Bioconlugate Chem. 1003, 4,353-357
363
Application of Monoclonal Antibodies against Cytosine Deaminase for the in Vivo Clearance of a Cytosine Deaminase Immunoconjugate David E. Kerr,' Ursula S. Garrigues, Philip M. Wallace, Karl Erik Hellstrom, Ingegerd Hellstrom, and Peter D. Senter Bristol-Myers Squibb Pharmaceutical Research Institute, 3005 First Avenue, Seattle, Washington 98121. Received April 12, 1993
The selective delivery of anticancer drugs to tumors vs normal tissue using targeted antibody-enzyme conjugates for prodrug activation is limited by the amount of drug generated by blood-borne enzyme. Clearance of non-tumor-associated conjugate would increase the tumor/blood conjugate ratio, and enable larger amounts of prodrugs to be administered. A method for clearing the monoclonal antibody (mAb) conjugate L6-cytosine deaminase (L6-CD) was established by using an antibody raised against CD. The mAb 102-26 was obtained by immunizing BALB/C mice with CD conjugated to keyhole limpet hemocyanin. 102-26 was able to precipitate purified CD from solution as assessed by radioimmune precipitation and recognized CD in Western blot analyses. Similar studies were used to establish that 102-26 also recognized CD when conjugated to the L6 and 1F5 mAbs. Selective removal of L6-CD from the circulation of nude mice bearing H2981 human lung adenocarcinoma (L6-antigen positive) was achieved by injecting 102-26 24 h after L6-CD administration. High T / B ratios were obtained by clearance of a L6-CD (38:l compared to 1.3:l without clearance).
INTRODUCTION Although 5-fluorouracil (5-FU)' is the drug most commonly used for treating colon cancer (11, it is rarely curative, and a great deal of research has been directed toward the improvement of 5-FU therapy. These methods include direct injections of 5-FU into the affected area (2),coadministration of the folate analog, leucovorin (31, and the immunostimulant, levamisole (4), inhibition of de novo pyrimidine synthesis with (phosphonoacety1)-Easpartate (5),implantation of CD-containing capsules for the generation of 5-FU a t the tumor (6),and transfecting tumor cells with the gene for CD to render cells sensitive to 5-FC (7). These treatment modalities are aimed a t increasing the local concentration of 5-FU, the effectiveness of the drug, or the protection of nontarget tissues from toxic side effects. We have previously demonstrated that a chemical conjugate of the anticarcinoma mAb L6 and the enzyme CD yields a molecule capable of binding to L6-positive tumor cells and enzymatically converting 5-FC to 5-FU (8). This provides a method for delivering 5-FU to areas where the conjugate is present. A problem recognized early in in vivo studies was that blood-borne L6-CD can convert 5-FC to 5-FU in the circulation and contribute significantly to generalized toxicity. mAb-enzyme conjugates are frequently retained in the circulation for extended periods of time, which can lead to low T / B conjugate ratios (9). Though the use of mAb fragments (e.g. F(ab')n, Fab', or sFv) conjugated to enzymes can
* Author to whom correspondence should be addressed. Telephone: 206-727-3523.Fax: 206-727-3603. Abbreviations used 5-FC, 5-fluorocytosine; 5-FU 5-fluorouracil; CD, cytosine deaminase; DEAE, (diethy1amino)ethyl; ELISA, enzyme-linkedimmunosorbent assay; HRP, horseradish peroxidase; ID, injected dose;KLH, keyhole limpet hemocyanin; mAb, monoclonal antibody; PBS, phosphate-bufferedsaline (10 mM sodium phosphate, 150 mM NaCl at pH 7.2);SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis;
SMCC, succinimidyl 4-(2-maleimidylethyl)cyclohexane-l-carboxylate; T/B, tumor to blood.
