Partial loss of Sorting Nexin 27 resembles age- and Down syndrome-associated T cell dysfunctions

SNX27 is ubiquitously expressed, complicating the analysis of its role on T cell functions even in heterozygous mice. Therefore, we generated a new transgenic mouse line bearing T cell-specific SNX27 silencing. SNX27 protein expression was depleted from the thymic double positive (DP) stage onwards by crossing mice displaying SNX27 floxed alleles (SNX27^flox/flox) with CD4⁺-Cre transgenic mice. The generated SNX27^+/- heterozygote mice (from now on CD4-Cre-SNX27^fl/+) were kept in a C57BL/6 genetic background. SNX27^flox/flox mice were used as controls (named SNX27^fl/fl). To examine that the inserted SNX27 floxed alleles were susceptible to CD4-Cre cleavage in vivo, we compared SNX27 expression in thymus and ex vivo-differentiated CTL between SNX27^fl/fl and CD4-Cre-SNX27^fl/+ mice. As expected, SNX27 was reduced in cell lysates from thymus and CTL in CD4-Cre-SNX27^fl/+ mice compared to SNX27^fl/fl ones (Fig. 1A). This confirmed the specific depletion of SNX27 in our mouse model. Consistent with reported data from heterozygous SNX27^+/- mice [4], SNX27^fl/fl and CD4-Cre-SNX27^fl/+ age-paired mice displayed similar body weight (Fig. 1B). No differences were found either when analyzing spleen and thymus weights and sizes (Fig. 1C), which correlated with comparable total cellularity in these organs (Fig. 1D). The analysis of thymic populations revealed no major differences in the differentiation of CD4⁺ and CD8⁺ pools (Fig. 1E) (gating strategy shown in Suppl Fig. 1). These results indicate that SNX27-specific deficiency in CD4-Cre-SNX27^fl/+ mice does not affect lymphoid organs development.

Blood analysis revealed normal hematological parameters (Suppl Fig. 2). Nonetheless, in-depth analysis using flow cytometry demonstrated slight but significant differences in the percentage of B cells in CD4-Cre-SNX27^fl/+ mice compared to SNX27^fl/fl (Fig. 2A). Albeit the percentage of total T cells did not change in CD4-Cre-SNX27^fl/+ mice compared to SNX27^fl/fl mice (Fig. 2B), a more detailed analysis revealed diminished percentage of CD4⁺ T cells (Fig. 2C) with a concomitant increase of CD8⁺ T cells (Fig. 2D), resulting in higher CD8⁺/CD4⁺ ratio (Fig. 2E). The rest of the analysis confirmed a similar percentage of classical and inflammatory monocytes as well as granulocytes in both mice genotypes (Fig. 2F-H). Analysis of bone marrow (Fig. 2I) and spleen (Fig. 2J) also showed increased CD8⁺/CD4⁺ ratios in CD4-Cre-SNX27^fl/+ mice compared to those observed in control mice. This finding was confirmed by confocal microscopy after staining and quantification of CD4⁺ cells in lymph nodes from SNX27^fl/fl and CD4-Cre-SNX27^fl/+ mice (Fig. 2K, L). These data resemble published data from individuals with DS and correspond with characteristic features in aging.

A deficit of CD4⁺ T cells could be the result of alterations in the naïve or memory pools. The relative numbers of naïve, central memory (Tcm) and effector/effector memory T cells were determined by the expression of the activation marker CD44 and the adhesion molecule CD62L by flow cytometry analysis [22]. Analysis of these populations in the different organs did not reveal significant differences in circulating blood (Fig. 3 A-C), bone marrow (Fig. 3D-F) or spleen (Fig. 3 G-I), which suggest that the reduction in CD4⁺ T cells expression in CD4-Cre-SNX27^fl/+ mice is not due to an alteration in the distribution of naïve or memory cells populations.

Antigen recognition by naïve T cells elicits metabolic reprogramming to support a rapid phase of cell growth and subsequent proliferation. Previous studies demonstrate that T lymphocytes from SNX27^−/− mice display impaired growth upon TCR triggering [17], consistent with the observation that full Snx27 deletion in mice resulted in smaller animals with reduced organ size [4, 6]. The size of splenocytes from SNX27^fl/fl or CD4-Cre-SNX27^fl/+ mice stimulated ex vivo with anti-CD3/CD28 antibodies for 1, 2, 6, or 7 days was monitored by flow cytometry using the forward scatter (FSC) parameter. Cells reached their maximum size after 2 days of stimulation with no gross differences between SNX27^fl/fl and CD4-Cre-SNX27^fl/+ T cells observed at any time point (Fig. 4A, B). This observation concurs with the normal size described in heterozygous SNX27^+/- mice [4] and suggests that defects in cell growth are associated to total SNX27 loss.

