Correlate with Their Function Levels of Membrane Lipid Order That T

Primary Human CD4+ T Cells Have Diverse Levels of Membrane Lipid Order That Correlate with Their Function

This information is current as Laura Miguel, Dylan M. Owen, Chrissie Lim, Christian of October 4, 2021. Liebig, Jamie Evans, Anthony I. Magee and Elizabeth C. Jury J Immunol 2011; 186:3505-3516; Prepublished online 9 February 2011; doi: 10.4049/jimmunol.1002980 http://www.jimmunol.org/content/186/6/3505 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2011/02/09/jimmunol.100298

Material 0.DC1 http://www.jimmunol.org/ References This article cites 53 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/186/6/3505.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on October 4, 2021

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2011 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Primary Human CD4+ T Cells Have Diverse Levels of Membrane Lipid Order That Correlate with Their Function

Laura Miguel,* Dylan M. Owen,†,1 Chrissie Lim,* Christian Liebig,†,2 Jamie Evans,* Anthony I. Magee,† and Elizabeth C. Jury*

Membrane lipid microdomains (lipid rafts) play an important role in T cell function by forming areas of high lipid order that facilitate activation. However, their role in regulating T cell differentiation and function remains controversial. In this study, by applying a new approach involving microscopy and flow cytometry, we characterize membrane lipid order in ex vivo primary human CD4+ T cells. We reveal that differential membrane lipid order dictates the response to TCR stimulation. T cells with high membrane order formed stable immune synapses and proliferated robustly, intermediate order cells had reduced proliferative ability accompanied by unstable immune synapse formation, whereas low order T cells were profoundly unresponsive to TCR activation. We also observed that T cells from patients with autoimmune rheumatic disease had expanded intermediate order Downloaded from populations compared with healthy volunteers. This may be important in dictating the nature of the immune response since most IFN-g+CD4+ T cells were confined within intermediate membrane order populations, whereas IL-4+CD4+ T cells were contained within the high order populations. Importantly, we were able to alter T cell function by pharmacologically manipulating membrane order. Thus, the results presented from this study identify that ex vivo CD4+ T cells sustain a gradient of plasma membrane lipid order that influences their function in terms of proliferation and cytokine production. This could represent a new mechanism to control T cell functional plasticity, raising the possibility that therapeutic targeting of membrane lipid order could direct http://www.jimmunol.org/ altered immune cell activation in pathology. The Journal of Immunology, 2011, 186: 3505–3516.

urrent evidence supports an important role for lipid The organization of plasma membrane sphingolipids and cho- microdomains (lipid rafts) in the formation of the im- lesterol into microdomains with relative liquid-order compared C munological synapse (IS) between T lymphocytes and with the surrounding disordered membrane is the basis of the lipid APC; this process involves the segregation and reorganization of raft hypothesis in mammalian cells (4). Some cell surface proteins membrane lipids and proteins and is dependent on the actin cy- preferentially associate with ordered lipid microdomains whereas toskeleton (1, 2). Although the functional outcome of T cell/APC others are excluded and diffuse freely in the more disordered interactions depends on the nature of IS formation (3), the im- membrane (5). A main issue when considering lipid microdomains by guest on October 4, 2021 portance of membrane microdomains in the regulation of IS de- is their visualization (6) since they are dynamic and of a size too velopment, cell differentiation, and function of primary human small to resolve using conventional microscopy (7). Original T cells is not fully understood. results based on the resistance of ordered lipid microdomains to solubilization with nonionic detergents (so called detergent-resistant membranes) (8) and cross-linking cell surface domains with multivalent probes such as cholera toxin subunit B (CTB) re- *Division of Medicine, Centre for Rheumatology Research, University College Lon- don, London W1P 4JF, United Kingdom; and †Section of Molecular Medicine, Na- vealed that capping of lipid microdomains at the IS following tional Heart & Lung Institute, Imperial College London, South Kensington, London TCR stimulation facilitates coordination, localization, and func- SW7 2AZ, United Kingdom tion of proteins residing proximal to the TCR (4, 5, 9). However, 1 Current address: Centre for Vascular Research, University of New South Wales, there is debate about whether these methods reliably identify or- Sydney, Australia. dered lipid microdomains as they exist in living cells (9, 10). 2Current address: Hertie-Institute for Clinical Investigation, Cell Biology and Neu- rological Disease, Tu¨bingen, Germany. A new approach to their analysis has been to observe ordered and disordered membranes in live T cells, using fluorescent membrane Received for publication September 3, 2010. Accepted for publication January 12, 2011. probes such as LAURDAN and di-4-ANEPPDHQ (ANE). ANE This work was supported by an Arthritis Research UK Career Development award partitions into both liquid-ordered (raft) and liquid-disordered to E.C.J. (18106) and a University College London Hospital Clinical Research and (nonraft) membranes and senses the environmental difference Development Committee project grant (GCT/2008/EJ). A.I.M. is supported by Med- between the two regions. It is water-soluble yet binds to lipid ical Research Council Grant G0700771. membranes with high affinity and is therefore easily loaded into Address correspondence and reprint requests to Dr. Elizabeth Jury, Centre for Rheu- matology Research, University College London, Windeyer Building, 46 Cleveland membranes (11). The incorporation of ANE into hydrophobic Street, London W1P 4JF, United Kingdom. E-mail address: [email protected] (more ordered) and hydrophilic (less ordered) membranes influ- The online version of this article contains supplemental material. ences its interaction with aqueous solution and its subsequent Abbreviations used in this article: ANE, di-4-ANEPPDHQ; CTB, cholera toxin sub- fluorescent emission spectra (11). When ANE is excited in the unit B; GP, generalized polarization; IRM, interference reflection microscopy; IS, blue spectral region with single-photon excitation it exhibits a 60- immune synapse; 7KC, 7-ketocholesterol; PKC, protein kinase C; pY, phosphotyrosine; RA, rheumatoid arthritis; rh, recombinant human; SEA, staphylococcal en- nm spectral blue shift between the disordered and ordered lipid terotoxin A; SEB, staphylococcal enterotoxin B; SLE, systemic lupus erythematosus; phases (11). The degree of membrane order can be calculated SMAC, supramolecular activation complex; SS, Sjo¨rgren’s syndrome; TIRF, total from the dye’s emission properties and expressed as a generalized internal reflection fluorescence. polarization (GP) value (a normalized intensity ratio of two dif- Copyright Ó 2011 by The American Association of Immunologists, Inc. 0022-1767/11/$16.00 ferent spectral channels) (12). This approach has provided evi- www.jimmunol.org/cgi/doi/10.4049/jimmunol.1002980 3506 MEMBRANE LIPID ORDER CORRELATES WITH T CELL FUNCTION dence confirming the importance of ordered lipid microdomains magnetic columns (Milteni Biotec) and labeled with CellTracker Blue in IS formation and T cell function (6, 13–16). CMHC (Invitrogen) following the manufacturer’s protocol before loading Using the lipid probe ANE to identify lipid microdomains in ex with superantigen (1 mg/ml staphylococcal enterotoxin E, 2 mg/ml staphy- + lococcal enterotoxin A [SEA]/staphylococcal enterotoxin B [SEB]) for 1 h at vivo human CD4 T cells, we reveal an array of plasma membrane 37˚C. FACS-sorted low, intermediate, and high order CD4+ Tcellswere lipid order, ranging from low, intermediate, to high order. We labeled with CellTracker Green CMFDA (Invitrogen). APC and T cells were demonstrate that this gradient of membrane lipid order dictates mixed in a ratio of 1:1, briefly centrifuged for 1 min at 100 3 g to form the outcome of CD4+ T cell responses to activation. Upon TCR conjugates, and then incubated in complete RPMI 1640 medium (100 ml) at 37˚C for 5 min. Thereafter, T cell/APC mixtures were fixed in PBS con- stimulation, high order T cells formed stable IS and proliferated taining 1% paraformaldehyde before analysis. The relative proportion of robustly, intermediate order cells had reduced proliferative ability orange, blue, and orange/blue events in each sample was determined. For accompanied by unstable IS formation, but low order T cells were manipulation of membrane order, 15 mg/ml 7-ketocholesterol (7KC) and profoundly unresponsive. Pharmacologically reducing T cell cholesterol (Avanti Lipids, Alabaster, AL) in ethanol were combined in a membrane order from high to intermediate induced unstable IS cholesterol-7KC ratio of 1:2. During 30 min, these were then added to a + solution of 50 mg/ml methyl-b-cyclodextrin (Sigma-Aldrich) in PBS at 80˚C formation and reduced proliferation. Furthermore, most IFN-g to a final sterol concentration of 1.5 mg/ml. Then, 15 ml/ml lipid solution CD4+ T cells were characterized by intermediate order, whereas was added to cell medium containing 1 3 106 cells at 37˚C for 30 min before IL-4+CD4+ T cells were within the high order population. In- labeling with ANE and analysis by flow cytometry and microscopy. terestingly, patients with autoimmunity had increased intermediate Differentiation of Th1 and Th2 phenotypes order T cell populations accompanied by reduced proliferation + 2 2 + and increased production of IFN-g compared with healthy con- FACS for CD4 CD8 CD25 CD45RA T cells were stimulated with anti- trols. Therefore, the results presented in this study identify a new CD3 and anti-CD28 (2 mg/ml each) and cultured in RPMI 1640, 10% (v/v)

