A cell surface molecule, JL1 - Nature

2 downloads 0 Views 109KB Size Report
Acute biphenotypic. 9. 6 (63.3 ± 18.2). 66.7 leukemia. CML in blast crisis. 4. 1 (42.0). 25.0. CLL of B cell lineage. 1. 0 (0). 0. Adult T cell leukemiab. 1. 1 (58.0).
Leukemia (1998) 12, 1583–1590  1998 Stockton Press All rights reserved 0887-6924/98 $12.00 http://www.stockton-press.co.uk/leu

A cell surface molecule, JL1; a specific target for diagnosis and treatment of leukemias WS Park1, YM BAE1,2, DH Chung1, TJ Kim1, EY Choi1, J-K Chung3, MC Lee3, SY Park4, MH Park5 and SH Park1 Departments of 1Pathology, 3Nuclear Medicine, 4Internal Medicine, and 5Clinical Pathology, Seoul National University College of Medicine, Seoul, Korea

We previously reported a novel differentiation antigen, which is specifically expressed in stage II double positive (CD4+CD8+) human cortical thymocytes (Park et al, J Exp Med 1993; 178: 1447–1451). This study was designed to investigate the expression pattern of JL1 in various types of leukemic cells from patients and normal hematopoietic cells to evaluate the possibility as a tool for diagnosis and treatment of leukemia. The expression of JL1 antigen was observed in 75.6% of leukemic cases (117 out of 154 leukemic patients tested) on flow cytometric analysis. The percentage of JL1-positive cases of T lineage acute lymphoblastic leukemia (T-ALL) (92.6%) was higher than that of other types of leukemias (75%). The presence of JL1 antigen was also confirmed by immunoblotting and immunoprecipitation. Since the JL1 antigen is selectively expressed on the surface of human leukemic cells but not on the mature human peripheral blood cells, normal bone marrow cells and various types of normal tissues, JL1 could be an excellent candidate for an immunodiagnostic and immunotherapeutic tool for hematopoietic malignancies such as leukemia. Keywords: JL1 antigen; T cell differentiation antigen; pan-leukemic marker; immunodiagnosis; immunotherapy

Since this glycoprotein of approximately 120 kDa was also found to be expressed in the T cell leukemic cell lines, extensive investigation by FACS for the expression of JL1 in cells obtained from 154 patients showing various types of leukemia has been done. To our surprise, this glycoprotein of approximately 120 kDa was highly expressed in a majority of tumor cells, but not in normal cells and tissues other than cortical thymocytes. Our immunoblotting analysis using anti-JL1 antibody confirmed the presence of the 120 kDa protein specific to JL1 in samples from leukemia patients, but not in those normal individuals. Our data strongly suggest that the JL1 mAb (monoclonal antibody) can be used as a reagent of choice in the routine diagnosis of hematopoietic malignancies and provide an excellent candidate for the treatment of these diseases. Materials and methods

Patients and controls Introduction Over the past decades, significant advances have been achieved in the immunodiagnosis and immunotherapy of leukemia and lymphoma. Ideally, the antibodies against tumorspecific antigens for immunodiagnosis and immunotherapy should specifically target tumor cells, and have minimal crossreactivity with normal tissues, because nonspecific bindings will reduce the diagnostic and therapeutic effects.1 Many investigators have searched for tumor-specific antigens, of which several such as CALLA, CD20, and CEA have been applied to diagnostic and therapeutic purposes.2–5 A few antigens which are specifically expressed in malignant melanoma have been suggested as being a candidate for the treatment of malignant melanoma.6–9 However, there is no single tumorspecific antigen of leukemia and lymphoma which completely satisfies all the requirements described above. Among the series of monoclonal antibodies (mAbs) against T cell surface antigens, those against CD1 family react specifically to immature cortical thymocytes and malignant lymphoblastic T cells.10,11 However, the antigens of CD1 family are known to also be expressed in extrathymic cells including Langerhans cells of the skin, dendritic cells in lymph nodes and B cell subsets.12–15 We have previously reported a novel human thymocyte differentiation antigen whose expression is restricted to stage II double-positive (CD4+CD8+) human cortical thymocytes.16 Correspondence: SH Park, Department of Pathology, Seoul National University College of Medicine, 28 Yongon-dong, Chongno-gu, Seoul 110–799, Korea; Fax: 011 822 763 6625 2 Current address: Section of Immunology, Yale School of Medicine, New Haven, CT, USA Received 1 December 1997; accepted 23 June 1998

One hundred and fifty-four patients who had been hospitalized and diagnosed as leukemia at Seoul National University Hospital were included in this study. The patients underwent diagnostic work that included peripheral blood examination, bone marrow aspiration and biopsy. Clinical information important for the diagnosis of leukemia was available from clinical records. At the time of diagnosis, information was obtained about presenting clinical features, tumor cell morphology, histochemistry and immunophenotype. The morphologic diagnoses of leukemia were based on the criteria of the French–American–British (FAB) cooperative group.17 All cases were studied at the time of diagnosis before treatment except for two cases of AML in relapse and four cases of CML in blast crisis. The control group consisted of peripheral blood samples from 186 normal blood donors and volunteers and 20 bone marrow samples from healthy donors for bone marrow transplantation and patients with malignancies such as neuroblastoma, Ewing’s sarcomas, and metastatic carcinomas with no tumor involvement. The protocol for sample collection was approved by the institutional review board at this hospital, and all patients and normal volunteers gave informed consent.

