Increased Activity of a Neutral Protease in Cytosol from Rat

Hepatoma Induced by N-2-Fluorenylacetamide

Kenji Wada, Hidenori Matsui and Kinji Tsukada
Cancer Res 1981;41:5130-5133. Published online December 1, 1981.

Updated Version

E-mail alerts
Reprints and
Subscriptions
Permissions

Access the most recent version of this article at:

http://cancerres.aacrjournals.org/content/41/12_Part_1/5130

Sign up to receive free email-alerts related to this article or journal.

To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at pubs@aacr.org.
To request permission to re-use all or part of this article, contact the AACR Publications
Department at permissions@aacr.org.

Downloaded from cancerres.aacrjournals.org on January 7, 2011

Copyright 1981 American Association for Cancer Research

[CANCER RESEARCH 41, 5130-5133,

0008-5472/81
/0041-OOOOS02.00

December 1981]

Increased Activity of a Neutral Protease in Cytosol from Rat Hepatoma

Induced by W-2-Fluorenylacetamide1
Kenji Wada, Hidenori Matsui, and Kinji Tsukada2
Department of Pathological Biochemistry,

Medical Research Institute. Tokyo Medical and Dental University, Kandasurugadai,

ABSTRACT

MATERIALS

A protease active with A/-a-benzoyl-DL-arginine-p-nitroanilide

with an optimum pH of 7.3 has been found in the cytosol of rat
liver. The activity of this protease increased in /V-2-fluorenylacetamide-induced
hepatoma as well as in fetal liver. It has
been purified from normal liver and hepatoma about 200-fold.
Its molecular weight is estimated by gel filtration to be about
200,000 in each tissue. The protease activity is unaffected by
chymostatin, pepstatin, soybean trypsin inhibitor, and p-chloromercuribenzoate.
Antipain, leupeptin, tosyl-L-lysine chloromethyl ketone, and phenylmethylsulfonyl
fluoride inhibit the
protease activity. This protease appears to be a serine pro
tease.

Materials

INTRODUCTION
The intracellular breakdown of proteins probably involves
the participation of proteases and peptidases, although its
degradation mechanism and regulation still remain to be clari
fied. Many intracellular proteases have been isolated from
various tissues and characterized (3). In mammalian tissues,
acid proteases are found in high concentration in lysosomes
(3, 4). The rate of degradation of cytosol proteins was found to
be less than 10% that of whole homogenate (5, 6), but it seems
likely that nonlysosomal protease may play an important role in
the degradation of certain cell proteins. Several proteases are
known to exist in liver; however, few have been purified and
characterized. Proteases with their major activity at neutral pH
have been demonstrated in nuclear (6, 10) and lysosomal (8,
18) fractions. Recently, several proteases, including high-mo
lecular-weight neutral proteases (7, 17) and Ca2*-dependent
neutral protease (19), are known to exist in the cytosol in
mammalian cells. In the course of studies on the distribution
and properties of a neutral, trypsin-like protease isolated to
homogeneity from rat intestine (22), the considerable rate of
BAPA3 hydrolysis at pH 7.3 was found in the high-speed
supernatant fraction from a rat hepatoma induced by a hepatocarcinogen. This communication describes that protease ac
tivity from rat hepatoma induced by 2-FAA increased in the
cytosol fraction and also describes a method for the partial
purification of this protease as well as some properties of this
protease.