improve T / B ratios (10, l l ) ,the absolute amounts of conjugate delivered to tumors with fragments is reduced compared to whole mAbs. In order to optimize therapy with mAb-CD conjugates in vivo, it is necessary to achieve high conjugate levels a t the tumor site while low enzyme levels are maintained in the blood. This could be accomplished using a clearance mechanism to selectively remove conjugate from the systemic system. This paper describes the isolation of a mAb against CD and the use of this anti-CD mAb for the clearance of L6-CD from the circulation of nude mice bearing human tumor xenografts. A dramatic increase in the T / B ratio of L6-CD is obtained via the administration of the anti-CD mAb. EXPERIMENTAL PROCEDURES CD was purchased from Calzyme Laboratories (San Luis Obispo, CAI. The mAbs used are L6 (IgG2a), selective for an antigen on human carcinomas (12);1F5 (IgGZa), which binds to CD-20 (13); and ME-20 (IgGd, selective for a melanoma antigen (14); all the mAbs are murine. The cell line H2981 is a human lung adenocarcinoma to which L6 binds strongly (12). SMCC, 2-iminothiolane hydrochloride, and ebiotinamidocaproate N-hydroxysuccinimide ester were purchased from Pierce (Rockford, IL). Preparation of KLH-CD.To a solution of CD in 50 mM sodium borate, pH 8.0, was added 24minothiolane hydrochloride (final concentration = 2 mM). After 90 min a t 4 "C, the protein was separated from the salts on a G-25 Sephadex column (Pharmacia, Piscataway, NJ). KLH in 25 mM potassium phosphate, pH 8.0 was modified with sulfo-SMCC (final concentration = 21 mM). After 1h a t room temperature the reaction was centrifuged and the supernatant applied to a G-25 Sephadex column to remove unreacted sulfo-SMCC. The modified proteins were combined (5 mg of CD plus 100 mg of KLH) and after 2 h a t room temperature the reaction mixture was purified on G-75 Sephadex (using 0.1 M ammonium bicarbonate, pH 7.0, as eluant) to remove unreacted CD. KLH-CD was lyophilized and reconstituted with PBS for injections.
1043-18Q2/93/29Q4-Q353$Q4.QQlQ 0 1993 Amerlcan Chemical Society
354
Bioconjugate Chem., Vol. 4, No. 5, 1993
Conjugation of CD to mAbs. CD was conjugated to L6 and 1F5 as previously described for penicillin-G amidase (15). Briefly, CD was modified with SMCC to yield 1-2 maleimide groups per enzyme. The mAbs were treated with dithiothreitol to reduce approximately one disulfide per antibody. The two proteins were mixed, purged with nitrogen, and reacted a t 4 "C overnight. N-Ethylmaleimide (0.1 mM) was then added and the conjugate was applied to an S-300 Sephacryl gel filtration column (Pharmacia, eluted with 20 mM Tris, 60 mM NaC1, pH 7.5) to remove aggregates and unconjugated CD. Ionexchange chromatography on DEAE Sephadex (Pharmacia) was used to remove unconjugated mAb. The DEAE Sephadex column was washed with the above Tris buffer until the absorbance a t 280 nm was close to zero. Conjugate was eluted with 20 mM Tris, 1 M NaC1, pH 7.5. SDSPAGE indicated that the conjugate consisted of primarily 1:l mAb:CD adducts. mAbs to CD. BALB/C mice were immunized with 100 pg per injection of KLH-CD in Ribi adjuvant system prepared with an equal volume of Montanide ISA 50 oil (Seppic, Paris, France). The first injection of antigen/ adjuvant mixture was given either intraperitoneally or subcutaneously, followed by two consecutive intraperitoneal injections of KLH-CD in oil a t 3-4 week intervals. A final boost of KLH-CD in saline was performed 3 days before the mice were sacrificed. Spleen cells of immunized mice were hybridized with Px63-Ag8.653 (ATCC) mouse myeloma cells (16). The initial selection of hybridomas was based on the reactivity of cell supernatants with purified CD using ELISA assays. Hybridoma cell lines 64-25 and 102-26, which produced mAb specific for CD, were obtained by this procedure, cloned, and expanded in vitro. Culture supernatants were purified by affinity chromatography on immobilized anti-K light-chain column (17). An ELISA isotyping assay was utilized to determine the class of each mAb produced (Fisher Biotech, Pittsburgh, PA). Radioimmune Precipitation of CD. CD was iodinated withiodogen and Na[12511as described (18)resulting in a specific activity of 2.2 pCi/mg of protein. Incubations were performed in precipitation buffer (20 mM phosphate, 0.5 M sodium chloride, 0.5% NP-40, pH 7.5) with 1251labeled CD (2.5 pg/mL) and 100 pg/mL of either 64-26, 102-26, or ME-20 (control). Immune complexes were precipitated with anti-K-chain mAb immobilized on Sepharose 4B gel and washed four times with precipitation buffer. Samples were heated in nonreducing sample buffer (0.125 M Tris, 20% glycerol, 4 % SDS, 0.02 % bromophenol blue, pH 6.8) and separated on a 16% SDS-PAGE gel (Novex, Encinitas, CA). Bands were detected by autoradiography. Biotinylation of mAbs. To 0.9 mL of 64-25,102-26, or ME-20 (control) (all a t 1mg/mL) in PBS was added 0.1 mL of 0.5 M sodium borate, pH 8.5. To each was added E- biotinamidocaproate N-hydroxysuccinimide ester (final concentration = 0.4 mM). After 1h, unreacted biotin was removed by gel filtration on a G-25 column (PD-10, Pharmacia) with PBS/O.O2% sodium azide as eluant. Western Blot Analysis of 64-25 and 102-26. CD was separated on a 16% SDS-PAGE gel and transferred to nitrocellulose a t 25 mA for 18 h in 12 mM Tris, 96 mM glycine, pH 8.3 with 20% methanol. The filter was placed in blocking buffer (2% powdered milk in PBS) for 1 h; incubated a t 4 "C with 0.5 pg/mL biotinylated 102-26, 64-25, or ME-20 (control) for 18 h; and washed 4 X 10 min with PBS/O.l % Tween. The blot was then incubated for 1h in avidin-HRP (Vector Laboratories, Burlingame, CA)
Kerr et el.
