Direct interaction between Fc␥RI (CD64) and periplakin controls receptor endocytosis and ligand binding capacity

Jeffrey M. Beekman*†, Jantine E. Bakema*†, Jan G. J. van de Winkel*‡§, and Jeanette H. W. Leusen*§¶

*Immunotherapy Laboratory, Department of Immunology, and †Medarex Europe, University Medical Center Utrecht, 3584 EA, Utrecht, The Netherlands; and ‡Genmab, Yalelaan 60, 3584 CM, Utrecht, The Netherlands

Edited by Jonathan W. Uhr, University of Texas Southwestern Medical Center, Dallas, TX, and approved May 24, 2004 (received for review February 20, 2004) Fc␥RI depends for its biological function on both the intracellular terminus (which can associate with filaments) from the domain of the ␣-chain and associated Fc receptor (FcR) ␥-chains. -interacting C terminus (17–19). Additionally, recent However, functional effectors of Fc␥RI’s intracellular do- studies suggested that PPL is involved in signaling through main have not been identified. In this study, we identified protein kinase B and G in nonimmune cells (20, 21). We periplakin (PPL) as a selective interacting protein for the intracel- found PPL expression in myeloid cells, and expression was lular tail of Fc␥RI but no other activatory FcRs. The interaction was increased by IFN-␥. Immunoprecipitation of transfected IIA1.6 confirmed by coimmunoprecipitation and blot-overlay assays. PPL cells and binding assays with recombinant proteins confirmed and Fc␥RI colocalized at the plasma membrane in monocytes and the interaction between Fc␥RI and PPL. Furthermore, we cell transfectants, and both were up-regulated by IFN-␥. By ex- assessed the role of PPL in Fc␥RI-mediated ligand binding, pressing C-terminal PPL in transfectants, we established a pivotal receptor modulation, and antigen presentation. role for this protein in Fc␥RI ligand binding, endocytosis, and antigen presentation. These data illustrate that intracellular pro- Materials and Methods tein interactions with a multisubunit FcR ␣-chain can confer unique Generated Constructs. cDNAs encoding the cytosolic tail se- properties to the receptor. quences of human Fc␥RI, Fc␥RIIa, Fc␥RIIIa, Fc␣RI, Fc␧RI, and murine (m) Fc␥RI were inserted in pGBT9. Casein kinase ␤ ␥ ntibody (Ig) complexes can induce potent immune effector II -subunit was cloned into pGAD-GH. Full-length (fl) Fc RI ␥ functions by crosslinking Fc receptors (FcRs) (reviewed in was cloned into pcDNA3, the cytosolic tail of Fc RI also in A ␥ refs. 1 and 2). Multiple FcRs exist as complexes of unique pGEX-2T (Amersham Biosciences). The mFcR -chain and ␥ Ig-binding ␣-chains that associate with promiscuous signaling mutant Y65F,Y76F were inserted in a pNUT vector described subunits harboring immunoreceptor tyrosine-based activation in ref. 10. cDNAs encoding PPL clone 3.4 were inserted in ͞ motifs. The FcR ␥-chain associates with a variety of FcR ␣-chains pGBT9, pcDNA3.1 His (Invitrogen, Leek, The Netherlands), that lack known signaling motifs, and it is crucial for induction and pGEX-2T, from which the GST tag had been replaced by a of phagocytosis and NADPH oxidase assembly, for example His tag (six histidine residues). fl-PPL was in pBluescript (a kind (3–5). These observations have led to the paradigm that multi- gift of F. M. Watt, Keratinocyte Laboratory, London Research subunit FcRs do not signal through the cytosolic domains (CYs) Institute, London). cDNAs generated by PCR were verified by of their ␣-chains, but signal through the FcR ␥-chain. However, sequence analysis. there is considerable evidence that CYs of particular FcR ␣-chains modulate FcR receptor function through as-yet- Yeast Two-Hybrid Screening. Oligo(dT)-primed HeLa (pGAD- undocumented signaling pathways. GH; gift of the Cell Biology Department of University Medical Fc␥RI is a high-affinity IgG receptor with a strict myeloid cell Center Utrecht) and bone marrow (pACT-2) cDNA libraries ␥ distribution that is up-regulated during inflammation (6, 7). were screened for Fc RI-CY-interacting proteins in yeast strain Its in vivo role is illustrated by Fc␥RIϪ/Ϫ mice that exhibit YGH1 (all from Clontech). Protein interactions were assessed by ␤ impaired Ab-dependent cellular processes such as bacterial growth of colonies on histidine-depleted medium, and -galac- ␥ clearance, phagocytosis, antigen presentation, and cytokine pro- tosidase activity was assessed with a replica filter assay. Fc RI- ␥ duction (8, 9). interacting proteins and Fc RI signaling motifs were identified by scanning cDNA sequences with BLAST (www.ncbi.nlm. Within the multisubunit FcR family, it has been shown that the ͞ Fc␥RI ␣-chain