Chiyoda-ku,

Tokyo 101, Japan

AND METHODS

BAPA, 2-FAA, TLCK, PCMB, PMSF and TPCK were obtained from
Sigma Chemical Co., St. Louis, Mo. Protease inhibitors from cultured
broth of Actinomycetes were supplied by Research Resources, Ministry
of Education, Science and Culture, Japan. DEAE-cellulose and Sephadex G-150 (superfine) were products of Pharmacia, Uppsala, Sweden.
Soybean trypsin inhibitor was purchased from Boehringer, Mannheim,
Mannheim, Federal Republic of Germany. Azocoll, azocasein, and
elastin-orcein were from Calbiochem-Behring Corp., La Jolla, Calif. All
other reagents were of analytical grade.
Animals
The rats used were an albino Wistar strain. The gestational age of
fetuses was calculated from the mating date, known within 12 hr. The
gestational period for this strain of rats was 22 days. Partial hepatectomy refers to the removal of about 70% of the liver (left lateral and
median lobes) (11).
Treatment of Animals
Male Wistar rats (Fuji Animal Farm, Tokyo, Japan) weighing 150 to
200 g were used. The basal diet (Oriental Yeast Co., Tokyo, Japan)
was the same as described previously (15). Animals were fed for 3
months with the basal diet containing 0.025% 2-FAA followed by a
basal diet after this treatment. The rats were sacrificed about 8 months
after the start of the experiment and about 5 months from the end of
the 2-FAA diet. The liver of each rat was examined macroscopically
and divided into nonhepatomatous and hepatoma areas. Tissue sam
ples were taken from an area adjacent to tissues used for biochemical
studies. For histolgica! studies, tissues were fixed in 10% neutral
buffered formaldehyde solution and stained with hematoxylin and
eosin. Histologically, hepatoma areas showed findings of typical hepatoceilular carcinoma. In this experiment, hepatoma areas were used
as hepatoma for the enzyme assay.
Determination

of Protease Activity

Activity in BAPA was measured at 410 nm in 0.05 M potassium

phosphate buffer (pH 7.3) as described by Erlanger ef al. (9). The
assay mixture consisted of 0.45 ml of enzyme, 0.5 ml of 2 mw BAPA,
and 0.05 ml of 1 M potassium phosphate buffer (pH 7.3). After incu
bation at 37for 1 to 10 hr, the reaction was terminated by addition of
0.1 ml of 10% sodium lauryl sulfate and 0.4 ml of 0.5 M Tris-HCI (pH
10.0). Blank determinations were carried out in the same way except
that distilled water replaced enzyme solution. A major absorption
difference of 8000 M : cm ' was used for all calculations. A unit of
enzyme activity is defined as .molof product liberated per hr.

' Supported in part by a Grant-in-Aid

for Cancer Research from the Ministry

of Education. Science and Culture. Japan.

* To whom requests for reprints should be addressed.
3 The abbreviations used are: BAPA. N-n-benzoyl-DL-arginme-p-nitroanilide;
2-FAA, W-2-fluorenylacetamide;
TLCK. tosyl-t-lysine
chloromethyl
ketone;
PMSF, phenylmethylsulfonyl
fluoride; PCMB. p-chloromercuribenzoate;
TPCK.
tosyl-L-phenylalanine chloromethyl ketone.
Received March 17. 1981; accepted September 10, 1981.

5130

Protein Determination
Protein was determined by the method of Lowry ef al. (14) with
bovine serum albumin taken as the standard.
Polyacrylamide Gel Electrophoresis
Disc electrophoresis

was performed at pH 4.5 and 7.5% polyacryl-

CANCER

Downloaded from cancerres.aacrjournals.org on January 7, 2011

Copyright 1981 American Association for Cancer Research

RESEARCH

VOL. 41

Increased Activity of a Protease in Hepatoma

amide as described by Reisfeld et al. (16). The gels were stained in
0.5% Coomassie Brilliant Blue and then destained.

