Epithelial Mucin-1 (MUCI) Expression and MA5 Anti-MUC1 Monoclonal Antibody Targeting in Multiple Myeloma I

Vol. 5, 3065s-3072s, October 1999 (Suppl.) Clinical Cancer Research 3065s

J. Burton, z D. Mishina, T. Cardillo, K. Lew, cGy/mCi of injected dose compared with 3099 cGy/mCi of A. Rubin, D. M. Goldenberg, and D. V. Gold tumor-absorbed dose delivered by nonspecific antibody. Garden State Cancer Center, Belleville, New Jersey 07109 [J. B., D. M., T. C., K. L., D. M. G., D. V. G.], and St. Joseph's Hospital and Introduction Medical Center, Paterson, New Jersey 07503 [A. R.] MM 3 is a B-cell malignancy that appears to result from the transformation and monoclonal expansion of a cell with char- acteristics of a plasma cell, i.e., a terminally differentiated B cell Abstract (1, 2). The expression by MM cells of certain non-B-cell anti- Multiple myeloma (MM) is the second most common gens also raises the possibility of the transformation of an hematological cancer in the United States. It is typically earlier, more multipotent lymphoid precursor cell. As with nor- incurable, even with myeloablative chemotherapy and stem- mal plasma cells, this degree of terminal differentiation is as- cell transplantation. The epithelial mucin-1 (MUC1) glyco- sociated with the complete or partial loss of certain B-cell- protein is expressed by normal and malignant epithelial cells associated antigens, such as surface immunoglobulin and CDs but has also been shown to be expressed by MM cells. MUC1 19-22 (CD19 is expressed on normal plasma cells; Ref. 3). In is a useful antigenic target in solid tumors for clinical diag- addition, there is the acquisition of selected non-B-cell-specific nostic and therapeutic monoclonal antibody (mAb)-based markers, such as CD28, CD38, and CD56; epithelial antigens approaches. The MA5 mAb, as well as other anti-MUC1 such as syndecan-1 (CD138) and epithelial glycoprotein-2; cer- mAbs reactive with the MUC1 variable number tandem tain myeloid antigens; and the antigen addressed in this report, repeat domain, exhibited moderate to strong reactivity with MUC1 (4-11). MUC1 is of particular interest because it is a both MM cell lines and clinical samples. To explore the diagnostic marker that exists as an integral membrane glyco- biochemical nature and potential of MUC1 as an antigenic protein and as shed forms and because it is a therapeutic target target in MM, studies were performed to: (a) compare the for both cell-mediated and antibody-based immunotherapeutic mRNA and the MUC1 glycoprotein species between epithe- strategies. Although no function has yet been determined for lial cancer and MM cell lines; and (b) develop and use a membrane-bound MUC1, a receptor/ligand-binding property human MM tumor xenograft model system to study the has been associated with neoplastic progression and cellular biodistribution of the MA5 mAb. MA5 mAb was strongly adhesion (12). Although most of the work to date with this target reactive with six of eight human MM cell lines by flow antigen has been in breast and pancreatic cancers, a few reports cytometry. In seven of eight MM patient samples (bone have shown that MUC1 can be detected in the serum of MM marrow and/or peripheral blood) reactivity was found in patients and, thus, may also represent a surrogate tumor marker 10-90% of the cells, whereas normal control (n = 5) and in MM (10). leukemia and lymphoma (n = 5) cells showed only 0-6% Progress in the therapy of MM has been achieved over the reactivity, lzsI-labeled MA5 whole-cell binding studies past 25 years with the introduction of chemotherapy regimens, showed quantitatively similar amounts of binding between resulting in objective antitumor effects (1, 2, 13-15). Some strongly positive MM lines and high-MUCl-expressing additional therapeutic benefit has resulted from the more recent breast carcinoma lines, mRNA expression was assessed by application of high-dose chemotherapy with autologous or al- Northern blotting and reverse transcription-PCR. MM cell logeneic stem-cell rescue (14, 15). Nonetheless, MM remains lines were positive by both methods, with strong similarity in fairly resistant to all of these approaches in the majority of cases. the sizes of the mRNAs and cDNAs that were obtained. Long-term survival, in fact, has not improved much since the Finally, biodistribution experiments were carried out with advent of chemotherapy with phenylalanine mustard (Melpha- 13~I-labeled MA5 versus a nonbinding control 12SI-labeled lan) and glucocorticoids. Drug resistance mechanisms via mul- mAb in a s.c. MM xenograft model. Selective MM tumor tidrug resistance protein 1 and lung resistance-related protein uptake of the MA5 mAb was demonstrated, with a potential have been described in MM and appear to be major mechanisms for delivering a tumor radiation absorbed dose of 8540 of resistance to therapy (16-18). Multidrug resistance protein 1 expression may be intrinsic to both malignant and nonmalignant

1 Presented at the "Seventh Conference on Radioimmunodetection and Radioimmunotherapy of Cancer," October 15-17, 1998, Princeton, NJ. 3 The abbreviations used are: MM, multiple myeloma; CD, cluster of Supported in part by USPHS Grants CA54425 and CA39841 from the differentiation; MUC1, epithelial mucin-1, polymorphic epithelial mu- NIH. cin or epithelial membrane antigen; RAIT, radioimmunotherapy; mAb, 2 To whom requests for reprints should be addressed, at Garden State monoclonal antibody; ATCC, American Tissue Culture Collection; RT- Cancer Center, 520 Belleville Avenue, Belleville, NJ 07109. Phone: PCR, reverse transcription-PCR; VNTR, variable number tandem re- 973-844-7024; Fax: 973-844-7020. peat; %iD/g, percentage of injected dose per gram.

