T. brucei transferrin receptor expressed from BES7 contains GPI-anchored ESAG7.

(A) Phylogenetic tree showing the relationship between ESAG6 and ESAG7 proteins from T. brucei Lister 427 (Hertz-Fowler et al., 2008). The tree was assembled in Geneious Prime (www.geneious.com) using an alignment by Muscle 5.1 followed by a neighbour joining tree with 100 bootstraps and no outgroup on residues 1 to 332 of the alignment. BES7 ESAG 6 and 7 are shown in red. (B) Variation in transferrin receptor expression in cell lines expressing VSGs from different bloodstream form expression sites (BES). BES3-VSG118, BES1-VSG221, BES7-VSG224 cell lines are unmanipulated and the BES7-VSG221 was made after exchanging VSG224 for VSG221 (see Figure 2A). Cell lysates were treated with PNGaseF and 2 x 106 cells equivalents were loaded. Western blots were probed with anti-TfR, and then overprobed with anti-VSG221. ESAG6 and ESAG7 plus GPI-anchor variants are indicated; BES3-VSG118 and BES1-VSG221 cell lysates contain ESAG7 without the GPI anchor. BES3-VSG118 and BES7-VSG224 cell lysates did not show any VSG221. The presence of VSG221 protein confirmed the switching of BES7 VSG224 to VSG221. Switching in VSG type did not affect the expression pattern of native TfR (ESAG6/7 plus GPI) protein. Coomassie blue stained SDS-PAGE gel to demonstrate loading of samples.

BES identity and the number of GPI-anchors determines the cellular TfR levels.

(A) Schematic illustrating the production of cell lines. For the addition and removal of ESAG7 GPI-anchor addition sequences, two constructs were used for each of BES1 and BES7. A description of the construct is provided in the methods. For BES7, the same process was used, one construct removing the GPI-anchor from ESAG7, the other leaving it intact. For BES1, integration of one resulted in modification of ESAG7 to add a GPI-anchor, integration of the second resulted in an unchanged ESAG7 polypeptide sequence. Restriction enzyme sites used in cloning: P, PacI; E, EcoRI; S, SacI; Bs, BstEII and Bb, BsrBI. The VSG in BES7-VSG224 cells was switched to VSG221 to remove any VSG-dependent variation. The targeting construct conferred G418 resistance. (B) Transferrin receptor expression in the following cell lines: BES7-VSG221 single (1-GPI) and double (2-GPI) GPI-anchored TfR and BES1-VSG221 single and double GPI-anchored TfR. Cell lysates were treated with PNGaseF and 2 x 106 cells equivalents were loaded per gel track. A western blot of the samples was probed with anti-TfR then over probed with anti-VSG221 and anti-SCD6. ESAG6 and ESAG7 plus GPI-anchor variants are indicated. (C) TfR quantitation estimates from a minimum of three biological replicates of each cell line, error bars display the standard deviation for each sample. Samples plus a dilution series from BES7-221 were minimally separated by SDS-PAGE to compress TfR components to a single band for analysis after western blotting, probed with anti-TfR and anti-SCD6. Donkey anti-rabbit Alexa 488 was used as secondary antibody and band intensity was calculated by ImageJ ensuring that signals taken for each reading were below saturation levels. Relative quantitation was determined by mapping values to those obtained from a dilution series. Measurements for SDC6 were used to normalise cell loading with the value for BES1-1GPI anchor samples given an arbitrary value of 1. Statistical analysis was used a two-tailed paired t-test, ∗∗p ≤ 0.01; ∗p ≤ 0.05.

Transferrin endocytosis is rapid, in both single and double GPI-anchored cells. Initial binding is to the cell surface and internalization is detected within 60 sec.

