(Ig) E Receptor ~X Chain Respond to Human Ige in Mast Cell

Home , FCER1, Fc receptor

Transgenic Mice Expressing the Human High-Affinity Immunoglobulin (Ig) E Receptor ~x Chain Respond to Human IgE in Mast Cell Degranulation and in Allergic Reactions ByWai-Ping Fung-Leung,** Jean De Sousa-Hitzler, * Armana Ishaque,* Lubing Zhou, Jesse Pang, Karen Ngo, * Juhe A. Panakos,* Erika Chourmouzis,$ Fu-Tong Liu,~ and CatherineY. Lau*

From the *R. l/ldJohnson Pharmaceutical Research Institute (La Jolla), San Diego, California 92121; CR. W. Johnson Pharmaceutical Research Institute (Toronto), Don Mills, Ontario M3 C 1L9, Canada; and SDepartment of Medicine and Experimental Medicine, The Scripps Research Institute, La Jolla, California 92037 Downloaded from http://rupress.org/jem/article-pdf/183/1/49/1107681/49.pdf by guest on 30 September 2021

Summary The high-affinity receptor for immunoglobulin (Ig) E (FceRl) on mast cells and basophils plays a key role in IgE-mediated allergies. FceRI is composed of one ix, one [3, and two ",/ chains, which are all required for cell surface expression of FceRI, but only the ci chain is involved in the binding to IgE. FceRI-IgE interaction is highly species specific, and rodent FceRI does not bind human IgE. To obtain a "humanized" animal model that responds to human IgE in allergic reactions, transgenic mice expressing the human FceRI ix chain were generated. The human FceRI ix chain gene with a 1.3-kb promoter region as a transgene was found to be sufficient for mast cell-specific transcription. Cell surface expression of the human FceRI ix chain was indicated by the specific binding of human IgE to mast cells from transgenic mice in flow cytometric analyses. Expression of the transgenic FceRI on bone marrow-derived mast cells was 4.7 • 104/ce11, and the human IgE-binding affinity was K d = 6.4 nM in receptor-binding studies using 12sI-IgE. The transgenic human FceRI ix chain was complexed with the mouse [3 and ",/chains in immunoprecipitation studies. Cross-linking of the transgenic FceRI with human IgE and antigens led to mast cell activation as indicated by enhanced tyrosine phosphorylation of the FceRI [3 and ~/chains and other cellular proteins. Mast cell degranulation in transgenic mice could be triggered by human lgE and antigens, as demonstrated by [3-hexosaminidase release in vitro and passive cutaneous anaphylaxis in vivo. The results demonstrate that the human FceRI ix chain alone not only confers the specificity in human IgE binding, but also can reconstitute a functional receptor by coupling with the mouse [3 and y chains to trigger mast cell activation and degranulation in a whole animal system. These transgenic mice "humanized" in IgE-mediated allergies may be valuable for development of therapeutic agents that target the binding of IgE to its receptor.

gE antibodies, mast cells, and basophils play a pivotal role variety of potent mediators, such as histamine, proteases, I in allergic responses (1, 2). Exposure of an individual to and arachidonic acid metabolites, accounts for many of the allergens induces the production of allergen-specific IgE symptoms in allergies (3, 4). antibodies. Mast cells and basophils bind monomeric IgE Fc~RI has been suggested to be the key receptor in IgE- via the high-affinity IgE receptor (FceRI) 1 (1, 2). Subse- mediated mast cell exocytosis (1, 2). This notion is sup- quent exposure to allergens results in cross-linking of reported by the finding that gene-targeted mice defective in ceptor-bound IgE on mast cells and basophils, leading to Fc~RI expression are resistant to IgE-mediated anaphylaxis cellular activation and degranulation (1, 2). The release of a (5). Fc~RI is a tetrameric receptor composed of one ix, one [3, and two disulfide-linked ~/ chains (1). Previous in vitro studies suggest that the binding of Fc~RI to IgE is mediated 1Abbreviations used in this paper: BMMC, bone marrow-derivedmast cells; by the ix chain (6-8). However, cell surface expression of DNP, dinitrophenyl; FceRI, high-affinityIgE receptor; NIP, nitro-iodo- FceRI requires the presence of all three subunits (9, 10). hydroxyphenyl. Fc~RI is expressed not only on mast cells, basophils, and