FCS, penicillin/streptomycin, and 2 mM L-glutamine. To generate differ- Downloaded from mechanism to control T cell functional plasticity, raising the pos- entiated Th cell responses, Th1 cultures were supplemented with rhIL- sibility that therapeutic targeting of membrane lipid order could 12p70 (10 ng/ml) and anti–IL-4 (10 mg/ml), and Th2 cultures were sup- direct altered immune cell activation in pathology. plemented with anti–IL-12 (10 mg/ml), rhIL-4 (100 U/ml), and anti–IFN-g (10 mg/ml). Cultures were supplemented with rhIL-2 (100 U/ml) on day 4 after the activation of cultures; on day 6 supernatants and cells were re- Materials and Methods covered. Th1/Th2 differentiation was confirmed by cytokine analysis of Cell isolation cell supernatants. http://www.jimmunol.org/ PBMC from 40 healthy donors were separated on Ficoll-Hypaque (Phar- T cell functional assays macia Biotech). PBMC from 40 healthy donors (mean age, 34.7 y; 10 male, 30 female) and 10 patients with active systemic lupus erythematosus T cells sorted for high, intermediate, and low membrane order were (SLE), 10 patients with rheumatoid arthritis (RA), and 5 patients with stimulated with either plate-bound anti-CD3 (2 mg/ml) and anti-CD28 (2 Sjo¨rgren’s syndrome (SS) attending the Rheumatology Clinic at University mg/ml) Abs, Human T-Activator CD3/CD28 beads (Dynabeads; Invi- College Hospital were included in this study after receiving informed trogen), or superantigen-coated (SEA/SEB) autologous APC. Three-day consent. The study was approved by the local ethics committee. Purified culture supernatants were analyzed for cytokine production by Cytokine CD4+ lymphocytes were obtained by negative selection using magnetic Bead Array (BD Biosciences). Intracellular staining for IFN-g, IL-4, IL-2, beads (Miltenyi Biotec) or by cell sorting using FACSAria (BD Bio- and IL-10 was carried out after incubation with 50 ng/ml PMA, 250 ng/ml sciences). ionomycin, and 2 mM monensin (Golgi-Plug; BD Biosciences) for 4 h before fixation/permeabilization. For proliferation, cells were stimulated by guest on October 4, 2021 Abs and reagents with anti-CD3/CD28 (2 mg/ml each) for 3 d then pulsed with [3H]thymidine for 16 h, followed by harvesting and scintillation analysis. Abs for flow cytometry included allophycocyanin-CD4, Pacific Blue-CD4, allophycocyanin-CD3, FITC-CD8, PE-Cy5-CD25, allophycocyanin-CD45RA, Interference reflection microscopy allophycocyanin-H7-CD27, allophycocyanin-CD69, allophycocyanin-CD25, allophycocyanin-Cy7-IFN-g, Pacific Blue-IL-10, allophycocyanin-IL-4, Glass coverslip cell culture chambers (Intracel, London, U.K.) were coated FITC-IL-2, PE-Ki67, allophycocyanin-T-bet, allophycocyanin-pSTAT6, with Abs to CD3/CD28 or isotype control (10 mg/ml) in PBS for a mini- allophycocyanin-annexin V, and fluorochrome-conjugated isotype controls mum of 2 h. Sorted high, intermediate, and low order cells were layered (all from BD Biosciences). Abs for confocal microscopy included anti- onto Ab-coated coverslips at 37˚C. Cells were imaged as they attached to CD3 (OKT3) and anti-phosphotyrosine (pY) (4G10) with secondary anti- the coverslip surface using a Zeiss confocal microscope and 488 nm argon- mouse IgG2a-Alexa Fluor 488 and anti-mouse IgG2b-Alexa Fluor 555 ion laser excitation and a 363 oil objective. Reflected and transmitted light from Invitrogen. For functional experiments and for coating coverslips, images were obtained at 30-s intervals. Images were merged using ImageJ purified Abs to CD3 (HIT3a), CD28 (CD28.2), or IgG isoype controls software to reveal areas of cell attachment. from BD Biosciences were used. For generation of polarized Th1/Th2 populations, anti–IL-4, anti–IL-12, anti–IFN-g (eBioscience), recombi- T cell/APC conjugates nant human (rh)IL-12p70, rhIL-2, and rhIL-4 (R&D Systems) were used. Sorted high, intermediate, and low order cells were cocultured with The membrane order probe di-4-ANEPPDHQ, CTB-biotin, and PP2 superantigen (1 mg/ml staphylococcal enterotoxin E, 2 mg/ml SEB/SEA)- were obtained from Invitrogen. DAPI was obtained from Sigma-Aldrich loaded Raji B cells for 10 min at 37˚C. Conjugates were attached to poly- (St. Louis, MO). L-lysine–coated coverslips and fixed with 2% paraformaldehyde before Flow cytometry permeabilization with 0.2% Triton X-100 for 10 min and blocking with 5% BSA/PBS for 2 h. Cells were stained with anti-CD3 and anti-pY Abs PBMC were stained for protein surface markers before washing and followed by appropriate secondary Abs before mounting onto glass slides resuspending with 4 mM ANE in PBS for up to 30 min at 37˚C. For each using anti-fade mountant (Invitrogen). Conjugates were imaged using 488 population the geometric mean fluorescence intensity for wavelengths at nm argon-ion, 633 nm helium, and 400 nm ultraviolet laser excitation and 570 nm (FL2 channel) and 630 nm (FL3 channel) were used to make the a 363 oil immersion lens using a Zeiss LSM-510 inverted laser-scanning GP calculation (see Equation 1 below). Staining for annexin V was per- confocal microscope. Approximately 10 conjugates were imaged from formed according to manufacturer’s instructions (BD Biosciences) fol- each sample, and 0.5-mm z-slices were acquired through each conjugate. lowing ANE labeling. Cells were analyzed without fixing using a BD LSR The Pearson correlation coefficient of anti-CD3 and anti-pY staining at the II flow cytometer (BD Biosciences) and FlowJo software (Tree Star). IS for each z-slice in the conjugates was measured using Zeiss LSM Staining for intracellular cytokines, Ki67, T-bet, and pSTAT6 was per- software. Three-dimensional reconstructed images were obtained from formed on FACS-sorted populations after fixation/permeabilization with deconvoluted images using Volocity software. either saponin- or methanol-based buffer according to the manufacturer’s Total internal reflection fluorescence (TIRF) microscopy was performed instructions (eBioscience). Cell viability was assessed by annexin V on a custom-built microscope with excitation at 473 nm and a 360, 1.45NA staining according to the manufacturer’s instructions and DAPI exclusion oil-immersion TIRF objective. Fluorescence was collected on an electron- by flow cytometry. For the flow cytometry-based conjugation assay, au- multiplying CCD camera (iXon; Andor, Belfast, U.K.) in the range 500– tologous APC were isolated from PBMC by removal of CD3+ cells using 593 and 600–680 nm using a two-channel imager (Dual-View; Optical The Journal of Immunology 3507

