ARTICLES
Induction of Programmed Cell Death in Kaposis
Sarcoma Cells by Preparations of Human Chorionic
Gonadotropin
Felipe Samaniego, Joseph L. Bryant, Ni Liu, Judith E. Karp, Anita L. Sabichi,
Alain Thierry, Yanto Lunardi-Iskandar, Robert C. Gallo
Background: Isolation of the first neoplastic acquired immu-
nodeficiency syndrome-related Kaposis sarcoma (KS) cell
line (KS Y-1) has furthered understanding of the pathogen-
esis of KS. Studies with KS Y-1 cells have indicated that
inhibition of KS cell proliferation occurs in early pregnancy
in mice and after treatment with certain commercial prepa-
rations of human chorionic gonadotropin (hCG, a pregnancy
hormone purified from urine). The activity of the commer-
cial preparations has been attributed to an hCG-associated
factor(s) (HAF). While several clinical benefits of HAF are
clearly evident, the basis for its anti-KS properties remains
unknown. We investigated the apoptosis-inducing effects of
HAF and the expression of apoptosis-related proteins in KS
cells. Methods: KS Y-1 and KS SLK cells were treated with
clinical-grade crude preparations of hCG, recombinant
hCG, or urine fractions exhibiting anti-KS activity and then
examined for features of apoptosis. Levels of proteins asso-
ciated with apoptosis were monitored by western blot analy-
sis, and cell DNA content was assessed by flow cytometry.
Tumors induced in mice by inoculation of KS Y-1 cells were
treated with preparations of hCG, and the tumors were ex-
amined for cell morphology and also for DNA fragmentation
by use of the terminal deoxynucleotidyl transferase-
mediated digoxigenin-deoxyuridine triphosphate nick-end-
labeling (TUNEL) assay. Results: The HAF present in some
preparations of hCG and in urine fractions has the ability to
induce apoptosis in KS cells in vitro and in vivo. HAF-
triggered apoptosis was preceded by increased levels of the
apoptosis-related proteins c-Myc and c-Rel and cell accumu-
lation in G0/G1 phase of the cell cycle. KS Y-1 cells trans-
fected with a c-Myc complementary DNA showed elevated
rates of apoptosis. Conclusion: The anti-KS activity of HAF
appears to induce apoptosis. Such activity suggests a role for
HAF in pregnancy-related regulation of cell death. [J Natl
Cancer Inst 1999;91:13543]
Kaposis sarcoma (KS) tumors are blood vessel-rich lesions
characterized by proliferation of cytokine-dependent spindle
cells, activated endothelial cells, immune cells (13), and the
presence of human herpesvirus 8 (HHV8)-infected cells (4).
These lesions express basic fibroblast growth factor (a potent
angiogenic factor), platelet-derived growth factor, and vascular
endothelial growth factor (5), as well as interleukin 6, hepato-
cyte growth factor, and interleukin 8 that may collectively con-
tribute to the abundantly vascularized appearance of these le-
sions (68). Nodular or late-stage forms of KS harbor malignant
Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999
cells, which proliferate independently of cytokines (911) and
are often clonal (12). We and others (11,12) have isolated tu-
morigenic cell lines (KS Y-1 and KS SLK) that are derived from
nodular KS lesions, which contained rearrangement in chromo-
some 3 at region 3p14, the location of the transformation-related
fragile histidine triad gene. Virtually all cases of KS tumors,
regardless of their epidemiologic type (acquired immunodefi-
ciency syndrome [AIDS]-related, endemic, classical, or iatro-
Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015
genic), contain HHV8 DNA sequences suggesting that it is a
necessary cofactor (13), and when present with human immu-
nodeficiency virus type 1 (HIV-1) infection, both contribute to a
progressive cancer.
Our initial study (10) of tumors induced by the inoculation of
KS Y-1 and KS SLK cells in nude mice revealed an unantici-
pated pronounced inhibition of tumor development during early
pregnancy. The anti-KS activity in mice was strikingly effective
during the first half of the gestation period when chorionic go-
nadotropin-like activity in sera from mice (and humans) is maxi-
mum (the murine equivalent of human chorionic gonadotropin
[hCG] has not yet been identified). Some clinical-grade crude
preparations of hCG initiated a regression of tumors induced by
KS Y-1 cells in nude mice and inhibited clonogenic KS Y-1 cell
foci formation (10). These animal experiments spurred human
trials to test local or systemic administration of hCG for KS.
Intralesional and systemic treatment with commercially avail-
able clinical-grade crude preparations of hCG induced signifi-
cant rates of regression of both cutaneous and of the more re-
sistant visceral KS lesions (14,15). Because commercial
preparations of hCG from various sources contain variable anti-
KS activity yet equivalent endocrine activity and because highly
purified hCG and recombinant hCG lack anti-KS activity, this
collectively suggested to us that the activity of the commercial
preparations of hCG was due to a co-purified factor (or factors)
that had the ability to induce regression of KS tumors. We have
termed this anti-KS factor as hCG-associated factor(s) (HAF)
that is present in some, but not all, commercial clinical-grade
Affiliations of authors: F. Samaniego, Institute of Human Virology, Medical
Biotechnology Center and Greenebaum Cancer Center, University of Maryland,
Baltimore; J. L. Bryant, N. Liu, Y. Lunardi-Iskandar, R. C. Gallo, Institute of
Human Virology, Medical Biotechnology Center, University of Maryland; J. E.
Karp, Greenebaum Cancer Center, University of Maryland; A. L. Sabichi, The
University of Texas M. D. Anderson Cancer Center, Houston; A. Thierry, Trans-
gene, France.
Correspondence to: Felipe Samaniego, M.D., Institute of Human Virology,
Medical Biotechnology Center, University of Maryland, 725 W. Lombard St.,
Rm. N454, Baltimore, MD 21201-1192 (e-mail: samanieg@umbi.umd.edu).
See Notes following References.
Oxford University Press
ARTICLES 135
crude preparations of hCG and that which is present in urine
during pregnancy. We have shown that anti-KS activity did not
reside in the hCG molecule or in one of its subunits (16). While
patients were undergoing treatment for KS with preparations of
hCG, we also noted an improvement in their HIV-1 load, he-
matopoietic parameters, and overall well-being. Furthermore,
macaques with clinical simian immunodeficiency virus infection
treated with preparations of hCG showed reduced viral load,
weight gain, and prevention of AIDS (16). The anti-HIV-1 and
anti-KS activities appear to co-separate in partial purification of
HAF, suggesting that these activities may stem from the same
molecule(s) (16) but are yet to be proven conclusively. There-
fore, HAF may have the ability to reverse some of the effects of
AIDS, in addition to its anti-KS activity. While several clinical
benefits are clearly evident with treatment with preparations of
hCG, the underlying basis for these improvements, including
anti-KS properties of HAF, remains unknown. To investigate the
anti-KS activity of HAF, we treated KS cells in vitro and KS
tumors in vivo with preparations of hCG and with partially pu-
rified fractions of HAF from the urine of pregnant women. We
describe apoptosis-related morphologic changes and protein ex-
pression.
METHODS
Treatment of tumors induced by KS Y-1 cells with preparations of hCG.
Tumors were first induced in beige nude mice (Harlem, Frederick, MD) by
subcutaneous inoculation of 1 106 KS Y-1 cells into the dorsal lateral flanks,
which consistently produced 5 5 mm2 tumors within 2 weeks. Several com-
mercial preparations of hCG (clinical-grade crude hCG at a dose of 500 IU [0.25
IU/g protein]) were evaluated, but a batch of Wyeth Ayerst preparations of
hCG (Wyeth Ayerst Laboratories, Philadelphia, PA) that induced apoptosis of
KS Y-1 cells in culture was used in all studies except where shown otherwise.
Preparations of hCG or an equivalent amount of hCG buffer (i.e., equivalent
concentration of salts in commercial hCG preparations) were administered daily