plasma cells because cell surface expression of this transporter Newark, N J). Mononuclear cells were isolated from bone mar- protein was found in 42 of 43 samples, which included 19 row by density gradient centrifugation (Cell-Sep; Larex, Inc., St. patients with either monoclonal gammopathy of undetermined Paul, MN). significance or amyloidosis (17). Lung resistance-related protein RT-PCR and Northern Blotting. The starting template expression was shown to be an important factor in determining material for RT-PCR was total RNA, which was isolated from response to standard-dose Melphalan-prednisone (18). The ex- PBS-washed cells that were solubilized with a guanidine iso- pression levels of the antiapoptotic protein BCL-XL has also thiocyanate-based buffer (Tri-Reagent; Sigma-Aldrich, St. been shown to play a role in determining chemoresistance, with Louis, MO). RNA was isolated according to a modification of increased levels leading to relative drug resistance (19). BCL-2 the method of Chomczynski and Mackey (29). Total RNA (5 may also play a role in chemoresistance in MM (20). Other Ixg) was used as the template for cDNA synthesis, using the factors, such as mutations in the ras oncogene or the p53 gene, First Strand kit of Novagen (Madison, WI) according to manu- are not common enough to play a major role in MM pathophys- facturer's instructions, with 4% of the resulting cDNA product iology (21, 22). The tendency toward chemoresistance, along being used as template for each PCR reaction. Primers and with the relative clinical radiosensitivity of MM (23-25), sug- dNTPs were added at standard concentrations, 0.5 and 200 lxM, gests that MM would be a favorable target for RAIT. The respectively. Thermostable DNA polymerase (0.5 p,l; KlenTaq; predominant involvement of the red marrow in MM lends Ab Peptides, St. Louis, MO) was added to each tube, and 35 further support to RAIT approaches because mAb uptake is both cycles of PCR were carried out under the following conditions: rapid and high in the bone marrow compartment. Indeed, pre- annealing temp of 65~ for 45 s, extension at 72~ for 30 s, and clinical experiments in immunotherapy and radioimmunotarget- denaturation at 94~ for 30 s (initial denaturation was at 94~ ing in a MM model system have shown potential for these for 2 rain). mAb-mediated approaches (26, 27). In addition, another ap- Total RNA (20 p~g) from selected cell lines was denatured proach has been developed to selectively deliver radiation doses in a 50% formamide-2.2 M formaldehyde solution, fractionated to the red marrow in MM through the use of a bone-seeking by electrophoresis through a 0.8% agarose gel containing 0.66 M radionuclide, 166H0. Initial clinical studies have demonstrated formaldehyde and transferred onto a nylon membrane (Hybond; that radiation doses of up to -5000 cGy can be safely delivered Amersham Corp., Chicago, IL) by capillary blotting. RNA was to the marrow cavity by this approach (28). fixed by baking at 80~ under vacuum. After prehybridization Given the above background, we sought in the studies at room temperature in a solution containing 5• SSC buffer reported herein to address selected questions regarding MUC1 plus 0.2 M sodium phosphate (pH 6.8), 1 • Denhardt's reagent, expression in MM that had not been fully addressed previously. 50% formamide, 100 txg/ml sheared salmon sperm DNA, 50 These questions are: (a) What mRNA species are expressed by pLg/ml yeast tRNA, and 20 Ixg/ml poly(A) n, the membrane was MM cell lines? (b) What is the general sequence structure of hybridized with the labeled MUC1 probe at 42~ for 14-16 h. MUCI cDNA in MM? (c) Is MUC 1 a target antigen that can be The probe was labeled with [oL-32p]-dCTP by the nick transla- effectively used in preclinical MM tumor xenograft biodistribution method, according to the manufacturer's instructions (Am- tion studies? ersham). The specific activity of the probe was -108 cpm/lxg. This MUC1 probe was a cDNA insert cut from pBS 42TR FMUC1 vector via digestion with EcoRI and BamHI followed Materials and Methods by gel purification of the desired fragment. This probe contained Cell Lines and Clinical Specimens. The MM cell lines 42 VNTR sequences (30). The probe was incubated with the RPMI 8226, U266, and MC/CAR were obtained from the Amer- blot at a concentration of 1 x 10 6 cpm/ml to 2 • 106 cpm/ml; ican Tissue Culture Collection (ATCC; Manassas, VA). The the membrane was subsequently washed under stringent condi- MM cell line, JJN-3, was kindly provided by Dr. M. Kuehl, tions prior to exposure to film. Navy-NCI Oncology Branch, Bethesda, MD. The KMS12-BM mAbs. The MA5 mAb was a murine IgG~, kindly pro- and KMS 12-PE cell lines were generously provided by Dr. T. vided by Immunomedics, Inc. (Morris Plains, NJ). This mAb Otsuki, Kawasaki Medical School, Okayama, Japan, and the has been characterized previously and has been studied clini- DUL4 cell line was kindly provided by the technical services cally as a radioantibody imaging agent for detection of breast department of Irvine Scientific (Santa Ana, CA). The MUC1 + cancer (31, 32). The other anti-MUC1 mAbs that were used breast carcinoma cell lines, MCF7, MDA-MB468, T47D, and (mostly for flow cytometry) were also murine IgGlS. These ZR-75-30, were obtained from ATCC for use as controls, as was were KC-4G3, H23 (obtained from ATCC), DF3-P (kindly the pancreatic carcinoma cell line, CaPan-1. The Ramos B- provided by Dr. D. Kufe, Dana-Farber Cancer Institute, Boston, lymphoma line was used a control cell line and was also ob- MA) and PAM4 mAb (33). The first three mAbs react with tained from ATCC. The ovarian carcinoma cell line, 2008, was epitopes within the VNTR region, whereas PAM4 reacts with an obtained from Dr. S. Howell (University of California, San epitope in the N-terminal, extracellular domain of MUC1. Diego, CA). These cell lines were maintained in RPMI 1640 Flow Cytometry. Cell lines were washed and set up at supplemented with 10% heat-inactivated FBS plus antibiotics, --0.5 X 106 cells/tube and incubated with 10 txg/ml control and except for the MC/CAR line, for which the medium preparation test mAbs for 30 rain at 4~ in flow buffer (PBS-3% FBS-0.1% recommended by ATCC was used. Clinical specimens consisted NAN3). After washing with the same buffer, the second step of either peripheral blood or bone marrow aspirates of patients reagent, pretitered F(ab')2 goat antimouse IgG-fluorescein iso- with active MM and were obtained through St. Joseph's Hos- thiocyanate (Biosource International, San Diego, CA), was pital, Paterson, N J, and Dr. A. Lippman (Beth Israel Hospital, added, and cells were incubated for an additional 30 rain at 4~

