Coronins From Unicellular Organisms

The identification of coronin in Dictyostelium was followed by the characterization of a coronin-related molecule in yeast.⁴ Unlike Dictyostelium that harbors three coronins, the single short coronin as well as a coronin 7 homologue (corA) and villidin (see Chapter 7), yeast contains a single coronin gene, termed crn1. Although, yeast coronin interacts with Arp2/3 in vitro,⁵ in contrast to the deletion of the short coronin in Dictyostelium, crn1-null mutants do not have any obvious phenotype and display normal F-actin dynamics.⁴

Coronins in Multicellular Invertebrate Organisms

Not much is known on the biology of coronins in the multicellular organisms Caenorhabditis elegans and Drosophila melanogaster. In C. elegans, a molecule named POD-1 that contains two stretches of WD repeats but no coiled coils, may be involved in the establishment of polarity in the developing embryo.⁶ In Drosophila, a Pod-1 homologue plays a role in axonal growth cone targeting.⁷ Besides Pod-1, both C. elegans as well as D. melanogaster contain a gene that encodes a protein (coro) that is far closer to Dictyostelium short coronin than Pod-1. However, while several Drosophila deletion mutants lacking coro show a rather pleiotropic phenotype, no information is available on a function for C. elegans coronin.

Mammalian Coronin 1

Coronin 1 (coronin 1A) is exclusively expressed in mammalian leukocytes, with no detectable expression in any other cell type (³ and data not shown). Several other coronin isoforms are expressed as well in leukocytes and therefore coronin 1 may have evolved to perform a highly specialized function in leukocytes. Interestingly, of all mammalian coronin isoforms, coronin 1 is the closest homologue to Dictyostelium short coronin. Whether or not the other coronin isoforms have evolved from coronin 1 or have arisen through a different evolutionary path is discussed to in Chapter 4.

Coronin 1 Has a Three-Domain Structure

A detailed sequence analysis revealed that coronin 1 (coronin 1A) is made up of three distinct domains (see Fig. 1A): The first, N terminal domain that also contains the 5 WD repeats is rich in β-sheet and is referred to as β–propeller (residues 1-355). The second domain is comprised of a region with little regular secondary structure and is referred to as linker domain (residues 356-429). Finally, the third domain is a coiled coil containing segment which is rich in α-helices (residues 430-461).⁸

Figure 1

Domain Structure of Coronin 1. A) The three domain structure of coronin 1 (cor- onin 1A). The N-terminal, 7-bladed beta propeller region consist of 5 typical WD repeats complemented with two stretches of sequence each forming four short β strands, (more...)

The Coronin 1 N-Terminal Domain Contains a 7-Bladed Propeller

One of the characteristic structure of all coronins are their central WD40 domains (see ref. 9 and Chapters 2 and 5). These repeats are characterized by a ~30-40 amino acid residue segment that are bordered by Gly-His (GH) and Trp-Asp (WD) peptide residues.^10,11 The WD repeat unit was first recognized in the beta subunit of the GTP binding protein transducin.¹² This domain seems to have evolved with the eukaryotic kingdom and may be involved in protein: protein interactions, although the precise function of the WD repeat domains in many proteins remains unknown.^13,14

Coronin-1 possesses 5 WD repeats and based on the homology with the G protein beta subunits it has been proposed that the WD repeat folds into a 5-bladed beta propeller.⁹ However, an extensive sequence analysis of coronin 1 revealed the presence of two additional sequence stretches of 46 and 44 residues, respectively, that flank the WD repeat-containing core sequence and are predicted to form four short β-strands and align with the corresponding β-strands of the five WD repeats.⁸ Since WD repeats are not strictly necessary to assert a propeller fold,^15-17 the prediction suggests that the coronin 1 propeller domain is, in fact, made up of at least seven blades instead of the previously proposed five blades.⁸ Consistent with this analysis, the crystal structure of coronin 1 indeed revealed the presence of a 7-bladed propeller¹⁸ (see Fig. 1B,C). Furthermore, the presence of a 7-bladed propeller in coronin 1 is consistent with the predicted similarity between the coronin 1 N-terminal domain and the yeast transcriptional repressor Tup1¹⁹ as well as the G protein β-subunit,²⁰ both WD repeat containing seven-bladed β-propeller proteins.

Coronin 1 Assembles into Coiled Coil Mediated Trimers

The C-terminal coiled coil in coronin 1 had been generally assumed to be involved in dimerization.⁹ However, observation of purified coronin 7 molecules isolated from macrophages by transmission electron microscopy revealed uniformly distributed particles with an apparent three-fold symmetry (Fig. 2). Further biochemical analysis suggested that the observed lobes apparent in Figure 2 correspond to the WD-repeat containing domains, that are assembled into trimers by virtue of the presence of the C-terminal coiled coil domain.⁸

Figure 2

Molecular organization of coronin 1. A) Transmission electron micrographs of affinity-purified and negatively stained coronin 1 (coronin 1A) complexes isolated from macrophages. The gallery shows multiple examples of the trimeric structure. Scale bar, (more...)

The assembly of the coiled coil domain into trimers, rather than the previously assumed dimer was further confirmed by X-ray crystallography.²¹ Interestingly, it turned out that the coronin 1 coiled coil contains a distinct structural motif, encompassing specific networks of surface salt bridges that is conserved among a variety of trimeric coiled coils.²¹ In this light it will be interesting to analyze the biophysical and biochemical characteristics of the coiled coil domains from other coronin isoforms.

