Lipid raft-based membrane order is important for antigen-specific clonal expansion of CD4⁺ T lymphocytes

We used DO.11 TCRαβ transgenic (a generous gift from Dr. Dennis Loh) [48] and Balb/c mice. Mice were housed at Villanova University vivarium with active sentinel program in accordance with approved internal IACUC protocols and guidelines. Research involving mice reported in the manuscript was approved by institutional IACUC committee as per guidelines set up by the Office of Laboratory Animal Welfare (OLAW) at National Institutes of Health.

Lymph nodes were dissected from mice and single cell suspension was generated by gently crushing them between frosted glass slides and harvesting in wash medium made with RPMI 1640 supplemented with 5% fetal bovine and 2 mM HEPES buffer (Invitrogen-Life Technologies, Grand Island, NY). CD4⁺ T cells were negatively selected from total LN cell preparations using magnetic beads according to the manufacturer’s protocol (Invitrogen-Life Technologies, Grand Island, NY).

Autologous spleen cells from Balb/c mice were used as a source of APCs. A single cell suspension generated from spleen was treated with Tris-ammonium chloride for 7 minutes at 37°C to lyse red blood cells. Spleen cells devoid of RBC were washed with wash media (RPMI 1640 supplemented with 5% fetal bovine serum and 2 mM HEPES) and re-suspended in culture media with cOva_323–339 peptide for presentation to CD4⁺ T cells.

Lymph node cells or purified DO.11 CD4⁺ T cells were cultured in the presence or absence of cOva_323–339 peptide (Sigma Genosys, Woodlands, TX). In some co-cultures syngeneic APC’s from Balb/c mice were included. To assess the proliferation of CD4⁺ T cells, lymph node cells or purified CD4⁺ T cells were labeled with CFSE (Molecular Probes-Life Technologies, Grand Island, NY) prior to their addition to the co-cultures as per previously published protocols with some modification [49]. Briefly, CFSE staining was carried out with 5×10⁶ cells at 600 nM final concentration in 1 ml volume for 10 minutes, with gentle shaking at ambient temperature. The staining was quenched by adding two volumes of ice cold wash media followed by incubation on ice for five minutes. The cells were pelleted by centrifugation and the washing step was repeated twice to remove excess CFSE dye. Cells were then re-suspended in culture media consisted of RPMI 1640 supplemented with 10% FBS, 2 mM of non-essential amino acids, 50 I.U/ml penicillin and 50 μg/ml streptomycin and 1 μg/ml fungizone, 2 mM HEPES (Invitrogen-Life Sciences, Grand Island, NY). 5x10⁵ CFSE labeled lymph node cells or 1 × 10⁵ CFSE-labeled CD4⁺ T cells were cultured with either the control cOva_324–334 or stimulatory cOva_323–339 peptide at 1 μM final concentration of the peptide for 72 hrs at 37°C in a 5% CO₂ incubator. For cultures with purified CD4⁺ T cells, 2.5×10⁵ syngeneic APC were added to present specific peptide. Additional co-cultures were set up with CD4⁺ T cells that were not labeled with CFSE to serve as a control for analysis by flow cytometer.

7-KC and MβCD complexes were generated and incorporated into the plasma membrane by following a previously published protocol [50]. Briefly, cells were treated with a mixture of an appropriate concentration of 7-KC (Sigma-Aldrich, St-Louis, MO) ranging from 70 μM to 17.5 μM and a fixed concentration (0.3 mM) of MβCD (Sigma-Aldrich, St-Louis, MO). 7-KC-MβCD complexes were added to the co-culture either at the beginning of the culture (time = 0) or after 5 min, 120 min or 24 hrs of initiating the co-cultures. mβCD - 7-KC complexes effectively targets 7-KC to the plasma membrane, but low concentration (0.3 mM) of MβCD does not disrupt lipid rafts [50]. In some experiments 7-KC-mβCD complexes were incubated with either lymph node cells or purified CD4⁺ T cells for 15 minutes and unbound complexes were washed twice by adding excess of wash media and centrifuging the cells. This unique experimental design allowed us to specifically examine effects of 7-KC on lymph node or CD4⁺ T cell membrane at higher concentrations without effects on cellular viability.

