Immunoglobulin Superfamily Receptor Translocation Associated 2 Protein on Lymphoma Cell Lines and Hairy Cell Leukemia Cells Detected by Novel Monoclonal Antibodies

Purpose: The immunoglobulin superfamily receptor translocation associated 2 (IRTA2) gene encodes a cell surface receptor homologous to the family of Fc receptors. Because of the restricted expression of mRNA in B cell–lineage cells, IRTA2 is a new potential target for the immunotherapy of B cell malignancies. To study the expression of the IRTA2 gene product, we produced monoclonal antibodies (MAbs) specific to IRTA2.

Experimental Design: A mouse used for cell fusion was DNA-immunized with an expression plasmid encoding the IRTA2 cDNA. The reactivity of MAbs secreted from the hybridomas were characterized with recombinant proteins of IRTA family members in an enzyme immunoassay and a fluorescence-activated cell sorter (FACS). Nineteen human lymphoma cell lines and blood cells from five patients with hairy cell leukemia (HCL) were analyzed with IRTA2 expression using FACS.

Results: Three MAbs (F25, F56, and F119) were selected based on their specific reactivity with recombinant IRTA2 and lack of cross-reactivity with other IRTA family members. In a FACS analysis, MAbs F56 and F119 detected IRTA2 expression in six of seven B cell non–Hodgkin's lymphoma and one of six Burkitt's lymphoma cell lines. Reverse transcriptase-PCR experiments and Western blotting using MAb F25 confirmed the expression profile. We also found that HCL cells from five patients expressed IRTA2.

Conclusions: Our results provide the first evidence that IRTA2 is expressed on the surface of human lymphoma cell lines and HCL cells. IRTA2 could be useful as a new target for immunotherapy.

The immunoglobulin (Ig) superfamily receptor translocation associated 2 (IRTA2) gene was identified as a gene differentially targeted by one of the 1q21 abnormalities which occur in B cell malignancies (1). Other research groups identified the same gene as BXMAS1 (2) and Fc receptor homologue 5 (FcRH5; ref. 3). This was accomplished by representational difference analysis of genes induced by anti-IgM crosslinking of a human B cell line, and by a BLAST query using a 32-amino acid consensus sequence in previous Fcγ receptors' Fc-binding site. A series of genes homologous to IRTA2/BXMAS1/FcRH5 have been identified in the same locus of the human genome. A total of five IRTA family members have been identified (4–6). These genes code for novel members of the Ig receptor family with three to nine Ig-like extracellular domains. Their cytoplasmic domains contain consensus immunoreceptor tyrosine-based activation motifs and/or immunoreceptor tyrosine-based inhibitory motifs, suggesting a role for IRTAs in cell differentiation and immune responses. Expression of IRTAs was investigated in human tissues at the RNA level by Northern blot, reverse transcriptase (RT)-PCR, and by in situ hybridization (1–3). mRNAs of IRTAs have been shown to be abundant in lymphoid tissues and each member is differentially expressed by mature B lineage cells in human tonsil. The IRTA proteins, however, have not been detected in cells or tissues except for one report with IRTA1 (7). The lack of specific antibodies limits detailed expression studies on different cell types and cancers.

Many groups are interested in developing immunotherapy using differentiation antigens on hematologic malignancies as targets (8). Target antigens include CD20 (9), CD22 (10), CD25 (11), and CD33 (12). Our own group has pursued an immunotoxin-based therapy (13, 14) for various target antigens including CD22, CD25, and CD30 (15–17). Recent clinical trials indicate that targeted therapy by recombinant immunotoxins show great promise in some types of hematologic malignancies, especially for treatment of hairy cell leukemia (HCL; refs. 18, 19). HCL is a malignancy of well-differentiated B lymphocytes (20, 21). HCL cells are detected in the blood as a CD103+, CD20+, CD11c+, and CD22+ cells (22).

Because of their potential as new targets for immunotherapy, we have produced monoclonal antibodies (MAb) to IRTAs and examined their expression on B cell malignancies as the first step in developing antibody-based therapeutics.