by injection into KS lesions and adjacent tissues for 5 consecutive days. The
dose of 500 IU of commercial hCG was determined as an adequate dose required
for inducing regression of the tumors, and a maximal tolerated dose was not
determined. Since the active principle with anti-KS activity is not totally char-
acterized, any chance of variation in anti-KS activity among different prepara-
tions was eliminated by using only one batch of hCG with reasonable activity.
Animal care was in accord with institutional guidelines. Bidirectional tumor
dimensions were noted before and after treatments, tumors were resected, and
hematoxylineosin-stained thin sections of tumors were examined for morphol-
ogy. To study the early effects of preparations of hCG-induced tumor regression,
tumors were treated intralesionally with preparations of hCG (500 IU) or buffer
for 2 days. The tumor sites and surrounding tissues were resected and fixed in
formalin. The terminal deoxynucleotidyl transferase-mediated digoxigenin-
deoxyuridine triphosphate nick-end-labeling (TUNEL) assay was performed by
treatment of thin tissue slide sections with terminal deoxynucleotidyl transferase
for in situ extension at DNA termini by incorporation of digoxigenin-
deoxyuridine triphosphate (Oncor Inc., Gaithersburg, MD). Cells with high lev-
els of DNA termini support incorporation of digoxigenin-deoxyuridine triphos-
phate, which was detected by peroxidase-conjugated anti-digoxigenin antibody.
Also, using the same tissue fixative method, thin tissue slides of control rat gut
were concurrently analyzed for apoptosis. Cells at the base and cells at the tip of
rat gut villi tissue served as reference material for live and apoptotic cells,
respectively.
Cells. KS Y-1 cells were isolated and established as a neoplastic cell line from
a pleural effusion of an HIV-1-infected male with KS involving the lung (10),
and KS SLK cells were derived from an oral KS lesion of an HIV-1-noninfected
kidney transplant recipient receiving immunosuppressive drugs (11). Both cell
lines bear endothelial markers and induce angiogenic tumors when inoculated in
nude mice. AIDS-KS4 cells are early passage (passages 35) hyperplastic cells
directly derived from human KS lesions that express tissue specific markers,
proliferate in response to inflammatory cytokine stimulation and induce forma-
tion of transient KS-like lesions when inoculated in nude mice (5,6,8). AIDS-
136 ARTICLES Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999
KS4 cells are considered to represent the hyperplastic spindle cells of KS lesions.
AIDS-KS4 cells as well as the two cell lines described above were maintained
in a gelatin (1.5%)-coated flask in RPMI-1640 medium containing the following:
15% fetal bovine serum (FBS), essential and nonessential amino acids, 1%
Nutridoma (Boehringer Mannheim Corp., Indianapolis, IN), penicillin G (100
U/mL), and streptomycin (100 g/mL). These cells (KS Y-1, KS SLK, and
AIDS-KS4) lack HHV8 and HIV-1 DNA (17). Human umbilical vein endothe-
lial (H-UVE) cells and breast cancer cells (Bt-483) were included as controls;
their propagation has been described by American Type Culture Collection
(Manassas, VA).
Partially purified HAF from urine of pregnant women. Urine from first-
trimester pregnancy (40 L) was filtered, concentrated, and then desalted on a
Sephadex G25 column that effectively reduced the volume to less than 500 mL
(16). The protein-containing peak was then separated on a Superdex 200 column
(26/60) (Pharmacia Biotech, Inc., Piscataway, NJ) in phosphate-buffered saline
(PBS). Fractions were measured for total protein levels and for concentrations of
heterodimeric hCG and the hCG -core fragment (a cleaved form of hCG) by
specific immunoassays. The hCG heterodimer elutes as a 70-kd molecule and the
hCG -subunit core elutes as a 10- to 25-kd peak on gel filtration (16). hCG-
immunoreactivity peaked between fractions 46 and 49 of pregnancy urine con-
centrates. The first and second peaks of anti-KS activity elute with molecular
ranges of 1530 and 24 kd and their respective peak positions are remote from
the elution positions of hCG or hCG -free subunit.
Treatment of cells with preparations of hCG and purified hCG (CR 127).
Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015
KS Y-1, KS SLK, and Bt-483 cells were seeded into T75 flasks at 20% con-
fluency in RPMI-1640 medium containing 15% FBS and grown for 1 day. The
cells were then synchronized by incubation for 24 hours in medium containing
1% FBS, pulsed for 24 hours with medium containing 10% FBS, and then
incubated in medium containing 1% FBS, preparations of clinical-grade hCG, or
equivalent amounts of CR 127 (1001000 IU/mL, National Hormone and Pitu-
itary Program and Center for Population Research, National Institutes of Health,
Bethesda, MD), thyroid-stimulating hormone (TSH), a hCG-related heterodimer
glycoprotein (Sigma Chemical Co., St. Louis, MO), or buffer containing the salts
in preparations of hCG. The CR 127 was available in limited quantities and was
used only in the preliminary experiments. To avoid cell death related to low FBS
levels, H-UVE cells were cultured in RPMI-1640 medium with 10% FBS for 24
hours, then in medium containing 15% FBS for 24 hours, and finally with 10%
FBS and preparations of hCG or buffer control for the indicated intervals de-
scribed above. Cells were harvested for genomic DNA isolation (Stratagene, La
Jolla, CA) or for monitoring protein expression.
Staining for actin and chromatin of cells treated with crude preparations
of hCG, CR 127, or partially purified HAF fractions. One hundred thousand
KS Y-1 cells, KS SLK or AIDS-KS4 cells, or 104 H-UVE cells were seeded onto
gelatinized glass chamber slides (200 mm2) and incubated for 24 hours with 500
IU/mL of preparations of hCG or buffer. The slides were fixed with formalin and
serially treated with Triton X-100 (0.01%) for 10 minutes at 25 C, with 0.4
g/mL of fluorescein isothiocyanate (FITC)-labeled Phalloidin, which binds
actin (Sigma Chemical Co.) for 30 minutes, washed with PBS, and then stained
DNA with propidium iodide (0.5 g/mL) (Sigma Chemical Co.) for 15 minutes
(18). The stained slides were washed and mounted with Slow Fade (Molecular
Probes Inc., Eugene, OR) and examined (1000) through a dual emission filter
XF 53 (Omega, Brattleboro, VA) by fluorescence microscopy. Also, KS Y-1
cells cultured in chamber slides were treated for 24 hours with clinical-grade
crude preparations of hCG and equivalent amounts of CR127, recombinant hCG,
TSH, or with chromatographic fractions 48, 67, and 76 (the latter two contain
anti-KS clonogenic activity) from human pregnancy urine concentrates (16), and
fixed according to instructions described above, but without Triton X-100. The
slides were then stained with propidium iodide and examined for chromatin and
cellular morphology under confocal microscopy.
Protein expression in cells undergoing apoptosis after treatment with
preparations of hCG. KS Y-1, KS SLK, and H-UVE cells were synchronized
by pulsing with FBS and treated with clinical-grade crude preparations of hCG
as described above. After 24 hours of incubation with preparations of hCG, the
cells were washed with PBS and lysed in buffer containing 1% Nonidet and the
protease inhibitors leupeptin and aprotinin. The lysate was clarified by centrifu-
gation (11 750g) at 4 C for 15 minutes. Clarified lysate (100 g protein) was
size separated by reducing 10% sodium dodecyl sulfatepolyacrylamide gel
electrophoresis and transferred to nitrocellulose. Filters were blotted with anti-
bodies to c-Myc (9E10) (Santa Cruz Biotech, CA), retinoblastoma (RB) protein
(Santa Cruz Biotech), Bcl-2 (Dakopatts, Carpenteria, CA), and c-Rel (Santa Cruz
Biotech) or with respective isotype control antibody followed by incubation with
peroxidase-conjugated second antibody, and detection by chemiluminesence
analysis (Amersham Life Science Inc., Arlington Heights, IL). To determine
whether HAF-induced cellular proteins can by themselves induce cell death,
