Human fibroblasts from cancer patients: lifespan and transformed phenotype in vitro and role of mesenchyme in vivo

Mutation Research, 199 (1988) 313-325

313

Elsevier MTR 02003

Human fibroblasts from cancer patients: lifespan and transformed phenotype in vitro and role of mesenchyme in vivo B. Azzarone 1, C. Chaponnier 3, p. Krief 1, M. Mareel and A. Macieira-Coelho 5

2, H.

Suarez 4

1 U268 Inserm, 94824 Villejuif (France), 2 Laboratory for Experimental Cancerology, Department of Radiotherapy and Nuclear Medicine, Ghent (Belgium), 3 D~partement de Pathologic, C.M.U., 1211 Geneoa (Switzerland), 4 LR.S.C., C.N.R., 94804 Villejuif (France), and 5 D~partement de Pathologie Cellulaire, INSERM, 94804 Villejuif (France)

(Accepted 23 September1987) Keywords: Human fibroblasts; Alterationin vitro.

Summary Human fibroblasts cultured in vitro can exhibit a different potential number of population doublings. In normal donors, the average number of population doublings is inversely related to the donor's age. An increased growth potential was detected in skin fibroblasts from breast cancer patients, independently of the donor's age. These cells responded in an abnormal way to 3 biological parameters: (1) colony formation in semisolid medium; (2) colony formation on monolayers of normal human epithelial cells; and (3) increase of saturation densities in overcrowded culture conditions. A third of these cultures, as well as skin fibroblasts from other cancer patients, at the plateau phase of growth exhibited a significant percentage of cells still synthesizing DNA. Exposure to overcrowding, limited in time, caused the selection of a cell subset which displayed new biological, biochemical and functional properties commonly found in transformed cells. The abnormal in vitro behavior of skin fibroblasts from breast cancer patients does not seem to be associated with the expression of oncofetal membrane markers (4F2, IL2 receptor) while the fibroblasts from patients with the adenomatosis of the colon and rectum (ACR) syndrome expressed the 4F2 antigen. This is the first time that the IL2 receptor is found on non-hematopoietic cells. Fibroblastic cells with abnormal characteristics, which may also present a decreased efficiency in organizing a primitive fibrin matrix, could represent in vivo an anarchistic milieu, favoring disturbed epithelial-stromal interactions and the emergence of the less structured tumor stromatic tissue.

Human normal fibroblasts invariably die after serial doublings in culture (Hayflick, 1965) and it was found that the potential number of cell generations is inversely related to the age of the donor

Correspondence: Dr. B. Azzarone, INSERM - Unit6 268, Oncogenrse Appliqure, 14 & 16, Avenue Paul-VaillantCouturier, 94800, ViUejuif(France).

(Martin et al., 1970). Fibroblasts from other species (e.g., rodents), on the contrary, can, even if with different efficiency, yield immortal cell lines (Todaro and Green, 1963; Terzi and Hawkins, 1975; Matsumura, 1983; Kraemer, 1986). It was also found that rodent cells are more susceptible to oncogenes, and to chemical, viral and physical carcinogens in comparison with normal human fibroblasts (for review see Sager, 1983). In most

0027-5107/88/$03.50 © 1988 ElsevierSciencePublishers B.V. (BiomedicalDivision)

314

instances, indeed, human fibroblasts treated with different carcinogens are not converted into fully malignant cells and only sometimes do they acquire a partially transformed phenotype (Freedman and Shin, 1977; Freeman et al., 1977, Milo and DiPaolo, 1978; Kakunaga, 1978; Sacks, 1981; Maher et al., 1982; Namba et al., 1985). Moreover, it seems that in cultured rodent fibroblastic cells expression and/or acquisition of chromosome instability are events associated with acquisition of the transformed and/or immortal phenotype. Indeed, transformation of freshly isolated Syrian hamster embryo fibroblasts after double transfection with c-ras plus c-myc was associated with important chromosome changes (Oshimura et al., 1985). Precrisis diploid murine fibroblasts, which invariably become spontaneously permanent cell lines, display high levels of recombinational events in the genome, such as bridges between chromosomes and radial figures, alterations that cannot be found in normal human diploid fibroblasts (Macieira-Coelho, 1980; Macieira-Coelho and Azzarone, 1986). These results raise the following questions: (1) do human fibroblasts exist which spontaneously escape partially or totally from a limited rife span, and (2) is there any relationship with the in vivo pathology of the donor and genomic stability of his somatic cells? In this paper we will try to offer some answers to these questions. We will describe the existence of fibroblasts, derived from patients with cancer, or at high genetic risk of cancer, which show a tendency for increased growth potential and/or for the expression of properties related with the transformed phenotype. We will also discuss the possible role of these cells in the emergence of tumors in vivo. Material and methods Cell lines

Two embryonic (964 S and ICIG 8) and 10 postnatal skin fibroblastic populations (NMS1, BMS2, BMS4, BMSs, BMS6, BMS7, RDF, AJ, LBB and LRB) were obtained from donors without any known neoplastic disease. Fibroblasts were also cultivated from biopsies of uninvolved skin from 8 patients with breast cancer (CM 3, CMs, CMs, CMlo , CMll , CM16, CM19, CMV) and from

