Stimulation of 5-fluorouracil metabolic activation by interferon-α in human colon carcinoma cells

Vol.

182,

No.

February

3, 1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

14, 1992

Pages

1232-l

239

STIMULATION OF 5-FLUOROURACIL METABOLIC ACTIVATION BY INTERFERON-u IN HUMAN COLON CARCINOMA CELLS’ Edward

L. Schwartz’,

Mark Hoffman, Carolyn and Scott Wadler

J. O’Connor

Department of Oncology, Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, New York 10467

Received

December

23,

1991

Interferon-o (IFNo) increases the cytotoxicity of 5fluorouracil (FUra) in vitro, and the combination has clinical efficacy against advanced colorectal cancer. IFNatreatment of HT-29 colon carcinoma cells induced a greater than two-fold increase in the intracellular levels of the active metabolite of FUra, FdUMP. Using cell extracts from HT-29 cells and FUra as substrate, IFNo produced a 1.9- and 8.7-fold increase, respectively, in the activities of uridine phosphorylase and pyrimidine nucleoside phosphorylase (PyNP). Furthermore, the effect was selective for the conversion of FUra to FdUMP, as IFNa did not increase the cellular levels of FUTP, nor did it change the extent of incorporation of FUra into RNA (or DNA). IFNo also had no effect on thymidine kinase activity, the second step in the activation of FUra. Hence the effect of IFNu on PyNP activity is likely a critical biochemical event that modulates the cytotoxicity of FUra. B 1992 Academic Press, Inc.

The fluorinated agent,

alone

pyrimidine,

or in combination

carcinoma.

As a single

combination

with

the aspartate

transcarbamylase

is markedly

also been shown

(FUra 1, is the most

other drugs,

FUra has only

modulating inhibitor,

agents

to synergistically

in combination

enhance

with advanced with

for the treatment modest

anticancer

commonly

employed

of advanced

colorectal

effects;

such as the reduced

folate,

N-(phosphonacetyl)-L-aspartate,

both in vitro and clinically

in clinical trials in patients employed

with

agent,

biochemical

enhanced,

5-fluorouracil

Recombinant

the cytotoxic

colorectal

FUra resulted

(1,2).

effects

carcinoma,

in objective

in

leucovorin,

or

the efficacy of FUra human

interferons

have

of FUra in vitrg (3-71, and

recombinant

clinical

however,

response

alfa-2a-IFN rates higher

(IFNu) than

’ Supported by grants from the American Cancer Society (CH-4791, the Mathers Foundation, and Cancer Center Support Grant P30 CA13330 from the National Cancer Institute. Dr. Wadler is a recipient of an American Cancer Society Career Development Award. ’ To whom

correspondence

should

be addressed.

Abbreviations used are: dAdo. deoxyadenosine; dThd, thymidine; FdUMP, 5fluorodeoxyuridylate; FdUrd, 5-fluorodeoxyuridine; FUra, 5-fluorouracik IFNu, recombinant human interferon-a2a; R-l-P, ribose-1 -phosphate. 0006-291X/92 $1.50 Copyright 0 1992 by Academic Press, Inc. All rights of reproduction in any form reserved.

1232

Vol.

182, No. 3, 1992

those expected

BIOCHEMICAL

with FUra alone (8-I 2). The mechanism

has not been determined, disposition

but might include

of interaction

alterations

between

in the cellular

uptake,

these agents metabolism

and

and three sites of action

have

of FUra.

The biochemistry been

AND BIOPHYSICAL RESEARCH COMMUNICATIONS

identified:

thymidylate

of FUra has been extensively

(a) binding

synthase

FUra into RNA (15-l

of the

FUra

(TS) with inhibition

metabolite,

fluorodeoxyuridylate

of the enzyme

9); and (c) incorporation

each of these to the effects

studied,

activity

(13-l

into DNA (20-21).

of FUra likely varies depending

(FdUMP),

to

4); (b) incorporation

of

The relative

importance

of

on the cell line and conditions

employed. The variety

IFNs have well

of normal

mediated

described

and transformed

by the binding

transcriptional

activation

an increase

in the activity

of several

action

nor how

it potentiates

agents

(reviewed

pathway

for the metabolic

in 22-23).

to cell surface

enzymes. Neither

the antitumor

in 24) have been determined.

in the activity

the active

cells (reviewed

of several genes, the appearance

by IFN treatment.

increase

and antiproliferative

of the polypeptide

are suppressed

FdUMP,

antiviral

of an enzyme activation

metabolite

which

The cellular receptors,

In addition,

actions which

the expression

of cytotoxic

cancer

a wide

of IFN are

leads to the proteins,

and

of other mRNAs

for IFN’s direct

In this study, we observed converts

against

of newly-synthesized

the mechanism activity

effects

antiproliferative

chemotherapeutic an IFNo-stimulated

FUra to FdUrd, the first step in one

of the fluoropyrimidine,

and in the cellular

levels of the

of FUra.

