| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Department of Biosciences and Nutrition, Karolinska Institute, Huddinge, Sweden;
Department of Medical Genetics, University of Helsinki, Helsinki, Finland; and
Department of Pathology, Tufts University School of Medicine, Boston, Massachusetts, USA
1Correspondence: Department of Biosciences and Nutrition, Hälsovägen 7–9, 141 57 Huddinge, Karolinska Institute, Stockholm, Sweden. E-mail: juha.kere{at}biosci.ki.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: complex disease • regulation of transcription • learning disability • reading impairment
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
.
Genetic linkage studies, whole-genome linkage scans, and association studies combined with identification of chromosomal translocation breakpoints in dyslexic families have suggested different regions of the genome to be involved in the etiology of dyslexia. Altogether, 9 loci, DYX1–DYX9, on chromosomes 1, 2, 3, 6, 11, 15, 18, and X, have been acknowledged by the Human Gene Nomenclature Committee. Since the publication of DYX1C1 as the first candidate dyslexia gene (2)
, significant progress has been made in the identification of new candidate genes. In only 3 yr, 5 genes have been associated with dyslexia, namely ROBO1 (3)
, KIAA0319 (4
, 5)
, DCDC2 (6
, 7)
, C2ORF3, and MRPL19 (8)
. Interestingly, ROBO1 is an axon and dendrite guidance gene, and studies using in utero RNA interference (RNAi) have shown that Dyx1c1, Kiaa0319, and Dcdc2 are necessary for the normal neuronal migration in mouse and rat neocortex (5
, 6
, 9)
. Most recent studies show also that RNAi of Dyx1c1 in rats may cause cortical and hippocampal malformations and impairments in auditory processing and spatial learning (10
, 11)
. The neuroanatomical findings are similar to small neocortical malformations in human dyslexic individuals (12
, 13)
. Little is known about the function of DYX1C1 and mechanisms by which it regulates neuronal migration or dyslexia. The DYX1C1 protein was suggested as a new chaperone or cochaperone in the ubiquitin pathway system that interacts with the U-box protein CHIP (encoded by the STUB1 gene), belonging to a family of ubiquitin-protein ligases (E3s) (14)
.
The DYX1C1 gene is located near to the region previously pointed in a number of linkage studies (15
16
17)
, and it was cloned by study of a Finnish family in which the translocation t(2;15)(q11;q21) cosegregated coincidentally with dyslexia (2
, 18)
. Sequence analysis of DYX1C1 cDNA in 20 dyslexic individuals revealed 8 polymorphisms, two of them associated with dyslexia; one of the changes (1249 G/T) introduced a stop codon truncating the protein by 4 amino acids. The second change was in the 5'UTR –3G/A (rs3743205), close to the translation initiation site. The genomic size of the DYX1C1 gene is 
78 kb, and DYX1C1 contains 10 exons with the start codon in exon 2, encoding a 420-aa protein with a molecular mass of 48 kDa and without major homology to other proteins of known function. However, three tetratricopeptide repeat domains (TPRs) were found toward the C-terminal and a p23 domain toward the N-terminal end, which are thought to play a role in protein-protein interactions (Fig. 1
). Cellular localization studies in COS-1 cells suggested that DYX1C1 is a nuclear protein (2)
, but later it has been found to be present also in the cytoplasm (9)
.
|
A number of attempts have been made to replicate the genetic association of DYX1C1 with dyslexia in other populations, with variable results (4
, 19
20
21
22
23)
. As many observations suggest that regulation, rather than coding changes, may be important for DYX1C1 function in humans, we resequenced the promoter region. In particular, two single-nucleotide polymorphisms (SNPs), rs12899331 and rs16787, in the positions –10567 T/C and –10310 C/A, respectively, were associated with dyslexia (F. Dahdouh et al., personal communication). In the present study, we characterize three possibly functional SNPs in the promoter of DYX1C1 and implicate three transcription factors in the regulation of DYX1C1.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. At 100% confluence, the cells were washed twice with cold PBS, and nuclear and total extracts were prepared essentially as described previously (43)
. For nuclear extracts, the cells from 75-mm-diameter dishes were collected by centrifugation and resuspended in approximately 3 times the packed cell volume in hypotonic buffer [10 mM Tris–HCl, pH 7.4; 10 mM NaCl; 6 mM MgCl2; 1 mM dithiothreitol (DTT); 0.4 mM PMSF; 10 mM NaF; 1 mM Na3VO4; and protease inhibitor cocktail] and incubated on ice for 10 min. The cells were lysed in a Dounce homogenizer (pestle B; Wheaton, Millville, NJ, USA), and the nuclei were collected by centrifugation followed by resuspension in 3 vol of high-salt buffer (20 mM HEPES, pH 7.9; 420 mM NaCl; 20% glycerol; 1.5 mM MgCl2; 0.2 mM EDTA; 0.2 mM PMSF; 1 mM DTT; 1 mM Na3VO4; 10 mM NaF; and protein synthesis inhibitors as above). After 30 min at 4°C, the pellets were removed by centrifugation, and supernatants containing the nuclear proteins were stored at –70°C.
