Article CCR5 deficiency impairs CD4+ T-cell memory responses and antigenic sensitivity through increased ceramide synthesis Ana Martín-Leal1,†, Raquel Blanco1,†, Josefina Casas2,3, María E Sáez4, Elena Rodríguez-Bovolenta5, Itziar de Rojas6, Carina Drechsler7,8,9, Luis Miguel Real10,11, Gemma Fabrias2,3, Agustín Ruíz6,12, Mario Castro13 , Wolfgang WA Schamel7,8,14 , Balbino Alarcón5 , Hisse M van Santen5 & Santos Mañes1,* Abstract Introduction CCR5 is not only a coreceptor for HIV-1 infection in CD4+ T cells, The C-C motif chemokine receptor 5 (CCR5) is a seven-trans- but also contributes to their functional fitness. Here, we show that membrane G protein-coupled receptor (GPCR) expressed on the by limiting transcription of specific ceramide synthases, CCR5 surface of several innate and adaptive immune cell subtypes, signaling reduces ceramide levels and thereby increases T-cell including effector and memory CD4+ T lymphocytes (Gonzalez- antigen receptor (TCR) nanoclustering in antigen-experienced Martin et al, 2012). CCR5 acts also a necessary coreceptor for mouse and human CD4+ T cells. This activity is CCR5-specific and infection by HIV-1. An HIV-resistant population served to identify independent of CCR5 co-stimulatory activity. CCR5-deficient mice a 32-bp deletion within the CCR5 coding region (ccr5D32), which showed reduced production of high-affinity class-switched anti- yields a non-functional receptor (Blanpain et al, 2002). Since bodies, but only after antigen rechallenge, which implies an ccr5D32 homozygous individuals are seemingly healthy, a radical impaired memory CD4+ T-cell response. This study identifies a body of thought considers that CCR5 is dispensable for immune CCR5 function in the generation of CD4+ T-cell memory responses cell function. and establishes an antigen-independent mechanism that regulates Experimental and epidemiological evidence nonetheless indicates TCR nanoclustering by altering specific lipid species. that CCR5 has an important role in innate and acquired immune responses. CCR5 and its ligands C-C motif ligand 3 (CCL3; also Keywords ccr5[delta]32; humoral response; membrane phase; sphingolipid; termed macrophage inflammatory protein [MIP]-1a), CCL4 (MIP- T-cell receptor 1b), CCL5 (regulated upon activation, normal T cell expressed and Subject Category Immunology secreted [RANTES]), and CCL3L1 have been associated with exacer- DOI 10.15252/embj.2020104749 | Received 18 February 2020 | Revised 12 May bation of chronic inflammatory and autoimmune diseases. Despite 2020 | Accepted 14 May 2020 | Published online 11 June 2020 varying information due probably to ethnicity effects (Lee et al, The EMBO Journal (2020) 39: e104749 2013; Schauren et al, 2013), further complicated in admixed popula- tions (Toson et al, 2017), epidemiological studies support the See also: C Matti & DF Legler (August 2020) 1 Department of Immunology and Oncology, Centro Nacional de Biotecnología (CNB/CSIC), Madrid, Spain 2 Department of Biological Chemistry, Institute of Advanced Chemistry of Catalonia (IQAC-CSIC), Barcelona, Spain 3 CIBER Liver and Digestive Diseases (CIBER-EDH), Instituto de Salud Carlos III, Madrid, Spain 4 Centro Andaluz de Estudios Bioinformáticos (CAEBi), Seville, Spain 5 Department of Cell Biology and Immunology, Centro de Biología Molecular Severo Ochoa (CBMSO/CSIC), Madrid, Spain 6 Alzheimer Research Center, Memory Clinic of the Fundació ACE, Institut Català de Neurociències Aplicades, Barcelona, Spain 7 Signaling Research Centers BIOSS and CIBSS, University of Freiburg, Freiburg, Germany 8 Department of Immunology, Faculty of Biology, University of Freiburg, Freiburg, Germany 9 Institute for Pharmaceutical Sciences, University of Freiburg, Freiburg, Germany 10 Unit of Infectious Diseases and Microbiology, Hospital Universitario de Valme, Seville, Spain 11 Department of Biochemistry, Molecular Biology and Immunology, School of Medicine, Universidad de Málaga, Málaga, Spain 12 CIBER Enfermedades Neurodegenerativas (CIBERNED), Instituto de Salud Carlos III, Madrid, Spain 13 Interdisciplinary Group of Complex Systems, Escuela Técnica Superior de Ingeniería, Universidad Pontificia Comillas, Madrid, Spain 14 Centre for Chronic Immunodeficiency (CCI), University of Freiburg, Freiburg, Germany *Corresponding author. Tel: +34 91 585 4840; Fax: +34 91 372 0493; E-mail: smanes@cnb.csic.es † These authors contributed equally to this work. ª 2020 The Authors. Published under the terms of the CC BY 4.0 license The EMBO Journal 39: e104749 | 2020 1 of 19 The EMBO Journal Ana Martín-Leal et al ccr5D32 allele as a marker for good prognosis for these overreactive T-cell function by inhibiting the synthesis of ceramides, which are immune diseases (Vangelista & Vento, 2017). In contrast, ccr5D32 identified here as negative membrane regulators of TCR nanoscopic homozygotes are prone to fatal infections by several pathogens such organization. as influenza, West Nile, and tick-borne encephalitis viruses (Lim & Murphy, 2011; Falcon et al, 2015; Ellwanger & Chies, 2019). The mechanisms by which the ccr5D32 polymorphism affects all these Results pathologies have usually been linked to the capacity of CCR5 to regulate leukocyte trafficking. For example, CCR5 deficiency reduces CCR5 deficiency impairs the CD4+ T-cell memory response recruitment of influenza-specific memory CD8+ T cells and acceler- ates macrophage accumulation in lung airways during virus rechal- To determine the role of CCR5 in CD4+ memory T-cell generation lenge (Dawson et al, 2000; Kohlmeier et al, 2008); this could lead to and/or function, we adoptively transferred congenic CD45.1 mice acute severe pneumonitis, a fatal flu complication. CCR5 nonethe- with lymph node/spleen cell suspensions from OT-II WT or CCR5/ less has migration-independent functions that maximize T-cell acti- mice (CD45.2) and subsequently infected them with OVA-encoding vation by affecting immunological synapse (IS) formation (Molon vaccinia virus; 5 weeks post-immunization, we analyzed spleen et al, 2005; Floto et al, 2006; Franciszkiewicz et al, 2009) as well as CD45.2+ donor cells from OT-II mice. CCR5 expression on OT-II cells T-cell transcription programs associated with cytokine production affected neither the total number of memory CD4+ T cells (Fig 1A (Lillard et al, 2001; Camargo et al, 2009). CCR5 and its ligands are and B) nor the percentage of CD4+ TEM (CD44hi; CD62L; Fig 1C) or also critical for cell-mediated immunity to tumors and pathogens, TCM (CD44hi; CD62L+; Fig 1D) cells generated. OT-II WT cells including HIV-1 (Dolan et al, 2007; Ugurel et al, 2008; González- nonetheless had stronger responses to antigenic restimulation than Martı́n et al, 2011; Bedognetti et al, 2013). OT-II CCR5/ memory T cells, as determined by the percentage of Whereas the role of CCR5 in T-cell priming is well established, interferon (IFN)c-producing cells after ex vivo stimulation with its involvement in memory responses has not been addressed in OVA323–339 (Fig 1E). depth. Only a single report suggested CCR5 involvement in CD4+ We also studied T cell-dependent B-cell responses in WT and T-cell promotion of memory CD8+ T-cell generation through a CCR5/ mice after immunization with the hapten 4-hydroxy-3- migration-dependent process (Castellino et al, 2006). It remains iodo-5-nitrophenylacetyl coupled to ovalbumin (NIP-OVA; Fig 1F). unknown whether CCR5 endows memory T cells with additional We detected no difference in the percentage or absolute number of properties. One such property is the elevated sensitivity of effector T follicular helper (Tfh) cells (CD4+, CD44hi, CXCR5+, PD1+) and memory (“antigen-experienced”) CD4+ and CD8+ T cells to between WT and CCR5/ mice at 7 days post-immunization their cognate antigen compared to naı̈ve cells (Kimachi et al, 1997; (Fig 1G–I). At day 30, half of the mice were boosted with the same Kersh et al, 2003; Huang et al, 2013). This sensitivity gradient NIP-OVA immunogen (OVA/OVA) and the other half received NIP (memory >> effector > naı̈ve) in CD8+ T cells is linked to increased conjugated with another carrier protein (OVA/KLH); levels of NIP- valency of preformed T-cell antigen receptor (TCR) oligomers at the specific high- and low-affinity immunoglobulins (Ig) were analyzed cell surface, termed TCR nanoclusters (Kumar et al, 2011). This anti- 15 days later. Comparison of the humoral responses between OVA/ gen-independent TCR nanoclustering (Schamel et al, 2005, 2006; OVA- and OVA/KLH-immunized mice would assess the effect of Lillemeier et al, 2010; Sherman et al, 2011; Schamel & Alarcon, memory CD4+ T cells specific for the first carrier protein on the 2013) enhances antigenic sensitivity by increasing avidity to multi- humoral response to NIP. There were no differences in high/low- meric peptide-major histocompatibility complexes (Kumar et al, affinity NIP-specific IgM production between WT and CCR5/ mice 2011; Molnar et al, 2012) and by allowing cooperativity between with either immunization strategy (Fig 1J and K). CCR5 deficiency TCR molecules (Martı́nez-Martı́n et al, 2009; Martı́n-Blanco et al, markedly impaired the generation of high-affinity class-switched 2018). TCRb subunit interaction with cholesterol (Chol) and the pres- anti-NIP antibodies specifically in OVA/OVA-immunized mice ence of sphingomyelins (SM) are both essential for TCR nanocluster- (Fig 1J and K). Since class switching was similar in WT and ing (Molnar et al, 2012; Beck-Garcia et al, 2015). Replacement of CCR5/ OVA/KLH-immunized mice, our results suggest that CCR5 Chol by Chol sulfate impedes TCR nanocluster formation and deficiency reduces the generation of high-affinity class-switched reduces CD4+CD8+ thymocyte sensitivity to weak antigenic peptides immunoglobulins due to deficient memory CD4+ T-cell function. (Wang et al, 2016). Whether antigen-experienced CD4+ T-cell sensi- tivity is linked to TCR nanoscopic organization and the homeostatic The CCR5 effect on antigen-experienced CD4+ T cells is factors that regulate TCR nanoclustering remains unexplored. cell-autonomous Given its co-stimulatory role in CD4+ T cells, we speculated that CCR5 signals would affect the antigenic sensitivity of CD4+ memory To test whether the in vivo memory defect associated with CCR5 T cells. To test this hypothesis, we analyzed the function of in vivo- deficiency was intrinsic to CD4+ T cells, we activated OT-II WT and generated memory CD4+ T cells in wild-type (WT) and CCR5/ CCR5/ spleen T cells with OVA323–339 antigen for 3 days; after mice, and the effect of CCR5 deficiency on CD4 T-cell help in the T- antigen removal, we cultured cells with IL-2 or IL-15. OT-II cells that dependent humoral response. We found that CCR5 is necessary for differentiated in exogenous IL-2 expressed CCL3, CCL4, CCL5, and a the establishment of a functional CD4 memory response through a functional CCR5 receptor, as determined by their ability to flux mechanism independent of its co-stimulatory role for the TCR Ca2+ and migrate after CCL4 stimulation (Appendix Fig S1A–D). signal. We show that CCR5 deficiency does not affect memory CD4 Like CD8+ T cells (Richer et al, 2015), OT-II cells cultured with T-cell generation, but reduces their sensitivity to antigen. Our data IL-15 showed a memory-like phenotype (Fig EV1); they were demonstrate an unreported CCR5 regulatory role in memory CD4+ smaller than IL-2-cultured cells and retained CD62L with reduced 2 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal A 10 5 B 10 C 100 p = 0.5 p = 0.4 CD4+ Vα2+ cells (x106) 104 8 80 TEM cells (%) 6 60 103 TEM TCM 4 40 102 2 20 WT 1 10 0 0 105 D 100 E 10 ** WT 80 CCR5-/- IFNγ+ cells (%) 104 8 TCM cells (%) 60 6 103 p = 0.5 CD44-PECy5 TEM TCM CD4-ef450 40 4 102 SSC 20 2 CCR5-/- 101 100 101 102 103 104 105 0 0 101 102 103 104 105 101 102 103 104 105 CD45.2-FITC Vα2-PE CD62L-APC G 105 H I F p = 0.3 104 10 4 p = 0.3 103 TFH cells/spleen (x105) 8 NIP-OVA 102 3 TFH cells (%) NIP-OVA 10 1 WT TFH 6 Electron 2 microscopy. 