ABSTRACT
Mutations in the coding region of the FOXP2 transcriptor factor gene are known to cause speech and language impairment. Chromosomal rearrangements with breakpoints downstream the gene have been hypothesised to impair speech and cognitive abilities via physical separation of distant regulatory DNA elements. In this study, we used highly efficient targeted chromosomal deletions induced by the CRISPR/Cas9 genome editing tool to characterise two functional enhancers: FOXP2-Eproximal and FOXP2-Edistal, located in the intergenic region between FOXP2 and its adjacent MDFIC gene. FOXP2-Edistal, separated from FOXP2 by a chromosomal rearrangement in a case of speech and language impairment, was demonstrated to be functional in a luciferase assay. Deletion of any of these two functional enhancers in a neuroectodermal tumor cell-line downregulates FOXP2 and decreases FOXP2 protein levels, conversely it upregulates MDFIC and increases MDFIC protein levels. We expect these findings contribute to a deeper understanding of FOXP2 and MDFIC may pace the development of speech and language in the brain.
INTRODUCTION
Mutations in FOXP2 gene are known to cause speech and language impairment (Vargha-Khadem et al. 2005; Zhao et al. 2010). Polymorphisms of the gene have been also associated to schizophrenia (Tolosa et al. 2010) and frontotemporal lobar degeneration (Padovani et al. 2010). FOXP2 has been hypothesised to regulate the development and function in the adult state of brain areas involved in human language processing (Lai et al., 2003; Fisher and Scharff 2009), because of its known role in neurogenesis, neuron differentiation and migration patterns in the developing telencephalon in mice (Tsui et al. 2013; Chiu et al. 2014; Garcia-Calero et al. 2016). Pathogenic mutations in humans have proven to impair auditory-motor association learning in mice (Kurt et al. 2012). Nonetheless, the exact role of FOXP2 in normal development is unknown. Common variants of the gene do not contribute appreciably to individual differences in language development (Mueller et al. 2016), nor in brain structure (Hoogman et al. 2014), although a FOXP2 polymorphism has been associated with enhanced procedural learning of non-native speech sound categories (Chandrasekaran et al. 2015). Less is known about how the expression of the gene is modulated. The promoter of FOXP2 contains four transcription start sites (Schroeder and Myers 2008), with multiple alternative splicing sites (Bruce and Margolis 2002). FOXP2 also contains six ultraconserved regions in its introns (Schroeder and Myers 2008), as well as enhancers for LEF1, a transcription factor that drives expression of the gene in the central nervous system during embryogenesis (Bonkowsky et al. 2008). Interestingly, several microRNAS bind the 3’UTR of the gene and regulate the expression of FOXP2 (Clovis et al. 2012; Shi et al. 2013; Fu et al. 2014b; Cuiffo et al. 2014).
Apart from gene mutations, chromosomal translocations involving the FOXP2-MDFIC intergenic region have been also associated to speech problems and cognitive impairment (Kosho et al. 2008). Microdeletions involving FOXP2 and/or MDFIC and the region between these two genes have been found in subjects with speech delay and cognitive impairment (DECIPHER patients 262086, 292652, and 301696). We have recently reported on a young female harbouring a genomic complex rearrangement involving chromosomes 7 and 11, who presents with severe expressive and receptive speech and language impairment in both Castilian Spanish and Valencian (Moralli et al. 2015). Although the FOXP2 coding region is intact, the breakpoint in 7q31.1 is located 205.5 kb downstream the 3’ end of FOXP2 and 22.8 kb upstream the 5’ region of MDFIC. Becker et al. (2015) found and characterized a functional enhancer located 2.5 kb downstream the breakpoint. In our proband this element was maintained in chromosome 7, whereas FOXP2 was rearranged to chromosome 11. A more robust approach, aimed at looking for changes in the expression level of the gene seems desirable in order to know if this enhancer regulates FOXP2 expression.
