Impairment of DET1 causes neurological defects and lethality in mice and humans

Significance

DET1, together with COP1, recruits specific transcription factors for polyubiquitination by the E3 ligase CRL4^COP1/DET1. This polyubiquitination targets the transcription factors for proteasomal degradation and ensures they only accumulate to promote transcription when signaling inactivates CRL4^COP1/DET1. Here, we illustrate the essential role of DET1 in neurodevelopment. Mice lacking Det1 died during embryogenesis, whereas mice lacking Det1 just in neural stem cells exhibited neurological defects and died after birth. We also identify an inactivating DET1 mutation that is associated with lethal developmental abnormalities in humans. This mutation caused aberrant expression of CRL4^COP1/DET1 substrates, abnormal gene expression, mitochondrial dysfunction, and enhanced cell death during neuronal differentiation. DET1 was also important in microglia, limiting their neurotoxicity.

Abstract

COP1 and DET1 are components of an E3 ubiquitin ligase that is conserved from plants to humans. Mammalian COP1 binds to DET1 and is a substrate adaptor for the CUL4A-DDB1-RBX1 RING E3 ligase. Transcription factor substrates, including c-Jun, ETV4, and ETV5, are targeted for proteasomal degradation to effect rapid transcriptional changes in response to cues such as growth factor deprivation. Here, we link a homozygous DET1^R26W mutation to lethal developmental abnormalities in humans. Experimental cryo-electron microscopy of the DET1 complex with DDB1 and DDA1, as well as co-immunoprecipitation experiments, revealed that DET1^R26W impairs binding to DDB1, thereby compromising E3 ligase function. Accordingly, human-induced pluripotent stem cells homozygous for DET1^R26W expressed ETV4 and ETV5 highly, and exhibited defective mitochondrial homeostasis and aberrant caspase-dependent cell death when differentiated into neurons. Neuronal cell death was increased further in the presence of Det1-deficient microglia as compared to WT microglia, indicating that the deleterious effects of the DET1 p.R26W mutation may stem from the dysregulation of multiple cell types. Mice lacking Det1 died during embryogenesis, while Det1 deletion just in neural stem cells elicited hydrocephalus, cerebellar dysplasia, and neonatal lethality. Our findings highlight an important role for DET1 in the neurological development of mice and humans.

Cullin-RING E3 ubiquitin ligases (CRLs) represent the largest family of E3 ubiquitin ligases, encompassing over 600 members in mammals and playing pivotal roles in diverse cellular processes (1–7). CRLs are multisubunit E3 ligases and they target various protein substrates for proteasomal degradation. COP1 and DET1 together form a substrate adaptor of the Cullin 4A (CUL4A), DNA damage-binding protein-1 (DDB1), and regulator of cullins-1 (RBX1) RING E3 ligase (CRL4). In plants, COP1 and DET1 regulate photomorphogenesis by targeting several transcription factors for proteasomal degradation (8). In mammals, CRL4^COP1/DET1 targets transcription factors of the AP-1, ETS, and c/EBP families. COP1 is widely expressed and essential during mouse embryogenesis (9). Restricting Cop1 loss to specific cell types in mice has highlighted the role of COP1 in tumor suppression (10), lung homeostasis (11), glucose metabolism (12), and brain development (13). One interesting facet of CRL4^COP1/DET1 biology is that the activity of the ligase is suppressed by ERK signaling, which allows its substrates to accumulate in cells exposed to growth factors (11, 14).

DET1 is located on chromosome 15 approximately 13 Mb away from COMMD4. A homozygous missense mutation (c.122T>G; p.L41R) in COMMD4 was recently associated with severe developmental abnormalities in three siblings born to a consanguineous Middle Eastern family (15). COMMD4 is a component of the Commander complex involved in endosomal trafficking (16–18). Mutations in CCDC22 or VPS35L, which encode distinct components of the Commander complex, cause the developmental malformation known as Ritscher–Schinzel syndrome (RSS) (19–24). The three siblings exhibited RSS-like features, but their disease was more severe than typically described (15).

