Modulation of the Surface Proteome through Multiple Ubiquitylation Pathways in African Trypanosomes

Abstract

Recently we identified multiple suramin-sensitivity genes with a genome wide screen in Trypanosoma brucei that includes the invariant surface glycoprotein ISG75, the adaptin-1 (AP-1) complex and two deubiquitylating enzymes (DUBs) orthologous to ScUbp15/HsHAUSP1 and pVHL-interacting DUB1 (type I), designated TbUsp7 and TbVdu1, respectively. Here we have examined the roles of these genes in trafficking of ISG75, which appears key to suramin uptake. We found that, while AP-1 does not influence ISG75 abundance, knockdown of TbUsp7 or TbVdu1 leads to reduced ISG75 abundance. Silencing TbVdu1 also reduced ISG65 abundance. TbVdu1 is a component of an evolutionarily conserved ubiquitylation switch and responsible for rapid receptor modulation, suggesting similar regulation of ISGs in T. brucei. Unexpectedly, TbUsp7 knockdown also blocked endocytosis. To integrate these observations we analysed the impact of TbUsp7 and TbVdu1 knockdown on the global proteome using SILAC. For TbVdu1, ISG65 and ISG75 are the only significantly modulated proteins, but for TbUsp7 a cohort of integral membrane proteins, including the acid phosphatase MBAP1, that is required for endocytosis, and additional ISG-related proteins are down-regulated. Furthermore, we find increased expression of the ESAG6/7 transferrin receptor and ESAG5, likely resulting from decreased endocytic activity. Therefore, multiple ubiquitylation pathways, with a complex interplay with trafficking pathways, control surface proteome expression in trypanosomes.

Fig 1. The effect of AP-1γ RNAi on markers of the endocytic pathway.

All experiments were performed in uninduced (-) or induced (+) cells for 18 hours. (A) Western blot and (B) immunofluorescence for GLP-1, CatL, MFST^12myc, (C) ISG65 and ISG75. In B and C scale bar = 2.5 μm, protein antigens are shown in red, DNA visualised with DAPI is in blue. (D) AP-1γ RNAi cells were subjected to RNAi knockdown followed by cycloheximide treatment. Cells were harvested at various time points and endogenous ISG65 and ISG75 levels were monitored by western immunoblotting. β-tubulin was used as a loading control. Lower panel shows quantification, with open symbols induced (+) and closed symbols uninduced (-) and representing the mean of three independent experiments, with the standard error indicated.

Fig 2. TbUsp7 but not TbVdu1 knockdown perturbs endocytosis.

(A) Knockdown of TbUsp7 and TbVdu1 in the presence (+, open bar) or absence (-, closed bar) of RNAi induction as measured by qRT-PCR. Results were normalised to β-tubulin and error bars denote the standard error of the mean (n = 3). (B) Phase contrast images showing the presence or absence of the BigEye phenotype under TbUsp7 and TbVdu1 knockdown. The enlarged flagellar pocket is indicated by a black arrowhead. Scale bar = 2.5 μm. (C) One hundred cells were analysed from uninduced or induced cultures of TbVdu1 (open bar) or TbUsp7 (closed bar) and scored for the appearance of an enlarged flagellar pocket, manifest as a phase-light vacuole at the posterior end of the cell. Results represent an average of four independent experiments with error bars denoting the standard error. (D) ConA uptake in the presence and absence of TbUsp7 knockdown. Following induction of RNAi, cells were incubated with FITC-conjugated ConA (green). Note that only an example of a cell with an enlarged flagellar pocket is presented for TbUsp7. Induced cells that do not display this morphology have normal ConA uptake. Scale bar = 2.0 μm. All cells were co-stained with DAPI (blue) to visualise nuclear and kinetoplast DNA.

Fig 3. Clathrin knockdown increases ubiquitylation of ISG75.

(A) Levels of clathrin heavy chain (CHC), ISG65 and ISG75 were monitored following RNAi knockdown of clathrin for 12 hours by western immunoblotting. β-tubulin was used as loading control and relative protein abundance was quantified by densitometry. (B) Endogenous ISG65 and ISG75 were immunoprecipitated using specific antibodies followed by western immunoblotting with anti-ubiquitin antibody. A western immunoblot from the same samples prior to immunoprecipation and probed with anti-β-tubulin was used as a loading control.

Fig 4. Knockdown of TbUsp7 and TbVdu1 differentially reduces steady state ISG65 and ISG75 levels.