The development of nuclease mediated genome editing tools, specially, of those based on clustering regularly interspaced short palindromic repeats (CRISPR) (Sakuma and Woltjen 2014; Torres-Ruiz and Rodriguez-Perales 2016), has emerged as a highly efficient way of inducing targeted chromosomal deletions and an accurate method to validate the functionality of enhancers (Cong et al. 2013; Mali et al. 2013). Here we report a detailed study of the intergenic region between the FOXP2 and MDFIC genes. We have found that this region contains, apart from the enhancer reported in Becker et al. (2015), a second functional enhancer, FOXP2-E proximal. We performed targeted deletions of each regulatory element by CRISPR-Cas9 and found that both affect the expression levels of FOXP2 and MDFIC in an opposite manner. We hypothesise therefore that the breakpoint in this case would cause FOXP2 to be anomalously downregulated by the separation of FOXP2-distal from FOXP2, while MDFIC to be anomalously upregulated by the separation of FOXP2-proximal from MDFIC. These changes in the expression levels of these two genes may account for the observed language deficits in this case. We expect these findings contribute to a better understanding of how FOXP2 is regulated.
MATERIALS AND METHODS
Cell culture and electroporation
Cells of the non neuronal cell-line HEK293A (CRL-1573, ATCC, USA) and the neuronal cell-line SK-N-MC (HTB-10, ATCC, USA) were maintained under standard conditions in Dulbecco's modified Eagle's medium (DMEM) (Lonza), supplemented with 10% foetal bovine serum (FBS) (Life Technologies), 1% GlutaMAX (Life Technologies), and 10mg/ml penicillin/streptomycin (Life Technologies). The neuronal cell-line SK-SY-5Y (CRL-2266, ATCC, USA) was cultured in a 1:1 mixture of Dulbecco's modified Eagle's medium (DMEM) (Lonza) and F12 medium (Lonza) supplemented with 10% FBS (Life Technologies), 1% GlutaMAX (Life Technologies), and 10mg/ml penicillin/streptomycin (Life Technologies). Cells were cultured at 37°C in a humidified incubator in a 5% CO2 + 20% O2 atmosphere.
For electroporation, we used the Neon Transfection System (Life Technologies). The manufacturer's protocols for HEK293A, SK-N-MC and SH-SY5Y cells were modified as follows. The three cell types were electroporated at 80% confluence. Cells were trypsinized and resuspended in R solution (Life Technologies). For SK-N-MC and SH-SY5Y, 10-µl tips were used to electroporate 2.5×106 cells with a single 50-ms pulse of 900 V. For HEK293A cells, 4×105 cells were electroporated with 10-µl tips using three 10-ms pulses of 1245 V. After electroporation, cells were seeded in a 24-well plate containing prewarmed medium. When required, cells were sorted 72 h post-transfection.
Construction of Double-Guide Cas9-Encoding Plasmids
The parental pLV-U6#1H1#2-C9G vector has been described elsewhere (Torres et al. 2014b). Two gBlocks gene fragments were synthesized to clone sgRNA#1 and sgRNA#2 flanking the FOXP2-Eproximal and FOXP2-Edistal enhancer regions in the backbone vector using BsrGI and SpeI target sites.
Flow Cytometry and Cell Sorting
72 hours after electroporation, cells were trypsinized and washed with DPBS twice, counted, and resuspended in an appropriate volume of sorting buffer (PBS containing 1% FBS and antibiotics) for flow cytometry analysis. Immediately before cell sorting, samples were filtered through a 70-µm filter to remove any clumps or aggregates. Cell sorting was carried out in a Synergy 2L instrument (Sony Biotechnology Inc.); flow cytometry was performed in a BD LSR Fortessa analyzer (BD Biosciences). Cells were sorted and seeded individually per well in a 96 well-plate.