49 J. Exp. Med. The Rockefeller University Press 0022-1007/96/01/49/08 $2.00 Volume 183 January 1996 49-56 eosinophils, but also on monocytes and Langerhans cells (2, phycoerythrin. The possible cross-reactivity of mouse IgE with 11-15), suggesting functions in addition to mediating de- human FceRI ~x chain (6) in transgenic BMMC was eliminated granulation. by pretreating BMMC with human IgE. Human IgE is highly species specific and does not cross- FcERI Receptor-binding Assay Using 12sI-IgE. Binding assays of react with rodent receptors (6). To obtain an animal model the transgenic human FceRI or the mouse FceRI on BMMC capable of responding to human IgE in allergic reactions, were performed using 12q-labeled human or mouse IgE, respec- we generated transgenic mice expressing the human FceRI tively. BMMC (4 X 106 cells/ml) from transgenic or C57BL/6J mice were incubated at 37~ for 2 h with various concentrations ci chain. Our study showed that the human oL chain of of 12sI-IgE. Cell-bound IgE was separated from free lgE by cen- FceRI alone can reconstitute a functional receptor in mice. trifugation of cells through 0.3 ml of an oil cushion (dioctyl These transgenic mice can respond to human igE in aller- phthalate/dibutyl phthalate, 2:3 vol/vol) at 10,000 g for 1 min. gic reactions. The dissociation constants and the receptor numbers of FceRI on BMMC were calculated from values of the bound and free 12Sl-Igt according to the one binding equation (GraphPad Prism; Materials and Methods GraphPad Software Inc.). Generation of Transgenic Mice Expressing the Human High-Affinity Identification of FcERI Subunits Complexed with the Human FceRl IgE Receptor oe Chain. The human genomic clone for the FceRI ee Chain. BMMC (107 cells) from transgenic or C57BL/6J mice c~ chain that we reported previously (16) was used to construct were incubated overnight with 1 Ixg/ml biotinylated human or the transgene. An 11.4-kb human genomic DNA fragment in the mouse IgE in 5 ml of culture medium. Cells were then solubi- Downloaded from http://rupress.org/jem/article-pdf/183/1/49/1107681/49.pdf by guest on 30 September 2021 clone covering the entire structural gene plus a 1.3-kb promoter lized in CHAPS buffer according to the method described by Ki- region and a 4.2-kb 3' flanking region was injected into mouse net et al. (19) with some modifications. In brief, cells were lysed embryos to generate transgenic mice. The transgene in mice was at 4~ in 0.2 ml of 10 mM CHAPS lysing buffer containing in- confirmed by DNA hybridization. Mouse tail DNA was digested hibitors for phosphatases and proteolytic enzymes (l mM sodium with EcoRI and hybridized to the previously reported human orthovanadate, 12.5 mM sodium fluoride, 0.1 mM zinc chloride, Fc~RI a chain cDNA (16). Transgenic founders identified by hy- 2 mM EDTA, 2 mM EGTA, 1 mM PMSF, 20 btg/ml aprotinin, bridization were then bred with C57BL/6J mice to obtain trans- and 10 Ixg/ml leupeptin). IgE-bound FceRI in cell lysates was genic offspring for phenotypic and functional studies. recovered by overnight incubation at 4~ with 0.1 ml packed Detection of Transgenic mRNA. Total RNA was isolated from volume of precleared avidin-conjugated agarose beads (Pierce bone marrow-derived mast cells (BMMC) and from mouse or- Chemical Co., Rockford, lL). The IgE-FceRI complex was gans according to the method described previously (16). The or- eluted in SDS-PAGE loading buffer, separated on 12.5% SDS- gans used in RNA extraction included heart, kidney, liver, polyacrylamide gels under reducing condition, and then trans- spleen, stomach, and skin. mRNA encoding the human FceRI 0~ ferred to membranes (PolyScreen; New England Nuclear, Bos- chain was detected by Northern blot hybridization using the pre- ton, MA). The mouse FceRI [3 and y chains were detected by viously reported human FceRI cr chain cDNA as a probe (16). the rat FceRI ~3 chain-specific mAb and the rabbit serum specific mRNA for the mouse FceRl c~ chain was detected by the oligo- for the mouse Fc~RI y chain (20), respectively. Antibody binding nucleotide 5'-TGTCAAAGGATCCATGGACTAAGATCATG- was detected with chemiluminescent reagents (ECL kit; Amer- 3', which is specific for exon 4 of the mouse FceRI o~ chain gene sham Corp., Arlington Heights, IL). and does not cross-react with the transgenic mRNA. Detection of Tyrosine Phosphorylation Triggered by Cross-linking of Mast Cell Cultures Derivedfrom Bone Marrows. Mast cells from FcERI. BMMC (2 X 106 cells) from transgenic or C57BL/6J transgenic or C57BL/6J mice were prepared from bone marrow mice were incubated either with 1 ~g/ml of anti-nitro-iodo- cells according to the method described by Razin et al. (17). In hydroxyphenyl (NIP) human IgE (Serotec Ltd.) or anti-dinitro- brief, bone marrow cells were collected from mouse femur and phenyl (DNP) mouse IgE (18) for 2 h at 37~ in PBS. Cells were tibia and cultured at 3 X 105 cells/ml in Razin's medium then rinsed with PBS and incubated for 2 min at 37~ with 40{} (RPMI-1640 medium with 10% FCS, 50 IxM 2-ME, 2 mM ng/ml NIP-BSA or DNP-BSA, respectively. BMMC incubated L-glutamine, 0.1 mM nonessential amino acids, 100 U/ml peni- with IgE but not with the corresponding antigens were used as cillin/streptomycin) supplemented with 20% WEHI-3B cell su- negative controls. Cells were lysed in SDS-PAGE loading buffer, pernatant (American Type Culture Collection, Rockville, MD). and proteins were separated and transferred to PolyScreen mem- Culture medium was changed weekly. After 3 wk in culture, the branes as mentioned above. Phosphotyrosine was detected by the majority of the bone marrow cells differentiated into mast cells. phosphotyrosine-specific mouse mAb 4G10 (Upstate Biotechnol- Mast cells were identified by the presence of basophilic granules ogy, Inc., Lake Placid, NY), and antibody binding was detected in the cytoplasm of the cells after May Grtinwald/Giemsa staining. with chemituminescent reagents from Amersham Corp. The Binding of IgE to Mast Cells Analyzed by Flow Cytometry. Detection of Tyrosine Phosphorflation on FcERI Subunits Triggered BMMC from transgenic and normal C57BL/6J mice were used by the Cross-link of FcERl. BMMC (107 cells) from transgenic or in flow cytometric analyses. 2 • 10 s BMMC in 50 IM of staining C57BL/6J mice were incubated overnight with 1 I~g/ml biotiny- buffer (PBS with 2% FCS, 0.1% sodium azide) were incubated lated anti-NIP human IgE or anti-DNP mouse IgE in 5 ml of for 1 h at 4~ with 2.5 Dg of biotinylated human IgE (Serotec culture medium. The human or mouse IgE-coated BMMC were Ltd., Oxford, UK) or mouse IgG (18), followed by incubation incubated with 1 Ixg/ml NIP-BSA or DNP-BSA, respectively, with streptavidin-phycoerythyrin (Becton Dickinson & Co., for 2 rain at 37~ in PBS to cross-link IgE-bound Fc~Rl. Mountain View, CA) for 20 min at 4~ Cell-bound IgE was an- BMMC were solubilized in CHAPS buffer as described earlier, alyzed via flow cytometry using the FACScan| program (Becton and IgE-bound FceRI in cell lysates was recovered by avidin- Dickinson & Co.). Specificity of human IgE binding was studied conjugated agarose beads. Proteins eluted from agarose beads by incubating BMMC with unbiotinylated human IgE before were separated by reducing S1)S-PAGE, transferred to Poly- biotinylated human IgE, which was detected by streptavidin- Screen membranes, and phosphotyrosines were detected. BMMC

