ICANCER RESEARCH 57, I 116-1123. March 15. 1997]
`
`LY231514, a Pyrrolo[2,3-d]pyrimidine-based Antifolate That Inhibits Multiple
`Folate-requiring Enzymes
`
`Chuan Shih,a Victor J. Chen, Lynn S. Gossett, Susan B. Gates, Warren C. MacKellar, Liilian L. Habeck,
`Katherine A. Shackelford, Lurane G. Mendelsohn, Daniel J. Soose, Vinod F. Patel, Sberri L. Andis,
`Jesse R. Bewley, Elizabeth A. Rayl, Barbara A. Moroson, G. Peter Beardsley, William Kohler, Manshan Ratnam,
`and Richard M. Schultz
`
`Lilly Research Laboratories, Eli Lilly and Company, Indianapolis, Indiana 46285 [C. S., V. J. C, L. S. G., S. B. G., W. C. M., L. L H., K. A. S., L. G. M., D. J. S., V. F. P., S. L. A.,
`J. R. B., R. M. S.]; Department of Pediatrics, Yale Universily, New Haven, Connecticut 06510 [E. A. R., B. A. M., G. P. B.]: and Department of Biochemistry and Molecular
`Biology. Medical College of Ohio, Toledo, Ohio 43699 [W. K., M. R.]
`
`ABSTRACT
`
`N.[4.[2-(2.amino-3,4-dihydro-4-oxo-7H-pyrrolo[2,3-d]pyrimidin-5-yl)-
`ethyl]-benzoyl]-L-glutamic acid (LY231514) is a novel pyrroio[2,3-
`d]pyrimidine-based anfifolate currently undergoing extensive Phase II
`clinical trials. Previous studies have established that LY231514 and its
`synthetic ~/-polyglutamates (glu3 and glus) exert potent inhibition against
`thymidylate synthase (TS). We now report that LY231514 and its poly-
`glutamates also markedly inhibit other key folate-requiring enzymes,
`including dihydrofolate reductnse (DHFR) and glycinamide ribonucle-
`otide formyltransferase (GARFT). For example, the Ks values of the
`pentaglutamate of LY231514 are 1.3, 7.2, and 65 nM for inhibition against
`TS, DIIFR, and GARFT, respectively. In contrast, although a similar high
`level of inhibitory potency was observed for the parent monoglutamate
`against DHFR (7.0 riM), the inhibition constants (Kl) for the parent mono-
`glutamate are significantly weaker for TS (109 riM) and GARFT (9,300
`riM). The effects of LY231514 and its polyglutamates on aminoimidazole
`carboxamide ribonucleoride formyltransferase, 5,10-methylenetetrahy.
`drofolate dehydrogenase, and 10-formyltetrahydrofolate synthetase were
`also evaluated. The end product reversal studies conducted in human cell
`lines further support the concept that multiple enzyme-inhibitory mech-
`anisms are involved in cytotoxicity. The reversal pattern of LY231514
`suggests that although TS may be a major site of action for LY231514 at
`concentrarions near the ICso, higher concentrations can lead to inhibition
`of DHFR and/or other enzymes along the purine de novo pathway. Studies
`with mutant cell lines demonstrated that LY231514 requires polygluta.
`marion and transport via the reduced folate carrier for cytotoxic potency.
`Therefore, our data suggest that LY231514 is a novel classical antifolate,
`the antitumor activity of which may result from simultaneous and multi-
`ple inhibition of several key folate-requiring enzymes via its polyglutama-
`ted metabolites.
`
`INTRODUCTION
`
`Several novel folate-based antimetabolites are currently being ac-
`tively investigated in clinical trials. These include lometrexol and
`2
`LY309887, which inhibit GARFT in the purine de novo biosynthetic
`pathway (1-3); edatrexate (4, 5) which acts on DHFR; and ZD1694
`(Tomudex; Refs. 6 and 7), AG337 (Thymitaq; Ref. 8), and
`BW1843U89 (9), which specifically target TS.
`
`Received 8/21/96: accepted 1/17/97.
`
`The costs of publication of this article were defrayed in part by the payment of page
`charges. This article must therefore be hereby marked advertisement in accordance with
`18 U.S.C. Section 1734 solely to indicate this fact.
`t To whom requests for reprints should be addressed, at Cancer Research Division,
`Lilly Research Laboratories, Drop 0540, Eli Lilly and Company, 307 E. McCarty St.,
`Indianapolis, IN 46285. Phone: (317)276-3520; Fax: (317)277-3652.
