGF in osteoblastic RCT-3 cells and rheumatoid synovial fibroblasts, 15 in human monocytic THP-1 cells, 16 in the rat lung, 17 and in cultured Müller cells. 18 Compounds that induce COX-2 expression and enzymatic activity also induce VEGF, and those that inhibit COX-2, such as indomethacin or methanesulfonamide, also inhibit the induction of VEGF. 19 These data suggest a key regulatory and potential priming role for COX-2 and PGE2 in the VEGF-expression pathways, vascular remodeling, and NV. In the experiments presented herein, in hypoxia- or IL-1β-treated monkey choroid-retinal endothelial (RF/6A) cells, the oxygen-sensing TFs HIF-1 and NF-κB (but not activator protein [AP]-1) were involved in driving COX-2 gene expression and subsequent VEGF gene induction. CGP43182, an enolized 1,3-dioxane-4,6-dione-5-carboxamide and potent secretory phospholipase (sPL)A2 antagonist 20 also inhibited NF-κB (but not HIF-1) DNA binding, COX-2 and VEGF gene activation, and PGE2 production. The activation of NF-κB, a known positive regulator of COX-2 gene expression, 8 9 10 14 precedes the detection of the maximum level of COX-2 RNA, followed by increased COX-2 protein and PGE2 levels, plus VEGF RNA message and protein abundance after 1 and 3 hours of hypoxia. CGP43182 also significantly reduced tube formation, an early event in angiogenesis, in cultured RF/6A endothelial cells, indicating that NF-κB DNA binding to target gene promoters may be a major molecular genetic event in the mobilization of pathogenic NV processes. These data together suggest that the earliest events before tube formation and NV may be mediated in part through dual interdependent gene expression pathways that involve both COX-2 and VEGF signaling. 
Materials and Methods
RF/6A Endothelial Cells and the Hypoxia Protocol
RF/6A (monkey choroid-retinal) cells (RF/6A cell line CRL-1780; American Type Culture Collection [ATCC], Rockville, MD) were placed in a normobaric chamber (model 116885 CO2 Incubator; Fisher Scientific, Pittsburgh, PA) containing humidified atmosphere and normal filtered air (controls) or in a hypoxic chamber (model MCO-175M Multigas Incubator; Sanyo Scientific, Itasca, IL) with humidified atmosphere. In the latter case, cells were perfused with 90% N2, 5% CO2, and 5% O2 during the period specified (0–3 hours). 21 The hypoxia-treated and age-matched control cells were kept under normoxic conditions (75% N2, 5% CO2, and 20% O2). There were no apparent changes in the gross morphology of the RF/6A cells during or after hypoxia treatment. 
Electrophoretic Mobility-Shift Assay, RT-PCR, and Western Immunoblot Analysis
Cellular and nuclear protein extracts (CPXTs, NPXTs) were rapidly isolated, and concentrations were quantified with a protein microassay (dot-Metric; sensitivity 0.3 ng/μL protein; Chemicon, Temecula, CA), as previously described. 6 8 9 Oligonucleotides containing DNA-bound and mutant consensus sequences of AP-1 (5′-TGACTCA-3′), 8 HIF-1 (5′-CGTG-3′), 4 5 6 7 and NF-κB (5′-GGACTT-3′) 9 10 11 were end labeled by T4 polynucleotide kinase and [γ-32P]adenosine triphosphate (ATP; 3000 Ci/mmol; Redivue; Amersham Pharmacia Biotech, Piscataway, NJ). Electrophoretic mobility-shift assay (EMSA) reactions, using 5 μg of input CPXT or NPXT, were run out on 3.5% or 5% TBE-acrylamide (90 mM Tris-borate [pH 8.4], 0.1 mM EDTA) gels 6 8 and analyzed and quantified at 50-μm resolution with a fluorescent image analyzer (FLA 2000; Fuji, Stamford, CT). Total RNA was isolated with extraction reagent (TRIzol Invitrogen, Carlsbad, CA). For NF-κB DNA binding analysis by EMSA, the p50/p65 and p50/p50 bands were quantified together. 9 11 RNA samples were quantified spectrophotometrically by using RNA chips (LabChips; Caliper Technologies, Mountain View, CA) and a bioanalyzer (model 2100; Agilent Technologies, Foster City, CA). 28S/18S ratios for total RNA were typically more than 1.5. One microgram of each RNA sample was fed directly into integrated RT-PCR reactions (One-Step RT-PCR Superscript System; Invitrogen). Relative levels of human COX-1, COX-2, and β-actin RNA message were quantitated by RT-PCR on 1.5% to 2.0% TBE-agarose gels, as previously described. 6 8 14 VEGF RNA message was quantified according to the methods of Ozaki et al. 22 COX-2 and VEGF protein abundance was determined by using primary antibodies SC-1745 and SC-152, respectively, in 35 μg of total CPXTs run on 20% acrylamide Tris-glycine SDS (TGSDS) gels, 8 22 with Western immunoblot analysis performed according to the manufacturer’s specifications (Santa Cruz Biotechnology, Santa Cruz, CA) and as previously described. 6 8 SC-1745 and SC-152 antibodies are known to be mouse, rat, and human reactive. The VEGF antibody SC-152 is known to recognize each of the 121-, 165- and 189-amino-acid splice variants of VEGF. 
PGE2 Assay
One of the primary prostaglandin products of COX-2 and arachidonic acid (AA), PGE2, was measured with an enzyme immunoassay (EIA) kit, using PGE2-acetylcholinesterase conjugate (PGE2 tracer) with a detection limit of 31 pg/mL (Cayman Chemicals, Ann Arbor, MI). 
Analysis of Tube Formation in RF/6A Cultures
RF/6A cells were grown at 37°C in a humidified atmosphere of normal air (controls) or under conditions of hypoxia (90% N2, 5% CO2, and 5% O2) in F12-K medium supplemented with 5% fetal calf serum (FCS), 100 U/mL penicillin, and 0.1 mg/mL streptomycin. Cells were seeded onto 24-well plates (6 × 104 cells/well) coated with collagen type I. When cells were 90% to 95% confluent, the medium was replaced with fresh F12-K medium containing 1% FCS or 1% FCS containing 0.1 to 1.0 μM CGP43182 and incubated overnight. Cells were then exposed to 10 ng/mL human recombinant IL-1β or to hypoxia for 1 or 3 hours and were compared to cells maintained under normoxic conditions. At these time points CPXTs, NPXTs, total RNA and total cellular protein were prepared as previously described. 6 8 9 To test the effects of CGP43182 on tube formation, plates were then transferred back into a standard incubator, and cells were incubated for an additional 72 hours, at which time five fields per sampling were photographed at 100× magnification by phase-contrast microscopy (Optiphot-2; Nikon Laboratories, Melville, NY). Image-analysis software (AnalySYS 3.0 software; AnalySYS, Münster, Germany) was used to measure the length of tubes in micrometers and the area of tight cellular aggregates in square micrometers on each image obtained by phase-contrast microscopy. In the later analysis, tube lengths were measured (in micrometers) using the “polygon length setting” of the image-analysis software, and the perimeter (in micrometers) and area (in square micrometers) of each aggregate was measured by using the “perimeter/area setting.” All data were saved on spreadsheets (Excel; Microsoft Corp., Redmond, WA). The average of each parameter per image was then calculated, and the average results obtained from five images characterized one well. The mean of eight wells characterized one group of cultured cells. Data on tube lengths and areas of cellular aggregates were statistically analyzed by one-way analysis of variance (ANOVA). 
DNA Sequence Analysis, Data, and Statistical Analysis
