CLIC4 and Schnurri-2

Abstract

CLIC4 is a highly conserved, multifunctional member of the chloride intracellular channel family of proteins. The protein is largely cytoplasmic but translocates to the nucleus upon a variety of stimuli including TGFβ, TNFα and etoposide. Nuclear resident CLIC4 causes growth arrest, terminal differentiation and apoptosis. Recently, it was discovered that TGFβ causes CLIC4 to associate with Schnurri-2 and together they translocate to the nucleus and dissociate thereafter. The nuclear function of CLIC4 was further illuminated by the discovery that CLIC4 enhances TGFβ signaling by associating with phospho-Smad2 and 3 and preventing their dephosphorylation. Enhanced TGFβ dependent gene expression and growth inhibition are downstream consequences of this activity of CLIC4. In this article, we speculate on other consequences of the CLIC4 relation to TGFβ signaling and the potential for CLIC4 to participate in other cellular functions related to normal homeostasis and disease.

The chloride intracellular channel family of proteins is comprised of seven family members-p64, CLIC1-CLIC5 and parchorin. CLIC1-CLIC5 proteins are similar in size and are highly homologous to the c-terminus of p64 and parchorin, while the N-termini of the latter are distinct. These proteins have a hydrophobic transmembrane domain that may be used to insert into the membrane and facilitate anion transport. Several CLIC family members are dimorphic and can transition between membrane bound or soluble forms in the cytoplasm.¹ Soluble CLIC proteins are structurally homologous to the glutathione s-transferase family of proteins and are subject to redox regulation of their three dimensional state.² CLIC family members are multifunctional and participate in a wide variety of signaling activities.

CLIC4 is the most extensively studied member of the CLIC family. It is a 28KD ubiquitously expressed protein with a particularly high expression in skin, kidney, brain and testes. Like several other CLIC proteins, CLIC4 is dimorphic. It has a transmembrane domain and is found in intracellular membranes including inner mitochondrial membrane, trans Golgi network and the endoplasmic reticulum. Soluble CLIC4 is abundant in the cytoplasm. CLIC4 also has a functional nuclear localization signal (NLS), phosphorylation sites for protein kinase C, protein kinase A and casein kinase II and several protein binding sites including those for 14-3-3, SH3 and SH2.³ It has putative chloride ion selective channel activity exhibiting a single channel conductance.⁴ CLIC4 is upregulated in response to p53 transcriptional activity and is essential for p53 and c-Myc mediated apoptosis.^5,6 Cytoplasmic CLIC4 translocates to the nucleus in multiple cell types, under conditions of metabolic stress, growth arrest, apoptosis and DNA damage.⁷ The presence of CLIC4 in the nucleus is strongly associated with growth arrest and apoptosis. Nuclear CLIC4 is also required for differentiation of keratinocytes,⁸ and CLIC4 expression is highly associated with adipocyte differentiation.⁹ In vivo, CLIC4 is nuclear in quiescent epithelial cells in most tissues and poorly expressed in the corresponding stromal compartment. In contrast, nuclear CLIC4 is lost from tumor epithelia and highly upregulated in tumor stroma.¹⁰ A potential functional role of CLIC4 in tumor stroma is supported by data showing that CLIC4 is the most upregulated transcript during conversion of fibroblasts to myofibroblasts,¹¹ and overexpression of CLIC4 enhances alpha smooth muscle actin (α-SMA) expression associated with fibroblast to myofibroblast transdifferentiation.¹⁰ CLIC4 knockdown abolishes TGFβ mediated α-SMA induction.¹² The recent revelation that Schnurri-2 is a CLIC4 binding protein suggested for the first time that CLIC4 might be involved in TGFβ superfamily signaling.

