CA2081281A1 - Activation of glucuronide prodrugs by ligand-beta-glucuronidase conjugates for chemotherapy - Google Patents

Abstract

Activation of Glucuronide Prodrugs By Ligand-.beta.-Glucuronidase Conjugates for Chemotherapy Abstract of the Disclosure A method for enhancing the specificity of chemotherapy is disclosed. A monoclonal antibody that can bind antigens preferentially expressed on the surface of malignant cells is used to form a conjugate with the enzyme .beta.-glucuronidase.
Glucuronide analogues of therapeutic drugs, possessing reduced toxicity and enhanced efficacy compared with the parent compound, are administered subsequent to the administration of the .beta.-glucuronidase-anitbody conjugate such that the glucuronide analogues are enzymatically converted to therapeutic agents at malignant cells that are able to bind .beta.-glucuronidase-monoclonal antibody conjugates.

Description

Activation of Glucuronide Prodrugs by Ligand~ Glucuronidase Con; ugates f or Chemotherapy DESCRIPTION OF THE INVEMTION
.

1~ Technical Field The pre~ent invention rela~es to the use of a `1 B -glucuronidase~antibody conjugate to convert glucuronide prodrugs to highly activs therapeutic agents preferentially at maliynant cells that are bound by the enzy~e-antibody conjugate for the purpose of increasing the effic~cy of chemotherapyO

2. Background Art A variety of methods in the prior axt have been employed to increase drug specificity by linking therapeutic agents to monoclonal antibodies tha~ bind to antigens or receptors pr~*erentially ~xpress~d on th~ sur~ace of malignant cells~
While direct conjugation of drugs to antibodies can increase sp~cific targeting of drugs to malignant cells, this method suffers ~rom several drawbacXs. Prac,tical limits on drug-loading or slow internaliza~ion o~ drug conjuga~e~ into malignant cells may result in poor e~icacy. In addition, cells expressing low levels of antigen in heterogeneous cell populations may also escape therapy.

Indirect dru~-targeting employing antibody targeted-enzyme activation o~ prodrugs has al~o been described in the literature, inter alia, by Bagshawe et al in Br. J. Cancer, Vol. 56,531-532 ~1987) and Br. J. Cancer, VolO 58-700-703 (1988); and by Senter el al i~ Proc. Natl. Acad. Sci., US~, Vol. 85,48~2-4846 (1988) and Cancer Res., Vol. 49:5789 5792 (1989). In this method, enzyme rather ~han drug is linked to an an~ibody that binds , .

:

c~

antigen preferentially expressed on the surface o~ malignant cells~ The antibody-enzyme complex is then targe~ed to malignant cells, thus allowing the conjugated enz~me to accumulate at the disease site. A latent, nontoxic "prodrug" i5 then introduced so that prodrug coming into contact with targeted enzyme at the site of the disease can be enzymatically converted to the active parent compound which can then xert its therapeutic effect.
This method may provide some advantages compared with chemoimmunoconjugates, including accumulation of higher drug concentrations at the site of the disease, less sensitivity to malignant cell antigen heterogeneity and the ability o~ creating defined immunoconjugates through genetic engineering.

Prior to this invention, the enzymes ~mployed -to ~orm the antibody-enz~me conjugates have been limited ~o ~acterial enzymes ! or to enzymes that are present at high concentrations in human serum. Bacterial enz~mes ara highly immunogenic and are known to induce strong immune responses in human pa~ian~s, limiting the j number of treatments that patients can receive. Enzymes present at high concentrations in human serum may prema~urely activate administered prodrug before it can reach the antibody-enzyme conjugate at the site of the disease, decreasing the specificity of treatment.
.
`~.9: The present in~ention employes ~ glucuronidase for the ; ~1

3 formation of an antibody-enzyme conjugate to activate glucuronide .~

~i , i 1 - :
9 , , " . ,.,. , . ,. . . . :, : ' ' ' ` ' ' ' ' ' . ' ' r3~ 5~ ~

derivatives of therapeutic agents. The use of ~-glucuronidase and glucu.ronide prodrugs for therapy offers improvements in several areas of the curre~ art. (1) The concentratio~ of g-glucuronidase in human serum i5 ~nown to be very low so that glucuronide prodruys should be stable in the blood af-ter i.v.
administration. (2) Mammalian tissues express the enzyme UDP-glucuronosyltransf~rase, a ~lass of xenobiotic-detoxi~ication enzymes that can reverse the reaGtion catalyzed by ~-glucurQnidase. Many therapeutic agents have been found to be converted to their glucuronide derivatives in vivo including aniline mu~tard, 7-hydro~yellip~icin~, 4'-epidoxorubicin, l-naphthol and AZT, 3'-azido-3' deoxyth~midine (Connors et al., Biochem, Pharmacol. 2~- 197~-19~0~ 1973; Matouh et al., Drug Metab. Dispos. 12~ 119, 1984; Weenen et al., EurO J. Cancer Clin. Oncol. 20: 919-9~6, 1984, Redegeld at al., J. Pharmacol.

