Stimulation of ovarian tumor cell proliferation with monocyte products including interleukin-l, interleukin-6, and tumor necrosis factor-a Shu Wu, MD, Kerry Rodabaugh, MD, Otoniel Martinez-Maza, PhD, Joanna M. Watson, PhD, David S. Silberstein, MD, Cinda M. Boyer, PhD, William P. Peters, MD,J. Brice Weinberg, MD,Jonathan S. Berek, MD, and Robert C. Bast, Jr., MD Durham, North Carolina, Los Angeles, California, and Boston, Massachusetts OBJECTIVE: We investigated whether monocyte-derived factors could stimulate the growth of ovarian cancer cells. STUDY DESIGN: Human peripheral blood monocytes or human monocyte-like cell lines THP-1 and U-937 were cultured with or without macrophage colony-stimulating factor, lipopolysaccharide, or phorbol myristate acetate. Culture supernatants or recombinant cytokines were assayed for growth stimulation of ovarian cancer cell lines by tritium-thymidine incorporation and direct cell counts followed by statistical analysis with Student t test. RESULTS: Conditioned medium from peripheral blood monocytes or from THP-1 or U-937 cells stimulated ovarian cancer cell growth. Interleukin-1a, tumor necrosis factor-a, and interleukin-6 also stimulated ovarian cancer cell growth, whereas macrophage, granulocyte, and granulocyte-macrophage colonystimulating factor did not. Concentrations of tumor necrosis factor, interleukin-1 , and interleukin-6 in conditioned medium could not account for all the growth stimulation, and activity remained after neutralization of tumor necrosis factor, interleukin-1, and interleukin-6 with antibodies. CONCLUSIONS: Interleukin-1, interleukin-6, tumor necrosis factor, and additional monocyte factor(s) could provide paracrine growth stimulation when monocytes are attracted to ovarian cancers that produce macrophage colony-stimulating factor. (AM J OSSTET GYNECOL 1992;166:997-1007.)

Key words: Monocyte, growth factor, ovarian cancer Epithelial ovarian cancer is the leading cause of death from gynecologic malignancy. I Macrophages are associated with ovarian cancer cells within solid tumor implants and ascites fluid. 2. 3 Understanding the interaction between ovarian tumors and host macro phages may aid in developing novel strategies for therapy. Previous work has indicated that ovarian cancer cells can produce macrophage colony-stimulating factor (CSF):' 5 which augments monocytopoiesis and acts as From the Departments of Medicine and Microbiology and Immunology, Duke Comprehensive Cancer Center, Duke University Medical Center, and the Veterans Administration Medical Center, Durham, the Departments of Obstetrics and Gynecology and Microbiology and Immunology, University of California, Los Angeles, School of Medicine, and the Department of Rheumatology and Immunology, Brigham and Womens Hospital, Boston. Supported in part fry grants 5-ROI-CA39930 and POI-CA4774I01 A2 from the National Cancer Institute, grant POI -AI-23308 from the National Institute of Allergy and Infectious Disease, grant P50AR-39162 from the National Institute of Arthritis and Rheumatism, Department of Health and Human Services. Funds were also provided fry the Veterans Administration Research Service. Received for publication May 30, 1991; revised September 10, 1991 ; accepted October 30, 1991. Reprint requests: Robert C. Bast Jr., MD, Box 3843, Duke University Medical Center, Durham, NC 27710. 611 /34744

a chemoattractant for monocytes.6 Macrophage CSF can also act on mature monocytes and macrophages, stimulating increased production of tumor necrosis factor-ex (TNF-ex).7 TNF-ex is cytotoxic or cytostatic for some but not all tumor cells. The growth of neoplastic B cells,· astrocytoma cells,9 and human osteosarcoma cells lo has been stimulated by TNF-ex. The growth of human cervical carcinoma cell lines is inhibited by high concentrations of tumor necrosis factor but is stimulated by low concentrations." Interleukin-I (IL-I), another monocyte product, can stimulate growth of a human astrocytoma cell line9 and malignant trophoblastic cells. I2 Interleukin-6 (IL-6) , produced by both macrophages and tumor cells, can stimulate growth of a human myeloma cell line and may act as an autocrine growth factor for human renal cell carcinomas. 13. .. Although it is known that monocytes or monocyte products can stimulate the growth of ovarian tumor cells and cell lines,I 5.17 the mechanism underlying this effect is not understood . The present study was undertaken to characterize monocyte-derived factor(s) that stimulate growth of ovarian carcinoma cells.

