Several more studied and a few lesser studied kinases have connections to neuronal development and disorders. While we will introduce the roles of members of the CDKL, NEK, and TTBK families in these processes in the next section, examples of other kinases from Table 1 with to that modulate ciliary function in the brain are included below. Table 2 serves as a quick summary of the ciliary kinases from Table 1 that have been linked to neurological issues and which are discussed in some detail herein.

Ciliary kinases regulate signaling pathways such as Hedgehog, Wnt, Hippo, and other pathways that are functionally linked to ciliogenesis and cancer. Many of the ciliary kinases discussed below are deregulated in tumors. Herein we will focus on a subset of ciliary kinases from Table 1 that are linked to cancer with the aim to promote efforts to understand their function. Table 2 provides a quick reference of the ciliary kinases from Table 1 with identified roles in cancer propagation and/or progression, all of which are highlighted in more detail in the following sections.

CDKL kinases are a family of five relatively underexplored serine/threonine human kinases: CDKL1, CDKL2, CDKL3, CDKL4, and CDKL5. This family has the highest sequence similarity to cyclin-dependent kinases (CDKs) (Canning et al., 2018; Ong et al., 2023). CDKLs contain a cyclin binding domain, although no cyclin-dependent functions have been ascribed. While little is known about CDKL1–4 regarding their function and role(s) in human biology, CDKL5 has been identified as a regulator of ciliogenesis and cilia length (Canning et al., 2018). Structural characterization of all CDKLs, except for CDKL4, has confirmed them to contain a conserved N-terminal kinase domain with variable C-termini (Canning et al., 2018). The C-termini of CDKL2 and CDKL3 were found to have an atypical αJ helix necessary for their catalytic functions, which is absent in CDKL1 and CDKL5 (Canning et al., 2018). Notably, for CDKL5, the C-terminus is involved in trafficking CDKL5 to its subcellular compartments during various developmental stages (Rusconi et al., 2008; Canning et al., 2018; Ong et al., 2023).

CDKL5, also known as serine/threonine kinase 9 (STK9), is the only member of the CDKL family with strong connections to ciliogenesis in humans. CDKL5 is ubiquitously expressed in human tissues, but is observed at higher levels in the hippocampus, cerebellar, striatum, and cortex regions of the CNS in humans and mice, consistent with its roles in dendritic spine growth, brain development, and excitatory synapse composition (Rusconi et al., 2008; Fagerberg et al., 2014; Canning et al., 2018; Ong et al., 2023). A study using RNAi targeting CDKL5, which caused reduced neurite growth and dendritic arborization in rat cortical neurons, confirmed that this kinase regulates neuronal morphogenesis (Chen et al., 2010).

CDKL5 localizes to the basal body of primary cilia, and when aberrantly overexpressed, can lead to jeopardized ciliogenesis (Canning et al., 2018). Hippocampal neurons from Cdkl5 deficient mice presented elongated primary cilia, however no changes in the levels of Wnt and Shh proteins were visualized by Western blot (Di Nardo et al., 2022). Missense mutations in the CDKL5 gene leads to a syndrome known as CDKL5-deficiency disorder (CDD). CDD is characterized by severe encephalopathy, noncanonical developmental physiology, and intellectual disability with intractable epilepsy onset from early age, reflective of a significant impact on brain function (Olson et al., 2019; Di Nardo et al., 2022; Castano et al., 2023; Ong et al., 2023).

To date, there have only been two high-quality chemical probes published by members of the Structural Genomics Consortium (SGC) for CDKL5 (Figure 3). SGC-CAF382-1 (B1 in the original publication) demonstrates a cell-free CDKL5 enzymatic IC₅₀ = 6.7 nM and engages with CDKL5 in cells with an IC₅₀ = 11 nM (Castano et al., 2023). SGC-CDKL5/GSK3-1 is a CDKL5 and GSK3 (GSK3⍺ and GSK3β) probe with a cell-free CDKL5 enzymatic IC₅₀ = 6.5 nM and a nearly equal in-cell CDKL5 target engagement IC₅₀ = 4.6 nM (Ong et al., 2023).