(1:1000),washed as above, and developed with 4-chloro1-naphthol. ELISA Assay for Anti-CD mAbs. L6-CD (1pg/mL in 50 mM carbonate buffer, pH 9.6) was plated in 96-well Immulon plates (Dynatech Labs, Chantilly, VA). After storage overnight a t 4 "C, the plates were blocked with specimen diluent (Genetic Systems, Redmond, WA) for 1 h. Serial dilutions of biotinylated mAb were added to the plates, incubated for 1h a t room temperature, washed 5 times with PBS/O.l % Tween, and then incubated for 30 min with avidin-HRP. The plates were washed five times with PBS/Tween, and 3,3/,5,5/-tetramethylbenzidine in buffered substrate (Genetic Systems) was added. The reaction was stopped after 15 min by adding 3 N HzS04. The absorbance was then measured at 450 nm on a Biotek Instruments EL 312 microplate reader (Winooski,VT). 1F5-CD (0.06-60 pg/mL) was incubated for 4 h with 0.6 pg/mL biotinylated 102-26. These mixtures were then added to plates coated with L6-CD as before. After 1h, the wells were washed as before, incubated with the avidinHRP reagent, and developed as before. This was repeated with L6-CD a t concentrations of 0.06-60 pg/mL, biotinylated 102-26 a t 0.6 pg/mL, and plates coated with 1F5CD. Activity of CD in the Presence of 102-26. CD (1.4 pg/mL, 42 nM) or L6-CD (6.7 pg/mL L6, 44 nM) were incubated for 30 min a t room temperature with 350 nM, 1.18 pM, or 3.5 pM 102-26 in a total volume of 0.9 mL. 5-FC (0.1 mL at 30 mM in H2O) was then added and the reaction was incubated a t 37 "C. Aliquota (50 pL) were removed a t various times and quenched in 0.1 N HCl(1 mL) and the concentrations of 5-FC to 5-FU were determined as previously described (8). Clearance of L6-CD Using 102-26. Groups of nude mice bearing subcutaneous H2981 tumor xenografts (three mice per group) were injected intravenously with 100 pg of L6-CD. After 24 h, 102-26 (25, 50, or 100 pg) was injected, also intravenously. After an additional 24 h, blood was removed from the orbital plexus into a heparinized tube and centrifuged, and the plasma was frozen at -70 "C until used. To determine intratumoral enzyme concentrations, mice were sacrificed a t 24 and 48 h and tumors were removed. The presence of L6-CD in tumors and blood was quantified by measuring CD activity levels. To 1mL of 5-FC in PBS (3 mM a t 37 "C) was added 50 pL of plasma. A t various times, 50-pL aliquots were removed, and the concentrations of 5-FC and 5-FU were determined as described before (8). Enzyme levels in tumors were measured by mincing and homogenizing the tumors in 0.45 mL PBS containing 1 pg/mL aprotinin, 30 pg/mL phenylmethyl sulfonyl fluoride, and 5 mM ethylenediaminetetraacetic acid. To this was added 50 pL of 30 mM I3H]-5-FC (2 pCi/sample), and the reactions were incubated a t 37 "C with periodic mixing. At various time intervals 50 pL of each mixture were removed and quenched in 0.5 mL of methanol. The samples were evaporated to dryness, and the residue suspended in 20 pL of methanol. After centrifugation, 5 pL of supernatant was applied to Baker-Flex IB-F TLC plates (J. T. Baker, Philipsburg, NJ), which were developed in 6:l CHCld MeOH (Rf 5-FC = 0.2, Rf 5-FU = 0.7). The sections containing 5-FC and 5-FU were cut out and placed into scintillation vials. The radioactivity was measured on a Beckman 3801 scintillation counter in Optifluor (Packard Instruments, Downers Grove, IL) and the amount of 5-FU generated per gram of tumor tissue was determined. All
Bioconjtgate Chem., Vol. 4, No. 5, 1993 355
Anti-CD Antibodies for Conjugate Clearance
*.I
94674330-
-
a-
*
94674330-
I
’*li
P
1.o
20.1-
14.4-