Cultured cells were split and one half of cell cultures were labeled with anti-CD4-PE (BioLegend, San Diego, CA) to enumerate proliferating CD4⁺ T cells and the other half was incubated with propidium Iodide (PI) (BD Biosciences, East Rutherford, NJ, USA) to assess the viability of cells as per previously published procedures from our laboratory [49]. Labeled and unlabeled (CFSE/anti-CD4-PE/PI) cells were analyzed by FACS Calibur flow cytometer (BD Biosciences, East Rutherford, NJ, USA) using CellQuestPro software. Moreover, cell cultures set-up with CFSE-labeled lymph node or purified CD4⁺ T cells but left untreated (absence of control and stimulatory c-Ova peptide and 7-KC) were used for setting up the quadrant gate to enumerate non-proliferating cells (upper right quadrant). For assessing the effect of peptide and 7-KC on CFSE-labeled cells, CFSE^highest non-proliferating cells present in the upper right quadrant were enumerated as it provides estimation of fraction of cells that did not undergo proliferation as opposed to cells that had undergone symmetrical and non-symmetrical cell division. Statistical significance between the 7-KC treated and untreated groups was determined by one-way analysis of variance tests using JMP software program. The Bonferroni procedure was then completed to determine which means were significantly different from the untreated groups.

To assess proximity of two proteins present within the lipid raft on the plasma membrane we carried out FRET analyses. Cells were labeled with a combination of acceptor antibody-conjugated fluorophore, anti-Thy-1-Alexa647 (FL3) (BD Biosciences, Franklin Lakes, NJ, USA), either in combination with donor antibody-conjugated fluorophore, anti-CD3ε mAb-PE (FL2) (eBiosciences, San Diego, CA, USA) or anti CD71-PE (FL2) (BD Biosciences, Franklin Lakes, NJ, USA) for 60 minutes on ice. Cells were washed twice with 1 ml of ice cold 0.1 M PBS and re-suspended in 250 μl of ice cold PBS for FRET analysis with a FACSCalibur flow cytometer (BD Biosciences, Palo Alto, CA) as per previously published procedures [51] with modifications described here. Donor fluorophore (anti-CD3ε-PE) was excited by 488 nm argon laser. Donor emission was detected in FL2 channel (575BP). Compensations were set up with donor-only labeled cells resulting in absence of detection for the donor emission in FL3 channel (620BP). Acceptor-only labeled cells were used to set up background and compensation for FL3 channel resulting in absence of observable emission detected in this channel when excited by 488 nm laser. Emission from the donor and acceptor fluorophore double labeled cells was used to enumerate FRET- positive events, where the emission from the donor (PE) when excited by 488 nm in turn excited the acceptor (AlexaFluor 647). The emission from the acceptor fluorphore was detected in FL3 channel. Cells left unlabeled were used to set-up the negative control gates of the quadrant to assess FRET positive events.

Membranes of isolated lymph node cells were stained with di-4-ANEPPDHQ (Invitrogen-Life Technologies, Grand Island, NY) at 0.5 μM final concentration for 20 min at room temperature as suggested by the vendor and previously published protocols [52]. Cells were co-stained with anti-CD90-AlexaFluor 647 and analyzed by FACSCalibur flowcytometer (BD Biosciences, East Rutherford, NJ, USA) using 488 nm and 630 nm (Red) excitation lasers. Emission from these flourophores was measured at wavelengths 570 (FL2 channel), 630 (FL3 channel) and 670 (FL4 channel) after appropriate compensation set-up to allow the detection of specific flourophore. To specifically assess the membrane order of T cells, Anti-CD90-AlexaFluor 647⁺ cells were gated for their assessment of Di-4-ANEPPDHQ staining (FL2 Vs FL3).