105 4 -/- Ig ELISA. CD44-APC WT / CCR5 104 1 2 PD1-ef780 FACS TFH NIP-KLH 103 Day 0 Day 7 Day 30 Day 45 102 0 0 WT CCR5-/- WT CCR5-/- 101 CCR5-/- TFH 101 102 103 104 105 101 102 103 104 105 CD4-PB CXCR5-PeCy7 J High affinity K Low affinity OVA/OVA OVA/KLH OVA/OVA OVA/KLH 3 * * *** WT 3 *** * CCR5-/- Absorbance (A.U.) Absorbance (A.U.) 2 2 1 1 0 0 M 1 2a 2b 3 A M 1 2a 2b 3 A M 1 2a 2b 3 A M 1 2a 2b 3 A G G Ig G G Ig Ig G G G G Ig G Ig Ig G Ig Ig G Ig Ig G Ig Ig G G G G Ig Ig Ig Ig Ig Ig Ig Ig Ig Ig Ig Ig Class-switched Igs Class-switched Igs Class-switched Igs Class-switched Igs Figure 1. CCR5 deficiency impairs CD4+ T-cell memory responses. A Representative plots of splenocytes from CD45.1 mice adoptively transferred with CD45.2 OT-II WT or CCR5/ lymph node cell suspensions, 5 weeks after infection with rVACV-OVA virus. The gating strategy used to identify the memory CD4+ T-cell subtypes is shown (n = 5). B Absolute number of OT-II cells recovered in spleens of mice as in A (n = 5). C, D Percentage of CD4+ TEM (C) and TCM (D) in the OT-II WT and CCR5/ populations (n = 5). E IFNc-producing OT-II WT and CCR5/ memory cells isolated from mice as in (A) and restimulated ex vivo with OVA323–339 (1 lM) (n = 4). F Immunization scheme for NIP-OVA and NIP-KLH in WT and CCR5/ mice. G–I Representative plots (G) and quantification of the frequency (H) and absolute number (I) of Tfh cells (CD4+CD44+PD-1+CXCR5+) in the spleen after primary immunization (day 7) with NIP-OVA (n = 7). J, K ELISA analysis of high- (J) and low-affinity (K) isotype-specific anti-NIP antibodies in sera from OVA/OVA- and OVA/KLH-immunized mice (day 15 post-challenge; n = 5 mice/group). Data representative of one experiment of two. Data information: (B–E, H–K), Data are mean  SEM. *P < 0.05, **P < 0.01, ***P < 0.001, two-tailed unpaired Student’s t-test. ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 3 of 19 The EMBO Journal Ana Martín-Leal et al activation marker expression (CD25, CD69, CD44) compared to IL- CCR5 modulates TCR nanoclustering in antigen-experienced 2-cultured T cells (Fig 2A). Findings were similar in OT-II WT and CD4+ T cells CCR5/ cells (Fig 2B), which reinforced the idea that CCR5 is not involved in CD4+ T memory cell differentiation. Restimulation of The high antigenic sensitivity of antigen-experienced CD8+ T cells IL-2- or IL-15-expanded OT-II lymphoblasts with the OVA323–339 was partially attributed to increased TCR nanoclustering (Kumar peptide nonetheless indicated that CCR5-expressing cells showed et al, 2011). To determine whether CCR5 deficiency influences TCR strong proliferation and higher IL-2 production at low antigen organization, we used electron microscopy (EM) to analyze surface concentrations than CCR5-deficient cells (Fig 2C–F), indicative of an replicas of OT-II WT and CCR5/ naı̈ve cells and lymphoblasts increased number of cells responding to antigenic stimulation. CCR5 after labeling with anti-CD3e antibody and 10 nm gold-conjugated might thus increase the antigenic sensitivity of antigen-experienced protein A; a representative image of a IL-15-expanded WT lympho- CD4+ T cells in a cell-autonomous manner. blast is shown (Fig EV2). We found no differences in TCR A B IL-2 IL-15 WT CCR5-/- 100 100 *** *** IL-2 IL-2 *** *** 80 80 CD62L+ cells (%) CD25+ cells (%) IL-15 IL-15 60 60 40 40 20 20 FSC 0 0 WT CCR5-/- WT CCR5-/- 100 300 *** *** * * CD44 expression (MFI) 80 250 CD69+ cells (%) 200 60 150 40 100 CD62-L-FITC 20 50 0 0 WT CCR5-/- WT CCR5-/- C E 25 IL-2-derived 40 IL-15-derived lymphoblasts * lymphoblasts ** [3H]-TdR (cpm x 103) [3H]-TdR (cpm x 103) 20 WT CCR5-/- 30 * 15 CD25-PE 20 10 5 ** 10 * 0 0 100 101 102 103 100 101 102 103 OVA323-339 (nM) OVA323-339 (nM) D F CD69-PeCy7 80 * *** 250 WT 200 CCR5-/- * 60 IL-2 (pg/ml) IL-2 (pg/ml) 150 40 100 20 50 0 0 CD44-APC 100 101 102 103 100 101 102 103 OVA323-339 (nM) OVA323-339 (nM) Figure 2. CCR5 increases the sensitivity of antigen-experienced CD4+ T cells. A, B Representative histograms and quantification of mean fluorescence intensity (MFI; A) or the percentage of cells positive for the indicated memory markers (B) in OT-II WT and CCR5/ lymphoblasts expanded in IL-2 or IL-15, as specified. Data shown as mean  SEM (n ≥ 3). The gating strategy is shown in Fig EV1. C–F IL-2- (C, D) and IL-15-expanded lymphoblasts (E, F) were restimulated with indicated concentrations of OVA323–339; cell proliferation (thymidine incorporation into DNA; C, E) and IL-2 production (by ELISA; D, F) were measured after 72 h. Data are presented as mean  SEM (n = 5). Data information: *P < 0.05, **P < 0.01, ***P < 0.001, two-way ANOVA (B) or two-tailed unpaired Student’s t-test (C–F). 4 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal nanoclusters between OT-II WT and CCR5/ naı̈ve cells, which CCR5-induced TCR nanoclustering is independent of its had a small percentage of TCR nanoclusters larger than 4 TCR in co-stimulatory activity both genotypes (Fig 3A). In contrast, there was a significant increase in TCR nanocluster number and size in WT compared to Since CCR5 has co-stimulatory functions in CD4+ T-cell priming CCR5/ lymphoblasts (Fig 3B and C). The number of TCR (Molon et al, 2005; González-Martı́n et al, 2011), it is of interest nanoclusters per cell analyzed in each condition is also indicated to know whether defective TCR clustering in CCR5/ lympho- (Appendix Table S1). As predicted, there was a gradient in TCR blasts is due to suboptimal primary activation of these cells. To nanoclustering of naı̈ve  IL-2- < IL-15-differentiated OT-II WT address this question, we treated OT-II WT cells with the CCR5 cells (Appendix Fig S1E), which coincided with increased anti- antagonist TAK-779 at various intervals throughout culture and genic sensitivity of the IL-15-expanded cells (Appendix Fig S1F analyzed TCR nanoclusters in IL-2-expanded T lymphoblasts. and G). These findings thus reinforce the IL-15-induced memory- TAK-779 addition during the priming phase (blockade of CCR5 co- like phenotype versus the IL-2-induced effector-like phenotype and stimulatory function) decreased the percentage of large TCR link TCR nanoclustering with increased sensitivity in antigen- nanoclusters compared to untreated controls (Fig 4A). TAK-779 experienced CD4+ T cells. The difference in TCR nanoclustering treatment did not alter TCR clustering in OT-II CCR5/ cells between WT and CCR5/ cells was nevertheless similar in IL-2- (Appendix Fig S4), which indicates that the TAK-779 effect on OT- and IL-15-expanded lymphoblasts, which indicates that CCR5 II cells is CCR5-specific. affects TCR nanoclustering in lymphoblasts independently of the To avoid interference with the CCR5 co-stimulatory activity, we cytokine milieu. primed OT-II WT cells in the absence of the inhibitor and added Using a Monte Carlo simulation, we applied data from surface TAK-779 only during IL-2-driven expansion of the CD4+ lympho- replicas of naı̈ve and IL-2-expanded OT-II lymphoblasts to deter- blasts. In these conditions, TAK-779 also reduced the percentage of mine whether the experimental frequency of cluster size was due large TCR nanoclusters (Fig 4B), which indicates that the CCR5 to random distribution of gold particles. In all cases, the cluster signals that control TCR organization are independent of those distributions observed experimentally differed significantly from involved in its co-stimulatory function. pure random proximity between clusters (Appendix Fig S2). To We next explored whether other chemokine receptors involved define the differences between OT-II WT and CCR5/ cells, we in T-cell activation control TCR nanoclusters in CD4+ T cells. used a model that accounts for receptor clustering dynamics (Cas- CXCR4 is a paradigmatic chemokine receptor that also provides co- tro et al, 2014), a Bayesian inference method that estimates the stimulatory signals (Kumar et al, 2006; Smith et al, 2013). We so-called clustering parameter, b. Based on this model, we primed OT-II WT cells in the presence of the CXCR4 antagonist concluded that the probability of a chance nanocluster distribution AMD3100 and analyzed TCR nanoclusters in IL-2-expanded T similar to that observed for naı̈ve and activated OT-II WT and lymphoblasts. Vehicle- and AMD3100-treated cells showed similar CCR5/ cells approaches 0% (Fig 3D and E). Posterior distribu- TCR nanocluster distribution (Fig 4D), which implies that CXCR4 tion analysis also showed that whereas the clustering parameter blockade does not interfere with TCR nanoclustering. was very similar between naı̈ve OT-II WT and CCR5/ cells (Fig 3D), there was clear separation in lymphoblasts (Fig 3E). CCR5 deficiency increases ceramide levels in CD4+ T cells These analyses provide a mathematical framework that validates the TCR nanoclustering differences between WT and CCR5/ We analyzed CCR5 regulation of TCR nanoclustering in CD4+ T cells, as determined by EM. cells and found no differences between OT-II WT and CCR5/ cells The differences in TCR oligomerization between OT-II WT and in TCR/CD3 chain mRNA levels or in cell surface expression of the CCR5/ lymphoblasts were also studied using blue-native gel TCRa chain (Fig EV3). These data suggest that the reduction in TCR electrophoresis (BN-PAGE) (Schamel et al, 2005; Swamy & clustering in CCR5/ cells is not due to decreased TCR expression. Schamel, 2009). Cell lysis with digitonin, a detergent that disrupts T-cell antigen receptor nanoclustering is dependent on plasma TCR nanoclusters into their monomeric components, showed that membrane Chol and SM (Molnar et al, 2012), two lipids also neces- WT and CCR5/ lymphoblasts expressed comparable TCR levels, sary for CCR5 signaling (Mañes et al, 2001). OT-II WT and CCR5/ as detected with anti-CD3f antibodies (Fig 3F). Cell lysis with lymphoblasts expressed comparable levels of total Chol and SM Brij96, which preserves TCR nanoclusters, showed a notable species (Fig 5A and B). OT-II CCR5/ lymphoblasts nonetheless reduction in large TCR complexes in CCR5/ compared to WT showed a significant increase in most ceramide (Cer) species and lymphoblasts (Fig 3F). Two independent techniques thus support their dihydroCer (dhCer) precursors (Fig 5C and D). These dif- a CCR5 role in TCR nanoscopic organization in antigen-experi- ferences were not observed in naı̈ve OT-II WT and CCR5/ cells enced CD4+ T cells. (Appendix Fig S5A), indicative that the Cer increase was specific to To determine whether CCR5 controls TCR nanoclustering in antigen-experienced cells. The increase in Cer species in CCR5/ in vivo-generated memory T cells, we analyzed TCR distribution in lymphoblasts was not linked to enhanced apoptosis compared to surface replicas of CD4+ memory T cells purified by negative selec- WT cells (Appendix Fig S5B). tion from OVA/OVA-immunized WT and CCR5/ mice (Appendix Fig S3). CD4+ memory cells from CCR5/ mice showed CCR5 deficiency upregulates specific ceramide synthases in fewer, smaller TCR nanoclusters than those from WT counterparts CD4+ T cells (Fig 3G; Appendix Table S1), which indicates that CCR5 promotes formation of large TCR nanoclusters in endogenously generated Our analysis of the mRNA levels of key enzymes involved in Cer CD4+ memory T cells. metabolism showed no differences in ceramidases (ASAH1, ACER 2, ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 5 of 19 The EMBO Journal Ana Martín-Leal et al A WT CCR5-/- D Naïve cells 0% in ROPE mean=0.03 mean=0.23 mean=0.31 100 Naïve cells Random WT 100 CCR5-/- Total gold particles (%) 80 80 0.0 0.2 0.4 0.6 0.8 1.0 60 60 Clustering parameter (b) 40 20 E IL-2-derived lymphoblasts 40 0 0% in ROPE 1 2 3 4 >4 mean=0.03 20 WT CCR5-/- 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 Gold particles per cluster mean=0.79 mean=0.44 B WT CCR5-/- 0.0 0.2 0.4 0.6 0.8 1.0 Clustering parameter (b) F WT CCR5-/- WT CCR5-/- 80 IL-2-derived lymphoblasts WT KDa 80 *** ** ** *** 3 CCR5-/- * Multi / Monomeric TCR Total gold particles (%) nanoclusters 60 60 TCR 40 2 40 20 880 0 1 Monomeric 1 2 3 4 >4 TCR 20 440 0 Digitonin Brij-96 Digitonin Brij-96 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 Gold particles per cluster C WT CCR5-/- G WT CCR5-/- 60 IL-15-derived lymphoblasts 60 60 60 ** * * ** Total gold particles (%) Total gold particles (%) 40 40 40 40 20 20 20 0 20 0 1 2 3 4 >4 1 2 3 4 >4 WT CCR5-/- 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 Gold particles per cluster Gold particles per cluster Figure 3. 