Table 1 MA5 (anti-MUC1) reactivity with cell lines (immunofluorescent flow cytometry). Cell line Cell type MA5 positive (%)a RPMI8266 Myeloma 98.3 ARH77 Myeloma 7.4 DUL4 Myeloma 4.0 H929 Myeloma 87.8 JJN-3 Myeloma 97.9 U266 Myeloma 84.3 Ramos B-cell lymphoma 0.5 Jurkat T-cell lymphoma 0.8 u:...,,, ..... MCF7 Breast carcinoma 57.4 T47D Breast carcinoma 96.6 MDA-MB-468 Breast carcinoma 98.5 a This value represents the percentage of cells positive for reactivity with MA5 after subtraction of value for binding of nonspecific antibody (Ag8).

Table 2 MA5 (anti-MUC1) reactivity with clinical specimens (immunofluorescent flow cytometry). MA5 positive Patient ID Diagnosis Specimen type (%)a 2 Myeloma Bone marrowb 25.5 6 Myeloma Bone marrow 14.3 1 Myeloma Peripheral bloodC 12.3 Fluorescence 2 Myeloma Peripheral blood 64.2 Fig. 1 Fluorescence histograms of cell populations stained with the 3 Myeloma Peripheral blood 18.9 MA5 mAb and an isotype/subclass-matched control. A, JJN-3 cell line 4 Myeloma Peripheral blood 14.9 with nonspecific control mAb; B, JJN-3 cell line with MA5 mAb; C, 5 Myeloma Peripheral blood 7.6 bone marrow aspirate from MM patient no. 2 with nonspecific control 7 Myeloma Peripheral blood 24.0 mAb: D, bone marrow aspirate from MM patient no. 2 with MA5 mAb. 8 Myeloma Peripheral blood 32.2 = 5 LymphomaJleukemia Peripheral blood 0.0-6.0 = 5 Normal Peripheral blood 0.0-0.2 a This value represents the percentage of cells positive for reactivity with MA5 after subtraction of value for binding of nonspecific (human serum albumin, the carrier protein used, served as antibody (Ag8). internal standard). 125I-labeled MA5 routinely showed <3% b Bone marrow mononuclear cells were isolated via density gradi- unbound iodide and >95% monomeric IgG with a specific ent centrifugation. c Buffy coats were isolated, and erythrocytes were lysed with 10 activity of 444-925 MBq/mg (12-25 mCi/mg). For cell-binding mM potassium bicarbonate buffer containing 154 mM ammonium chlo- experiments, 1 • 106 cells/tube in 100 ~1 of buffer were used. ride and 0.1 mM EDTA. To each tube, either binding buffer (RPMI 1640 supplemented with 10% FBS) or a 150-fold molar excess of unlabeled MA5 was added. Tubes were incubated for 15 min at 4~ after which the 125I-labeled MA5 was added to a final concentration of 0.75 The cells were then fixed with 1.5% paraformaldehyde in PBS Ixg/ml (saturating binding conditions). The binding reaction was and analyzed on a Becton Dickinson (Mountain View, CA) carried out for 90 min at 4~ after which cell-bound label was FACScan. separated from free label by centrifuging the cells (7000 • g for Radiotraeer Binding to Cell Lines. MA5 mAb was 75 s) through a 200-~1 cushion of 80% dibutyl phthalate-20% iodinated using Nal25I (New England Nuclear, Boston, MA). olive oil (Sigma-Aldrich, St. Louis, MO) in a 0.4-ml polyeth- The ~25I was activated with precoated Iodogen tubes (Pierce, ylene tube. The tips of the tubes were excised and counted in a Rockford, IL), according to the method of Chizzonite et al. (34). gamma counter. Briefly, 100 Ixl of 0.2 M sodium phosphate buffer (pH 7.4) was Western Blotting. After washing with PBS, approxi- added to a 1 mCi stock vial of Naa25I. The entire solution mately 10 v cells were solubilized with buffer consisting of 0.2% volume was transferred to a precoated Iodogen tube and acti- Zwittergent 3-12 in 0.5 M Tris-HC1 (pH 8.0) containing 0.15 M vated for 7 min at room temperature, after which the Na~25I NaC1, 2 mu phenylmethylsulfonyl fluoride, and 0.5 mM N-tosyl- solution was transferred to a 1.5-ml tube containing 20 Ixg of L-phenylalanine chloromethyl ketone. After solubilization at MA5 mAb in 25 Ixl of PBS. The iodination reaction was allowed 4~ for 60 min, nuclei and debris were pelleted at 10,000 • g to proceed for 7 min at room temperature, after which bound and for 5 min. Aliquots of these whole-cell lysates were loaded onto free iodine were separated on a PD-10 column (Pharmacia, a 4% SDS-PAGE gel and separated. The gel was then electro- Piscataway, NJ), which was equilibrated and eluted with PBS blotted (24 V for 16 h) onto a nitrocellulose membrane containing 0.2% gelatin. Radiolabeled mAb was analyzed by (Schleicher & Schuell, Keene, NH). After the membrane was both instant TLC and gel filtration high-performance liquid blocked with blocking/incubation buffer (PBS containing 0.5% chromatography, with radioactive and absorbance monitoring BSA and 0.2% Tween-20), MUC1 was detected by incubating

30000

25000

20000

II1 Fig. 2 Bar graph showing the specific cpm 15000 of 125I-labeledMA5 bound at saturation. The specific cpm are obtained by subtracting the cpm of a matched tube containing excess cold t,~ 10000 MA5 from the total cpm.

