MCB
Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Other Versions of this Article:
MCB.00052-07v1
27/12/4207    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Copyright Information
Right arrow Books from ASM Press
Right arrow MicrobeWorld
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bandyopadhyay, S.
Right arrow Articles by DiCorleto, P. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bandyopadhyay, S.
Right arrow Articles by DiCorleto, P. E.
Molecular and Cellular Biology, June 2007, p. 4207-4216, Vol. 27, No. 12
0270-7306/07/$08.00+0     doi:10.1128/MCB.00052-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.

HOXA9 Participates in the Transcriptional Activation of E-Selectin in Endothelial Cells{triangledown}

Smarajit Bandyopadhyay, Mohammad Z. Ashraf, Pamela Daher, Philip H. Howe, and Paul E. DiCorleto*

Department of Cell Biology, Lerner Research Institute and Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, Cleveland Clinic Foundation, Cleveland, Ohio 44195

Received 10 January 2007/ Returned for modification 14 February 2007/ Accepted 3 April 2007


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The homeobox gene HOXA9 has recently been shown to be an important regulator of endothelial cell (EC) differentiation and activation in addition to its role in embryonic development and hematopoiesis. In this report, we have determined that the EC-leukocyte adhesion molecule E-selectin is a key target for HOXA9. The depletion of HOXA9 protein in ECs resulted in a significant and specific decrease in tumor necrosis factor alpha (TNF-{alpha})-induced E-selectin gene expression. In addition, HOXA9 specifically activated the E-selectin gene promoter in ECs. Progressive deletional analyses together with site-specific mutagenesis of the E-selectin promoter indicated that the Abd-B-like HOX DNA-binding motif, CAATTTTATTAA, located in the proximal region spanning bp –210 to –221 upstream of the transcription start site was crucial for the promoter induction by HOXA9. Both HOXA9 in EC nuclear extract and recombinant HOXA9 protein bound to this sequence in vitro. Moreover, we showed that HOXA9 binds temporally, in a TNF-{alpha}-dependent manner, to the region containing this Abd-B-like element in vivo. We have thus identified a novel and functionally critical cis-regulatory element for TNF-{alpha}-mediated transient expression of the E-selectin gene. Further, we provide evidence that HOXA9 acts as an obligate proinflammatory factor by mediating cytokine induction of E-selectin.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The endothelium, comprising the inner lining of blood vessels, plays a key role in maintaining vascular homeostasis. However, in response to inflammation and during cell-mediated immune responses, quiescent endothelial cells (EC) become activated, synthesizing new growth regulatory and vasoreactive molecules absent in resting EC (12, 15). One such proinflammatory molecule, E-selectin, an EC-leukocyte adhesion molecule, is an inducible glycoprotein that is expressed exclusively and transiently on the surface of activated EC in response to various agonists, including tumor necrosis factor alpha (TNF-{alpha}), interleukin-1, lipopolysaccharides, and thrombin (4, 46, 47). In addition to taking part in the complex process of leukocyte recruitment (25), E-selectin also promotes metastasis by enhancing the adhesion of circulating tumor cells to the endothelium (28).

Promoter analysis studies have established the importance of the transcription factors NF-{kappa}B, ATF-2/Jun, and high-mobility-group protein I(Y) [HMG-I(Y)] in the transient induction of the E-selectin gene in EC (26, 34, 42). In addition, EC-specific histone modifications and chromatin remodeling at the E-selectin promoter have recently been implicated in the induction process (14). However, the kinetics of E-selectin gene expression reveal that even in the continuous presence of TNF-{alpha}, mRNA levels are transiently maintained for 8 h following transcriptional initiation, though NF-{kappa}B and ATF-2 proteins are still present in the nucleus at 24 h post-cytokine induction (6). Based on these observations, it is postulated that continuous stimulation of EC may lead to down-regulation and/or posttranslational modification of other transcriptional activators/coactivators necessary for the induction of the E-selectin gene.

Previous in vitro studies have implicated several members of the HOX gene cluster of homeobox transcription factors in the expression of cell adhesion molecules, such as E-cadherin in activated T cells, neural cell adhesion molecule, and liver cell adhesion molecule (19, 24, 33). However, little is known regarding the role of this gene family in the induction of EC-leukocyte adhesion molecules during EC activation. The highly homologous HOX genes are known to contribute significantly in specifying body plan and cell fate during development and in hematopoiesis (8, 9, 18, 21, 22, 30). They have also been reported to play a regulatory role in vascular cell differentiation and proliferation (3, 20). For example, studies have described the important contributions of HOX genes in establishing the angiogenic phenotype in EC (5, 39) and have identified HOXB5 as a potential transcriptional regulator of EC differentiation from mesoderm-derived precursor cells (54).

HOXA9, one of the Abd-B HOX family members, is required for normal development of the vertebral lumbar region and has also been implicated both in normal hematopoiesis and in leukemogenesis (17, 29, 32, 50, 52, 55). HOXA9 has been reported to function as a transcription factor, both alone and as a multiprotein complex with binding partner Pbx and/or Meis homeodomain protein (31, 41, 48). Moreover, posttranslational modification of HOXA9 protein by phosphorylation has been reported to modulate its DNA-binding activity and HOXA9-mediated gene transcription during myeloid differentiation (2, 51). HOXA9 expression levels are not only transcriptionally down-regulated in differentiated and mature hematopoietic cells (32, 44) but also posttranslationally down-regulated through a ubiquitin-dependent proteolytic mechanism (58). Recently, HOXA9 has been identified as a key regulator of endothelial commitment of adult progenitor cells (43). In addition, in previous studies, we have demonstrated a rapid down-regulation of the message level of one of the two splice variants (27) of the HOXA9 gene, HOXA9-EC/B in EC in response to stimulation with TNF-{alpha}, suggesting that HOXA9 is a potential regulator of EC homeostasis (41). HOXA9 has also recently been shown to exert its proangiogenic activity in EC by modulating the expression of EphB4 receptor tyrosine kinase (7, 13).

In the current study, we present evidence that HOXA9 in EC is required for the transcriptional induction of E-selectin by TNF-{alpha}, and this induction is mediated through the binding of HOXA9 to a specific Abd-B-like sequence, located at bp –210 to –221 from the transcription start site. These results are the first to reveal a proinflammatory role of HOXA9 in the induction of an "activation gene" in EC.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
cDNA construction and protein expression. Both full-length HOXA9 and Hoxa2 cDNAs were subcloned into pcDNA3 vector (Invitrogen) for expression in mammalian cells as previously described (41). A FLAG-tagged epitope sequence was introduced upstream of the initiating ATG codon in both the HOXA9 and Hoxa2 cDNAs in pcDNA3 by PCR. Once expressed in mammalian cells, the resulting protein contained a FLAG epitope tag (MDYKDDDDK) at its N terminus. For the electrophoretic mobility shift assay (EMSA), the FLAG-tagged full-length HOXA9 in pcDNA3 was synthesized in vitro using a TNT-coupled in vitro translation system (Promega) in the presence of [35S]methionine. Rabbit reticulocyte lysates were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) for the presence of the full-length labeled product. HOXA9 deletion constructs were generated by PCR using the full-length HOXA9 cDNA as a template and specific primer sets with the appropriate restriction sites at the 5' and 3' ends. Following restriction digestions, the products were subcloned into the pcDNA3 vector. The constructs were designated HOXA9 {Delta}N70, HOXA9 {Delta}N150, and HOXA9 {Delta}N205, containing only the homeodomain and the short C-terminal tail of HOXA9. The constructs were checked for protein translation in vitro as described above. Reporter luciferase constructs containing a 510-bp fragment of the human E-selectin promoter 5' flanking region (E-510) and 400 bp of the human platelet-derived growth factor B (PDGF-B) chain promoter 5' flanking region (P-400) were generated by subcloning into the pGL3-Basic vector (Promega) as previously described (16, 45). The mutant promoter construct (E-510m) was created by mutating two bases at bp –219 to –216 (AATA to CAGA) to disrupt the HOXA9-binding core sequence at the Abd-B-like site by PCR and subcloning subsequently into the pGL3-Basic vector. The deletion promoter constructs were generated by suitable restriction enzyme digestions or by PCR followed by subcloning. The plasmid pCMVß (Clontech) containing the ß-galactosidase gene driven by the cytomegalovirus (CMV) promoter was used as a control in the transfection experiments. Details of vector construction and primer sequences are available upon request.