`2The abbreviations used are: LY231514, N-[4-[2-(2-amino-3,4-dihydro-4-oxo-7H-
`
`pyrrolo[2,3-dlpyrimidin-5-yl)ethyl]-benzoyll-L-glutamic acid; r, recombinant; h, human;
`m, murine; TS, thymidylate synthase (EC 2.1.1.45); DHFR, dihydrofolate reduclase (EC
`1.5.1.3); GARFT, glycinamide ribonucleotide formyltransferase (EC 2.1.2.2); AICA,
`5-aminoimidazole-4-carb~xamide; AICARFT, aminoimidazole carboxamide ribonucle-
`otide formyltransferase (EC 2.1.2.3); CI-S, Cl tetrahydrofolate synthase; FPGS, folyl-
`polyglutamate synthetase (EC 6.3.2.17); RFC, reduced folate carrier; FBP-a, folate
`
`binding prot~in-~t; M’FF, 3-[4.5-dimethylthiazol-2yl]-2,5-dipbenyltetrazolium bromide;
`NADPH./3-NADP+. reduced form; 6R-MTHF. 6[R]-5,10-methylene-J,6,7,8-tetrahydro-
`folate.
`
`LY231514 is a structurally novel antifolate antimetabolite that pos-
`sesses the unique 6-5-fused pyrrolo[2,3-d]pyrimidine nucleus (10, 11)
`instead of the more common 6-6-fused pteridine or quinazoline ring
`structure (Fig. 1). Previous studies have demonstrated that LY231514 is
`one of the best substrates that is known for the enzyme FPGS (Km = 1.6
`
`l~ and V,~IK= = 621; Ref. 12). It is likely that polyglutamafion and the
`polyglutamated metabolites of LY231514 play profound roles in deter-
`mining both the selectivity and the antitumor activity of this novel agent
`(ll, 12). Whereas LY231514 only moderately inhibited TS (Ki = 340
`riM, recombinant mouse), the pentaglutamate of LY231514 was 100-fold
`more potent (Ki = 3.4 hi; Ref. 11), making LY231514 one of the most
`potent folate-based TS inhibitors known today (13).
`Preliminary cell culture end product reversal studies in human CCRF-
`CEM and murine LI210 leukemia cells have demonstrated that thymi-
`dine (5 p.i) alone was not able to fully reverse the cytotoxic action of
`LY231514 ( l 1). Both thymidine (5 pay0 and hypoxanthine (100/~) were
`required to fully protect cells from the growth-inhibitory activity exerted
`by LY231514. This reversal pattern is significantly different from other
`TS inhibitors, such as ZD1694 (6) and BWI843U89 (9). Cell culture
`experiments showed that the andproliferative activity of LY231514 was
`completely reversed by the addition of leucovorin (0.05--16 bu~t) in a
`competitive manner (11), suggesting that LY231514 competed with
`natural reduced folate cofactors both at transport and intracellular folate
`levels and acted as a pure folate antagonist.
`Promising antitumor responses have recently been observed in the
`Phase I trials of LY231514. Moreover, patients who had previously
`failed to respond to ZD1694 and 5-fluorouracil/leucovorin treatment
`responded to LY231514 (14). This pattern of clinical response, to-
`gether with the aforementioned observations of partial protection by
`thymidine in cell culture, suggest that inhibition of TS by LY231514
`may not solely account for the overall antitumor effect of this novel
`antifolate. LY231514 and its polyglutamates may inhibit other folate-
`requiring enzymes, such as DHFR, or enzymes along the de novo
`purine biosynthetic pathway. LY231514 may thus act as a multitar-
`geted antifolate, with multiple mechanisms of action affecting the
`intracellular folate pools and cellular pyrimidine/purine biosynthesis.
`We now summarize our findings of LY231514 and its polygluta-
`mates (glu3 and glus) against various folate-requiring enzymes, in-
`cluding human TS, DHFR, AICARPT, 5,10-methylenetetrahydrofo-
`late dehydrogenase, and 10-formyltetrahydrofolate synthetase
`activities of CI-S and murine GARFT. In addition, we report a
`detailed comparison of cell culture reversal patterns observed in
`several human cell lines between compounds LY231514 and ZD 1694.
`Finally, we examine the role of polyglutamation and transport (via the
`RFC) in the cytotoxicity of LY231514.
`
`MATERIALS AND METHODS
`
`Materials. LY231514 and ZD1694 were prepared according to published
`methods and procedures (7, ! 1). The syntheses of the 7-glutamyl derivatives of
`LY231514 were by the method of Pawelczak et al. (15). For in vitro studies,
`
`1116
`
`Downloaded from cancerres.aacrjournals.org on October 16, 2015. © 1997 American Association for Cancer
`Research.
`
`Sandoz Inc.
`Exhibit 1021-0001
`
`JOINT 1021-0001
`
`

`
`LY231514, A MULTITARGETED ANTIFOLATE
`
`0
`
`COOH
`=
`
`o
`
`HN~ "~’COOH
`
`Fig. I. Structure of LY2315 ! 4. N-[4-[2-(2-amino-3,4-dihydro-4-oxo-TH-pyrrolo[2.3-
`d]pyrimidin-5-yl)ethyl ] -benzoyl]-L-glutamic acid.