5000

MDA-MB468 JJN3 U266 H929 ARH-77 Cell Lines

the membrane successively with 5 Ixg/ml MA5 mAb in blocking extrapolated from the exponential curve. The resulting integral buffer for 1 h at room temperature, followed by three washes for each organ was converted to cGy/mCi (cGy/MBq) using S with blocking buffer. The blot was then incubated for 1 h with values appropriate for the radionuclide and organ weight. These a 1:5000 dilution of antimouse IgG-peroxidase (Jackson Labs, S values were derived by assuming uniformly distributed activ- West Grove, PA). The blot was then developed with the Super- ity in small, unit-density spheres (35). Signal system (Pierce), according to the manufacturer's instructions, and exposed to X-ray film. MM Xenograft Model and Biodistribution Studies. Results s.c. tumors were established by two approaches. One of these Anti-MUC1 VNTR-Region mAbs React with MM Cell was to expand the JJN-3 cell line in vitro, wash and resuspend Lines and Clinical Specimens. A panel of cell lines, consist- the cells in fresh medium (RPMI 1640 + 10% FBS), and inject ing of both hematopoietic cell lines and breast carcinomas, was the cells s.c. into the flanks of immunodeficient mice (10 • 106 analyzed by flow cytometry for reactivity with anti-MUC1 cells/0.3 ml to 15 • 10 6 cells/0.3 ml injected into each mouse). mAbs (Table 1). Six of eight MM cell lines tested demonstrated The mouse strains that were used were C.B-17 SCID and strong reactivity with the MA5 mAb as well as with other Cr:NIH(S) nu/nu. The other approach for establishing s.c. tu- anti-MUC1 mAbs with anti-VNTR specificity, including the mors was to excise established JJN-3 tumors (0.6-1 cm 3) and DF-3P mAb. The latter mAb is notable for the fact that it was dissect viable regions into individual pieces of --40 mg. These raised against a MUC1 VNTR-peptide immunogen. Testing of pieces were then implanted s.c. into anesthetized mice of the clinical samples (bone marrow and/or peripheral blood) from same strain followed by a skin-closing suture. When tumors MM patients with active disease showed moderate to strong derived from either approach reached a size of 0.3-0.6 cm 3, reactivity with the MA5 mAb in 88% (seven of eight), whereas tumor-bearing mice were co-injected with a mixture of 25 txCi of ~3~I-labeled MA5 and 10 txCi of a nonbinding control mIgG~ none of five normal control blood specimens showed reactivity [either P3X63Ag8 (MOPC-21) or anti-ot-fetoprotein] labeled with this antibody (Table 2). In addition, five leukemia and with 125I. Groups of five mice were injected and sacrificed at 1, lymphoma patient specimens (bone marrow and/or blood) gave 4, and 7 days post injection, at which time tumors and normal only weak (1-6%) positive cell responses. Representative his- tissues were harvested and counted, mAb uptake values were tograms are shown in Fig. 1. On a semi-quantitative basis, the determined and expressed as the %ID/g. Localization indices degree of fluorescence shift (i.e., the mean channel) was similar were calculated according to the formula: localization index -- to that observed with the breast carcinoma cell lines. (uptake of 131I-labeled MA5 in tumor/uptake of 131I-labeled Radiolabeled MA5 Shows Specific Binding to MM Cell MA5 in blood) + (uptake of control mAb in tumor/uptake of Lines. To confirm the findings from flow cytometry, MM lines control mAb in blood), with uptake units being %ID/g. Radia- as well as other MUC 1 + and MUC1- cell fines from the flow tion dose estimates were also determined from the biodistribu- cytometry panel were tested for the binding of 125I-labeled MA5 at tion data. Calculations were performed by first integrating the a saturating mAb concentration. A good correlation was found trapezoidal regions for tumors or the exponential regions for the between the mean fluorescence intensity in flow cytometry and normal tissues, as defined by the time-activity data. To avoid absolute specific radioantibody binding. The specific 125I-labeled overestimation of the cumulative radiation absorbed dose to MA5 cpm bound varied between MM lines. Those with the highest tumor, a zero time value of zero was assumed for this trapezoi- levels of binding had levels similar to those observed in high- dal fit method. For normal tissues, the zero time point was expressing MUC1 + breast carcinoma cell lines (Fig. 2).

M 123 1 2 3 4 5

2000bp 4.4kb >

2.0kb >

Fig. 3 Expression of MUC1 mRNA as assessed by RT-PCR. PCR products from cell lines that were separated by agarose gel electrophore- Fig. 4 Expression of MUC1 mRNA as assessed by Northern blotting. sis and stained with ethidium bromide are shown. Lane M, molecular mRNA from selected cell lines was separated by formaldehyde-agarose size markers; Lane 1, MDA-MB468 breast cancer; Lane 2, JJN-3 electrophoresis, blotted onto a membrane, and probed with a MUCI myeloma; Lane 3, U266 myeloma. VNTR cDNA probe. The resulting autoradiogram is shown. Lane I, Ramos B-lymphoma; Lane 2, MDA-MB468 breast cancer; Lane 3, JJN-3 myeloma; Lane 4, U266 myeloma; Lane 5, ARH-77 myeloma.