Cell culture and transfection. Human umbilical vein EC were maintained in MCDB 107 medium supplemented with 15% fetal bovine serum, 150 µg/ml endothelial cell growth supplement, and 90 µg/ml heparin as previously described (45). Dulbecco modified Eagle medium containing 5% fetal bovine serum was used for the maintenance of bovine aortic EC. Lipofectin-mediated transfections (Life Technologies, Inc.) of effector, reporter, and control plasmids were carried out in cultured cells at 80 to 90% confluence with endotoxin-free DNA for 6 h using OptiMEM, and cells were allowed to recover in serum-containing media for 16 to 20 h before harvesting or TNF-{alpha} treatment. Luciferase activity was measured for reporter expression in triplicate and corrected for transfection efficiency by cotransfection with the plasmid vector pCMVß and measurement of the ß-galactosidase activity with the Galactolight Plus detection system (Tropix). All results were verified by multiple independent experiments using at least two different batches of endotoxin-free plasmid DNA.

Real-time PCR assay. First-strand cDNA was synthesized from total RNA (1 µg), isolated from untreated EC or EC treated with TNF-{alpha} (2 ng/ml), in a 50-µl reaction mixture, using the SuperScript first-strand synthesis system for reverse transcription-PCR (Life Technologies) according to the manufacturer's instructions. Five microliters of this reaction mixture was used to perform PCR with specific primer pairs corresponding to a particular gene of interest. Real-time PCR was performed using SYBR green PCR core reagents (PE Applied Biosystems, United Kingdom) and a PerkinElmer ABI PRISM 7700 sequence detector according to the manufacturers’ instructions. Details of primer sequences are available upon request.

Western blot analysis. The level of protein expression, either endogenous or recombinant, was assessed by conventional immunoblotting. Both rabbit reticulocyte lysates and cell extracts from EC were separated on 10 to 12% SDS-PAGE gels by electrophoresis and then transferred to Immobilon-P membranes. The blots were incubated with protein-specific primary antibodies followed by horseradish peroxidase-conjugated specific secondary antibodies according to the manufacturer's instructions. Immunocomplexes were detected by enhanced chemiluminescence (Amersham Pharmacia Biotech). Anti-HOXA9 obtained from Upstate Biotechnology (Lake Placid, NY) and anti-E-selectin antibody obtained from Santa Cruz Biotechnology (Santa Cruz, CA) were used. Monoclonal anti-ß-actin and monoclonal antibody (M2) to the FLAG epitope were obtained from Sigma (St. Louis, MO).

RNA interference. Specific small interfering RNAs (siRNAs) for the HOXA9 target sequence (AACTACTACGTGGACTCGTTC) and scrambled sequence (AAGCCTCAGATAGTCTTCTCG) were designed and synthesized using a Silencer siRNA construction kit (Ambion).

Transfection experiments with siRNA were performed using Targefect F-2 and peptide enhancer (Targeting Systems, CA) according to the manufacturer's instructions with a minor modification. Targefect F-2 (5 µl/ml), peptide enhancer (5 µl/ml), and 50 nM siRNA were successively added to Dulbecco modified Eagle medium, thoroughly mixed, and kept at 37°C for 30 min. Human EC (>90% confluent) were incubated with the transfection mixture for 4 to 5 h. Cells were replenished with fresh EC culture medium and, following a 24-h incubation period, treated with TNF-{alpha} (2 ng/ml) for both 90 min and 3 h. Cells were harvested for subsequent RNA and protein analyses.

EMSA and UV cross-linking. EMSA was performed with both nuclear extract (20 µg) and in vitro-translated protein as described previously (41) with the following modifications. Rabbit reticulocyte lysate (5 µl) expressing either FLAG-tagged HOXA9 protein or the control was used in a 25-µl final reaction volume in a buffer containing 20 mM HEPES (pH 7.5), 60 to 75 mM NaCl, 2 mM MgCl2, 1 mM dithiothreitol, 500 µg/ml bovine serum albumin, 5% glycerol, and 2.5 µg of poly(dI-dC). DNA probes labeled using T4 polymerase and [{gamma}-32P]ATP were subsequently incubated with this reaction mixture at room temperature for 20 min. DNA-protein complexes were resolved on a 6% polyacrylamide gel in 0.25% Tris-buffered saline buffer, followed by autoradiography or phosphorimaging analysis. For competition or supershift assays, the appropriate oligonucleotides or antibodies were used. The sequences of E-selectin promoter DNA probes used for the EMSA were as follows: Abd-B-like site forward, GTATATGCAATTTTATTAATAT, and Abd-B-like site reverse, ATATTAATAAAATTGCATATAC. UV cross-linking was performed as previously described (1). Nuclear extracts (20 µg) prepared from human EC were incubated with 32P-labeled double-stranded DNA (Abd-B-like site of E-selectin promoter) for 20 min at 4°C. The labeled covalent protein-DNA complexes were resolved on a 10% SDS gel by electrophoresis and detected by autoradiography.

ChIP assay. The chromatin immunoprecipitation (ChIP) assay was performed using a kit (Upstate Biotechnology, NY) according to the manufacturer's instructions with some modifications. Human EC (5 x 106) were treated with TNF-{alpha} (2 ng/ml) for 1 h and 3 h. Both control untreated and TNF-{alpha}-treated cells were cross-linked by adding 1% formaldehyde directly to the culture medium for 10 min at 37°C. The cells were then lysed with the lysis buffer supplied in the kit and sonicated at 4°C for five 10-s periods with the setting at 3 (Branson Sonifier 450). Approximately 1% of the resulting sample was kept to serve as an input genomic DNA control, and the rest was precleared with salmon sperm DNA-protein A. Cell lysates were incubated with HOXA9 antibody (N-20; Santa Cruz Biotechnology) overnight at 4°C with rotation, followed by a 1-h incubation with protein A-agarose for ChIP. HOXA9 in the immunoprecipitate was detected by Western blotting. The precipitated chromatin-HOXA9 complex was washed and eluted, and cross-links were reversed by heating at 65°C for 4 h. The genomic DNA fragments, thus isolated, were purified by phenol-chloroform extraction and ethanol precipitation and amplified by PCR. The PCR products were separated on a 2% agarose gel. Three primer sets for PCR were designed to cover different regions of the E-selectin promoter. The product of the first set of primers spanned the sequence from bp –133 to +5 (138 bp) covering the HOXA9-binding site located in the TATA-box area in the E-selectin minimal promoter. The primer sequences were 5'-GCATCGTGGATATTCCCGGGAAAG-3' and 5'-CAGCTGAACACTACTTCGGCTGAGG-3'. The second product spanned the sequence from bp –326 to –110 (217 bp) covering the Abd-B-like site. The primer sequences were 5'-CTACCACAACTACATGAGAGACACTAC-3' and 5'-CTTTCCCGGGAATATCCACGATGC-3'. The third product spanned the sequence from bp –515 to –320 (195 bp), serving as a control. The primer sequences were 5'-ATCTACCTTGTGAGTCATTC-3' and 5'-TAGTTGTGGTAGTAATTAGAAT-3'. The primer sequences used for the PCR amplification of a region of the EphB4 promoter in ChIP analysis were described earlier (7).

Flow cytometric analysis. Human EC were transiently cotransfected with pEGFPC-3 and a plasmid expressing either FLAG-HOXA9 or FLAG-Hoxa2, using Targefect F-2 (Targeting Systems) at a molar ratio of 1:2 according to the manufacturer's protocol with some modification. Transfections were carried out in triplicate in cultured 80 to 90% confluent EC with endotoxin-free DNA (0.5 µg/ml), using Targefect F-2 (5 µl/ml) and peptide enhancer (5 µl/ml) for 4 h. Cells were allowed to recover in serum-containing media for 16 to 20 h prior to treatment with TNF-{alpha} (2 ng/ml) for 1, 5, 15, or 30 h. EC were collected by centrifugation after being dislodged from the dish by using cell dissociation solution (Sigma). To detect the surface expression of E-selectin, cells were incubated with anti-human E-selectin (CD62E) monoclonal antibody (R&D Systems, MN) at a 1:100 dilution for 30 min at room temperature, washed, and subsequently incubated for 30 min with phycoerythrin-conjugated secondary antibody at a 1:200 dilution. Cells, either fresh or fixed with 2% paraformaldehyde, were analyzed by fluorescence-activated cell sorting (FACS; Becton Dickinson and Co., NJ) for the detection of both green fluorescent protein (GFP) and E-selectin protein.

