Modulation of Fc (cid:1) RI (CD64) Ligand Binding by Blocking Peptides of Periplakin*

Fc (cid:1) RI requires both the intracellular domain of the (cid:2) -chain and associated leukocyte Fc receptor (FcR) (cid:1) -chains for its biological function. We recently found the C terminus of periplakin to selectively interact with the cytoplasmic domain of the Fc (cid:1) RI (cid:2) -chain. It thereby enhances the capacity of Fc (cid:1) RI to bind, internalize, and present antigens on MHC class II. Here, we characterized the domains involved in Fc (cid:1) RI-periplakin interaction using truncated and alanine-substituted Fc (cid:1) RI mutants and randomly mutagenized periplakin. This allowed us to design TAT peptides that selectively inter-fered with endogenous Fc (cid:1) RI-periplakin interactions. The addition of these peptides to Fc (cid:1) RI-expressing cells modulated Fc (cid:1) RI ligand binding, as assessed by eryth-rocyte-antibody-rosetting. These data support a domi-nant-negative role of C-terminal periplakin for the molecular between determined

Fc␥RI (CD64, Fc␥RIa1) is unique among multisubunit FcR due to a high affinity binding to human IgG, its limited myeloid cell distribution, and a relatively large intracellular domain (7,8). Products of related genes include Fc␥RIb and Fc␥RIc isoforms, but these specify low affinity IgG receptors if functionally expressed at all (9 -12). Besides a role in antigen clearance, Fc␥RI (a1) can potently enhance MHC class I and II antigen presentation in vitro and in vivo (13)(14)(15)(16). These properties make Fc␥RI a candidate target for immunotherapy, and concepts are being developed to modulate immune responses by Fc␥RI-directed agents (16 -19). The potential of such therapeutic approaches supports further work to enlarge our knowledge of Fc␥RI biology.
The FcR ␥-chain has been studied in great detail and is critically important for Fc␥RI function; it stabilizes Fc␥RI ␣-chain surface expression in vivo (20) and mediates several key functions that require ITAM signaling motifs (21)(22)(23)(24). In addition, recent data show that the cytosolic domain of the Fc␥RI ␣-chain (Fc␥RI-CY) could transduce signals leading to cellular effector functions (25,26). MHC class II antigen presentation assays using IIA1.6 cells co-expressing truncated Fc␥RI-CY mutants and "signaling-dead" FcR ␥-chains indicated a motif for antigen presentation in the membrane proximal ϳ34 aa of Fc␥RI-CY (25). Deletion of Fc␥RI-CY in the presence of functional FcR ␥-chain lowered the kinetics of endocytosis and phagocytosis and abolished interleukin-6 production (26). Fc␥RI-CY signaling likely involves (de)phosphorylation of its serine residues and may include other mechanisms of post-translational modification (27). Thus far no protein effectors have been described that control Fc␥RI function by Fc␥RI-CY interaction. Filamin A (ABP-280) has been shown to bind Fc␥RI-CY, but no functional consequences are known for this interaction (28).
We recently found periplakin to selectively bind Fc␥RI-CY and to modulate ligand binding, receptor modulation, and antigen presentation via Fc␥RI. 2 Periplakin represents a 195-kDa protein implicated in cornified envelope assembly and structural stability of epithelia (29 -32). Like other members of the plakin family (for review, see Refs. 33 and 34) periplakin associates with the actin and intermediate filament cytoskeleton (32,35,36). Periplakin has recently been suggested to be involved in signaling of protein kinase B (37) and G-proteins located downstream of the -opioid receptor in neurons (38).