Genomic DNA Extraction and PCR Analysis
Genomic DNA was extracted using standard procedures (Torres et al. 2014a). Briefly, 5-10×106 cells were either trypsinized or scraped, washed in PBS, pelleted, and lysed in 100mM NaCl, Tris (pH 8.0) 50mM, EDTA 100mM, and 1% SDS. After overnight digestion at 56°C with 0.5 mg/ml of proteinase K (Roche Diagnostics), the DNA was cleaned by precipitation with saturated NaCl, and the clear supernatant was precipitated with isopropanol, washed with 70% ethanol, air-dried, and resuspended overnight at room temperature in 1xTE buffer. Serial DNA dilutions were quantified with a NanoDrop ND 1000 Spectrophotometer (NanoDrop Technologies).
Standard PCR was performed in a Veriti 96-well Thermal Cycler (Applied Biosystems) under the following conditions: template denaturation at 95°C for 1 min followed by 30 cycles of denaturation at 94°C for 30 s, annealing at 62.5°C for 30 s, extension at 72°C for 60 s, and a final extension of 5 min at 72°C. Primers used are listed in Supplementary Table 1.
RNA extraction and PCR
Total RNA was extracted from tissues and cell cultures using Trizol (Sigma-Aldrich), followed by treatment with RNase-free DNAse (Roche Applied Science). cDNA was synthesized from 500 ng of total RNA using the Superscript III First Strand cDNA Synthesis Kit (Life Technologies). Specific mRNAs were quantified by qRT-PCR using an ABI Prism 7900 HT Detection System (Applied Biosystems) and TaqMan detection. PCR was performed in 96-well plate microtest plates with TaqMan master mix (Thermo Fisher) for 40 cycles. In all experiments, mRNA amounts were normalized to the total amount of cDNA by using amplification signals for hGUSB. Each sample was determined in triplicate, and at least three independent samples were analysed for each experimental condition.
Western Blot
Proteins were extracted by standard procedures as previously described (Rodriguez-Perales et al. 2015) in the presence of Complete Protease Inhibitor Cocktail Tables (Roche Applied Science). Proteins were transferred to PVDF using TransFi (Invitrogen; Life Technologies), and membranes were probed for FOXP2 or MDFIC with monoclonal mouse anti-human FOXP2 or MDFIC antibodies (1/1000 or 1/500; BD Pharmigen) or for GAPDH (AbCam), with antibodies diluted 1/2500 in PBS/0.1% Tween-20 (PBS-T). Secondary antibodies were HRP-conjugated with goat anti-mouse IgG (1/1000) and goat anti-Rabbit (1/500; Dako, Barcelona, Spain), and blots were developed with ECL (GE Healthcare).
RESULTS
In silico search of enhancer regions
We first hypothesised that the breakpoint in 7q31.1 (114,888,284 hg38) affected the expression of FOXP2 by physically separating some cis-acting distant element with an enhancer role. Accordingly, we searched in silico for putative enhancers in the intergenic region between FOXP2 and MDIFC looking for the following hallmarks: DNAsa clusters, presence of histones with specific post-translational modifications (specifically histone H3, lysine 4 monomethylation (H3K4me1) and H3 lysine 27 acetylation (H3K27ac)), and ChIP-seq data provided by ENCODE of regions recruiting co-activators and co-repressors as revealed by chromatin immunoprecipitation followed by deep sequencing. We found two putative enhancers located at 120kb and 203.5kb downstream the end of FOXP2, respectively (Figure 1A and 1B). These putative enhancers (referred hereafter as FOXP2-Eproximal and FOXP2-Edistal) span 6264bp (chr7:114,817,431-114,823,694 hg38) and 2300bp (chr7:114,900,989-114,903,302 hg38 equivalent to 114,541,370114,542,201 hg19), respectively. FOXP2-Edistal is the one previously validated by luciferase assay by Becker et al. (2015); FOXP2-Eproximal is a new putative regulatory element.
