Induction of Thymidine Phosphorylase Activity and ......Vol. 4, 1013-1019, April 1998 Clinical Cancer Research 1013 Induction of Thymidine Phosphorylase Activity and Enhancement of

Vol. 4, 1013-1019, April 1998 Clinical Cancer Research 1013

Induction of Thymidine Phosphorylase Activity and Enhancement

of Capecitabine Efficacy by Taxol/Taxotere in Human

Cancer Xenografts

Noriaki Sawada, Tohru Ishikawa, Yu Fukase,

Miwa Nishida, Takashi Yoshikubo, and

Hideo Ishitsuka’Cytostatics Group, Nippon Roche Research Center, Kamakura-city,Kanagawa 247-0853, Japan

ABSTRACT

Thymidine phosphorylase (dThdPase) is an essential

enzyme for the activation of the cytostatics capecitabine

(N�’-pentyloxycarbonyl-5’-deoxy-5-fluorocytidine) and its in-

tennediate metabolite [5’-deoxy-5-fluorouridine (5’-dFUrd)] to

5-fluorouracil in tumors. We have tried to identify the best

partners of capecitabine in combination therapy, such as

dThdPase up-regulators, which may enhance the efficacy of

this compound. Among various cytostatics studied with the

WiDr human colon cancer xenograft model, Taxol, Taxo-

tere, and mitomycin C greatly increased levels of human

dThdPase in tumors, and cyclophosphamide slightly in-

creased the enzyme level. These cytostatics simultaneously

increased the levels of human tumor necrosis factor a

(TNFa), which is an up-regulator of dThdPase. In cultures

of the WiDr cells, however, Taxol did not up-regulate TNFa

to a detectable level and only slightly enhanced levels of

dThdPase. These results suggest that Taxol might indirectly

elevate TNFa in tumor cells, which in turn up-regulated

dThdPase in the tumor cells in the WiDr cancer xenograft.

In the combination therapy, the efficacy of Taxol and Taxo-

tere with either capecitabine or 5’-dFUrd was more than

just additive. In contrast, Taxol and either 5-fluorouracil or

UFT (a mixture of tegafur and uracil) in combination

showed only additive activity. Taxol and Taxotere might

enhance the efficacy ofcapecitabine and 5’-dFUrd, probably

by modulating dThdPase activity in tumor tissues.

INTRODUCTION

Capecitabine generates 5-FUra2 selectively in tumors by

three enzymes located in the liver and in tumors; the final step

Received 9/5/97; revised 1/14/98;accepted 1/16/98.The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to

indicate this fact.I To whom requests for reprints should be addressed, at CytostaticsGroup, Nippon Roche Research Center, Kajiwara 200, Kamakura-city,Kanagawa 247-0853, Japan. Phone: 81-467-47-2220; Fax: 81-467-45-

6782.2 The abbreviations used are:5-FUra, 5-fluorouracil; dThdPase, thymi-dine phosphorylase; 5’-dFUrd, 5’-deoxy-5-fluorouridine; MMC, mito-mycin C; DXR, doxorubicin; CPA, cyclophosphamide; VCR, vincris-

is the conversion of the intermediate metabolite 5’-dFUrd to

5-FUra by dThdPase in tumors (1, 2). The enzyme is now

known to be an angiogenic factor (3, 4), and its activity is higher

in various tumor tissues than in normal tissues adjacent to the

tumors (5, 6). dThdPase is essential for the efficacy of capecit-

abine and 5’-dFUrd. The susceptibility of tumor cells to 5’-

dFUrd is reported to be enhanced by the transfection of the

dThdPase gene (7, 20), whereas the activity of 5’-dFUrd was

inhibited by dThdPase inhibitors (8). Therefore, it is expected

that dThdPase up-regulators would enhance the efficacy of

capecitabine and 5’-dFUrd. In fact, the efficacy of 5’-dFUrd has

been reported to be enhanced through the up-regulation of this

enzyme activity by various factors, such as IL-la, TNFa, IFN-y

(9), basic fibroblast growth factor, platelet-derived growth fac-

tor (10), and IFNot (11, 12). In addition, we preliminarily

reported that IL-12 enhanced the efficacy of 5’-dFUrd and

capecitabine in the in vivo tumor models indirectly through the

up-regulation of IFN�y, a dThdPase up-regulator (13, 14).

