Expression constructs encoding for human flotillin-1 or flotillin-2 were a gift from V. Niggli (University of Bern). TCRζ-PSCFP2 was provided by Prof. K. Gaus (University of New South Wales). PA-mCherry expression backbone was obtained from Clontech. TCRζ PA-mCherry was made as described in . mCherry-Rab5 was a gift from Gia Voeltz (Addgene plasmid # 49,201). mCherry-Rab11a was a gift from Michael Davidson (Addgene plasmid # 55,124). pBGPa-CMV-GFP-OSBP PH domain was a gift from Tim Levine (Addgene plasmid # 58,724). LACT-C2 and 2xFYVE expression constructs were a gift from Prof. Rob Parton (University of Queensland).
Jurkat T cells (Clone E6.1) and flotillin-1 and flotillin-2 knock-out Jurkat cell lines were cultured in RPMI 1640 medium (Pan Biotech) supplemented with 10% (vol/vol) FCS (Gibco). Flotillin-1 and flotillin-2 knock-out Jurkat cell lines were generated as described in . Cells were transfected with 1 μg DNA per 200,000 cells, 18–20 h prior to imaging using the Neon electroporation kit (Invitrogen).
Before imaging, cells were incubated for 10 min at 37 °C on 18-mm glass-coated surfaces (Marienfeld, #0,117,580) that were prepared by incubating with poly-L-lysine (Sigma, #P8920) for 30 min at room temperature, then 10 μg/ml anti-CD3ε (eBioscience, #16–0037) and anti-CD28 (eBioscience, #16–0289) antibodies for T cell activation. For live cell imaging, cells were imaged from 10 to 40 min after their deposition on the coverslips.
For imaging transferrin internalisation, transfected Jurkat T cells were activated for 5 min, transferred onto ice, media exchanged for fresh cold media with 25 μg/ml transferrin-Alexa488 (Jackson ImmunoResearch, #009–540-050) or transferrin-Alexa647 (Jackson ImmunoResearch, #009–600-050) added, incubated for 5 min, media exchanged twice with fresh cold media, then transferred to 37 °C and imaged.
Anti-CD3ε (clone OKT-3; eBioscience, #16–0037), anti-CD28 (clone CD28.2, eBioscience, #16–0289), anti-pPAK1/2 (1:1000, Cell Signaling Technologies (CST #2601 T)), anti-pZap70 (1:1000, CST #2701P), anti-pLAT (1:1000, CST#3584S), anti-pSrc family (1:1000, CST#2101), anti-pSLP-76 (1:1000, Abcam (#ab75829)), anti-pVav1 (1:1000, #ab47282), anti-CD11/CD18-FITC (1 μg/ml, #ab13219), anti-beta-actin (1:5000, #ab6276), anti-CD69-PE (1:100, Serotec), anti-CD25-APC (1:100, Biolegend, #302,610), anti-TfR-biotin (clone OKT-9; eBioscience, #13–0719-82), anti-TfR (clone M-A712, 10 μg/ml, BD Biosciences #555,534), anti-CD18-PE (1:50, BD Biosciences #555,924) and anti-CD29-PE (1:75, clone HUTS-21; BD Biosciences, #556,049).
Preparation of primary human cells
Blood donation for research purposes was approved by the local ethics committee and individual donors gave written consent. PBMCs were enriched using the Vacutainer CPT system (BD Biosciences, #362,782). Then, T cells were enriched by negative magnetic sorting using the Pan T cell Isolation Kit (Miltenyi Biotec, #130–096-535). Pan T cells were used for experiments directly after isolation or cultured in RPMI, supplemented with 1% Penicillin/Streptomycin and 5% FCS if not indicated differently. For experiments involving expanded primary T cells, 1 × 106 isolated T cells per ml were incubated with anti-CD3/anti-CD28-coated Dynabeads (Invitrogen, #11161D) in 1:1 ratio and 30 U/ml IL-2 were added. After 3 days, Dynabeads were magnetically removed and 1 × 106 cells per ml were cultured for 4 more days with 30 U/ml IL-2.