Statistical analysis. Data from at least three independent experiments were analyzed and expressed as means ± standard deviations (SD). Statistical analysis was performed by Student's t test to compare two data sets and analysis of variance for serial analyses.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
HOXA9 participates in TNF-{alpha}-induced expression of the EC-leukocyte adhesion molecule E-selectin. In previous studies, we demonstrated that TNF-{alpha} regulated EC expression of the HOXA9 gene, suggesting a role for this transcription factor in modulating EC-specific genes during cytokine activation (41). One such gene is the leukocyte adhesion molecule E-selectin, which is transiently expressed on the surface of cytokine-activated EC. Direct analysis of the E-selectin promoter sequence revealed one highly conserved potential Abd-B HOX DNA-binding site consisting of the HOX consensus core element, the TTAT sequence. This site consists of a nearly identical inverted abdominal B (Abd-B) sequence (CAATTTTATTAA) from base pairs –210 to –221. The E-selectin promoter also contains a HOX DNA-binding consensus site (TTTACTGG) from base pairs –82 to –75 and an inverted HOX recognition sequence within the TATA-box region (TTTTATAGG) from base pairs –24 to –32 from the transcriptional start site.

We directly addressed the role of HOXA9 in the induction of E-selectin in TNF-{alpha}-treated EC by using HOXA9-specific siRNA to ablate HOXA9 protein levels. Transfection of EC with HOXA9 siRNA led to efficient suppression (>60%) of TNF-{alpha}-induced E-selectin mRNA expression compared to transfection of EC with scrambled siRNA (Fig. 1A). In contrast, the suppression of HOXA9 protein levels had no effect on TNF-{alpha}-induced intracellular adhesion molecule 1 (ICAM-1) expression in EC (Fig. 1A). Simultaneously, the TNF-{alpha}-induced E-selectin protein, which is absent in resting EC, was significantly reduced (~60%) in EC transfected specifically with HOXA9 siRNA (Fig. 1B). Multiple E-selectin protein bands with molecular masses of ~115 kDa suggest the posttranslational modification of E-selectin, as has previously been reported (40, 56). The HOXA9 protein (~32 kDa) was constitutively expressed in EC, and a 3-h TNF-{alpha} treatment had little, if any, effect on the protein level. Most importantly, the results demonstrated that the siRNA was highly specific in attenuating HOXA9 protein levels in both control and TNF-{alpha}-treated EC (Fig. 1C). Although HOXA9 message was reduced in TNF-{alpha}-treated EC at 4 h, as reported previously (41), the steady-state level of HOXA9 protein was modestly changed, which is consistent with the HOXA9 half-life of >3 h as reported by others (58). These results thus demonstrate that HOXA9 protein is constitutively expressed in human EC and that it is required for TNF-{alpha} induction of the E-selectin gene.


Figure 1
View larger version (32K):
[in this window]
[in a new window]

  		FIG. 1. HOXA9 regulates TNF-{alpha}-induced expression of E-selectin. Human EC were transfected with HOXA9 siRNA or scrambled RNA for 24 h and treated with TNF-{alpha} (2 ng/ml) for an additional 90 min or 3 h, respectively, as described in Materials and Methods. E-selectin or ICAM-1 mRNA, isolated from both untreated EC and EC treated with TNF-{alpha} for 90 min, was measured by real-time PCR (A); the percent change was calculated relative to that of mock-transfected EC. Each mRNA expression level was normalized to that of GAPDH (glyceraldehyde-3-phosphate dehydrogenase) mRNA expression. Data represent the means ± SD for three replicate experiments. *, P < 0.05 versus mock. E-selectin (B) or HOXA9 (C) protein levels were detected in EC whole-cell lysates of both untreated EC and EC treated with TNF-{alpha} for 3 h by Western blot analysis. The blots were subsequently reprobed for ß-actin as an internal loading control. The protein levels were quantified by densitometry; the percent change was calculated relative to that of mock-transfected EC. Data represent the means ± SD for three replicate experiments. *, P < 0.05 versus mock (B); **, P < 0.01 versus mock untreated or versus mock TNF-{alpha}-treated EC (C).

 
E-selectin promoter is transactivated by HOXA9. We and others have previously demonstrated that a 510-bp fragment of the 5' flanking region of the well-characterized E-selectin promoter (E-510) is sufficient to drive reporter gene expression in response to the appropriate agonists in EC (11, 16, 23, 38). We wished to determine whether HOXA9 could regulate E-selectin promoter activity in cultured EC, both human and bovine. We transiently cotransfected a CMV-driven HOXA9 cDNA vector together with the 510-bp E-selectin reporter construct into bovine EC. As a control, we cotransfected the HOXA9 vector with a PDGF-B chain promoter-luciferase construct (P-400) consisting of a previously characterized 400-bp fragment of the PDGF-B chain 5' flanking region (45). As shown in Fig. 2A, the overexpression of HOXA9 led to a dose-dependent increase in E-selectin promoter activity. Cotransfection of 1.5 µg/ml of HOXA9 plasmid led to an ~4-fold induction of E-selectin promoter activity. The expression of another HOX gene, Hoxa2, also in a CMV-driven promoter construct, had no influence on E-selectin promoter transactivation. In addition, HOXA9 overexpression failed to transactivate the P-400 PDGF-B chain promoter (Fig. 2A). Similar results were obtained using human EC (data not shown). The results in Fig. 2B demonstrate the relative expression levels of the transfected FLAG-tagged HOXA9 and Hoxa2 proteins in bovine EC and are consistent with the concentrations of transfected plasmids.


Figure 2
View larger version (42K):
[in this window]
[in a new window]

  		FIG. 2. (A) E-selectin promoter activity is up-regulated by HOXA9. FLAG-HOXA9 or FLAG-Hoxa2 cDNA expression constructs in the pcDNA3 vector were cotransfected with the promoter-luciferase reporter construct (0.5 µg/ml) pGL3-E-510 (containing the fragment spanning bp –510 to +1 of the human E-selectin promoter) or pGL3-P-400 (containing the fragment spanning bp –400 to +1 of the human PDGF-B chain promoter) into bovine EC. The reporter luciferase activity was normalized by cotransfecting pCMV ß-galactosidase (20 ng/ml). Relative normalized luciferase activity is expressed as n-fold induction compared to that of control vector transfection. Data represent the means ± SD for three replicate experiments (*, P < 0.05 versus control; **, P < 0.01 versus control). (B) Western blot analyses of both FLAG-tagged Hoxa2 and HOXA9 protein levels in corresponding cDNA-transfected bovine EC. Cell lysates of transfected EC were analyzed for protein levels by SDS-PAGE and immunoblotting. A monoclonal anti-FLAG antibody was used for the immunodetection of both FLAG-tagged HOXA9 and Hoxa2. Each lane in the figure indicates the plasmid concentration used for transfection. Control lanes represent vector transfection.

 
Having demonstrated that HOXA9 overexpression, in and of itself, was sufficient to transactivate the E-510 E-selectin promoter, we wished to determine whether HOXA9 could cooperate with other TNF-{alpha}-induced factors (37) in the induction of the E-selectin promoter. We cotransfected the E-selectin promoter reporter (E-510) with control (empty CMV vector), HOXA9, or Hoxa2 expression vectors in bovine EC and studied reporter activity following treatment with TNF-{alpha}. As shown in Fig. 3, TNF-{alpha} elicited an ~8-fold induction of E-selectin promoter activity when either control or Hoxa2 expression vectors were cotransfected. As noted above, HOXA9 expression alone was sufficient to elicit an ~4-fold induction in E-selectin promoter activity, and TNF-{alpha} treatment resulted in an additional increase in E-selectin promoter activity. These results, therefore, support our earlier observation and further implicate HOXA9 as a factor that participates positively with other TNF-{alpha}-induced factors in activating the E-selectin promoter.


