Novel antibodies detect nucleocytoplasmic O-fucose in protist pathogens, cellular slime molds, and plants

Abstract

Cellular adaptations to change often involve post-translational modifications of nuclear and cytoplasmic proteins. An example found in protists and plants is the modification of serine and threonine residues of dozens to hundreds of nucleocytoplasmic proteins with a single fucose (O-Fuc). A nucleocytoplasmic O-fucosyltransferase (OFT) occurs in the pathogen Toxoplasma gondii, the social amoeba Dictyostelium, and higher plants, where it is called Spy because mutants have a spindly appearance. O-fucosylation, which is required for optimal proliferation of Toxoplasma and Dictyostelium, is paralogous to the O-GlcNAcylation of nucleocytoplasmic proteins of plants and animals that is involved in stress and nutritional responses. O-Fuc was first discovered in Toxoplasma using Aleuria aurantia lectin, but its broad specificity for terminal fucose residues on N- and O-linked glycans in the secretory pathway limits its use. Here we present affinity purified rabbit antisera that are selective for the detection and enrichment of proteins bearing fucose-O-Ser or fucose-O-Thr. These antibodies detect numerous nucleocytoplasmic proteins in Toxoplasma, Dictyostelium, and Arabidopsis, as well as O-Fuc occurring on secretory proteins of Dictyostelium and mammalian cells, although the latter are frequently blocked by further glycosylation. The antibodies label Toxoplasma, Acanthamoeba, and Dictyostelium in a pattern reminiscent of O-GlcNAc in animal cells including nuclear pores. The O-fucome of Dictyostelium is partially conserved with that of Toxoplasma and is highly induced during starvation-induced development. These antisera demonstrate the unique antigenicity of O-Fuc, document conservation of the O-fucome among unrelated protists, and will enable the study of the O-fucomes of other organisms possessing OFT-like genes.

Keywords: Acanthamoeba castellanii; Arabidopsis thaliana; Dictyostelium discoideum; O-fucose; SPY; Spindly; Toxoplasma gondii; fucosyltransferase; mono-glycosylation; nuclear pore; nucleocytoplasmic glycosylation; parasites; protists; social amoeba.

Fig. 1.

Anti-fucopeptide antisera. (A) Workflow of antibody generation and application to O-fucome studies. The fucopeptide library immunogens (1) were injected into rabbits (2). After boosting, serum was collected and passed over a peptide library column (3), which was eluted with low pH buffer (4) to produce an agnostic reference antiserum (5). The flow-through from the peptide column was applied to a fucopeptide library column (6), which was eluted with low pH (7). The resulting affinity purified antiserum was assayed by ELISA(8), and used for immunofluorescence analysis of cells (9), western blotting of cell extracts (10), and for affinity capture of O-Fuc proteins (11) for proteomic analysis using nLC-MS/MS (12). (B) Description of the fucopeptide libraries, containing FOS or FOT, and peptide libraries containing Ser or Thr, that were incorporated via their N-terminal Cys residues into a tripartite immunogen (1), to columns for negative or positive affinity purification (3, 6), to BSA for ELISA assays (8), or to beads for affinity capture (11). (C) A dilution series of a representative bleed of affinity purified anti-FOS antisera was assayed for binding to the FOS library or the unmodified Ser-peptide library, using an ELISA method based on alkaline phosphatase conjugated goat-anti-rabbit IgG. (D) Titration of a representative affinity-purified bleed a rabbit immunized with the FOT peptide library. Error bars represent S.D. of 2 technical replicates.

Fig. 2.

