AKAPs in lipid rafts are required for optimal antigen presentation by ...

Immunology and Cell Biology (2011) 89, 650–658 & 2011 Australasian Society for Immunology Inc. All rights reserved 0818-9641/11 www.nature.com/icb

ORIGINAL ARTICLE

AKAPs in lipid rafts are required for optimal antigen presentation by dendritic cells Robynn V Schillace1,2, Casey L Miller1 and Daniel W Carr1,3 Dendritic cell (DC) maturation and antigen presentation are regulated by activation of protein kinase A (PKA) signaling pathways, through unknown mechanisms. We have recently shown that interfering with PKA signaling through the use of anchoring inhibitor peptides hinders antigen presentation and DC maturation. These experiments provide evidence that DC maturation and antigen presentation are regulated by A-kinase anchoring proteins (AKAPs). Herein, we determine that the presence of AKAPs and PKA in lipid rafts regulates antigen presentation. Using a combination of western blotting and immuno-cytochemistry, we illustrate the presence of AKAP149, AKAP79, Ezrin and the regulatory subunits of PKA in DC lipid rafts. Incubation of DCs with the type II anchoring inhibitor, AKAP-in silico (AKAP-IS), removes Ezrin and RII from the lipid raft without disrupting raft formation. Addition of a lipid raft disruptor, methyl-b-cyclodextrin, blocks the efficacy of AKAP-IS, suggesting that the lipid raft must be intact for AKAP-IS to inhibit antigen presentation. Ezrin and AKAP79 are present in the lipid raft of stimulated KG1 cells, but Ezrin is not present in the lipid raft of unstimulated KG1 cells and AKAP79 levels are greatly diminished, suggesting that Ezrin and AKAP79 may be the key AKAPs responsible for regulating antigen presentation. Immunology and Cell Biology (2011) 89, 650–658; doi:10.1038/icb.2010.148; published online 11 January 2011 Keywords: AKAP; antigen presentation; human dendritic cell; lipid raft; signal transduction

Incubation of human dendritic cells (DCs) with prostaglandin E2 induces DC maturation and antigen presentation.1,2 DC maturation is accompanied by increased expression of costimulatory molecules, secretion of pro-inflammatory cytokines and DC migration.3,4 Increased antigen presentation is associated with increased cyclic AMP (cAMP) levels and expression of major histocompatibility complex (MHC) class II and decreased antigen uptake.5 Increases in cAMP levels activate signaling via the protein kinase A (PKA) pathway.6 The serine threonine protein kinase PKA is a dimer of regulatory subunits each bound to a catalytic subunit. The regulatory subunits have four isoforms, RIa, RIb, RIIa and RIIb, while the catalytic subunits have three isoforms, Ca, b and g. PKA is inactive as a holoenzyme and is activated when an elevation in cAMP levels promotes cAMP binding to the regulatory subunits initiating release and activation of the catalytic subunits. Signaling via PKA is regulated and targeted through interactions with A-kinase anchoring proteins (AKAPs).7 The interaction between AKAPs and the regulatory subunits of PKA occurs via an amphipathic helix in the AKAPs, which positions PKA to respond to increases in cAMP levels in a spatio-temporally appropriate manner. Regulation of immune cell signaling by AKAPs is a developing field.8 AKAPs in general and the AKAP Ezrin in particular have been shown to participate in regulation of B-cell and T-cell signaling.9,10 In addition, we have generated

evidence for the importance of AKAPs in regulation of DC signaling and antigen presentation.11 Disruption of PKA/AKAP interaction using the PKA anchoring inhibitor peptide Ht31 suggests that AKAPs are important regulators of DC maturation and antigen presentation.11 Ht31 is a peptide, which forms an amphipathic helix mimicking that found in AKAPs.12 The helix binds to the regulatory subunits of PKA disrupting localization with all AKAPs. Anchoring inhibitor peptides that are selective for disruption of RI/AKAP interactions, the RI anchoring disruptor (RIAD) or RII/AKAP interactions, AKAP-in silico (AKAP-IS) have been developed using bioinformatics and spot peptide arrays.13,14 These peptides and their controls have now been used to demonstrate regulation of a variety of cellular events via RI/AKAP interactions or RII/AKAP interactions.11,15–17 Lipid rafts are membrane microdomains that have been shown to be essential for regulating immune synapse formation and antigen presentation.18,19 These microdomains are enriched in cholesterol and other lipid modified proteins leading to their insolubility in nonionic detergents. A significant proportion of MHC class II resides in lipid rafts as demonstrated by immunofluorescence and detergent insolubility.20 Disruption of lipid rafts and perturbation of MHC class II association with the rafts inhibits antigen presentation to T cells.21 In this study, we sought to determine if AKAPs and PKA are present in

1Portland Veterans Affairs Medical Center, Portland, OR, USA; 2Department of Neurology, Oregon Health and Science University, Portland, OR, USA and 3Department of Endocrinology, Oregon Health and Science University, Portland, OR, USA Correspondence: Dr DW Carr, Portland Veterans Affairs Medical Center, VAMC RD-8, 3710 US Veterans Hospital Road, Portland, OR 97239, USA. Email: [email protected] Received 11 August 2010; revised 7 October 2010; accepted 8 November 2010; published online 11 January 2011

