Cullin-RING Ubiquitin Ligases in Neurodevelopment and Neurodevelopmental Disorders

1. Introduction

Ubiquitination is dynamic post-translational modifications that play critical roles in maintaining cellular function and homeostasis by modulating protein stability, activity, and localization [1,2]. Ubiquitination refers to the covalent attachment of ubiquitin molecules to a substrate protein. This process is carried out through a sequential enzymatic cascade involving three key enzymes. First, E1 ubiquitin-activating enzyme activates ubiquitin in an ATP-dependent manner, forming a high-energy thioester bond. Next, the activated ubiquitin is transferred to E2 ubiquitin-conjugating enzyme. Finally, E3 ubiquitin ligase facilitates the transfer of ubiquitin from the E2 enzyme to the target protein, determining the substrate specificity [3]. The transfer of ubiquitin to the substrate is orchestrated by either a RING (really interesting new gene) finger domain, a HECT (homologous to the E6-AP carboxyl terminus) domain, or a RBR (RING-between-RING) domain, leading to the classification of ubiquitin ligases into RING-type, HECT-type or BRB-type based on the domain they possess [4,5]. Ubiquitin ligases function as either single-protein entities which contain both E2- and substrate-binding domains, or multi-subunit complexes which harbor a protein responsible for binding E2, and a substrate receptor protein.

The largest subclass of ubiquitin ligases is Cullin-RING ubiquitin ligase (CRL) in which Cullin protein serves as a scaffold to bind both RING domain protein ROC1/2 (also called RBX1/2), and substrate receptor protein directly or indirectly through adaptor protein [6,7]. Mammalian cells contain eight (CUL1, 2, 3, 4A, 4B, 5, 7, 9) cullin proteins, each of which builds their specific ubiquitin ligase complexes[6,7]. For example, CUL1-based ubiquitin ligase uses F-box proteins as a substrate receptor which is linked to CUL1 via SKP1 adaptor protein [8], while CUL4A or CUL4B employs DDB1 adaptor protein which recruits DCAF substrate receptor proteins [9,10] (Figure 1). In contrast to CUL1, CUL2, CUL3, CUL4A, CUL4B, and CUL5, recent structural investigations have demonstrated that CUL7 and CUL9 directly interact with their substrates, with the ubiquitylation module being supplied by CUL1 in the case of CUL7 [11]. CUL9, on the other hand, employs its intrinsic RBR domain to help catalyze ubiquitylation [12]. All Cullin proteins are thought to use ROC1 as a RING domain protein, except for CUL5 which uses ROC2 [13] (Figure 1). In addition to these Cullin-based complexes, APC/C (anaphase promoting complex/cyclosome) ubiquitin ligase constitutes a similar structure in which APC2 serves as a main scaffold, APC11 as a RING domain protein, and CDC20 or CDH1 as substrate receptor [14] (Figure 1). Due to this structural similarity and its significant functions in neurodevelopment [15], we include and discuss APC/C ubiquitin ligase in this review.

In neurodevelopment, ubiquitination has been demonstrated to play a significant role in neurodevelopmental processes such as neural stem cell proliferation/differentiation, axonal/dendritic growth, migration, synaptogenesis, and synaptic function (Figure 2). Thus, dysregulation of these processes has been linked to several neurodevelopmental disorders (NDDs), including autism spectrum disorder (ASD), intellectual disability (ID), and attention-deficit/hyperactivity disorder (ADHD). Focusing on CRLs, this review aims to explore the role of ubiquitination in neurodevelopment and their implications for NDDs, providing insights into their potential as therapeutic targets.