Exp. Ther. 224: 263-267, 1988; and Blum et al., Am. J. MedO 85:

189-194, 19883. These studies show that glucuronide prodru~s are common metabolites in vivo and that glucuronide prodrugs should be resistant to premature activation by endogenous ~-glucuronidase in vivo. These studi~ also suggest that ', activated prodrug not taken up by tumor cells may be reconverted :, to the glucuronide prodrug af~er passing through organs ~ontaining high UDP~glucuronosyltransferase activi~ies.
~1 (3) ~-glucuronidase can be obtained from a wide variety of , mammalian and prokaryotic sources. Conjugates can thera~ore be i ~orm~d employing ~-glucuronidase from dif~erent species. This '~

.1 4 `:~

provide~ two major advantages over other m~hods in the current art. First, human ~glucuronidase can be used in the antibody-enzyme conjugate, thus minimizing the possibili-ty of inducing a host i~mune response against the enzyme porticn of the conjugate. Second~ a series of conjuga~es could constructed, each one using ~-glucuronidase from a different species. Upon induction of a host i~mune response against ~-glucuronidase fr~m one species, therapy could still be continued by using a conjugate employing ~-glucuronidase from a~other species~ The ability to administer multiple treatmenks of conjugate and prodrug offer~ a signi~icant advantage over me~hods limited to one or two treatments due to the induction o~ a host immune response against the conjugats. (4) ~-glucuronidase is highly speci~ic for the glucuronyl residue of glucuronidP prodrugs but has little specificity for the conjugated therapeutic agent. A
wide variety o~ glucuronide prodrugs can be synthesized and employed for therapy. This allows ~lexibility in apply the most ef~ective therapeutic agent ~or therapy. In addition, combinatlon chemotherapy, with its well known advantages, can easily be carried out by simultaneously administering several different glucuronide prodrugs. (~) Glucuronide prodrugs are ,highly hydrophilic and are limited i~ their abili~y to cross the cell membrane. ConvQrsion of a tox.ic tharapeutic agent to a glucuronide prodrug greatly reduce~ its toxis:ity~

,; . , :

~ .
~. ' ` ' ' ~7 DISCLOSURE OF THE INVENTION

The present inventlon dls~_loses a method for enhancing the specificlty of chemotherapy. Glucuronic acid analogues of therapeutic agents are synthesized in order ~to crea~_e glucuronide prodrugs with reduced toxicity. A conjug,~te between an antibody or other cell blnding ligand and the enzyme ,L4,-glucuronidas~ie is also created. The antibody or ligand portion of th~, con~ugate is selectel such that it binds to antiyens or receptors pref'erentially expressed on the surf'ace of malignant cells. Th~e conjugate is all_w~d to bind to the malignant cells and the glucuronide p~odrug is subsequently administered so th,2,t it can be selectively converted to the paxenk therapeutic agent by ,~,-glucuronidase at th~e malignant cells.

, :!
. Brief description of figures:
.i FIG. 1 de,picts the cytotoxic eff'ects of ~M, B~AMG or -the combined eff!ect of ~,HAMG and R-glucuronidas~e ~o human cells.

Panel A shows results using COLO 205 (human colon carcinoma) ~" cells and panel B shows results With WISH (human "mniotic3 cells~

. ~ .
' , ~

2 ~ ~? ~

FIG. 2 depicts the in vitro activation of BHAMG prodrug by ~ -glucuronidase-a~tibody conjugate. Panel A indicates COLO 205 cells and panel B indicates WISH cells. Cells were preincubated with (a,b) medium alone, (c) 20 ~g/ml ~-glucuronidase-antibody conjugate or (d) 50 ~g/ml conjugate. After washing the cells, they were incubated with (a) 200 ~g/ml HAM or ~b,c,d) 200 ug/ml B~MG for 24 h. Cellular protein synthesis rate was measured 24 h later.