Insights). Data were processed using custom software (LabVIEW; Na- corresponding with the preference for ANE to incorporate into tional Instruments, Austin, TX). The sample was maintained at 37˚C using disordered membranes (13). In contrast, ANE labeling was con- a stage and objective heater. Quantification of image data was performed fined mainly to the plasma membrane in the high order cells, with by measuring average GP values (a normalized intensity ratio of two different spectral channels: green [ordered] and red [disordered]) where intracellular membranes becoming labeled only after 17 min (Fig. high GP (increased green fluorescence compared with red) represents high 1G, arrow). These results were confirmed by flow cytometry (Fig. membrane lipid order (12): 1H), showing that even after 1 min after ANE staining the three distinct populations were revealed and did not change significantly I500 2 570 2 I620 2 750 GP ¼ : ð1Þ over time. Finally, we related the relative level of plasma mem- I500 2 570 þ I620 2 750 brane lipid order to the expression of lipid microdomain-associated Confocal microscopy lipids, cholesterol, GM1, and GM3. Intermediate and high order Sorted high, intermediate, and low order cells were labeled with 4 mM ANE cells had increased expression of cholesterol compared with low for 30 min and applied to coverslip chambers at 37˚C. Attachment of cells order cells (assessed by filipin binding), and a significant positive was imaged at 5-min intervals up to 30 min from addition of cells to the chamber, using single-photon excitation confocal microscopy on an correlation was seen between global lipid order and membrane inverted laser-scanning confocal microscope (TCS SP2; Leica Micro- levels of cholesterol (Fig. 1I, upper panels). Alternatively, GM1 systems, Wetzlar, Germany; or Zeiss LSM-510) with a 363, 1.25NA oil- (measured by CTB binding) was upregulated in the intermediate immersion objective lens. Excitation was at 488 nm from an argon-ion and low order cells, and a significant negative correlation was seen laser. Fluorescence detection was in the wavelength bands 500–580 and between membrane order and CTB binding (Fig. 1I, middle pan- 620–750 nm using internal photomultiplier tubes. GP values were calculated from the two wavelength channels according to Equation 1, where I els). Interestingly, glycosphingolipid GM3 was not differentially is the intensity of emission. The z-slices were acquired every 0.5 mm. expressed between the subsets (Fig 1I, lower panels). These results reveal differences between measurement of global membrane lipid Downloaded from Statistics order and the traditional markers for assessing membrane lipid All values are expressed as mean 6 SEM. We performed analysis of microdomains. significance in Prism (GraphPad Software) by the Mann–Whitney U test, Student t test, or paired t test as appropriate. Differential membrane order is associated with distinct patterns of IS formation

Results http://www.jimmunol.org/ + Because recent work has shown that high lipid order is important Distinct CD4 T cell subsets identified based on plasma for stable IS formation (14, 16), the effect of differential global membrane lipid order plasma membrane order on IS development was examined by To investigate the role of membrane lipids in ex vivo human CD4+ interference reflection microscopy (IRM). FACS-sorted low, in- T cells, we used a new approach; that is, cells were labeled with termediate, and high order T cells were applied to coverslips the lipid probe ANE and analyzed by confocal microscopy. The coated with anti-CD3/CD28 or isotype control and visualized for degree of membrane order was calculated and expressed as a GP up to 10 min. Low order cells demonstrated limited and transient value where high GP represents high membrane lipid order (Fig. attachments compared with partial interactions by intermediate 1A) (12). We observed heterogeneous membrane order in ex vivo order cells (Fig. 2A). In contrast, high order cells exhibited robust by guest on October 4, 2021 negatively purified CD4+ T cells from healthy volunteers. This and symmetrical attachments (Fig. 2A). These results were con- was depicted by a wide spectrum of intensity of ANE staining, firmed by confocal microscopy (Supplemental Fig. 1A). Quanti- revealing cells with relative low (green/blue in the GP image), tation of imaging data verified that high order cells formed a more intermediate (orange/green in the GP image), and high (red in the substantial contact surface area with the coverslip compared with GP image) plasma membrane lipid order. To confirm these find- intermediate and low order cells (Fig. 2B). All three populations ings and to examine a larger number of cells from many indi- possessed similar expression levels of CD3, excluding the possi- viduals, a method was developed combining ANE labeling with bility that differences in CD3 expression might explain the altered multiparameter flow cytometry (Fig. 1B). Again, three populations patterns of synapse formation (Supplemental Fig. 1B). were defined with relative low (1.4 6 1.3% SD), intermediate Given that increased membrane order at the IS facilitates T cell (8.28 6 5.1% SD), and high (90.47 6 5.7% SD) plasma mem- activation (14, 15, 18), a more accurate assessment of lipid order brane order (Fig. 1B,1C,1D). Intermediate and high order cells at the cell/coverslip interface in FACS-sorted intermediate and were viable by DAPI exclusion and annexin V staining (Fig. 1E). high order populations was made using TIRF microscopy (the Low order cells had comparatively reduced viability with ∼50% more transient nature of low order cell interactions made it diffi- being annexin V positive, indicating that many were preapoptotic. cult for them to be visualized using this technique). Representative Furthermore, when FACS-sorted low, intermediate, and high order GP/TIRF images (Fig. 2C) depict areas of highest membrane or- T cells were cultured overnight and relabeled with ANE, they der in red and lowest order in blue/green. High order cells had the maintained their order (Fig. 1F). highest average order at the IS compared with intermediate order The ability of flow cytometry and ANE labeling to distinguish cells (Fig. 2D). The apparent differential nature of IS formation in between plasma compared with intracellular membranes (which are the three populations was also tested in a more physiological largely low order) (13, 17) was verified by assessing ANE labeling T cell/APC system. FACS-sorted low, intermediate, and high order of CD4+ T cells over time. ANE was rapidly incorporated into cells were cocultured with superantigen-loaded Raji B cells, and cellular membranes; plasma membrane order was observed at T cell/APC conjugates were stained with anti-CD3 (green) and early time points and remained unchanged over time as observed anti-pY (red) and analyzed by confocal microscopy. Representa- in the representative merged intensity images of intermediate and tive images are shown in Fig. 2E. CD3 was patched at the IS in high order cells (Fig. 1G). The accompanying RGB profiles rep- both intermediate and high order cells (Fig. 2E,2H) accompanied resent ANE emission spectra detected in 500–570 (green) and by pY accumulation (Fig. 2E,2G,2I). CD3 did not accumulate at 620–750nm (red) channels obtained from equatorial cross-sections the interface between low order T cells and APC (Fig. 2E,2F, of the intermediate and high order cells. They reveal that ANE was 2H), and no pY accumulation was seen (Fig. 2F,2G,2I). Although rapidly incorporated into the plasma and intracellular membranes both the intermediate and high order cells formed functional of the intermediate (Fig. 1G) and low order cells (data not shown), synapses in terms of pY accumulation, differences were seen in 3508 MEMBRANE LIPID ORDER CORRELATES WITH T CELL FUNCTION Downloaded from http://www.jimmunol.org/ by guest on October 4, 2021