DNA sequence and structure analyses identifying putative cis-acting DNA regulatory elements located between −1100 bp and +100 bp of the COX-1, COX-2, and VEGF gene promoters (GenBank accession nos. D64068 [M31812], AF276953, and AF095785, respectively; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD) were performed using sequence analysis software (DNASIS ver. 2.6; Hitachi Genetic Systems, Alameda, CA; and Transfac, ver. 5.2; Biobase, Wolfenbuttel, Germany). 8 14 Functional AP-1, HIF-1- and NF-κB-DNA consensus sequences within the COX-2 and VEGF gene promoters are described in GenBank AF276953, AF095785 and in this study (see Fig. 9 ). EMSA, RT-PCR, Western immunoblot, and EIA data were quantitated by phosphorescence imaging analysis and data-acquisition/statistical-analysis software provided with the molecular imager (model GS250; Bio-Rad, Hercules, CA) or a fluorescence image analyzer (model FLA-2000; Fuji). Tables and figures were generated on computer (Designer ver. 6.0 and Excel ver. 5.0; Microsoft). Statistical significance of EMSA, RT-PCR, or EIA data were analyzed by using a completely randomized design in a two-way factorial analysis of variance. After the finding of a significant overall F-test and a significant interaction F-test, individual interaction means (each represented by a histogram bar in the figures) were compared with protected t-tests. The α levels of these tests were corrected for the number of comparisons made, using a simulation-based multiple-comparison method. 23 All statistical procedures were performed with the programs and procedures in a commercial statistical-analysis program (Statistical Analysis System software; SAS Institute, Cary, NC). 
Results
Effect of Hypoxia on Rapid Gene Signaling in RF/6A Cells
Data in Figures 1 through 6 indicate that the time points of 1 and 3 hours after hypoxia were the most useful in preliminary analyses of the earliest genetic events that involve TF translocation, HIF-1 and NF-κB DNA binding, and signal coupling of these oxygen-sensitive TFs to gene targets in monkey choroid-retinal (RF/6A) nuclei. 6 14 At these time points, TF DNA binding was determined by EMSA, using AP-1–, HIF-1–, and NF-κB-specific mutant and consensus TF DNA target-containing oligonucleotides (see Figs. 1 2 4 ). Gene expression was quantitated for COX-1, COX-2, and VEGF RNA message levels by RT-PCR, using β-actin as an internal control RNA message marker in the same retinal endothelial cell sample (see Figs. 3 5 ). 6 8 9 10 In these sample sets, COX-1, COX-2, and VEGF protein (see Fig. 6 ) and PGE2 levels (see Fig. 7 ) were quantified by Western immunoblot analysis and enzyme-immunoassay (EIA), respectively. 
EMSA of AP-1, HIF-1, and NF-κB DNA Binding and COX-2 and VEGF Gene Expression
The kinetics of the HIF-1 and NF-κB DNA-binding activities correlated both positively and temporally with COX-2 (but not COX-1) RNA message levels, and this was followed by increasing levels of VEGF RNA message in the same sample. Figure 1 shows the levels of AP-1, HIF-1, and NF-κB DNA binding in RF/6A NPXTs. These increased to 1.5-, 4-, and 3-fold, respectively, over time 0 controls, after only 1 hour of hypoxia, suggesting that nuclear HIF-1 and NF-κB, but not AP-1, activations are rapid TF-signaling events. Figure 2 shows the cellular and nuclear protein extracts (CPXT, NPXT, respectively) of HIF-1 in hypoxic RF/6A cells, as determined by EMSA. CPXT levels of AP-1, HIF-1, and NF-κB were only slightly elevated after 3 hours of hypoxia. In contrast, NPXT levels of HIF-1 were sharply elevated after 1 and 3 hours of hypoxia, implying that the major changes in HIF-1 DNA-binding levels are confined to the nuclear compartment during this time course (as are NF-κB-, but not AP-1 DNA-binding profiles; data not shown). Figure 3 shows RT-PCR results for β-actin, COX-1, and COX-2 RNA message levels in RF/6A cells after 1 and 3 hours of hypoxia. The basal order of RNA message abundance was β-actin >>> COX-1 ∼ COX-2 ∼ VEGF RNA at time 0, changing to β-actin >>> COX-2 > VEGF ≫ COX-1 RNA message at 3 hours. COX-2 (but not β-actin or COX-1) RNA message levels rose to five- and fourfold over the control values after 1 and 3 hours of hypoxia, respectively. During this same 1- and 3-hour time course and at 3 hours after hypoxia, the latter being the longest time point reported in the study, the level of VEGF RNA was still increasing. 
Hypoxia Triggering Effect of IL-1β
Figure 4 shows that the levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from hypoxic RF/6A cells were emulated by 1-and 3-hour treatment of normoxic cells with 10 ng/mL of human recombinant IL-1β. As discussed more fully later, application of the proinflammatory cytokine IL-1β has been shown to simulate hypoxia treatment, through induction of both HIF-1 and NF-κB DNA binding in human adult neural cells 14 and in human gingival and synovial fibroblasts, 24 suggesting that the TFs NF-κB and HIF-1 both have roles in inflammatory signaling, possibly in cellular pathways that mediate oxidative homoeostasis. 4 5 6 7 12 13 14 24 25 Patterns of COX-2 and VEGF gene activation, at the RNA and protein levels in RF/6A cells, were observed (Figs. 5 6) after treatment with normobaric hypoxia (90% N2, 5% CO2, and 5% O2) for 1 and 3 hours. Further experiments were performed in parallel with IL-1β-triggered RF/6A cells (data not shown). Comparison of the effects between hypoxia- or IL-1β-treated groups on AP-1, HIF-1, and NF-κB DNA binding and on COX-2 and VEGF gene expression patterns over 1 and 3 hours showed statistically nonsignificant differences (P > 0.05). 
Effects of CGP43182 on TF DNA Binding and COX-1, COX-2, and VEGF Expression
Figure 4 shows that NF-κB (but not AP-1 or HIF-1) DNA binding in hypoxia-triggered RF/6A cells was significantly reduced in the presence of the enolized 1,3-dioxane-4,6-dione-5-carboxamide, CGP43182, a known sPLA2 antagonist (Ottlecz A, et al. IOVS 2001;42:ARVO Abstract 515). 20 In the same cell homogenates, the levels of COX-2 and VEGF (but not that of COX-1 or β-actin) RNA message was reduced by approximately 50% over untreated controls in the presence of CGP43182 (Fig. 5) . In control cells, increases in COX-2 and VEGF RNA message species translated into marked upregulation of COX-2 and VEGF (but not of COX-1) protein abundance in the same sample; however, the levels of COX-2 and VEGF proteins were again markedly attenuated when treated with CGP43182 (Fig. 6) . Increased levels of COX-2 RNA and protein correlated with those of PGE2 levels, as detected by EIA (Figs. 5 6 7) . Further, when cells were pretreated overnight with 1μM CGP43182 and exposed to hypoxic conditions, PGE2 levels were attenuated after 1 and 3 hours of hypoxia when compared with 0-hour control cultures (Fig. 7) . As mentioned, quantitatively similar repressive effects on PGE2 levels were found after 1 and 3 hours of IL-1β treatment in the presence of CGP43182. 
Effects of CGP43182 on Cellular Aggregate and Tube Formation
In the presence of CGP43182 during hypoxia, the area occupied by cellular aggregates was reduced from 79,788 ± 8,759 μm2 to 21,016 ± 3,415 μm2, representing a decrease in cellular aggregate area of 73% (P < 0.01; Fig. 8 ). Tubular morphogenesis, an indicator of NV, was also significantly inhibited at 72 hours after hypoxia. The mean plus 1 SD for tube lengths was 575 ± 92 μm for hypoxia-treated and 212 ± 32 μm for CGP43182+hypoxia-treated cells, when compared with control cultures. This represented a decrease of 63% in tube length in the presence of CGP43182 during hypoxia (P < 0.01; Fig. 8 ). 
Discussion