For total extracts, cells were harvested in 1 ml RIPA buffer (50 mM Tris-HCl, pH 7.4; 1% Triton X-100; 150 mM NaCl; 5 mM EDTA; 1 mM PMSF; 1 mM Na3VO4; 1 mM DTT; and protease inhibitors) for 30 min (43)
. Cell lysates were centrifuged at 14,000 g for 15 min, and the resulting supernatants were collected. Protein concentrations of the nuclear and total extracts were measured with the protein assay (Bio-Rad Laboratories, Hercules, CA, USA) with BSA as standard (44)
.
Cell culture and transfections
Human neuroblastoma SH-SY5Y cells were cultured in modified essential medium (MEM) with Earle’s salts and glutaMAX-I, supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, and 100 µg/ml streptomycin at 37°C in a humidified atmosphere of 5% (v/v) CO2/air. For the transfections used for quantitative reverse transcriptase-polymerase chain reaction (qRT-PCR), cells were cultured in phenol red-free MEM medium with 4% of dextran-charcoal-treated serum. Cell culture reagents were obtained from Gibco Invitrogen Corporation (Täby, Sweden). Untransfected cells and cells transfected with empty vectors were used as controls in all the experiments. The transfections were performed either with FuGene6 (Roche, Indianapolis, IN, USA) or Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA), following the recommendations from the manufacturers.
For cellular localization experiments, SH-SY5Y cells were transiently transfected with DYX1C1-V5 fusion construct and stained with mouse anti-V5 antibody (Invitrogen) and Alexa Fluor 555-conjugated anti-mouse-IgG (Invitrogen). Nuclei were stained with 4',6-diamidino-2-phenylindole (DAPI) (Sigma, St. Louis, MO, USA). Fluorescently labeled samples were mounted with ProLong Gold Antifade Reagent (Invitrogen). The immunofluorescence images were acquired with the Meta LSM 510 confocal imaging system and an Axiovert 200 M Zeiss microscope (Carl Zeiss, Oberkochen, Germany), using an x63 1.4 numerical aperture objective and multichannel scanning in frame mode.
Electrophoretic mobility shift assay (EMSA)
The EMSA was performed according to standard protocols (45)
. Binding reactions were performed by preincubating 10 µg of SH-SY5Y of nuclear or total extracts with 0.5 µg of poly (dI-dC) and 1 µl of 100 mM DTT in a 30 µl reaction with a 4x buffer containing 80 mM HEPES (pH 7.6), 2 mM EDTA, 200 mM NaCl, 20% glycerol, 1.2 mg BSA, and 40 mM DTT for 15 min at room temperature. 32P-end-labeled double-stranded probes were used, and the mixture was incubated for 15 min at room temperature. In competition assays before the addition of the probes, unlabeled probes were incubated with a 100- and 200-M excess for 15 min at room temperature. The samples were electrophoresed on 5% nondenaturing polyacrylamide gels in 1x TBE (0.09 M Tris borate, 2 mM EDTA) at 240 V at 4°C for 3 h. The radioactive pattern was visualized by autoradiography using PhosphorImager scanning (Fuji Photo Film Co., Ltd., Stamford, CT, USA). The oligonucleotides used in the EMSA were as follows: rs3743205.G: 5'-CTC CCG TTG CTA CCG GAA TGC CTC TTC AGG-3', 5'-CCT GAA GAG GCA TTC CGG TAG CAA CGG GAG-3'; rs3743205.A: 5'-CTC CCG TTG CTA CCA GAA TGC CTC TTC AGG-3', 5'-CCT GAA GAG GCA TTC TGG TAG CAA CGG GAG-3'; rs12899331.T: 5'-AAT ACA CCT GTC AGT CTC CTT CAT GCT TTC-3', 5'-GAA AGC ATG AAG GAG ACT GAC AGG TGT ATT-3'; rs12899331.C: 5'-AAT ACA CCT GTC AGC CTC CTT CAT GCT TTC-3', 5'-GAA AGC ATG AAG GAG GCT GAC AGG TGT ATT-3'; rs16787.C: 5'-TCA AAA AAT ACC TAT CGG GCC GGG CGC AGT-3', 5'-ACT GCG CCC GGC CCG ATA GGT ATT TTT TGA-3'; rs16787.A: 5'-TCA AAA AAT ACC TAT AGG GCC GGG CGC AGT-3', 5'-ACT GCG CCC GGC CCT ATA GGT ATT TTT TGA-3'. Probes for competition assays were as follows: SRE: 5'-AAT TCT CCT TTA CAC AGG ATG TCC ATA TTA GGA CAT CTC-3', 5'-GAG ATG TCC TAA TAT GGA CAT CCT GTG TAA AGG AGA ATT-3'; and Sp1 consensus probe (Santa Cruz Biotechnology, Santa Cruz, CA, USA): 5'-ATT CGA TCG GGG CGG GGC GAG C-3', 5'-G CTC GCC CCG CCC CGA TCG AAT-3'