MUC1 mRNA Is Present as Assessed by Both RT-PCR and Northern Blotting. Total RNA was isolated from spec- 1 2 3 4 imens and was then analyzed by both methods. RT-PCR was performed initially with primers designated 262-forward (5'- TTG-AAT-GCT-CAC-AGC-CCC-GGT-TCA-GGC-TCC-3') and 1017-reverse (5'-TTT-GAA-TTC-CTA-TTC-AGA-AAT- GTG-TCT-CTG-3'), designed from GenBank accession no. J05582, human pancreatic mucin mRNA, complete coding se- 220 kDa --~ quence. These sequences are outside the VNTR and are far enough 5' and 3', respectively, of the flanking VNTR- 116 kDa degenerate sequences of MUC1 to allow for amplification of one predominant PCR product (Fig. 3). This amplicon was 1.8 kb and was present in all of the MUCI+ breast, ovarian, pancreatic, and MM cell lines tested. Other hematopoietic cell lines with no protein expression by flow cytometry and/or mAb tracer-binding studies were also negative by RT-PCR. In addi- Fig. 5 Expression of MUC1 glycoprotein as assessed by Western tion, PCR reactions were performed with a primer pair (both blotting. Cell lysates were fractionated by SDS-PAGE, electroblotted, 20-mers) spanning the 5' region of MUC1 from the translation probed with the MA5 mAb, and detected by chemiluminescence. Lane 1, Ramos B-lymphoma; Lane 2, ARH-77 myeloma; Lane 3, U266 initiation codon up to nucleotide 262. These PCR reactions myeloma; Lane 4, JJN-3 myeloma. likewise showed the presence of a single amplicon of the correct size with cDNA derived from the same cell lines (data not shown). kDa was observed, whereas cell lines that were negative for A Northern blot was performed using total RNA that was MUC1 mRNA expression or cell surface protein expression (by isolated from MM and other selected cell lines. This blot flow cytometry) were also negative by this method. The pan- showed bands at --7, 4.4, and 2.1 kb in both the MM and creatic carcinoma cell lines tested had strong signals in Northern MUC1 + epithelial cell lines, with no hybridization signal ob- blotting and RT-PCR, but these cell lines were nonreactive with served in cell lines that were negative by flow cytometry (Fig. MA5 in flow cytometry and in Western blot analysis. 4). The mRNA species that were observed were very similar in MA5 mAb Showed Selective Uptake into MM Xe- size to those described previously for MUC1. nografts in Nude Mice. A s.c. MM model using the JJN-3 Western Blot Analysis of MUC1 from MM Revealed cell line growing as a xenograft in the athymic nude mouse was Species of Similar Molecular Mass to Those Expressed by used in these experiments. Tissue uptake of radiolabeled anti- Epithelial MUCI+ Cell Lines. MM and control cell lines body, expressed as %ID/g, is shown in Fig. 6 for day 4, at which were detergent solubilized, and aliquots of the soluble extracts time point the maximum tumor uptake of MA5 (11.8% ID/g) were separated by SDS-PAGE followed by electroblotting onto was observed. Tumor/nontumor ratios for MA5 were always nitrocellulose membranes for immunodetection with the MA5 greater than for the nonspecific control P3• mAb. A mAb (Fig. 5). In all of the breast, ovarian, and MM cell lines tumor localization index of 3.2 was observed for a31I-labeled tested thus far, a predominant immunoreactive band of -400 MA5 at day 4, with nontumor tissue localization indices in the

[ 9 MA5 [] Ag8!

1 i J

o e~ 8 Fig. 6 MA5 uptake into JJN-3 (MM) xenografts in nude mice. The %ID/g for 131I-labeled MA5 "E' and 125I-labelednonbinding control isotype/sub- m 6 class-matched Ag8 mAb in JJN-3 tumors and ,.a e- normal organs at 4 days post injection are shown. 8 g~ 4