induces effector functions. The Fc␥RI cytosolic nih.gov BLAST). ␥ domain (Fc RI-CY) mediated MHC class II antigen presenta- ␥ tion without active FcR ␥-chain signaling (10), whereas deletion RT-PCR. Fc RI was amplified from oligo(dT)-primed cDNA by 25 ␥ PCR cycles (30 s at 95°C, 30 s at 57°C, and 45 s at 72°C) by using of Fc RI-CY retarded kinetics of endocytosis and phagocytosis ␥ Ј and abrogated Fc␥RI-triggered IL-6 secretion (11). Thus far, primers annealing to Fc RI-CY (5 -CTGTTCTCTGGGTGA- CAATACG-3Ј and 5Ј-AGAACTGTGTGTCTCATGGTAT- only actin-binding protein 280 (filamin A) has been described to Ј bind Fc␥RI-CY under some conditions, but effects on receptor 3 ). PPL PCRs were performed for 35 cycles (30 s at 95°C, 30 s function have not been shown (12). Here, we identified periplakin (PPL) to selectively interact This paper was submitted directly (Track II) to the PNAS office. with Fc␥RI-CY in yeast two-hybrid screens. PPL is a cytosolic Abbreviations: CY, cytosolic domain; EA, erythrocyte–antibody; FcR, Fc receptor; fl, full- protein of 195 kDa and a member of the cytoskeletal-associated length; m, murine; OVA, ovalbumin; PPL, periplakin; RIPA, radioimmunoprecipitation plakin family (reviewed in ref. 13). Plakins contribute to the assay. structural integrity of epithelia by, e.g., tethering cytoplasmic §J.G.J.v.d.W. and J.H.W.L. contributed equally to this work. domains of adhesive receptors to the . PPL is ¶To whom correspondence should be addressed at: Department of Immunology, Lundlaan associated with desmosomes in keratinocytes and is involved in 6, University Medical Center Utrecht, 3584 EA, Utrecht, The Netherlands. E-mail: cornified envelope assembly (14–16). PPL’s central rod domain [email protected]. mediates homo- and heterodimerization and separates the N © 2004 by The National Academy of Sciences of the USA

10392–10397 ͉ PNAS ͉ July 13, 2004 ͉ vol. 101 ͉ no. 28 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0401217101 Downloaded by guest on September 30, 2021 at 52°C, and 1 min at 72°C) with primers annealing to human and Fc␥RI mAb H22 (23) (Medarex) for 60 min at 37°C and stained mouse PPL (5Ј-AGCCAAGAGATCCAG-3Ј and 5Ј-CTACT- for PPL as indicated above. Twenty-four hours after electropo- TCTGCCCAG-3Ј). The presence of cDNA in each sample was ration of PPL, stably (truncated) Fc␥RI-transfected IIA1.6 cells verified by GAPDH PCR. were separated by Ficoll density gradient centrifugation, and viable cells were processed for immunofluorescence as in ref. 22. Intracellular PPL Staining. Human peripheral blood monocytes were isolated from buffy coats of healthy volunteers. Mononu- Immunoprecipitation. IIA1.6 cells were stably transfected with clear cells were collected from Ficoll gradients and cultured with Fc␥RI and FcR ␥-chain as described in ref. 24, and C-terminal Iscove’s modified Dulbecco’s medium containing L-glutamine PPL was transiently transfected by electroporation. After 48 h, ͞ (GIBCO BRL) supplemented with 10% FCS, penicillin, and homogenates of 107 cells per immunoprecipitation in 500 ␮lof Ϸ streptomycin. After 3hat37°C, nonadherent cells were radioimmunoprecipitation assay (RIPA) buffer (150 mM NaCl͞ removed and adherent cells were incubated overnight with 300 ⅐ ͞ ͞ ͞ ͞ ␥ ␥ 100 mM Tris HCl, pH 8.3 0.1% SDS 0.5% deoxycholate 1% IU ml IFN- (IFN- 1b, Boehringer Ingelheim, Biberach, Ger- Triton X-100) were spun for 10 min at 1,500 ϫ g, and superna- many). Monocytes and IIA1.6 cells were fixed in PBS with 3.3% tants were incubated with protein A beads preabsorbed with paraformaldehyde and stained in PBS containing 0.2% BSA, 7445 serum (PPL-specific rabbit serum, B. Burgering; ref. 20). 0.1% saponin, 5% mouse serum, and 5% goat serum. PPL was After a 1-h rotation at 4°C, beads were washed four times with detected by rabbit serum 5117 (characterized in ref. 20 and RIPA buffer and boiled in reducing Laemmli sample buffer. kindly provided by B. Burgering, University Medical Center Samples were analyzed by Western blotting by using rabbit Utrecht) diluted 100-fold and subsequently by goat-anti-rabbit- serum to detect Fc␥RI (no. 3532, kindly provided by R. Kim- FITC (Jackson ImmunoResearch). Preimmune serum was used as negative control. berly, University of Alabama at Birmingham, Birmingham) and rabbit serum 5117 to detect PPL. Confocal Immunofluorescence Microscopy. Monocytes were iso- Coprecipitation of GST-Fc␥RI and His-PPL. Fusion protein of lated as described above, resuspended in PBS with 2 mM EDTA ␥ ␥ and 0.1% BSA, washed in medium, adhered to poly(L-lysine) Fc RI-CY and GST (GST-Fc RI) or His-tagged PPL