The present studies focused on the activation and binding kinetics to DNA of AP-1, a composite dimeric TF implicated in growth control and neurodegeneration 8 26 ; NF-κB, a known proinflammatory TF and positive regulator of the COX-2 gene in many cells and tissue types 6 10 11 ; and HIF-1, a master sensor for intracellular oxygen homeostasis and a key transcriptional regulator of hypoxia-sensitive genes. 4 5 7 14 22 27 28 These activation events were analyzed by EMSA, a highly sensitive radiolabel-based electrophoretic technique that can monitor real-time TF-DNA interactions in promoter DNA sequences that immediately precede transcription from that gene. 6 8 9 14 Proximal promoter DNA sequence analysis (−1100 to +100 bp) showed that HIF-1 and NF-κB occupy three sites centered at −442, −190, and −160 bp in the human COX-2 (but not in the human COX-1) gene promoter. 8 9 12 13 14 The human VEGF promoter (GenBank accession no. AF095785) contains no obvious NF-κB DNA-binding consensus sequence in this 1200-bp region, but it has at least 5 HIF-1 DNA-binding sites centered at positions −920, −440, −360, −205, and −80 bp, relative to the start of transcription at +1 3 22 (Fig. 9) . HIF-1- and NF-κB- (but not AP-1) DNA binding increased three- to fourfold after only 1 hour of hypoxia (or IL-1β) treatment (Fig. 1) . This suggests that part of the induction of VEGF by NF-κB DNA binding may be indirect and perhaps is mediated by NF-κB DNA activation of the COX-2 promoter, the latter being a known trigger for COX-2 gene transcription. 8 9 10 11 29 30 31 This would result in subsequent COX-2 gene induction, followed by the activation of VEGF gene expression through upregulation of COX-2-generated PGs, such as PGE2, a known and potent activator of VEGF activity 14 15 16 17 18 19 29 (Figs. 7 9) . In parallel, COX-2, but not β-actin or COX-1, RNA message was increased by several-fold over controls after only 1 hour of hypoxia (or IL-1β) stimulation. These phenomena were accompanied by increasing levels of VEGF RNA message and protein abundance in the same retinal cell samples, even after 3 hours of hypoxia (Figs. 5 6) . We have reported that tube formation and NV are significantly inhibited (26%–44%) by CGP43182 at 0.1 to 1.0 μM (Ottlecz A, et al. IOVS 2001;43:ARVO Abstract 515). In the present experiments, CGP43182 also inhibited NF-κB DNA binding, COX-2 and VEGF RNA and protein abundance, PGE2 levels, and tube formation in hypoxia-treated RF/6A cells (Figs. 4 5 6 7 8) . These data collectively support the hypothesis that NF-κB is one of the key regulators of cellular aggregation and tube formation leading to pathoangiogenesis and NV in hypoxia (or IL-1β)-stimulated retinal-choroid cells. 
Another emerging hypothesis substantiated by the present study is that, like NF-κB, HIF-1 DNA binding is involved in proinflammatory TF signaling and is an important transcriptional activator element in the induction of both the COX-2 and VEGF genes. 6 9 14 24 25 26 27 28 32 33 HIF-1 DNA binding increased at 1 and 3 hours after hypoxia in the NPXTs of RF/6A cells, whereas changes in HIF-1 DNA-binding activities were significantly lower in the CPXT fractions (Fig. 2) . These findings suggest nuclear- and not cytoplasmic-mediated genetic events in HIF-1-regulated gene expression. 4 5 6 7 8 During related experiments in mixed cultures of human neural progenitor cells in primary culture and over a time course of 1 to 3 hours, the inflammatory cytokine peptide IL-1β was also noted to induce significant NF-κB and HIF-1 DNA binding and COX-2 gene expression. 14 In doing so, it simulated the 1- and 3-hour periods of hypoxia treatment in RF/6A cells (Fig. 4) . In further support of this finding, several independent laboratories have recently demonstrated this aspect of HIF-1 activation by hypoxia or IL-1β in human neural primary cells, 14 in human proximal tubular cells, 34 in the human hepatoma cell line HepG2, 35 36 in human fetal astrocytes, 37 and in cultured rat neonatal cardiac myocytes. 38 In fact, in RF/6A cells, treatment with 10 ng/mL of human recombinant IL-1β simulated the hypoxic effect on HIF-1 and NF-κB DNA binding during the time course of 1 to 3 hours after its application (Fig. 4) , and this was also reflected in the upregulation of both COX-2 and VEGF RNA message and protein abundance over the same time course (data not shown). It is therefore becoming increasingly clear that because conditions of hypoxia potently induce IL-1β production and release in a wide variety of cell types, direct addition of IL-1β to the test systems studied would circumvent this first hypoxia- and cytokine-mediated gene activation step. 14 34 35 36 38  
Hypoxia-triggered AP-1, HIF-1, and NF-κB DNA binding was also assessed by EMSA in RF/6A cells pretreated overnight with CGP43182. No significant effects were noted in AP-1 or HIF-1 DNA binding; however, NF-κB DNA binding decreased to approximately 50% of that of hypoxia- or IL-1β-treated controls (Fig. 4) . These latter results suggest that CGP43182 not only prevents the formation of certain AA metabolites produced by COX-2 and certain lipoxygenases by reducing the release of AA from membrane phospholipid stores, but also suppresses the genetic mechanism involving NF-κB activation. The physiologic roles of COX-2 may not be as simple as previously thought, because COX-2 may possess both pro- and anti-inflammatory properties through the generation of different forms of PGs. 8 9 29 As has been widely reported, NF-κB is a major regulator of COX-2 gene transcription, 8 9 10 29 30 31 whereas cyclopentenone PGs can inhibit NF-κB activation and its nuclear accumulation through the inhibition of IκB kinase. 11 20 38 The mode of action of CGP43182 is that it prevents the phosphorylation and proteolytic degradation of the endogenous NF-κB inhibitor (I)κB, a process necessary for NF-κB activation. 11 20 This results in the cellular persistence of the IκB-NF-κB complex and suppression of NF-κB-sensitive gene activation, as we have demonstrated in these experiments (Figs. 4 5 6) . Antioxidants such as pyrrolidine dithiocarbamate have also been shown to prevent NF-κB activation and in doing so suppress tube and capillary formation both in vitro and in vivo in microvascular endothelial cells and in the mouse retina. 29 30 31 The present data also suggest that CGP43182 antagonizes NF-κB DNA binding without exerting any inhibitory effect on that of HIF-1 under the same treatment conditions. Further, overnight pretreatment of RF/6A cells with CGP43182 produced no significant effects on β-actin or COX-1 RNA message induction in comparison to COX-2 and VEGF RNA message levels after 1 and 3 hours of hypoxia (Fig. 5) . Taken together, these data suggest that the anti-tube-forming and antiangiogenic effects of CGP43182 may not be exerted through the direct inhibition of the VEGF gene, but rather through the suppression of NF-κB-driven target gene activation, such as through the COX-2 gene and PGE2 signaling pathways (Figs.5 6 7) . These in turn appear to modulate VEGF expression negatively at both the RNA and protein levels of VEGF and also restrict the cellular aggregation and tube formation associated with angiogenesis and NV (Figs. 8 9)
In summary, HIF-1 and NF-κB should both be considered novel targets for antiangiogenic and anti-NV drugs, because disruption of their binding to proangiogenic promoters may selectively diminish the activation of NF-κB- and HIF-1-driven target genes. These data also indicate that the administration of IL-1β is an effective and a more easily controllable trigger, compared with the application of hypoxia, for studies of HIF-1 and NF-κB DNA binding, COX-2 and VEGF gene expression, and PGE2 generation in RF/6A cells. IL-1β may therefore be useful for the simulation of hypoxia-triggered gene signaling alterations in models of retinal NV and thus for testing the efficacy of pharmacological compounds for the potential management of angiogenesis and NV. Finally, these studies support the idea that VEGF expression may be regulated in part through at least two interdependent mechanisms: one involving HIF-1-VEGF promoter activation directly and the other indirectly through NF-κB-mediated COX-2 gene expression and PGE2 production. 
 