2. Human Genetics of CRL in NDD Patients

Exome sequencing and chromosomal microarray testing of NDD patients identified de novo heterozygous loss-of-function mutations in CUL3 on chromosome 2 [16,17], homozygous or compound heterozygous loss-of-function mutations in CUL7 on chromosome 6, and hemizygous loss-of-function mutations in CUL4B on X chromosome [18,19] in NDD patients. Furthermore, variants of genes encoding CRL1 substrate receptors FBXO10 [20], FBXO11[21,22], FBXO28 [22,23], FBXO31 [24], FBXO47 [25], FBXL3 [26], FBXL4 [27,28], FBXL10 (also known as KDM2B) [29], β-TrCP1 (also known as FBXW1) [30,31,32], β-TrCP2 (also known as FBXW11) [31], FBXW7 [33,34,35], as well as CRL3 substrate receptors KLHL15, KLHL17 (also known as actinfilin) [36], KLHL20 [37], KCTD7 [38], KCTD13 [39], and CRL4 substrate receptors DCAF1 (also known as VprBP) [40], DCAF14 (also known as PHIP, BRWD2 and RepID) [41,42], COP1 [43], and CRBN [44,45], and APC/C substrate receptor CDH1 [46] have been implicated in NDD patients. These findings suggest the involvement of these genes in NDDs, thereby prompting further research to elucidate how these mutations contribute to the pathogenesis of NDDs.

3. Behavioral Phenotypes of Mice with CRL Mutations

Mouse models have been developed to investigate the mechanisms underlying the pathogenesis of NDDs [47,48,49]. To assess the suitability of these models for NDD research, various behavioral evaluations have been designed and conducted [50,51]. For instance, communication deficits are quantified through open field and three-chamber social interaction tests, while impairments in learning and memory are assessed using a range of maze and contextual fear conditioning paradigms [50,51]. Additionally, increased distance traveled serves as an indicator of hyperactivity, and anxiety is primarily evaluated through open field and elevated plus-maze tests [50,51]. Several CRL mutant mouse models have been shown to exhibit these NDD-associated behaviors, as summarized in Table 1. These findings suggest that mutations in these genes are most likely causal, rather than merely correlative, in the etiology of NDDs.

4. CRLs in Neural Stem Cell Proliferation and Differentiation

In the following sections, we will provide an overview of the current understanding of the molecular mechanisms underlying neurodevelopment and NDDs, as derived from studies on CRLs.

Neural stem cells (NSCs) are multipotent progenitor cells that give rise to neurons, astrocytes, and oligodendrocytes during brain development [62,63]. NSCs first give rise to radial glial cells (RGCs), which serve dual roles as neural progenitors and as scaffolds that guide migrating neurons. RGCs divide asymmetrically, producing either neurons directly or intermediate progenitors (IPs) as an intermediary step in neurogenesis. IPs undergo a limited number of divisions, amplifying the production of neurons to meet developmental demands. Mature neurons are generated from the differentiation of IPs or directly from RGCs in a highly regulated process. In the early stages of development, RGCs primarily produce neurons directly; however, as development progresses, they increasingly generate IPs to enhance neuronal numbers. This sequential progression from NSCs to RGCs, then to IPs, and finally to mature neurons ensures the proper formation and organization of the nervous system. By maintaining the progenitor pool, NSCs and RGCs regulate the balance between self-renewal and differentiation, while IPs contribute to amplifying neurogenesis without depleting the pool of progenitors. This hierarchical relationship is essential for the precise development of neural structures and circuits [64,65,66]. The tightly regulated processes of NSC proliferation and differentiation are fundamental for proper neurodevelopment and rely heavily on the dynamic control of protein expression and degradation. CRLs play a pivotal role in these processes by regulating key signaling pathways and transcription factors (Figure 3).

NSC proliferation is governed by the precise regulation of cell cycle progression, which is controlled by cyclins, cyclin-dependent kinases (CDKs), and their inhibitors. Ubiquitination ensures the timely degradation of these cell cycle regulators, thus maintaining proper cell cycle transitions. For example, APC/C ubiquitin ligase recognizes mitotic regulators such as cyclin B through CDH1 substrate receptor (APC/C^CDH1), and targets them for ubiquitination, promoting the progression of mitosis [67,68]. In Cdh1-deficient neural stem cells, aberrantly elevated cyclin B-associated CDK activity induces DNA damage-mediated apoptotic cell death [68], potentially by disrupting the precise regulation of S phase timing, thereby eliciting DNA replication stress [67]. Another key regulator is the CUL1-based ubiquitin ligase complex (CRL1, also known as SKP1-CUL1-F-box SCF ubiquitin ligase), which ubiquitinates cyclin E for degradation with the use of FBXW7 substrate receptor (CRL1^FBXW7). ASPM suppresses this SCF^FBXW7-mediated ubiquitination, thereby stabilizing cyclin E. This stabilization facilitates the shortening of the G1 phase in the cell cycle, promoting NSC proliferation [69].