We now report that the siblings also harbored a homozygous mutation in DET1. Both parents were healthy, heterozygous carriers of the DET1 p.R26W mutation. Using Det1-deficient mice and human-induced pluripotent stem cells (iPSCs) expressing DET1^R26W, we show that DET1 is essential for normal neuronal development. Mechanistically, the p.R26W mutation compromised the interaction between DET1 and DDB1, resulting in the aberrant accumulation of known CRL4^COP1/DET1 substrates and increased neuronal cell death.

Results

A Homozygous DET1 Mutation in Siblings with Neonatal Lethality.

Three siblings with a homozygous mutation in COMMD4 exhibited a severe RSS-like congenital multisystem disorder characterized by dysmorphic facial features, cardiac abnormalities, gastrointestinal complications, and neurological impairments, including microcephaly (15). To determine whether mutations in additional genes might explain their disease severity, we performed further sequencing. A second homozygous mutation in DET1 (c.76C>T, p.R26W) was identified in the siblings, and their parents were heterozygous carriers (Fig. 1 A and B). DET1 and COMMD4, both found on chromosome 15, are predicted to cosegregate. The heterozygous DET1 missense mutation is exceedingly rare in the general population, occurring at a frequency of 0.00005257 in the Genome Aggregation and 0.0001867 in the United Kingdom Biobank. The homozygous mutation has not been reported previously. These findings suggest that homozygous mutations in both DET1 and COMMD4 could contribute to the more severe clinical phenotype of the siblings, beyond what is typically seen with mutations in the Commander complex alone.

Fig. 1.

Identification of human patients bearing mutant DET1 and the generation of mice lacking Det1 in neural stem and progenitor cells. (A) Patient pedigree chart. Mutation screening was done for the patients as well as their parents and healthy siblings. Patients are indicated by red filled symbols. (B) DNA sequence of the DET1^R26W missense patient mutation. (C) Littermates aged 4 wk. (D) Hematoxylin and eosin-stained brain sections from mice aged 4 wk. Results representative of 5 Det1^{fl/fl or fl/−} Nestin.cre and 4 Det1^{fl/+ or +/+} Nestin.cre mice aged 3 to 4 wk. (Scale bars: cortex, 800 μm; cerebellum, 200 μm.) (E) Immunolabeling of c-JUN (brown) in brain sections from mice aged 4 wk. (Scale bars: cortex, 100 μm; cerebellum, 200 μm.) Results representative of 3 Det1^{fl/fl or fl/−} Nestin.cre and 3 control (Det1^fl/+Nestin.cre or Det1^+/+) mice aged 4 wk.

To further examine the role of DET1 during development, we studied Det1^−/− mice (1). No Det1^−/− pups were born to Det1^+/− parents (SI Appendix, Fig. S1A), indicating that DET1, like COP1 (9), is essential for mouse embryogenesis. Det1^−/− embryos were identified between embryonic days 9.5 and 11.5, but were smaller than their wild-type (WT) and heterozygous littermates (SI Appendix, Fig. S1A). A similar phenotype was reported for Cop1^−/− embryos, which were lost by E12.5 (9). These data underscore the pivotal role of CRL4^COP1/DET1 in early mouse development.

Det1 Deficiency in Neural Stem and Progenitor Cells Causes Neonatal Lethality in Mice.