Figure 1.
 
Levels of transcription factors AP-1, HIF-1, and NF-κB DNA binding were quantified in nuclear protein extracts prepared from normoxic RF/6A cells and in cells incubated under hypoxic conditions for 1 and 3 hours. Nuclear binding was analyzed on 3.5% TBE-acrylamide gels. The results are the mean plus 1 SD of results in five replicate cell experiments. Results from the normoxic and hypoxic cells at the different times were compared with a two-way factorial analysis of variance, with multiple comparisons conducted using a simulation-based method for α level correction. 23 *P < 0.03; **P < 0.05.
Figure 1.
 
Levels of transcription factors AP-1, HIF-1, and NF-κB DNA binding were quantified in nuclear protein extracts prepared from normoxic RF/6A cells and in cells incubated under hypoxic conditions for 1 and 3 hours. Nuclear binding was analyzed on 3.5% TBE-acrylamide gels. The results are the mean plus 1 SD of results in five replicate cell experiments. Results from the normoxic and hypoxic cells at the different times were compared with a two-way factorial analysis of variance, with multiple comparisons conducted using a simulation-based method for α level correction. 23 *P < 0.03; **P < 0.05.
Figure 2.
 
Level of transcription factor HIF-1 DNA binding in cellular and nuclear protein extracts (CPXTs, NPXTs) isolated from RF/6A cells subjected to hypoxia for 1 and 3 hours, as determined by EMSA. HIF-1 DNA binding was analyzed on 3.5% TBE-acrylamide gels and the results are the means plus 1 SD of results in three replicate cell experiments. *P < 0.03.
Figure 2.
 
Level of transcription factor HIF-1 DNA binding in cellular and nuclear protein extracts (CPXTs, NPXTs) isolated from RF/6A cells subjected to hypoxia for 1 and 3 hours, as determined by EMSA. HIF-1 DNA binding was analyzed on 3.5% TBE-acrylamide gels and the results are the means plus 1 SD of results in three replicate cell experiments. *P < 0.03.
Figure 3.
 
Levels of RNA message for β-actin, COX-1, COX-2, and VEGF in RF/6A cells incubated under hypoxic conditions for 1 and 3 hours. COX-2 (but not β-actin or COX-1) RNA message was elevated after 1 and 3 hours of hypoxia to five- and fourfold over control cells. VEGF RNA message abundance was still increasing at 3 hours but consistently lagged behind COX-2 RNA message at equivalent time points. RT-PCR products were analyzed on 1.5% TBE-agarose gels, and the results are the mean plus 1 SD of results in three independent experiments. *P < 0.03.
Figure 3.
 
Levels of RNA message for β-actin, COX-1, COX-2, and VEGF in RF/6A cells incubated under hypoxic conditions for 1 and 3 hours. COX-2 (but not β-actin or COX-1) RNA message was elevated after 1 and 3 hours of hypoxia to five- and fourfold over control cells. VEGF RNA message abundance was still increasing at 3 hours but consistently lagged behind COX-2 RNA message at equivalent time points. RT-PCR products were analyzed on 1.5% TBE-agarose gels, and the results are the mean plus 1 SD of results in three independent experiments. *P < 0.03.
Figure 4.
 
Left panel: Levels of transcription factor AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from normoxic RF/6A cells and in cells incubated under hypoxic conditions for 1 and 3 hours. Middle panel: Similar alterations in the levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from hypoxic cells were achieved with the addition of 10 ng/mL of IL-1β to normoxic cells incubated over the same time points. The results shown are from four independent experiments. Significance over 0-hour control cells: *P < 0.03; **P < 0.05. Right panel: Levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from RF/6A cells incubated under hypoxic conditions for 1 and 3 hours after overnight pretreatment with 1 μM CGP43182. NF-κB, but neither AP-1 nor HIF-1, DNA binding was significantly reduced in the presence of CGP43182. TBE-acrylamide gels (5%). The results shown are from four independent experiments; *P < 0.03.
Figure 4.
 