Here we describe the production of MAbs to the extracellular domain of IRTA2 and their use in the identification of the protein. It was previously reported that four isoforms of IRTA2 exist (1). The three major mRNA isoforms (ITRA2a, ITRA2b, and ITRA2c) encode proteins that share the first six extracellular Ig-like domains. The longest isoform, IRTA2c, which is the target previously discussed, has three additional extracellular Ig-like domains (nine total) with eight potential N-linked glycosylation sites, a 23-residue transmembrane region, and a 104-residue cytoplasmic domain. IRTA2a is predicted to be a secreted glycoprotein with eight extracellular Ig-like domains, all of which are shared with IRTA2c. IRTA2b has 32 additional residues in addition to the six common Ig-like domains and is supposedly attached to the plasma membrane with a GPI anchor. For the purpose of this paper, we will simply call the IRTA2c isoform IRTA2. The nine extracellular Ig-like domains in IRTA2 show high similarity to Ig domains in other IRTA members. In addition, the six domains near the plasma membrane of IRTA2 also show high homology with each other. Therefore, the cross-reactivity of anti-IRTA2 MAbs to other IRTAs was carefully examined.

In this paper, we describe the production of IRTA2-specific MAbs by the use of a recently developed DNA-immunization method (23). We show that the IRTA2 protein is expressed on the plasma membrane of many B cell lymphoma cell lines as well as on cells from patients with HCL.

Cells. We used 19 human hematologic cell lines in this study. They are JD38 (24), derived from B cell non–Hodgkin's lymphoma (B-NHL, undifferentiated); Karpas 1106 (25), B-NHL (mediastinal lymphoblastic); OCI-Ly2 (26), B-NHL (diffuse large cell type); OCI-Ly7 (26), B-NHL (diffuse large cell type); OCI-Ly10 (26), B-NHL (immunoblastic); SU-DHL-5 (27), B-NHL (diffuse large cell, noncleaved cell type); SU-DHL-6 (27), B-NHL (diffuse, mixed small and large cell type); CA46 (28), Burkitt's lymphoma (BL); Daudi, BL, American Type Culture Collection (ATCC, Manassas, VA, 20208, http://www.atcc.org) number CCL-213; NAMALWA, BL, ATCC CRL-1432; NC-37, BL, ATCC CCL-214; Raji, BL, ATCC CCL-86; Ramos, BL, ATCC CRL-1596; CCRF-SB, lymphoblastic leukemia, ATCC CCL-120; Karpas 299 (29), anaplastic large cell lymphoma; L540 (30), Hodgkin's disease (HD); MOLT-4, T-cell lymphoma, ATCC CRL-1582; RPMI 6666, HD, ATCC CCL-113 and U-937, histiocytic lymphoma, ATCC CRL-1593.2.

Daudi cells were grown in RPMI 1640 medium supplemented with 20% fetal bovine serum (FBS; HyClone, Logan, UT). Karpas 299 and L540 cells were grown in Iscove's modified Dulbecco's medium supplemented with 10% FBS. OCI-Ly2, OCI-Ly7, OCI-Ly10, and Karpas 1106 cells were kindly provided by Dr. Louis M. Staudt (National Cancer Institute, NIH) and maintained in Iscove's modified Dulbecco's medium supplemented with 15% normal human serum. The other cell lines were grown in RPMI 1640 medium supplemented with 10% FBS. For the plasmid transfection experiment, 293T cells were grown in DMEM supplemented with 10% FBS. Sp2/0-neo myeloma cells (23) were used as the fusion partner for hybridoma formation and maintained in Iscove's modified Dulbecco's medium with 15% FBS. Blood cells were obtained from patients with HCL via an approved protocol.

Primers. The DNA oligo primers are listed in Fig. 1A. The primers were synthesized by Lofstrand Laboratories (Gaithersburg, MD).