Figure 1. Radioimmunoprecipitation and Western analyses of 102-26 and 64-25. (A) *%I-labeledCD was immunoprecipitated with 102-26 (lane 2), an antibody-free solution (lane 3), 64-26 (lane 4), or ME-20 (lane 5). The pellets were heated in sample buffer, centrifuged,and applied to a 16% SDS-PAGE gel. Lane 1 is ‘%I-labeledCD (100 ng) alone. (B) CD was run on a 16% SDS-PAGE gel, transferred to nitrocellulose, and probed with biotinylated 102-26 (lane 2), 64-26 (lane 3), and ME-20 (lane 4). Avidin-HRP was added, and the bands were visualized with 4-chloro-1-naphthol. Lane 1is a CoomassieBlue stain of purified CD.
0.1
0.01
1
pg/ml mAb
Figure 2. ELISA assays of the binding of 102-26 and 64-25 to CD: 102-26 (e),64-25 (A), ME-20 (A). CD was plated on immulon 96-well microtiter plates at 1 pg/mL. Biotinylated mAb’s at varying concentrations were added, incubated 1h, and washed, and avidin-HRP was added for 30 min and washed. Color was visualized with 3,3’,5,5’-tetramethylbenzidine.
reactions were linear for the time required to consume 1 pmol of 5-FC. RESULTS
mAb-CD Conjugates. The enzyme CD was conjugated to the mAbs L6 and 1F5 as previously described for the preparation of L6-penicillin-G amidase conjugates (15). This involved partial reduction of the mAb disulfides with dithiotheitol and reaction with maleimidyl-substitutedCD to yield thioether-linked conjugates. After purification, the conjugates were devoid of unconjugated mAb or CD, and consisted primarily of 1:l mAb:CD adducts. The conjugates retained enzymatic and binding activities according to assays as previously described (8). mAbs to CD. In preliminary experiments, it was determined that unconjugated CD was not an adequate immunogen to generate specific antibodies. As a result, CD was attached to KLH to render it more immunogenic. Conjugation of CD to KLH yielded a high molecular weight molecule free of unbound CD which could hydrolyze 5-FC to 5-FU. This was used with adjuvant to immunize BALB/C mice using the treatment schedule outlined in the Experimental Procedures section. Hybridization of spleen cells from treated mice with mouse myeloma cells and subsequent selection and cloning yielded two positive hybridomas secreting mAbs which recognized CD. These were designated 64-25 (IgG1) and 102-26 (IgGSb). The mAbs were purified using an affinity column with immobilized anti-rc-chain mAb (17). The anti-CD mAbs were tested for their specificities and abilities to precipitate the enzyme in radioimmune precipitation studies. Figure 1A shows that both 64-25 and 102-26 were able to precipitate CD from solution, whereas ME-20 was not. This shows that both mAbs recognize CD in solution. Western blot analysis of the binding of these mAbs to purified CD (Figure 1B) shows that both 64-25 and 102-26 bind to CD, but do not recognize the minor impurities present in the CD preparation (Figure lA, lane 1). This provides evidence that these antibodies are specific for CD. The ability of 102-26 and 64-25 to bind to CD was also measured in an ELISA assay. CD (1pg/mL) was plated in 96-well plates and biotinylated 102-26,64-25, or ME20 (control) was titrated into the wells. As can be seen in Figure 2,102-26 was specific for CD, whereas 64-25 was no better than ME-20 a t binding to CD. The reduced binding of 64-25 is consistent with the data shown in Figure
0
I
I
I
20
40
60
pg/ml mAb
Figure 3. Blocking of the binding of 102-26 to LG-CDand IF’& CD. LG-CD was plated on Immulon 96-well microtiter plates at 1pg/mL. 1F5-CD (v)at varying concentrations was mixed with biotinylated 102-26 at 0.6 pg/mL. These mixtures were added 4 h later to the microtiter wells. After 1h, the plate was washed with PBS/Tween, and incubated with avidin-HRP for 30 min and washed. Color was visualized with the 3,3’,5,5’-tetramethwas incubated with biotinylated 102ylbenzidine. L6-CD (0) 26 for 4 h and added to a 96-well plate coated with IF5-CD at 1pg/mL. The plate was washed and treated with avidin-HRP and 3,3’,5,5’-tetramethylbemidine. 