Cell line and samples A tumor cell line, Molt-4 was obtained from ATCC (American Type Culture Collection, Rockville, MD, USA). The cells were cultured in RPMI 1640 medium containing 10–20% (v/v) fetal calf serum (FCS), 2 mm glutamine, penicillin (100 units/ml), and streptomycin (100 mg/ml) at 37°C in a humidified atmosphere of 5% CO2/95% air and used for following study at optimal growth condition. A total of 154 leukemic cases were collected and the diag-

Leukemia-associated antigen, JL1 WS Park et al

1584

noses and numbers of cases are summarized in Table 1. All leukemia cases were diagnosed by the morphological, immunological, and cytochemical results. For immunophenotyping of leukemia, mononuclear cells from peripheral blood or bone marrow were stained with a panel of monoclonal and polyclonal antibodies as follows: TdT (BRL, Gaithersburg, MD, USA), Smlg (Cappel, Malvern, PA, USA), Leu9 (CD7; Becton Dickinson, San Jose, CA, USA), T1 (CD5; Coulter, Hialeah, FL, USA), T11 (CD2; Coulter), 12 (HLA-DR; Coulter), B4 (CD19; Coulter), J5 (CD10; CALLA, Coulter), MY9 (CD33; Coulter), MY7 (CD13; Coulter) and GP IIIa (CD61; Dako, Carpinteria, CA, USA). Peripheral blood from 186 healthy blood donors and 20 bone marrow samples from normal healthy donors and patients with solid tumors and no marrow involvement were also included. Fresh cell suspensions of mononuclear cells were prepared from bone marrow or peripheral blood specimens (EDTA anticoagulated) with the use of Ficoll–Hypaque (Pharmacia, Piscataway, NY, USA) density gradient technique. These fresh cell suspensions were used for flow cytometric or biochemical analysis immediately or after preservation in liquid nitrogen. Some of the peripheral blood cells from healthy donors were cultured in vitro as described above for variable periods with or without mitogen (PHA (GIBCOBRL) 10 ␮g/ml or PWM (Boehringer Mannheim Biochemicals, Frankfurt, Germany) 3 ␮g/ml) and underwent flow cytometric analysis. Additionally, fresh thymuses were collected from the patients who had portions of their thymuses removed during corrective cardiac surgery. After the specimens were placed in RPMI 1640 medium, blood clots and fibrous capsules were carefully removed and single cell suspensions were prepared.

Immunofluorescence staining and flow cytometric analysis Fresh cell suspensions from patients and healthy donors were submitted for the flow cytometric analysis of JL1 expression using anti-JL1 mAb. They were stained by the indirect immunofluorescence method with purified anti-JL1 mAb, followed by FITC-conjugated goat anti-mouse immunoglobulin. Indirect immunofluorescence study was performed by incubating 106 cells with saturating amounts of the purified antiJL1 mAb or controlled immunoglobulin in phosphate-buffered saline containing bovine serum albumin (0.2%) and sodium azide (0.05%) for 1 h at 4°C. After three washes, 50␮l of the fluorescein-conjugated goat anti-mouse IgG (Cappel) was added and incubated for 30 min at 4°C. For two-color flow cytometric analysis, this procedure was followed by the incubation of cells with PE-conjugated mAb. Analysis was performed with FACScan (Becton Dickinson). In order to keep the setting and histograms uniform during the studies, the instrument was calibrated with calibration beads (CaliBRITE beads; Becton Dickinson), and all positive-negative cutoff values set at the 99th percentile of isotype-matched negative control immunoglobulin. The data were recorded on a Hewlett-Packard computer (Meriden, CT, USA), histograms were generated and the percentage of fluorescent cells as well as the mean fluorescence intensity (mean channel number) was calculated. Cells were considered positive for the JL1 antigen when fluorescence intensity was above the upper limit of the negative control. Cases with over 20% JL1-positive cells were regarded as positive cases based on the immunoblotting results.