Plasmids. We cloned five different IRTA cDNAs (IRTA1, IRTA2, IRTA3, IRTA4, and IRTA5) and constructed a pair of expression vectors for each. One encodes the full-length protein under cytomegalovirus promoter and the other encodes the extracellular domain genetically fused with human IgG1 Fc portion. The Genbank accession numbers of reference sequences of IRTA1-5 are AF343659, AF343662, AF459027, AF459633, and AF459634, respectively.

The IRTA DNAs were amplified by PCR from cDNA of human spleen (Clontech, Palo Alto, CA) or from Daudi cells (prepared from total RNA using random hexamer and Superscript III Reverse Transcriptase; Invitrogen, Carlsbad, CA). The primers used for PCR are listed in Fig. 1A. The cDNA fragments were digested with the restriction enzymes listed in Fig. 1B and ligated into the multicloning site of the pcDNA3 plasmid (Invitrogen). For IRTA2, IRTA3, and IRTA4, two or three DNA fragments were cloned into the vector and connected in tandem in the same vector. Finally, full-length cDNAs of five different IRTAs were cloned under a cytomegalovirus promoter in pcDNA3.

For Fc-fusion proteins, a NotI-XbaI fragment of pRB-2k1-CD30 (23), which contains the hinge and Fc portion of IgG1 with three introns, was subcloned into pcDNA3. To prevent its possible interaction with IRTAs, the first eight amino acids of the constant region 2 of the Fc (amino acids 231-238) were changed from APELLGGP to APPVAGP by the use of GeneTailor Site-Directed Mutagenesis System (Invitrogen). These mutations (E233P, L234V, L235V, and deletion of G237) reduce the affinity of IgG1 for FcγRI to a nondetectable level (31). DNA fragments that encode the extracellular domains of the IRTAs, including the signal peptides, were amplified by PCR, and then inserted upstream of the human Fc in pcDNA3. The plasmids for the expression of human CD30 (pHRm30c; ref. 32) or the extracellular domain of the CD30-Fc fusion protein (pRB-2k1-CD30; ref. 17) were used as controls in some experiments.

Transfection of 293T Cells With the Plasmids. In a typical experiment, 5 × 10⁵ of 293T cells were seeded in a 10-cm dish (BD Biosciences, Bedford, MA) 24 hours prior to the transfection. Four micrograms of plasmid DNA were transfected by Lipofectamine and Plus reagent (Invitrogen) according to the manufacturer's instruction. After 48 to 72 hours transfection, the cells transfected with the plasmids encoding full-length IRTAs or CD30 were used for fluorescence-activated cell sorter (FACS) analysis, cell-ELISA, Western blot or immunofluorescence.

Preparation of the Fc-fusion Proteins. The 293T cells transfected with plasmids encoding Fc-fusion proteins secreted the Fc-fusion proteins in the supernatants. From days 2 to 8 after transfection, the Fc-fusion proteins in the media were harvested daily. The concentration of the Fc-fusion proteins was measured by a human IgG-specific sandwich ELISA using purified CD30-Fc as the standard. Several of the Fc-fusion proteins were purified with protein A-Sepharose 4 Fast Flow (Amersham Biosciences, Piscataway, NJ).

Production of MAbs. A 6-week-old female BALB/c mouse was DNA-immunized five times with the IRTA2 plasmid as previously described (23). The mouse was boosted with 293T cells transfected with the same plasmid, and 3 days later, the spleen cells were fused with SP2/0-neo myeloma cells as described previously (23). The hybridomas were screened for secretion of specific MAbs in an ELISA using IRTA2-Fc antigen. After multiple rounds of cell cloning by limiting dilution, the established hybridomas were grown to a high density to harvest the MAbs in the culture supernatants. The isotype of the MAbs was determined by mouse MAb isotyping reagents (ISO2; Sigma-Aldrich, St. Louis, MO). Ig concentrations in the culture supernatants were determined by a sandwich ELISA as previously described (33).

