Analysis of close associations of uropod-associated proteins in human T-cells using the proximity ligation assay

Materials and suppliers

• Stromal cell derived factor 1 (SDF-1): Peprotech, Paris, France.

• Bovine serum albumin (BSA): Serva, Germany.

• Gey’s solution contained 138 mM NaCl, 6 mM KCl, 100 µM EGTA, 1 mM Na₂HPO₄, 5 mM NaHCO₃, 5.5 mM glucose and 20 mM Hepes (pH 7.4).

Antibodies

Monoclonal murine antibodies directed against flotillin-2 (Cat. No. E35820) and PSGL-1 (Cat. No. 556053) were obtained from Transduction Laboratories/BD Pharmingen, Germany. A polyclonal rabbit antibody directed against phospho ezrin (Thr567)/radixin (Thr564)/moesin (Thr558) (Cat. No. 3141) was from Cell Signaling Technology. Polyclonal rabbit antibodies directed against flotillin-2 (Cat. No. F1680) or flotillin-1 (Cat. No. F1180), a monoclonal murine antibody recognizing the hemaglutinin tag (HA) (clone HA-7) and a FITC-conjugated antibody directed against murine IgG (Cat. No. F5387) were obtained from Sigma. A monoclonal murine antibody specifically recognizing β-cytoplasmic actin was kindly provided by C Chaponnier (Dugina et al., 2009).

Construct

For preparation of N-terminally HA-tagged PIPKIγ90, a construct coding for wild-type PIP5KIγ661 N-terminally tagged with EGFP in a pcDNA3.1 vector (Lokuta et al., 2007) was used as a PCR template. Then the PCR product was cloned into the phCMV2 Vector (Genlantis).

Isolation of human T-lymphocytes

Resting T-lymphocytes were isolated from buffy coats of healthy donor blood using the Pan T Cell Isolation Kit II (Miltenyi Biotec) and separation on LS columns (Miltenyi Biotec) as described previously (Martinelli et al., 2013). The buffy coats were obtained from the Central Laboratory of the Swiss Red Cross, Bern, Switzerland. The resulting cell suspension contained >95% T-lymphocytes as assessed using anti-CD3 staining. The cells were used after overnight incubation in RPMI with 10% FCS at 37°C and 5% CO₂.

Immunofluorescence staining

T-cells were incubated as described in the figure or table legends, followed by fixation with TCA or PFA and staining for the indicated proteins as described (Affentranger et al., 2011).

Transient transfections of T-lymphocytes

For transfections, 3–6 × 10⁶ freshly isolated T-lymphocytes were resuspended in 100 µl human T cell nucleofector solution (Amaxa, Köln, Germany) diluted 1:2 with PBS and 1 µg of plasmid DNA was added. Then, the cell suspension was transferred to a cuvette and nucleofection was carried out (Amaxa Nucleofector, program U-14). Immediately, 500 µl of medium with 20% FCS was added and the cells were transferred to a prewarmed 12-well plate containing 2.5 ml of medium with 20% FCS, followed by incubation at 37°C in a CO₂ incubator for 4 h. Transfected cells were subsequently washed, resuspended in Gey’s solution and used for experiments.

Proximity ligation assay

The in situ PLA (kit obtained from Olink Bioscience; www.olink.com) was used to detect protein-protein interactions in fixed cells. Nontransfected or transfected cells were treated without or with SDF-1, as described in the figure legends, followed by fixation with TCA or PFA as described (Affentranger et al., 2011), incubation with primary antibodies, incubation with the PLA probes (anti murine and anti-rabbit IgG antibodies conjugated with oligonucleotides), ligation and amplification according to the manufacturer’s instructions. Imaging was performed on fixed samples with a confocal laser scanning microscope Olympus Fluoview FV1000-IX81, 60 × oil immersion objective. For determination of the fraction of cells with at least one red fluorescent dot per cell 100 cells were evaluated per sample and experiment by microscopical analysis. For the evaluation of the specific number of dots per cell and the number of dots per uropod, pictures of representative optical sections of cells with positive PLA were used (the numbers of the cells analyzed per experiment and condition are indicated in the text).

