[CANCER RESEARCH 43. 5222-5227.
November 1983]
Molecular and Biological Characterization of Anticarcinogenic
and Tumor Cell Growth-inhibitory Activities
of Syrian Hamster Lymphotoxin
Janet H. Ransom1 and Charles H. Evans
Tumor Biology Section, Laboratory of Biology. Division ot Cancer Cause and Prevention, National Cancer Institute, Bethesda, Maryland 20205
ABSTRACT
Syrian hamster lymphotoxin has three identified anticancer
activities: (a) cytolytic activity measured as the release of 3H from
[3H]thymidine-labeled murine «L-929cells; (b) cytostatic activity
assessed by the inhibition of cell colony formation of tumorigenic
Syrian hamster cells; and (c) anticarcinogenic activity measured
by inhibition of chemical carcinogen or radiation-induced mor
phological transformation of Syrian hamster fetal cells in vitro
and in vivo. Lymphotoxin cytolytic, cytostatic, and anticarcino
genic activities have similar molecular weights of approximately
45,000 by ACA-44 Ultrogel chromatography. A three-step puri
fication of (a) diafiltration and concentration, (b) column isoelectric
focusing in a pH 4 to 6 gradient, and (c) ACA-44 Ultrogel column
chromatography yields 25% recovery of lymphotoxin cytolytic
and anticarcinogenic activities. The three-step purified lympho
toxin is free of detectable Interferon, macrophage migration
inhibitory factor, and mitogenic factors inducing hamster T-lymphocyte proliferation and is anticarcinogenic in vivo. The purified
lymphotoxin had a specific anticarcinogenic activity of 58,500
units/mg of protein and a molar specific activity of 2,900 units/
nmol. Although homogenous in molecular size, the purified lym
photoxin is comprised of several molecular species. Cytolytic
activity is associated with molecules having a single major isoe
lectric point of 5.0, whereas the cytostatic and anticarcinogenic
activities comigrate with molecules with major isoelectric points
of 5.0 and 4.6. This is the first evidence that the anticancer
actions of lymphotoxin are properties of similarly sized but
several different electrically charged molecules and that the
several lymphotoxin species have differing anticancer activities.
INTRODUCTION
Lymphotoxin, an immunological hormone, has been credited
with antitumor cell and anticarcinogenic activities (5, 7, 10, 15,
25-27, 31). The antitumor cell activity may be either cytolytic or
cytostatic, the latter being demonstrable by reversible inhibition
of cell division in the absence of cell lysis (9). Lymphotoxin
containing lymphokine preparations or partially purified lympho
toxin has been shown to be anticarcinogenic in hamsters both
in vitro and in vivo (7, 26, 27). The anticarcinogenic activity is
irreversible and occurs without inhibiting the growth of normal
cells (7). Of the known lymphokines, 2 have been shown to
directly affect carcinogenesis, lymphotoxin (7) and Interferon (2).
Additional lymphokines with potential anticarcinogenic action
when assessed in vivo may be macrophage-activating factor or
1Present address: Litton Institute of Applied Biotechnology,
Lane, Kensington, Md. 20895.
Received April 11, 1983; accepted August 4. 1983.
5222
5516 Nicholson
MIF2 and the T-lymphocyte mitogenic factors, IL-1 and IL-2,
because these lymphokines activate leukocytes to become tumoricidal (11, 14, 18, 23). Interferon is not a concern in Syrian
hamster lymphokine preparations, as interferon is not detectable
in hamster lymphokine preparations induced by in vitro stimula
tion of leukocytes with PHA (5). The purpose of this investigation
was to characterize the lymphokine species possessing the
anticancer activities ascribed to lymphotoxin.
MATERIALS AND METHODS
Animals. All animals used in this study were Syrian golden hamsters
(NIH/N strain) obtained from the NIH animal production unit (8).
Preparation of Lymphotoxin. Ten ml of mineral oil were injected i.p.
into adult male hamsters. Three days later, peritoneal cells were collected
asceptically, washed 3 times by centrifuging at 280 x g for 10 min with
RPMI 1640, and plated at a density of 2 x 107 cells per 20 ml of RPMI
1640 containing 10 ^g of PHA (leukoagglutinin type IV; Sigma Chemical
Co., St. Louis, Mo.) per ml in 100-mm Petri dishes (7). The peritoneal
cells consisted of 60% macrophages, 20% lymphocytes, and 20%
polymorphonuclear leukocytes. After 24 hr, the cell-free culture medium
was collected, diafiltered against PBS-0.1% PEG (M, 4000), and concen
trated 40-fold with an Amicon diafiltration apparatus containing a 10,000
nominal molecular weight-excluding YM10 membrane (Amicon Corp.,
Danvers, Mass.) (28). Each preparation consisted of the concentrated
pooled-culture medium after stimulation of cells from 6 animals. The
concentrated-culture medium from cells cultured in the absence of PHA
did not display any lymphotoxin cytolytic, cytostatic, or anticarcinogenic
activities.
Quantitation of Lymphotoxin Cytolytic, Cytostatic, and Anticarcin
ogenic Activities. Lymphotoxin cytolytic activity was determined by the
release of 3H from [3H]dThd-labeled murine «L-929 cells (9). One lym
photoxin cytolytic unit causes the release of 50% of the 3H from 1 x 104
3H-labeled «L-929 cells using a standard guinea pig lymphotoxin prepa
ration. Typically, 30,000 to 90,000 cytolytic units were obtained in the
concentrated lymphotoxin preparations.