CDKL1, the only other CDKL kinase with connections to ciliogenesis, has only been studied in model systems (zebrafish and C. elegans). The human and C. elegans CDKL1 proteins share 77% identity, and both have similar phosphorylation regulatory sites: Thr¹⁴, Tyr¹⁵, and Thr¹⁶¹ in humans and Ser¹⁴, Tyr¹⁵, and Thr¹⁵⁹ in zebrafish (Hsu et al., 2011). CDKL1, however, has been shown to influence cilia length and Hedgehog signaling in zebrafish, a pathway that is also linked to primary cilia (Hsu et al., 2011; Bangs and Anderson, 2017; Park et al., 2021). In lieu of a high-quality chemical tool for CDKL1, genomics, siRNA, and other techniques have been used to begin to uncover the functions of this kinase. In C. elegans, CDKL1 localizes to the transition zone at the base of the cilium, in a CEP-290-dependent manner, and regulates the length of the growing cilia by interacting with IFT anterograde kinesin motor proteins involved in axoneme formation of the cilia (Li et al., 2016; Park et al., 2021). Due to the high sequence identity, known roles of orthologs, and strong connection between Hedgehog signaling and primary cilia, it is proposed that CDKL1 might have an underlying role in human ciliogenesis (Bangs and Anderson, 2017).

The NIMA-related kinase family comprises 11 relatively understudied kinases (NEK1–NEK11) that play roles in many important biological processes and have been linked to several diseases. Recent reviews discuss their biological and disease relevance and provide information about inhibitors (Fry et al., 2012; Meirelles et al., 2014; Wells et al., 2018; Pavan et al., 2021; Panchal and Evan Prince, 2023b; Nguyen et al., 2023). These and other reviews highlight that mutations in all NEK family members have been identified in different cancers and the functions of each NEK in cancer progression delineated (Moniz et al., 2011; Panchal and Evan Prince, 2023a). NEK1 mutations have also been identified as a genetic cause of amyotrophic lateral sclerosis (Kenna et al., 2016). Members of the NEK family play various roles in ciliogenesis and are implicated in distinct ciliopathies (Table 1). Statistical analysis has linked the evolution of the NEKs and centrioles, both of which are responsible for organization of microtubule development as well as the basal bodies of cilia (Quarmby and Mahjoub, 2005).

NEK1 is localized to the basal body region and centrosomes (Shalom et al., 2008). Nek1 mutant mice exhibited a range of defects including progressing PKD (Upadhya et al., 2000), which has a well-established tie to ciliary dysfunction (White and Quarmby, 2008). NEK1 serves as a coordinator between ciliogenesis and cell cycle progression, a role that is speculated to involve signaling between the nucleus and the primary cilium (White and Quarmby, 2008). Remarkably, both overexpression of NEK1 and the complete removal of NEK1 severely decreased the percentage of cells bearing a primary cilium, suggesting that ciliogenesis depends on tightly regulated NEK1 expression (Shalom et al., 2008). The authors proposed that this observation could be due to either NEK1 blocking the upregulation of ciliary proteins or due to it physically interfering with the formation of complexes at the centrosome (White and Quarmby, 2008). NEK1 has also been shown to bind KIF3A, a kinesin motor protein crucial for ciliogenesis, further underscoring its role in directly impacting IFT (Quarmby and Mahjoub, 2005; White and Quarmby, 2008).

While fewer papers have been written about its role in ciliogenesis, NEK2 has a confirmed function in cilia homeostasis. NEK2 phosphorylates C-Nap1, Cep68, and Rootletin at the beginning of mitosis, allowing for the formation of the mitotic spindle (Fry et al., 1998; Bahe et al., 2005). This kinase is required for centriole splitting and is involved in the HDAC6 pathway, which promotes tubulin deacetylation (Endicott et al., 2015; DeVaul et al., 2017). NEK2 has been specifically shown to phosphorylate KIF24, ensuring that cilia resorption occurs prior to mitosis, and increased levels of NEK2 facilitate cilia depolymerization (Mirvis et al., 2018). Conversely, reduction of NEK2 leads to centriole defects and deficiencies in cilia biogenesis (Endicott et al., 2015).

