Bioconjugate Chem. 1002, 3, 549-553

Determination of Immunoreactivity of Doxorubicin Antibody Immunoconjugates by a Ley Competitive RIA Huiru Zhao,’ David Willner, Jeffrey S. Cleaveland, Gary R. Braslawsky, and Joseph P. Brown Bristol-Myers Squibb Company, Pharmaceutical Research Institute, Wallingford, Connecticut 06492. Received August 11,1992 Many monoclonal antibody-drug immunoconjugates have been evaluated for their ability to deliver cytotoxic drugs to tumors. It is essential to establish that the ability of the conjugates to bind antigen, i.e. their immunoreactivity, is not adversely affected by the drug conjugation procedure. We have described herein a measurement of the immunoreactivity of BR96-DOX, a conjugate comprised of BR96, a chimeric monoclonal antibody specific for the Ley tetrasaccharide commonly expressed on human carcinomas, and doxorubicin, an anticancer agent in widespread clinical use. We have employed a competitive RIA, in which microtiter wells were coated with synthetic Ley conjugated to human serum albumin and then incubated with 1251-labeledantibody BR96 in the presence of test conjugate or intact BR96 mAb. The test conjugates were found to compete as effectively as unconjugated BR96. This aseay is highly applicable to QC processes with the intra-assay CV = 2.0% and the interassay CV = 4.3%.

Doxorubicin (DOXl)-monoclonal antibody conjugates have been developed to increase a drug’s specificity and selectivity to tumor cells and to decrease its toxicity, particularly its dose-limiting cardiac toxicity and myelosuppression (1-5). In these targeted delivery systems, DOX was coupled to monoclonal antibodies (mAbs) with an acid-labilehydrazone linker. In general, effectivecancer chemotherapy utilizing antibody-drug conjugates requires that the mAbs transport DOX in an antigen-specific manner. Once bound to tumor cells the conjugate is internalized and free DOX is released from lysosomes to function as a DNA alkylator (6,7).In our study, chimeric BR96 mAb was selected as a drug carrier for the following features: high tumor reactivity, rapid internalization, and direct cytotoxicity on antigen-positive tumor cells as well as mediation of ADCC and CDC (8). Preliminary studies have indicated that at least a portion of the epitope recognized by BR96 is a Ley carbohydrate chain (8). Animal studies have established that BR64-DOX immunoconjugates, in which BR64 binds the same antigen as BR96, kill tumor cells in an antigen-specific manner in vitro and in vivo against human breast, colon, lung tumor xenograft (9). All those results imply that BR96-DOX immunoconjugates have the potential to be clinically effective antineoplastic drugs in tumor chemotherapy. Antibodies may undergo structural changes after conjugation and this may affect their binding characteristics and their efficacy as selective carrier vehicles (IO). In most circumstances, the more drug molecules that were directly coupled to a mAb via amide or ester bonds, the lower the immunoreactivity (binding avidity) of the conjugate (11). Therefore, our major concern is to determine the immunoreactivity of the drug-modified BR96 mAb as compared to that of intact BR96 mAb.

* Author to whom correspondence should be addressed.

1 Abbreviations used DOX,doxorubicin;BR96 mAb, chimeric BR96 monoclonal antibody; ADCC, antigen-dependentcellular cytotoxicity; CDC, complement-dependent cytotoxicity; BR96DOX, doxorubicin-BR96 antibody immunoconjugates; RIA, radioimmunoassay; D-PBS, Dulbecco’s phosphate-buffered saline; HSA, human serum albumin; Ley, Lewisy hapten; FBS, fetal bovine serum; FITC, fluorescein isothiocyanate.

Furthermore, there is substantial need for a precise assay to probe structural varintions and bi.iding avidities in the conjugates. In this study, a 96-well microtiter plate was coated with synthetic Ley conjugated to human serum albumin. This plate was then incubated with 1251-labeledantibody BR96 in the presence of test conjugates or standard BR96 to evaluate BR96-DOX immunoconjugates. Moreover, a comparative study has also been carried out with fluorescence-activated cell sorter on MCF, cells. There is a 10% difference in the binding avidity between these two assays. However, the ease, ?peed, and reproducibility of microtiter-based assay overcome some of the drawbacks which occur in cell-bas ?d immunoassays such as their manipulative inconvenience and widely variant results. EXPERIMENTAL PROCEDURES