Capecitabine is being assessed in clinical trials worldwide,

and 5’-dFUrd is being marketed in Japan and in other Asian

countries for the treatment of breast, colorectal, gastric, and

other cancers. It is therefore of interest to know what factors and

drugs used in cancer treatment other than cytokines could either

directly or indirectly affect the levels of dThdPase. Among

cytostatics, Taxol has been reported to elevate levels of TNFa,

a potential dThdPase up-regulator, in macrophages (15-17). In

the present study, we examined various cytostatics, including

Taxol, to determine whether they affected dThdPase levels and

enhanced the efficacy of capecitabine and 5’-dFUrd in combi-

nation therapy. We found that several cytostatics, including

Taxol and Taxotere, up-regulated the enzyme level in tumors,

and these drugs in combination with either capecitabine or

5’-dFUrd showed additive to the synergistic activity. The mech-

anism of the up-regulation is also discussed.

MATERIALS AND METHODS

Animals. Four- or 5-week-old male and female BALB/c

nu/nu mice were obtained from SLC Japan, Ltd. (Shizuoka,

Japan).

Tumors. The human cancer lines used were obtained

from the following institutions: colon cancer WiDr from the

American Type Culture Collection (Rockville, MD) and MX-l

from Dr. T. Tashiro (Cancer Chemotherapy Center, Japanese

Foundation for Cancer Research, Tokyo, Japan). WiDr was

maintained in vitro cultures with Eagle’s MEM containing 10%

tine; VDS, vindesine; VLB, vinblastine; CDDP, cisplatin; MTX,methotrexate; TNFa, tumor necrosis factor a; IL, inter!eukin; MTh,maximum tolerated dose.

Association for Cancer Research. by guest on September 1, 2020. Copyright 1998 Americanhttps://bloodcancerdiscov.aacrjournals.orgDownloaded from

https://bloodcancerdiscov.aacrjournals.org

1014 Enhancement of Capecitabine Efficacy by Taxanes

Table 1 Effect of cytostatics on up-regulation of dThdPase and tumor growth inhibition in the WiDr human colon cancer xenograft

Tumor volume and body weight were measured 8 days after the single administration of cytostatics, and then tumor tissues (n = 3) were excised

for measuring the levels of dThdPase, human TNFa, and mouse TNFa.

dThdPase(units/mg protein) Human TNFa’� Mouse TNFa Carcass body

Dose (pg/g tissue), (ng/g tissue), Tumor volume change” weight change’ (g),

Treatment (mg/kg) mean ± SD fold mean ± SD mean ± SD (mm3), mean ± SD mean ± SD

Experiment 1Not treated 2.2 ± 0.5 100 ± 30 1.18 ± 0.42 332 ± 51 -0.7 ± 0.4

Saline 2.3 ± 0.6 1.0 <80 0.81 ± 0.15 328 ± 77 -0.7 ± 0.5

Taxol 100 17.6 ± 2.Y’ 8.0 576 ± 181 0.92 ± 0.24 -31 ± 53 -0.4 ± 0.9

Taxotere 15 13.5 ± 0.2” 6.1 373 ± 55 1.01 ± 0.07 214 ± 62 -3.7 ± 1.2

CDDP 10 5.0 ± 2.2 2.3 119 ± 27 1.24 ± 0.15 226 ± 53 -2.8 ± 1.02.5 2.5 ± 0.3 1.1 <90 1.19 ± 0.29 325 ± 109 -0.9 ± 0.6

DXR 7.5 2.7 ± 0.1 1.2 <109 1.30 ± 0.10 252 ± 48 -2.0 ± 1.4

1.5 2.3 ± 0.3 1.0 <90 1.25 ± 0.21 258 ± 50 -0.1 ± 1.1

MMC 5.0 16.9 ± 36d 7�7 265 ± 77 0.89 ± 0.10 76 ± 57 -2.3 ± 0.4

1.0 3.2 ± 1.2 1.5 <108 1.22 ± 0.49 301 ± 39 -0.9 ± 0.6VCR 1.5 3.1 ± 0.9 1.4 <80 0.80 ± 0.10 97 ± 165 -2.6 ± 0.4

0.3 2.5 ± 0.4 1.1 <80 1.16 ± 0.56 276 ± 81 -0.2 ± 0.6VLB 3.0 4.5 ± 08d 2.0 < I 14 0.94 ± 0.26 145 ± 100 -4.0 ± 0.6