In contrast with defects in antigen-dependent growth, T cells from SNX27^−/− mice showed enhanced proliferation upon stimulation [17]. To determine if partial SNX27 loss affected proliferation, splenocytes from SNX27^fl/fl or CD4-Cre-SNX27^fl/+ mice were stained with CFSE, and proliferating cells were identified by flow cytometry analysis of dye dilution upon stimulation. After 2 days of CD3/CD28 stimulation, we detected a higher proliferation rate in CD4⁺ CD4-Cre-SNX27^fl/+ cells compared to SNX27^fl/fl cells that was even greater for the CD8⁺ T cell pool (Fig. 4C). Detailed quantification confirmed a lower percentage of CD4⁺ and CD8⁺ T cells at the initial stages of division, indicative of faster proliferation (Fig. 4D). A significative increase in the proliferation index was observed for CD4⁺ and CD8⁺T lymphocytes (Fig. 4E). These data indicate that partial SNX27 deficiency in T lymphocytes does not impair initial T cell growth but enhances proliferation.

Accelerated proliferation of CD4-Cre-SNX27^fl/+ T cells could result from enhanced IL-2 production. Notwithstanding, the amount of IL-2 determined in the supernatant of CD4-Cre-SNX27^fl/+ splenocytes 48 h after activation was significantly reduced compared to that determined in SNX27^fl/fl ones (Fig. 4F). Analysis of IFN γ also revealed a significant decrease, although the variability between mice was very high (Fig. 4G). IL-2 is mostly produced by CD4⁺ T cells, reduced IL-2 abundance in the supernatant of total splenocytes could thus result from enhanced consumption by actively growing CD8⁺ T cells. The analysis of purified naïve CD4⁺ T cells confirmed a tendency toward diminished IL-2 (Fig. 4H) and enhanced IFN γ (Fig. 4G) production in CD4-Cre-SNX27^fl/+ CD4⁺ cells compared to controls. These findings indicate that SNX27 partial deficiency moderately affects the capacity of mouse T cells to secrete cytokines under our ex vivo stimulation conditions.

The previous analysis indicates that enhanced proliferation in the absence of augmented IL-2 production could favor expansion of the CD8 pool. Next, we examined the consequences of partial SNX27 deletion during ex vivo T cell activation on the simultaneous expansion of CD4⁺ and CD8⁺ T cells. Splenocytes from SNX27^fl/fl and CD4-Cre-SNX27^fl/+ mice were stimulated with plate-bound anti-CD3 in combination with soluble anti-CD28 and the percentage of CD4⁺ and CD8⁺ T cells was determined by flow cytometry at the indicated times. Both genotypes showed a similar trait with a reduction in the CD4⁺ pool and concomitant massive expansion of the CD8⁺ T cells, as it is generally observed in the absence of exogenous addition of IL-2 (Fig. 5A, B, C). However, the reduction in CD4⁺ cells and the increment of CD8⁺ cells was greater in the CD4-Cre-SNX27^fl/+ mice compared to the SNX27^fl/fl mice (Fig. 5B, C). As a result, the CD8⁺/CD4⁺ ratio was significantly increased at the end of the experiment (Fig. 5D).