997

998 Wu et al.

March 1992 Am J Obstet Gyneco1

c: 0

~ ...

120 100

Oeo)

0..

" '.0 0 -

80

~)(

.- 60 =oE .- 0. 40 EO Q)

c: -

>--

.....r:: I

l:

eo)

20 0

OVCA 420

OVCA 429

OVCA 432

OVCA 433

Fig. 1. 'H-thymidine incorporation by ovarian carcinoma cell lines in medium conditioned by peripheral blood monocytes. Each ovarian carcinoma cell line was incubated with 0, 1%, 10% and 100% of peripheral blood monocyte-conditioned medium (bars, from left to right). Significant stimulation (asterisk) was observed with OVCA 429 (p = 0.03), OVCA 432 (p = 0.01), and OVCA 433 (p = 0.03).

Material and methods Growth factors. Recombinant macrophage CSF (2 x 106 U I mg) was generously provided by Genetics Institute (Cambridge, Mass.). Recombinant granulocyte-macrophage CSF (1 x 108 U I mg) and granulocyte CSF (1 x 108 U/mg) were provided by AMGEN (Thousand Oaks, Calif.). Recombinant TNF-a (2 x 107 U I mg) was purchased from Genzyme (Boston). Recombinant IL-Ia (5 x 108 U/mg) was purchased from R&D systems (Minneapolis). IL-6 (1 x 105 U Iml) was provided by Drs. T . Kishimoto and T. Hiramo (Osaka). Chemicals. Phorbol 12-myristate-13-acetate, lipopolysaccharide from Escherichia coli 0127: B8, and phenyl-Sepharose CL-4B were purchased from Sigma Chemical Company (St. Louis). Tritiated thymidine was supplied by New England Nuclear (Boston). Cell lines. Four ovarian carcinoma cell lines were used in this study. OVCA 420, OVCA 429, OVCA 432, and OVCA 433 cells were established from ascites tumor cells of patients with ovarian carcinoma. 18 The cells were grown in Eagle's minimal essential medium that was supplemented with 10% fetal bovine serum, 2 mmol/L L-glutamine, 100 U/ml penicillin, 100 IJ-g/ml streptomycin, 1 mmol/L sodium pyruvate, and 1% nonessential amino acid mixture. Medium was changed every 3 days and tumor cells were subcultured once a week. For subculture and experiments, monolayers were detached with 0.25% trypsin and 0.02% ethylenediaminetetraacetic acid. The human acute monocytic leukemia cell line THP-l and the human histiocytic lymphoma U-937 were obtained from the American Type Culture Collection (Bethesda, Md.). THP-l and U-937 cells were grown in RPMI 1640 medium with 10% fetal bovine serum and antibiotics. In

the case of THP-l, medium was supplemented with 5 x 10- 5 mol/L ~-mercaptoethanol. Monocytes and monocyte-conditioned media. Human peripheral blood mononuclear cells were obtained by Ficoll-Hypaque discontinuous density-gradient centrifugation and washed three times in Hanks' balanced salt solution. Mononuclear cells were incubated for 2 hours in T150 tissue culture flasks at a density of 1 x 106 cells per milliliter in culture medium consisting of RPMI 1640 supplemented with 10% fetal bovine serum and antibiotics. The nonadherent cells were removed from flasks by washing three times with Hanks' balanced salt solution. The residual adherent population (peripheral blood monocytes) contained >90% monocytes judged by morphologic features after Giemsa staining and by the ability to phagocytize 1 IJ-m latex particles. Adherent cells were incubated in culture medium or RPM I 1640 with 2 mmol/L L-glutamine and antibiotics but without fetal bovine serum. Peripheral blood monocytes were treated with different concentrations of M-CSF for 24, 48, and 72 hours or with 10 IJ-g/mllipopolysaccharide for 48 hours at 37° C in 5% carbon dioxide and 95 % humidified air. The medium was collected, centrifuged at 205g for 10 minutes, and passed through 0.20 IJ-m filters. Supernatants were stored at - 20° C until used for experiments. Media conditioned by monocyte-derived cell lines. THP-l and U-937 cells were grown in the media described above with 10% fetal bovine serum. Cells were sedimented by centrifugation at 205g for 10 minutes at 20° C, washed three times with Hanks' balanced salt solution, and treated with 10 IJ-g/mllipopolysaccharide and 10 ngl ml of phorboI12-myristate-13-acetate in culture medium or serum-free medium at a density of 5 X 105 cells I milliliter. After 2 days the supernatants