0.6 3.3 ± 0.8 1.5 <80 0.80 ± 0.12 351 ± I 12 -2.0 ± 0.5VDS 5.0 4.7 ± 0.8” 2.1 102 ± 14 1.1 1 ± 0.22 49 ± 202 -3.5 ± 2.7

1.0 3.1 ± 1.0 1.4 <80 0.96 ± 0.31 272 ± 67 - 1.6 ± 0.5

Experiment 2Not treated 3.3 ± 0.8 <80 1.55 ± 0.19 408 ± 55 -0.1 ± 0.4

Taxotere 15 13.9 ± 15d 4.2 372 ± 79 1.52 ± 0.20 166 ± 18 -2.0 ± 0.8

CPA 200 9.2 ± I 3d 2.8 248 ± 22 1 .5 1 ± 0.06 263 ± 37 - 1 .9 ± 0.3

40 3.7 ± 0.6 1.1 121 ± 19 2.09 ± 0.26 299 ± 77 0.2 ± 0.6MIX 50 2.9 ± 0.1 0.9 98 ± 25 1.44 ± 0.26 268 ± 77 0.4 ± 0.8

10 3.3 ± 0.7 1.0 <82 1.33 ± 0.08 300 ± 101 0.4 ± 0.7

MMC 5.0 11.0 ± 7.3 3.3 <200 1.67 ± 0.21 142 ± 34 -0.1 ± 0.7

2.5 3.8 ± 1.0 1.2 96 ± 17 1.49 ± 0.16 243 ± 98 0.7 ± 0.3

1.0 4.0 ± 0.5 1.2 <113 1.62 ± 0.13 283 ± 139 -0.3 ± 1.0

a The detection limit of human INFa level is 80 pg/g tissue.b Measured tumor volume on day 8 minus that on day 0.

C Measured as carcass body weight on day 8 minus that on day 0.d Significantly different from not treated: P < 0.05.

FCS and nonessential amino acids. MX- 1 was maintained by

continuous passage in BALB/c nu/nu mice.

Human Cancer Xenograft Model. A single-cell sus-

pension (7-8 X 106 cells/mouse) of WiDr or small pieces of

MX- 1 was inoculated s.c. into nude mice. To evaluate the

antitumor effect of the cytostatics tested, we measured tumor

size and body weight when tumor tissues were excised for

measuring dThdPase levels. The experiments for WiDr and

MX- 1 were started when the tumor volume reached approxi-

mately 0.5 and 0.7 cm3, respectively. The tumor volume was

estimated by using the following equation, V = ab2/2, where a

and b are tumor length and width, respectively. For the evalu-

ation of cytostatic efficacy, tumor volume change was expressed

by the formula VT V0. where VT is the mean volume on any

given day and V0 is the mean volume at the start of adminis-

tration. Carcass body weight was calculated by subtracting the

tumor weight, which was estimated from tumor volume, from

the body weight (18). To evaluate the antitumor effect of flu-

oropyrimidines Taxol and Taxotere, tumor size and body weight

were measured twice a week.

Chemicals. Capecitabine was synthesized by the method

described elsewhere (19). 5’-dFUrd was synthesized at Hoff-

mann-La Roche (Basel, Switzerland). The other cytostatic drugs

were purchased from various suppliers: 5-FUra, MMC, and

DXR from Kyowa Hakko Co. (Tokyo, Japan); UFT (a mixture

of tegafur and uracil) from Taiho Pharmaceutical Co. (Tokyo,

Japan); Taxol and Cremophor EL from Sigma; Taxotere from

Rh#{244}ne-Poulenc Rorer (Antony, France); CPA, VCR, and VDS

from Shionogi Pharmaceutical Co. (Osaka, Japan); VLB from

Kyorin Pharmaceutical Co. (Tokyo, Japan); CDDP from Nippon

Kayaku Co. (Tokyo, Japan); and MTX from Takeda Pharma-

ceutical Co. (Osaka, Japan).

Capecitabine, 5’-dFUrd, and UFT were dissolved or sus-

pended in 40 mM citrate buffer (pH 6.0) containing 5% gum

arabic as the vehicle and then administered p.o. to mice. 5-FUra,

CPA, CDDP, DXR, MMC, VLB, VCR, VDS, and MTX were

dissolved in saline and given intraperitoneally. Taxol was dis-

solved in equal volumes of Cremophor EL and ethanol; this

dissolution was facilitated by sonication for 3-5 mm. The stock

solution of Taxol prepared in this manner waskept refrigerated.

The stock solution of Taxol was further diluted 1 :9 as needed

with saline so that the final working solution was 5% ethanol

and 5% Cremophor EL in saline. A diluted Taxol solution was

given to 2-3 mice i.v. within 5 mm of preparation. This proce-

dure allowed us to successfully administer Taxo! at a dosage of

100 mg/kg in a volume corresponding to 0.02 mug of body



0Vw

y�J

I� #{149}

A

E0,D)CCe

.C0a,E

0>

0E

I-

y= 24.819 + 22.881xn =48r’= 0.720p < 0.0001

U

0

.

. 0

Do

0

U

ci�(1)

(I)

0)

0.

U-zI-C(0EI

A

0

0 5 10 15 20 25

0

30

dThdPase (unit/mg protein)

Fig. 1 Correlation between dThdPase levels in tumor tissues and tumorvolume change. Data points, data from individual mice tested in theexperiment shown in Table 1 (experiment 1). V. not treated; V. saline;� , Taxol; #{149}, Taxotere; 0, CDDP; #{149},DXR; EL MMC; U, VCR; i�,

VLB; A, VDS.

