Proteomic analysis yields an unexpected trans-acting point in control of the human sympathochromaffin phenotype

Abstract

Background: The secretory protein chromogranin A (CHGA) plays a necessary role in formation of catecholamine storage vesicles and gives rise to a catecholamine release-inhibitory fragment. Because genetic variation in the proximal human CHGA promoter predicts autonomic function and blood pressure, we explored how a common genetic variant alters transcription of the gene.

Methods and results: Bioinformatic analysis suggested that the common G-462A promoter variant (rs9658634) may disrupt as many as 3 transcriptional control motifs: LEF1, COUP-TF, and PPARγ-RXRα. During electrophoretic mobility shifts, chromaffin cell nuclear proteins bound specifically to the A (though not G) allele of CHGA promoter G-462A. On oligonucleotide affinity chromatography followed by electrospray ionization followed by 2-dimensional (tandem) mass spectrometry analysis of A allele eluates, the transcription factor LEF1 (lymphoid enhancer-binding factor-1) was identified. Interaction of LEF1 with the A allele at G-462A was confirmed by supershift. On cotransfection, LEF1 discriminated between the allelic variants, especially in chromaffin cells. Allele specificity of trans-activation by LEF1 was transferable to an isolated G-462A element fused to a heterologous (SV40) promoter. Because β-catenin (CTNNB1) can heterodimerize with LEF1, we tested the effect of cotransfection of this factor and again found A allele-specific perturbation of CHGA transcription.

Conclusions: Common genetic variation within the human CHGA promoter alters the interaction of specific factors in trans with the promoter, with LEF1 identified by proteomic analysis and confirmed by supershift. Coexpression experiments show functional effects of LEF1 and CTNNB1 on CHGA promoter. The findings document a novel role for components of the immune and WNT pathways in control of human sympathochromaffin phenotypes.

Figure 1

Identification of binding motifs spanning position G-462A in the human CHGA promoter. A. Putative factors binding the CHGA promoter at the G-462A SNP position. A region of 31-bp spanning the -462A SNP in the variant CHGA promoter was analyzed using the software Mapper (http://bio.chip.org/mapper) for putative binding motifs. Shown are the hits with a score >3, the associated model and its properties, as well as the motif alignments (modelled motifs are represented between *-> <-*, with partial matches shown by +). Vertical arrows indicate the position of the G-462A variant (in italics). On the minus strand, the same variant is C/T. B. Schematic representation of the putative binding motifs and factors. Only the + strand (-479 to -449) of the region spanning the -462 SNP (bold) is represented, together with the putative motifs on the + strand (+, above the sequence) or on the reverse complement (−, below the sequence). C. LEF-1 homology match at human CHGA promoter variant G-462A across primate species.

Figure 1

Identification of binding motifs spanning position G-462A in the human CHGA promoter. A. Putative factors binding the CHGA promoter at the G-462A SNP position. A region of 31-bp spanning the -462A SNP in the variant CHGA promoter was analyzed using the software Mapper (http://bio.chip.org/mapper) for putative binding motifs. Shown are the hits with a score >3, the associated model and its properties, as well as the motif alignments (modelled motifs are represented between *-> <-*, with partial matches shown by +). Vertical arrows indicate the position of the G-462A variant (in italics). On the minus strand, the same variant is C/T. B. Schematic representation of the putative binding motifs and factors. Only the + strand (-479 to -449) of the region spanning the -462 SNP (bold) is represented, together with the putative motifs on the + strand (+, above the sequence) or on the reverse complement (−, below the sequence). C. LEF-1 homology match at human CHGA promoter variant G-462A across primate species.

Figure 2

In vitro binding specificity of the two alleles of human CHGA G-462A region with PC12 nuclear extract. A. Sequences of the probes used for EMSA experiments. The SNP is indicated in red. B. Electro-Mobility Shift Assay (EMSA) of the two alleles G-462 and -462A with nuclear extract of PC12 cells. Biotinylated probes from (A) were incubated with nuclear extract and/or non-biotinylated competitor as indicated and the resulting complexes analyzed as described in M&M. PC12 nuclear extract alone shows a non-specific (NS) band (arrow). Protein:DNA complexes (asterisk) and free probes (brackets) are indicated.

