|Home | About | Journals | Submit | Contact Us | Français|
Myosin II disassembly in Dictyostelium discoideum is regulated by three structurally related myosin heavy chain kinases (myosin II heavy chain kinase A [MHCK-A], -B, and -C). We show that the WD repeat domain of MHCK-C is unique in that it mediates both substrate targeting and subcellular localization, revealing a target for regulation that is distinct from those of the other MHCKs.
The ability of a cell to undergo highly specific modifications in shape during processes such as cytokinesis, cell migration, cell adhesion, and receptor capping is dependent, in large part, on the proper control of where and when myosin II contracts actin filaments in the cell (3, 4). In Dictyostelium discoideum, myosin II filament disassembly is regulated by at least three myosin II heavy chain kinases (myosin II heavy chain kinase A [MHCK-A], MHCK-B, and MHCK-C). The Dictyostelium MHCKs possess alpha kinase domains and carboxyl-terminal WD repeat domains (11, 13, 17). The WD repeat domains of MHCK-A and MHCK-B facilitate myosin II heavy chain phosphorylation by these kinases by binding directly to myosin II filaments (14, 15). However, the WD repeat domains play no detectable role in determining the subcellular localization of these kinases. Similar functions for the WD repeat domain of MHCK-C have not been explored, and there is nothing known about the signaling events regulating MHCK-C localization and activity, thus limiting comparisons among the MHCKs that could ultimately reveal distinct functions and mechanisms of regulation for these seemingly redundant enzymes.
We explored the hypothesis that the predicted structural similarity of the WD repeat domain of MHCK-C with those of MHCK-A and -B reflects a shared function, such as substrate targeting (7). To this end, we examined the ability of FLAG-tagged full-length MHCK-C or a truncation lacking the WD repeat domain (MHCK-C-Δ-WD) to phosphorylate MHC substrate (0.42 μM) as well as to phosphorylate a peptide substrate (MH-1; 50 μM) that is phosphorylated by other alpha kinases in a WD repeat domain-independent manner (Fig. 1A and B) (15). Full-length MHCK-C phosphorylated MHC in a time-dependent manner and with a stoichiometry of approximately 2 moles of phosphate transferred per mole of MHC (Fig. 1C). In comparison, the MHCK-C-Δ-WD truncation was unable to phosphorylate MHC to a significant level, even after 20 min of incubation (Fig. 1C). In contrast, both MHCK-C and MHCK-C-Δ-WD phosphorylated the MH-1 peptide robustly and to essentially the same level (Fig. 1D), demonstrating that the catalytic activity of the MHCK-Δ-WD truncation is intact even in the absence of the WD repeat domain, and by extension, that the WD repeat domain is required for MHCK-C phosphorylation of MHC.
To examine further the role of the WD repeat domain in defining the activity of MHCK-C, Dictyostelium cell lines overexpressing FLAG-tagged full-length MHCK-C (MHCK-C++ cells) or truncations either lacking the WD repeat domain (MHCK-C-Δ-WD++ cells) or containing only the WD repeat domain (MHCK-C-WD++ cells) were monitored for growth in suspension culture. These cell lines exhibited expression levels at least 14-fold-higher than that of endogenous MHCK-C (see Fig. S1A and S1B in the supplemental material). We found that MHCK-C++ cells exhibited dramatically reduced rates of proliferation in suspension culture (Fig. 2A), accompanied by increased cell multinuclearity (Fig. 2B and C) and loss of cytoskeleton-associated myosin II (see Fig. S1C in the supplemental material). In contrast, the MHCK-C-Δ-WD++ cells and MHCK-C-WD++ cells grew at essentially the same rate as AX2 and MHCK-C-null cells, with very few multinucleated cells (Fig. 2). We also found that MHCK-C-Δ-WD++ cells, unlike MHCK-C++ cells and myosin II-null cells (6), completed the developmental cycle and formed fruiting bodies containing viable spores (data not shown).
To gain insight into the signaling pathways involved in regulating MHCK-C activity and localization, we examined the effect of cyclic AMP (cAMP) stimulation on the localization of green fluorescent protein (GFP)-tagged MHCK-C (GFP-MHCK-C). It is noteworthy that the expression of GFP-MHCK-C for these experiments (see Fig. S2A in the supplemental material) did not affect the levels of cytoskeleton-associated myosin II (see Fig. S2C in the supplemental material) and did not induce the myosin II-null phenotype observed with MHCK-C++ cells (see Fig. S2B in the supplemental material). For localization experiments, these cells were prepared and then stimulated with cAMP (50 μM) as described by us previously (16). In unstimulated cells, GFP-MHCK-C exhibits a relatively diffuse distribution (Fig. 3A, 0-s panel). About 40 seconds after stimulation, GFP-MHCK-C becomes detectably enriched throughout the cell cortex, with peak enrichment observed about 60 s after stimulation (Fig. 3A, white arrows). Chemoattractant-induced translocation was undetectable in cells lacking myosin II or treated with latrunculin A (5 μM, 20 min) (see Fig. S3 in the supplemental material). Complementary Triton X-100 cytoskeleton fractionation assays (performed exactly as described previously ) revealed that cAMP stimulation of Dictyostelium AX2 cells expressing only endogenous MHCK-C leads to a greater than 3-fold increase in the amount of kinase associated with the cytoskeleton-enriched fraction, with a timing that lags slightly behind that of myosin II (Fig. 3B and C). Taken together, these data are the first to demonstrate that the activation of chemoattractant signaling pathways can induce MHCK-C recruitment to the cell cortex.
We next examined the structural determinants required for proper localization of MHCK-C in the cell. We found that GFP-C-Δ-WD, which lacks the WD repeat domain, is distributed diffusely throughout the cell, with no cortical enrichment upon cAMP stimulation (Fig. 4A). In contrast, GFP-C-WD, which contains the WD repeat alone, displayed robust cortical recruitment upon uniform stimulation of cells (Fig. 4A). GFP-C-WD also localizes specifically to the rear of cells undergoing chemotaxis or random migration (Fig. 4B), thus demonstrating that the WD repeat domain contains all of the determinants required for normal localization of MHCK-C in the cell. As with the full-length kinase, no detectable translocation of GFP-C-WD was observed in cells lacking MHC or cells that were treated with latrunculin A (see Fig. S3 in the supplemental material).
Collectively, our results are the first to show that a WD repeat domain can function both in directing kinase phosphorylation of a substrate and in mediating subcellular localization and thus reveal an important distinction from the functions of the same domain of the MHCK-A isoform. In the case of MHCK-A, the WD repeat domain shares substrate targeting, but not localization, activity with that of MHCK-C (14, 15). Instead, MHCK-A localization to the actin-rich anterior of the cell is mediated by an extended region at the amino terminus of the kinase that is predicted to form mainly coiled-coil structures (12). The findings reported here, along with those showing MHCK-C recruitment to the cleavage furrow (11), support a model in which the assembly of myosin II into the cell cortex is the chief determinant for the MHCK-C translocation to the rear of cells undergoing chemotaxis or to the cleavage furrow during cytokinesis (11). At this time, there are a number of candidate effectors of the MHCKs in Dictyostelium; however, none have been shown to alter kinase activity or localization via a direct interaction or by phosphorylation (or other modification) (2, 5, 10). Our results suggest that such signals could influence both MHCK-A and MHCK-C, but with very different responses from the kinases due to their very different localization domains.
This work was supported by an NIH grant to P.A.S. (2R15GM066789-02).
†Supplemental material for this article may be found at http://ec.asm.org/.
Published ahead of print on 11 December 2009.