0 5 10 15 20

dlhdPase (unit /mg protein)

25 30

Fig. 2 Correlation between dThdPase and human INFa levels in tu-

mor tissues. Data points, data from individual mice tested in the exper-iment shown in Table 1 (experiment 1). Significance was calculated

assigning the human TNFa level below the detection limit as 80 pg/g of

tissue.

Clinical Cancer Research 1015

weight. Taxotere was dissolved in saline containing 2.5% eth-

anol and 2.5% polysorbate 80 and given iv. To evaluate the

antitumor effect of the cytostatics tested, the cytostatics were

given at doses ranging from #{189}to 1/is of the single LD50, except

for Taxol.

dThdPase Assays. Tumor tissues were homogenized in

10 mM Tris buffer (pH 7.4) containing 15 mtvi NaCl, 1.5 mrsi

MgCl2, and 50 p.M potassium phosphate with a glass homoge-

nizer, whereas tumor cells cultured were sonicated in 10 mrsi

PBS solution (pH 7.4). The homogenate was then centrifuged at

10,000 X g for 20 mmat 4#{176}C,and the supernatants were stored

at - 80#{176}Cuntil use. For the evaluation of Taxol and Taxotere on

WiDr in the time course experiments (see Fig. 3), the homoge-

nate was centrifuged at 105,000 X g for 90 mm. at 4#{176}C.The

supernatants were then dialyzed overnight against 20 nmt po-

tassium phosphate buffer containing I mt�i 3-mercaptoethanol

and stored at - 80#{176}Cuntil use. The protein concentration was

determined by the methods of Lowry et a! (21). The dThdPase

level was measured by ELISA with monoclonal antibodies

specific to human dThdPase as described previously by Nishida

et a!. (5). The monocional antibodies were not cross-reacted

with mouse dThdPase. One unit corresponds to the dThdPase

level of the standard enzyme solution (extracts of human colon

cancer xenograft HCT 116), which phosphorolyzes 5’-dFUrd to

5-FUra at a rate of 1 �i.g of 5-FUra/h (5).

TNFa Assays. The supernatants prepared from cul-

tured cells and tumor tissues were examined for their levels

of TNFes with ELISA method. Human TNFes produced in

tumor cells in human cancer xenografts was measured using

QUATIKINETM ELISA kit (R&D Systems), and mouse

TNFa produced by host cells was measured using a mouse

TNFa ELISA kit (Endogen). These ELISA assays are spe-

cific to only human TNFa and mouse TNFa, respectively.

Statistical Analysis. Tumor volume change and the dif-

ference of carcass body weight were analyzed using the

ANOVA test. The up-regulation of dThdPase was analyzed

using the Student’s t test. Differences were considered signifi-

cant when P was <0.05.

RESULTS

dThdPase Up-Regulation. Table 1 shows dThdPase

levels in tumor tissues, which were measured 8 days after the

test drugs were administered to mice bearing the WiDr human

colon cancer xenograft. Because the dThdPase levels were de-

tected by the ELISA with monoclonal antibodies specific to

human dThdPase, the levels of dThdPase measured corre-

sponded to those of the enzyme produced in the human cancer

cells. Taxol, Taxotere, and MMC greatly enhanced the levels of

dThdPase in the tumors, whereas CPA slightly enhanced the

enzyme level. In contrast, other cytostatics, such as CDDP,

DXR, VCR, VLB, VDS, and MTX, did not enhance dThdPase

levels.

Positive Correlation between dThdPase and Human

TNFa Up-Regulations. To obtain an insight to mechanisms

of the dThdPase elevation by the cytostatics, we compared the

degree of the dThdPase elevation with that of the antitumor

activity of the cytostatics and with that of the human TNFa

elevation in each mouse treated. There appears to be no corre-

lation between the dThdPase up-regulation and the efficacy

(Table 1 and Fig. 1). Only Taxol, Taxotere, and MMC elevated

dThdPase levels at doses showing the antitumor activity. There

was also no correlation between the up-regulation of dThdPase

and that of mouse TNFa, which is produced by host stromal

cells in the tumor tissue. In contrast, the dThdPase and human

TNFa elevation in tumors correlated with each other (Fig. 2).

