Src is a kinase that plays pivotal roles in many signaling processes (Yeatman, 2004). CSK (C-terminal Src kinase) phosphorylates Src at its C-terminal tyrosine (Y530 in human) (Avraham et al., 1995; Chong et al., 2004), inactivating Src, playing a key role in controlling tumorigenic properties of Src and other physiological processes.

Human CHK (CSK-homologous kinase) is a kinase composed of 527 amino acids that has highest homology with human CSK, sharing 53% amino acid identity. The genomic structure of CHK is identical to that of CSK as reflected in the organization of its exons (Hamaguchi, et al., 1994). CHK has also been reported by different research groups as HYL, MATK, CTK, LSK, NTK, and BATK (Bougeret et al., 2001). Similar to CSK, CHK consists of SH2, SH3 and kinase (SH1) catalytic domains.

However, there is evidence to indicate that CHK and CSK have different biological roles. For example, CHK expression in brain increases postnatally whereas the CSK expression decreases with age (Hamaguchi et al., 1996). While similar to CSK, CHK has been shown to phosphorylate and inactivate Src, it can also inhibit Src in a phosphorylation-independent mechanism, different from CSK (Chong et al., 2004).

Moreover, potential substrates such as β-synuclein have been identified for CHK, specifically (Ia, et al., 2011). Expression of CHK has been shown to be regulated epigenetically recently (Chüeh et al., 2021; Zhu et al., 2021). However, the physiological and pathological roles of CHK as well as the regulation of CHK are still largely unknown.

In this review, we summarize the current knowledge about CHK, including regulation, molecular targets of CHK, as well as its biological functions, especially in the development of cancer. We also discuss challenging tasks in this field and its promising future.

The human and murine CHK gene codes for four and five splicing mRNA isoforms, respectively. However, CHK proteins of only two molecular weights (52 kDa and 56 kDa) are produced in each species (Zhu et al., 2008) (Figure 1A). The 52 kDa CHK isoforms are predominantly expressed in mouse, whereas the 56 kDa isoforms are predominantly expressed in human (Chow et al., 1994a; Zhu et al., 2008). Besides the similar structure in SH1, SH2 and SH3 domains, both CSK and CHK lack a myristoylation signal, the autophosphorylation site (Y419), and the carboxyl-terminal tyrosine (Y530), all present in Src (family members) (Figure 1B).

On the other hand, although CHK protein is homologous to CSK, CHK has a unique N-terminus, and its SH3 and SH2 domains share only 30% and 59% amino acid identity approximately to that of CSK, respectively (Hirao et al., 1997). In addition, part of the structural basis for the functional difference between SH2 Domains of CHK and CSK has been determined, including Glu127 in CSK, and Ile167 in CHK (Ayrapetov et al., 2005). Therefore, CHK has unique structure that supports its potentially different biological roles from CSK.

For many years, expression of CHK is believed to be limited in the brain, and in most types of hematopoietic cells of bone marrow, spleen and thymus (except erythroid cells) (Hamaguchi et al., 1996), and weakly expressed in the testis (germ cells) (Chow et al., 1994b; Kaneko et al., 1995). Surprisingly, many years after, we found that CHK was also expressed in normal colon tissues (Zhu et al., 2008), and Clemmons lab found that CHK was also expressed in smooth muscle cells (Radhakrishnan et al., 2011).

The restricted expression of CHK to certain tissues, in contrast to the ubiquitous expression of CSK (Okada et al., 1991), suggests specific roles of CHK in certain tissues. Besides, expression of CHK and CSK may differ developmentally, or with the same cytokine stimulation: in the developing mouse brain, expression of CHK increases postnatally, while the expression of CSK decreases with age (Brinkley et al., 1995); in human monocytes, IL-4 and IL-13, but not IFN-γ, induce CHK, in contrast to CSK (Musso et al., 1994). The different expression of CHK in comparison to CSK also suggests different roles CHK may play.