Figure 3
View larger version (22K):
[in this window]
[in a new window]

  		FIG. 3. HOXA9 coexpression augments TNF-{alpha} activation of the E-selectin promoter. The pcDNA3-HOXA9 or Hoxa2 cDNA expression construct (1.5 µg/ml) was cotransfected into bovine EC with the pGL3-E-510 E-selectin promoter reporter construct (0.5 µg/ml). The reporter luciferase activity was normalized by cotransfecting pCMV ß-galactosidase (20 ng/ml). Relative normalized luciferase activity is expressed as n-fold induction compared to that of control vector transfection. Data represent the means ± SD for three replicate experiments (*, P < 0.05 versus TNF-{alpha}-treated control or Hoxa2).

 
To further characterize the role of HOXA9 in E-selectin promoter transactivation, we generated a series of HOXA9 N-terminal deletion constructs to test their ability to transactivate the E-selectin promoter. These constructs, designated {Delta}N70, {Delta}N150, and {Delta}N205, when expressed using an in vitro translation system, produce HOXA9 proteins with deletions representing 70, 150, and 205 amino acid residues, respectively, from the N terminus (data not shown). Cotransfection of the full-length HOXA9 expression plasmid with the E-selectin promoter-luciferase reporter resulted in the induction of the E-selectin promoter in bovine EC; however, cotransfection of the HOXA9 deletion constructs yielded little promoter activation (Fig. 4). Deletion of the first 70 amino acid residues from the N terminus resulted in nearly complete abrogation (~80%) of E-selectin promoter activation compared to full-length HOXA9. Further deletion of HOXA9, {Delta}N150 or {Delta}N205, consistently maintained this loss in transactivation. In addition, the expression of a heterologous chimeric construct in which the HOXA9 DNA-binding homeodomain was replaced with the GAL-4 DNA-binding domain was also completely ineffective in mediating E-selectin promoter transactivation (not shown). Taken together, these results demonstrate that the expression of a fully functional HOXA9 protein, containing both its DNA-binding and transactivation domains, is required for E-selectin promoter activation.


Figure 4
View larger version (24K):
[in this window]
[in a new window]

  		FIG. 4. Both the transactivation and DNA-binding domains of HOXA9 are required for E-selectin promoter up-regulation. The pGL3-E-510 E-selectin promoter reporter construct (0.5 µg/ml) was cotransfected into bovine EC with control pcDNA3 vector, the full-length HOXA9 construct, or the HOXA9 {Delta}N70, {Delta}N150, or {Delta}N205 deletion construct (1.5 µg/ml). The reporter luciferase activity was normalized by cotransfecting pCMV ß-galactosidase (20 ng/ml). Relative normalized luciferase activity is expressed as n-fold induction compared to that of control vector transfection. Data represent the means ± SD for three replicate experiments (*, P < 0.05 versus HOXA9; **, P < 0.01 versus HOXA9).

 
Both recombinant and EC nuclear extract-derived HOXA9 proteins bind to the Abd-B-like site in vitro. Since sequence analysis of the E-selectin promoter revealed the presence of an Abd-B-like HOX-binding site, we wished to determine whether HOXA9 could bind to this site. We first generated FLAG-tagged HOXA9 protein by using a rabbit reticulocyte translation system. As shown in Fig. 5A, when synthesized in vitro, a specific 35S-labeled band corresponding to the appropriate molecular weight of FLAG-HOXA9 was readily detectable following SDS-PAGE and autoradiographic analysis. Further, the in vitro-synthesized FLAG-tagged HOXA9 was confirmed by an anti-FLAG immunoblot analysis (Fig. 5B). We next performed EMSAs using this FLAG-tagged HOXA9 protein and a radiolabeled double-stranded 22-mer oligonucleotide probe corresponding to the Abd-B-like sequence in the E-selectin promoter. As shown in Fig. 5C, a specific protein-DNA complex was formed with reticulocyte lysate containing FLAG-tagged HOXA9 (Fig. 5C, lane 3) but not with control reticulocyte lysate (Fig. 5C, lane 2). The specificity of the complex was further established by the disappearance of this protein-DNA complex with an anti-FLAG-specific monoclonal antibody (Fig. 5C, lane 4) but not with an unrelated mouse anti-immunoglobulin G (IgG) antibody (Fig. 5C, lane 5).


Figure 5
View larger version (42K):
[in this window]
[in a new window]

  		FIG. 5. (A) In vitro translation of HOXA9. [35S]methionine-labeled HOXA9 was synthesized using a pcDNA3 construct in rabbit reticulocyte lysate with T7 polymerase in a coupled transcription-translation system (Promega). Autoradiogram of SDS-polyacrylamide gel showing the presence of an ~32-kDa FLAG-tagged HOXA9 protein (lane 2) and control reaction products containing pcDNA3 vector only (lane 1). (B) Western blot analysis of in vitro-translated FLAG-tagged HOXA9. Rabbit reticulocyte lysates were analyzed for translation products by SDS-PAGE and immunoblotting. A monoclonal anti-FLAG antibody was used for the immunodetection of FLAG-tagged HOXA9. The vector lane represents the control reaction. (C) Specific binding of HOXA9 to the Abd-B-like site in the E-selectin promoter. An EMSA was performed by incubating in vitro-translated FLAG-tagged HOXA9 in rabbit reticulocyte lysate and a double-stranded 32P-labeled Abd-B-like site oligonucleotide, as described in Materials and Methods. Lane 1, no-lysate control; lane 2, control lysate; lane 3, lysate expressing FLAG-tagged HOXA9; lane 4, incubation with monoclonal anti-FLAG antibody; lane 5, incubation with mouse IgG.

 
To test more directly whether endogenous HOXA9 protein can bind to the Abd-B-like site identified above, we performed HOXA9/DNA-binding analysis using human EC nuclear extracts. The specific double-stranded 22-mer oligonucleotide probe consisting of the Abd-B-like HOX consensus site was incubated with EC nuclear extracts and UV cross-linked. As shown in Fig. 6, multiple protein-DNA complexes were formed when UV cross-linking was performed with the Abd-B-like site oligonucleotide (Fig. 6, lane 2). The relative sizes of these protein/DNA complexes were consistent with the combined sizes of the Abd-B HOX proteins (~25 to 35 kDa) and the 22-mer oligonucleotide (~15 kDa). The complexes were also specific for the Abd-B-like HOX consensus sequence, as revealed by competition with a 10-fold excess of the unlabeled wild-type oligonucleotide (Fig. 6, lane 3). One such complex appeared to be highly specific for HOXA9, since it was not affected by nonspecific IgG (Fig. 6, lane 4), but disappeared in the presence of HOXA9-specific antibody (Fig. 6, lane 5). These results suggest that both the endogenous and the in vitro-translated HOXA9 protein generate a highly specific protein-DNA complex with the Abd-B-like site of the E-selectin promoter. However, no HOXA9-specific protein-DNA complex was detected when either of the two other HOX-binding motifs, as described earlier, was used in this assay (data not shown).


Figure 6
View larger version (52K):
[in this window]
[in a new window]

  		FIG. 6. HOXA9 present in EC nuclear extracts binds to the Abd-B-like site. Human EC nuclear extracts were incubated and UV cross-linked to a double-stranded 32P-labeled oligonucleotide (Oligo) containing the Abd-B-like site (GTATATGCAATTTTATTAATAT), as described in Materials and Methods. Reaction products were analyzed by 10% SDS-PAGE. Lane 1, binding reaction in the absence of UV treatment; lane 2, binding reaction following UV treatment; lane 3, binding reaction including a 10-fold molar excess of an unlabeled Abd-B-like site oligonucleotide; lane 4, binding reaction including normal rabbit IgG (2 µg); lane 5, binding reaction including rabbit polyclonal anti-HOXA9 antibody (2 µg).