This positive correlation was also observed in the time course

experiments, in which dThdPase and human TNFa levels were



2 4 6 8 10 2 4 6 8 10

Days after Taxol administration Days after Taxotere administration

C�8

0

0.0)

E

C

Fig. 4 Effect of Taxol and Taxotere on human

TNFa of the WiDr human colon cancer tumor innude mice. The method is described in the legendto Fig. 3. Columns, mean human TNFa values;

bars, SD. Human TNFa levels below the detectionlimit were assigned as 80 pg/g of tissue. a: 0, notreatment; �, vehicle; 0, 15 mg/kg Taxol; U, 100mg/kg Taxol. b: D, no treatment; �, vehicle; 0,3.75 mg/kg Taxotere; U, 15 mg/kg laxotere.

a. Taxol b. Taxotere

Days afterTaxol administration2 4 6 8 10 2 4 6 8 10

Days afterlaxotere administration


(U

II

Fig. 3 Effect of Taxol or Taxotere on dThdPase

up-regulation of the WiDr human colon cancertumor in nude mice. Mice bearing WiDr tumorsare randomized into groups of three mice each,and Taxol or Taxotere was administered iv. Micewere sacrificed and tumor tissues were excised atvarious days after administration. Columns, meandThdPase value; bars, SD. *, significantly differ-

ent from vehicle at each time point; P < 0.05. a:

0, no treatment; �, vehicle; 0, 15 mg/kg Taxol;U, 100 mg/kg Taxol. b: 0, no treatment; �, ye-

hide; 0, 3.75 mg/kg Taxotere; U, 15 mg/kg Taxo-

tere.

measured at several time points after Taxol and Taxotere were

given to mice bearing the WiDr human cancer xenograft (Figs.

3 and 4). Similar time course kinetics of dThdPase elevation was

observed in mice bearing other human cancer xenograft line,

MX-1 breast cancer (data were not shown).

dThdPase and TNFa Levels in Cell Cultures. To in-

vestigate whether Taxol directly up-regulates TNFa and

dThdPase, the WiDr human cancer cells were exposed to Taxol

in vitro for 4 h, and levels of TNFct and dThdPase were

measured after cultures of 3 days. The level of dThdPase in the

cultured WiDr cells was much lower than that observed in the

WiDr cancer xenograft (0.19 versus 2.2 units/mg of protein). In

addition, Taxol at concentrations of 150 nt�i (IC50 in 4 h drug-

exposure and the following 3-day culture) and 1000 nt�i en-

hanced the level of dThdPase by only 2.0- and 2. 1-fold and did

not induce TNFa up to a detectable level. Taxol might indirectly

induce TNFa and dThdPase in the human cancer xenograft

model.

Combination Therapy. The efficacy of fluoropyrimi-

dines in combination with Taxol or Taxotere was compared in

the WiDr human colon cancer xenograft, which is refractory to

fluoropyrimidines, and the MX-! human breast cancer xe-

nograft. Fluoropyrimidines were given at their MTh (WiDr) or

#{189}MTD (MX-1), whereas taxanes were given over a range of #{189}

MTD to MTD. Fig. 5, a and b, shows that the efficacy of Taxol

in combination with either capecitabine or 5’-dFUrd was more

than just additive, and the tumor regressed. In contrast, Taxol

and either 5-FUra or UFT in combination showed only additive

activity (Fig. 5, c and d). On the other hand, toxicity in terms of

weight loss does not appear to be synergistic (Fig. 5). The

additive to synergistic activity was also observed in combination

therapy with Taxotere and capecitabine in the same tumor

model (Fig. 6) and in the MX-1 human breast cancer xenograft

model (Fig. 7). No increase or additive effect in the toxicity was

observed in the combination with the taxanes and capecitabine,

and the degree of the increase was similar to that of the corn-

bination with the taxanes and 5-FUra (Figs. 6 and 7).

DISCUSSION

dThdPase, which converts 5’-dFUrd to 5-FUra, is essential

for the activation of capecitabine and 5 ‘-dFUrd in human cancer

cells (1, 2). The present study showed that several cytostatics,

such as Taxol, Taxotere, MMC, and CPA, elevated this enzyme

level in tumor cells in a human cancer xenograft. In addition,

both Taxol and Taxotere have thus far showed beuer efficacy in

combination therapy with 5’-dFUrd or capecitabine than with

5-FUra or UFF. The major difference between 5’-dFUrd/cape-

citabine and 5-FUra is that the efficacy of the former com-

pounds is greatly modulated by dThdPase. Elevating these en-

zyme levels by gene transfection or by dThdPase up-regulators

made tumor cells more susceptible to 5’-dFUrd (7, 20). Taxol

and Taxotere might enhance the efficacy of 5’-dFUrd and cape-

citabine by up-regulating the dThdPase levels in tumors.