In the present study we characterized the molecular interaction between Fc␥RI and periplakin. We determined the periplakin binding domain of Fc␥RI and vice versa by progressive truncations and alanine-scanning mutagenesis of Fc␥RI-CY and random mutagenesis of periplakin. Peptides of these binding domains and the membrane-translocating TAT sequence (39) were designed to interfere with Fc␥RI-periplakin interactions in IIA1.6 cells. Fc␥RI ligand binding was studied in EA-rosette assays. tants (see Fig. 1) were generated by PCR and EcoRI/SalI-cloned into pGBT9 (Clontech, Palo Alto, CA). Alanine replacement of Fc␥RI residues 311-325 was achieved by PCR-based cloning techniques (primer mutations and overlap extension PCR) and EcoRI/SalI insertion into pGBT9. pGAD-GH (Clontech) contained C-terminal periplakin (Gen-Bank TM accession number AF001691) clone 2.2 (bp 4620 -5361) or 3.4 (bp 4207-5361). Glycine replacement of aspartic acid 1694 (D1694G) of periplakin clone 2.2 was achieved by substitution of periplakin bp Ala-5081 for Gly by PCR using an adjacent unique SpeI site of periplakin and a mutated primer. Mammalian expression constructs WT-Fc␥RI and truncated mutants were cloned with HindIII/XbaI restriction into pcDNA3 (Invitrogen) as described in Van Vugt et al. (25). Periplakin clones 2.2, 3.4, and the mutated 2.2 (D1694G) were subcloned into pcDNA3.1 HISABC (Invitrogen). The murine FcR ␥-chain with mutated ITAM (Y65F,Y76F) was expressed from pNUT (40,41). PCR reagents were from PerkinElmer Life Sciences (Nieuwerkerk a/d IJssel, The Netherlands) except for primers (Isogen Bioscience, Maarssen, the Netherlands). All construct were verified by dideoxy sequencing using BigDye Terminators (Applied Biosystems, Warrington, UK) and analyzed on an ABI Prism® 3100 Genetic Analyzer (Applied Biosystems).
Random Mutagenesis of C-terminal Periplakin-pGAD-GH clone 2.2 served as a PCR template in the presence of dITP and limiting amounts of dATP or dATP and dTTP to generate a pool of randomly mutated periplakin PCR products. This method was adapted from a protocol described by Spee et al. (43). Specific PCR characteristics were: 10 ng of pGAD-GH clone 2. colonies were lifted and tested for loss of interaction by the absence of ␤-galactosidase activity. Plasmids were prepared from ␤-galactosidasenegative colonies and sequenced. Sequences were aligned using BioEdit software (www.mbio.ncsu.edu/BioEdit/bioedit.html).
EA-rosetting-Human erythrocytes were prepared by Ficoll/ Hypaque density centrifugation, stored in sterile Alsever at 4°C, and used within 2 weeks. Erythrocytes were fluorescently labeled using the PKH26 fluorescent cell linker kit (Sigma) according to the manufacturer's protocol and opsonized by hybridoma supernatant containing mIgG2a anti-human glycophorin A for 1 h at 2 ϫ 10 8 erythrocytes/ml at 4°C (44). Erythrocytes were washed twice with Hepes-buffered RPMI 1640 medium (Invitrogen) at 4°C. Subsequently, 5 ϫ 10 6 erythrocytes were resuspended with 1 ϫ 10 5 cells in 50 l of RPMI in round-bottom 96-well plates, incubated for 60 min at 4°C, and resuspended after a 30-min incubation. EA-rosettes were fixed by the addition of 3% paraformaldehyde for 30 min. Cells were diluted 2-3-fold in Hepesbuffered RPMI 1640 medium and analyzed by flow cytometry. Cells and free erythrocytes were distinguished by their scatter patterns and autofluorescence in the FL1 channel. The percentage of cells that were FL-2-positive was expressed as percentage of EA-rosettes. Fc␥RI surface expression was measured on a FACScalibur flow cytometry system (BD Biosciences) using the F(abЈ) 2 fragment of CD64 mAb H22 (Ref. 45; a kind gift of Dr. T. Keler, Medarex, Annandale, NJ) and goat F(abЈ) 2 anti-human k-light chain-FITC (Southern Biotech, Birmingham, AL).
TAT Peptides-Fusion peptides of the protein transduction domain of TAT (39,46) and the binding domains of Fc␥RI for periplakin (TAT-Fc␥RI) or vice versa (TAT-PPL) were from Eurogentec (Herstal, Belgium). The sequence of TAT-Fc␥RI is YGRKKRRQRRRGVTIR-KELKRKKKWDLEI (29-mer), and that of TAT-PPL is YGRKKRRQR-RRGKLRSQECDWEEISVK (27-mer). Peptides were Ͼ95% pure and had standard N and C termini. Control TAT-peptide (YGRKKRRQR-RRG) consisted of the TAT sequence above and was a kind gift of Dr. P. Coffer (Dept. of Pulmonary Diseases, University Medical Center Utrecht). For transduction, cells were washed twice in Hepes-buffered RPMI 1640 medium containing 5 mM EDTA (EDTA buffer) to remove free extracellular calcium. Cells were incubated for 30 min in 10 M TAT peptide at 37°C in EDTA buffer at 5 ϫ 10 6 cells/ml. Cells were diluted in ice-cold EDTA buffer, washed at 4°C in buffer without EDTA, and used in EA-rosette assays. Aliquots were analyzed by flow cytometry for Fc␥RI surface expression and cell viability by annexin V and propidium iodide staining (Roche Applied Science). Periplakin expression was assessed by Western blot using periplakin-recognizing rabbit serum 5117 after transient expression of full-length periplakin (construct kindly provided by Dr. F. Watt, Keratinocyte Laboratory, London Research Institute). For control experiments, TAT peptides were FITClabeled (Molecular Probes, Leiden, The Netherlands), pulsed as described above, and analyzed by confocal microscopy. Surface IgG was stained by an anti-mouse IgG mAb CY-3 conjugated (Jackson).