Lymphotoxin cytostatic activity was determined by measuring the
inhibition of cell colony formation by Syrian hamster tumorigenic 7997
cells (10), essentially as described by Evans and Heinbaugh (9), with the
exception that these cells were grown in Dulbecco's minimal essential
medium with 10% FBS. The 7997 cells are susceptible to lymphotoxin's
growth-inhibitory
activity but are not lysed when incubated in medium
containing up to 100 cytolytic lymphotoxin units/ml.
The anticarcinogenic activity of lymphotoxin was assessed both in
vitro (7) and in vivo (26, 27). The end point in both assays was the
quantitation of morphologically transformed Syrian hamster fetal cell
colonies. Morphological transformation is a phenotypic trait of cells
undergoing carcinogenesis, because morphologically transformed colo2 The abbreviations used are: MIF, macrophage migration-inhibitory factor(s); IL1, interteukin 1; IL-2, interteukin 2; PHA, phytohemagglutinin; RPMI 1640, Roswell
Park Memorial Institute Medium 1640; PBS, phosphate-buffered saline (0.12 M
NaCI-0.01 M sodium phosphate, pH 7.4); PEG, polyethylene glycol of molecular
weight 4000; dThd, thymidine; FBS, fetal bovine seurm; IEF, isoelectric focusing;
SOS. sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on July 31, 2017. © 1983 American Association for Cancer Research.
VOL. 43
Lymphotoxin Anticarcinogenic
nies of cells can be isolated and expanded in culture and can produce
tumors when injected into hamsters or athymic nude mice (4, 6, 24). In
the in vitro assay, 12- to 13-day-gestation Syrian golden hamster fetal
cells exposed transplacentally 2 days earlier to 3 mg of diethylnitrosamine
per 100 g of maternal body weight were X-irradiated with 250 rads and
plated at a density of 9 x 103/cells/60-mm culture dish in 8 ml of modified
Dulbecco's medium containing 10% FBS, 3 x 10~6 M dThd, 3 x 10~5
hypoxanthine (Calbiochem-Behring
Corp., La Jolla, Calif.) and 50 ^g of
endothelial cell growth supplement/ml (Collaborative Research, Inc.,
Waltham, Mass.) with or without the addition of a lymphotoxin sample
(26). After 7 days of growth, colonies were fixed, stained, and examined
microscopically for the appearance of morphologically transformed cells.
Twenty-four dishes were examined for each experimental point. Lym
photoxin inhibition of carcinogenesis was reported as the percentage of
reduction in transformed colonies compared to a non-lymphotoxintreated control. The number of lymphotoxin anticarcinogenic units was
determined by quantitating the percentage of inhibition of a lymphotoxin
preparation diluted over a 100-fold range and is defined as the reciprocal
of the dilution that causes 50% inhibition of the transformation frequency.
The total average number of transformed colonies in 24 dishes after
carcinogen exposure was 26, which is a transformation frequency of 1.2
x 10~Vcell seeded.
In the in vivo anticarcinogenesis assay, lymphotoxin was injected i.v.
in the foreleg of an 11- to 12-day-gestation Syrian hamster immediately
following an i.v. injection of 250 nC't of "Tc/kg
of maternal weight in the
contralateral foreleg. "Te
is a ^-emitted
radionuclide with a 6-hr half-
life, which has been shown previously to morphologically transform
Syrian hamster fetal cells (27). After 24 hr, the fetal cells were prepared
(3) and plated at a density of 2 x 103 cells/60-mm dish in 8 ml of the
and Antitumor Cell Activities
per ml (GRS reagent solution; Schering Corp., Kenilworth, N. J.). The
cells were centrifuged at 280 x g for 10 min, resuspended in 1 ml of the
medium, and passaged through glass and nylon wool columns to obtain
the nonadherent fraction enriched for T-lymphocytes (there remained a
0.4% nonspecific esterase staining cell contamination) (1). The T-lym
phocytes were suspended to 2 x 106 cells/ml and 100 ^l were plated
per well in 96-well flat-bottomed microtiter plates (Costar 3596; Costar,
Cambridge, Mass.). One hundred n\ of test sample diluted over a 100fold range or medium were added to each well in quadruplicate. The
plates were incubated at 37°for 72 hr; then, 0.5 ^Ci of [3H]dThd in 50
n\ RPMI 1640-10% FBS was added per well. After 4 hr, the amount of
3H incorporated into the cells was determined by harvesting cells onto
filter paper with a Titertek cell harvester (Flow Laboratories, Rockville,
Md.) and counting radioactivity in a scintillation counter (21). The data
are expressed in cpm (which varied between 5 and 10% for replicate
samples) or in units of biological activity in which units were calculated
by probit analysis of [3H]dThd incorporation (13) with reference to a
standard preparation containing a defined number of units (100 units/
ml). The reference standard was a hamster lymphokine preparation that
had been diafiltered, concentrated, and frozen at -20° in aliquots. One
unit of factor activity is defined as the amount giving half-maximal activity
(13). The T-lymphocytes did not have enhanced [3H]dThd uptake when
2 to 10 Mg PHA per ml was added. Only when lymphokine was added
was there enhanced incorporation, and lymphokine enhancement was
the same with or without the addition of PHA.