AKAPs in DC lipid rafts RV Schillace et al 651

DC lipid rafts and if so, which AKAP(s) may be responsible for regulating antigen presentation. RESULTS Inhibition of antigen presentation by PKA anchoring inhibitor peptides In previous studies, we illustrate that anchoring inhibitor peptides disrupt antigen presentation.11 Incubation of DCs with anchoring inhibitor peptides followed by a wash step before incubation with T cells confirmed that the effect of the anchoring inhibitor peptides is on the DCs and not on the T cells. In addition, anchoring inhibitor peptides inhibit antigen presentation using protein or peptide antigen suggesting that the effects are on antigen presentation not processing, and anchoring inhibitor peptides disrupt antigen presentation in an MLR, demonstrating that the effect is not antigen dependent.11 To facilitate identification of the AKAP(s) potentially involved in regulating antigen presentation, we incubated day 5 DC with the indicated concentrations of AKAP-IS or RIAD anchoring inhibitor peptides and the scrambled peptide controls (SCR and RSCR) and analyzed the of the DCs ability to stimulate T cells. As in our previous studies, we used clonal T cells isolated against Mycobacterium tuberculosis antigens as tools to study antigen presentation and T-cell activation.11 Thus, we measured antigen presentation by activation of CD4+ clonal T cells in response to protein antigen in an interferon-g (IFN-g) ELISPOT assay, and to reduce the possibility that the anchoring inhibitor peptides are acting on the T cells, the DCs were rinsed after incubation with anchoring inhibitor peptides and before addition of antigen. As illustrated in Figure 1, the RII selective anchoring inhibitor peptide (AKAP-IS) appears to be more potent at inhibiting antigen presentation than the RIAD (panel A hatched bars compared with panel B hatched bars). AKAP-IS inhibits antigen presentation B55% at 10 mM compared with untreated and SCR control (Unt, black bars, SCR control, white bars), while RIAD needed to be used at 100 mM to elicit 30% inhibition compared with untreated and RSCR control (Unt, black bars, RSCR control, white bars). These results in DC are in stark contrast to those observed in lymphocytes in which RIAD blocks Lck phosphorylation at concentrations as low as 15 mM.13 In addition, we conducted experiments to determine if the effects of the type I and type II anchoring inhibitor peptides on antigen presentation are additive or exclusive. The data indicate that RI and RII have independent roles in regulating antigen presentation (Supplementary data, Figure 1). Taken together, these results suggest that while both RI and RII regulate antigen presentation, AKAPs that anchor RII PKA may be more important in regulating antigen presentation in human DCs than AKAPs that anchor RI PKA. Presence of AKAPs and PKA in DC lipid rafts In light of the critical role that lipid rafts have in regulating antigen presentation, we thought it prudent to determine if AKAPs and PKA subunits are present in lipid rafts. Day 5 DCs were solubilized in MES, NaCl buffer containing 1% Brij-98 (see Methods section) and lysates were subjected to discontinuous sucrose gradient fractionation. Fractions were collected from the top and analyzed by SDS-polyacrylamide gel electrophoresis, RII overlay and western analyses. As confirmed by the presence of human leukocyte antigen (HLA)-DR and Lyn, detergent-resistant membranes were found in fractions 1–4 (Figure 2). RII overlay analyses revealed bands at 50 and 80 kD (data not shown), leading to western analysis with Ezrin, AKAP79 and regulatory subunit antibodies. AKAP 79, Ezrin, PKA RI and PKA RII were all found in fractions corresponding with HLA-DR and Lyn, demonstrating the presence of AKAPs and PKA subunits in DC lipid rafts

Figure 1 Regulation of antigen presentation by type II PKA. Allogeneic day 5 DCs were incubated in the absence (Unt, black bars) or presence of (a) the indicated concentrations of type II anchoring inhibitor (IS, hatched bars) or the control (SCR, open bars), or (b) the indicated concentrations of type I anchoring inhibitor (RIAD, dotted bars) or the control (RSCR, open bars) for 1 h at 37 1C, 5% CO2. The cells were collected, rinsed and replated. Then 0.03 mg ml–1 of protein antigen (CFP10) was added for 1 h at 37 1C, 5% CO2, and antigen-specific CD4+ clonal T cells were added. IFN-g production was assayed after 16 h by ELISPOT. Data are an average of three donors. Error bars are s.e.m. Student’s t-test reveal significance, *Po0.05.

Figure 2 AKAPs and PKA subunits are found in DC lipid rafts. Day 5 DCs solubilized in MES, NaCl buffer containing 1% Brij-98 (see Methods section) were subjected to discontinuous sucrose gradient fractionation. Fractions were collected from the top and analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and western analyses. Immunodetection of HLA-DR and Lyn were used as markers for lipid rafts (indicated by bars). One representative set of westerns from three experiments is shown.

(Figure 2). However, not all proteins are found in lipid rafts as illustrated by the absence of glyceraldehyde 3-phosphate dehydrogenase in fractions 1–4 (Figure 2, bottom panel). Immunology and Cell Biology


These data are supported by immunofluorescence confocal microscopy of day 5 DC using fluorescein isothiocyanate-conjugated cholera toxin (ChTx) and fluorescein isothiocyanate-conjugated HLA-DR to stain lipid rafts. Figure 3 illustrates overlapping staining patterns for AKAP79 (red) and ChTx (green) (Figure 3a, top row, merge), AKAP79 (red) and HLA-DR (green) (Figure 3a, second row, merge), Ezrin (red) and ChTx (green) (Figure 3b, top row, merge) and Ezrin (red) with HLA-DR (green) (Figure 3b, second and third

rows, merge). Staining patterns in the membrane (Figure 3, all rows, arrows) and in zymosan-induced phagosomes (Figure 3a, second row, arrow heads) are in agreement with lipid raft staining seen in DCs in other studies.25 The images indicate distinct and overlapping (yellow) staining patterns, consistent with the idea that sub-populations of the proteins are in lipid rafts. Quantitation of colocalization within lipid rafts was performed on 10–40 cells from separate experiments using different donor DCs. Figure 3c illustrates the colocalization coefficients for overlap of AKAP79 and Ezrin with HLA-DR and ChTx. The Manders overlap coefficient for each region quantitated was 40.65. These values provide additional evidence that AKAP79 and Ezrin are found in the lipid rafts of human DCs.24 Regulation of antigen presentation in KG-1 cells Lipid raft isolations require large numbers of cells. To conserve DCs we explored the use of the DC-like cell line KG-1 for additional lipid raft experiments. KG-1 cells are a highly endocytic, human erythromylogenous leukemia line with the capacity to mature into macrophages and dendritic-like cells.26 KG-1 cells can be produced in high numbers with consistent purity providing an in vitro system in which to examine DC-specific processes. Stimulation with phorbol myristate acetate and ionomycin or phorbol myristate acetate and tumor necrosis factor-a differentiates KG-1 cells into dendritic-like cells characterized by the extension of cellular processes and the upregulation of MHC class I and II, co-stimulatory molecules, and DC-specific markers27 although with a lower efficiency than in vitro matured DCs.28 To confirm that KG-1 cells are an acceptable model cell line for use in these experiments, we assayed antigen presentation by phorbol myristate acetate/tumor necrosis factor-a stimulated KG-1 cells in the presence and absence of Ht31 and Ht31p (Figure 4a), AKAP-IS and SCR (Figure 4b). Stimulated KG-1 cells were treated or not with Ht31, AKAP-IS or the control peptides then loaded with protein (Figure 4a) or peptide (Figure 4b) antigen and incubated with antigen-specific clonal T cells. ELISPOT analyses of IFN-g production from the clonal T cells indicate that, as seen in human DCs, Ht31 (Figure 4a, hatched bars) and AKAP-IS (Figure 4b, hatched bars) inhibit antigen presentation by stimulated KG-1 cells. Although the 20–30% inhibition illustrated here is modest, it is statistically significant compared with untreated (Figure 4, black bars) and Ht31p or SCR control peptides (Figure 4, white bars), thus was sufficient to establish KG-1 cells as a model system for lipid raft experiments. Presence of PKA/AKAPs in KG1 cell lipid rafts Studies on KG-1 cells illustrate that unstimulated KG-1 cells are less able to activate T cells than stimulated KG-1 cells.27,28 To determine if this phenomenon is associated with changes in AKAP fractionation to lipid rafts, stimulated and unstimulated KG-1 cells were solubilized in