Cultures of single (1-GPI) or double (2-GPI) GPI-anchored TfR expressing BES7-VSG221 and BES1-VSG221 cells were supplemented with 500 nM Alexa568 labelled holo-transferrin (TfA568) and incubated for 15 sec, 60 sec or 600 sec and fixed by addition of formaldehyde to 1% for 5 min. Following washing with PBS containing 1% bovine serum albumin cells were visualized under Zeiss Axio Imager.Z2 with X100 objective. Images in all these experimental time points for both 1- and 2GPI TfR cells were captured with same intensity of excitation and exposure time. (A) Images showing the location of bound or endocytosed transferrin in representative cells in populations and at time points as indicated. Scale bar is 5 µm. Images were processed with ZEN Blue 3.4 software without any deconvolution. Larger numbers of cells are shown in Supplementary Figure 3. (B) The variation between individual cells in transferrin binding and uptake and endocytosis shown by measurements of total fluorescence intensity of TfA568 in BES7 VSG221 1- and 2-GPI TfR cells at 15 sec, 60 sec and 600 sec. Measurements were obtained using Zen Blue 3.4 software and normalizing the background fluorescence. Scatter plots of measurements of individual cells are shown for 1-GPI (grey) and 2-GPI (purple) TfRs. The average and standard deviation are shown for each cell line. (C) TfA568 fluorescence intensity distribution across transverse sections of randomly selected cells was measured using ImageJ software. The plots are shown for both BES7 1- and 2-GPI cell lines, showing the pattern of individual cells in different colours at 60 sec and 600 sec. (D) (E) and (F) Same as (A) (B) and (C) but with BES1-VSG221 1- and 2GPI TfR expressing cells.

Immunofluorescence localisation of single and double GPI-anchored transferrin receptors to the entire plasma membrane.

Cell lines expressing VSG221 and either 1- or 2-GPI TfR variants were fixed in culture with final concentrations of 4% formaldehyde and 0.2% glutaraldehyde. Immunofluorescence localisations were performed with rabbit anti-TfR followed by donkey anti-rabbit Alexa 488 (green). Cells were visualized using a Zeiss Axio Imager.Z2 with X100 objective. Images in all these experimental sets were captured with same intensity of excitation and exposure time. (A) Localization in cell lines as indicated. Scale bar is 5 µm. Images were processed with ZEN Blue 3.4 software. A larger number of cells is shown in Supp. Figure 5. (B) Fluorescence distribution across transverse sections of the cells was measured using ImageJ software. Plots are shown for both BES7-VSG221 TfR cell lines, each cell in a different colour. (C) Mean anti-TfR fluorescence intensity in individual cells of both BES1 VSG221 and BES7 VSG221 1- and 2-GPI TfR cell population was measured using ZEN Blue 3.4 and normalized against the background fluorescence. Bar diagram shows the fluorescence intensity of either 1-GPI (grey) or 2-GPI (purple) TfR from individual cells. The columns represent the average and the error bars show the standard deviation.

Localization of single and double GPI-anchored transferrin receptors under 3D structured illumination microscope.

Cells expressing either 1- or 2-GPI-anchored TfR were fixed in culture with final concentrations of 4% formaldehyde and 0.2% glutaraldehyde. Immunofluorescence localisation was performed with rabbit anti-TfR and donkey anti-rabbit Alexa 488. Cells were visualized using a Zeiss Elyra7 microscope with X63 objective. Images in all these experimental sets were captured with same intensity of excitation and exposure time. (A) Localization of TfR in different cell lines as indicated. Scale bar is 5 µm. (B) 3D projection generated using ZEN Blue 3.4 software example cells from the cell lines as indicated. These images show the localization pattern of TfR on the cell surface as well as within intracellular compartments. X-Y-Z axis is indicated.

Haptoglobin-haemoglobin endocytosis assay. Initial binding is to the cell surface followed by rapid endocytosis.

Cells were supplemented with 500 nM Alexa 568 labelled HpHb, incubated for 15 sec, 60 sec or 600 sec and fixed in culture with 1% formaldehyde. Following washing with PBS containing 1% bovine serum albumin. Cells were visualized under Zeiss Axio Imager.Z2 with x100 objective. Images in all these experimental time points were captured with same excitation intensity and exposure time. (A) Images showing the location of HpHb in +/+ and -/- cells at different time points in representative cells as indicated. Images were processed with ZEN Blue 3.4 software. Scale bar is 5 µm. A greater number of cells is shown in Supp. Figure 6. (B) HpHb fluorescence intensity distribution over transverse section of cells at 15 sec measured using ImageJ software. Individual cells are shown in different colours.