Hamster PHA-induced thymocyte mitogenic responses were assayed
using a method similar to that described to measure PHA-induced
mitogenic responses of murine thymocytes (20), except that thymocytes
were obtained from 6-week-old male Syrian golden hamsters. Unfrac-
aforementioned medium without the addition of lymphotoxin. Control
animals received i.v. injections of PBS and 250 ¿iCiof "Te/kg
of
tionated Syrian hamster lymphokine preparations, however, enhanced
incorporation of [3H]dThd no more than 1.5-fold by PHA-stimulated
maternal weight. Seven days after plating, colonies were fixed, stained,
and examined microscopically for morphological transformation. A total
of 120 dishes were examined from each animal.
MIF Assay. MIF was determined by measuring macrophage migration
under agarose (32). Adult male Syrian golden hamsters were given i.p.
injections of 10 ml of mineral oil. Three days later, the peritoneal cavity
was lavaged with 40 ml of RPM11640, and peritoneal cells were washed
3 times by centrifuging at 280 x g for 10 min with RPMI 1640 and
suspended at a concentration of 2 x 108 cells/ml of RPMI 1640 with
hamster thymocytes. Because the hamster thymocytes displayed a
relatively weak response, selected lymphokine samples were kindly
analyzed in a standard IL-1 assay (20) using murine C3H/HeJ thymocytes
10% horse serum (Grand Island Biological Co., Grand Island, N. Y.). The
cells were mixed with an equal volume of test sample or PBS-0.1% PEG
as a control, and 5 n\ was added to each of six 2.5-mm wells punched
in an agarose dish (5 ml of 0.75% SeaKem agarose (Marine Colloid, Inc.,
Portland, Maine) in RPMI 1640 with 10% horse serum). The migration
was allowed to proceed overnight at 37°. The cells were fixed and
stained, and the percentage of migration was determined by measuring
the diameter of the migration and calculating:
% of migration
_ Mean area of migration in wells with added sample x 100
Mean area of migration in wells with added medium
Each experimental point was performed in triplicate and varied less than
5%. MIF units were expressed as the reciprocal of the sample dilution
(diluted over a 100-fold range) causing a 50% inhibition of migration.
Total MIF units varied from 0 to 40 in the concentrated lymphokine
preparations. Preparations containing undetectable MIF activity did have
typical amounts of lymphotoxin cytolytic activity.
Mitogenic Factor Assay. T-lymphocyte mitogenic activity was meas
ured as the enhancement of [3H]dThd uptake using 3 different target
cells: (a) Syrian hamster lymph node T-lymphocytes; (b) Syrian hamster
PHA-induced thymocytes; or (e) C3H/HeJ murine PHA-induced thymocytes (20). For hamster T-lymphocyte mitogenic activity, the axillary,
inguinal, and cervical lymph nodes were aseptically removed from 4 adult
male Syrian golden hamsters and pressed through a stainless steel sieve
into RPMI 1640 medium containing 10% FBS and 50 g of gentamicin
NOVEMBER
1983
by Dr. John Schmidt (Laboratory of Biochemistry, National Heart, Lung,
and Blood Institute).
Endotoxin Assay. The role of endotoxin (lipopolysaccharide) in the
bioassay for lymphotoxin was determined. Lipopolysaccharide with serotype numbers 055:B5 and 0111:B4 obtained from Sigma Chemical
Company was added to the cytolytic, cytostatic, and carcinogenesis
assays at concentrations of 0.01, 0.1, 0.5, 1.0, and 2.0 ng/ml in the
appropriate medium without the addition of lymphotoxin. The effect on
«L-929cell lysis, hamster tumorigenic 7997 cell cytostasis, and hamster
fetal cell carcinogenic transformation was determined as described
above. Also, selected samples were assayed for endotoxin using the
limulus amebocyte lysate test kit obtained from M. A. Bioproducts
(Walkersville, Md.). Water and PBS-0.1% PEG buffer were rendered
pyrogen free by collecting the filtrates from an Amicon cell with a YM10
membrane that had been rendered pyrogen free by passing through 200
ml of 0.1 M NaOH.
Interferon Assay. Over 24 hamster lymphokine preparations have
been assayed for us by Biofluids, Inc. (Rockville, Md.) using the inhibition
of bovine vesicular stomatitus virus cytopathic effect method (5, 7). No
detectable Interferon was evident in any preparation. As a positive
control, hamster Interferon induced in vivo by injection of Newcastle
disease virus was run with each lymphokine assay.
IEF. Preparative column IEF of the concentrated lymphokine samples
was performed initially within a pH 3.5 to 10 ampholine (LKB Instruments,
Inc., Rockville, Md.) gradient in a LKB 110-ml IEF column (LKB-Produckter AB, Bromma, Sweden) to determine the isoelectric points of lympho
toxin, MIF, and T-lymphocyte mitogenic factors (28). Three-mi fractions
were collected, the pH of each fraction was measured, and fractions
approximately within consecutive 0.5 pH units were pooled, diafiltered
as described above, and assayed for lymphotoxin cytolytic and cytostatic
activities, MIF, and T-lymphocyte mitogenic factors. Lymphotoxin activi
ties were further characterized by column IEF of concentrated lympho
kine preparations within a pH 4 to 6 ampholine gradient (28). Individual
5223
Downloaded from cancerres.aacrjournals.org on July 31, 2017. © 1983 American Association for Cancer Research.
J. H. Ransom and C. H. Evans
3-ml fractions between pH 4.3 and 5.5 were diafiltered, concentrated,
and assayed for lymphotoxin cytolytic, cytostatic, and anticarcinogenic
activities.
Gel Filtration Chromatography.