Schnurri-2, a transcription factor, is important for Decapentaplegic (Dpp) mediated dorsal patterning and growth in Drosophila.^13–15 The Drosophila genome has one Schnurri (Shn) gene that contains two metal fingers with significant homology to the metal finger regions of vertebrate Schnurri proteins. In vertebrates, the Schnurri family consists of three homologs¹⁶ which share homology in their zinc finger domains and molecular weight of approximately 300 kD but have little sequence similarity and display distinct functions. Murine Schnurri-2 is required for thymogenesis^17–19 and bone remodeling.²⁰ Schnurri-2 deficient mice have a defect in T cell development, have reduced white adipose tissue, a defect that is attributed to defective BMP dependent adipogenesis²¹ and are also hypersensitive to stress.²² The Schnurri-2 protein has been previously linked to BMP signaling.^13,14,21,23 Schnurri-2 undergoes nuclear translocation upon BMP-2 stimulation. At least three nuclear localization signals (NLS) in the Schnurri-2 sequence indicate a high propensity to traffic into the nucleus. Previous studies have implied a role for Schnurri-2 in TGFβ signaling in mouse bone where genetic ablation of Schnurri-2 impairs osteoclast function and osteoblast differentiation, changes also produced by targeting a dominant negative type 2 TGFβ receptor to osteoblasts.^20,24 Our recent findings directly confirmed a role for mammalian Schnurri-2 in the canonical TGFβ pathway and elucidated a crucial function of Schnurri-2 as a cotransporter of CLIC4 to the nucleus in response to TGFβ.²⁵

The CLIC4-Schnurri-2 Interaction

TGFβ enhances the expression and association of both CLIC4 and Schnurri-2 and promotes their nuclear translocation.²⁵ Once inside the nucleus, they dissociate. Nuclear CLIC4 binds to phosphorylated Smads 2 and 3 and prevents their dephosphorylation by Smad phosphatases, specifically by PPM1a, a recently identified Smad phosphatase.²⁶ By inhibiting Smad dephosphorylation, nuclear CLIC4 prolongs the nuclear residence of p-Smad2 and 3 and enhances the TGFβ downstream functions including growth arrest and transcriptional activation.

While CLIC4 has previously been shown to associate with dynamin I, actin, tubulin and 14-3-3 isoforms,²⁷ binding to Schnurri-2 is the first reported CLIC4 interaction with proteins known to function as transcription factors. Aa 814–1167 of Schnurri-2 interact with aa 121–197 of CLIC4. This region of Schnurri-2 encodes one of its putative nuclear localization sequences. The 121–197 region of CLIC4 contains several consensus phosphorylation sites (casein kinase II, protein kinase C), an N-myristoylation site and the 14-3-3 binding site.²⁷ It is not clear if modification of CLIC4 is required for interaction with Schnurri-2, but the absence of direct association of full length recombinant and in vitro translated CLIC4 and Schnurri-2,²⁵ respectively suggests cell based processing is involved. Smad proteins are not required for the TGFβ stimulated association of Schnurri-2 and CLIC4.²⁵ This suggests that additional TGFβ dependent mechanisms must regulate these events. Furthermore, the nuclear localization signal (NLS) domain of CLIC4 is not required for interaction with Schnurri-2 since the CLIC4 deletion construct lacking the NLS domain (aa 1–197) still associates with Schnurri-2. Previous data indicates that the CLIC4 NLS is required for nuclear translocation in response to cell stress,⁷ suggesting that aside from TGFβ stimulation, there are additional mechanisms regulating CLIC4 nuclear translocation independent of Schnurri-2.

Since there is no detectable direct interaction between Schnurri-2 and CLIC4 in the nucleus and nuclear CLIC4 can enhance TGFβ signaling in the absence of Schnurri-2, it appears that a major requirement of the Schnurri protein to enhance TGFβ signaling is to facilitate CLIC4 nuclear translocation. Binding of CLIC4 to a Schnurri-2 NLS region and the presence of three NLSs in the Schnurri-2 protein would make Schnurri-2 an effective chaperone for CLIC4, and potentially other proteins as well, to be carried into the nucleus. The absence of interaction of Schnurri-2 and CLIC4 within the nucleus is consistent with other nuclear importing complexes where carrier proteins discharge their cargo once translocation occurs. Nevertheless, an indirect interaction between CLIC4 and Schnurri-2 in the nucleus such as through adaptor proteins cannot yet be ruled out. If such interactions occur, they may impact the functions of nuclear CLIC4 or nuclear Schnurri-2. Further understanding of these mechanisms could reveal novel pathways to interrupt TGFβ signaling.