50 Transgenic Human High-Affinity IgE Receptor o~ Chain in Allergic Reactions treated similarly with unbiotinylated human or mouse IgE did not show any signals in phosphotyrosine detection, confirming the specificity of the immunoprecipitations. IgE-mediated BMMC Degranulation Detected by the Release of [3-Hexosaminidase. Mast cell degranulation triggered by IgE and antigens was carried out according to the methods described by Yen et al. (21). BMMC (5 X I0 s cells) in 0.5 ml Tyrode's buffer were sensitized with 1 p~g/ml anti-NIP human IgE or anti-DNP mouse lgE for 1 h at 37~ After two rinses with Tyrode's buffer to remove unbound IgE, degranulation was triggered by 10-rain incubation with 50 ng/ml of the antigens, DNP, or NIP conjugated to BSA. Degranulation was assayed by measuring the en- zyme activity of released [3-hexosaminidase using a colorimetric assay (22). Spontaneous 13-hexosaminidase release was measured in samples not treated with IgE or antigen. Total 13-hexosaminidase was measured after lysing cells by sonication. Percentage of 13-hexosaminidase release was calculated as: (13-hexosaminidase in Figure 1. Expression of the human Fc{RI ~x chain transgene. (A) The supernatant/total [3-hexosaminidase) • 100. transgene was an 11.4-kb human genomic DNA fragment containing the

Passive Cutaneous Anaphylaxis. Passive cutaneous anaphylaxis entire structural gene with 5 exons (black blocks), a 1.3-kb promoter re- Downloaded from http://rupress.org/jem/article-pdf/183/1/49/1107681/49.pdf by guest on 30 September 2021 was performed in mice according to the method described by gion, and a 4.2-kb 3' flanking region. The location of the codons ATG Saloga et al, (23). Transgenic mice and normal C57BL/6J mice and TGA, indicating the start and the end sites of translation, respectively, are shown. The transgene for embryo injection was recovered by SacI di- were injected intradermally on the dorsal side with 500 ng NIP- gestion. The SacI site at the 5' end of the DNA fragment is derived from specific human lgE in 25 txl saline, and on the ventral side with the plasmid vector. (B) Transgenic mice were identified by Southern blot 25 Ixl saline for comparison. 2 h later, NIP-BSA conjugates (0.5 hybridization. Tail genomic DNA samples were digested with EcoRI, mg) plus 1% Evans blue in 250 txl saline i.v. was injected. Mice and the transgene in transgenic (TG) mice was detected by a human were killed 20 rain later, and cutaneous anaphylaxis was assessed Fc~RI (x chain cDNA probe (16). The specificity of the DNA probe was visually by the blue dye leakage from blood vessels into the skin. confirmed by the positive DNA bands shown in the DNA sample from a Mouse IgE-mediated anaphylaxis was carried out similarly using human lymphoblastic cell line (IM9), but not in that from the nontrans- DNP-specific mouse IgE and DNP-BSA. genic C57BL/6J mouse (B6).