c-Myc was expressed in cells using plasmid LTRhmyc. KS Y-1 cells (1 107)
were transfected with pLTRhmyc or vector alone (19) by electroporation (Gene
Pulser; Bio-Rad Laboratories, Hercules, CA) and monitored for c-Myc expres-
sion by western blot analysis (using 9E10 antibody), and for cell death by
morphology as described for preparations of hCG-treated cells. Cells were also
synchronized with FBS as described above and analyzed for apoptosis after
incubation for 48 h with either 0100 ng/mL anti-Fas antibody or isotype control
antibody (Upstate Biotechnology, Inc., Lake Placid, NY).
Fluorescence flow cytometry. Preparations of hCG- or buffer-treated KS Y-1
cells were dislodged from monolayer cultures by gentle pipetting with cell
dissociation buffer (Life Technologies, Inc. [GIBCO BRL], Gaithersburg, MD).
To measure DNA content, cells were washed in PBS and fixed in 2% paraform-
aldehyde for 10 minutes at 4 C followed by ethanol/PBS (70%) for 2 hours at
20 C. The fixative was removed by centrifugation and the cells were washed
twice in PBS, incubated with ribonuclease A (10 g/mL) (Sigma Chemical Co.)
for 20 minutes at 37 C, stained with propidium iodide (5 g/mL) at 28 C, and
cells scanned by flow cytometry (see below). For detection of proteins, the
paraformaldehyde (2%) fixed and ethanol-treated cells were incubated with an-
tibodies targeting c-Myc (9E10) (Santa Cruz Biotech), RB (Santa Cruz Biotech),
Bcl-2 (Dakopatts), c-Rel (Santa Cruz Biotech), Fas (CH-11, Upstate Biotech-
nology, Inc.), or their respective isotype control antibody for 30 minutes in PBS
and then incubated with FITC-labeled second antibody (Boehringer Mannheim
Corp.). Baseline activity was assigned by staining cells with isotype control
antibody. Analysis for DNA content and protein expression was accomplished
on a FACScan flow cytometer using Lysis II software (Becton Dickinson, Ruth-
erford, NJ).
Statistical analysis. Where appropriate, the means are shown with standard
deviation and representative experiments presented. The Students t test was
applied to estimate the statistical significance of difference (20). All P values
were two-sided.
RESULTS
Oligonucleosomal DNA Cleavage and In Situ Apoptosis in
KS Tumors After Treatment With Preparations of hCG
To demonstrate the effects of clinical-grade crude hCG (HAF
in the preparations of hCG), KS tumors of equal size (5 5 mm)
induced by the inoculation of KS Y-1 cells in beige nude mice
were treated intralesionally with preparations of hCG (500 IU)
or buffer daily for 5 days. The buffer-treated lesions in three
mice showed continued tumor progression (>5 5 mm),
whereas three mice treated with intralesional preparations of
hCG showed near-complete tumor regression (except that acel-
lular bands of tissue remained) confirming the anti-KS effect of
preparations of hCG observed in previous studies (10). To ex-
amine the early effects of preparations of hCG on tumors, KS
Y-1 and KS SLK cell-induced tumors and surrounding mouse
tissues were treated with preparations of hCG (500 IU) or with
buffer for 2 days. On the third day the tumor sites and surround-
ing tissues and other control tissues were resected and genomic
DNA extracted. DNA isolated from buffer-treated tumors and
from endothelium-rich tissue (aorta) treated with preparations of
hCG (Fig. 1) showed high-molecular-weight DNA (>23 kilo-
base) as determined on ethidium bromide-stained agarose gel.
The DNA of tumors treated with preparations of hCG showed
DNA in an oligonucleosomal fragmentation pattern typi-
cal of programmed cell death (Fig. 1). Some of the preparations
of hCG-treated tumors showed minimal detectable fragmented
DNA that was not easily visualized (Fig. 1, left lane; KS Y-1
with 500 IU) while the TUNEL assay showed enhanced apop-
tosis (see below), indicating that detection of apoptosis by ge-
nomic DNA examination was a less-sensitive technique. Thus,
Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999
Fig. 1. DNA isolated from tumors
treated with preparations of human
chorionic gonadotropin (hCG) or
with buffer. Tumors were induced in
beige nude mice by the inoculation of
Kaposis sarcoma (KS) Y-1 and KS
SLK cells. Genomic DNA was iso-
lated from tumors and mouse tissues
48 hours after intralesional treatment
with preparations of hCG or buffer.
Five micrograms of genomic DNA
was size fractionated in a 1.2% aga-
rose gel and stained with ethidium
bromide.
there was evidence that the regression of KS tumors induced by
Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015
the preparations of hCG was associated with oligonucleosomal
DNA fragmentation.
To analyze the early mechanism leading to tumor regression,
formalin-fixed slides of tumor tissues were stained with hema-
toxylineosin for morphologic analysis. Tumors treated with
preparations of hCG showed hypocellular areas (Fig. 2, A, X)
with a few densely staining cells, but no inflammatory cell in-
filtration. In contrast, surrounding normal mouse cells (Fig. 2, A,
Z) did not show depletion of cells or did not show cell mor-
phology suggestive of an occurrence of apoptosis despite
equivalent exposure of tissues to preparations of hCG. To further
assess the mechanism of tumor regression induced by prepara-
tions of hCG, tumors were analyzed for the presence of DNA
termini by using digoxigenin-deoxyuridine triphosphate nucleo-
tide incorporation in TUNEL assay (Fig. 2, B, X) (9). Peroxidase
activity (indicative of the presence of DNA termini) was mark-
edly evident within residual small cells in hypocellular areas of
tumors treated with preparations of hCG. Notably, there was no
extracellular staining, indicating that the DNA termini generated
were retained in cells or intact cell fragments, which is consis-
tent with apoptosis and not necrosis, to account for depletion of
cells. The aforementioned cell-associated high peroxidase activ-
ity and morphologic changes were not observed in mouse tissues
similarly treated with preparations of hCG (Fig. 2, B, Z) or in
buffer-treated tumors (Fig. 2, C). Specifically, the treated tumors
contain areas showing on average (standard deviation [SD])
6.5% (1.3%), 7.3% (2.2%), 7.5% (1.9%), and 8.0% (1.4%)
positive staining cells per high-power field, whereas prepara-
tions of hCG-treated adjacent mouse tissues showed 1%
(0.8%), 1% (0.8%), 0.2% (0.5%), and 0.5% (1%) positive
cells per high-power field, indicating significantly higher apop-
totic rates in treated tumors (P<.05). Buffer-treated tumors con-
tained 2.5% (1.3%) and 2% (1.4%) positive cells per high-
power field (Fig. 2, C). Control rat gut tissue sections
concurrently stained in TUNEL assay showed the presence and
absence of apoptosis indicated by peroxidase labeling in cells
residing in the tip and the base of villi, respectively (Fig. 2, D).
Thus, treated tumors showed significantly higher rates of apop-
tosis over control buffer-treated tumors. These observations ac-
count for HAF-related KS lesion regression.
ARTICLES 137
Fig. 2. Tumors induced in nude mice by inocu-
lation of Kaposis sarcoma (KS) Y-1 cells un-
dergo apoptosis after treatment with preparations
of human chorionic gonadotropin (hCG). Repre-
sentative micrographs of sections of preparations
of hCG-treated and buffer-treated KS Y-1 tumor
in nude mice are shown. A) Shows hypocellular
area (X) with pyknotic nuclei. These morpho-
logic changes were not observed in unaffected
mouse tissue (Z). B) The similarly treated tis-
sues analyzed for the presence of DNA termini
by terminal deoxynucleotidyl transferase-
mediated digoxigenin deoxyuridine triphosphate
nick-end-labeling (TUNEL) assay show some
tumor cells (X) staining strongly positive with
the smallest cell bodies containing more intense
activity, whereas preparations of hCG-treated
mouse tissues (Z) did not show activity. C)
Buffer-treated tumor and mouse tissues showed
much reduced rates of apoptosis by the TUNEL
assay. D) For positive and negative results, rat
gut in TUNEL assay shows positive and nega-
tive cells at the tip and at the base of villi, re-
spectively. The brown immunoperoxidase reac-

Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015

tion product indicates positive detection of
significant levels of DNA termini.

DNA Fragmentation in KS Y-1 and KS SLK Cells Treated
With Preparations of hCG
To determine whether the anti-KS activity of preparations of
hCG is a direct effect, tumor and control cells in culture were
treated with preparations of hCG (400 IU/mL) for 48 hours.
Genomic DNA was isolated from cells and size-fractionated by
electrophoresis. The DNA from KS Y-1, KS SLK, and AIDS-
KS4 cells, after treatment with preparations of hCG, was frag-
mented in an oligonucleosomal pattern, whereas DNA from
similarly treated control H-UVE (21) and Bt 483 cells, as well as
buffer-treated KS cells, was of high molecular weight (Fig. 3).
Treatment of KS Y-1 and KS SLK cells with highly purified
hCG (CR 127) did not induce cell death (see below) suggesting
selective HAF activity in clinical-grade crude preparations of
hCG was not due to hCG.
Morphologic Changes Induced by Preparations of hCG
To characterize (with enhanced resolution) the early steps
accompanying apoptosis, cells were monitored by epifluores-
cence microscopy 24 hours after treatment with preparations of
hCG. Fig. 4 shows representative morphology of the buffer-
treated cells, which contain a relatively high, cytoplasmic-to-
Fig. 3. Preparations of human chorionic
gonadotropin (hCG) induce oligonucleo-
somal DNA fragmentation of Kaposis sar-
coma (KS) cells in culture. KS Y-1 and KS
SLK cells, acquired immunodeficiency
syndrome (AIDS)-KS and control primary
cells, human umbilical vein endothelial (H-
UVE) and breast cancer (Bt-483) cells
were treated for 48 h with preparations of
hCG 400 IU/mL (+) or buffer (). Geno-
mic DNA was isolated and electrophoreti-
cally separated in a 1.2 % agarose gel and
stained with ethidium bromide.
nuclear size ratio (Fig. 4, AD). Buffer-treated cells showed
sharply defined nucleoli and well-defined filamentous actin of
the cytoskeleton that was distributed widely throughout the cy-
toplasm. In contrast, KS Y-1, KS SLK, and AIDS-KS4 cells
treated with preparations of hCG (Fig. 4, EG) exhibited overall
smaller cell size and contained small nuclei with focal areas of
intense homogeneous nuclear staining. Similarly treated H-UVE
cells did not show the morphologic changes found in KS Y-1
cells treated with preparations of hCG (Fig. 4, H) as observed by
others (21). This acquired morphology is found in the earliest
recognizable phases of apoptosis (18). In addition, cells treated
with preparations of hCG displayed a low cytoplasmic-to-
nuclear size ratio and a reduced content of filamentous actin that
was poorly defined and preferentially retracted to the perinuclear
area as observed during early apoptosis (18). Moreover, exami-
nation by confocal microscopy, revealed a twofold to fourfold
reduction in the cross-sectional nuclear area and in the overall
cell area of treated KS Y-1 cells compared with buffer-treated KS
Y-1 cells. Thus, HAF in clinical-grade crude preparations of hCG
induces prototypical early cell morphology changes of apoptosis.
To begin to identify HAF that is present in human pregnancy
urine, KS Y-1 cells were selected for testing the limited avail-
able HAF material isolated in chromatography-purified fractions
from urine. The KS Y-1 cells were treated with HAF fractions
for 48 hours and cells labeled with propidium iodide. Commer-
cial preparations of hCG (Fig. 5, B) and urine HAF fractions 67
(Fig. 5, E) and 76 (Fig. 5, F) induced reduction in cell size and
diffuse staining and clumping of chromatin, which is consistent
with apoptosis. Insufficient material in HAF fractions 67 and 76
prevented further evaluation of these fractions. In contrast,
buffer treatment (Fig. 5, A), CR 127 (Fig. 5, C), and urine
fraction 48 (Fig. 5, D) did not acquire morphology of apoptotic
cells even with a much higher concentration (188 g/mL), sug-
gesting specific anti-KS activity in fractions 67 and 76. In ad-
ditional experiments, treatment of KS SLK and KS Y-1 cells
with CR 127, recombinant hCG, or TSH did not induce changes
in cell number over 5 days, indicating lack of anti-KS activity.