FIG. 3 shows the enzyme and antihody activities o~ RH1-~&~ The enzyme activity of ~ ~glucuronidase before and after coupling to Mab RHl are indicated in panel A. Enzyme~linked immunosorbent assay measurement o~ antigen-bindins activitiss of ~ab R~ 1 or RHl ~G to antigen-positive A5-3OD cells or antigen negative HepG2 cells are shown in panel B. Combined immunosorbent and glucuronidase activity measurement of ~ G is shown in panel '~
FIG. 4 depicts the in vitro growth inhibition o* hpa-toma cells by HAM and BHAMG. Protein synthesi~ rate of AS-30D rat hepatoma (pan~1 A) and HepG2 human hepatoma ~panel ~) cells are shown ,a~ter exposure o~ cells to HAM, BHAMG, or BHAMG plus 10 units ~-glucuronida~e.
,, ~
FIG. 5 depicts the in vitro activation of BHAMG by RHl~ ~G.
AS-30D~A) or ~epG2~B) cells were prt~incubatetl with different _ ' ., ,.~ , ; , , ~ . .

concentratlons of RHl~~G for 30 min at room tempera~ure~ Cells were then washed and exposed ~o B~G for 24 h. Cellular protein synthesis rate was measured 24 h later and is co~pared to prot2in synthesis of control cells expose~ ~o 90~M BHAMG only.

~IG. 6 depicts the speci~icity of RH1-~G for AS~30D cells.
AS-3OD cells (p~nels a and b~ and HepG2 cQlls (panels c and d) were preincubated with ~Hl- ~G alone (panels a and c) or RH1- ~G
plus excess RHl F~ab I )2 (panels b and d) for 1 h at room temperature. Cells were washed, exposed to 90 ~M BH~MG for 24 h and incubated an additional ~4 h in fresh medium be~ore measuring cellular protein synthesis rat~. Results are expressed as p,~rc nt protein sy~thesis in treated cells compar~d to control cells exposed to ~0 ~M ~HAMG without addition o~ RH~G.

.

FI~. 7 depicts the specificity of BHAMG ac~ivation at antigen-positive tumor c~lls. AS~30~ (panel A) and HepG2 (panal B~ cells were pxeincubated with R~ G or medium~ washed and incubated for 24 h in the presence of HAM or BHAMG.

FIG. 8 depicts the survival o~ rats bearing AS-30D ascites a~ter receiving vaxious treatments.

.1 .

l EX~MPI;E 1 `i ' !
' 8 .

~, .
:

(i) Synthesis of drug and glucuronide prodrug Using methods known in the art, p-hydroxyaniline mustard ~HAM) and its glucuronide prodrug, [t~tra n-buty] ammonium salt of (p-2-chloroethylaminophenol- ~ ~glucopyranosid) uronic acid)]
~BHAMG), were synthesized and their structures were con~irmed by N~R and melting point dete~mination. The chemical structures of HAM and ~AMG are shown as follows:

(a) HAM

-3 HC~ ~ N ~ CH2C}~f2C:i H2~ H2C~
,j ,.j .:~

CC)O~ M (i~

~ N ~ ~ ~2~ H2GI
OH~ S;H2~;H2 C) H

'"I `~:
, ,.
g _ ! ~ .
;~ .
'`l;` :
is ~

Mab 12.8 is a murine IyG monoclonal antibody that binds strongly to COLO 205 human colon carcinoma cells but does not react with WISH human amniotic cells. COLo 205 and WISH cell lines were obtained from American Type Culture Collectlon, Rockville, MD. Mab 12.8 also reacts positively with about 40% of human colon carcinoma tissues but has limited reaction with normal human tissues (Hwa, M.S. Thesis, National Defense Medical Center, Taipei, Taiwan, 19~6). Ma~ 1~.8 F(ab')~ fragments were generated by pepsin digestion.

;

-glucuronidase (EC 3.2.1.3~) from E. coli (Sigma Chemical Company, St. Louis, MO) was linked to Mab 12.8 F(ab')z ~ia a disulfide bond with the heterobifunctional cr~s-linking agent SPDP ~N-~uccinimidyl-3(2-pyri~yldithio) propionate] (Pharmacia LKB Biotechnology, Uppsala, Sweden) to create the con~ugat~ 12.8 F(ab' )a ~ ~G, having the ~ollowing ~ormula, BG-S S-Mab. An average of 2.4 and 2.0 2-pyridyl groups were introduced into Mab 12.8 F(ab')~ and ~-glucuronidase molecules, respectively.
Derivatized enzyme was reduced with 1 mM dithiothrQitol and a~ter removing excess reducing agent by gel filtrati~n, deri~itized enzyme and antibody were mixed and incubated overnight at room temperature.

Enzyme-antibody conjugate was puri~ied by HPLC gel filtration on two 7.5 mm x 30 cm SW 300 columns (Waters) in series. ~he ~-glucuronidase activity o~ ~he conjugate was ~ .
~ -- 1 0 --~1~

-.,.. , ... .. , .. . , ,,, :, . , . .. : . . . : ~ .. ,....... ~ . .