clone 3.4 slides, and processed for immunofluorescence as described in (His-PPL) were purified from nondenatured Escherichia coli ref. 22. Aspecific interactions were blocked by the addition of 5% lysates with glutathione Sepharose 4B (Amersham Biosciences) mouse serum and 5% goat serum during staining. Fc␥RI was and Ni-NTA Spin columns (Qiagen, Hilden, Germany), respec- stained with mAb m22-FITC (Medarex, Annandale, NY) or tively. Tosyl-activated M-280 Dynabeads (Dynal, Oslo) were 10.1-FITC (Serotec), and PPL was stained with rabbit serum precoated with anti-GST Ab (Amersham Biosciences) according 5117 (4,000-fold diluted) or preimmune serum as control. Goat- to the manufacturers’ instructions. Beads were incubated over- anti-rabbit CY3 (Jackson ImmunoResearch) was used as a night with 1 ␮g of GST-Fc␥RI or GST and 1 ␮g of His-PPL in secondary Ab. For internalization experiments, monocytes ad- 500 ␮l of RIPA buffer. Beads were washed three times in RIPA, hered to poly(L-lysine) slides were incubated with 50 ␮lof boiled in Laemmli sample buffer, and analyzed by SDS͞PAGE medium containing 10 ␮g͞ml FITC-conjugated humanized anti- and Western blotting. IMMUNOLOGY

Fig. 1. PPL binds the Fc␥RI cytosolic tail in yeast two-hybrid screens. (A) Protein sequences of the cytosolic tails of Fc␥RI, Fc␥RIIa, Fc␥RIIIa, Fc␣RI, Fc␧RI, and mFc␥RI are depicted; putative PKC (dotted line) and casein kinase II (solid line) recognition motifs are indicated for Fc␥RI. Amino acids are color-coded as follows: green, hydrophobic͞aromatic; blue, basic͞hydrophilic; red, acidic; yellow, aliphatic. Predicted regions (PredictProtein at www.embl-heidelberg.de) are indicated. TM, transmembrane. Numbers at the top refer to amino acids of Fc␥RI, and numbers at the bottom indicate the size of the predicted cytoplasmic domain. (B–D) Yeast two-hybrid analyses. For each combination of cDNA, three independent colonies were transferred to histidine-lacking plates and tested for ␤-galactosidase activity (indicated by blue staining). (B) Protein interactions of Fc␥RI-CY with PPL cDNA clones 2.2, 1.77, 3.4, 4.27, and 91 and casein kinase II. (C) Protein interactions of PPL clone 3.4 with Fc␥RI-CY that lacks residues 311–313 (-VTI), Fc␥RI with a spacer of six glycines between Gal4-BD and Fc␥RI-CY (6ϫGly-Fc␥RI), and PPL 3.4. (D) Protein interactions of PPL clone 3.4 with a panel of activatory FcRs. One representative experiment of three is shown. (E) A hypothetical drawing of a PPL dimer. The five independent hits from yeast two-hybrid screens are indicated below the drawing.

Beekman et al. PNAS ͉ July 13, 2004 ͉ vol. 101 ͉ no. 28 ͉ 10393 Downloaded by guest on September 30, 2021 Blot Overlay for GST-Fc␥RI and His-PPL. Purified fractions of His- of OVA (100 ␮g͞ml) was added to each IIA1.6 cell line, which tagged PPL clone 3.4 were separated by SDS͞PAGE and was set at 100%. [3H]Thymidine incorporation was expressed as transferred to poly(vinylidene difluoride) membranes. Mem- percentage of this maximum level. branes were blocked with 5% low-fat milk powder in PBS and incubated overnight with 50 nM purified GST-Fc␥RI or GST in Erythrocyte–Antibody (EA) Rosette Assay. Human erythrocytes 3 ml of RIPA buffer supplemented with 0.1% BSA and 0.5% were prepared by Ficoll͞Hypaque density centrifugation and low-fat milk powder. Membranes were washed in RIPA buffer, stored in sterile Alsever’s solution at 4°C. Erythrocytes were and binding of GST-Fc␥RI was detected by incubation with opsonized by serial dilutions of hybridoma supernatants con- anti-GST Ab (Amersham Biosciences) and rabbit anti-goat IgG taining mIgG2a anti-human glycophorin A mAb (30) for1hat Ab conjugated to horseradish peroxidase (Pierce), followed by 2 ϫ 108 erythrocytes per ml at 4°C. Erythrocytes were washed, enhanced chemiluminescence and autoradiography. Methodol- and 5 ϫ 106 erythrocytes were resuspended with 1 ϫ 105 ogy was adapted from refs. 25 and 26. transfected IIA1.6 cells in 50 ␮l of RPMI medium 1640 in round-bottom 96-well plates and incubated for 60 min at 4°C. EA Modulation of Fc␥RI Expression. Rabbit IgG–ovalbumin (OVA) rosettes were resuspended after 30 and 60 min and fixed by complexes were generated as described (10). A total of 2.5 ϫ 105 addition of 3% paraformaldehyde for 30 min. Cells were diluted transfected IIA1.6 cells were transferred to 96-well plates in 100 2- to 3-fold in Hepes-buffered RPMI medium 1640, and bound ␮l of RPMI medium 1640 with 10% FCS and increasing con- erythrocytes were counted by light microscopy. Expression levels ␥ Ј centrations of rabbit IgG–OVA complexes. Samples were incu- of Fc