Journal: bioRxiv

Article Title: Stoichiometric transcription factor partnerships specify GABAergic neuron subtype identity

doi: 10.64898/2026.05.25.727662

Figure Lengend Snippet: A , Genomic distribution of SP9 ChIP-seq peaks based on data from ( ; ). Peaks localize predominantly to enhancer-like regions (77%), with smaller fractions at promoters (6.2%) and other regions (16.8%). B , Motif enrichment within SP9 peaks. Promoter-associated peaks (left) are enriched for GC-rich motifs recognized by KLF and SP factors, whereas enhancer-like peaks (right) are enriched for TAATT motifs recognized by DLX homeobox factors. Within each block: motif logos (left), − log 10 𝑃 -value and target/background percentage of peaks containing the motif (middle), and log 2 enrichment (right). C , Schematic of the motif scan-based classification of SP9 peaks. Peaks were grouped based on the presence of an SP motif only (direct binding), a DLX motif only (indirect binding via DLX), or both (cobinding). D , Coarse genomic annotation (enhancer, transcription start site (TSS), or other) of the four SP9 peak categories defined in (C): SP9+DLX (both motifs present), DLX (DLX motif only), Other, and SP9 (SP motif only). E , Fine genomic annotation (exon, intron, intergenic, promoter, UTR, other) of the same four peak categories. F , Intersection of DLX2 and SP9 ChIP-seq peaks from E13.5 mouse GE, based on data from ( ; ; ). Bar plot shows the number of overlapping and non-overlapping peaks per category (DLX2 only, overlap, SP9 only). G , Whole-mount X-Gal staining (blue) of E11.5 transgenic mouse embryos carrying VISTA enhancer reporters hs119 (near Arx ), hs170 (near Fign ), hs883 (near Sox6 ), and hs298 (near Dlx6os1 ) used in the luciferase assays in panels K-N and shown as genome browser tracks in panel H. Images from . H , Genome browser tracks showing ChIP-seq signal for SP9, DLX2, and DLX5 in E13.5 mouse GE at the Sp9 and Six3 promoters (highlighted in grey) and the VISTA enhancers hs119, hs170, hs883, and hs298 (highlighted in blue). Numbers below each track indicate the count of SP and DLX motifs within the selected regulatory element. Data from ( ; ; ). I-N , Luciferase reporter activity in N2A cells transfected with combinations of Sp9 and Dlx2 , Dlx5 , or Dlx6 as indicated, driven by the Sp9 promoter ( I ), Six3 promoter ( J ), or enhancer regions hs119 of Arx ( K ), hs170 of Fign ( L ), hs883 of Sox6 ( M ), and hs298 of Dlx6os1 ( N ). Bars represent mean ± s.e.m. of 9 (I,J,L-N) or 12 (K) replicates from 3 (I,J,L-N) or 4 (K) independent batches performed in triplicate; points indicate batch means. Statistical significance was assessed by two-way ANOVA with Tukey’s honestly significant difference (HSD) post hoc test. Exact 𝑃 -values are provided in Table S6.

Article Snippet: Mouse N2A neuroblastoma cells (Neuro2A cells, ECACC, 89121404), human HEK293FT (Thermo Fisher Scientific, R70007), mouse 3T3-L1 (CLS Cell line service, 400103) were cultured in Dulbecco’s modified Eagle medium (DMEM, Sigma, D6429) supplemented with 10% (v/v) fetal bovine serum (FBS, Sigma, F9665) and containing 1% (v/v) antibiotics (100 Uml −1 penicillin, 100 μgmL −1 streptomycin, Sigma, P0781).

