| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Molecular Genetics of Cardiovascular Disease, Leibniz-Institute for Arteriosclerosis Research, University of Münster, Münster, Germany;
University Hospital Münster, Internal Medicine and Nephrology D, Münster, Germany;
INSERM, UMR S 525, Paris, France; and
Faculty of Health, Medicine, and Life Science, Maastricht University, Maastricht, The Netherlands
1 Correspondence: Leibniz-Institute for Arteriosclerosis Research at the University of Münster, Department of Molecular Genetics of Cardiovascular Disease, Domagkstraße 3, 48149 Münster, Germany. E-mail: brandher{at}uni-muenster.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and interferon consensus site-binding protein/interferon-regulating factor 8 interacted only with the C allelic promoter portion. Furthermore, genotyping in a case-control study for essential hypertension (n=745 hypertensive patients; n=769 normotensive control subjects) for this variant revealed an odds ratio for hypertension of 0.73 (95% confidence interval 0.58–0.91, P=0.006) associated with the T allele, and normotensive subjects carrying the protective T allele displayed a significant decrease in diastolic (P=0.036) and systolic (P=0.024) blood pressure levels. We here report detection of a functional enhancer module in the upstream IGF1 promoter region, which might play a key role in local IGF1 bioavailability. Whether –1411C>T is also associated with other IGF1-related disease phenotypes should be evaluated further in population studies.—Telgmann, R., Dördelmann, C., Brand, E., Nicaud, V., Hagedorn, C., Pavenstädt, H., Cambien, F., Tiret, L., Paul, M., Brand-Herrmann, S.-M. Molecular genetic analysis of a human insulin growth factor 1 promoter P1 variation.
Key Words: gene expression • hypertension • bandshift assay • reporter gene • transcription factor • functional analysis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. With respect to vascular physiology, IGF1 has complex properties (2
, 3)
: it increases nitric oxide generation in an endothelium-dependent way (4
, 5)
, prevents apoptosis (6)
, and stimulates proliferation as well as migration of vascular smooth muscle cells (VSMCs) (7
, 8)
. Impaired IGF1 protection against apoptosis may promote VSMC loss and plaque instability in atherosclerosis (9
, 10)
. In VSMCs, IGF1 induces expression of the angiotensin II type 1 receptor, a paracrine effect linking IGF1 to the renin-angiotensin-aldosterone system (11)
.
The molecular basis of tissue- and stimulation-specific IGF1 gene expression is only partly understood. IGF1 is encoded by a single copy gene, residing on chromosome 12 (12
, 13)
and consisting of six exons and five introns (14
, 15)
. Its transcription is controlled by a complex system comprising two alternative promoters, P1 and P2 (located in intron 1), and individual splicing events, which affect the stability of the transcript and the subcellular localization of the product. Both promoters are TATA-less, lacking defined transcriptional start points and also GC-rich areas or CpG-islands. For many TATA-less promoters, parts of the 5'-untranslated region (UTR) of exon 1 are essential for effective gene activation (16)
. Distal portions of P1 contain elements required for cell type-specific and inducible control (17)
, but very little is known about the identity and cooperation of the factors involved. Because the –1411C>T variant resides in a region of several putative consensus binding sites, we performed an in-depth analysis of the molecular function of this variant in reporter gene studies and bandshift assays under cAMP and phorbol 12-myristate 13-acetate (PMA) stimulatory regimens and in cell lines including HEK293T, HepG2, SaOs-2, and differentiated and nondifferentiated THP-1. As there is accumulating evidence for a role of IGF1 in blood pressure regulation (18
, 19)
, we assayed the molecular functionally relevant –1411C>T polymorphism in the distal promoter region of IGF1 P1 in a large case-control association study for essential hypertension [Projet d’Etude des Gènes de l’Hypertension Artérielle Sévère à modérée Essentielle (PEGASE), n=1514].
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and TRANSFAC 7.0 database. Sequence homology and conservation, conserved transcription factor binding sites (TFBSs), and repetitive motifs were analyzed by use of the University of California Santa Cruz genome browser (http://genome.ucsc.edu/) and the Evolutionary Conserved Regions browser (http://ecrbrowser.dcode.org).