induced by 2-FAA. At pH 7.3, the activity

of the protease

increased at about 5 months, when the tissue is shown to be

hyperplastic and premalignant, after onset of feeding 2-FAA,
Partial Purificationof Protease
and markedly increased 8 months after treatment. A neutral
Step 1. Preparationof Crude Extracts. To purifya neutralprotease protease activity in regenerating rat liver 24 and 48 hr after
from rat hepatoma, 30 g of the tissues were cut up with scissors and operation increased approximately 2-fold compared with that
homogenized in 4 volumes of 0.25 M sucrose. This homogenate was from normal liver (Table 1). On the 19th day of gestation,
centrifugea at 10,000 x g, and the supernatant fraction was subjected markedly increased protease activity was observed in fetal rat
to centrifugation at 105,000 x g for 90 min to obtain the cytosol liver just as it was in hepatoma (Table 1). The effects of
fraction.
leupeptin, chymostatin, TLCK, TPCK, and PMSF on each crude
Step 2. pH 5.0 Supernatant. The cytosolfractionwas adjustedto
enzyme preparation in Table 1 were shown to be almost same
pH 5.0 by careful addition of 1 N acetic acid with vigorous stirring, and
as those of partially purified protease as described in Table 3.
the supernatant fraction was obtained after centrifugation. The pH 5.0
The purification results from hepatoma tissues are summa
supernatant fraction was adjusted to pH 7.0 with 1 N NH4OH.
rized in Table 2. The enzyme was purified about 200-fold with
Step 3. Calcium PhosphateGel Treatment. To thissolution(about
a yield of 12%, but it was not homogeneous, judging from the
850 mg protein) were added 100 ml of calcium phosphate gel (20 mg/
results that the activity and absorbance at 280 nm did not
ml), and the mixture was stirred for 15 min. After stirring, the suspen
coincide in the fractions from Sephadex G-150 (data not
sion was centrifugea to obtain the supernatant fraction.
Step 4. AmmoniumSulfate Fractionation.To the supernatantfrac
shown). From analytical disc electrophoresis of the partially
tion, solid ammoniumsulfate was added to 80% saturation. After being purified protease on polyacrylamide gel, several bands were
stirred for 20 min, the suspension was centrifugea for 20 min at 12,000 detected by protein staining, one of which was also active on
x g. The precipitate was dissolved in 0.01 M potassium phosphate
buffer (pH 7.0)-0.25 M sucrose-1 mw 2-mercaptoethanol (Buffer A) in BAPA (data not shown). The protease was also purified about
200-fold from normal rat livers with the yield of about 10%.
order to obtain a final concentration of approximately 40 mg of protein
With
the partially purified protease, the rate of degradation of
per ml and was dialyzed against Buffer A.
Step 5. DEAE-CelluloseChromatography. The dialyzed solution the substrate proceeded linearly for more than 4 hr and was
was applied to a column of DEAE-cellulose (1.8 x 8.5 cm) equilibrated
with Buffer A. The chromatogram was developed with a linear gradient
from 0 to 0.5 M KCI-containing Buffer A, and fractions of 2.5 ml were
collected. The fractions were assayed for protein by measuring absorbance at 280 nm and for enzyme activity (Chart 1). The fractions
(Tubes 24 to 28) containing the protease activity, eluted at 0.06 M KCI
as a sharp peak, were combined and brought to 80% ammonium
sulfate saturation. After centrifugation, the precipitate was dissolved in
0.02 M potassium phosphate (pH 7.5M3.25 M sucrose-1 mw 2-mer
captoethanol (Buffer B) and dialyzed against Buffer B.
Step 6. Sephadex G-150 Chromatography.The dialyzedsolution
was applied on a Sephadex G-150 (1.5 x 54 cm) equilibrated with
Buffer B. The elution was performed with the same buffer at a flow rate
of 15 ml/hr, and fractions of 2.55 ml were collected. The fractions
(Tubes 17 to 19) containing the protease activity were collected.
RESULTS AND DISCUSSION
The protease activity from cytosol fraction of normal liver
and hepatoma was assayed at various pH values. Chart 2
shows the protease activity in soluble fractions from hepatoma

Chart 2. Protease activity of normal rat liver and hepatoma at various pH

values. The cytosol fractions (105,000 x g supernatant after centrifugation for
90 min) from normal rat livers and 2-FAA-induced hepatoma were adjusted to pH
5.0 by careful addition of 1 N acetic acid, and the supernatant after centrifugation
was adjusted to pH 7.0 with 1 N NH.,OH This solution (about 0.5 mg protein)
from each tissue was used as enzyme preparation and assayed for 10 hr in 0.05
M concentrations of various buffers at the various pH values indicated: pH 5.2,
acetate buffer; pH 6.3 and 7.3, potassium phosphate buffer; pH 8.0 and 8.5,
Tns-HCl buffer. ,normal liver; O and O, 5 and 8 months after onset of feeding
2-FAA, respectively. Pooled hepatoma areas from 4 to 5 rats were used for assay
of protease activity, mil, milliunits.
Table 1
Comparison of protease activity from normal, fetal, and neoplastia rat livers
The cytosol fractions from normal, regenerating, and fetal livers and hepatoma
(8 months after treatment) were treated by the procedures described in Chart 2;
assay conditions were also described in Chart 2. Each value is the mean of 3 to
6 determinations.
TissuesNormal

IO

20
30
40
Fraction number

50

Chart 1. DEAE-cellulose column Chromatography of the protease from hepa

toma. The dialyzed solution (141 mg protein) from the ammonium sulfate fractionation was applied to a column of DEAE-cellulose equilibrated with Buffer A
and was eluted as described in the text. The salt concentration was determined
by conductance measurements. Fractions were analyzed for protein at 280 nm
(),protease activity (O), and KCI concentration (x).