The neurological abnormalities and microcephaly associated with the homozygous DET1 p.R26W mutation prompted us to investigate the role of DET1 in the brain. We generated Det1^fl/− Nestin.cre and Det1^fl/fl Nestin.cre mice because the Nestin.cre transgene converts floxed (fl) alleles to knockout (−) alleles in neural stem and progenitor cells (25). Genotyping of the cerebral cortex, cerebellum, brainstem, and tail of Det1^fl/− Nestin.cre mice confirmed successful Det1 deletion in the brain (SI Appendix, Fig. S1B). Breeding of Det1^+/- Nestin.cre mice with Det1^fl/fl mice yielded Det1^fl/− Nestin.cre offspring, but they were underrepresented at 1 to 7 d of age at approximately 50% of the expected frequency (SI Appendix, Fig. S1C). Det1^{fl/− or fl/fl} Nestin.cre mice failed to thrive (Fig. 1C and SI Appendix, Fig. S1D), exhibited trembling, and had a median survival of only 22 d (SI Appendix, Fig. S1E). Histological analyses revealed hydrocephalus and cerebellar dysplasia in the Det1^{fl/− or fl/fl} Nestin.cre brains (Fig. 1D). Immunolabeling of c-JUN in the cortex and cerebellum showed that Det1^{fl/− or fl/fl} Nestin.cre mice had more strongly positive cells than control mice (Fig. 1E), consistent with inactivation of CRL4^COP1/DET1 causing an accumulation of c-JUN. Collectively, these data indicate that DET1, like COP1 (13), is essential for normal mouse brain development.

DET1^R26W Is Detrimental to the Development of Human Neurons.

We studied the homozygous DET1 p.R26W mutation in human iPSCs. Neurons derived from two independent clones of DET1^R26W cells were compared to their WT counterparts using a Neurogenin-2 (NGN2)-driven iPSC-to-neuron differentiation protocol (26, 27) (Fig. 2A and SI Appendix, Fig. S2A). WT and DET1^R26W iPSCs expressed equivalent amounts of DET1 (Fig. 2B), indicating that the p.R26W mutation does not compromise DET1 protein stability. DET1^R26W iPSCs contained more ETV4 and ETV5 protein than WT iPSCs despite expressing less ETV4 and ETV5 mRNA (Fig. 2 B and C), consistent with DET1^R26W disabling CRL4^COP1/DET1 ligase activity. During neuronal differentiation, DET1 mutant iPSCs exhibited less proliferation (Fig. 2D) and released more lactate dehydrogenase (LDH), indicative of cell death, than control iPSCs (Fig. 2 E and F). Cell death in both the WT and mutant cultures was accompanied by cleavage of caspase-3 and its substrate PARP (SI Appendix, Fig. S2B), consistent with apoptosis induction. The pan-caspase inhibitor emricasan suppressed this cell death (Fig. 2G and SI Appendix, Fig. S2 C and D), resulting in more viable neurons (Fig. 2H). In line with these observations, DET1 mutant neuronal cultures contained fewer MAP2⁺ or TUBB2A⁺ neurons on day 14 than WT cultures (Fig. 3A). These data indicate that DET1^R26W is a loss-of-function mutant that impairs the development and morphogenesis of human neurons.

Fig. 2.

Differentiation of WT and DET1^R26W iPSCs into neurons. (A) Representative WT and DET1^R26W cells during NGN2-driven iPSC-to-neuron differentiation. (Scale bar, 100 μm.) (B) Western blots of two clones of DET1^R26W iPSCs. (C) ETV4 and ETV5 mRNA expression in iPSCs. Symbols represent biological replicates. Lines indicate median values. (D) Cell confluency during NGN2-driven iPSC-to-neuron differentiation. Initial seeding density = 4,000 cells/well. Data are the mean ± SEM (n = 4 biological replicates). ****P < 0.0001 by the unpaired t test. (E) LDH detected in the media of differentiating iPSCs. Data are the mean ± SEM. n = 3 biological replicates. ****P < 0.0001 by the unpaired t test. (F) Percentage of LDH released from differentiating iPSCs at day 7. Data are the mean ± SEM. n = 4 biological replicates for WT and DET1^R26W. (G) Number of dead cells staining with Cytotox Red (4,632, Sartorius) at day 7. Data are the mean ± SEM. n = 6 to 18 biological replicates. Emr, emricasan. (H) Cell viability determined by CellTiter-Glo assay (G7570, Promega) at day 14. Initial seeding density = 6,000 cells/well. Data are normalized to confluency and are mean ± SEM (n = 6 to 18 biological replicates).