Cell culture
HepG2, SaOs-2, and HEK293T cells were maintained in Dulbecco’s modified Eagle’s medium (Sigma, Geissendorf, Germany), 10% fetal calf serum (FCS) (PAA, Cölbe, Germany), 100 U/ml penicillin, and 100 µg/ml streptomycin (Sigma). In this case, cells were stimulated with 0.5 mM 8-bromo-cAMP (Biolog, Bremen, Germany) or 10–8 M PMA (Fluka, Seelze, Germany). THP-1 cells were maintained in RPMI 1640, 10% FCS, 100 U/ml penicillin, 100 µg/ml streptomycin, and 1x modified Eagle’s medium amino acid solution (Sigma). Differentiation of THP-1 cells was performed by stimulation with 10–8 M PMA. Cells were harvested when they were fully adhesive and had changed morphologically into a macrophage-like phenotype.
Diagnostic polymerase chain reaction (PCR)
Crude RNA (5 µg) was used for generation of cDNA (Fermentas, St. Leon-Rot, Germany). Primers for exon 1 (S1: TCCAACCCAATTATTTAAGTG), exon 2 (S2: CCACCCACAAAGCAGCACATG), exon 3 (S3: TGCACACCATGTCCTCCTCG; A3: CCACACACGAACTGAAGAGC), and exon 4 (A4: AGAGCGAGCTGACTTGGCAG) were used for semiquantitative diagnostic PCR (REDTaq polymerase; Sigma) of IGF1 gene expression (Fig. 1A
). Intactness of cDNAs was controlled by diagnostic PCR for ribosomal protein 27 (sense primer: CCAGGATAAGGAAGGAATTCCTCCTG; antisense primer: CCAGCACCACATTCATCAGAAGG).
|
Reporter constructs and expression vectors
Plasmids P1-IGF1-C/luc and P1-IGF1-T/luc were constructed as follows. Human genomic DNA from patients bearing either the C or the T allele was used for amplification. Primers were located at position –1593 5' (TCTTAGTGTGCTCTTTGGGTC) and immediately upstream of the start ATG (ATTAACCACACAGAAGACTC); hence, fragments contained 165 bp of 5'-UTR and 1428 bp of 5'-flanking region. Amplicons were subcloned into vector pGL3-basic (Promega, Mannheim, Germany). For vectors p1411Cprom and p1411Tprom, plasmids P1-IGF1-C/luc and P1-IGF1-T/luc were digested with SacI and excised fragments (–1593/–1350) were subcloned into pGL3-promoter (pGL3-prom; Promega). Vectors pSG5-C/EBP
, pSG5-C/EBPβ, pSG5-C/EBP, and pSG5-hEts2 were a kind gift from Dr. B. Gellersen (Endokrinologikum, Hamburg, Germany). Generation of vectors pIRF1 and pIRF2 was described elsewhere (21)
. Vector pCMV-E1A-F was a kind gift from Prof. K. Yoshida (Sapporo, Japan) (22)
.
Transient transfection assays
HEK293T and monocytic THP-1 cells were transfected with Lipofectamine 2000 (Invitrogen, Karlsruhe, Germany) and SaOs-2 and HepG2 cells with Effectene (Qiagen, Valencia, CA, USA), and luciferase activity was determined using a Sirius luminometer (Berthold, Pforzheim, Germany). Differentiated THP-1 cells were transfected with Nanofectin (PAA). Cotransfection experiments were performed with Lipofectamine at a reporter to effector plasmid ratio of 3:1.