MATERIALS

AND METHODS

Cell Culture. HT-29 human colon carcinoma cells (25) were maintained in RPM1 1640 with 10% fetal bovine serum (GIBCO) in 5% CO,, and were free of Mvcoplasma contamination. [6-3HlFUra metabolism. HT-29 cells were incubated with 5.5 PM f6-3HlFUra (1.8 Ci/mmol; Moravek Biochemicals) with or without 500 U/ml rlFNu-2a (500 U/ml) (Hoffman-LaRoche, Nutley, NJ). At intervals up to 24 hrs, aliquots were removed, washed with phosphatebuffered saline (PBS), and extracted with 0.5 N perchloric acid (PCA). After a brief centrifugation, the supernatants were removed and neutralized as described (26) with 1.5 vol of alaminelfreon, lyophilized, and reconstituted in HPLC mobile phase. An aliquot was analyzed by HPLC; fractions were collected and radioactivity determined by liquid scintillation counting. Recovery of radioactive material was 87-95%. Separation of fluoropyrimidine bases, nucleosides and nucleotides (27) was done using an Adsorbosphere Cl8 column (Alltech) with the following conditions: 5 mM tetrabutylammonium hydrogen sulfate in 5 mM potassium phosphate, pH 7 for 10 min at 2 ml/min, then a gradient of 0 to 5% methanol in the same buffer over 5 minutes (at 1.5 ml/min), followed by a gradient of 5% to 72% methanol over 25 min. at 1.5 mllmin. Identity of radioactive peaks was determined using authentic standards (with retention times indicated): FUra, 3.3 min; FUrd, 6.5 min; FdUrd, 8.5 min; FUMP, 18 min; FdUMP, 22 min; FUDP, 33 min; FdUDP, 33 min; FdUTP, 35 min; FUTP, 36 min. Because of inadequate resolution of FUDP from FdUDP and FdUTP from FUTP, an aliquot of reconstituted cell extract was treated with sodium periodate followed by methylamine to digest deoxyribonucleotides (28). allowing for the determination of the ribonucleotides. A radioactive peak was observed immediately after the FdUMP peak in all experiments. This peak was completely abrogated by periodate 1233

Vol.

182, No. 3, 1992

oxidation followed of total metabolite,

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

by methylamine cleavage. This ribonucleotide was most likely FUDP sugars, as described

peak, accounting previously (29).

for 7-l 2%

Assays of enzyme activity. HT-29 cells were treated with IFNo for 24 hrs (500 U/ml), washed with PBS, and resuspended in 50 mM Tris-HCI (pH 7.5), 1 mM EDTA. The cells were sonicated on ice and stored at -70° C until assayed. For determination of thymidine kinase activity, cell extracts (100 pug protein) were incubated in a total volume of 150 ~1 containing 0.1 M Tris-HCI, pH 7.5, 10 mM MgCI,, 10 mM ATP, 4 mM dithiothreitol, 1 mglml albumin, and 0.1 mM 3H-dThd (40,!Xi/ml). At intervals up to 1 hour, duplicate aliquots were removed, EDTA was added (33 mM final concentration) to stop the reactions, and the samples were spotted on DE81 (Whatman) filter paper. Filters were washed with cold 5% trichloroacetic acid (TCA) and acid-insoluble radioactivity determined. Pyrimidine nucleoside phosphorylase (PyNP) activity was assayed as described (30) by incubating cell extracts (50 I.rg) in 30 mM Tris-HCI, pH 7.4, 1 mM EDTA, 5 mM MgCI,, 2.5 mM deoxyribose-l-phosphate (dR-l-P), and 0.25 mM [6-3Hl-FUra (20 &i/ml). For analysis of uridine phosphorylase activity, R-l-P was substituted for dR-1-P. At times up to 30 minutes, reactions were stopped by boiling, briefly centrifuged, and supernatants were spotted on silica gel thin layer chromatography plates. Separation of 3H-FUra from 3H-FUra-nucleosides was as previously described (31). Incorporation of [6-3HIFUra into RNA and DNA. various concentrations of [6-3Hl-FUra (2 ,&i/ml) precipitated with PCA, and RNA was hydrolyzed reprecipitated with PCA and radioactivity in both by liquid scintillation counting.