NEK4 is localized to the base of primary cilium in RPE cells, suggesting that its role may include cilia stabilization (Basei et al., 2015). This kinase interacts with RPGRIP1 and RPGRIP1L, two proteins associated with ciliopathies (Patnaik et al., 2015). In a study from Coene et al., NEK4 knockdown resulted in a decrease in ciliated cells. The authors hypothesized that the interaction between RPGRIP1/RPGRIP1L and NEK4 forms a scaffold for the assembly of cilium-related kinases and substrates (Coene et al., 2011).

NEK8 is involved in the regulation of ciliary physiology and associated human ciliopathies (Choi et al., 2013; Grampa et al., 2016; Claus et al., 2022). In a mouse model of juvenile cystic kidney disease, mice bearing mutations in Nek8 display defects in ciliary localization that were potentially causal in the emergence of nephronophthisis (Otto et al., 2008). In the same mouse model, mice were observed to have an abnormal interaction between Nek8 and the polycystin complex (Valkova et al., 2005). During the transition from the cell cycle stage to ciliogenesis, NEK8 is activated and then ultimately degraded (Zalli et al., 2012). Mutations in NEK8, including missense and loss-of-function mutations, affected the regulation of signaling in the Hippo pathway through its main effector, YAP (Grampa et al., 2016).

NEK9 has roles in cell division as studies link phospho-NEK9 with microtubule and centrosomal organization, cytokinesis, and centrosome maturation (Roig et al., 2002; Roig et al., 2005; Sdelci et al., 2012; Meirelles et al., 2014). NEK9 has been implicated in the pathologies of Perthes disease, upward gaze palsy, arthrogryposis, congenital contracture syndrome, fetal akinesia, skeletal ciliopathies, meningiomas, and nevus comedonicus (Casey et al., 2016; Levinsohn et al., 2016; Dunn et al., 2019; Abraham et al., 2022; Liu et al., 2022). Ciliogenesis is bidirectionally regulated by selective autophagic cellular processes, whereby ciliogenesis promotes autophagy and vice-versa. NEK9 has been shown to have a role in upregulating autophagy via its binding to ATG8 proteins, which regulate autophagy (Behrends et al., 2010). ATG8 proteins are regulated via a domain on NEK9 called LC3 interacting region (LIR). Negative regulators of autophagy in the ATG8 family include GABARAP and GABARAPL1. Finally, autophagy of NEK9 is required for cilia formation, as NEK9 regulates ciliogenesis by interacting with autophagy adaptor MYH9 and myosin IIA, a suppressor of ciliogenesis (Yamamoto et al., 2021).

NEK10 has a role in promoting optimal cilia length during post-mitotic cilia assembly via interactions with pericentriolar matrix protein 1 (PCM1). NEK10 also stimulates ciliary transport as well as ciliary number and structure (Chivukula et al., 2020). Removal of NEK10 led to a decline in ciliated cells and a decrease in NEK10-promoted cilia resorption (Al Mutairi et al., 2020). Loss-of-function mutations in NEK10 resulted in PCD in humans (Al Mutairi et al., 2020). NEK10 forms a trimeric complex with PCM1 and RIIβ. A protein complex that includes NEK10 is found at centriolar satellites, and the role of NEK10 in this complex makes it essential for ciliogenesis (Porpora et al., 2018).