Molecular weight estimations and
purification of lymphotoxin activities were performed by gel filtration
Chromatography in a 1.6- x 78-cm ACA-44 (LKB) agarose bead column
with an upward flow rate of 24 ml of PBS-0.1% PEG/hr at 4°.Either 1
ml of an unfractionated
sample or an IEF-fractionated
sample with pHs
between 4.5 and 5.2 that had been pooled and concentrated were
applied to the ACA-44 column. For molecular weight estimation, 2-ml
fractions were collected after elution and assayed for lymphotoxin cyto
lytic, cytostatic, and anticarcinogenic activities. For purification, Fractions
37 to 41 were pooled after Chromatography of an IEF-separated lymphokine sample, concentrated to 2 ml, and sterile filtered through a millexGV 0.22 filter (Millipore Corp., Bedford, Mass.) that had been prewetted
with media containing 10% FBS. Prewetting of the filter is critical for
sample recovery in its purified form. The purified material was subject to
analysis for lymphotoxin cytolytic, cytostatic, and anticarcinogenic activ
ities, MIF, and T-lymphocyte mitogenic factors. Molecular weight stand
ards used to calibrate the column were bovine serum albumin, (M, 68,500)
ovalbumin (M, 45,000), chymotrypsinogen A (M, 25,000), and RNase A
(M, 13,700) (Pharmacia Fine Chemicals, Piscataway, N. J.).
The amount of protein contained within samples during the purification
procedure was estimated by measuring the absorbance of the sample
at 280 nm and determining the amount of protein using a standard curve
of the absorbances of bovine serum albumin solutions from 100 to 300
jig/ml. Protein determination was performed on the highly purified lym
photoxin before sterile filtration to avoid possible contamination of serum
protein from the prewetted filter.
SOS-PAGE. Both preparative and analytical SDS-PAGE were per
formed in 12% polyacrylamide gels with 5% stacking gels (17) using a
Bio-Rad flatbed gel apparatus with 0.9-mm spacers. In the analytical gel,
cytolytic activity possessing an isoelectric point of 4.9 to 5.0 and
another with an isoelectric point at 4.6.
Analysis of the molecular weight of lymphotoxin cytolytic,
cytostatic, and anticarcinogenic activities by gel filtration Chro
matography showed that each had the same apparent molecular
weight of 35,000 to 45,000 (Chart 3). Thus, based on the
molecular characterization by size and charge, a 3-step purifi
cation was performed which consisted of diafiltration and con
centration, IEF within the pH range of 4 to 6, and ACA-44 column
Chromatography (Table 1). After diafiltration, 98% of the cytolytic
activity was recovered, with a 7.7-fold increase in specific activ
ity. IEF yielded a 95% recovery of lymphotoxin cytolytic activity
for a 23-fold increase in specific activity. When IEF was followed
by gel filtration Chromatography, a 25% recovery of the original
lymphotoxin cytolytic activity was obtained in the fractions within
the range of M, 35,000 to 55,000, with a final 35-fold increase in
specific activity. No lymph node T-lymphocyte mitogenic factor
or MIF activities were detectable in the purified lymphotoxin
(Table 1). There was also no IL-1 activity as assayed on C3H/
HeJ thymocytes in the purified lymphotoxin that initially had 300
total IL-1 units in the diafiltered and concentrated sample. Injec
tion of 2000 to 3000 cytolytic units of 3-step purified lymphotoxin
into pregnant hamsters inhibited the MmTc-/n wVo-induced trans
formation of fetal cells by 56% (Table 2).
The molecular size and protein content of the IEF-gel filtrationpurified lymphotoxin was analyzed by SDS-PAGE (Chart 4). A
single silver-stained protein band with a molecular weight of
50,000 dallons was present. When the starting unpurified-lym-
50 liters of an unfractionated sample containing 900 cytolytic lymphotoxin
units or 50 n\ of a sample purified by IEF and ACA-44 Chromatography
and concentrated to 0.5 ml containing 400 cytolytic lymphotoxin units
were boiled for 3 min in 0.5 M Tris-HCI (pH 7.0), 2% SDS, 5% mercaptoethanol, and 0.1% bromophenol blue and electrophoresed until the
tracking dye reached the bottom of the gel. The gels were removed,
fixed in 50% methanol-12% acetic acid overnight, and stained for protein
with silver stain (19). For preparative SDS-PAGE, a 1-ml unfractionated
sample containing 43,000 lymphotoxin cytolytic units was prepared and
electrophoresed as described above. After electrophoresis, the gel was
sliced into 4-mm strips. SDS was eluted by incubating each slice for 1
hr in 5 ml of RPMI 1640-10% FBS at 37°. Lymphotoxin was eluted by
4000
Lymphotoxin Cytolytic Activity
<fl 2000-
Lymphotoxin Cytostatic Activity
O
so
ÕÕ 40
additional incubation in 5 ml of fresh RPMI 1640-10% FBS for 18 hr at
37°. Molecular weight markers included in both the analytical and pre
parative gels were bovine serum albumin, ovalbumin, and cytochrome c
(M, 12,500) (Pharmacia).