Electrophoretic mobility shift assay (EMSA) and supershift assay
Nuclear protein extracts were harvested as described elsewhere (23)
. Oligonucleotides harboring either the C or the T allele (AGTCCCCTGAGAGTCATGCGGAAAAAAAAAAA or AGTCCCCTGAGAGTCATGTGGAAAAAAAAAAA, respectively) were synthesized at a coupling efficiency of >98.5% and purified twice by HPLC (Iba, Göttingen, Germany). Single-stranded oligonucleotides were TdT-labeled with 11-Biotin-dUTP (Fermentas) and annealed overnight. Efficiency of labeling and annealing was ascertained by PAGE. Per EMSA reaction, 5-µg nuclear protein extracts were incubated with poly[dI·dC] (Amersham, Braunschweig, Germany) and, where appropriate, a 200-fold molar access of unlabeled oligonucleotide as a specific competitor. After native PAGE and blotting onto polyvinylidene difluoride membranes (Millipore Corporation, Billerica, MA, USA), bands were visualized by the streptavidin-horseradish peroxidase reaction (LightShift kit; ThermoFisher, Bonn, Germany). For supershift assays, EMSA reactions were extended by incubation with specific antibodies (2 µg of IgG per sample of anti-CCAAT/enhancer-binding protein (C/EBP) , anti-C/EBPβ (C-19), anti-C/EBP (C-22), anti-c-Jun N, anti-Ets1/Ets-2 (C-275), or anti-interferon consensus site-binding protein (ICSBP) (C-19) (St. Cruz Biotech, Heidelberg, Germany).
Study population
The population has been described in detail previously (24)
(see also http://genecanvas.ecgene.net/news.php). In brief, PEGASE was performed as follows. Male and female patients with essential hypertension were recruited in 15 regions of France by 139 general practitioners of the EURAXI network, which has a long experience in the field of cardiovascular drug trials. Criteria for inclusion in the study was a diastolic blood pressure 
105 mmHg without antihypertensive treatment or 100 mmHg with treatment (92% were treated), a creatinine level <120 mM, kaliemia >3.7 mM, and the absence of proteinuria. Patients with secondary hypertension were excluded. Blood pressure was measured three times using a mercury sphygmomanometer according to the World Health Organization/International Society of Hypertension recommendations and the mean of the three values was used in the analyses. Participants had to be aged <60 yr, to be of European origin, and to have both parents born in France and four grandparents born in Europe; they had to be free of myocardial infarction, stroke, or heart failure. Normotensive control subjects (no antihypertensive treatment; exclusion of subjects with systolic blood pressure 160 or diastolic blood pressure 95) were recruited in Centres for Preventive Medicine all over metropolitan France and were matched to hypertensive patients for sex, region of birth, and strata of age of 5 yr. Their parents had to be born in France and their four grandparents had to have been born in Europe. Ethics approval for the study was obtained from the relevant committees, and each subject gave consent for the study.
Genotyping
For IGF1 –1411C>T, genotyping of the PEGASE study population was performed using allele-specific oligonucleotides, as described previously (25)
. Technical details, including oligonucleotide sequences and hybridization conditions, are available online (http://genecanvas.ecgene.net/news. php).
Statistical methods
Hardy-Weinberg equilibrium was tested by a 
2 test (1 degree of freedom) separately in hypertensive and normotensive subjects. Allele frequencies were calculated from genotype frequencies by allele counting. The odds ratio (OR) and 95% confidence interval (CI) for hypertension associated with carrying of the T allele were calculated by logistic regression adjusted on age and sex. Homogeneity of the OR was tested in males and females and for age younger and older than the median (45 yr) by introducing the corresponding interaction term. Systolic and diastolic blood pressure mean levels were calculated in normotensive subjects for each genotype, and the IGF1 –1411C>T effect was tested assuming an additive effect of alleles (i.e., CC was coded as 0, CT as 1, and TT as 2 in this model). Transfection experiments were repeated at least 3 times in triplicate for every plasmid. Significance was calculated by an unpaired, 2-tailed t test (CI=95%); significance levels were set at P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
), with both promoters P1 and P2 active in parallel in several cell lines (Fig. 1B
). Analyses of the C allele- or T allele-carrying promoter P1 sequences (Fig. 1C
) with the net-based program AliBaba2.1 predicted several binding sites for different transcription factor families. Both alleles showed a sequence similarity for members of the bZIP family of activator protein 1 (AP1) factors, directly adjacent 5' to the position of the SNP. For both alleles, the SNP is located in a predicted binding site for C/EBP transcription factors (consensus sequence A/GTTGCGC/TAAC/T). For the C allele, the sequence was predicted to be a putative binding site for interferon-consensus site binding protein/interferon-regulating factor 8 (ICSBP/IRF8), a sequence also recognized by other IRFs of the class of tryptophane-clustered factors within the helix-turn-helix family. This binding pattern is lost for the T allele. Odd-skipped, a factor predicted for the T allele, is a member of the Drosophila Krueppel-like factors. Recently, a related human homologue (OSR1) was identified with expression strictly restricted to adult colon, small intestine, prostate, testis, and fetal lung (26)
and was therefore omitted from further experiments. Both alleles revealed sufficient similarity for binding of hepatic nuclear factor 3 (HNF-3), a group of more than 30 members of the forkhead/winged helix family, 3' of the SNP position. No program predicted the presence of a putative binding site for members of the Ets-like proteins (GxGGAA), which was visually accounted for. Homology conservation between species was analyzed by use of the ECRbrowser (http://ecrbrowser.dcode.org) (Fig. 1D, E
).