CLIC4 Inhibits PPM1a Dependent Dephosphorylation

CLIC4 employs an unprecedented mechanism for TGFβ signal enhancement, by protection of phospho-Smad2 and phospho-Smad3 from the action of phosphatases that function to dampen the TGFβ signal. The wide variety of cellular processes regulated by TGFβ signaling depends on a delicate balance between Smad phosphorylation and dephosphorylation.^28,29 TGFβ binds to its cell surface receptors, facilitating the phosphorylation of Smad2 and Smad3. Activated Smads form a complex with Smad4 and translocate to the nucleus where they regulate the expression of target genes leading to a specific biological response. The TGFβ signal is terminated via several mechanisms including reduction in active TGFβ ligand, degradation of receptors, induction of inhibitory Smads and Smad dephosphorylation. A common theme among all these processes is regulation of phospho-Smad levels at a particular time. Dephosphorylation of Smad proteins shuttles them out of the nucleus, terminating the signal. Thus the duration of nuclear presence of Smads and therefore the length of the TGFβ stimulus is dependent on the phosphorylation state of the Smads. While there are other proteins that alter nuclear levels of phospho-Smads, protection of phosphorylated Smads from phosphatase action as a way to prolong their activity is so far unique to CLIC4. This mechanism provides a new target for therapeutic intervention which may now be explored for other TGFβ signal enhancers as well. It also presents CLIC4 as a novel modifier of TGFβ signaling, reiterating the fact that a delicate balance between the activity of R-Smad kinases and phosphatases seems to fine tune the complex and critical TGFβ signaling pathway. Although only PPM1a has been examined, the possibility of CLIC4 modifying the ability of other, hitherto undetected Smad phosphatases cannot be ruled out. PPM1a (also known as PP2Cα) has also been implicated as a phosphatase for Smad1,³⁰ p-IKKbeta,³¹ phospho-Cdks³² and MAPK³³ proteins. The binding site of CLIC4 on p-Smad2/3 is not known yet, but if CLIC4 is a higher affinity competitor for binding to the same Smad2/3 motif as PPM1a and if this motif is similar in all PPM1a binding proteins, CLIC4 could influence a number of other pathways as well. For example, CLIC4 could promote TNFα induced NFκB activation by inhibiting PPM1a/PPM1b mediated IKKβ dephosphorylation. It is intriguing that TNFα causes CLIC4 upregulation and nuclear translocation⁵ which may have some bearing on a potential role of CLIC4 in augmenting TNFα responses in a positive feedback loop. Similarly, CLIC4 could influence cell cycle by modulating the activity of cyclin dependent kinases or alter cellular stress response by increased phosphorylation of JNK and p38. These possibilities are purely speculative and whether the protection afforded by CLIC4 is unique to Smad2/3 or is common to all PPM1a targets remains to be elucidated. However, it does present a possibility for CLIC4 to be implicated in a wide variety of signaling pathways and therefore influence an array of biological functions.

Schnurri-2 as a TGFβ Signaling Regulator

The nuclear translocation of CLIC4 adds another dimension to the participation of Schnurri-2 as a regulator of TGFβ signaling. Whether Schnurri-2 regulates TGFβ signaling apart from its modulation of CLIC4 nuclear translocation is not clear yet. Nevertheless, several functions of Schnurri-2 may now be attributed to its direct involvement in CLIC4 function and TGFβ signaling.

The role of Schnurri-2 in T cell development may be linked to TGFβ signaling. Schnurri-2 knockout mice have defective T cell development in the thymus.^17–19 These mice have impaired positive selection for CD4⁺ and CD8⁺ T cells and also exhibit enhanced differentiation of peripheral T cells into Th2 cells. TGFβ is a critical regulator of T cell development and differentiation.³⁴ TGFβ promotes differentiation of thymic T cells and survival of peripheral CD4⁺ and CD8⁺ T cells.³⁴ Interestingly, TGFβ blocks Th1 and Th2 cell differentiation in peripheral T cell populations.³⁴ The mechanism of TGFβ dependent thymic T cell differentiation is yet unknown, but the TGFβ mediated inhibition of Th2 cell differentiation occurs in a Smad dependent manner.³⁴ Whether the defect in T cell development in Schnurri-2^-/- mice arises from altered TGFβ signaling is not known yet, but given the importance of both TGFβ and Schnurri-2 in this process and their ability to influence each other, it would not be surprising if Schnurri-2 modulates these processes by participating in TGFβ signaling.

Similarly the role of Schnurri-2 in brain function may be related with TGFβ activity. Schnurri-2^-/- mice display hyperactivity and hypersensitivity to stress.²² Anxiety and stress induced corticosterone levels were higher in these mice. TGFβ superfamily (TGFβ/BMP/activin) signaling is critical for development of the nervous system.³⁵ These pathways may also be important in neuronal network formation, neuroplasticity and synapse function in the adult brain.³⁵ The hypothesis that Schnurri-2 might affect brain responses through TGFβ signaling²² now finds further support. This might involve CLIC4 as well, as discussed in the following section.