1,indicating that 64-25 immunoprecipitated less CD from solution than 102-26 (Figure lA, lane 4 vs lane 2) and displayed weaker binding in a Western blot (Figure lB, lane 3vs lane 2). Subsequent experiments were, therefore, performed using 102-26. The ability of 102-26 to recognize CD bound covalently to mAbs was tested by incubating L6-CD (0.06-60pg/mL) for 4 h with biotinylated 102-26 (0.6 pg/mL) and the binding of 102-26 to 1F5-CD in an ELISA format was measured. This was repeated with 1F5-CD (0.06-60pg/mL) incubated with biotinylated 10226 a t 0.6 pg/mL, and the binding of 102-26 to L6-CD was assessed. Figure 3 shows that 102-26 did bind to both LG-CD and 1 F M D and that the binding could be blocked by the opposite immunoconjugate. Effect of 102-26 on CD Activity. The effects of 10226 on the enzymatic activity of free CD or L M D was determined. CD and L6-CD were each incubated with an 8-, 24-, and 75-fold molar excess of 102-26. The ability of CD to facilitate the conversion of 5-FC to 5-FU was then measured. There was virtually no decrease in activity of CD in the presence of 8 times as much mAb, while the
358
Bloconlugete Chem., Vol. 4, No. 5, 1993
Kerr et al.
DISCUSSION
I
I
0
20
40
60
80
Molar Ratio 102-26/CD
Figure 4. Effect of 102-26 on CD activity. CD (0) or L6-CD (m) at 42 and 44 nM, respectively, were incubated for 30 min with 102-26 at 8-, 24-, and 75-fold molar excess. CD activity was then assessed as outlined in Experimental Procedures. Points are the average of duplicate determinations which differ by less than 5%. Table I. Clearance of LGCD by 102-26. time postinjection 102-26 of LG-CD (h) (bg) 24 0 48 0 48 25 48 50 48 100
% ID/g of tumor
% ID/mL of blood
4.3 f 0.8 3.9 f 1.1 3.9f 2.8 2.7 f 0.5 1.9f 1.0
5.4 f 0.2 2.9 f 0.6 1.5* 0.3 0.3* 0.1 38
a Mice (three per group) were injected with 100 pg per mouse of L M D . 24 h later, 102-26 was administered. At 48 h, blood and tumors were taken and CD activity was measured. Below the limit of detection.
activity of L6-CD was reduced 20%. A t a 75-fold excess of 102-26, the reduction of either CD or L6-CD was 30 % (Figure 4). Clearance of L 6 4 D by 102-26. The ability of 10226 to reduce the levels of blood-borne CD activity in mice treated with L6-CD was tested in nude mice bearing subcutaneous H2981 tumors. The levels of L6-CD in the tumors and blood of mice that received 100 pg of L6-CD did not undergo dramatic change between 24 and 48 h (Table I). At 24 h the T / B ratio was 0.8, while after 48 h this ratio had only increased to 1.3. To test the ability of 102-26 to increase T / B ratios, mice were injected with L6-CD, followed 24 h later by varying amounts of 102-26. The T/B ratios of L6-CD were determined 24 h after the administration of 102-26. As shown in Table I, 102-26 was able to increase the T / B ratio of L6-CD in a dosedependent manner. Administration of 25 pg of 102-26 increased the T / B ratio to 2.6, which was twice as high as in untreated mice. By increasing the dose of 102-26 to 100 pg, the T / B ratio was raised to >38. At this dose, the level of CD activity in the blood was below the limit of detection. The increase in the T / B ratio after 102-26 administration was not accompanied by corresponding decreases in the levels of tumor-associated conjugate. Treatment of mice with 100 pg of 102-26 resulted in a 2.1-fold decrease in enzymatically active L6-CD (from 3.9 to 1.9% ID/g), whereas blood levels were reduced greater than 58-fold. The individual mice that received 100 pg of 102-26 retained 3.0,1.2, and 1.4% ID/g intratumoral CD activity. The reduced amount of CD activity measured in the tumors may actually reflect the clearance of L6-CD from the tumor vascular bed and not removal of antigenbound conjugate. Thus, we have shown that 102-26 dramatically reduces the levels of CD activity in the blood without severely affecting tumor levels.