`
`compounds were dissolved in either DMSO or 5% sodium bicarbonate at an
`inidal concentration of 1-50 raM, and dilutions were made in either enzyme
`assay buffer or cell culture medium (RPMI 1640 with 10% dialyzed FCS). The
`final DMSO concentration never exceeded 0.5%. Vehicle controls confirmed
`that there was no effect of DMSO at this concentration. A water-soluble form
`of the disodium salt of LY231514 was used in some investigations. The
`recombinant enzymes used were all obtained in purified form from the fol-
`lowing sources: rhTS from Dr. D. V. Sand (Universiq/of California at San
`Francisco, San Francisco, CA; Ref. 16); trifunctional mGARFT from Dr. R. G.
`Moran (Medical College of Virginia, Richrnond, VA; Ref. l?); rhDHFR from
`Dr. M. Rarnam of Medical College of Ohio, Toledo, Ohio (lg) and Anatrace
`Co. (Maumee, OH). Two forms of rhCI-S were obtained frorn Dr. R. E.
`Mackenzie (McGill University, Montreal, Quebec, Canada; Ref. 19): (a) the
`M, I01,000 full-length enzyme of C l-S containing 5,10-methylenetetrahydro-
`folate dehydrogenase, 5,10-methenyltetrahydrofolate cyclohydrolase and 10-
`formyltetrahydrofolate synthetase (EC 6.3.4.3) activities; and (b) the protein
`domain of CI-S containing the 5,10-methylenetetrahydrofolate dehydrogenase
`(EC 1.5.1.5) and 5,10-methenyRetrahydrofolate cyclohydrolase activities (the
`M, 35,000 truncated version of CI-S, which contained only the dehydrogenase
`and cyclohydro|ase activities). Human AICARFF was purified as described by
`Rayl et aL (20). 10-Formyl-[6R,S]-5,6,?,8-tetrahydrofolate was prepared by a
`method similar to that of Rowe (21). 10-Formyl-5,8-dideazafolic acid and
`aJ3-glycinamide ribonucleotide were prepared as described previously (2).
`6R-MTHF for use in the TS assay was obtained from Eprova AG (Schaff-
`hausen, Switzerland); the ~-ihydrochloride salt of [6R,S]-5,6,?,8-tetrahydrofo-
`late and the magnesium salt of [6R,S]-5,10-methylene-5,6,?,8-tetrahydrofolate
`for use in the CI-S dehydrogenase assay were obtained from Dr. B. Schircks
`Laboratories (Jona, Switzerland). 5-Aminoimidazole-4-carboxamide ribonu-
`cleotide, AICA, folic acid. folinic acid, ?,8-dihydrofolate, NADPH, bypoxan-
`thine, methotrexate, IVlTl’, and thymidine were purchased from Sigma Chem-
`ical Company (St. Louis, MO). Dialyzed fetal bovine serum was purchased
`from Hyclone (Logan, UT). Regular and folate-free RPMI 1640 with 25 mM
`HEPES buffer were purchased from Whittaker Bioproducts (Walkersville,
`MD). The ENZFITTER microcomputer package was obtained from Biosoft
`(Ferguson, MO). CCRF-CEM cells were obtained from St. Jude Children’s
`Research Hospital (Memphis, TN). HCT-8 cells were purchased from the
`American Type Culture Collection (Rockville, MD). CRl 5, a 5,10-dideazatet-
`rahydrofolic acid-resistant CCRF-CEM subline, was described by Pizzoroo et
`aL (22). ZR-?5-1 human breast carcinoma cell sublines with differing folate
`transport properties were generously provided by Dr. K. Cowan (NCI, Be-
`thesda, MD; Ref. 23). The GC3/CI cell line was developed by Dr. L Houghton
`(St..lude Children’s Research Hospital, Memphis, TN; Ref. 24). CCRF-CEM,
`HCT-8, CRIS, ZR-?5-1, MTXaZR-?5-1, and GC3/C1 cells were routinely
`cultured in RPMI 1640 medium containing L-glutamine and 25 m~ HEPES
`buffer and supplemented with 10% dialyzed FCS. ZR-?5-1 cells expressing
`FBP-a (MTXRBB3-FR+ and 2FR+AA6) were cultured in folic acid-free
`RPMI 1640 containing L-glutamine, 25 r~ HEPES buffer, 2 nM folinic acid,
`and 10% dialyzed FCS.
