Harding, Charlton, and Evans

23. Hay WW. Fetal glucose metabolism. Semin Perinatol 1979;3:157-76. 24. Hay WW, Sparks jW, Quissell Bj, Battaglia FC, Meschia G. Simultaneous measurements of umbilical uptake, fetal utilization rate, and fetal turnover rate of glucose. Am j Physiol 1981 ;240:E662-8. 25. Reid DL, Neaves N, Siotten P, Phernetton TM, Rankin

February 1992 Am J Obstet Gynecol

jHG. Effects of antipyrine on electrocortical state and regional blood flows in the mature sheep fetus. j Dev PhysioI1989;11:11-4. 26. Morriss FH, Boyd RDH, Makowski EL, Meschia G, Battaglia FC. Glucose/oxygen quotients across the hind limb of fetal Iambs. Pediatr Res 1973;7:794-7.

Regulation of growth of normal ovarian epithelial cells and ovarian cancer cell lines by transforming growth factor - ~ Andrew Berchuck, MD: Gus Rodriguez, MD: George Olt, MD: Regina Whitaker, BS: Matthew P. Boente, MD: Bradley A. Arrick, MD/ Daniel L. Clarke·Pearson, MD: and Robert C. Bast, Jr., MDb , c Durham, North Carolina, and South San Francisco, California OBJECTIVE: The purpose of this study was to study the role of transforming growth factor-(3 in regulation of proliferation of normal and malignant ovarian epithelial cells. STUDY DESIGN: We examined production of and responsiveness to transforming growth factor-(3 in primary monolayer cultures of epithelial cells from five normal human ovaries and in five ovarian cancer cell lines. RESULTS: In normal ovarian epithelial cells, proliferation always was inhibited by transforming growth factor-(3 (>40%) (p < 0.01). Among the cancer cell lines, proliferation of one was markedly inhibited (>95%) (p < 0.01), two were only modestly inhibited (15% to 20%) (p < 0.05), and two were unaffected. In addition, we found that all of the normal ovarian epithelial cells and four of five ovarian cancer cell lines produce transforming growth factor-(3 ribonucleic acid and protein. CONCLUSIONS: These data suggest that transforming growth factor-(3 may act as an autocrine growth inhibitory factor for normal ovarian epithelium in vivo. Because most of the ovarian cancer cell lines are relatively resistant to the growth inhibitory effect of transforming growth factor-[3 and because one cell line does not produce transforming growth factor-(3, it is possible that loss of the transforming growth factor-(3 pathway may playa role in the development of some ovarian cancers. (AM J OSSTET GVNECOL

1992;166:676-84.)

Key words: Ovarian cancer, growth inhibitor, transforming growth Transforming growth factor-~ (TGF-~) initially was discovered in the culture media of virally transformed cells. It was called a transforming growth factor due to its ability to act in concert with TGF -ex to elicit malignant transformation of some normal non transformed cells. Subsequently, however, it has been demonstrated that TGF-~ also is produced by a wide range of normal cells. From the Division of Gynecologic Oncology, Department of Obstetrics and Gynecology, a and the Departments ofMedicine' and M icrobiologyImmunology,' Duke University, and the Department of Developmental Biology/ Genentech, Inc. Supported by National Cancer Institute grant CA 39930 and American Cancer Society Career Development Award 91-199 to A.B. Received for publication June 4,1991; revised September 3,1991; accepted September 13, 1991. Reprint requests: Andrew Berchuck, MD, Duke University Medical Center, P.O. Box 3079, Durham, NC 27710. 6/1/33780

factor-~

Molecular cloning of TGF-~ has revealed three closely related forms (TGF-~" TGF-~2' TGF-~3)' all of which are composed of two amino acid chains held together by disulfide bonds. In addition, it has been shown that TGF -~ is secreted in an inactive form noncovalently bound to a portion of its precursor, from which it must be released to produce biologically active TGF-~.2 Although TGF-~ initially was discovered due to its ability to induce the transformed phenotype in some normal cells, the predominant effect of TGF -~ on most normal cells in vitro is inhibition of proliferation. It has been proposed that TGF-~ may also act as a growth inhibitory factor in vivo; however, the sites of origin and physiologic role ofTGF-~ in various organ systems is not yet entirely clear. In addition, it has been proposed that the unrestrained proliferation of some cancers may, in part, be due to loss of normal growth