To begin to investigate the role of membrane order in clonal expansion of CD4⁺ T cells we first sought to examine whether 7-KC did indeed disrupt the integrity of lipid rafts on the plasma membrane. To accomplish this, we designed an antibody FRET experiment for which we used a FRET pair of an AlexaFluor 647-conjugated antibody against the membrane raft marker Thy1 (acceptor) and a PE conjugated antibody against CD3ε, a component of TCR/CD3 signaling complex (donor). While Thy-1 is known to reside in lipid rafts, the TCRαβ/CD3 complex is also present in these nanodomains [17],[42]. In the absence of either the donor (anti-CD3ε PE – FL2 channel) or the acceptor (anti-Thy-1-AlexaFluor647 – FL3 channel) we did not observe cells with productive FRET signal (upper right quadrant) (Figure 1A, controls 1 & 2) by flow cytometer. These negative control experiments were used to set up the gates to enumerate cells with productive FRET signal (events in the upper right quadrant of the two color histogram). To examine the productive FRET signal we stained primary CD4⁺ T cells with anti-CD3ε PE and anti-Thy1-AlexaFluor-647. (Figure 1, no 7-KC group). Productive FRET signal in these cells is indicated by detection of emission from AlexaFluor647 (cells in upper right quadrant). Exposure of CD4⁺ T cells to 7-KC diminished the number of cells with productive FRET signal. Inhibition of FRET signals by incorporating 7-KC in the membrane was dependent on the dose of 7-KC, with a strong FRET quenching observed at 70 μM of 7-KC (>90% inhibition) and low (~50% quenching observed with 17.5 μM 7-KC (Figure 1A and C). To examine the specificity of inhibition by 7-KC we performed a control FRET experiments with anti-CD71 (anti-transferrin receptor), a non-raft marker. CD71 protein is homogenously distributed along the plasma membrane, but excluded from lipid rafts [53]. As expected, we did not observe significant alterations in number of cells emitting FRET signals between CD71-PE and Thy1-AlexaFluor647 in either the presence or absence of 7-KC (Figure 1B and C) at all the 7-KC concentrations tested. Quantitatively the FRET signal was similar to the background FRET with either the donor only (anti-CD3ε-PE) or acceptor (Thy-1 AlexaFluor647) only fluorophore (Figure 1, control 1 and Control 2 groups). These control experiments were also used to establish appropriate gates to set up the thresholds to quantify FRET positive cells (Figure 1A).

We next sought to examine if this disruption of lipid raft-based membrane order was reversible. To test that we carried out antibody FRET on CD4⁺ T cells previously exposed to 7-KC after their culture under cholesterol-rich culture conditions (culture media with 10% FBS) for 20 hrs. The FRET signal was restored in cells re-exposed to cholesterol during the 20 hour incubation period (Figure 2). CD4⁺ T cells that were not exposed to 7-KC and cultured under similar culture conditions served as controls and showed a strong FRET signal (Figure 2). These data indicate that the effects of 7-KC on the plasma membrane are reversible with almost complete restoration of the membrane order after 20 hours and therefore suggesting the dynamic nature of these nanodomains. Taken together, these data suggest that the addition of 7-KC to the T cell membrane specifically prevents interaction between the two raft molecules, Thy1 and CD3ε in a reversible way.

To assess alterations in membrane fluidity by 7-KC we chose di-4-ANEPPDHQ dye, this dye senses the membrane order and reports it on FL2 (570 nm) and FL3 (630 nm) channels when excited by 488 nm laser [52],[54]. Figure 2 shows that gated (Thy-1⁺) lymph node T cells (Figure 3A) showed altered membrane order after their exposure to 7-KC in a concentration dependent manner (Figure 3C-E) when compared to the untreated cell controls (Figure 3B). Alterations of membrane order by 7-KC were significant, the highest effect was observed at two higher concentrations of 7-KC (70 μM & 35 μM) (Figure 3B-E & G). The observed effect of 7-KC was specific as the cells treated with mβCD vehicle control (Figure 3F & G) did not show alterations in membrane order when compared to cells that were left untreated (Figure 3B & G). Taken together these data are consistent with the FRET data and suggest that CD4⁺ T cell membranes exposed to 7-KC alter membrane order.