According to the chemical probes portal and SGC databases there are no high-quality chemical probes targeting any member(s) of the NEK family. To date, there are several lead compounds, sourced from the literature, that represent candidates for optimization. Examples of potential tool molecules are summarized in Figure 4. BAY 61-3606 was originally reported as a SYK inhibitor (Yamamoto et al., 2003), but interesting biological activity led to a deeper dive into its kinase target profile, revealing that NEK1 was also a target, with a IC₅₀ of 159 nM (Lau et al., 2012). NEK2 is the most well-studied member of the NEK family, and, as such, there are several inhibitors available, including CRUK ICR compound (R)-21 with a NEK2 IC₅₀ = 22 nM (Innocenti et al., 2012; Wells et al., 2018). This compound shows good selectivity over PLK1, an often seen off-target of NEK2 inhibitors. Some additional NEK2 inhibitors can be found in a NEK-family review (Wells et al., 2018). Abbott compound 17 was originally identified as a MAP3K8 inhibitor (George et al., 2008), and subsequently was found to bind to NEK4 with an IC₅₀ = 50 nM (Metz et al., 2011). No further structure–activity relationship (SAR) studies on this scaffold for NEK4 have been reported. The BRAF inhibitor dabrafenib has been approved by the FDA for melanoma. Because dabrafenib showed activity against some cancer lines that differentiated it from other BRAF inhibitors, further kinase profiling was performed. These experiments demonstrated useful and consequential inhibition of CDK16 and NEK9 (Phadke et al., 2018). While some of these candidates bind to their associated NEK with high affinity, the modest kinome-wide selectivity of these compounds has thus far precluded them from being considered chemical probes (Arrowsmith et al., 2015). Through informed SAR campaigns, it should be possible to refine these leads into chemical probes, which may prove to be valuable tools in understanding the role of the NEKs in ciliary biology.

Tau tubulin kinases are a family of serine/threonine/tyrosine kinases that belong to the larger casein kinase superfamily (CMGC) (Sato et al., 2006; Nozal and Martinez, 2019; Bashore et al., 2023). There are two isoforms of TTBK, tau tubulin kinase 1 and tau tubulin kinase 2, and their kinase domains share 88% identity and 96% similarity. Their catalytic residues, K63 and D164 for TTBK1 and K50 and D141 for TTBK2, are similar; however, their non-catalytic domains are distinct from one another (whole sequence similarity: 63% and 35% identity) (Nozal and Martinez, 2019; Bashore et al., 2023). TTBK2 regulates the initiation of ciliogenesis by acting at the distal end of the mother centriole to remove capping protein CP110 from the mother centriole (Čajánek and Nigg, 2014) and facilitate the recruitment of IFT proteins needed for the subsequent assembly of the ciliary axonemal microtubules (Čajánek and Nigg, 2014; Bashore et al., 2023; Binó and Čajánek, 2023). TTBK2 is not only highly expressed in the granular cell layer, cerebellum Purkinje cells, hippocampus, midbrain, and substantia nigra regions of the brain, but also ubiquitously expressed in most human tissues (Houlden et al., 2007; Bashore et al., 2023). As TTBK2 is essential for initiating cilia assembly, it is not surprising that a pathogenic mutation of TTBK2 leads to a neurological disorder known as spinocerebellar ataxia 11 (SCA11), characterized by atrophy of the Purkinje cells of the cerebellum (Bowie et al., 2018; Bashore et al., 2023). This truncated gene product is lacking the C-terminus, which is involved in trafficking TTBK2 to the basal body of primary cilium and also disrupts its interaction with CEP164, a key substrate that leads to initiation of ciliogenesis (Bowie et al., 2018; Bashore et al., 2023). Defects in ciliary assembly, stability, and function are observed in mice carrying the mutated alleles (Bowie et al., 2018). Another study showed cerebellar degeneration, altered intracellular levels of calcium, and loss of VGLUT2⁺ synapses in Ttbk2 mutated mice (Bowie and Goetz, 2020).

TTBK1 expression is confined to hippocampal, cortical, and entorhinal cortex neurons (Nozal and Martinez, 2019; Bashore et al., 2023). In human pluripotent TTBK2 knockout stem cells, TTBK1 was able to compensate for the loss of TTBK2 and regulate the assembly of primary cilia during neural rosette formation (Binó and Čajánek, 2023). Binó et al. further showed that in a TTBK2 rescue experiment, TTBK1 activity and expression levels increase enough to compensate for the loss of TTBK2 during neural rosette formation (Binó and Čajánek, 2023). Although TTBK1 lacks the C-terminal CEP164-binding domain that directs TTBK2 to the mother centriole, TTBK1 can still phosphorylate the key substrates of ciliogenesis outside of the mother centriole (Binó and Čajánek, 2023). This study brings to light the first indication that TTBK1 is a regulatory kinase of ciliogenesis. Furthermore, TTBK1 is known to act on neuropathic proteins like tau at pathogenically relevant sites and is overexpressed in Alzheimer’s disease (AD) with single nucleotide polymorphisms associated with late-onset AD (Ikezu and Ikezu, 2014; Bashore et al., 2023). TTBK1 also co-localizes with other neuropathic proteins and neurofibrillary tangles associated with amyotrophic lateral sclerosis and frontotemporal dementia (Ikezu and Ikezu, 2014; Bashore et al., 2023).