NSC differentiation requires the fine-tuned suppression of self-renewal programs and the activation of lineage-specific transcriptional networks. Ubiquitination modulates this process by targeting signaling pathway components and transcription factors that drive differentiation. For instance, the Notch signaling pathway, which maintains NSC and RGC self-renewal, is also regulated by the ubiquitin ligase CRL1^FBXW7, which ubiquitinates and targets Notch intracellular domain (NICD) for proteasomal degradation [70]. This degradation facilitates the transition from NSC maintenance to astrocytic differentiation [70]. APC/C also plays a critical role in NSC differentiation. APC/C^CDH1 ubiquitinates and facilitates the degradation of the CDK activator CDC25A and SKP2, a substrate receptor of SCF ubiquitin ligase that targets CDK inhibitors for degradation [68]. This process prevents cell cycle re-entry, thereby promoting neuronal differentiation [68]. Similarly, the Wnt/β-catenin signaling pathway, a critical regulator of NSC maintenance, is modulated by ubiquitination. The E3 ligase CRL1^β-TrCP ubiquitinates β-catenin, leading to its degradation and attenuation of Wnt signaling [71,72].

In addition to cell cycle and signaling pathways, ubiquitination directly regulates transcription factors critical for NSC identity and differentiation. SOX2, a core transcription factor that maintains NSC pluripotency, is subject to ubiquitination. In human pluripotent stem cell-derived NSC, CUL4A-based Cullin-RING ubiquitin ligase with DET1-COP1 heterodimer as a substrate receptor (CRL4A^DET1-COP1) ubiquitinates SOX2 for degradation, leading to neural differentiation [73].

Dysregulation of ubiquitination in NSC proliferation and differentiation has profound implications for NDDs. Mutations in genes encoding CRL-related proteins (please refer to section 2) and FBXW7 regulator ASPM [74,75] are identified in NDD patients and these could be caused by aberrant NSC proliferation of differentiation.

5. CRLs in Neuronal Polarization

Immature neurons extend processes known as neurites, which subsequently differentiate into axon, the intracellular signal-transmitting structure, or dendrites, the primary signal-receiving components of neurons. The developmental process that defines this distinction is known as neuronal polarization, signifying the establishment of distinct subcellular compartments to enable precise signal transmission and reception [76]. Neuronal polarization is initiated by extracellular symmetry-breaking cues, followed by cytoskeletal rearrangements [77], with actin filaments (F-actin) guiding axon/dendrite extension and microtubules playing an essential role in axon stabilization [78,79]. By modulating protein functions, CRLs ensure the spatial and temporal coordination of neuronal polarization (Figure 4).

The protein kinase AKT facilitates axon formation by phosphorylating GSK-3β to negate its function, leading to the activation of microtubule-binding proteins and subsequent stabilization of microtubules [80]. AKT activity is negatively regulated by PTEN [81], necessitating PTEN inactivation within the axon. Notably, the HECT-type ubiquitin ligase NEDD4 is reported to facilitate PTEN degradation through ubiquitination, thereby promoting neurite outgrowth [82] and axon branching [83]. However, the decrease in ubiquitination and accumulation of PTEN were not detected in NEDD4-deficient neurons [84], indicating that the E3 ubiquitin ligase for PTEN in developing axon awaits to be revealed. Although not in neuronal cells, CRL4B^DCAF13 [85,86] and SCF^FBXO22 [87] are reported to ubiquitinate PTEN for degradation, providing the possibility that Cullin-RING ubiquitin ligases are involved in axon formation through PTEN regulation.

The activity of a small G protein RhoA is critical for dendrite formation through promoting actin arc formation, which prevents microtubule protrusion and axon growth [88]. RhoA is stabilized in dendrites by the ubiquitin ligase APC/C^CDH1 that ubiquitinates the RhoA targeting HECT-type ubiquitin ligase SMURF1 for degradation [89]. In the axon, the activity of RhoA is downregulated by a mechanism in which RhoA activator PDZ-RhoGEF is targeted for degradation by the ubiquitin ligase CUL3 with KLHL20 serving as a substrate receptor (CRL3^KLHL20) [90]. A symmetry-breaking signaling by BDNF induces phosphorylation of PDZ-RhoGEF, potentiating its ubiquitination and subsequent degradation [90], facilitating the neuronal polarization.