Role of CLIC4 in Schnurri-2 Responses

Schnurri-2 translocates to the nucleus following BMP-2 treatment and, together with Smad1/4 and CEBPα, regulates the expression of PPARγ2, a key transcription factor for adipocyte differentiation.³⁶ Quite expectedly then, Schnurri-2 knockout mice have reduced white adipose tissue and MEFs from these mice have an impaired ability to differentiate into adipocytes. It is possible that CLIC4 translocates together with Schnurri-2 here as well and protects Smad1 from PPM1a phosphatase activity (PPM1a acts as a Smad1 phosphatase as well). This would enhance the presence of Smad1 on the PPARγ2 promoter and increase PPARγ2 expression, making CLIC4 an important player in Schnurri-2 dependent adipocyte differentiation. Giving credence to this possibility is the report that CLIC4 expression is highly upregulated during differentiation of 3T3-L1 fibroblasts into adipocytes.⁹ Although not studied for PPARγ, it is interesting that mice that are knockout for PPARβ/δ (also implicated in lipid metabolism) have reduced expression of CLIC4.³⁷ The presence of a feedforward loop is likely, where CLIC4 enhances the expression of PPAR isoforms which in turn upregulate CLIC4 expression during adipogenesis.

The role of Schnurri-2 in the brain, described earlier, may also involve CLIC4. Brain is one of the highest CLIC4 expressing organs. Rat brain CLIC4 is localized to large dense core vesicles in neurosecretory cells.³⁸ Brain CLIC4 directly associates with dynamin I and 14-3-3 isoforms.²⁷ Dynamin I is essential for synaptic vessel endocytosis and neurite formation³⁹ while certain 14-3-3 isoforms are also localized in synaptosomes and synaptic membranes.⁴⁰ It has been hypothesized that CLIC4 might be associated with synaptic vessel endocytosis. Based on the strong association between TGFβ superfamily signaling and nervous system establishment and function, we hypothesize that the large amount of CLIC4 present in the brain might affect neuronal function by altering TGFβ signaling in a Schnurri-2 dependent manner. Alternatively, the importance of Schnurri-2 in brain function might be linked to a requirement for it in CLIC4 modification of TGFβ signaling.

CLIC4 and p53 Signaling

The discovery that CLIC4 is an important participant in TGFβ signaling and acts through a nuclear mechanism could explain some of the diverse activities of this multifunctional protein. CLIC4 is a direct response gene for p53 mediated transcription⁵ and is required for p53 mediated apoptosis⁴¹ but a direct link has not been elucidated. However, p53 and TGFβ have several target genes that are co-regulated, with Mix2, PAI-1, p21, MMP2, SM22 and Smad7 being a few among 200 coregulated genes that mainly fall in either of the two categories of growth inhibition or extracellular matrix remodelling genes.^42,43 p53 acts as a master regulator of TGFβ signaling through binding to the R-Smad-Smad4 complex on DNA and enhancing Smad-DNA binding activity.⁴⁴ It also modulates transcription of TβRII.⁴² p53 regulation of CLIC4 expression could be mediated via its effect on TGFβ dependent gene transcription. A need for CLIC4 during p53 dependent apoptosis might be due to a necessity for enhanced TGFβ signaling that the presence of CLIC4 would ensure. p53 can be crucial for TGFβ induced growth arrest, and lack of p53 or expression of dominant loss of function mutant p53 hampers the TGFβ cytostatic effect and impairs the induction of p21 expression by TGFβ in some cell types.⁴³ Since p53 regulates CLIC4 expression, a requirement for intact p53 function in several TGFβ responses, might reflect, at least in part, a need for CLIC4 in the TGFβ program.