(A) Steady state levels of endogenous ISG65 and ISG75 in the presence (+) or absence (-) of TbUsp7 and TbVdu1 knockdown were assessed by western immunoblotting. β-tubulin was used as a loading control. Bar graphs represent the mean of three independent experiments normalised to uninduced cells at 100% (dotted line), with the standard error indicated. (B) The location and levels of ISG65 and ISG75 were assessed by immunofluorescence in the presence or absence of TbUsp7 or TbVdu1 knockdowns (+ or -, respectively). Cells with enlarged flagellar pockets (“BigEye”) in the presence of TbUsp7 RNAi are also shown. Cells were stained with DAPI (blue) to visualise nuclear and kinetoplast DNA, while ISG65 and ISG75, visualised with specific polyclonal antibodies, are shown in red. Scale bar = 2.5 μm.

Fig 5. ISG65 and ISG75 turnover under TbUsp7 and TbVdu1 knockdown.

(A) TbUsp7 and (B) TbVdu1 RNAi knockdown was followed by cycloheximide treatment. Cells were harvested at various time points and endogenous ISG65 and ISG75 levels were monitored by western immunoblotting. β-tubulin was used as a loading control. Closed symbols indicate uninduced (-) and open symbols induced (+) cultures. Graphs represent the mean of three independent experiments, with the standard error indicated.

Fig 6. Assessment of ISG65 and ISG75 transcription and biosynthesis under TbUsp7 and TbVdu1 knockdown.

(A) ISG65 and ISG75 transcript levels were determined in the presence (+, open bars) or absence (-, closed bars) of TbUsp7 and TbVdu1 knockdown by qRT-PCR normalised to β-tubulin. Error bars denote standard error of the mean. (B) Biosynthesis of ISG65 and ISG75 were monitored by immunoprecipitation using specific polyclonal antibodies, in the presence (+) or absence (-) of TbUsp7 and TbVdu1 knockdown. Bar graphs represent the mean of three independent knockdown experiments (open bars) normalised to uninduced (closed bars) cells, with the standard error indicated. Statistical analysis: Student’s t-test; *p<0.05, **p<0.01.

Fig 7. DUB-knockdown mediated expression level changes depend on cytoplasmic domain sequences.

BiPN-ISG75L is a chimeric reporter construct containing the BiP ATPase N-terminal domain fused to an HA epitope tag followed by the trans-membrane and cytoplasmic domain of ISG75. Cells were grown in the presence (+) or absence (-) of TbUsp7 RNAi for 48 hours followed by treatment with cycloheximide for the indicated times. The levels of HA-BiPN-ISG75L and endogenous ISG75 were assessed at different time points by western immunoblotting with anti-HA or anti-ISG75 antibodies. β-tubulin was used as a loading control.

Fig 8. Volcano plots of protein abundance changes.

Normalised SILAC ratios, averaged from duplicates (TbUsp7) or triplicates (TbVdu1), are plotted against the respective -log10 transformed standard deviation. Data points representing protein groups significantly shifted after 48 hours are labeled (see also Table 1). ISG65 and ISG75 paralogs are highlighted in green and orange, respectively.

Fig 9. Model integrating suramin-sensitivity pathways, trafficking and ISG turnover.

A simplified schematic of the trypanosome endomembrane system is shown, with the flagellar pocket at top left. Teal arrows indicate ISG degradative and recycling trafficking routes, red putative AP-1-mediated transport from the Golgi complex to the lysosome [14, 27] and gray exocytic/biosynthetic pathways. The predominant locations of ISG75, ESAG6/7 (the transferrin receptor) and p67 (the major lysosomal protein) are indicated by icons. Evidence suggests that ISG75 is ubiquitylated at, or close to the surface (magenta) and deubiquitylation by TbUsp7 and/or TbVdu1 is proposed to take place prior to the sorting step at the early endosome that selects for the recycling or degradative arm of the endocytic system. TbVdu1 is known to associate with structures in this region, whilst TbUsp7 is likely cytosolic. Approximate percent changes in the proportion of ISG75 transiting the different arms of the endocytic system are calculated from differences in half-life derived for observed ISG75 turnover changes upon TbUsp7 knockdown (Fig 5) and assuming a recycling cycle time of ~10 min. AP-1 is proposed to mediate pathways required to deliver components to the lysosome required for suramin to translocate to the cytosol.