To examine the role of lipid raft-based membrane order in antigen-specific clonal expansion, lymph node cells from DO11 TCR transgenic mice were first labeled with CFSE followed by incubation with different concentrations of 7-KC along with either a stimulatory peptide c-OVA_323–339, a control c-OVA _324–334 peptide or stimulatory anti-CD3ε monoclonal antibody. Proliferation of CD4⁺ T cells in the culture was examined by FACS after staining with anti-CD4 antibody to enumerate CD4⁺ T cells. Day-3 cultures also were stained with propidium iodide to assess viability of cells in these cultures. Whole range of concentrations for 7-KC was assessed for its effect on cell viability (data not shown). Based on these initial experiments 35 μM and lower concentrations of 7-KC was used for the experiments here. Figure 4A shows that 35 μM 7-KC showed a significant inhibition of CD4⁺ T cell proliferation induced by c-OVA_323–339 (Figure 4A and B) and anti-CD3ε (Figure 4C). In the presence of non-stimulatory peptide cOVA_324–334 (Figure 4A and B) or anti-CD3ε (Figure 4C) about 90% cells expressed non-dividing phenotype (CD4⁺CFSE^highest cells in upper right quadrant), as expected. In contrast, as many as 45% or 76% of cells had not undergone cell division (CD4⁺CFSE^highest) in the presence of specific stimulatory peptide or anti-CD3ε mAb respectively. Minimal inhibition was observed in the presence of 17.5 μM 7-KC as only 17.3% and 41.2% non-dividing CD4⁺CFSE^highest T cells remained in the cultures stimulated with cOva_323–339 and anti-CD3ε respectively (Figure 4A and B). In contrast, we observed clonal expansion of untreated CD4⁺ T cells and CD4⁺ T cells treated with vehicle control (mβCD) only in the presence of cOva_323–339 (Figure 4A and B) or anti-CD3ε antibody (Figure 4C), as expected. Only 21% (for cOva_323–339 stimulated cultures) and 31% (for anti-CD3ε stimulated cultures) of day-3 CD4⁺ T cells had not undergone cell division. To examine the downstream effects of altered membrane order we next examined the effects of 7-KC on cytokine production. CD4⁺ T cells exposed to 7-KC showed reduced IFN-γ production and these effects were directly proportional to the concentration of 7-KC treatment over the three day period (Additional file 1: Figure S1). Highest inhibition was observed in cultures with 70 μM and 35 μM 7-KC. These inhibitory effects were specific as the cytokines produce in cultures with vehicle control, mβCD, generated similar levels of cytokines as cell that were not treated with 7-KC.

To examine the specificity of 7-KC treatment on clonal expansion of CD4⁺ T cells we assessed the viability of day-3 cell cultures. Figure 5 shows that cells treated with lower 7-KC concentrations (35 & 17.5 μM) did not affect cell viability (Figure 5A-B) compared to cells treated with cOVA_324–334 control peptide alone or treated with vehicle control (mβCD), but the clonal expansion of CD4⁺ T cells in response to c-OVA_323–339 peptide and anti-CD3ε monoclonal antibody was significantly reduced in the presence of 35 μM 7-KC (Figure 4). Reduced viability of normal CD4⁺ T cells in controls on prolonged cell culture is well known and previously documented by us as well [49]. Taken together these data suggest that disruption of lipid raft-based membrane order with insertion of 7-KC in the membrane inhibits antigen receptor-driven clonal expansion of CD4⁺ T cells and cytokine production without compromising the viability of the cells.