ELISA. In a typical indirect ELISA for the screening of hybridomas, microtiter plates (MaxiSorp; Nalge Nunc, Rochester, NY) were coated with 100 ng per 50 μL/well of goat anti-human IgG in PBS for 2 hours at room temperature. Next, 50 ng per 100 μL/well of IRTA2-Fc in blocking buffer (25% DMEM, 5% FBS, 25 mmol HEPES, 0.5% bovine serum albumin, 0.1% sodium azide in PBS) were added to each well and incubated for 2 hours at room temperature. After washing with PBS containing 0.05% Tween 20, 50 μL of the hybridoma supernatants were added and incubated for 2 hours at room temperature. After washing, the bound MAbs were detected by a 2-hour incubation with horseradish peroxidase–labeled goat anti-mouse IgG (Jackson ImmunoResearch, West Grove, PA) followed by tetramethylbenzidine substrate kit (Pierce, Rockford, IL). In some experiments, other IRTA-Fc fusion proteins were used instead of IRTA2-Fc.

Affinity Determination. Affinities (dissociation constant, K_d) of the anti-IRTA2 MAbs to IRTA2-Fc were determined by ELISA as previously described (34). Briefly, each MAb was incubated with various concentrations of IRTA2-Fc at room temperature for 20 hours to reach equilibrium, then the amount of MAb remained unbound at equilibrium was measured by indirect ELISA as described above.

Cell-ELISA. 293T cells that had been transfected with the IRTA2 expression plasmid or a CD30 expression plasmid 48 hours prior to use were seeded into 96-well plates (2 × 10⁴ well). After overnight culture, the cells were fixed with 3.7% formalin in the culture medium at room temperature overnight. Incubation with a dilution of MAbs, secondary antibody, and substrate was done as described in the ELISA protocol except that 1% FBS in PBS was used for washing.

FACS Analysis. Various lymphoma cell lines and transfected 293T cells were used in the FACS analysis. Typically, 4 × 10⁵ cells were incubated with 250 ng of MAb in 100 μL of PBS containing 5% FBS and 0.1% sodium azide. After incubation for 1 hour at 4°C, the cells were washed twice with the same buffer and incubated with 1:100 dilution of R-phycoerythrin (R-PE)–labeled goat anti-mouse IgG F(ab′)₂ (BioSource, Camarillo, CA) for 1 hour. After washing twice, the cells were suspended in 1 mL buffer, and the fluorescence associated with the live cells was measured using a FACSCalibur Flow Cytometer (Becton Dickinson, Franklin Lakes, NJ).

For blood samples of patients with HCL or normal donors, whole peripheral blood mononuclear cells (PBMC) were isolated using Ficoll-Paque (Amersham Biosciences) and stained with mixtures of F56-R-PE (200 ng/mL, specially made by Molecular Probes, Eugene, OR), CD20 PerCP-Cy5.5, and CD22-FITC, or CD103-FITC (Becton Dickinson). In some experiments, IRTA2-Fc or CD30-Fc was added at 20 μg/mL with the antibodies as the competitors. Compensations of fluorescence were set prior to each analysis using calibrated beads (CaliBRITE beads, Becton Dickinson). In some experiments, the fluorescence associated with the cells was converted to the number of IRTA molecules bound to the cell using QuantiBRITE beads (Becton Dickinson).

Immunofluorescence Antibody. 293T cells were transfected with an IRTA2 or a CD30 plasmid; 48 hours after transfection, the cells were detached, counted, and combined in a 1:1 ratio. A total 5 × 10⁴ cells were seeded in each well of a four-well chamber slide (Nalge Nunc). After 24 hours, the cells were washed with PBS and fixed in 3.7% formaldehyde for 5 minutes. After washing with PBS twice and blocking with 5% normal goat serum, the cells were incubated with a mixture of anti-IRTA2 MAb F56 (IgG1) and anti-CD30 MAb Ki-1 (IgG3) at 10 μg/mL each. After washing, a mixture of Alexa 488-conjugated goat anti-mouse IgG1 (1:1,000 dilution; Molecular Probes) and Alexa 555-conjugated goat anti-mouse IgG3 (1:1,000 dilution; Molecular Probes) was added to the cells. The primary antibodies, F56 and Ki-1, on the slides were probed by the different secondary reagents with different wavelengths of the emission based on the different subclasses. Nuclei were stained with 4′,6-diamidino-2-phenylindole (1 μg/mL; Roche Applied Science, Indianapolis, IN). Negative controls were done without antibodies.

