AUTS2 Controls Neuronal Lineage Choice Through a Novel PRC1-Independent Complex and BMP Inhibition

Abstract

Despite a prominent risk factor for Neurodevelopmental disorders (NDD), it remains unclear how Autism Susceptibility Candidate 2 (AUTS2) controls the neurodevelopmental program. Our studies investigated the role of AUTS2 in neuronal differentiation and discovered that AUTS2, together with WDR68 and SKI, forms a novel protein complex (AWS) specifically in neuronal progenitors and promotes neuronal differentiation through inhibiting BMP signaling. Genomic and biochemical analyses demonstrated that the AWS complex achieves this effect by recruiting the CUL4 E3 ubiquitin ligase complex to mediate poly-ubiquitination and subsequent proteasomal degradation of phosphorylated SMAD1/5/9. Furthermore, using primary cortical neurons, we observed aberrant BMP signaling and dysregulated expression of neuronal genes upon manipulating the AWS complex, indicating that the AWS-CUL4-BMP axis plays a role in regulating neuronal lineage specification in vivo. Thus, our findings uncover a sophisticated cellular signaling network mobilized by a prominent NDD risk factor, presenting multiple potential therapeutic targets for NDD.

Keywords: Autism susceptibility candidate 2 (AUTS2); BMP signaling; Embryonic stem cell; Neurodevelopmental disorders; Neuronal differentiation; SKI.

FIGURE 1.. Auts2 is required for neuronal differentiation of mouse embryonic stem cells.

a, Immunoblotting of wild-type (WT) mouse embryonic stem cells (ESC), embryoid bodies (EB), and neuronal progenitor cells (NPC), using antibodies as indicated. The long and short isoforms of Auts2 are labeled as Auts2-L and Auts2-S, respectively. * indicates a non-specific protein. b, Immunoprecipitation (IP) with FLAG antibody-conjugated M2 beads from nuclear extract (NE) of HEK293T cells transfected with N-terminal FLAG and HA-tagged AUTS2 long-form (NFH-AUTS2-L), short-form (NFH-AUTS2-S), or control plasmid (Vector). Bound proteins were resolved on SDS-PAGE and detected by immunoblotting for the antigens indicated. 5% of input was loaded in all cases unless otherwise indicated. c, Immunofluorescence staining of Nestin (Green) and Auts2 (Red) in mouse ESCs differentiated at NPC stage. DAPI is in Blue. The bottom panel shows the blowup of the box region as indicated. Auts2−/−, complete Auts2 knockout; Auts2-L−/−, Auts2-L specific knockout; Auts2-S−/−, Auts2-S specific knockout. d, Quantification of average Nestin immunofluorescence intensity by Image J. Each value is the mean of three independent measurements with error bars representing standard error. ** indicates p<0.01 by two-sided t-test.

FIGURE 2.. Auts2 controls the proper expression of lineage-specific genes.

a, Heatmap of transcriptomic analysis in WT and Auts2−/− ESCs and NPCs. RNA-seq analyses were performed on duplicate samples. The processed FPKM (reads per kilobase of exon per million reads mapped) values for each gene were used to calculate z scores used to generate the heatmap (See Supplemental Information for details). b, GO analysis of genes that are down- or up-regulated in Auts2−/− NPCs compared with WT. The x-axis (in logarithmic scale) corresponds to the binomial raw P values. c, Whisker-box plot shows the expression of genes targeted by SMAD1/5 or SMAD2/3. Fold change of RPKM values between WT and Auts2−/− NPCs from all expressed genes, down- or up-regulated genes, down- or up-regulated SMAD1/5-targeted genes, and down- or up-regulated SMAD2/3-targeted genes are plotted. SMAD1/5 and SMAD2/3 target genes are obtained from (Morikawa et al., 2011). d, Expression of BMP responsive genes during differentiation, measured by quantitative RT-PCR. All mean values and standard deviations were calculated from three independent measurements. ** indicates p<0.01 by two-sided t-test.

FIGURE 3.. Auts2 is required for the inhibition of BMP signaling during neuronal differentiation.

a, Immunoblotting of WT and Auts2−/− mouse ESC, EB, and NPC using antibodies as indicated. b, WT and Auts2−/− ESCs were stimulated with BMP4 at 25 μM for the time as indicated. Activation of BMP signaling was measured by the level of Smad1/5/9 phosphorylation by immunoblotting. c, Expression of selected marker genes (neuroectoderm: Pax6, Nes, and Neurod1; mesoderm: Gata4; endoderm: Gata6) in Noggin-treated or untreated NPC of WT or Auts2−/−, measured by quantitative RT-PCR. All mean values and standard deviations were calculated from three independent measurements. * p<0.05, ** p<0.01, n.s. not significant.