138 ARTICLES Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999
Fig. 4. Preparations of human chorionic gonadotropin (hCG) induce morpho- with preparations of hCG (H) contain a high cytoplasm to nuclear ratio and
logic alterations consistent with apoptosis. Cells were seeded in chamber slides well-defined filamentous actin distributed widely throughout cytoplasm. In con-
and incubated with buffer or preparations of hCG for 24 hours. The cells were trast, Kaposis sarcoma (KS) Y-1, KS SLK, and acquired immunodeficiency

Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015

fixed and serially stained for actin and chromatin with fluorescein isothiocya- syndrome (AIDS)-KS4 cells treated with preparations of hCG (E-G) exhibited
nate-labeled Phalloidin and propidium iodide, respectively, and observed under contraction of cytoplasm, accumulation of poorly defined filaments in the peri-
fluorescence microscopy as outlined in the Methods section. The buffer- nuclear area, and a lower cytoplasm to nucleus ratio, all changes characteristic
treated cells (A-D) and human umbilical vein endothelial (H-UVE) cells treated of early apoptotic changes (18).

Fig. 5. Pregnancy urine-derived human chorionic gonadotropin

(hCG)-associated factor (HAF), preparations of hCG (clinical-
grade crude), but not highly purified hCG, induce apoptosis of
Kaposis sarcoma (KS) Y-1 cells as visualized through confocal
microscopy. A) KS Y-1 cells were treated with buffer, B) with
clinical-grade preparations of hCG (Wyeth Ayerst Laboratories,
Philadelphia, PA), C) CR 127, D) or partially purified HAF from
urine fractions 48, E) fraction 67, or F) fraction 76, and stained with
propidium iodide. Panels show normal heterogeneous DNA stain-
ing of large nuclei in cells treated with buffer, CR 127 and urine
fraction 48. In contrast, cells treated with preparations of hCG,
partially purified HAF from urine fractions 67 or 76 showed ho-
mogeneous DNA staining in a clump pattern in small nuclei. After
incubation of cells for 24 hours with the different treatments, the
cells were fixed and stained with propidium iodide and visualized
under confocal microscopy as outlined in the Methods section.

Fractions 67 and 76 also abrogated formation of KS Y-1 cell foci
in clonogenic assays, as shown previously (16). Fractions 67 and
76 do not contain hCG heterodimer, hCG subunits, or -core of
hCG. Thus, partial purification of HAF from urine of pregnant
women contains apoptosis-inducing factor(s) that are not attrib-
uted to hCG, its subunits, or -core of hCG.
Protein Expression in Apoptosis Induced by Preparations
of hCG
Because preparations of hCG are known to stimulate the
growth of some cells and induce apoptosis in others, we antic-
ipate that preparations of hCG-stimulated proliferation-
associated pathways in KS cells preferentially operate in these
cells to accomplish programmed cell death (2126). After treat-
ment with clinical-grade crude preparations of hCG (400 IU/
mL) for 24 hours, extracts of KS Y-1 and KS SLK cells were
examined for proliferation-associated protein expression by
western blot analysis. While buffer-treated cells and cells with
no additional treatment contained undetectable levels of c-Myc,
both KS Y-1 and KS SLK cells when treated with preparations
of hCG showed markedly enhanced levels of c-Myc (Fig. 6, A).
In parallel studies, treatment of KS Y-1 cells with preparations

Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999 ARTICLES 139

Fig. 6. Preparations of human chorionic gonadotropin (hCG) induce expression
of c-Myc and c-Rel and accumulation of the hypophosphorylated form of the
retinoblastoma (RB) protein. A) Treatment of cells with preparations of hCG
(400 IU/mL) induced elevated c-Myc protein levels (arrow) in the Kaposis
sarcoma (KS) Y-1 and KS SLK cell lines but not in endothelial cells (data not
shown). B) Preparations of hCG treatment (400 IU/mL) also induced KS Y-1
cells to express c-Rel protein. Similar analyses show that preparations of hCG
induced a reduction of the levels of Bcl-2 protein and no changes in the p53
protein levels (data not shown). Preparations of hCG treatment also induced
hypophosphorylation of the RB protein as shown by the preferential loss of the
higher molecular weight band (RB). KS Y-1 and KS SLK cells were treated with
preparations of hCG (200 or 400 IU/mL) or buffer for 24 hours and lysates of
cells were analyzed by western blots using the indicated antibodies as described
in the Methods section.
of hCG (400 IU/mL) induced c-Rel protein expression, indicat-
ing a coordinated expression of at least two proto-oncogenes
known to participate in the same signaling pathway (23) (Fig. 6,
B). Additionally, KS Y-1 cell expression of RB protein did not
show changes in the overall levels of RB after treatment with
preparations of hCG (data not shown). However, there was a
preferential reduction in the hyperphosphorylated form (higher
molecular weight of the two bands) of RB in KS Y-1 cells (Fig.
6, B) after treatment with preparations of hCG. The diminished
hyperphosphorylated levels of RB is consistent with accumula-
tion of cells in G0/G1 phase and this type of cell cycle regulation
is characteristic of certain apoptotic pathways (27). Moreover,
western blot analysis showed that treatment with preparations of
hCG lowered Bcl-2 levels (by 50%), whereas the levels of p53
remained unchanged (data not shown), as commonly noted in
programmed cell death induced by the nongenotoxic type of
stimuli (22). Thus, these data indicate HAF ultimately induces
apoptosis following expression of proliferation-associated pro-
teins as observed with other mitogen-triggered apoptosis.
To better determine c-Myc expression in the apoptotic pro-
cess induced by preparations of hCG, we examined whether
c-Myc expression in itself was sufficient to stimulate pro-
grammed cell death. To test this, KS Y-1 cells were electropor-
ated with pLTRhmyc and cells evaluated for c-Myc expression
(at 24 hours) by western blot analysis. These pools of cells were
140 ARTICLES Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999
examined for cell death by following the same protocols for
preparations of hCG treatment of cells. Whereas the three
LTRhmyc-transfected pools contained an average (SD) of 23
(13.1) cells per high-power field with morphology consistent
with apoptosis, the vector-alone transfected cells and mock-
transfected cells contained 5.2 (6.4) and 4 (8.1) apoptotic cells
per high-power field, respectively. Thus, induced c-Myc expres-
sion stimulates apoptosis, suggesting that c-Myc may participate
in programmed cell death in KS cells after treatment with prepa-
rations of hCG.
KS Y-1 and KS SLK cells (Fig. 7, A and B, respectively)
were found to express modest levels of Fas, as shown by the
shift in black peak from white peak (isotype antibody control)
with anti-fas antibody. We tested whether these cells were sus-
ceptible to Fas-mediated cell killing by treating cells with anti-
Fas antibody (0100 L of CH-11) that induces apoptosis but
did not observe enhanced apoptosis (<2% apoptotic cells in both
antibody-treated cells and isotype control antibody-treated cells).
Given these negative results, it is likely that HAF-mediated apop-
tosis involves mechanisms independent of Fas receptor stimulation.
In a second strategy to analyze protein expression in HAF-
Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015
induced apoptosis, cells were monitored for DNA and proto-
oncogene content by flow cytometry analysis. KS Y-1 cells
treated with preparations of hCG and stained with propidium
iodide contained a subset of cells with hypodiploid levels of
DNA content (Fig. 7, D) that was absent in buffer-treated cells
(Fig. 7, C). The cells with hypodiploid DNA content were of
relatively smaller size as shown by the epifluorescence studies
described above and findings are consistent with an apoptotic
mechanism. The cells treated with preparations of hCG also
preferentially accumulated in G0/G1 phase (Fig. 7, D). These
observations are in agreement with accumulation of hypophos-
phorylated RB protein (Fig. 6, B). Antibody staining of cells
treated with preparations of hCG exhibited a higher number of
cells expressing c-Myc (37%78%), c-Rel (27%48%), and un-
changed levels of RB protein (Table 1). In addition, treated cells
contained fewer cells (by 50%) expressing Bcl-2. These results
indicate coordinated protein expression as well as cell cycle
modulation in HAF-induced programmed cell death.
DISCUSSION
Our initial observation linking anti-KS activity with factors
associated with early pregnancy in mice (10) has led to these
studies that attribute novel apoptotic HAF activity present in
some clinical-grade crude preparations of hCG (which is com-
mercially prepared from urine of pregnant women) and in urine
concentrates from pregnant women prepared in our laboratory.
Here, we show HAF induces KS cell programmed cell death in
cell culture and in vivo as demonstrated by oligonucleosomal
DNA fragmentation and the selective accumulation of abundant
DNA termini in individual tumor cells in culture and in a KS
tumor in an animal model. Cells treated with preparations of
hCG undergo typical cellular morphologic changes that are fea-
tures of cells that undergo apoptosis. This includes nuclear con-
densation and global cell shrinkage with dissolution of the pe-
ripheral actin cytoskeleton (18). Apoptosis follows up-regulated
c-Myc and c-Rel protein levels and a simultaneous reduction of
Bcl-2 levels that were observed both by western blot analysis
and flow cytometry analysis. Aside from our initial studies de-
scribing the anti-KS effect of clinical-grade crude preparations
of hCG, others (21) have noted KS cell apoptosis with nonclini-
Fig. 7. Kaposis sarcoma (KS) Y-1 and KS SLK
cells express Fas and preparations of human chori-
onic gonadotropin (hCG) induce apoptosis by flow
cytometry analysis. A) KS Y-1 cells and B) KS SLK
cells express modest levels of Fas by flow cytometry
analysis. The open peak indicates activity with iso-
type control antibody whereas the black (filled peak)
indicates activity with anti-Fas antibody. Prepara-
tions of hCG treatment reduce the DNA content of
KS Y-1 cells (C and D). Whereas buffer-treated
cells (C) contain more than 95% of cells with at least
2N DNA complement, the preparations of hCG-
treated cells (D) exhibit a substantial higher number
of cells with lower DNA content. Cells treated with
preparations of hCG also preferentially accumulate
in G0/G1. For Fas expression, the cells were fixed
and stained with anti-Fas antibody (CH-11) and sub-
sequently with fluorescein isothiocyanate-labeled
antibody as described in the Methods section.
For DNA content analysis, cells were treated with
buffer or preparations of hCG for 48 hours and were
fixed and incubated with propidium iodide as de-
scribed in the Methods section. The cells were
then monitored for DNA content by flow cytometry

Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015

analysis. For panels C and D, the vertical axis indi-
cates the relative cell number and the horizontal axis
(left to right) is the relative DNA content of prop-
idium stained cells (low to high). Stained cells were-
analyzed by flow cytometry for expression of other
proteins as shown in Table 1.