FIGURE 1. Identification of distinct CD4+ T cell subsets based on membrane lipid order. Negatively selected CD4+ T cells were labeled with ANE. A, Confocal microscopy showing intensity and GP images, revealing cells with low (***), intermediate (**), and high (*) order. Scale bar, 10 mm. B, Flow cytometry detecting ANE in FL2 (570 nm) and FL3 (630 nm) channels. Three cell populations were gated according to FL2 versus FL3 intensity; a representative dot plot is shown. Cumulative data from 18 healthy volunteers showing (C) percentage CD4+ cells in each population and (D) corresponding GP value depicting membrane order. E, Cell viability determined by DAPI exclusion (left panel) and annexin V binding (right panel) in CD4+ T cells from six healthy donors. F, FACS-sorted ANE-labeled high, intermediate, and low order T cells from three healthy volunteers were cultured for 18 h, relabeled with ANE, and GP was assessed by flow cytometry. G, ANE incorporation into plasma and intracellular membranes. ANE was added to CD4+ T cells and imaged under physiological conditions for 30 min. Representative merged intensity images (green, 500–570 nm and red channels, 620–750 nm) from intermediate and high order cells. Arrows indicate intracellular membrane staining. Scale bars, 5 mm. Right panels show corresponding RGB profiles (green and red channels) obtained in ImageJ software from equatorial cross-section through the cell. H, Negatively isolated CD4+ T cells were labeled with ANE and analyzed by flow cytometry at 1, 7, and 30 min. Results from three healthy control samples showing number of cells (%) in low, intermediate, and high populations. I, FACS-sorted high, intermediate, and low order T cells from 10 healthy individuals were labeled with either filipin, CTB, or anti-GM3 to detect levels of surface lipids. Cumulative results are show in bar graphs. **p # 0.002, *p = 0.05 low versus high. Lipid expression was assessed in whole The Journal of Immunology 3509 Downloaded from http://www.jimmunol.org/ by guest on October 4, 2021

FIGURE 2. Differential IS formation in T cells with low, intermediate, and high membrane order. A, FACS-sorted high, intermediate, and low order CD4+ T cells from six healthy donors interacted with anti–CD3/CD28- or isotype-coated coverslips and imaged by IRM; images were taken at 30-s intervals for 10 min. Representative images of cell/coverslip attachment at 5 min (dark/black color), with unattached cell body outlined in intermediate and low order cells. Scale bar, 5mm. B, Cell contact area (pixels) was measured; cumulative results from ∼50 cells/condition from six individuals are shown. ***p = 0.0001 IgG versus high, *p = 0.01 IgG versus intermediate. C, Cell/coverslip interaction analyzed by TIRF microscopy; representative GP images from high and intermediate order cells 30 min after application to coverslips are shown. Red, high order; blue/green, low order. Scale bars, 5 mm. D, Cumulative TIRF/GP values from ∼25 cells/condition in five individuals. *p = 0.01. E, FACS-sorted ANE-labeled high, intermediate, and low order CD4+ T cells from three individuals interacted with superantigen-loaded Raji B cells. Cells stained for anti-CD3 (green) and anti-pY (red) and imaged by confocal microscopy (∼10 conjugates/sample) are shown. Representative conjugates show differential interference contrast and deconvoluted images for anti-CD3, anti-pY, and merged intensity. Scale bars, 5 mm. F, Representative three-dimensional–rendered images from merged CD3/pY confocal z-stacks at the IS. Scale bars, 1 mm G, Representative Pearson correlation coefficient (R) of red/green colocalization at the IS shown in E and F. H, Pattern of CD3 accumulation at IS for each conjugate; cumulative results showing percentage of cells with CD3 patched at IS. *p = 0.03 high versus intermediate and low order cells. I, Cumulative data of red/green colocalization across each IS (∼10 conjugates/sample). **p # 0.006, high and intermediate versus low order cells. the pattern of CD3 arrangement; significantly fewer intermediate Plasma membrane order is associated with distinct T cell cells formed synapses with distinct CD3 patching (Fig. 2H), and function most intermediate order T cells displayed a characteristic diffuse The dramatic differences in IS formation suggested that membrane pattern of CD3/pY colocalization at the IS (Fig. 2E,2F). Thus, lipid order may influence T cell function following TCR stimu- these results suggest that in ex vivo T cell populations, hetero- lation. IS stability was correlated with the speed of T cell/APC con- geneous membrane order may influence the nature and stability of jugate formation. The three membrane order subsets were FACS- IS formation with APC. sorted and interacted with superantigen-loaded APC and the

CD4+ T cell populations from eight healthy controls by correlating total CD4+ GP with either filipin or CTB binding or staining with anti-GM3 Abs. **r2= 0.0713, p = 0.009; *r2 = 0.728, p = 0.01. 3510 MEMBRANE LIPID ORDER CORRELATES WITH T CELL FUNCTION kinetics of conjugate formation were observed by flow cytometry 3E,3F). A more detailed assessment of cytokine production in (Fig. 3A,3B). These experiments revealed that high order T cells FACS-sorted high, intermediate, and low order T cells revealed form conjugates more slowly than intermediate and low order that TCR stimulation of intermediate order cells preferentially cells, although as shown above, the IS is more stable once it is induced INF-g and IL-6 production compared with high order formed. Changes in IS formation and stability were associated cells that produced more IL-4 and IL-10 (Fig. 3G). CD4+ T cells with differences in cell function. Assessment of low, intermediate, characterized by low lipid order did not proliferate or produce IL- and high membrane order T cells following 3 d coculture with 2 in response to TCR stimulation; furthermore, they were more superantigen-loaded APC revealed that high order T cells dis- prone to cell death following activation (Fig. 3H) but were rescued played elevated levels of the proliferation marker Ki-67 (Fig. by exogenous IL-2 or stimulation by PMA/ionomycin (Fig 3I), 3C), compared with intermediate and low order cells. Differential suggesting an unresponsive or tolerized phenotype (19). proliferation between the three populations was confirmed by Phenotypic characterization of the three populations in terms thymidine incorporation assays following stimulation with anti- of memory and activation marker expression using a panel of sur- CD3/CD28 (Fig. 3D), and it was found to be associated with re- face markers (CD45RA, CD27, CD25, and CD69) and multipara- duced production of IL-2 in the intermediate order T cells (Fig. meter flow cytometry is shown in Supplemental Fig. 2. The results Downloaded from http://www.jimmunol.org/ by guest on October 4, 2021

FIGURE 3. Plasma membrane lipid order affects speed of T cell/APC conjugate formation potency of T cell proliferation and cytokine production. FACS-sorted high, intermediate, and low order CD4+ T cells from four individuals were labeled with CellTracker Orange and cocultured with superantigen- loaded APC labeled with blue CellTracker for 5 min. A, Representative dot plots showing rate of T cell/APC conjugate formation (%) and (B) cumulative data for the three populations. *p = 0.015. C, Representative dot plot showing T cell expression of proliferation marker Ki-67 following 72 h coculture with superantigen-loaded APC. D, FACS-sorted high, intermediate, and low order CD4+ T cells from eight individuals were cultured with or without anti-CD3/ CD28 for 72 h. Cell proliferation was determined by thymidine incorporation. **p = 0.007 high versus intermediate order. E, Representative dot plot showing IL-2 production in T cells following 72 h coculture with superantigen-loaded APC. F, Cumulative results from four experiments showing IL-2 production in high, intermediate, and low order cells.*p = 0.05. G, Combined results from eight individuals comparing the percentage difference in cytokine production by intermediate order cells compared with high order cells. *p $ 0.05 IL-4 and IL-10 compared with both INF-g and IL-6 ,**p = 0.007 IL-2 compared with INF-g. H, FACS-sorted cells cultured with or without TCR stimulation for 24 h and analyzed for cell viability. I, Viability of low order cells cultured in the presence of IL-2 and PMA. *p = 0.02. The Journal of Immunology 3511