3 patients with melanoma, hepatoma and osteosarcoma tumors, respectively. The donors had not previously received any chemotherapy, hormone therapy or radiotherapy. Details concerning clinical data and growth culture characteristics of the above cell fines have been described elsewhere (Azzarone et al., 1984a; Azzarone and Macieira-Coelho, 1987). Thoracic and abdominal biopsies from the superficial skin region (epidermis and papillary layer) were placed in plastic petri dishes and covered with the nutrient medium (MEM supplemented with 10% fetal calf serum and glutamine 2 mM). After an initial outgrowth of epithelial-like cells, migration and proliferation of fibroblast-like cells occurred. Confluency was reached within 30 days, then the primary cultures were transferred by trypsinization into plastic bottles at 1:2 split ratio. When cells formed a confluent sheet, no mitoses could be observed and cell counts did not increase for 3 days, they were subcultivated at 1 : 2 sprit ratio. The subculture interval was 5-6 days, without any change of medium. The skin fibroblast cell lines derived from patients with adenomatosis of the colon and rectum (GM3901, GM3314, GM2694A, GM3900, GM6855) or from normal adult donors (AG-4151, AG-4353, AG-4445 and AG-4059) were purchased from the Human Genetic Mutant Cell Repository (Cadmen, N J, U.S.A.). The human embryonic lung fibroblast cell lines employed in the immunofluorescence assays (WI38, IMR-90, MRC-5, 964 L, ICIG-7 and ICIG-9) have been characterized elsewhere (Azzarone et al., 1984b). The human embryonic lung cell lines (GM-1380, GM-1604 and GM-1375) were purchased from the Human Genetic Mutant Cell Repository (Cadmen, N J, U.S.A.). The origin and the characteristics of the human sarcoma cell lines HT1080 (fibrosarcoma). Te-85 (osteosarcoma), RD (rhabdomyosarcoma), OHA (osteosarcoma) and SHAC (fibrosarcoma) have been reported elsewhere (Azzarone et al., 1984b, 1986). Increase of saturation densities in overcrowded culture conditions

Confluent cultures were trypsinized, counted and divided into 2 groups. In one, each culture

315 was subcultivated into 2 new bottles (1:2 split ratio); in the other, each culture was subcultivated into 1 new bottle (1 : 1 split ratio). When the cells formed a confluent sheet, no mitoses could be observed and cell counts on duplicate cultures, performed with an electronic Coulter Counter, did not vary for at least 3 days, cells were trypsinized, resuspended in growth medium and subcultivated at 1 : 2 or 1 : 1 split ratios, until cell densities of the 1 : 1 group did not increase further.

Anchorage dependence Colony formation in soft agar was performed as follows: single cell suspensions were diluted in 0.3% agarose (Difco or IBF, France) in growth medium (MEM supplemented with 20% fetal calf serum and glutamine 2 mM) and 2 x 105 cells per 30-mm petri dish were added to 2 ml of a prehardened base containing 0.5% agar in growth medium. After 2 weeks, colony formation was determined in at least 3 petri dishes with the help of an inverted microscope. Only aggregates with more than 30 cells were considered to be colonies; at the end of the experiment cell viability was ascertained by the ability of single colonies picked up from agar to spread in liquid medium onto a plastic surface. Results are the means of 3 different assays. Colony formation The assay was performed according to a previously described technique (Smith et al., 1976). Confluent monolayers of the normal human embryonic intestinal epithelial cell line Int. 74 (Owens et al., 1976; Azzarone et al., 1983) were prepared in 60-ram diameter tissue-culture dishes. This cell line was a gift of Dr. W. Nelson-Rees (Naval Biomedical Research Laboratory, Oakland, CA, U.S.A.). Cultures to be tested were trypsinized and seeded at 103 cells per dish and the medium was changed twice a week for 21 days; they were then fixed in methanol and stained with May-Griinwald Giemsa. Results are the means of 2 different assays. Invasiveness in organotypic culture We used a method based on the confrontation of spheroidal aggregates of the cells to be tested with precultured fragments from chick embryo

hearts in shaker cultures. Aggregates of CM-11 (1 : 2), CM-11 (1 : 1) and embryonic fibroblasts were prepared by incubation of 6 x 105 trypsinized ceils in 6 ml culture medium (MEM REGA I, Gibco plus 10% fetal calf serum and 0.05% Lglutamine) on a gyratory shaker at 70 rpm during 3 days, as described previously (Mareel, 1979; Mareel et al., 1979). Cultures were examined under a stereomicroscope ( x 25) and individual aggregates of each cell culture, of about 0.2 mm, were then cocultivated in liquid medium with a fragment of 9-day-old chick cardiac muscle on the top of a semisolid agar layer. After attachment, individual cocultivated spheroids were further incubated in liquid medium on a gyratory shaker at 120 rpm. Triplicate cocultivated spheroids were fixed in Bouin Hollande's solution after 4, 7 and 14 days, embedded in paraffin, completely sectioned and processed for microscopic observation. Further details of the organ-culture assay and its relevance for evaluating invasion in vitro have been described elsewhere (Mareel, 1979; Mareel et al., 1979).

Fibrin clot retraction assay (FCR) The degree of a plasma clot compaction induced by cultured cells was determined and quantified according to a previously described technique (Azzarone et al., 1981a,b, 1983; Azzarone and Macieira-Coelho, 1984). Briefly, 96-h cultures were trypsinized, washed, resuspended in Tyrode buffer and tested for FCR at a final concentration of 1 x 106 cells/ml. The assay was performed in 5-mm internal diameter glass tubes with a 2-ram glass rod standing in the middle. 200 #1 of cell suspension, 100/~I of bovine platelet-free plasma (Gibco) and 50/~1 of thrombin at 20 I U / m l were added in rapid succession. Immediately after the addition of thrombin, the content of the tubes was mixed with the glass rod which was left in the tube. The system was incubated at 37°C in a plastic water bath through which the diameter of the clot around the glass rod was measured with a microcaliper applied to the outer surface. Retraction values were calculated as a percentage of complete clot retraction; the data reported below are obtained from the means of triplicate samples and from 3 different experiments. At the end of the assay, clots were recovered, microscopically

316 scored and photographed for the cell fibrin interactions.