Figure 3

Purification of sequence-specific DNA-binding proteins by affinity chromatography and identification by LC-MS/MS. A. Schematic representation of the DNA sequence used for the purification experiment. The sequence is identical to that of Figure 2, but contains a single 3′-TEG-Biotin tag. SNP G-462A allele is indicated in bold. B. Coomassie blue SDS-PAGE showing the elution fractions from both alleles, as well as beads alone, used for analysis by LC-MS/MS. The arrow indicates an M_r ~64 kDa band seen only in the “A” lane; of note, LEF1 typically migrates with a apparent M_r ~55 kDa (e.g. http://www.scbt.com/datasheet-8591-lef-1-n-17-antibody.html). C. Summary of the peptides identified by LC-MS/MS from the elution fractions. D. LC-MS/MS spectra identifying the peptide ESAAINQILGR, corresponding to amino acid residues 225–235 in rat LEF1 <http://www.uniprot.org/uniprot/Q9QXN1>. Arrows indicate the sequence reads from MS/MS.

Figure 4

Identification of the binding factor of the A-allele by EMSA. A. Supershift experiment using the biotinylated A-allele probe from Figure 2. The probe was sequentially incubated with PC12 nuclear extract and antibodies directed against PPARγ, COUP-TF or LEF1 (Santa Cruz Biotechnology; see Methods). An antibody directed against Pax6 was used as a negative control (sc-32766X). The Protein:DNA complexes (asterisk, *) and the Protein:DNA:antibody complexes (arrowheads) are indicated. B. G and A allele -462 region and recognition motifs of PPARγ-RXRalpha, COUP-TF and LEF1/TCF1. Wide, thin and dotted lines indicate respectively a perfect match with the base in the motif, a second choice base, and third or fourth choice base. The variable base is presented in bold. WebLogo profiles of consensus base preference are from Chip-Mapper <http://mapper.chip.org/>.

Figure 5

Effect of LEF1 on the regulation of CHGA promoter transcriptional activity in chromaffin (rat PC12) cells. A. Luciferase constructs containing the 1.2-kbp CHGA proximal promoter (left panel) with the -462 SNP A allele (Hap-1/-462A) or G allele (Hap-1/-G462) or the SV40 constructs illustrated in panel B were co-transfected with a plasmid expressing the LEF1 transcription factor or an empty plasmid control (No TF; pcDNA3.1) in PC12 cells. 24h post transfection, the regulatory effects of LEF1 were analyzed by gene reporter assay in as described in Methods. Each transfection experiment included a transfection efficiency control plasmid, pRL-TK (Promega). CHGA promoter transcriptional activity is expressed as the ratio of Firefly luciferase/Renilla luciferase activity with co-transfection of LEF1 plasmid over the negative control pcDNA3.1 (in %). The bars represent standard errors. B. Schematic representation of the constructs used in the assays. The 31-bp fragments spanning the position -462 (red or blue) of the CHGA promoter were inserted 19 bp upstream of the SV40p, in the plasmid pGL3-Promoter.

Figure 6

Effect of beta-catenin (CTNNB1) on CHGA promoter transcriptional activity in chromaffin (rat PC12) cells. Luciferase constructs containing the 1.2-kbp CHGA proximal promoter (left panel) with the -462 SNP A-allele (Hap-1/-462A) or G-allele (Hap-1/-G462) were co-transfected with plasmids expressing either CTNNB1 or LEF1 in PC12 cells. 24h post transfection, the regulatory effects of LEF1 were analyzed by gene reporter assay in as described in M&M. Each transfection experiment included a transfection efficiency control plasmid, pRL-TK (Promega). CHGA promoter transcriptional activity is expressed as the ratio of Firefly luciferase/Renilla luciferase activity with co-transfection of CTNNB1 or CTNNB1 and LEF1 normalized to the basal state of the promoter (co-transfection with pcDNA3.1). The bars represent standard errors.