Materials. HSA-Ley antigen was purchased from ChemBiomed, Alberta, Canada. The 96-well microtiter plates (Immulon-2) were purchased from Dynatech Laboratory Inc., Chantilly, VA; BSA, EDTA, Tris, and sucrose were from Sigma Chemical Co., St. Louis, MO; Dulbecco’s phosphate-buffered saline (D-PBS, 1X)was from GIBCO Laboratories, Grand Island, NY; sodium carbonate was from EM Science, Gibbstown, NJ; Tween-20 was from Bio-Rad, Richmond, CA; and chimeric BR96 mAb (IgG1) was obtained from Bristol-Myers Squibb Biotechnology Purification Pilot Plant, Syracuse, NY. FITC-labeled goat anti-human Ig K chains F(ab)’z was purchased from Boehringer Mannheim Corp., Indianapolis, IN. All other reagents were purchased from Fisher Scientific, Springfield, NJ. Equipment. A Waters (Milford, MA) HPLC system, equipped with a 600 E system :ontroller, 490E programmable multiwavelength detector, 700 Satellite WISP autosampler, and Maxima 820 computer-controlled data processor was employed for analytical study. A Supelco 7.8 X 300 mm size-exclusion HPLC column (Rohm and Haas Co.), a Progel-TSK G3000 SWXL column, and a guard column of the same type of packing (6.0 X 40.0 mm) were used for determining the quality of the immunoconjugates. 0 1992 American Chemlcal Society

550

Bloconlugete Chem., Vol. 3, No. 6, 1992

Immobilization of Ley Antigen Plates. The 96-well microtiter plates were coated by treatment with 100 pL of Ley antigen at a concentration of 15-20 pg/mL in 0.1 M sodium bicarbonate/carbonate buffer (pH 9.0) for 1620 h a t 4 "C. The antigen-coated plates were then blocked with 200 pL of 2 5% BSA/l mM EDTA/5% sucrose in Tris buffer (pH 7.2) and were incubated at ambient temperature for 2 h. The blocked plates were dried on absorbent paper and stored dessicated at 4 "C for subsequent use. Selection of Optimal Conditions for Competitive RIA. Microtiter plates were coated with 100pL of a serial dilution of Ley antigen solution in 0.1 M bicarbonate/ carbonate buffer, pH 9.0, at concentrations of 3.12,6.25, 12.5,25.00,50.00, and 100.00pg/mL. 2 % BSA and D-PBS buffer were used as controls. The plates were immobilized and dried as described above. Each point was measured in quadruplicate. To wells containing the same concentration of Ley antigen, four series of 1251-labeledchimeric BR96 mAb at 1.3,2.6,6.6, and 13.0 pg/mL in 2% BSA phosphate buffer were added. 1251-labeledBR96 mAb was prepared by the chloramine-T method (12) with a specific activity of 6.3 mCi/mg. The plates were sealed and incubated a t 37 "C for 2 h. Plates were then washed four times with 0.05% Tween-eO/PBS and dried on absorbent paper at ambient temperature. To each well was added 1 M sodium hydroxide. The solution in each well was then transferred into individual scintillation tubes. lZ5Icounts bound per well (cpm) were recorded on a LKB gamma counter and a correction was made for nonspecific binding. From the above experiment, optimum concentration for 1251-labeledBR96 mAb was selected as 5 pg/mL. Several different parameters were selected to search for the optimum conditions for the binding study. Among these, two densities of Ley antigen at concentrations of 10 and 2.5 pg/mL were chosen, and two competition reaction periods of 2 and 3 h at 37 "C were adopted. The standard unmodified chimeric BR96 mAb was diluted 2-fold serially with 1% BSA/PBS to concentrations of 200,100,50,25, 12.5, and 3.12 pg/mL. The above standard mAb solution was mixed with an equal volume of 1251-labeledBR96 mAb and was added to the immobilized Ley antigen plates. The plates were sealed and incubated at 37 OC for 2 or 3 h. The repetitive sequences were carried out as described above. Competitive RIA for the Evaluation of BR96-DOX Immunoconjugates. Based on the competitive RIA studies described above, optimum concentrations for Ley antigen and lZ5I-BR96mAb were selected as 10 and 5 pg/ mL, respectively. The competitive reaction period was selected as 2 hours. Antibody conjugates with a drug load of between 6 and 8 mol of DOX and standard unmodified chimeric BR96 mAb were diluted 2-fold serially with 1% BSA/PBS to concentrations of 200,100,50,25,12.5, and 6.25 pg/mL. The immunoreactivities were determined as described above. All assays were performed in duplicate or triplicate. Additional controls were used to determine the maximal cpm bound as well as the specificity of binding by using 1%BSA in PBS, and polyclonal human IgGDOX and free DOX in 1%BSA phosphate buffer, respectively. The percent avidity was determined as the concentration of BR96 mAb at 50% bound (ICw) divided by the concentration of the conjugate at 50% bound. The purity of the BR96-DOX immunoconjugates was analyzed on a Progel-TSK G3000 SWXL HPLC column (7.8 X 300 mm) with a guard column (6.0 X 40.0 mm) of the same type of packing. The eluant for the HPLC system was 50 mM phosphate buffer. A series of molecular weight