DECEMBER

liver
Regenerating liver
6 hr after operation
24 hr after operation
48 hr after operation
Fetal liver, 3 days before birth
0.58" HepatomaProtease
Mean S.D.

1981

activity (milliunits/mg
protein)0.56

0.07a0.63

1.19
1.20
2.92
8.21

0.05
0.09
0.10
0.25

5131

Downloaded from cancerres.aacrjournals.org on January 7, 2011

Copyright 1981 American Association for Cancer Research

K. Wada et al.
proportional to the amount of the protease added to the assays.
The molecular weight of the protease from the cytosol of
hepatoma and liver was estimated by gel filtration on Sephadex
G-1 50 column as approximately 200,000 in each tissue, by the
method of Andrews (1 ), as shown in Chart 3. The pH depend
ency of the activity of partially purified protease was similar to
that of crude preparations as shown in Chart 2. Some reagents
for general protein substrates, e.g., azocoll, azocasein, and
elastin-orcein, were not hydrolyzed by the protease.
The thermal stability of the partially purified protease from
hepatoma was determined by assaying the activity remaining
after incubation for 30 min at various temperatures between
30and 70in 0.05 M buffer, pH 5.0 or 7.0, indicated in Chart
4A. The results indicated that the activity of the protease was
stabilized at pH 5.0 compared to pH 7.0, although the activity
was almost completely destroyed at 70in both pH. When the
protease was heated at 50 in 0.05 M potassium phosphate
buffer (pH 7.0) with the indicated concentrations of ammonium
sulfate, the activity of the protease was prevented from destroy
ing by the addition of 0.2 M ammonium sulfate as shown in
Chart 4B. Almost the same results for thermal stability were
obtained from the partially purified protease from normal rat
liver.
The effects of protease inhibitors are presented in Table 3.
Among the protease inhibitors obtained from cultured broths
of Actinomycetes, the protease was inhibited only by antipain
and by also leupeptin, which is a strong competitive inhibitor of
proteolysis by plasmin, trypsin, or papain (2). The protease
was also inhibited by a specific inhibitor of trypsin, TLCK, and
could be identified as serine protease as it was inhibited by
Table 2
Purification

of protease from rat hepatoma 8 months after onset of feeding 2FAA

IO
20
3C>
Time ( min )

40 50 60
Temperature(*C)

Chart 4. A. thermal stability of the most purified protease from hepatoma. The
enzyme (10 fig protein) in 0.2 ml of 0.05 M potassium phosphate buffer, pH 7.0
(O), or 0.05 M acetate buffer, pH 5.0 (),was heated for 30 min at the indicated
temperature and cooled in ice after the addition of 0.25 M distilled water. Assay
conditions are described in "Materials and Methods." B, stabilizing effect of
ammonium sulfate on the protease from hepatoma. The enzyme (25 fig protein)
in 1.5 ml of 0.05 M potassium phosphate buffer, pH 7.0, was heated at 50
without or with ammonium sulfate at the indicated concentrations ().An aliquot
of 0.2 ml was taken at the indicated time; for protease activity determinations,
see "Materials and Methods."

Table 3
Effects of several protease inhibitors and chemical reagents on the activity of
protease from normal rat liver and hepatoma
The most purified enzyme preparations

in Table 2 were used.

Activity remaining

Inhibitors
reagentsLeupeptinAntipainChymostatmPepstatinSoybean
and chemical

liver11.214.496.899.696.296.098.240.8
toma9.810.397.594.797

fig/ml10
fig/ml10
fig/ml10
fig/ml10
inhibitorW-EthylmaleimidePCMBTLCKTPCKPhenylmethylsulfonyl
trypsin
fig/ml1
mM0.25
mM0.25
mM0.25
mM1
mM1
fluorideEDTAConcentration10
mMNormal