Monoclonal antibodies The murine monoclonal antibody, anti-JL1, was produced as previously described.16 The ascitic form of anti-JL1 mAb was produced after injection of hybridoma cells into pristane-pretreated Balb/c mice. The antibody was purified using Qsepharose (Pharmacia) and hydroxylapatite (Bio-gel HTP gel; Pharmacia) column chromatography. Anti-CD16 antibody, phycoerythrin (PE)-conjugated antiCD34 antibody, and the antibodies for immunotyping were commercially purchased (Becton Dickinson). Table 1 FACS profile of anti-JL1 mAb immunofluorescence in leukemias

Type of leukemia

T-ALL non-T-ALLa AML Acute biphenotypic leukemia CML in blast crisis CLL of B cell lineage Adult T cell leukemiab Total

No. of cases tested

27 50 62 9 4 1 1 154

No. of Percent JL1-positive cases of JL1positive cases 25 41 43 6

(83.1 ± 16.6)c (72.7 ± 19.3) (61.5 ± 29.7) (63.3 ± 18.2)

92.6 82.0 69.4 66.7

1 (42.0) 0 (0) 1 (58.0)

25.0 0 100

117

75.6

a CALLA-positive or negative cases (one case was CALLA-negative and JL1-positive). b HTLV-related case. c Average percent of positive cells ± standard deviation.

Immunoblotting Fresh cell suspensions of bone marrow or peripheral blood samples from leukemic patients and some healthy donors, thymocytes (positive control), and Molt-4 cells were lysed with 1% NP-40 in 50 mM Tris-HCl, pH 7.4, 50 mM EDTA, and 1 mM PMSF (phenyl methyl sulfonyl fluoride). The lysates were mixed gently by slowly inverting the tube, incubated at 4°C for 30 min and centrifuged at 13 000 g for 15 min to remove the unlysed fraction. The supernatant was subjected to 8% SDS-PAGE under the reduced condition, with appropriate molecular weight markers. The electrophoretic transfer of proteins to the nitrocellulose paper was done at 45 V for 16 h. After the proteins transfer, the nitrocellulose was blocked with 5% skim milk in Tris-buffered saline (10 mM Tris-HCl, 150 mM NaCl, pH 7.6) containing 0.05% Tween-20 (TBST). The antigens were detected by incubating the nitrocellulose with antiJL1 mAb. After washing with TBST, it was incubated for 1 h at room temperature with peroxidase-conjugated goat antimouse IgG (Zymed, San Francisco, CA, USA) diluted 1:5000 in blocking solution. After washing with TBST, the bound peroxidase was visualized using the Amersham chemiluminescence (ECL) detection system (Amersham, Arlington Heights, IL, USA). Restaining of the nitrocellulose was done as above after detaching the bound antibodies by treating the nitrocellulose with 62.5 mM Tris buffer (pH 6.7) containing 100 mM ␤-mercaptoethanol and 2% SDS for 30 min at room temperature, followed by 0.2 mM glycine buffer (pH 2.8) containing 500 mM NaCl, 0.05% NaN3 for 30 min at 55°C.

Leukemia-associated antigen, JL1 WS Park et al

Immunoprecipitation To examine the expression pattern of JL1 antigen on the cell surface of leukemic cells, bone marrow cells and PHA-activated lymphocytes, a series of immunoprecipitations was performed as reported previously.16 Briefly, the cells were surface-radiolabeled with 125I using lactoperoxidase-catalyzed iodination, and were lysed in the buffer (50 mM Tris-HCl, pH 7.4, 50 mM EDTA, 1 mM PMSF) containing 1% (v/v) Nonidet P-40. After preclearing by mixing the lysate with irrelevant immunoglobulin-conjugated protein A-sepharose CL-4B beads, immunoprecipitations were done by mixing lysate with anti-JL1 mAb-conjugated protein A-sepharose CL-4B beads. The mixture was incubated overnight at 4°C. The immunoprecipitates were electrophoresced in 8% SDS-PAGE. The dried gels were exposed to Kodak XAR X-ray film (Kodak, Rochester, NY, USA) for 3 days at −70°C.

Scatchard analysis on binding affinity of Anti-JL1 mAb The affinity constant of the antibody was determined by Scatchard analysis, using a cell binding assay. In 200 ␮l, one million Molt-4 cells, which express JL1, were incubated with 3.6–300 ng of 125I-labeled antibody. Nonspecific binding was measured by adding 25 ␮g of unlabeled antibody. After 2 h incubation, cells were spun down and the cell-bound radioactivity was counted. Activity of the specific binding was calculated by subtracting nonspecific binding from the total binding. The slope of a plot is a measure of the affinity constant (or equilibrium constant) of the reaction of the antibody with target antigen. From the x-axis intercept, the number of binding sites of the antibody per cell was calculated.18,19

lated from the z value. Two-tailed test results were considered statistically significant at P ⬍ 0.01. Results

Selective expression of JL1 antigen in leukemic cells Samples of acute lymphoblastic leukemias and acute myelogenous leukemias showed positive immunofluorescence of variable degrees. We defined JL1-positive leukemia when more than 20% of cells from patients expressed JL1, which was confirmed by immunoblotting. Since weakly JL1-positive cells usually made a single peak in the histogram, we concomitantly performed immunofluorescent microscopic examination with several weak positive cases. Almost all leukemic cells were weakly positive for JL1 while control PBL samples were completely negative. Therefore, we thought that most leukemic cells express JL1 even in weak positive cases. Most cases of T-ALL showed the JL1 positivity (92.6%), and the average percentage of positive cells was 83.1 ± 16.6% in JL1positive T-ALL. The proportion of JL1-positive cases in the acute non-T lymphoblastic leukemia (non-T-ALL) (82.0%) was slightly lower than that of T-ALL. The average percent of positive cells in non-T-ALL was 72.7 ± 19.3%. Of 50 non-T-ALL cases, 49 cases were CALLA-positive and one CALLA-negative case was positive for JL1. Seventy percent of AML cases was JL1-positive. The results of the flow cytometric analysis of the cells from leukemic patients are summarized in Table 1 and the representative FACS profiles are shown in Figure 1. There