Fixed and live-cell confocal microscopy were performed on a Leica SP5 (Leica Microsystems) or Zeiss LSM780 laser-scanning confocal microscope (Zeiss) that are equipped with an argon laser (405, 488 nm), a diode pump solid state laser (561, 647 nm) and a live-cell incubation chamber (Pecon). GFP constructs and Alexa488-labelled proteins were excited using the 488-nm line of the argon laser source, while PA-mCherry- and mCherry-tagged proteins were excited with the 561-nm laser line. Images were acquired with a 100 × 1.4NA DIC M27 Apo-Plan oil immersion objective (Zeiss) and GaAsP-PMTs in simultaneous, bidirectional scanning mode. For each channel, the pinhole was set to 1 Airy Unit. Confocal photoactivation was achieved by illuminating a region of the cell outer membrane with a 7.2 μW 405-nm laser pulse with 12.24 μs pixel dwell time.
Live-cell TIRF images for vesicle (co-)fusion analysis were acquired on an ELYRA total internal reflection fluorescence microscope (Zeiss) with a 100 × oil-immersion objective with a numerical aperture of 1.46. Images were captured with a 20-ms exposure time.
Live-cell TIRF images for quantification of transferrin in flotillin-compartment were acquired on a DMi8 (Leica Microsystems) equipped with an Infinity TIRF module, a 405-nm diode laser, a 488-nm solid state laser, a 561-nm diode pumped solid state laser and a 638-nm solid state laser. Fast two-channel live cell imaging was performed using W-VIEW GEMINI image splitting optics with filters for spectral separation of GFP and mCherry (Hamamatsu, # A12801-01). Images were acquired with a HC PL APO 100 × /1.47 oil-immersion objective (Leica Microsystems) and fluorescence was detected with a DFC9000GTC sCMOS camera (Leica Microsystems). GFP was excited with the 488-nm laser; transferrin-Alexa647 was excited with the 638-nm laser. TIRF angle was adjusted to 100-nm penetration depth for 488-nm excitation light. Signals from GFP and transferrin-Alexa647 were separated with the W-VIEW GEMINI image splitting optics and simultaneously recorded on one half of the camera chip.
Vesicle count and cross channel nearest neighbour distances for confocal experiments were determined with a custom Matlab vesicle tracking and cross-channel nearest-neighbour distance evaluation software as detailed in . A GUI application and all source code for this analysis is freely available from https://github.com/PRNicovich/PAVesT.git.
Vesicle co-fusion events were visually identified and counted as containing transferrin-Alexa488 if a decrease in mean maximum intensity from the last three frames was greater than 10% compared to the maximum intensity of the initial frame as quantified by the analysed particles measure in FIJI.
Mean dispersion was calculated using a custom FIJI plugin, which evaluates endosome dispersion by calculating the intensity weighted measure of the average pixel distance from the centre of mass of the cell.
Transferrin entry into Rab5/11 endosomal compartments and transferrin content in flotillin-positive compartment was quantified in FIJI by creating a thresholded mask of endosomal compartments using the Rab or flotillin channel, respectively. The background set to zero, then the transferrin channel divided by the Rab/flotillin mask to calculate the mean transferrin intensity present in the desired compartments.
Treatment for transferrin-binding inhibition
Jurkat T cells and expanded primary T cells were treated for 16 h prior to the start of experiments involving transferrin-binding inhibition. Treatment was continued during the experiments. First, cells were transferred into reduced-serum medium (Jurkat: 2% FCS, primary T cells: 5% FCS) to lower the amount of available iron-loaded transferrin. Additionally, in order to block the transferrin receptor (TfR, CD71) and inhibit binding of iron-loaded transferrin, cells were treated with 10 μg/ml anti-TfR (clone M-A712).