The expansion of naïve CD8⁺ cells triggers the onset of the cytotoxic program, which facilitates their differentiation into CTL, key players of the adaptive immune system in the clearance of infected cells [23]. To act against infection, activated CTL release cytotoxic granules that contain a pore-forming protein called perforin, as well as proteases termed granzymes, like granzyme A and B [24]. Given the higher expression of CD8⁺ cells as the result of SNX27 deficiency, we next investigated the consequences of SNX27 partial depletion on CTL differentiation and cytotoxic function. To obtain CTL, CD8⁺ T cells were stimulated with anti-CD3/CD28 and differentiated during 6 days with IL-2. Then, the ex vivo-differentiated CTL from SNX27^fl/fl and CD4-Cre-SNX27^fl/+ mice were stimulated with soluble anti-mouse CD3 and CD28 for the indicated times and cell lysates were analyzed by western blot. Quantification of phosphorylation of ERK at its T202/Y204 residue and phosphorylation of mTORC-1 dependent S6 kinase (S6K) at its Ser240/244 residue showed similar ERK and slight decreased mTOR activation in CD4-Cre-SNX27^fl/+ mice (Fig. 6A). This observation partially resembles the decreased S6K phosphorylation observed in SNX27-silenced Jurkat T cells and splenocytes derived from SNX27 knockout (KO) mice upon CD3/CD28 and Concanavalin A stimulation [17]. Next, we studied the mRNA expression of lytic enzymes and pro-inflammatory cytokines by qPCR in ex vivo-differentiated CTL in resting conditions and 24 h after stimulation with CD3/CD28. CD4-Cre-SNX27^fl/+ CTL showed upregulated expression of Granzyme A compared to SNX27^fl/fl ones (Fig. 6B). Although mRNA levels of Granzyme B, Perforin and Il-2 mRNA also appeared to be increased, these changes were not statistically significant (Fig. 6B). Analysis of Tnfα expression revealed a significant increment in the transcription of this cytokine before stimulation in CD4-Cre-SNX27^fl/+ CTL compared to SNX27^fl/fl CTL (Fig. 6B). Studies in individuals with DS demonstrate enhanced expression of Eotaxin, Mp1 and T-bet by CD8 naïve T cells after stimulation with CD3/CD28 [18, 19], important mediators in inflammation and immunity against pathogens. Analysis of the expression of these molecules showed higher mRNA expression in basal conditions in SNX27 heterozygous CTL (Fig. 6B). Next, we sought out to determine whether the increased translation of lytic enzymes observed in CD4-Cre-SNX27^fl/+ CTL resulted in an altered cytotoxicity. To do so SNX27^fl/fl or CD4-Cre-SNX27^fl/+ CTL were stimulated with plate-bound anti-CD3 and soluble CD28 for 3 h and CD107a surface expression was assessed by flow cytometry, which did not show significant differences between both mice genotypes (Suppl Fig. 7A, B). Analysis of the killing capacity of SNX27^fl/fl and CD4-Cre-SNX27^fl/+ CTL using a lactate dehydrogenase-release colorimetric assay was also comparable between both mice genotypes (Suppl Fig. 7C). These results demonstrate that partial SNX27 deficiency alters the CTL cytotoxic program with enhanced transcription of cytolytic and pro-inflammatory genes but does not affect their killing capacity under our ex vivo experimental setup.

The analysis of CD4-Cre-SNX27^fl/+ mice reveals a loss of peripheral CD4⁺ T cell pool that alters the CD4⁺/CD8⁺ ratio. Early studies demonstrated that long-term maintenance of CD4⁺ T cells in the periphery requires continuous contact with MHC-II molecules that provide survival signals [25]. Studies in human T cell lines demonstrated that SNX27 silencing impairs microtubule organization at the IS [21]. Thus, we reasoned that polarization defects in naïve CD4⁺ T cells could be related to the observed alterations on this population. To mimic the interaction between T cells and antigen-presenting cells, isolated naïve CD4⁺ T cells from SNX27^fl/fl or CD4-Cre-SNX27^fl/+ mice were incubated 15 min with anti-CD3-coated P815 target cells and stained with antibodies that recognize pericentrin and phalloidin, for visualization of the microtubule organizing center (MTOC) and F-actin respectively. Formation of an actin ring at the cell–cell interface was indicative of the establishment of an IS (Fig. 7A). When in contact with target cells, purified CD4⁺ T cells from SNX27^fl/fl mice showed MTOC polarization to the contact area (Fig. 7A, left panel). On the contrary, CD4-Cre-SNX27^fl/+ CD4⁺ T cells failed to polarize the MTOC (Fig. 7A, right panel). Analysis of MTOC polarity indexes confirmed significantly greater values in CD4-Cre-SNX27^fl/+ CD4⁺ T cells (Fig. 7B). While an average of 67% of control CD4⁺ T cells bound to P815 displayed a translocated MTOC, only 23% of CD4-Cre-SNX27^fl/+ CD4⁺ T cells did (Fig. 7C).

The structure of the cytotoxic synapse between a CTL and its target cell resembles the one formed between a CD4⁺ and an antigen presenting cell (APC) [26]. Efficient TCR stimulation leads to targeted polarization towards the cell–cell junction of cytotoxic granules, which are later released into the synaptic cleft [27, 28]. We next investigated the organization of cytotoxic IS by incubating ex vivo-differentiated CTL and anti-CD3-coated P815 target cells (Fig. 7D). In contrast to that observed for naïve CD4⁺ T cells, CTL derived from SNX27^fl/fl mice or CD4-Cre-SNX27^fl/+ were equally able to position the MTOC at the cell–cell interface (Fig. 7D, E, F). These data confirm the relevance of SNX27 for MTOC positioning in naïve CD4⁺ T cells and demonstrate that this requirement is not necessary in CTL.