Techniques: ChIP-sequencing, Blocking Assay, Binding Assay, Staining, Transgenic Assay, Luciferase, Activity Assay, Transfection

Journal: bioRxiv

Article Title: Stoichiometric transcription factor partnerships specify GABAergic neuron subtype identity

doi: 10.64898/2026.05.25.727662

Figure Lengend Snippet: A , Schematic of C-terminally V5-tagged SP9 constructs: wild-type (WT) SP9, an N-terminal deletion mutant (NΔSP9), and a zinc finger (ZF) domain deletion mutant (SP9ΔZF). The nine-amino-acid transactivation domain (9aaTAD), the three ZF domains, and the position of the patient-derived SP9*378 variant are indicated. B , Co-immunoprecipitation (CoIP) from nuclear lysates of HEK293FT cells co-transfected with DLX5-FLAG and either SP9-V5 WT, SP9*378-V5, or NΔSP9V5. Left: input fractions probed with anti-SP9 (top) and anti-DLX5 (bottom). Right: CoIP using anti-FLAG beads, probed with anti-SP9 (top) and anti-DLX5 (bottom). C , CoIP from whole-cell lysates of HEK293FT cells co-transfected with DLX5-FLAG and either SP9-V5 WT or SP9ΔZF-V5. Left: input fractions probed with anti-V5 (top) and anti-DLX5 (bottom). Right: CoIP using anti-FLAG beads, probed with anti-V5 (the anti-SP9 epitope lies within the deleted ZF region, requiring V5 detection; top) and anti-DLX5 (bottom). D , Luciferase reporter activity driven by the hs298 enhancer of Dlx6os1 in N2A cells co-transfected with Dlx5 and either Sp9 , Sp9*378 , N Δ Sp9 , or Sp9 Δ ZF . Bars represent mean ± s.e.m. of9replicatesfrom3independentbatchesperformedintriplicate; pointsindicatebatchmeans. Statistical significance was assessed by two-way ANOVA with Tukey’s honestly significant difference (HSD) post hoc test. Exact 𝑃 -values are provided in Table S6. E , Label-free quantitative mass spectrometry (MS) of proteins captured by DNA pull-down using a 46-bp fragment of the hs298 enhancer of Dlx6os1 versus the same fragment with the TAATT motifs scrambled, from E14.5 ganglionic eminence (GE) nuclear lysates. Proteins with p ≤ 0.05 and log 2 LFQ FC ≥ 1.5 were considered significantly enriched (n = 25). Transcription factors (TFs) are indicated in black. F , Volcano plot of proteins enriched in anti-SP9 CoIP-MS from E14.5 GE nuclear lysates, relative to IgG controls. Proteins with log 2 FC ≥ 1.5 and p ≤ 0.05 were considered significantly enriched (n = 223). Components of HDAC1/2-containing complexes are highlighted in blue; TFs are indicated in black. 𝑃 -values were calculated using a two-sided Student’s t-test with permutation-based false discovery rate (FDR) correction. G , Gene Ontology (GO) molecular function enrichment analysis of proteins identified by anti-SP9 CoIP-MS. H , Heatmap of ChIP-seq signal intensity across SP9-bound regions for SP9 ( ; ), DLX2, histone modifications (H3K27ac, H3K4me1, H3K4me3) , GTF2I , and NuRD complex subunits (MBD3, CHD4, RBBP4, RBBP7, HDAC1, HDAC2) . Signal is plotted over a ±5 kb window centered on SP9 peak summits and organized by k-means clustering (5 clusters), separating SP9-bound regions by their co-binding patterns with DLX2, NuRD subunits, and active histone marks. All datasets are from mouse E13.5 GE except GTF2I, which is from E13.5 whole brain.

Article Snippet: Mouse N2A neuroblastoma cells (Neuro2A cells, ECACC, 89121404), human HEK293FT (Thermo Fisher Scientific, R70007), mouse 3T3-L1 (CLS Cell line service, 400103) were cultured in Dulbecco’s modified Eagle medium (DMEM, Sigma, D6429) supplemented with 10% (v/v) fetal bovine serum (FBS, Sigma, F9665) and containing 1% (v/v) antibiotics (100 Uml −1 penicillin, 100 μgmL −1 streptomycin, Sigma, P0781).