 
The Abd-B-like site is critical for E-selectin promoter transactivation by HOXA9. To functionally characterize the HOXA9-binding element identified above, we made a series of E-selectin deletion reporter constructs by serially truncating the 5' flanking promoter regions (Fig. 7). Previous studies have shown that the E-selectin promoter becomes less responsive to TNF-{alpha} when the ATF-2 site is deleted (construct at bp –143) and completely TNF-{alpha}-unresponsive upon deletion of the first NF-{kappa}B site (construct at bp –119). We performed transient cotransfection studies with HOXA9 and the various truncated promoter reporter constructs. As shown in Fig. 7, deletion to position –271 had little effect on HOXA9-induced E-selectin promoter transactivation. Further deletion of the 5' sequences to position –188, resulting in the elimination of the Abd-B-like site, led to a significant reduction (>60%) in HOXA9-mediated promoter induction. This result clearly demonstrates the critical role of the region spanning positions –271 to –188 containing the Abd-B-like site in HOXA9-induced E-selectin promoter transactivation. However, successive deletions to position –143, –117, –73, or –52 did not show any additional change in promoter induction. We were not able to observe a complete loss of HOXA9-mediated promoter induction. This may be due to the fact that the TATA-box region, still present in our –52 deletion construct, contains a consensus HOXA9-binding site.


Figure 7
View larger version (15K):
[in this window]
[in a new window]

  		FIG. 7. The Abd-B-like site contributes to E-selectin promoter up-regulation by HOXA9. The indicated luciferase reporter constructs (0.5 µg/ml) were transiently cotransfected in triplicate into bovine EC with or without 1.5 µg/ml of a HOXA9 expression vector. The presence of defined response elements relative to the reporter constructs used is indicated on the left. The reporter luciferase activity was normalized by cotransfecting pCMV ß-galactosidase (20 ng/ml). Relative normalized luciferase activity is expressed as induction compared to that of control vector transfection.

 
To further characterize the role of the Abd-B-like site in E-selectin promoter induction, we created an Abd-B site mutant promoter construct in which two bases of the core binding element were altered. We then cotransfected either the wild type (E-510) or the mutant promoter construct (E-510m) with a HOXA9 expression vector in bovine EC. Cotransfection resulted in an ~6.5-fold induction of the E-selectin promoter (Fig. 8); however, mutation in the Abd-B-like site substantially abrogated promoter activation (>70% reduction compared to that of the wild-type promoter). This result, therefore, confirms a crucial role for the Abd-B-like site in the transactivation of the E-selectin promoter.


Figure 8
View larger version (25K):
[in this window]
[in a new window]

  		FIG. 8. Mutation of the Abd-B-like site significantly reduces HOXA9-mediated induction of the E-selectin promoter. The pGL3-E-510 or pGL3-E-510m (mutated in the Abd-B-like site) E-selectin promoter reporter construct (0.5 µg/ml) was transiently cotransfected with either the pcDNA3 control or the HOXA9 expression construct (1.5 µg/ml) into bovine EC. The reporter luciferase activity was normalized by cotransfecting pCMV ß-galactosidase (20 ng/ml). Relative normalized luciferase activity is expressed as n-fold induction compared to that of control vector transfection. Data represent the means ± SD for three replicate experiments (*, P < 0.05 versus E-510/HOXA9).

 
The Abd-B-like site is responsible for endogenous HOXA9 binding in vivo. In addition to the in vitro DNA-binding analysis, we wished to determine whether HOXA9 was associated with the endogenous E-selectin promoter upon stimulation by TNF-{alpha}. We, therefore, performed ChIP assays with a HOXA9-specific antibody by using cellular lysates from TNF-{alpha}-treated human EC. As shown in Fig. 9A (upper panel), with primers amplifying sequences spanning from positions –326 to –110 of the E-selectin promoter, which includes the Abd-B-like site, a 216-bp DNA fragment was amplified from anti-HOXA9 chromatin precipitates from 1-h TNF-{alpha}-treated cells. Anti-HOXA9 precipitates from untreated control or 3-h TNF-{alpha}-treated EC resulted in little or no DNA amplification. These results demonstrate that in EC, TNF-{alpha} induces the transient binding of HOXA9 to the Abd-B-like site. When the appropriate primers for the region spanning from positions –133 to +5, containing the two other consensus HOX DNA-binding sites, were used to amplify chromatin immunoprecipitated with anti-HOXA9, no DNA fragment was amplified from either control or TNF-{alpha}-treated EC (Fig. 9A, middle panel). These results confirm the data for HOXA9 DNA binding with the Abd-B-like site (Fig. 5C and 6). Finally, no DNA fragment was amplified from chromatin immunoprecipitated with anti-HOXA9 antibodies from either TNF-{alpha}-treated or untreated EC with primers amplifying the sequence from positions –515 to –320 (Fig. 9A, lower panel). With all three sets of primers, no DNA was amplified from chromatin immunoprecipitated with control IgG (data not shown). This transient, TNF-{alpha}-dependent interaction of HOXA9 with the E-selectin promoter differs from the case of another important EC gene, that for the EphB4 promoter (7), which constitutively bound HOXA9 independent of TNF-{alpha} (Fig. 9B).


Figure 9
View larger version (32K):
[in this window]
[in a new window]

  		FIG. 9. HOXA9 binds transiently to the E-selectin promoter region containing the Abd-B-like site in response to TNF-{alpha}. ChIP assays were performed, using anti-HOXA9-specific antibodies, with human EC treated in the absence or presence of TNF-{alpha} for the indicated times. Three distinct sets of primers spanning different regions of the E-selectin promoter were used to amplify the immunoprecipitated (IP) DNA. (A) The upper panel represents amplified DNA, using primers spanning the E-selectin sequence from bp –326 to –110, containing the Abd-B-like site; the middle panel represents amplified DNA, using primers spanning the E-selectin sequence from bp –133 to +5, containing the TATA-box site; and the lower panel represents amplified DNA, using primers that span the E-selectin sequence from bp –515 to –320. The PCR products of both input and ChIP samples were loaded as indicated and analyzed using a 2% agarose gel. Data represent one of three similar sets of experiments. (B) The PCR products of both input and ChIP samples of the immunoprecipitated EphB4 promoter were loaded as indicated and analyzed using a 2% agarose gel. All the amplified immunoprecipitated DNA fragments were sequenced for identification following gel purification (data not shown).

 
Overexpression of HOXA9 does not change the TNF-{alpha}-dependent transient expression pattern of the E-selectin gene. Since we postulated that the down-regulation of HOXA9 may represent an "off switch" to attenuate E-selectin expression (41), we examined whether the overexpression of HOXA9 resulted in a sustained level of E-selectin protein on the EC surface. Both a GFP expression vector and either the HOXA9 or Hoxa2 construct were cotransfected into EC, and the surface expression of E-selectin protein was monitored by FACS analysis. As known previously, the majority of EC exhibited surface expression of E-selectin in response to TNF-{alpha} treatment for 5 h (Fig. 10A). Further, in spite of the continuous presence of TNF-{alpha}, E-selectin protein levels were reduced significantly at 15 h and essentially to basal levels at 30 h. However, the TNF-{alpha}-dependent transient surface expression profile of E-selectin remained unaltered in EC overexpressing both GFP and either HOXA9 or control Hoxa2. This was independently supported by Western blot analysis of E-selectin protein levels in total cell lysates of control as well as transfected EC (data not shown). Although the overexpressed protein levels were determined to be comparable (Fig. 10B), ChIP analysis failed to show HOXA9 binding to the Abd-B-like site beyond 1 h following TNF-{alpha} treatment as noted above (Fig. 10C). Thus, temporal binding of HOXA9 to the E-selectin gene promoter in spite of its persistent presence in EC appears to be important for E-selectin gene expression in EC, but continued overexpression does not negate the "turning off" of E-selectin expression at later times following TNF-{alpha} stimulation.


Figure 10
View larger version (25K):
[in this window]
[in a new window]

  		FIG. 10. FACS analysis of E-selectin surface expression kinetics. Early-passage human EC were cotransfected for 24 h with the constructs expressing GFP and either HOXA9 or Hoxa2. Following treatment of both control and cotransfected cells with TNF-{alpha} (2 ng/ml) for 0, 5, 15, or 30 h, cells were dislodged and assayed for E-selectin surface expression by FACS analysis using a human E-selectin-specific antibody, as described in Materials and Methods. Data were expressed as the percentage of the total EC population for the control and the percentage of GFP-positive EC for transfected cells expressing surface E-selectin (A). Results are the means for two individual experiments. Following 24 h of transfection, FLAG-HOXA9 or FLAG-Hoxa2 protein levels were detected by Western blot analysis of whole-cell lysates of both untreated and TNF-{alpha}-treated EC for the indicated times (B). For the ChIP assay, primers spanning the E-selectin sequence from bp –326 to –110, containing the Abd-B-like site, were used for PCR amplification of both input and immunoprecipitated (IP) DNA as described in the legend to Fig. 9. The PCR products of both input and ChIP samples of the immunoprecipitated E-selectin promoter were loaded as indicated and analyzed using a 2% agarose gel (C).