6 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal ◀ Figure 3. CCR5 increases TCR nanoclustering in antigen-experienced CD4+ T cells. A–C Analysis of TCR nanoclustering by EM in OT-II WT and CCR5/ naïve cells (A; n = 6 cells/genotype; WT: 3,427, CCR5/: 3,528 particles), and IL-2- (B; WT, n = 8 cells, 15,419 particles; CCR5/, n = 6 cells, 5,410 particles) or IL-15-expanded lymphoblasts (C; WT, n = 8 cells, 27,518 particles; CCR5/, n = 7 cells, 22,696 particles). A representative small field image at the top of each panel shows gold particle distribution in the cell surface replicas of anti-CD3e-labeled cells; at bottom, quantification (mean  SEM) of gold particles in clusters of indicated size in WT (gray bars) and CCR5/ cells (red). Insets show the distribution of clusters of one, two, three, four, or more than four particles, and statistical analysis. D, E Posterior distribution in naïve (D) and IL-2-expanded lymphoblasts (E) of the clustering parameter b for WT (gray) and CCR5/ cells (red); randomly generated distributions of receptors are shown in blue. The mean value of the b parameter is indicated for each condition. The probability of a chance distribution similar to that determined in cells is nearly 0% by the ROPE. F Comparison of TCR oligomer size using BN-PAGE and anti-CD3f immunoblotting in day 10, IL-2-expanded WT and CCR5/ OT-II lymphoblasts lysed in buffer containing digitonin or Brij-96. The marker protein is ferritin (f1, 440 and f2, 880 kDa forms). The ratio of TCR nanoclusters to monomeric TCR in each lysis condition was quantified by densitometry (right; n = 5). G Top, representative small field EM images showing gold particle distribution in the cell surface replicas of CD4+ T cells isolated from OVA/OVA-immunized WT and CCR5/ mice. Bottom, quantification (mean  SEM) of gold particles in clusters of the indicated size (WT, gray bars; n = 5 cells, 14,680 particles; CCR5/, red; n = 7 cells, 15,374 particles). Insets show the distribution between clusters of one, two, three, four, or more than four particles, and statistical analysis. Data information: (A–C, F, G), Data are mean  SEM. *P < 0.05, **P < 0.01, ***P < 0.001, one-tailed unpaired Student’s t-test. Scale bar, 50 nm (A–C, G). ACER 3) and sphingomyelinases (SMPD1–4) between OT-II WT and Ceramide levels control TCR nanoclustering CCR5/ naı̈ve cells or lymphoblasts (Fig EV4). mRNA levels of the ceramide synthases (CerS) CerS2, CerS3, and CerS4 were nonethe- We used a synthetic biology approach to determine whether cera- less upregulated in OT-II CCR5/ lymphoblasts (Fig 5E); CerS5 and mide content affects TCR nanoclustering. Large unilamellar vesicles CerS6 were unaltered, and the nervous system-specific CerS1 isoen- (LUV) were prepared at different molar ratios of PC, Chol, SM, and zyme was not detected. CerS2, CerS3, and CerS4 levels were compa- Cer (Fig 6A) and then reconstituted with a streptavidin-binding- rable in naı̈ve CD4+ WT or CCR5/ cells (Fig 5E), which again peptide-tagged TCR purified in its native state from murine M.mf- associate the CCR5 transcriptional effect on these genes with activa- SBP (streptavidin-binding peptide) T cells (Swamy & Schamel, tion. 2009). The proteoliposomes were analyzed by BN-PAGE after solu- We sought to validate the CerS isoforms upregulated by CCR5 bilization in 0.5% Brij96 to maintain TCR nanocluster integrity or in deficiency at the protein level. In accordance with mRNA analy- 1% digitonin to disrupt TCR clusters. As anticipated (Molnar et al, ses, CerS2 protein levels were significantly higher in CCR5/ 2012; Wang et al, 2016), TCR was monomeric in PC-containing than in WT lymphoblasts (Fig 5F); CerS3 and CerS4 were unde- LUV, whereas it formed nanoclusters when reconstituted in PC/ tectable or only barely detectable by immunoblot. This is Chol/SM liposomes (Fig 6B and C). The inclusion of ceramides in consistent with the fact that CerS2 has the highest expression these LUV (PC/Chol/SM/Cer liposomes) reduced TCR nanocluster- level and the broadest substrate specificity in other cell types ing in a dose-dependent manner. This effect was not due to differen- (Laviad et al, 2008). Chromatin immunoprecipitation (ChIP), tial TCR reconstitution in Cer-containing LUV, since digitonin followed by amplification of a region of the CerS2 promoter treatment rendered equivalent levels of monomeric TCR in each enriched in CpG islands, showed that binding of the transcrip- condition (Fig 6B). These data suggest that Cer membrane content tional activation marker acetylated histone H3K9 (H3K9Ac) was impairs TCR nanoclustering. higher in CCR5/ than in WT lymphoblasts (Fig 5G). Moreover, To test whether this effect also occurs in live cells, we treated blockade of CCR5 signaling with pertussis toxin (PTx; an inhi- OT-II WT lymphoblasts with recombinant sphingomyelinase bitor of the Gai subunit) also increased CerS2 mRNA expression (SMase), which hydrolyzes SM to ceramide (Kitatani et al, 2008). (Fig 5H). SMase treatment of WT OT-II blasts increased Cer levels robustly To further study CCR5 transcriptional regulation of CerS, we (Fig 6D), but did not compromise cell viability (Appendix Fig S6). scanned for transcription factors with putative binding sites in the Analysis of membrane replicas from these cells showed that CerS2, CerS3, and CerS4 promoters, which are transcriptionally SMase treatment reduced the number of high valency TCR upregulated in CCR5/ lymphoblasts, but not represented in the nanoclusters compared to controls (Fig 6E; Appendix Table S1), CerS6 promoter, which is not CCR5-regulated. We selected two which indicates that high Cer levels hinder TCR nanoclustering in regions; region 1 comprised 5 kb to the 50 UTR, and region 2 CD4+ T cells. encompassed the 50 UTR to the first coding exon (Fig 5I). This bioin- formatic approach identified GATA-1 and NF-IB (nuclear factor-1B) CerS2 silencing restores TCR nanoclustering after CCR5 as putative transcription factors involved in the differential expres- functional blockade sion of the CerS2 isoform (Fig 5J and K). We focused on GATA-1, since it is implicated in the differentia- To correlate increased CerS2 expression with the impaired TCR tion of some CD4+ T-cell subtypes (Sundrud et al, 2005; Fu et al, nanoclustering in OT-II CCR5/ T cells, we attempted to silence 2012). Immunofluorescence analyses showed increased nuclear CerS2 expression by lentiviral transduction of primary lymphoblasts levels of the phosphoSer142-GATA-1 form in OT-II CCR5/ with short-hairpin (sh) RNA for CerS2 or control (shCtrl). In the compared to WT lymphoblasts (Fig 5L and M), which correlated most successful experiments, we were only able to transduce ~ 20% with enriched GATA-1 binding to the CerS2 promoter in CD4+ of the lymphoblasts, which did not lead to solid CerS2 mRNA silenc- CCR5/ lymphoblasts (Fig 5N). CCR5 deficiency might induce ing (Appendix Fig S7A and B). Despite the low efficiency, antigenic CerS2 transcription through GATA-1. restimulation tended to promote stronger responses in shCerS2- ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 7 of 19 The EMBO Journal Ana Martín-Leal et al A TAK-779 B TAK-779 Day3 Day10 Electron Electron microscopy microscopy Day10 OT-II WT Antigen OT-II WT Antigen removal removal Medium TAK-779 Medium TAK-779 80 80 80 80 ** *** * Total gold particles (%) Total gold particles (%) 60 60 60 60 40 40 40 20 40 20 0 0 1 2 3 4 >4 1 2 3 4 >4 20 20 Medium Medium TAK-779 TAK-779 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 Gold particles per cluster Gold particles per cluster C AMD3100 80 Day3 Day10 80 Electron Total gold particles (%) microscopy 60 60 OT-II WT Antigen removal 40 Vehicle AMD3100 40 20 0 1 2 3 4 >4 20 Vehicle AMD3100 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 Gold particles per cluster Figure 4. CCR5-induced TCR nanoclustering is specific and independent of its co-stimulatory activity. A OT-II WT cells were activated with OVA323–339, alone or with TAK-779. After 3 days, antigen and TAK-779 were removed and lymphoblasts expanded in IL-2-containing medium. TCR nanoclustering was analyzed in anti-CD3e-labeled surface replicas of day 10 lymphoblasts. Top, representative small field EM images showing gold particle distribution in the cell surface replicas of WT CD4+ T cells alone or with TAK-779. Bottom, quantification of gold particles in clusters of the indicated size. Inset, distribution of gold particles between clusters of one, two, three, four, or more than four particles in vehicle- (gray bars; n = 5 cells, 11,266 particles) and TAK-779-treated cells (black; n = 6 cells, 5,138 particles). B OT-II WT cells were activated with OVA323–339, and TAK-779 was added at days 3, 5, and 7 after antigen removal. Analysis as above, untreated (gray bars; n = 5 cells, 6,400 particles) and TAK-779-treated cells (black; n = 6 cells, 7,153 particles). Inset shows the distribution between clusters of one, two, three, four, or more than four particles, and statistical analysis. C OT-II WT naïve cells were activated with antigen in the presence or not of the CXCR4 inhibitor AMD3100. Left, representative EM images showing gold particle distribution in the cell surface replicas. Right, analysis of gold particles in clusters as above, vehicle- (gray bars; n = 6 cells, 12,339 particles) and AMD3100-treated cells (black; n = 7 cells, 17,059 particles). Inset shows the distribution between clusters of one, two, three, four, or more than four particles, and statistical analysis. Data information: (A–C), Data are mean  SEM. *P < 0.05, **P < 0.01, ***P < 0.001, one-tailed unpaired Student’s t-test. Scale bar, 50 nm. than in shCtrl-transduced cells (Appendix Fig S7C and D). The low 779 treatment increased CerS2 levels and impaired TCR efficiency also precluded analysis of TCR nanoclusters in membrane nanoclustering (Appendix Fig S8). The data suggest that CCR5 replicas, as transduced cells could not be distinguished from non- effects on TCR nanoclustering and CerS2 induction are not exclu- transduced cells. sive to the OT-II system and that TAK-779-treated 2B4 cells To overcome these difficulties, we used the 2B4 CD4+ T-cell mimic the functional findings in OT-II CCR5/ lymphoblasts. line. We verified that 2B4 cells expressed CCR5 and that TAK- 2B4 cells were transduced efficiently by lentiviruses and, after 8 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal 3 days of antibiotic selection, 100% of the cells expressed the TAK-779, CD69 upregulation was higher in CerS2-deficient than shRNA; this led to strong silencing of CerS2 mRNA and protein in shCtrl-cells (Fig 6J). in shCerS2- compared to shCtrl-transduced cells (Fig 6F–H). Analysis of TCR organization showed recovery