Immunofluorescent staining: microfilaments and sensitivity to plasma Gelsofin Gelsolin is an actin-binding protein, which fragments actin filaments, nucleates actin filaments and blocks their fast-growing end (Chaponnier et al., 1986). Its ability to sever actin filaments may be useful in recognizing in immunofluorescent assays different degrees of actin polymerization and hence the level of the microfilament network organization; sensitivity to Gelsolin is indeed expressed with a disappearance of the microfilament network (Low et al., 1981). So we determined the sensitivity of cytoplasmic actin to plasma Gelsolin of CMll cultures maintained at different final saturation densities. Confluent CMxa cultures maintained at 1 : 2 and 1 : 1 split ratios were trypsinized and then seeded onto glass coverslips at the same inoculum (3 × 105 cells/cm2). Four hours later, when spreading was accomplished, coverslips were rinsed twice with phosphate-buffered saline and fixed with absolute ethanol (30 s at room temperature). Half of the coverslips from both groups were incubated with purified plasma Gelsolin (100/~g/rnl for 30 min), then all the coverslips were processed for immunofluorescent staining with purified antiactin antibodies as previously described (Low et al., 1981).

Immunofluorescence: oncofetal membrane antigens Immunofluorescent staining of adherent living cells was performed as previously described (Azzarone et al., 1984b). 4F2 monoclonal antibody was a gift of Dr A. Fauci ( N I H , Bethesda, MD, U.S.A.); anti-IL2 receptor monoclonal antibody was purchased from Immunotech (Marseille, France). 500 cells of each cell line were examined in at least 3 independent assays. Results

Relationship between lifespan in vitro, donor's age and pathology We have plotted in Fig. 1 the average population doublings (PDs) versus the age of the donor

50. Z ~40.

[ts64 s~.Ms L_¢,,%o:,,o

ocM3 oTM %cM, %,, OC~16

BMS 6 .BMS5

~ 20'

R=0.9052

BM$4 BMS2

JlO.

iO

20

30 40 50 60 DONOR'S AGE (yeors)

70

80

90

I00

Fig. 1. Age of the donor versus average population doublings of the respective skin fibroblast strains derived from a human embryo and from patients with benign mammary lesions (e) or breast cancers (O). The population doublings were calculated by the evaluation Log2 ×(N/No), where N is the final cell number and N Othe initial one. The straight line is the regression fitting the plot corresponding to the fibroblasts originating from donors with benign lesions. The correlation coefficient is indicated on the chart (Azzarone et al., 1984).

in in vitro skin fibroblasts from patients with benign and malignant breast lesions. The data indicate that in patients with benign lesions or having undergone surgery for non-neoplastic disease there is an inverse correlation between average PDs and donor's age. In contrast, in patients with breast cancer the relationship between average PDs and age is lost, the plot forms an horizontal line and the lifespan of these cells approaches that found in h u m a n embryonic skin fibroblasts.

Skin fibroblasts, donor's pathology and spontaneous expression of abnormal phenotype We further characterized skin fibroblasts from patients with breast cancer, which express an increased growth potential, for the expression of biological properties commonly found in tumor cells. Analysis of Table 1 shows that all the skin fibroblasts derived from cancer patients formed colonies in soft agar, on contact inhibited monolayers of h u m a n epithelial ceils and were able to increase their saturation densities in overcrowded culture conditions. In contrast, the fibroblast cultures obtained from patients with benign lesions and the embryonic h u m a n fibroblasts were unable to do so, with the exception of the cell line BMS5 (Azzarone et al., 1984b).

317 TABLE 1 BIOLOGICAL CHARACTERIZATION IN VITRO OF THORACIC SKIN FIBROBLASTS DERIVED FROM PATIENTS WITH MALIGNANT OR BENIGN BREAST LESIONS Cell line

Correlation between number of doublings in vitro and donor's age

Number of colonies in soft agar per 2 × 105 cells plated

Number of colonies on contact inhibited human epithelial sheets per 1 × 103 cells plated

% Change in saturation densities after the first 1 : 1 split

+ + + + + + +

24 (+4) 12 (+ 3) 23 (5-5) 27 (+3) 150 (5:25) 50 (5:9) 62 (+9)

9(+2) N.T. 9(+1) 11 (+1) 24 (+2) 8(5-1) 12 (5:2)

+40 + 50 +50 +45 +50 +40 +70

0 0 14 (+6) 0 0 0 0

0 0 12(+1) 0 0 0 0

+10 +8 +58 -10 +10 -3 -3

Malignant: CM 3 CM5 CM s CM10 CMll CM16 CM19

Benign: BMS2 BMS4 BMS5 BMS6 BMS7 NMS 1 964S

Moreover, analysis of T a b l e 2 shows t h a t within the g r o u p of b r e a s t cancer p a t i e n t s it was p o s s i b l e to define a s u b g r o u p whose f i b r o b l a s t s e x h i b i t e d a significant n u m b e r of cells still synthesizing D N A w h e n g r o w t h curves r e a c h a plateau. T h e p h e n o m TABLE 2 PERCENTAGE OF LABELED INTERPHASES ON THE SECOND DAY OF THE PLATEAU PHASE OF GROWTH IN HUMAN SKIN FIBROBLASTS DERIVED FROM NORMAL DONORS OR FROM CANCER PATIENTS Numbers within parentheses indicate the standard errors. Donors

%

Age

Sex

Controls a (12 cell lines)

From embryo to postnatal (20-65 years)

M and F

Melanoma Hepatoma Osteosarcoma Breast cancer BMS5 CMs CMV

30 61 20

M M F

30 ( 5-4.5) 30 (+ 4.5) 20 (+ 3.0)

30 58 35

F F F

22 (+ 3.3) 25 ( 5: 3.7) 28 (+ 4.2)

5 (5: 0.8)

" Details on the control cell lines are reported in Material and methods.

e n o n c o u l d also b e d e t e c t e d in skin f i b r o b l a s t s f r o m p a t i e n t s with o t h e r t y p e s o f cancer. F o r each cell line g r o w t h a n d D N A synthesis curves were p e r f o r m e d with d a i l y cell counts a n d a u t o r a d i o g r a p h i c analysis o n triplicate samples; experim e n t s were a r r e s t e d w h e n n o further increase in cell n u m b e r was o b s e r v e d for 3 days. I n the h e p a t o m a p a t i e n t , d e t e c t i o n o f this defect prec e d e d the d i s c o v e r y of the disease ( A z z a r o n e a n d M a c i e i r a - C o e l h o , 1987).