Digestion with endoglycosidase To evaluate the presence of N-glycosylation, the immunoprecipitates from leukemic cells, thymocytes, and Molt-4 cells were eluted by boiling with 100 mM potassium phosphate buffer (pH 7.0) containing 0.2% SDS. After adding the same amount of 200 mM potassium phosphate buffer (pH 7.0) containing 40 mM EDTA, 2% NP-40 and 20 mM beta-mercaptoethanol, 0.25 U endoglycosidase-F (endo-F; Boehringer Mannheim Biochemicals) was added. The mixtures were incubated for 18 h at 37°C. The digests were mixed with sample buffer, separated on 8% SDS-PAGE gels, and transferred to nitrocellulose for Western blotting with anti-JL1 mAb. For the evaluation of the presence of O-glycosylation, the immunoprecipitates were treated with 10 milliunits neuramidase (Boehringer Mannheim Biochemicals) and 100 milliunits O-endoglycosidase (Boehringer Mannheim Biochemicals) simultaneously. To remove the N-linked and O-linked glycosylation, the immunoprecipitates were treated with three enzymes simultaneously.

Statistical analysis The ␹2 test was used to compare ectopic expression of lineage-specific markers between JL1-positive and -negative cases and the odds ratio was calculated. Expression frequency of JL-1 and other lineage-specific markers in each type of leukemia was compared by z-test and significance was calcu-

Figure 1 Representative flow cytometric analysis of leukemic cell suspensions from patients with T-ALL (a), non-T-ALL (b) and AML (c). Dotted lines are all negative controls using isotype-matched irrelevant monoclonal antibody. The fluorescence intensity of T-ALL is slightly stronger than that of AML or CALLA(+) B-ALL but not in general.

1585

Leukemia-associated antigen, JL1 WS Park et al

1586

was a lineage-dependent difference in JL1 expression between types of leukemia in terms of fluorescence intensity. We found a consistently higher mean fluorescence intensity for staining in T-ALL, whereas non-T-ALL and AML showed fluctuation in their fluorescence intensity. JL1 staining in 76% (19 out of 25) of T-ALL cases had a mean fluorescence channel intensity of over 100, which has been regarded as a high intensity group, as compared to 43.9% (18/41) and 30.2% (13/43) of JL1-positive non-T-ALL and AML cases, respectively. A case of chronic myelogenous leukemia in blast crisis, of which three were in myeloid and one in megakaryocytic crisis, was JL1-positive (25%). One HTLV-related adult T cell leukemia was also JL1-positive. To confirm the flow cytometric results biochemically, immunoblotting was performed for cells from selected leukemic cases. As shown in our previous results, normal human thymocytes expressed a protein band of approximately 120 kDa (lanes 1 and 2, Figure 2a). Fifteen representative leukemic cases, in which leukemic cells were available for immunoblotting, were analyzed. Of them three cases (lanes 7, 12 and 16), which individually represents type of leukemia, showed high relative antigen index on flow cytometric analysis and the others low or intermediate. Leukemic cells showed clear bands with a slight variation of their sizes around 120 kDa, indicating the existence of JL1 molecule (lanes 3–7; CALLA-positive non-T-ALLs, lanes 8–14; AMLs, lanes 15–17; T-ALLs, Figure 2a). Three samples from one AML and two TALL cases (lanes 11, 15 and 17) showed a very faint protein band. Immunoprecipitation of those samples of weak JL1 density on immunoblotting, using a large amount of radiolabeled cell lysates to concentrate this molecule, again confirmed the 120 kDa bands corresponding to JL1 antigen (Figure 2b). These findings strongly support the fact that the positive flow cytometric data even in leukemic cells of the lowest intensity were not due to nonspecific binding to the other cell surface molecules such as Fc receptor but by engagement of JL1 molecules with this antibody.