–1411C>T resides in a cell- and differentiation-specific enhancer region
In the human embryonic kidney cell line HEK293T (Fig. 2A
), transient transfection was very effective, and the shuttle vector pGL3-prom (containing a minimal promoter) was also activated to comparably high levels. Despite that condition, both allelic promoters revealed a strong transcriptional activation over the shuttle vector (P=0.003 for the C allele; P
0.0001 for the T allele). This strong enhancer quality was irrespective of the orientation of the fragment, in accordance with the definition of enhancer regions (data not shown). In this poorly differentiated cell line, the T allele showed significantly elevated transcriptional activity compared with the C allele (P=0.0097).
|
In SaOs-2 osteosarcoma cells, activation of the positive but not the negative control vector reached levels comparable to those in HEK293T, suggesting that in more differentiated cells IGF1 transactivation is more specifically regulated. The enhancer portion was also used in this cell line, but in contrast to HEK293T cells, the activity of the –1411T allele was significantly lower than that of the C allele (P=0.0001 without stimulation, P<0.0001 with cAMP, and P<0.0001 with PMA) (Fig. 2B-D
). Furthermore, in unstimulated as well as in cAMP-stimulated cells, the T allele exerted mild activation over the shuttle vector pGL3-prom (P=0.0002 and P<0.0001, respectively), whereas in PMA-stimulated cells, its activity was lower than that of the negative control (P=0.0002).
In the hepatoma cell line HepG2, the C allele was strongly active, but the presence of the T allele completely abrogated the transcriptional activity, irrespective of the stimulatory regimen applied (P<0.0001 without stimulation and with cAMP and P=0.0003 with PMA) (Fig. 3
). C allele enhancer activity responded to stimulation. There was no effect on utilization of the C allele in cAMP-stimulated cells but a noticeably less strong activation over the shuttle vector was seen in PMA-stimulated cells (Fig. 3C
). The T allele was associated with down-regulated basal transcriptional activity conveyed by the minimal promoter of pGL3-prom, indicating the presence of an active silencing capacity.
|
–1411C>T alters the interaction of HepG2 nuclear proteins with the IGF1 P1 enhancer fragment
To identify DNA-protein interactions underlying the differential allelic activity, we concentrated on HepG2 cells in which these differences were most distinct. Oligonucleotides of either allelic portion, comprising the 5' AP1 site, the 3' HNF-3 site, and the flanked region of different binding sequences harboring the SNP, were used as nuclear probes in EMSA experiments. Nuclear proteins from HepG2 cells were able to interact with both alleles but resulted in profoundly different binding patterns (Fig. 4
), indicating that this portion of the IGF1 promoter was bound by more than one nuclear factor and that the presence of the SNP interfered with their composition.
|
The composition of factors assembling at a transcriptional module depends on the stimulatory patterns in a given pathophysiological context. HepG2 cells were, therefore, maintained under stimulation with cAMP and PMA for 24 h. These stimuli altered the binding pattern individually but differentially for both alleles, indicating a specific response to external stimuli, which resulted in altered levels of IGF1 production from each allele.