mAbs are currently the subject of a great deal of research for the delivery of drugs, toxins, and radionuclides to tumor cell populations (19,ZO). While some promising preclinical data have been obtained (21-23), there is a number of limitations to this form of cancer therapy, particularly for the treatment of solid tumors. These include the limited amount of drug that can be attached to the mAb, the heterogeneity of antigen expression in tumors (24, 25), and the fact that macromolecules cannot easily penetrate into solid tumors (24,261. We have worked on an approach that is designed to address these limitations (27). Briefly, a mAb-enzyme conjugate is administered and allowed to bind to tumor cells and clear from the circulation, a t which time a relatively nontoxic prodrug is given systemically. The targeted enzyme is selected for its ability to convert the prodrug into a potent anticancer drug. This approach amplifies the number of drugs delivered to the tumor because of the catalytic conversion of prodrug to drug. Drugs that are released on the outside of antigen-expressing cells should be able to enter into antigen-negative cells and penetrate to interior portions of the tumor, otherwise inaccessible to the conjugate. Many example of this approach have been described recently (27-29). It is apparent that optimal therapy using mAb-enzyme for prodrug activation will require highT/B ratios of conjugate. In this study we have addressed this issue by isolating a mAb that can bind to the enzyme component of a mAbenzyme conjugate and showing that the antienzyme mAb accelerates the clearance of the conjugate in the systemic circulation. The mAb 102-26 recognizes both free and conjugate-bound CD. Upon binding to CD, 102-26 leads to minimal loss in enzyme activity (Figure 4). The use of such a clearance mechanism was necessitated by the observation that L6-CD is not rapidly cleared, and has a very modest T / B ratio (Table I). 102-26 was able to dramatically increase the T / B ratio of L6-CD from 1.3 to >38. This was achieved mainly by reducing the blood level of conjugates by a factor of a t least 58. The finding that the tumor level was reduced 2.1-fold may be due to both loss of the conjugate from the tumor vascular bed, and to some loss of antigen-bound conjugate. The amount remaining in the tumor (1.9% ID/g) is comparable to that obtained with other targeting strategies (301, and may be sufficiently high to generate cytotoxic quantities of 5-FU intratumorally . The approach described here is related to a recently published report by Sharma and co-workers, who have shown that the anti-carboxypeptidase G2 (CPG-2) mAb SB43 can accelerate the clearance of a mAb-CPG-2 conjugate (9). However, SB43 inactivated CPG-2 such that there was approximately 95% inactivation with a 6-fold excess of SB43. As a result, it was necessary to modify SB43 with galactose to accelerate its clearance, in order to avoid the inactivation of tumor-associated CPG2. With the mAb against CD described here, this was not necessary. The use of an antienzyme mAb for conjugate clearance complements other recently described clearance mechanisms which include modification of conjugates with sugars (301,use of anti-idiotypes (28,31,32), plasmapheresis (331, and clearance of biotinylated mAbs with avidin (28,34). The advantages of the approach described here is that potentially any enzyme conjugate can be cleared; and free enzyme that may either contaminate the conjugate preparation or results from conjugate fragmentation can be cleared as well. By lowering the level of non-tumor-