Complete absence of JL1 in normal peripheral blood and bone marrow cells On the immunofluorescence and flow cytometric analysis, 183 peripheral blood preparations of 186 healthy donors were negative to the anti-JL1 mAb immunofluorescence. It was not possible to perform direct immunofluorescence study due to frequent localization of lysine residue in the Fab portion of anti-JL1 immunoglobulin, which was confirmed by RT-PCR and DNA sequencing analysis of anti-JL1 antibody coding

gene in hybridoma cell line. This is why the evaluation of the presence of JL1 antigen in the remaining three samples from healthy donors which showed weak positive reaction to the anti-JL1 mAb (Figure 3a) were performed by immunoblotting and immunofluorescence analysis. Instead of the 120 kDa molecule, bands of about 50 kDa were clearly visualized (Figure 3b) and these molecules were proved to correspond to the Fc receptor by detaching the antibodies and restaining of the nitrocellulose paper with anti-CD16 mAb (Figure 3c). The results of immunoblotting analysis disclosed that the weak positivity was due to nonspecific binding of anti-JL1 mAb to Fc receptors on the lymphocytes. The peripheral blood mononuclear cells were also JL1-negative even after a long-term in vitro culture of more than 20 days with mitogenic stimulation. Twenty bone marrow samples from nonleukemic patients were all negative for JL1 antigen, which has also been confirmed by immunoprecipitation and immunoblotting as well as flow cytometry. These data, however, could not rule out the possibility that JL1-positive cells actually constitute the fractional percentage of CD34-positive stem/progenitor cells which usually escape the single histogram-flow cytometric detection. It was clearly demonstrated in a reproducible way that JL1-positive cells were not present within the CD34-positive cell compartment which is a minor population of total bone marrow cells (Figure 4).

Ectopic expression of other markers in leukemic cells and their relationship with JL1 antigen expression The ectopic expression of lineage-specific markers in acute leukemias is well known, especially in AML.20–27 Therefore, we evaluated the ectopic expression of the lineage-specific markers in the leukemic cases. Twenty-one (48.3%) of 43 JL1positive AML cases, and nine (47.4%) of 19 JL1-negative cases expressed ectopic lymphoid markers which were CD2, CD5, CD7, CD10, CD19 or TdT (terminal deoxytransferase). The ectopic expression of T cell or myeloid markers in non-T-ALLs was observed in 11 cases (22%). No significant correlation was found in the ectopic expression of lineage-specific markers between JL1-positive and -negative cases (P = 0.57 in TALL, P = 0.64 in non-T-ALL and P = 0.87 in AML). The most frequently expressed ectopic markers were CD7 (8%) and CD33 (8%) in non-T-ALL and CD7 (21.0%) in AML. However, the expression of JL1 in non-T-ALL and AML cases was significantly higher than that of the ectopically expressed markers (z values; 13.6 in non-T-ALL and 7.3 in AML) (Table 2).

Figure 2 Immunoblots of bone marrow cell suspensions from CALLA(+) non-T-ALL (lanes 3–7), AML (lanes 8–14), and T-ALL (lanes 15–17) stained by anti-JL1 mAb, showing bands of very varied molecular weights from 120 to 160 kDa (a). Molt-4 and thymocytes (lanes 1 and 2, 130 kDa and 120 kDa, respectively) also show different bands. The protein (JL1) bands are absent or faint in lanes 11, 15 and 17. Immunoprecipitation of 125I-labeled cell lysates of those samples with low JL1 density on immunoblotting was performed (b), and prominent bands corresponding to JL1 antigen are demonstrated in all three samples.

Leukemia-associated antigen, JL1 WS Park et al

1587

Figure 3 The weak JL1-positivity of three healthy donors on immunofluorescence study was due to non-specific binding of antibody to Fc receptors but not presence of this molecule. The flow cytometric profile of the three healthy donors are presented in (a). The numbers 3, 4 and 5 represent the patients of lanes 3–5 in (b) and (c), respectively. Lanes 1 and 2 were Molt-4 and thymocytes, respectively, as positive control. The nitrocellulose papers were stained with anti-JL1 mAb (a) and anti-CD16 Ab (b).

Table 2 Comparison of JL1 expression with other ectopically expressed markers in non-T-ALL and AML

Type of leukemia

JL1 TdT CD2 CD5 CD7 CD10 CD19 CD13 CD33 CD61

non-T-ALL (50 cases)

AML (62 cases)

41

43 8 10 5 13 0 5

2 4 3 4 0

Figure 4 No JL1 expression in CD34-positive stem/progenitor bone marrow cells. Bone marrow cells from a healthy donor were analyzed for JL1 expression in CD34-positive stem/progenitor cell population by double-color immunofluorescence. The percentages shown represent the proportion of cells in each quadrant of the fluorescence gates.

Glycosylation of JL1 protein causing size variation To evaluate the possibility that the size variation of JL1 antigen in leukemic cells was due to the differences in glycosylation, immunoprecipitation and digestion with endoglycosidase were done in three selected cases. When digested with endoF, the molecular weight of JL1 antigen was slightly reduced but still showed differences in their molecular sizes (Figure 5a). These differences were also the same in cases of simultaneous treatment of endo-F and O-glycosidase as well as in O-glycosidase treatment only (Figure 5b, c).

Figure 5 Immunoprecipitation and immunoblotting of bone marrow cells from leukemic patients (samples from patients of lanes 9, 12 and 13 in Figure 3). The immunoprecipitates were treated with endo-F (a), neuramidase and O-glycosidase (b) or the three enzymes simultaneously (c). After treatment with the three enzymes, there are slight shiftings of the bands but they still do not show the same molecular weight.