RI were measured by flow cytometry by using F(ab )2 of ␥ Ј ␬ bated at 37°C for 16 h, washed, and stained with CD64 mAb 32.2 anti-Fc RI mAb H22 and goat F(ab )2 anti-human light Ј (Medarex; ref. 27), followed by phycoerythrin-labeled F(ab )2 of chain-FITC (Southern Biotechnology Associates). goat anti-mouse IgG1 antiserum (Jackson ImmunoResearch). Modulation of Fc␥RI was determined by flow cytometry as the Results and Discussion reduction in receptor expression relative to untreated transfec- Fc␥RI Cytosolic Tail Interacts with PPL. Alignment of intracellular tants (28). domains of FcR did not show high sequence similarity (Fig. 1A). Fc␥RI-CY was found relatively large for non-tyrosine- Antigen Presentation Assays. Antigen presentation assays were containing CYs and shared highest sequence similarity with carried out as described in ref. 29. Briefly, IIA1.6 transfectants Fc␥RIIA (only 20%, compared with 19% for mFc␥RI and 12% were incubated with different concentrations of immune com- for Fc␥RIIIA). No functional signal sequences were recognized plexes and cocultured with OVA-specific DO11.1 T cells that within Fc␥RI-CY, but two frequently observed phosphorylation produce IL-2 after T cell receptor triggering. IL-2 in superna- sites of casein kinase II and one of PKC were predicted in tants is indicated by the proliferative response of IL-2-dependent Fc␥RI-CY. However, interactions of Fc␥RI-CY with the ␤- CTLL-2 cells measured by [3H]thymidine incorporation. To subunit of casein kinase II were not observed (Fig. 1B). assess receptor-independent antigen presentation, a large excess We identified PPL as a selective Fc␥RI-CY-interacting pro-

Fig. 2. Expression of PPL in leukocytes. (A) RT-PCRs were performed for Fc␥RI, PPL, and GAPDH with RNA of MCF7, IIA1.6, Jurkat, Raji, and U937 cell lines, primary human polymorphonuclear cells, and Fc␥RI [transgenic (Tg) and nontransgenic (NTg) littermates] murine bone marrow-derived dendritic cells (DC) (39). IFN-␥ (300 units͞ml) stimulations were overnight. (B) Flow cytometric analysis of PPL expression in primary monocytes. Dotted trace, preimmune serum background staining; open solid trace, PPL levels in freshly isolated monocytes; shaded trace, PPL levels after overnight IFN-␥ stimulation. (C) Confocal microscope analyses of monocytes, cultured overnight with or without IFN-␥, adhered to poly(L-lysine) coated slides, and fluorescently labeled for Fc␥RI (mAb m22-FITC) and PPL (rabbit serum 5117͞goat anti-rabbit-CY3). Merged pictures show colocalization in yellow. Alternatively, IFN-␥ stimulated monocytes were incubated with the FITC-labeled anti-Fc␥RI mAb H22 for 1 h and stained for PPL. (D) Subcellular distribution of Fc␥RI and PPL in IIA1.6 cells. PPL (fl or clone 3.4) was transiently expressed in IIA1.6 cells stably transfected with WT Fc␥RI or mutant Fc␥RI-⌬332 (which lacked residues 333–374) and WT FcR ␥-chain (␥)or␥Y65F,Y76F. Fc␥RI was stained with 10.1-FITC. PPL is shown in red, and merged pictures of two representative cells show colocalization in yellow (n ϭ 3).

10394 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0401217101 Beekman et al. Downloaded by guest on September 30, 2021 Fig. 3. Physical interaction between Fc␥RI and PPL. (A) PPL clone 2.2 was transiently expressed in IIA1.6 cells stably transfected with Fc␥RI with (Left) or without (Right) the FcR ␥-chain. Cells were lysed in RIPA buffer after 48 h, and homogenates were incubated with protein A beads preabsorbed with rabbit anti-PPL antiserum (5117) or control serum. Pre- and post-PPL immunoprecipitated supernatants, anti-PPL, and control beads were subjected to SDS͞PAGE, blotted onto poly(vinylidene difluoride) membranes, and stained for PPL (Upper) and Fc␥RI (Lower). One representative experiment of three is shown. (B and C) Recombinant Fc␥RI-CY (GST-Fc␥RI) and PPL clone 3.4 (His-PPL) were purified and assessed for interaction by immunoprecipitation (B) or blot overlay (C). (B) His-PPL was precipitated with anti-GST beads in the presence of GST-Fc␥RI but not GST tag alone. Membranes were incubated with anti-His Ab and subsequently with anti-GST Ab. (C) Blot-overlay analysis. Lanes 1 and 2, Coomassie stainings of purified His-PPL fractions containing Ϸ1 and 0.3 ␮g of His-PPL. Lanes 3 and 4, His-PPL was transferred to membranes and stained by anti-His Ab. Lanes 5–8, Blots were overlaid with GST-Fc␥RI (lanes 5 and 6) or GST tag alone (lanes 7 and 8), and binding was detected by subsequent incubations with anti-GST Ab. Lanes 7 and 8 were overexposed compared with lanes 5 and 6. One representative experiment of three is shown.