Identification of the Periplakin Interaction Domain within
Fc␥RI-To pinpoint the binding domain of Fc␥RI for periplakin, we tested progressive truncations of Fc␥RI-CY for interaction with C-terminal periplakin in yeast two-hybrid binding assays (Fig. 1). The minimal binding domain consisted of the membrane-proximal 17 amino acids of Fc␥RI-CY, as demonstrated by growth of yeast colonies on histidine-depleted media after transformation with Fc␥RI-⌬327 and periplakin (residues 1372-1756). However, in ␤-galactosidase assays a slightly larger domain of Fc␥RI (22 residues, Fc␥RI-⌬332) was required for periplakin interaction in yeast cells. Removal of the N-terminal valine (residue 311) or more (312 ϩ 313) completely abrogated periplakin interaction. However, residues of the GAL4 DNA binding domain directly upstream of the N terminus of Fc␥RI-CY did not contribute to the interaction, as a stretch of six glycines in between the GAL4 DNA binding domain and Fc␥RI-CY left binding to periplakin intact (data not shown).
We next mapped the binding domain of Fc␥RI for periplakin in IIA1.6 transfectants. Stable transfectants of WT-Fc␥RI, Fc␥RI-⌬342, Fc␥RI-⌬332, and Fc␥RI-⌬315 were co-transfected with C-terminal periplakin clone 3.4 and assessed after 48 h without further stimulation. WT-Fc␥RI and Fc␥RI-⌬342 colocalized with c-terminal periplakin, showing both proteins to FIG. 2. Alanine-scanning mutagenesis of Fc␥RI and interaction with periplakin. A, site-directed mutagenesis was applied to Fc␥RI residues 311-325, and mutants were tested for interaction with periplakin clone 3.4 in yeast cells (shown in triplicate). Growth of yeast cells on medium lacking leucine and tryptophan (-LT), leucine, tryptophan, histidine (-LTH), and ␤-galactosidase activity are shown. One representative experiment is shown (n ϭ 2). B, sequence alignment of Fc␥RI from five different species. Both human and mouse Fc␥RI are truncated at the C terminus, where no significant alignment was observed. Alanine substitutions that abrogate or enhance the interaction are indicated by Ϫ or ϩ, respectively. Green residues indicate hydrophobicity/aromaticity; blue indicates hydrophilicity/basicity; red indicates acidity; yellow indicates aliphaticity. be present at similar sites in cells. Although periplakin localized to the (sub)plasma membrane area, co-localization of Fc␥RI-⌬332 and tail-less Fc␥RI (Fc␥RI-⌬315) with C-terminal periplakin was abrogated, suggesting a loss of interaction.
To assess the relative contribution of residues within Fc␥RI-CY for periplakin interaction, alanine-scanning mutagenesis was applied to Fc␥RI-CY residues 311-325, and proteins were assessed for interaction with periplakin in yeast cells. Notably, substitutions were found that either abrogated or improved the interaction between Fc␥RI and periplakin ( Fig.  2A). Alanine replacement of plasma membrane-proximal residues abrogated the interaction, except for Thr-312, whereas single substitutions of the stretch of positively charged residues (KRKKK) and Asp-324 apparently led to a better interaction, except for the last lysine. The increase in ␤-galactosidase activity may reflect a better-stabilized interaction with periplakin. By alanine substitution of the proximal part of Fc␥RI-CY, we targeted a sequence that is largely conserved from mouse to man (Fig. 2B). However, we did not detect any interaction between murine Fc␥RI-CY and human periplakin. The absence of Glu-316 and Trp-323 in the mouse sequence might contribute to the differences between these species.