Leukemia-associated antigen, JL1 WS Park et al

1588

Cell binding affinity of anti-JL1 mAb The results of Scatchard analysis are shown in Figure 5. The affinity constant of the 125I-labeled anti-JL1 mAb was 1.69 × 109 l/mol. The binding sites of the antibody per Molt4 cell were 0.54 × 104 (Figure 6). Discussion We have shown that JL1 is specifically expressed on the surface of leukemic cells, but not on that of peripheral lymphocytes and bone marrow cells from healthy donors. In addition, complete absence of this antigen has been confirmed through extensive screening of red blood cells, platelets and natural killer cells (data not shown). Our FACS analysis showed that more than 75% of leukemia patients with T- and non-T-ALL, AML and acute biphenotype leukemia strongly reacted with anti-JL1 antibody. The percentage of JL1-positive cases was higher in T-ALL (92.6%) than in non-T-ALL (82.0%) or AML (69.4%) (Table 1). Comparing the mean fluorescence intensity, T-ALL showed high mean fluorescence channel intensity of ⬎100 in 76% of cases as compared to 43.9% and 30.2% in non-T-ALL and AML, respectively. Fluctuation of mean fluorescence intensity in cases of non-T-ALL and AML was more evident than that in cases of T-ALL which had consistently high mean fluorescence intensity. It is known that none of the surface antigens hitherto studied are exclusively leukemia-specific. JL1 is unique among various leukocyte markers in that the proportion of JL1-positive cases in non-T-ALL and AML is rather higher than that of leukemic cases expressing other ectopic markers (Table 2), and also that a majority of JL1-positive T-ALL and some non-T-ALL and AML express high levels of this antigen. In addition, two parameter flow cytometric data addressing the complete absence of JL1-positive cells within the CD34-positive com-

partment clearly indicate that JL1-positive cells have no stageor lineage-relationship with CD34-positive stem/precursor cells (Figure 2). These data suggested that, at least in some cases, the expression of this gene may have a role in the evolution of leukemia. The molecular weights of JL1 antigen recognized by antiJL1 mAbs in several leukemic patients were widely variable from 120 to 160 kDa. This observation suggested that the variation of molecular weight would be caused by either the complexity of glycosylation of this molecule, which might be different among the various leukemic cells, or the differences in sizes of protein backbone of JL1 antigen, which could be associated with alternative splicing or genetic mutations during leukemogenesis. There was a slight reduction of molecular weight after deglycosylation using Endo-F, but there still remained differences in their size in the Molt-4 and thymocytes. Even after deglycosylation with combined treatment of Endo-F and O-endoglycosidase, the molecular weights of JL1 antigens in leukemic cells were not the same. Although it was not the purpose of this study to elucidate the pathogenetic implication of the size variation of JL1 molecule in individual leukemia cases, these observations support the idea that variation of the JL1 molecule in individual leukemic cases might not be due to the post-translational event but due to the actual differences in protein core. However, the molecular analysis of JL1 is to be pursued further. The advent of monoclonal antibody technology in the 1980s led the way to massive research efforts to generate panels of monoclonal reagents specific for both normal differentiation-associated as well as tumor-specific antigens. Monoclonal antibodies have been shown to have clinical utility in the histologic assessment and classification of malignant diseases, particularly for leukemias and lymphomas, and have applicability in screening procedures and monitoring clinical disease status. In addition, mAbs offer the possibility of selective delivery of antineoplastic agents to the tumor site, thereby

Figure 6 Determining the affinity (equilibrium) constant and binding site of radiolabeled anti-JL1 monoclonal antibody. The proportion of antibody specifically bound to Molt-4 cell is plotted against the concentration of bound antigen in a Scatchard analysis. From the slope of this plot, the affinity (equilibrium) constant can be obtained. The binding site of the antibody per cell is calculated from the value of x-intercept.

Leukemia-associated antigen, JL1 WS Park et al

reducing systemic exposure and toxicity to normal tissue components. Therefore, mAbs alone or mAbs conjugated to various radioisotopes, toxins or drugs may have potential therapeutic applications.1–9,28–31 However, the clinical utility of many mAbs is debatable, because virtually all apparent tumor antigens could be found either to be expressed on subsets of normal cells at a particular developmental stage or ubiquitously on normal cells at lower levels than were expressed on tumor cells. To overcome these limitations, many investigators have made efforts to search for the tumorspecific antigens exclusively expressed on the tumor cells. For effective immunotherapy using radiolabeled antibodies1 or immunotoxins such as angiogenin28 or ricin29 the specificity and affinity of mAbs plays a key role. Anti-JL1 mAb showed an excellent binding affinity with an affinity constant in Molt-4 of 1.69 × 109 l/mol. We believe that JL1 could be used as an excellent target for diagnosis and therapy of leukemias. As our results demonstrated, the JL1 antigen is highly and specifically expressed on the cell surfaces of most human leukemic cells but not on normal human tissues and cells, with a sole exception of the immature cortical thymocytes. To our knowledge, there has been no surface protein showing this high specificity for tumor cells. Anti-JL1 mAb is selectively reactive with the antigen in spite of the differences in the molecular weight which might be associated with structural variation of the antigen, and the binding affinity of anti-JL1 mAb is relatively high. Currently, we are not only exploring the possibility of using this antibody for therapeutic purposes, but also searching for other mAbs against JL1 antigen with high affinity.