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
The "activation" or "dysfunctional" state of EC leads to the expression of a variety of EC-leukocyte adhesion molecules on the EC surface, including E-selectin. E-selectin is precisely lineage specific and required for the firm adhesion of leukocytes to the endothelium during the initial stages of inflammation. Multiple transcription factors have been demonstrated to play a role in cytokine-mediated induction of E-selectin, including the Rel family member NF-{kappa}B (10), a cyclic AMP-independent ATF family member, ATF-2 (26), and HMG-I(Y) (34). Recent results have suggested a model for EC-specific induction of the E-selectin gene in which multiple coactivators, including p300 and I{kappa}B kinase {alpha}, are required and the core promoter chromatin architecture is specifically remodeled in EC (14). In addition, other unidentified regulatory factors that may specify lineage-specific expression in the endothelium have also been proposed to play a role in the induction of the E-selectin gene. Here, we have identified for the first time a role for the TNF-{alpha}-modulated homeobox gene HOXA9, the expression of which is restricted principally to cells of endothelial and hematopoietic origin, in the regulated transcription of E-selectin. The active participation of HOXA9, an important regulator of organ development, in the induction of E-selectin is consistent with the finding that E-selectin is differentially regulated and expressed in endothelial lineage cells during development (36).

In this report, we have demonstrated that the depletion of HOXA9 by specific siRNA resulted in a significant inhibition of E-selectin protein expression in TNF-{alpha}-activated EC, though the induction of another important EC-leukocyte adhesion molecule, ICAM-1, was unaffected. It should be noted that while the E-selectin promoter contains multiple HOX-binding elements, ICAM-1 lacks these elements. The Abd-B-like HOX proteins, including HOXA9, recognize DNA-binding motifs with the core sequence TTTAT/C; however, HOXA9 has also been reported to bind motifs with a core TAAT in either inhibiting osteopontin expression in transforming growth factor ß-stimulated epithelial cells (49) or inducing EphB4 expression in EC (7). Moreover, in myeloid cells, HOXA9 activates the transcription of the CYBB gene by binding to the Hox/Pbx consensus-like element (ATGATTAT) together with Pbx1 (2). DNA-binding functional trimeric complexes containing HOXA9 and TALE proteins, Pbx and Meis, respectively, are present in leukemic cells and bind to consensus sequence A/GTGATTT/AAT/CG (48).

The physiological relevance of multiple HOX-binding elements within the E-selectin proximal promoter has not been explored previously. Here, we report on the existence of three motifs containing the core recognition sequence (TTTAT/C) for Abd-B-like HOX proteins within the E-selectin promoter at bp –210 to –221, –82 to –75, and –24 to –32. However, the Abd-B-like site at bp –210 is highly homologous with the Drosophila Abd-B consensus sequence (AAATTTTATG/TG/AC), whereas the other two sites have flanking nucleotides that do not resemble the consensus sequence. Our results from promoter deletion experiments demonstrate that neither the –82 site nor the –24 site has a significant impact on HOXA9-induced E-selectin promoter activity.

The Abd-B-like site, on the other hand, is functionally important in HOXA9-induced E-selectin promoter transactivation, as evidenced by both deletional analysis and site-directed mutagenesis. HOXA9 protein, either endogenous or recombinant, directly binds to the Abd-B-like site, as shown by in vitro protein-DNA interaction studies. In vivo ChIP assays have also identified specific and temporal HOXA9 binding to this Abd-B-like site in a TNF-{alpha}-dependent manner. Therefore, specific biologic processes must impart selectivity toward the Abd-B-like site among the three motifs in the E-selectin promoter. In addition, the lack of binding at later time points following TNF-{alpha} treatment is in contrast to the findings that HOXA9 constitutively interacts with the EC tyrosine kinase receptor EphB4 promoter (7) and remains associated in the continual presence of TNF-{alpha}. Both the TNF-{alpha}-induced E-selectin-specific "histone code" and posttranslational modification of HOXA9, as reported earlier (2, 14), may account for this difference. Future study is needed to elucidate the role, if any, of the other two sites in the induction of the E-selectin gene by other agonists, considering the presence of other Abd-B-like HOX proteins in EC.

Previous studies have indicated that the activation of two distinct but highly regulated parallel pathways involving both sequence-specific DNA binding and associations among transcription factors, including NF-{kappa}B, ATF-2/Jun, and HMG-I(Y), is required for the transcriptional induction of the E-selectin promoter (53). Such interactions have been proposed to be important for E-selectin promoter looping to bring those factors in close proximity to each other and/or for the transcription initiation site in controlling transcriptional activity (35, 37). It is possible that as a critical site-specific transcription factor, HOXA9 alters the spatial configuration of the promoter, resulting in activated transcription by binding to the Abd-B-like site in response to TNF-{alpha}. Alternatively, HOXA9 DNA binding may promote the stabilization of a multimeric transcriptional complex by bringing the Abd-B-like site close to the TATA box through promoter looping. However, we did not observe any physical association between HOXA9 and the transcription initiation factor TF-IID (data not shown), though such interaction could be highly transient in nature. Moreover, HOXA9, in addition to binding DNA, may associate with the crucial architectural factor HMG-I(Y), which has been proposed to be a general cofactor in HOX-mediated transcriptional activation (57). We are currently investigating such a possibility in the E-selectin induction process.

Finally, results based on our current study as well as previous studies (41) suggest a possible explanation for the active down-regulation of E-selectin gene induction following TNF-{alpha} treatment. We propose that both the spatial and temporal chromatin binding and subsequent down-regulation of HOXA9, not alone but concomitantly with other functionally critical factors, including the chromatin modifying factor SWI/SNF, as reported earlier (14), may constitute the "off-switch" circuitry for the E-selectin gene. Further supporting evidence comes from our finding using FACS analysis that even sustained overexpression of HOXA9 alone, in conformity with its temporal chromatin binding only in the presence of TNF-{alpha}, did not increase the persistence of E-selectin on the EC surface.

In summary, we have demonstrated that HOXA9 is involved in the transcriptional up-regulation of the E-selectin gene in EC by binding to a specific Abd-B-like sequence in the E-selectin promoter. Moreover, our results suggest that the binding of HOXA9 to this Abd-B-like site in the E-selectin promoter is temporally mediated by TNF-{alpha}. These results, therefore, identify a specific role for HOXA9 in EC activation in which it acts as a potential proinflammatory factor in the induction of EC-leukocyte adhesion molecules, such as E-selectin.


   ACKNOWLEDGMENTS
 
This work was supported by National Institutes of Health grant HL29582 (P.E.D.).

We thank E. Poptic, P. Hoang, and L. Mavrakis for cell culture assistance. Human umbilical vein endothelial cells were isolated from the cords collected by the Cleveland Clinic Birthing Services and the Perinatal Clinical Research Center (supported by National Institutes of Health Research Center award RR-00080) at the Cleveland MetroHealth Hospital.


