Clin. Exp. Metastasis, 1992, 10, 259-266

Correlation between spontaneous metastatic potential and type I collagenolytic activity in a human pancreatic cancer cell line (SUIT-2) and sublines Shoji Taniguchi, T a k e s h i I w a m u r a and T a k e t o Katsuki

First Department of Surgery, Miyazaki Medical College, Miyazaki, Japan (Received 2 January 1992; accepted 6 April 1992)

A human pancreatic cancer cell line (SUIT-2) and four sublines cloned in vitro ($2-007, $2-013, $2-020 and $2-028) were inoculated into nude mice for assessment of metastatic potentials. After 16 weeks of subcutaneous injection, the parent SUIT-2 line metastasized to the lungs and lymph nodes in three of six mice. $2-007 cells presented the highest metastatic potential in pulmonary (5/6) and lymph node (2/6) metastases among the four sublines. No metastasis was found in $2-028. The incidence of spontaneous pulmonary metastasis was correlated with that of pulmonary colonization after intravenous (i.v.) injection of cell clusters (r = 0.87, P = 0.056). Pulmonary colonization potential using single cells, however, did not always reflect a spontaneous metastatic ability. Type I collagenolytic activity in serum-free conditioned media of these cells was correlated effectively with the incidence of spontaneous pulmonary metastasis (r = 0.92, P = 0.026) and pulmonary colonization after i.v. injection of cell clusters (r = 0.95, P = 0.013). Thus, type I coilagenolytic activity may possibly be essential to spontaneous cancer metastasis. Keywords: cell line, human pancreatic cancer, nude mouse, metastasis, type I collagenase

Introduction Recent data regarding the mechanisms of cancer metastasis suggest metastatic heterogeneity in clonal cells from a cancer [1-3], and using high metastatic cells, the relationship between metastasis and proteases, which degrade the extracellular matrix, has been investigated [4-7]. Degradation of the extracellular matrix, including collagen, is a necessary step for invasion and metastasis by cancer cells [8]. Collagenase secreted by cancer cells may degrade collagen, and collagen degradation products may enhance the migration of cancer cells [9, 10]. A human pancreatic cancer cell line (SUIT-2) Address for correspondence: Dr S. Taniguchi, Dept. Surg. Koga General Hospital 2 - 5 - 7 Takachiho-Dori, Miyazaki-City, Miyazaki 880, Japan. Fax: (+81) 985 28-6861.

1992 Rapid Communications of Oxford Ltd

from a liver metastasis of a human pancreatic cancer has been shown to metastasize to regional lymph nodes [11] and the lungs in nude mouse xenograft, and 28 sublines have been cloned from SUIT-2 by soft agar culture [12]. In the present study, metastatic heterogeneity and the relationship between metastatic potential and type I collagenolytic activity in SUIT-2 and its sublines is discussed.

Materials and methods

Mice Eight- to 16-week-old nude mice (BALB/c/nu/nu Sic) were used.

Clinical & Experimental Metastasis Vol 10 No 4 259

S. Taniguchi et al. Cells A human pancreatic cancer cell line (SUIT-2) and four sublines cloned in vitro--a clone producing high levels of carcinoembryonic antigen (CEA) and low levels of carbohydrate antigen 19-9 (CA19-9) [$2-007], a high CA19-9 and low CEA clone ($2-028), and clones high in both ($2-013) and low in both (S2-020)--were established and maintained for over 4 years [11, 12]. All cells were cultured in Dulbecco's modified Eagle's medium (DMEM; Flow Laboratories, Irvine, UK) supplemented with 10% fetal bovine serum (Flow, McLean, VA, USA), penicillin (100U/ml) and streptomycin (100/~g/ml) (full medium) at 37°C in a humidified incubator with an atmosphere of 5% CO2 in air.

Xenografts Cell clusters were obtained from confluent culture dishes by trypsinization (0.25% trypsin with 1 mM EDTA; Gibco Laboratories, Grand Island, NY, USA) for 5 min. The 'cell clusters' were comprised of cell clumps of 2-5 cells in 50-70%, 6-10 cells in 5-20%, and 10-30% of single cells in all cell lines except $2-028, in which clumps of 2-3 cells and single cells were proportioned almost evenly. In SUIT-2, $2-007 and $2-013, however, the 'cell clusters' contained large cell clumps of more than 11 cells in a few per cent of cell clusters. The 'single cells' were obtained from thinly seeded dishes by trypsinization for 1 min, with the composition of single cells more than 90%. The harvested cells were rinsed first with a full medium, then twice with Ca 2+- and Mg2+-free phosphate-buffered saline [PBS(-)], and resuspended in PBS(-). The cells were injected subcutaneously (s.c.) into the flank (1 x 107 cells/0.2 ml/mouse) or lateral tail vein (i.v.; 5 × 105 cells/0.1 ml/mouse). The size of the resulting tumor after s.c. injection was measured and volume estimated by the formula V = L x W × HI2, where V is the volume, L the length, W the width and H the height.