Acknowledgements This study was supported by a grant from HanWha Corporation Pharmaceutical Division.

9

10

11 12

13

14

15

16 17 18

References 1 Scott AM, Welt S. Antibody-based immunological therapies. Curr Opin Immunol 1997; 9: 717–722. 2 Oshimi K, Seto T, Oshimi Y, Masuda M, Okumura K, Mizoguchi H. Increased lysis of patient CD10-positive leukemic cells by T cells coated with anti-CD3 Fab⬘ antibody cross-linked to antiCD10 Fab⬘ antibody. Blood 1991; 77: 1044–1049. 3 Kaminski MS, Zasadny KR, Francis IR, Fenner MC, Ross CW, Milik AW, Estes J, Tuck M, Regan D, Fisher S, Glenn SD, Wahl RL. Iodine-131-anti-B1 radioimmunotherapy for B-cell lymphoma. J Clin Oncol 1996; 12: 1974–1981. 4 Mach JP, Pelegrin A, Folli S, Buchegger F. Radiolabeled monoclonal antibodies as anti-tumor missiles, their diagnostic success and therapeutic potential. Bull Acad Natl Med 1992; 176: 879– 889. 5 Begent RH, Verhaar MJ, Chester KA, Casey JL, Green AJ, Napier MP, Hope-Stone LD, Cushen N, Keep PA, Johnson CJ, Hawkins RE, Hilson AJ, Robson L. Clinical evidence of efficient tumor targeting based on single-chain Fv antibody selected from a combinatorial library. Nature Med 1996; 2: 979–984. 6 Houghton AN. Cancer antigens: immune recognition of self and altered self. J Exp Med 1994; 180: 1–4. 7 Real FX, Mattes MJ, Houghton AN, Oettgen HF, Lloyd KO, Old LJ. Class I (unique) tumor antigens of human melanoma. Identification of 90 000 dalton cell surface glycoprotein by autologous antibody. J Exp Med 1984; 160: 1219–1233. 8 Kawakami Y, Eliyahu S, Sakaguchi K, Robbins PF, Rivoltini L, Yannelli JR, Appella E, Rosenberg SA. Identification of the immunodominant peptides of the MART-1 human melanoma antigen recog-