FIGURE 4.. A novel protein complex formed by Auts2, Wdr68, and Ski.

a, Tandem affinity purification (TAP) followed by quantitative mass spectrometry analysis (MS) revealed a novel binding partner for Wdr68 in NPC, shown by a volcano blot. An E14 cell line with engineered Wdr68 locus inserted with N-terminal FLAG and HA tags (E14: NFH-Wdr68) was either cultured at ESC or differentiated to NPC stages. TAP was performed in triplicate samples of ESCs and NPCs, followed by label-free quantitative MS analysis (see Experimental Procedures). The X-axis shows the mean ratio of peptide intensity between NPC and ESC, and the Y-axis corresponds to P-value. b, IP from NE of E14: NFH-Wdr68 cells at the ESC or NPC stages, using FLAG M2 beads. Bound proteins were resolved on SDS-PAGE and detected by Western blotting for the antigens indicated. c, IP from NE of HEK293T cells transfected with NFH-AUTS2-L or NFH-AUTS2-S or vector control, using FLAG M2 beads. Bound proteins were resolved on SDS-PAGE and detected by Western blotting for the antigens indicated. d, Mapping of AUTS2 domains required for interaction with indicated proteins or protein complexes. Plasmids expressing NFH-AUTS2 with various lengths were expressed in HEK293T cells followed by IP with M2 beads to detect their interaction with PRC1 components, WDR68 and SKI, as indicated on the right (See Experimental Procedures for details). The deduced domains required for specific interaction are highlighted and indicated at the bottom. e, Glycerol gradient (15–35%) analysis of FLAG-purified NFH-AUTS2-S (See Experimental Procedures). Every other fraction was resolved on SDS-PAGE followed by immunoblotting for the antigens indicated. The fractions containing AUTS2, WDR68, and SKI simultaneously indicate the presence of a complex formed with these proteins (AWS). f, Schematic model of AUTS2-containing complexes. See text for details.

FIGURE 5.. The AWS complex promotes degradation of pSMAD1/5/9 through CUL4 E3.

a, IP in HEK293T cells transfected with NFH-AUTS2 and HA-DDB1, using FLAG M2 beads. Bound proteins were resolved on SDS-PAGE and detected by Western blotting for the antigens indicated. b, in vivo poly-ubiquitination assay. HEK293T cells were transfected with HA-ub and FLAG-SMAD1. Two days after transfection, cells were lysed in the presence of 1% SDS and immunoprecipitated with HA beads, followed by immunoblotting with SMAD1 antibody. c, Cullin E3 negatively regulates the level of pSMAD1/5/9. HEK293T cells were stimulated with BMP4 at various times as indicated with or without prior treatment with MLN4924, a pan inhibitor for CUL E3, followed by immunoblotting with antibodies as indicated. d, Two days after treatment of control or siRNA for DDB1, HEK293T cells were stimulated with BMP4 at various times as indicated, followed by immunoblotting. e, Two days after transfection with a dominant-negative form of CUL4 (DN-CUL4B), HEK293T cells were stimulated with BMP4 at various times as indicated, followed by immunoblotting. f, HEK293T cells were transfected with AUTS2 and SKI, individually or combinatorically, then stimulated with BMP4, followed by immunoblotting.

FIGURE 6.. Ski mediates Auts2 regulation on neuronal differentiation.

a, RNA-seq reads confirm the expression of human-specific short form AUTS2 (hAUTS2) with or without deletion between 600 to 700 amino acids (hAUTS2-S-Δ and hAUTS2-S) through lentiviral infection in Auts2−/− mouse ESCs. A schematic of the deleted region of AUTS2 that promotes AUTS2 -WDR68 and AUTS2 -SKI interaction is shown at the bottom. b, Comparison of gene expression levels based on RNA-seq analysis from WT and Auts2−/− NPCs with lentiviral infection with mock, hAUTS2-S and hAUTS2-S-Δ. The x-axis is the log₂ value of the average FPKM of a gene and the y-axis is the log2 value of fold changes of FPKM of a gene between two groups. Genes with an average FPKM below one are filtered out. c, Heatmap of transcriptomic analysis in WT and Auts2−/− ESCs and NPCs, with lentiviral infection as indicated. Z scores were calculated using processed FPKM values for each gene as in Fig. 2a. d, Expression of selected marker genes (pluripotency: Oct4 and Nanog; neuroectoderm: Pax6, Nes, and Neurod1; mesoderm: Hand1 and Gata4; endoderm: Gata6) in ESC or NPC of WT or Ski−/−, measured by quantitative RT-PCR. All mean values and standard deviations were calculated from three independent measurements. * p<0.05, ** p<0.01, n.s. not significant. e, Immunoblotting of WT and Ski−/− mouse ESC and NPC using antibodies as indicated.

Figure 7.. Auts2 is required for proper gene expression and BMP signaling in mouse primary cortical neurons

a, Genotyping PCR in WT and Auts2+/− mouse primary cortical neurons. b, Relative expression of neuron markers between WT and Auts2+/− mouse cortical neurons. c, Relative expression of mesoderm and endoderm marker genes. (b and c) Each value is the mean of three independent measurements with error bars representing standard error. * p<0.05, ** p<0.01. d, WT cortical neurons were treated with MLN4924 or DMSO for 1 hour, then treated with BMP4 for 0, 15, and 30 minutes, followed by immunoblotting. e, Model of AWS/CUL4 E3-mediated regulation of BMP signaling during neuronal differentiation. See text for details.