Table 1. Preparations of human chorionic gonadotropin (hCG) modulates
proto-oncogene levels in Kaposis sarcoma Y-1 cells undergoing apoptosis*
Cells expressing protein, %
Protein Buffer Preparations of hCG
c-Myc 37 78
c-Rel 27 48
Bcl-2 46 22
Retinoblastoma protein 86 82
*Proto-oncogene expression in KS Y-1 cells treated with preparations of hCG
were examined by flow cytometry. KS Y-1 cells were treated with preparations
of hCG (400 IU/mL) or buffer for 24 hours. The cells were fixed and stained
with antibodies for targeting proteins indicated and examined by flow cytometry
analysis as described in the Methods section. The results are from represen- teolytic cleavage in vivo and can potentially produce peptides
tative studies and indicate the percentage of positive staining cells over signal
generated with control isotype antibodies.
cal-grade preparations of hCG, as well as reduced cell prolifera-
tion (28).
The newly described apoptotic effect attributed to HAF re-
sembles the apoptosis pathway best followed by a number of
growth factors related to programmed cell death. In various cell
systems, extracellular ligands and hormones that stimulate cell
growth pathways can also, under the restricted conditions, in-
duce apoptosis. Because preparations of hCG can induce prolif-
eration and/or induce c-Myc in some normal and neoplastic cells
(26,29), it was anticipated to precipitate apoptosis, perhaps by
diverting an activated cell proliferation pathway to a cell death
end point (2325). Here, HAF-induced c-Myc and c-Rel expres-
sion, as well as induced c-Myc expression in transfectants, was
followed by KS cell apoptosis, suggesting participation of these
transcription factors in HAF-induced cell death. Preparations of
hCG-stimulated c-Myc and c-Rel concur with results from other
laboratories showing regulation of AP-1 complex formation in
HAF-induced apoptosis (28,30). Even though c-Myc-related ap-
optosis has extensively been described in cell culture, a study
(25) implicates c-Myc in apoptosis as a mechanism in cancer
regression in vivo.
The identity of HAF is unknown. Fractions of preparations of
hCG containing HAF activity are distinct from hCG, its sub-
units, or nick forms and is characterized as a protein by dena-
turing and protein digest experiments (16). Others (31) have
found ribonuclease isolated from preparations of hCG to contain
anti-KS activity; however, the purified fractions with HAF pre-
pared in our laboratory lack ribonuclease. Fractions containing
HAF-induced apoptosis of KS cells were not blocked by antiri-
bonuclease antibodies. Interestingly, hCG is susceptible to pro-
that activate receptors. The subunit of hCG can be
potentially cleaved to form hCG AA 3757. Synthetic forms of
this peptide form amphipathic helical structures that can
mimic the hCG heterodimer in activating its cognate receptor,
albeit at several-fold higher concentrations (32).
Since HAF activity is found in urine in early pregnancy (16),
its presence may be linked to the substantially lower rates of
AIDS-KS in women versus men (33). HIV-1-infected women
have lower rates of HHV8 infection and KS compared with
HIV-1-infected men. Even though nonpregnant women already
exhibit lower rates of KS, one would predict that pregnancy via
HAF would offer enhanced protection against development of
KS or HHV8 infection. However, a single retrospective study by
Rabkin et al. (34) in a small number of HIV-1-infected African
women found that pregnancy was not associated with lower KS
rates. For other types of cancers, such as mammary cancer in
humans and rats, the pregnancy state does confer subsequent
lower rates of tumor occurrence. During pregnancy, breast tissue
undergoes differentiation that renders breast tissue cells resistant
to transformation (35). In contrast, the HAF effect we observed

Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999 ARTICLES 141
with KS cells and preparations of hCG is primarily an apoptotic
mechanism. Although stimulated cell differentiation can subse-
quently include apoptosis in some cells (36) and a transient
cell differentiation effect in our studies cannot be ruled out,
the overwhelming evidence suggests that HAF induces an
immediate and primarily apoptotic effect. Indeed, apoptosis
appears to be the basis for the demonstrated high remission rates
(>80%) of intralesional preparations of hCG in a phase II trial of
human KS (14) as well as high remission rates of the more
advanced and resistant visceral KS in response to systemic treat-
ments with preparations of hCG (15). The apoptosis activity of
preparations of hCG indicates that a human-derived factor,
HAF, is directly involved in KS cell death and holds promises
for biologic-based anticancer therapy and further suggests the
existence of apoptosis regulation that has previously been unap-
preciated.
REFERENCES
(1) McNutt NS, Fletcher V, Conant MA. Early lesions of Kaposis sarcoma in
homosexual men. An ultrastructural comparison with other vascular pro-
liferations in skin. Am J Pathol 1983;111:6277.
(2) Ruszczak Z, Mayer-Da Silva A, Orfanos CE. Kaposis sarcoma in AIDS.
Multicentric angioneoplasia in early skin lesions. Am J Dermatopathol
1987;9:38898.
(3) Zhang YM, Bachmann S, Hemmer C, van Lunzen J, von Stemm A, Kern
P, et al. Vascular origin of Kaposis sarcoma. Expression of leukocyte
adhesion molecule-1, thrombomodulin, and tissue factor. Am J Pathol
1994;144:519.
(4) Staskus KA, Zhong W, Gebhard K, Herndier B, Wang H, Renne R, et al.
Kaposis sarcoma-associated herpesvirus gene expression in endothelial
(spindle) tumor cells. J Virol 1997;71:7159.
(5) Samaniego F, Markham PD, Gendelman R, Watanabe Y, Kao V, Kowalski
K, et al. Vascular endothelial growth factor and basic fibroblast growth
factor present in Kaposis sarcoma (KS) are induced by inflammatory
cytokines and synergize to promote vascular permeability and KS lesion
development. Am J Pathol 1998;152:143343.
(6) Samaniego F, Markham PD, Gallo RC, Ensoli B. Inflammatory cytokines
induce AIDSKaposis sarcoma-derived spindle cells to produce and
release basic fibroblast growth factor and enhance Kaposis
sarcoma-like lesion formation in nude mice. J Immunol 1995;154:
358292.
(7) Samaniego F, Markham PD, Gendelman R, Gallo RC, Ensoli B. Inflam-
matory cytokines induce endothelial cells to produce and release basic
fibroblast growth factor and to promote Kaposis sarcoma-like lesions in
nude mice. J Immunol 1997;158:188794.
(8) Samaniego F, Gallo RC. Immunopathogenesis of Kaposis sarcoma. In:
Gupta S, editor. Immunology of HIV infection. New York (NY): Plenum
Press; 1996. p. 43750.
(9) Lunardi-Iskandar Y, Gill P, Lam VH, Zeman RA, Michaels F, Mann DL,
et al. Isolation and characterization of an immortal neoplastic cell line (KS
Y-1) from AIDS-associated Kaposis sarcoma. J Natl Cancer Inst 1995;87:
97481.
(10) Lunardi-Iskandar Y, Bryant JL, Zeman RA, Lam VH, Samaniego F, Bes-
nier JM, et al. Tumorigenesis and metastasis of neoplastic Kaposis sar-
coma cell line in immunodeficient mice blocked by a human pregnancy
hormone [published erratum appears in Nature 1995;376:447]. Nature
1995;375:648.
(11) Siegal B, Levinton-Kriss S, Schiffer A, Sayar J, Engelberg I, Vonsover A,
et al. Kaposis sarcoma in immunosuppression. Possibly the result of a dual
viral infection. Cancer 1990;65:4928.
(12) Popescu NC, Zimonjic DB, Leventon-Kriss S, Bryant JL, Lunardi-Iskandar
Y, Gallo RC. Deletion and translocation involving chromosome 3 (p14) in
two tumorigenic Kaposis sarcoma cell lines. J Natl Cancer Inst 1996;88:
4505.
(13) Moore PS, Chang Y. Detection of herpesvirus-like DNA sequences in
142 ARTICLES
Kaposis sarcoma in patients with and without HIV infection. N Engl J Med
1995;332:11815.
(14) Gill PS, Lunardi-Iskandar Y, Louie S, Tulpule A, Zheng T, Espina BM, et
al. The effects of preparations of human chorionic gonadotropin on
AIDS-related Kaposis sarcoma [published errata appear in N Engl J Med
1997;336:670 and 1997;336:1115]. N Engl J Med 1996;335:12619.
(15) Hermans P, Clumeck K, Picard O, Van Vooren JP, Duriez P, Zucman D,
et al. AIDS-related Kaposis sarcoma patients with visceral manifestations:
response to human chorionic gonadotropin preparations. J Hum Virol 1998;
1:829.
(16) Lunardi-Iskandar Y, Bryant JL, Blattner WA, Hung CL, Flamand L, Gill P,
et al. Effects of a urinary factor from women in early pregnancy on HIV-1,
SIV and associated disease. Nature Med 1998;4:42834.
(17) Flamand L, Zeman RA, Bryant JL, Lunardi-Iskandar Y, Gallo RC. Ab-
sence of human herpesvirus 8 DNA sequences in neoplastic Kaposis sar-
coma cell lines. J Acquir Immune Defic Syndr Hum Retrovirol 1996;13:
1947.
(18) Endresen PC, Prytz PS, Aarbakke J. A new flow cytometry method for
discrimination of apoptotic cells and detection of their cell cycle speci-
ficity through staining of F-actin and DNA. Cytometry 1995;20:
16271.
(19) Kelekar A, Cole MD. Tumorigenicity of fibroblast lines expressing the
adenovirus E1a, cellular p53, or normal c-myc genes. Mol Cell Biol 1986;
6:714.
Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015
(20) Daniel WW. Comparisons of means. In: Biostatistics: a foundation for
analysis in the health sciences. New York (NY): John Wiley & Sons; 1978;
p. 12059.
(21) Lang ME, Lottersberger C, Roth B, Bock G, Recheis H, Sgonc R. Induction
of apoptosis in Kaposis sarcoma spindle cell cultures by the subunits of
human chorionic gonadotropin. AIDS 1997;11:133340.
(22) Kyprianou N, English HF, Davidson NE, Isaacs JT. Programmed cell death
during regression of the MCF-7 human breast cancer following estrogen
ablation. Cancer Res 1991;51:1626.
(23) Lee H, Arsura M, Wu M, Duyao M, Buckler AJ, Sonenshein GE. Role of
Rel-related factors in control of c-myc gene transcription in receptor-
mediated apoptosis of the murine B cell WEHI 231 line. J Exp Med 1995;
181:116977.
(24) Hermeking H, Eick D. Mediation of c-Myc-induced apoptosis by p53.
Science 1994;265:20913.
(25) Alarcon RM, Rupnow BA, Graeber TG, Knox SJ, Giaccia AJ. Modulation
of c-Myc activity and apoptosis in vivo. Cancer Res 1996;56:43159.
(26) Yoshimura M, Nishikawa M, Mori Y, Yoshikawa N, Horimoto M, Toyoda
N, et al. Human chorionic gonadotropin induces c-myc mRNA expression
via TSH receptor in FRTL-5 rat thyroid cells. Thyroid 1992;2:
3159.
(27) Dou QP, An B, Will PL. Induction of a retinoblastoma phosphatase activity
by anticancer drugs accompanies p53-independent G1 arrest and apoptosis.
Proc Natl Acad Sci U S A 1995;92:901923.
(28) Kachra Z, Guo WX, Sairam MR, Antakly T. Low molecular weight com-
ponents but not dimeric HCG inhibit growth and down-regulate AP-1 tran-
scription factor in Kaposis sarcoma cells. Endocrinology 1997;138:
403841.
(29) Gillott DJ, Iles RK, Chard T. The effects of beta-human chorionic gonado-
trophin on the in vitro growth of bladder cancer cell lines. Br J Cancer
1996;73:3236.
(30) Colotta F, Polentarutti N, Sironi M, Mantovani A. Expression and involve-
ment of c-fos and c-jun protooncogenes in programmed cell death induced
by growth factor deprivation in lymphoid cell lines. J Biol Chem 1998;
267:1827883.
(31) Griffiths SJ, Adams DJ, Talbot SJ. Ribonuclease inhibits Kaposis sarcoma
[letter]. Nature 1997;390:568.
(32) Keutmann HT, Charlesworth MC, Mason KA, Ostrea T, Johnson L, Ryan
RJ. A receptor-binding region in human choriogonadotropin/lutropin beta
subunit. Proc Natl Acad Sci U S A 1987;84:203842.
(33) Safai B, Johnson KG, Myskowski PL, Koziner B, Yang SY, Cunningham-
Rundles S, et al. The natural history of Kaposis sarcoma in the ac-
quired immunodeficiency syndrome. Ann Intern Med 1985;103:
74450.
(34) Rabkin CS, Chibwe G, Muyunda K, Musaba E. Kaposis sarcoma in preg-
nant women [letter]. Nature 1995;377:212.
Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999
(35) Thordarson G, Jin E, Guzman RC, Swanson SM, Nandi S, Talamantes F.
Refractoriness to mammary tumorigenesis in parous rats: is it caused by
persistent changes in the hormonal environment or permanent biochemical
alterations in the mammary epithelia? Carcinogenesis 1995;16:
284753.
(36) Martin SJ, Bradley JG, Cotter TG. HL-60 cells induced to differentiate
towards neutrophils subsequently die via apoptosis. Clin Exp Immunol
1998;79:44853.
Journal of the National Cancer Institute, Vol. 91, No. 2, January 20, 1999
NOTES
Supported in part by a research supplement to R01AI38192-02S1 from the
National Institute of Allergy and Infectious Diseases, National Institutes of
Health, Department of Health and Human Services (F. Samaniego).
We thank Anna M. Mazzuca and Stacy Cline for their editorial assistance.
Manuscript received August 28, 1998; revised November 13, 1998; accepted
November 18, 1998.
Downloaded from http://jnci.oxfordjournals.org/ by guest on April 15, 2015
ARTICLES 143