Together, these results pointed to a periplakin binding domain of Fc␥RI-CY in juxtaposition to the plasma membrane. Periplakin binding to Fc␥RI-CY in yeast required a minimal motif of 17 residues within Fc␥RI-CY, but significant co-localization was only observed when 32 residues of Fc␥RI-CY were present. This discrepancy was observed consistently and indicated that both systems differ in Fc␥RI-periplakin binding requirements. For further studies with blocking peptides, we utilized the membrane-proximal 17 residues of Fc␥RI as a blocking domain for Fc␥RI-periplakin interaction.
Identification of the Interaction Domain of Periplakin for Fc␥RI-Random mutagenesis of periplakin (schematically shown in Fig. 3A) was chosen as a tool to define the periplakin domain that interacts with Fc␥RI. We prepared a library of mutated clone 2.2 cDNAs by PCR amplification in the presence of dITP and screened for loss of interaction with Fc␥RI in yeast cells (Table I). A 2-fold increase of colonies that acquired a loss-of-interaction phenotype was achieved by the addition of dITP in the PCR. Fc␥RI-periplakin interaction was lost in 43% (68/158), and 57.3% (55/96) of yeast colonies by lowering dATP and dATP/dTTP concentrations, respectively. By (partially) sequencing plasmids from 41 colonies, a total of 65 missense mutations and 8 premature stop codons were identified. The most C-terminal-located stop codon was introduced after residue 1688, indicating that residues 1510 -1688 of periplakin do not contain a full binding site for Fc␥RI. Only one of the missense mutants had a single amino acid substitution (aspartic acid at position 1694 into glycine (D1694G)) that resulted in a disturbed interaction (Fig. 3B). We confirmed this observation by site-directed mutagenesis of D1694G in the original periplakin 2.2 construct. Interactions were mitigated both in yeast cells (data not shown, n ϭ 2) and transfected IIA1.6 cells (Fig. 3C). The D1694G-mutated clone 2.2 staining was more dominant in the cytoplasm and sometimes followed a filamentous pattern.
Six substitutions were located within the 15 residues adjacent to Asp-1694 (1687-KLRSQECDWEEISVK-1701). At four positions residue charge was affected, and two times a proline was inserted (K1687E,K1687D, L1688P, D1694G, E1696G, S1699P, K1701E; Fig. 3B). When this domain was aligned to the proximal part of Fc␥RI-CY (in opposite direction), several parts exhibited clear opposing electrostatic charges that may well contribute to the interaction between both proteins (Fig.  3B). However, it is hard to assign specific periplakin residues besides Asp-1694 of being located at the binding interface, as the tertiary structure of these domains is unknown. These data supported periplakin 1687-KLRSQECDWEEISVK-1701 to be part of the Fc␥RI binding domain of periplakin, and we hypothesized that a peptide with this sequence may block such interaction.
Flow Cytometric Analysis of EA-rosetting-Our previous work documented the ligand binding capacity of Fc␥RI to be increased by C-terminal periplakin transfection. Here, we developed a quick assay to measure Fc␥RI-ligand binding via the use of EA-rosetting by flow cytometry (i.e. the percentage of cells that bound mIgG2a-sensitized erythrocytes, upper right panel in Fig. 4B). Non-bound cells and FL2-labeled erythrocytes were found in the lower right and upper left quadrants, respectively. We observed 5% background binding to untransfected IIA1.6 cells or unsensitized erythrocytes (data not shown). A 2.5-fold increase in binding of mIgG2a-sensitized erythrocytes to Fc␥RI was observed upon co-expression of Cterminal periplakin in IIA1.6 cells (Fig. 4C). Binding appeared independent of small differences in surface levels of Fc␥RI (Fig.  4A) and FcR ␥-chain signals. Similar differences were observed when EA-rosetting was scored by light microscopy (data not shown, n ϭ 4).
Modulation of EA-rosetting by TAT-Periplakin-Peptides containing a TAT motif and the postulated binding domains of Fc␥RI and periplakin were designed to block intracellular Fc␥RI-periplakin interaction. Fc␥RI surface expression and periplakin levels were unaffected by the addition of TAT pep- tides (Fig. 5, A and B; n ϭ 3). Similarly, cell viability assessed by annexin V and propidium iodide-staining remained intact (data not shown, n ϭ 3). To visualize cell transduction with the TAT peptides, FITC-conjugated TAT-Fc␥RI 311-327 (TAT-Fc␥RI), TAT-periplakin 1687-1701 (TAT-PPL), and non-TAT peptides were incubated with IIA1.6 cells (n ϭ 3). Intracellular accumulation of FITC was observed only with the TAT peptides, consistent with their intracellular delivery (Fig. 5C).