Cells were treated for 24 hours. Cells with 1 M KOH (37O the RNA and DNA

with 500 U/ml were washed for 16 hours). fractions was

IFNo and with PBS, DNA was measured

Thymidine salvage. Cells were treated with IFNo (500 U/ml) for 24 hrs, and were then incubated with 1 pCi/ml of either 3H-dAdo or 3H-dThd for an additional hour. Cells were trypsinized, counted, briefly centrifuged, treated with ice-cold 5% TCA, and washed twice with TCA. The radioactivity in the pellet was determined by liquid scintillation counting. Incorporation of radioactivity into RNA was eliminated in cells treated with 3H-dAdo by incubation with 0.2 M NaOH (1 hr at 37”) prior to the addition of the TCA.

RESULTS Although are dependent the effect

several

biochemical

upon the metabolic

mechanisms activation

of IFNo on the levels of cellular

of action

have been described

of FUra to a nucleotide metabolites

for FUra, all

(Fig. 1). We determined

of [6-3H]FUra,

including

the active

oyFi[=FU,DP~FUTP-RNA

FUra

-

FUrd

TK FdUrd

_

FdUDP

=

FdUTP

-c

DNA

iI FdUMP

d

7\ FdUMP

-

CH,THF

Fia. 1 1 Schematic diagram of the metabolism of FUra. Abbreviations used are: CH,THF, 5,10-methylene tetrahydrofolate; OPRT, orotate phosphoribosyl transferase; PyNP, pyrimidine nucleoside phosphorylase; TK, thymidine kinase; TS, thymidylate synthase; UrdP, uridine phosphorylase. 1234

Vol.

182,

No.

3, 1992

TABLE

1. Effect

BIOCHEMICAL

AND

BIOPHYSICAL

of IFNa

Levels

of Cellular

on the omol/lO’

Metabolite FUra FUMP FdUMP FUDP FUTP FU-sugar Total FU-3H

Control 0.83 f 0.72 f 0.82 f 1.33 f 6.77 f 1.02 + 11.6 f

RESEARCH

Metabolites

cells IFNu-treated 0.73 f 0.10 0.49 f 0.12 1.86 f 0.09 1.07 f 0.38 3.90 f 0.57 0.62 f 0.06 8.74 rt 0.65

0.16 0.29 0.09 0.46 1.91 0.24 2.46

anabolism increase carcinoma instances hexose,

FUTP of FUra was cell

FdUMP.

line,

and

Incubation

to FdUMP

by

seen as early

produced FUTP,

and

SW480 a decrease

not in the

(Table

2.0 -

2.4-fold

IFNa

in HT-29

for

cells

1).

shown).

The

levels

h selectively 1 and

IFNo

of radiolabelled

IFNu-induced

with or without rlFNu-2a were separated by HPLC * indicates significantly

Fig.

also observed

In contrast,

cellular

24

(Table

as 12 h (Fig. 2). and was (data

FdUrd

with

of FUra

% Control 88% 68% 227%’ 80% 58% 61% 75%

HT-29 cells were incubated with 5.5 PM IB-“HlFUra (1.8 Cilmmol) (500 U/ml) for 24 hours. Cells were extracted and FURa metabolites as described in Methods. Data are means f SEM of 3 experiments. different from control, p < 0.05.

metabolites

COMMUNICATIONS

increase

enhanced 2).

the

A significant

in a second

colon

did not

increase,

and

in some

FUra,

FUMP,

FUDP,

FUDP-

in cellular

FdUMP

was

A

1.6 10

5 0.4

-

0 0

4

8

12

18

20

24

0

Time

4

8

12

16

20

24

Ihours)

&J, Effect of IFNo on the metabolism of FUra to FdUMP. HT-29 cells were incubated with 5.5 NM [6-3HlFUra with (closed symbols) or without (open symbols) IFNo (500 U/ml) for the indicated times. Cells were extracted and FUra metabolites separated by HPLC as described in Methods. Levels of I’HI-FdUMP (0.0; panel A), FUra (a,~) and total radioactivity 6Il.W) (panel 8) are shown. Data are means f SEM for 3 experiments. 1235

more

Vol.