specificity in the binding to human IgE (Fig. 3 E). 125I-human Results IgE was also used in receptor-binding assays to determine the binding affinity of transgenic FcERI toward human Mast Cell-specific Transcription of the Transgene Encoding the IgE. As shown in Table 1, the number of transgenic FceRI Human FcERI ce Chain. We previously cloned the human on BMMC was 4.7 X 104/ce11, and the human IgE dissoci- FceRI ci chain gene and showed that the human gene is ation constant was 6.4 nM. These values were comparable similar in genomic structure to that of the rodent (16). The to those of the mousee Fc{RI (1.8 X 10S/cell, Ka = 2.19 human FceRI 0r chain gene spans over a region of 5.9 kb nM) determined in the same experiment using mouse IgE and contains 5 exons. As shown in Fig. 1, an 11.4-kb DNA (Table 1), and were also consistent with the values reported fragment covering the entire human cx chain gene with an previously for human or mouse FceRI on mast cells (24, 1.3-kb promoter region and a 4.2-kb 3' flanking region 25). Taken together, the specific binding of human IgE to was used to generate transgenic mice. Transcription of the BMMC from transgenic mice as shown in flow cytometric transgene in different organs and in mast cells from trans- analyses and in 12sI-IgE-binding assays confirms the cell genic mice was studied by Northern blot hybridization. surface expression of the human FceRI oL chain. Mast cells used in the study were BMMC. Transgenic mRNA was detected only in BMMC, but not in the different organs from transgenic mice (Fig. 2 A). Tissue distri- bution of transgenic mRNA was identical to that tran- Figure 2. Tissue-specific expres- scribed by the endogenous mouse Fc~RI ci chain gene sion of FceRI cl chain mRNA from (Fig. 2 B). The data suggest that the 1.3-kb promoter re- the human transgene. (A) Total RNA samples from different organs (t5 p~g) gion of the human Fc~RI cl chain gene is sufficient for reg- and from BMMC (7.5 ~zg) of trans- ulating proper tissue-specific expression. genic mice were detected by a human Cell Surface Expression of the Human FcERI cr Chain on FceRI cx chain cDNA probe. (B) Transgenic Mast Cells. Cell surface expression of the trans- Mouse FceRI c~ chain mRNA was genic human Fc~RI ci chain on mast cells was assessed by detected by an oligonucleotide specific for the mouse gene within exon 4. its specific binding to human IgE. Human IgE was shown RNA samples (35 ~g) from the to bind to BMMC from transgenic mice, but not to that human KU812 mast cell line (41) from C57BL/6J mice in flow cytometric analyses (Fig. 3, C and from BMMC of nontransgenic and D). The binding of biotinylated human IgE to trans- C57BL/6J mice (B6) were used as positive controls for hmnan and genic BMMC was completely abrogated when cells were mouse FceRl 0t chain mRNA, respec- pretreated with unbiotinylated human lgE, suggesting the tively.

51 Fung-Leung et al. g." j A C Figure 3. Cell surface expression of the transgenic human Fc~RI ~x chain on mast cells C57BL/6 from transgenic mice. (A and B) Histograms of I cell-bound mouse IgE on BMMC from transgenic (Tg) mice and nontransgenic C57BL/6J mice are shown. These BMMC samples have

~ " ~6, " ..... i~ ":":"'~, been treated with 2.5 p.g unbiotinylated human IgE for 20 min at 4~ before binding to mouse IgE. (C and D) Histograms of cell-bound human IgE on BMMC are shown. (E) The histogram of cell-bound human IgE on the trans- Tg genic BMMC pretreated with unbiotinylated human IgE is shown. Cells stained with streptavidin-phycoerythrin alone were used as negative controls (gray histogram) to compare to specific IgE binding (black histogram). Cell-bound IgE was analyzed by flow cytometry using the mlgE-B hlgE-B hlgE + hlgE-B FACScan| program. Downloaded from http://rupress.org/jem/article-pdf/183/1/49/1107681/49.pdf by guest on 30 September 2021

Mouse FceRI on mast cells was assayed similarly using transgenic human FceRI or the endogenous mouse FceRI mouse IgE. Mast ceils from both transgenic and C57BL/6J was immunoprecipitated via its binding to human or mouse mice were shown to bind to mouse IgE in flow cytometric IgE, respectively. The mAb specific for the rat Fc~RI [3 analyses (Fig. 3, A and B). However, the amount of cell- chain detected the mouse 13 chain in mouse FceRI immu- bound mouse IgE on transgenic mast cells was less than that noprecipitates with .an apparent molecular mass of 30 kD on the nontransgenic BMMC. Consistent with this finding, (Fig. 4). Interestingly, the mouse FceRI 13 chain was also 125I-IgE-binding studies also showed that mouse FceRI re- present in the transgenic human Fc{RI immunoprecipitates ceptor number on transgenic BMMC was reduced, whereas (Fig. 4). The mouse Fc~RI ",/chain, which was ~10 kD in its mouse IgE-binding affinity was unchanged (Table 1). size, was detected with rabbit sera specific for the mouse The results suggest that the cell surface expression of mouse Fc~RI ~/ chain (20). Similarly, the mouse y chain was FceRI was reduced on BMMC from transgenic mice. This found in both the immunoprecipitates of the transgenic decrease in mouse FceRI level on transgenic mast cells human FceRI and the mouse FceRI (Fig. 4). The results could be due to the competition between the human and demonstrated that the human Fc~RI et chain is coupled mouse (x chains in coupling with the mouse y and/or 13 with both the mouse 13 and "y chains to form a chimeric chains. The possibility that the human FceRI 0t chain forms human/mouse receptor. a chimeric receptor with mouse y and/or [3 chains has also Normal Tyrosine Phosphorylation Triggered by Cross-linking been suggested by previous transfection studies (9, 10, the Transgenic FceRI. Activation and degranulation of mast 26-28). cells via FceRI aggregation has been shown to involve The Transgenic Human FcERI ee Chain Is Complexed with early signaling events including activation of protein ty- the Mouse [3 and 3, Chains. The human FceRI 0L chain has rosine kinases (29-33). The function of the transgenic FceRI been shown in transfected cells to require at least the ot and in initiating tyrosine phosphorylation in BMMC was inves- y chains for cell surface expression (10, 26-28). In this tigated. Transgenic human Fc~RI was cross-linked by first transgenic model, proteins that are complexed with the binding to anti-NIP human IgE (34), followed by interact- transgenic human FceRI cx chain were analyzed in immu- ing with the multivalent antigen NIP-BSA. Mouse endog- noprecipitation studies. To maintain the integrity of the enous Fc~RI was aggregated similarly by anti-DNP mouse FcERI receptor complex, BMMC were lysed with CHAPS IgE (18) and DNP-BSA. Significant induction in tyrosine buffer according to the method of Kinet et al. (19). The