C/EBP and ICSBP/IRF8 specifically interact with the C allele of the IGF1 promoter
In supershift experiments using HepG2 nuclear proteins, antibodies against C/EBP family members failed to interact with the T allele (data not shown), but in unstimulated cells, addition of an antibody against C/EBP resulted in a significant alteration of the binding pattern (Fig. 5A
). For the C allele, a faster migrating band remained unchanged, whereas a second, slower moving, band was significantly reduced in the presence of the antibody and appeared to be strongly repressed. This result further strengthens the suggestions that the C allele constitutes a transcriptional module bound by more than one protein and that addition of the anti-C/EBP antibody removes this protein from a complex, leading to its higher electrophoretic mobility.
|
In addition, antibodies against ICSBP/IRF8 supershifted a specific complex binding to the C allele (Fig. 5B
), which was exclusively observed under PMA stimulation and which was not present in the T allele-containing promoter (data not shown). In addition, antibodies against c-Jun supershifted a band exclusively when the T allele was used as a probe (Fig. 5C
), suggesting that this variant alters the affinity toward different components of the AP1 protein complex.
The P1 enhancer region integrates the transcriptional interaction of different stimulatory pathways
To investigate the effect of transcription factors for which consensus binding sites were predicted but remained undetected in HepG2 or SaOs-2 cells (data not shown), we performed cotransfection experiments with HEK293T cells in which the –1411T allele did not exert a transcriptionally repressive action. For these overexpression studies, we concentrated on factors whose putative binding sites were either affected by the presence of the SNP (IRF1 and IRF2) or not affected, but in which the variant is located [Ets-like family: Ets-2 and E1A-F (PEA3); and C/EBP family: C/EBP, C/EBPβ, and C/EBP].
Overexpression of IRF1 had no differential effect on allelic promoter usage (Fig. 6A
); IRF2 reduced the activity of the T allele, but this association failed to reach statistical significance (Fig. 6B
). Hence, factors of the interferon responsive family are able to functionally interfere with this portion of the IGF1 promoter. Overexpression of E1A-F (PEA3) had no effect (Fig. 6C
), whereas coexpression of Ets-2 resulted in a drastic, significantly different, activation by strongly favoring the C allele (Fig. 6D
). Thus, the enhancer portion was able to differentially interact with members of the Ets-like family and discriminate between members sharing similar binding specificities.
|
Overexpression of three members of the C/EBP family, one of which (C/EBP) can either be expressed as a transcriptional activator (liver-enriched activatory protein), or as a transcriptional repressor (liver-enriched inhibitory protein), did not reveal isoform-specific differences. The T allele, which was slightly more active in untransfected cells, was neither significantly activated nor repressed by overexpression of either transcription factor in HEK293T cells (data not shown).
Use of the IGF1 enhancer region and differential binding by THP-1 monocytes and macrophages
Given the broad range of IGF1-expressing tissues and cell types, we further investigated whether the enhancer region was also differentially used in monocytes and derived macrophages. In contrast with any of the other cell lines examined, in undifferentiated THP-1 monocytes the –1411C allele of the enhancer fragment of IGF1 P1 failed to exert any activatory effect (P=0.482). However, as in other cells, the presence of the T allele significantly reduced it (=0.037), indicating an active repressing effect of the variant in monocytes (Fig. 7A
). Confirming these transfection results, proteins from undifferentiated THP-1 monocytes completely failed to bind to the C allele but complexed with the T allele, possibly causing its transcriptional repression. Conversely, differentiated THP-1 macrophages acquired the ability to interact with both alleles in a similar pattern of two slow-migrating bands, as shown in EMSA (Fig. 7C
) without losing the T allele-specific interaction. The appearance of these proteins in macrophages resulted in a faint but noticeable activation of the C allele but failed to rescue the T allele from its transcriptional repression (Fig. 7B
). This finding indicates that the influence of protein binding to the T allele is strong enough to impede the activational effect in THP-1 macrophages.