CRISPR deletion of FOXP2-Eproximal and FOXP2-Edistal
We then tested in vitro the functionality of FOXP2-Eproximal and FOXP2-Edistal. Since both putative enhancers are located in an intergenic region, we aimed at characterizing that both of them are functional with respect to FOXP2 or/and MDFIC. We relied on a CRISPR genome editing approach to delete the entire predicted sequence of each enhancer. Accordingly, we designed two couples of sgRNAs targeting the flanking regions of either FOXP2-Eproximal or FOXP2-Edistal (Figure 1C). Each sgRNA pair was cloned in the pLV-U6#1H1#2-C9G (Torres et al. 2014b) in order to couple the expression of the sgRNAs to the expression of Cas9 and a GFP reporter. Afterwards, we tested if the sgRNAs were able to induce the expected deletions. HEK293A cells were nucleofected with 2ug of pLV-U6#1H1#2-C9G plasmid targeting either FOXP2-Eproximal or FOXP2-Edistal. After 72 h, the DNA was isolated and analyzed. After designing PCR oligos that span the deleted regions (Supplementary Table 1), PCR assays were performed. They revealed efficient targeted deletions of the 6.2kb or the 2.3kb regions containing the entire sequence of FOXP2-Eproximal and FOXP2-Edistal, respectively (data not shown).
Neuronal cell lines defective in FOXP2-Eproximal or FOXP2-Edlstal
We next used CRISPR to delete FOXP2-Eproximal or FOXP2-Edistal in two cell lines: SK-N-MC, a metastatic cell line derived from the supra-orbital area, which expresses FOXP2 constitutively (although it does not express MDFIC at the same level), and SH-SY5Y, a neuroblastomic cell line which expresses neither FOXP2 nor MDFIC. Cells were electroporated with 2ug of either pLV-U6#1H1#2-C9G-Eproximal, pLV-U6#1H1#2-C9G-Edistal, or with empty plasmids. After 72h the DNA was extracted and analysed. PCR and Sanger sequencing analyses confirmed the deletion of the 6.2kb or the 2.3kb fragment (Figures 2A and 2B). We then generated two clonal cell lines (one for each putative enhancer) by sorting GFP positive cells into 96-well plates for single cell colony expansion (data not shown). We confirmed by PCR that the cellular clones had expanded. These two cell lines were used for further expression analyses.
FOXP2 and MDFIC expression analyses
We next aimed to characterize in more detail the functionality of FOXP2-Eproximal and FOXP2-Edistal. We used RT-qPCR to determine the amount of FOXP2 mRNA in the SK-N-MC cells transduced with either pLV-U6#1H1#2-C9G-Eproximal, or pLV-U6#1H1#2-C9G-Edistal, or an empty plasmid as a control. The expression of FOXP2 was significantly reduced (2.9 fold change) compared to that of the control when FOXP2-Eproximal was deleted (Figure 2C, left). Likewise, FOXP2 expression was decreased (2 fold change) when FOXP2-Edistal was deleted (Figure 2C left). We then measured the levels of expression of MDFIC after deletion of each enhancer. As shown in Figure 2C right, the expression of MDFIC was significantly increased when either FOXP2-Eproximal or FOXP2-Edistal were deleted (8.6 and 7.5 fold change, respectively). The experiment was replicated in SH-SY5Y cells, but no change of expression of FOXP2 or MDFIC was detected (data not shown).
We next analysed by Western blot the amount of FOXP2 and MDFIC proteins in the SK-N-MC cells transduced with either pLV-U6#1H1#2-C9G-Eproximal, or pLV-U6#1H1#2-C9G-Edistal, or an empty plasmid, used as a control (Figure 2D). In line with our mRNA data, the deletion of FOXP2-Eproximal or FOXP2-Edistal was found to reduce the level of FOXP2 (Figure 2D top) and to increase the level of MDFIC (Figure 2D bottom).