Figure 4. The transgenic human Table 1. Number of Human FcERI on Transgenic BMMC and FceRI (x chain is complexed with Its Binding AflTnity to Human IgE the mouse ~ and 3' chains. Mouse FceRI from C57BL/6J (Bt) BMMC or human Fc~RI from transgenic BMMC IgE Receptor/cell Kd (Tg) BMMC was immunoprecipitated with mouse IgE (m) or human nM IgE (h) as described in Materials and Transgenic Human 4.7 X 104 6.40 Methods. The mouse FceRI ~ chain was detected with anti-rat 13 mAb Mouse 4.1 X 10 4 2.93 (20), whereas the 3' chain was de- C57BL/6J Mouse 1.8 X 10 s 2.19 tected by using the 3' chain-specific rabbit serum (21).

52 Transgenic Human High-Affinity IgE Receptor c~ Chain in Allergic Reactions phosphorylations was detected 2 rain after cross-linking Figure 6. Tyrosinephosphor- of mouse FceRI on BMMC from transgenic or control ylation of the mouse ~ and ~/ chains in the transgenic FceRI C57BL/6J mice (Fig. 5). Human IgE and NIP-BSA also upon receptor aggregation. triggered pronounced tyrosine phosphorylations in trans- Fc~RI on BMMC from C57BL/ genic BMMC but not in C57BL/6J BMMC, suggesting 6J (B6) or transgenic (Tg) mice that the human IgE-induced signaling events were medi- were cross-linked (X) by mouse IgE (m) or human IgE (h) plus ated specifically by the transgenir human FceRI (Fig. 5). their specific antigens. FceRI The patterns of phosphorylated proteins induced by aggre- was then immunoprecipitated, gation of the transgenic human FceRI or the endogenous and phosphotyrosines were de- mouse FceRI appeared to be similar (Fig. 5). tected as described in Materials and Methods. Uncross-linked Recent reports have shown that aggregation of Fc~RI mouse Fc~RI was also used as a leads to significant tyrosine phosphorylation of the 13 and ~/ control for comparison. subunits, which recruit other signaling molecules for cellular activation (29, 31). Tyrosine-phosphorylated proteins in immunoprecipitates of the transgenic FceRI were therefore specific mouse IgE and DNP-BSA induced degranulation examined. Proteins of ~30 and 10 kD, which correspond of BMMC from both transgenic and nontransgenic mice by size to the FceRI [3 and y chains, respectively, were (Fig. 7). The results indicate that the FceRI with the trans- Downloaded from http://rupress.org/jem/article-pdf/183/1/49/1107681/49.pdf by guest on 30 September 2021 barely detectable in immunoprecipitates of uncross-linked genic human o~ chain is responding to human IgE, and is mouse FceRI (Fig. 6). Upon aggregation of mouse Fc{RI, functional in mediating murine mast cell exocytosis. Mast these proteins were dramatically enhanced in tyrosine cells from transgenic mice, however, can still respond to phosphorylation (Fig. 6). Cross-linking of the transgenic mouse IgE in exocytosis. FceRI also showed significant tyrosine phosphorylation of Passive Cutaneous Anaphylaxis of Transgenic Mice Induced by the 30 and 10 kDa proteins (Fig. 6). Human IgE does not Human IgE and Antigens. To assess whether the transgenic bind to mouse Fc~RI and therefore had no effect on Fc~RI Fc~RI with the human o~ chain can respond to human lgE on nontransgenic BMMC. Taken together, the data sug- in mast cell degranulation in vivo, passive cutaneous ana- gest that the transgenic human Fc~RI is functional in trig- phylaxis was carried out in transgenic mice. Mice were in- gering mast cell activation. The coupling of mouse [3 and y tradermally injected wtih human or mouse IgE, followed chains with the transgenic human ot chain appears not only by intravenous injection of antigens and Evans blue. In important for cell surface expression, but also for the proper transgenic mice, human IgE and its antigen NIP-BSA in- function of the chimeric receptor. duced a specific leak of Evans blue into the skin at the site Degranulation of Transgenic Mast Cells Triggered by Cross- ofintradermal injection. This blue dye leakage was not ob- linking the Human FcERI ee Chain. With this transgenic served in control C57BL/6J mice (Fig. 8). The extravasa- model, we address the question whether the human/mouse tion of the dye is a result of blood vessel dilation, which is chimeric receptor is competent in eliciting mast cell de- induced by mediators released from mast cells in degranula- granulation. Mast cells were prepared from transgenic mice, tion. Mouse monoclonal IgE and its antigen DNP-BSA and degranulation was triggered by first allowing the bind- caused a similar leak of Evans blue in both transgenic mice ing of IgE to Fc~RI on mast cells, followed by cross-link- and control C57BL/6J mice (Fig. 8). The results provide a ing receptor-bound IgE with a multivalent antigen. Release of [3-hexosaminidase upon degranuladon was measured. The 20 NIP-specific human IgE and NIP-BSA triggered a specific release of [3-hexosaminidase from transgenic BMMC, but not from control C57BL/6J BMMC (Fig. 7). The DNP-