Left panel: Levels of transcription factor AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from normoxic RF/6A cells and in cells incubated under hypoxic conditions for 1 and 3 hours. Middle panel: Similar alterations in the levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from hypoxic cells were achieved with the addition of 10 ng/mL of IL-1β to normoxic cells incubated over the same time points. The results shown are from four independent experiments. Significance over 0-hour control cells: *P < 0.03; **P < 0.05. Right panel: Levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from RF/6A cells incubated under hypoxic conditions for 1 and 3 hours after overnight pretreatment with 1 μM CGP43182. NF-κB, but neither AP-1 nor HIF-1, DNA binding was significantly reduced in the presence of CGP43182. TBE-acrylamide gels (5%). The results shown are from four independent experiments; *P < 0.03.
Figure 5.
 
RNA message levels for β-actin, COX-1, COX-2, and VEGF in normoxic RF/6A cells (0 hour) and in cells incubated under hypoxic conditions for 1 and 3 hours. After 1 hour, the levels of COX-2 and VEGF RNA message were reduced 1.6- and 1.9-fold, respectively, in the presence of the NF-κB DNA binding inhibitor CGP43182. No significant effects were noted on either β-actin or COX-1 RNA message levels (data for β-actin RNA message not shown). TBE-agarose gels (2%). The results shown are the mean plus 1 SD of results in three independent experiments. *P < 0.03; **P < 0.05.
Figure 5.
 
RNA message levels for β-actin, COX-1, COX-2, and VEGF in normoxic RF/6A cells (0 hour) and in cells incubated under hypoxic conditions for 1 and 3 hours. After 1 hour, the levels of COX-2 and VEGF RNA message were reduced 1.6- and 1.9-fold, respectively, in the presence of the NF-κB DNA binding inhibitor CGP43182. No significant effects were noted on either β-actin or COX-1 RNA message levels (data for β-actin RNA message not shown). TBE-agarose gels (2%). The results shown are the mean plus 1 SD of results in three independent experiments. *P < 0.03; **P < 0.05.
Figure 6.
 
Left panel: Abundance of COX-1, COX-2, and VEGF protein isolated from normoxic RF/6A cells (0 hour) and in RF/6A cells incubated under hypoxic conditions for 1 and 3 hours. Right panel: Equivalent to left panel, except after overnight treatment with 1 μM CGP43182. TGSDS gels (20%). The results are the means plus 1 SD of results in three independent experiments. *P < 0.03; **P < 0.05.
Figure 6.
 
Left panel: Abundance of COX-1, COX-2, and VEGF protein isolated from normoxic RF/6A cells (0 hour) and in RF/6A cells incubated under hypoxic conditions for 1 and 3 hours. Right panel: Equivalent to left panel, except after overnight treatment with 1 μM CGP43182. TGSDS gels (20%). The results are the means plus 1 SD of results in three independent experiments. *P < 0.03; **P < 0.05.
Figure 7.
 
Prostaglandin E2 (PGE2) release in normoxic RF/6A cells (0 hour) and in cells incubated under hypoxic conditions for 1 and 3 hours and in the absence and presence of 1 μM CGP43182. The results represent the means plus 1 SD of results in three independent experiments. **P < 0.05.
Figure 7.
 
Prostaglandin E2 (PGE2) release in normoxic RF/6A cells (0 hour) and in cells incubated under hypoxic conditions for 1 and 3 hours and in the absence and presence of 1 μM CGP43182. The results represent the means plus 1 SD of results in three independent experiments. **P < 0.05.
Figure 8.
 
Typical cord and cellular aggregate formation in RF/6A cells. Cells were exposed to 3 hours of hypoxia and then to 72 hours of normoxia. (A) Normoxic control with no exposure to hypoxia, where no tube formation and only a few small cellular aggregates were apparent. Arrowhead: cellular aggregate. The hypoxic condition (B) reflects the typical situation of cellular aggregate and tube formation. Arrow: formed tube. The CGP43182-treated hypoxic condition shown in (C) demonstrates the inhibitory effect of CGP43182 (1 μM) on tube formation. Arrow: small formed tube. Phase-contrast photomicrographs. Magnification, ×100; scale bar, 400 μm.
Figure 8.
 
Typical cord and cellular aggregate formation in RF/6A cells. Cells were exposed to 3 hours of hypoxia and then to 72 hours of normoxia. (A) Normoxic control with no exposure to hypoxia, where no tube formation and only a few small cellular aggregates were apparent. Arrowhead: cellular aggregate. The hypoxic condition (B) reflects the typical situation of cellular aggregate and tube formation. Arrow: formed tube. The CGP43182-treated hypoxic condition shown in (C) demonstrates the inhibitory effect of CGP43182 (1 μM) on tube formation. Arrow: small formed tube. Phase-contrast photomicrographs. Magnification, ×100; scale bar, 400 μm.
Figure 9.
 
Dual, interdependent mechanisms for COX-2 and VEGF gene activation in hypoxia- or IL-1β-triggered retinal endothelial cells: proposed scheme for the temporal activation of COX-2 followed by VEGF gene activation. The −1100- to +100-bp 5′ flanking region from the immediate COX-2 and VEGF gene promoters were analyzed with GenBank AF276953 and AF095785, respectively. As these and previous studies indicate, there is a large hypoxia- or IL-1β-induced increase in COX-2 gene activation that immediately precedes VEGF gene expression. This fact suggests at least a partial mediation of VEGF expression through prostaglandin (PG), the rapidly synthesized primary catalytic product of COX-2, in addition to VEGF activation by the hypoxia-generated TF HIF-1. Transcription factor IID (TFIID) binding at the TATA box is recognized by the RNA polymerase II basal transcription complex. 8 14 Actual positions of AP-1, HIF, and NF-κB DNA binding sites may be slightly skewed due to space constraints. (⊕) A positively acting regulatory element; dashed lines: putative pathways and sites of interaction for CGP43182.
Figure 9.
 