Western blot analysis
Protein extracts derived from the SH-SY5Y cells were separated on 12% Bis-Tris SDS-PAGE gels (Invitrogen) in NuPAGE N-morpholino propane sulfonic acid (MOPS) sodium dodecyl sulfate (SDS) running buffer and electroblotted to polyvinylidene difluoride (PVDF) Hybond-P transfer membranes (GE Healthcare Bio-Sciences, Little Chalfont, UK). After transfer of proteins, the filters were blocked for unspecific protein binding by 1 h incubation at RT of 5% nonfat dry milk in 1% Tween-PBS followed by incubation of anti-TFII-I antibody (Transduction Laboratories, Lexington, KY, USA; 1:2500). Filters were washed 4 times in Tween-PBS for 15 min and 3 times for 5 min before incubation with ECL anti-mouse horseradish peroxidase (HRP) (GE Healthcare) (1:2x106). Filters were again washed at the same conditions as above, and detection was performed using the ECL Advance Western blotting detection kit (GE Healthcare). Anti-β-actin antibodies (Abcam, Cambridge, UK; 1:5000) were used for normalization.
Protein purification and protein sequencing
To purify the proteins bound to the EMSA probe, we used the DNA-binding protein purification kit (Roche, Indianapolis, IN, USA). In brief, the rs3743205G probe (double 60-mer) was concatenated by PCR and attached to magnetic beads by ligation to tethered oligonucleotides. The oligonucleotides were bound to the beads by streptavidin-biotin affinity. Total and nuclear extracts from SH-SYS5 cells were incubated with the beads followed by several washing steps. The extractions were performed several times, and the extracts eluted were microdialysed to remove the high concentrations of salts. To monitor the purification and the amount of proteins, the extractions were SDS-PAGE electrophoresed and stained with GelCode SilverSNAP Stain (Pierce Biotechnology, Rockford, IL, USA). The final extracts were further separated into a 10% SDS-PAGE to distinguish them from residual contaminating proteins and stained with GelCode Blue Stain (Pierce Biotechnology). Two bands were isolated and further analyzed by Ingel digestion followed by peptide sequencing. Briefly, stained protein bands were cut from the gel, and the pieces were digested using the MassPREP robotic protein handling system (Waters, Millford, MA, USA); washing was carried out in 50 mM ammonium bicarbonate containing 50% acetonitrile. The protein was reduced (DTT) and alkylated (iodoacetamide) followed by ingel digestion with 0.3 µg trypsin (modified, Promega, Madison, WI, USA) in 50 mM ammonium bicarbonate for 5 h at 40°C. The tryptic peptides were extracted with 1% formic acid/2% acetonitrile, followed by 50% acetonitrile twice. The acetonitrile was evaporated, and the peptide extract was concentrated under a stream of nitrogen to 5–10 µl for injection onto the LC-MS/MS system (CapLC chromatography system and Q-Tof Ultima API mass spectrometer, Waters, Millford, MA, USA). The peptides were separated on a Waters Atlantis C18 column, and the effluent was electrosprayed into the mass spectrometer via a PicoTip emitter (New Objective, Woburn, MA, USA). Data were collected using Data Dependent Acquisition (DDA, Waters) over the mass range 300–2000 m/z. Data analysis was performed using ProteinLynx Global SERVER 2.1 (PLGS 2.1, Waters) software and MassLynx peptide sequence software 4.0 (Waters), combined with use of the BLAST search engine (National Center for Biotechnology Information, Bethesda, MD, USA). Primers for probe rs3743205G, 60-mer: 5'-CTC CCG TTG CTA CCG GAA TGC CTC TTC AGG CTC CCG TTG CTA CCG GAA TGC CTC TTC AGG-3'; 5'-CCT GAA GAG GCA TTC CGG TAG CAA CGG GAG CCT GAA GAG GCA TTC CGG TAG CAA CGG GAG-3'.