Modification of behavior of skin fibroblasts from breast cancer patients after overcrowding W e tried to u n d e r s t a n d w h e t h e r the initial changes in s a t u r a t i o n densities, o b t a i n e d b y v a r y ing the split ratios, are c o n s t a n t t h r o u g h the lifesp a n . Therefore, p o s t c o n f l u e n t CM11 cultures r o u t i n e l y m a i n t a i n e d w i t h the 1 : 2 splits were d i v i d e d i n t o 2 g r o u p s at different p a s s a g e levels: one g r o u p was c a r r i e d o u t with the 1 : 2 splits a n d the o t h e r o n e with 1 : 1 splits. Fig. 2 shows t h a t m u l t i p l e s u b c u l t i v a t i o n s at 1 : 1 splits c a u s e d a progressive increase in cell c o u n t s which within 3 weeks r e a c h e d significantly h i g h e r values t h a n those f o u n d in p a r a l l e l cultures m a i n t a i n e d at 1 : 2 split ratio. This p r o p e r t y was

318

4xlos" CMII 3x,o~2x,o5

i

,o..~--=/7~'~ '= ~....=..= /'/ / .P"=~"~fo.-o--=

/

/~/~/~

S x 104"

2xlO 4

IO 4

•

•

•

5

•

•

•

•

.

•

•

•

,

.

.

,

,

o

.

.

,

.

•

.

.

•

.

,

.

I0

15 ZO 25 30 35 40 45 NUMBER OF SUBCULTURES Fig. 2. Maximal cell densities before subcultivation of C M n cells maintained at 1 : 2 (e) and 1 : 1 ( O ) split ratios (Azzarone et al., 1984).

detected throughout the entire'lifespan of the CMax cells, even if the maximal cell densities at different passage levels decreased with the age of the cultures (Azzarone et al., 1984b). To see whether the acquisition of higher saturation densities is associated with the selection of a subpopulation with different characteristics, we tested CM11 cultures, propagated in parallel at 1 : 2 or at 1 : 1 sprit ratios for an identical number of subcultures, for several biochemical, biological and functional parameters. Cultures were analyzed at the end points of the first, second and third serial subculture at 1 : 1 split (Fig. 2). Analysis of Table 3 shows that multiple subcultivations of the CM~I ceils in overcrowded culture conditions caused the selection of a cellular subset with: (1) faster entrance into the S phase (15 h after seeding only 0.6% of CM 1 : 2 ceils synthesized DNA, whereas 6% of CM11 1:1 cells are able to do so; (2) reduced capacity to organize by compaction a primitive fibrin matrix and to anchor to it; (3) increased sensitivity to plasma Gelsolin (Fig. 3); (4) different patterns of fibronectin secretion during spreading; (5) increased cloning efficiency on epithelial sheets; (6) these cells became able to invade foreign tissues in

vitro (Fig. 4). Alterations concerning points 1, 2, 3 and 5 were checked and confirmed, in at least 2 independent assays, in CM cell lines serially subcultivated with the same protocol. Fig. 3 shows that, 4 h after seeding, cultures derived from the 1 : 2 (3a) and from the 1:1 groups (3b) exhibit a well-developed microfilament network. This structure is totally disassembled by pre-exposure to plasma Gelsolin in the 1:1 group (3d) but not in the 1:2-derived one (3c). A different sensitivity to Gelsolin is maintained throughout the culture lifetime. Fig. 4 shows that after a 4-day confrontation of a precultured chick heart fragment with an aggregate of CMll cells from 1 : 2 splits, these latter cells have adhered without penetrating (4a) whereas the cells from the 1 : 1 splits have acquired the capacity of invading the host tissue (4b). Finally we observed that normal adult skin fibroblasts express maximal values (100%) at the end of the F C R assay (24 h), microscopic observation of the clots shows parallel arrays of wellspread cells organized along the longitudinal axis of the clots. Whereas the majority of CM cultures exhibit values of about 70% and a less wellorganized alignment within the clots, CM 1 : 1-de-

319

c Fig. 3. Immunofluorescent staining of CMll cultures with anti-actin antibodies 4 h after seeding at 3)<10 4 cells/cm 2. A well-developed network of actin cables is evident in (a) CMll (1:2)- and (b) Cmll (1 : 1)-derived cultures, while pre-incubation with Gelsolin causes the disappearance of the microfilaments (d) in the CMll (1:1)- but not (c) in the CMll (1:2)-derived cultures.

Fig. 4. (a) I-Iistologic section from a 4-day confrontation of a precultured chick heart fragment with an aggregate of CM u (1 : 2 split ratio) ceUs. CM11 ceUs have adhered to host tissue without penetrating. The arrow indicates the human fibroblasts. (b) Histologic section from a 4-day confrontation of a precultured chick heart fragment with an aggregate of CM11 (1 : 1 split ratio) cells. CMll cells invaded the host tissue. The arrow indicates the human fibroblasts (Azzarone et al., 1984).

320 TABLE 3

TABLE 4

D I F F E R E N C E S BETWEEN SKIN FIBROBLASTS IN C U L T U R E ( C M n ) DERIVED F R O M THE M A M M A R Y G L A N D OF A PATIENT WITH BREAST C A N C E R A N D MAINTAINED AT D I F F E R E N T SPLIT RATIOS (1 : 1, 1 : 2)

B I N D I N G OF 4F2 A N D anti-IL2 RECEPTOR mAbs TO EMBRYONIC, A D U L T A N D M A L I G N A N T FIBROBLASTIC CELLS SPREAD ONTO A S U R F A C E 4F2 binding IL2 receptor binding

Cell line Saturation density (cells/cm × 2z) D N A synthesis 15 h after seeding at the same density (3×10×4z cells/era x 2z) Fibrin clot retractile activity after 4 h Sensitivity to plasma Gelsolin of microfilament network Presence of fibroneetin network 4 h after seeding Colony efficiency on contact-inhibited human epithelial cells (per 1000 seeded cells) Invasiveness in organotypic culture

CMla (1 : 1)

CM H (1 : 2)

4 x 10 5

1.1 × 10 5

6%

0.6%

< 1%

51%

+

-

+

-

50 + 2

24 + 2

+

-

rived cultures express final values ranging between 25 and 50%; cellular distribution within the clots is highly disorganized.