Figure 5. Tyrosine phosphoryla- am tion of cellular proteins induced by ,a' cross-linking of transgenic FceRl. FceRl on BMMC from C57BL/6J (B6) and transgenic (Tg) mice were h IgE rn IgE Spont. cross-linked (X) by mouse IgE (m) or human IgE (h) plus their specific Figure 7. Human IgE plus antigens mediated degranulation of mast antigens as described in Materials cells from transgenic mice. ~-Hexosaminidase released from transgenic and Methods. Uncross-linked con- BMMC (blackcolumn) and nontransgenic C57BL/6J BMMC (graycolumn) trols were samples treated with IgE were induced by human or mouse monoclonal lgE plus their specific an- but not the antigens. Total cell ly- tigens, NIP-BSA or DNP-BSA, respectively. The value of ]3-hex- sates were prepared 2 rain after osaminidase release is the average of three experiments. Error bars indicate cross-linking the receptors, and SEM. In nontransgenic BMMC, 13-hexosaminidase release induced by phosphotyrosines were detected on human IgE is not significantly different from spontaneous release (p = Western blots. 0.41 by Wilcoxin rank sum test).

53 Fung-Leung et al. Figure 8. Human IgE plus antigen induced passive cutaneous anaphylaxis in transgenic mice. Passive cutaneous anaphylaxis in transgenic (Tg) and C57BL/6J control (C57BL/6) mice was triggered by mouse IgE (mlgE)

or human lgE (hlgE) plus their Downloaded from http://rupress.org/jem/article-pdf/183/1/49/1107681/49.pdf by guest on 30 September 2021 specific antigens as described in Materials and Methods. clear in vivo demonstration that transgenic mice can re- ity, whereas the 13 and y chains are responsible for signaling spond to human IgE in anaphylaxis. events (6, 11, 29, 31, 38, 39). The extracellular portion of the FceRI ~ chain has been shown to display less conserva- tion between species (10). This structural diversity could be Discussion responsible for the species specificity seen in IgE-FceRI in- Human FceRI expression on the cell surface is known to teraction. require at least the ot and y chains (10, 26-28). iThe ability The proper assembly and coordination of the different of the "y chain to allow surface expression of the Fc~RI subunits of FceRI is essential for its function in triggering chain is analogous to the role of the ~ chains in the expres- mast cell degranulation (1). In this regard, the FceRI c~ sion of the TCR-CD3 complex, although the molecular chain receives external signals through IgE binding, whereas mechanism of the FceRI assembly is still unclear. Both the the FceRI y chain plays a critical role in signal transduction FceRI c~ and y chains are highly conserved at the trans- (11, 29, 31, 38, 39). Although the human FceRI o~ chain in membrane portions, which are suggested to be important association with mouse y chains has been shown to trigger for proper assembly of the FceRI receptor complex (10, limited early signaling events in mast cell lines, the induc- 28). In contrast, the role of the 13 chain for cell surface ex- tion of cellular degranulation has not been demonstrated pression and activation of human FceRI has been unclear. (26, 39, 40). We showed with this transgenic model that Questions have therefore been raised regarding to the im- the human/mouse chimeric FceRI is functional in mediat- portance of the Fc~RI 13 chain in humans and the possible ing mast cell activation and degranulation. The results sug- difference in FceR1 receptor complex between humans gest that the proper conformation of the receptor complex and rodents. We have demonstrated here in a transgenic is maintained in the transgenic FceRI. model that the transgenic FceRI, which elicits normal mast Allergic diseases affect 20% of the population and are an cell activation and degranulation, is composed of the hu- important cause of morbidity and mortality. The cell type- man c~ chain and the mouse 13 and y chains. The data sug- specific expression and the key role in mast cell exocytosis gest that the 13 chain may play a role in the biological func- make FceRI an attractive drug target for treatment of IgE- tions of FceRI. The importance of the 13 chain has also dependent allergies. However, animal studies are limited by been suggested by the finding that it is also associated with the species specificity in IgE-FceRI interaction. We have the IgG Fc receptor type III, which mediates effector func- demonstrated here in a transgenic model that the human tions similar to that of FceRI (35). A correlation of muta- FceRI o~ chain not only confers the specificity in human tions on the FceRI [3 chain gene with atopic dermatitis has IgE binding but also can reconstitute a functional FceRI to also been reported recently (36, 37). trigger mast cell degranulation in vitro and in vivo. These The transgenic FceRI in which only the ot chain was of transgenic mice with a "humanized" FceRI could be a human origin was shown to bind to human IgE specifically. valuable model for developing drugs that block human IgE This is consistent with previous in vitro studies showing from binding to its receptor. that the ot chain confers entirely the IgE-binding specific-

We thank A. Dalloul for critical comments, A. Yen for helpful advice in bone marrow cultures, K. Kishi for providing the KU812 cell line, J. Ravetch for the mouse laceRI y chain~pecific antiserum, B. Baird and

54 Transgenic Human High-Affinity IgE Receptor a Chain in Allergic Reactions D. Holowka for the rat FceRI [3 chain~pecific antibody, L. Traeger and J. Booth for preparation of the manuscript, and A. Branco andJ. Doctolero for maintenance of the animal facility.