|
Population genetic data
Supplemental Table 1 depicts the characteristics of the participants in the PEGASE study. Genotype and allele frequencies are shown in Supplemental Table 2. There was no departure from Hardy-Weinberg equilibrium in normotensive control subjects. There was a significant difference in genotype frequency between hypertensive and normotensive control subjects (P=0.019), leading to an OR for hypertension of 0.73 (95% CI 0.58–0.91; P=0.006) associated with carrying of the T allele. This protective effect was seen in males and in females and for age younger and older than 45 yr (P=0.86 and P=0.76 for interaction, respectively). More specifically, in normotensive control subjects, there was a significant decrease in diastolic (P=0.036) and systolic (P=0.024) blood pressure level with the number of T alleles (Supplemental Table 3). There was no significant association between allele frequency and parental history of myocardial infarction or stroke in hypertensive subjects.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, HepG2 (17)
, and HeLa cells and RAW 264.7 macrophages (27)
. Congruently, exon 1 and its directly adjacent 5'-flanking region are highly conserved among species [position chromosome 12: 101398080–101398709 is still 79.2% homologous and 84.4% for the first 255 bp of 5'-flanking region compared with the frog (Xenopus naevis)]. In mammals, including mice, rats, cows, and canines, this conserved region extends to 500–750 bp upstream of the 5'-UTR with very high homologies (85.3, 87.2, 84.9, and 87.2%, respectively). Further upstream, between chromosome positions 12:101399137 and 101399757, containing the position of the –1411C>T variant and covering the part of the IGF1 promoter P1 that was reported to be transcriptionally active in serial deletion analyses, especially in response to activation of the protein kinase C pathway by PMA (17)
, a second region of increased homology is present in mice (74.6%), rats (76.1%), cows (77.4%), and canines (77.6%). Despite the high homology of the embedding region, the extended stretch of 19 adenosine residues, interspersed by one G, residing directly 3' of and containing the –1411C>T variant at its flank, is found exclusively in primates (Macaca mulatta and Pan troglodytes) (Fig. 1E
). It is likely that the enhancer region identified contributes to the fine-tuning of spatial rather than systemic human IGF1 promoter activity and is involved in control of the tightly conducted local IGF1 expression and its autocrine and paracrine functions, enabling the respective tissues to respond quickly and efficaciously to physiological stimuli independent of circulating IGF1 levels. The importance of this control was recently supported by two findings: the "liver-specific targeted IGF1 deletion" mouse model, which despite presenting 
75% less circulating IGF1 levels, showed unchanged peripheral IGF1 mRNA levels and normal postnatal growth (28)
; and Klover and Hennighausen (29)
demonstrated recently that deletion of skeletal muscle signal transducer and activator of transcription 5 in Stat5MKO mice led to a 60% reduction in IGF1 mRNA in this tissue and merely a modest decline in circulating IGF1; this small decrease in circulating IGF1 obviously did not contribute to the observed growth retardation of Stat5MKO mice.
In the present study, we assayed the molecular functional –1411C>T variant in an appropriately powered and well phenotyped and matched Caucasian case-control study for essential hypertension. The strength of the population study originates from its very specific blood pressure phenotype in that hypertensive subjects had to be free of hypertension-associated diseases such as myocardial infarction, stroke, or heart failure. This exclusion ensures that we are specifically studying the hypertensive component of the complex disease phenotype and not associated diseases. It is noteworthy that we were able to show not only a significant "protective" case-control difference in carrying of the –1411T allele (qualitative association) but also decreased systolic as well as diastolic blood pressure levels in –1411T allele carriers (quantitative association). These findings are in line with a previous report by Nagy et al. (30)
, who demonstrated that polymorphic microsatellite markers at the IGF-1 gene locus were linked to phenotype systolic blood pressure and that the linkage with blood pressure was remarkably strong because the QTL effect was demonstrable in all IGF-1 positions.