Monocyte factors stimulate ovarian cancer 999

Volume 166 Number 3

20

c

0 :;:;

*

CO

'-

0(') ~I

15

'-0 0..-

(,)

)(

.5 10 CD c:cE .EO >-- 5 .c ~

lI

J:

(')

0

Control

M-CSF

PBM undiluted supernatant

PBM+M-CSF undiluted supernatant

Fig. 2. Effect of macrophage CSF (M-CSF) on ability of peripheral blood monocytes (PBM) to stimulate growth of OVCA 433 cells. Peripheral blood monocyte were incubated with or without 100 ng/ml of macrophage CSF. 3H-thymidine incorporation by OVCA 433 cellls was measured in presence of each undiluted conditioned medium. Macrophage CSF (100 ng/ml) was incubated with OVCA 433 as control. Aslerisk, Significant stimulation (p < 0.05).

were harvested and stored under the same conditions described for monocyte-conditioned media. 'H-Thymidine incorporation and cell counts. Four ovarian carcinoma cell lines were incubated at a density of 5 x 10' cells per 200 .....1 of tissue culture medium in 96-welI, flat-bottomed microtiter plates. After 24 hours, the medium was removed from each well. For experiments requiring serum-free conditions, cells were washed three times with Hanks' balanced salt solution. Different concentrations of conditioned medium, cytokines, lipopolysaccharide, or phorbol 12-myristate13-acetate diluted with tissue culture medium were added in a final volume of 200 ..... 1. All assays were performed in quadruplicate. Cells were cultured for an additional 24 hours. Proliferation of ovarian tumor cells was measured by the addition of 3H-thymidine (l .....Ci per well) for the final 6 hours before harvest. After cells were labeled, culture medium was removed from each well, and cells were washed with Hanks' balanced salt solution and solubilized with 2 moll L sodium hydroxide. Cell-associated radioactivity was measured with a Packard j3-counter. For cell counts, cohort wells of ovarian carcinoma cells were incubated with growth factors and monocyte supernatant, trypsinized, and counted with a hemocytometer at 72 hours. Biochemical characterization of the monocyte-derived growth factors. Aliquots of conditioned medium were exposed to different experimental conditions, and the residual growth-stimulating activity was measured. Heat stability was evaluated by boiling the conditioned medium for 5 minutes. Acid stability was tested by treating conditioned medium with a final concentration of IN acetic acid (pH 2.6). Sensitivity to reduction or denaturation was assessed by subjecting aliquots of con-

ditioned medium to a final concentration of 5% 2mercaptoethanol, 0.1 % sodium dodecyl sulfate, or 2 moIlL guanidine hydrochloride. Conditioned medium was exposed to acid or to reduction overnight at 40 C and dialyzed for 24 hours against RPMI 1640 with a dialysis membrane that retained molecules of> 12 to 14 kd. The dialyzed media were diluted with culture medium and assayed for their ability to stimulate 'Hthymidine incorporation by human ovarian carcinoma cell lines. Measurement of IL-I, IL-6, and TNF-a in conditioned media. Concentrations of IL-l were measured by bioassay with mouse thymocytes as described. 19 IL-6 was detected with enzyme-linked immunsorbent assay as described!O TNF-ex was detected with a radioimmunoassay as described by Genzyme. Neutralization of tumor necrosis factor, IL-I, and IL-6. Diluted conditioned medium was incubated for 2 hours at room temperature with 10' to 105 neutralizing units of polyclonal rabbit antihuman TNF -ex from Genzyme or 250 to 2500 neutralizing units of goat antihuman IL-Iex from R&D Systems or 10 to 100 neutralizing units of rabbit antihuman IL-6 from Genzyme. The effect on growth of ovarian tumor cells lines was measured with the 'H-thymidine incorporation assay. The ability of anti -tumor necrosis factor, anti-IL-l, and anti-IL-6 to neutralize authentic tumor necrosis factor, IL-l, or IL-6 was measured in the same experiments. Partial purification of the monocyte-derived growth factor. Ammonium sulfate was added to conditioned culture medium to provide a final concentration of 20% before separation on a phenyl-Sepharose column that had been equilibrated with 20% ammonium sulfate.