The interplay between F-actin and microtubules, mediated by coupling proteins, is also implicated in axon and dendrite formation [91]. One such coupling protein, DCX, is subject to negative regulation by CRL3^KLHL15 [92], CRL4A^CRBN [93], and CRL4B^CRBN [93]. The ubiquitination and subsequent proteasomal degradation of DCX by these CRLs have been shown to attenuate axonal and dendritic complexity and length [92,93].

An additional regulatory layer in the process of neuronal polarization is provided by transcriptional mechanisms, whose activities are also modulated by ubiquitination. For instance, APC/C^CDH1 facilitates the ubiquitination of SnoN [94], a transcription factor that promotes the expression of positive regulators of axon growth, and ID2 [95], an inhibitor of axon growth repressors, thereby suppressing axon extension. We have also demonstrated that the ubiquitin ligase CUL4B regulates NGF-induced neurite extension via transcriptional modulation [96,97]. CUL4B mediates the ubiquitination and subsequent degradation of WDR5 [96], a key component of the histone H3 lysine 4 methyltransferase transcriptional complex [98], in rat neuroblastoma PC12 cells, downregulating the neuronal gene expression [99]. These findings underscore the critical role of transcriptional regulation in neuronal process formation.

Dysregulation of ubiquitination in neuronal polarization has also profound implications for NDDs. Mutations in genes encoding CRL-related proteins (please refer to section 2), as well as DCX [100,101], SMURF1 [36,102], and PTEN [103,104] are identified in NDD patients and these could be caused by aberrant neuronal polarization.

6. CRLs in Neuronal Migration

NSCs are generated and remain adjacent to the brain ventricle, while differentiating neurons migrate towards the cortical surface, with later-differentiated neurons passing by earlier-generated neurons during corticogenesis [66,105]. This inside-out layering of the cortex is guided by various mechanisms, some of which are regulated by CRLs (Figure 5).

The most well-characterized regulator of neuronal migration is Reelin/DAB1 signaling [106]. Reelin binds to its receptors and activates tyrosine kinases, leading to the phosphorylation of DAB1 [107]. Phosphorylated DAB1 then interacts with several proteins, including LIS1, a regulator of microtubule-based dynein-dynactin motor proteins, which facilitates cytoskeletal reorganization essential for migration [108]. CUL5 with SOCS proteins as substrate receptors (CRL5^SOCS) targets phosphorylated DAB1 for ubiquitination and degradation, thus preventing DAB1 hyperactivation [109]. CUL5 knockdown-induced accumulation of phosphorylated DAB1 results in excessive migration and abnormal superficial positioning [109]. Furthermore, CUL4B was shown to bind to LIS1 [110], suggesting that CRL4B may also play a role in regulating neuronal migration through Reelin signaling.

In contrast, heterozygous deletion of Cul3, a condition that mirrors NDD in humans, leads to impaired migration of embryonic neurons and cortical lamination abnormalities in mice [53]. Proteomic analysis has identified PLS3, an actin-bundling protein, as a critical substrate of CUL3, and the accumulation of PLS3 caused by Cul3 haploinsufficiency disrupts actin filament organization and results in abnormal adhesion with impaired migration [53]. The substrate receptor responsible for PLS3 recognition remains unidentified.

The generation of the cerebellum and hippocampus also depends on proper neuronal migration [111,112]. FBXO41, a neuron-specific substrate receptor of CRL1 ubiquitin ligase, has been shown to promote neuronal migration in the cerebellum (from the external to the internal granule layer) [113], and hippocampus (from the hilus to the granule cell layer) [114]. It localizes to the centrosome and disassembles primary cilia, an antenna-like structure that receives several signaling ligands [115]. However, the involvement of primary cilia and the substrates of SCF^FBXO41 in neuronal migration remain unclear.

In addition to CUL3 and CUL4B, mutations in genes encoding Reelin [116,117], DAB1 [118], and LIS1 [119] have been identified in patients with NDDs, indicating that defects in neuronal migration may, at least in part, contribute to the pathogenesis of these disorders.