DISCUSSION
In this paper we have characterised in detail the role of two functional regulatory elements located downstream FOXP2, a gene important for speech and language (Fisher and Scharff 2009, Graham and Fisher 2013). Both elements affect the expression of FOXP2, but also that of the adjacent gene, MDFIC, a gene associated to developmental language and cognitive impairment (DECIPHER patients 262086, 292652, and 301696). MDFIC is highly expressed in the cerebellum during human embryonic development and in the thalamus after birth (Human Brain Transcriptome http://hbatlas.org/). These two brain regions seem to play an important role in language processing interacting in a dopaminergic cortico-striato-thalamic loop (Vargha-Khadem et al. 2005). Interestingly the cerebellum and thalamus of those bearing the R553H mutation in FOXP2 associated to speech and language impairment exhibit changes in their grey matter suggesting that the modulation of brain volume may impact in sensorimotor performance (Watkins et al. 2002b). One of these enhancers, FOXP2-Edistal, had been previously found to be functional in a luciferase assay (Becker et al. 2015). We have been able to prove further that if deleted FOXP2 becomes downregulated and the levels of FOXP2 protein are reduced. Conversely, its deletion upregulates MDFIC and increases the levels of MDFIC protein in the same SK-N-MC neuroectodermal tumor cell-line. The second enhancer, FOXP2-Eproximal was previously unknown. We have now found that it also upregulates FOXP2 and downregulates MDFIC. These findings are coherent with previous reports of two genes being regulated by the same enhancer (Gould et al., 1997; Tsujimura et al. 2010), which in some cases has proven to facilitate recruitment of RNA polymerase II to promoters of both genes (Collins et al. 2012).
Our findings in a neuronal cell line give support to the view that the breakpoint in our proband which separated these two functional enhancers may have altered the expression levels of both FOXP2 and MDFIC contributing to the observed speech and language deficits (Moralli et al 2015). Accordingly, we expect the expression of FOXP2 to be downregulated and the expression of MDFIC to be upregulated because of the displacement of both enhancers with respect to both genes (whereas FOXP2-E distal and MDFIC remained in chromosome 7, FOXP2-Eproximal and FOXP2 were rearranged to chromosome 11). The knockdown of FoxP2 in zebra finch results in a shorter window for song learning and in less accurate song imitation and performance (Haesler et al., 2007). This resembles the inability that those carrying the R553H mutation of FOXP2 show in repeating words and pseudowords (Watkins et al., 2002a). We expect that our findings also contribute to a better understanding of the role that this region may have played in the evolution of language. Differences in the expression levels of both FOXP2 and MDFIC are expected between extinct hominins and modern humans, plausibly accounting for some of the presumed differences in their language abilities. Neanderthals bear the ancestral allele of a binding site for POU3F2 within intron 8 of FOXP2, which is more efficient in activating transcription (Maricic et al. 2013). Accordingly, higher levels of FOXP2 are expected in this hominin species. Likewise, the MDFIC locus is among the top five percent S score regions in modern humans (Green et al. 2010, table S37). Finally, both genes are functionally related to RUNX2, which encodes an osteogenic factor that controls the closure of cranial sutures and several aspects of brain growth, and that has been related to the changes that brought about our more globular brain (case) and our species-specific mode of cognition, including language (Boeckx and Benitez-Burraco 2014; Benitez-Burraco and Boeckx 2015).
We expect that our study, together with new available data about seed sequences of miRs in the 3’UTR region of FOXP2 (Clovis et al. 2012; Shi et al. 2013; Fu et al. 2014a; Cuiffo et al. 2014), contributes to a deeper understanding of how FOXP2, but also MDFIC, are regulated, and how they each contribute to the development of brain function underlying language.
ACKNOWLEDGMENTS
This project was supported by funds from The University of Oxford John Fell OUP Research Grant [121/435] awarded to its Principal Investigator Paloma Garcia-Bellido. This study was supported in part by funds from the Spanish National Research and Development Plan, Institute de Salud Carlos III, and FEDER (FIS project no. PI14/01884 to Sandra Rodriguez-Perales), and in part by funds from the Spanish Ministry of Economy and Competitiveness (grant numbers FFI2014-61888-EXP and FFI-2013-43823-P to Antonio Benitez-Burraco). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.