Transferrin uptake assay
Jurkat T cells were collected and re-suspended in serum-free RPMI containing 30 μg/ml transferrin-Alexa647 (Jackson ImmunoResearch, #009–600-050) and supplemented (activated) or not (resting) with 1.5 μg/ml anti-CD3ε (OKT-3) + 1 μg/ml anti-CD28 (CD28.2). Cells were then incubated for 0 min or 30 min at 37 °C, cooled down on ice and washed twice with cold FACS buffer (PBS + 2% FCS) to remove unbound transferrin-Alexa647. Subsequently, fluorescence of at least 1 × 104 cells per sample was analysed at a LSR II flow cytometer (BD Biosciences). Samples were kept on ice at all times before measuring. SYTOX Blue (0.1 μM; Invitrogen, #S11348) was added directly before measuring for identification of dead cells. The ratio of MedianFI of activated/resting cells was calculated for 0 min and 30 min timepoints.
To assess efficacy of anti-TfR treatment, Jurkat and primary T cells were treated or not with 10 μg/ml anti-TfR (M-A712) and activated for 20 h on plate-bound anti-CD3ε (OKT-3) and anti-CD28 (CD28.2). Transferrin-Alexa647 (30 μg/ml) was added for the last 90 min of incubation. Subsequently, activated cells were re-suspended, washed twice with PBS and at least 1 × 104 cells per sample were analysed using a LSR II flow cytometer (BD Biosciences).
Jurkat T cells were cultured overnight in RPMI + 2% FCS supplemented or not with 10 μg/ml anti-TfR (M-A712). Next, Jurkat T cells were collected and re-suspended to 1.5 × 106 cells per sample (200 μl) using the corresponding supernatant (ctrl versus anti-TfR). Similarly, primary T cells were pre-treated (RPMI + 5% FCS ± anti-TfR), collected and re-suspended to 4 × 106 cells per sample (200 μl). Activating antibodies (Jurkat: 1.5 μg/ml anti-CD3ε (OKT-3) + 1 μg/ml anti-CD28 (CD28.2); primary T cells: 5ug/ml anti CD3ε (OKT-3) + 5 μg/ml anti-CD28 (CD28.2)) were added to the samples on ice. For activation, samples were incubated at 37 °C in a thermomixer with shaking (300 rpm) for the indicated times. Subsequently, pre-heated 5 × sample buffer (225 mM Tris–HCl pH 6.8, 5% SDS, 50% glycerol, 0.05% bromophenol blue, 4% β-ME) was added (50 μl) to lyse the cells for 10 min at 95 °C. Samples were frozen and thawed once. SDS-PAGE was followed by blotting using the Trans-Blot® Turbo™ Transfer System (BioRad) with 1A/25 V for 30 min onto Nitrocellulose Blotting Membranes (Amersham Protran 0.45 μm, GE Healthcare, #GE10600002). After washing with PBS-T (PBS + 0.02% Tween20), membranes were blocked for 1 h with ROTI-Block solution (Carl Roth, #A151.2). Depending on the target proteins, membranes were cut and subsequently incubated with primary antibody diluted in PBS with 0.05% Tween20 and 3% BSA overnight at 4 °C. Next, membranes were washed 3 times with PBS-T and then incubated for 1 h at RT with the corresponding secondary HRP-conjugated antibody diluted in PBS-T with 5% milk powder. Finally, membrane (-parts) were washed 3 times with PBS-T and then developed using Clarity Western ECL Substrate (Bio-Rad, #1,705,060) or SuperSignal West Femto Substrate (Thermo Scientific, #34,096). Band intensities were quantified relative to the brightest band on the blot and normalised to corresponding beta actin band intensity. Raw images of all used developed membranes are shown in Additional file 2, uncropped blots.