tion(-fold)11.63.95.126.8191.5
PMSF. The protease

was not inhibited by soybean trypsin

inhibitor, PCMB, or TPCK. Apparently, no different properties
1.2.3.4.5.6.StepCrude
extractspH
were observed between the normal and neoplastic hepatic
supernatantCalcium
5.0
enzymes, except that the specific activity of the latter showed
gel(NH4)2SO,DEAE-celluloseSephadex
phosphate
a remarkably higher increase than that of the former.
Leupeptin markedly inhibited tumorigenesis in mouse skin
G-1 50Protein(mg)1086.4512.7149.5141.410.60.7Activity(units)9.616.995.096.362.511.18Specificactivity(units/mgprotein)0.00880.0140.0340.0450.2361.685Purifica
induced by a single, noncarcinogenetic
dose of 7,12-dimethylbenzanthracene followed by repeated application of cro
30ton oil, and it also inhibited the activity of p-toluenesulfonyl-Larginine methyl ester esterase in the skin of animals treated
with croton oil (12). It was reported that synthetic inhibitors of
proteases markedly depressed tumorigenesis in mouse skin
induced by painting the skin with 7,12-dimethylbenzanthracene and croton oil (13, 21). Therefore, they suggested that
certain proteases might be involved in tumorigenesis.
\2 1.4 1.6 I.8 2.0 22
In contrast to other intracellular proteases in liver, this en
V*/Vo
zyme is shown to be located in cytosol. On the basis of
Chart 3. Estimation of the molecular weight of the protease from normal rat
evidence presented in this study, this protease is shown to be
liver and hepatoma by gel filtration of Sephadex G-150. The column (1.5 x 54
distinct from lysosomal proteases as to optimum pH and the
cm; V. 32.5 ml) was equilibrated with Buffer B-0.5 M NaCI. The resulting
concentration of enzyme solution (6 mg protein of each) from DEAE-cellulose
effect of several protease inhibitors. The protease, cathepsin
was dialyzed against Buffer B-0.5 M NaCI and applied to the column. Elution was
B active with BAPA in lysosomal fraction from rat liver, was
performed with Buffer B-0.5 M NaCI at a flow rate of 15 ml/hr, and fractions of
crystallized by Towatari et al. (20). The activity of this protease
1.45 ml were collected. The column was calibrated with molecular weight
markers. The following marker proteins were used: Cat, catalase (M w 244,000);
was maximal at pH 6.0 and decreased markedly above pH 7.O.
LDH. lclate dehydrogenase (M.W. 140,000); BSA. bovine serum albumin (M.W.
It is a thiol protease with a molecular weight of 26,000 by gel
68,000); Ova, ovalbumin (M.W. 45.000). O, protease of normal liver and hepa
filtration (20). The protease described here is distinct from
toma at eighth month after treatment. Ve, elution volume; Vo, void volume.
5132

CANCER

Downloaded from cancerres.aacrjournals.org on January 7, 2011

Copyright 1981 American Association for Cancer Research

RESEARCH

VOL. 41

Increased Activity of a Protease in Hepatoma

other proteolytic enzymes in liver cytosol that have been re
ported recently (7, 17). The protease reported by Rose ef al.
(17) is a high-molecular-weight
(M.W. >400,000)
protease
active with labeled globin at pH 7.5 in soluble fractions from
mouse tissues and is most active in liver. This protease is
inhibited by PCMB and is not inhibited by TLCK and TPCK
(17). The enzyme reported by DeMartino and Goldberg (7) is
alkaline endoprotease with an apparent molecular weight of
550,000 the activity of which is stimulated by ATP and other
nucleotides as well as by PPi.
We have shown here that the neutral protease activity in
creased in hepatoma as well as in fetal liver, but the protease
may be characteristic of the cancer cells. These proteases may
be related to growth, highest in hepatoma (8.21) and next in
fetal liver (2.92), regenerating liver (1.20) compared to normal
liver (0.56). The identification of this enzyme may lead to an
indicator enzyme for liver carcinogenesis.

3712-3715, 1979.

8. Destree, O. H. J., D'Adehart-Toorop,

9.

10.
11.

12.

13.

14.

15.

16.