Taxol and Taxotere up-regulated dThdPase levels at doses

inhibiting tumor growth; however, the enzyme up-regulation

does not always appear to correlate with tumor growth inhibi-

tion. CDDP, DXR, VCR, VLB, VDS, and MTX did not enhance

the enzyme level, although some of them inhibited tumor

growth. Levels of mouse TNFa produced in host stromal cells

in tumor tissues were not changed by the cytostatics. They were

elevated without Taxol or Taxotere probably by tumor environ-



a. Capecitabine b. 5’-dFUrd0.8

0.6

0.4

0.2

0

EU

U,0)C(U

U(UE0

0

EI-

0.8

0.6

0.4

0.2

0

1-Ii

Fig. 5 Antitumor activity of thecombination of Taxol and fluoropy-rimidines on the growth of WiDrcolon cancer xenograft in nudemice. Taxol was administered iv.weekly at the optimal dose (15 mg/kg), and fluoropyrimidines were ad-ministered p.o. or i.p. five times aweek at the MID. Data points,

mean values of tumor volumechange and carcass body weight;

bars, SD. *, significantly differentfrom vehicle; P < 0.05. 0, control;., Taxol; & capecitabine; A, cape-citabine + Taxol; D, 5’-dFUrd; U,5’-dFUrd + Taxol; V, 5-FUra; V.5-FUra + Taxol; c� , UFT; #{149}, UFT+ Taxol.

�a�

2U)Cl)CU

�

�15 2b 25 30 35 40 45

�24

�*�j,,,.“�20 1816

4’

1�5 20 25 30 35 40 45

5015

2422201816

20 25 30 35 40

�1

45

15 20 25 30 35 40 4550

C.)Days after tumor inoculation Days after tumor inoculation

5’O

c. 5-FUra d. UFT0.8

0.6

0.4

0.2

0

EU

C)0)CCU

U

(UE0>

0

EI-

2?

1

I

i 20 25 3b 35 40 45

0.8

0.6

0.4

0.2

0

-0.2

� ‘�5

Days after tumor inoculation

2422201816

.C0)(U

>��00

.0Cl)Cl)CUUCU

C-)

50 5 20 25 30 35 40 45 50

15 20 25 30 35 40 45


50


mental stimuli or by infiltration ofhost cells producing TNFes to the

tumor tissues, such as macrophages. In contrast, the elevation of

dThdPase levels in tumors by the cytostatics accompanied by

increased levels of human TNFa, which is produced by human

tumor cells, in the human cancer xenograft. We previously ob-

served that TNFa, as well as IL-la and IFN’y, up-regulated dTh-

dPase activity in tumor cells and made the cells more susceptible to

5’-dFUrd (9). Taxol and Taxotere might increase dThdPase levels

indirectly through up-regulation of TNFa or by another dThdPase

up-regulator though unknown mechanisms.

Taxol and Taxotere enhanced the levels of TNFa and

dThdPase in the WiDr human cancer xenograft starting at day 4

or 6 after treatment started. However, Taxol is known to induce

TNFa in 1 day in cultures of macrophages (16), whereas we

observed that TNFes up-regulated dThdPase in the WiDr tumor

cells also in 24 h (9). It is not yet clarified why there was a long

delay between the single dose of TaxolfFaxotere and the induc-

tion of dThdPase in the tumors. The present study in vitro

indicates that Taxol, however, did not induce any detectable

levels of TNFa in the WiDr cells. In addition, dThdPase levels

in the tumor cells in cultures were very low as compared with

those in the human cancer xenograft, and Taxol enhanced

dThdPase levels only 2-fold in 3-day cultures. These results

indicate that the taxanes might indirectly enhance human TNFes

and dThdPase in the tumor cells in tissues. Taxanes might affect

host stromal cells in tumor tissues, resulting in the up-regulation

of TNFa in tumor cells and consequently that of dThdPase.

These processes may take a certain number of days.



a. Capecitabine b. 5-FUra

1.0

0.8

0.6

0.4

0.2

0

10 15 20 25 30 35 40

:�

U�0).�>,.�.0��

0

26

22

18

14

Ilv.

10

.

45-U.� ‘ ‘ , , ,

10 15 20 25 30 35,

40,

45

. 101’45 10 15 20 25 30 35 40 45


15 20 25 30 35 40


b. 5-FUra6.0

5.0

E.i�. 4.0U)0)C

.� 3.0U

I 2.0j 1.0

�11ii

a. Capecitabine 6.0

5.0

4.0

3.0

2.0

1.0

0

182226303438 42 46 14

()

140)

0)U)

.�

.0Cl)Cl)CUUCU

0

14

18 22 26 30 34 38 42 46

14 18 22 26 30 34 38 42 46 14


18 22 26 30 34 38 42 46



1.0

C� 0.8U

U)�) 0.6CU.� 0.4

E

� 0.2

Fig. 7 Antitumor activity of the

combination of Taxotere and cape-

citabine or 5-FUra on the growth of

MX- 1 mammary cancer xenograft

in nude mice. Taxotere was admin-

istered iv. weekly at V8 of the MID,

which had been obtained in this

model (3.75 mg/kg), and fluoropy-

rimidines were administered p.o. ori.p. five times a week at the optimaldose (V2 of the MID). Data points,

mean values of tumor volume

change and carcass body weight;

bars, SD. *, significantly different

from vehicle; P < 0.05. 0, control;

., Taxotere; EL capecitabine; U,capecitabine + laxotere; �,

5-FUra; A, 5-FUra + laxotere.