150000

-

0

20

40

60

60

100

1 0

Concentration of HSA-Le7 antlgen ( p g / m L) Figure 1. Direct binding RIA with a serial dilution of HSA-Ley antigen on Immulon-2 microtiter plates. At each concentration of antigen, 1251-BR96mAb was selected at concentrations (pg/ mL) of 1.3 (O),2.6 (A),6.6 (O), and 13.0 (m).

standards were used to calibrate the HPLC column. The peaks were monitored at two wavelengths, 220 and 495 nm, at which only DOX has absorbance. The peak heights or areas of the aggregates and the monomeric conjugates were measured which had absorbance at both 220 and 495 nm. There was less than 5 % aggregates in the synthesized conjugates. The concentration of BR96 mAb in BR96-DOX conjugates was calculated spectrophotometrically at two wavelengths, 280 and 495 nm (5). The amount of conjugated DOX in antibody was determined by absorbance at 495 nm (e495 = 8030). To correct for the overlap of DOX absorbance at 280 nm, the following equation was used: mAb (mg/mL) = [A,, - (0.72 X A495)l/l.4 Determination of Immunoreactivities of BR96DOX on MCF7 Cells by FACS. Assays were performed by fluorescenceas previouslydescribed (20). Briefly,target cells were harvested in the logarithmic phase using EDTA in calcium- and magnesium-free PBS. The cells were washed twice in PBS containing 1% bovine serum albumin and resuspended to 1 X 107/mL in PBS containing 1% BSA. Cells (0.05mL) were mixed with standard BR96 mAb, BR96-DOX with a drug load of 6.9, and nonbinding human IgG at concentrations of 66, 33, 6.66, 3.33, 0.66, 0.33, and 0.066 nM. The cells were then incubated for 45 min at 4 "C. The cells were washed two times in 1% FBS/ PBS buffer and resuspended in 0.1 mL of an appropriate concentration of FITC-labeled goat anti-human IgG K chains F(ab)'z. Cells were incubated for 45 min, washed two times, and kept on ice until analyzed on a Coulter EPICS 753 fluorescence-activated cell sorter. Data are expressed as the mean channel number of specific versus control antibody. RESULTS

Selection of the Optimum Conditions in the Competitive RIA. The standard curve of direct binding lz5IBR96 mAb to HSA-Ley antigen on microtiter plates is shown in Figure 1. There was no significant increase of counts bound at antigen excess with antibody concentrations above 6.6 pg/mL. As compared by assays at different antigen densities in immobilization, a LeY antigen

Bloconjugete Chem., Voi. 3, NO. 6, l Q Q 2 611

Ley Competitive R I A 120 ,

100

i

8 .-

so

-

U

60

-

40

-

20

-

v)

N

a

m m

c

a 0

o ! 1

10

. .

1

1000

100

.

.

10

. . . . . . I

.

. ..... 11 30

100

Log (C) of BRS6 mAb (ug/ml)

Log (C) of BRS6 mAb (ug/ml)

12000 10000

. . . . . . I

Figure 3. Competitive RIA for standard BR96 mAb (e),human IgG-DOX (drug load = 8.2) (W), BR96-DOX (drug load = 7.8) (A), and free DOX ( 0 )under the same conditions as in Figure 2b with a 2-h reaction period.

0

5000

f

6000

3 ul r N

-

4000

U

a

2000

m a

a

0 0

20

40

60

so

100

120

(C) of BR96 mAb (ug/ml)

0

40001 2000

j

1

Figure 2. (a) Competitive RIA for standard BR96 mAb with HSA-Ley antigen in plate at concentrations of 10 pg/mL (m) and 2.5 pg/mL (+). The lZ6I-BR96concentration was 5 pg/mL. (b) Competitive RIA for standard BR96 mAb at HSA-Ley antigen in plate (10 pg/mL) with an immunoreaction period of 2 h ( 0 ) and 3 h (A). The rest parameters were same as above. The X-axis represents the concentration of competing BR96 mAb while the Y-axis represents percent the counts bound of lZ5IBR96 mAb.