Volume 166 Number 2

inhibitory pathways such as TGF-I3. 2 This hypothesis also needs to be tested rigorously. In the rodent ovary, it has been shown that both granulosa and theca cells produce TGF-13 and that TGF-13 regulates both proliferation and responsiveness to gonadotropins of granulosa cells. 3.6 Production of and responsiveness to TGF-13 in ovarian epithelial cells, from which the majority of ovarian cancers in the United States arise, has not previously been studied, however. In addition, although it is appreciated that epithelial ovarian cancers spread intra peritoneally due to implantation and growth of cells exfoliated by the primary ovarian neoplasm, the factors that regulate the growth of ovarian cancers remain poorly understood. We have performed a series of experiments designed to investigate the role of TGF-13 in growth regulation and transformation of human ovarian epithelium. Material and methods

Cell culture. Five epithelial ovarian cancer cell lines previously established in our lab (OVCA 420, OVCA 429, OVCA 432, OVCA 433, DOV 13) were routinely cultured in modified Eagle's medium supplemented with 10% heat-inactivated fetal bovine serum. Primary monolayer cultures of normal human ovarian epithelial cells from five patients were established as previously described. 7 . 8 Briefly, the surface of the ovary was scraped gently, and the epithelial cells were then plated in a 1: 1 mixture of MCDB 105/M199 medium supplemented with 15% heat-inactivated fetal bovine serum and epidermal growth factor (10 ng/ml). Cells were cultured at 37° C in 5% carbon dioxide and 95% humidified air. Collection of conditioned culture medium. Each of the ovarian cancer cell lines and normal ovarian epithelial cultures were grown to confluence in their respective culture media containing 10% fetal bovine serum. After washing the cells with Hank's buffered salt solution, the cells were grown for 48 hours in serumfree medium. The cancer cell lines were grown in Richter's serum-free medium,9 and the normal epithelial cells were grown in serum-free MCDB 105/M199 medium without epidermal growth factor. After 48 hours, the conditioned medium was removed and stored at - 70° C. Before testing conditioned media for the presence ofTGF-I3, media from each ofthe cells was divided into two aliquots; one of which was heated at 90° C for 10 minutes while the other was untreated. The conditioned media were then dialyzed, lyophilized, and reconstituted at a fivefold increased concentration in medium containing 2% fetal bovine serum. The reconstituted media each were added to six replicate wells of CCL-64 cells, which are known to be growth inhibited by TGF-I3,10 to test for the presence of active TGF-I3. Serum-free media that had not been exposed to cells was processed in an identical manner and used as a

TGF-13 in normal and malignant ovarian epithelium

677

negative control. Authentic porcine TGF-131 (R & D Systems, Minneapolis, Minn.) was used as a positive control. Immunohistochemistry. Frozen sections were obtained from three normal human ovaries that were removed at the time of hysterectomy for benign gynecologic diseases. Immunohistochemical staining for TGF-13 was performed with the Vectastain ABC kit (Vector Laboratories, Burlingame, Calif.) as described previously. I I Monoclonal antibody ID 11.16 (Collagen Corporation, Palo Alto, Calif.) which recognizes the mature forms of both TGF-131 and TGF-132, was used as the primary antibody (5 fLg/ml).'2 Purified mouse immunoglobin G specific for nonhuman tissue (Coulter Immunology, Hialeah, Fla.) was used as a negative control (1 : 100 dilution). Anticytokeratin AE 1I AE3 (Boehringer Mannheim Biochemicals, Indianapolis, Ind.) was used as a positive control. The slides were developed for 4 minutes with the enzyme substrate diaminobenzidine (0.5% diaminobenzidine in 0.05% Tris buffer, 0.6% hydrogen peroxide). Slides then were rinsed in water, counterstained with methyl green, dehydrated, and mounted. [,H]thymidine assay. Each of the five ovarian cancer cell lines and CCL-64 cells were grown in supplemented modified Eagle's medium containing 2% fetal bovine serum in 96-well culture plates (5 x 104 cells/well). Normal ovarian epithelial cells were grown in MCDB 105/M199 medium containing 2% FBS. Authentic porcine TGF-131 (0, 0.2, 2,10, or 20 ng/ml) (R & D Systems, Minneapolis, Minn.) or antibody was added at the time of plating to six replicate wells containing CCL-64 cells. Anti-TGF-13 antibodies used included a polyclonal antisera (R & D Systems) and monoclonal antibody IDl1.6 (Collagen Corporation). Both of these reagents neutralize TGF-131 and TGF-132' In several experiments, A2C6 (Centocor, Malvern, Pa.), a murine monoclonal antibody raised against a human hepatitis B virus, was used as a negative control. The cells were grown for 48 hours. During the last 4 hours of culture, 1 fLCi of [3H]thymidine (6.7 Cil mmoll L) was added to each well. The culture medium was then removed. Cells were washed three times with phosphate-buffered saline solution and harvested. Incorporated radioactivity was measured in a Packard liquid scintillation counter. Ribonuclease protection assay. Cells were grown to confluence in medium containing 10% fetal bovine serum. Ribonucleic acid (RNA) was prepared after lysis of cells in a hypotonic solution with the nonionic detergent Nonidet P-40 supplemented with 200 fLg/ml heparin (Sigma Chemical Co., St. Louis, Mo.) as a ribonuclease inhibitor. RNA was quantitated by absorbance at 260 nm. Analysis of total RNA by the ribonuclease protection assay was performed according to published protocol, with some optimizing changes. '3 Briefly, 10 fLg cellular RNA plus 20 fLg transfer RNA