of large TCR CCR5 modulates TCR nanoclustering in human CD4+ T cells nanoclustering in TAK-779-treated, shCerS2-transduced cells compared to controls (Fig 6I; Appendix Table S1); after restimu- Finally, we tested whether CCR5 deficiency also impairs TCR orga- lation with plate-bound anti-CD3e antibody in the presence of nization in human CD4+ T cells. Approximately 1% of the A p = 0.9 B C D 25 300 WT 150 2.5 * dhCer (pmol /106 cells) CCR5-/- Cer (pmol /106 cells) SM (pmol /106 cells) Chol (μg /106 cells) 20 * 2.0 200 100 15 1.5 * 1.0 10 100 50 ** ** * * ** 5 * * ** ** ** 0.5 0 0 0 0 C :0 C :0 C :1 C :0 C 1 C :0 C :1 C :0 C :1 C :0 C 1 :2 :0 :0 :0 :0 :0 :1 C :0 C :0 C :1 C :0 C :1 C :0 C :1 C :0 C :1 C :0 C :1 C :2 :3 : : 14 16 18 20 22 24 16 18 20 22 24 16 18 20 22 24 24 24 14 16 18 20 22 24 16 18 20 22 24 24 24 C C C C C C C C SM species Cer species dhCer species E F 15 ** 20 CerS2 * 20 CerS3 20 CerS4 WT CCR5-/- CerS2 / β-actin ratio WT Naïve Blasts Naïve Blasts 10 CCR5-/- 15 15 15 p = 0.08 CerS2 10 10 10 *** * 5 β-actin mRNA expression (Rq) 5 5 5 0 Naïve Blasts 0 0 0 Naïve Blasts Naïve Blasts Naïve Blasts G H 25 2.5 CerS5 p = 0.7 3 CerS6 Y 1.5 * * signal relative to input 20 *** CerS2 mRNA (Rq) 2.0 WT 2 CCR5-/- ChIP 1.0 15 H3K9Ac 1.5 CerS2 promoter CpG 10 1.0 1 p = 0.4 -2,500 +1 0.5 5 0.5 0 0 0 0 0 10-3 10-2 10-1 Naïve Blasts Naïve Blasts WT CCR5-/- PTx (μg/ml) TSS I L promoter WT CCR5-/- 5’UTR 3’UTR Exon GATA-1 M N Y -5 kb +5 kb Region 1 Region 2 Intron (pSer142) ChIP J Cers3 Cers4 T A-1 4 7 GA 142 pSer -GATA-1 CerS2 promoter 300 Integrated density (x103) 30 5 0 5 ** -2,500 -260 GATA-1 30 15 0 NF-IB * signal relative to input 200 24 5 0 Cers6 Cers2 25 23 10 52 100 K DAPI Cers3 Cers4 5 23 36 0 10 15 WT CCR5-/- 1 0 11 0 2 54 WT CCR5-/- Cers2 1 Cers6 10 4 8 1 9 Figure 5. ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 9 of 19 The EMBO Journal Ana Martín-Leal et al ◀ Figure 5. CCR5 deficiency increases Cer levels by upregulating specific CerS. A Total Chol levels in WT and CCR5/ OT-II lymphoblasts (day 10, IL-2-expanded) as determined by a fluorometric assay (n = 6). B–D SM (B), Cer (C) and dhCer (D) levels in WT and CCR5/ OT-II 10-day lymphoblasts, as determined by UPLC-TOF MS. Values, after normalization with C17 standards and cell number in each sample, are the mean of two independent experiments (n = 6). E RT–qPCR determination of CerS mRNA levels in naïve and IL-2-expanded WT and CCR5-/ OT-II 10-day lymphoblasts (n = 3–5). F Representative immunoblot showing CerS2 protein levels in naïve and WT and CCR5/ OT-II 10-day lymphoblasts, and densitometric quantification of blots as above (n = 10). G ChIP analysis of the CerS2 promoter using an anti-H3K9Ac antibody. Scheme of the CerS2 promoter showing CpG islands and primers used for amplification. Relative ChIP of the CerS2 promoter in WT and CCR5/ OT-II 10-day lymphoblasts (n = 3). H Relative CerS2 mRNA level in CD4 T cells treated with PTx (n = 3). I Scheme of a canonical CerS gene to illustrate the in silico strategy used to search for CerS-specific transcription factors. J, K Venn diagrams showing the number of transcription factors with putative binding sites in the indicated CerS genes in regions 1 (J) and 2 (K). The red circle highlights the transcription factors shared by CerS2, CerS3, and CerS4 promoters, but not present in the CerS6 promoter. L Representative immunofluorescence images showing pSer142-GATA-1 staining (green) of OT-II WT and CCR5/ lymphoblasts. The green channel (top) and the merge with nuclear DAPI staining (blue; bottom) are shown. Scale bar, 10 lm. M Quantification of nuclear staining of the cells plotted as integrated density fluorescence intensity in DAPI-stained area (n ≥ 50 cells/condition). N Top, basic scheme of the CerS2 promoter, indicating the putative GATA-1 binding site (blue) and location of the primers used for amplification in ChIP assays (black arrows). Bottom, relative anti-GATA-1 ChIP levels in OT-II WT and CCR5/ lymphoblasts (n = 5). Data information: (E, G, H, N), Each data point is the average of triplicates in an independent experiment. (A–H, M, N), Data shown as mean  SEM of triplicates; *P < 0.05, **P < 0.01, ***P < 0.001, two-tailed unpaired Student’s t-test. Source data are available online for this figure. Spanish population bears the ccr5D32 polymorphism in homozy- super-resolution microscopy (Schamel et al, 2005; Hu et al, 2016; gosity (Mañes et al, 2003). Purified CD4+ T cells from healthy Jung et al, 2016; Pageon et al, 2016). We demonstrate here differen- WT or ccr5D32 homozygous donors were activated with anti-CD3 tial TCR nanoclustering in WT and CCR5-deficient cells using two and anti-CD28 antibodies for 3 days and maintained for five addi- complementary approaches (EM and BN-PAGE), based on different tional days with IL-2. We found that ccr5D32 lymphoblasts had a conceptual principles. In EM, TCR nanoclusters were defined as gold lower percentage of large TCR nanoclusters than WT cells (Fig 7A; particle aggregates at < 10 nm distance from one another. Previous Appendix Table S1); concomitantly, the fraction of monomeric analyses showed that this criterion permits identification of TCR TCR was increased in the former. Sphingolipid analysis of these nanoclusters formed by TCR-TCR interactions (Kumar et al, 2011); CD4+ lymphoblasts showed an increase in saturated 24-carbon these tightly associated TCR nanoclusters would allow inter-TCR Cer (C24:0) and its precursor (dhCer C24:0) in cells derived from cooperativity for pMHC binding (Martı́n-Blanco et al, 2018). We ccr5D32 donors, whereas SM levels were comparable between therefore intentionally considered TCR not to be in the same both genotypes (Fig 7B). This increase in Cer levels was associ- nanocluster if the gap between them was > 10 nm; this excludes ated with upregulation of CerS2 mRNA in ccr5D32 lymphoblasts considering more loosely associated TCR as nanoclusters, but the compared to WT controls (Fig 7C); expression of other enzymes strict definition allowed association of TCR nanoclusters to a T-cell involved in Cer metabolism was unchanged in both genotypes biological function. (Fig EV5). These results indicate that, as found in mouse CCR5/ It is also important to clarify that gold particle counts do not lymphoblasts, antigen-experienced human CD4+ T cells from necessarily correspond to the number of TCR molecules in a ccr5D32 homozygotes show defective TCR nanoclustering associ- nanocluster. BN-PAGE defines neither the exact size nor the ated with increased Cer levels and upregulated CerS2. Moreover, abundance of TCR nanoclusters. Direct comparison of CCR5/ they indicate that these CCR5 effects are not restricted to specific with WT cells using both methods nonetheless allowed us to T-cell clones, but can be observed in a polyclonal T-cell detect relative differences and determine the promoter effect of repertoire. CCR5 in TCR nanoclustering. Application of Monte Carlo simula- tions further indicated that the nanoclusters observed in EM are not the result of random proximity of gold particles. The esti- Discussion mated clustering parameter (b) for randomly distributed particles was virtually zero. Here, we show that CCR5 signaling is largely dispensable for Since CCR5 provides positive signals during activation of memory CD4+ T-cell differentiation, but provides specific signals naı̈ve CD4+ T cells (Molon et al, 2005; Nesbeth et al, 2010; that improve the functional fitness of memory cells after antigen re- González-Martı́n et al, 2011), we attempted to clarify whether encounter. The CCR5 signals optimize TCR nanoclustering and anti- TCR nanoclustering impairment in CCR5/ cells is solely an gen sensitivity by triggering a CD4+ T-cell-specific transcription effect of this defective priming. This is unlikely, since TCR clus- program that regulates Cer metabolism. This CCR5 program oper- tering was inhibited when CCR5 was blocked during expansion ates in murine and human CD4+ T cells, which suggests physio- of fully activated WT lymphoblasts. This effect on TCR nanoclus- pathological relevance. tering during lymphoblast expansion was modest compared to A central observation of our study is that CCR5 expression that observed in the priming phase, but is probably the result of enhances the degree of TCR nanoclustering in resting antigen- insufficient CCR5 inhibition during lymphoblast expansion. CCR5 experienced T cells, both in vitro and in vivo. The presence of TCR is not only upregulated shortly after activation, but is maintained nanoclusters in resting T cells was shown by BN-PAGE, EM, and in memory CD4+ T cells, which are highly susceptible to 10 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal A B C 150 0.1 PC + + + + + + + + Polydispersity index ( ) Extrusion TCR Chol/SM - + + + - + + + 2.0 * p=0.2* Nanocluster/Monomeric LUV size (nm; ) LUV formation purification Cer - - 0.08 0.14 - - 0.08 0.14 PC 100 KDa PC/chol/SM 1.5 nanoclusters PC/chol/SM/Cer(0.08) SBP 0.05 PC/chol/SM/Cer(0.14) TCR Lipid mix 1.0 50 880 880 Monomeric 0.5 TCR- Proteoliposomes 440 TCR 0 0 440 PC 1 0.7 0.62 0.56 Chol/SM - 0.3 0.3 0.3 0 Brij-96 Digitonin Digitonin Brij-96 Cer - - 0.08 0.14 D E Untreated (control) SMase-treated F G H 500 1.5 ** 2 rS Cer (pmol /106 cells) CerS2 mRNA (Rq) * rl Ce 400 Ct sh 1.0 sh 300 Total gold particles (%) 80 Untreated 80 100% 0.5 200 60 SMase 60 ** * ** 0.0050 anti-CerS2 100 40 0.0025 40 20 0 0 Control SMase anti-β-actin 4 rl 2 0 GFP 2B Ct rS 20 1 2 3 4 >4 sh Ce sh 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 Gold particles per cluster I J TAK-779-treated 2B4 cells 80 80 80 ** * shCtrl Total gold particles (%) 60 60 CD69+ cells (%) 60 ** * 40 20 40 40 0 1 2 3 4 >4 20 shCerS2 20 shCtrl shCtrl shCerS2 shCerS2 0 0 0.2 0.4 0.6 0.8 1.0 0 Anti-CD3 antibody (μg/ml) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 Gold particles per cluster Figure 6. Cer levels determine the grade of TCR nanoclustering. A Scheme of the strategy used to form TCR proteoliposomes, and size of LUV generated at the indicated lipid molar ratio. Polydispersity index values are shown as black squares for each condition (n = 3). SBP, streptavidin-binding-peptide-tagged TCR. B Representative immunoblots comparing TCR nanocluster sizes via BN-PAGE and anti-CD3f immunoblotting in TCR proteoliposomes lysed in the presence of Brij-96 or digitonin. The marker protein is ferritin (f1, 440 and f2, 880 kDa forms). C The ratio of the nanocluster and monomeric TCR in each lysis condition was quantified by densitometry from immunoblots as in (B) (n ≥ 4). D Cer levels in OT-II 10-day lymphoblasts, untreated or treated with SMase (n = 4). E Representative small field images showing gold particle distribution, and quantification (mean  SEM) of gold particles in clusters of the indicated size in cell surface replicas from untreated (gray bars; n = 5 cells, 8,126 particles) and SMase-treated (1 h) OT-II lymphoblasts (cyan; n = 6 cells, 8,457 particles) after CD3e labeling, as determined by EM. The inset shows distribution between clusters of one, two, three, four, or more than four particles. F GFP expression in shCtrl- (black) and shCerS2 (orange)-transduced 2B4 cells after puromycin selection, as determined by FACS. Non-transfected 2B4 cells (gray). G Relative CerS2 mRNA levels in TAK-779-treated 2B4 cells as in (F). Values were normalized to those obtained in untransduced TAK-779-treated 2B4 cells (n = 3). H Representative immunoblot with anti-CerS2 antibody to determine CerS2 protein levels in shCtrl and ShCerS2-transduced 2B4 cells as in (G). Filters were rehybridized with b-actin as loading control. I TCR nanoclustering of shCtrl- and shCerS2-transduced 2B4 cells in the presence of TAK-779 as determined by EM. Representative small field images and quantification (mean  SEM) of gold particles in clusters of indicated sizes in cell surface replicas shCtrl (black bars; n = 6 cells, 12,337 particles) and shCerS2 2B4 lymphoblasts (orange; n = 7 cells, 13,456 particles). Inset, distribution between clusters of indicated size and statistical analysis. J Percentage of CD69+ shCtrl (black) and shCerS2 (orange) 2B4 cells restimulated with plate-bound anti-CD3e antibody in the presence of TAK-779 (n = 3). Data information: (A, C–E, G–J), Data are shown as mean  SEM. *P < 0.05, **P < 0.01, one-tailed (E, I) or two-tailed unpaired Student’s t-test. Scale bar, 50 nm. Source data are available online for this figure. ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 11 of 19 The EMBO Journal Ana Martín-Leal et al A B WT ccr5Δ32 500 WT SM (pmol /106 cells) 400 ccr5Δ32 300 200 100 80 80 * * ** 0 :0 :0 :1 :0 :1 :0 :0 :1 :2 Total gold particles (%) :0 14 16 16 18 18 20 22 24 24 24 60 60 SM species 40 40 p=0.1 Cer (pmol /106 cells) 20 40 30 0 1 2 3 4 >4 20 20 WT ccr5Δ32 10 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 >15 0 Gold particles per cluster C 16:0 22:0 24:0 24:1 Cer species 2.5 15 * * dhCer (pmol /10 cells) CerS2 mRNA (Rq) 2.0 10 6 1.5 1.0 5 0.5 0 0 WT ccr5Δ32 16:0 22:0 24:0 dhCer species Figure 7. CCR5 controls TCR nanoclustering and Cer metabolism in human CD4+ cells. A Analysis of TCR nanoclustering in lymphoblasts from healthy WT and ccr5D32 homozygous donors by EM. Top, representative small field image showing gold particle distribution in cell surface replicas of anti-CD3e-labeled cells; bottom, quantification (mean  SEM) of gold particles in clusters of the indicated size in the WT (gray bars; n = 5 cells, 17,689 particles) and D32/D32 cells (light red; n = 4 cells, 16,938 particles). Insets show the distribution between clusters of one, two, three, four, or more than four particles, and statistical analysis. B Normalized SM, Cer, and dhCer levels in lymphoblasts obtained as in (A). A representative experiment is shown (n = 3 donors/genotype; n = 2 independent experiments). C Relative CerS2 mRNA levels in day 8 WT and ccr5D32 lymphoblasts. Each data point is the average of a technical triplicate from three donors in two independent experiments (n = 6). Data information: Data are shown as mean  SEM (B, C). *P < 0.05, **P < 0.01, two-tailed unpaired Student’s t-test. Scale bar, 50 nm. infection by R5-tropic HIV-1 strains (Nie et al, 2009). Our results essential for TCR nanoclustering in lymphoblasts. CCR5 homo- and showed increased CCR5 mRNA expression during lymphoblast heterodimers are thought to associate differently with Ga subunits; expansion. These lymphoblasts also expressed CCR5 ligands, homodimers signal through the PTx-sensitive Gai, whereas heterodi- suggesting autocrine CCR5 stimulation during this phase. We mers generate PTx-resistant responses (Mellado et al, 2001). CCR5 thus propose that TCR nanoclustering is regulated by CCR5 co-stimulatory signals in the IS are PTx-resistant (Molon et al, signals transduced during lymphoblast differentiation rather than 2005), consistent with CCR5/CXCR4 heterodimerization during during priming. priming. PTx potentiates transactivation of the CerS2 promoter Another feature that distinguishes CCR5 effects on priming and (Fig 5H), however, which suggests involvement of the CCR5- on TCR nanoclustering is the role of CXCR4 in these events. During induced Gai pathway in TCR nanoclustering. CCR5 thus appears to priming, CCR5 and CXCR4 are recruited to and accumulate as trigger distinct signaling pathways for co-stimulation and TCR heteromeric complexes at the IS of CD4+ T cells; AMD3100 (a nanoclustering in CD4+ T cells. Since chemokine receptors can form CXCR4 antagonist) prevented not only CXCR4 but also CCR5 accu- nanoclusters (Martinez-Muñoz et al, 2018), it would be of interest mulation (Contento et al, 2008), which indicates necessary coopera- to study potential feedback loops between TCR and CCR5 nanoclus- tion between CCR5 and CXCR4 for full T-cell activation. In contrast, ters. AMD3100 did not affect TCR nanoclusters in CCR5-expessing cells, Cholesterol and SM are two lipids essential for CCR5 signaling which suggests that CXCR4/CCR5 heterodimer signaling is not and TCR nanoclustering (Mañes et al, 2001; Molnar et al, 2012). In 12 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal resting T cells, these receptors nonetheless partition in different responses of CCR5-deficient memory CD4+ T cells generated by membrane phases, liquid-ordered (lo) for CCR5 (Molon et al, 2005) vaccination were impaired after ex vivo stimulation. In a second and liquid-disordered (ld) for TCR (Beck-Garcia et al, 2015). This model that involves T:B-cell cooperation, we show that CCR5 defi- differential phase segregation argues against direct CCR5/TCR inter- ciency impaired class switching of high-affinity antibodies after re- action as a mechanism that influences TCR nanoclustering. Our exposure to a T cell-dependent antigen. Affinity maturation and results suggest instead that increased levels of long-chain Cer class switching depend on recruitment of Tfh cells to GC (Vinuesa species cause defective TCR nanoclustering in CCR5/ lympho- et al, 2016). This Tfh cell confinement is a result of CXCR5 expres- blasts. Indeed, elevation of Cer levels in TCR-reconstituted proteoli- sion and the downregulation of other homing receptors, including posomes and in live cells by SM hydrolysis impaired nanoscopic CCR5 (Crotty, 2011), which could explain the lack of difference in TCR organization. Although Cer levels were higher in lymphoblasts class switching between OVA/KLH-immunized WT and CCR5/ than in naı̈ve cells, which supports a role for Cer in T-cell activation mice. There were also no differences between WT and CCR5/ (Sofi et al, 2017), CCR5 deficiency further increased Cer levels effector Tfh cells after the first OVA immunization. Once GC resolve, specifically in lymphoblasts. The Cer increase in CCR5/ lympho- however, some Tfh cells are reported to enter the circulation as Tfh blasts did not cause spontaneous apoptosis (Appendix Fig S6B), central memory-like cells (Vinuesa et al, 2016). These circulating, which coincides with the non-apoptotic and preventive effects of antigen-experienced Tfh cells express CCR5 and are very susceptible long-chain Cer in this process (Stiban & Perera, 2015). to HIV-1 infection (Xu et al, 2017). We found that the frequency of Although Cer levels increased in antigen-experienced CD4+ T large TCR nanoclusters increased in memory T cells from WT cells from CCR5/ mice and ccr5D32 homozygotes, there were dif- compared to CCR5/ OVA/OVA-immunized mice, which suggests ferences between the mouse and the human cell ceramidome. This increased antigenic sensitivity. could depend on many factors, including the overall species-specific We hypothesize that following re-exposure to antigen, CCR5- enzymes involved in sphingolipid metabolism, or the lipid composi- expressing memory pre-Tfh cells will have a more efficient response tion of the diet. Membrane phase segregation properties of cera- than CCR5-deficient cells, which would support robust antibody mides are dependent not only on the saturation, but also on the responses after their differentiation to GC-Tfh cells. In humans, func- length of their acyl chain and the lipid microenvironment that tional CCR5 deficiency does not cause strong immune suppression, surrounds them (Alonso & Goñi, 2018). Although C24:1 is the most but ccr5D32 homozygosity was associated with four times more fatal abundant Cer in mouse WT and CCR5/ T cells, the biggest dif- infections than average during the 2009–2011 influenza season in ferences associated with CCR5 deficiency were observed for C20:0 Spain (Falcon et al, 2015) and fatal infections by the West Nile and C22:0 Cer. The differential membrane phase segregation proper- Virus in the United States (Lim & Murphy, 2011). Our results ties of C24:0 and C24:1 Cer might be subtle in an environment provide a conceptual framework on which to base clinical trials to enriched in other long-chain saturated Cer and dhCer. evaluate CD4+ T-cell memory responses in CCR5-deficient humans, Several settings can be hypothesized that explain Cer effects on and suggest caution regarding the risks associated with genetic abla- TCR nanoclustering. In model membranes, Cer has strong segrega- tion of CCR5 as a preventive strategy to block HIV-1 infection. tion capacity, which might affect their lateral organization (Alonso & Goñi, 2018). When the Chol concentration is saturating, however, as is the case of cell membranes and our proteoliposomes, Cer- Material and Methods enriched domains are not formed, due to the ability of Chol and Cer to displace one another (Castro et al, 2009). This argues against the Resource Identification Portal accession numbers for antibodies, cell idea that Cer impairs TCR clustering by promoting a general reduc- lines, animals, and other reagents used in the study are provided in tion in membrane lateral diffusion. Our mathematical model also Appendix Table S2. predicted that cluster distribution is independent of TCR diffusivity. Interaction of the TCRb subunit with Chol/SM complexes is critical Antibodies and reagents for TCR nanoclustering (Molnar et al, 2012; Wang et al, 2016). High levels of Cer or their precursors (dhCer) could transform Chol/SM Antibodies used to characterize mouse cells by flow cytometry were into Chol/SM/Cer domains with specific physicochemical proper- anti-Va2TCR-PE (B20.1), anti-CD25-PE (PC61), anti-CD45.2-FITC ties, which might hinder TCR nanocluster formation physically or (104), anti-CD62L-FITC/APC (MEL-14), anti-CD69-PeCy7 (H1.2F3), thermodynamically. For instance, high dhSM levels inhibit CCR5- and biotinylated anti-CXCR5 (2G8) from BD Biosciences; anti- mediated HIV-1 infection by rigidifying CCR5-containing lo domains human biotinylated CD3 (OKT3), anti-CD4-PeCy7/eFluor450/Pacific (Vieira et al, 2010). Chol/SM interaction depends on the hydrogen Blue (RM4.5), anti-IFNc-APC (XMG1.2), and anti-PD1-eFluor780 bond generated by the amide group of the SM molecule and the 3- (J43) from eBioscience; and anti-CD44-PeCy5/APC (IM7) from hydroxyl group of Chol (Ramstedt & Slotte, 2002), but the SM amide BioLegend. Biotinylated and purified anti-CD3e (145-2C11; BD Bios- group can also form hydrogen bonds with the Cer hydroxyl group ciences) were used for EM and T-cell activation, respectively. Anti- (Garcia-Arribas et al, 2016). It is thus possible that, rather than mouse CerS-2 (1A6; Novus Biologicals), anti-CD3f (449, purified forming SM/Chol/Cer domains, small increases in Cer levels would from hybridoma), and anti-b-actin (AC-15; Sigma-Aldrich) were increase the mutual displacement of three lipids. This could lead to used for immunoblot. Anti-mouse GATA-1 (D52H6; Cell Signaling) replacement of SM/Chol by SM/Cer complexes with gel-like and anti-mouse phospho-GATA1pSer142 (Thermo Fisher Scientific) biophysical properties (Sot et al, 2008). were used for immunofluorescence. Anti-GATA1 (ab11852, Abcam), Our results indicate that the CCR5 effect on TCR clustering is anti-histone H3Lys9 (CS200583), and purified IgG rabbit (PP64B; biologically meaningful. In a first model, we show that the EMD Millipore) were used for ChIP. ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 13 of 19 The EMBO Journal Ana Martín-Leal et al The OVA323–339 peptide was synthesized at the CNB Proteomics using the EasySep Human CD4+ Enrichment kit (Stem Cell Tech- facility. TAK-779, AMD-3100, poly-L-lysine, pertussis toxin, Cer nologies) and stimulated with anti-CD3-coated magnetic beads (bovine spinal cord), and sphingomyelinase (Bacillus cereus) were (Dynabeads M-450 tosyl-activated, Thermo Fisher) for 3 days. from Sigma-Aldrich. mCCL4, mIL-2, and mIL-15 were from Pepro- Beads were removed with a magnet, and cells were incubated with Tech; NIP-KLH, NIP-OVA, NIP(7)-BSA, and NIP(41)-BSA were from IL-2 (5 ng/ml) to generate lymphoblasts. The ccr5D32 polymor- Biosearch Technology. Soybean phosphatidylcholine Chol, egg SM, phism (rs333) was genotyped by PCR (AriaMx Real-time; Agilent C12 Cer (d18:1/12:0), C16 Cer (d18:1/16:0), C18 Cer (d18:1/18:0), Technologies) as described (Mañes et al, 2003). C24 Cer (d18:1/24:0), C24:1 Cer (d18:1/24:1(15Z)), C16 dhCer (d18:0/16:0), C18 dhCer (d18:0/18:0), C24 dhCer (d18:0/24:0), Flow cytometry C24:1 dhCer (d18:0/24:1(15Z)), C12:0 SM (d18:1/12:0), C16:0 SM (d18:1/16:0), C18:0 SM (d18:1/18:0), C24:0 SM, C24:1 SM, and the For cell surface markers, cell suspensions were incubated (20 min, Cer mix from bovine spinal cord were from Avanti Polar Lipids. 4°C) with the indicated fluorochrome-labeled or biotinylated mono- Lentiviral pGIPZ containing shRNAs for murine Cers 2 clonal primary antibodies in phosphate-buffered saline with 1% (V3LMM_454307, V3LMM_454309, and V3LMM_454311 clones), BSA and 0.02% NaN3 (PBS staining buffer). For intracellular label- and the mismatched control were from Dharmacon. ing, cells were fixed and permeabilized with IntraPrep (Beckman Coulter), followed by intracellular staining with indicated antibod- Mice and cell lines ies. Cells were analyzed on Cytomics FC500 or Gallios cytometers (both from Beckman Coulter) and data analyzed using FlowJo soft- C57BL/6J WT and CCR5/ mice were from The Jackson Labora- ware. tory. TCR transgenic OT-II CCR5/ mice, recognizing OVA323–339 (ISQAVHAAHAEINEAGR; I-Ab MHC class II molecule), have been Immunization and adoptive transfer described (González-Martı́n et al, 2011). B6-SJL (Ptprca Pepcb/ BoyJ) mice bearing the pan-leukocyte marker allele CD45.1 were Spleen and lymph node cell suspensions from OT-II WT or CCR5/ used for adoptive transfer experiments. CD3e-deficient mice (DeJar- cells were adoptively transferred (5 × 106 cells/mouse) into CD45.1 nette et al, 1998) were used as a source of antigen-presenting cells mice. The following day, recipient mice were infected intravenously for restimulation assays. Mice were maintained in SPF conditions in with rVACV-OVA virus (2 × 106 pfu). Mice were sacrificed 35 days the CNB and CBM animal facilities, in accordance with national and later, and splenocyte suspensions were obtained as described EU guidelines. All animal procedures were approved by the CNB above. and the Comunidad de Madrid ethical committees (PROEX 277/14; C57BL/6J or CCR5/ mice were immunized (i.p.) with NIP- PROEX 090/19). Human embryonic kidney HEK-293T cells and the OVA (200 lg) in alum (100 ll) diluted 1:1 in PBS. At 7 days post- murine 2B4 hybridoma and its derivative M.mf-SBP (which immunization, spleens from three mice of each genotype were expresses a SBP-tagged form of CD3f) (Swamy & Schamel, 2009) harvested and analyzed by flow cytometry to detect Tfh cells. On were cultured in standard conditions. day 30, mice were randomized in a blind manner and half of the mice in each group were re-immunized with NIP-OVA/alum (as Isolation and culture of mouse and human primary T cells above); the other half received NIP-KLH (200 lg)/alum. Mice were sacrificed 15 days later, and serum anti-NIP antibodies were deter- Spleen and lymph nodes from 6- to 12-week-old OT-II WT and mined by ELISA. Plate-bound NP(7)-BSA and NP(41)-BSA (5 lg/ CCR5/ mice were isolated and cell suspensions obtained using ml) were used to measure high- and low-affinity Ig, respectively. 40-lM pore filters. Erythrocytes were lysed with AKT lysis buffer Sera from NIP-OVA- and NIP-KLH-immunized mice were diluted (0.15 M NH4Cl, 10 mM KHCO3, 0.1 mM EDTA), and cells were acti- 1:175, and, after several washing steps, anti-NIP antibody binding vated with the appropriate OVA peptide for 3 days. Antigen was was developed with the SBA Clonotyping System-HRP (Southern removed, and cells were cultured with IL-2 (5 ng/ml) or IL-15 Biotech). Absorbance at 405 nm was determined in a FilterMax F5 (20 ng/ml). For some experiments, naı̈ve OT-I and OT-II cells were microplate reader (Molecular Devices). Memory cells from NIP- obtained by negative selection using the Dynabeads Untouched OVA- and NIP-KLH-immunized mice were purified as indicated and Mouse CD4 Cell Kit (Thermo Fisher). Flow cytometry indicated processed for EM. > 85% enrichment in all cases. Memory CD4+ T cells, generated in vivo after NIP-OVA or NIP-KLH immunization (see below), were Immunogold labeling, replica preparation, and EM analysis isolated by negative selection with the Mouse Memory T cell CD4+/ CD62L/CD44hi Column Kit (R&D Systems). Immunogold-labeled cell surface replicas were obtained as Blood samples from ccr5D32 homozygous and WT healthy described (Kumar et al, 2011). Briefly, T cells were fixed in 1% donors were from the Fundació ACE (Barcelona, Spain) and paraformaldehyde (PFA) and labeled with anti-mouse CD3 mAb obtained with informed consent of the donors. No personal data (145-2C11) or anti-human CD3 mAb OKT3, followed by 10 nm were registered, and all procedures using these samples were in gold-conjugated protein A (Sigma-Aldrich). Labeled cells were accordance with the standards approved by the Ethics Committee of adhered to poly-L-lysine-coated mica strips and fixed with 0.1% the Hospital Clinic Barcelona (HCB/2014/0494 and HCB/2016/ glutaraldehyde. Samples were covered with another mica strip, 0659). Human peripheral blood mononuclear cells (PBMC) were frozen in liquid ethane (KF-80, Leica), and stored in liquid nitro- isolated from Vacutainer Cell Preparation Tubes by separation on a gen. Cell replicas were prepared with a Balzers 400T freeze frac- Ficoll gradient. CD4+ T cells were obtained by negative selection ture (FF) unit, mounted on copper grids, and analyzed on a 14 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal JEM1010 electron microscope (Jeol, Japan) operating at 80 kV. Bio-Rad; overnight, 4°C). Proteoliposomes were collected by ultra- Images were taken with a CCD camera (Bioscan, Gatan, Pleasan- centrifugation (180,000 g, 4 h, 4°C), lysed, and analyzed by BN- ton, CA) and processed with TVIPS software (TVIPS, Gauting, PAGE as above. DE). EM images were collected by two researchers, one of them blind to the experiment. Gold particles were counted on the Sphingolipid and Chol quantification computer. When distance between gold particles was smaller than their diameter (10 nm), they were considered part of the same Total Chol level was measured with the Amplex Red Cholesterol cluster. Assay Kit (Invitrogen) after lysis (50 mM Tris–HCl pH 8; 150 mM NaCl, 1% NP-40). Sphingolipid determinations were performed by BN-PAGE analysis of TCR clustering an external researcher blind to the experimental groups. Prior to sphingolipid quantification, calibration curves were prepared with Membrane fractions from OT-II WT and CCR5/ cells (20 × 106) mixtures of C12Cer, C16Cer, C18Cer, C24Cer, C24:1 Cer, were prepared with a Dounce homogenizer, followed by incubation C16dhCer, C18dhCer, C24dhCer, C24:1dhCer, C12SM, C16SM, in hypotonic buffer (10 mM HEPES pH 7.4, 42 mM KCl, 5 mM C18SM, C24SM, and C24:1SM. For sphingolipid determination, MgCl2, protease inhibitors). Membranes were recovered by ultracen- cell pellets (1 × 106) were mixed with internal standards (N-dode- trifugation (100,000 g, 45 min, 4°C) and lysed in 150 ll BN lysis canoyl-sphingosine, N-dodecanoylglucosylsphingosine, N-dodeca- buffer (500 mM Bis-Tris 40 mM pH 7.0, 1 mM e-aminocaproic acid, noyl-sphingosylphosphorylcholine, C17-sphinganine, and C17- 40 mM NaCl, 20% glycerol, 4 mM EDTA, and 0.5% Brij96 or 1% sphinganine-1 phosphate; 0.2 nmol each; Avanti Polar Lipids) in a digitonin) with protease inhibitors. BN-PAGE gradient gels (4–8%) methanol:chloroform solution. Sphingolipids were extracted as were prepared and used as described (Swamy & Schamel, 2009), described (Merrill et al, 2005), solubilized in methanol, and using ferritin 24-mer and 48-mer (f1, 440 kDa; f2, 880 kDa) as analyzed by ultra-performance liquid chromatography (UPLC; protein markers. Proteins were transferred to PVDF membranes and Waters, Milford, MA) connected to a time-of-flight detector (TOF; probed with anti-CD3f antibody. LCT Premier XE) controlled by Waters/Micromass MassLynx soft- ware. Lipid species were identified based on accurate mass Restimulation assays measurement with an error < 5 ppm, and their LC retention time was compared with the standard ( 2) (Muñoz-Olaya et al, Splenocytes from CD3e/ mice were irradiated (15 Gy), seeded 2008). (0.6 × 105 cells/well), and loaded (2 h, 37°C) with different concentrations of OVA323–339 peptide. After centrifugation (300 g, SMase treatment 5 min), isolated lymphoblasts (0.75 × 105 cells/well) were co- cultured for 48 h. Supernatants were collected to measure IL-2 by All experiments (sphingolipid quantification, apoptosis, and TCR ELISA (ELISA MAX Deluxe, BioLegend) and proliferation was nanoclustering) were performed by incubating OT-II WT and assessed by methyl-3[H]-thymidine (1 lCi/well) incorporation into CCR5/ cells (0.2 × 106) with recombinant sphingomyelinase from DNA, in a 1450 Microbeta liquid scintillation counter (Perki- Bacillus cereus (0.5 U/ml; 1 h, 37°C) in serum-free medium. Cells nElmer). were washed and processed immediately for EM analysis or for sphingolipid quantification as above. TCR purification Quantitative RT–PCR analyses The TCR fused to streptavidin-binding peptide was purified from M.mf-SBP cells. Briefly, 100 × 106 cells were lysed in lysis buffer Total RNA was extracted from human or murine cells using the (20 mM Bis-Tris pH 7, 500 mM f-aminocaproic acid, 20 mM NaCl, RNeasy Mini Kit (Qiagen), and cDNA was synthesized from 1 lg 2 mM EDTA, 10% glycerol, 1% digitonin). After incubation of the total RNA (High Capacity cDNA Reverse Transcription Kit, lysate with streptavidin-conjugated agarose (overnight, 4°C), the Promega). Quantitative RT–PCR was performed using FluoCycle II TCR was eluted by incubating samples with 2 mM biotin in lysis SYBR Master Mix (EuroClone) with specific primers buffer (30 min, 4°C). (Appendix Table S3) in an ABI 7300 Real-Time PCR System (Applied Biosystems). Results were analyzed using SDS2.4 soft- Preparation of large unilamellar vesicles and TCR reconstitution ware. Large unilamellar vesicles with a custom lipid composition were CerS2 silencing prepared by the thin film method (Molnar et al, 2012), followed by extrusion through polycarbonate membranes with a pore size Lentiviruses were produced in HEK-293T cells after co-transfection of 200 nm (21 times) and 80 nm (51 times). The diameter of the with pGIPZ-shRNA-CerS2 or control plasmids, pSPAX2 and pMD2.G resulting LUV was determined by dynamic light scattering (Zeta- (VSV-G protein) using LipoD293tm (SignaGen). Supernatants were master S, Malvern Instruments). The LUV preparation (2 mM) concentrated by ultracentrifugation and supplemented with poly- was mixed with purified TCR (0.1 lg) in 100 ll saline-phosphate brene (8 lg/ml). Lymphoblasts (3 days post-activation) or 2B4 cells buffer with 0.02% Triton X-100, and 40 ll 0.01% Triton X-100 (1.5 × 106 cells/ml) were resuspended in lentiviral supernatant and was added. Samples were agitated (30 min, 4°C), and detergent centrifuged (900 g, 90 min, 37°C). Transduction efficiency was was removed by adsorption to polystyrene Bio-Beads SM-2 (3 mg; analyzed after 24 h by FACS. In the case of 2B4 cells, transduced ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 15 of 19 