Figure 4. Determination of immunoreactivities of BR96-DOX immunoconjugatesby competitiveRIA under the same conditions as described previously. 0 represents BR96 mAb. Drug loads in BR96-DOX are 6.9 (A), 7.0 (O), 7.7 (U), and 8.0 (A). The log-scale plot is inset.

concentration of 10pg/mL would provide more information content (Figure 2a). Thus, the Ley antigen concentration for immobilization was selected as 10 pg/mL, and the labeled antibody concentration as 5 pg/mL. From Figure 2b, we see that equilibrium for the competition reaction has been reached at 2 h. There is no significant difference in results between 2 and 3 h. Therefore, 2-h incubation at 37 O C was selected for the following RIA. Assay Sensitivity and Reproducibility. As observed in the competitive assay for BR96 mAb (Figure 2a), the standard dilutions ranged from 6.25 to 100 pg/mL. The curve was linear between 10 and 30 pg/mL on the log plot and this portion of the plot can be utilized for more quantitative determination of immunoconjugates. The sensitivity of this assay is as low as 3 pg/mL. The intraassay coefficient of variation (cpm) was 2.0%, and the interassay coefficient of variation was 4.3 % . The titer of the assay is lo4 at an antigen concentration of 10 pg/mL. The titer of the assay will increase as the concentration of antigen goes up in the plate.

Antibody Specificity. The competitive RIA for intact and modified BR96 mAb is displayed in Figure 3. The human polyclonal antibody IgGDOX was selected as a negative control as also was free DOX. Ley antigen, which is the tumor-cell-associated antigen (13), competes with BR96 mAb and modified BR96 conjugate (drug load = 8.0). The binding affinity constant of BR96 mAb to Ley antigen is in the range of 104-10-7 M, which corresponds to the same level as that of BR96 mAb to H3396 breast adenocarcinoma line (8). Competitive RIA Evaluation of BR96-DOX Immunoconjugates. The binding curve for modified BR96 mAb conjugates with a doxorubicin load of 6-8 mol/mol of antibody is shown in Figure 4. The percent avidities were determined as the concentration of control (BR96) at 50% bound divided by the concentration of the conjugate at 50 5% bound. These results are shown in Table I. The ICs0 values for both unmodified BR96 and BR96DOX were obtained from the log scale regression of experimental data by using the Cricket graph program.

0

20

40

60

80

100

120

(C) of BRS6 mAb (ug/ml)

Zhao et el.

552 Ebconlugete Chem., Vol. 3,No. 6, 1992

Table I. Binding Avidities of BR96-DOX Immunoconjugates with Drug Loads of 6-8 mol/mol of mAba IC50 binding IC50 binding BR96-DOX BR96-DOX values avidity drug values avidity drug load (rg/mL) ( % ) load (rra/mL) ( % ) 6.9 16.2 80 7.7 13.1 98 7.0 14.0 92 8.0 14.3 90 ~

a These data were calculated from a regressionof the experimental data (in log scale) of Figure 4 by using the Cricket graph program.

140

120

-

100

-

so

-

60

40

i z

20

1

I

Log (C) of BR96 mAb

Figure 5. FACS profile for BR96 mAb

(nM)

( 0 ) and BR96-DOX

with a drug load of 6.9 ( 0 )on MCF, cells. The X-axis represents the concentration of mAb on a log scale (nM) while the Y-axis represents the mean channel fluorescence.

Table 11. Determination of Immunoreactivity of BR96-DOX with a Drug Load of 6.9 with Comparison of Commtitive RIA and FACS binding avidity ( % ) ICW(MalmL) 80 RIA 16.2 71 0.38 FACS ~~

Comparative Study of Two Assays in Determination of the Immunoreactivity of BR96-DOX. The binding profile of BR96-DOX on MCF, cells by FACS is shown in Figure 5. In this assay the binding avidity was expressed as the concentration of the conjugate at 50% bound divided by the concentration of BR96 mAb a t 50% bound. The binding avidities of BR96-DOX were determined and found to be 71% in the FACS assay. Whereas in competitiveRIA the binding avidity for BR96DOX (drug load = 6.9) was 80%. There was about 10% difference between these two assays. The comparative data are listed in Table 11. DISCUSSION