7. CRLs in Synaptogenesis and Synaptic Function

Elongating axons extend towards the dendrites of target neurons, where synapses are established [120,121]. At the presynaptic terminal, neurons secrete neurotransmitters such as glutamate, dopamine, acetylcholine (ACh), and gamma-aminobutyric acid (GABA), which bind to their corresponding receptors in the specialized postsynaptic density (PSD) region of the postsynaptic neurons [122], thereby enabling communication and the transmission of information. CLRs play a significant role in synaptogenesis and synaptic function (Figure 6).

Presynaptic development is transcriptionally regulated by APC/C^CDC20, in which the ubiquitination and degradation of the transcription factor NeuroD2 leads to the downregulation of Complexin II expression, a negative regulator of presynaptic differentiation, thus promoting the formation of presynaptic structures [123]. Both the presynaptic active zones and PSDs are supported by scaffolding proteins, the abundance of which is frequently regulated by ubiquitin ligases [124]. For example, FBXL20 (also known as SCRAPPER), a distinctive F-box protein characterized by a C-terminal CAXX motif essential for membrane localization, localizes at the presynaptic terminal, where it targets RIM1, a Ca²⁺-sensing synaptic vesicle regulator, for ubiquitination and degradation, ensuring the proper release of glutamate [125].

Glutamate stimulates postsynaptic neurons by activating four types of ion channels -AMPA, kainate, NMDA, and GluD receptors- [126], and the G-protein-coupled metabotropic receptor (mGluR) [127]. Chronic synaptic activity triggers the degradation of the AMPA receptor, a mechanism critical for preventing toxic hyperactivation of postsynaptic neurons, known as excitotoxicity. APC/C^CDH1 plays a pivotal role in the downregulation of the AMPA receptor by targeting the GluA1 (also known as GluR1) subunit for ubiquitination and degradation in response to synaptic stimuli [128]. Periodic synaptic activation induces long-term potentiation (LTP) or long-term depression (LTD) mediated through glutamate receptors, which are the foundations of synaptic plasticity essential for learning and memory and processes frequently dysregulated in NDD patients [129]. APC/C^CDH1 is involved in mGluR-mediated LTD by synaptic stimulus-dependent ubiquitination of FMRP1, a negative regulator of LTD, leading to its degradation [130]. Accordingly, loss of CDH1 in excitatory neurons of the mouse brain impairs LTD, resulting in sustained cell surface expression of the AMPA receptor and defective degradation of FMRP1 [130]. The levels of the NMDA receptor are also regulated by CRL3^KCTD13 which targets GluN1 for ubiquitination and degradation [131]. In patients with epileptic seizures, the expression of KCTD13 is diminished, likely leading to heightened activation of excitatory synaptic transmission and an increased susceptibility to epilepsy [131].

Dysfunction of the dopamine system is implicated in the phenotypes associated with NDDs [132,133]. Consistently, mice with a heterozygous Cul3 deletion in DA neurons exhibit hyperactivity of DA neurons accompanied by the behavioral abnormalities such as increased locomotion, impaired working memory, and deficits in sensorimotor gating (please refer to Table 1), all of which were reversed by the forced inactivation of DA neuron [55]. The hyperexcitability of DA neurons was shown to be caused by the accumulation of HCN2 channels, a substrate of CUL3 [55]. These findings confirm the critical role of elevated DA activity in NDDs.

Potassium channels are essential for maintaining neuronal activity by regulating membrane potential [134]. The potassium channel Kv10.1 (also known as EAG1 or KCNH1), whose gene is mutated in NDDs [135,136], undergoes ubiquitination-mediated degradation, a process driven by CRL7^FBXW8 [137]. Consequently, overexpression of CUL7 diminishes potassium currents, whereas reduced CUL7 expression leads to enhanced potassium currents in non-neuronal cells that exogenously express Kv10.1 [137]. The extent to which these findings can be extrapolated to neuronal function requires further investigation.