678

Berchuck et al.

February 1992 Am J Obstet Gynecol

..... c::

....Q !U

Q Q,

...

-

120

OVCA 429

100

'.

~

....

....Q

DOV 13 .....................•

................................ . t:.::·.··..·.........................

'0 u ... c:: ... 80 •. .y,.' ... -III Qc:: ..., -'-, ....c:: U 60 Q

~':.:.......

OVCA 432

.

'·----------·---·-------:~ ~·~·r.~-·---·--1 ----------OVCA 433

.,

e~

.c:>-~

-'0,

40

.•.•.•.•.•.•.•.•.•.•.•

20

o

......

o

OVCA 420 ......................................•

10

2

TGF{3 (ng/mI)

Fig. 1. Effect of TGF-j3 on proliferation of ovarian cancer cell lines (OVCA 420, P < 0.01; OVCA 432, OVCA 433, P < 0.05; OVCA 429, DOV 13, not significant).

c:

100

~

80

.2 +-' as o a.

'.'. "...,.,., :,..,

.:..., ....,

I... .,

.~

0(5 c: ~ 60

c:

:c 'E

c:

40

.s::.

.....

20

......

o

A

• ....

'" '.

.........

'.a.c.-.-.- .-.••

~

>--

+-'

.,

.~:

+-'

0 0

. .

...~

0_ (J)

-..

'~'. ...•• & ••••••••••••••

~

".

' •• _

.....

·0····· o

.2

2

TGF{3 (ng/ml)

"

.'.

IS

·-·-t CL

WK

::-:....... j EV

HN

20

Fig. 2. Effect of TGF-j3 on proliferation of normal human ovarian epithelial cells (p < 0.01). was combined with phosphorus 32-labeled complementary RNA (1.5 to 2 x 105 counts/min) in 30 fJ-l hybridization buffer, heated to 85° C for 5 minutes, and then incubated at 53° C overnight. Ribonuclease digestion was carried out by the addition of 350 fJ-l of ice-cold digest buffer, containing 0.24 units/ml ribonuclease A and 80 units/ml ribonuclease TI> and proceeded for 30 minutes at 35° C. Ribonuclease A and T I and transfer RNA were from Boehringer Mannheim (Indianapolis, Ind.). Samples were then treated with proteinase K, extracted with phenol/chloroform, and precipitated with ethanol. Samples were dissolved in gel-loading buffer and fractionated by electrophoresis with 4.5% polyacrylamide-6 mol/L urea sequencing gels, each lane containing the equivalent of 6 fJ-g of inital cellular RNA. Control samples, processed con-

currently with authentic samples, contained transfer RNA alone (negative control) or transfer RNA plus a mixture of sense complementary RNAs prepared by in vitro transcription of complementary DNA for TGF131> 132, and 133 positioned downstream of the SP6 promoter (positive control). The use of sense complementary RNAs containing full-length coding sequences for each of the three human TGF-l3s demonstrated the lack of cross hybridization between the three TGF-13 probes (data not shown). The preparation of TGF-13 probes for the ribonuclease protection assay has been described previously.14 RNA probes were stored in water at - 70° C and used within 4 to 7 days of preparation. For ribonuclease protection assays, probes for TGF-131 and 132 were mixed together (1.5 x 105 counts/min each/sample) since the diagnostic 32P-labeled fragments