182, No. 3, 1992

TABLE

BIOCHEMICAL

AND BIOPHYSICAL

of IFNa on the Pyrimidine

2. Effect

Substrate

Activitv’ Uridine

Pyrimidine phospiorylase Thymidine kinase Thymidine

salvage’

Biosynthetic

Control

FUra/R-1 -P FUraldR-1 -P dThd FdUrd dThd

ohosohorvlase

RESEARCH COMMUNICATIONS

Pathway

IFNu-treated 0.60 5.68’

0.31 0.65

1.09

0.90

2.03 8.53

1.69 3.70’

% Control 194% 874% 82% 83% 41%

Cells were treated with IFNa (500 U/ml) for 24 hours. Cells extracts were used for measurement of enzyme activities, and intact cells for thymidine salvage. ’ Enzyme activities are expressed as pmollpg proteinlhr. Data are from 3 experiments. ’ Thymidine salvage is expressed as the ratio of dpmlcell for 3H-dThd incorporation divided by the dpmlcell for 3HdAdo incorporation into DNA. The average decrease in incorporation with IFNatreatment was significantly different from control by 68% for 3H-dThd and 21% for 3H-dAdo. * Indicates paired t-test, p < 0.05.

striking

when

corrected

Fig.

which

presumably

2),

Metabolic

was

activation

1), and we sought substrate,

for the overall decline

there

to discern

cells compared

to extracts

have an increased

thymidine

kinase.

found,

on thymidine

incorporation

activation

data were corrected treatment inhibition carcinoma

in the direct

that IFNa-treated

conversion

cellular

the thymidine

acid-insoluble

FUMP,

material.

We next

of FUra to FdUMP,

FUDP or FUTP levels,

in the incorporation

pathway

effect

others

We determined

obtained

reflected

of IFNa effects on DNA

salvage

by 59%

(data not shown). 1236

and

into these nucleic

kinase activity,

in HT-29

dThd

synthesis

incorporation

for at least 72 hours and was also observed and Hep-2/500

on

of FUra into RNA.

of FUra concentrations

(32-33).

of 3H-deoxyadenosine

thymidine

its effect

have if this

the effect of IFNa on the incorporation of

The values

but were independent

for 24 hours reduced

cell lines, SW480

salvage

cells by examining

for the incorporation

by IFNu persisted

cells

there was no significant

we saw no effect of IFNa on thymidine

and degradation,

of uridine

from IFNa-treated

that IFNa had no effect on the extent of FUra incorporation

into cellular

Using FUra as

kinase activity.

was also the case for the HT-29 3H-dThd

(Fig.

FUra to ribo- and deoxyribonucleosides. enzyme

into RNA and DNA over a range

that IFN can inhibit

pathways

in the activities

(PyNP) in extracts

2). This suggests

could have led to an increase

acids (Fig. 3). Although reported

cells (Table

to convert

of the second

respectively,

phosphorylase

IFNa did not increase

phosphorylase

however,

nucleoside

1 and

for the IFNa effect.

Using both FdUrd and dThd as substrates,

of IFNa treatment

We measured

might be responsible

from control

capacity

the activity

uridine

which

(Table

on FUra uptake.

to an effect

was a 1.9 and 8.7 fold increase, and pyrimidine

measured

FURa-metabolites

of FUra to FdUMP can occur by three interrelated

phosphorylase

Although

due

in total cellular

uptake, since

into DNA.

cells

(Table

2).

the

lFNa The

in two other colon

Vol.

182, No. 3, 1992

BIOCHEMICAL

AND BIOPHYSICAL

RESEARCH COMMUNICATIONS

0.11 1

10 FUra

Cont.

50

(JAI)

Fia. 3, Effect of IFNa on the incorporation of FUra into RNA and DNA. Cells were treated with (closed symbols) or without (open symbols) 500 U/ml IFNo and the indicated concentrations of [6-3Hl-FUra (2 ,&i/ml) for 24 hours. Radioactivity incorporation into RNA and DNA

(A.4)

(0.0)

was

determined

as described

in Methods.