BR96 monoclonal antibody is specific for Ley antigen, which is highly expressed in tumor-associated cells. BR96 and ita immunoconjugates might be expected to be more potent and more effective cancer chemotherapeutics. The assay established here is based on the interaction of BR96 mAb with Ley antigen. As observed in Figure 1,HSA-Ley M. antigen has high binding affinity to BR96 mAb at This antigen is also readily immobilized onto Immulon-2 microtiter plates through hydrophobic interaction. Data derived from sucrose-coated aged plates (2 weeks old) stored desiccated at 4 "C was consistent with that obtained with freshly prepared plates. The antigen plates can stay dessicated for 1 month without losing potency. The

reported procedure for binding antigen and blocking nonspecific binding was a minor modificationof a previous immobilization protocol employed in our laboratory; Le., the antigen plates were blocked with 2 % BSA, washed with 0.05 % Tween-20, and dried on absorbent paper for immediate use. Our present blocking technique is described in the Experimental Procedures section. Results from both protocols are in agreement, allowing for preparation of large numbers of antigen-coated 96well plates. The data in Figure 3indicate that L@antigen specifically competes with BR96 mAb and ita modified conjugates. Ley antigen was expressed at higher levels in tumorassociated cells (such as MCF7 cells and RCA cells) even though it is also present to a lesser degree in many other tissues in the form of glycolipid and glycoproteins derivatives (16,17). The binding affinity between Ley antigen and human IgGDOX is very low as evidenced by the noncompetitive nature of IgGDOX conjugates in this assay. Also, free DOX, which might be generated during storage, did not interfere with this assay. Therefore, this assay provides a generic protocol for the determination of the immunoreactivity of BR96 mAb conjugates including drug-antibody conjugates, immunotoxin conjugates, and radionuclides conjugates. Other considerations in assay development include both time and economic efficiency. The traditional cell-based immunoassay (11)requires at least one entire day for the analysis of one or two conjugates and large numbers of cells are required for five or six samples. The microtiterbased RIA reported here is efficient and allows for higher throughput. It requires about 4-5 h to evaluate approximately six samples with potential for the evaluation of many more. Thus this assay allows for the efficient evaluation of many more samples in less time than previous methodologies. Most of the synthetic immunoconjugateshave preserved reasonable binding avidity as displayed in Table I. Higher loads of drug up to 8.0 mol/antibody do not structurally alter the antibody binding avidity. This implies that doxorubicin would not be conjugated to the antigenic binding sites (VH region) of the antibody. It is very interesting to notice that there is some difference in the binding avidity between competitive RIA and FACS in assessment of BR96-DOX immunoconjugates. FACS assay is based on direct binding mAb on the target MCF7 cells, in which the Ley antigen was expressed at higher density. The binding ability to the cells was detected by the second antibody, F(ab)Z-FITC. Whereas in competitive RIA, the labeled antibody and test conjugates competed for the excess amount of purified Ley antigen. The data from these assays were all acceptable as evidenced by antigen-specific cell killing in vitro for this batch of the immunoconjugate (unpublished results). Thus, BR96-DOX with a drug load of 6.9 has relatively preservable immunoreactivity. With respect to speed, simplicity, and replicability, the competitive RIA is more suitable as a quality control process. ACKNOWLEDGMENT

The authors wish to express appreciation to Dr. Ned Heindel, Lehigh University, for revising the manuscript. LITERATURE CITED (1) Kaneko, T., Willner, D., Monkovic, I., Knipe, J. O., Braslaw-

sky, G. R., Greenfield, R. S., and Vyas, D. M. (1991) New hydrazone derivatives of adriamycin and their immunocon-