RESULTS
Initial characterization of the lymphokine preparations by IEF
revealed lymphotoxin cytolytic and cytostatic activities to have
isoelectric points between 4.3 and 5.5 (Chart 1). Lymph node Tlymphocyte mitogenic activity had 3 maxima with isoelectric
points of 4 to 4.5 and 6.7 and a major peak at 8.3. MIF had
activity peaks with isoelectric points of 4.3 and 5.9. Therefore,
by selecting a narrower pH gradient of pH 4 to 6, the majority of
the mitogenic activity and MIF activities could be separated by
lymphotoxin. Chart 2 shows the IEF separation of lymphotoxin
activities in a pH 4 to 6 gradient with the majority of the
lymphotoxin activity focusing in the range of pH 4.5 to 5.2.
Lymphotoxin cytolytic activity possessed an isoelectric point of
4.9 to 5.1. Lymphotoxin cytostatic and anticarcinogenic activi
ties, however, displayed 2 maxima, one coinciding with the
5224
Lymphocyte
Mitogenic Activity
800-
£
O 400o
Macrophage
Migration Inhibition
80-
40
4.3 4.5 5.3 5.9 6.7 7.5 8.3 9.2
pH
Chart 1. IEF characterization of lymphokines in a pH 3.5 to 10 ampholine
gradient. Two ml of a diafiltered and concentrated lymphokine sample containing
10,000 lymphotoxin cytolytic units, 500 units of mitogenic factor, and 20 MIF units
were focused. Three-mi fractions were collected, and fractions within consecutive
0.5 pH unit were pooled, diafiltered, concentrated to 3 ml, and assayed. The pH
reported for each fraction is the mean pH of the fraction pool.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on July 31, 2017. © 1983 American Association for Cancer Research.
VOL. 43
Lymphotoxin Anticarcinogenic
photoxin-containing
lymphokine
resed on an SDS-polyacrylamide
preparation was electrophogel and cytolytic lymphotoxin
and Antitumor Cell Activities
800
100
activity was eluted and assayed, a single peak at 50,000 daltons
was obtained. The recovery of lymphotoxin activity after boiling
the lymphokine preparation with SDS-mercaptoethanol sample
buffer before electrophoresis was 0.4%, which does not make
this method useful for purification of lymphotoxin. It does, how
ever, verify that lymphotoxin has an apparent molecular weight
of 50,000 daltons on SDS-PAGE. It also suggests that the
apparent molecular weight and heterogeneity displayed by gel
filtration chromatography may have been influenced by interac
tions of lymphotoxin with the gel matrix, causing a retardation
of lymphotoxin elution and, therefore, a lower apparent molecular
weight.
Six hamster lymphokine preparations were tested for the
presence of endotoxin using the limulus amebocyte lysate assay.
All 6 gave a positive test result; however, PBS-PEG buffer that
had been filtered through a M, 10,000 cutoff membrane also
gave a positive test result. Water prepared by the membrane
20000
15000-
z
^
10000-
0»
t
z
22
9
o
27
32
37
42
47
FRACTION NUMBER
52
Charts. Analysis of lymphotoxin cytolytic (•),cytostatic (O), and anticarcinogenic (O) activities by ACA-44 gel filtration chromatography. Two replicate experi
ments were performed in which 2 different lymphotoxin preparations were assayed.
In Experiments 1 and 2, 70.000 and 50,000 cytolytic units, respectively, were
chromatographed with a 27 and 25% recovery of activities. Cytostatic activity is
reported as the percentage of inhibition of control Syrian hamster tumorigenic cell
colony formation [55 ±2.3 (S.E.) colonies/dish in Experiment 1). Anticarcinogenic
activity is reported as the percentage of inhibition of the diethylnitrosamine-Xirradiation-induced transformed Syrian hamster cell colony formation frequency
(which was 0.92 ±0.08 transformed colonies/dish in Experiment 1). Analysis of
variance, followed by the Dunette test to compare an experimental group to the
control (36), showed that the percentage of inhibitions greater than 20% was
statistically significantly different from that of the control (p < 0.05).
p
>
o
>
o
z
X
o
Io
X
o.
4.2 4.4 4.6 4.8 5.0
PH
Chart 2. Analysis of lymphotoxin cytolytic (•),cytostatic (O), and anticarcinogenic (Ü)activities after lEF in a pH 4 to 6 ampholine gradient. Two replicate
experiments were performed in which 2 different lymhotoxin preparations were
assayed. In Experiments 1 and 2, 71,000 and 135,000 cytolytic units, respectively,
were focused with 95 and 97% recovery of the activities. After focusing, 3-ml
fractions were collected and diafiltered. Cytostatic activity is reported as the
percentage of inhibition of control Syrian hamster tumorigenic cell colony formation
[which was 71 ±3.7 (S.E.) colonies/dish in Experiment 1]. Anticarcinogenic activity
is reported as the percentage of inhibition of the diethylnitrosamine-X-irradiationinduced transformed Syrian hamster cell colony formation frequency [which was
1.2 ±0.2 transformed colonies/dish in Experiment 1]. Analysis of variance followed
by the Dunette test to compare an experimental group to the control (36) showed
that the percentage of inhibitions greater than 20% was statistically significantly
different from that of the control (p < 0.05).
NOVEMBER
1983
filtration method gave a negative test result. Therefore, the PEG
in the buffer and in the 6 hamster lymphokine preparations may
have caused a false-positive test. In order to determine what
biological effects endotoxin may have had in our experimental
system, the following experiments were performed.