Western Blotting. Typically, 40 μg of cell lysates or 20 ng of Fc-fusion proteins for each well or strip were separated onto 4% to 20% SDS polyacrylamide gels (Bio-Rad, Hercules, CA) under reducing conditions. Proteins were transferred to a polyvinylidene difluoride membrane (0.2 μm, Immuno-Blot; Bio-Rad) in transfer buffer [25 mmol Tris-HCl, 192 mmol glycine, 20% (v/v) methanol (pH 8.3)] at 4°C for 1 hour at 250 mA. After blocking with 5% skim milk in PBS, the membrane was incubated with 2.5 to 10 μg/mL of MAbs. The bound MAbs were detected with alkaline phosphatase-labeled goat anti-mouse IgG (BioSource) and 5-bromo-4-chloro-3-indolyl phosphate/p-nitroblue tetrazolium chloride substrate (Pierce).

RT-PCR. The RNA expression of IRTA2 in the cell lines was examined by a RT-PCR method. Total RNA was prepared from each cell line using Trizol LS reagent (Invitrogen). cDNA was made from 4 μg of the total RNA by Superscript III Reverse Transcriptase (Invitrogen) with random hexamer priming according to the manufacturer's instructions. PCRs were done using cDNA from 100 ng RNA and two sets of primers for β-actin and IRTA2 (c isoform–specific) listed in Fig. 1A one with 26 cycles of amplification (95°C, 15 seconds; 55°C, 30 seconds; 72°C, 1 min).

Cloning of IRTA cDNAs and Preparation of Fc-Fusion Proteins. Because of the high homology of IRTA2 to other family members, it was important to prepare all possible IRTA proteins to evaluate the cross-reactivity of anti-IRTA2 antibodies. All cDNAs of these IRTAs whose mRNA expression had been reported were amplified by RT-PCR and cloned into a pair of vectors, one for the expression of the full-length of the genes under a cytomegalovirus promoter, and the other for the expression of the extracellular domains as fusion proteins using the Fc portion of human IgG1. Figure 1B summarizes the structure of each IRTA and the plasmid construction. All full-length proteins were expressed on 293T cells transiently transfected with each plasmid as shown by cross-reactive MAbs (see below). All the Fc-fusion proteins were secreted into the culture supernatants of 293T cells transiently transfected with each plasmid, although the production level varied (0.4-10 μg/mL). The purified Fc-fusion proteins migrated in SDS-PAGE with the expected sizes (Fig. 1C).

Production of MAbs. A BALB/c mouse was DNA-immunized with the full-length IRTA2 expression plasmid and boosted with 293T cells transfected with the same plasmid (23). The fusion experiment resulted in 36 stable hybridomas secreting MAbs that reacted to the IRTA2 extracellular domain-Fc but not to CD30-Fc. We obtained 4 (11%) IgG1s, 24 (67%) IgG2as, and 8 (22%) IgG2bs. All the MAbs reacted to IRTA2-Fc with high affinities (1.1-18.7 nmol/L of K_ds, 5.4 nmol/L in average). In addition, all the MAbs reacted to IRTA2 expressed on 293T cells transfected with IRTA2 expression plasmid without background reactivity to 293T cells transfected with a CD30 plasmid (data not shown). We conclude that all the MAbs recognize the native conformation of IRTA2 on the plasma membrane with high affinity. We chose three MAbs (F25, F56, and F119) for further study based on their high affinities, different subclasses, lack of cross-reactivity with other IRTA family members, and reactivity in immunocytochemistry and Western blotting (see below). The characteristics of the three MAbs are summarized in Table 1 

Cross-reactivity of the Anti-IRTA2 MAbs. To examine the cross-reactivity of the MAbs with other IRTAs, we tested the reactivity of each MAb at a saturated concentration to other IRTA-Fc fusion proteins in an ELISA and to native IRTAs expressed on transfected 293T cells. Overall results indicate that 25 out of 36 MAbs specifically reacted with IRTA2, whereas 11 MAbs showed cross-reactivity with other IRTAs to various extents (data not shown). As shown in Fig. 2, the binding of MAbs, F25, F56, and F119 was specific to IRTA2 in both assays.