Dual, interdependent mechanisms for COX-2 and VEGF gene activation in hypoxia- or IL-1β-triggered retinal endothelial cells: proposed scheme for the temporal activation of COX-2 followed by VEGF gene activation. The −1100- to +100-bp 5′ flanking region from the immediate COX-2 and VEGF gene promoters were analyzed with GenBank AF276953 and AF095785, respectively. As these and previous studies indicate, there is a large hypoxia- or IL-1β-induced increase in COX-2 gene activation that immediately precedes VEGF gene expression. This fact suggests at least a partial mediation of VEGF expression through prostaglandin (PG), the rapidly synthesized primary catalytic product of COX-2, in addition to VEGF activation by the hypoxia-generated TF HIF-1. Transcription factor IID (TFIID) binding at the TATA box is recognized by the RNA polymerase II basal transcription complex. 8 14 Actual positions of AP-1, HIF, and NF-κB DNA binding sites may be slightly skewed due to space constraints. (⊕) A positively acting regulatory element; dashed lines: putative pathways and sites of interaction for CGP43182.
The authors thank Jim Hill, Department of Ophthalmology, Louisiana State University Eye Center (New Orleans, LA), and Gregg Semenza, Institute of Genetic Medicine, Johns Hopkins University (Baltimore, MD), for critical evaluation of the manuscript and Darlene Guillot, Carole Lachney, and Aileen Pogue for expert technical assistance. 
Larrivee, B, Karsan, A. (2000) Signaling pathways induced by vascular endothelial growth factor (review) Int J Mol Med 5,447-456 [PubMed]
Hasan, J, Jayson, GC. (2001) VEGF antagonists Expert Opin Biol Ther 1,703-718 [CrossRef] [PubMed]
Giles, FJ. (2001) The vascular endothelial growth factor (VEGF) signaling pathway: a therapeutic target in patients with hematologic malignancies Oncologist 6(suppl 5),32-39 [CrossRef] [PubMed]
Semenza, GL. (2001) HIF-1, O(2), and the 3 PHDs: how animal cells signal hypoxia to the nucleus Cell 107,1-3 [CrossRef] [PubMed]
Semenza, GL. (2000) HIF-1: mediator of physiological and pathophysiological responses to hypoxia J Appl Physiol 88,1474-1480 [PubMed]
Lukiw, WJ, Gordon, WC, Rogaev, EI, Thompson, H, Bazan, NG. (2001) Presenilin-2 (PS2) expression upregulation in a model of retinopathy of prematurity and pathoangiogenesis Neuroreport 12,53-57 [CrossRef] [PubMed]
Semenza, GL. (1998) Hypoxia-inducible factor 1: master regulator of O2 homeostasis Curr Opin Genet Dev 8,588-594 [CrossRef] [PubMed]
Lukiw, WJ, Pelaez, RP, Martinez, J, Bazan, NG. (1998) Budesonide epimer R or dexamethasone selectively inhibit platelet-activating factor-induced or interleukin 1beta-induced DNA binding activity of cis-acting transcription factors and cyclooxygenase-2 gene expression Proc Natl Acad Sci USA 95,3914-3919 [CrossRef] [PubMed]
Lukiw, WJ, Bazan, NG. (1998) Strong nuclear factor-kappa B-DNA binding parallels cyclooxygenase-2 gene transcription in aging and in sporadic Alzheimer’s disease superior temporal lobe neocortex J Neurosci Res 53,583-592 [CrossRef] [PubMed]
Epstein, FH. (1997) Nuclear factor-κB: a pivotal transcription factor in chronic inflammatory diseases N Engl J Med 336,1066-1071 [CrossRef] [PubMed]
Baeuerle, PA, Baltimore, D. (1996) NF-κB: Ten years after Cell 87,13-20 [CrossRef] [PubMed]
Hierholzer, C, Harbrecht, BG, Billiar, TR, Tweardy, DJ. (2001) Hypoxia-inducible factor-1 activation and cyclo-oxygenase-2 induction are early reperfusion-independent inflammatory events in hemorrhagic shock Arch Orthop Trauma Surg 121,219-222 [CrossRef] [PubMed]
Schmedtje, JF, Jr, Ji, YS, Liu, WL, DuBois, RN, Runge, MS. (1997) Hypoxia induces cyclooxygenase-2 via the NF-kappa B p65 transcription factor in human vascular endothelial cells J Biol Chem 272,601-608 [CrossRef] [PubMed]
Lukiw, WJ, Bazan, NG. (2002) COX-2 and PS1 gene expression induced by IL-1beta and Abeta 42 is potentiated by hypoxia in primary human neural cells J Biol Chem 277,30359-30367 [CrossRef] [PubMed]
Inoue, H, Takamori, M, Nagata, N, et al (2001) An investigation of cell proliferation and soluble mediators induced by interleukin 1 beta in human synovial fibroblasts: comparative response in osteoarthritis and rheumatoid arthritis Inflamm Res 50,65-72 [CrossRef] [PubMed]
Hoper, MM, Voelkel, NF, Bates, TO, et al (1997) Prostaglandins induce vascular endothelial growth factor in a human monocytic cell line and rat lungs via cAMP Am J Respir Cell Mol Biol 17,748-756 [CrossRef] [PubMed]
Voelkel, NF, Hoeper, M, Maloney, J, Tuder, RM. (1996) Vascular endothelial growth factor in pulmonary hypertension Ann NY Acad Sci 796,186-193 [CrossRef] [PubMed]
Cheng, T, Cao, W, Wen, R, Steinberg, RH, LaVail, MM. (1998) Prostaglandin E2 induces vascular endothelial growth factor and basic fibroblast growth factor mRNA expression in cultured rat Müller cells Invest Ophthalmol Vis Sci 39,581-591 [PubMed]
Nie, D, Lamberti, M, Zacharek, A, et al (2000) Thromboxane A(2) regulation of endothelial cell migration, angiogenesis, and tumor metastasis Biochem Biophys Res Commun 267,245-251 [CrossRef] [PubMed]
Scholz-Pedretti, K, Eberhardt, W, Rupprecht, G, et al (2000) Inhibition of NFkappaB-mediated pro-inflammatory gene expression in rat mesangial cells by the enolized 1,3-dioxane-4,6-dione-5-carboxamide, CGP-43182 Br J Pharmacol 130,1183-1190 [CrossRef] [PubMed]
Mukhopadhyay, D, Tsiokas, L, Zhou, X-M, Foster, D, Brugge, JS, Sukhatme, VP. (1995) Hypoxic induction of human vascular endothelial growth factor expression through c-Src activation Nature 375,577-581 [CrossRef] [PubMed]
Ozaki, H, Yu, AY, Della, N, et al (1999) Hypoxia inducible factor-1α is increased in ischemic retina: temporal and spatial correlation with VEGF expression Invest Ophthalmol Vis Sci 40,182-189 [PubMed]
Edwards, D, Berry, JJ. (1987) The efficiency of simulation-based multiple comparisons Biometrics 43,913-928 [CrossRef] [PubMed]
Thornton, RD, Lane, P, Borghaei, RC, Pease, EA, Caro, J, Mochan, E. (2000) Interleukin-1 induces hypoxia-inducible factor 1 in human gingival and synovial fibroblasts Biochem J 350,307-312 [CrossRef] [PubMed]
Bazan, NG, Gordon, WC, DeCoster, MA, Mukherjee, PK. (1997) Experimental models and their use in studies of diabetic retinal microangiopathy Therapie 52,447-451 [PubMed]
Imler, JL, Wasylyk, B. (1989) AP1, a composite transcription factor implicated in abnormal growth control Prog Growth Factor Res 1,69-77 [CrossRef] [PubMed]
Royer, C, Lachuer, J, Crouzoulon, G, et al (2000) Effects of gestational hypoxia on mRNA levels of Glut3 and Glut4 transporters, hypoxia inducible factor-1 and thyroid hormone receptors in developing rat brain Brain Res 856,119-128 [CrossRef] [PubMed]
Jung, F, Palmer, LA, Zhou, N, Johns, RA. (2000) Hypoxic regulation of inducible nitric oxide synthase via hypoxia inducible factor-1 in cardiac myocytes Circ Res 86,319-325 [CrossRef] [PubMed]
Poligone, B, Baldwin, AS. (2001) Positive and negative regulation of NF-kappaB by COX-2: roles of different prostaglandins J Biol Chem 276,38658-38664 [CrossRef] [PubMed]
Stoltz, RA, Abraham, NG, Schwartzmann, ML. (1996) The role of NF-κB in the angiogenic response of coronary microvessel endothelial cells Proc Natl Acad Sci USA 93,2832-2837 [CrossRef] [PubMed]
Yoshida, A, Yoshida, S, Ishibashi, T, Kuwano, M, Inomata, H. (1998) Suppression of retinal neovascularization by the NF-κB inhibitor pyrrolidine dithiocarbamate in mice Invest Ophthalmol Vis Sci 39,1624-1629
Marti, HJ, Bernaudin, M, Bellail, A, et al (2000) Hypoxia-induced vascular endothelial growth factor expression precedes neovascularization after cerebral ischemia Am J Pathol 156,965-976 [CrossRef] [PubMed]
O’Rourke, JF, Pugh, CW, Bartlett, SM, Ratcliffe, PJ. (1996) Identification of hypoxically inducible mRNAs in HeLa cells using differential-display PCR: role of hypoxia-inducible factor-1 Eur J Biochem 241,403-410 [CrossRef] [PubMed]
El Awad, B, Kreft, B, Wolber, EM, et al (2000) Hypoxia and interleukin-1 beta stimulate vascular endothelial growth factor production in human proximal tubular cells Kidney Int 58,43-50 [CrossRef] [PubMed]
Hellwig-Burgel, T, Rutkowski, K, Metzen, E, Fandrey, J, Jelkmann, W. (1999) Interleukin-1 beta and tumor necrosis factor-alpha stimulate DNA binding of hypoxia-inducible factor-1 Blood 94,1561-1567 [PubMed]
Stiehl, DP, Jelkmann, W, Wenger, RH, Hellwig-Burgel, T. (2002) Normoxic induction of the hypoxia-inducible factor 1alpha by insulin and interleukin-1beta involves the phosphatidylinositol 3-kinase pathway FEBS Lett 512,157-162 [CrossRef] [PubMed]
Stanimirovic, D, Zhang, W, Howlett, C, Lemieux, P, Smith, C. (2001) Inflammatory gene transcription in human astrocytes exposed to hypoxia: roles of the nuclear factor-kappaB and autocrine stimulation J Neuroimmunol 119,365-376 [CrossRef] [PubMed]
Petrova, TV, Akama, KT, Van Eldik, LJ. (1999) Cyclopentenone prostaglandins suppress activation of microglia: down-regulation of inducible nitric-oxide synthase by 15-deoxy-Delta12,14-prostaglandin J2 Proc Natl Acad Sci USA 96,4668-4673 [CrossRef] [PubMed]
Figure 1.
 