`Enzyme Assays and Methods. TS activity was assayed using a spectro-
`photometric method described by Greene et al. (25), which involved monitor-
`ing the increase in absorbance at 340 nm resulting from formation of the
`product, 7,8-dihydrofolam. The assay buffer contained 50 mM N-tris[hy-
`droxymethyl]methyl-2-aminoethanesulfonic acid, 25 mM MgCI2, 6.5 mM
`formaldehyde, I mM EDTA, and 75 mM 2-mercaptoethanol, pH 7.4. The
`concenwations of deoxyuridylate monophosphate, 6R-MTHF, and hTS were
`100 p~, 30 p2a, and 30 nM (1.7 milliunits/ml), respectively. (One milliunit of
`enzyme activity is defined as 1 nmol of product produced per min.) At the
`
`6R-MTHF concentration, an uninhibited reaction and six concentrations of
`
`inhibitor were assayed. Ki apparent (Ki ,pp) values were determined by fitting
`the data to the Morrison equation (26) using nonlinear regression analysis with
`
`the aid of the program ENZFITTER. K~ values were calculated using the
`equation: Ki aw = Ki (l + [S]/Km), where IS] is equal to 30 p,M and Km is equal
`
`to 3 ~.M.
`DHFR activity was assayed spectrophotometrically by monitoring the dis-
`appearance of the substrates NADPH and 7,8-dihydrofolate (combined
`
`~ = 12 x l03 M-t cm-~) at 340 nm. The reaction took place at 25°C in 0.5
`
`ml of 50 mM potassium phosphate buffer, which contained 150 mM KCI and
`
`10 mM 2-mercaptoethanol, pH 7.5, and 14 nM (0.34 milliunil/ml) DHFR. The
`NADPH concentration was l0 p.M, and 7,8-dihydrofolate was varied at 5, 10,
`
`or 15 p.M. At each 7,8-dihydrofolate concentration, an uninhibited reaction and
`seven concentrations of inhibitor were assayed. The ENZFITFER microcom-
`
`puter program was used to obtain Ki app values by fitting the data to the
`Morrison equation by nonlinear regression analysis. K~ values were calculated
`
`using the equation: Kiapp = Ki (! + [S]/Km), where IS] is equal to the
`concentration of 7,8-dihydrofolate used and Km of 7,8-dihydrofolate is 0.15
`
`~ (27, 28).
`GARFT activity was assayed speca’ophotometrically as described previ-
`
`ously (2) by monitoring the increase of absorbance resulting from formation of
`
`the product 5,8-dideazafolate at 295 nm. The reaction solvent contained 75 rn~
`HEPES, 20% glycerol, and 50 m~ a-thioglygerol, pH 7.5, at 25"C. The
`
`concentrations of substrates and enzyme used were l0 ~M a,/3-glycinamide
`
`dbonucleotide, 0-10 p,M 10-formyl-5,8-dideazafolic acid, and l0 nM (!.9
`
`milliunits/ml) GARFT. K~ values were calculated using the Enzyme Mecha-
`
`nism program of the Beckman DU640 spectrophotometer, which uses nonlin-
`
`ear regression analysis to fit data to the Michaelis-Menten equation for com-
`petitive inhibition.
`
`AICARFT inhibition assays were carried out at room temperature by
`
`monitoring the formation of [6S]-5,6,7,8-tea’ahydrofolate from 10-formyl-
`
`[6R,S]-5,6,7,8-tetrahydrofolate at A298. All solutions were purged with N2 gas
`
`prior to use. The reaction solution contained 33 m~ Tris-Ci, pH 7.4, 25 m~
`
`KCI, 5 m~ 2-mercaptoethanol, 0.05 m~ AICA dbonucleotide, and 16 n~ (2.0
`
`milliunitsiml) of AICARPT. 10-Formyl-[6R,S]-5,6,7,8-tetrahydrofolate con-
`centrations of 0.037, 0.074, and 0.145 mM were used (0.61, 1.23, and 2.45
`
`times its Km value, respectively). LY231514 was tested as an inhibitor at
`0.080-0.800 m~ (four concentrations). When the tri- and pentaglutamates of
`
`LY231514 were used as inhibitors, the concena’ations were 0.0005-0.009 mM
`(eight concentrations). Enzyme assays were initiated by the addition of en-
`zyme. Data was analyzed using the ENZFITTER program for competitive
`
`inhibition.
`The method of Tan et al. (29) was used for assaying the activities of CI -S.
`
`This involves quenching the reaction by acid and subsequent speca’ophoto-
`mea’ic quantitation of the amount of 5,10-methenyltetrahydrofolate produced
`
`at 350 rim. Accordingly, the dehydrogenase assay was conducted in a pH 7.3
`reaction mixture containing 0.084 M potassium phosphate, 0.12 M 2-mercap-
`toethanol, 0.17 m~ NADP, 5.75-168 p.M [rR,S]-5,10-methylene-5,6,7,8-tet-
`
`rahydrofolate, and 0.96 nM (0.7 milliunit/ml) protein domain of C l-S contain-
`
`ing the 5,10-methylenetetrahydrofolate dehydrogenase (EC 1.5.1.5) and 5,10-
`
`rnethenylteu’ahydrofolate cyclohydrolase activities. The synthetase assay was
`conducted in a pH 8.0 reaction mixture containing 0.1 M triethanolamine, 0.14
`
`M 2-mercaptoethanol, 0.05 M KCI, 0.04 M sodium formate, 1.0 m~ MgCl2, 1.0
`m~ ATP, 62.5-2000 p.M [6R,S]-5,6,7,8-tetrahydrofolate, and 1.4 nM (0.7
`
`milliuniffml) full-length enzyme of CI-S containing 5,10-methylenetetrahy-
`drofolate dehydrogenase, 5,10-methenyltetrahydrofolate cyclohydrolase and
`lO-formyltetrahydrofolate synthetase (EC 6.3.4.3) activities. The concentra-
`
`tions of LY231514 and its polyglutaraates studied in each activity were from
`
`about 0.5 × Ki to 3 x Ki. All reactions were conducted at ambient temperature
`
`(23"C) in a final volume of 0.475 ml and quenched with 0.025 ml of 0.4 M HCI.