Table I. Characteristics of high, intermediate, and low order CD4+ T cells

Membrane Order High Intermediate Low Area of T cell/APC contact ++ ++ + Speed of T cell/APC conjugate formation + +++ +++ Accumulation of CD3/pY at T cell/APC IS +++/+++ ++/+++ 2/2 Proliferation +++ ++ 2 IL-2 production +++ ++ ND IL-4/IFN-g production +++/+ +/+++ ND Apoptosis 22+ Associated membrane lipids Cholesterol +++ +++ ++ GM1 + ++ +++ GM3 ++ ++ ++ +++, strong; ++, intermediate; +, weak; ND, not detected. revealed that high order cells were predominantly naive/ by the addition of protein tyrosine kinase inhibitor PP2 (Supple- nonactivated cells whereas intermediate order cells had a more mental Fig. 3C), indicating that signaling via TCR-associated activated/effector memory phenotype (Supplemental Fig. 2C–E). kinases mediates an effect on membrane order. Analysis of the low order population revealed that a high per- The results described above suggest that membrane lipid order Downloaded from centage were CD45RA+, but they also revealed a more activated could represent a phenotypic marker that reflects the functional phenotype expressing high levels of CD69 and CD25 (Supple- capability of T cells with a gradient ranging from T cells with high mental Fig. 2E), further supporting their “anergic-like” profile membrane order with a naive phenotype, which proliferate robustly (20). In support of the phenotyping profiles, in vitro TCR trig- and produce IL-2 and IL-4, to intermediate order cells that have gering using soluble anti-CD3 at a range of concentrations together a memory phenotype, proliferate weakly, but produce IFN-g, and with costimulation with anti-CD28 (1 mg/ml) reduced global finally to low membrane order T cells that have an activated http://www.jimmunol.org/ membrane order and led to an increase in the intermediate and low phenotype, are unresponsive to TCR stimulation, and are more populations (Supplemental Fig. 3A,3B). This effect was inhibited prone to apoptosis (Table I). by guest on October 4, 2021

FIGURE 4. Expansion of intermediate order T cell population in patients with autoimmune rheumatic disease. A, Membrane order was assessed in ex vivo T cells from healthy volunteers (n = 10) and patients with active SLE (n = 10), RA (n = 10) and SS (n = 5). Repre- sentative dot plots and (B) cumulative data are shown. ***p # 0.0008. CD4+ T cells isolated from healthy controls and patients with SLE, RA, and SS were assessed for strength of proliferation by thymidine incorporation (C) and IFN-g production by intracellular staining (D). 3512 MEMBRANE LIPID ORDER CORRELATES WITH T CELL FUNCTION

Differential membrane order in patients with autoimmune To examine this possibility we cultured FACS-sorted naive rheumatic disease T cells under Th1- or Th2-polarizing conditions (Fig. 5A) and then + labeled the differentiated cells with ANE to evaluate their plasma To relate CD4 T cell membrane order to in vivo function in + humans, we investigated autoimmune T cells, characterized by membrane order. The vast majority of naive T cells (CD4 CD45RA+CD252) had relatively high membrane order (Fig. 5B). chronic activation, isolated from patients with RA, SLE, and SS, + compared with healthy volunteers. The representative dot plots However, after 6-d differentiation, naive CD4 T cells cultured and cumulative data (Fig. 4A,4B) show that patients with active under Th1-polarizing conditions assume an intermediate order lupus and RA have a significant expansion of the intermediate phenotype (lower GP) (Figs. 5B,5C,6A). Successful Th1 differ- order population compared with healthy volunteers. Interestingly, entiation was confirmed by the production of high levels of IFN-g and expression of the transcription factor T-bet (Fig. 5D). In those patients with expanded intermediate order populations had + lower proliferation (Fig. 4C) and increased production of INF-g in contrast, CD4 T cells cultured under Th2-polarizing conditions response to in vitro TCR stimulation (Fig. 4D) compared with retained a relative high membrane order (higher GP) (Figs. 5B, healthy controls. These results imply that membrane order could 5C,6A), produced high levels of IL-4, and exhibited increased play an important role in determining the response of T cells to levels of phosphorylated STAT6, an established Th2 marker (23) TCR stimulation in vivo. (Fig. 5E). The link between membrane order and Th cell function was Th cell phenotypes are associated with specific plasma strengthened when IS formation and proliferation were assessed membrane lipid order as described previously. Th1 cells with intermediate lipid order Previous reports examining polarized Th cell populations revealed formed less stable IS compared with Th2 cells with higher lipid differences in IS formation between Th1 and Th2 phenotypes (21, order (Fig. 6A), thus confirming the aforementioned findings in Downloaded from 22). We reasoned that the distinct patterns of IS formation and Fig. 2A. Furthermore, Th1 cells formed T cell/APC conjugates alterations in proliferation and cytokine production seen in ex vivo more quickly (Fig. 6C,6D) but had reduced proliferation (Fig. 6E) intermediate and high order CD4+ populations could be connected compared with the Th2 cells, again corroborating the results de- with different function in human Th cell subsets. scribed previously (Fig. 3A–D). http://www.jimmunol.org/ by guest on October 4, 2021

FIGURE 5. Th1 and Th2 cells are associated with different relative levels of plasma membrane lipid order. A, Naive T cells (CD4+CD82 CD45RA+CD252) were sorted from four individuals and cultured under Th1- or Th2-polarizing conditions. B, Representative dot plot showing polarized Th1 and Th2 cells labeled with ANE. C, Cumulative results from ﬁve experiments showing percentage of cells with intermediate and high order membranes, *p # 0.05 Th1 versus Th2. To conﬁrm phenotype differentiation, cells and supernatants from 6 d polarizing cultures were analyzed for production of IFN-g and expression of T-bet (D)or production of IL-4 and expression of pSTAT6 (E). Results shown are from four individual experiments. *p # 0.05. The Journal of Immunology 3513

reduce global membrane order in primary T cells, we tested its ability to influence T cell function in FACS-sorted ex vivo CD4+ intermediate and high order populations. Cells were treated with 7KC or left untreated for 30 min and IS formation was assessed by IRM as before. 7KC treatment of intermediate order T cells did not influence their membrane order or their ability to form attachments to Ab-coated coverslips (Fig. 7A,7B). However, when high order T cells were treated with 7KC, their membrane order was reduced (shown in the GP images in Fig. 7A), accompanied by a significant reduction in the area of cell/coverslip contact (Fig. 7A,7B). Despite this, 7KC treatment did not significantly influence the number of cells within each population interacting with anti–CD3/CD28-coated coverslips (Supplemental Fig. 4D). Interestingly, by reducing membrane order in high order T cells (using 7KC) we were able to induce a phenotype similar to the intermediate order population; namely, proliferation was inhibited (Fig. 7C) and production of IFN-g was increased (Fig. 7D,7E), although no significant differences in IL-4 production were seen (data not shown). 7KC did not influence cytokine production by intermediate order T cells. Taken together, these results provide Downloaded from evidence that changes in membrane order could influence T cell plasticity in terms of cytokine production and proliferation.