REFERENCES
1. Andrews, P. Estimation of the molecular weights of proteins by Sephadex
gel-filtration. Biochem. J.. 97. 222-232. 1964.
2. Aoyagi. T., Miyata. S , Nanba. M , Kojima, F.. Matsuzaki, M . Ishizuka, M.,
Maeda, K., and Umezawa, H. Biological activities of leupeptins. J. Antibiot.
(Tokyo) Ser. A. 22. 558-568, 1969.
3. Barrett, A. J. Proteases in Mammalian Cells and Tissues. Amsterdam:
Elsevier/North Holland BiomdicalPress, Inc., 1977.
4. Barrett, A. J., and Dingle, J. T. Tissue Proteases. Amsterdam: Elsevier/
North Holland Biomdical Press, 1977.
5. Brostrom, C. O.. and Jeffay, H. Protein catabolism in rat liver homogenates:
a pre-evaluation of the energy requirement for protein catabolism. J. Biol.
Chem., 245: 4001-4008,
1970.
6. Brostrom, C. O., and Jeffay, H. Protein catabolism in rat liver nuclei. Biochim.
Biophys. Acta, 278. 15-27, 1972.
7. DeMartino, Q. N., and Goldberg, A. L Identification and partial purification
of an ATP-stimulated alkaline protease in rat liver. J. Biol. Chem., 254:

DECEMBER

17.
18.

19.

20.
21.

22.

H. A., and Charles, R. Cytoplasmic

origin of the so-called nuclear neutral histone protease. Biochim. Biophys.

Acta. 378. 450-458, 1975.
Erlanger, B. F., Kokowsky, N., and Cohen, W. The preparation and proper
ties of two new chromogenic substrates of trypsin. Arch. Biochem. Biophys.,
95. 271-278, 1961.
Furlan, M . and Jericijo. M. Protein catabolism in thymus nuclei. Biochim.
Biophys. Acta. 147: 135-153, 1967.
Higgins, G. M., and Anderson, R. M. Experimental pathology of the liver I.
Restoration of the liver of the white rat following partial surgical removal.
Arch. Pathol., 12: 186-202, 1931.
Hozumi. M., Ogawa, M., Sugimura, T., Takeuchi, T., and Umezawa, H.
Inhibition of tumorigenesis in mouse skin by leupeptin, a protease inhibitor
from Actinomycetes. Cancer Res., 32. 1725-1728.
1972.
Klassen, A., Janoff. A., and Troll. W. Possible role of a neutral protease in
tumor-promotion by phorbol ester. Proc. Am. Assoc Cancer Res., 11: 44,
1970.
Lowry, O. 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.
Marugami, M., Ito, N., Konishi. Y., Hiasa, Y., and Farber, E. Influence of 3methylcholanthrene on liver carcinogenesis in rats ingesting DL-ethionine,
3'-methyl-4-dimethylaminoazobenzene
and N-2-fluorenylacetamide.
Cancer
Res., 27. 2011-2019.
1967.
Reisfeld, R. A., Lewis, U. J., and Williams, D. E. Disk electrophoresis of
basic proteins and peptides on polyacrylamide gels. Nature (Lond.), 195.
281-283. 1962.
Rose, I. A., Warms, J. V. B . and Hershko, A. A high molecular weight
protease in liver cytosol. J. Biol. Chem., 254. 8135-8138,
1978.
Stellwagen, R. H.. Reid, B. R.. and Cole, R. D. Degradation of histones
during the manipulation of isolated nuclei and deoxyribonucleoprotein.
Biochim. Biophys. Acta. 755. 581-592, 1968.
Takai, Y., Yamamoto, M., Inoue, M., Kishimoto, A., and Nishizuka. Y. A
preenzyme of cyclic nucleotide-independent
protein kinase and its activation
by calcium-dependent
neutral protease from rat liver. Biochem. Biophys.
Res. Commun., 77. 542-550. 1977.
Towatari, T., Kawabata, Y., and Katunuma, N. Crystallization and properties
of cathepsin B from rat liver. Eur. J. Biochem., 702. 279-289, 1979.
Troll, W., Klassen, A., and Janoff, A. Tumorigenesis in mouse skin: inhibition
by synthetic inhibitors of proteases. Science (Wash. D.C.), 769: 12111213, 1971.
Wada, K., Sawai, Y.. and Tsukada, K. A protease from rat intestine. Biochim.
Biophys. Acta, 672. 253-261, 1980.

1981

5133

Downloaded from cancerres.aacrjournals.org on January 7, 2011

Copyright 1981 American Association for Cancer Research