Functional analysis of -351 interleukin-9 promoter polymorphism reveals an activator controlled by NF-kappaB

Abstract

Genetic studies have shown linkages for asthma to the chromosomal region 5q31-q33 in humans that includes the IL-9 gene. An A-to-G base substitution has been identified at bp -351 in the IL-9 promoter. The role of this polymorphism in IL-9 promoter function was assessed utilizing CD4+ T cells purified from individuals with one or two of the G alleles in comparison to those homozygous for the wild-type A. The presence of an A at -351 (A allele) increased mitogen-stimulated IL-9 transcription twofold in comparison to subjects with one or two G alleles at this position. Binding of nuclear extract proteins from IL-9-producing human cell lines to DNA sequences including this base exchange demonstrated specific binding of the transcription factor NF-kappaB. Binding of NF-kappaB to the IL-9 promoter was confirmed in vivo using the chromatin immunoprecipitation assay. Recombinant NF-kappaB bound to a promoter fragment with the A allele with fivefold higher affinity than it did to a promoter with the G allele. Individuals carrying the A allele of the IL-9 promoter display increased synthesis of IL-9, which may result in strong Th2 immune responses and a modulation of their susceptibility to infectious, neoplastic, parasitic or atopic disease.

Figure 1

Diagram depicting the IL-9 -351 polymorphism showing the wild type (‘AT’) and base exchanged (‘GC’) forms of the promoter.

Figure 2

IL-9 mRNA production from purified CD4+ T cells. T cells were enriched from peripheral blood of subjects, containing either the A/A or G/A alleles for the IL-9 -351 polymorphism, using magnetic bead affinity purification and cultured for 24 hrs in the presence or absence of PMA (1.0 μg/ml), PHA (1.0 μg/ml) or both. Cells were collected and RNA isolated. IL-9 mRNA levels were measured using quantitative PCR and separated according to IL-9 genotype: A/A (n=3) and G/A (n=3). P<0.05 is considered statistically significant.

Figure 3

NF-κB-specific consensus competition. Human PMA-stimulated T cell (Jurkat) nuclear extract was allowed to interact with ³²P-labeled oligomers. The labeled probes represent the wild type (‘AT’) and base exchanged (‘GC’) forms of the IL-9 promoter at base -351. Non-specific binding was mitigated by addition of poly dIdC. FP represents free probe without addition of nuclear extract. Competition of shifted bands was performed with increasing concentrations (100-1000 fold) of unlabeled oligonucleotide. Bands were eliminated by competition with excess unlabeled NF-κB consensus probe, but not a mutated form of the NF-κB probe.

Figure 4

Supershift assay demonstrates both p50 and p65 are part of the NF-κB complex on the IL-9 promoter. Jurkat nuclear extract was allowed to interact with ³²P-labeled oligomers. The labeled probes represent the wild type (‘AT’) and base exchanged (‘GC’) forms of the IL-9 promoter at base ^-351. Anti-NF-κBp50 and anti-NF-κBp65 supershift the higher molecular weight band and anti-NF-κBp50 also supershifts the lower molecular weight band. An isotype matched anti-NFAT antibody did not alter any of the shifted complexes.

Figure 5

Binding of NF-κB to the IL-9 promoter from extracts derived from CD4+ T cells. CD4 cells were isolated from peripheral blood and stimulated with PMA/PHA and nuclear proteins isolated. Nuclear extract was allowed to interact with ³²P-labeled oligomers. The labeled probes represent the wild type (‘AT’) and base exchanged (‘GC’) forms of the IL-9 promoter at base ^-351. Competition of shifted bands was performed with increasing concentrations (100-1000 fold) of unlabeled oligonucleotide. Bands were eliminated by competition with excess unlabeled NF-κB consensus probe, but not a mutated form of the NF-κB probe. Anti-NF-κBp50 and anti-NF-κBp65 supershift the higher molecular weight band and anti-NF-κBp50 also supershifts the lower molecular weight band. An isotype matched anti-NFAT antibody did not alter any of the shifted complexes.

Figure 6

Scatchard analysis of recombinant NF-κB p50 binding to the IL-9 promoter. Labeled probes were used representing the wild type (‘AT’) and base exchanged (‘GC’) forms of the IL-9 promoter at base ^-351. Increasing concentrations of recombinant NF-κB p50 were added to each probe and the percentage of each probe bound was calculated. From this the relative Ka of recombinant NF-κB p50 binding to each probe was calculated.

Figure 7

Chromatin immunoprecipitation assay performed on A) Jurkat T cells and B) purified human CD4+ T cells to measure in vivo NF-κB p50 and NF-κB p65 binding. Proteins were cross-linked to the DNA with formaldhyde and antibodies directed against NF-κB p50, NF-κB p65 or Stat6 (antibody control) were added to precipitate any protein-DNA complexes. The -control lanes were processed according to the protocol, but did not have any antibody added to the samples. The + control lane was DNA that went through sonication and reversal of the protein-DNA contacts followed by ethanol precipitation. PCR were performed on isolated DNA using primers that span the -351 site in the IL-9 promoter and analyzed by agarose gel electrophoresis.

Figure 8

Transient transfection assays were performed in the human T cell line Jurkat. A. Using electroporation, 1×10⁷Jurkat cells were transfected with 10μg of each plasmid construct. Following incubation for 48 hrs, cells were collected and luciferase activity measured from cell lysates. Basal transcription of each IL-9 promoter template was normalized relative to the activity of the -346 construct. B. Transfection was performed as described above with the exception that Jurkat T cells were stimulated with PMA/PHA for the final 6 hrs to induce IL-9 transcription. Data was normalized to the unstimulated sample for each construct. Data points represent the average of five independent experiments with error bars being calculated as standard errors of the mean.

Figure 8

Transient transfection assays were performed in the human T cell line Jurkat. A. Using electroporation, 1×10⁷Jurkat cells were transfected with 10μg of each plasmid construct. Following incubation for 48 hrs, cells were collected and luciferase activity measured from cell lysates. Basal transcription of each IL-9 promoter template was normalized relative to the activity of the -346 construct. B. Transfection was performed as described above with the exception that Jurkat T cells were stimulated with PMA/PHA for the final 6 hrs to induce IL-9 transcription. Data was normalized to the unstimulated sample for each construct. Data points represent the average of five independent experiments with error bars being calculated as standard errors of the mean.