6~Q~ ?"

measured (Stahl PD and Fishman WH, ~ D-glucuronidase. In:
Methods of Enzymatic Analysis, 3rd Edition. (Ed. Bergmeyer HU), vol 4, pp. 246~256, Verlag Chemie, Weinhei~l, 1984). The conjugate retained an enzymatic activit~ of 0025 U/~ g (1 U
liberates 1.0 ~ mole p-nikrophenol from p-nitrophenol glucuronide per hr at 37~, abou-t 20% o~ original ~-glucuronidase activity. The conjugate retained full anti~ody activity compared with unconjugated antibody as measured by enzyme immunoassay (EIA) using fixed COL0 205 cells as antigon source.

(iii) Activity of drug and prodrug The cytotoxicity of ~L~M and BXAMG to cells was deter~ined by measuring incorporation o~ [3~-leucine into cellular protein of cells exposed to these drugs for 24 h. COL0 205 or WISH cells were plated overnight in 96 well microtiter pla~es at 20,000 cells per well. Cells were maintained in RPM~ 1640 medium (Gibco BRL, Grand Island, NY) supplemented with ~% heat-inactivated f~tal c~lf serUm, 1000 U/ml penicillin and 100 ~g/ml streptomycin. Serial dilutions of ~M or BHAMG in RP~I medium containing 5~ fetal calf serum were added to cells ~or 24 h at 37. Cells were subsequently washed once with sterile phosphate buf~ered saline (PBS), incubat2d an additional 24 h in fresh medium and then puls~d for 2 h with rjH]-leucine (1 ~Ci per well) in leuci~e-fxee medium. Radioactivity of trichloroacetic ~ I ~

:1 .
.J

acid precipitated protein was measured in a Beckman LS 6000 series liquid scintillation counter.

BHAMG conversion to cytotoxic HAM was tested by incubating BHAMG and ~-glucuronidase (10 U per well) for 2~ h at 37. Cells were washed once with PBS and a~ter an a~ditional 24 h incubation, the rate of protein synthesis was measured. The in vitro activation of ~H~MG by Mab 12.8 F(ab')~ glucuronidase conjugate was tested by preinc~bating pla-ted cells with the indicated concantrations of conjugate for 1 h at room : t2mperature. A~ter washing cells once with PBS, 200 ~g/ml BHAMG
was added and the assay carried out as described above. All experiments were performed in triplicate.

FIG~ 1 demons~rates that B ~G was less toxic than H~ to human cells. Results are presente~ at % protein syn~hesis relative to control cells, measured as [~]-leucine incorporation , (200,000 and 290,000 cpm for COLO 205 and WISH control cells, respectively). COLO 205 and WIS~ cells were about equally 1 sensitive to H~M with IC5~ values ~concentration of drug that inhibits cellular protein synthesis by 50%) of 34 and 49 ~U~
~, respecti~ely. The water soluble prodrug BHAMG was less toxic to both COLO 205 and WISH cell lines with IC~o values of 1890 and 1320 ~M, rsspectively. Simultaneous addition of 10 U
B-glucuronidase and BHAMG to cells resulted in a cytotoxic effect il ~ as great as addition of HAM alone, indicating that enzymatic '`l - 12 -i - , ,1 A, . ~ . ~ . ., i ~J ~ , f~ ~ ~
cleavage of the water soluble glucuronide group coverts ~H~MG to the cytotoxlc alkylating agent H~M~ Addition of ~-glucuroni.dase alone had no effect o~ cells.

(iv) Specific therapy of human tumox cells.

12.8 F(ab')z -~G, a conjugate formed betwee~ Mab 12.8 F(abl)~
and ~-glucuronidase, can bind to antigen-positive COLO 205 cells but is unable to ~ind antigen-negative WIS~ cells. FIG. 2 showns the e~fect of pretreaking COLO 205 (panel A) or WIS~I (panel B) cells with enzyme-antibody conjugate. Cells were preincubated with (a,b) medium alone, (G) 20 ~g/ml ~-glucuronidilse~antibody conjuga~e or (d) 50 ~g/ml conjugzte fox 1 hr at room temperatureO
The cells were then washed and i.ncubated with (a) 200 ~g/ml HAM
or (b,c,d) 200 ~ /ml BHAMG ~or 2~ h Cellular protein synthesic rate was meisured 24 h laterA Reiults are. presented as %
protein synthesis relative to control cells, measured as [~H]-leucine incorporation (140,000 and 290,000 cpm for COLO 205 and WISH control cells, respectively) and represent the mean of results ~rom 3 separate wells. Error bars show the standard error of the mean. Pretreatment o~ COLO 205 cells with conjugate greatly increased th~ cyto~oxic ef~ect of BH~
Antigen-negatlve WISH cells, in contrast, were still resistant to BH~M~ after pretreatment with conjugate. The addition of s antibody, enzyme or conjugate alone had no e~ect on cells.

