Molecular networks implicated in speech-related disorders: FOXP2 regulates the SRPX2/uPAR complex

Abstract

It is a challenge to identify the molecular networks contributing to the neural basis of human speech. Mutations in transcription factor FOXP2 cause difficulties mastering fluent speech (developmental verbal dyspraxia, DVD), whereas mutations of sushi-repeat protein SRPX2 lead to epilepsy of the rolandic (sylvian) speech areas, with DVD or with bilateral perisylvian polymicrogyria. Pathophysiological mechanisms driven by SRPX2 involve modified interaction with the plasminogen activator receptor (uPAR). Independent chromatin-immunoprecipitation microarray screening has identified the uPAR gene promoter as a potential target site bound by FOXP2. Here, we directly tested for the existence of a transcriptional regulatory network between human FOXP2 and the SRPX2/uPAR complex. In silico searches followed by gel retardation assays identified specific efficient FOXP2-binding sites in each of the promoter regions of SRPX2 and uPAR. In FOXP2-transfected cells, significant decreases were observed in the amounts of both SRPX2 (43.6%) and uPAR (38.6%) native transcripts. Luciferase reporter assays demonstrated that FOXP2 expression yielded a marked inhibition of SRPX2 (80.2%) and uPAR (77.5%) promoter activity. A mutant FOXP2 that causes DVD (p.R553H) failed to bind to SRPX2 and uPAR target sites and showed impaired down-regulation of SRPX2 and uPAR promoter activity. In a patient with polymicrogyria of the left rolandic operculum, a novel FOXP2 mutation (p.M406T) was found in the leucine-zipper (dimerization) domain. p.M406T partially impaired the FOXP2 regulation of SRPX2 promoter activity, whereas that of the uPAR promoter remained unchanged. Together with recently described FOXP2-CNTNAP2 and SRPX2/uPAR links, the FOXP2-SRPX2/uPAR network provides exciting insights into molecular pathways underlying speech-related disorders.

Figure 1.

qRT–PCR experiments. Wild-type FOXP2 protein down-regulates the native SRPX2 and uPAR genes. qRT–PCR experiments were performed from HEK293T cells transfected with pcDNA4/HisMax-FOXP2 construct. Expression changes are given as the mean of log₂ expression ratios of cells transfected with pcDNA4/HisMax-FOXP2 compared with cells transfected with non-recombinant pcDNA4/HisMax vector and are normalized for equal expression of the GAPDH internal control. Values represent the mean of comparisons of five independent cDNA syntheses. P-values were calculated using two-tailed unpaired t-test. *P< 0.05 and **P < 0.01.

Figure 2.

Luciferase reporter assays. Wild-type FOXP2 protein down-regulates the activities of the SRPX2 (A) and of the uPAR (B) promoters. Either of the SRPX2 (pGL3-SRPX2) and uPAR (pGL3-uPAR) promoter constructs was co-transfected into HEK293T cells with FOXP2-containing (pcDNA4/HisMax-FOXP2) or with empty (pcDNA4/HisMax) vector and with a β-galactosidase expression vector (pHSV-LacZ) for the normalization of transfection efficiencies. In parallel, an empty (promoterless) vector (pGL3-basic) was transfected with pcDNA4/HisMax-FOXP2 and with pHSV-LacZ. Transcriptional activities were determined by quantifying the luciferase activity of cellular extracts prepared 48 h after transfection. Data show the mean ± SD relative activity from three experiments done in triplicate. Statistical significances were determined by two-tailed unpaired t-test. ***P < 0.0005.

Figure 3.

Electrophoretic mobility shift assays. FOXP2 directly binds sites identified within the SRPX2 and uPAR promoters. EMSAs were used to determine whether direct interactions occurred between FOXP2 and the promoter regions of the SRPX2 and uPAR genes. DNA probes were designed based on the presence of predicted FOX, FOXP and FOXP2-binding sites (Table 1, Supplementary Material, Table S1). Radiolabeled probes that had shown binding in earlier experiments (Supplementary Material, Fig. S2) were incubated with nuclear extracts taken from HEK293T cells transfected with either an empty vector (pcDNA4), wild-type FOXP2 or FOXP2.R553H. The binding of FOXP2 to a known consensus binding site (CONSENSUS) is shown as a positive control. Binding assays were also performed in the presence of competition from unlabeled wild-type (wt), mutant (mut) and irrelevant (NFK) probes. In each case, FOXP2 binding to the labeled wild-type probe was efficiently impaired via competition with unlabeled wild-type competitor probe, but not by mutant or irrelevant competitor probes, displaying the specificity of the interaction. Addition of an antibody directed to FOXP2 (FOXP2 Ab) caused a supershift to occur in each case, an effect not observed when an irrelevant control antibody (IgG) was added. This confirmed that the identity of the protein causing gel retardation is indeed FOXP2.