Fibroblasts from cancer patients and expression of oncofetal markers Skin fibroblasts from patients with adenomatosis of the colon and rectum (ACR), a group at high genetic risk of cancer, or from breast cancer patients, seem to display biological properties which are usually expressed by embryonic cells (Pfeffer et al., 1976; Schor et al., 1985). It was therefore suggested that these adult cells present alterations in the differentiation pathway leading to the maintenance of the oncofetal phenotype. We tested this hypothesis, analyzing the presence of oncofetal antigens at the surface of spread living fibroblasts from breast cancer patients and from patients with ACR. We have previously shown that the 4F2 monodonal antibody binds to the surface of living fibroblasts, of embryonic and neoplastic origin, spread onto a coverslip, but not

Embryonic fibroblasts 12 cell lines Sarcoma cells 5 cell lines Adult fibroblasts from normal donors 15 cell lines Adult fibroblasts from breast-cancer patients 5 cell lines Adult fibroblasts from ACR patients 5 cell lines

+

+

+

+

+

to the surface of normal adult ones (Azzarone et al., 1984b, 1985b, 1986). Table 4 summarizes the results obtained in different types of fibroblasts, analyzed by indirect immunofluorescence for the presence of the 4F2 antigen. 100% of cells from different embryonic cell lines (Fig. 5a) expressed the 4F2 antigen, whereas none of the adult fibroblast cell lines, derived from normal or from breast cancer donors, were 4F2-positive (Fig. 5b). One exception is the skin fibroblasts from adult ACR patients, which, spread onto a surface, were also 4F2-positive (Fig. 5c). Sarcoma ceils behave as the embryonic ones (not shown). Interestingly, we detected a different behavior in the immunofluorescent staining when the fibroblast cells were incubated with different monoclonal antibodies directed against the IL2 receptor. Indeed, we observed the presence of the IL2 receptors on the surface of embryonic (Fig. 6a) and sarcoma cells (Fig. 6c) but not on the surface of adult fibroblasts derived either from normal donors or from breast cancer and ACR patients (Fig. 6b). This result was confirmed by the analysis in Northern blots of the specific mRNAs. Data showed, in the embryonic and sarcoma cells but not in the adult ones, the presence of 2 classes of mRNA whose size was similar to that previously described in activated T cells (3.6 and 1.7 kbp, respectively). This is the first

321

\

Fig. 5. Immunofluorescent staining of human fibroblasts spread onto a glass coverslip with the 4F2 mAb. (a) Embryonic fibroblasts (100% positive cells), (b) human adult fibroblasts (no cells stained), (c) ACR fibroblasts (100% positive cells). time that the IL2 r e c e p t o r was d e t e c t e d on n o n h e m a t o p o i e t i c cells. These results suggest that the a b n o r m a l b e h a v i o r d e t e c t e d in skin f i b r o b l a s t s f r o m b r e a s t c a n c e r

Fig. 6. Immunofluorescent staining with anti-IL2 receptor mAb of human fibroblasts spread onto a surface. (a) Embryonic cells (100% positive), (b) adult fibroblasts from normal, breast-cancer and ACR donors (no cells stained), (c) fibrosarcoma cells (100%).

p a t i e n t s is n o t a s s o c i a t e d with the m e m b r a n e organization characteristic of human embryonic fibroblasts.

322 Discussion

In this paper we show the existence of human fibroblasts spontaneously exhibiting a different behavior for the potential number of doublings performed in vitro and the expression of abnormal properties. The expression of an altered phenotype seems to depend on the donor's pathology, since it was only detected in mammary-cancer patients. Similar data were reported by 2 different groups (Durning et al., 1984; Schor et al., 1986; Lynch et al., 1984). The former described the existence of abnormal cell-collagen interactions in fibroblasts from patients with breast cancer and the familial transmission of the defect (Schor et al., 1986), the latter reported that skin fibroblasts from breast cancer patients with a family history of the disease frequently present karyotypic alterations (Lynch et al., 1984). In patients with breast cancer we observed that the inverse relationship between in vitro lifespan and donor's age, usually found in normal donors (Martin et al., 1970), was lost, since their growth potential approached that found in human embryonic skin fibroblasts (Fig. 1). Moreover, a prolongation of the lifespan in vitro of fibroblasts derived from donors at high risk of cancer or bearing a cancer has also been obtained, through variations in the culture condition (Diatloff-Zito and Macieira-Coelho, 1982) or following exposure to ionizing radiations (Azzarone et al., 1980; Diatloff and Macieira-Coelho, 1979). In these experiments the increase in lifespan seemed to be correlated with the potential for chromosome rearrangements of the untreated target cells (Macieira-Coelho and Azzarone, 1986). Finally, we also obtained immortal and tumorigenic fibroblast cell lines from some patients with lung cancer but not from normal donors (Azzarone and Pedulla, 1976; Azzarone et al., 1976). However, these cultures presented, from the primary culture on, high levels of aneuploidy and in some instances specific loss of the chromosomes of the G group. Moreover, emergence of the malignant phenotype was only obtained in cultures subjected to a strong and continuous selective pressure, induced by prolonged maintenance in overcrowded conditions (Azzarone and Pedulla, 1976; Azzarone