The complement factor H receptor localises to the cell surface.

The observation that abundant FHR expression occurred in cells grown in rabbit serum but not in FBS supplemented medium was used to investigate its localisation. (A) Western blot analysis of FHR expression in T. brucei EATRO1125 cells grown in medium supplemented with either foetal bovine serum or rabbit serum. Cell lysates were not treated with PNGaseF and 2×106 cells equivalent were loaded. The blot was probed with rabbit anti-FHR and then overprobed with anti-SCD6 for loading. (B) Immunofluorescence determination of localisation of FHR in cell lines as indicated. Cells were fixed in culture with final concentration of 4% formaldehyde and 0.2% glutaraldehyde. Rabbit anti-FHR antibody was followed by donkey anti-rabbit Alexa 488 secondary antibody. A set of cells grown in RbS, fixed similarly and stained only with secondary antibody was used as a control for residual rabbit IgG from the growth medium and cells grown in FBS for non-expressers. Cells were visualized using a Zeiss Axio Imager.Z2 with x100 objective. Scale bar is 5 µm. A larger number of cells is shown in Supp. Figure 7 (C) Localisation of FHR by 3D structured illumination microscopy. Cells fixed as above were visualised using a Zeiss Elyra7 with x63 objective. Images were processed with ZEN Blue 3.4 software. Two representative cells images showing FHR in white and merged panel with DNA in blue. Scale bar is 5 µm.

Relationship between ESAG7 variants.

A. Clustal W alignment of the nucleotide sequences of ESAG7s showing the translation termination codon in red. The sequence for GPI-addition is present in all genes but a series of small deletions between nucleotides 980 and 1080 resulted in a frame shift and a new stop codon that effectively removed the C-terminal ∼50 residues from the open reading frame. B. Same alignment but just including ESAG7 from BES1 and BES7 showing the conservation of the coding potential to the 3’ side of the stop codon in BES1 ESAG7 indicating the shorter ESAG7 genes probably arose from a longer BES7-like variant. Functional stop codons are in red.

Relationship between ESAG 6 and 7.

Clustal W alignment of BES7 ESAG6 and 7 nascent polypeptide sequences. Predicted processing sites are indicated by yellow highlighting. The N-terminal signal sequence cleavage site was predicted using SignalP 6.0 and GPI-anchor addition sites by NetGPI 1.1 as detailed below the alignment.

Transferrin endocytosis is rapid, in both single and double GPI-anchored cells. Initial binding is to the cell surface and internalization is detected within 60 sec.

Larger fields of cells from the experiments shown in Figure 3. Scale bar is 20 µm.

T. brucei BES7 VSG221 1- and 2-GPI transferrin receptor does not bind canine transferrin.

(A) Clustal W alignment of the polypeptide sequences of cow and dog transferrin. (B) Cultures were supplemented with 500 nM Alexa568 labelled canine holo-transferrin and incubated for 60 sec and fixed in culture by addition of formaldehyde to 1%. Following washing with PBS containing 1% bovine serum albumin cells were visualized under Zeiss Axio Imager.Z2 with x100 objective. Images for both BES7 VSG221 1- and 2-GPI TfR cells were captured with same intensity of excitation and exposure time. Images showing the absence of any bound or endocytosed canine transferrin in around forty cells in population of both 1- and 2GPI TfR cells. Scale bar is 20 µm. Images were processed ZEN Blue 3.4 software.

Immunofluorescence localisation of single and double GPI-anchored transferrin receptors to the entire plasma membrane.

Larger fields of cells from the experiments shown in Figure 4. Scale bar is 20 µm.

Haptoglobin-haemoglobin endocytosis assay. Initial binding is to the cell surface followed by rapid endocytosis.

Larger fields of cells from the experiments shown in Figure 6. Scale bar is 20 µm.

The complement factor H receptor localises to the cell surface.

Larger fields of cells from the experiments shown in Figure 7B. Scale bar is 20 µm.