Fig. 3.

Enhanced neurotoxicity of Det1-deficient microglia in vitro. (A) Immunolabeling of MAP2 and TUBB2A in day 14 iPSC-derived neurons. (Scale bars, 50 μm.) Data are the mean ± SEM. n = 6 to 16 biological replicates. **P < 0.01 by the unpaired t test. (B) Cytokines and chemokines secreted by mouse microglia. Cells were treated with 100 ng/mL LPS overnight. Circles represent independent experiments. WT untreated, n = 2; WT LPS treated, n = 3; KO untreated, n = 3; KO LPS treated, n = 4. Bars, mean ± SEM. **P-value < 0.01, ***P-value < 0.005, ****P-value < 0.001 by the t test. (C) Representative cocultures of iPSC-derived neurons (red) and mouse microglia (green). (Scale bars, 100 μm.) MAP2 neuronal staining (red) is graphed. Bars indicate the mean ± SEM (n = 2 to 6 biological replicates). *P < 0.05, **P < 0.01 and ***P < 0.005 by the unpaired t test.

The DET1^R26W mutation might also impact nonneuronal cells, including microglia. Cop1-deficient microglia have a proinflammatory phenotype and are more toxic than WT microglia to cocultured neurons (28). Det1-deficient microglia also produced more chemokines and cytokines than WT microglia (Fig. 3B and SI Appendix, Fig. S2E). WT and DET1^R26W neurons exhibited reduced viability when cultured with Det1-deficient microglia as compared to WT microglia (Fig. 3C). WT microglia themselves did not appear to impact neuronal viability. Therefore, a global deficit in DET1 function might impact neuronal function directly by compromising neuronal differentiation and indirectly by exacerbating microglial toxicity.

RNA sequencing (RNAseq) and proteomic analyses of WT and DET1^R26W iPSCs before and during NGN2-driven differentiation illuminated the processes compromised by DET1^R26W. The transcriptomes of undifferentiated WT and DET1^R26W iPSCs were markedly different (30% of genes exhibited differential expression), but cells of both genotypes that survived the neuronal differentiation protocol to day 7 up-regulated neuronal differentiation markers (SI Appendix, Fig. S3 A–D). Gene set enrichment analysis of the iPSCs highlighted the epithelial-to-mesenchymal transition (EMT), with WT iPSCs expressing genes associated with the epithelial state (q value of 0.004) and DET1^R26W iPSCs showing similarity to the mesenchymal state (q value of 0.00018) (SI Appendix, Fig. S3 E and F). The EMT signature alone was sufficient to clearly distinguish genotypes at both the iPSC stage and day 7 (SI Appendix, Fig. S3G). Exploring the top 100 differentially expressed genes in WT and DET1^R26W iPSCs revealed dysregulated expression of genes pivotal for developmental processes (Fig. 4A). For example, ZEB1 and ZEB2, encoding the transcriptional repressors that drive EMT (29), were up-regulated in DET1^R26W iPSCs, as were the transcriptional regulators NR2F1, NR2F2, SOX1, and SOX6. Genes that were down-regulated in DET1^R26W iPSCs encoded metallothioneins involved in metal detoxification, and nuclear proteins, including PRAC1, PRAC2, and histones (Fig. 4A).

Fig. 4.