CHK knock-out mice appeared to be normal (Hamaguchi et al., 1996), in contrast to the CSK knock-out mice, which is embryonic lethal due to defects in neural tube formation (Imamoto and Soriano, 1993; Nada et al., 1993). CHK deficient mice also showed normal hematopoiesis. This includes blood counts, white cell differential counts, bleeding tendency, the sizes of spleen and thymus, the distribution patterns of hematopoietic stem cells, monocytes/macrophages, B-cells, and T-cells, the colony formation of bone marrow and spleen cells, as well as megakaryocyte counts in bone marrow (Hamaguchi et al., 1996; Samokhvalov et al., 1997). The absence of severe abnormalities in phenotypes in CHK knock-out mice is likely due to the compensatory effects of CSK, which is ubiquitously expressed. However, a more extensive study found that stimulation of the hematopoietic cells of these mice with IL-7, or injection of these mice with antigen (TNP-ovalbumin), led to physiologic responses, very different from that of CHK+/+ mice, suggesting that CHK may be involved in certain immune responses that CSK is not capable of (Lee et al., 2006).

CHK is often recruited to the plasma membrane via binding to ErbB2, c-Kit and TrkA.

In breast cancer cells, CHK has been shown to be associated with ErbB-2 (via CHK SH2 domain) upon heregulin stimulation (Zrihan-Licht et al., 1997). This leads to the suppression of Src kinase activity by CHK, followed by the attenuation of the activated receptor signaling and oncogenic properties (Zrihan-Licht et al., 1997; Bougeret et al., 2001; Kim et al., 2002).

In megakaryocytic cells, CHK is translocated to the plasma membrane by association of its SH2 domain to autophosphorylated Tyr^568/570 in c-Kit upon stimulation by stem cell factor/kit ligand (SCF/KL) (Price et al., 1997). By associating with c-Kit, CHK is likely to suppress activity of SFK (which also binds activated c-Kit), and the signaling pathways involving PI3K, PLCγ-1, and ras-GAP (Jhun et al., 1995). c-Kit signaling is important for hematopoiesis, and some other functions (Lennartsson and Rönnstrand, 2012). In the megakaryocytic cell line Dami, elevated CHK expression is able to suppress activation of Lyn kinase and VLA5-mediated Dami cell spreading that is dependent of Lyn activation (Hirao et al., 1998).

In neuronal cells, CHK was found to be involved in TrkA signaling. CHK, via its SH2 domain, is associated with phosphorylated TrkA receptors (Tyr-785) upon NGF stimulation. CHK overexpression activates the mitogen-activated protein kinase (MAPK) pathway upon NGF stimulation, and contributes to the neurite outgrowth of the cells (Yamashita et al., 1999). Further studies have shown that CHK induces formation of a complex containing SHP-2 and Grb2, leading to the activation of Ras and Raf/MEK/MAPK pathway. Interestingly, activation MAPK by CHK was independent of presence of NGF or the inhibition of Src kinase by CHK (Zagozdzon et al., 2006).

In colon epithelial cells, CHK colocalizes with Src at perinuclear region (Zhu et al., 2008). There is evidence to suggest that newly synthesized Src is autophosphorylated first, followed by binding to CHK and thus Src becomes inactive at perinuclear region, where it is associated with cytoskeleton structure (Zhu et al., 2008; Figure 2). Inactive perinuclear Src transits to the plasma membrane through the cytoskeleton structure and loses its interaction with CHK and becomes active. Active Src induces phosphorylation of Cbp in the membrane, and phosphorylated Cbp recruits CSK to the membrane in the proximity of Src (Kawabuchi et al., 2000). Src becomes phosphorylated by CSK at Y530 and stays inactive. Under appropriate conditions, PTP1B etc., starts to dephosphorylate Src at Y530 and to activate Src (Zhu et al., 2008). Active Src then phosphorylates its substrates and initiates positive signaling pathways. Activated Src also triggers activation-dependent ubiquitination and subsequent degradation of Src as negative feedback (Oda et al., 1999; Pan et al., 2011). In colon cancer cells, the loss of CHK expression allows Src to be active at perinuclear region, leading to the tumorigenicity (Zhu et al., 2008).