The EMBO Journal Ana Martín-Leal et al cells were selected with puromycin (2 lg/ml) for 3 days prior to properties of the receptor on the membrane, but not existing clus- analyses. ter size. That is, Immunofluorescence analyses q q q q q q 1 2 3 ... n1 n n þ 1. . . qþ qþ qþ qþ qþ qþ OT-II 10-day WT or CCR5/ lymphoblasts were plated in poly-L- lysine-coated coverslips (Nunc Lab-Tek Chamber Slide, Thermo The “affinity” of the process is given by b = q+/q, which we Scientific; 50 lg/ml, overnight, 4°C). After adhesion (1 h, 37°C), also refer to as the clustering or affinity parameter. In the steady cells were fixed in 4% PFA (10 min), Triton-X100-permeabilized state, we can calculate analytically the fraction of clusters of a given (0.3% in PBS, 15 min), and blocked with BSA 0.5% in PBS. size n: Samples were incubated (overnight, 4°C) with anti-mouse phospho- GATA1pSer142 antibody (1/200), followed by anti-rabbit Ig Alexa-488 bn1 ð1  bÞ pn ¼ secondary antibody (1 h). Coverslips were mounted in Fluoro- ð1  bNmax Þ mount-G with DAPI (Southern Biotech); images were acquired with a Zeiss LSM710 and analyzed by a blind observer with NIH ImageJ with software. b \ 1; n  f1; 2; 3; . . .; Nmax g: ChIP assay The model was fitted using the Bayesian JAGS code (Kruschke, Chromatin immunoprecipitation assays were performed with the 2014) (see Appendix Supplementary Methods). The histograms for EZ-ChIP Kit (Millipore). In brief, OT-II WT or CCR5/ lympho- the number of clusters of a given size n (Nn) were modeled as a blasts (2 × 107) were fixed (1% PFA, 10 min, RT) and quenched multinomial distribution with the number of observations, N, given (125 mM glycine, 5 min, RT). Cells were harvested (1 × 107 cells/ by the total count per experiment, and probabilities pn given by the ml), lysed (15 min, 4°C), and DNA sheared by sonication (45 formulas above. The priors for the clustering parameter b are beta cycles; 30 s on/off; Bioruptor Pico, Diagenode) in aliquots (0.2 ml). distributions with shape parameters A and B with non-informative Of each lysate, 1% was stored as input reference, and the remain- uniform priors. Specifically, ing material was immunoprecipitated (14 h, 4°C, with rotation) with antibodies to GATA1, histone H3-Lys9, or purified IgG (con- Nn ~ Multinomial (pn, N) trol). Immune complexes were captured using Protein G Magnetic b ~ Beta (A, B) Beads (Bio-Rad) and, after washing, eluted with 100 mM NaHCO3, A ~ Uniform (0, 1,000) 1% SDS; protein/DNA bonds were disrupted with proteinase K B ~ Uniform (0, 1,000) (10 lg/ll, 2 h, 62°C). DNA was purified using spin columns, and Cers2 gene promoter sequences were analyzed with specific Posterior distribution of the estimated clustering parameter, b, primers (Appendix Table S3). The relative quantity of amplified is given with the so-called Region of Practical Equivalent (ROPE), product in the input and ChIP samples was calculated (Mira et al, defined as the probability of a parameter from a dataset to be 2018). explained by another dataset. ROPE quantifies the probability that the observed clustering parameter (and distribution of clus- Cell migration and Ca2+ flux assays ters) in the experiment can be obtained by pure random proxim- ity. OT-II WT or CCR5/ lymphoblasts (106) were added to the upper At the molecular level, the kinetic rates q+ and q can be þ= chamber of a Transwell (3-lm pore; Corning) and allowed to expressed in terms of the diffusion rates of the receptors,kd , the migrate toward 100 nM CCL4 for 4 h. Migrating cells were quanti- receptor size (a), the mean distance between receptors (s), and the fied by flow cytometry (Cytomics FC500). Mobilization of intracellu- correct receptor–receptor binding rates, k+/, through the equations lar Ca2+ stores after CCL4 (100 nM) stimulation was measured as (Lauffenburger & DeLisi, 1983): reported (Gómez-Moutón et al, 2015). kþ dk þ k dk  kþ kþ qþ ¼ q ¼ ) b ¼ d  Mathematical and Bayesian analyses kþ d þ k kþ dþk  kd þ k To analyze cluster size distribution, we used a standard chi-square with test to compare the fraction of clusters of a given size (1, 2, 3, etc.) in each dataset. In all plots, “Random” refers to synthetic distribu- 4pD 2pD tions of receptors generated randomly. kþ d ¼ k d ¼ : logðs=aÞ  3=4 s2 ðlogðs=aÞ  3=4Þ To quantify the mechanistic relevance of cluster size between random distributions of clusters and clusters in WT and in CCR5/ CD4+ T cells, we used a Bayesian inference model on The clustering parameter b is independent of the TCR diffusivity top of a mechanistic model (Castro et al, 2014). The model (as D is canceled), so the observed TCR nanoclustering differences assumes that TCR aggregates by incorporating one receptor at a for WT and CCR5/ cells would be due to TCR-TCR interactions, time, with on and off rates that depend on the diffusion as previously reported (Beck-Garcia et al, 2015). 16 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal Identification of transcription factors in CerS promoters and RB wrote the manuscript. All authors read, discussed, and edited the manuscript. Cers2, Cers3, Cers4, and Cers6 gene coordinates were obtained from the UCSC Genome browser (https://genome.ucsc.edu/; mouse Conflict of interest genome version GRCm38/mm10). Known transcription factors for The authors declare that they have no conflict of interest. these genes were identified at GTRD v17.04 (http://gtrd17-04.b iouml.org/). Venn diagrams were constructed to identify common and specific transcription factors for ceramide synthase genes. References Statistical analyses Alonso A, Goñi FM (2018) The physical properties of ceramides in membranes. Annu Rev Biophys 47: 633 – 654 For comparison between two conditions, data were analyzed using Beck-Garcia K, Beck-Garcia E, Bohler S, Zorzin C, Sezgin E, Levental I, parametric Student’s t-tests, paired when different treatments were Alarcon B, Schamel WW (2015) Nanoclusters of the resting T cell applied to the same sample, or unpaired with Welch’s correction. antigen receptor (TCR) localize to non-raft domains. Biochim Biophys Multiple parametric comparisons were analyzed with one-way Acta 1853: 802 – 809 ANOVA with Bonferroni’s post-hoc test. The chi-square test was Bedognetti D, Spivey TL, Zhao Y, Uccellini L, Tomei S, Dudley ME, Ascierto ML, used to analyze overall distribution of gold particles. F test was used De Giorgi V, Liu Q, Delogu LG et al (2013) CXCR3/CCR5 pathways in to compare variances. All analyses were performed using Prism 6.0 metastatic melanoma patients treated with adoptive therapy and or 7.0 software (GraphPad). Differences were considered significant interleukin-2. Br J Cancer 109: 2412 – 2423 when P < 0.05. Blanpain C, Libert F, Vassart G, Parmentier M (2002) CCR5 and HIV infection. Receptors Channels 8: 19 – 31 Camargo JF, Quiñones MP, Mummidi S, Srinivas S, Gaitan AA, Begum K, Data and code availability Jimenez F, VanCompernolle S, Unutmaz D, Ahuja SS et al (2009) CCR5 expression levels influence NFAT translocation, IL-2 production, and This study includes no data deposited in external repositories. The subsequent signaling events during T lymphocyte activation. J Immunol authors confirm that all relevant data and materials supporting the 182: 171 – 182 findings of this study are available on reasonable request. This Castellino F, Huang AY, Altan-Bonnet G, Stoll S, Scheinecker C, Germain RN excludes materials obtained from other researchers, who must (2006) Chemokines enhance immunity by guiding naive CD8+ T cells to provide their consent for transfer. The Bayesian JAGS code gener- sites of CD4+ T cell-dendritic cell interaction. Nature 440: 890 – 895 ated in the study is provided as supplementary information in the Castro BM, Silva LC, Fedorov A, de Almeida RF, Prieto M (2009) Cholesterol- Appendix Supplementary Methods. rich fluid membranes solubilize ceramide domains: implications for the structure and dynamics of mammalian intracellular and plasma Expanded View for this article is available online. membranes. J Biol Chem 284: 22978 – 22987 Castro M, van Santen HM, Ferez M, Alarcon B, Lythe G, Molina-Paris C (2014) Acknowledgements Receptor pre-clustering and T cell responses: insights into molecular We thank D Sancho and JW Yewdell for rVACV-OVA virus, RM Peregil for technical mechanisms. Front Immunol 5: 132 assistance, MC Moreno and S Escudero for flow cytometry services (CNB), MT Contento RL, Molon B, Boularan C, Pozzan T, Mañes S, Marullo S, Viola A Rejas and M Guerra for EM service (CBM-SO), and C Mark for excellent editorial (2008) CXCR4-CCR5: a couple modulating T cell functions. Proc Natl Acad assistance. Fundació ACE would like to thank patients, controls, and the staff Sci USA 105: 10101 – 10106 who participated in this project. This work was funded by grants from the Span- Crotty S (2011) Follicular helper CD4 T cells (TFH). Annu Rev Immunol 29: ish Ministerio de Ciencia, Innovación y Universidades (SAF2017–83732-R to SM; 621 – 663 FIS2016-78883-C2-2-P to MC; CTQ2017-85378-R; AEI/FEDER, EU), the Instituto de Dawson TC, Beck MA, Kuziel WA, Henderson F, Maeda N (2000) Contrasting Salud Carlos III (ISCIII) (PI13/02434, PI16/01861 to AR), the Comunidad de Madrid effects of CCR5 and CCR2 deficiency in the pulmonary inflammatory (B2017/BMD-3733; IMMUNOTHERCAN-CM to SM), and the Merck-Salud Founda- response to influenza A virus. Am J Pathol 156: 1951 – 1959 tion (to SM). WWS and CD were supported by the Deutsche Forschungsgemein- DeJarnette JB, Sommers CL, Huang K, Woodside KJ, Emmons R, Katz K, Shores schaft (DFG) through BIOSS-EXC294 and CIBSS-EXC 2189 (Project 390939984), EW, Love PE (1998) Specific requirement for CD3epsilon in T cell SCHA976/7-1, and SFB1381. The Genome Research @ Fundació ACE project development. Proc Natl Acad Sci USA 95: 14909 – 14914 (GR@ACE) is supported by the Fundación Bancaria La Caixa, Grifols SA, Fundació Dolan MJ, Kulkarni H, Camargo JF, He W, Smith A, Anaya JM, Miura T, Hecht ACE, and ISCIII (Ministry of Health, Spain). Fundació ACE is a participating center FM, Mamtani M, Pereyra F et al (2007) CCL3L1 and CCR5 influence cell- in the Dementia Genetics Spanish Consortium (DEGESCO). mediated immunity and affect HIV-AIDS pathogenesis via viral entry- independent mechanisms. Nat Immunol 8: 1324 – 1336 Author contributions Ellwanger JH, Chies JAB (2019) Host immunogenetics in tick-borne SM conceived the study. AM-L and RB designed, performed most experiments, encephalitis virus infection-The CCR5 crossroad. Ticks Tick Borne Dis 10: and interpreted data. JC and GF performed lipid analyses, and CD performed 729 – 741 liposome experiments. ER-B performed analysis of TCR chains. MC carried out Falcon A, Cuevas MT, Rodriguez-Frandsen A, Reyes N, Pozo F, Moreno S, mathematical simulations and Bayesian analyses. MES carried out bioinfor- Ledesma J, Martinez-Alarcon J, Nieto A, Casas I (2015) CCR5 deficiency matic analyses. IR, LMR, and AR performed ccr5D32 genotyping and provided predisposes to fatal outcome in influenza virus infection. J Gen Virol 96: samples. WWAS, HMS, and BA contributed ideas and technical support. SM 2074 – 2078 ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 17 of 19 The EMBO Journal Ana Martín-Leal et al Floto RA, MacAry PA, Boname JM, Mien TS, Kampmann B, Hair JR, Huey OS, Laviad EL, Albee L, Pankova-Kholmyansky I, Epstein S, Park H, Merrill AH Jr, Houben EN, Pieters J, Day C et al (2006) Dendritic cell stimulation Futerman AH (2008) Characterization of ceramide synthase 2: tissue by mycobacterial Hsp70 is mediated through CCR5. Science 314: distribution, substrate specificity, and inhibition by sphingosine 1- 454 – 458 phosphate. J Biol Chem 283: 5677 – 5684 Franciszkiewicz K, Le Floc’h A, Jalil A, Vigant F, Robert