`
`Activity data collected with a range of substrate and drug concentrations were
`
`fit to the Michaelis-Menten equation for competitive inhibition by nonlinear
`
`regression with the aid of the GRAFIT computer program (30).
`In Vitro Cell Culture Studies. Dose-response curves were generated to
`
`determine the concentration required for 50% inhibition of growth (lCso). Test
`
`compounds were dissolved initially in DMSO at a concentration of 4 mg/ml
`
`and further diluted with cell culture medium to the desired concentration.
`CCRF--CEM leukemia cells in complete medium were added to 24-well
`
`Cluster plates at a final concentration of 4.8 × 104 cells/well in a total volume
`
`1117
`
`Downloaded from cancerres.aacrjournals.org on October 16, 2015. © 1997 American Association for Cancer
`Research.
`
`Sandoz Inc.
`Exhibit 1021-0002
`
`JOINT 1021-0002
`
`

`
`LY231514, A MULTITARGETED ANTIFOLATE
`
`0.60
`
`0.50
`
`0.20
`
`0.10
`
`0
`
`16.
`
`0
`
`o s lO 15 20 25 30
`
`i~rmF] ~
`
`o. 2
`
`o.b6 o.i 6 o.lo o.12
`
`[LY231514-glu3] (p.M)
`
`of 2.0 ml. Test compounds at various concentrations were added to duplicate
`wells so that the final volume of DMSO was 0.5%. The plates were incubated
`for 72 h at 37°C in an atmosphere of 5% CO: in air. At the end of the
`incubation, cell numbers were determined on a ZBI Coulter counter. Control
`wells usually contained 4 × 10~ to 6 × l0"s cells at the end of the incubation.
`For several studies, ICsos were determined for each compound in the presence
`"~ 0.40
`of either 300/xM AICA, 5 p.M thymidine, 100 p.M hypoxanthine, or combina-
`~
`g.~
`don of 5 bLm thymidine plus 1130 bl,m hypoxanthine,
`0.30
`For adherent tumor cells, we used a modification of the original MTT ~ o
`colorimetric assay described by Mosmann (31) to measure cell cytotoxicity.
`~
`The human tumor cells were seeded at I × 104 cells in 100 bd of assay
`>
`medium/well in 96-well flat-bottomed tissue culture plates (Costar, Cam-
`bridge, MA). The assay medium contained folic acid-free RPMI 1640 supple-
`mented with 10% FCS and either 2 nM folinic acid or 2.3 bLM folic acid as the
`sole folate source. Well IA was left blank (100 p~l of growth medium without
`cells). Stock solutions of antifolates were prepared in Dulbecco’s PBS at 1
`mg/ml, and a series of 2-fold dilutions were subsequently made in PBS. Ten-~l
`aliquots of each concentration were added to triplicate wells. Plates were
`incubated for 72 h at 37°C in a humidified atmosphere of 5% COz-in-air. MTIF
`was dissolved in PBS at 5 mg/ml, 10 p,I of stock MTT solution were added to
`each well of an assay, and the plates were incubated at 37°C for 2 additional
`h. Following incubation, 1130 /~l of DMSO were added to each well. After
`thorough formazan solubilization, the plates were read on a Dynatech MR600
`reader, using a test wavelength of 570 nm and a reference wavelength of 630
`nm. The ICso was determined as the concentration of drug required to inhibit
`cell growth by 50% compared to an untreated controls.