Techniques: Construct, Mutagenesis, Derivative Assay, Variant Assay, Immunoprecipitation, Transfection, Luciferase, Activity Assay, Mass Spectrometry, ChIP-sequencing, Binding Assay

Journal: bioRxiv

Article Title: Stoichiometric transcription factor partnerships specify GABAergic neuron subtype identity

doi: 10.64898/2026.05.25.727662

Figure Lengend Snippet: A , Schematic of the transcription factor (TF) overexpression strategy in N2A cells. Cells were transfected with SP9 alone or co-expressed with WT DLX5 (overexpression experiment 1, OE1) or with the N-terminal deletion mutant NΔDLX5 (OE2), followed by CUT&RUN profiling. B , Domain organization of SP9, DLX5, and NΔDLX5, highlighting the zinc finger (ZF) and homeobox domains. C , Heatmaps of SP9 CUT&RUN signal (anti-V5) centered on SP9 peak regions (±5 kb), clustered into four groups (C1-C4) by k-means. Conditions: SP9 alone, SP9+DLX5 (OE1), SP9+NΔDLX5 (OE2). D , Aggregate signal plots for each cluster (C1-C4) showing normalized SP9 binding profiles at peak centers under the three conditions in (C). E , Top enriched motifs identified by motif analysis of peaks in each cluster, with associated 𝑃 -values. F , Genomic annotation of SP9-bound peaks across clusters. Left: distribution across promoter, exon, intron, intergenic, and other regions. Right: distance to transcription start sites (TSS). G , Genome browser tracks at three representative loci showing SP9 binding under OE1 and OE2 conditions, illustrating loss of SP9 occupancy upon DLX5 coexpression at SP-motif-containing sites and rescue with NΔDLX5, as well as gain of SP9 occupancy at DLX-motif-containing sites with WT DLX5. H , Luciferase reporter activity driven by the Six3 promoter in N2A cells transfected with combinations of Sp9 , Dlx5 , and N Δ Dlx5 as indicated. I , Volcano plot of co-immunoprecipitation followed by mass spectrometry (CoIP-MS) using anti-V5 in N2A cells co-transfected with Sp9-V5 and Dlx5-FLAG , relative to IgG controls. Proteins with log 2 FC > 1.5 and p < 0.05 were considered significantly enriched (n = 2254). NuRD complex components are highlighted in blue, transcription factors (TFs) in black, and AP-1 family TFs in green. J , Volcano plot comparing the anti-V5 CoIP-MS interactomes between SP9+DLX5 and SP9+NΔDLX5 conditions in N2A cells, using the same cutoffs as in (I). I,J , 𝑃 -values were calculated using a two-sided Student’s t-test with permutation-based false discovery rate (FDR) correction. K , Luciferase reporter activity driven by the Six3 promoter in N2A cells transfected with varying amounts (ng plasmid DNA) of SP9 (S) and DLX5 (D), showing dose-dependent transcriptional activation. H-K , Bars represent mean ± s.e.m. of 9 replicates from 3 independent batches performed in triplicate; points indicate batch means. Statistical significance was assessed by two-way ANOVA with Tukey’s honestly significant difference (HSD) post hoc test. Exact 𝑃 -values are provided in Table S6. L , Proposed mechanism. When SP9 levels exceed DLX (SP9≫DLX), SP9 binds directly at GC-rich SP motifs. When SP9 and DLX are comparable or DLX is in excess (SP9≈DLX or SP9 < DLX), DLX5 sequesters SP9 to TAATT-containing DLX motifs through the DLX5 N-terminal domain, away from its direct SP-motif binding sites. NΔDLX5, which cannot interact with SP9, fails to redirect SP9 to DLX motifs, restoring SP9 binding at SP motifs.

Article Snippet: Mouse N2A neuroblastoma cells (Neuro2A cells, ECACC, 89121404), human HEK293FT (Thermo Fisher Scientific, R70007), mouse 3T3-L1 (CLS Cell line service, 400103) were cultured in Dulbecco’s modified Eagle medium (DMEM, Sigma, D6429) supplemented with 10% (v/v) fetal bovine serum (FBS, Sigma, F9665) and containing 1% (v/v) antibiotics (100 Uml −1 penicillin, 100 μgmL −1 streptomycin, Sigma, P0781).

Techniques: Over Expression, Transfection, Mutagenesis, Binding Assay, Luciferase, Activity Assay, Immunoprecipitation, Mass Spectrometry, Plasmid Preparation, Activation Assay