tein in four independent yeast two-hybrid screens of HeLa Plasma Membrane Localization of Fc␥RI and PPL. Next, we studied (clones 2.2 and 3.4) and bone marrow cDNA libraries (clones the subcellular localization of both proteins in primary mono- 1.77, 4.27, and 4.91) (Fig. 1 B–D). Notably, deletion of the first cytes and transfectants. Cells were stained for Fc␥RI (green) three amino acids of our Fc␥RI-CY construct (-VTI) abrogated and PPL (red) and examined by using confocal scanning laser the interaction (Fig. 1C), implicating these to be part of the PPL microscopy. Both proteins colocalized (indicated by the yellow binding site of Fc␥RI. This finding would suggest a larger color) at the plasma membrane in specific patches in monocytes ␥ cytosolic region than is indicated by computation. Interactions (Fig. 2B). IFN- incubation enhanced detection of colocaliza- were not disturbed when a glycine kinker was cloned between tion and increased PPL staining intensity, as was indicated by Gal4-binding domain (Gal4-BD) and Fc␥RI-CY but were abol- intracellular stainings of PPL by flow cytometry (Fig. 2B). IMMUNOLOGY ␥ Intracellular staining of Fc␥RI and to some extent PPL was also ished when Gal4-BD was expressed without the Fc RI insert ␥ (Fig. 1D, empty). No binding to any other FcR was detected, observed but did not specifically colocalize. When Fc RI was ␥ triggering with the FITC-labeled anti-CD64 mAb H22, distinct including mFc RI (Fig. 1D), which is 32% identical and 56% ␥ similar in the proximal part of Fc␥RI-CY. Furthermore, we staining patterns were observed for internalized Fc RI and PPL confirmed the capacity of PPL to form homodimers as has that relocalized to the cells’ interior (Fig. 2C). These data suggested that PPL acts before or at early moments of Fc␥RI been suggested (14). The topology of PPL and location of the triggering. IFN-␥ treated polymorphonuclear cells were also five clones identified by yeast two-hybrid screens are indicated stained, but high background fluorescence prevented specific in Fig. 1E. detection of PPL (data not shown). In IIA1.6 transfectants (Fig. 2D), fl-PPL and C-terminal PPL PPL Is Expressed in Myeloid Cells. The cellular distribution of PPL (c-PPL) colocalized with Fc␥RI at the plasma membrane as in was determined by RT-PCR and was found highly similar to ␥ ␥ ␥ monocytes. Active or mutated ( Y65F,Y76F) FcR -chain did Fc RI (Fig. 2A). U937 cells, activated human granulocytes not affect staining patterns. However, deletion of Fc␥RI resi- (polymorphonuclear cells), and bone marrow-derived dendritic ␥ ⌬ ␥ dues 333–374 (Fc RI- 332) induced clear green and red patches cells from Fc RI transgenic mice (31) expressed transcripts, but at the plasma membrane, suggesting that this mutant receptor these were hardly, if at all, detected in lymphoid cell lines (Jurkat does not interact with PPL. and Raji). The myeloid͞lymphoid IIA1.6 cells (24) expressed some endogenous PPL (Fig. 2A; see also Fig. 6, which is Fc␥RI and PPL Coimmunoprecipitate in Immune Cells and Interact published as supporting information on the PNAS web site). Directly. The B cell͞macrophage cell line IIA1.6 (24) is devoid of Increased levels of Fc␥RI and PPL were found upon overnight endogenous FcRs (34) and has been used previously in func- IFN-␥ treatment (Fig. 2 A and B). An Sp1 site is present in the tional assays for Fc␥RI (10). We performed transient transfec- promoter region of PPL and could regulate its myeloid expres- tions with the C-terminal part of PPL in IIA1.6 cells stably sion, because Sp1 is involved in expression of myeloid markers expressing Fc␥RI, with or without FcR ␥-chains (Fig. 3A). such as CD11c (32, 33). Immunoprecipitations were performed with Abs to PPL, and