`
`Fig. 2. Morrison Analysis of tight-binding inhibition of rhTS by LY231514-glu3. A
`velocity versus inhibitor concentration curve is shown from a representative experiment
`illustrating the concentration-dependent inhibition of rhTS (29 nra) in the presence of
`6R-MTHF (15 p.M) and 100/XM deoxyuridylate monophosphate. Inset, Ki app values were
`determined by the nonlinear fitting of data collected at three concentrations of 6R-MTHF
`to the Morrison equation using the ENZIqTIPdR microcomputer package. The K~ value
`(1.3 riM) was determined from the slope of the graph K,
`for 6R-MTHF of 3.0
`
`RESULTS
`
`Enzyme Inhibition Studies. The inhibition of rhTS by LY231514
`and its polyglutamates is summarized in Table 1. The parent mono-
`glutamate LY231514 inhibited rhTS with a Ki of 109 nM when the
`monoglutamated form of the substrate (6R-MTHF) (6[R]-5,10-meth-
`ylenetetrahydrofolate) was used. This is in good agreement with the Ki
`value generated earlier for rmTS (K~ = 340 riM; Ref. 11). The
`longer-chain 7-glutamyl derivatives of LY231514 demonstrated sig-
`nificantly enhanced affinity to rhTS. The addition of two extra 3,-glu-
`tamyl residues (glu3) to LY231514 resulted in 68-fold reduction of the
`Ki value. Further extension of the glutamate tail (LY231514-glu5) did
`not result in any significant enhancement of inhibitory potency toward
`rhTS. In comparison, ZDI694 was less dependent on polyglutama-
`tion. A 5-fold increase in affinity was observed for ZD1694 polyglu-
`tamates toward rhTS. It has been well recorded that mammalian TS
`showed a strong preference for polyglutamated folate substrates. A
`similar effect had been reported by Jackman et al. (6, 33) and Sikora
`et al. (32) in their studies of the quinazoline antifolates CB3717,
`ZD!694, and their polyglutamates by using partially purified LI210
`murine TS. In both cases, the corresponding triglutamate derivatives
`demonstrated 87- and 56-fold reductions in Ki values, respectively,
`compared to the parent compounds. In a separate study by Cheng et
`al. (34), CB3717-glu3 was approximately 20-fold more potent than
`the parent monoglutamate compound in inhibiting human TS isolated
`from HeLa $3 and KB cells. The quantitative differences in the
`
`Table 1 Inhibitory activity of LY231514, ZDI694, and their polyglutamates against
`rhTS, rhDHFR, and rmGARFT"
`
`Ki value (rim +-- SE; n > 3)
`
`Compound rhTS
`
`LY231514
`LY231514-(glu)3
`LY231514-(glu)s
`
`109-+ 9(n = 4)
`1.6 +- 0.1
`1.3 -+ 0.3
`
`ZDI694
`ZDI694-(glu)3
`ZDI694-(glu)5
`
`6.0 -+ 0.9
`I.I - 0.3
`1.4 ___ 0.1
`
`rhDHFR
`
`7.0--. 1.9
`7.1 +-. 1.6
`7.2 _+ 0.4
`
`45 -+ 3
`37 _+ 7
`30 - 3
`
`See "’Materials and Methods" for assay procedures.
`
`rmGARFT
`
`9,300+_.690
`380 --. 92
`65 -+ 16
`
`424,000 (336,000, 513,000)
`104,000 (81,000, 127,000)
`132,000 (124,000, 141,000)
`
`reported degree of enhancement in potency as a result of polygluta-
`mation are likely due to a combination of the variation in enzyme
`source used, as well as the inherent difficulty in obtaining Ki estimates
`for very tightly bound compounds. LY231514 and its polyglutamates
`inhibited rhTS in a competitive fashion with respect to the natural
`substrate [6R]-5,10-methylenetetrahydrofolate. The data of
`LY231514-glu3 against rhTS is shown in Fig. 2. The Ki values
`reported in Table 1 are calculated assuming competitive inhibition for
`ZD 1694.
`LY231514 was found to be a very potent inhibitor when tested
`against recombinant human DHFR. Tight binding analysis showed
`that LY231514 inhibited rhDHFR in a competitive fashion with a Ki
`of 7.0 nM (Table 1). In contrast to rhTS, attachment of additional
`-y-glutamyl residues to LY231514 had little effect on the inhibition
`toward rhDHFR (the glu3 and glu5 of LY231514 exhibited identical
`Ki values against rhDHFR). The polyglutamates of LY231514 also
`showed a competitive inhibition pattern toward rhDHFR (data not
`shown). It was reported that CB3717 had a Ki of 250 nM on DHFR
`isolated from human KB/6B cells (34) and that ZD1694 inhibited rat
`liver DHFR with a K~ of 93 nM (6). Likewise, polyglutamation of
`CB3717 and ZD 1694 did not enhance affinity to DHFR. In our hands,
`ZD1694 and its polyglutamates also inhibited rhDHFR but were
`7-fold less potent than LY231514. The polyglutamates of ZD1694
`showed slight enhancement of affinity toward rhDHFR.