Figure 3 Confocal analysis of AKAPs in DC lipid rafts. Day 5 DC were harvested and plated on poly-L-lysine coated coverslips overnight. Cells were kept on ice and fixed, permeabilized with saponin and stained for (a) AKAP79 or (b) Ezrin (Texas red) and fluorescein isothiocyanate (FITC)conjugated HLA-DR or FITC-conjugated cholera toxin (ChTx). In row two cells were incubated for 30 min with zymosan before fixation. Arrowheads indicate zymosan phagosomes and arrows indicate membrane lipid raft staining. Images are representative of three different experiments using DCs from different donors. Scale bar is 25 mm. (c) Quantitation of confocal images using Volocity software (Perkin Elmer, Waltham, MA, USA). The colocalization coefficient My represents the overlap of AKAP79 or Ezrin with HLA-DR or ChTx. Each dot represents a different region of lipid raft quantitation of 10–40 regions from 4 to 40 cells was conducted. Immunology and Cell Biology


MES NaCl buffer containing 1% Brij-98 and subjected to discontinuous sucrose gradient fractionation as above for the DCs. Fractions were taken from the top of the gradient and analyzed by SDSpolyacrylamide gel electrophoresis and western techniques. Similar to DCs, the data illustrate the presence of AKAP79, Ezrin, RI and RII with HLA-DR in the raft fraction of stimulated KG-1 cells (Figure 5a).

Figure 4 Anchored PKA is required for optimal antigen presentation by KG1 cells. (a) Phorbol myristate acetate (PMA) (10 ng ml–1)-ionomycin (100 ng ml–1) stimulated KG-1 cells were pretreated without or with 100 mM Ht31, Ht31p, then CFP10 peptide antigen (0.1 mg ml–1) and D454 E12 CD4+ clonal T cells were added. (b) PMA-ionomycin stimulated KG-1 cells were pretreated 25 mM IS or SCR, then MSL peptide antigen (0.03 mg ml–1) and D454 MEG CD8+ T cells were added. IFN-g production was measured by ELISPOT. Data presented are an average of four experiments with values normalized to untreated. Student’s t-test reveal significance, *Po0.05.

We also detect AKAP149 but not AKAP95 in the lipid rafts of stimulated KG-1 cells. The rafts in KG-1 cells appear to be more diffuse spanning seven fractions (3–9) compared with five fractions (2–6) as seen in DCs, and the AKAPs and PKA subunits in KG-1 cells appear to primarily localize to fractions 4–9 with the exception of small amounts of AKAP79 and RI found in fraction 3. Strikingly, in lipid rafts isolated from unstimulated KG-1 cells, Ezrin is absent from fractions 3–8 and AKAP79 levels are dramatically reduced (Figure 5b). AKAP149, RI and RII are still found in the raft with HLA-DR, and AKAP95 is not in the rafts. Quantitation of three independent fractionations is presented in Figure 5c and illustrates statistical significance between the amount of AKAP79 and Ezrin in the lipid rafts of stimulated versus stimulated cells. It is interesting to speculate that the presence of Ezrin and AKAP79 in the lipid rafts from stimulated KG-1 cells _ENREF_18may be necessary for optimal antigen presentation. Anchoring inhibitors disrupt localization of AKAP/PKA to lipid raft To determine whether disrupting PKA anchoring would alter the presence of PKA subunits in lipid rafts we treated stimulated KG-1 cells with 100 mM AKAP-IS or the SCR control peptide for 15 min, then lysed the cells in detergent-resistant membrane buffer and isolated lipid rafts fractions as above. Figure 6 illustrates that treatment of stimulated KG-1 cells with AKAP-IS does not globally disrupt the lipid raft as HLA-DR is localized to the raft in untreated (Unt), AKAPIS (IS) treated and SCR-treated cells, and Lyn is also localized to the raft in IS-treated cells (Figure 6, bottom panels). As anticipated, AKAP-IS (IS) disrupted RII from lipid raft localization, while the SCR control did not. Interestingly, AKAP-IS (IS) also disrupted Ezrin from the lipid raft fractions. These data suggest that PKA signaling is responsible for localizing Ezrin to the lipid rafts of stimulated KG-1 cells. Disruption of RII and Ezrin targeting to lipid rafts by AKAP-IS also implicates Ezrin as having a role in regulating antigen presentation. Small interfering RNA experiments to address the involvement of Ezrin and AKAP79 in regulating antigen presentation are in progress. Also of interest is the slightly shifted localization of Lyn protein compared with HLA-DR, AKAPs and PKA. Further analysis into light and dense fractionation of KG-1 lipid rafts and the associated proteins in these fractions may yield insight into the biological relevance of this shift.

Figure 5 Presence of PKA and AKAPs in KG-1 cells differs in stimulated versus unstimulated cells. (a) Phorbol myristate acetate (PMA)/tumor necrosis factor (TNF) stimulated KG-1 cells (b) unstimulated KG-1 cells were collected, solubilized in MES, NaCl buffer containing 1% Brij-98 (see Methods section) and subjected to discontinuous sucrose gradient fractionation. Fractions were collected from the top and analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and western analyses. Immunodetection of HLA-DR was used as a marker for lipid rafts (indicated by bars). One representative set of westerns from three experiments is shown. (c) Quantitation of three sets of westerns from independent fractionations. Errors bars are s.e.m. Student’s t-test indicate significance when *Po0.05. Immunology and Cell Biology


lipid raft formation and then treated the DC with AKAP-IS to disrupt PKA signaling or the SCR control. Antigen and T cells were added and IFN-g production was assayed by ELISPOT. Figure 7 illustrates that when lipid rafts are intact (first three bars), treatment of the DCs with AKAP-IS inhibits antigen presentation (hatched bar), but when the raft has been disrupted (second set of bars), AKAP-IS is no longer able to inhibit antigen presentation (stippled bar). These data support the hypothesis that the presence of AKAPs in lipid rafts is necessary for regulating antigen presentation.

Figure 6 Anchoring inhibitors disrupt PKA and AKAP localization to lipid rafts. Phorbol myristate acetate (PMA)/tumor necrosis factor (TNF) stimulated KG-1 cells were incubated with or without 100 mM AKAP-IS or the SCR control for 15 min at 37 1C, 5% CO2. Cells were collected and fractionated as in Figure 5.

Figure 7 Lipid rafts must be intact for regulation by anchoring inhibitor. Day 5 DCs were incubated in the absence or presence of 5 mM methyl-bcyclodextrin (MCD) for 15 min at 37 1C, 5% CO2, then incubated without or with 25 mM AKAP-IS or the SCR control. Antigen was loaded for 1 h at 37 1C, 5% CO2 and T cells were added for 18 h. IFN-g production was assayed by ELISPOT. Data presented are an average of four experiments normalized to untreated.27 Student’s t-test reveal significance, *Po0.05.