Tumor Liver Spleen Kidney Lungs Muscle Blood

range of 0.9-1.0. Biodistribution data were used to estimate protein. This variation in molecular mass stems mostly from potential radiation doses to the tumor of 8540 cGy/mCi (230.8 allelic variation in the number of VNTR units. MUC1 alleles are cGy/MBq) of injected dose for MA5 and 3099 cGy/mCi (83.8 usually expressed codominantly, which results in the presence cGy/MBq) for nonspecific control P3• a 2.8-fold ad- of two mRNA species of between 4 and 7 kb and two corre- vantage for MA5. Potential blood doses of 6479 cGy/mCi sponding glycoprotein species with molecular masses of 300 to (175.1 cGy/MBq) and 7706 cGy/mCi (208.3 cGy/mCi) of in- >1000 kDa. This additional size heterogeneity of the mature jected dose were noted for MA5 and P3• respectively. glycoprotein species results from heterogeneous glycosylation, Potential radiation doses to other tissues were all less than 3000 mostly of the O-type (41, 42). cGy/mCi (81.8 cGy/MBq). Our RT-PCR and Northern blotting results show that there is both sequence and structural relatedness of MUC1 mRNA from Discussion MM and MUC1 + epithelial cell lines. Whereas DNA sequence Consistent with previous reports using other anti-MUC1 data from the PCR amplicons obtained from MM lines are not yet mAbs, we showed reactivity of the MA5 anti-MUC 1 mAb with available, our data show that the resulting PCR fragments resulting MM cell lines and clinical specimens. Because this mAb reacts from both the 5' and the trans-VNTR primer pairs were very with the VNTR repeat sequence of MUC1, we considered it similar in MM and pancreatic, breast, and ovarian carcinomas. formally possible that this sequence may be shared by a related Northern blot analysis showed that the mRNA species of MM and gene that is expressed in MM and possibly other cancer types. carcinoma cell lines were of very similar molecular weights when We therefore proceeded to characterize both the MUC1 mRNA probed with a MUC1 VNTR probe. Likewise, Western blot anal- and glycoprotein species from MM and compared it to the yses showed similarity of the MUC1 glycoprotein species between corresponding MUC1 species in carcinomas, such as breast, MM and epithelial cell lines, using the MA5 mAb as probe. ovarian, and pancreatic. Quantitative estimation of MUC 1 expression showed that some of In earlier efforts to generate anti-epithelial tumor mAbs the MM lines showed absolute specific binding of 125I-labeled using tumor-derived immunogen preparations, MUC 1 proved to MA5 of a similar magnitude as high-expressing MUCI+ carci- be very antigenic. Thus, many anti-MUC1 mAbs have been noma cell lines at saturating MA5 mAb protein concentrations. generated. The majority of these mAbs recognize epitopes in the Mean fluorescence channel values were, likewise, similar between VNTR; Ref. 31). Subsequent cloning and sequencing of MUC1 the high MUC 1-expressing MM and the carcinoma cell lines. Thus, cDNAs from different cell sources by several groups showed high-level expression of apparently canonical MUC1 mRNA and that the protein-coding sequence consists of three general re- glycoprotein species were seen in both MM cell fines as well as gions, a 5' upstream region, a central VNTR region, and 3' clinical specimens. downstream region (36-40). Cloning and sequencing of MUC1 Several prior reports have shown that MUC1 may be ex- from breast and pancreatic sources shows a very high level of pressed on lymphoid cell populations, particularly MM cell lines sequence identity (>99%; Ref. 40). The VNTR region contrib- (8-11). Another report showed that MM cells and clinical samples utes substantially to the size and the properties of the MUC1 react with a mAb, MUSE11, which is VNTR-reactive. It was also glycoprotein. It is composed of repeating units of 60 nucleo- shown that MM cells appeared to release MUC1 because levels tides, with from --40-90 such units being present in the mature above the cutoff value for normal controls were observed in 12 of

25 (48%) MM serum samples tested (10). Circulating MUC1 was 3. King, M. A., and Nelson, D. S. Tumor cell heterogeneity in multiple compared between patients with MM and breast and pancreatic myeloma: antigenic, morphologic, and functional studies of cells from carcinoma. MM and breast carcinoma showed predominantly one blood and bone marrow. Blood, 73: 1925-1935, 1989. band of --300 kDa, whereas serum MUC1 from pancreatic carci- 4. Epstein, J., Xiao, H. Q., and He, X. Y. Markers of multiple hematopoietic-cell lineages in multiple myeloma. N. Engl. J. Med., 322: noma showed multiple bands of -300-450 kDa. The latter finding 664-668, 1990. is consistent with the tendency of pancreatic carcinoma-derived 5. Ruiz-Arguelles, G. J., and San Miguel, J. F. Cell surface markers in MUC1 to have a higher degree of O-glycosylation compared with multiple myeloma. Mayo Clin. Proc., 69: 684-690, 1994. breast carcinoma (39). Another major finding of this report was that 6. Bergsagel, P. L., Victor-Kobrin, C., Timblin, C. R., Trepel, J., and a normal T-lymphocyte bulk cell line could be generated from Kuehl, W. M. A murine cDNA encodes a pan-epithelial glycoprotein repeated in vitro stimulation of peripheral blood mononuclear cells that is also expressed on plasma cells. J. Immunol., 148: 590-596, 1992. from a MM patient with MUC1 + tumor cell lines. This cell line, 7. Robillard, N., Jego, G., Pellat-Deceunynck, C., Pineau, D., Puthier, which exhibited a CD3 +/TCR od[3 +/CD8 + phenotype, possessed D., Mellerin, MP., Barille, S., Rapp, M. J., Harousseau, J. L., Amiot, M., and Bataille, R. CD28, a marker associated with tumoral expansion in an ability to lyse both MUC 1 + MM cell lines as well as one of four multiple myeloma. Clin. Cancer Res., 4: 1521-1526, 1998. breast carcinoma cell lines tested while having no effect on 8. Delsol, G., Gatter, K. C., Stein, H., Erber, W. N., Pilford, K. A. F., MUCI - cell lines (including the natural killer cell-sensitive K562 Zinne, K., and Meason, D. Y. Human lymphoid cells express epithelial cell line). These results are consistent with our findings with respect membrane antigen: implications for diagnosis of human neoplasms. to the epitope reactivity in breast cancer and MM lines displayed by Lancet, 2: 1124-1129, 1984. the anti-MUC1 mAbs used. In addition, our results indicate very 9. Duperry, C., Klein, B., Durie, G. M., Zhang, X., Jourdan, M., Poncelet, P., Favier, F., Vincent, C., Brochier, J., Lenoir, G., and similar MUC l glycoprotein species between breast carcinoma and Bataille, R. Phenotype analysis of human myeloma cell lines. Blood, 73: MM cell lines. Our results extended these findings with respect to 566-569, 1989. comparisons of MUC1 mRNA and RT-PCR species (both trans- 10. Takahashi, T., Makiguchi, Y., Hinoda, Y., Kakiuchi, H., Nakagawa, VNTR and 5' region amplicons). N., hnai, K., and Yacht, A. Expression of MUC1 on myeloma cells and The expression of MUC1 at high levels on a majority of induction of HLA-unrestricted CTL against MUC1 from a multiple myeloma specimens may provide a new target and approach for myeloma patient. J. Immunol., 153: 2102-2109, 1994. therapy, namely RAIT. Radiolabeled antibodies have been used 11. Kamoshida, S., and Tsutsumi, Y. Expression of MUC- 1 glycoprotein in plasma cells, follicular dendritic ceils, myofibroblasts and pert- with some success in the detection, diagnosis, staging, and neural cells: immunohistochemical analysis using three monoclonal therapy of many different types of malignancies, including antibodies. Pathol. Int., 48: 776-785, 1998. MUCl-expressing tumors. However, to the best of our knowl- 12. Hilkens, J., Ligtenberg, M. J. L., Vos, H. L., and Litvinov, S. V. edge, this treatment modality has not been explored in MM Cell membrane-associated mucins and their adhesion-modulating prop- patients. Our studies have demonstrated the potential for apply- erty. Trends Biochem. Sci., 17: 359-363, 1992. ing radiolabeled MA5 as a therapeutic agent for this disease. At 13. Barlogie, B., Jagannath, S., Tricot, G., Desikan, K. R., Fassas, A., and Siegel, D. Advances in the treatment of multiple myeloma. Adv. least in this tumor model (JJN-3 in athymic nude mice), 1311- Intern. Med., 43: 279-320, 1998. labeled MA5 has demonstrated the potential to provide an 14. Attal, M., Harousseau, J. L., Stoppa, A. M., Sotto, J. J., Fuzibet, effective radiation dose to the tumor. Although a high radiation J. G., Rossi, J. F., Casassus, P., Maisonneuve, H., Facon, T., Ifrah, N., dose was also shown for the blood, with the current widespread Payen, C., and Bataille, R. A prospective, randomized trial of autolo- use of autologous or allogeneic stem cells myelotoxicity can be gous bone marrow transplantation and chemotherapy in multiple my- significantly reduced. Hepatotoxicity would be the next major eloma. N. Engl. J. Med., 335: 91-97, 1996. concern; however, the radiation dose potentially delivered by 15. Barlogie, B., Jagannath, S., Vesole, D. H., Naucke, S., Cheson, B., Mattox, S., Bracy, D., Salmon, S., Jacobson, J., Crowley, J., and Tricot, 13~I-labeled MA5 to this organ was only 2355 cGy/mCi (63.6 G. Superiority of tandem autologous transplantation over standard ther- cGy/MBq; thus, the total liver dose was 706 cGy at the murine apy for previously untreated multiple myeloma. Blood, 89: 789-793, maximum tolerated dose) of administered dose, which is well 1997. below the hepatotoxic level of ~1500 cGy usually observed in 16. Grogan, T. M., Spier, C. M., Salmon, S. E., Matzner, M., Rybski, J., mice (this dose threshold is 3000-3500 cGy in humans), a We Weinstein, R. S., Scheper, R. J., and Dalton, W. S. P-Glycoprotein expression in human plasma cell myeloma: correlation with prior chem- are currently performing RAIT studies to examine the antibody otherapy. Blood, 81: 490-495, 1993. form and radionuclide versus 9Oy) to provide conclusive (131I 17. Mongkonsrflxagoon, W., Kimlinger, T., Ahmann, G., and Greipp, evidence of a therapeutic potential in MM and the rationale for P. R. Is multidrug resistance (P-glycoprotein) an intrinsic characteristic a clinical study. of plasma cells in patients with monoclonal gammopathy of undetermined significance, plasmacytoma, multiple myeloma and amyloidosis? Leuk. Lymphoma, 29: 577-584, 1998.