1000 Wu et al.

March 1992

Am J Obstet Gynecol

c::

90

o

80

o~

70

:;

a..

o 52

60

~ ..->< 50 Q)

c::- 40

:sE

'E 13

30

~

20

>--

±

("I)

10

o

OVCA 420

OVCA 429

OVCA 432

OVCA 433

Fig. 3. 3H-thymidine incorporation by human ovarian carcinoma cells in medium conditioned by peripheral blood monocytes grown in presence of lipopolysaccharide. Peripheral blood monocytes were treated with 10 IJ-g/ml lipopolysaccharide. Each of four cell lines was grown in presence of 0%, 1%, 10%, and 100% conditioned medium, indicated by bars from left to right. Asterisk, Significant stimulation (p < 0.05).

c::

50 ,-------------------------------------,

0

:;:::

as 40 0("1) a. . ~

.... 0

0..-

>< 30 .E .,.... Q) c::-

>-~ 10

±

("I)

o

I

*

;' *

:s E 20

'E 13

J-·--·-l ,;""';' ,'" .J...

()

;,' .;,'

.;'

*

.;'

Control

1

10

100

Percentage of Conditioned Medium Fig. 4. 3H-thymidine incorporation by OVCA 433 cells in serum-free medium conditioned by peripheral blood monocytes in the presence (dashed line) and absence (solid line) of 10 IJ-g/mllipopolysaccharide. Asterisk, significant stimulation (p < 0.05).

After addition of the sample, the column was washed with 50 ml of 20% ammonium sulfate in phosphatebuffered saline solution and 50 ml of deionized water. Subsequently, material was eluted with 50 ml of 5% and 50 ml of 25% ethylene glycol (Sigma). Aliquots of the eluate were dialyzed in phosphate-buffered saline solution for 24 hours at 4° C, diluted with tissue culture medium, and tested for their ability to stimulate proliferation of ovarian cancer cell lines. Analysis. The Student t test was used to compare control and experimental groups. Data were expressed as mean ± SD. All assays were performed with at least four replicates. Results Growth stimulation of ovarian carcinoma cell lines by peripheral blood monocyte-conditioned media. Initial experiments were carried out to assess the effect

of peripheral blood monocyte-conditioned medium on the growth of four ovarian carcinoma cell lines, including OVCA 420, OVCA 429, OVCA 432. and OVCA 433. The growth of OVCA 429, OVCA 432. and OVCA 433 was stimulated by peripheral blood monocyte-conditioned medium (Fig. 1). The growth-stimulating activity of the peripheral blood monocyte supernatant was not enhanced by treatment of peripheral blood monocytes for 1 to 3 days with macrophage CSF (Fig. 2). Direct addition of from 1 to 1000 ng/ml macrophage CSF to ovarian cancer cell lines did not affect growth of the ovarian cancer cells and did not enhance the stimulatory effect of peripheral blood monocyte supernatants. Neither granulocyte-macrophage nor granulocyte CSF stimulated or inhibited the growth of the four ovarian carcinoma cell lines tested (data not shown). Enhancement of the growth-stimulating activity of

Monocyte factors stimulate ovarian cancer

Volume 166 Number 3

120

"?

0 ,... )( ,...

-

E 0

Q.

c: 0 :;:::;

... ...0

-

.J::.

l1:

C')

50

o

OVCA 420

OVCA 429

OVCA 432

OVCA 433

Fig. 5. A, 'H-thymidine incorporation by ovarian carcinoma cell lines in medium conditioned by THP-I cells grown in presence of 10 jJ.g/mllipopolysaccharide and 10 nglml phorboI12-myristate13-acetate. Each of four cell lines was grown in presence of 0%, 1%, 10%, or 100% conditioned medium, indicated by bars from left to right. Asterisk, Significant stimulation or inhibition (p < 0.05). B, 'H-thymidine incorporation by ovarian carcinoma cell lines with growth medium conditioned by U-937 in presence of 10 jJ.g/mllipopolysaccharide and 10 nglml phorbol 12-myristate-13-acetate. All four cell lines could be stimulated with 1% or 10% conditioned medium, indicated by bars from left to right.