AntiCD Antibodies for Conjugate Clearance
associated conjugate it should be possible to increase the amount of prodrug that can be administered. Such studies are now underway. ACKNOWLEDGMENT We would like to thank George Schreiber for valuable discussions and Ana Wieman for preparation of the manuscript. LITERATURE CITED (1) Kemeny, N., Lokich, J. J., Anderson, N., and Ahlgren, J. D. (1993)Recent advancesin the treatment of advanced colorectal cancer. Cancer 71,9-18. (2) Speyer, J. L., Collins, J. M., Dedrick, R. L., Brennan, M. F.,
Buckpitt, A. R., Londer, H., DeVita Jr., V. T., and Myers, C. E. 11980)Phase I and Dharmacoloeical studies of 5-fluorouracil administered intraperitoneally. kancer Res. 40,567-572. (3) Leichman, C. G., Leichman, L., Spears, C. P., Rosen, P. J., Jeffers, S., and Groshen, S. (1993) Prolonged continuous infusion of fluorouracil with weekly bolus of leucovorin: a phase I1study in patients with disseminated colorectal cancer. J. Natl. Cancer Inst. 85,41-44. (4) Grem, J. L. (1990)Levamisole as a therapeutic agent for colorectal carcinoma. Cancer Cells 2, 131-137. (5) Morrell, L. M., Bach, A., Richman, S. P., Goodman, P., Fleming, T. R., and MacDonald, J. S. (1991)A phase I1 multiinstitutionaltrial of low-doseN-(phosphonoacety1)-1-aspartate and high-dose 5-fluorouracil as a short-term infusion in the treatment of adenocarcinomaof the pancreas. Cancer 67,363366. (6) Nishiyama, T., Kawamura, Y., Kawamoto, K., Matsumura, H., Yamamoto, N., Ito, T., Ohyama, A., Katsuragi, T., and Sakai, T. (1985)Antineoplastic effects in rats of 5-fluorocytosine in combination with cytosine deaminase capsules. Cancer Res. 45,1753-1761. (7) Mullen, C. A., Kilstrup, M., and Blaese, R. M. (1992)Transfer of the bacterial gene for cytosine deaminase to mammalian cells confers lethal sensitivity to 5-fluorocytosine: A negative selection system. Proc. Natl. Acad. Sci. U.S.A. 89,33-37. (8) Senter, P. D., Su, P. C., Katsuragi, T., Sakai, T., Cosand, W. L., Hellstrom, I., and Hellstrom, K. E. (1991)Generation of 5-fluorouracil from 5-fluorocytosine by monoclonal antibodycytosine deaminase conjugates. Bioconjugate Chem. 2,447451. (9) Sharma, S.K., Bagshawe,K. D., Burke, P. J., Boden, R. W., and Rogers, G. T. (1990)Inactivation and clearance of an antiCEA carboxypeptidase G2 conjugate in blood after localisation in a xenograft model. Br. J. Cancer 61, 659-662. (10)Yokota, T., Milenic, D. E., Whitlow, M., and Schlom, J. (1992)Rapid tumor penetration of a single-chain Fv and comparison with other immunoglobulin forms. Cancer Res. 52,3402-3408. (11) King, D. J.,Mountain, A., Adair, J. R., Owens,R. J., Harvey, A., Weir, N., Proudfoot, K. A., Phipps, A., Lawson, A., Rhind, S. K.. Pedlev. B.. Boden. J.. Boden. R.. Beeent. R. H. J.. and Yarronton, G. T. (1992)Tumor locdiza~onof engineered antibody fragments. Antibody, Immunoconjugates, Radiopharm. 5,159-170. (12)Hellstrom, I.,Horn, D., Linsley,P., Brown, J. P.,Brankovan, V., and Hellstrom, K. E. (1986)Monoclonal mouse antibodies raised against human lung carcinoma. Cancer Res. 46,39173923. (13) Clark, E.A,, Shu, G., and Ledbetter, J. A. (1985)Role of the Bp35 cell surface polypeptide in human B-cell activation. Proc. Natl. Acad. Sci. U.S.A. 82,1766-1770. (14)Hellstrom, I., Marquardt, H., Maresh, G., Wan, H., Anderson, J., and Hellstrom, K. E. An internalizing murine monoclonal antibody (ME20) binds to an antigen with high specificity for human melanoma. In preparation.