Plasmid constructs
The luciferase assay constructs were prepared using the same sequences for the EMSA probes but duplicated to enhance the binding of the factors regulating the luciferase expression (see
Fig. 3
). Double-stranded probes from both alleles of the rs3743205, rs12899331, and rs16787 (60-mer) were blunt-end cloned into the pGL3 basic and pGL3 promoter vectors (Promega). The pcDNA3-flag-ERa expression construct was a kind gift of Eckard Treuter and the TFII-I 
and TFII-I β were cloned in mammalian expression vector pEBG, as described previously (46)
. The cloning of DYX1C1 in pcDNA3.1 has been described previously (2)
. Full-length cDNA from poly (ADP-ribose) polymerase 1 (PARP1) and splicing factor proline and glutamine-rich (SFPQ) were cloned into pcDNA3.1 expression vector (Invitrogen). To obtain the –3A DYX1C1 construct, the QuickChange II Site-Directed Mutagenesis kit was used according to manufacturer’s instructions (Stratagene, La Jolla, CA, USA).
|
|
Real-time qRT-PCR
Total RNA was extracted from SH-SY5Y cells at 6, 12, and 24 h after transient transfection, using the RNAeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. For the cDNA synthesis, 0.5 µg of total RNA in a volume of 20 µl was used as template, with RNase H treatment using the Superscript III RT-PCR kit (Invitrogen). For qRT-PCR, TaqMan gene expression assays were used to quantitate DYX1C1 (Hs00608519_m1) and TFII-I (Hs. Hs01073660_m1) (Applied Biosystem, Foster City, CA, USA). PARP1, SFPQ, and HPRT mRNA were quantitated using Power SYBR Green PCR Master Mix (Applied Biosystems). Relative expression of the different gene transcripts was calculated by the delta-delta-Ct (DDCt) method and converted to the relative expression ratio (2–DDCt). All data were normalized to the endogenous reference gene hypoxanthine-guanine phosphoribosyltransferase (HPRT) expression. The following primers were used: PARP1-F 5'-CCCAAAGGAATTCCG AGAAA-3', PARP1-R 5'-TCCTTTTTGGTGCTGATG-3'; SFPQ-F 5'-GGGCTGT TGTAATAGTGGATGA-3', SFPQ-R 5'-CCAAAGACTTCCATCGCTGA-3'; HPRT-F 5'-TCAGGCAGTATAATCCAAAGATGGT-3', HPRT-R 5'-AGTCTGGCTTATATCC AACACTTCG-3'.
Luciferase reporter assays
The neuroblastoma SH-SY5Y cells were seeded in 6-well plates at 90% confluence (2.5x106/ml). After 24 h, the cells were cotransfected with 1 µg or 500 ng of plasmid containing the variations (see luciferase reporter constructs in pGL3) and 50 ng of the pRLTK vector containing the herpes simplex thymidine kinase promoter linked to a constitutively expressing renilla luciferase reporter gene (Promega). The pGL3 empty vectors were used as negative controls and GFP plasmid was used as positive control for transfections. At 24 h after transfection, the cells were harvested, proteins were extracted, and luciferase activity was determined using the Dual-Luciferase® Reporter Assay System (Promega) and a Tecan Infinite luminescence reader (Tecan AG, Männedorf, Switzerland). All data were normalized to renilla luciferase. The statistical significance of the experiments was measured by Student’s t test with a threshold of P 
0.05.