References 18. Raaijmakers, H. G., Izquierdo, M. A., Lokhorst, H. M., de Leeuw, 1. Bataille, R., and Harousseau, J. L. Multiple myeloma. N. Engl. C., Belien, J. A., Bloem, A. C., Dekker, A. W., Scheper, R. J., and Sonneveld, P. Lung-resistance-related protein expression is a negative J. Med., 336: 1657-1664,1997. predictive factor for response to conventional low but not to intensified 2. Greipp, P. R., and Witzig, T. Biology and treatment of myeloma. dose alkylating chemotherapy in multiple myeloma. Blood, 91: 1029- Curr. Opin. Oncol., 8: 20-27, 1996. 1036, 1998. 19. Tu, Y., Renner, S., Xu, F., Fleishman, A., Taylor, J., Weisz, J., Vesciom, R., Rettig, M., Berenson, J., Ka'ajewski, S., Reed, J. C., and Lichtenstein, A. BCL-X expression in multiple myeloma: possible in- dicator of chemoresistance. Cancer Res., 58: 256-262, 1998. 4 Dr. R. Blumenthal, Garden State Cancer Center, Belleville, NJ, per- 20. Tu, Y., Xu, F. H., Liu, J., Vescio, R., Berenson, J., Fady, C., and sonal communication. Lichtenstein, A. Upregulated expression of BCL-2 in multiple myeloma