TGF-13 in normal and malignant ovarian epithelium

Volume 166 :-.lumber 2

679

Fig. 3. Immunohistochemical staining for TGF-I3. A, Placenta (positive control). B, OVCA 420 cells. C, Normal human ovary.

could be easily distinguished by size. The probe for TGF-P3 was hybridized with samples separately (2 x 10' counts/min/sample). Statistics. The Student t test was used to analyze experiments in which mean values were compared. Results

The effect ofTGF-p on [3H]thymidine incorporation of the five ovarian cancer cell lines and normal ovarian epithelial cells is shown in Figs. 1 and 2. TGF-j3 markedly inhibited [3H]thymidine incorporation (>90%) in OVCA 420 cells (p < 0.01), whereas OVCA 432 and OVCA 433 cells were only modestly inhibited (15% to 20%) (p < 0.05). Proliferation of OVCA 429 and DOV 13 cells was unaffected by TGF-p (Fig. 1). In normal ovarian epithelial cells from five patients, TGF-p inhibited incorporation of [3H]thymidine by >40% in all cases (p < 0.01) (Fig. 2). In all experiments, TGF-j3 (2 to 10 ng/ml) inhibited [3H]thymidine incorporation of CCL64 cells (the positive control) by >95% (data not shown). Immunohistochemical staining for TGF-j3 with a monoclonal antibody reactive with both TGF-p, and

TGF-j32 is shown in Fig. 3. In Fig. 3, A, staining for TGF-j3 is seen in placenta, which is known to produce large amounts of TGF-p (positive control). In Fig. 3, B, immunohistochemically detectable TGF-j3 is seen in a cytospin preparation of OVCA 420 cells. Staining for TGF-j3 also was seen in each of the other ovarian cancer cell lines except OVCA 432 (data not shown). In Fig. 3, C, immunoreactive TGF-p is seen in the epithelium in a frozen section of normal human ovary. Immunohistochemically detectable TGF-j3 also was seen in cytospin preparations of cultured normal ovarian epithelial cells. Expression ofTGF-j3 messenger RNA by the ovarian cancer cell lines is shown in Figs. 4 and 5. TGF-j3! and TGF-j32 messenger RNA were present in OVCA 420, OVCA 429, OVCA 433, and DOV 13 cells (Fig. 4). In contrast, OVCA 432 cells do not contain TGF-p, messenger RNA and only a faint TGF-P2 band is seen. In Fig. 5, TGF-p, and TGF-j32 transcripts are seen in RNA derived from normal ovarian epithelial cells from two patients. In addition, the appropriate positive and negative controls that were described in the material and methods section are shown in Fig. 5. Although a TGF-

680

Berchuck et al.

February 1992 Am J Obstet Gynecol

-+ -

~~

00 CC

Fig. 4. TGF-j3 messenger RNA expression in ovarian cancer cell lines. 133 band was seen in the lane representing the positive control, none of the ovarian cancer cell lines or normal ovarian epithelial cells were found to produce TGF-133 messenger RNA (data not shown). Previously, we have shown that among the five ovarian cancer cell lines, only OVCA 433 cells secrete biologically active TGF _13. 15 In the current study, we examined whether the ovarian cancer cell lines secrete latent TGF-13 that can be activated by heating. The effect of fivefold concentrated heated conditioned culture medium from each of the ovarian cancer cell lines except DOV 13 (not done) on [3H]thymidine incorporation of CCL64 cells is com pared with that of unheated medium (Fig. 6). Heating resulted in generation of TGF-I3-like activity in conditioned media from all but one of the ovarian cancer cell lines relative to unheated media, which is respresented as 100% of control in Fig. 6 (p < 0.05). OVCA 432, which does not produce appreciable amounts of TGF-13 messenger RNA transcripts or immunohistochemically detectable TGF-I3, also was not found to produce latent TGF-13 with this bioassay. We also tested heated and unheated serum-free conditioned medium from normal ovarian epithelial cultures from two patients for the presence of active and latent TGF-I3. We found that fivefold concentrates of unheated medium conditioned by epithelial cells from two normal ovaries consistently decreased [3H]thymidine incorporation of CCL64 cells (p < 0.05). When the medium was heated before exposure to CCL64 cells, to activate latent TGF-I3, an even greater inhibition of proliferation was noted (Fig. 7). In addition, the inhibition caused by the conditioned medium was