19 20

21

22

23

24

25

nized by the majority of HLA-A2 restricted tumor infiltrating lymphocytes. J Exp Med 1994; 180: 35–42. Brichard V, Van Pel A, Woelfel T, Wolfel C, De Plaen E, Lethe B, Coulie P, Boon T. The tyrosinase gene codes for an antigen recognized by autologous cytolytic T lymphocytes on HAL-A2 melanomas. J Exp Med 1993; 178: 489–495. Cattoretti G, Berti E, Parravicini C, Buscaglia M, Cappio F, Caputo R, Cerri A, Crosti L, Delia D, Gaiera G, Polli N. Expression of CD1 molecules on dendritic cells: ontogeny, epitope analysis on normal and malignant cells, and tissue distribution. In: Knapp W, Do¨rken B, Gilks WR, Rieber EP, Schmidt RE, Stein H, Kr von dem Borne AEG (eds). Leukocyte Typing IV. White Cell Differentiation Antigens. Oxford University Press: London, 1989, pp 262–264. Park SH, Mahoney RJ, Given SR, Fajardo MA, Dubey DP, Yunis EJ. Serological identification of thymocyte differentiation antigens. Hum Immunol 1988; 22: 151–162. Small TN, Keever CA, Knowles RW, O’Reilly RJ, Flomenberg N. CD1c expression during normal B-cell ontogeny. In: Knapp W, Do¨rken B, Gilks WR, Rieber EP, Schmidt RE, Stein H, Kr von dem Borne AEG (eds). Leukocyte Typing IV. White Cell Differentiation Antigens. Oxford University Press: London, 1989, pp 265–266. Ringler DJ, Hancock WW, King NW, Murphy GF. Characterization of nonhuman primate epidermal and dermal dendritic cells with monoclonal antibodies. A study of Langerhans cells and indeterminate cells in the rhesus monkey. Lab Invest 1987; 56: 313– 320. Amiot M, Bernard A, Raynal B, Knapp W, Deschildre C, Boumsell L. Heterogeneity of the first clusters of differentiation: characterization and epitope mapping of three CD1 molecules on normal human thymus cells. J Immunol 1986; 136: 1752–1758. Bernard A, Boumsell L, Reinherz EL, Nadler LM, Ritz J, Coppin H, Richard Y, Valensi F, Dausset J, Flandrin G, Lemerle J, Schlossman SF. Cell surface characterization of malignant T cells from lymphoblastic lymphoma using monoclonal antibodies: evidence for phenotype differences between malignant T cells from patients with acute lymphoblastic leukemia and lymphoblastic lymphoma. Blood 1981; 57: 1105–1110. Park SH, Bae YM, Kwon HJ, Kim TJ, Kim J, Lee SJ, Lee SK. JL1, a novel differentiation antigen of human cortical thymocyte. J Exp Med 1993; 178: 1447–1451. Bennett JM, Catovsky D, Daniel MT, Flandrin G, Galton DA, Gralnick HR, Sultan C. Proposals for the classification of the acute leukemias. Br J Haematol 1976; 33: 451–458. Eckelman WC. The development of single-photon emitting receptor-binding radiotracers. In: Nunn AD (ed). Radiopharmaceuticals, Chemistry and Pharmacology. Marcel Dekker: New York, 1992, pp 167–219. Scatchard G. The attractions of proteins for small molecules and ions. Ann NY Acad Sci 1949; 52: 660–672. Pui CH, Raimondi SC, Head DR, Schell MJ, Riversa GK, Mirro J Jr, Crist WM, Behm FG. Characterization of childhood acute leukemia with multiple myeloid and lymphoid markers at diagnosis and at relapse. Blood 1991; 78: 1327–1337. Mirro J, Zipf TF, Pui CH, Kitchingman G, Williams D, Melvin S, Murphy SB, Stass S. Acute mixed lineage leukemia: clinicopathologic correlations and prognostic significance. Blood 1985; 66: 1115–1123. Kuerbitz SJ, Civin CI, Krischer JP, Ravindranath Y, Steuber CP, Weinstein HJ, Winich N, Ragab AH, Gresik MV, Crist WM. Expression of myeloid-associated and lymphoid-associated cellsurface antigens in acute myeloid leukemia of childhood: a pediatric oncology group study. J Clin Oncol 1992; 10: 1419–1429. Uckun FM, Muraguchi A, Ledbetter JA, Kishimoto T, O’Brien RT, Roloff JS, Gajl-Peczalska K, Provisor A, Koller B. Biphenotypic leukemic lymphocyte precursors in CD2+CD19+ acute lymphoblastic leukemia and their putative normal counterpart in human fetal hematopoietic tissues. Blood 1989; 73: 1000–1115. Kasparu H, Koller U, Kreiger O, Nowotny H, Tuchler H, Lutz D. Significance of gp40/CD7 or TdT positivity in AML patients. In: Knapp W, Do¨rken B, Gilks WR, Rieber EP, Schmidt RE, Stein H, Kr von dem Borne AEG (eds). Leukocyte Typing IV. White Cell Differentiation Antigens. Oxford University Press: London, 1989; pp 936–937. Schwonzen M, Niestroj W, Seckler W, Diehl V, Pfreundschuh M. Clinical significance of T-cell-related antigens in acute myelomon-

1589

Leukemia-associated antigen, JL1 WS Park et al

1590

ocytic and monocytic leukemias. In: Knapp W, Do¨rken B, Gilks WR, Rieber EP, Schmidt RE, Stein H, Kr von dem Borne AEG (eds). Leukocyte Typing IV. White Cell Differentiation Antigens. Oxford University Press: London, 1989, p 938. 26 Claxton DF, Reading CL, Nagarajan L, Tsujimoto Y, Andersson BS, Estey E, Cork A, Huh YO, Trujillo J, Deisseroth AB. Correlation of CD2 expression with PML gene breakpoints in patients with acute promyelocytic leukemia. Blood 1992; 80: 582–586. 27 Pui CH, Schell MJ, Vodian MA, Kline S, Mirro J, Crist WM, Behm FG. Serum CD4, CD8, and interleukin-2 receptor levels in childhood acute myeloid leukemia. Leukemia 1991; 5: 249–254. 28 Haagen IA, Geerars A, de Lau WB, Clark MR, van de Griend RJ, Bast BJ, de Gast BC. Killing of autologous B-cell lineage malignancy using CD3×CD19 bispecific monoclonal antibody in end stage leukemia and lymphoma. Blood 1994; 84: 556–563.

29 van Oosterhout YV, Preijers FW, Wessels HM, de Witte T. Cytotoxicity of CD3-ricin a chain immunotoxins in relation to cellular uptake and degradation kinetics. Cancer Res 1992; 52: 5921– 5925. 30 Rybak SM, Hoogenboom HR, Meade HM, Raus JC, Schwartz D, Youle RJ. Humanization of immunotoxins. Proc Natl Acad Sci USA 1992; 89: 3165–3169. 31 Jurcic JG, Caron PC, Nikula TK, Papadopolous EB, Finn RD, Gansow OA, Miller WH Jr, Geerlings MW, Warrell RP Jr, Larson SM, Scheinberg DA. Radiolabeled anti-CD33 monoclonal antibody M195 for myeloid leukemias. Cancer Res 1995; 55 (Suppl.): 5908s–5910s.