`We also studied drug inhibition against the folate-requiring en-
`zymes along the purine de novo biosynthetic pathway. LY231514
`only moderately inhibited rmGARFT (Ki = 9.3 /ZM). Through earlier
`studies of 5,10-dideazatetrahydrofolates, it was discovered that
`GARFT inhibition is highly dependent upon the polyglutamation
`status of inhibitors (2). The triglutamate and pentaglutamate of
`LY231514 had significantly enhanced inhibitory activity against
`GARFT, with Ki values of 380 nM (24-fold) and 65 nM (144-fold),
`respectively. This makes the pentaglutamate of LY231514 a poten-
`tially potent inhibitor of purine de novo biosynthesis. In comparison,
`ZDI694 and its polyglutamates showed extremely weak inhibitory
`activity against GARFT. The Ki values of ZD 1694, ZD1694-glu3, and
`ZD1694-glu5 were 424, 104, and 132 ~M, respectively (Table 1). This
`result demonstrates that polyglutamyl derivatives of LY231514 are
`300-2000-fold more effective than ZD1694 in inhibiting GARFT, an
`
`1118
`
`Downloaded from cancerres.aacrjournals.org on October 16, 2015. © 1997 American Association for Cancer
`Research.
`
`Sandoz Inc.
`Exhibit 1021-0003
`
`JOINT 1021-0003
`
`

`
`LY231514, A MULTITARGETED ANTIFOLATE
`
`Table 2 Inhibitory activity of LF231514 and its polyglutamates against hAICARFT and
`the dehydrogenase and synthetase activities in CI-S~
`
`gi value (~)
`
`5.10-
`methylenetetrahydrofolate
`dehydrogenase
`Compound of CI synthase
`
`10-formyltelrahydrofolate
`synthetase of
`CI synthase
`
`AICARFT
`
`LY231514
`9.5 ± 0.9b
`LY231514-(glu)3 3.7
`LY23 ! 514-(glu)5
`5.0
`a See "Materials and Methods" for procedures.
`b__. SD;n = 3.
`
`364
`25
`1.6
`
`3.58
`0.48
`0.26
`
`important enzyme along the purine de novo biosynthetic pathway. The
`second folate-requiring enzyme along the purine de novo biosynthetic
`pathway is AICARFT, which uses the same folate cofactor as
`GARFT, 10-formyl-tetrahydrofolate, as the one carbon donor in pu-
`fine biosynthesis. A similar trend of enhancement of affinity was
`observed for LY231514 and its polyglutamates toward hAICARF/’.
`The Ki values observed were 3.58/xM, 480 nM (7.5-fold), and 265 nu
`(13.5-fold) for the mono-, tri-, and pentaglutamyl derivatives of
`LY231514, respectively (Table 2).
`Finally, LY231514 and its polyglutamates were also found to be
`competitive inhibitors against both the 5,10-methylenetetrahydrofo-
`late dehydrogenase and 10-formyltetrahydrofolate synthetase activi-
`ties of C1-S (Table 2). The Ki values for the mono-, tri-, and penta-
`glutamyi derivatives of LY231514 were 9.5, 3.7, and 5.0
`respectively, for dehydrogenase and 364, 25, and 1.6/xu for synthe-
`tase. This demonstrates that the effect of polyglutamation of
`LY231514 on inhibition of dehydrogenase activity is marginal, but is
`quite significant for inhibition of synthetase activity. This observation
`is consistent with previous reports on the sensitivity of these two
`enzymes to polyglutamation status of their respective folate cofactors
`(35, 36). Based on the Ki values of LY231514 and its polyglutamates,
`the importance of CI-S as a potential target will be dependent upon
`the intracellular concentration of drug achieved (see below).
`Cell Cdture End Products Reversal Studies. Previous studies
`demonstrated that the antiproliferative activity of LY231514 was
`prevented by leucovorin but incompletely reversed by thymidine (I 0,
`11). This suggested that aside from TS, additional enzymatic targets
`for this antifolate compound exist. We have now further characterized
`the reversal pattern of LY231514 and ZD1694 in various human
`tumor cell lines, including CCRF-CEM leukemia, GC3/C1 colon
`carcinoma, and HCT-8 ileocecal carcinoma. It was observed that 5
`thymidine fully protected these cells from cytotoxicity with ZD1694
`(Table 3). In sharp contrast, similar treatment with thymidine (5
`only increased the IC5o of LY231514 versus CCRF-CEM cells by
`5.5-fold, GC3/C1 by 18.7-fold, and HCT-8 by 15-fold. It is interesting
`to note that thymidine alone produced its greatest protective effect at
`or near the IC5o of LY231514 (Fig. 3). In contrast, higher drug
`
`concentrations of LY231514 required the combination of both thymi-
`dine (5 p~M) plus hypoxanthine (100 p~) to protect CCRF-CEM cells.
`Moreover, the combination of thymidine plus hypoxanthine totally
`reversed the cytotoxicity exerted by LY231514 in all three cell lines
`(IC5o values > 40/~r~; Table 3). Hypoxanthine (100 #M) or amino-
`imidazole carboxamide (300 /xM) alone did not markedly influence
`cytotoxicity by LY231514 (except for HCT-8 cells, in which a 5-fold
`
`1°°t
`
`100’~
`
`80-’
`
`60"
`
`40"
`
`20"
`
`100-
`
`80’
`
`60’
`
`40,
`
`20’
`
`0~0001 ........ ~ ........ ’ ......... ’ ........ ~ ........ ’ ’’
`0.001 0.01
`0.1
`1
`10 30
`Conc. (~)
`Fig. 3. End product reversal studies of LY231514, ZD1694, and methotrexate in
`CCRF-CEM human leukemia cells. The indicated concentrations of these compounds
`were incubated with cells for 72 h in the presence of a source of purines and/or
`thymidylate: no additions (©); 5 p.M thymidine (0); 100 ~M hypoxanthine (A); 300 p.~
`AICA (1); or 5 p.M thymidine plus 100 /~M hypoxanthine (¯). The reversal study of
`LY231514 was disclosed previously (11).