Beekman et al. PNAS ͉ July 13, 2004 ͉ vol. 101 ͉ no. 28 ͉ 10395 Downloaded by guest on September 30, 2021 Fig. 5. Effect of PPL on Fc␥RI-mediated EA rosetting. (A)Fc␥RI expression Fig. 4. Effect of PPL on Fc␥RI modulation and antigen presentation. (A) levels measured by flow cytometry with F(abЈ) of anti-Fc␥RI mAb H22 and Modulation of Fc␥RI expression in subcloned IIA1.6 transfectants stably ex- 2 goat F(abЈ) anti-human ␬ light chain-FITC or F(abЈ) of an Ab isotype control pressing Fc␥RI, FcR ␥-chain (WT or immunoreceptor tyrosine-based activation 2 2 (dotted line). The thin black line, thick black line, and shaded histogram plot motif-mutated), and PPL clone 3.4 by rabbit IgG–OVA immune complexes represent Fc␥RI expression of IIA1.6 cells transfected with Fc␥RI and FcR (OVA–IgG) after overnight incubation. Fc␥RI expression without OVA–IgG was ␥-chain; Fc␥RI and ␥Y65F,Y76F; or Fc␥RI, ␥Y65F,Y76F, and C-terminal PPL clone set at 100%, and reduction of cell-surface Fc␥RI expression levels by increasing 3.4, respectively. (B) EA rosettes are shown of IIA1.6 transfectants incubated OVA–IgG concentrations was measured by flow cytometry and plotted as with erythrocytes (RBC) sensitized with 0.5 ␮g͞ml mIgG2a anti-glycophorin A percentage modulation. (B) Cells were treated with OVA–IgG and coincu- mAb. (C) Ab dose-dependent binding of RBC to IIA1.6 cells. Percentages of bated with OVA-specific DO11 T cells for 16 h. IL-2 production of DO11 cells cells interacting with at least one (Left) or three (Right) RBC were scored. was measured by [3H]thymidine incorporation of IL-2 responsive cells (CTLL-2) indicative of antigen presentation. [3H]Thymidine incorporation after treat- ␮ ment with 100 g of OVA was set at 100% for each transfectant. Two ␥ independent stable transfectants for Fc␥RI and ␥Y65F,Y76F (designated 1 and by plasma membrane reduction of Fc RI after 16 h of incubation 2) were compared with two stable transfectants for Fc␥RI, ␥Y65F,Y76F, and with immune complexes and is indicative of lysosomal trafficking C-terminal PPL clone 3.4. Untransfected IIA1.6 cells were used as control. of Fc␥RI͞antigen complexes (28). Typically, Ϸ20–30% of the Measurements were in triplicate, error bars mark standard deviations, and receptor disappeared after overnight incubation with 1 ␮g͞ml one representative experiment of three is shown. Groups were compared by immune complexes, independent of active or signaling ‘‘dead’’ Ͻ Student’s paired t test, which indicated statistical difference (P 0.05). (␥Y65F, Y76F) FcR ␥-chains (Fig. 4A). Stable expression of C-terminal PPL in Fc␥RI transfectants (confirmed by Western blot; data not shown) increased levels of internalized receptors Western blots of these were stained for PPL and Fc␥RI. PPL was to Ϸ50–60% (P Ͻ 0.05, paired Student’s t test; n ϭ 4). pulled down by specific rabbit serum and coprecipitated Fc␥RI In this transfection model, Fc␥RI-CY mediates antigen pre- independent of the FcR ␥-chain in these cells. sentation in the presence of ␥Y65F,Y76F, although antigen These data extended our yeast two-hybrid results but did not presentation is suboptimal (Ϸ60% left) when compared with the discriminate between direct and indirect types of interaction. WT FcR ␥-chain (10). We confirmed the capacity of Fc␥RI to Therefore, we performed immunoprecipitations and blot- mediate antigen presentation by comparing Fc␥RI-dependent overlay assays with purified recombinant proteins. Anti-GST antigen presentation to MHC class II presentation after pino- beads coprecipitated His-PPL clone 3.4 when incubated with cytosis of a large excess of free OVA (set at 100%; Fig. 4B). GST-Fc␥RI but not with GST tag alone (Fig. 3B). Similarly, ␥ Stable coexpression of C-terminal PPL in these transfectants anti-His beads that pulled down PPL captured GST-Fc RI but increased antigen presentation (P Ͻ 0.05, paired Student’s t test; not GST (data not shown). In blot-overlay assays (Fig. 3C), n ϭ 3), as was shown for Fc␥RI modulation. Untransfected cells ␥ GST-Fc RI or GST tag alone were incubated with membranes exhibited no response to these concentrations of OVA-immune containing His-PPL. GST signals were detected at the correct complexes, implicating Fc␥RI dependency. Together, these re- height of PPL when membranes were incubated with GST- sults suggested that PPL regulates downstream effector func- ␥ Fc RI but not with GST tag alone. This assay was not successful tions of Fc␥RI-CY. vice versa, suggesting that a linear