CD4⁺-specific-SNX27 deficient mice were generated by backcrossing SNX27-floxed mice (SNX27^flox/flox), a kind gift from Dr. Paul Slesinger (University of California, USA), with CD4-Cre transgenic mice. CD4⁺-Cre-SNX27 heterozygotes mice (CD4-Cre-SNX27^fl/+) were kept in a C57BL/6 genetic background. SNX27^flox/flox mice were used as controls. Validation of SNX27-specific depletion was performed by PCR using the following primers: 5′-ACCATAGGGTTGATGAGATGGCCAA-3′ and 3′ ACAGCATGACTTGGCTGACATACGTG-5´. Mice were bred in pathogen-free conditions in the animal facility of the National Centre for Biotechnology and handled according to European and Spanish regulations (2010/63/UE; RD 53/2013).

SNX27^fl/fl and CD4-Cre-SNX27^fl/+ male and female mice were sacrificed at 9–12 weeks of age. Spleen, thymus, and bone marrow were harvested. Spleen and thymus weight and size were determined, and cells from these organs were extracted and total cellularity was analyzed using a cell counter. Bone marrow cells were isolated from the hind legs of the mice (femur and tibia bones) by flushing 10 ml of RPMI medium through the bone using a syringe with 25-gauge needle, and naïve, memory, and effector cell subsets were analyzed.

For lymphocyte isolation, splenocytes were mechanically disaggregated in phosphate-buffered saline (PBS), cells were collected using a 40 µm cells strainer (BD Biosciences) and treated with erythrocyte lysis buffer. For CD4⁺ T cells isolation MojoSort Mouse CD4 T Cell Isolation Kit was used, and cells were cultured in mouse medium: RPMI supplemented with 10% FBS, 2 mM L-glutamine, 1 mM sodium pyruvate, 1% nonessential amino acids, 50 µM β-mercaptoethanol, 10 mM HEPES, and penicillin–streptomycin (100 U/ml), and further used in the experiments. To analyze CTL in CD4-Cre-SNX27^fl/+ mice, mouse CD8⁺ cells were purified from isolated splenocytes using the MojoSort Mouse CD8 T cell Isolation Kit. Then CD8⁺ cells were cultured for 2 days in plates coated with anti-mouse CD3 (2.5 μg/ml), soluble anti-mouse CD28 (1.25 μg/ml) and IL-2 (100 U/ml). Then fresh IL-2 (100 U/ml) was added every two days, and experiments were performed after 6 days.

Individual mice were analyzed separately, and no significant gender effect was observed in any of the experiments performed. Data from male and female mice were pooled for each experimental condition.

The mouse mastocytoma cell line P815, kindly donated by Dr. Andrés Alcover, was used as tumor target cell.

For stimulation of primary T cells, cells were incubated in complete mouse medium (1 × 10⁶ cells/ml) with plate-bound anti-mouse CD3 (2.5 µg/ml) supplemented with anti-mouse CD28 (1.25 µg/ml).

For conjugate formation with CD4⁺ T cells and CTL, the mouse mastocytoma cell line P815 was used to generate stimulatory cells. P815 were stained in complete DMEM medium with 10 μM CMAC (1 h, 37 °C, darkness). Labeled P815 were washed twice in PBS and were coated with anti-mouse CD3 antibody (2.5 µg/ml) (45 min, 37 °C). Coated P815 were then added at 1:1 ratio on top of primary T cells previously plated on poly-L-lysine-coated coverslips. Cells were incubated for the indicated times (37 °C, 5% CO2), after which immunofluorescence was performed.

For analysis of CTL activation, cells (1 × 10⁶ cells/ml) were stimulated with soluble anti-mouse CD3 and CD28 (1 μg/ml). After the incubation period, cells were collected and analyzed by western blot.

Mouse blood was collected from the heart using a syringe with 21-gauge needle and introduced in polypropylene tubes containing ethylene diamine tetra acetic acid (EDTA). Aliquots of whole blood were then analyzed using an Abacus Junior Vet, automated hematological analyzer (CVM SL.). The parameters analyzed were: white blood cells (WBC), lymphocytes, monocytes and granulocytes count and percentage, red blood cells (RBC) counts, hemoglobin concentration, hematocrit (HCT) and platelets count.