   FOOTNOTES
 
* Corresponding author. Mailing address: Department of Cell Biology, Lerner Research Institute and Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, Cleveland Clinic Foundation, Cleveland, OH 44195. Phone: (216) 444-5849. Fax: (216) 444-3279. E-mail: dicorlp{at}ccf.org Back

{triangledown} Published ahead of print on 23 April 2007. Back


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
1. Bandyopadhyay, S., and R. M. Gronostajski. 1994. Identification of a conserved oxidation-sensitive cysteine residue in the NFI family of DNA-binding proteins. J. Biol. Chem. 269:29949-29955.[Abstract/Free Full Text]

2. Bei, L., Y. Lu, and E. A. Eklund. 2005. HOXA9 activates transcription of the gene encoding gp91Phox during myeloid differentiation. J. Biol. Chem. 280:12359-12370.[Abstract/Free Full Text]

3. Belotti, D., N. Clausse, D. Flagiello, Y. Alami, M. Daukandt, C. Deroanne, B. Malfoy, E. Boncinelli, A. Faiella, and V. Castronovo. 1998. Expression and modulation of homeobox genes from cluster B in endothelial cells. Lab. Investig. 78:1291-1299.[Medline]

4. Bevilacqua, M. P., S. Stengelin, M. A. Gimbrone, Jr., and B. Seed. 1989. Endothelial leukocyte adhesion molecule 1: an inducible receptor for neutrophils related to complement regulatory proteins and lectins. Science 243:1160-1165.[Abstract/Free Full Text]

5. Boudreau, N., C. Andrews, A. Srebrow, A. Ravanpay, and D. A. Cheresh. 1997. Induction of the angiogenic phenotype by Hox D3. J. Cell Biol. 139:257-264.[Abstract/Free Full Text]

6. Boyle, E. M., Jr., T. T. Sato, R. F. Noel, Jr., E. D. Verrier, and T. H. Pohlman. 1999. Transcriptional arrest of the human E-selectin gene. J. Surg. Res. 82:194-200.[CrossRef][Medline]

7. Bruhl, T., C. Urbich, D. Aicher, A. Acker-Palmer, A. M. Zeiher, and S. Dimmeler. 2004. Homeobox A9 transcriptionally regulates the EphB4 receptor to modulate endothelial cell migration and tube formation. Circ. Res. 94:743-751.[Abstract/Free Full Text]

8. Capecchi, M. R. 1996. Function of homeobox genes in skeletal development. Ann. N. Y. Acad. Sci. 785:34-37.[Medline]

9. Capecchi, M. R. 1997. Hox genes and mammalian development. Cold Spring Harbor Symp. Quant. Biol. 62:273-281.[Medline]

10. Collins, T., M. A. Read, A. S. Neish, M. Z. Whitley, D. Thanos, and T. Maniatis. 1995. Transcriptional regulation of endothelial cell adhesion molecules: NF-kappa B and cytokine-inducible enhancers. FASEB J. 9:899-909.[Abstract]

11. Collins, T., A. Williams, G. I. Johnston, J. Kim, R. Eddy, T. Shows, M. A. Gimbrone, Jr., and M. P. Bevilacqua. 1991. Structure and chromosomal location of the gene for endothelial-leukocyte adhesion molecule 1. J. Biol. Chem. 266:2466-2473.[Abstract/Free Full Text]

12. DiCorleto, P. E., and M. A. Gimbrone, Jr. 1996. Vascular endothelium, p. 387-399. In V. Fuster, R. Ross, and E. J. Topol (ed.), Atherosclerosis and coronary artery disease, vol. 22. Lippincott-Raven Publishers, Philadelphia, PA.

13. Diehl, F., L. Rossig, A. M. Zeiher, S. Dimmeler, and C. Urbich. 2007. The histone methyltransferase MLL is an upstream regulator of endothelial-cell sprout formation. Blood 109:1472-1478.[Abstract/Free Full Text]

14. Edelstein, L. C., A. Pan, and T. Collins. 2005. Chromatin modification and the endothelial-specific activation of the E-selectin gene. J. Biol. Chem. 280:11192-11202.[Abstract/Free Full Text]

15. Endemann, D. H., and E. L. Schiffrin. 2004. Endothelial dysfunction. J. Am. Soc. Nephrol. 15:1983-1992.[Abstract/Free Full Text]

16. Faruqi, T. R., and P. E. DiCorleto. 1997. IFN-gamma inhibits double-stranded RNA-induced E-selectin expression in human endothelial cells. J. Immunol. 159:3989-3994.[Abstract]

17. Fromental-Ramain, C., X. Warot, S. Lakkaraju, B. Favier, H. Haack, C. Birling, A. Dierich, P. Doll e, and P. Chambon. 1996. Specific and redundant functions of the paralogous Hoxa-9 and Hoxd-9 genes in forelimb and axial skeleton patterning. Development 122:461-472.[Abstract]

18. Gaufo, G. O., S. Wu, and M. R. Capecchi. 2004. Contribution of Hox genes to the diversity of the hindbrain sensory system. Development 131:1259-1266.[Abstract/Free Full Text]

19. Goomer, R. S., B. D. Holst, I. C. Wood, F. S. Jones, and G. M. Edelman. 1994. Regulation in vitro of an L-CAM enhancer by homeobox genes HoxD9 and HNF-1. Proc. Natl. Acad. Sci. USA 91:7985-7989.[Abstract/Free Full Text]

20. Gorski, D. H., and K. Walsh. 2000. The role of homeobox genes in vascular remodeling and angiogenesis. Circ. Res. 87:865-872.[Abstract/Free Full Text]

21. Grier, D. G., A. Thompson, A. Kwasniewska, G. J. McGonigle, H. L. Halliday, and T. R. Lappin. 2005. The pathophysiology of HOX genes and their role in cancer. J. Pathol. 205:154-171.[CrossRef][Medline]

22. Hombría, J. C., and B. Lovegrove. 2003. Beyond homeosis—HOX function in morphogenesis and organogenesis. Differentiation 71:461-476.[CrossRef][Medline]

23. Hooft van Huijsduijnen, R., J. Whelan, R. Pescini, M. Becker-Andre, A. M. Schenk, and J. F. DeLamarter. 1992. A T-cell enhancer cooperates with NF-kappa B to yield cytokine induction of E-selectin gene transcription in endothelial cells. J. Biol. Chem. 267:22385-22391.[Abstract/Free Full Text]

24. Jones, F. S., E. A. Prediger, D. A. Bittner, E. M. De Robertis, and G. M. Edelman. 1992. Cell adhesion molecules as targets for Hox genes: neural cell adhesion molecule promoter activity is modulated by cotransfection with Hox-2.5 and -2.4. Proc. Natl. Acad. Sci. USA 89:2086-2090.[Abstract/Free Full Text]

25. Kansas, G. S. 1996. Selectins and their ligands: current concepts and controversies. Blood 88:3259-3287.[Free Full Text]

26. Kaszubska, W., R. Hooft van Huijsduijnen, P. Ghersa, A.-M. DeRaemy-Schenk, B. P. Chen, T. Hai, J. F. DeLamarter, and J. Whelan. 1993. Cyclic AMP-independent ATF family members interact with NF-{kappa}B and function in the activation of the E-selectin promoter in response to cytokines. Mol. Cell. Biol. 13:7180-7190.[Abstract/Free Full Text]

27. Kim, M. H., H. H. Chang, C. Shin, M. Cho, D. Park, and H. W. Park. 1998. Genomic structure and sequence analysis of human HOXA-9. DNA Cell Biol. 17:407-414.[Medline]

28. Krause, T., and G. A. Turner. 1999. Are selectins involved in metastasis? Clin. Exp. Metastasis 17:183-192.[CrossRef][Medline]

29. Kroon, E., J. Krosl, U. Thorsteinsdottir, S. Baban, A. M. Buchberg, and G. Sauvageau. 1998. Hoxa9 transforms primary bone marrow cells through specific collaboration with Meis1a but not Pbx1b. EMBO J. 17:3714-3725.[CrossRef][Medline]

30. Krumlauf, R. 1994. Hox genes in vertebrate development. Cell 78:191-201.[CrossRef][Medline]

31. LaRonde-LeBlanc, N. A., and C. Wolberger. 2003. Structure of HoxA9 and Pbx1 bound to DNA: Hox hexapeptide and DNA recognition anterior to posterior. Genes Dev. 17:2060-2072.[Abstract/Free Full Text]

32. Lawrence, H. J., C. D. Helgason, G. Sauvageau, S. Fong, D. J. Izon, R. K. Humphries, and C. Largman. 1997. Mice bearing a targeted interruption of the homeobox gene HOXA9 have defects in myeloid, erythroid, and lymphoid hematopoiesis. Blood 89:1922-1930.[Abstract/Free Full Text]

33. Lawrence, H. J., G. Sauvageau, R. K. Humphries, and C. Largman. 1996. The role of HOX homeobox genes in normal and leukemic hematopoiesis. Stem Cells 14:281-291.[Abstract]

34. Lewis, H., W. Kaszubska, J. F. DeLamarter, and J. Whelan. 1994. Cooperativity between two NF-{kappa}B complexes, mediated by high-mobility-group protein I(Y), is essential for cytokine-induced expression of the E-selectin promoter. Mol. Cell. Biol. 14:5701-5709.[Abstract/Free Full Text]