,1 There results indicate that suf~icient ~-glucuronidase can be .1 , -- 13 --., ,' i '' : ' .' ~ '; . ' ' ., '. ., ' ' i , ' " ' . ' , : ' ' . ',' ' ' , : , ' j . . , ~ 3~ J~
targeted to tumor cells to convert glucuronide prodrugs ~o actlve - anti-neoplastic agents. In addition, the lack of ~oxicity to a~tigen-negative cells indicates that prodrug activation is specif ic .

:`
, ~:~
The drug HAM and glucuronide prodrug ~AMG were synthesized as before~

':.' (i) Conjugation of ~-glucuronidase ~o antibody Mab RHl is a murine IgG~ monoclonal antibody developed in our laboratory that binds strongly to AS~30D cells (provided by , Dr. J.P. Chang, ~nstitute o~ Zoology, ~c~demia Sinica, RoOoC~ ) but does not bind HepG2 human hepatoma cells (obtained ~rom the ~ American Type Culture Collection~.

:i, RHl- G was formed by linking ~rglucuronidase ~o Mab RHl via `' a thioether bond. A maleimido group was first introduced into !~ the immunoglobulin molecule with the heterobifunctional .~, crosæ-linking agent SMCC ~Succinimidyl 4-(N-maleimidomethyl) I cyclohexane-l carboxylate, Pierce Chemica:L Company, Rockord, I IL). A sevenfold molar excess of SMCC dissolved in dioxane (3 !1 mg/ml) was ~dded to Mab RHl ~5-10 mg/ml) in PBS ~or 45 min a-t 37 ~ - 14 -C. Excess SMCC was removed by gel filtration on Sephadex G-25, Modlfied R~l antibody was then reacted with ~hiol groups present in ~-glucuronidase. Lypholized ~glucuronidase was dissolved in PBS (3 mg/ml) and passed through Sephadex G 25. ~-glucuronidase was mixed With derivatized ~gG, concentr~ted by ultrafiltration and r~acted overnight at 4C. All coupl.ing reactio~s were performed in PBS containing 1 mM EDT~, deoxygenated by boiling and sparging with nitrogen.

RH1-~G was p~ified in a two step processO Uncoupled B-glucuronidase was removed from the conjugate by protein A-Sepharose affinity chromatographyO Free Mab ~Hl was then removed by ion ~xchange chromatography on a DEA~ 5 PW
high-performance liquid chromatography ~olumn (Waters) by eluting with a linear gradient of NaCl in 20 m~ Bis-tris, p~ 600. Eluted conjugates were concentrated by ultrafiltratian and after adding 1 mg/ml human serum albumin they wer~ filter sterilized and stored at -70 C. Sodium dodecyl sul~ate-polyacrylamide gel electrophoresis and Western blot analysis revealed that RH1- BG
consisted o~ a major band with molecular weight of 221 kD, corresponding to a ConjUgate containing 1 molecul~ each of Mab ~Hl and 6-glucuronidase.

~ (ii) Activity of Mab RHl-~G conjugate 1~
I T~e antigen bindlng and enzymatic activities of Mab RHl- ~G
,.:

', . .
., were determined. FIG. 3 (panel ~) shows tha~ R~ retained almost complete enzymatic activity measured by monitoring the release of p-nitrophenol ~rom p-nitrophenyl ~-D-glucuronide at 405 nm. Panel B shows that RHl ~G also retained antigen-binding activity as measured by enzy~e-linked immunosorbent assay.
RHl-~G did not bind to control HepG2 cells. Enzyme and antibody activity of RHl- ~G was also simultaneously assayed by first allowing the conjugate to bind to AS-30D cells and then assaying for bound ~-glucuronidase activity (Panel ~). RH1-~G was active at concentrations less than 200 ng/ml. These results show that the RH1-~G conjugate xetained both antibody and enzyme activitles as well as specificity for antigen positive AD-30D cells.

(iii) Cytotoxic effect of HAM and BEA~G to ~umor cells The effect of BHAMG and ~AM on several tumor cell lines was determined by measuring [3~] leucine incorporation into cellular protein of cells after drug exposure. FIG. 4 depicts the effect of HAM, BHA~G and BHAMG plus 10 units ~glucuronidase (one unit of R-glucuronidase can hydrolyze 1 ~mole p~nitrophenyl ~-glucuronide in one hour at 37C) on AS-30D (panel A) and ~epG2 (panel B) calls. Cells were exposed to drugs for 24 h, washed with PBS, and then incubated in ~resh medium for an additional 24 h. Cellular protein synthesis rate of drug-treated cells is compared to untreated-control cells at hour 48. Comparison of IC
values revealed that ~HA~G was over 1000 times less toxic than . ~ .
- 16 ~

, , .