Lipopolysaccharide with serotypes 055:B5 and 0111 :B4 were
added at concentrations of 0.01, 0.1, 0.5,1.0, and 2.0 ng/ml to
the appropriate medium for the cytolytic, cytostatic, and carcinogenesis assays. Neither of these 2 endotoxins at any of the
concentrations tested lysed aL-929 cells, inhibited the growth of
tumorigenic 7997 hamster cells, or inhibited the growth of normal
or carcinogen-transformed hamster fetal cells. Because endo
toxin has a molecular weight of greater than 2,000,000, the
absence of any of the 3 biological activities from the void volume
of the gel filtration column suggests along with the lack of
bioactivity by either of the endotoxin preparations tested that
endotoxin is not mediating any of the biological activity attributed
to lymphotoxin.
5225
Downloaded from cancerres.aacrjournals.org on July 31, 2017. © 1983 American Association for Cancer Research.
J. H. Ransom and C. H. Evans
Table 1
Three-step purification of lymphotoxin activity
MIF
Lymphotoxin activity
Mitogenic factor
CytolyticFractionation
stepUnfractionatedDiafiltered
ac
ac
activity
tivity8 (units/
tivity (units/
Total
(units/mg
protein
Total
protein)a50,000
units
mg
units
protein)—
mgprotein)1,64011,00034,50058,800AnticarcinogenicSpecific
units42,00040,00038,00010,000Specific
(mg)25.63.61.10.17Total
13,900—
and con
centratedIEF.
+ ACA-44 gel
«trattaiTotal
—, not measured.
6 No activity detected.
—32
—218
8.9<1»
61—
<1<1Specific
—10,000
—<1*
4.5-5.2IEF
pH
activity
Total
(units/mg
units
protein)—
58,800Specific
Table 2
Inhibition of carcinogenesis in vivo by purified lymphotoxin
Pregnant hamsters at 11 to 12 days of gestation were given i.v. injections of
250 »jCiTc/kg
of maternal weight and in the contralateral foreleg with PBS or
diafiltered and concentrated lymphotoxin purified by IEF, pH 4 to 6, and ACA-44
gel filtration. After 24 hr, fetuses were removed and trypsinized, and cells were
plated for colony formation. After 7 days, the number of morphologically trans
formed colonies was quantitated.
Lymphotoxin cytolytic
units
i.v.None
injected
colonies/
tions/cells
seeded (x
10-4)7.0
dish1.
40 ±0.39a
(control)
20-30
0.93 ±0.20
4.7
0.62 ±0.10°Transforma
2000-3000Transformed
3.1%
* Mean ±S.E.; 6 animals were examined in each group,
"p < 0.05 compared by Student's f test.
Molecular
Weight
Lymphotoxin
Units/
ml
20 40 60
of inhibi
tion of trans
formation3456
68,500
>
45.000
>
DISCUSSION
This investigation demonstrates that interferon-free Syrian
hamster lymphotoxin purified by a 3-step diafiltration, IEF, and
gel filtration procedure is devoid of detectable MIF and T-lymphocyte mitogenic factors and inhibits carcinogenesis in vitro
and in vivo. Furthermore, the 3-step purified lymphotoxin mi
grates as a single band on SDS-polyacrylamide gels with an
approximate molecular weight of 50,000. A unique finding of this
investigation is that lymphotoxin's anticarcinogenic and tumor
cell cytostatic activities had one shared isoelectric point (5.0) in
common with and one isoelectric point (4.6) distinct from lym
photoxin cytolytic activity. It is important to note, however, that
hamster target cells were used to assay the anticarcinogenic
and cytostatic activities, and a murine target cell was used to
test for cytolytic activity. Differences in the isoelectric points of
classes of glycoproteins are frequently due to the number of
charged species in the carbohydrate chain (30). Also, carbohy
drate residues have also been implicated as dominant or critical
domains in receptors for lymphotoxin binding (16). Therefore,
the anticarcinogenic and tumor-inhibitory yet noncytolytic lym
photoxin activities with an isoelectric point of 4.6 could be due
to differences in carbohydrate side chains, resulting in slightly
different cellular recognition and associated activity. Differences
in glycosylation may result from postranslational intracellular
glycosyl transferase activity or from deglycosylation during puri
fication of the lymphokine and warrant further study.
This is the first characterization, to our knowledge, of the
isoelectric points of hamster MIF and T-lymphocyte mitogenic
5226
12,500>
Chart4. SDS-PAGE of a diafiltered ÇA)and 3-step purified (B) lymphotoxin
sample. The chart demonstrates the results when an unfractionated preparation
(Lane A) was electrophoresed, the gel was sliced and eluted. and the eluates were
tested for lymhotoxin cytolytic activity. Each point represents the relative position
of the gel slice compared to the molecular weight standards.
factors. Hamster MIF has isoelectric points similar to those of
guinea pig MIF (29). Guinea pig MIF has been extensively char
acterized and appears to be indistinguishable from macrophage
activating factor (22, 29). Macrophage activating factor activates
macrophages to become tumoricidal (23). Unfractionated ham
ster lymphokine preparations had low levels of hamster thymocyte mitogenic activity (IL-1); therefore, IL-1 was assessed in
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on July 31, 2017. © 1983 American Association for Cancer Research.
VOL.
43
Lymphotoxin Anticarcinogenic and Antitumor Cell Activities
unfractionated and purified lymphotoxin samples by enhance
ment of PHA-induced mitogenic responses of murine thymocytes
(20). Although there was significant IL-1 activity in unfractionated
lymphokine preparations, purified lymphotoxin contained no de
tectable IL-1. In addition to IL-1 activity, unfractionated lympho
kine also induced hamster lymph node T-lymphocyte [3H]dThd
incorporation. This mitogenic activity induced [3H]dThd incorpo
ration into freshly derived unstimulated T-lymphocytes and is,
therefore, functionally different than IL-2. This lymphocyte mito
genic factor, like IL-2 in humans and mice, may augment the
tumoricidal activity of lymphocytes in vitro (14) and may sustain
the longevity of antitumor cytotoxic T-lymphocytes in vivo (34).