Levels of transcription factors AP-1, HIF-1, and NF-κB DNA binding were quantified in nuclear protein extracts prepared from normoxic RF/6A cells and in cells incubated under hypoxic conditions for 1 and 3 hours. Nuclear binding was analyzed on 3.5% TBE-acrylamide gels. The results are the mean plus 1 SD of results in five replicate cell experiments. Results from the normoxic and hypoxic cells at the different times were compared with a two-way factorial analysis of variance, with multiple comparisons conducted using a simulation-based method for α level correction. 23 *P < 0.03; **P < 0.05.
Figure 1.
 
Levels of transcription factors AP-1, HIF-1, and NF-κB DNA binding were quantified in nuclear protein extracts prepared from normoxic RF/6A cells and in cells incubated under hypoxic conditions for 1 and 3 hours. Nuclear binding was analyzed on 3.5% TBE-acrylamide gels. The results are the mean plus 1 SD of results in five replicate cell experiments. Results from the normoxic and hypoxic cells at the different times were compared with a two-way factorial analysis of variance, with multiple comparisons conducted using a simulation-based method for α level correction. 23 *P < 0.03; **P < 0.05.
Figure 2.
 
Level of transcription factor HIF-1 DNA binding in cellular and nuclear protein extracts (CPXTs, NPXTs) isolated from RF/6A cells subjected to hypoxia for 1 and 3 hours, as determined by EMSA. HIF-1 DNA binding was analyzed on 3.5% TBE-acrylamide gels and the results are the means plus 1 SD of results in three replicate cell experiments. *P < 0.03.
Figure 2.
 
Level of transcription factor HIF-1 DNA binding in cellular and nuclear protein extracts (CPXTs, NPXTs) isolated from RF/6A cells subjected to hypoxia for 1 and 3 hours, as determined by EMSA. HIF-1 DNA binding was analyzed on 3.5% TBE-acrylamide gels and the results are the means plus 1 SD of results in three replicate cell experiments. *P < 0.03.
Figure 3.
 
Levels of RNA message for β-actin, COX-1, COX-2, and VEGF in RF/6A cells incubated under hypoxic conditions for 1 and 3 hours. COX-2 (but not β-actin or COX-1) RNA message was elevated after 1 and 3 hours of hypoxia to five- and fourfold over control cells. VEGF RNA message abundance was still increasing at 3 hours but consistently lagged behind COX-2 RNA message at equivalent time points. RT-PCR products were analyzed on 1.5% TBE-agarose gels, and the results are the mean plus 1 SD of results in three independent experiments. *P < 0.03.
Figure 3.
 
Levels of RNA message for β-actin, COX-1, COX-2, and VEGF in RF/6A cells incubated under hypoxic conditions for 1 and 3 hours. COX-2 (but not β-actin or COX-1) RNA message was elevated after 1 and 3 hours of hypoxia to five- and fourfold over control cells. VEGF RNA message abundance was still increasing at 3 hours but consistently lagged behind COX-2 RNA message at equivalent time points. RT-PCR products were analyzed on 1.5% TBE-agarose gels, and the results are the mean plus 1 SD of results in three independent experiments. *P < 0.03.
Figure 4.
 
Left panel: Levels of transcription factor AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from normoxic RF/6A cells and in cells incubated under hypoxic conditions for 1 and 3 hours. Middle panel: Similar alterations in the levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from hypoxic cells were achieved with the addition of 10 ng/mL of IL-1β to normoxic cells incubated over the same time points. The results shown are from four independent experiments. Significance over 0-hour control cells: *P < 0.03; **P < 0.05. Right panel: Levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from RF/6A cells incubated under hypoxic conditions for 1 and 3 hours after overnight pretreatment with 1 μM CGP43182. NF-κB, but neither AP-1 nor HIF-1, DNA binding was significantly reduced in the presence of CGP43182. TBE-acrylamide gels (5%). The results shown are from four independent experiments; *P < 0.03.
Figure 4.
 