''~ ;

HAM to AS-30D rat hepatoma cells and about 150 fold less toxic to HepG2 human hepatoma cell~ after 24 h drug exposure.
Simultanaous addition of ~ -~lucuronidase (10 units/well) and B~AMG to tumor cells resulted in a cytotoxic effect equal to that of ~L~M alone, indicating that cleavage o~ the glucuronide functional group converted BHAMG to ~Mo Addition of ~-glucuronidase alone did not affec~ t~H]-leucine incorporation into cellular protein.

TABLE 1 summarizes ths e~fects of HAM and BHA~G on cellular protain synthesis in several cell lines. CaSki human cervical carcinoma cells were provided by Dr. R. A~ Pattillo, Medical College of Wisconsin, Milwaukee, WI. HepG2 human hepatoma an~
COLO 205 human colon carclnoma cells were obtained from ATCC~
Rockville, MD. Human cells were maintained in RPMI 1640 medium (Gibco BRLJ Grand Island, NYj supplemented with 5~
heat inactivated fetal bovine serum, 100 U/ml penicillin and 100 ~g/ml streptomycin. AS-30D cells were cultured in DM~M (Gibco BRL) supplemented as above. The efect of p-hydroxyaniline mustard (HAM) and its glucuronide prodrug (B~M~) on pxo~ein synthesis in AS- OD rat hepatoma and human hep~2 hepatoma, Colo 205 colon carcinoma or CaSki cervical carcinoma cells was calculated by interpolation of dose-response cur~es.

. j ,, ;~

~ 17 -, .

~ ~J~ ~

Table 1 In vitro effect of HAM and BHAMG

Mean IC~o(~M~ Latency Cell Llne ~M BHA~G BHAMG~-glucuronidase (B~IAMG/IIAM) _~A _ _ __ _ ___ r __ ____________ ~__ _ _______ _ _ __ _ ~_ _ _ _ _ _ __ _ _ _ _ _ _ _ __ _ _ _ __ ~S-30~ 0 ~ 85+0 ~ 15 (~;) 1090+1~30 (4) 0 ~ ~9~0 ~ 25 (4) 128 (1 h exposure) 1.18+0~08(2) 1370~200(2) 0~5s~0.09(2) 116 (24 h exposure) 0.52~0.12(3) 809+9 (2) 0.82+0.58(2) 130 ; HepG2 7.9~1.6(8)1185~138(7) ~0.8~2.9(9) 15 Colo 205 37+6.4(3)1880+18(3) 15.g~2.5~7) 51 CaSki 53.5~2.2(3) 2790~190(2) 126+42(3) 52 . ~
.. ~Unless otherwise indicated, cells wer~ exposed to drugs for 24 h. Numbers in parentheses indicate the number of independent assays, each carried out in triplicate, used ko determine mean values. Standard arrors of the mean are also lndicated.

bLatency is the ratio of m an IC50 Yalues for B~ G to HAM.

.~ .
1 As shown in Table 1, AS-30D cells were most sensitive to H~M

with a mean ICso value of 0.85 ~M. Other cell lines were more ~! resistant to H~M, with CaSki human cervical carcinoma cells being ~ the most resistant (ICso value of 57 ~M). Prodrug lat~ncy, a i ~ : m~a~ure of the difference in toxiciti~s between prodrug and the ' parent compound, was also greatest for AS-30D cells; BXA~G was an :~ :
.~ , . - 18 .,:

:' ~ g (,~j3 average of 12i30 times less toxic than HAM to AS-30D cells. The ~ffect of drug exposure time on cell cytotoxicity was also examined in ~S-3OD cells. HAM and BHAMG were both about 2 times more toxic in a 2 4 h exposure assay compared to 1 h exposure.
Drug latency, however, was relati~ely insensitive to d~ug exposure time (1300 vs~ 1160 for 1 and 24 h exposure times, respectively). These results show that th~i glucuronide prodrug BHAMG is much less toxic to human cells than the parent compound H~M. In addition, BHAMG can be enzymatically con~erted to H~M by the addition o~ B-glucuronidase to the cells.