Raft chelation disrupts regulation of antigen presentation by AKAP/PKA Finally, we tested the hypothesis that the presence of AKAPs in lipid rafts is essential for regulating antigen presentation. As we wanted to examine the effect of lipid raft disruption on AKAP-mediated regulation of antigen presentation, we desired conditions under which rafts are disrupted, but antigen presentation still occurs. Methyl-bcyclodextrin is a lipid raft disruptor that inhibits antigen presentation at low antigen concentrations but not at high antigen concentrations.21 We therefore assayed antigen presentation to antigen-specific clonal T cells at high antigen concentrations (0.3 mg ml–1 peptide antigen). We treated day 5 DC with methyl-b-cyclodextrin to disrupt Immunology and Cell Biology

DISCUSSION Engagement of MHC class II molecules during antigen presentation results in the elevation of cAMP levels.5 Elevated cAMP activates PKA leading to increased antigen presentation and DC maturation.1,2 As PKA activity in other cell types is targeted and regulated via interactions with AKAPs,7 we postulated that AKAPs regulate antigen presentation in DCs. We have recently illustrated the presence of AKAPs in human DCs and demonstrated that disrupting PKA anchoring to AKAPs inhibited antigen presentation and DC maturation.11 In this study, we test the hypothesis that PKA anchoring to DC lipid rafts is essential for regulation of antigen presentation. Lipid rafts are small, heterogeneous, dynamic membrane microdomains that have been shown to be essential for regulating immune synapse formation and antigen presentation.18,19 These microdomains are enriched in cholesterol and other lipid modified proteins leading to their insolubility in nonionic detergents. Disruption of lipid rafts and perturbation of MHC class II association with the rafts inhibits antigen presentation to T cells.21 Although resistance to solubilization by Triton X-100 has been widely utilized as a basis for isolating raft membranes,29,30 Triton X-100-resistant membrane complexes could only be found at low temperatures, only small amounts of MHC I and MHC II were associated with rafts,31 and endoplasmic reticulum-associated proteins were found in the rafts32 suggesting that lipid rafts isolated by Triton X-100 extraction may not represent raft domains in cell membranes under physiological conditions. Drevot et al.33 searched to find a detergent that could discriminate more ordered domains from their phospholipid environments at 37 1C and selected Brij-98. Brij detergents are now routinely being used to isolate lipid rafts.31,34–36 Brij extraction of lipid rafts results in significantly more MHC I and MHC II in the lipid raft fractions, the rafts are stable and have more resolution from soluble molecules.31,37 The amount of MHC II we find in lipid rafts using Brij-98 solubilization closely resembles that of other published Brij-98 extractions and coincides with fractions containing an additional raft marker Lyn (Figure 2). RII overlay analysis of DC lipid raft fractions revealed bands at 50 and 80 kD (data not shown). We confirmed the likely identity of these proteins by western analysis to be the regulatory subunits of PKA (RI and RII) (50 kD) and the AKAPs, Ezrin and AKAP79 (80 kD) (Figure 2). Additional western analyses demonstrated the presence of AKAP149, but not AKAP95 in lipid rafts (Figures 2 and 5). We find western analysis to be more sensitive than RII overlays and as such imagine additional AKAPs are present in DC lipid rafts. Furthermore, we feel that a longer extraction time might result in a greater separation of soluble and raft fractions, and may be useful to determine if AKAPs and the regulatory subunits shift from high-density to low-density lipid raft fraction as seen for HLA-DR.31 Using confocal immunofluorescence, we confirmed the codistribution of Ezrin and AKAP79 with HLA-DR and GM1 (through ChTx binding) in lipid rafts (Figure 3). Quantitation of these images


reveals Manders overlap coefficients 40.6 and colocalization coefficients (My) 40.5 indicating colocalization of AKAPs with HLA-DR and GM1 in lipid rafts of human DCs. However, consistent with western analysis not all of AKAP79 or Ezrin are found in the lipid rafts. As lipid raft experiments require so many cells, we used KG-1 cells for analysis of the effects of anchoring inhibitors on AKAP and PKA localization to lipid rafts (Figures 4 and 6). KG-1 cells are a highly endocytic, human erythromylogenous leukemia line with the capacity to mature into macrophages and dendritic-like cells.26 We demonstrate that anchoring inhibitor peptides modestly inhibit the ability of stimulated KG-1 cells to stimulate antigen presentation. Furthermore, because it has been shown that unstimulated KG-1 cells are less able to activate T cells than stimulated KG-1 cells,27,28 we examined the presence or absence of AKAPs and PKA subunits in unstimulated versus stimulated KG-1 cells (Figure 6). Strikingly, Ezrin was not detected in lipid raft fractions 2–8 and detection of AKAP79 was dramatically reduced. Interestingly, the R subunits of PKA were equally detectable in the lipid rafts of unstimulated and stimulated KG-1 cells. These results suggest that some of the R subunits are maintained in lipid rafts via interactions with other AKAPs, such as AKAP149, which is detected in the lipid rafts of stimulated and unstimulated KG-1 cells. Whether PKA moves from one AKAP to another within the lipid rafts, or additional PKA molecules are brought to the lipid rafts bound to Ezrin or AKAP79 on stimulation remains to be determined. Although several reports have now been published on the role of Ezrin in lipid rafts, Ezrin was initially studied as a regulator of cytoskeletal movement and actin polymerization. In fact it is now believed that interactions between Ezrin and actin regulate the association of Ezrin with the lipid raft.10,38 Rho GTPases also regulate actin and are regulated by PKA.39 Rho GTPase activities are regulated by interactions between Ezrin and Rho GDP dissociation inhibitors and Ezrin and Rho guanine nucleotide exchange factors.40 Thus, it is likely that the localization of Ezrin within the lipid raft of stimulated KG-1 cells is accompanied by localization of Rho GTPase regulators and changes actin dynamics. The role of PKA anchoring in regulating Ezrin interactions with actin and Rho signaling in human DCs is currently under investigation. To our knowledge, this is the first study to be published on the presence of AKAP79 in lipid rafts. AKAP79 associates with the b-adrenergic G-protein-coupled receptor, facilitating phosphorylation of the receptor by PKA and a switch from Gs-coupled activation of adenylyl cyclase to Gi-coupled activation of the mitogen-activated protein kinase pathway.41 Phosphorylation of the b-adrenergic receptor by PKA also promotes recycling and resensitization of the receptor.42 b2-agonists inhibit interleukin-12 production from DCs stimulated with CD40–CD40L interaction,43 suggesting that b-adrenergic receptors are present in DCs and may be regulated by AKAP79. It is exciting to consider the prospect that AKAP79 localization to DC lipid rafts with MHC II may regulate antigen presentation by controlling cAMP production via G-protein-coupled receptors localized in DC lipid rafts. It is a common finding that membrane initiated signaling events require lipid raft formation to proceed.44 Prostaglandin E2 activation of the EP3 receptor is lipid raft dependent.45 MHC class II signaling to Src kinases is disrupted in the absence of lipid rafts.20 T-cell activation at low concentrations of antigen is lipid raft dependent.21 However, not all membrane signaling events are raft dependent; for example, in B cells MHC class II signaling to ERK1/2 and PKCb are not raft dependent.46,47 To determine whether regulation of antigen presenta-