peripheral blood monocytes by treatment with lipopolysaccharide. When peripheral blood monocytes were treated with 10 f.Lg/mllipopolysaccharide, supernatants once again enhanced growth in three of four ovarian carcinoma cell lines by 55% to 159% (Fig. 3). When tested in serum-free medium, greater stimulation was noted in the presence of lipopolysaccharide than in its absence (Fig. 4). Incubation of ovarian carcinoma cell lines with 10 f.Lg/mllipopolysaccharide produced slight enhancement of ' H-thymidine incorporation by the OVCA 420 (10%), OVCA 432 (25%), and OVCA 433 (19%) cell lines but not by OVCA 429 (data not shown). Growth stimulation of ovarian carcinoma cell lines with media conditioned by THP-l or U-937 in the presence of lipopolysaccharide and phorbol 12-myristate-13-acetate. The THP-I and U-937 cell lines were established from human neoplasms of monocytic lineage. To enhance production of ovarian cancer growth-stimulating activity, both cell lines were treated with phorbol 12-myristate-13-acetate and lipopolysaccha ride. The growth of all four ovarian carcinoma cell lines was significantly stimulated by low concentrations of THP-I and U-937 products (Fig. 5, A and B). High

concentrations of conditioned medium could, however, inhibit the growth of all four ovarian carcinoma cell lines (Fig. 5, A and B). In serum-free medium the growth of the four cell lines was stimulated and supernatants were active at a higher dilution (data not shown). Since phorbol 12-myristate-13-acetate and lipopolysaccharide had been added to the conditioned medium, the effect of these agents on ovarian tumor cell growth was determined in the same experiment. High concentrations of phorbol 12-myristate-13-acetate (10 ng/ml) in combination with lipopolysaccharide (10 f.Lg/ml) could stimulate growth of OVCA 429, OVCA 432, and OVCA 433 but inhibited growth of OVCA 420 (Fig. 6). Lower concentrations of 0.1 to I f.Lg / mllipopolysaccharide in combination with 0.1 to I ng/ ml phorbol I 2-myristate-13-acetate failed to affect tumor cell growth . The effect of TNF-a, IL-l, and IL-6 on growth of ovarian carcinoma cell lines. TNF-o: and IL-I stimulated the growth of OVCA 429, OVCA 432, and OVCA 433 (Fig. 7, A and B). To stimulate growth of OVCA 433 required> 100 U I ml for TNF -0: and 50 U I ml for IL-I . OVCA 429 and OVCA 432 were much more sensitive than OVCA 433 to TNF-o: and IL-t. No greater

1002 Wu et al.

March 1992 Am J Obstet Gynecol

c:

100

o

:;

0(')

80

a. .

.... 0

~:;

60

CD

c: -E 40 =0 .- a. EO >-.c: 20 l-

i

(')

o

OVCA

420

OVCA

429

OVCA

432

OVCA

433

Fig. 6. Effect of phorbol 12-myristate-13-acetate and lipopolysaccharide on 3H-thymidine incorporation by ovarian carcinoma cell lines. Each of four cell lines was incubated with diluent, 0.1 fJ-g/ml lipopolysaccharide plus 0.1 ng/ml phorbol l2-myristate-13-acetate, 1 fJ-g/ml lipopolysaccharide plus 1 ng/ml phorbol 12-myristate-13-acetate, or 10 fJ-g/ml lipopolysaccharide plus 10 ng/ml phorbol 12-myristate-13-acetate, indicated by bars from left to right. Three of four cell lines, OVCA 429 (P = 0.0004), OVCA 432 (P = 0.03), and OVCA 433 (P = 0.009), were significantly stimulated, but OVCA 420 (P = 0.0000003) was inhibited by lO fJ-g/mllipopolysaccharide plus lO ng/ml phorbol 12-myristate-13-acetate.

stimulation was found at higher concentrations of TNF -a (up to 10' DI ml) or IL-l (50 to 5000 U I ml). All four cell lines were slightly stimulated by 1 U I ml of IL-6 (Fig. 7, C). When ovarian cancer cell lines were treated with all three growth factors in combination, no additive or superadditive effects were observed (data not shown). No growth inhibition was observed after 24 hours incubation with any concentration of TNF-a, IL-l, or IL-6. Growth of ovarian tumor cell lines was studied in the presence of 10% fetal bovine serum or under serumfree conditions. Three of the four cell lines could be stimulated with TNF-a, IL-l or peripheral blood monocytes in either medium. Under serum-free conditions, lower concentrations of the factors were required. In the presence of serum, growth of the fourth cell line, OVCA 420, was not affected by TNF-a, IL-l, or peripheral blood monocyte-conditioned medium. In serum-free medium proliferation of OVCA 420 could be stimulated by IL-l (Fig. 8), TNF-a, and peripheral blood monocyte-conditioned medium (data not shown). Stimulation of the growth of ovarian tumor cells was also confirmed by direct cell counts (Table I). Concentrations ofTNF.a, IL.1, and IL·6 in medium conditioned by peripheral blood monocyte and mono· cyte-derived cell lines. Relatively low concentrations of IL-l (4.3 to 27.3 U/ml) and TNF-a (10 to 22 U/ml) were detected in all peripheral blood monocyte-conditioned medium (Table II). In the absence of lipopolysaccharide, IL-6 was not detected in supernatants from peripheral blood monocyte-conditioned medium