et al., 1976). It is interesting to note that an increased lifespan was observed in groups exhibiting different degrees of chromosome instability. Two other reports appeared in the literature dealing with spontaneous transformation of apparently normal adult skin fibroblasts: the former concerned a patient bearing a melanoma (Mukherji et al., 1984), the latter a patient with xeroderma pigmentosum (Thielman et al., 1983). Further characterization of fibroblasts from breast cancer patients (Table 1) showed that all the cultures were positive for 3 biological parameters (loss of anchorage dependence, colony formation on contact-inhibited epithelial sheets and increase of saturation densities). These properties can be shared by tumor ceils. In cell lines from patients with benign lesions and in embryonic cells (Table 1), these alterations were absent (Azzarone et al., 1984a). Moreover, within this group we could define a subgroup whose fibroblasts exhibit a significant fraction of cells still synthesizing DNA when the growth curves reach the stationary phase of growth (Table 2). The phenomenon was also detected in skin fibroblasts from patients with other types of cancer and in 2 instances the detection of the defect preceded the discovery of the disease (Azzarone and MacieiraCoelho, 1987), allowing early detection in 1 patient (BMSs), whose mother had developed breast cancer (Azzarone et al., 1984a). Data published elsewhere show that the high percentage of labelled interphases is not due to an increased duration of the S phase relative to the total cell cycle duration, but that these cells probably represent a subset delayed in the G2 phase (Azzarone and MacieiraCoelho, 1987). This defect in the cell cycle progression could be germane to previous observations dealing with the higher sensitivity to radiation in the G2 period of fibroblasts from high-risk cancer patients (Parshad et al., 1985). Overcrowding in this type of cells, even if it did not induce immortalization, caused the selection of a cell subset with new properties commonly found in tumor cells (Table 3). Increased microfilament instability (Rubin et al., 1978), decreased capacity to organize a provisional fibrin matrix by compaction (Azzarone et al., 1981a,b, 1983, 1985a) and acquisition of invasive potential (Marcel, 1979)

323 are properties of tumor cells and could represent coordinate steps leading to the acquisition of an increased autonomy. * Several groups have reported the existence of abnormal properties a n d / o r genetic instability in somatic cells from different types of cancer patients or from patients at high genetic risk of cancer (Chanduri et al., 1975; Smith et al., 1976; Schor et al., 1985; Duming et al., 1984; Bauer et al., 1979; Hsu et al., 1985; Krizman et al., 1987; Dosaka et al., 1987; Pfeffer et al., 1976; Kopelovich, 1982; Delhanty et al., 1983). Moreover, it was suggested that the skin fibroblasts from cancer patients exhibiting abnormal interactions with collagen (Schor et al., 1985) or from ACR patients (Kopelovich, 1982) could represent cells displaying a fetal phenotype. For these reasons we investigated the presence of oncofetal membrane markers in both types of cells (Azzarone et al., 1986). Our results suggest that fibroblasts from breast cancer patients do not exhibit the membrane organization characteristic of embryonic cells (expression of 4F2 and TAC antigens), whereas fibroblasts from ACR patients display a 4F2 cellular distribution similar to that found in embryonic fibroblasts (Fig. 5) but maintain the loss of the IL2 receptor typical of normal adult cells (Fig. 6). In vivo tumor stroma displays, when compared with the normal counterpart, an altered tissular organization. This involves vessel architecture (Denekamp and Hobson, 1982), cell-fibrin interactions with a non-self limited formation of new provisional matrix (Dvorack, 1986) or the presence of markers which are not found in the normal counterparts (Bartal et al., 1986). Data presented in this paper showing that fibroblasts from cancer patients may develop altered mechanisms for actin polymerization and fibrin compaction (Table 3) and previous results concerning disorganization of actin cables (Kopelovich, 1982) and reduction of FCR activity (Azzarone and Macieira-Coelho, 1984) in ACR patients, as well as the abnormal cell-collagen interactions reported by Schor et al. (1985), could be the in vitro counterpart of some of the defects leading to the formation of the tumor stroma. Thus the ubiquitous presence in some cancer patients of abnormal fibroblast cells could open

new issues in the interpretation of the physiopathology of human cancer. These cells could, on the one hand, cause persistent dysfunctions in the normal m e s e n c h y m a l - e p i t h e l i a l interactions, which are necessary for the correct differentiation of epithelial organs (Drew and Drew, 1977; Durnberger and Kratochwil, 1980; Kratochwil and Schwartz, 1976) and, on the other hand, favor the formation of the less-organized stromal tumor tissue. Both mechanisms, acting as a basic systemic disorder, may in certain individuals influence the development of an epithelial tumor. Finally our results suggest that a better understanding of the initial steps leading to immortalization and neoplastic transformation could be obtained by the contemporary study of the modified 'normal' somatic cells from cancer patients in addition to tumor cells, which have inevitably undergone profound changes. Acknowledgements

This work has been funded with contracts from ARC 3055 and Ligue pour la lutte contre le cancer Dept. Yvelines. References

Azzarone, B., and D. Pedulla (1976) Spontaneous transformation of human fibroblast cultures derived from bronchial carcinomata, Eur. J. Cancer, 12, 557-561. Azzarone, B., and A. Macieira-Coelho(1984) Role of cytoskeletal elements in the retractile activity of human skin fibroblasts, Exp. Cell. Re*., 155, 299-304. Azzarone, B., and A. Macieira-Coelho(1987) Further characterization of the defects of skin fibroblasts from cancer patients, J. Cell. Science, 87, 155-162. Azzaxone, B., D. Pedulla and C.A. Romanzi (1976) Spontaneous transformation of human skin fibroblasts derived from neoplastic patients, Nature (London), 262, 74-75. Azzarone, B., C. Diadoff-Zito, C. Billard and A. MacieiraCoelho (1979) Effect of low dose rate irradiation on the division potential of cells in vitro, In Vitro, 16, 634-638. Azzarone, B., D. Brouty-Boy6and A. Macieira-Coelho(1981a) Relationship between fibrin clot retraction and tumorigenesis in C3H/10T1/2 cells, Int. J. Cancer. 28, 799-803. Azzarone, B., C. Curatolo, G. Carloni, M.B. Doanti, L. Morasca and A. Macieira-Coelho(1981b) Fibrin clot retraction in normal and transformed avian fibroblasts, J. Natl. Cancer Inst., 67, 89-94. Azzarone, B., A. Macieira-Coelho,L. Curatolo, D. Brouty-Boy6, O. Varnier, M.B. Donati and L Morasca (1983) Fibrin clot retractile activity in normal, established, and tumorigenic

