Mapping Functional Domains of Chloride Intracellular Channel (CLIC) Proteins in Vivo

Abstract

Chloride intracellular channel (CLIC) proteins are small proteins distantly related to the omega family of glutathione S-transferases (GSTs). CLIC proteins are expressed in a wide variety of tissues in multicellular organisms and are targeted to specific cellular membranes. Members of this family are capable in vitro of changing conformation from a globular, soluble state to a membrane-inserted state in which they provide chloride conductance. The structural basis for in vivo CLIC protein function, however, is not well understood. We have mapped the functional domains of CLIC family members using an in vivo assay for membrane localization and function of CLIC proteins in the nematode Caenorhabditis elegans. A < 70 amino acid N-terminal domain is a key determinant of membrane localization and function of invertebrate CLIC proteins. This domain, which we term the ‘‘PTM‘‘ domain, named after an amphipathic putative transmembrane helix contained within it, directs distinct C. elegans CLIC homologs to distinct subcellular membranes. We find that within the PTM region, the cysteine residues required for GST-type activity are unnecessary for invertebrate CLIC function, but that specific residues within the proposed transmembrane helix are necessary for correct targeting and protein function. We find that among all tested invertebrate CLIC proteins, function appears to be completely conserved despite striking differences in the charged residues contained within the amphipathic helix. This indicates that these residues do not contribute to anion selectivity as previously suggested. We find that outside the PTM region, the remaining three-quarters of CLIC protein sequence is functionally equivalent not only among vertebrate and invertebrate CLIC proteins, but also among the more distantly related GST-omega and GST-sigma proteins. The PTM region thus provides both targeting information and CLIC functional specificity, possibly adapting GST-type proteins to function as ion channels.

Introduction

Chloride intracellular channel (CLIC) proteins are small proteins of approximately 240 residues that are distantly related to the omega family of glutathione S-transferases (GSTs).¹ CLIC proteins are expressed in a wide variety of tissues in multicellular organisms. Distinct CLIC proteins localize to distinct cellular membranes, including the plasma membrane, Golgi membrane, mitochondria, the nuclear membrane, dense core vesicles, lysosomal membranes, cell–cell junctions, and the luminal membrane of the excretory canal cell of the nematode Caenorhabditis elegans.2., 3., 4., 5., 6., 7. CLIC proteins have been shown to play roles in diverse processes including bone resorption,⁸ regulation of cell motility,⁹ apoptosis,4., 10., 11., 12. β-amyloid-induced neurotoxicity,¹³ and tubulogenesis⁶ (Figure 1).

Based on several lines of evidence, it has been proposed that CLIC proteins may function as chloride channels. The first CLIC protein to be identified, p64, was isolated by its ability to bind the known chloride channel inhibitor indanyloxyacetic acid (IAA).14., 15., 16. Since then, many CLIC proteins have been shown to confer chloride channel activity in transfected cells.2., 17., 18., 19. While CLIC proteins lack an identifiable signal sequence, epitope tagging experiments have demonstrated that transfected HsCLIC1 and HsCLIC4 can adopt an integral membrane conformation, spanning the membrane an odd number of times with the N terminus on the extracellular side.5., 20. Moreover, it has been shown that purified recombinant HsCLIC1 can insert into artificial lipid bilayers and confer chloride channel activity.21., 22., 23.Finally, proteins of the CLIC family have been associated with physiological processes requiring anion flux, such as water secretion and bone resorption.6., 8., 24.

In spite of these lines of evidence, the molecular function of CLIC proteins is still poorly understood, both on a biochemical and physiological level. Recent work indicates that, in vitro, rat CLIC1 may form non-selective pores rather than anion-selective channels.²⁵ Furthermore, it remains possible that CLIC proteins function in vivo as accessory proteins to anion channels, or play an altogether different role, such as interacting with the actin cytoskeleton.7., 26., 27., 28. Thus, several open questions about CLIC proteins remain. First, is the physiological role of CLIC proteins indeed that of an ion channel and if so, an anion channel or a non-selective pore? One strategy to address this issue at least in part is to define domains and residues required for CLIC translocation and conductance in vitro, and to determine if these are also required for the in vivo activity of CLIC proteins. Second, if CLIC proteins do function as anion channels or non-selective pores, what is the biophysical mechanism of conduction? Given that CLIC proteins are much smaller than and show no sequence homology with the well characterized ClC family of chloride channels,²⁹ they may utilize a novel mechanism of ion conductance. Third, given that there are six vertebrate CLIC homologs which are expressed in different cell types and localized to different cell membranes, do distinct CLIC proteins have distinct functions, or do they differ only in their expression patterns and membrane targeting properties? Similarly, does the function of vertebrate and invertebrate CLIC proteins differ? Lastly, how are CLIC proteins, which do not contain recognizable signal or sorting sequences, targeted to intracellular membranes?

We have previously described the identification of the genes exc-4 and exl-1, which encode the two CLIC protein orthologs in C. elegans.⁶ Genetic removal of exc-4 causes a cystic defect in the unicellular excretory canal cell of C. elegans⁶ (Figure 1). This defect has been proposed to model polycystic kidney disease.30., 31. We have shown that EXC-4 is localized to and required at the luminal membrane of the excretory canal cell to establish tubular architecture following cell hollowing and to maintain tubular architecture following development⁶ (Figure 1). These results have enabled us to establish an in vivo assay for CLIC protein localization and function, by expressing wild-type or mutated CLIC protein homologs in the C. elegans excretory canal cell and asking whether these correctly localize to the intracellular luminal membrane and provide rescuing activity in exc-4 null mutant animals.

We report here the results of using this in vivo assay to define residues and functional domains specifically required for CLIC protein targeting and function within a physiologically relevant, cellular context in C. elegans. We show that a minimal 66 amino acid N-terminal domain, termed the PTM domain, named after a putative transmembrane helix contained within it, is a key determinant of both membrane localization and CLIC-specific protein function. Within this region, an amphipathic α-helix containing positively charged residues has been proposed to function as a pore-lining helix.^19,25,32 We describe a mutational analysis of this helix using a functional in vivo assay. Moreover, we demonstrate that outside the PTM region, the C-terminal three-quarters of CLIC proteins are functionally equivalent among vertebrate and invertebrate CLIC proteins and, surprisingly, also among the more distantly related GST-omega and GST-sigma proteins. Our in vivo assay for CLIC function in C. elegans thus reveals that the PTM region provides CLIC-type functional specificity in addition to directing specific membrane targeting. Finally, we show that differences in membrane targeting by functionally equivalent CLIC proteins in C. elegans are determined by both cellular context and differences in protein sequence.