Interaction of flotillin-1 and -2 in freshly isolated human T-cells analyzed with the proximity ligation assay

Endogenous flotillin-1 and -2 showed marked colocalization in resting and chemokine-stimulated freshly isolated human T-cells (Fig. 1), as shown previously (Affentranger et al., 2011; Baumann, Affentranger & Niggli, 2012). Resting cells were mainly spherical, with a punctate, membrane-associated location of both flotillin-1 and -2. Upon stimulation of cells with the chemokine SDF-1 for 15 min, the majority of the cells polarized, correlating with exclusive location of both flotillins at the plasma membrane at the tips of the uropods (Fig. 1A). Our previous results using FRET in cells transfected with tagged flotillin-1 and -2 suggest that preformed flotillin-1 and -2 heterooligomers coalesce in chemokine-stimulated T-cells (Baumann, Affentranger & Niggli, 2012). In the present study, we have investigated interactions of endogenous flotillin-1 and -2, using the PLA, as a positive control. The in situ PLA is based on fixed cells incubated with primary murine and rabbit antibodies reacting for example with flotillin-1 and -2 respectively, followed by incubation of samples with modified secondary antibodies reacting with murine or rabbit IgG antibodies. These secondary antibodies are conjugated with oligonucleotides (PLA probe minus and PLA probe plus). Annealing of the probes occurs when the target proteins are in close proximity (less than 30–40 nm distance), which then initiates the amplification. The amplicons can be detected as red dots by fluorescence microscopy (Söderberg et al., 2006). We obtained a strong PLA signal in the majority of the T-cells using primary murine and rabbit antibodies to flotillin-2 and -1 respectively. This signal was located randomly along the plasma membrane in resting cells (range: 1–11 dots per cell; mean: 4 ± 1 dots per cell; 58 cells analyzed in 3 experiments), and aligned linearly around the entire border of the uropod (at least 4 dots per uropod) in 65% (n = 2; 86 cells analyzed) of the stimulated cells, corresponding to the location of endogenous flotillins (Fig. 1B; top panels: lower magnification; lower panels; higher magnification). These data are in agreement with our FRET studies indicating heterooligomerization of tagged flotillin-1 and -2 (Baumann, Affentranger & Niggli, 2012). Very few cells with one red dot corresponding to a positive PLA reaction per cell were detected when the samples were only incubated with the flotillin-1 antibody (Fig. 1C).

Interactions of P-ERM with PSGL-1 and of flotillins with PSGL-1 and P-ERM in T-cells studied using PLA

We studied in situ interactions of endogenous flotillins with the adhesion receptors PSGL-1 and activated phosphorylated ERM (P-ERM) proteins, and of PSGL-1 with P-ERM in fixed human T-cells. Immunofluorescence pictures indeed show partial or extensive colocalization of PSGL-1 with P-ERM (Fig. 2A) and of flotillins with PSGL-1 (Fig. 3A) and P-ERM (Fig. 4A) in resting T-cells and in the uropod of stimulated T-cells. We then analysed whether these colocalizations correlate with close interactions using PLA in human T-cells. As a positive control we studied the well established direct interaction between PSGL-1 and P-ERM using primary antibodies specifically recognizing PSGL-1 and P-ERM respectively which work well in immunofluorescence (Fig. 2A). As expected from previous findings (Ivetic & Ridley, 2004), we obtained positive PLA signals for PSGL-1 and P-ERM in 94 ± 2% of resting and 87 ± 3% (n = 3) of chemokine-activated cells (Fig. 2B). In resting cells the dots indicating close proximity of the proteins were randomly located at the cell periphery (range: 4–20 dots per cell; mean: 12 ± 1 dot per cell, analysed in 60 cells derived from 3 experiments). In stimulated cells the dots lined the entire border of the uropod in 55 ± 5% (n = 3) of the polarized PLA-positive cells (a total of 248 cells analysed). The remainder of the polarized PLA-positive cells featured 1–2 dots/uropod. A negative control where the samples were only incubated with the P-ERM antibody is shown in Fig. 2C.