Subsequently, transfectants were pulsed with TAT peptides and incubated with mIgG2a-sensitized erythrocytes to assess the modulating capacity of these peptides on Fc␥RI-ligand interaction. TAT-PPL increased the interaction of IIA1.6 cells with mIgG2a-sensitized erythrocytes to levels observed in Cterminal periplakin transfection experiments (Fig. 5D, n ϭ 3). TAT-Fc␥RI slightly increased the capacity of Fc␥RI to form EA-rosettes, but differences were not significant, possibly due to an inability of this peptide to transduce cells as efficiently as TAT-PPL or to adapt to an appropriate tertiary structure. Similarly, a peptide consisting of only the TAT sequence did not affect EA-rosetting ability. When C-terminal periplakin was co-expressed with Fc␥RI none of the peptides modulated EArosetting. These data support that the interaction between Fc␥RI and endogenous periplakin was effectively blocked by C-terminal periplakin.

DISCUSSION
In this report we studied the molecular interaction between Fc␥RI and periplakin. Minimal binding domains were defined and generated as TAT peptides to disrupt intracellular Fc␥RIperiplakin interactions. TAT-PPL transduction enhanced the capacity of Fc␥RI to form EA-rosettes in transfected IIA1.6 cells without affecting receptor expression levels (Fig. 5). Because this effect mimicked stable transfection of C-terminal periplakin, it is likely that C-terminal periplakin and TAT-PPL regulate Fc␥RI by preventing Fc␥RI-CY binding to endogenous periplakin (Fig. 6). This suggests that endogenous periplakin somehow decreases Fc␥RI-ligand binding.
We showed that the proximal part of Fc␥RI-CY binds periplakin. Receptor truncation experiments in yeast cells indicated Fc␥RI C-terminal residues 333-374 to be fully dispen- Two screens were performed to define the Fc␥RI binding site within periplakin. Random PCR mutagenesis was achieved by including an artificial fifth nucleotide dITP in PCR in combination with lowered amounts (indicated by the down arrow) of dATP (screen 1) or dATP and dTTP (screen 2). "Control PCR" represents co-transfections of pGBT9 WT Fc␥RI-CY and linearized pGAD-GH with periplakin inserts generated by PCR without dITP. "Mutagenesis PCR" represents co-transfections of pGBT9 WT Fc␥RI-CY and linearized pGAD-GH with periplakin inserts generated by PCR with dITP. Percentages and absolute numbers (between parentheses) of colonies that lost interaction between Fc␥RI and periplakin are indicated. The locations of introduced stop codons (amino acid position) by random mutagenesis PCR are indicated in the last column.  2 anti-human light chain FITC by flow cytometry. Black lines represent Fc␥RI staining, and dotted lines represent isotype control. B, erythrocytes (RBC) were FL2-labeled and sensitized with mIgG2a ␣-glycophorin A mAb. Subsequently, Fc␥RI transfectants and sensitized erythrocytes were incubated, and EA-rosetting was assessed by flow cytometry as the percentage of cells bound to mIgG2aerythrocytes. Lower right quadrants contain non-bound transfectants (that differ in size and FL-1 autofluorescence from erythrocytes), upper left quadrants contain non-bound erythrocytes, upper right quadrants contain transfectants bound to erythrocytes. C, analysis of EA-rosetting of transfectants using erythrocytes sensitized with two concentrations of mIgG2a ␣-glycophorin A mAb. One representative experiment out-of-four is shown; error bars indicate S.D.
sable for the interaction of Fc␥RI-CY with periplakin. Fc␥RI-⌬327, albeit less efficient than WT-Fc␥RI, bound periplakin, as indicated by growth on histidine-depleted media showing the 17 membrane-proximal residues of Fc␥RI to be minimally required for interaction. In transfected cells, Fc␥RI-CY-depend-ent co-localization with periplakin was observed. Notably, however, a larger motif of Fc␥RI (Fc␥RI-⌬342) was required than in the yeast system. The discrepancy between yeast and mammalian cells for the minimal requirements of interaction between Fc␥RI and periplakin might be explained by differences in protein folding in the two systems or associated molecules like the FcR ␥-chain or filamin A in IIA1.6 cells. This might account for the non-functional TAT-Fc␥RI peptide as well. The presence of the WT or ITAM-mutated FcR ␥-chain did not influence co-localization under the conditions tested.