tion by PKA type II anchoring is dependent on the presence of lipid rafts, we disrupted the rafts with the lipid raft chelator methyl-bcyclodextrin and assayed antigen presentation in the presence and absence of AKAP-IS inhibitor peptide. Our results indicate that the ability of the DCs to form lipid rafts is critical to regulation of antigen presentation by PKA/AKAP (Figure 7). It will be interesting to determine whether the presence of AKAPs/PKA in the rafts promotes a direct interaction with MHC class II or regulates signaling from MHC through indirect mechanisms. As DCs and T cells are both involved in our antigen presentation assay, it is important to emphasize that in these and previous studies we confirmed that the effects of anchoring inhibitors on antigen presentation are mediated via the DCs and not the T cells.11 In addition, it is useful to compare our results in DCs with those published in T lymphocytes. Experiments in T lymphocytes illustrate the presence of Ezrin, AKAP149, and the RI, RII and catalytic subunits of PKA in lipid rafts, but not AKAP79.9 Although no functional role was identified for AKAP149 in regulating T-cell activation, Ezrin has been implicated as a key player in T cells. Ezrin colocalizes PKA and Csk by forming a supramolecular signaling complex consisting of PKA, Ezrin, Ezrin/radixin/moesin-binding protein of 50 kD (EBP50), phosphoprotein associated with glycosphingolipid-enriched membrane microdomains and Csk.48 This complex regulates phosphorylation and activation of Csk and subsequent T-cell activation. As seen with other lipid raft proteins, disruption of the actin cytoskeleton increased the amount of Ezrin in the lipid raft, consistent with the idea that localization with the actin cytoskeleton prevents localization with lipid rafts thus controlling signaling events.9 These experiments in T cells also support the hypothesis that the RI subunit of PKA is primarily responsible for regulating T-cell signaling proximal to TCR activation. The data presented here in DCs and KG1 cells indicate an important role for Ezrin in regulating antigen presentation, but in DCs, the type II specific anchoring inhibitor (AKAP-IS) is more potent at inhibiting antigen presentation than the type I anchoring inhibitor (RIAD) suggesting that the type II PKA (RII) has a more prominent role in regulating antigen presentation (Figure 1). Experiments in which the type I and type II anchoring inhibitors are added together do indicate that RI and RII are contributing separately to regulation of antigen presentation (Supplementary data, Figure 1), but we caution readers that this experiment cannot really be conducted ideally as we either halve the individual concentrations to have equal amounts of total peptide, or double the total peptide concentration to have equal amounts of individual peptides. Additional differences between DCs and T cells are the presence of AKAP79 in DC lipid rafts, and the apparent means by which Ezrin is localized to the rafts. In T cells, treatment with anchoring inhibitors disrupts localization of PKA to the raft, but does not affect AKAP localization. In the DC-like cell line KG-1, incubation with the anchoring inhibitor AKAP-IS disrupts RII and Ezrin localization to the rafts (Figure 6). The anchoring inhibitor RIAD had no effect on RI localization to KG-1 lipid rafts (RVS, unpublished data). The Ezrin/radixin/moesin family of proteins is known to interact with lipid rafts and the interaction of Ezrin with the rafts requires the phosphorylated active conformer, which tethers the plasma membrane to the cytoskeleton.49 Activation occurs on conformational changes triggered by binding of phosphatidylinositol 4,5-bisphosphate (PIP2) in the N-terminus and phosphorylation of a conserved threonine residue (Ezrin T567) in the C-terminus. These events release interactions between the N-and C-termini. Several kinases have been shown to phosphorylate the conserved threonine in Ezrin/radixin/moesin proteins. PKA has been shown to phosphorylate S66 in Ezrin and this phosphorylation regulates Immunology and Cell Biology


NK cell death.50,51 PKA also phosphorylates merlin, a tumor-suppressor protein with high homology to Ezrin/radixin/moesin proteins, at S66 and S518. Phosphorylation of merlin S518 promotes heterodimerization between Ezrin and merlin, an event suggested to convert merlin from a growth suppressive to a growth permissive state.38 These studies are consistent with a hypothesis in which disrupting PKA anchoring to Ezrin changes the phosphorylation state of Ezrin thereby altering the conformation of Ezrin and/or interaction of Ezrin with other proteins leading to removal from the lipid raft. This hypothesis is also consistent with studies performed in B cells (Ramos cells), which illustrate that Ezrin is dephosphorylated after BCR crosslinking, inducing a closed conformation that detaches from actin and lipid rafts.10 The data presented here reveal that PKA and AKAPs reside in DC lipid rafts. Localization of PKA, Ezrin and AKAP79 to the raft is anchoring dependent, and regulation of antigen presentation by anchored PKA is dependent on lipid raft integrity. Determining how PKA anchoring is regulating antigen presentation will require a better understanding of the molecules in the lipid raft that are responsible for activating PKA and acting as substrates for PKA. METHODS Cells Human subject’s protocols and consent forms were approved by the Oregon Health and Science University Institutional Review Board (IRB). Apheresis product collected in house was obtained with written consent from donors. Apheresis product was also purchased from Key Biologics (Memphis, TN, USA), and shipped overnight for morning delivery at room temperature. Peripheral blood mononuclear cells were isolated as previously described22 by ficoll gradient. Immature DCs (day 5 DC) were selected by plastic adherence. Peripheral blood mononuclear cells were plated in flasks in RPMI 2% human serum for 1 h, media was removed and the flask was rinsed gently three times with phosphate-buffered saline (PBS). Cells remaining adhered to the plastic were cultured for 5 days with 10 ng ml–1 granulocyte-macrophage colonystimulating factor and 10 ng ml–1 interleukin-4 in RPMI 1640 with 10% human serum (a generous gift of Drs David and Deborah Lewinsohn). DCs were collected by incubation with cell dissociation media as previously described and used as antigen-presenting cells.11 Day 5 DC matured by adherence result in an impure population of DC consisting of 45–65% DCs, 10–30% T/B cells and 10–20% dead cells. KG-1 cells were purchased from ATCC (Manassas, VA, USA) (ccl246) and cultured in Iscoves modified Dulbecco’s medium 20% fetal calf serum with 1% penicillin/streptomycin. KG-1 cells were stimulated by incubation with 10 ng ml–1 phorbol myristate acetate and 10 ng ml–1 tumor necrosis factor-a for 18 h at 37 1C, 5% CO2. Day 5 unstimulated KG-1 cells are suspension cells and were therefore collected by centrifugation. Stimulated KG-1 cells are collected like day 5 DC.