We next sought to test the time at which 7-KC exerted its effect on CD4⁺ T cell clonal expansion in response to a specific antigen. The early events within seconds and minutes of engagement of the antigen receptor are associated with activation of membrane kinases and phosphorylation events as well as generation of secondary messengers. The later events (after gene transcription) are associated with mitosis primarily driven by the growth factor IL-2. To examine whether the presence of 7-KC affected clonal expansion by altering lipid raft-based membrane order at an early or later phase of clonal expansion, we added this compound at different time points after engaging the antigen receptor. Figure 6A and C & Additional file 2: Figure S2 show that addition of 7-KC at 5 minute time point after the cultures were set up. Addition of 7-KC at these time points dramatically inhibited the clonal expansion of CD4⁺ T cells. In the presence of 35 μM 7-KC (a concentration at which the viability of CD4⁺ T cells is not affected) we observed significant reduction in clonal expansion of CD4⁺ T cells in response to stimulation through the antigen receptor. Quantifying day-3 cell cultures that were set-up in the presence of 35 μM 7-KC indicated that about 80% and 70% of CD4⁺ cells remained unresponsive (not undergone cell division) to c-OVA_323–339 peptide and anti-CD3ε, respectively (Figure 6A and C & Additional file 2: Figure S2A & B) which was similar to the control cultures set up with control cOVA_324–334 peptide (Figure 6A & Additional file 2: Figure S2A) or left untreated (Figure 6C & Additional file 2: Figure S2B) showing approximately 80% CD4⁺CFSE^highest non-proliferating cells. Proportionally lower inhibition was observed with 17.5 μM 7-KC with about 15% and 20.5% of non-divided cells remaining in the cultures with cOva_323–339 peptide and anti-CD3ε antibody stimulated cells respectively. Fraction of non-proliferating cells in the cultures was not significantly different from the cultures stimulated with cOVA_323–339 in the presence (Figure 6A and C & Additional file 2: Figure S2) or absence (data not shown) of vehicle mβCD. Similar results were obtained when 7-KC was added at 2 hours after initiating the co-cultures (Additional file 2: Figure S2). In contrast, addition of 7-KC at 24 hrs time point after the engagement of antigen receptor did not affect clonal expansion of CD4⁺ T cells (Figure 6B and D & Additional file 2: Figure S2). These results suggest that intactness of membrane order is critical during early signaling events. Taken together, these data suggest that lipid raft-based membrane order contributes to clonal expansion during the early phase (occurring within 2 hrs.) of the response triggered through engagement of the antigen receptor.

Next we sought to examine if lipid raft-based order specifically on the plasma membrane CD4⁺ T cells contributed to diminished clonal expansion of these cells in response to antigen. To test this we incorporated 7-KC in the membrane of purified CD4⁺ T cells for brief period (excess of 7-KC was removed) and tested their clonal expansion in response to c-OVA_323–339 presented by APCs that were not exposed to 7-KC. Brief treatment (15 min) with 7-KC allowed us to examine the effects of 7-KC at higher concentration (70 μM). Figure 7 shows that CD4⁺ T cells with 7-KC incorporated in the plasma membrane showed reduced clonal expansion in a dose-dependent fashion. A maximum effect was observed with 70 μM of 7-KC and minimal inhibition. As much as, 44.2% of CD4⁺ T cells incubated with 70 μM 7-KC and 18% of CD4⁺ T cells incubated with 35 μM 7-KC did not proliferate in response to c-OVA_323–339 peptide. In the absence of 7-KC and in the presence of vehicle control mβCD, only 10.6% and 15% of CD4⁺ cells showed lack of proliferation in the OVA_323–339 stimulated cultures (Figure 7). These data indicate that disturbance of membrane order on CD4⁺ T cell membrane contributes to their diminished clonal expansion in response to the antigen receptor.