RNA sequencing and proteomics of iPSC-to-neuron differentiation. (A) Volcano plot comparing the transcriptomes of WT and DET1^R26W iPSCs. The FDR-based top 50 significantly up- or down-regulated genes are highlighted. Data collected from three biological replicates. (B) Normalized abundances of the proteins in cluster 1. Plot displays the median (line), and the lower (0.25) and upper (0.75) quartiles (gray area). Data collected from three biological replicates. (C) Overrepresentation analysis of the proteins in cluster 1. Proteins were annotated with GO terms for Biological Process, Molecular Function, and Cellular Component. −log10 FDR and enrichment scores calculated using Fisher’s exact test. The top five most significant terms for each category are shown. (D) Oxygen consumption rates (OCR) in neurons at baseline and under stress. Data are the mean ± SD (n = 4 biological replicates per genotype). **P < 0.01 by the unpaired t test. (E and F) Overrepresentation analysis of transcripts and proteins of WT and DET1^R26W neurons. Proteins were annotated with GO Biological Process terms. −log10 Benjamini–Hochberg adjusted P-values (FDR) and scores determined using the 1D enrichment test. (G) Transcription factor enrichment analysis of the FDR-based top 100, 200, and 500 differentially expressed genes in DET1^R26W versus WT iPSCs. Heatmap illustrates the enrichment scores of the top 10 overlapped transcription factors.

More than 80% of the WT proteome underwent significant changes during iPSC-to-neuron differentiation, and DET1^R26W cells exhibited substantial differences from WT cells at all time points (SI Appendix, Fig. S4A). Hierarchical clustering of the proteins expressed differentially in WT and DET1^R26W cells (ANOVA, FDR < 0.01) revealed their differing trajectories (SI Appendix, Fig. S4B). The largest distinct cluster (cluster 1, n = 162) indicated that DET1^R26W cells express fewer mitochondrial proteins at day 7 than WT cells (Fig. 4B). This cluster was significantly enriched for cellular processes associated with mitochondria (Fig. 4C). The transition from the predominantly glycolytic progenitor state to mitochondrial aerobic respiration in mature neurons is a crucial phase of neuronal differentiation (30). Seahorse analyses confirmed that mitochondrial function was impaired in DET1^R26W neurons; they had a lower baseline OCR than their WT counterparts, as well as a diminished response to mitochondrial stress (Fig. 4D and SI Appendix, Fig. S4 C and D). Therefore, impaired mitochondrial homeostasis may contribute to the neurological defects associated with DET1^R26W.

The fold changes of genes and proteins identified as common across both the proteomics and transcriptomics datasets showed a positive correlation for both iPSCs and neurons (SI Appendix, Fig. S5 A and B). The pathways down-regulated significantly in DET1^R26W neurons were associated with lipid metabolism, intracellular and axonal transport, and synaptic signaling, including exocytosis and secretion (Fig. 4 E and F). For example, DET1^R26W cells expressed low levels of Synapsin I, a protein regulating synaptic vesicle trafficking (See SI Appendix, Fig. S4B, Cluster 11). Mitochondrial positioning, which is crucial in neurons for supplying energy to far-reaching neuronal processes, was another pathway down-regulated significantly in DET1^R26W neurons (Fig. 4 E and F). For example, Syntaphilin, an anchor protein involved in mitochondrial positioning (31, 32) was less abundant in DET^R26W neurons compared with WT neurons. Negative regulation of the cell cycle and RNA/transcription-related processes were enriched in DET1^R26W neurons (Fig. 4 E and F), the former aligning with the slower rate of proliferation in these cultures. DET1^R26W neurons also had low expression of neuronal-specific beta-tubulins (TUBB2A, TUBB2B, TUBB4A), which are essential for axonal guidance and structural maintenance, as well as the multifunctional neuronal protein APOE, a genetic risk factor for Alzheimer’s disease (See SI Appendix, Fig. S4B, Cluster 9 to 10).

