Punctuate Structures Research Articles

Characterisation of Toxoplasma gondii engineered to express mouse interferon-gamma

Recent studies have shown the feasibility of using Toxoplasma gondii as an expression system for heterologous protein. For better understanding of the mechanism of interferon-gamma (IFN-γ) dependent immunity to T. gondii, the parasites were stably transfected with IFN-γ gene, under control of the GRA1 promoter. Immunofluorescence analyses showed that recombinant mouse IFN-γ localised to discrete punctuate structures consistent with dense granules and secreted into the vacuolar space. The production of IFN-γ was detectable in both extracellular parasites and the parasite-infected cells. Growth of the recombinant parasites was inhibited in the mouse macrophage cell line (J774A.1 cells), but not in monkey kidney adherent fibroblasts (Vero cells), demonstrating the species-specificity of IFN-γ. Addition of anti-mouse IFN-γ antibody resulted in growth recovery of the recombinant parasites, suggesting that IFN-γ, secreted from the parasitised host cells across the parasitophorous vacuole membrane, acted in a paracrine manner. Reverse transcription (RT)-PCR analysis revealed significant expression of inducible nitric oxide synthase mRNA and high levels of nitric oxide production in recombinant parasite-infected J774A.1 cells. A competitive inhibitor of the l-arginine-dependent effector pathway, N G-monomethyl- l-arginine, inhibited the reduction of recombinant parasite growth in J774A.1 cells. Taken together, our data suggest that the T. gondii expression system may provide a new tool for cytokine gene expression and that parasites engineered to express a cytokine gene may be rationally designed for use in studies on immune responses to T. gondii.

The Dynamin-like GTPase DLP1 Is Essential for Peroxisome Division and Is Recruited to Peroxisomes in Part by PEX11

Peroxisome division involves the conserved PEX11 peroxisomal membrane proteins and in yeast has been shown to require Vps1p, a dynamin-like protein. We show here that DLP1, the human homolog of the yeast DNM1 and VPS1 genes, plays an important role in peroxisome division in human cells. Disruption of DLP1 function by either RNA interference or overexpressing dominant negative DLP1 mutants causes a dramatic reduction in peroxisome abundance, although overexpression of functional DLP1 has no effect on peroxisome abundance. Overexpression of PEX11 induces peroxisome division in a multistep process involving elongation of preexisting peroxisomes followed by their division. We find that DLP1 is dispensable for the first phase of this process but essential for the second. Furthermore, we show that DLP1 associates with peroxisomes and that PEX11 overexpression recruits DLP1 to peroxisome membranes. However, we were unable to detect physical interaction between PEX11 and DLP1, and the stoichiometry of PEX11 and peroxisome-associated DLP1 was far less than 1:1. Based on these and other aspects, we propose that DLP1 performs an essential but transient role in peroxisome division and that PEX11 promotes peroxisome division by recruiting DLP1 to peroxisome membranes through an indirect mechanism.

The deletion of CaVPS34 in the human pathogenic yeast Candida albicans causes defects in vesicle-mediated protein sorting and nuclear segregation.

A Candida albicans null mutant of the phosphatidylinositol (PI) 3-kinase gene (CaVPS34) involved in virulence was examined by different microscopical techniques. We observed that vacuoles of the Cavps34 null mutant were considerably enlarged and electron-transparent. An interesting result obtained by transmission electron microscopy analysis of Cavps34 mutant cells was the aberrant patch-like accumulation of vesicles, which were localized in the periplasm close to the plasma membrane. We assume that the vesicles result from missorted prevacuolar compartments. In contrast to the accumulations of the specific endocytic dye FM4-64 in the vacuole membrane in C. albicans wild-type strains (ring staining pattern), the Cavps34 mutant strain showed a staining of punctuate structures, possibly multivesicular bodies (MVB), that are scattered all over the cell. This defect indicates a late block in endocytic vesicle transport. Measurement of the total activity of carboxypeptidase Y revealed significantly lower activity in Cavps34 mutant cells. This may indicate that carboxypeptidase Y is not properly activated as a result of mislocalization due to the lack of Vps34p. The deletion of the CaVPS34 gene caused disturbance of normal nuclear migration, which suggests that in the Cavps34 mutant the cell-size mediated control process of cell division is affected.

JEM-1, a novel nuclear co-factor: localisation and functional interaction with AP-1.