DISCUSSION IFNa synergistically carcinoma

increases

cells in vitro (4,5,7),

may be a direct

antitumor activity

Previous studies

have indicated

cells without extended further

demonstrated

that IFN treatment

that the action

increased

the endogenous relevant

of IFNa is selective

showing

an increase

human

not increase

in activity

of FUra catabolic

to have

concentrations would

the cellular

produced

be expected

of FdUMP

of the FUra. in leukemia (34).

colon carcinoma with

with IFNa treatment.

enzymes,

dR-I-P,

including

consistent

We have cells, and

none of the other This suggests

effect

on FdUMP.

therefore,

to be the mechanism

event is likely responsible

Similarly,

and phosphatases,

In the absence

the IFNu-mediated

for the elevation

of limiting

increase

in cellular

is significant

for the synergistic

that

with our observation

5’-nucleotidases

The effect of IFNu on FdUMP levels and PyNP activity 1) an FdUMP-mediated

IFNa alone has

levels of either FUra or total FUra-metabolites.

the selective

of the co-substrate

in which

for FdUMP,

colon

of the combination

the cytotoxicity

antagonist

model,

and other

FdUMP accumulation

or efflux of FUra is not the basis for the IFNa action,

are unlikely

activity

under conditions

that the IFN likely enhances

to a clinically

of FUra measured

that IFNadid changes

suggests

of FUra in HT-29

that part of the clinical efficacy

That this occurs

levels of dUMP,

this observation

metabolites uptake

affecting

suggesting

effect.

no antiproliferative

the cytotoxicity

FdUMP

in PyNP levels.

for several reasons:

cytotoxic

action

of IFNu

in this cell line, since IFNa did not alter the incorporation

of FUra into RNA or DNA; 2) cellular

FdUMP levels are correlated

of TS, assuming

concentration become

limiting

of reduced

with the extent of inhibition folates

(35,361;

due to an IFN-mediated

3) under those conditions increase 1237

in FdUMP,

there is an adequate

in which

addition

reduced

of leucovorin

folates should

Vol.

BIOCHEMICAL

182, No. 3, 1992

further

increase

been demonstrated

the activity

of the FUra/lFN

anabolism;

5) the stimulation

interferons

can increase to mediate

does not appear when

compared

mechanism

combination,

an action that would

of PyNP activity

the expression

appears

of several

IFN’s antiproliferative

to fit in this category, to non-neoplastic

of this IFNo action

would

(40).

(7); 4) it has to an increase

the effect of IFNa on FUra

to be a novel action of IFNo. Although

genes and proteins,

since its activity

therefore

of FURa leading

complement

or antiviral

tissues

RESEARCH COMMUNICATIONS

as has been reported

that IFNo can alter the pharmacokinetics

in plasma levels of FUra (37-39).

been thought

AND BIOPHYSICAL

actions

these increases

(22-23).