However, thymus-derived lymphocytes can have either positive
or negative effects on carcinogenesis (33). Therefore, the fact
that the 3-step purified lymphotoxin did not contain MIF or T-cell
mitogenic factors suggests that activation of macrophages or Tlymphocytes is not a direct result of lymphotoxin's effect on
carcinogenesis. It also demonstrates that the lymphokine inhib
iting carcinogenesis in vivo is a form of lymphotoxin.
The specific anticarcinogenic activity of purified lymphotoxin
is 58,500 units per mg of protein, which, based on a molecular
weight for lymphotoxin of 50,000, yields a molar specific activity
of 2,900 units/nmol. As 2000 to 3000 units of lymphotoxin
caused a greater than 50% inhibition of carcinogenesis in vivo,
then approximately 0.7 nmol of lymphotoxin causes this degree
of inhibition. Most hormones are biologically active at physiolog
ical concentrations of less then 10 nw (35). In addition, lympho
toxin has been induced in the peritoneal cavity of hamsters after
injection with either PHA or antigen at levels up to 1000 total
units or 0.3 nmol (12).
Thus, inducible levels in the hamster are within the dosage
range of in vivo anticarcinogenic effectiveness, suggesting that
lymphotoxin has the potential to be active at physiological con
centrations. Should a carcinogenic alteration of cells occur such
that an immunological response by the host is induced, lympho
toxin may function as a natural homeostatic regulatory molecule
of carcinogenesis.
REFERENCES
1 Altman. A., and Rapp. H. J. Natural cell-mediated cytotoxicity in guinea pigs:
properties and specificity of natural killer cells. J. Immunol., 727: 2244-2252,
1978.
2. Brouty-Boyé, D.. and Little, J. B. Enhancement of X-ray-induced transformation
in C3H/10TV2 cells by interferon. Cancer Res., 37: 2714-2716,1977.
3. DiPaolo, J. A. Quantitative in vitro transformation of Syrian golden hamster
embryo cells with the use of frozen stored cells. J. Nati. Cancer Inst., 64:
1485-1489, 1980.
4. DiPaolo, J. A., Nelson, R. L, Donovan, P. J., and Evans, C. H. Host-mediated
in vivo-in vitro assay for chemical carcinogenesis. Arch. Pathol., 95. 380-385.
1982.
5. Evans, C. H. Lymphotoxin—an immunologie hormone with anticarcinogenic
and antitumor activity. Cancer Immunol. Immunother., 12:181-190,1982.
6. Evans, C. H., and DiPaolo, J. A. Comparison of nude mice with the host
species for evaluation of the tumorigenicity of guinea pig and hamster cells
transformed in vitro by chemical carcinogens. Cancer Res., 36: 128-131,
1976.
7. Evans, C. H., and DiPaolo, J. A. Lymphotoxin: an anticarcinogenic lymphokine
as measured by inhibition of chemical carcinogen or ultraviolet irradiationinduced transformation of Syrian hamster cells. Int. J. Cancer, 727: 45-49.
1981.
8. Evans, C. H., DiPaolo, J. A., Heinbaugh, J. A., and DeMarinis, A. J. Immunemodulation of the lymphoresponsive phases of carcinogensis: mechanisms of
natural immunity. J. Nati. Cancer Inst., 69: 737-740, 1982.
9. Evans. C. H., and Heinbaugh, J. A. Lymphotoxin cytotoxicity, a combination
of cytolytic and cytostatic cellular responses. Immunopharmacology, 3: 347359, 1981.
NOVEMBER
10. Evans, C. H., Heinbaugh, J. A., and DiPaolo, J. A. Comparative effectiveness
of lymphotoxin anticarcinogenic and tumor cell growth inhibitory activities. Cell.
Immunol., 76: 295-303, 1983.
11. Fidler, l. J. Therapy of spontaneous métastases by intravenous injection of
liposomes containing lymphokines. Science (Wash. D. C.), 208: 1469-1470,
1980.
12. Fuhrer, J. P., Evans. C. H., Ransom, J. H., and Heinbaugh, J. A. In vivo and in
vitro identification of (i-lymphotoxin as the molecular class of Syrian hamster
lymphotoxin with antineoplastic activity. Cell. Immunol., 75: 63-70, 1983.
13. Gillis, S., Perm, M. M., Ou, W., and Smith, K. A. T-cell growth factor: parameters
of production and a quantitative microassay for activity. J. Immunol., 720:
2027-2032, 1978.
14. Grimm, E. A., Mazumder, A., Zhang, H. Z., and Rosenberg. S. A. Lymphokineactivated killer cell phenomenon lysis of natural killer-resistant fresh solid tumor
cells by interieukin 2-activated autologous human periferal blood lymphocytes.
J. Exp. Med., 755: 1823-1830, 1982.
15. Khan, A., Martin, E. S., Webb, K., Weldon, D., Hilt, N. 0., Duvall, J., and Hill,
J. M. Regression of malignant melanoma in a dog by local injections of alphalymphotoxin. Proc. Soc. Exp. Bid. Med., 769: 291-294, 1982.
