The endocytic activity of the flagellar pocket in Trypanosoma brucei is regulated by an adjacent phosphatidylinositol phosphate kinase

Abstract

Phosphoinositides are spatially restricted membrane signaling molecules. Phosphatidylinositol 4,5-bisphosphate [PI(4,5)P2]--a phosphoinositide that is highly enriched in, and present throughout, the plasma membrane--has been implicated in endocytosis. Trypanosoma brucei has one of the highest known rates of endocytosis, a process it uses to evade the immune system. To determine whether phosphoinositides play a role in endocytosis in this organism, we have identified and characterized one of the enzymes that is responsible for generating PI(4,5)P2. Surprisingly, this phosphoinositide was found to be highly concentrated in the flagellar pocket, the only site of endocytosis and exocytosis in this organism. The enzyme (designated TbPIPKA, annotated as Tb927.10.1620) was present at the neck of the pocket, towards the anterior-end of the parasite. Depletion of TbPIPKA led to depletion of PI(4,5)P2 and enlargement of the pocket, the result of impaired endocytosis. Taken together, these data suggest that TbPIPKA and its product PI(4,5)P2 are important for endocytosis and, consequently, for homeostasis of the flagellar pocket.

Keywords: Endocytosis; Flagellar pocket; PI(4,5)P2; PIP kinase; Trypanosoma brucei.

Fig. 1.

Multiple sequence alignment of PIP kinase domains. The sequences of ScMss4p, HsPIPKIα and HsPIPKIIα were aligned with the sequences of the T. brucei proteins Tb427.04.1620, Tb427.10.4770 and Tb427.10.3890 using MAFFT software. Only the PI4P 5-kinase Pfam domain is shown here. Conserved residues are highlighted with a gray background. Residues implicated in ATP-binding are indicated with a red background, amino acids thought to be in the catalytic core are marked in green. The region within the activation loop, which determines substrate specificity, is marked with a red box and suggests that Tb427.04.1620 is more closely related to type I, rather than type II, PIP kinases. Tb427.04.1620 is designated TbPIPKA throughout the manuscript.

Fig. 2.

Overexpression of TbPIPKA increases PIP₂ levels. PCF cells were stably transfected with pLew100-Ty1-TbPIPKA. Expression of the fusion proteins was induced using the indicated amounts of doxycycline (Dox) for 15 hours. (A) Tagged proteins from total cell lysates were detected using monoclonal antibodies against Ty1. (B) Lipids were extracted from 8×10¹⁰ cells and subjected to HPLC analysis. The peak area ratio of PIP₂ to PIP was calculated and an increase in the PIP₂∶PIP ratio was observed following increased expression of Ty1–TbPIPKA. The basal ratio in the absence of doxycycline is likely caused by the activity of endogenous PIP kinases. Results are presented as the mean of three independent experiments±s.e.m. (C) Overlaid output traces obtained during HPLC analysis of lipids from cells that had been transfected with pLew-Ty1-TbPIPKA growing in the absence (−Dox) or presence (+Dox) of 100 ng/ml doxycycline. Note that the increase in PIP₂ is accompanied by a decrease in PIP, suggesting that Tb427.04.1620 is a PIP kinase.

Fig. 3.

Depletion of TbPIPKA by RNAi. (A) TbPIPKA depletion (+Dox) caused a growth arrest in PCF and BSF cells. Data points are the means±s.e.m., n = 3. (B) Depletion of TbPIPKA does not cause a cell-cycle progression defect in PCF (left panel) or BSF (right panel) cells. Samples of control (−Dox) and induced-RNAi (+Dox) from PCF and BSF cells were taken after 2 and 3 days. Nuclear (N) and kinetoplast (K) DNA was labeled with DAPI to determine the cell cycle state. At least 300 cells were analyzed for each timepoint across three independent experiments. Values are mean±s.e.m., n = 3. 1N1K, one nucleus, one kinetoplast etc. (C) TbPIPKA mRNA depletion following induction of RNAi. RNA was isolated from PCF (top panel) and BSF (bottom panel) cells after growth in the absence or presence of 100 ng/ml doxycycline for 3 days. Tubulin served as the loading control, taq (tubulin) was used to determine DNA contamination. (D) Extracted lipids from TbPIPKA-depleted (+Dox) PCF cells were analyzed by HPLC and the PIP₂∶PIP peak area ratio was plotted. Three independent experiments, values are mean±s.e.m.

Fig. 4.

TbPIPKA localization in PCF and BSF cells. (A) Schematic representation (not to scale) of the flagellar neck region showing the location of marker proteins and TbPIPKA. Top view, left: A single flagellum, surrounded by its own flagellar membrane (FM), nucleates from the basal body (BB) and exits the cell body through the flagellar pocket (FP) at a site termed the flagellar pocket collar (FPC). The plasma membrane (PM) subdomain between the flagellar pocket and that part of the plasma membrane exposed to the environment is termed the flagellar pocket neck (FPN). Top view, middle and right: TbBILBO1 localizes to the flagellar pocket collar. The two arms of the bilobe are labeled by TbMORN1 and TbCentrin4, respectively. TbCentrin4 also localizes to the basal bodies. The flagellar pocket can be labeled with tetramethylrhodamine–Concanavalin A (rhodamine–ConA). Top view, right: The flagellar attachment zone (FAZ) tracks the long axis of the flagellum. (B) PCF cells and (C) BSF cells stably expressing YFP–TbPIPKA from its endogenous promoter. YFP–TbPIPKA was labeled using antibodies against GFP. (a–e) Cells were co-labeled with (a) rhodamine–ConA or antibodies against (b) TbMORN1, (c) TbCentrin4, (d) TbBILBO1 or (e) the FAZ-specific monoclonal antibody L3B2. Nuclear and kinetoplast DNA was labeled with DAPI (blue). Right column in each panel: fluorescence images from the areas indicated were superimposed on differential interference contrast (DIC) images. YFP–TbPIPKA appears to localize to the flagellar pocket neck region, mostly posterior to TbCentrin4, above TbBILBO1 and at the tip of the FAZ. Immunofluorescence images are compressed z-stacks. Scale bars: 5 µm.