Immunofluorescence analysis of anti-FOS specificity using Toxoplasma infected HFFs. Monolayers of human foreskin fibroblasts were infected with parasites, fixed, and imaged using Super-Resolution Structured Illumination Microscopy (SR-SIM). (A) HFFs containing wild-type RHΔΔ or TgSPYΔ parasites were probed with affinity purified rabbit anti-FOS (1 μg/ml) and mouse anti-SAG1 followed by Alexa Fluor-488 goat anti-rabbit IgG and Alexa Fluor-594 goat anti-mouse IgG to localize O-Fuc and outline the parasites, respectively. DAPI (blue) was used to visualize parasite and host cell nuclei (an example is labeled). An instance of 2 parasites occupying a parasitophorous vacuole within an HFF is outlined with a white circle. In the lower row, RHΔΔ samples were probed with anti-FOS in the presence of 0.2 M αMeFuc. Maximum intensity projections are shown. (B) Same as panel A, except that 1 μg/ml AAL-biotin and Alexa Fluor-594 streptavidin were used in place of anti-FOS. Scale bars: 5 μm. See Fig. S1 for corresponding probing with anti-FOT. (C) Same as above, except that samples were probed with anti-FOS and either AAL-biotin or anti-PLP6 (1:5000), a marker of nuclear epichromatin. Representative single z-slices of the nuclear regions are shown. See Fig. S2 for additional images with anti-FOS/T. (D) Pearson’s correlation coefficients for colocalization probes in panel C. Each point represents a single cell (n>30), and means are represented with a horizontal bar. The difference was significant using an unpaired, two-tailed t-test (****p< 0.0001). Scale bars: 5 μm.

Fig. 3.

Western blot analysis of anti-FOS and anti-FOT specificity in Toxoplasma. Parasites were harvested from hTERT fibroblasts as a mixture of intracellular and freshly spontaneously lysed out cells. In parallel with hTERT fibroblasts, samples were dissolved in SDS and analyzed by SDS-PAGE/western blotting. Post-blot gels were stained with Coomassie blue as loading controls. (A) Samples from RHΔΔ (wild-type) and SPYΔ strains, and host cells, were probed using 1 μg/ml AAL-biotin and Alexa Fluor-680 streptavidin. (B) The same samples were probed with 1 μg/ml anti-FOS and Alexa Fluor-680 goat-anti rabbit IgG, either with or without 0.2 M αMeFuc. (C) Same as panel B using 0.3 μg/ml anti-FOT. (D) Strips from a single western blot were probed with AAL-biotin, anti-FOS, anti-FOT or no primary probe (secondary) as indicated. Dashed lines represent boundaries between strips.

Fig. 4.

Immunofluorescence localization of anti-FOT and anti-FOS in Dictyostelium amoebae. Growth stage amoebae were allowed to adhere to coverslips, fixed in cold methanol, blocked in BSA, probed with the indicated reagents, and imaged with SR-SIM as in Fig. 2. (A) Wild-type (w/t, strain Ax3) amoebae were probed with anti-FOT and AAL-biotin, followed by Alexa Fluor-488 goat anti-rabbit IgG and Alexa Fluor-594 streptavidin. DAPI (blue) was used to visualize nuclei. Multinuclearity results from inefficient cytokinesis when grown in shaking culture. Maximum projection images are shown. (B) As in panel A, except that w/t and Ddspy-KO (spy⁻) amoebae were probed with anti-FOS and murine anti-actin followed by Alexa Fluor-488 goat anti-rabbit IgG and Alexa Fluor-594 goat anti-mouse IgG. See Fig. S3 for corresponding probing with anti-FOT. Scale bars: 5 μm. (C) As in panel A, except that amoebae were probed with anti-FOS and anti-PLP6 (1:5000), followed by Alexa Fluor-488 goat anti-rabbit IgG and Alexa Fluor-594 goat anti-mouse IgG. (D) Comparison of anti-FOS and anti-FOT with nuclear envelope proteins. Growth stage amoebae whose Nup62, Nup210, or Src1 genomic loci were C-terminally tagged with mNeon were probed with anti-FOS followed by Alexa Fluor-594 goat anti-rabbit IgG, and co-imaged with intrinsic mNeon fluorescence and DAPI. Single slice images are shown. See Fig. S3D for data for anti-FOT. (E) Pearson’s correlation coefficients for colocalization of anti-FOS and anti-FOT with Nup62-mNeon, Nup210-mNeon, and Src1-mNeon, calculated from super-resolution images. Each point represents a single cell, and means are indicated with a horizontal bar. Analyses were performed with ≥30 cells (except FOT – Nup210, which was not statistically evaluated). Significance was evaluated using an unpaired, two-tailed t-test (****p<0.0001, ***p<0.001, **p<0.01, *p<0.05, ns=not significant). Scale bars: 5 μm.

Fig. 5.