Bioconjugate Chem., Vol. 4, No. 5, 1993
357
(15) Bignami, G. S.,Senter, P. D., Grothaus, P. G., Fischer, K. J., Humphreys, T., and Wallace, P . M. (1992) N-(4'-hydroxyphenyacetyl)palytoxin: A palytoxin prodrug
that can be activated by a monoclonal antibody-penicillin-G amidase conjugate. Cancer Res. 52,5749-5764. (16)Kohler, G. and Milstein, C. (1975)Continuous cultures of fused cells secreting antibody of predefined specificity. Nature 256,495-497. (17) Yelton, D. E., Desaymard, C., and Scharff, M. D. (1981) The use of monoclonal anti-mouse immunoglobulin to detect mouse antibodies. Hybridoma 1, 5-11. (18)Markwell, M. A. K. and Fox, C. F. (1978)Surface-specific iodination of membrane proteins of viruses and eucaryotic Biocells using 1,3,4,6-tetrachloro-3a,6a-diphenylgycouril. chemistry 17,4807-4817. (19) Pimm, M. V. (1988)Drug-monoclonalantibody conjugates for cancer therapy: potentials andlimitations. Crit. Reu. Ther. Drug Carrier Syst. 5,189-227. (20) Pietersz, G. A. and McKenzie, I. F. C. (1992)Antibody conjugates for the treatment of cancer. Zmmunol. Rev. 129, 57-81. (21) Johnson,D. A., Baker, A. L., Laguzza,B. C.,Fix, D. V., and Gutowski, M. C. (1990) Antitumor activity of L/lC2-4desacetylvinblastine-3-carboxyhydrazide immunoconjugate in xenografts. Cancer Res. 50, 1790-1794. (22) Beaumier, P. L., Venkatesan, P., Vanderheyden, J.-L., Burgua, W. D., Kunz, L. L., Fritzberg, A. R., Abrams, P. G., and Morgan, A. C., Jr. (1991)lWReradioimmunotherapy of small cell lung carcinoma xenografts in nude mice. Cancer Res. 51,676-681. (23) Friedman, P. N., Mcandrew, S. J., Gawlak, S. L., Chace, D., Trail, P. A., Brown, J. P., and Siegall, C. B. (1993)BR96 sFvPE40, A potent single-chain immunotoxin that selectivelykills carcinoma cells. Cancer Res. 53,334-339. (24) Weinstein, J. N.andvan Osdol,W. (1992)Earlyintervention in cancer using monoclonal antibodies and other biological ligands: Micropharmacology and the "binding site barrier". Cancer Res. 52,27475-2751s. (25) Cobb, P. W., and LeMaistre, C. F. (1992)Therapeutic use of immunotoxins. Semin. Oncol. 29,6-13. (26) Jain, R. K. (1991)Haemodynamic and transport barriers to the treatment of solid tumours. Znt. J . Radiat. Biol. 60, 85-100. (27) Senter, P. D., Wallace, P. M., Svensson, H. P., Vrudhula, V. M., Kerr, D. E., Hellstrom, I., and Hellstrom, K. E. (1993) Generation of cytotoxic agents by targeted enzymes. Bioconjugate Chem. 4,3-9. (28) Bamias, A. and Epenetos, A. A. (1992)2-Step strategies for the diagnosis and treatment of cancer with bioconjugates. Antibody Immunoconjugates, Radiopharm. 5,385-395. (29) Bagshawe,K. D. (1990)Antibody-directed enzyme/prodrug Trans. 18,750-752. therapy (ADEPT). Biochem. SOC. (30) Mattes, M. J., Sharkey, R. M., Goldenberg, D. M., and Ong, G. L. (1991)Manipulation of Blood Clearance of Antibodies by Galactose Conjugation. Antibody, Zmmunoconjugates, Radiopharm. 4,877-884. (31) Sharkey, R. M., Boerman, 0. C., Natale, A., Pawlyk, D., Monestier, M., Losman, M. J., and Goldenberg, D. M. (1992) Enhanced clearance of radiolabeled murine monoclonal antibody by a syngeneicanti-idiotype antibody in tumor-bearing nude mice. Znt. J . Cancer 51,266-273. (32) Pedley, R. B., Dale, R., Boden, J. A., Begent, R. H., Keep, P. A., and Green, A. J. (1989)The effect of second antibody clearance on the distribution and dosimetry of radiolabeled anti-CEAantibody in a human colonictumor xenograft model. Int. J . Cancer 43, 713-718. (33) Henry, C. A., Clavo,A. C., and Wahl, R. L. (1991)Improved monoclonalantibody tumor/background ratios with exchange transfusions. Int. J . Radiat. Appl. Znstrum. [B] 18,565-567. (34) Paganelli, G., Malcovati,M., and Fazio, F. (1991)Monoclonal antibody pretargeting techniques for tumour localization: The avidin-biotin system. Nucl. Med. Commun. 12,211-234.