inhibitor peptide. Plated cells were loaded with protein antigen (0.1 mg ml–1 CFP10 protein or 100 mg ml–1 tuberculosis cell wall) for 1 h at 37 1C, 5% CO2. CFP10 is an immunodominant tuberculosis antigen used here as a tool to observe antigen presentation and T-cell activation. Tuberculosis cell wall contains CFP10 and can therefore be used as a substitute for CFP10, but much higher concentrations are needed to elicit a response. D454 E12 CD4+ clonal T cells specific for CFP10 were then added (2104 per well), and the plate was incubated overnight at 37 1C, 5% CO2. Plates were washed 3 with PBS/0.05% Tween and 100 ml of 1 mg ml–1 horseradish peroxidase-conjugated secondary anti-IFN-g mAb (7B6–1-horseradish peroxidase; Mabtech) was added and incubated for 2 h at room temperature. Plates were washed again 6 with PBS/ 0.05% Tween 20, incubated 15 min in PBS/0.05% Tween, washed 3 with PBS, then 100 ml of 3-amino-9-ethylcarbazole substrate (Vectastain AEC substrate kit; Vector, Burlingame, CA, USA) was added. After 4–10 min, the colorimetric reaction was stopped by washing thoroughly with distilled water, and plates were air dried. Spots were quantitated using an AID ELISPOT High Resolution Reader System (Cell Technology, Inc., Columbia, MD, USA).

Lipid Rafts Six million day 5 DCs, unstimulated KG-1 cells, stimulated KG-1 cells or stimulated KG-1 cells treated with 100 mM AKAP-IS, SCR for 15 min at 37 1C, 5% CO2, were lysed on ice for 30 min in 400 ml lysis buffer (20 mM MES pH6.5, 150 mM NaCl, 1% Brij-98, 1:100 protease inhibitor cocktail (Sigma-Aldrich, St Louis, MO, USA), 5 mM EDTA, 1 mM NaVO4, 25 mM NaF). Total cell lysate was mixed with an equal volume (400 ml) 90% sucrose in (20 mM MES pH6.5, 150 mM NaCl) and layered at the bottom of an ultracentrifuge tube (Beckman 331372, Miami, FL, USA), 1 ml of 30% sucrose (in 20 mM MES pH6.5, 150 mM NaCl) was layered on next, followed by 530 ml 5% sucrose (in 20 mM MES pH6.5, 150 mM NaCl). Samples were centrifuged for 16 h at 100 000g 4 1C (Beckman Optima L70 K Ultra SW41 TI rotor), and 14170 ml fractions were collected from the top. Fractions 1–9 were concentrated twofold by speedy-vac and fractions 10–14 were diluted twofold before SDS-polyacrylamide gel electrophoresis and western analyses.

Western analysis Samples were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to PVDF membrane (Immobilon-P, Millipore). Membranes were blocked in 5% milk TTBS for 1 h. Primary antibodies were incubated at the following dilutions: 1 h RT in TTBS: Ezrin (1:1000) (Thermoscientific, Waltham, MA, USA), glyceraldehyde 3-phosphate dehydrogenase (1:1000) (Imgenex, San Diego, CA, USA) AKAP79 (1:250), AKAP149 (1:250), AKAP95 (1:100), RI (1:500) and RII (1:1000) (BD Biosciences, San Jose, CA, USA), Lyn (1:1000, overnight at 4 1C in 5% bovine serum albumin TTBS) (Cell Signaling, Danvers, MA, USA), HLA-DRa (1:1000) (FL254, Santa Cruz Biotecnology, Santa Cruz, CA, USA). Secondary antibodies, goat anti-rabbit and goat anti-mouse horseradish peroxidase conjugates, were used at 1:5000 for 1 h at RT (Santa Cruz Biotechnology). NEN Renaissance chemiluminescence was used for detection.

IFN-c detection by elispot

Immunofluorescence

Antigen-specific activation of T cells was detected by measuring IFN-g production by ELISPOT as previously described.23 Briefly, 96-well nitrocellulose-backed plates (MAHA S4510; Millipore, Bedford, MA, USA) were coated as recommended by the manufacturer with 10 mg ml–1 capture mouse anti-IFNg (1-D1K; Mabtech, Nacka, Sweden) overnight at room temperature. Plates were then washed and blocked and untreated or treated day 5 DC (1–2104 per well) or KG-1 cells (2.5–5103 per well) were plated. Cells were treated by addition of 100 mM Ht31 or Ht31p (Ht31p is the control peptide in which a proline substitution in the middle kinks the helix thwarting the ability to bind AKAPs and disrupt interaction with PKA) (Promega, Madison, WI, USA), or the indicated concentrations of IS, SCR, RIAD, RSCR (IS; AMAQIEYLAKQI VDNAIQQAKG-RRRRRRRRRRR, SCR; AMAQDVEIQLKAAYNQKLIAIG-RR RRRRRRRRR, RIAD; LEQYANQLADQIIKEATEK-RRRRRRRRRRR, RSCR; IE KELAQQYQNADAITLEK-RRRRRRRRRRR (Biomatik, Cambridge, Ontario, Canada)) 20 min before addition of antigen. In DC ELISPOTs, the DCs were rinsed before plating and addition of antigen to remove any excess anchoring

Day 5 DC were generated by plating DC onto poly-L-lysine (Sigma-Aldrich) coated round coverslips in a 12-well plate and incubating in RPMI 10% human serum for 3 h overnight at 37 1C, 5% CO2. The DC were fixed with 2% paraformaldehyde in PBS for 1 h, washed with PBS and blocked with 2% human serum, 2% goat serum and 0.5% fetal calf serum in PBS with 0.1% saponin 15 min to overnight at 4 1C. DC were washed 3 with 1 ml of PBS+0.1% saponin and incubated with a 1:100 dilution of anti-AKAP79 or anti-Ezrin. DC were washed 3 with 1 ml of PBS+0.1% saponin and incubated with 1:500 dilution donkey anti-mouse Texas-red-conjugated secondary antibody in PBS+0.1% saponin in a moist staining chamber for 1 h. DC were washed again and incubated with 1:00 dilution of fluorescein isothiocyanateconjugated HLA-DR antibody (BD Biosciences) or 1:50 dilution of fluorescein isothiocyanate-conjugated ChTx in PBS+0.1% saponin in a moist staining chamber for 1 h. Finally, DC were washed two times with PBS and one time with milliQ H20, mounted using Fluoromount G (Southern Biotech, Birmingham, AL, USA), and analyzed by confocal microscopy (Leica TCS-NT confocal

Immunology and Cell Biology

AKAPs in DC lipid rafts RV Schillace et al 657 imaging system with a 40X Pl apo oil emersion objective). Staining of cells with secondary antibody alone was used to eliminate detection of background staining.