(iv) Specific Activatiorl o:~ Prodrug Specific activation of BH~MG at antigen positive AS~30D

cells was examined by ~irst incubating cells with different ~, concentrations of RH1-~G ~or 30 min, washing cells and then exposing the cells to gO ~M BHAMG for 24 h. FI~o 5/ panel A
i shows that protein synthesis was reduced by up to 95% in AS~30~
, cells preincubated with R~ Ç and then exposed to B~L~MG. Even : at a ~H1-~G conCentration of only 60 ng~ml, protein synthesis of BHAMG-treated AS-30D cells was inhibited by 44% compared to cells ] not exposed to RHl-BG. RHl- ~ G activation of B~A~G was specific for antigen-positivei cells; preincubation of an~igen-nega_ive HepG2 cells with RHl-~G did no~ increase the toxicity of BHAMG to .. . .
~ these cells (FIG. 5, panel B~. .
,j .

.1 .

i .

~J ~ J3 ~
RH1-~G specificity or ~S-30D cells was further verifi.ed by a competition assay. ~dditicn of soA~g~ml ~ab R~I1 F(ab')~ during preincu~ation of AS-30D cells with 1 ~g/ml R~ G protected the cells from sHAMGi cellular pro~ein synthesis was inhibited by only 30% (FIG. 6, Lane b) compared to 90% inhibition of protein synthesis in the absence of competing antibody frayment (FIC. 6, Lane a~. Blocking of R~ BG with excess Mab R~Il F(ab')~ did not affect protei~ synthesis o~ antigen negative ~epG2 cells exposed to B~L~MG (FIG. 6, Lanes c and d)~

.. ..

The ability of RH1- ~G to specif.ically a~tivate sxAMG at antigen~positive AS 30D c211s was also examined by preincubatiny cells with 1 or 1OlLg/ml RHl~G and suksequently exposing cells to varying concentrations of ~HAM~ ~or 24 h~ Preincubation of AS~30D cells with 1 ~g/ml RHl-~G decreased the IC vallle of ~HAMG by about 200 ~old (FIG. 7, panel A). In contxast RH1-was ineffec~ive at potentiating ~he acti~ity of BHAM~ at ~IepG2cells (FIG. 6, panel B). TABLE Z shows that B~M~ toxicity to AS-30D cells was further increasPd by raising RH1- ~ G
concentration to 10~ y/ml. ~ this concentration of R~ G, BH~MG was as p~tent ~s ~M with ~n TCso vDlue of <0.75 M.

~ .

i ~i ~,:: .. .. : : ~ ` , ~ . . ': ' . ' . ` : , .

f.,) ~ f .' "~

Table 2 Selective acti~ation o:~ BH~G by R~l~f3G ~t AS-30D cells IC~o (,~lM~)Selectivity Cell Line RH1-~G -~ BHAMG
(llg/ml ) HAM BHAMG ( B~MG ~ RH~ -,BG ) _ _ _ _ _ _ _ _ _ ._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ~ _ _ _ _ _ ._ _ _ _ _ ,. _ _ .. _ _ _ _ _ _ _ _ _ _ _ _ _ AS-30D 0 0 . 41 >770 1 3 . 6 >210 ,, ~0 . 75 >1000 F~epG2 0 3 . 6>770 ., .

'1 1 >770 !
, 10 >770 ., --____~_~_~ ____________ i ' Selectivity i5 de~ined as th . ratio of IC~jo values :Eor B~I~MG
: befoxe and a~ter incubation of cells with RHl-~BG conjugate~
.j .
i~
~ -- 21 (v) In vivo therapy of experimenkal hepatocellular carcinoma The efficacy of BHAMG in combination with RH1-~ G for treating cancer in vivo was examined in SD rats~ Rats were injected in the peritoneal cavity with 107 AS 30D cells on day 1.
Five days later rats received either 150 ~g RH1 ~G, 150 ~g control Mab-~-glucuronidase (which can not bind to ~S-30D cells) or PBS. Two days later, rats were ~reated with two doses of either 5 mg/Xg BH~ or PBS, FI~. 8 shows that 4/5 tumor bearing rats receiving only PBS died within 3 weeks due to massive ascites formation. Similarly~ 4/5 rats receiving BHAMG but not RHl-~G alss died within 3 weeks. Also 3/5 rats receiving control antibody-~-glucuronidase conjugate in combina~ion with BH~MG also died. In contrast, 5/5 rats receiving RH1-~ followed by BHAMG
treatment survived and wQre apparen~ly free of tumor cells~ This result ~hows that RHl-BG can localize at tumor cells in vi~o ind subsequently activate B~u~G coming into contac~ with the tumor cells.

The decreased toxicity o~ BH~MG compared ~o HAM was also . examined by injec~ing these drugs into the peritoneal cavity of Balb/c mice (Table 3) or SD rats (Table 4).

, .