16. Kobayashi, Y., Sawada, J., and Osawa, T. Isolation and characterization of an
inhibitory glycopeptide against guinea pig lymphotoxin from the surface of Lcells. Immunochemistry, 75: 61-69, 1978.
17. Laemelli, U. K. Cleavage of structural proteins during the assembly of the head
of bacteriophage T4. Nature (Lond.), 227: 680-683,1970.
18. Lotze, M. T., Grimm, E. A., Mazumder, A., Strausser, J. L., and Rosenberg,
S. A. In vitro growth of cytotoxic human lymphocytes. IV. Lysis of fresh and
cultured autologous tumors by lymphocytes cultured in T-cell growth factor.
Cancer Res., 41: 4420-4425. 1982.
19. Merril, C. R., Goldman, D., Sedman, S. A., and Eben, M. H. Ultrasensitive
stain for proteins in polyacrylamide gels shows regional variation in cérébrospinal fluid proteins. Science (Wash. D. C.), 277: 1437-1439, 1981.
20. Mizel, S. B., Oppenheim, J. J., and Rosenstreich, D. L. Characterization of
lymphocyte-activating
factor (LAP) produced by the macrophage cell line
P388D. I. Enhancement of LAF production by activated T-lymphocytes. J.
Immunol.. 720: 1497-1503,1978.
21. Oppenheim, J. J.. and Schecter, B. Lymphocyte transformation. In: N. Rose
and H. Friedman (eds.), Manual of Clinical Immunology, pp. 81-123. Washing
ton, D. C.: American Society for Microbiology, 1976.
22. Pick, E. Lymphokines: physiologic control and pharmacological modulation of
their production and action. In: J. W. Hadden, R. C. Coffey, and F. Spreofico
(eds.). Comprehensive Immunology, Vol. 3, pp. 178-198. New York: Plenum
Publishing Corp., 1977.
23. Piessens, W. F., Churchill, W. J., and David, J. R. Macrophages activated in
vitro with lymphocyte mediators kill neoplastic but not normal cells. J. Immunol.,
774:293-297,
1975.
24. Quaries. J. M., Sega. M. W., Schenley, C. K.. and Lijinsky, W. Transformation
of hamster fetal cells by nitrosated pesticides in a transplacental assay. Cancer
Res., 39: 4525-4533, 1979.
25. Ransom. J. H., and Evans, C. H. Lymphotoxin enhances the susceptibility of
neoplastic and preneoplastic cells to natural killer cell mediated destruction.
Int. J. Cancer, 29: 451-458, 1982.
26. Ransom, J. H., Evans, C. H., and DiPaolo, J. A. Lymphotoxin prevention of
diethylnitrosamine carcinogenesis in vivo. J. Nati. Cancer Inst., 69: 741-744,
1982.
27. Ransom, J. H., Evans, C. H.. Jones, A. E.. Zoon, R. A., and DiPaolo, J. A.
Control of the carcinogenic potential of Technetium
by the immunologie
hormone lymphotoxin. Cancer Immunol. Immunother., in Press, 1983.
28. Ransom, J. H., Rundell, J. O., Heinbaugh, J. A., and Evans, C. H. Biological
and physiochemical characterization of keyhole limpet hemocyanin-induced
guinea pig lymphotoxin. Cell. Immunol., 67: 1-13, 1982.
29. Remold, H. G.. McCarthy, P. L., and Mednis, A. D. Purification of guinea pig
pH 3 migration inhibitory factor. Proc. Nati. Acad. Sei. U. S. A., 78: 40884098,1981.
30. Remold, H. G., and Mednis, A. D. The effect of tunicamycin on the molecular
heterogeneity of guinea pig migration inhibitory factor. Cell. Immunol., 66:171181, 1982.
31. Rundell. J. O., and Evans, C. H. Species specificity of guinea pig and human
lymphotoxin colony inhibitory activity. Immunopharmacology, 3: 9-18, 1981.
32. Spitter, L. E., and von Muller, C. Leukocyte migration inhibition in agarose. In:
B. R. Bloom, and J. R. David (eds.), in Vitro Methods in Cell-mediated and
Tumor Immunity, pp. 645-650. New York: Academic Press, Inc., 1976.
33. Stutman, O. Immunologie surveillance and cancer. In: H. Waters (ed.), The
Handbook of Cancer Immunology. New York: Garland Publishing, Inc.
34. Valentine, F., Golomb, F., and Fazzini, E. Effects in vivo of autologous
lymphokines (LK) on injected and on noninjected metastatic nodules of human
malignant melanoma. Int. J. Immunopharmacol., 4: 375-400, 1981.
35. Weite, K., Wang, C. V., Mertelsmann, R., Venuta, S., Feldman, S. P., Moore,
M. A. S. Purification of human interieukin 2 to apparent homogeneity and its
molecular heterogeneity. J. Exp. Med., 756: 454-468, 1982.
36. Winer, B. J. In: Statistical Principles in Experimental Design, pp. 201-204.
New York: McGraw Hill Book Co., 1971.
1983
Downloaded from cancerres.aacrjournals.org on July 31, 2017. © 1983 American Association for Cancer Research.
5227
Molecular and Biological Characterization of Anticarcinogenic
and Tumor Cell Growth-inhibitory Activities of Syrian Hamster
Lymphotoxin
Janet H. Ransom and Charles H. Evans
Cancer Res 1983;43:5222-5227.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/43/11/5222
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 [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on July 31, 2017. © 1983 American Association for Cancer Research.