For analysis of circulating immune populations in blood samples aliquots of whole blood were incubated with a mix of LIVE/DEAD violet fixable cell stain and mouse-directed anti-CD45-APC, anti-CD3-APC-eFluor780, anti-B220-PECy7, anti-CD8-Bviolet605, anti-CD4-PerCPCy5.5, antiLy6C-FITC and anti-Ly6G-PE antibodies (20 min, RT). Versalyte solution was then added to lyse red cells (10 min, RT, darkness) and the percentage of cells of the following populations was analyzed by flow cytometry: B cells (CD45⁺ CD3⁻ B220⁺), T cells (CD45⁺ CD3⁺ B220⁻), CD4⁺ and CD8⁺ (CD45⁺ B220⁻ CD3⁺), monocytes (CD45⁺ CD3⁻ B220⁻ Ly6C⁺), inflammatory monocytes (CD45⁺ CD3⁻ B220⁻ Ly6C⁺ high) and granulocytes (CD45⁺ CD3⁻ B220⁻ Ly6G⁺). Gating strategy is shown in Suppl Fig. 3.

For analysis of splenocytes, bone marrow cells and thymocytes cells (gating strategies are depicted in Suppl Figs. 1, 4 and 5) were collected in ice-cold PBS and cell surface proteins stained with saturating concentrations of the indicated fluorophore-conjugated primary antibodies in PBS staining buffer (1% FBS, 0.5% BSA, 0.01% sodium azide/PBS) 30 min at 4 ºC. The purity of the isolated CD4⁺ or CD8⁺ T cells was analyzed by flow cytometry by staining with anti-mouse CD3-FITC, anti-mouse CD19-PE, anti-mouse CD8-APC and anti-mouse CD4-PECy7 antibodies. Cells were washed and fixed using 1% PFA and maintained at 4 ºC for flow cytometry using Cytomix FC500, Galios Cytometer or Cytoflex (Beckman Coulter). Cells and singlets were gated using forward and side scatter parameters. LIVE/DEAD violet dead fixable cell strain was used when compatible with the fluorophore panel. Data were analyzed using FlowJo software (TreeStar).

Flat-bottom 96 well-plates were prepared with or without anti-mouse CD3 coating (2.5 μg/mL). Splenocytes were stained in warm PBS with 0.5 μM CFSE (15 min, 37 °C, darkness). Unbound dye was quenched by adding five times the staining volume of FBS-containing medium (5 min, 37 °C). Cells were then washed twice in FBS-containing medium. 100,000 splenocytes were seeded per well and were additionally incubated with soluble anti-mouse CD28 (1.25 μg/mL). Afterwards, cells were washed twice in cold PBS and stained with LIVE/DEAD violet fixable stain, followed by staining with mouse-directed anti-CD4-PECy7, anti-CD8-APC and anti-CD19-PE antibodies. CD4⁺ or CD8⁺ T cells were gated on live, CD19⁻ splenocytes and dye dilution analysis of T cell division was carried out by flow cytometry. Gating strategy is shown in Suppl Fig. 6.

To assess IL-2 and IFNγ cytokine production, total splenocytes or isolated naïve CD4⁺ T cells from SNX27^fl/fl or CD4-Cre-SNX27^fl/+ mice were stimulated 48 h with anti-CD3 (2.5 µg/ml) and soluble anti-CD28 (1.25 µg/ml). Afterwards, the concentration of the depicted cytokines in the supernatant was assessed by sandwich ELISA (Biolegend) according to the manufacturer´s instructions.

For western blot analysis, cells were washed in ice-cold PBS and lysed in NP40 lysis buffer (10 mM HEPES pH 7.5, 15 mM KCl, 1 mM EGTA, 1 mM EDTA, 1% NP40, 10% glycerol) containing protease and phosphatase inhibitors (2 nM NaF, 40 mM β-glycerophosphate, 20 μM leupeptin, 1.5 μM aprotinin, 1 mM PMSF and 1 mM Na₃VO₄). Lysates were clarified and quantified using the Pierce 660 nm Protein Assay. A similar amount of protein per sample was analyzed by SDS-PAGE, transferred to nitrocellulose membranes (BioRad) and incubated with each of the indicated primary antibodies. Detection of fluorescent-conjugated secondary antibodies was carried out on a ChemiDocTM MP Imaging System (BioRad).