, , - 2~ -' "'' ~ ' ", ', ' '. .. . ', ' " , . . :
' .. ' . ' '.' ' ' ', ..... '' ' , . ' ,'" . '' :

Table 3 comparison of B~L~MG and H~M toxicity to ~alb/c mice ___ ___ ___~____ ____~ ______ ~__________ _____ B~AMG Dea ths per EIAMDeaths per (mg/kg) group (mgJkg) group 22 0/2 ~ . 62/;~

:` 100 0/2 ~, . .
06 2~2 , :

_ ..
,.

,~
, .
, ~ Female Balb/c mice received a single i O p . inj ection o~ the .~. indicated compounds.
;;.`~i, .i, ~, , :
.
1 , , -- 23 , . ~ .

~ 2 :

~; Table 4 BH~IG toxicity to SD rats _ _ _ _ ~ BHAM(; Deaths per .~ (mg/kg) group ' 5~ OJ2 lo~ o/2 . . îSO 0/2 , ., Male SD rats received a single i~p. injection ~f B~G at ;:~ the indicated doses.

:!
Table 3 shows that HAM was toxic to mice at a dose o~ 9 . 6 ;~ mg/kg while B~G was n~ntoxic to mica at 100 mg/kg. Rats were .. not affected by BHAMG at a dose of 150 mg/kg.
,.j From the foregoing it will b~ appreciated that, although speci~ic embodiments of ~h~ invention have been described herein . . ~
.~: for the purposes of illustration, YariOU5 modi~ications may be ., .J

.,,: .. ,' ,' ', ', ' . . ',i' ~' '. ' .. ' , ' ' . "' ' . '', ," ' '' ';', ' ' made without deviating from the spirit and scope of the invention. Accordingly, -the invention is not to be limited except as by the appended claims.

.

~3 .~ `
'~ .
.~
i ., `

i'.

3~

:j :

t

Claims (24)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A method for enhancing the specificity and efficacy of chemotherapy comprising:
administration of a conjugate between a ligand and beta-glucuronidase to allow accumulation of the conjugate at malignant cells which are targeted by the ligand;
and subsequent administration of a glucuronide prodrug.

2. The method of claim 1 in which the glucuronide prodrug is BHAMG.

3. The method of claim 1, in which the targeted maligant cells are tumor cells.

4. The method of claim 1, claim 2 or claim 3, in which the ligand is an antibody.

5. The method of claim 1, claim 2 or claim 3, in which the ligand is an monoclonal antibody.

6. The method of claim 1, claim 2 or claim 3, in which the ligand is a human monoclonal antibody.

7. The method of claim 1, claim 2 or claim 3, in which the glucuronidase is human beta-glucuronidase.

8. A conjugate between a ligand and beta-glucuronidase in which the ligand targets to maligant cells.

9. The conjugate of claim 8 in which the ligand is an antibody.

10. The conjugate of claim 8 in which the ligand is a monoclonal antibody.

11. The conjugate of claim 8 in which the ligand is a human monoclonal antibody.

12. The conjugate of claim 8, claim 9, claim 10 or claim 11 that is useful for activating a glucuronide prodrug of chemotherapeutic agents.

13. The conjugate of claim 8, claim 9, claim 10 or claim 11 in which the targeted malignant cells are tumor cells.

14. The conjugate of claim 8, claim 9, claim 10 or claim 11 that is administered prior to the administration of a glucuronide prodrug.

15. A composition for use in a method of treating malignant cells with a glucuronide prodrug comprising a conjugate of a ligand and beta-glucuronidase in which the ligand targets to malignant cells.

16. The composition of claim 15 in which the ligand is an antibody.

17. The composition of claim 15 in which the ligand is a monoclonal antibody.

18. The composition of claim 15 in which the ligand is a human monoclonal antibody.

19. The composition of claim 15, claim 16, claim 17 or claim 18 in which the targeted malignant cells are tumor cells.

20. The composition of claim 15, claim 16, claim 17 or claim 18 that is administered prior to the administration of the glucuronide prodrug.

21. A unit package comprising a conjugate of claim 8, claim 9, claim 10 or claim 11 and a glucuronide prodrug.

22. A process for preparing a conjugate between a ligand and beta-glucuronidase in which the ligand targets to malignant cells comprising:
forming a conjugate by chemically linking beta-glucuronidase to the ligand.

23. The process of claim 22 in which the linking reagent is selected from the group consisting of succinimidyl 4-(N-maleimidomethyl) cyclohexane-1-carboxylate and N-succunimidyl-3(2-pridyldithio) propionate.

24. A process for preparing a conjugate between a ligand and beta-glucuronidase in which the ligand targets to malignant cells comprising:

forming a conjugate by creatng a fusion protein between the ligand and beta-glucuronidase.