Figure 4.

Regulation of, and binding on, the SRPX2 and uPAR promoters is altered by a pathogenic p.R553H FOXP2 mutation. (A) qRT–PCR experiments showing loss of down-regulation of the native SRPX2 and uPAR expressions by mutant FOXP2. Expression changes are given as the mean of log₂ expression ratios of HEK293T cells transfected with pcDNA4/HisMax-FOXP2 (wild-type FOXP2) or with pcDNA4/HisMax-R553H (mutant p.R553H FOXP2), when compared with HEK293T cells transfected with non-recombinant pcDNA4/HisMax vector, and are normalized for equal expression of the GAPDH internal control. Values represent the mean of comparisons of five independent cDNA syntheses. P-values were calculated using two-tailed unpaired t-test. *P < 0.05, **P < 0.01 and ***P < 0.001. (B) Luciferase reporter assays showing loss of down-regulation of the SRPX2 and uPAR promoter activities by mutant p.R553H FOXP2. Either of the SRPX2 (pGL3-SRPX2) and uPAR (pGL3-uPAR) promoter constructs was co-transfected into HEK293T cells with FOXP2-containing (pcDNA4/HisMax-FOXP2 or pcDNA4/HisMax-R553H) or with empty (pcDNA4/HisMax) vector and with a β-galactosidase expression vector (pHSV-LacZ) for the normalization of transfection efficiencies. In parallel, the empty (promoterless) vector (pGL3-basic) was transfected with pcDNA4/HisMax-FOXP2 and with pHSV-LacZ. Transcriptional activity was determined by quantifying the luciferase activity of cellular extracts prepared 48 h. after transfection. Data show the mean ± SD relative activity from three experiments done in triplicate. Statistical significances were determined by two-tailed unpaired t-test. ****P < 0.0001.

Figure 5.

Novel p.M406T FOXP2 mutation in a patient with polymicrogyria of the left rolandic operculum partially impairs proper regulation of SRPX2. (A) MRI (axial inversion recovery section) of the patient with left opercular polymicrogyria. (B) c.T1591C mutation of FOXP2. The nucleotide sequences and translations are shown above the direct sequencing trace from PCR-amplified fragment. Top: Section of a wild-type sequence in the unaffected mother (white circle). Middle: Section of a sequence with the mutation in the carrier father (dotted square). Bottom: Section of a sequence with the mutation in the affected proband (full-blackened circle). Two more siblings were carriers, but did not show obvious neurological problems, although no MRI was performed (sequence traces not shown). (C) p.R553H and p.M406T mutant FOXP2 proteins display altered intracellular localizations. HEK293T cells were transfected with pcDNA4/HisMax (mock) vector (top, left), with pcDNA4/HisMax-FOXP2 (wt) vector (bottom, left), with pcDNA4/HisMax-R553H (p.R553H) vector (top, right) or with pcDNA4/HisMax-M406T (p.M406T) vector (bottom, right). Wild-type FOXP2 displayed unambiguous nuclear localization. Mutant p.R553H FOXP2 showed both nuclear and cytoplasmic localizations, as already described (18), and p.M406T FOXP2 also showed altered nuclear targeting. HisMax-tag fusion protein FOXP2 was detected using an antibody to the N-terminal Xpress™ tag (green). Blue: nuclear DAPI staining. (D) Luciferase reporter assays showing partial loss of down-regulation of the SRPX2 promoter activity by mutant p.M406T FOXP2. Activity of the uPAR promoter remained unchanged. Either of the SRPX2 (pGL3-SRPX2) and uPAR (pGL3-uPAR) promoter constructs was co-transfected into HEK293T cells with FOXP2-containing (pcDNA4/HisMax-FOXP2 or pcDNA4/HisMax-M406T) or with empty (pcDNA4/HisMax) vector and with a β-galactosidase expression vector (pHSV-LacZ) for the normalization of transfection efficiencies. In parallel, an empty (promoterless) vector (pGL3-basic) was transfected with pcDNA4/HisMax-FOXP2 and with pHSV-LacZ. Transcriptional activity was determined by quantifying the luciferase activity of cellular extracts prepared 48 h after transfection. Data show the mean ± SD relative activity from three experiments done in triplicate. Statistical significances were determined by two-tailed unpaired t-test. NS, not significant (P > 0.05), ***P < 0.001 and ****P < 0.0001.