Ley Competlthre R I A

jugates-A correlation between acid stability and cytotoxicity. Bioconjugate Chem. 2, 133-144. (2) Yang, H. M., and Reisfeld, R. A. (1988) Doxorubicin conjugated with a monoclonal antibody directed to a human melanoma-associated proteoglycan suppresses the growth of established tumor xenografts in nude mice. h o c . Natl. Acad. Sci. U.S.A. 85,1189-1193. (3) Dillman, R. O., Jojnson, D. E., Shawler, D. L., and Koziol, J. A. (1988) Superiority of an acid-labile daunorubicinmonoclonal antibody immunoconjugatecompared to free drug. Cancer Res. 48,6097-6102. (4) Braslawsky, G. R., Edson, M. A., Pearce, W., Kaneko, T., and Greenfield, R. (1990) Antitumor activity of adriamycin (hydrazone-linked) immunoconjugates compared with free adriamycin and specificity of tumor cell killing. Cancer Res. 50,6608-6614. ( 5 ) Greenfield,R., Kaneko, T., Daves, A., Edson, M. A., Frtzgerald, K. A,, Olech, L. J., Grattan, J. A., Spitalny, G. L., and Braslawsky, G. R. (1990) Evaluation in vitro of adriamycin immunoconjugates synthesized using an acid-sensitive hydrazone linker. Cancer Res. 50, 6600-6607. (6) deDuve,C. (1983)Lysosome revisited. Eur. J.Biochem. 137, 391-397. (7) Pastan, I. H., and Willingham, M. C. (1985) Pathway of Endocytosis. In Endocytosis (I. H. Pastan, and M. C. Willingham, Eds.) pp 1-44, Plenum Publishing Corp., New York. (8) Hellstrom, I., Garrigues, H. J., Garrigues, U., and Hellstrom, K. E. (1990) Highly tumor-reactive, internalizing, mouse monoclonal antibodies to Ley-related cell surface antigens. Cancer Res. 50, 2183-2190. (9) Trail, P. A., Willner, D., Lasch, S. J., Henderson, A. J., Greenfield, R. S., King, D., Zoeckler, M. E., and Braslawsky, G. R. (1992) Antigen-specific activity of carcinoma reactive BR64-Adriamycinconjugates evaluated in vitro and in human tumor xenograft models. Cancer Res. In press. (10) Tyle, P., and Ram, B. P. (1990) Targeted Therapeutic System (P. Tyle and B. P. Ram, Eds.) pp 9-10, Marcel Dekker, Inc., New York. (11) Pierce, D. L., Heindel, N. D., Schray, K. J., Jetter, M. M., Emrich, J. G., and Woo, D. V. (1990) Misonidazoleconjugates

Bloconlugete Chem., Vol. 3, No. 8, 1992 669

of the colorectal tumor associated monoclonal 17-1A. Bioconjugate Chem. 1 , 314-318. (12) Hunter, R. M., and Greenwood, F. C. (1962) Preparation of iodine-131 labeled human growth hormone of high specific activity. Nature 194, 495-496. (13) Shestowsky, W., Fallavollita, L., and Brodt, P. (1990) A monoclonal antibody to lewis lung carcinoma variant H-59 identifies a plasma membrane protein with apparent relevance to lymph node adhesion and metastasis. CancerRes.50,19481953. (14) Hunger, H. D., Flachmeier, C., Schmidt, G., Behrendt, G., and Coutelle, Ch. (1990) Ultrasensitive enzymatic radioimmunoassay using a fusion protein of protein A and neomycin phoephotransferase I1 in two-chamber-wellmicrotiter plates. Anal Biochem. 187,89-93. (15) Cleaveland, J. S., Kiener, P. A,, Hammond, D. J., and Schacter,B. Z. (1990)Amicrotiter-basedassay for thedetection of protein tyrosine kinase activity. Anal Biochem. 190,249253. (16) Waldock, A., Ellis, I. O., Armitage, N., Turner, D. R., Hardcastle, J. D., and Embleton, J. (1989) Differential expression of the lewis Y antigen defined by monoclonal antibody C14/1/46/10 in colonic polyps. CancerRes. 64,414421. (17) Bara, J., Mollicone, R., Herrero-Zabaleta, E., Gautier, R., Daher, N., and Oriol, R. (1988) Ectopic expression of the Y (Ley)antigen defined by monoclonalantibody 12-4LEin distal colonic adenocarcinomas. Znt. J. Cancer 41, 683-689. (18) Heindel, N. D., Zhao, H., Egolf, R. A., Chang, C. H., Schray, K. J., Emrich, J. G., Mclaughlin, J. P., and Woo, D. V. (1991) A novel heterobifunctional linker for formyl to thiol coupling. Bioconjugate Chem. 2, 427-430. (19) de Rosario, R. B., Wahl, R. L., Brocchini, S. J., Lawton, R. G., and Smith, R. H. (1990) Sulfhydryl site-specific crosslinking and labeling of monoclonalantibodies by a fluorescent equilibrium transfer alkylation cross-link reagent. Bioconjugate Chem. 1, 51-59. (20) Hellstrom, I., Hom, D., Linsley, P., Brown, J. P., Brankovan, V., and Hellstrom, K. E. (1986) Monoclonalmouse antibodies raised against human lung carcinoma. Cancer Res. 46,39173923.