Investigating the transcription factors that were identified in both datasets revealed that Zeb1 and Zeb2 are regulated at both the protein and transcript levels (SI Appendix, Fig. S5A), while CEBPB—a known target of the COP1/DET1 complex—was regulated exclusively at the protein level, with no corresponding changes at the transcript level (SI Appendix, Fig. S5 C and D). Transcription factor enrichment analysis of differentially expressed genes based on previously annotated target genes from various sources identified c-JUN and ETV4, known targets of CRL4^COP1/DET1 (Fig. 4G). These findings suggest that the accumulation of CRL4^COP1/DET1 substrates contributes to the transcriptomic and proteomics changes in DET1^R26W iPSCs and neurons.

DET1^R26W Does Not Support CRL4^COP1/DET1 Ligase Activity.

We characterized DET1^R26W biochemically in human A375 melanoma cells, which express constitutively active BRAF^V600E as well as the components of CRL4^COP1/DET1 (SI Appendix, Fig. S6A). MEK-ERK signaling downstream of BRAF^V600E suppresses CRL4^COP1/DET1 ligase activity (11), resulting in ETV5 accumulation. However, ETV5 is degraded in the proteasome upon MEK inhibition with cobimetinib (SI Appendix, Fig. S6 B and C). DET1 KO A375 cells lack CRL4^COP1/DET1 ligase activity and express more ETV5 protein than their WT counterparts (SI Appendix, Fig. S6B). Reconstitution of DET1 KO A375 cells with WT DET1 (untagged or Flag-tagged) restored ETV5 degradation in response to cobimetinib, but Flag-tagged DET1^R26W did not (Fig. 5A). Global proteomics indicated a good correlation (Pearson r = 0.8) between the changes induced by DET1^R26W and cobimetinib (SI Appendix, Fig. S6D). These data are consistent with DET1^R26W disabling CRL4^COP1/DET1 ligase activity. Accordingly, DET1^R26W-expressing cells exhibited elevated expression of ETV1, ETV4, and ETV5 proteins, but ETV1, ETV4, and ETV5 mRNA expression was not increased owing to posttranslational stabilization of the transcription factors (SI Appendix, Fig. S6E).

Fig. 5.

DET1^R26W fails to degrade substrates and interact with DDB1. (A) Western blots of A375 DET1 KO cells stably reconstituted with DET1 variants, treated with cobimetinib (1 µM, 1 h). Results representative of two independent experiments. (B) Workflow for the identification of DET1-associated proteins. (C) Affinity-enrichment MS analysis quantifying proteins associated with WT DET1 versus DET1^R26W. Data collected from three biological replicates. (D) Western blots of DET1-containing complexes immunoprecipitated from reconstituted A375 DET1 KO cells. Cells treated with DMSO or cobimetinib (1 µM, 1 h). Results representative of two independent experiments. (E) Cryo-EM structure of the DDB1:DDA1:DET1 ternary complex at an average resolution of 2.9 Å. DDB (ivory). DDA1 (teal). DET1 (magenta). DET1 forms a partial beta propeller while two amino terminal helices project into the DDB1 cleft between BPA and BPC domains. (F) The Hbox motif of human and mouse DET1 with the patient variant highlighted red. (G) A cutaway view of the DET1 Hbox helix (residues 17 to 27) highlights interactions between key residues of DET1 and DDB1. Salient hydrogen bonds are indicated with dashed lines. Some foreground sections of DDB1 cartoons have been hidden for clarity. (H) An example of the density map quality at the region of interest where local resolution of the maps approaches 2.5 Å. Here, key DET1:DDB1 interaction is mediated in part by the patient variant position R26.

DET1^R26W Fails to Interact with DDB1.

To determine whether DET1^R26W is incorporated into CRL4^COP1/DET1 complexes, we affinity purified DET1 complexes from the reconstituted A375 DET1 KO cells (Fig. 5B). WT DET1 co-immunoprecipitated with DDB1, DDA1, COP1, CUL4A, CUL4B, RBX1, UBE2E1, UBE2E2, UBE2E3, NEDD8, and members of the COP9 signalosome (CSN), but DET1^R26W did not (Fig. 5 C and D and SI Appendix, Fig. S6F). Thus, DET1^R26W is not recruited to CRL4^COP1/DET1 complexes.