Jurkat cells were seeded at 4 × 105 per ml in RPMI with 2% FCS. Primary human T cells were seeded at 1 × 106 cells per ml in RPMI with 1% Penicillin/Streptomycin and 5% FCS. Before activation cells were left untreated or treated for 24 h with 10 μg/ml anti-TfR (M-A712). To activate Jurkat or primary human cells, they were seeded in flat bottom 96 wells previously coated for 1 h with anti-CD3ε (10 μg/ml; clone OKT-3) and anti-CD28 (10 μg/ml; clone CD28.2). Non-activated (resting) controls were seeded in uncoated wells. After a defined time of activation or incubation in resting conditions, cells were re-suspended, washed once with PBS and stained for 20 min at 4 °C with anti-CD69-PE (1:100), anti-CD25-APC (1:100) or a combination thereof, diluted in FACS-buffer (PBS, 2% FCS). Samples were washed and at least 3 × 103 cells were measured on a LSR II flow cytometer (BD Biosciences) equipped with violet (405 nm), blue (488 nm) and red (633 nm) lasers. SYTOX Blue (Invitrogen, # S34857) was added directly before measuring for identification of dead cells. Lymphocytes were gated based on scatter signals. Then, live cells were selected in a FSC/SYTOX blue dot-plot and median fluorescence intensities of PE and APC channels were determined. For each condition, the MedianFI of the corresponding unstained sample was subtracted from the mean of technical replicates before analysis.
To assess upregulation of CD25 after 96 h, corresponding “resting” cells were stained for surface CD25 and measured prior to activation after 24 h-treatment with anti-TfR. “Activated” cells were measured after 96-h activation on anti-CD3ε (OKT-3)- and anti-CD28 (CD28.2)-coated 96 wells.
TfR surface level detection
Jurkat cells were collected and re-suspended in cold RPMI + 2% FCS to a density of 2 × 106 cells/ml. For “activated” condition, activating antibodies (1.5 μg/ml anti-CD3ε (OKT-3) and 1 μg/ml anti-CD28 (CD28.2)) were added to the cells in a twofold concentrated solution, resulting in a cell number of 1 × 106 per ml during activation. Plain RPMI + 2% FCS was added in the “resting” samples to yield 1 × 106 cells per ml final concentration. In experiments with expanded primary T cells (d7), activation was performed with using 5 μg/ml and anti-CD3ε (OKT-3) and anti-CD28 (CD28.2) final concentration. Activation of cells was allowed for defined durations (5–60 min) by transferring them to 37 °C. Corresponding “resting” cells were incubated at 37 °C as well. Zero-minute timepoint samples were kept on ice. After the incubation at 37 °C, surface TfR was stained using different strategies depending on experimental setup. For experiments presented in Fig. 1, PFA (Polysciences, #18,814–20) was added to the samples to a final concentration of 3.7% and cells were fixed for 20 min at RT. Cells were washed twice and then surface TfR was stained with 2 μg/ml anti-TfR-biotin (OKT-9). Subsequently, samples were washed twice and stained with Streptavidin-Pacific Blue (4 μg/ml; Invitrogen, #S11222). Alternatively, (Fig. 3) cells were washed twice with cold serum-free RPMI and stained subsequently for 30 min in 30 μg/ml labelled transferrin (transferrin-Alexa647) diluted in serum-free RPMI. (Figs. 1 and 3) Before measuring at least 1 × 104 cells with flow cytometry, cells were washed twice with PBS and re-suspended in FACS buffer.