324 human epithelial cells in culture, J. Natl. Cancer Inst., 71, 879-885. Azzarone, B., M. Mareel, C. Billard, P. Scemama, C. Chaponnier and A. Macieira-Coellio (1984a) Abnormal properties of skin fibroblasts from patients with breast cancer, Int. J. Cancer, 33, 759-764. Azzarone, B., H. Suarez, M.-C. Mingari, L. Moretta and A.S. Fauci (1984b) 4F2 monoclonal antibody recognizes a surface antigen on spread human fibroblasts of embryonic but not of adult origin, J. Cell. Biol., 98, 1133-1137. Azzarone, B., G. Carloni, M. Mareel, O.E. Vamier and A. Macieira-Coelho (1985a) Role of retroviral oncogenes in the attachment of cells to natural substrata, Humana Press, pp. 473-483. Azzarone, B., Y. Malpitce, P. Zaech, L. Moretta, A. Fauci and H. Suarez (1985b) Analysis of the expression of the 4F2 surface antigen in normal and neoplastic fibroblastic human cells of embryonic and adult origin, Exp. Cell. Res., 159, 451-462. Azzarone, B., P. Eid, Y. Andeol, A.S. Fauci and H. Suarez (1986) Evolution in the structure and distribution of 4F2antigen from the oncofetal to the adult phenotype of human fibroblasts, Int. J. Cancer, 38, 177-181. Bartal, A.H., C. Lichtig, C.C. Cardo, C. Feit, E. Robinson and Y. Hirshaut (1986) Monoclonal antibody defining fibroblasts appearing in fetal and neoplastic tissues, J. Natl. Cancer Inst., 76, 415-419. Bauer, E.A., J. Uitto, R.C. Waiters and A.Z. Eisen (1979) Enhanced eollagenase production by fibroblasts derived from human basal cell carcinoma, Cancer Res., 39, 4594-4599. Chanduri, S., L. Koproska and J. Rowinski (1975) Different agglutinability of fibroblasts underlying various precursor lesion of human uterine cervical carcinoma, Cancer Res., 35, 2350-2354. Chaponnier, C., P.A. Janmey and H.L. Yin (1986) The actin filament-severing domain gelsolin, J. Cell. Biol., 103, 1473-1481. Delhanty, J.D., M.B. Davis and J. Wood (1983) Chromosome instability in lymphocytes, fibroblasts and colon epitheliallike cells from patients with familial polyposis coli, Cancer Gen. Cytogen., 8, 27-50. Denekamp, J., and B. Hobson (1982) Endothelial-cell proliferation in experimental tumours, Br. J. Cancer, 46, 711-720. Diatloff, C., and A. Macieira-Coellio (1979) Effect of lowdose-rate irradiation on the division potential of cells in vitro. V. Human skin fibroblasts from donors with a high risk of cancer, J. Natl. Cancer Inst., 63, 55-59. Diatloff-Zito, C., and A. Macieira-Coelho (1982) Effect of growth arrest on the doubling potential of human fibroblasts in vitro: a possible influence of the donor, In Vitro, 18, 606-610. Dosaka, H., A. Synti, S. Motomichi, S. Miyamoto and Y. Kawakami (1987) Sister chromatid exchange induction by benzo[a]pyrene in cultured peripheral blood lymphocytes of lung cancer patients and healthy individuals with or without familial history of neoplasms, Int. J. Cancer, 39, 329-332.

Drews, U., and U. Drews (1977) Regression of mouse mammary gland anlagen in recombinants of Tfm and wild-type tissues: telgosterone acts via the mesenchyme, Cell, 10, 401-404. Durnberger, H., and K. Kratochwil (1980) Specificity of tissue interaction and origin of mesenchymal cells in the androgen response of the embryonic mammary gland, Cell, 19, 465-471. Durning, P., S.L. Schor and R.A.S. Sellwood (1984) Fibroblasts from patients with breast cancer show abnormal migratory behaviour in vitro, Lancet, i, 890-892. Dvorack, H.S. (1986) Tumours: wounds that do not heal. Similarities between tumor stroma generation and wound healing, New Engl. J. Med., 315, 1650-1659. Freedman, V.H., and S. Shin (1977) Isolation of human diploid cell variants with enhanced colony-forming efficiency in semisolid medium after a single-step chemical mutagenesis, J. Natl. Cancer Inst., 58, 1873-1875. Freeman, A.E., R.S. Lake, H.J. Igel, L. Gernand, M.R. Pezzuti, J.M. Malone and W.F. Benedict (1977) Heteroploid conversion of human skin cells by methylcholantrene, Proc. Natl. Acad. Sci. (U.S.A.), 74, 2451-2455. Hayflick, L. (1965) The limited in vitro lifetime of human duplo'fd strains, Exp. Cell. Res., 35, 614. Hsu, T.C., L.M. Cherry and N.A. Samaan (1985) Differential mutagen susceptibility in cultured lymphocytes of normal individuals and cancer patients, Cancer Genet. Cytogenet., 17, 307-313. Kakunaga, T. (1978) Neoplastic transformation of human diploid fibroblast cells by chemical carcinogens, Proc. Natl. Acad. Sci. (U.S.A.), 75, 1334-1338. Kopelovich, L. (1982) Hereditary adenomatosis of the colon and rectum: relevance to cancer promotion and cancer control in humans, Cancer Genet. Cytogenet., 3, 333-351. Kraemer, P.M., F.A. Ray, A.R. Brothman, M.F. Bartholdi and L.S. Cram (1986) Spontaneous immortalization rate of cultured Chinese hamster cells, J. Natl. Cancer Inst., 76, 703-709. Kratochwil, K., and P. Schwartz (1976) Tissue interaction in androgen response of embryonic mammary rudiment of mouse: identification of target tissue for testosterone, Proc. Natl. Acad. Sci. (U.S.A.), 73, 4041-4044. Krizrnan, D.B., S. Pathak, N.A. Samaan and R.C. Hickey (1987) Genetic instability in fibroblasts of patients with thyroid cancers (TC), Int. J. Cancer, 39, 179-181. Low, R.B., C. Chaponnier and G. Gabbiani (1981) Organization of actin in epithelial cells during regenerative and neoplastic conditions, Lab. Invest., 44, 359-367. Lynch, H.T., W.A. Albano, B.S. Danes, M.A. Layton, W.A. Kimberling, J.F. Lynch, S.C. Cheng, K.A. Costello, G.M. Mulchahi, C.A. Wagner and S.L. Tindall (1984) Genetic predisposition to breast cancer, Cancer, 53, 612-622. Macieira-Coelho, A. (1980) Implications of reorganization of the cell genome for aging and immortalization of dividing cells in vitro, Gerontology, 26, 276-282. Macieira-Coelho, A., and B. Azzarone (1986) Aging, cancer, and the life span of cells in vitro, In: D. Platt (Ed.), Drugs and Aging, Springer Verlag, Heidelberg, pp. 35-44.