JEM-1 is a novel gene whose mRNA expression in acute promyelocytic leukemia (APL) is induced by retinoid treatments. The gene product, a 45 kDa basic nuclear factor containing a leucine repeat, was transiently expressed in HeLa or COS-7 cells and immunocharacterized within the nuclei in fine punctuated structures which increase in size after cell transfection. Jem-1 was not expressed in the nucleoli. Experimental deletion of peptide domains of Jem-1 (JemDelta331-400 and Jem DeltaL179-206) showed that its C-terminal sequence (Thr331 --> Leu400) is required for nuclear translocation, while the leucine repeat domain (Arg179 --> Glu206) has no influence on subcellular localization. The Jem-1 protein was not detected in the PML-containing nuclear bodies or in speckled structures containing the splicing factor SC-35. In contrast it was localized in the nucleus in structures containing activator protein-1 (AP-1). DNA mobility shift assays showed that the in vitro translated Jem protein interacts neither with the DNA binding site of AP-1, nor directly with in vitro co-translated c-Fos or/and c-Jun proteins bound to this specific sequence. Interestingly, Jem-1-1 increased substantially the transcriptional activity of c-Jun (three-fold) and more strongly that of ectopically co-expressed c-Fos and c-Jun (five- to six-fold), as measured by a CAT reporter gene driven by a heterologous promoter containing the AP-1 binding site of the human collagenase gene. These synergistic effects were strongly Jem-1 dose-dependent. However, Jem-1 alone showed no activity on the collagenase promoter. A deletion of the leucine repeat of Jem-1 (Arg179 --> Glu206) did not diminish the enhancer capacity of Jem-1 on AP-1 activity. In contrast, the enhanced AP-1 activity was abrogated when Jem-1 was deleted of its C-terminus (Thr331 --> Leu400). We conclude that the 45 kDa nuclear product of the JEM-1 gene has features of a novel transcription cofactor, which is enhancing AP-1 activity without directly interacting with c-Jun or c-Fos proteins. Possible implications of these findings for APL cell maturation are discussed.

Actin-filaments localize on the sorting endosomes of 3Y1 fibroblastic cells.

By immunofluorescence microscopic observation, monoclonal and polyclonal antibodies against a synthetic actin C-terminal peptide were found to stain too colloguial, ambiguous punctuate structures distributed throughout the cytoplasm of 3Y1 cells, independently of actin stress fibers. Antibody against rab5, a small GTP binding protein of the sorting endosome, and anti-actin antibody co-stained these punctuate structures. On the other hand, transferrin receptor, a well characterized maker of the sorting and recycling endosomes, colocalized with actin on the vesicular structures at the cell peripheral region but not at the perinuclear area where the recycling endosome localized. These observations suggest that actin molecules localize on the sorting endosomes. Tropomyosin, F-actin binding protein, also colocalized with actin on the sorting endosomes. From these results, we proposed that actin-filaments with tropomyosin constitute the membrane skeleton on the sorting endosome surface. This article is the first report to show that actin-filaments localize on the intact endosomes.

Distinct cellular and subcellular patterns of expression imply distinct functions for the Drosophila homologues of moesin and the neurofibromatosis 2 tumor suppressor, merlin.

Interest in members of the protein 4.1 super-family, which includes the ezrin-radixin-moesin (ERM) group, has been stimulated recently by the discovery that the human neurofibromatosis 2 (NF2) tumor suppressor gene encodes an ERM-like protein, merlin. Although many proteins in this family are thought to act by linking the actin-based cytoskeleton to transmembrane proteins, the cellular functions of merlin have not been defined. To investigate the cellular and developmental functions of these proteins, we have identified and characterized Drosophila homologues of moesin (Dmoesin) and of the NF2 tumor suppressor merlin (Dmerlin). Using specific antibodies, we show that although these proteins are frequently coexpressed in developing tissues, they display distinct subcellular localizations. While Dmoesin is observed in continuous association with the plasma membrane, as is typical for an ERM family protein, Dmerlin is found in punctuate structures at the membrane and in the cytoplasm. Investigation of Dmerlin cultured cells demonstrates that it is associated with endocytic compartments. As a result of these studies, we propose that the merlin protein has unique functions in the cell which differ from those of other ERM family members.

CaBP1, a calcium binding protein of the thioredoxin family, is a resident KDEL protein of the ER and not of the intermediate compartment

A cDNA encoding rat CaBP1 has been isolated and sequenced. The deduced polypeptide chain consists of 440 amino acids including two internal thioredoxin-like domains and a C-terminal KDEL retention/retrieval signal. Regarding the high degree of identity to the hamster protein P5, CaBP1 is considered to be the homologous rat protein. Previous work has suggested that CaBP1 is a resident luminal protein of the intermediate compartment (Schweizer, A., Peter, F., Nguyen Van, P., Söling, H.D. and Hauri, H.P. (1993) Eur. J. Cell Biol. 60, 366-370). Our conclusion that CaBP1 is a resident protein of the endoplasmic reticulum and not of the intermediate compartment is based on three different approaches: subcellular fractionation, indirect immunofluorescence and overexpression of CaBP1. Subcellular fractionation of Vero cells in a velocity controlled step gradient led to copurification of CaBP1-containing vesicles and several marker proteins for the ER including calreticulin and alpha-SSRP. The intermediate compartment, as defined by a monoclonal antibody against the marker protein p53 (ERGIC-53), could be separated from these ER markers. Double immunofluorescence analysed by laser scanning microscopy showed no significant colocalization between CaBP1 and p53, but between CaBP1 and calreticulin. In addition experiments, Vero cells were infected with VSV tsO45. At 15 degrees C the VSV-G protein accumulated in punctuate structures representing the intermediate compartment, while CaBP1 maintained its original reticular localization. Even after high-level overexpression in COS cells, CaBP1 was not detected in the intermediate compartment, but was efficiently retained in the ER as judged by light microscopy.