Evaluation of spontaneous metastasis and pulmonary colonization The mice were killed at 12 or 16 weeks after s.c. injection, at 8 weeks after i.v. injection, or when they were in the process of dying. The lungs were dissected after injection of about 2 ml of 15% black India ink through the trachea and fixed in Fekete's solution (a mixture containing 100 ml of 70% alcohol, 10 ml of formalin, and 5 ml of glacial acetic acid). White nodules on the black lung surface were counted on the following day under a

260

Clinical & Experimental Metastasis Vol 10 No 4

dissecting microscope [13] and subsequently ascertained histologically as metastatic loci. Regional and mediastinal lymph nodes were fixed with 10% phosphate-buffered formalin and examined histologically.

Preparation of serum-free conditioned medium Cells were cultured in an 80 mm 2 culture flask (Nunc, Roskilde, Denmark) in 15 ml of full medium. The numbers of seeding cells were determined preliminarily so that the cells would be confluent at the time of collecting serum-free conditioned medium (SFCM). On the second day, the cells in the flask were rinsed twice with PBS(-), and the medium was replaced with 15 ml of a mixture of DMEM and Ham's F-12 (Flow) (1:1) without any supplements (DME/Ham12). On the third day, the medium was replaced again with 15 ml of DME/Ham12. After 24 h of culture, the SFCM was collected and centrifuged at 500 g for 10 min to remove any floating cells. The supernatant was concentrated about 5-fold using a Centriprep-10 (Amicon Division, Grace Co., Danvers, MA, USA) and stored at -30°C. The cell number was counted at the time, collecting the SFCM.

Type I collagenolytic activity Type I collagenolytic activity was estimated by the Collageno-kit (Collagen Gijutsu Kenshukai, Tokyo, Japan). Sample (200 #l) and FITC-labeled type I collagen in soluble form (200/zg/200 #l) were incubated at 35°C for 2 h, and reincubated at 35°C for l h after reaction was stopped with 80 mM o-phenanthrorin (10/~l) so as to denature only the digested collagen, denaturable below 35°C. Ethanol (400/~l) was added for extracting denatured products. After precipitating the intact collagen by centrifugation (3000 rev/min for 10 min), the fluorescent intensity (FI) of the supernatant was measured with excitation and emission at 495 and 520 nm. Type I collagenolytic activity of the sample was estimated as the percentage of FI of the positive control whose collagen was denatured by incubation at 80°C for 10 min [14, 15]. Concentrated DME/Ham12 was used as the negative control. To determine the activation of the latent collagenase, a sample was incubated with 4/~g of trypsin (type XIII; Sigma Chemical Co., St Louis, MO, USA) for 10 min at 37°C and then with 30/~g of soybean trypsin inhibitor (SBTI, type II-S; Sigma) for 10 min at 37°C. The reaction time was 10 min, as determined from a time course curve. Concentrated DME/Ham12 containing the same

Metastasis and type I collagenolytic activity amount of trypsin/SBTI was used as the negative control. Each assay was performed in triplicate. Type I collagenolytic activities of SFCMs (unit/101° cells. 24 h) were compared. Type I collagenolytic activity of SUIT-2 was confirmed as collagenase activity through the detection of type I collagen degradation products peculiar to digestion by vertebrate collagenase in SDS-PAGE.

S D S - P A G E analysis of type IV collagenolytic activity One hundred ~g/100/zl of type IV collagen from the E n g e l b r e t h - H o l m - S w a r m transplantable mouse tumor (Biomedical Products Division, Collaborative Research Incorporated, Two Oak Park, Bedford, MA, USA), were incubated at 37°C for 10 h with 100 #1 of trypsin-activated type IV collagenase in SFCMs of SUIT-2 and its sublines, dissolved in 0 . 5 M Tris-HC1 (pH 6.8)/5% 2mercaptoethanol/2% sodium dodecyl sulfate, and analysed electrophoretically in an 8% separating gel (TEF Co., Nagoya, Japan). Proteins in the gel were stained with 0.1% Coomassie Brilliant Blue. The gel was photographed through green filter and

dried using a gel-dry kit (TEF Co.). Concentrated D M E / H a m 1 2 was used as the negative control. One hundred /zg/100/zl, 75/zg/100/zl, 50/zg/100 pl and 25/~g/100/zl of type IV collagen were mixed with 100/tl of a buffer solution and electrophoresed in the same gel as the standards.