Reactivity of the MAbs to Denatured IRTA2. We next examined the reactivity of the MAbs in a cell-ELISA, by immunocytochemistry, and by Western blotting. In these methods, most of the epitopes on IRTA2 protein were probably denatured.

First, we tested the reactivity of the MAbs to formalin-fixed antigen on transfected 293T cells in a cell ELISA. About two-thirds of MAbs (21/36) lost their reactivity to formalin-fixed IRTA2 (titers were < 1), but 15 MAbs bound to the formalin-fixed antigen to some extent (data not shown). Only MAb F56 specifically reacted to IRTA2 with a good titer (Table 1). We then did an immunofluorescence study using MAb F56 for the detection of IRTA2 in the transfected cells (Fig. 3A). Two dishes of 293T cells were separately transfected with IRTA2 or CD30 plasmids and were mixed at a 1:1 ratio and reseeded. The cells were fixed in formaldehyde and double-stained with a mixture of anti-IRTA2 MAb F56 (IgG1) and anti-CD30 MAb Ki-1 (IgG3). Four views with different filters of the same field are shown (Fig. 3A). The cells expressing IRTA2 and stained with F56 showed a green fluorescence signal on the plasma membrane (anti-IRTA2 F56). In contrast, different cells expressing CD30 and stained with anti-CD30 showed a red signal on the plasma membrane (CD30). The nuclei of the cell were visualized with 4′,6-diamidino-2-phenylindole. The results indicate that MAb F56 can be used in immunofluorescence staining of cells after fixation with formalin to detect recombinant IRTA2 protein expressed on the plasma membrane.

Next, we examined the reactivity of the MAbs to SDS-denatured antigen in Western blotting. A mixture of IRTA2-Fc and CD30-Fc fusion proteins were blotted on the membrane and incubated with each MAb. Only with MAbs, F25, and F20, was a 160-kDa band of IRTA2-Fc visualized (Fig. 3B). These MAbs did not react with CD30-Fc on the same blots. The CD30-Fc band around 110-kDa was stained with an anti-human Fc antibody and an anti-CD30 MAb (Ber-H2), but not with F20 or F25. Therefore, the reactivity of F20 and F25 on the Western blot is specific. None of the other 34 MAbs were able to react with IRTA2 in Western blotting, indicating that they could detect native but not SDS-denatured IRTA2.

Expression of IRTA2 on Human Cell Lines. We surveyed IRTA2 expression on a variety of human cancer cell lines by FACS using MAbs F56 and F119 (Fig. 4A). IRTA2 was detected on seven B cell lymphoma lines, six were derived from different types of non–Hodgkin's lymphoma (NHL). Only one of six Burkitt's lymphoma cell lines was positive and this weakly expressed IRTA2. All cells were stained with the appropriate CD markers used as the positive controls and were not stained with mouse IgG1 as the negative control. To verify that the staining was specific, a 100-fold excess of IRTA2-Fc or CD30-Fc was used as the competitors of the staining. IRTA2-Fc was able to block staining of SU-DHL-5 and OCI-LY7 cells, whereas CD30-Fc was ineffective (data not shown).

The expression of IRTA2 mRNA in cells was examined by RT-PCR (Fig. 4B). A set of primers (2cF and 2cR Fig. 1A) for IRTA2 located on separate exons was designed to give a 432-bp PCR fragment. This primer set should detect only the spliced mRNA for the full-length of the transmembrane type of IRTA2 (IRTA2c) that was used as the antigen. As shown in Fig. 4B, the IRTA2 primers generated DNA fragments with expected size from the RNA of the cells positive for IRTA2 expression in FACS. In contrast, no IRTA2 cDNA was amplified from the RNA of FACS-negative cells, except for the trace amounts of the PCR product in Ramos cells. The actin primers generated similar amounts of the DNA fragment from all cells.