Fc␥RI-CY harbors an antigen presentation motif present in Fc␥RI-⌬342 and not in Fc␥RI-⌬332 (25). In IIA1.6 cells, we observed co-localization of Fc␥RI-⌬342, but not Fc␥RI-⌬332, with periplakin, consistent with a possible regulatory role for periplakin in MHC class II antigen presentation. We recently found that IFN-␥, a cytokine regulating proteins important for MHC class I and II antigen presentation (for review, see Ref. 47), up-regulated both periplakin and Fc␥RI expression in monocytes and PMN. Recently, also PMN were shown to functionally express MHC class II (48). If periplakin is involved in antigen presentation, its role may be to "fine-tune" responses, as tailless Fc␥RI was shown capable of mediating antigen presentation, although less efficiently when a functional FcR ␥-chain was present (25).
Alanine substitution of individual Fc␥RI-CY residues 311-325 that affected binding to periplakin were found largely conserved from mouse to man (Fig. 2). Residues located directly adjacent to the plasma membrane and more downstream (Lys-322, Trp-323, and Leu-325) abrogated Fc␥RI binding to periplakin. Remarkably, most single substitutions of the large positive KRKKK stretch and Asp-324 seemed to facilitate binding between Fc␥RI and periplakin. Overall sequence similarity of Fc␥RI with mouse Fc␥RI or other activable human FcR receptors is low (maximally 20%), and no obvious overlapping domains are present. These could not interact with periplakin in yeast two-hybrid studies, 2 although the membrane ϭproximal region of mouse Fc␥RI shares significant similarity with human Fc␥RI.
In mice, seven allelic variants of Fc␥RI have been described, from which three have altered amino acids in the proximal part of the intracellular domain (49). Although no functional polymorphisms have been assigned to the cytosolic tail of human Fc␥RI, amino acid substitutions that influence periplakin interaction might have an effect on Fc␥RI function in vivo. Notably, the human Fc␥RI b and c isoforms have identical cytosolic tails to Fc␥RIa1 but contain asparagine at position 324 instead of aspartic acid (11). This cytosolic tail variant exhibited increased interaction with periplakin in yeast two-hybrid binding studies (data not shown). However, the functional relevance of these isoforms is not known at present.
The random mutagenesis PCR suggested periplakin residues 1687-1701 to be part of the Fc␥RI binding domain within periplakin (Fig. 3). Furthermore, the effect of TAT-PPL on Fc␥RI function implies this region to be essential (Fig. 5). Within this sequence aspartic acid at position 1694 diminished binding to Fc␥RI-CY. The negatively charged residues in this region of periplakin align with the cationic residues in Fc␥RI-CY, and mutations found in periplakin align with critical residues of Fc␥RI in yeast (Fig. 3). This part of periplakin also binds vimentin and is a conserved structure among plakin family members (36). However, periplakin residues 1687 and 1689 are exclusively found in periplakin and might facilitate selective binding of Fc␥RI to periplakin in immune cells. In addition, protein kinase B and the -opioid receptor can bind to the C terminus of periplakin (37,38). No clear sequence similarity exists between Fc␥RI and other periplakin-binding proteins. Thus far, it is not known whether other proteins beside Fc␥RI can compete for binding to periplakin in immune cells or bind simultaneously to form complexes.
Thus far we have not succeeded in demonstrating modulation of Fc␥RI binding to monomeric IgG by periplakin. Although we observed a small increase in a stable transfectant containing C-terminal periplakin, TAT-PPL did not consistently modulate monomeric IgG binding to Fc␥RI (data not shown, n ϭ 3). COS cells express considerable levels of endogenous periplakin, and Fc␥RI monomeric IgG binding assays in these cells showed that Fc␥RI-CY lowered ligand affinity by ϳ2-fold (50), supporting the hypothesis that Fc␥RI-CY-periplakin binding decreases Fc␥RI interaction with monomeric IgG.
This report describes Fc␥RI function to be enhanced by reagents that target the receptor intracellular tail. This might lead to the development of selective reagents that regulate Fc␥RI function for immunotherapy. Fc␥RI-directed immunotherapy may be enhanced by TAT-PPL, resulting in increased treatment efficacies when applied in combination.