Fig. 6 Antitumor activity of thecombination of Taxotere and cape-

citabine or 5-FUra on the growth of

WiDr colon cancer xenograft innude mice. Taxotere was adminis-

tered iv. weekly at the MID, which

had been obtained in this model

(3.75 mg/kg), and fluoropyrimi-dines were administered p.o. or i.p.

five times a week at the MID. Data

points, mean values of tumor vol-ume change and carcass body

weight; bars, SD. *, significantly

different from vehicle; P < 0.05. 0,control; #{149},Taxotere; L�, capecitab-

me; A, capecitabine + laxotere; V.5-FUra; V. 5-FUra + Taxotere.

In the present study with human cancer xenografts, we

measured tumor levels of human dThdPase only by using an

ELISA with monoclonal antibodies specific to human

dThdPase. It is likely that Taxol and Taxotere also enhance

the toxicity of capecitabine and 5’-dFUrd by up-regulating

the enzyme in host organs in the tumor model. However, the

toxicity of Taxol/Taxotere and capecitabine does not appear

to be synergistic, although the efficacy of these compounds in

combination was additive to synergistic, and the degree of the

increase was similar to that of the combination with the

taxanes and 5-FUra. In addition, we have observed that

neither Taxol nor Taxotere elevated the enzyme activity,

which converts 5’-dFUrd to 5-FUra, in the intestinal tract of

mice thus far tested (data were not shown). Therefore, it is

unlikely that Taxol or Taxotere greatly enhances the enzyme

activity in normal organs. The advantage of the combination

with either Taxol or Taxotere and capecitabine should be

pursued further. If Taxol or Taxotere enhances the enzyme

activity mainly in tumors in patients, clinical assessment of

the combination would be warranted.




ACKNOWLEDGMENTS

We thank Dr. Y. Tanaka, M. Endo, and N. Inagaki for helpful

discussion.

REFERENCES1. Kono, H., Hara, Y., Sugata, S. Karube, Y., Matsushima, Y., and

Ishitsuka, H. Activation of 5’-deoxy-5-fluorouridine by thymidine phos-

phorylase in human tumor tissues. Chem. Pharm. Bull. (Tokyo), 31:

175-178, 1983.

2. Miwa, M., Ura, M., Nishida, M., Sawada, N., Ishikawa, I., Mon. K.,Shimma, N., Umeda, I., and Ishitsuka, H. Design of a novel oral

fluoropyrimidine carbamate, capecitabine, which generates 5-fluoroura-cil selectively in tumors by enzymes concentrated in human liver and

cancer tissue. Eur. J. Cancer, in press, 1998.

3. Miyadera, K., Sumizawa, I., Haraguchi, M., Yoshida, H., Konstanty,

W., Yamada, Y., and Akiyama, S. Role of thymidine phosphorylaseactivity in the angiogenic effect of platelet-derived endothelial cell

growth factor/thymidine phosphorylase. Cancer Res., 55: 1687-1690,

1995.

4. Fox, S. B., Westwood, M., Moghaddam, A., Comley, M., lurley, H.,

Whitehouse, R. M., Bicknell, R., Gatter, K. C., and Harris, A. L. Theangiogenic factor platelet-derived endothelial cell growth factor/thymi-

dine phosphorylase is up-regulated in breast cancer epithelium and

endothelium. Br. J. Cancer, 73: 275-280, 1996.

5. Nishida, M., Hino, A., Mori, K., Matsumoto, I., Yoshikubo, I., and

Ishitsuka, H. Preparation of anti-human thymidine phosphorylase mono-

clonal antibody useful for detecting the enzyme levels in tumor tissues.

Biol. Pharm. Bull., 19: 1407-1411, 1996.

6. Takebayashi, Y., Akiyama, S., Akiba, S., Yamada, K., Miyadera, K.,Sumizawa, I., Yamada, Y., Murata, F., and Aikou, I. Clinicopathologicand prognostic significance of an angiogenic factor, thymidine phos-phorylase, in human colorectal carcinoma. J. Natl. Cancer Inst., 88:

1110-1117, 1996.