Despite being understudied, recent efforts have been made towards developing inhibitors of TTBK1/2. Biogen, Bristol-Myers Squibb, and AstraZeneca have published co-crystal structures of TTBK1 bound inhibitors (AZ1, AZ2, BMS1, and BGN18, Figure 5) (Xue et al., 2013; Kiefer et al., 2014; Halkina et al., 2021; Bashore et al., 2023). In 2013, AstraZeneca reported TTBK1 inhibitors AZ1 and AZ2, with Kd values of 0.24 µM and 4.1 µM, respectively (Xue et al., 2013). Later, a group optimized this same scaffold to yield compound 29, with selectivity for TTBK1 over TTBK2 (TTBK1 IC₅₀ = 0.24 µM, TTBK2 IC₅₀ = 4.22 µM) (Figure 5) (Nozal et al., 2022). Compound 29 was brain penetrant and lowered levels of phosphorylated TDP-43 in vitro and in vivo in a TDP-43 transgenic mouse model (Nozal et al., 2022). In 2014, Bristol-Myers Squibb published BMS1 and co-crystallized it in complex with TTBK1 (Kiefer et al., 2014). This compound has cell-free IC₅₀ values for TTBK1 of 120 nM and for TTBK2 of 170 nM (Kiefer et al., 2014). In 2021, Biogen published a series of brain penetrant, azaindazole-based TTBK1/2 inhibitors: BGN8, BGN18, and BGN31 (Figure 5) (Halkina et al., 2021). Azaindazole BGN8 had a TTBK1 biochemical IC₅₀ = 60 nM and in-cell TTBK1 IC₅₀ = 571 nM (Halkina et al., 2021). This compound was further optimized to yield BGN18, which demonstrated TTBK1/2 biochemical IC₅₀ values of 13–18 nM and TTBK1 in-cell potency of 259 nM (Halkina et al., 2021). Finally, the campaign yielded an azaindole-based analog, BGN31, which is the most advanced inhibitor of the series (Halkina et al., 2021). This compound has single-digit nM TTBK1/2 biochemical IC₅₀ values and displays TTBK1 in-cell potency of 315 nM (Halkina et al., 2021). BGN31 was advanced to in vivo studies of tau phosphorylation, utilizing a mouse model of hypothermia and a developmental rat model, and was shown to reduce phosphorylation of tau at disease relevant sites (Halkina et al., 2021; Bashore et al., 2023).

The results of an academic effort by the SGC at the University of North Carolina to produce TTBK1/2 inhibitors was published in 2023 (Bashore et al., 2023). The indole scaffold of these inhibitors was discovered from an investigation of the off-target activity of a published Amgen NF-κB inhibitor, AMG28, which included inhibition of TTBK1/2 (Bashore et al., 2023). Compounds 9 and 10 were the most advanced leads from that campaign and these analogs demonstrated TTBK1/2 enzymatic IC₅₀ values of 384 nM and 175 nM, respectively, for compound 9 and TTBK1/2 enzymatic IC₅₀ values 579 nM and 258 nM, respectively, for compound 10 (Figure 5) (Bashore et al., 2023). Both compounds have been shown to inhibit ciliogenesis in human pluripotent stem cell-based models and phenocopy what is observed due to genetic editing of TTBK2 (Bashore et al., 2023; Binó and Čajánek, 2023). The available TTBK1/2 inhibitors in Figure 5 are mostly devoid of isoform selectivity, and many of the inhibitors lack overall kinome-wide selectivity data. The high sequence homology between the kinase domains of TTBK1 and TTBK2 suggests that selectivity between these kinases may be very difficult to achieve. To date, no potent and selective chemical probe for either TTBK1 or TTBK2 exists, but delivery of such a compound will facilitate deconvolution of the importance and roles of these kinases in ciliogenesis, canonical signaling pathways, and disease.