Next, we examined IRTA2 protein expression in several cell lines by Western blotting using MAb F25 (Fig. 4C). The IRTA2 protein was detected in the three B cell NHL (B-NHL) cell lines tested (SU-DHL-5, SU-DHL-6, and JD38) as bands around 150 kDa, which is the same size as the recombinant IRTA2 protein expressed in 293T cells transfected with an IRTA2 plasmid. MAb F25 also reacted to IRTA2-Fc in the same blot. In contrast, no IRTA2 bands were detected in Karpas 299 and L540 cell lysates.

Figure 4D shows a summary of the detection of IRTA2 by different methods in various cell lines. There is a clear correlation in IRTA2 expression detected by FACS, mRNA expression detected by RT-PCR, and IRTA2 protein detected by Western blots.

Expression of IRTA2 on HCL Cells from Patients. To determine if the MAbs might be useful for examining cells from patients, we analyzed IRTA2 expression in PBMCs of five patients with HCL and of two normal donors by FACS. In these experiments, the PBMCs were three color–stained with anti-IRTA2 F56 and CD20 and CD22 or with F56 and CD20 and CD103. We found that the HCL cells (CD20+, CD22+, and CD103+) from all five patients significantly expressed IRTA2, whereas we did not detect IRTA2 on PBMCs of normal donors (on CD14+, CD3+ or CD19+ cells). Figure 5 shows typical results of a FACS analysis of two patients with HCL. Staining with anti-IRTA2 F56 revealed that 27% or 39% of the PBMCs from the two patients expressed significant amounts of IRTA2 (Fig. 5A,, histograms). IRTA2 was not detected in PBMCs from normal donors (data not shown). The IRTA2+ cells also expressed CD20, CD22 and CD103 (Fig. 5A,, blue dot plots), indicating that IRTA2 positive cells represent HCL cells. On the other hand, HCL-maker positive cells from the same patients (CD20+/CD22+ or CD20+/CD103+) expressed IRTA2 (Fig. 5B). These data show that all HCL cells in these patients express IRTA2. The IRTA2 signal was inhibited by the addition of a 100-fold excess amount of IRTA2-Fc (red lines, Fig. 5B) but not by CD30-Fc (blue lines), indicating that IRTA2 staining of the patient samples was antigen-specific. The number of HCL cells in PBMCs of the other three patients was low (4-8% of the PBMCs) but the levels of IRTA2 expression were similar to the two samples presented (data not shown).

We have identified for the first time the intrinsic IRTA2 protein on the surface of many human lymphoma cell lines and on HCL cells from patients by the use of newly developed MAbs specific for IRTA2. Our results indicate that these MAbs may be useful for leukemia diagnosis and raise the possibility of IRTA2-targeted immunotherapy.

Although we isolated many MAbs reacting with IRTA2, we mainly used MAb F56 in the FACS analysis. The staining of MAb F56 was always in accordance with that of MAb F119, although these MAbs belong to different subclasses (F56, IgG1; F119, IgG2b) and react with different epitopes (formalin-resistant epitope for F56 and formalin-sensitive epitope for F119). These differences in subtype and epitope recognition reduce the chance of artifacts specifically related to a single MAb. In addition, the staining was specifically inhibited by IRTA2-Fc, confirming the specificity of the staining.

By using highly specific staining, we detected IRTA2 expression on leukemia cells from five out of five HCL patients, whereas no IRTA2-positive cells were detected in PBMC from normal donors. Also, all IRTA2-positive cells expressed the markers for HCL (CD20+, CD22+, and CD103+) and all HCL cells were positive for IRTA2. This complete accordance between IRTA2-positive cells and HCL cells imply that IRTA2-targeted diagnosis and immunotherapy of HCL could be highly specific. We did not detect any significant IRTA2+ cell population in PBMC from normal donors (data not shown), which is consistent with the interpretation that only small subsets of B cells such as follicular mantle naive-B, memory B and plasma cells in the germinal center light zone and in intraepithelial and interfollicular regions of lymph node expressed IRTA2 mRNA (3, 5). The restricted expression of IRTA2 on the small population of normal cells along with its high expression on the malignant cells could be useful for IRTA2-targetted immunotherapy.