DET1 is thought to bridge the DDB1–CUL4–RBX1 complex and the substrate adaptor COP1 (33). To understand how the patient variant R26W impacts the DET1:DDB1 interaction, we determined an experimental cryo-electron microscopy (cryo-EM) structure of the DDB1:DDA1:DET1 complex (SI Appendix, Fig. S7). Density was robustly defined for DDB1 propeller domains BPA and BPC, much of DDA1, and the interface-proximal regions of DET1. The BPB domain in DDB1 and some distal portions of DET1 were indicated by weaker density, but have been excluded from the final model. We generated an AlphaFold2-multimer model that confirmed our model and extended the interpretation of less contiguous DET1 density regions in the EM data (Fig. 5E and SI Appendix, Fig. S7F).

Many DDB1 partner proteins form WD-repeat beta propeller folds whose top surface faces DDB1, forming primary interactions via a sequence-degenerate extended helical Hbox motif (34). DET1 features a similar helical motif, but forms a broken beta propeller, maintaining only five of the typical seven blades. The full ring is interrupted by a helical-loop segment as a kind of wedge (SI Appendix, Fig. S7G). The broader curvature partial propeller is reminiscent of the DCAF15 structure. However, the DET1 C-terminal “wedge” is instead a partner-protein binding pocket in DCAF15 (35–37). This difference has potential ramifications for how DET1 may occlude or regulate its substrate interactions. A recent cryo-EM structure of the DDB1:DDA1:DET1 complex with bound UBE2E2 shows how DET1 residues 257 to 328, which are disordered in our particle classes, form a helical cup motif mediating interaction to UBE2E2 (38). Given this finding, the patient mutation at residue 26 is unlikely to impact the interaction between DET1 and UBE2E2. Modeling and mutagenesis also suggest that our propeller-interrupting wedge region (residues 381 to 582) interacts with the WD40 domains of a COP1 dimer (38). Our shared observation of several additional classes of particles implies dynamic flexibility within DET1 and DDA1.

DET1 residues 17 to 45 form the Hbox motif that extends into the DDB1 BPA:BPC cleft (Fig. 5F). Residue R26 is deeply buried in the interaction cavity and forms hydrogen bonds to the backbone carbonyls of DDB1 residues R722 and V360 (Fig. 5 G and H). A substitution to a bulky tryptophan would be poorly tolerated in the confined space of this pocket. Moreover, DDB1 M910 and L912 form significant VanDerWaals packing against the DET1 helix, and Y913 additionally makes a hydrogen bond with DET1 R25 (Fig. 5 G and H). To further confirm the interaction between DET1 and DDB1 through the Hbox helix, we mutated DDB1 R722/K723 and L912/Y913, which contact the α-helix motif of DET1. DDB1^R722A/K723A was expressed poorly in 293T cells (SI Appendix, Fig. S6 G–I). DDB1^L912N/Y913F and WT DDB1 were expressed at comparable levels, and both proteins co-immunoprecipitated with endogenous DDA1 (SI Appendix, Fig. S6G). However, only WT DDB1 co-immunoprecipitated with Flag-tagged DET1 (SI Appendix, Fig. S6 H and I), indicating the importance of DDB1 L912 and/or Y913 to the DDB1:DET1 interaction. Mutating the broader DDB1 MALY motif (residues 910 to 913), which interacts with various DCAFs (39), also prevented DET1 binding (SI Appendix, Fig. S6 H and I). In summary, our findings suggest that the DET1^R26W patient variant disrupts the Hbox motif of DET1 and prevents its interaction with DDB1, thereby rendering the CRL4^COP1/DET1 complex inactive.