Jurkat T cells or expanded primary T cells (d7) were collected, re-suspended in RPMI + 2% FCS to a density of 2 × 106 cells/ml and incubated for 90 min with biotinylated anti-TfR (2 μg/ml, OKT-9) at 37 °C, to allow uptake of biotinylated anti-TfR (feeding). Subsequent steps, except recycling phase, were performed on ice with cold solutions. After the “feeding” phase cells were collected, washed with RPMI + 2% FCS twice and subsequently incubated with unlabelled Streptavidin (4 μg/ml in PBS) to block surface exposed biotin-anti-TfR. Per treatment one sample was not treated with unlabelled streptavidin to assess surface levels of TfR at t = 0 min. After 15 min, Alexa-488-labelled biotin (1 μg/ml; Sigma, #30,574-1MG-F) was added to the samples to block all remaining accessible binding sites of streptavidin. Then, cells were washed twice and re-suspended in cold RPMI + 2% FCS (“resting”) or in RPMI + 2% FCS, supplemented with activating antibodies (1.5 μg/ml anti-CD3ε (OKT-3) and 1 μg/ml anti-CD28 (CD28.2)) (“activated”). Next, to enable recycling of antibody-labelled TfR, cells were incubated for 5–40 min at 37 °C, or kept on ice (0 min recycling). Cells were washed twice with cold PBS and re-suspended in cold PBS containing Pacific Blue-labelled Streptavidin (4 μg/ml; Invitrogen, # S11222). “Unstained” samples were re-suspended in cold PBS without labelled streptavidin. Staining was performed 20 min at 4 °C. After washing with cold PBS, the samples were resuspended in cold FACS buffer and at least 1 × 104 cells were measured at a LSR II flow cytometer (BD Biosciences). Samples were kept on ice at all times before measuring. Prior to measuring, 0.1 μM TO-PRO-3 (Invitrogen, #T3605) was added for identification of dead cells. For Fig. 1B, the ratio of Pacific Blue MedianFI of activated/resting cells was calculated for all timepoints. To compare TfR recycling between WT and FlotKO cell lines upon activation, Pacific Blue MedianFI relative to t = 0 min was plotted in Fig. 3C.
LIP measurement (Calcein-AM)
Jurkat T cells were incubated overnight in RPMI + 2% FCS supplemented or not with 10 μg/ml anti-TfR. Next, Jurkat T cells were collected, re-suspended to 1 × 107 cells per ml and stained with 1 μM Calcein-AM (Invitrogen, C1430) for 20 min. Then, samples were washed twice with PBS and re-suspended to 1 × 106 cells per ml in RPMI + 2% FCS supplemented or not with 10 μg/ml anti-TfR. Prior to measuring fluorescence intensity of at least 1 × 104 cells at a LSR II flow cytometer (BD Biosciences) using the GFP filter set, 10 μg/ml of anti-CD3ε (OKT3) and anti-CD28 (CD28.2) were added for activation. Dead cells were excluded by addition of 0.1 μM TO-PRO-3 (Invitrogen, #T3605). Median fluorescence intensity of anti-TfR-treated samples was normalised to the intensity of the corresponding control-treated sample.
Jurkat T cells were incubated overnight in RPMI + 2% FCS supplemented or not with 10 μg/ml anti-TfR (M-A712). Next, Jurkat T cells were collected, re-suspended to 2 × 106 cells per ml in the corresponding supernatant and stained with 5 μM CellROX green (Invitrogen, #C10444) for 30 min or left unstained. For activation, 1.5 μg/ml of anti-CD3ε (OKT3) and 1 μg/ml anti-CD28 (CD28.2) were added for the final 5 min of staining. Then, cells were fixed with 4% PFA (Polysciences, #18,814–20) for 15 min, washed twice with PBS and at least 1 × 104 cells were analysed at a LSR II flow cytometer (BD Biosciences) using the GFP filter set.
Adhesion assay on ICAM-I and VCAM-I
Jurkat cells were pre-treated with anti-TfR or left untreated as described above. For the experiment, 96 wells were coated with ICAM-I-Fc (2 μg/ml, diluted in PBS, R&D Biosystems, # 720-IC-050) or VCAM-I-Fc (4 μg/ml, diluted in PBS, Biolegend, #553,706) for 1 h at 37 °C. To determine the exact number of cells added per well, one well was left uncoated to recover all cells after the experiment (input). Coated wells were washed twice carefully with PBS. Subsequently, 1 × 105 cells were added per well in 100 μl and incubated for 15 min to equilibrate temperature and pH. Then, activating mAb cocktail (anti-CD3ε + anti-CD28, 1.5 μg/ml and 1 μg/ml final, respectively), or plain medium or MnCl2 (1 mM final) was added into corresponding wells and incubated 30 min at 37 °C. Input samples were collected for counting directly after incubation. Next, remaining wells were carefully washed twice with medium and adherent cells were dissociated with enzyme-free dissociation buffer (Gibco, #13,151,014). Cells were then resuspended and counted using a LSR Fortessa flow cytometer (BD Biosciences).