We detected IRTA2 expression of eight lymphoma cell lines out of 19 cell lines examined; these are mainly B-NHL cells. Only one B-NHL cell line, OCI-Ly2, did not express IRTA2 by unknown reasons. The expression profile was confirmed by RT-PCR and Western blotting, indicating that the FACS analysis using the MAbs is highly specific. The expression of IRTA2 in OCI-LY7, OCI-LY10, and Karpas 1106 is consistent with mRNA expression detected by DNA microarray.²

The five cell lines included in our panel (U937, MOLT-4, Ramos, NAMALWA, and Daudi) had been previously tested for mRNA expression by Northern blotting (1). This data is almost consistent with ours except that they detected IRTA2 mRNA expression in Ramos and CA46 cells. We detected trace expression of IRTA2 mRNA in Ramos cellsby RT-PCR but not by FACS; thus, the discrepancy maybe explained by a difference in sensitivity of different methods.

The importance of IRTA2 in human cancer remains to be studied. IRTA2 has one potential immunoreceptor tyrosine-based activation motifs and two immunoreceptor tyrosine-based inhibitory motifs on the cytoplasmic domain, which are involved in intracellular kinase signaling cascade for cell growth and death (3–6). Also, IRTA2 is on chromosome 1q21, where one of the loci that is frequently rearranged in B lymphoid malignancy is found (3–6). The rearrangement could alter IRTA2 protein expression and affect cell growth.

According to AceView http://www.ncbi.nih.gov/IEB/Research/Acembly/index.html, National Center for Biotechnology Information, on 3/30/2004), five different protein isoforms of IRTA2 were suggested by 72 cDNA sequences. We detected the longest form, IRTA2c, but we have not yet determined the localization of epitopes of our MAbs. Epitope mapping of these MAbs may enable us to develop assays for the secreted isoforms of IRTA2, such as a sandwich ELISA, and may clarify the presence and role of the soluble forms.

We produced 36 anti-IRTA2 MAbs in this study. Out of the 36, 11 cross-reacted with other family members. This high ratio of cross-reactive MAbs was expected based on the high homology in the alignment of IRTAs by ClustalW (35). We used all IRTA recombinant proteins to check the cross-reactivity; 25 IRTA2-specific MAbs do not cross-react with other family members. Among the 25 IRTA2-specific MAbs, MAbs F56 (IgG1) and F25 (IgG2a) are particularly useful because they detect formalin-fixed antigen in cell-ELISA and immunofluorescence antibody, and SDS-denatured antigen in Western blots, respectively. Using an immunofluorescence technique, F56 specifically stained the plasma membranes of the 293T cells transfected with full-length IRTA2 plasmid (Fig. 3A). This result directly shows the plasma membrane localization of IRTA2 protein which had been predicted by the amino acid sequence.

Because we detected IRTA2 protein expression on HCL cells and many different lymphoma cell lines, we are now examining IRTA2 expression on cells and tissues of patients with lymphomas and other hematologic malignancies. We expect that IRTA2 will be established as a new target for immunotherapy.

We thank Drs. Louis M. Staudt and Eric R. Davis, National Cancer Institute, NIH, for the cell lines, OCI-LY2, OCI-LY7, OCI-LY10, and Karpas 1106 and for unpublished microarray data of IRTA2 expression in various lymphomas; Wyndham Wilson, National Cancer Institute, NIH, for critical reading of the manuscript; Susan Garfield and Stephen Wincovitch, Confocal Microscopy Core Facility, National Cancer Institute, NIH, for technical assistance in confocal microscopy; Inger Margulies and Karen Bergeron for their assistance in collection and preparation of blood cells from patients; Gail McMullen for her assistance in animal work; and Anna Mazzuca for editorial assistance.