Bioinformatics
For the analysis of transcription binding sites, the transcription element search system (TESS; http://www.cbil.upenn.edu/tess/) and the Genomatix package (http://www.genomatix.de) were used. For the gene coexpression data, the TMM (http://microarray.cpmc.columbia.edu/cgi-bin/find-links.cgi) and Gemma databases (http://bioinformatics.ubc.ca/gemma) were used.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
was stained with monoclonal anti-V5 and Alexa Fluor 555-conjugated anti-mouse-IgG. The nuclei were stained with DAPI. The staining showed protein localization in both cytoplasm and nucleus (Fig. 1)
.
Prediction of DYX1C1 cis-regulatory elements and allele-specific differential retardation in electrophoretic mobility shift assays
We chose three promoter SNPs as particularly interesting for characterization because they appeared to form a haplotype associated with dyslexia (F. Dahdouh et al., personal communication). Comparison of TESS analyses of both alleles for the three SNPs, –3G/A (rs3743205), –10567T/C (rs12899331), and –10310C/A (rs16787), suggested that these polymorphisms may alter the binding affinity for several transcription factors. The sequences containing the –3A polymorphism but not the –3G allele were predicted to bind the TFII-I and Elk1 transcription factors. The –10567T/C and –10310C/A polymorphisms were predicted to have many differences in binding affinity, including the Sp1 transcription factor family.
As the factors predicted by TESS are nuclear factors, we used nuclear extracts from the SH-SY5Y cells to test by EMSA. We synthesized double-stranded DNA probes containing the variations and used these to screen for differences in protein factor binding by EMSA. We observed allele-specific differential retardation for each of the three SNPs, rs3743205, rs12899331, and rs16787 (Fig. 2
A).
As TFII-I had been suggested to bind to the –3A probe by TESS, we tested the binding of TFII-I to our probe by overexpressing two isoforms of TFII-I ( and β) in SH-SY5Y cell culture and used these extracts in EMSAs. We detected enhanced binding affinity of both alleles –3G and –3A in these extracts, confirming the true binding of TFII-I to this site (Fig. 3
A). Others have shown that TFII-I regulates gene expression of some certain responsive genes in the presence of ER
(24)
; therefore, we tested to enrich the SH-SY5Y cell extracts with ER and could observe an increase in the binding affinity in total cell extracts. We also tested to overexpress ER together with TFII-I and TFII-Iβ. Here we observed a decrease of the TFII-I binding to both A and G probes, whereas in the case of TFII-Iβ it enhanced the binding to the G probe and decreased binding to the A probe (Fig. 2A
).
In addition, we performed competition assays. We used probes from the well-studied c-fos promoter, where the serum response element (SRE) is known to bind TFII-I, and the sequence of this consensus is unrelated to our 30-bp probe (25)
. We observed a reduction of TFII-I binding to the labeled –3A and –3G probes. To compete for the Sp1 binding site, we used a commercially available consensus probe (see Materials and Methods). The competition assays confirmed the predicted factors in all cases.
Identification of additional proteins retarded in EMSA: DYX1C1 trans-acting proteins
To further characterize the proteins bound differentially to the DYX1C1 promoter in EMSA, we performed a series of protein extraction experiments. By using a concatenated probe corresponding to the –3G allele that showed differential binding of proteins in the EMSA (Fig. 2A
), we enriched proteins that were resolved in SDS-PAGE. We used silver staining of gels to monitor the protein purification. The concatenated and biotinylated probe was bound to avidin-coated magnetic beads, and the beads were incubated with total cellular protein and nuclear extracts from the neuroblastoma cell line SH-SY5Y. PAGE resolution of the protein extracts revealed four bands, and we chose to proceed with further analysis of the two strongest and largest molecular weight bands by ingel digestion followed by matrix-assisted laser desorpton/ionization-time of flight mass spectrometry (MALDI-TOF-MS) and protein sequencing by the liquid chromatography mass spectrometry (LC-MS/MS) system (Fig. 3)
. We identified three trans-acting proteins that bind to the –3G allele in the 5'UTR of the DYX1C1 gene. Proteins isolated from the upper band were identified as TFII-I and PARP1 and from the lower band as SFPQ/PSF (Fig. 3A
).