Raji B- and (control or anti-TfR treated) Jurkat T cells were stained with 5 μM CellTrace™ Violet (Invitrogen, #C34557 (1 × 106 cells per ml)) or CFSE (Invitrogen, #C34554 (1 × 107 cells/ml)) diluted in PBS for 20 min, respectively, with dye switching between experiments. Next, Raji B cells (1 × 106/ml) were pulsed (+ SEE) or not (− SEE) for 30 min with 2 μg/ml Staphylococcal Enterotoxin E (SEE/Superantigen) (Toxin Technology, #ET404). Subsequently, cells were washed three times with RPMI + 2% FCS. For conjugate formation, stained Jurkat T cells and Raji B cells were mixed in a 1:1 ratio, pelletised for 4 min at 300 × g and incubated at 37 °C for 10 min. For analysis, pellets were carefully resuspended in 100 μl RPMI + 2% FCS, diluted in 200 μl PBS, and at least 1 × 104 events were analysed using a LSR II flow cytometer (BD Biosciences). Because a 1:1 ratio of Jurkat T:Raji B cells was not exactly observed during flow cytometric analysis, the number of conjugates, defined as CTV + / CFSE + double-positive events, are displayed relative to the total cell number (single cells + conjugates) of the cell type which was in minority in the particular experiment. Unspecific conjugates, defined as % conjugates in samples with un-pulsed Raji B cells (− SEE), were subtracted from the corresponding + SEE values.
Staining of integrins
To determine surface levels of LFA-I and VLA-4 in low- and high-affinity conformation, 1 × 105 cells were distributed in round bottom 96 wells in 100 μl and incubated for 15 min to equilibrate temperature and pH. Then, directly labelled antibodies detecting CD18 (PE, Clone 6.7, 1:50 final) or high-affinity-LFA-I (FITC-mAb24 CD11/CD18, 1 μg/ml final) or high affinity CD29-PE (PE-HUTS-21, 1:75 final) were added together with plain assay medium or activating mAb cocktail anti-CD3ε (1.5 μg/ml) + anti-CD28 (1 μg/ml)) or MnCl2 (1 mM final) into corresponding wells and incubated for 10 min at 37 °C. Then, cells were rapidly cooled down on ice and pelletised in a pre-cooled centrifuge (5 min, 350 × g, 4 °C). Supernatant was discarded and cells were washed once with ice-cold PBS. After washing, cells were resuspended in FACS buffer and fluorescence of at least 1 × 104 cells was analysed using a LSR Fortessa flow cytometer (BD Biosciences).
All statistical analyses excepting time of divergence analysis were performed using GraphPad software (Prism v9). Statistical significance between datasets was determined by performing two-tailed, unpaired non-parametric Student’s T-tests, one-way ANOVA, two-way ANOVA or one-sample t-test. Graphs show mean values for either single cells or independent repeats as indicated, and error bars represent the SEM. In statistical analysis, p > 0.05 is not marked or indicated as not significant (n.s.), whereas statistically significant values are indicated by asterisks as follows: *p ≤ 0.05, **p < 0.01, ***p < 0.001, ****p < 0.001.
Time of divergence analysis was performed using a custom MatLab script. To determine the timepoint of divergence, pooled fluorescence intensity distributions of WT and FlotKO#1 were compared by means of the non-parametric Wilcoxon rank-sum test at each timepoint. Briefly, the null hypothesis that WT and FlotKO#1 data sets are from continuous distributions with equal medians is tested against the alternative hypothesis they are not. For each timepoint, a p-value is calculated, with the timepoint of divergence defined as the first timepoint after which all subsequent p-values are equal to or smaller than the considered significance level. *p ≤ 0.05, **p < 0.01.