The predicted domains of SP85 are depicted in Fig. . In a previous study (), expression of DNA encoding the N domain with the celA signal peptide and a c-myc epitope tag, under control of the prespore-specific cotB promoter (pVSBN) (Fig. ), had resulted in a protein that was targeted properly to the PSV and subsequently secreted but was not incorporated into the coat. To examine the functional properties of the C1 and C2 domains, they were separately fused downstream of the N domain and its C-terminal myc tag (Fig. ), to direct targeting to the PSV as for native SP85. An alternative approach to express C1 and C2 alone has thus far been unsuccessful in vegetative cells, with most protein remaining intracellular and insoluble in extracts (P. Zhang and C. M. West, unpublished data). The resulting plasmids, pVSBNC1 and pVSBNC2, were electroporated as described in Materials and Methods into the normal strain Ax3, the SP85-null strain B1, and the cellulose synthase (dcsA)-null strain DG1099. The results from one representative high-expressing clone of each strain are presented here.
Photo provided by Flickr
As described in the Discussion, the C1 domain exhibits cellulose-binding activity, which might be involved in regulating cellulose synthesis. To test for this, NC1 was expressed in the dcsA-null strain DG1099, which lacks the cellulose synthase catalytic subunit (). Although true spores are not formed by the parental strain DG1099 owing to the lack of cellulose, they produce spherical, highly refractile, osmotically sensitive pseudospores enveloped by a diffuse precoat of protein (). Sporulation was induced by dissociating 24-h slug cells in 20 mM 8-Br-cAMP on coverslips, as sporulation of DG1099 in vivo is delayed and asynchronous compared to that of wild-type cells (). The appearance of normal pseudospores is shown in Fig. . Expression of NC2 in DG1099 (strain DNC2) did not affect pseudospore differentiation (Fig. ). In contrast, DNC1 produced only amoeboid cells that remained attached to the coverslip and occasional stalklike cells recognized by their vacuolation (Fig. ). The cells exhibited only slight swelling when incubated in water for 15 min (data not shown), suggesting that they are more like prespore cells than pseudospores, which swell extensively under these conditions (). The somewhat more amoeboid appearance of DNC1 (Fig. ) than of ANC1 cells (Fig. ) might have been because DNC1 cells were differentiated in osmotically balanced buffer rather than in the hypertonic environment of the sorus (). As in strain ANC1, SpiA expression was normal (Fig. , lanes K to M), and immunolabeling showed that coat proteins were localized exclusively at the cell surface (Fig. ) as seen in DG1099 and DNC2 pseudospores (Fig. ). Therefore, DNC1 cells exocytosed their PSVs normally but were unable to form pseudospores, which involves cell shrinkage, suppression of osmotic regulation, and acquisition of a spherical shape (). Thus, NC1 appeared to interrupt the entire terminal differentiation program that occurs after exocytosis (i.e., stage II [see Fig. ]) not just cellulose synthesis per se. NC1 action does not depend on cellulose binding and does not act by inhibiting intracellular cAMP accumulation, as the effect was observed even though the cells were bathed in the membrane permeant cAMP analog 8-Br-cAMP.
spore coat polysaccharide biosynthesis protein SpsI
Photo provided by Flickr
The effect of NC1 appears to be attributable to native functions of the C1 and N domains and not to misfolded or otherwise aberrant protein configurations. NC1 is expressed at a level comparable to that of other coat proteins and accumulates at the cell surface coordinately with other coat proteins (Fig. ). The C1 domain confers retention of the fusion protein at the cell surface since the N domain, expressed alone, is released in soluble form (). The fraction of NC1 that is secreted in soluble form by induced dcsA-null cells is competent to bind Avicel cellulose (L. Kaplan and C. M. West, unpublished data), indicating that its folding is not grossly aberrant. Moreover, C4C motifs are expressed in a variety of sequence contexts in other coat proteins (), suggesting that their local environments might not be critical for folding. Indeed, C4C-folding may be intrinsically driven as the absence of N-glycosylation motifs indicates that it is not subject to glycosylation-dependent quality control in the rER (), and independent folding seems to be a general feature of EGF-like modules (). NC1-mediated arrest occurs at a discrete step many hours after accumulation of NC1, arguing that NC1 does not nonspecifically interfere with rER or Golgi processes. Finally, the effect of NC1 depends on the presence of endogenous SP85 (Fig. ), which also argues that it is not simply due to an accumulation of denatured NC1.