A positive PLA reaction was also observed for PSGL-1 and flotillin-2 in resting and chemokine-activated T-cells, confirming and extending the data obtained in human neutrophils using co-immunoprecipitation of flotillin-2 and PSGL-1 (Rossy et al., 2009). Here we obtained positive PLA signals in 83 ± 2% (n = 4) of the resting cells and 88 ± 2% (n = 4) of the chemokine-stimulated T-cells (Fig. 3B), with fluorescent dots located at the plasma membrane of the resting cells (range: 1–11 dots per cell; mean: 4 ± 1 dots per cell analysed in 30 cells derived from 3 experiments), and along the entire uropod border in 67% (n = 2; 198 cells analysed) of the polarized, PLA-positive stimulated cells. The remainder of the polarized PLA-positive cells featured 1–2 dots/uropod. Negative controls with only the anti-PSGL-1 antibody are shown in Fig. 3C.

The PLA of flotillin-2 and P-ERM was also positive in 88 ± 1% of the resting T-cells (range: 1–6 dots per cell; mean: 3 ± 1 dots per cell analysed in 59 cells derived from 2 experiments), and in 54 ± 8% of the stimulated cells. Especially in the stimulated cells the number of dots per cell was clearly lower as compared to the result obtained for PSGL-1 and flotillin, indicating weaker, possibly transient interactions (Fig. 4B). Only 15 ± 7% (n = 3) of polarized, PLA-positive cells featured dots lining the border of the entire uropod (a total of 261 cells analysed). The remainder of the polarized PLA-positive cells featured 1–2 dots /uropod. Negative controls using only the anti-P-ERM antibody are shown in Fig. 4C.

As a negative control we also applied the PLA assay to β-actin and flotillin-2 (Fig. S1). β-Actin is mainly located in protrusions at the front of polarized T-cells. Only small amounts of β-actin are detectable in the uropod (Fig. S1A). Indeed we observed only in 22 ± 5% of the resting cells (n = 3) and in 33 ± 5% (n = 3) of the stimulated cells weakly positive PLA signals (1 ± 0 dots/cell; analysed in 68 cells with positive PLA derived from 3 independent experiments) for β-actin and flotillin-2. Both the % of cells with positive PLA and the number of dots per cell are thus clearly lower for this pair of antibodies as compared to the data shown in Figs. 1–5. In stimulated cells these dots were located outside of the uropod in 67 out of 68 inspected cells with positive PLA, derived from 3 independent experiments (Fig. S1B). Considering the uropod, this is thus a control showing zero response. The few dots detected outside of the uropod may be explained by an incidental close contact of occasional flotillin molecules located at the plasma membrane outside of the uropod with the abundant β-actin. The negative control with only the anti-β-actin antibody is shown in Fig. S1C.

Interactions of PIPKIγ90 with P-ERM studied using PLA

GFP-tagged PIPKIγ90 accumulates in the uropod of murine T-cells (Lokuta et al., 2007; Mathis et al., 2013). We expressed an N-terminally HA-tagged PIPKIγ90 construct in human T-cells and observed colocalization of this construct with P-ERM in the uropod (Fig. 5A). We studied possible interactions of PIPKIγ90 with P-ERM using the PLA. As we have no highly specific antibodies to PIPKIγ90 available which work well in immunofluorescence staining, we transfected cells with PIPKIγ90 tagged with HA. The PLA was carried out with murine anti-HA antibodies and rabbit antibodies reacting with P-ERM. To visualize the transfected cells, we incubated the fixed cells with a FITC-tagged anti-murine IgG antibody that detects the HA antibody, together with the PLA probes. The results are shown in Fig. 5B (top panels: lower magnification; bottom panels: high magnification). We observe an extensive close association of PIPKIγ90 with P-ERM at the plasma membrane in 98 ± 1% of the transfected resting cells (range: 9–24 dots per cell; mean: 15 ± 1 dots per cell derived from 23 cells in 4 experiments). Similarly the PLA was positive in 89 ± 7% (n = 3) of the transfected polarized cells, mostly restricted to the uropods and lining the uropods in 75 ± 7% (n = 3) of the cells (Fig. 5B). The remainder of the polarized PLA-positive cells featured 1–2 dots/uropod. Dots indicating a positive PLA occurred exclusively in the transfected cells as can be appreciated from the top panels, where both transfected and untransfected cells are shown in the same picture (Fig. 5B). Negative controls carried out only with the anti-HA antibody are shown in Fig. 5C. A less extensive interaction was detected for PIPKIγ90 and flotillin-2 (data not shown).