Discussion

We report the critical role of DET1 in human and mouse neurodevelopment. DET1 binding partners, DDB1 and COP1, are also essential for normal mouse brain development (13, 40), consistent with DET1 mediating its effects as part of the CRL4^COP1/DET1 E3 ligase complex. The DET1 R26W patient mutant, which perturbs the differentiation of human iPSCs into neurons, appears to disrupt the CRL4^COP1/DET1 complex because it cannot interact with DDB1 normally. However, we suspect DET1 p.R26W is a hypomorphic mutation because efforts to generate DET1 knockout iPSCs were unsuccessful, the implication being that iPSCs completely lacking DET1 are not viable.

The relative contributions of the DET1 p.R26W and COMMD4 p.L41R homozygous mutations to disease is unclear. COMMD4 L41R is reported to be less stable than WT COMMD4, but still able to form the commander complex important for endosomal cargo retrieval and recycling (15). Features of the disorder resemble those in RSS patients with mutations affecting distinct components of the commander complex (19–24), consistent with the deleterious nature of the Commd4 p.L41R mutation. However, the severity of the disorder when taken together with the deleterious effects of DET1 deficiency in mice or human iPSCs, strongly suggests that the linked DET1 mutation also contributes to disease.

Microcephaly was observed in two of the siblings with DET1 p.R26W and COMMD4 p.L41R homozygous mutations (15). Reduced cell proliferation and increased cell death contribute to the pathogenesis of microcephaly (41), and both phenotypes were observed when DET1^R26W iPSCs were differentiated into neurons. Another mechanism contributing to microcephaly and abnormal brain development is the dysregulation of microglial cells (42, 43). Cop1-deficient microglia exhibit increased c/EBPβ-dependent transcriptional programs that promote inflammation and neuronal cell death (28). Det1-deficient microglia were also more neurotoxic than WT microglia, indicating that the deleterious effects of the DET1 p.R26W mutation might stem from the dysregulation of multiple cell types.

The impaired differentiation of DET1^R26W iPSCs into neurons might reflect their aberrant expression of both positive and negative regulators of neurogenesis. Elevated expression of ZEB1, perhaps driven by an abundance of ETV4 (44, 45), may repress neuronal differentiation (46). Abnormally high expression of SOX6 may also inhibit neuronal differentiation (47). Surviving DET1^R26W neurons exhibited reduced mitochondrial gene expression and function, which is a feature of many neurodegenerative diseases (48). Decreased expression of mitochondrial respiratory chain complexes and impaired ATP production probably contribute to the poor viability of DET1^R26W neurons. Collectively, our data indicate that dysregulation of the transcription factors targeted by CRL4^COP1/DET1 elicits complex alterations that impair neurodevelopment. The broader relevance of the transcriptomic and proteomic changes associated with DET1 deficiency to human physiology requires further exploration, including studies to distinguish the primary effects of DET1 inactivation from downstream consequences.

Materials and Methods

Patient genome sequencing was performed by their local genome diagnostic laboratory and processed according to standardized diagnostic pipelines (15). Written informed consent was obtained from the parents and the family was enrolled under Sidra Medicine IRB approved protocol (IRB 1636872).

All mouse studies complied with relevant ethics regulations and were approved by the Genentech Institutional Animal Care and Use Committee in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC)-accredited facility in accordance with the Guide for the Care and Use of Laboratory Animals and applicable laws and regulations. Det1^+/− and Det1^fl/fl mice were generated using gene targeted C57BL/6 embryonic stem cells described previously (11). The latter were bred to Nestin.cre mice (25). Detailed methods and reagents are provided in SI Appendix, Materials and Methods.

Supplementary Material

Appendix 01 (PDF)

Data, Materials, and Software Availability

RNA-Seq, proteomics, and cryo-EM structure data have been deposited in GEO, MassIVE, and PDB [GSE271764 (49), GSE271767 (50), MSV000095151 (51), and 9DHD (52)]. All other data are included in the manuscript and/or SI Appendix. Some study data are available; unique reagents generated in this study may be requested through Genentech’s MTA program.