PyNP activation

has been found to be higher Studies

to understand

have both biologic

have

in tumor

the molecular

and therapeutic

impact.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

Evans, R.M., Laskin, J.D. and Hakala, M.T. (1981) Cancer Res. 41:3288-3295. Grem, J.L., King, S.A. and O’Dwyer, P.J., Leyland-Jones, B. (1988) Cancer Res. 48~4441-4454. Miyoshi, T., Ogawa, S., Kanamori, T., Nobuhara, M. and Namba, M. (1983) Cancer Lett. 17:239-247. Elias, L. and Crissman, H.A. (1988) Cancer Res. 48:4868-4873. Wadler, S., Wersto, R., Weinberg, V., Thompson, D. and Schwartz, E.L. (1990) Cancer Res. 50:5735-5739. Chu, E., Zinn, S., Boarman, D. and Allegra, C.J. (1990) Cancer Res. 50:5834-5840. Houghton, J.A., Adkins, D.A., Rahman, A. and Houghton, P.J. (1991) Cancer Commun. 3:225-231. Wadler, S., Schwartz, E.L., Goldman, M., Lyver, A., Rader, M., Zimmerman, M., Itri, L., Weinberg, V. and Wiernik, P.H. (1989) J. Clin. Oncol. 7:1769-1775. Kemeny, N., Younes, A., Seiter, K., Kelsen, D., Sammarco, P., Adams, L., Murray, P. and Houston, C. (1990) Cancer 66:2470-2475. Pazdur, R., Ajani, J.A., Patt, Y.Z., Winn, R., Jackson, D., Shepard, B., DuBrow, R., Campos, L., Quaraishi, M., Faintuch, J., Abbruzzese, J.L., Gutterman, J. and Levin, 8. (1990) J. Clin. Oncol. 8:2027-2031. Fornaseri, A., Daniele, O., Ghiotti, C., Aversa, S.M.L., Morandi, P. and Fiorentino, M.V. (1990) Tumori 76:385-388. Wadler, S., Lembersky, B., Atkins, M., Kirkwood, J. and Petrelli, N. (1991) J. Clin. Oncol. 9:1806-1810. Santi, D.V. and McHenry, C.S. (1972) Proc. Natl. Acad. Sci. (USA) 69:1855-1857. Danenberg, P.V., Langenbach, R.J. and Heidelberger, C. (1974) Biochem. 13:926930. Houghton, J.A., Houghton, P.J. and Wooton, R.S. (1979) Cancer Res. 39:2406-2413. Spiegelman, S., Nayak,R., Sawyer, R., Stolfi, R. and Martin, D. (1980) Cancer Res. 45:1129-1134. Kufe, D.W. and Major, P.P. (1981) J. Biol. Chem. 256:9802-9805. Glazer, R.I. and Lloyd L.S. (1982) Molec. Pharm. 21:468-473. Dolnick, B.J. and Pink, J.J. (1983) J. Biol. Chem. 258:13299-13306. Lonn. U. and Lonn, S. (1984) Cancer Res. 44:3414-3418. Shuetz, J.D. and Biasio, R.B. (1985) Bioch. Biophys. Res. Comm. 133:361-367. Lengyel, P. (1982) Ann. Rev. Biochem. 51:251-282. Pestka, S., Langer, J.A., Zoon, K.C. and Samuel, C.E. (1987) Ann. Rev. Biochem. 58:727-777. Wadler, S. and Schwartz, E.L. (1990) Cancer Res. 50:3473-3486. Fogh, J. and Trempe, G. New human tumor cell lines. In: J Fogh, ed. Human Tumor Cells In Vitro, Plenum Press, New York, 1975, pp 1 15-I 59. Khym, J.X. (1975) Clin. Chem. 21:1245-l 252. Grem, J.L. and Fischer, P.H. (1986) Cancer Res. 46:6191-6199. 1238

Vol. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39.

40.

182, No. 3, 1992

BIOCHEMICAL

AND

BIOPHYSICAL

RESEARCH

COMMUNICATIONS

Garrett, C. and Santi, D.V. (1979) Anal. Biochem. 99:268-273. Poglotti, A.L., Nolan, P.A. and Santi, D.V. (1981) Anal. Biochem. 117:178-186. Peters, G.J., Laurensse, E., Leyva, A., Lankelma, J. and Pinedo, H.M. (I 986) Cancer Res. 46:20-28. Sobrero, A.F., Moir, R.D., Bertino, J.R. and Handschumacher, R.E. (1985) Cancer Res. 45:3155-3160. Brouty-Boye, D. and Tovey, M.G. (1978) Intervirol. 9:243-252. Divizia, M. and Baglioni, C. (1984) Virol. 133:216-219. Elias, L. and Sandoval, J.M. (1989) Biochem. Biophys. Res. Comm. 163, 867-874. Heidelberger, C., Danenberg, P.V. and Moran, R.G. (1982) Adv. Enzymol. 54:57-l 19. Houghton, J.A., Maroda, S.J., Phillips, J.O. and Houghton, P.J. (I 981) Cancer Res. 41:144-149. Schuller, J., Czejka, M., Miksche, M., Fogl, U. and Schernthaner, G. (1991) Proc. Am. Sot. Clin. Oncol. 10:98. Lindley, C., Bernard, S., Gavigan, M. et. al. (1990) J. Interferon Res. lO:Sl32. Grem, J.L., McAtee, N., Murphy, R.F., Balis, F.M., Steinberg, S.M., Hamilton, J.M., Sorenson, J.M., Sartor, O., Kramer, B.S., Goldstein, L.J., Gay, L.M., Caubo, K.M., Goldspiel, B. and Allegra, C.J. (1991) J. Clin. Oncol. 9:181 I-1820. Yoshimura, A., Kuwazuru, Y., Furukawa, T., Yoshida, H., Yamada, K. and Akiyama, S. (I 990) Biochim. Biophys. Acta 1034: 107-I 13.

1239