Developmental regulation of the Dictyostelium O-fucome. (A-C) The indicated strains of growth stage (vegetative) amoebae were subjected to SDS-PAGE, western blotting, and immunoprobing as in Fig. 3. The blots were probed with anti-FOS (A), anti-FOT (B), or AAL-biotin (C) with secondary Abs or streptavidin. Excerpts from the Coomassie blue stained blotted gels are shown as loading controls. (D, E). Growth stage amoebae (0 h) were deposited on non-nutrient agar and harvested after the indicated number of h for western blotting with anti-FOS (D) or anti-FOT (E). Multicellular slugs are formed by 14 h, Culmination begins around 18 h, and fruiting bodies are formed by 24 h. Molecular weight marker positions are shown in kDa. (F-I) Proteomics analysis of the O-fucome. Proteins were captured from CHAPS-solubilized whole cell extracts from wild-type (strain Ax3) or Ddspy-KO cells with anti-FOS or anti-FOT cross-linked to Protein A/G magnetic beads. Proteins eluted with αMeFuc in 8 M urea were reduced and alkylated, converted to peptides using endo Lys-C and trypsin, and analyzed by nLC using an Orbitrap mass analyzer. Proteins were identified based on detection of ≥2 peptides (1% FDR) at <5% protein FDR. The Volcano plots report on the fold enrichment in wild-type vs. spy-KO cells (abscissa) vs. confidence of fold-change for 3 biological replicates each with 3 technical replicates for a total of 9 data pairs for each protein. Data from anti-FOS and anti-FOT pulldowns of vegetative stage extracts are shown in panels F and G, and similarly for slug stage extracts in panels H and I. The ‘DDB_’ prefix is removed from the labels in Table 2, which lists wild-type enriched proteins (upper red quadrant, in red) along with which fractions they were found. See Table S1 for an analysis of proteins over-enriched in spy-KO extracts (upper left quadrant). Proteins for whose over-enrichment was supported in complementary pulldowns are in blue, and those without support are in white (see text for explanation). (J-O) Summary of O-Fuc proteins detected. Venn diagrams show distribution of O-Fuc candidates between anti-FOS and anti-FOT pulldowns, and between vegetative (veg) and slug stage cells, as indicated. Categories are illustrated in Table 2. Panel M includes candidates from a previous AAL-pulldown (7).

Fig. 6.

Recognition of O-Fuc proteins from Acanthamoeba, Arabidopsis, and mammalian cells. (A) Localization of AAL, anti-FOS and anti-FOT in Acanthamoeba castellanii, based on confocal microscopy. 2D=optical section; 3D=maximum projection). Note: DAPI labels nuclei and hundreds of mitochondria, which have an unusually large genome. (B,C) Western blot analysis of Arabidopsis seedlings. Protein was extracted from whole plant seedlings of wild-type (Columbia) and spy-3 and spy-4 strains, western blotted, and probed with AAL-biotin, anti-FOS, or anti-FOT, followed by streptavidin or secondary Abs as above. The post-blot gels were stained with Coomassie blue as loading controls. In panel C, w/t proteins were run on the same gel and slices were probed with anti-FOS or anti-FOT, with or without αMeFuc. Vertical dashed lines indicate slicing of western blots for probing with different Abs. Positions of molecular weight markers are shown in kDa. (D,E) anti-FOT and anti-FOS were used to detect O-Fuc on mammalian Notch-1 and Thrombospondin-1. HEK293T cells were transiently transfected to overexpress epitope-tagged human Notch1 (FLAG-N1-myc), and co-transfected with the myc-tagged LFNG capping glycosyltransferase as indicated. FLAG- and myc-tags are on the extracellular and intracellular domains of FLAG-N1-myc, respectively. Western blots of total cell extracts, or material captured by immunoprecipitation with anti-FLAG (FLAG-IP), were probed with anti-myc for FLAG-N1-myc, or with anti-FOT or anti-FOS for O-Fuc, followed by Alexa Fluor-800 or Alexa Fluor-680 fluorescent secondary Abs for the protein or O-Fuc, respectively (D). Partial processing of FLAG-N1-myc generated the myc-tagged intracellular domain (NICD) and the fucosylated extracellular domain (NECD). At high laser scanning intensity used to search for anti-FOS labeling, non-specific binding was observed (*). A truncated version of N1 consisting of 5 EGF repeats, or truncated TSP1 with 3 TSR repeats, both epitope-tagged, were transiently overexpressed in normal, POFUT1-null or B3GLCT-null HEK293T cells (E). Secreted truncated proteins were captured from culture media with anti-myc Ab and subjected to western blot analysis as in panel D.