cells induced by exposure to doxombicin, etoposide, and hydrogen labeled monoclonal antibody (MA5) against a mucin-like antigen. Can- peroxide. Blood, 88: 1805-1812, 1996. cer Res., 50: 927s-931s, 1990. 21. Portier, M., Moles, J. P., Mazars, G. R., Jeanteur, P., Bataille, R., 33. Gold, D. V., Lew, K., Malinak, M., and Cardillo, T. Characteriza- Klein, B., and Theillet, C. p53 and RAS gene mutations in multiple tion of monoclonal antibody PAM4 reactive with a pancreatic cancer myeloma. Oncogene, 7: 2539-2543, 1992. mucin. Int. J. Cancer, 57: 204-210, 1994. 22. Neri, A., Baldini, L., Trecca, D., Cro, L., Polli, E., and Maiolo, A. T. 34. Chizzonite, R., Truitt, T., Desai, B. B., Nunes, P., Podlaski, F. J., p53 gene mutations in multiple myeloma are associated with advanced Stern, A. S., and Gately, M. K. IL-12 receptor: I. Characterization of the forms of malignancy. Blood, 81: 128-135, 1993. receptor on PHA-activated human lymphoblasts. J. Immunol., 148: 23. Mill, W. B., and Griffith, R. The role of radiation therapy in the 3117-3124, 1992. management of plasma cell tumors. Cancer (Phila.), 45: 647- 652,1980. 35. Siegel, J. A., and Stabin, M. G. Absorbed fractions for electrons and 24. Dimopoulos, M. A., Goldstein, J., Fuller, L., Delasalle, K., and [3 particles in spheres of various sizes. J. Nucl. Med., 35: 152-156, Alexanian, R. Curability of solitary bone plasmacytoma. J. Clin. Oncol., 1994. 10: 587-590, 1992. 36. Gendler, S. J., Burchell, J. M., Duhig, T., Lamport, D., White, R., 25. Liebross, R. H., Ha, C. S., Cox, J. D., Weber, D., Delasalle, K., and Parker, M., and Taylor-Papadimitriou, J. Cloning of partial cDNA Alexanian, R. Solitary bone plasmacytoma: outcome and prognostic encoding differentiation and tumor-associated mucin glycoproteins ex- factors following radiotherapy. Int. J. Radiat. Oncol. Biol. Phys., 41: pressed by human mammary epithelium. Proc. Natl. Acad. Sci. USA, 1063-1067, 1998. 84: 6060-6064, 1987. 26. Ozaki, S., Kosaka, M., Wakatsuki, S., Abe, M., Koishihara, Y., and 37. Siddiqui, J., Abe, M., Hayes, D., Shani, E., Yunis, E., and Kufe, D. Matsumoto, T. Immunotherapy of multiple myeloma with a monoclonal antibody directed against a plasma cell-specific antigen, HMI.24. Isolation and sequencing of a cDNA coding for the human DF3 breast carcinoma-associated antigen. Proc. Natl. Acad. Sci. USA, 85: 2320- Blood, 90: 3179-3186, 1997. 2323, 1988. 27. Ozaki, S., Kosaka, M., Harada, M., Nishitani, H., Odomi, M., and Matsumoto, T. Radioimmunodetection of human myeloma xenografts 38. Ligtenberg, M. J. L., Vos, H. L., Gennissen, A. M. C., and Hilkens, with a monoclonal antibody directed against a plasma cell specific J. Episialin, a carcinoma-associated mucin, is generated by a polymor- antigen, HM1.24. Cancer (Phila.), 82: 2184-2190, 1998. phic gene encoding splice variants with alternative amino termini. J. Biol. Chem., 265: 5573-5578, 1990. 28. Bayouth, J. E., Macey, D. J., Kasi, L. P., Garlich, J. R., McMillan, K., Dimopoulos, M. A., and Champlin, R. E. Pharmacokinetics, dosim- 39. Lan, M. S., Hollingsworth, M. A., and Metzgar, R. S. Molecular etry and toxicity of holmium-166-DOMTP for bone marrow ablation in characterization of a mucin-type antigen associated with human pancre- multiple myeloma. J. Nucl. Med., 36: 730-737, 1995. atic cancer. The DU-PAN-2 antigen. J. Biol. Chem., 262: 12863-12870, 29. Chomczynski, P., and Mackey, K. Substitution of chloroform by 1987. bromo-chloropropane in the single-step method of RNA isolation. Anal. 40. Lan, M. S., Batra, S. K., Qi, W-N., Metzgar, R. S., and Hollings- Biochem., 225: 163-164, 1995. worth M. A. Cloning and sequencing of a human pancreatic tumor 30. Batra, S. K., Kern, H. F., Worlock, A. J., Metzgar, R. S., and Holling- mucin cDNA. J. Biol. Chem., 265: 15294-15299, 1990. sworth, M. A. Transfection of human Mucl mucin gene into a poorly 41. Gendler, S. J., Spicer, A. P., Lalani, E. N., Duhig, T., Peat, N., differentiated human pancreatic tumor cell line, PANCI: integration, ex- Burchell, J., Pemberton, L., Boshetl, M., and Taylor-Papadimitriou, J. pression and ultrastmctural changes. J. Cell. Science, 100: 841-849, 1991. Structure and biology of a carcinoma-associated mucin, MUC1. Am. 31. Dion, A. S., Smorodinsky, N. I., Williams, C. J., Wreschner, D. H., Rev. Respir. Dis., 144: $42-$47, 1991. Major, P. P., and Keydar, I. Recognition of peptidyl epitopes by poly- 42. Nishimori, I., Perini, F., Mountjoy, K. P., Sanderson, S. D., morphic epithelial (PEM)-specific monoclonal antibodies. Hybridoma, Johnson, N., Cerny, R. L., Gross, M. L., Fontenot, J. D., and 10: 595-610, 1991. Hollingsworth, M. A., N-Acetylgalactosamine glycosylation of 32. Major, P. P., Dion, A. S., Williams, C. J., Mattes, M. J., Wang, T., MUCI tandem repeat peptides by pancreatic cell extracts. Cancer and Rosenthall L. Breast tumor radioimmunodetection with a J~In- Res., 54: 3738-3744, 1994.

Downloaded from clincancerres.aacrjournals.org on October 1, 2021. © 1999 American Association for Cancer Research. Epithelial Mucin-1 (MUC1) Expression and MA5 Anti-MUC1 Monoclonal Antibody Targeting in Multiple Myeloma

J. Burton, D. Mishina, T. Cardillo, et al.

Clin Cancer Res 1999;5:3065s-3072s.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/5/10/3065s

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/5/10/3065s. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.