Discussion One of the determining properties of lipid raft microdomains is that they form areas of lipid order in biological membranes that help to http://www.jimmunol.org/ compartmentalize signaling molecules (18, 25). Lipid order affects membrane fluidity; relative disordered membranes support free/ random movement of protein molecules within the membrane, whereas movement of molecules within more ordered membrane may be restricted (26). However, controversy has surrounded the role, and very existence, of these regions mainly because they have been difficult to visualize by microscopy (9, 10). The use of FIGURE 6. IS formation and proliferation in Th1- and Th2-differenti- membrane probes such as ANE and LAURDAN (12, 27) has gone by guest on October 4, 2021 ated cells mirror the characteristics of intermediate and high membrane some way to resolving some of these difficulties, and the impor- order populations. Naive T cells were sorted from four individuals and tance of membrane order in the formation of stable IS in both cultured under Th1- or Th2-polarizing conditions. A, Th1- and Th2-dif- T cell lines and in primary T cells has been established (13, 14, ferentiated cells were analyzed by IRM. Representative images show the 16). In this study, by labeling human CD4+ T cells with ANE, we pattern of cell/coverslip attachment with representative GP image showing go further and reveal that ex vivo T cells are heterogeneous in membrane order. Scale bars, 5 mm. B, Cell contact area (pixels) from the IRM images; a representative experiment of four is shown. ***p = 0.001. terms of global plasma membrane lipid order, and three distinct Th1- and Th2-differentiated cells and superantigen-loaded APC were la- populations can be discerned. Importantly, we found that not only beled with CellTracker and cocultured for 5 min. Representative dot plots can the overall level of T cell plasma membrane lipid order predict show (C) T cell/APC conjugates and (D) cumulative data from four the stability and pattern of IS formation, but it can also predict individuals. *p # 0.05. E, Th1- and Th2-differentiated cells were cocul- T cell function in terms of proliferative ability and profile of cy- tured with superantigen-loaded APC for 48 h before either staining for Ki- tokine production. Moreover, we were able to manipulate T cell 67 expression (left panel) or assessing for thymidine incorporation (right function by changing membrane lipid order. These results offer panel). Results are expressed as percentage change in proliferation from a possible mechanism by which T cells could control functional Th1 to Th2 populations. *p = 0.03 plasticity and raise the possibility that therapeutic targeting of membrane lipid order could direct immune cell activation, helping to correct abnormal immune responses in conditions such as au- Manipulation of membrane order modulates T cell function toimmunity. The results described so far suggest that the ability of T cells to High lipid order at the IS has been shown by several groups, respond to TCR stimulation might be associated with mechanisms mainly in Jurkat T cell lines. Combining imaging of the IS with that control membrane order. Therefore, to further understand the LAURDAN labeling has shown that, at the T cell/APC interface, relationship between lipid order and T cell function we used the lipid order is high compared with low order in the surrounding oxysterol 7KC, which has been shown to reduce lipid order at the membrane (14, 15). These results corroborate the wealth of data T cell/APC IS and inhibit T cell proliferation and IL-2 production associating lipid microdomains with T cell activation (1, 2, 4, 18, (14, 24). First, we confirmed this finding in primary T cells; CD4+ 28, 29). Additionally, the results indicate that global membrane T cells cultured with 7KC show a rapid reduction in membrane lipid order, not just order at the IS, plays a role in shaping the order as shown by reduced GP values (Supplemental Fig. 4A) and interaction between a T cell and APC. This could provide an an increase in the number of intermediate and low order cells additional mechanism influencing how T cells relate with other (Supplemental Fig. 4B). TCR-induced proliferation (Supplemental cells in their immediate environment and may ultimately de- Fig. 4C) and IL-2 production (data not shown) were also inhibited termine T cell differentiation and function. High order T cells as described previously (14). Having established that 7KC could formed IS with a large surface area that resembled the classical 3514 MEMBRANE LIPID ORDER CORRELATES WITH T CELL FUNCTION

FIGURE 7. Manipulation of membrane order modiﬁes T cell function. FACS-sorted CD4+ intermediate and high order cells from three healthy volunteers were cultured for 30 min either without sterols or with 7KC. A, Cells were labeled with ANE and analyzed by IRM as described in Fig. 2A. Repre- sentative IRM and corresponding GP images. GP panels: red, high order; green/blue, low order. Scale bars, 5 mm. B, Cumulative results from three experiments showing the contact area (pixels) of each cell from the IRM images. *p = 0.02 high order no treat- Downloaded from ment versus 7KC. C, Proliferation in untreated and 7KC-treated intermediate and high order T cells assessed by thymidine incorporation. Cytokine production assessed by intracellular staining for IFN-g in intermediate and high order T cells treated with and without 7KC. Representative dot plots (D) and cumulative data from ﬁve experiments are shown (E). **p = http://www.jimmunol.org/ 0.001 high versus high7KC, p = 0.008 high versus intermediate. by guest on October 4, 2021

bulls-eye pattern of IS formation, with peripheral supramolecular ential membrane lipid order in T cells could regulate these (and activation complex (SMAC) and distal SMAC regions of attach- other) key signaling events by controlling the clustering of ment, whereas intermediate order cells did not form distinct distal membrane proteins at the IS, thereby shaping subsequent T cell SMAC regions, had a smaller area of attachment, and interactions function. It is known, for example, that accumulation of PKC-u at were more rapid. Low order cells did not form a distinct IS pattern the IS is essential for Th2 cell development and production of of attachment visible by IRM, although they did form conjugates IL-4 (33), and that altered localization of key signaling molecules with APC (3). It is possible that differences in global membrane at the IS, including PKC-u and CTLA-4, controls regulatory order could discriminate between cells that form long-lived stable T cell function (34, 35). It remains to be determined whether IS that last several hours and the more dynamic and transient molecules such as PKC-u preferentially accumulate at the IS in contacts lasting a few minutes, also known as kinapses (30). These the high order cells, thereby contributing to more robust pro- differences could also account for the striking variation in proliferation. The unresponsive nature of low order cells, together liferative ability between T cells with low, intermediate, and high with their inability to partition CD3 and the accompanying low lipid order since stable IS formation is required for full T cell levels of protein tyrosine phosphorylation at the IS, is suggestive activation (31), although further work is needed to confirm this of a tolerized or terminally differentiated phenotype (19, 36). proposal. In vivo-tolerized cells are able to form conjugates with APC but The different patterns of IS formation in the three lipid order are unable to mobilize TCR, PKC-u, or lipid raft clustering at the populations were characterized by a distinctive arrangement of immune synapse and do not induce tyrosine phosphorylation of CD3 and accumulation of tyrosine phosphorylated proteins at the signaling proteins (37). It seems likely therefore that low plasma T cell/APC interface and subsequently by altered T cell function. membrane order could be a hallmark of tolerized T cells. These These differences most likely reflect altered strength and organi- cells were also more prone to apoptosis. It is plausible that dif- zation of TCR-associated signaling events at the IS (31). Stability ferential membrane order could influence multiple T cell intra- of IS formation in naive, effector, and regulatory T cell subsets is cellular signaling pathways, including cell survival and apoptosis partly determined by differential accumulation of signaling pro- (38), as indicated by recent work showing that T cell apoptosis is teins, including protein kinase C (PKC)-u and Wiskott–Aldrich associated with differential localization of PKC family proteins syndrome protein, that influence actin polymerization in the pe- with membrane lipid microdomains (39). It remains to be seen ripheral and distal SMAC regions of the synapse (3, 32). Differ- whether lipid order plays a role in such mechanisms, and detailed The Journal of Immunology 3515 biochemical analysis of downstream signaling pathways following membrane order is associated with higher levels of membrane immune activation will reveal important information to answer cholesterol as might be expected since cholesterol gives structure these questions. Taken together, the results presented in this study to the membrane by allowing lipids to become tightly packed (50, go some way to support the original lipid microdomain hypothesis 51). However, CTB binding to GM1 was associated with the less in that they provide a “platform” to support or influence differ- ordered populations, suggesting that membrane lipid order does ential T cell signaling (4). The concept that T cells have different not reflect lipid microdomains as they are described in the litera- levels of membrane lipid order provides an extra level of subtlety ture. More detailed analysis of membrane lipid content using more to the role that they play during T cell activation. However, note accurate methods to determine lipid species will help improve our that further work is needed to fully establish the relationship be- understanding of the role played by membrane lipids in T cell tween membrane lipid microdomains and plasma membrane or- activation and function. der. The relevance of T cell membrane order was assessed in a range Differential patterns of IS formation have been described pre- of patients with autoimmune rheumatic disease. Patients with lupus viously in helper T cell subsets (40). Specifically, Th1-polarized and RA are characterized by chronic immune cell activation. We cells preferentially cocluster TCR and IFN-g receptor with lipid reveal that these patients have increased numbers of T cells with rafts (41) and form a more compact IS compared with Th2 cells intermediate membrane order, which was associated with reduced that favor multifocal IS formation (21, 22). However, although we proliferation and increased production of proinflammatory cyto- found a correlation between Th1/Th2 IS formation and membrane kine IFN-g. Lupus T cells are characterized by altered expression order, it was difficult to relate these differences to the earlier of raft-associated lipids, defective raft-associated signaling, and reports (21, 22). We show that the human Th2-like, high order cells altered patterns of IS formation when compared with T cells from form IS that more closely resemble a classical bulls-eye pattern as healthy volunteers (52, 53). The enrichment of the intermediate Downloaded from opposed to Th2 cells generated in TCR-transgenic mice that order population may reflect the altered lipid composition of T cell formed a multifocal IS. These disparities may be explained by membranes in patients with lupus. Whether this is a result of differences in methodology used and differences in lipid content chronic activation by endogenous Ags is unclear. A more detailed between mouse and human lymphocytes (42); furthermore, the examination of lipids in T cell membranes from patients with pattern of IS formation in the transgenic Th2 model was dependent autoimmune disease is underway. on the concentration of antigenic stimulation, which was not con- To conclude, we show in this study that membrane lipid order http://www.jimmunol.org/ sidered in this study since it is difficult to assess in ex vivo hu- is heterogeneous in primary T cells and associated with distinct man T cells (22). We also report differences in proliferation be- Th1-like, Th2-like, and tolerogenic characteristics. However, how tween the Th1-polarized/intermediate order T cells that proliferate plasma membrane lipid order relates to the previously described weakly compared with Th2-polarized/high order T cells that ex- lipid raft microdomains remains unclear. Artificial manipulation of hibit strong proliferation. Interestingly, increased proliferation in lipid order influenced IS formation, which in turn may influence the Th2-polarized cells has been identified previously (43). ability of membrane proteins to cluster at the IS and shape T cell Although we could establish a clear relationship between mem- function. This raises the possibility that therapeutic targeting of brane order and T cell cytokine production in polarized T cell molecules controlling membrane lipid order could regulate im- by guest on October 4, 2021 populations, in ex vivo, nonmanipulated CD4+ cells, the pattern mune activation and help control abnormal immune responses in was similar although less clear-cut. Intermediate order cells pro- conditions such as autoimmunity (29). duced more IFN-g whereas high order cells produced more IL-4 and IL-10. One possibility is that changes in membrane lipid order Acknowledgments could represent a mechanism by which T cells fine-tune their so- We thank the Facility for Imaging by Light Microscopy at Imperial College. called “functional plasticity” (44, 45). Indeed, when we artificially We also thank Claudia Mauri, Panagiotis Kabouridis, and David Isenberg manipulated membrane order we were able to influence IS for- for discussions, input, and critical reading of the manuscript. mation and proliferation, and we skewed cytokine production to- ward IFN-g production in high order cells. Interestingly, although Disclosures we were able to influence the function of high order cells by re- The authors have no financial conflicts of interest. ducing membrane order with 7KC, intermediate order cells were relatively unaffected. We were also unable to influence the function of intermediate order cells by increasing cholesterol content References (data not shown). This could be because membrane order does not 1. Dykstra, M., A. Cherukuri, H. W. Sohn, S.-J. Tzeng, and S. K. Pierce. 2003. Location is everything: lipid rafts and immune cell signaling. Annu. Rev. change by simply adding cholesterol exogenously, or it may in- Immunol. 21: 457–481. dicate that factors other than cholesterol content are important in 2. Hancock, J. F. 2006. Lipid rafts: contentious only from simplistic standpoints. these cells. Interestingly, cholesterol and molecules that control Nat. Rev. Mol. Cell Biol. 7: 456–462. 3. Fooksman, D. R., S. Vardhana, G. Vasiliver-Shamis, J. Liese, D. A. Blair, cellular cholesterol homeostasis have been recognized recently to J. Waite, C. Sacristań, G. D. Victora, A. Zanin-Zhorov, and M. L. Dustin. 2009. affect T cell function (46); therefore, more detailed analysis ex- Functional anatomy of T cell activation and synapse formation. Annu. Rev. amining the balance between cellular cholesterol and oxidized Immunol. 28: 79–105. forms of cholesterol in membranes (including 7KC) may prove 4. Simons, K., and D. Toomre. 2000. Lipid rafts and signal transduction. Nat. Rev. Mol. Cell Biol. 1: 31–39. important in understanding T cell functional plasticity. Alterna- 5. Janes, P. W., S. C. Ley, A. I. Magee, and P. S. Kabouridis. 2000. The role of lipid tively, glycosphingolipid species have distinct roles in T cell rafts in T cell antigen receptor (TCR) signalling. Semin. Immunol. 12: 23–34. function, are integral components of lipid microdomains, and play 6. Owen, D. M., M. A. Neil, P. M. French, and A. I. Magee. 2007. Optical tech- a role in T cell signaling events (38, 47, 48). We reveal differential niques for imaging membrane lipid microdomains in living cells. Semin. Cell Dev. Biol. 18: 591–598. glycosphingolipid expression in low, intermediate, and high order 7. Zacharias, D. A., J. D. Violin, A. C. Newton, and R. Y. Tsien. 2002. Partitioning T cells that may contribute to their functional heterogeneity (49), of lipid-modified monomeric GFPs into membrane microdomains of live cells. and thus further dissection of the lipid content of these populations Science 296: 913–916. 8. Shogomori, H., and D. A. Brown. 2003. Use of detergents to study membrane could help identify mechanisms linking plasma membrane order to rafts: the good, the bad, and the ugly. Biol. Chem. 384: 1259–1263. T cell function. The results presented in this study show that higher 9. Munro, S. 2003. Lipid rafts: elusive or illusive? Cell 115: 377–388. 3516 MEMBRANE LIPID ORDER CORRELATES WITH T CELL FUNCTION