Total RNA was isolated from cell cultures (2–5 × 10⁶ cells) using Trizol reagent. Complementary DNA (cDNA) was synthesized from isolated RNA (2 μg) after retrotranscription with Randrom Primers (Applied Biosystems) and Multiscribe Reverse Transcriptase (Applied Biosystems). qPCR reactions were performed with Evagreen qPCR mix plus (Cultek molecular Bioline) in 8 μl final volume. PCR conditions were as follows: 95ºC (10 min), 40 cycles at 95ºC (15 s) and 60ºC (1 min), continued by 95ºC (15 s), 60ºC (1 min) and 95ºC (15 s). Reactions were run in triplicate with the Applied Biosystems QuantStudio 5 System. β-actin was used as control because this gene was stably expressed in all cases. All samples were subjected to a melting curve analysis to verify the single amplification product. Relative expression of each gene was calculated using the ΔΔCt method. qPCR reactions were run in triplicates on each sample, and three biological replicates were performed. Primer sets used were:

Differentiated mouse CTL were stimulated for the indicated time points with coated anti-mouse CD3 (5 μg/ml) and soluble anti-mouse CD28 (2.5 μg/ml) in the presence of CD107a-PE (LAMP1) (2.5 μg/ml). At each time point, cells were harvested in cold PBS and incubated on ice. At the end of the time course, cells were labeled with LIVE/DEAD violet dead cell stain followed by anti-mouse CD8-APC staining. Events were acquired on the flow cytometer. Singlets were gated for live cells, in which surface marker expression was analyzed.

Cytotoxicity was assessed using the Cytotox96 Non-Radioactive Cytotoxic Assay. Briefly, differentiated mouse CTL were resuspended in RPMI supplemented with 3% FBS, 2 mM L-glutamine and 10 mM HEPES. CTL were then co-cultured (4 h, 37 °C) at the indicated ratios with P815 target cells previously coated (45 min, 37 °C) with anti-mouse-CD3 (2.5 μg/ml). Lactate dehydrogenase released by dead target cell was measured at 490 nm using an absorbance microplate reader. The percentage of target cell lysis was calculated following manufacturer´s instructions.

For immunofluorescence labeling, cells were fixed with 2% PFA (15 min, RT). After washing twice with PBS, cells were blocked and permeabilized for 30 min at RT (1% BSA, 0.1% Triton X-100 in PBS). This buffer was also used throughout the procedure as staining buffer. Cells were incubated with the corresponding primary antibodies (1 h, RT), washed with PBS, and incubated with the corresponding fluorophore conjugated secondary antibodies (30 min, RT). Coverslips were washed twice in PBS and mounted on glass slides using ProLong Gold.

Confocal images were acquired using a Leica TCS SP8 confocal microscope equipped with a Plan-Apochromat HCX PL APO 63 × 1.4 NA oil immersion objective or using the Olympus Fluoview FV1000 confocal laser-scanning microscope equipped with a uPlan-Apochromat 60 × 1.4 NA oil objective lens. Images were collected with LAS X (Leica) or the FV10-ASW4.2 (Olympus) acquisition softwares. Sets to be compared were acquired using the same acquisition settings. To quantify protein, fluorescence images were processed using ImageJ software (NIH).

For CD4 localization assays, z-series optical Sects. (0.5 μm intervals) were collected and maximal intensity Z-projections of 4–6 contiguous optical planes, which included the three-dimensional fluorescence information, were stacked. The mean fluorescence intensity (MFI) was measured in the whole cell in basal conditions.

To measure the ability of the MTOC to translocate to the IS, cell z-stacks (0.2-µm intervals) were acquired, and maximum intensity projections were generated using the Fiji software [41]. The geometric center of MTOC (MTOC^C) as well as the IS region were determined. The polarity index was computed by dividing the distance between MTOC^C to the IS (“a” distance) by the distance from the IS to the distal pole of the T cell (“b” distance). This allowed polarity indexes to be normalized to cell size and shape. We considered polarization to occur when the polarity index was < 0.25. Polarity index values are represented in graphs as dot plots, where each dot represents an individual cell.

Statistical analysis was performed with GraphPad Prism 8 software, and samples were assumed to fit a normal distribution. Details about the data presentation, the experimental replication, and the statistical tests used are included in the individual figure legends. Briefly, an unpaired student’s t-test was used to analyze differences between two conditions. Two-way ANOVA with a Bonferroni post-test was applied for multiple comparisons. The level of statistical significance is represented by * p < 0.05; ** p < 0.01; *** p < 0.001; ****p < 0.0001. Data are shown as mean ± standard error of the mean (SEM) unless otherwise specified.