(Table II). Peripheral blood monocyte-conditioned medium from three of the four donors significantly stimulated growth of three ovarian cancer cell lines. Peripheral blood monocytes from donor A in Table II, for example, stimulated growth of OVCA 429, OVCA 432, and OVCA 433 in Fig. 1. In the presence of lipopolysaccharide, peripheral blood monocytes produced significant amounts of IL-6, achieving concentrations of 19.5 to 27 U I ml (Table II). Low concentrations of IL-6 (1 U / ml) could stimulate each of the four cell lines (Fig. 7, C). Supernatants from peripheral blood monocytes and from lipopolysaccharide-treated THP-l and U-937 cells contained detectable quantities of TNF-a (6 and 90 Dlml), IL-I (1003 and 1294 U/ml), and IL-6 (0.7 and 9.5 U/ml). Antisera to TNF -a, IL-I, or IL-6 could neutralize authentic tumor necrosis factor, IL-I, and IL-6 but failed to neutralize the growth-stimulating activity of lipopolysaccharidetreated peripheral blood monocytes (Table III) or THP-l-conditioned medium (data not shown). Characterization of stimulatory factor(s) from pe· ripheral blood monocytes and THp·1. Several approaches were taken to characterize the factor(s) in conditioned media that were responsible for stimulating growth of ovarian carcinoma cell lines. Stimulatory activity in the supernatants of peripheral blood monocytes and THP-l cells were destroyed by boiling for 5 minutes but remained intact after incubation overnight at pH 2.6 or in the presence of reducing and denaturing agents (Table IV). Preservation of activity after reduction and dialysis suggests that factors of > 12 to 14

Monocyte factors stimulate ovarian cancer

Volume 166 Number 3

120

A

*

100

1003

80 60 40

,

C')

20

..... .....>
.

.s::. ~

•

~

C')

B

20 0 160 140 120 100 80 60 40 20 0

OVCA 420

OVCA 429

OVCA 432

OVCA 433

Fig. 7. A, Effect of TNF-cx on ' H-thymidine incorporation by ovarian carcinoma cell lines. Ovarian cancer cell lines were incubated with 0, 10, 100, and 1000 V Iml TNF-o-oS:

I-;"

40

J:

C')

30

o

5

50

500

IL-1 a (U/ml) Fig. 8. Effect of IL-I on 3H-thymidine incorporation by OVCA 420 cells in serum-free medium (dashed line) or in presence of 10% fetal bovine serum (solid line). Asterisk, Significant stimulation (P < 0.05).

Table I. Comparison of 3H-thymidine incorporation and cell number of ovarian cancer treated with IL-I, TNF-a, IL-6, and monocyte (THP-I) supernatant OVCA 429

OVCA 420 Conditions

Counts/min

Medium IL-I (500 U/ml) IL-I (50 Ulml) TNF (1000 U/ml) TNF (100 Ulml) THP-I (100%) THP-l (10%)

126,744 151,695* 149,521 * 167,573* 224,610* 195,136* 157,797*

I Cells

X

4.5 10.0 8.5 5.5 8.5 8.5 7.5

104

Counts/min

9,954 33,448* 28,329* 64,055* 75,759* 80,421* 31,379*

OVCA 432

I Cells x 10

4

3.0 5.0 3.5 7.5 9.5 9.5 6.0

Counts/min

217,631 264,601* 256,781* 259,567* 264,913* 252,696* 270,960*

I Cells

OVCA 433 X

5.5 10.0 5.0 10.0 6.5 7.5 6.5

104

Counts/min

15,926 17,353 14,728 39,501* 35,609* 36,353* 17,178

I Cells x 10

4

2.5 3.5 2.0 4.0 5.5 6.0 2.4

Ovarian cancer cells (2 x 104 ) were incubated in quadruplicate with above growth factors in serum-free medium, and 'Hthymidine incorporation was performed at 24 hours (counts per minute, mean ± SD). SD was