Fig. 5. TGF-j3 messenger RNA expression in normal ovarian epithelial cells from two patients along with positive and negative controls. reversible if anti-TGF-13 monoclonal antibody (10 IJ-gmlml) was added to the cultures. Serum-free medium that had not been exposed to cells but which had been processed in an identical manner had no effect on [3H]thymidine incorporation of CCL-64 cells (data not shown). In Fig. 8, the effect of polyclonal anti-TGF-13 antisera on [3H]thymidine incorporation of four of the ovarian cancer cell lines is shown. Proliferation of three ovarian cancer cell lines that either do not express TGF-13 (OVCA 432) or express only latent TGF-13 (OVCA 420, OVCA 429) was not affected by the antibody. On the other hand, proliferation of OVCA 433 cells, which both secrete active TGF-13 and are inhibited by exogenous TGF-I3, was stimulated by the anti-TGF-13 antisera (p < 0.05). Similar results were obtained using the monoclonal anti-TGF-13 antibody (data not shown). In addition, the monoclonal anti-TGF-13 antibody significantly stimulated (p < 0.01) [3H]thymidine incorporation of normal ovarian epithelial cells from one patient (Fig. 9), whereas ovarian epithelial cells derived from another patient were not stimulated. These experiments were repeated several times with different concentrations of serum (0% to 10%), different cell densities (confluent, subconfluent), and different incubation times (24 to 48 hours). Optimal stimulation of proliferation was noted in confl uent cells after 48 hours in serum-free medium. The negative control antibody A2C6 had no effect on [3H]thymidine incorporation of

TGF-j3 in normal and malignant ovarian epithelium

Volume 166 Number 2

..

0 ....= ...0

120

= !U

...o -0

-....= = 0

U

60

0

40

.....

....

"tS

Heated 90°C

80

l:

GI

Unheated •

100

Q.

\,I



681

e~ >. ......

..c: E-c

20

::r:

0

.......

......

Of)

OVCA 420

OVCA 429

OVCA 432

OVCA 433

Fig. 6. Production of latent TGF-j3 by epithelial ovarian cancer cell lines (OVCA 420, OVCA 429, OVCA 433, P < 0.05; OVCA 432, not significant).

t:::

...

,g

30

..c

01

25

CI.

~~ v,

...

t:::

I:) ~

20

QI

t::: X

;; Ei 15 Ei CI. v ».c: 10 foe

~

...

:I:

5

o

control

authentic TGF-{j

unheated unheated

(lng/mI)

+

anti-TGF- {j antibody

heated

heated +

anti-TGF-{j antibody

Fig. 7. Production of latent and active TGF-j3 by human ovarian epithelial cells (heated, unheated,

p < 0.05).

the cancer cell lines or the normal epithelial cells (data not shown).

Comment Although TGF-(3 was discovered due to its ability to act in concert with TGF-()( to transform some types of normal cells in culture, I TGF-(3 has been shown to inhibit proliferation of many normal epithelial cells. 2 In addition, TGF-13 also frequently induces differentiation of epithelial cells in culture. For example, bronchial epithelial cells undergo terminal differentiation to a squamous phenotype after exposure to TGF _13. 16 The mechanism of action of TGF-13 is poorly understood, however. TGF-13 receptors have not been well characterized and the intracellular events that are triggered after TGF-j3 binding have not been defined. Although in vitro evidence suggests that TGF-(3 might be a phys-

iologically important growth inhibitory substance, the role of TGF-(3 in normal processes of growth and differentiation remains unclear. The potential role of TGF-13 as a growth inhibitory substance has been studied extensively in breast epithelium. It has been shown that normal breast epithelial cells produce TGF-13 and that TGF-13 acts to inhibit proliferation of these cells. 17 . 18 In addition, it has been shown that steroid hormones that inhibit growth of normal breast epithelium, such as progestins and antiestrogens, stimulate production of TGF _13. 19 Conversely, estrogens, which stimulate growth, act to inhibit TGF-13 production. 14. 19 It has been proposed that the effect of circulating steroid hormones on proliferation of breast epithelium may be due to induction or suppression of production ofTGF-13 and other peptide growth factors.

682

Berchuck et al.

February 1992 Am J Obstet Gyneco1

150 OVCA 433

OVCA 420 OVCA 432 OVCA 429

SO

o

o

5

10

15

20

Anti-TGF- {J Neutralizing Ab (J..lg/mI) Fig. 8. Effect of polyclonal anti-TGF-13 antisera on proliferation of epithelial ovarian cancer cell lines (OVCA 433, P < 0.05).