Quantitation of colocalization Quantitation of confocal images was performed with Volocity cellular imaging software (Perkin Elmer). The colocalization coefficient My describes the proportion of red pixels that overlap with green pixels compared with the total number of red pixels or the fluorescence overlap of the AKAP (either AKAP79 or Ezrin) with the lipid raft control (either ChTx or HLA-DR). Coefficient values can range between 1 and 0, with 1 being high colocalization, 0 being none; My values 40.5 are indicative of colocalization.24 Manders overlap coefficient (R) was also calculated and values 40.6 are indicative of colocalization.24

Statistics Student’s t-test were performed in Microsoft Excel or Graphpad prism (La Jolla, CA, USA). Statistical significance is considered when Po0.05.

ACKNOWLEDGEMENTS This research was supported by Department of Veterans Affairs, Veterans Health Administration, Office of Research and Development, Biomedical Laboratory Research and Development Service (DWC). We thank Sarah Fiedler and Dr Sonemany Salinthone for helpful comments during paper preparation.

1 Rubio MT, Means TK, Chakraverty R, Shaffer J, Fudaba Y, Chittenden M et al. Maturation of human monocyte-derived dendritic cells (MoDCs) in the presence of prostaglandin E2 optimizes CD4 and CD8T cell-mediated responses to protein antigens: role of PGE2 in chemokine and cytokine expression by MoDCs. Int Immunol 2005; 17: 1561–1572. 2 Krause P, Singer E, Darley PI, Klebensberger J, Groettrup M, Legler DF. Prostaglandin E2 is a key factor for monocyte-derived dendritic cell maturation: enhanced T cell stimulatory capacity despite IDO. J Leukoc Biol 2007; 82: 1106–1114. 3 Steinbrink K, Paragnik L, Jonuleit H, Tuting T, Knop J, Enk AH. Induction of dendritic cell maturation and modulation of dendritic cell-induced immune responses by prostaglandins. Arch Dermatol Res 2000; 292: 437–445. 4 Scandella E, Men Y, Gillessen S, Forster R, Groettrup M. Prostaglandin E2 is a key factor for CCR7 surface expression and migration of monocyte-derived dendritic cells. Blood 2002; 100: 1354–1361. 5 Cambier JC, Newell MK, Justement LB, McGuire JC, Leach KL, Chen ZZ. Ia binding ligands and cAMP stimulate nuclear translocation of PKC in B lymphocytes. Nature 1987; 327: 629–632. 6 Taylor SS, Buechler JA, Yonemoto W. cAMP-dependent protein kinase: framework for a diverse family of regulatory enzymes. Annu Rev Biochem 1990; 59: 971–1005. 7 Skroblin P, Grossmann S, Schafer G, Rosenthal W, Klussmann E. Mechanisms of protein kinase A anchoring. Int Rev Cell Mol Biol 2010; 283: 235–330. 8 Torgersen KM, Aandahl EM, Tasken K. Molecular architecture of signal complexes regulating immune cell function. Handb Exp Pharmacol 2008; 186: 327–363. 9 Ruppelt A, Mosenden R, Gronholm M, Aandahl EM, Tobin D, Carlson CR et al. Inhibition of T cell activation by cyclic adenosine 5¢-monophosphate requires lipid raft targeting of protein kinase A type I by the A-kinase anchoring protein Ezrin. J Immunol 2007; 179: 5159–5168. 10 Gupta N, Wollscheid B, Watts JD, Scheer B, Aebersold R, DeFranco AL. Quantitative proteomic analysis of B cell lipid rafts reveals that Ezrin regulates antigen receptormediated lipid raft dynamics. Nat Immunol 2006; 7: 625–633. 11 Schillace RV, Miller CL, Pisenti N, Grotzke JE, Swarbrick GM, Lewinsohn DM et al. A-kinase anchoring in dendritic cells is required for antigen presentation. PLoS ONE 2009; 4: e4807. 12 Carr DW, Stofko Hahn RE, Fraser ID, Bishop SM, Acott TS, Brennan RG et al. Interaction of the regulatory subunit (RII) of cAMP-dependent protein kinase with RII-anchoring proteins occurs through an amphipathic helix binding motif. J Biol Chem 1991; 266: 14188–14192. 13 Carlson CR, Lygren B, Berge T, Hoshi N, Wong W, Tasken K et al. Delineation of type I protein kinase A-selective signaling events using an RI anchoring disruptor. J Biol Chem 2006; 281: 21535–21545. 14 Alto NM, Soderling SH, Hoshi N, Langeberg LK, Fayos R, Jennings PA et al. Bioinformatic design of A-kinase anchoring protein-in silico: a potent and selective peptide antagonist of type II protein kinase A anchoring. Proc Natl Acad Sci USA 2003; 100: 4445–4450. 15 Torheim EA, Jarnaess E, Lygren B, Tasken K. Design of proteolytically stable RIanchoring disruptor peptidomimetics for in vivo studies of anchored type I protein kinase A-mediated signalling. Biochem J 2009; 424: 69–78.