CLIC4 in Cancer

It is well documented that tumors harbour a variety of p53 mutations and it has been proposed that loss of p53 and p53 mutations could contribute to TGFβ refractoriness in tumors.^43,44 The reduced expression and exclusion of CLIC4 from the nucleus of tumor cells¹⁰ may be associated with loss of TGFβ responsiveness and p53 mutations commonly seen in tumors. The progressive decline in nuclear CLIC4 with increasing tumor grade further strengthens this possibility as later stage cancers develop greater resistance to TGFβ. At the same time, decreased nuclear CLIC4 in tumor cells could in turn lead to a further reduction in TGFβ signaling for lack of inhibition of Smad dephosphorylation that CLIC4 presence would have afforded (Fig. 1). Loss of CLIC4 could aid in tumorigenesis and tumor progression by inhibiting differentiation. CLIC4 antisense inhibits keratinocyte differentiation in a PKCδ regulated manner.⁸ Since the activity of PKCs is partly regulated by TGFβ signaling, this effect of CLIC4 on PKC dependent keratinocyte differentiation could be related to its involvement in TGFβ signaling.

Figure 1.

Loss of CLIC4 in tumor cells reduces sensitivity to TGF-β. (A) A schematic representation of the mechanism of enhancement of TGF-β signaling by CLIC4 in a normal cell where TGF-β causes nuclear translocation of CLIC4 in conjunction with Schnurri-2. Subsequently, nuclear CLIC4 inhibits binding of PPM1a and other phosphatases to p-Smads 2 and 3 and thus delays their dephosphorylation and prolongs the TGF-β signal. (B) We hypothesize that lack of CLIC4 in tumor cells would prevent this stabilization of p-Smads and contribute to decreased TGF-β responsiveness in these cells.

While TGFβ signaling components play an important role in tumor suppression in cancers of epithelial origin, in certain contexts and depending on the stage of cancer, TGFβ acts to enhance tumor progression, increasing invasion and metastasis of tumor cells. Epithelial growth inhibition by CLIC4 and enhanced tumor growth in xenografts using CLIC4 overexpressing fibroblasts¹⁰ suggests that such a dual role might exist for CLIC4 as well. An important aspect of the pro-tumor activities of TGFβ is its contribution to making a rich stroma to support tumor growth and invasion. Increased tumor growth with CLIC4 overexpressing fibroblasts along with the association of CLIC4 and alpha smooth muscle actin expression in the transition of fibroblasts to myofibroblasts in tumor stroma of multiple target sites^10,11 may now be linked to regulation of transdifferentiation by TGFβ signaling.⁴⁵ Yao et al. recently showed that TGFβ causes CLIC4 upregulation via production of reactive oxygen species during fibroblast to myofibroblasts transdifferentiation in ovarian cancer stroma.¹² Furthermore, CLIC4 expression in tumor stroma increases with tumor progression. This supports its importance in fibroblast transdifferentiation and suggests that CLIC4 might also contribute to TGFβ dependent extracellular matrix generation. CLIC4 also plays an important role in multiple steps of vasculogenesis and angiogenesis. CLIC4 promotes endothelial cell proliferation and regulates endothelial morphogenesis.⁴⁶ CLIC4^-/- mice have defective tubulogenesis in vitro and angiogenesis in vivo⁴⁷ and impaired collateral formation in skeletal muscle and brain.⁴⁸ The CLIC4 orthologue EXC-4 is necessary for proper development of the C. elegans excretory canal.⁴⁹ It is well known that the TGFβ family is critical in endothelial cell differentiation, vascular network formation and maintenance of vessel wall integrity. Several studies on mice deficient in TGFβ1, TβRII, ALK5 and ALK1 show major defects in vascular morphogenesis and angiogenesis.^50,51 In this light, it is highly likely that the critical role that CLIC4 plays is attributed to its influence on TGFβ signaling. Although, some of these possibilities are speculative presently and open for testing, they do point to an important role of CLIC4 in cancer development.

The revelation of their involvement in TGFβ signaling presents Schnurri-2 and CLIC4 as a unique combination of novel players in a critical pathway controlling diverse aspects of cell behavior. Data suggest that CLIC4 may be an attractive target both in cancer stages where TGFβ signal augmentation or inhibition is required due to responses in separate tissue compartments. Whether Schnurri-2 would aid CLIC4 function in both these aspects remains to be determined. Apart from cancer therapy, these two proteins may also serve as targets for modifying TGFβ signaling under conditions where this pathway is involved in other pathological changes.

Commentary to: Shukla A, Malik M, Cataisson C, Ho Y, Friesen T, Suh KS, Yuspa SH. TGFβ signalling is regulated by Schnurri-2-dependent nuclear translocation of CLIC4 and consequent stabilization of phospho-Smad2 and 3. Nat Cell Biol. 2009;11:777–784. doi: 10.1038/ncb1885.