Fig. 5.

Distribution of PI(4,5)P₂ using YFP–2xPH^PLCδ as a sensor. (A–C) PCF cells were stably transfected with YFP–2xPH^PLCδ under the control of the strong constitutive PARP promoter. (A) Cells were incubated with 3 µg/ml Hoechst stain (blue) for 25 minutes and imaged live. YFP–2xPH^PLCδ localization (green) was predominantly observed at the flagellar pocket membrane. Some YFP–2xPH^PLCδ was also detected at the plasma membrane. (B) Cells were co-labeled with rhodamine–ConA before fixation. Note that YFP–2xPH^PLCδ localization surrounds the flagellar pocket, as marked by rhodamine–ConA. Note the apparent diffusion of YFP–2xPH^PLCδ in fixed cells relative to living cells. (C) Cryosections of control cells and cells that had been stably transfected with YFP–2xPH^PLCδ. Sections were labeled with antibodies against GFP (10-nm gold). Note the labeling of the flagellar pocket membrane, and the absence of labeling over the ER or plasma membrane (arrowheads, middle panel). The right-hand panel shows that PI(4,5)P2 is not present at the ER exit site (ERES) and Golgi. Labeling of multivesicular bodies (MVBs) was variable (dashed lines). Arrows in the right-hand panel indicate gold particles marking the position of YFP–2xPH^PLCδ. (D) PCF cells stably expressing HA–TbPIPKA from its endogenous promoter and YFP–2xPH^PLCδ from the PARP promoter, labeled after fixation by using an antibody against HA. Note that TbPIPKA is anterior to the YFP–2xPH^PLCδ-labeled pocket. Nuclear and kinetoplast DNA was labeled with DAPI (blue). Scale bars: 5 µm (A,B,D); 500 nm (C).

Fig. 6.

Depletion of TbPIPKA alters the levels of PI(4,5)P₂ at the flagellar pocket. (A,B) TbPIPKA-RNAi cells were stably transfected with YFP–2xPH^PLCδ expressed under the control of the PARP promoter. TbPIPKA depletion was induced for 3 days (+Dox). Cryosections of control (−Dox; left hand panel) and induced (+Dox; right hand panel) cells were labeled with antibodies against GFP (10-nm gold). FP: flagellar pocket. Scale bars: 500 nm. (C–E) Flagellar pocket profiles from control (n = 73) and TbPIPKA-depleted (n = 73) cells were analyzed. Statistical significance (P-values) is indicated in the histograms. The box represents the 25th–75th percentiles, the line represents the median and the whiskers encompass 95% of the data points. (C) The number of gold particles per flagellar pocket profile was counted manually. (D) The circumference of flagellar pocket profiles was measured using ImageJ software. Depletion of TbPIPKA led to a small, but statistically significant, increase in the number of cells that had enlarged flagellar pockets. (E) The ratio of gold particles∶circumference (µm) was plotted. A strong decrease in the linear density (gold particles/µm) of the labelling of YFP–2xPH^PLCδ at the flagellar pocket was observed.

Fig. 7.

Depletion of TbPIPKA causes an increase in flagellar pocket volume. (A–E) PCF and (F–J) BSF TbPIPKA-RNAi cells were induced for 3 days and analyzed by serial block-face scanning electron microscopy. Single slices of (A) control (–Dox) and (B) induced (+Dox) PCF cells are shown. To estimate the enlargement of the flagellar pocket, a representative flagellar pocket in each reconstruction was outlined in red. The flagellar pocket circumference was tracked using the IMOD 3Dmod software in 300 acquired sections: 45 flagellar pockets from control cells (C) and 57 flagellar pockets from TbPIPKA-depleted cells (D) were modeled. (E) The volume of the flagellar pockets was calculated using 3Dmod software and plotted as a histogram. Single slices of control (F) and induced (G) BSF cells are shown. The flagellar pocket volumes of 115 control cells (H) and 158 TbPIPKA-depleted cells (I), from a total of 1700 sections, were modeled and plotted as a histogram. (J) 6% of TbPIPKA-depleted cells with extremely large flagellar pocket (nine cells) were removed from the data set and the remaining values plotted as a histogram. Scale bars: 2 µm.

Fig. 8.

TbPIPKA depletion negatively affects endocytosis but not export. (A) Endocytosis of AF633–transferrin in BSF cells. TbPIPKA-RNAi was induced for three days. The ligand uptake was calculated by determining the relative fluorescence mean values of the FACS data. The relative fluorescence values for control cells (−Dox) were set to 100% and the fluorescence values for the depleted cells (+Dox) were normalized to these values. Six independent RNAi experiments were performed. The values shown are mean±s.e.m. The induction of TbPIPKA-RNAi resulted in a reduced rate of receptor-mediated endocytosis of transferrin. (B) Export of VSG in BSF cells. Radiolabeled surface and internal VSG was recovered using ConA–Sepharose beads. The data were compiled from three independent experiments; the values shown are the mean±s.e.m. The depletion of TbPIPKA (right panel) did not significantly affect the half-time for VSG delivery to the cell surface.