10. Shaw, A. S. 2006. Lipid rafts: now you see them, now you don’t. Nat. Immunol. 33. Marsland, B. J., T. J. Soos, G. Spa¨th, D. R. Littman, and M. Kopf. 2004. Protein 7: 1139–1142. kinase C u is critical for the development of in vivo T helper (Th)2 cell but not 11. Jin, L., A. C. Millard, J. P. Wuskell, X. Dong, D. Wu, H. A. Clark, and Th1 cell responses. J. Exp. Med. 200: 181–189. L. M. Loew. 2006. Characterization and application of a new optical probe for 34. Zanin-Zhorov, A., Y. Ding, S. Kumari, M. Attur, K. L. Hippen, M. Brown, membrane lipid domains. Biophys. J. 90: 2563–2575. B. R. Blazar, S. B. Abramson, J. J. Lafaille, and M. L. Dustin. 2010. Protein kinase 12. Parasassi, T., G. De Stasio, A. d’Ubaldo, and E. Gratton. 1990. Phase fluctuation C-u mediates negative feedback on regulatory T cell function. Science 328: 372–376. in phospholipid membranes revealed by Laurdan fluorescence. Biophys. J. 57: 35. Flores-Borja, F., E. C. Jury, C. Mauri, and M. R. Ehrenstein. 2008. Defects in 1179–1186. CTLA-4 are associated with abnormal regulatory T cell function in rheumatoid 13. Owen, D. M., S. Oddos, S. Kumar, D. M. Davis, M. A. A. Neil, P. M. W. French, arthritis. Proc. Natl. Acad. Sci. USA 105: 19396–19401. M. L. Dustin, A. I. Magee, and M. Cebecauer. 2010. High plasma membrane 36. Thomas, S., R. Kumar, A. Preda-Pais, S. Casares, and T.-D. Brumeanu. 2003. A lipid order imaged at the immunological synapse periphery in live T cells. Mol. model for antigen-specific T-cell anergy: displacement of CD4-p56lck signal- Membr. Biol. 27: 178–189. osome from the lipid rafts by a soluble, dimeric peptide-MHC class II chimera. J. 14. Rentero, C., T. Zech, C. M. Quinn, K. Engelhardt, D. Williamson, T. Grewal, Immunol. 170: 5981–5992. W. Jessup, T. Harder, and K. Gaus. 2008. Functional implications of plasma 37. Ise, W., K. Nakamura, N. Shimizu, H. Goto, K. Fujimoto, S. Kaminogawa, and membrane condensation for T cell activation. PLoS ONE 3: e2262. S. Hachimura. 2005. Orally tolerized T cells can form conjugates with APCs but 15. Gaus, K., E. Chklovskaia, B. Fazekas de St Groth, W. Jessup, and T. Harder. are defective in immunological synapse formation. J. Immunol. 175: 829–838. 2005. Condensation of the plasma membrane at the site of T lymphocyte acti- 38. Gombos, I., E. Kiss, C. Detre, G. La´szlo´, and J. Matko´. 2006. Cholesterol and vation. J. Cell Biol. 171: 121–131. sphingolipids as lipid organizers of the immune cells’ plasma membrane: their 16. Harder, T., C. Rentero, T. Zech, and K. Gaus. 2007. Plasma membrane segre- impact on the functions of MHC molecules, effector T-lymphocytes and T-cell gation during T cell activation: probing the order of domains. Curr. Opin. death. Immunol. Lett. 104: 59–69. Immunol. 19: 470–475. 39. Solstad, T., E. Bjørgo, C. J. Koehler, M. Strozynski, K. M. Torgersen, K. Tasken, 17. Owen, D. M., P. M. Lanigan, C. Dunsby, I. Munro, D. Grant, M. A. Neil, and B. Thiede. 2010. Quantitative proteome analysis of detergent-resistant P. M. French, and A. I. Magee. 2006. Fluorescence lifetime imaging provides membranes identifies the differential regulation of protein kinase C isoforms enhanced contrast when imaging the phase-sensitive dye di-4-ANEPPDHQ in in apoptotic T cells. Proteomics 10: 2758–2768. model membranes and live cells. Biophys. J. 90: L80–L82. 40. Maldonado, R. A., D. J. Irvine, R. Schreiber, and L. H. Glimcher. 2004. A role 18. Zech, T., C. S. Ejsing, K. Gaus, B. de Wet, A. Shevchenko, K. Simons, and for the immunological synapse in lineage commitment of CD4 lymphocytes. T. Harder. 2009. Accumulation of raft lipids in T-cell plasma membrane domains Nature 431: 527–532. Downloaded from engaged in TCR signalling. EMBO J. 28: 466–476. 41. Maldonado, R. A., M. A. Soriano, L. C. Perdomo, K. Sigrist, D. J. Irvine, T. Decker, 19. Carlin, L. M., K. Yanagi, A. Verhoef, E. N. Nolte-’t Hoen, J. Yates, L. Gardner, and L. H. Glimcher. 2009. Control of T helper cell differentiation through cytokine J. Lamb, G. Lombardi, M. J. Dallman, and D. M. Davis. 2005. Secretion of IFN- receptor inclusion in the immunological synapse. J. Exp. Med. 206: 877–892. g and not IL-2 by anergic human T cells correlates with assembly of an im- 42. Marwali, M. R., J. Rey-Ladino, L. Dreolini, D. Shaw, and F. Takei. 2003. mature immune synapse. Blood 106: 3874–3879. Membrane cholesterol regulates LFA-1 function and lipid raft heterogeneity. 20. Sechi, A. S., J. Buer, J. Wehland, and M. Probst-Kepper. 2002. Changes in actin Blood 102: 215–222. dynamics at the T-cell/APC interface: implications for T-cell anergy? Immunol. 43. Bird, J. J., D. R. Brown, A. C. Mullen, N. H. Moskowitz, M. A. Mahowald,