`
`Table 3 End products reversal studies with LY231514 and ZDI694a
`
`Alone
`
`+5/~M dThd
`
`ICsn of compound
`
`+ 100 p.M
`hypoxanthine
`
`25
`34
`220
`
`15
`4
`65
`
`138
`637
`3104
`
`>40,000
`>40,000
`>40,000
`
`32
`34
`1077
`
`13
`4
`44
`
`Cell line
`
`LY231514
`CCRF-CEM
`GC3/CI
`HCT-8
`ZD1694
`CCRF-CEM
`GC3/CI
`HCT-8
`
`’~ Cytotoxicity determined by MTT analysis after 72 h exposure to drug.
`SE of triplicate determinations did not exceed 10% of mean.
`
`1119
`
`+ dTHd and
`hypoxanthine
`
`>40,000
`>40,000
`>40,000
`
`>40,000
`>40,000
`>40,000
`
`Downloaded from cancerres.aacrjournals.org on October 16, 2015. © 1997 American Association for Cancer
`Research.
`
`Sandoz Inc.
`Exhibit 1021-0004
`
`JOINT 1021-0004
`
`

`
`[] ZD1694
`[] LY23~514
`
`1OO0O0
`
`1O0OO
`
`10
`
`LY231514. A MULTITARGETED ANTIFOLATE
`
`Transport Mechanimlns for Cytotoxic Activity. The roles of the
`RFC and FBP-a in the cytotoxic activity of LY231514 and ZD1694
`were determined by using ZR-75-1 human breast carcinoma sublines
`that differ in expression of RFC and FBP-a (23). Wild-type ZR-75-1
`cells express RFC as the major transport route for natural reduced
`folate cofactors and antifolate compounds and do not express detect-
`able levels of FBP-a. The predominant role of RFC in transport of
`these compounds is illustrated by the fact that wild-type ZR-75-1 cells
`with or without transfected FBP-a were much more sensitive to drug
`cytotoxicity than sublines resistant to methotrexate through decreased
`RFC expression (Table 5). These results indicated that both
`LY231514 and ZD1694 are less dependent on FBP-a as the major
`route for internalization.
`
`0
`
`0.3
`
`1 3
`Thymidine (pM)
`
`10
`
`DISCUSSION
`
`Fig. 4. Effect of thymidine on cytotoxicity of LY231514 and ZD 1694 against GC3/C 1
`human colon carcinoma cells.
`
`Table 4 Antiproliferative activity of various antifolates against a 5,10-
`dideazatetrahydrofolate-resistant CCRF-CEM subline (CR15) and relative efficiency as
`substrate for FPGS
`
`Compound
`
`LY231514
`ZD 1694
`LY249543 (lometrexol)
`Methotrexate
`
`CCRF-CEM
`IC5o (riM)a
`
`CR 15
`
`Its0 (I1M)a’b
`
`FPGS
`(VmaxlKm)c
`
`25.4
`15.3
`9.7
`4.2
`
`> 200,000
`30,657
`>200,000
`336
`
`549
`495
`60
`4
`
`’~ Cytotoxicity determined by M’lq" analysis after 72 h exposure to drug. SE of
`triplicate determinations did not exceed 10% of mean.
`bCRI5 cells, generously provided by Dr. G.P. Beardsley (Yale University, New
`Haven, CT), have markedly diminished capacity to accumulate 5,10-didcazatetrahydro-
`folic acid polyglutamates (their FPGS activity is approximately 10% of the wild type).
`c Hog liver FPGS data taken from Ref. 17.
`
`decrease in potency was observed with the addition of 100 ]IJ~M of
`hypoxanthine). In GC3/C1 cells, the physiological concentration of
`thymidine in mouse plasma (1.0 ~tM; Ref. 24) was significantly more
`effective in reversing the cytotoxicity of ZD1694 than LY231514
`(Fig. 4). The distinctively different reversal pattern exerted by thymi-
`dine on these two agents indicates that whereas TS may be the sole
`target for ZD1694, it is likely that there are other inhibitory sites for
`LY231514. The cell culture reversal pattern of LY231514 in CCRF-
`CEM cells was also distinctly different from that of methotrexate (no
`protection by thymidine alone; Fig. 3) and the GAR~I’ inhibitor
`Iometrexol (strong protection by hypoxanthine alo