35. Meacock, S., R. Pescini-Gobert, J. F. DeLamarter, and R. Hooft van Huijsduijnen. 1994. Transcription factor-induced, phased bending of the E-selectin promoter. J. Biol. Chem. 269:31756-31762.[Abstract/Free Full Text]

36. Milstone, D. S., P. E. O'Donnell, G. Stavrakis, R. M. Mortensen, and V. M. Davis. 2000. E-selectin expression and stimulation by inflammatory mediators are developmentally regulated during embryogenesis. Lab. Investig. 80:943-954.[Medline]

37. Min, W., and J. S. Pober. 1997. TNF initiates E-selectin transcription in human endothelial cells through parallel TRAF-NF-kappa B and TRAF-RAC/CDC42-JNK-c-Jun/ATF2 pathways. J. Immunol. 159:3508-3518.[Abstract]

38. Montgomery, K. F., L. Osborn, C. Hession, R. Tizard, D. Goff, C. Vassallo, P. I. Tarr, K. Bomsztyk, R. Lobb, J. M. Harlan, et al. 1991. Activation of endothelial-leukocyte adhesion molecule 1 (ELAM-1) gene transcription. Proc. Natl. Acad. Sci. USA 88:6523-6527.[Abstract/Free Full Text]

39. Myers, C., A. Charboneau, and N. Boudreau. 2000. Homeobox B3 promotes capillary morphogenesis and angiogenesis. J. Cell Biol. 148:343-351.[Abstract/Free Full Text]

40. Nishiwaki, Y., T. Yokota, M. Hiraoka, M. Miyagishi, K. Taira, M. Isobe, H. Mizusawa, and M. Yoshida. 2003. Introduction of short interfering RNA to silence endogenous E-selectin in vascular endothelium leads to successful inhibition of leukocyte adhesion. Biochem. Biophys. Res. Commun. 310:1062-1066.[CrossRef][Medline]

41. Patel, C. V., R. Sharangpani, S. Bandyopadhyay, and P. E. DiCorleto. 1999. Endothelial cells express a novel, tumor necrosis factor-alpha-regulated variant of HOXA9. J. Biol. Chem. 274:1415-1422.[Abstract/Free Full Text]

42. Read, M. A., M. Z. Whitley, S. Gupta, J. W. Pierce, J. Best, R. J. Davis, and T. Collins. 1997. Tumor necrosis factor alpha-induced E-selectin expression is activated by the nuclear factor-kappaB and c-JUN N-terminal kinase/p38 mitogen-activated protein kinase pathways. J. Biol. Chem. 272:2753-2761.[Abstract/Free Full Text]

43. Rossig, L., C. Urbich, T. Bruhl, E. Dernbach, C. Heeschen, E. Chavakis, K. Sasaki, D. Aicher, F. Diehl, F. Seeger, M. Potente, A. Aicher, L. Zanetta, E. Dejana, A. M. Zeiher, and S. Dimmeler. 2005. Histone deacetylase activity is essential for the expression of HoxA9 and for endothelial commitment of progenitor cells. J. Exp. Med. 201:1825-1835.[Abstract/Free Full Text]

44. Sauvageau, G., P. M. Lansdorp, C. J. Eaves, D. E. Hogge, W. H. Dragowska, D. S. Reid, C. Largman, H. J. Lawrence, and R. K. Humphries. 1994. Differential expression of homeobox genes in functionally distinct CD34+ subpopulations of human bone marrow cells. Proc. Natl. Acad. Sci. USA 91:12223-12227.[Abstract/Free Full Text]

45. Scarpati, E. M., and P. E. DiCorleto. 1996. Identification of a thrombin response element in the human platelet-derived growth factor B-chain (c-sis) promoter. J. Biol. Chem. 271:3025-3032.[Abstract/Free Full Text]

46. Scholz, D., B. Devaux, A. Hirche, B. Potzsch, B. Kropp, W. Schaper, and J. Schaper. 1996. Expression of adhesion molecules is specific and time-dependent in cytokine-stimulated endothelial cells in culture. Cell Tissue Res. 284:415-423.[CrossRef][Medline]

47. Shankar, R., C. A. de la Motte, E. J. Poptic, and P. E. DiCorleto. 1994. Thrombin receptor-activating peptides differentially stimulate platelet-derived growth factor production, monocytic cell adhesion, and E-selectin expression in human umbilical vein endothelial cells. J. Biol. Chem. 269:13936-13941.[Abstract/Free Full Text]

48. Shen, W.-F., S. Rozenfeld, A. Kwong, L. G. Kömüves, H. J. Lawrence, and C. Largman. 1999. HOXA9 forms triple complexes with PBX2 and MEIS1 in myeloid cells. Mol. Cell. Biol. 19:3051-3061.[Abstract/Free Full Text]

49. Shi, X., S. Bai, L. Li, and X. Cao. 2001. Hoxa-9 represses transforming growth factor-beta-induced osteopontin gene transcription. J. Biol. Chem. 276:850-855.[Abstract/Free Full Text]

50. Thorsteinsdottir, U., A. Mamo, E. Kroon, L. Jerome, J. Bijl, H. J. Lawrence, K. Humphries, and G. Sauvageau. 2002. Overexpression of the myeloid leukemia-associated Hoxa9 gene in bone marrow cells induces stem cell expansion. Blood 99:121-129.[Abstract/Free Full Text]

51. Vijapurkar, U., N. Fischbach, W. Shen, C. Brandts, D. Stokoe, H. J. Lawrence, and C. Largman. 2004. Protein kinase C-mediated phosphorylation of the leukemia-associated HOXA9 protein impairs its DNA binding ability and induces myeloid differentiation. Mol. Cell. Biol. 24:3827-3837.[Abstract/Free Full Text]

52. Wang, N., H. G. Kim, C. V. Cotta, M. Wan, Y. Tang, C. A. Klug, and X. Cao. 2006. TGFbeta/BMP inhibits the bone marrow transformation capability of Hoxa9 by repressing its DNA-binding ability. EMBO J. 25:1469-1480.[CrossRef][Medline]

53. Whelan, J., P. Ghersa, R. Hooft van Huijsduijnen, J. Gray, G. Chandra, F. Talabot, and J. F. DeLamarter. 1991. An NF kappa B-like factor is essential but not sufficient for cytokine induction of endothelial leukocyte adhesion molecule 1 (ELAM-1) gene transcription. Nucleic Acids Res. 19:2645-2653.[Abstract/Free Full Text]

54. Wu, Y., M. Moser, V. L. Bautch, and C. Patterson. 2003. HoxB5 is an upstream transcriptional switch for differentiation of the vascular endothelium from precursor cells. Mol. Cell. Biol. 23:5680-5691.[Abstract/Free Full Text]

55. Yan, J., Y. X. Chen, A. Desmond, A. Silva, Y. Yang, H. Wang, and X. Hua. 2006. Cdx4 and menin co-regulate Hoxa9 expression in hematopoietic cells. PLoS ONE 1:e47.[CrossRef]

56. Yoshida, M., W. F. Westlin, N. Wang, D. E. Ingber, A. Rosenzweig, N. Resnick, and M. A. Gimbrone, Jr. 1996. Leukocyte adhesion to vascular endothelium induces E-selectin linkage to the actin cytoskeleton. J. Cell Biol. 133:445-455.[Abstract/Free Full Text]

57. Zappavigna, V., L. Falciola, M. Helmer-Citterich, F. Mavilio, and M. E. Bianchi. 1996. HMG1 interacts with HOX proteins and enhances their DNA binding and transcriptional activation. EMBO J. 15:4981-4991.[Medline]

58. Zhang, Y., G. Morrone, J. Zhang, X. Chen, X. Lu, L. Ma, M. Moore, and P. Zhou. 2003. CUL-4A stimulates ubiquitylation and degradation of the HOXA9 homeodomain protein. EMBO J. 22:6057-6067.[CrossRef][Medline]

Molecular and Cellular Biology, June 2007, p. 4207-4216, Vol. 27, No. 12
0270-7306/07/$08.00+0     doi:10.1128/MCB.00052-07
Copyright © 2007, American Society for Microbiology. All Rights Reserved.




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Other Versions of this Article:
MCB.00052-07v1
27/12/4207