T, Vergnon I, Lee YH, Bae SC, Song GG (2013) Association between the chemokine receptor Mackiewicz A, Benihoud K, Validire P, Chouaib S et al (2009) Intratumoral 5 delta32 polymorphism and rheumatoid arthritis: a meta-analysis. Mod induction of CD103 triggers tumor-specific CTL function and CCR5- Rheumatol 23: 304 – 310 dependent T-cell retention. Cancer Res 69: 6249 – 6255 Lillard JW Jr, Boyaka PN, Taub DD, McGhee JR (2001) RANTES potentiates Fu W, Ergun A, Lu T, Hill JA, Haxhinasto S, Fassett MS, Gazit R, Adoro S, antigen-specific mucosal immune responses. J Immunol 166: 162 – 169 Glimcher L, Chan S et al (2012) A multiply redundant genetic switch ‘locks Lillemeier BF, Mortelmaier MA, Forstner MB, Huppa JB, Groves JT, Davis MM in’ the transcriptional signature of regulatory T cells. Nat Immunol 13: (2010) TCR and Lat are expressed on separate protein islands on T cell 972 – 980 membranes and concatenate during activation. Nat Immunol 11: 90 – 96 Garcia-Arribas AB, Alonso A, Goñi FM (2016) Cholesterol interactions with Lim JK, Murphy PM (2011) Chemokine control of West Nile virus infection. ceramide and sphingomyelin. Chem Phys Lipids 199: 26 – 34 Exp Cell Res 317: 569 – 574 Gómez-Moutón C, Fischer T, Peregil RM, Jiménez-Baranda S, Stossel TP, Mañes S, Lacalle RA, Gómez-Moutón C, del Real G, Mira E, Martínez A-C Nakamura F, Mañes S (2015) Filamin A interaction with the CXCR4 third (2001) Membrane raft microdomains in chemokine receptor function. intracellular loop regulates endocytosis and signaling of WT and WHIM- Semin Immunol 13: 147 – 157 like receptors. Blood 125: 1116 – 1125 Mañes S, Mira E, Colomer R, Montero S, Real LM, Gomez-Mouton C, Jimenez- González-Martín A, Gómez L, Lustgarten J, Mira E, Mañes S (2011) Maximal T Baranda S, Garzon A, Lacalle RA, Harshman K et al (2003) CCR5 expression cell-mediated antitumor responses rely upon CCR5 expression in both influences the progression of human breast cancer in a p53-dependent CD4+ and CD8+ T cells. Cancer Res 71: 5455 – 5466 manner. J Exp Med 198: 1381 – 1389 Gonzalez-Martin A, Mira E, Mañes S (2012) CCR5 in cancer immunotherapy: Martín-Blanco N, Blanco R, Alda-Catalinas C, Bovolenta ER, Oeste CL, Palmer more than an “attractive” receptor for T cells. Oncoimmunology 1: E, Schamel WW, Lythe G, Molina-París C, Castro M et al (2018) A window 106 – 108 of opportunity for cooperativity in the T Cell Receptor. Nat Commun 9: Hu YS, Cang H, Lillemeier BF (2016) Superresolution imaging reveals 2618 nanometer- and micrometer-scale spatial distributions of T-cell receptors Martínez-Martín N, Risueño RM, Morreale A, Zaldivar I, Fernández-Arenas E, in lymph nodes. Proc Natl Acad Sci USA 113: 7201 – 7206 Herranz F, Ortiz AR, Alarcón B (2009) Cooperativity between T cell Huang J, Brameshuber M, Zeng X, Xie J, Li QJ, Chien YH, Valitutti S, Davis MM receptor complexes revealed by conformational mutants of CD3epsilon. Sci (2013) A single peptide-major histocompatibility complex ligand triggers Signal 2: ra43 digital cytokine secretion in CD4(+) T cells. Immunity 39: 846 – 857 Martinez-Muñoz L, Rodriguez-Frade JM, Barroso R, Sorzano COS, Torreno- Jung Y, Riven I, Feigelson SW, Kartvelishvily E, Tohya K, Miyasaka M, Alon R, Pina JA, Santiago CA, Manzo C, Lucas P, Garcia-Cuesta EM, Gutierrez E Haran G (2016) Three-dimensional localization of T-cell receptors in et al (2018) Separating actin-dependent chemokine receptor relation to microvilli using a combination of superresolution microscopies. nanoclustering from dimerization indicates a role for clustering in CXCR4 Proc Natl Acad Sci USA 113: E5916 – E5924 signaling and function. Mol Cell 70: 106 – 119.e10 Kersh EN, Kaech SM, Onami TM, Moran M, Wherry EJ, Miceli MC, Ahmed R Mellado M, Rodriguez-Frade JM, Vila-Coro AJ, Fernandez S, Martin de Ana A, (2003) TCR signal transduction in antigen-specific memory CD8 T cells. J Jones DR, Toran JL, Martinez A-C (2001) Chemokine receptor homo- or Immunol 170: 5455 – 5463 heterodimerization activates distinct signaling pathways. EMBO J 20: Kimachi K, Croft M, Grey HM (1997) The minimal number of antigen-major 2497 – 2507 histocompatibility complex class II complexes required for activation of Merrill AH Jr, Sullards MC, Allegood JC, Kelly S, Wang E (2005) naive and primed T cells. Eur J Immunol 27: 3310 – 3317 Sphingolipidomics: high-throughput, structure-specific, and quantitative Kitatani K, Idkowiak-Baldys J, Hannun YA (2008) The sphingolipid salvage analysis of sphingolipids by liquid chromatography tandem mass pathway in ceramide metabolism and signaling. Cell Signal 20: spectrometry. Methods 36: 207 – 224 1010 – 1018 Mira E, Carmona-Rodriguez L, Perez-Villamil B, Casas J, Fernandez-Acenero Kohlmeier JE, Miller SC, Smith J, Lu B, Gerard C, Cookenham T, Roberts AD, MJ, Martinez-Rey D, Martin-Gonzalez P, Heras-Murillo I, Paz-Cabezas M, Woodland DL (2008) The chemokine receptor CCR5 plays a key role in the Tardaguila M et al (2018) SOD3 improves the tumor response to early memory CD8+ T cell response to respiratory virus infections. chemotherapy by stabilizing endothelial HIF-2alpha. Nat Commun 9: 575 Immunity 29: 101 – 113 Molnar E, Swamy M, Holzer M, Beck-Garcia K, Worch R, Thiele C, Guigas G, Kruschke JK (2014) Doing Bayesian data analysis: a tutorial with R, JAGS, and Boye K, Luescher IF, Schwille P et al (2012) Cholesterol and sphingomyelin Stan. Burlington, MA: Academic Press, Elsevier drive ligand-independent T-cell antigen receptor nanoclustering. J Biol Kumar A, Humphreys TD, Kremer KN, Bramati PS, Bradfield L, Edgar CE, Chem 287: 42664 – 42674 Hedin KE (2006) CXCR4 physically associates with the T cell receptor to Molon B, Gri G, Bettella M, Gomez-Mouton C, Lanzavecchia A, Martinez AC, signal in T cells. Immunity 25: 213 – 224 Mañes S, Viola A (2005) T cell costimulation by chemokine receptors. Nat Kumar R, Ferez M, Swamy M, Arechaga I, Rejas MT, Valpuesta JM, Schamel Immunol 6: 465 – 471 WW, Alarcon B, van Santen HM (2011) Increased sensitivity of antigen- Muñoz-Olaya JM, Matabosch X, Bedia C, Egido-Gabas M, Casas J, Llebaria A, experienced T cells through the enrichment of oligomeric T cell receptor Delgado A, Fabriás G (2008) Synthesis and biological activity of a novel complexes. Immunity 35: 375 – 387 inhibitor of dihydroceramide desaturase. ChemMedChem 3: 946 – 953 Lauffenburger D, DeLisi C (1983) Cell surface receptors: physical chemistry Nesbeth YC, Martinez DG, Toraya S, Scarlett UK, Cubillos-Ruiz JR, and cellular regulation. Int Rev Cytol 84: 269 – 302 Rutkowski MR, Conejo-Garcia JR (2010) CD4+ T cells elicit host immune 18 of 19 The EMBO Journal 39: e104749 | 2020 ª 2020 The Authors Ana Martín-Leal et al The EMBO Journal responses to MHC class II-negative ovarian cancer through CCL5 domains, and generation of gel regions in giant lipidic vesicles. FEBS Lett secretion and CD40-mediated licensing of dendritic cells. J Immunol 582: 3230 – 3236 184: 5654 – 5662 Stiban J, Perera M (2015) Very long chain ceramides interfere with C16- Nie C, Sato K, Misawa N, Kitayama H, Fujino H, Hiramatsu H, Heike T, ceramide-induced channel formation: a plausible mechanism for Nakahata T, Tanaka Y, Ito M et al (2009) Selective infection of CD4+ regulating the initiation of intrinsic apoptosis. Biochim Biophys Acta 1848: effector memory T lymphocytes leads to preferential depletion of memory 561 – 567 T lymphocytes in R5 HIV-1-infected humanized NOD/SCID/IL- Sundrud MS, Vancompernolle SE, Eger KA, Bruno TC, Subramaniam A, 2Rgammanull mice. Virology 394: 64 – 72 Mummidi S, Ahuja SK, Unutmaz D (2005) Transcription factor GATA-1 Pageon SV, Tabarin T, Yamamoto Y, Ma Y, Nicovich PR, Bridgeman JS, Cohnen potently represses the expression of the HIV-1 coreceptor CCR5 in human A, Benzing C, Gao Y, Crowther MD et al (2016) Functional role of T-cell T cells and dendritic cells. Blood 106: 3440 – 3448 receptor nanoclusters in signal initiation and antigen discrimination. Proc Swamy M, Schamel WW (2009) Purification of the T cell antigen receptor Natl Acad Sci USA 113: E5454 – E5463 and analysis by blue-native PAGE. Methods Mol Biol 514: 135 – 150 Ramstedt B, Slotte JP (2002) Membrane properties of sphingomyelins. FEBS Toson B, Dos Santos EJ, Adelino JE, Sandrin-Garcia P, Crovella S, Louzada- Lett 531: 33 – 37 Junior P, Oliveira RD, Pedroza LS, de Fatima Lobato Cunha Sauma M, de Richer MJ, Pewe LL, Hancox LS, Hartwig SM, Varga SM, Harty JT (2015) Lima CP et al (2017) CCR5Delta32 and the genetic susceptibility to Inflammatory IL-15 is required for optimal memory T cell responses. J Clin rheumatoid arthritis in admixed populations: a multicentre study. Invest 125: 3477 – 3490 Rheumatology (Oxford) 56: 495 – 497 Schamel WW, Arechaga I, Risueno RM, van Santen HM, Cabezas P, Risco C, Ugurel S, Schrama D, Keller G, Schadendorf D, Brocker EB, Houben R, Zapatka Valpuesta JM, Alarcon B (2005) Coexistence of multivalent and M, Fink W, Kaufman HL, Becker JC (2008) Impact of the CCR5 gene monovalent TCRs explains high sensitivity and wide range of response. J polymorphism on the survival of metastatic melanoma patients receiving Exp Med 202: 493 – 503 immunotherapy. Cancer Immunol Immunother 57: 685 – 691 Schamel WW, Risueño RM, Minguet S, Ortiz AR, Alarcon B (2006) A Vangelista L, Vento S (2017) The expanding therapeutic perspective of CCR5 conformation- and avidity-based proofreading mechanism for the TCR- blockade. Front Immunol 8: 1981 CD3 complex. Trends Immunol 27: 176 – 182 Vieira CR, Muñoz-Olaya JM, Sot J, Jimenez-Baranda S, Izquierdo-Useros N, Schamel WW, Alarcon B (2013) Organization of the resting TCR in nanoscale Abad JL, Apellaniz B, Delgado R, Martinez-Picado J, Alonso A et al (2010) oligomers. Immunol Rev 251: 13 – 20 Dihydrosphingomyelin impairs HIV-1 infection by rigidifying liquid-ordered Schauren JS, Marasca JA, Veit TD, Monticielo OA, Xavier RM, Brenol JC, Chies membrane domains. Chem Biol 17: 766 – 775 JA (2013) CCR5delta32 in systemic lupus erythematosus: implications for Vinuesa CG, Linterman MA, Yu D, MacLennan IC (2016) Follicular helper T disease susceptibility and outcome in a Brazilian population. Lupus 22: cells. Annu Rev Immunol 34: 335 – 368 802 – 809 Wang F, Beck-Garcia K, Zorzin C, Schamel WW, Davis MM (2016) Inhibition Sherman E, Barr V, Manley S, Patterson G, Balagopalan L, Akpan I, Regan CK, of T cell receptor signaling by cholesterol sulfate, a naturally occurring Merrill RK, Sommers CL, Lippincott-Schwartz J et al (2011) Functional derivative of membrane cholesterol. Nat Immunol 17: 844 – 850 nanoscale organization of signaling molecules downstream of the T cell Xu Y, Phetsouphanh C, Suzuki K, Aggrawal A, Graff-Dubois S, Roche M, Bailey antigen receptor. Immunity 35: 705 – 720 M, Alcantara S, Cashin K, Sivasubramaniam R et al (2017) HIV-1 and SIV Smith X, Schneider H, Kohler K, Liu H, Lu Y, Rudd CE (2013) The chemokine predominantly use CCR5 expressed on a precursor population to establish CXCL12 generates costimulatory signals in T cells to enhance phos- infection in T follicular helper cells. Front Immunol 8: 376 phorylation and clustering of the adaptor protein SLP-76. Sci Signal 6: ra65 Sofi MH, Heinrichs J, Dany M, Nguyen H, Dai M, Bastian D, Schutt S, Wu Y, License: This is an open access article under the Daenthanasanmak A, Gencer S et al (2017) Ceramide synthesis regulates T terms of the Creative Commons Attribution 4.0 cell activity and GVHD development. JCI Insight 2: e91701 License, which permits use, distribution and reproduc- Sot J, Ibarguren M, Busto JV, Montes LR, Goni FM, Alonso A (2008) Cholesterol tion in any medium, provided the original work is displacement by ceramide in sphingomyelin-containing liquid-ordered properly cited. ª 2020 The Authors The EMBO Journal 39: e104749 | 2020 19 of 19