.....

....o= IU

o

...Q,. ,.......

400 350

300 oIJ '0 ...

= .. 250 -= U0= 200 .... QI

:g'Q

=*

150

..c:>-- 100 I-c

••••••

•.•............

.' .'.'

•••••••••.

..................• 2 %serum

SO

o

0.2

2.0

20.0

Anti-TGF-{J antibody (~g/mI) Fig. 9. Effect of monoclonal anti-TGF-13 antibody on proliferation of normal ovarian epithelial cells (serum-free, 2% serum, p < 0.05).

In the present study we found that, like normal breast epithelium, normal ovarian epithelium produces both TGF-j3, and TGF-j32 and is growth-inhibited by exogenous TGF-j3. These data suggest that TGF-j3 might act in vivo as an autocrine growth inhibitory substance that prevents inappropriate proliferation of ovarian epithelial cells, which normally are thought to proliferate relatively slowly. At the time of ovulation, this inhibitory pathway might be turned off to allow proliferation of epithelial cells to cover the defect that is caused by extrusion of the ovum through the ovarian surface. Because the levels of estrogens and progesterone in the ovary vary markedly during the cycle, we are investigating the possibility that TGF-j3 production by ovarian epithelial cells is regulated by steroid hormones in a manner similar to that observed in breast epithelium.

In addition, our groupS and Siemens and Auersperg20 have shown that epidermal growth factor stimulates proliferation of normal human ovarian epithelial cells. Proliferation of ovarian epithelial cells after ovulation, or at other times, might be due to both a decrease in growth inhibitory pathways such as TGF-j3 and increased availability of mitogenic peptide growth factors. We found that normal human ovarian epithelial cells sometimes, but not always, appeared to activate the latent TGF-j3 that they secreted into their culture medium. Further studies are needed to determine whether TGF-j3 produced by ovarian epithelial cells is activated in vivo. In addition, the mechanism of activation and the cell types involved also must be identified. In this regard, it is known that TGF-j3 is secreted from cells in a biologically inactive form noncovalently

Volume 166 Number 2

bound to a portion of the precursor molecule from which it is cleaved during processing. 2 Furthermore, TGF-[3 in serum is bound to a2-macroglobulin}' Although TGF-[3 can be activated by low pH or heating in vitro!2 these are not thought to be physiologically relevant mechanisms of TGF-[3 activation. There is some evidence to suggest that proteases and glycosidases may play an important growth regulatory role in vivo by liberating active TGF-[3 from its precursor. 23 . 24 If a cell that produces TGF-[3 and responds to TGF-[3 also produced the enzymes that activate TGF-[3, it would possess all of the components of an autocrine growth inhibitory loop. On the other hand, if a cell lacked the enzymes needed to activate TGF-[3, it is possible that TGF-[3 still might be activated in the local environment if these enzymes were produced by other nearby cells. This latter hypothesis is particularly appealing because the ovary and other human organs are comprised of several different cell types that presumably must communicate to act in a coordinated fashion. Although increased proliferation associated with cancer may be due to inappropriate production of growth stimulatory factors, loss of growth inhibitory pathways, such as TGF-[3, also could result in increased proliferation. In this regard, we found that all but one of the ovarian cancer cell lines were either partially or completely resistant to the growth inhibitory effect ofTGF[3 relative to normal ovarian epithelial cells. Similarly, several other transformed cells and human cancer cells that are resistant to TGF-[3 have been described?5-2' In addition, although normal ovarian epithelial cells and most of the ovarian cancer cell lines produced TGF-[3, and TGF-[32' one of the ovarian cancer cell lines produced little, if any, TGF-[3. Finally, only one of the ovarian cancer cell lines was able to activate the TGF[3 that it produced. These findings suggest that loss of various components of the TGF-[3 pathway is common in ovarian cancer cell lines. It is possible that loss of this growth inhibitory pathway might playa role in the development of some ovarian cancers. REFERENCES 1. Anzano MA, Roberts AB, SmithJM, Sporn MB, DeLarco JE. Sarcoma growth factor from conditioned medium of virally transformed cells is composed of both type

Regulation of growth of normal ovarian epithelial cells and ovarian cancer cell lines by transforming growth factor-beta.

The purpose of this study was to study the role of transforming growth factor-beta in regulation of proliferation of normal and malignant ovarian epit...
2MB Sizes 0 Downloads 0 Views