16 Li Y, Sroubek J, Krishnan Y, McDonald TV. A-kinase anchoring protein targeting of protein kinase A and regulation of HERG channels. J Membr Biol 2008; 223: 107–116. 17 Gold MG, Lygren B, Dokurno P, Hoshi N, McConnachie G, Tasken K et al. Molecular basis of AKAP specificity for PKA regulatory subunits. Mol Cell 2006; 24: 383–395. 18 Dykstra M, Cherukuri A, Pierce SK. Rafts and synapses in the spatial organization of immune cell signaling receptors. J Leukoc Biol 2001; 70: 699–707. 19 Lingwood D, Simons K. Lipid rafts as a membrane-organizing principle. Science 2010; 327: 46–50. 20 Huby RD, Dearman RJ, Kimber I. Intracellular phosphotyrosine induction by major histocompatibility complex class II requires co-aggregation with membrane rafts. J Biol Chem 1999; 274: 22591–22596. 21 Anderson HA, Hiltbold EM, Roche PA. Concentration of MHC class II molecules in lipid rafts facilitates antigen presentation. Nat Immunol. 2000; 1: 156–162. 22 Salinthone S. Lipoic acid attenuates inflammation via cAMP and protein kinase A signaling. J Nutr Biochem 2010 (in press). 23 Heinzel AS, Grotzke JE, Lines RA, Lewinsohn DA, McNabb AL, Streblow DN et al. HLA-E-dependent presentation of Mtb-derived antigen to human CD8+ T cells. J Exp Med 2002; 196: 1473–1481. 24 Zinchuk V, Zinchuk O. Quantitative colocalization analysis of confocal fluorescence microscopy images. Curr Protoc Cell Biol 2008 Chapter 4:Unit 4 19. 25 Xu S, Huo J, Gunawan M, Su I-H, Lam K-P. Activated dectin-1 localizes to lipid raft microdomains for signaling and activation of phagocytosis and cytokine production in dendritic cells. J Biol Chem 2009; 284: 22005–22011. 26 Koeffler HP, Golde DW. Acute myelogenous leukemia: a human cell line responsive to colony-stimulating activity. Science 1978; 200: 1153–1154. 27 St Louis DC, Woodcock JB, Franzoso G, Blair PJ, Carlson LM, Murillo M et al. Evidence for distinct intracellular signaling pathways in CD34+ progenitor to dendritic cell differentiation from a human cell line model. J Immunol 1999; 162: 3237–3248. 28 Teobald I, Dunnion DJ, Whitbread M, Curnow SJ, Browning MJ. Phenotypic and functional differentiation of KG-1 into dendritic-like cells. Immunobiology 2008; 213: 75–86. 29 Simons K, Ikonen E. Functional rafts in cell membranes. Nature 1997; 387: 569–572. 30 Brown DA, London E. Functions of lipid rafts in biological membranes. Annu Rev Cell Dev Biol 1998; 14: 111–136. 31 Knorr R, Karacsonyi C, Lindner R. Endocytosis of MHC molecules by distinct membrane rafts. J Cell Sci 2009; 122: 1584–1594. 32 Alfalah M, Wetzel G, Fischer I, Busche R, Sterchi EE, Zimmer KP et al. A novel type of detergent-resistant membranes may contribute to an early protein sorting event in epithelial cells. J Biol Chem 2005; 280: 42636–42643. 33 Drevot P, Langlet C, Guo XJ, Bernard AM, Colard O, Chauvin JP et al. TCR signal initiation machinery is pre-assembled and activated in a subset of membrane rafts. Embo J 2002; 21: 1899–1908. 34 Munõz P, Navarro M-d-C, Pavoń EJ, Salmeroń J, Malavasi F, Sancho J et al. CD38 signaling in T cells is initiated within a subset of membrane rafts containing Lck and the CD3-z subunit of the T cell antigen receptor. J Biol Chem 2003; 278: 50791–50802. 35 Caparros E, Munoz P, Sierra-Filardi E, Serrano-Gomez D, Puig-Kroger A, Rodriguez-Fernandez JL et al. DC-SIGN ligation on dendritic cells results in ERK and PI3 K activation and modulates cytokine production. Blood 2006; 107: 3950–3958. 36 Karacsonyi C, Knorr R, Fulbier A, Lindner R. Association of major histocompatibility complex II with cholesterol- and sphingolipid-rich membranes precedes peptide loading. J Biol Chem 2004; 279: 34818–34826. 37 Poloso NJ, Muntasell A, Roche PA. MHC class II molecules traffic into lipid rafts during intracellular transport. J Immunol 2004; 173: 4539–4546. 38 Alfthan K, Heiska L, Gronholm M, Renkema GH, Carpen O. Cyclic AMP-dependent protein kinase phosphorylates merlin at serine 518 independently of p21-activated kinase and promotes merlin-Ezrin heterodimerization. J Biol Chem 2004; 279: 18559–18566. 39 Shurin GV, Tourkova IL, Chatta GS, Schmidt G, Wei S, Djeu JY et al. Small Rho GTPases regulate antigen presentation in dendritic cells. J Immunol 2005; 174: 3394–3400. 40 Louvet-Valleé S. ERM proteins: from cellular architecture to cell signaling. Biol Cell 2000; 92: 305–316. 41 Cong M, Perry SJ, Lin FT, Fraser ID, Hu LA, Chen W et al. Regulation of membrane targeting of the G protein-coupled receptor kinase 2 by protein kinase A and its anchoring protein AKAP79. J Biol Chem 2001; 276: 15192–15199. 42 Gardner LA, Tavalin SJ, Goehring AS, Scott JD, Bahouth SW. AKAP79-mediated targeting of the cyclic AMP-dependent protein kinase to the beta1-adrenergic receptor promotes recycling and functional resensitization of the receptor. J Biol Chem 2006; 281: 33537–33553. 43 Panina-Bordignon P, Mazzeo D, Lucia PD, D’Ambrosio D, Lang R, Fabbri L et al. Beta2agonists prevent Th1 development by selective inhibition of interleukin 12. J Clin Invest 1997; 100: 1513–1519. 44 Patel HH, Murray F, Insel PA. G-protein-coupled receptor-signaling components in membrane raft and caveolae microdomains. Handb Exp Pharmacol 2008; 186: 167–184. 45 Yamaoka K, Yano A, Kuroiwa K, Morimoto K, Inazumi T, Hatae N et al. Prostaglandin EP3 receptor superactivates adenylyl cyclase via the Gq/PLC/Ca2+ pathway in a lipid raft-dependent manner. Biochem Biophys Res Commun 2009; 389: 678–682.


AKAPs in DC lipid rafts RV Schillace et al 658 46 Guo W, Castaigne JG, Mooney N, Charron D, Al-Daccak R. Signaling through HLA-DR induces PKC beta-dependent B cell death outside rafts. Eur J Immunol 2003; 33: 928–938. 47 Bouillon M, El Fakhry Y, Girouard J, Khalil H, Thibodeau J, Mourad W. Lipid raftdependent and -independent signaling through HLA-DR molecules. J Biol Chem 2003; 278: 7099–7107. 48 Cornez I, Tasken K. Spatiotemporal control of cyclic AMP immunomodulation through the PKA-Csk inhibitory pathway is achieved by anchoring to an Ezrin-EBP50-PAG scaffold in effector T cells. FEBS Lett 2010; 584: 2681–2688.

49 Prag S, Parsons M, Keppler MD, Ameer-Beg SM, Barber P, Hunt J et al. Activated Ezrin promotes cell migration through recruitment of the GEF Dbl to lipid rafts and preferential downstream activation of Cdc42. Mol Biol Cell 2007; 18: 2935–2948. 50 Zhou R, Cao X, Watson C, Miao Y, Guo Z, Forte JG et al. Characterization of protein kinase A-mediated phosphorylation of Ezrin in gastric parietal cell activation. J Biol Chem 2003; 278: 35651–35659. 51 Wang S, Guo Z, Xia P, Liu T, Wang J, Li S et al. Internalization of NK cells into tumor cells requires Ezrin and leads to programmed cell-in-cell death. Cell Res 2009; 19: 1350–1362.

The Supplementary Information that accompanies this paper is available on the Immunology and Cell Biology website (http://www.nature.com/icb)