Besides its perinuclear localization, we have also shown nucleus localization of CHK in colon cells (Zhu et al., 2008; Figure 2B). This has confirmed earlier report that overexpression of CHK in COS-1 cells showed nuclear localization and growth inhibition of CHK (Yamaguchi et al., 2001). In human immature myeloid KMT-2 cells, Lyn and overexpressed CHK are associated with mitotic chromosome scaffolds and spindles. CHK overexpression induced a decrease in Lyn activation and phosphorylation of some associated proteins. CHK appears to participate in metaphase chromosome dynamics, since CHK overexpression caused aberrant chromosome movement. Probably due to insufficient formation of mitotic spindles, this leads to multinucleation, causing perturbation of normal cell division process, eventually leading to growth inhibition (Yamaguchi et al., 2001). Expression of CHK in the nucleus prolonged S phase of the cell cycle along with the nuclear multi-lobulation, suggesting that S phase may also be involved in the growth inhibition (Nakayama and Yamaguchi, 2005). Further experiments have shown tyrosine phosphorylation of a variety of proteins in the nucleus at the tyrosine residues upon CHK expression, suggesting there are probably potential substrates of CHK in the nucleus yet to be identified, besides SFKs (Nakayama et al., 2006). Structurally, N-terminal unique domain of CHK appears to be involved in the enhanced protein tyrosine phosphorylation in the nucleus and the induction of the multi-lobulation (Nakayama and Yamaguchi, 2005; Nakayama et al., 2006).

In addition, CHK could also relocate to cytoskeleton from the plasma membrane, which happens in platelets upon thrombin stimulation (Hirao et al., 1997), suggesting a role of CHK in platelet activation.

Last but not least, the level of CHK is regulated during development of germ cell, suggesting that CHK may be involved in the regulation of differentiation of male germ cells (Kaneko et al., 1995). In summary, CHK can be recruited to the nucleus, perinuclear region, and plasma membrane to perform different roles under different circumstances.

Whereas CSK is constitutively expressed, CHK expression is induced in T cells upon activation (McVicar et al., 1994). CHK expression is regulated by IL-4 family cytokines (IL-4, IL-13, IL-3, GM-CSF) in human monocytes (Musso et al., 1994; Hiremath et al., 2004), and by stem cell factor (SCF) and PMA in the human megakaryoblastic cell line MO7e (Grgurevich et al., 1997). The consistent increases in CHK mRNA and protein levels induced by the cytokines demonstrate that, in both blood cell types, transcriptional control of CHK stimulated by cytokines may play important roles in regulating CHK levels (Grgurevich et al., 1997). However, CHK induction can be inhibited or reversed by treatment of IFN-γ, which controls protein synthesis (Hiremath et al., 2004).

In colon cancer cells, the downregulation of CHK expression was associated with a high level of promoter methylation of CHK, which has been shown to be associated with enhanced levels of DNA methyltransferases (DNMTs). The hypermethylation of CHK promoters by DNMT promotes the oncogenic properties of colon cancer cells (Chüeh et al., 2021; Zhu et al., 2021).

Unlike CSK, the roles that CHK plays are still largely unknown. Exploring potential CHK substrate(s) and verifying its roles in the signaling pathways are important in defining the exact physiological roles CHK plays. Furthermore, it is important to define roles of different isoforms of CHK. Depending on the cellular context, CHK may act as a tumor suppressor, possibly via inhibiting SFK signaling pathways, or act as a protooncogene, activating MAPK signaling pathways. Besides, CHK absence or its promoter methylation in colon cancer and brain cancer suggests its potential as a molecular biomarker in these cancer types and holds promise to become a novel cancer diagnosis standard alone or in combination with other molecules based on further clinical investigations. The ability of CHK to suppress tumorigenicity in a variety of cancers also holds promise for using CHK as an effective therapeutic intervention.