Left panel: Levels of transcription factor AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from normoxic RF/6A cells and in cells incubated under hypoxic conditions for 1 and 3 hours. Middle panel: Similar alterations in the levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from hypoxic cells were achieved with the addition of 10 ng/mL of IL-1β to normoxic cells incubated over the same time points. The results shown are from four independent experiments. Significance over 0-hour control cells: *P < 0.03; **P < 0.05. Right panel: Levels of AP-1, HIF-1, and NF-κB DNA binding in NPXTs isolated from RF/6A cells incubated under hypoxic conditions for 1 and 3 hours after overnight pretreatment with 1 μM CGP43182. NF-κB, but neither AP-1 nor HIF-1, DNA binding was significantly reduced in the presence of CGP43182. TBE-acrylamide gels (5%). The results shown are from four independent experiments; *P < 0.03.
Figure 5.
 
RNA message levels for β-actin, COX-1, COX-2, and VEGF in normoxic RF/6A cells (0 hour) and in cells incubated under hypoxic conditions for 1 and 3 hours. After 1 hour, the levels of COX-2 and VEGF RNA message were reduced 1.6- and 1.9-fold, respectively, in the presence of the NF-κB DNA binding inhibitor CGP43182. No significant effects were noted on either β-actin or COX-1 RNA message levels (data for β-actin RNA message not shown). TBE-agarose gels (2%). The results shown are the mean plus 1 SD of results in three independent experiments. *P < 0.03; **P < 0.05.
Figure 5.
 
RNA message levels for β-actin, COX-1, COX-2, and VEGF in normoxic RF/6A cells (0 hour) and in cells incubated under hypoxic conditions for 1 and 3 hours. After 1 hour, the levels of COX-2 and VEGF RNA message were reduced 1.6- and 1.9-fold, respectively, in the presence of the NF-κB DNA binding inhibitor CGP43182. No significant effects were noted on either β-actin or COX-1 RNA message levels (data for β-actin RNA message not shown). TBE-agarose gels (2%). The results shown are the mean plus 1 SD of results in three independent experiments. *P < 0.03; **P < 0.05.
Figure 6.
 
Left panel: Abundance of COX-1, COX-2, and VEGF protein isolated from normoxic RF/6A cells (0 hour) and in RF/6A cells incubated under hypoxic conditions for 1 and 3 hours. Right panel: Equivalent to left panel, except after overnight treatment with 1 μM CGP43182. TGSDS gels (20%). The results are the means plus 1 SD of results in three independent experiments. *P < 0.03; **P < 0.05.
Figure 6.
 
Left panel: Abundance of COX-1, COX-2, and VEGF protein isolated from normoxic RF/6A cells (0 hour) and in RF/6A cells incubated under hypoxic conditions for 1 and 3 hours. Right panel: Equivalent to left panel, except after overnight treatment with 1 μM CGP43182. TGSDS gels (20%). The results are the means plus 1 SD of results in three independent experiments. *P < 0.03; **P < 0.05.
Figure 7.
 
Prostaglandin E2 (PGE2) release in normoxic RF/6A cells (0 hour) and in cells incubated under hypoxic conditions for 1 and 3 hours and in the absence and presence of 1 μM CGP43182. The results represent the means plus 1 SD of results in three independent experiments. **P < 0.05.
Figure 7.
 
Prostaglandin E2 (PGE2) release in normoxic RF/6A cells (0 hour) and in cells incubated under hypoxic conditions for 1 and 3 hours and in the absence and presence of 1 μM CGP43182. The results represent the means plus 1 SD of results in three independent experiments. **P < 0.05.
Figure 8.
 
Typical cord and cellular aggregate formation in RF/6A cells. Cells were exposed to 3 hours of hypoxia and then to 72 hours of normoxia. (A) Normoxic control with no exposure to hypoxia, where no tube formation and only a few small cellular aggregates were apparent. Arrowhead: cellular aggregate. The hypoxic condition (B) reflects the typical situation of cellular aggregate and tube formation. Arrow: formed tube. The CGP43182-treated hypoxic condition shown in (C) demonstrates the inhibitory effect of CGP43182 (1 μM) on tube formation. Arrow: small formed tube. Phase-contrast photomicrographs. Magnification, ×100; scale bar, 400 μm.
Figure 8.
 
Typical cord and cellular aggregate formation in RF/6A cells. Cells were exposed to 3 hours of hypoxia and then to 72 hours of normoxia. (A) Normoxic control with no exposure to hypoxia, where no tube formation and only a few small cellular aggregates were apparent. Arrowhead: cellular aggregate. The hypoxic condition (B) reflects the typical situation of cellular aggregate and tube formation. Arrow: formed tube. The CGP43182-treated hypoxic condition shown in (C) demonstrates the inhibitory effect of CGP43182 (1 μM) on tube formation. Arrow: small formed tube. Phase-contrast photomicrographs. Magnification, ×100; scale bar, 400 μm.
Figure 9.
 
Dual, interdependent mechanisms for COX-2 and VEGF gene activation in hypoxia- or IL-1β-triggered retinal endothelial cells: proposed scheme for the temporal activation of COX-2 followed by VEGF gene activation. The −1100- to +100-bp 5′ flanking region from the immediate COX-2 and VEGF gene promoters were analyzed with GenBank AF276953 and AF095785, respectively. As these and previous studies indicate, there is a large hypoxia- or IL-1β-induced increase in COX-2 gene activation that immediately precedes VEGF gene expression. This fact suggests at least a partial mediation of VEGF expression through prostaglandin (PG), the rapidly synthesized primary catalytic product of COX-2, in addition to VEGF activation by the hypoxia-generated TF HIF-1. Transcription factor IID (TFIID) binding at the TATA box is recognized by the RNA polymerase II basal transcription complex. 8 14 Actual positions of AP-1, HIF, and NF-κB DNA binding sites may be slightly skewed due to space constraints. (⊕) A positively acting regulatory element; dashed lines: putative pathways and sites of interaction for CGP43182.
Figure 9.
 
Dual, interdependent mechanisms for COX-2 and VEGF gene activation in hypoxia- or IL-1β-triggered retinal endothelial cells: proposed scheme for the temporal activation of COX-2 followed by VEGF gene activation. The −1100- to +100-bp 5′ flanking region from the immediate COX-2 and VEGF gene promoters were analyzed with GenBank AF276953 and AF095785, respectively. As these and previous studies indicate, there is a large hypoxia- or IL-1β-induced increase in COX-2 gene activation that immediately precedes VEGF gene expression. This fact suggests at least a partial mediation of VEGF expression through prostaglandin (PG), the rapidly synthesized primary catalytic product of COX-2, in addition to VEGF activation by the hypoxia-generated TF HIF-1. Transcription factor IID (TFIID) binding at the TATA box is recognized by the RNA polymerase II basal transcription complex. 8 14 Actual positions of AP-1, HIF, and NF-κB DNA binding sites may be slightly skewed due to space constraints. (⊕) A positively acting regulatory element; dashed lines: putative pathways and sites of interaction for CGP43182.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×