Coronin 1 in Macrophages

One of the first reports on a role for coronin 1 in leukocytes implicated coronin 1 in the intracellular survival of pathogenic mycobacteria.³ While most other bacteria upon internalization in macrophages are rapidly transferred from phagosomes to lysosomes followed by their destruction, mycobacteria, once internalized, actively block the fusion of phagosomes with lysosomes.^25-27

A search for host molecules possibly involved in mediating the block in phagosome-lysosome fusion, identified coronin 1 (then known as TACO, for Tryptophan Aspartate containing Coat protein) as the sole detectable protein exclusively retained on mycobacterial phagosomes.³ Coronin 1, which in non-infected macrophages distributes equally between the cytosol and the membranes, is exclusively retained on phagosomes of macrophages infected with live mycobacteria, whereas in macrophages infected with dead mycobacteria, coronin 1 is initially co-internalized but rapidly dissociates from phagosomes. Following dissociation, the noncoated phagosomes fuse with or mature into lysosomes, resulting in the subsequent degradation of the internalized mycobacteria. This suggested an essential role for coronin 1 in preventing the fusion of mycobacterial phagosomes with lysosomes.^3,28 Moreover, mycobacteria are effectively destroyed within Kupffer cells, the resident macrophages of the liver that do not express coronin 1.³ Interestingly, virulent strains of the human pathogen H. pylori are equally capable to actively retain coronin 1 at the phagosomal membrane upon internalization, suggesting that the coronin 1-mediated block in lysosomal fusion might be utilized also by other pathogens.²⁹ However, whether or not pathogen specific molecules are involved in the active phagosomal retention of coronin 1 remains is as yet unclear.

How coronin 1 mediates the survival of pathogenic mycobacteria has remained obscure. While a direct contribution of coronin 1 to the internalization of mycobacteria could be excluded early on, its molecular role remained enigmatic. The analysis of a genetic model for coronin 1, however, has recently shed some light on the molecular activities of coronin 1 in macrophages. It turned out that, while being fully dispensible for all F-actin-mediated functions analyzed (phagocytosis, macropinocytosis, motility), coronin 1 is required for the activation of the Ca²⁺ dependent phosphatase calcineurin.³⁰ In wild type macrophages, upon internalization of mycobacteria this phosphatase becomes activated, thereby blocking phagosome-lysosome fusion by an as yet unknown mechanism and allowing mycobacterial survival (see also Fig. 3). In the absence of coronin 1, calcineurin activation does not occur resulting in phagosome-lysosome fusion and intracellular killing of the internalized mycobacteria. Strikingly, the genetic depletion of coronin 1 can be phenocopied by the addition of the calcineurin inhibitors cyclosporin A and FK506. Thus, it appears that coronin 1 has evolved to activate Ca²⁺ dependent signaling reactions in macrophages thereby promoting the survival of pathogenic mycobacteria.³⁰ How, exactly, coronin 1 mediates the activation of calcineurin remains to be analyzed and may be related to its activity in modulating the association of membranes with the cytoskeleton.⁸

Figure 3

Model for the Activity of Coronin 1 in Macrophages. In resting macrophages, coronin 1 (coronin 1A) is distributed between the cytoplasm as well as the cell cortex. Upon the entry of pathogenic mycobacteria, coronin 1 is recruited and actively retained (more...)

A role for coronin 1 in the regulation of signaling reactions in leukocytes is also consistent with its localization within cholesterol-enriched domains^31,32 as well as to a possible role for protein kinase C in the modulation of its localization.^33,34 Cholesterol-enriched domains are known to harbor a subset of molecules involved in signaling.^35-37 Since in the absence of cholesterol mycobacteria are not even phagocytosed,^31,38 it is well possible that mycobacteria have evolved mechanisms that ensure their uptake in a cholesterol dependent manner such that once they are internalized, they can reside within coronin 1-coated phagosomes. Consequently, the recruited coronin 1 then activates calcineurin resulting in blocking phagolysosome fusion and preventing degradation of the bacilli, thereby allowing the mycobacteria to survive intracellularly (Fig. 3).

Importantly, corroborating the original observation that coronin 1 was not involved in mycobacterial uptake, macrophages devoid of coronin 1 were perfectly capable to internalize mycobacteria as well as a range of other phagocytic cargo.³⁰ Thus, in contrast to the situation in Dictyostelium, coronin 1 is dispensable for phagocytosis in macrophages. Also, several other actin-dependent processes, such as macropinocytosis, cell motility, cell spreading and membrane ruffling were unaffected in macrophages lacking coronin. These results strongly suggest that in macrophages, coronin 1 is dispensable for F-actin mediated processes.

Acknowledgement

I thank the members of my laboratory for many stimulating discussions, Benoit Combulazier, Rajesh Jayachandran, Philipp Mueller and Varadha Sundaramurthy for critical reading of the manuscript and Annette Roulier for assistance in the artwork. Research in my laboratory is supported by the Swiss National Science Foundation, the Roche Research Foundation, the Olga Mayenfisch Foundations as well as the Swiss Lung Liga and the Swiss Life Jubileum Stiftung.