Fc␥RI binding motif is present in PPL but that the PPL binding site of Fc␥RI depends on C-Terminal PPL Enhances Fc␥RI Ligand Binding. Because PPL seemed secondary and͞or tertiary structures. to interact with Fc␥RI without receptor crosslinking (Figs. 2C and 3A), we tested the ability of C-terminal PPL to affect Fc␥RI C-Terminal PPL Enhances Downstream Effector Functions of Fc␥RI-CY. ligand binding. Ligand binding was increased by C-terminal PPL, Upon C-terminal PPL transfection, no appreciable differences in as illustrated by enhanced EA rosette scores (Fig. 5), although cell-surface expression were observed in flow cytometric and Fc␥RI expression levels were similar. These experiments were microscopic analyses (data not shown). Next, we tested the performed at 4°C to inhibit cellular activity and thus indicated capacity of C-terminal PPL to affect Fc␥RI function using that PPL controls Fc␥RI function by mechanisms that were subcloned IIA1.6 transfectants. Receptor modulation is defined preexperimentally active in nonstimulated cells. To accredit

10396 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0401217101 Beekman et al. Downloaded by guest on September 30, 2021 temperature sensitivity of protein interactions, similar experi- cellular activation, indicating that lower affinity might be ben- ments were performed at a physiological temperature (37°C) eficial for immune complex triggering (8). Because cells with yielding identical results, also in the presence of okadaic acid, intact Fc␥RI–PPL interaction exhibit lower capacity to form EA which prevents Fc␥RI-mediated phagocytosis (35) (see Fig. 7, rosettes, PPL expression might also skew interactions of immune which is published as supporting information on the PNAS web cells to more intensely opsonized antigens. site). The effect of C-terminal PPL is likely by functional blockade of Concluding Remarks. Signal transduction by multichain FcRs has ␥ Fc RI–endogenous PPL interaction in IIA1.6 cells. Although we been considered to be exclusively mediated by immunoreceptor could not detect endogenous PPL by Western blot, flow cytometric tyrosine-based activation motif-containing subunits. However, analysis indicated PPL protein in IIA1.6 (Fig. 6). Similarly, trans- we found PPL to selectively modulate Fc␥RI function through duction of cells with PPL peptides containing a minimal Fc␥RI ␣ ␥ its intracellular -chain. Enhanced ligand binding through in- binding domain enhanced Fc RI ligand binding comparable to the side-out signaling has been described for Fc␣RI (38). Here, we effect of C-terminal PPL (36). Thus, endogenous mPPL (clone 3.4 ␥ ␥ document a similar mechanism for Fc RI and the protein that shared 94% similarity with its murine counterpart) bound Fc RI, is involved in this inside-out mechanism. and this was blocked by C-terminal PPL. This finding would imply that endogenous PPL lowers ligand binding. Miller et al. (37) found ␥ We thank J. van de Linden for technical assistance, Dr. L. Meyaard for that Fc RI-CY removal increases ligand affinity, concordant with critically reading the manuscript, the Sanquin Bloedbank Midden Ned- our hypothesis. erland for preparations of buffy coats, Dr. B. Nagelkerken (University In vivo, regulation of PPL function might lower ligand affinity Medical Center Utrecht) for cDNA constructs, Dr. B. Burgering and Dr. to facilitate replacement of bound monomeric IgG for antigen- R. Kimberly for rabbit sera, and Dr. F. M. Watt for the fl-PPL construct. Ϫ Ϫ complexed IgG. Fc␥RI / mice illustrated that serum IgG This work was supported by Medarex Europe (J.M.B. and J.E.B.) and the functionally competes with ligand complexes for induction of Dutch Science Foundation (NWO Grant 901-07-229 to J.H.W.L.).