Address correspondence to Dr. Wai-Ping Fung-Leung, The. R.W. Johnson Pharmaceutical Research Insti- tute, 3535 General Atomics Court, Suite 100, San Diego, CA 92121.

Receivedfor publication 5 August 1994 and in revisedform 2June 1995.

References 1. Metzger, H. 1992. The receptor with high affinity for IgE. Fung-Leung, C.Y. Lau, F.-T. Liu, and L. Zhou. 1993. Char- Immunol. Rev. 125:37-48. acterization of the gene for the human high affinity IgE re- 2. Metzger, H., G. Alcaraz, R. Hohman, J.P. Kinet, V. Pri- ceptor (FceRl) oL-chain.J. Immunol. 151:6166-6174. bluda, and R. Quarto. 1986. The receptor with high affinity 17. Razin, E., C. Cordon-Cardo, and R.A. Good. 1981. Growth for immunoglobulin E. Annu. Rev. Immunol. 4:419-470. of a pure population of mouse mast cells in vitro with condi- 3. Serafin, W.E., and Austen, K.F. 1987. Mediators of immedi- tioned medium derived from concanavalin A-stimulated sple- ate hypersensitivity reactions. N. Engl. J. Med. 317:30-34. nocytes. Proc. Natl. Acad. Sci. USA. 78:2559-2561. 4. Stevens, R.U, and K.F. Austen. 1989. Recent advances in 18. Liu, F.-T., J.W. Bohn, E.L. Ferry, H. Yamamoto, C.A. Mo- the cellular and molecular biology of mast cells. Immunol. To- linaro, L.A. Sherman, N.R. Klinman, and D.H. Katz. 1980. Downloaded from http://rupress.org/jem/article-pdf/183/1/49/1107681/49.pdf by guest on 30 September 2021 day. 10:381-386. Monoclonal dinitrophenyl-specific murine IgE antibody: prep- 5. Dombrowicz, D., V. Flamand, K.K. Brigman, B.H. Koller, aration, isolation and characterization.J. Immunol. 124:2728- and J.-P. Kinet. 1993. Abolition of anaphylaxis by targeted 2737. disruption of the high affinity immunoglobulin E receptor cx 19. Kinet, J.-p., G. Alcaraz, A. Leonard, S. Wank, and H. chain gene. Cell. 75:969-976. Metzger. 1985. Dissociation of the receptor for immuno- 6. Hakimi, J., C. Seals, J.A. Kondas, L. Pettine, W. Danho, and globulin E in mild detergents. Biochemistry. 24:4117-4124. J. Kochan. 1990. The ot subunit of the human IgE receptor 20. Kurosaki, T., andJ.V. Ravetch. 1989. A single amino acid in (FceR1) is sufficient for high affinity IgE binding. J. Biol. the glycosylphosphatidylinositol attachment domain deter- Chem. 265:22079-22081. mines the membrane topology of Fc~/RIII. Nature (Lond.). 7. Riske, F., J. Hakimi, M. Mallamaci, M. Griffin, B. Pilson, N. 342:805-807. Tobkes, P. Lin, W. Danho, J. Kochan, and R. Chizzonite. 21. Yen, A., F.-T. Liu, K.E. Barrett, and I. Gigli. 1993. Alter- 1991. High affinity human IgE receptor (FceR1). J. Biol. ations in FceRI induced by photoporthyrin plus long wave- Chem. 266:11245-11251. length ultraviolet light in mouse bone marrow derived mast 8. Blank, U., C. Ra, and J.P. Kinet. 1991. Characterization of cells.J. Immunol. 151:1003-1011. truncated ot chain products from human, rat and mouse high 22. Schwartz, L.B., K.F. Austen, and S.I. Wasserman. 1979. Im- affinity receptor for immunoglobulin E.J. Biol. Chem. 266: munologic release of [3-hexosaminidase and [3-glucuronidase 2639-2646. from purified rat serosal mast cells. J. Immunol. 123:1445- 9. Blank, U., C. Ra, L. Miller, andJ.P. White. 1989. Complete 1450. structure and expression in transfected cells of high affinity 23. Saloga, J., H. Renz, G. Lack, K.L. Bradley, J.I. Greenstain, IgE receptor. Nature (Lond.). 337:187-189. G. Larsen, and E.W. Gelfand. 1993. Development and trans- 10. Ra, C., M.H. Jouvin, andJ.P. Kinet. 1989. Complete struc- fer of immediate cutaneous hypersensitivity in mice exposed ture of the mouse mast cell receptor for IgE (FceR1) and sur- to aerosolized antigen.J. Clin. Invest. 91:133-140. face expression of chimeric receptors (rat-mouse-human) on 24. Capron, A., J.P. Dessaint, M. Capron, M. Joseph, J.C. transfected cells.J. Biol. Chem. 264:15323-15327. Ameisen, and A.B. Tonnel. 1986. From parasites to allergy: a 11. Ravetch, J.V., andJ.P. Kinet. 1991. Fc receptors. Annu. Rev. second receptor for IgE. Immunol. Today. 7:15-18. Immunol. 9:457-492. 25. Ravetch, J., and J.-P. Kinet. 1991. Fc receptors. Annu. Rev. 12. Gounni, A.S., B. Lamhiared, K. Ochiai, Y. Tanaka, E. De- Immunol. 9:457-492. laporte, A. Capron, J.P. Kinet, and M. Capron. 1994. High 26. Miller, L., G. Alber, N. Varin-Blank, R. Ludowyke, and H. affinity IgE receptor on eosinophils is involved in defence Metzger. 1990. Transmembrane signalling in p815 mastocy- against parasites. Nature (Lond.). 367:183-186. toma cells by transfected IgE receptors. J. Biol. Chem. 265: 13. Maurer, D., A. Rieger, O. Kilgus, B. Reininger, K. Ochiai, 12444-12553. J.P. Kinet, and G. Stingl. 1994. Expression of functional high 27. Kuster, H., H. Thompson, and J.P. Kinet. 1990. Character- affinity immunoglobulin E receptors (FceRI) on monocytes ization and expression of the gene for the human Fc receptor ofatopic individuals.J. Exp. Med. 179:745-750. gamma subunit. Definition of a new gene family. J. Biol. 14. Wang, B., A. Rieger, O. Kilgus, K. Ochiai, D. Maurer, D. Chem. 265:6448--6452. Fodinger, J.P. Kinet, and G. Stingl. 1992. Epidermal Langer- 28. Varin-Blank, N., and H. Metzger. 1990. Surface expression hans cells from normal human skin bind monomeric IgE via of mutated subunits of the high affinity mast cell receptor for FceR1.J. Exp. Med. 175:1353-1365. IgE.J. Biol. Chem. 265:15685-15694. 15. Bieber, T., H. de la Salle, A. Wollenberg, J. Hakimi, R. 29. Eiseman, E., andJ.B. Bolen. 1992. Engagement of the high- Chizzonite, J. Ring, D. Hanau, and C. de la Salle. 1992. Hu- affinity IgE receptor activates src protein-related tyrosine ki- man epidermal Langerhans cells express the high affinity re- nases. Nature (Lond.). 355:78-80. ceptor for irrmmnoglobulin E (Fc~R1). J. Exp. Med. 175: 30. Minoguchi, K., M. Benhamou, W.D. Swain, Y. Kawakami, 1285-1290. T. Kawakami, and R.P. Siraganian. 1994. Activation of pro- 16. Pang, J., G.R. Taylor, D.G. Munroe, A. Ishaque, W.-P. tein tyrosine kinase p75 ~yk by Fc~RI aggregation in rat baso-