325 Maher, V.M., L.A. Rowan, K.C. Silinskas, S.A. Kateley and J.J. McCormick (1982) Frequency of UV-induced neoplastic transformation of diploid fibroblasts is higher in xeroderma pigmentosum cells than in normal cells, Proc. Natl. Acad. Sci. (U.S.A.), 79, 2613-2617. Mareel, M. (1979) Is invasiveness in vitro characteristic of malignant cells?, Cell. Biol. Int. Reports, 3, 637. Mare.l, M., T. Kim and C. Meyvisch (1979) Methods of study of the invasion of malignant C3H mouse fibroblasts into embryonic chick heart in vitro, Virchows Arch. B., 30, 95. Martin, G.M., C.A. Spragne and C.J. Epstein (1970) Replicative life span of cultivated human cells. Effects of donor's age, tissue and genotype, Lab. Invest., 23, 86-92. Matsumura, T., M. Hayashi and R. Konishi (1985) Immortalization in culture of rat cells: a genealogic study, J. Natl. Cancer Inst., 74, 1223-1232. Milo, G.E., and DiPaolo, J.A. (1978) Neoplastic transformation of human diploid cells in vitro after chemical carcinogen treatment, Nature (Lond.), 275, 130-132. Mukherji, B., T.J. Macalister, A. Guha, C.G. Chilies, C. Jeffers and S.K. Slocum (1984) Spontaneous in vitro transformation of human fibroblasts, J. Natl. Cancer Inst., 73, 583-593. Namba, M., K. Nishitani, F. Hyodoh, F. Fukishima and T. Kimoto (1985) Neoplastic transformation of human diploid fibroblasts (KMST-6) by treatment with 6°Co gamma rays, Int. J. Cancer, 35, 275-280. Oshimura, M., T.M. Gilmer and J.C. Barrett (1985) Nonrandom loss of chromosome 15 in Syrian hamster tumours induced by v-Ha-ras plus v-myc oncogenes, Nature (Lond.), 316, 636-639. Parshad, R., K.K. Sanford and G.M. Jones (1985) Chromosomal radiosensitivity during the G2 cell-cycle period of skin fibroblasts from individuals with familial cancer, Proc. Natl. Acad. Sci. (U.S.A.), 82, 5400-5403.

Pfeffer, L.M., and L. Kopelovich (1977) Differential genetic susceptibility of cultured human skin fibroblasts to transformation by Kirsten's murine sarcoma virus, Cell, 10, 313. Rubin, R.W., R.H. Warren, D.S. Lukeman and E. Clements (1978) Actin content and organization in normal and transformed cells in culture, J. Cell. Biol., 78, 28. Sacks, Jr., G.H. (1982) Human cell transformation by Simian virus 40, a review, In vitro, 17, 1-19. Sager, R. (1983) Resistance of human cells to oncogenic transformation, Cancer Cells, 2, 487-493. Schor, S.L., J.A. Haggie, P. Durning, A. Howell, R.A.S. Sellwood and D. Crowther (1986) Occurrence of a fetal fibroblast phenotype in familial breast cancer, Int. J. Cancer, 37, 831-836. Schor, S.L., A.M. Schor, P. Durning and G. Rushton (1985) Skin fibroblasts obtained from cancer patients display foetal-like migratory behaviour on collagen gels, J. Cell. Sci., 73, 235-244. Smith, H.S., R.B. Owens, A.J. Hiller, W.A. Nelson-Rees and J.O. Johnston (1976) The biology of human cells in tissue culture. I. Characterization of cells derived from osteogenic sarcomas, Int. J. Cancer, 17, 219-234. Terzi, M., and T.S.C. Hawkins (1975) Chromosomal variation and the establishment of somatic cell lines in vitro, Nature (Lond.), 253, 361-362. Thielmann, H.W., E. Fischer, R.T. Dzarlieva, D. Komitowski, O. Popanda and L. Edler (1983) Spontaneous in vitro malignant transformation in a xeroderma pigmentosum fibroblast line, Int. J. Cancer, 31, 687-700. Todaro, G.J., and H. Green (1963) Quantitative studies of the growth of mouse embryo cells in culture and their development into established lines, J. Cell. Biol., 17, 299-313.