Alternative promoters are a structural element of the majority of human protein coding genes and represent the molecular basis for tissue-specific expression or differentiation-dependent gene expression (31)
. Our data show that –1411C>T in the upstream promoter P1 resides within an enhancer region containing a cluster of several partly overlapping TFBSs, capable of integrating transcriptionally active stimuli from different signal transduction pathways in a cell type-specific manner. This cluster can be divided into three regions: a downstream site for tissue-specific factors HNF-3, an upstream site for AP1 proteins, and, in its center, linking HNF-3 and AP1 in a relay-like fashion, triplicate binding sites for factors specific for IRF or involved in inflammatory signal integration (Ets-like and C/EBP proteins) (Fig. 8
). The pathophysiological activity of C/EBP transcription factors within a given cellular context depends on the actual availability of these factors, their interaction with other partners, and their activational status (32)
.
|
In our supershift experiments in HepG2 cells, antibodies against C/EBP and ICSBP/IRF8 interfered with a protein complex binding to IGF1 P1, but exclusively to the C allele. ICSBP/IRF8 sites also serve as binding sites for the IRF family (33)
. IRF1, a transcriptional activator mediating IFN-stimulated transcription, did not show significant activation of either construct nor did the constructs differ in their response. Conversely, IRF2, a transcriptional repressor and a physiological antagonist to IRF1, noticeably reduced the activity of the T allele-bearing promoter. It is, therefore, tempting to assume that the ICSBP/IRF8 site, predicted by AliBaba2.1 even at high stringency presets and residing adjacent to the polymorphic site, conveys a transcriptional susceptibility for the IRF pathway to the IGF1 promoter P1 in nonimmune system cell lineage. The core ICSBP/IRF8 site itself is affected by the presence of the variant and cotransfection experiments with IRF1 and IRF2 suggest that its affinity for members of the IRF family is altered, which could in turn result in a differential transcriptional performance under respective physiological conditions.
Whereas cotransfections with E1A-F (PEA3) did not reveal any differential allelic effect, Ets2 overexpression, being highly expressed in mature macrophages on stimulation (34)
, led to a significant reduction of transcriptional activity related to the T allele-bearing promoter. Therefore, the IGF1 promoter fragments are able to interact with and discriminate between members of the Ets family. There is ample evidence of interactions of members of the C/EBP family with AP1 binding proteins, especially with c-Jun (35)
, and it is conceivable that AP1 binding proteins either bind to the AP1 consensus site independently of –1411C>T or that the altered recruitment of C/EBP proteins has a direct effect on recruitment and/or function of AP1 proteins.
The –1411C allele was inactive in THP-1 monocytes and mildly, but significantly, activated when THP-1 cells differentiate into macrophages. Conversely, the T allele was inactive in monocytes, possibly due to a T allele-specific interaction with monocyte nuclear proteins and failed to reach C allele expression levels in differentiated cells. We, therefore, suggest that the T allele might lead to reduced IGF1 expression in macrophages with possible consequences of attenuating autocrine effects. Given the involvement of inflammatory factors such as C/EBP, ICSBP/IRF8, Ets2, and IRF2, it is tempting to suggest that while retaining sufficient IGF1 availability, the T allelic promoter might dampen inflammatory- or stress-induced local IGF1 overexpression. Less transcriptional activity of the T allele does not equal a less protective spatial effect of IGF1 in the progression of vascular deleterious phenotypes: IGF1 was recently shown to be incapable of reverting myofibrillar protein loss, induced by the proinflammatory cytokines tumor necrosis factor- and IFN
(36)
, while it retained its full activational effect on FOXO and glycogen synthetase kinase-3β-phosphorylation.
In conclusion, –1411C>T is located within a transcriptional module exerting partly drastic functional, allele-specific effects that lead to a differential DNA-protein binding pattern. The observed protective effect of the –1411T allele with respect to the blood pressure level in normotension and essential hypertension in a case-control design should be further evaluated in appropriately designed and powered prospective cohorts.
|
ACKNOWLEDGMENTS |
|---|
Received for publication July 22, 2008. Accepted for publication November 26, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Hagedorn, R. Telgmann, C. Dordelmann, B. Schmitz, S. Hasenkamp, F. Cambien, M. Paul, E. Brand, and S. M. Brand-Herrmann Identification and Functional Analyses of Molecular Haplotypes of the Human Osteoprotegerin Gene Promoter Arterioscler Thromb Vasc Biol, October 1, 2009; 29(10): 1638 - 1643. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