The postsynaptic sites in dendrites often form specialized protruded structures known as spines [138]. KLHL17 has been shown to enlarge spines and facilitate synaptic activity. Since this function requires the BTB domain, which is essential for association with CUL3, CRL3^KLHL17 likely contributes to this process, although the specific substrate remains to be identified [57]. Dendrites are dynamic structures that must be maintained to ensure the proper functioning of the brain. One destabilizing factor of dendrites is ROCK2, whose levels are regulated by APC/C^CDH1 [61]. Therefore, loss of CDH1 leads to the accumulation of ROCK2, resulting in dendritic spine disruption and learning deficits, which can be rescued by a ROCK inhibitor [61].

8. Therapeutic Implications and Future Directions

The critical roles of Cullin-RING ubiquitin ligases (CRLs) in neurodevelopment and their involvement in NDDs, as summarized in Table 2, highlight the potential for therapeutic targeting. Given that mutations in genes encoding CRL-related proteins have been identified in NDD patients (as discussed in Section 2), modulating ubiquitination pathways represents a promising approach to mitigate disease phenotypes. One potential therapeutic strategy involves the use of small-molecule inhibitors to regulate CRL activity. For instance, MLN4924, a NEDD8-activating enzyme (NAE) inhibitor disrupts all CRL functions, leading to the stabilization of CRL substrates [139]. Given the role of CRLs in neuronal differentiation and synaptic function, controlled inhibition of specific CRLs may hold therapeutic potential for certain NDDs. However, systemic CRL inhibition may lead to broad cellular dysfunction, necessitating the development of highly selective inhibitors targeting specific CRL complexes or substrate receptors.

Gene-editing technologies, including CRISPR/Cas9 and base editing, provide potential avenues for correcting pathogenic mutations in CRL-associated genes. For example, CRISPR-mediated correction of loss-of-function mutations could restore normal ubiquitination processes, thereby ameliorating NDD-related phenotypes. Additionally, targeted modulation of gene expression using CRISPR activation (CRISPRa) or inhibition (CRISPRi) may help fine-tune the activity of specific CRL components without permanent genomic alterations [140]. A particularly intriguing target is the CRL1^FBXW7 and APC/C^CDH1 complexes, the best characterized CRLs in relation to tumorigenesis, which also regulates key neurodevelopmental factors. Pharmacological stabilization of FBXW7 or its substrates may provide therapeutic avenues for NDDs characterized by disrupted Notch or Wnt signaling. Modulating APC/C^CDH1 activity could influence synaptic function and plasticity, with potential implications for learning and memory disorders [130].

Not all of the CRL-related proteins discussed in this review have been identified as mutated, but the increasing sample sizes may reveal mutations in these genes. In relation, patient-specific genetic and molecular profiling can guide therapeutic interventions for NDDs. Whole-genome, transcriptome, and proteomic analyses could aid in identifying individual CRL-related mutations, enabling the selection of personalized therapeutic strategies. In addition, patient-derived induced pluripotent stem cells (iPSCs) offer an invaluable platform for modeling disease mechanisms and testing novel therapeutics in a personalized context [141].

Despite significant progress in understanding the role of CRLs in neurodevelopment, several key questions remain unanswered. Future research should focus on (1) identifying substrate specificity: Many CRL substrate receptors remain uncharacterized. Understanding their target specificity is essential for developing selective therapeutic interventions; (2) developing selective CRL modulators: current pharmacological tools lack specificity for individual CRL complexes, though CUL-level inhibitors are emerging, as exemplified by DI-1548 and DI-1859 for CUL3 [142], and 33-11 and KH-4-43 for CUL4 [143]. Advances in structure-based drug design may enable the development of highly selective CRL inhibitors or activators; (3) exploring non-degradative ubiquitination: while ubiquitination is often associated with protein degradation, non-degradative ubiquitin modifications also play critical roles in signaling pathways. Investigating these roles could reveal novel therapeutic targets.

9. Concluding Remarks

The ubiquitination pathway, particularly through the action of CRLs, plays a fundamental role in neurodevelopment and the pathophysiology of NDDs. Advances in molecular and pharmacological approaches have provided promising avenues for targeting ubiquitin signaling in therapeutic contexts. However, further research is needed to refine these strategies, ensuring both efficacy and safety in clinical applications. As our understanding of ubiquitin-mediated neurodevelopmental regulation continues to expand, new opportunities for treating NDDs are likely to emerge, bringing us closer to precision medicine-based interventions for these complex disorders.