Effect of TFII-I, PARP1, and SFPQ in the regulation of DYX1C1 in the neuroblastoma cell line SH-SY5Y
To study the effect of the three protein factors TFII-I, PARP, and SFPQ that bind to DNA around rs3743205 in the transcriptional regulation of DYX1C1, we transiently transfected expression constructs of DYX1C1 (–3G or –3A), PARP1, SFPQ, and TFII-I ( or β) to SH-SY5Y cells. Using a DYX1C1 TaqMan assay, we measured the expression levels of both –3G and –3A alleles of DYX1C1. Cell lysates were prepared at 6, 12, and 24 h after transient transfection (Fig. 4
). The levels of the –3G allele were significantly higher than the levels of –3A allele, independent of which variant of TFII-I was overexpressed ( or β) in combination with the PARP1 and SFPQ. We observed in three independent assays that the levels of the –3G allele are significantly higher than the levels of the –3A allele. We also confirmed the genotype of rs3743205 in the SH-SY5Y cells as –3G allele. The overexpression of each of the factors was confirmed in each experiment by qRT-PCR (data not shown).
|
Functional verification of the effects of allele-specific binding differences
To further verify the effects of the DYX1C1 5'UTR and promoter polymorphisms and to confirm the roles of the transcription factors, we prepared constructs using a tandem repeat of the EMSA probes (Fig. 5
) cloned into the pGL3 basic and pGL3 promoter vectors (Promega). The inserts were made separately for both alleles of the investigated SNPs. These constructs were transiently transfected into SH-SY5Y. Probes cloned into pGL3 promoter vector containing the upstream SV40 promoter yielded significant differences in luciferase activity with all three SNPs: for rs16787 and rs3743205, P < 0.001; and for rs12899331, P < 0.05 (Fig. 5)
.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 6
, 9)
. The accumulated information may well suggest that each of the implicated genes can increase the risk for dyslexia if its function is compromised, even though genetic association studies may be underpowered or true population differences may exist, explaining the ambiguous results. As in many other examples, even rare genetic defects may unveil an important pathogenic pathway, and therefore these poorly characterized genes deserve further functional studies.
We undertook this study to understand better the regulation of DYX1C1 and to search for explanations for genetic association results. We have adopted a working hypothesis that even slightly altered expression of DYX1C1, rather than coding changes, may influence the risk of dyslexia. Of note, dyslexia was associated with haploinsufficiency of DYX1C1, cosegregating with a chromosome translocation that disrupted one copy of the gene, whereas the other copy remained intact (2)
.
Our results showed that polymorphisms in the promoter and 5' UTR of DYX1C1 are functional. Furthermore, for the DNA segment, including the rs3743205 (–3G/A) polymorphism, we have identified a complex of three proteins that may be highly relevant in the context of brain development and that have not previously been reported to act together in gene regulation. These three factors are discussed in more detail below.
TFII-I is a 120 kDa protein, initially characterized for its binding to initiator sites of various promoters functioning as a basal transcription factor. Later, it was shown that it could also interact with upstream regulatory factors, including USF1 and c-myc (26)
. TFII-I has been suggested to play a role in signal transduction as well as in transcription initiation. Interestingly, TFII-I maps to 7q11.3, and deletions of this region are associated with the neurodevelopmental Williams-Beuren syndrome (WBS) in humans. WBS has multisystemic manifestations, including supravalvular aortic stenosis, hypercalcemia in infancy, mild to moderate mental retardation, cognitive defects, and characteristic facial features. WBS has also been associated with dyslexia (27)
. In addition, TFII-I is expressed in a variety of tissues, with the highest expression in the brain.
PARP1 is an intracellular protease that plays an important role in the development of neuronal cell death, apoptosis, and necrosis and in the sequelae of cerebral ischemia. PARP1 activation is involved in a number of pathological conditions, including ischemic/reperfusion injury, traumatic brain injury, Parkinson’s disease, and Alzheimer’s disease. PARP1 has also been implicated in the regulation of gene expression through modification of transcription factors (modifying histones to alter chromatin structure) or functioning as part of enhancer/promoter binding complexes in conjunction with other DNA binding factors and coactivators (28)
. Interestingly, PARP1 has been shown to be involved in learning and long-term memory in the sea slug Aplysia californica (29)
and in rats (30)
. Working-memory deficits are characteristic in some dyslexic individuals (31)
The SFPQ or polypyrimidine tract-binding (PTB) protein-associated splicing factor (PSF) is a nuclear factor initially identified as a splicing factor but later shown to have multifunctional characteristics in a variety of nuclear processes (32)
. PSF contains both DNA-binding domain and RNA-binding domain, making it able to bind both DNA and RNA. Studies in mouse suggest that PSF could be involved in the control of neuronal-specific splicing events occurring at particular stages of neuronal differentiation and maturation (33)
.