1. Daeron, M. (1997) Annu. Rev. Immunol. 15, 203–234. 22. van Vugt, M. J., Reefman, E., Zeelenberg, I., Boonen, G., Leusen, J. H. W. & 2. Ravetch, J. V. & Bolland, S. (2001) Annu. Rev. Immunol. 19, 275–290. van de Winkel, J. G. J. (1999) Eur. J. Immunol. 29, 143–149. 3. Durden, D. L., Liu, Y. B., Strzelecka, A., Kwiatkowska, K. & Sobota, A. (1994) 23. Wallace, P. K., Keler, T., Coleman, K., Fisher, J., Vitale, L., Graziano, R. F., Blood 84, 2102–2108. Guyre, P. M. & Fanger, M. W. (1997) J. Leukocyte Biol. 62, 469–479. 4. Zheng, L., Nibbering, P. H., Zomerdijk, T. P., van Furth, R., Durden, D. L., Liu, 24. Van Vugt, M. J., Van den Herik-Oudijk, I. E. & Van de Winkel, J. G. J. (1998) Y. B., Strzelecka, A., Kwiatkowska, K. & Sobota, A. (1994) Infect. Immun. 62, Clin. Exp. Immunol. 113, 415–422. 4296–4303. 25. Carr, D. W. & Scott, J. D. (1992) Trends Biochem. Sci. 17, 246–249. 5. Strzelecka, A., Kwiatkowska, K. & Sobota, A. (1997) FEBS Lett. 400, 11–14. 26. Leusen, J. H. W., Fluiter, K., Hilarius, P. M., Roos, D., Verhoeven, A. J. & 6. Van de Winkel, J. G. J. & Anderson, C. L. (1991) J. Leukocyte Biol. 49, 511–524. Bolscher, B. G. (1995) J. Biol. Chem. 270, 11216–11221. 7. Van de Winkel, J. G. J. & Capel, P. J. (1993) Immunol. Today 14, 215–221. 27. Guyre, P. M., Graziano, R. F., Vance, B. A., Morganelli, P. M. & Fanger, M. W. 8. Ioan-Facsinay, A., de Kimpe, S. J., Hellwig, S. M., van Lent, P. L., Hofhuis, 143, F. M., van Ojik, H. H., Sedlik, C., da Silveira, S. A., Gerber, J., de Jong, Y. F., (1989) J. Immunol. 1650–1655. et al. (2002) Immunity 16, 391–402. 28. Wallace, P. K., Keler, T., Guyre, P. M. & Fanger, M. W. (1997) Cancer 9. Barnes, N., Gavin, A. L., Tan, P. S., Mottram, P., Koentgen, F. & Hogarth, P. M. Immunol. Immunother. 45, 137–141. (2002) Immunity 16, 379–389. 29. Van den Herik-Oudijk, I. E., Ter Bekke, M. W., Tempelman, M. J., Capel, P. J. 10. van Vugt, M. J., Kleijmeer, M. J., Keler, T., Zeelenberg, I., van Dijk, M. A., & Van de Winkel, J. G. J. (1995) Blood 86, 3302–3307. Leusen, J. H. W., Geuze, H. J. & van de Winkel, J. G. J. (1999) Blood 94, 30. Boot, J. H., Geerts, M. E. & Aarden, L. A. (1989) J. Immunol. 142, 1217–1223. 808–817. 31. Heijnen, I. A., van Vugt, M. J., Fanger, N. A., Graziano, R. F., de Wit, T. P., 11. Edberg, J. C., Yee, A. M., Rakshit, D. S., Chang, D. J., Gokhale, J. A., Indik, Hofhuis, F. M., Guyre, P. M., Capel, P. J., Verbeek, J. S. & van de Winkel, Z. K., Schreiber, A. D. & Kimberly, R. P. (1999) J. Biol. Chem. 274, J. G. J. (1996) J. Clin. Invest. 97, 331–338. 30328–30333. 32. Aho, S., Rothenberger, K., Tan, E. M., Ryoo, Y. W., Cho, B. H., McLean, W. H. 12. Ohta, Y., Stossel, T. P. & Hartwig, J. H. (1991) Cell 67, 275–282. & Uitto, J. (1999) Genomics 56, 160–168. 12, 13. Leung, C. L., Green, K. J. & Liem, R. K. (2002) Trends Cell Biol. 37–45. 33. Noti, J. D., Reinemann, B. C. & Petrus, M. N. (1996) Mol. Cell. Biol. 16, IMMUNOLOGY 14. Ruhrberg, C., Hajibagheri, M. A., Parry, D. A. & Watt, F. M. (1997) J. Cell Biol. 2940–2950. 139, 1835–1849. 34. Jones, B., Tite, J. P. & Janeway, C. A., Jr. (1986) J. Immunol. 136, 348–356. 15. Aho, S., McLean, W. H., Li, K. & Uitto, J. (1998) Genomics 48, 242–247. 35. Edberg, J. C., Qin, H., Gibson, A. W., Yee, A. M., Redecha, P. B., Indik, Z. K., 16. Steinert, P. M. & Marekov, L. N. (1999) Mol. Biol. Cell 10, 4247–4261. Schreiber, A. D. & Kimberly, R. P. (2002) J. Biol. Chem. 277, 41287–41293. 17. DiColandrea, T., Karashima, T., Maatta, A. & Watt, F. M. (2000) J. Cell Biol. 36. Beekman, J. M., Bakema, J. E., van der Linden, J., Tops, B., Hinten, M., van 151, 573–586. 18. Karashima, T. & Watt, F. M. (2002) J. Cell Sci. 115, 5027–5037. Vugt, M., van de Winkel, J. G. J. & Leusen, J. H. W. (2004) J. Biol. Chem., 19. Kazerounian, S., Uitto, J. & Aho, S. (2002) Exp. Dermatol. 11, 428–438. in press. 20. Van den Heuvel, A. P., de Vries-Smits, A. M., van Weeren, P. C., Dijkers, P. F., 37. Miller, K. L., Duchemin, A. M. & Anderson, C. L. (1996) J. Exp. Med. 183, de Bruyn, K. M., Riedl, J. A. & Burgering, B. M. (2002) J. Cell Sci. 115, 2227–2233. 3957–3966. 38. Bracke, M., Lammers, J. W., Coffer, P. J. & Koenderman, L. (2001) Blood 97, 21. Feng, G. J., Kellett, E., Scorer, C. A., Wilde, J., White, J. H. & Milligan, G. 3478–3483. (2003) J. Biol. Chem. 278, 33400–33407. 39. Heijnen, I. A. & Van de Winkel, J. G. J. (1995) J. Hematother. 4, 351–356.

Beekman et al. PNAS ͉ July 13, 2004 ͉ vol. 101 ͉ no. 28 ͉ 10397 Downloaded by guest on September 30, 2021