55 Fung-Leung et al. philic leukemia cells.J. Biol. Chem. 269:16902-16908. 36. Ravetch, J.V. 1994. Atopy and Fc receptors: mutation is the 31.Jouvin, M.-H.E., M. Adamczewski, R. Numero(, O. Le- message? Nature Genetics. 7:117-118. tourneur, A. Valle, andJ.-P. Kinet. 1994. Differential control 37. Shirakawa, T., A. Li, M. Dubowitz, J.W. Dekker, A.E. of the tyrosine kinases Lyn and Syk by the two signaling Shaw, J.A. Faux, C. Ra, W.O.C.M. Cookson, and J.M. chains of the high affinity immunoglobulin E receptor. J. Hopkin. 1994. Association between atopy and variants of the Biol. Chem. 269:5918-5925. [3 subunits of the high-affinity immunoglobulin E receptor. 32. Yamashita, T., S.-Y. Mao, and H. Metzger. 1994. Aggrega- Nature Genetics. 7:125-129. tion of the high-affinity IgE receptor and enhanced activity 38. Kinet, J.P. 1992. The gamma-zeta dimers of Fc receptors as of p53/56 ly'1 protein-tyrosine kinase. Proc. Natl. Acad. Sci. connectors to signal transduction. Curr. Opin. Immunol. 4:43- USA. 91:11251-11255. 48. 33. Fukamachi, H., N. Yamada, T. Miura, T. Kato, M. Ishikawa, 39. Alber, G., L. Miller, C.L. Jelsema, N. Varin-Blank, and H. E. Gulbins, A. Altman, Y. Kawakami, and T. Kawakami. Metzger. 1991. Structure-function relationships in the mast 1994. Identification of a protein, SPY75, with repetitive he- cell high affinity receptor for IgE.J. Biol. Chem. 266:22613- lix-turn-helix motifs and an SH3 domain as a major substrate 22620. for protein tyrosine kinases activated by FceRI cross-linking. 40. Taudou, G., N. Varin-Blank, H. Jouin, F. Marchand, A. J. Immunol. 152:642-652. Weyer, and U. Blank. 1993. Expression of the alpha chain of 34. Neuberger, M.S., G.T. Williams, E.B. Mitchell, S.S. Jouhal, human FceRI in transfected rat basophilic leukemia cells: J.G. Flanagan, and T.H. Rabbitts. 1985. A hapten-specific functional activation after sensitization with human mite-spe-

chimeric IgE antibody with human physiological effector cific IgE. Int. Arch. Allergy Immunol. 100:344-350. Downloaded from http://rupress.org/jem/article-pdf/183/1/49/1107681/49.pdf by guest on 30 September 2021 function. Nature (Loud.). 314:268-270. 41. Kishi, K. 1985. A new leukemia cell line with Philadelphia 35. Kurosaki, T., I. Gander, U. Wirthmueller, and J.V. Ravetch. chromosome characterized as basophil precursors. Leuk. Res. 1992. The [~ subunit of the FceRI is associated with the 9:381-390. Fc',/RIII on mast cells.J. Exp. Med. 175:447-451.

56 Transgenic Human High-Affinity IgE Receptor o~ Chain in Allergic Reactions