Both TFII-I and PARP1 can regulate gene expression. TFII-I often activates transcription but it has also been shown that TFII-I can down-regulate some estrogen-responsive genes by interacting with ER (24)
. Our EMSA results suggested that these factors trans-activate the wild-type allele –3G, but when the variation –3A was present, the binding of the factors was weaker (Fig. 3)
. Combined with information that our luciferase reporter assays gave significant differences between the alleles, it is possible that TFII-I acted as a repressor in this context. When the A allele was present and the binding in EMSA was weaker, the expression of the –3A allele was higher (Fig. 5)
. However, we have tested the direct effect of the three factors identified above by overexpressing them together with the two alleles of (–3G/A) DYX1C1. We show a direct effect of TFII-I, PARP1, and SFPQ in the regulation of DYX1C1 and also show that the effect is allele dependent (Fig. 4)
.
Transcriptional regulation is a complex phenomenon in which many proteins may play roles in gene regulation, and the direction of the effect may depend on balance of the tested factors and additional cellular factors. At least TFII-I may act either as enhancer or suppressor depending on the context of other factors.
As for SFPQ/PSF, splicing and transcription are coupled nuclear processes, and PSF also has the ability to bind to DNA. Therefore, PSF should also be considered an interesting candidate molecule in the regulation of the DYX1C1 gene. Interestingly, the SFPQ gene lies in a region that has been linked to speech sound disorder, language impairment, and dyslexia on chromosome 1p34-p36 by several groups (34
35
36
37)
. Recently, it has been shown that sfpq in zebra fish is expressed in all tissues, but the strongest expression is in the brain during neuronal development, suggesting its importance for neuronal development. Interestingly, a mutant for sfpq in zebra fish, whitesnake (38)
, shows abnormal shape of midbrain and hindbrain.
Finally, using bioinformatic tools to explore the coexpression of genes based on microarray data by TMM and Gemma, we could confirm coexpression between TFII-I, SFPQ, and PARP1. This finding supports the concept that these pairs of proteins might have coregulatory roles on target genes.
Microarray studies of gene-expression differences in human and nonhuman primate brains have shown an excess of genes with higher expression levels in the brain of humans compared to that of chimpanzees (39
, 40)
. These differences were observed exclusively in the brain and not in other tissues, such as liver or heart (41)
. These results suggested that the genetic differences are specific to the brain and are likely associated with our more complex cognitive functions. Interestingly, when genes with the most consistent expression changes in human brain were listed, the TFII-I gene appeared at the top of the list, with an average fold change of 2.5–4.2 (41)
. Elucidation and understanding of the biology behind this gene and the other genes related to dyslexia will open new pathways in the study of brain functions.
|
ACKNOWLEDGMENTS |
|---|
expression construct was kindly provided by Eckardt Treuter (Department of Biosciences and Nutrition, Karolinska Institute, Stockholm, Sweden). Ingel digestion, peptide extraction, LC-MS/MS, and database searches for protein identification were performed at the Protein Analysis Center, Karolinska Institutet. A.L.R. is supported by grants from the U.S. National Institutes of Health (HD046034). Work in J.K.’s laboratory is supported by the Swedish Royal Bank Tercentennial Foundation, the Knut and Alice Wallenberg Foundation, the Swedish Research Council, the Sigrid Jusélius Foundation, the Academy of Finland, and the Folkhälsan Institute of Genetics. Received for publication January 7, 2008. Accepted for publication April 2, 2008.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. Massinen, K. Tammimies, I. Tapia-Paez, H. Matsson, M.-E. Hokkanen, O. Soderberg, U. Landegren, E. Castren, J.-A. Gustafsson, E. Treuter, et al. Functional interaction of DYX1C1 with estrogen receptors suggests involvement of hormonal pathways in dyslexia Hum. Mol. Genet., August 1, 2009; 18(15): 2802 - 2812. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