Rev. 189: 98–110. J. R. Sider, T. F. Gajewski, C. R. Wang, and S. L. Reiner. 1998. Helper T cell http://www.jimmunol.org/ 21. Balamuth, F., D. Leitenberg, J. Unternaehrer, I. Mellman, and K. Bottomly. differentiation is controlled by the cell cycle. Immunity 9: 229–237. 2001. Distinct patterns of membrane microdomain partitioning in Th1 and th2 44. O’Shea, J. J., and W. E. Paul. 2010. Mechanisms underlying lineage commitment cells. Immunity 15: 729–738. and plasticity of helper CD4+ T cells. Science 327: 1098–1102. 22. Thauland, T. J., Y. Koguchi, S. A. Wetzel, M. L. Dustin, and D. C. Parker. 2008. 45. Zhou, L., M. M. W. Chong, and D. R. Littman. 2009. Plasticity of CD4+ T cell Th1 and Th2 cells form morphologically distinct immunological synapses. J. lineage differentiation. Immunity 30: 646–655. Immunol. 181: 393–399. 46. Bensinger, S. J., M. N. Bradley, S. B. Joseph, N. Zelcer, E. M. Janssen, 23. Zhu, J., L. Guo, C. J. Watson, J. Hu-Li, and W. E. Paul. 2001. Stat6 is necessary M. A. Hausner, R. Shih, J. S. Parks, P. A. Edwards, B. D. Jamieson, and and sufﬁcient for IL-4’s role in Th2 differentiation and cell expansion. J. P. Tontonoz. 2008. LXR signaling couples sterol metabolism to proliferation in Immunol. 166: 7276–7281. the acquired immune response. Cell 134: 97–111. 24. Massey, J. B., and H. J. Pownall. 2005. The polar nature of 7-ketocholesterol 47. Gupta, G., and A. Surolia. 2010. Glycosphingolipids in microdomain formation determines its location within membrane domains and the kinetics of membrane and their spatial organization. FEBS Lett. 584: 1634–1641.

microsolubilization by apolipoprotein A-I. Biochemistry 44: 10423–10433. 48. Sorice, M., A. Longo, T. Garofalo, V. Mattei, R. Misasi, and A. Pavan. 2004. by guest on October 4, 2021 25. Zurzolo, C., G. van Meer, and S. Mayor. 2003. The order of rafts: conference on Role of GM3-enriched microdomains in signal transduction regulation in microdomains, lipid rafts and caveolae. EMBO Rep. 4: 1117–1121. T lymphocytes. Glycoconj. J. 20: 63–70. 26. Brown, D. A., and E. London. 1998. Functions of lipid rafts in biological 49. Schade, A. E., and A. D. Levine. 2002. Lipid raft heterogeneity in human pe- membranes. Annu. Rev. Cell Dev. Biol. 14: 111–136. ripheral blood T lymphoblasts: a mechanism for regulating the initiation of TCR 27. Gaus, K., E. Gratton, E. P. W. Kable, A. S. Jones, I. Gelissen, L. Kritharides, and signal transduction. J. Immunol. 168: 2233–2239. W. Jessup. 2003. Visualizing lipid structure and raft domains in living cells with 50. Tani-ichi, S., K. Maruyama, N. Kondo, M. Nagafuku, K. Kabayama two-photon microscopy. Proc. Natl. Acad. Sci. USA 100: 15554–15559. J.-i. Inokuchi, Y. Shimada, Y. Ohno-Iwashita, H. Yagita, S. Kawano, and 28. Janes, P. W., S. C. Ley, and A. I. Magee. 1999. Aggregation of lipid rafts A. Kosugi. 2005. Structure and function of lipid rafts in human activated T cells. accompanies signaling via the T cell antigen receptor. J. Cell Biol. 147: 447–461. Int. Immunol. 17: 749–758. 29. Jury, E. C., F. Flores-Borja, and P. S. Kabouridis. 2007. Lipid rafts in T cell 51. Tavano, R., G. Gri, B. Molon, B. Marinari, C. E. Rudd, L. Tuosto, and A. Viola. signalling and disease. Semin. Cell Dev. Biol. 18: 608–615. 2004. CD28 and lipid rafts coordinate recruitment of Lck to the immunological 30. Dustin, M. L. 2008. T-cell activation through immunological synapses and synapse of human T lymphocytes. J. Immunol. 173: 5392–5397. kinapses. Immunol. Rev. 221: 77–89. 52. Jury, E. C., P. S. Kabouridis, F. Flores-Borja, R. A. Mageed, and D. A. Isenberg. 31. Huppa, J. B., M. Gleimer, C. Sumen, and M. M. Davis. 2003. Continuous T cell 2004. Altered lipid raft-associated signaling and ganglioside expression in receptor signaling required for synapse maintenance and full effector potential. T lymphocytes from patients with systemic lupus erythematosus. J. Clin. Invest. Nat. Immunol. 4: 749–755. 113: 1176–1187. 32. Sims, T. N., T. J. Soos, H. S. Xenias, B. Dubin-Thaler, J. M. Hofman, J. C. Waite, 53. Krishnan, S., M. P. Nambiar, V. G. Warke, C. U. Fisher, J. Mitchell, N. Delaney, T. O. Cameron, V. K. Thomas, R. Varma, C. H. Wiggins, et al. 2007. Opposing and G. C. Tsokos. 2004. Alterations in lipid raft composition and dynamics effects of PKCu and WASp on symmetry breaking and relocation of the im- contribute to abnormal T cell responses in systemic lupus erythematosus. J. munological synapse. Cell 129: 773–785. Immunol. 172: 7821–7831.