7. Haraguchi, M., Furukawa, I., Sumizawa, I., and Akiyama, S. Sen-sitivity of human KB cells expressing platelet-derived endothelial cell

growth factor to pyrimidine antimetabolites. Cancer Res., 53: 5680-

5682, 1993.

8. Miwa, M., Cook, A., and Ishitsuka, H. Enzymatic cleavage of various

fluorinated pyrimidine nucleosides to 5-fluorouracil and their antipro-liferative activities in human and murine tumor cells. Chem. Pharm.

Bull. (Tokyo), 34: 4225-4232, 1986.

9. Eda, H., Fujimoto, K., Watanabe, S., Ura, M., Hino, A., Tanaka, Y.,Wada, K., and Ishitsuka, H. Cytokines aiduce thymidine phosphorylaseexpression in tumor cells and make them more susceptible to 5 ‘-deoxy-

5-fluorouridine. Cancer Chemother. Pharmacol., 32: 333-338, 1993.

10. Eda, H., Fujimoto, K., Watanabe, S., Ishikawa, I., Ohiwa, I.,

latsuno, K., Tanaka, Y., and Ishitsuka, H. Cytokines induce uridinephosphorylase in mouse colon 26 carcinoma cells and make the cellsmore susceptible to 5’-deoxy-5-fluorouridine. Jpn. J. Cancer Res., 84:

341-347, 1993.

1 1. Tevaerai, H. I., Laurent, P. L., Suardet, L., Eliason, J. F., Give!,

J. C., and Odartchenko, N. Interactions of interferon-a2a with 5 ‘-deoxy-5-fluoroundine in colorectal cancer cells in vitro. Eur. J. Cancer, 28:

368-372, 1992.

12. Schwartz, E. L., Hoffman, M., O’Connor, C. J., and Walder, S.Stimulation of 5-fluorouracil metabolic activation by interferon-a2a in

human colon carcinoma cells. Biochem. Biophys. Res. Commun., 182:

1232-1239, 1992.

13. Tanaka, Y., Endo, M., Ouchi, K., Mori, K., and Ishitsuka, H.Augmentation of antitumor activity of 5 ‘-deoxy-5-fluorouridine byIL-l2 through the up-regulation of pyrimidine nucleoside phosphorylasein munne tumor models. Eur. J. Cancer, 33 (Suppl. 8): Sl76-S177,1997.

14. Ishitsuka H., Ishikawa, I., Fukase, Y., Sawada, N., Tanaka, Y.,

Ouchi, K., Yoshikubo, I., and Nishida, M. Capecitabine and the

dThdPase up-regulators IFN-y or laxol showed synergistic activity in

human cancer xenografts. Proc. Am. Assoc. Cancer Res., 37: 2766, 1996.

15. Ding, A. H., Porteu, F., Sanchez, E., and Nathan C. N. Shared

actions of endotoxin and Taxol on INF receptors and TNF release.

Science (Washington DC), 248: 370-372, 1990.

16. Bogdan, C., and Ding, A. Taxol, a microtubule-stabilizing antineo-

plastic agent, induces expression of tumor necrosis factor a and inter-

leukin-l in macrophages. J. Leukocyte Biol., 52: 1 19-121, 1992.

17. Burkhart, C. A., Berman, J. W., Swindell, C. S., and Horwitz, S. B.

Relationship between the structure of Taxol and other taxanes on in-

duction of tumor necrosis factor-a gene expression and cytotoxicity.

Cancer Res., 54: 5779-5782, 1994.

18. Tanaka, Y., Edit, H., Tanaka, I., Udagawa. I., Ishikawa. I., Horii, I.,and Ishitsuka, H. Experimental cancer cachexia induced by transplantablecolon 26 adenocarcinoma in mice. Cancer Res., 50: 2290-2295, 1990.

19. Arasaki, M., Ishitsuka, H., Kuruma, I., Miwa, M., Murasaki, C.,Shimma, N., and Umeda, I. N-Oxycarbonyl substituted 5’-deoxy-5-

fluorocytidines. European Patent Application No. 92121538.0, 1992.

20. Eda, H., Hasegawa, M., Tatsuno, K., Matsumoto, I., Watanabe, S.,Uchida, I., and Ishitsuka, H. Characteristics of human colon cancer

cells transfected with thymidine phosphorylase gene (in Japanese).

J. Jpn. Soc. Cancer Ther., 29: 164, 1994.

21. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.Protein measurement with the Folin phenol reagent. J. Biol. Chem., 193:

265-275, 1951.



Documents

Induction of Thymidine Phosphorylase Activity and ......Vol. 4, 1013-1019, April 1998 Clinical Cancer Research 1013 Induction of Thymidine Phosphorylase Activity and Enhancement of