Results

Tumor growth after s.c. injection Doubling times of the s.c. implanted tumors were 8.7, 9.7, 11.5, 10.8 and 21.9 days and cutaneous ulceration over the tumors was observed at rates of 8/11, 11/16, 4/11, 8/11 and 1/11 in SUIT-2, $2-007, $2-013, $2-020 and $2-028, respectively.

Spontaneous metastasis after s.c. injection Lymph nodes and pulmonary metastases from the s.c. tumor of SUIT-2 are shown in Figure 1. The incidence of pulmonary metastasis was highest in $2-007 both at 12 (63%) and 16 weeks (83%) after s.c. injection (Table 1). The incidence of lymph node metastasis was also highest in $2-007 among the four sublines but less than that of the parent

Figure 1. Lymph nodes (A, B) and pulmonary metastasis (C, D) from a s.c. inoculated tumor of SUIT-2. (A) Macroscopic appearance. Skin around the tumor has been removed. Lymph vessels between giant s.c. tumor and axillary lymph nodes are parallel to the blood vessels. (B) Histological appearance. The lymph node has been virtually replaced by tumor cells. Tumor cells can be seen in perinodal lymph vessels. H&E, ×30. (C) Macroscopic appearance. The lung was treated by the method of H. Wexler. White metastatic nodules on the black lung surface are evident. (D) Histological appearance. A colony of cancer cells with large irregular shaped nuclei is present in the lung. H&E, x135.

Clinical& ExperimentalMetastasisVo110No 4 261

S. Taniguchi et al. Table 1. Pulmonary and lymph node metastasis after s.c. injection of the tumor cells Cell line

SUIT-2 $2-007 $2-013 $2-020 $2-028

Experiment 1~

Experiment 2 b

Experiment

Lung (%)

Lymph node (%)

Lung (%)

Lymph node (%)

1/5(20) 2/3(67) 0/5(0) 1/5(20) 0/5(0)

2/5(40) 0/3(0) 0/5(0) 1/5(20) 0/5(0)

3/6(50) 5/6(83) 2/6(33) 2/6(33) 0/6(0)

3/6(50) 2/6(33) 1/6(17) 0/6(0) 0/6(0)

3Q

Lung (%)

Lymph node (%)

3/5(60)

2/5(40)

The cells were injected s.c. into the flank of nude mice [1 x 107 cells/0.2 ml PBS(-)/mouse]. The mice were killed at 12 or b 16 weeks after injection. SUIT-2 cell line. In contrast, no metastatic lesion could be detected in $2-028.

Colonization after i.v. injection All mice had p u l m o n a r y colonies after i.v. injection of SUIT-2 or $2-007 cell clusters, but only one of ten mice following i.v. injection of $2-028 cell clusters. W h e n single cells were i.v. injected instead of the cell clusters, the incidence of pulmonary colonization was reduced except in the instance of $2-028 (Table 2). P u l m o n a r y colonies of miliary pattern could be seen in $2-020. Metastases in the adrenal glands in $2-007 and $2-020 and mediastinal lymph nodes in SUIT-2 were also observed.

Type I collagenolytic activity of SFCM Native and activated type I collagenolytic activity was significantly low in $2-028 ( P < 0 . 0 1 ) , the lowest metastatic subline (Table 3). Sixty-three per cent of activated type I collagenolytic activity was inhibited by the addition of 2.5 m M E D T A in SUIT-2 and 16% by 0.5 m M E D T A .

SDS-PAGE analysis of type IV collagenolytic activity Figure 2 shows the reduction of two o~ chains of type I V collagen by treatment with type IV collagenase in SFCMs from the cells. $2-028 showed lower type I V collagenolytic activity than the other sublines.

2. Pulmonary colonization after i.v. injection of the tumor cells

Discussion

Cell line

Incidence (%)

No. of lung nodules

Experiment 1 SUIT-2:c Q $2-007:c $2-013:c $2-020:c $2-028:c

4/4(100) 6/6(100) 5/6(83) 3/6(50) 0/6(0)

86, 112, 151, 173 1, 10, 32, 40, 172, 182 0, 2, 3, 5, 30, 35 0, 0, 0, 2, 18, >200 0, 0, 0, 0, 0, 0

Considerable attention has been directed to differences in biological behavior between high- and low-metastatic clonal cancer cells for clarification of the mechanisms of cancer metastasis [6, 16-18].

Experiment 2 SUIT-2:s b $2-007:s $2-013:s $2-020:s $2-028:s

4/6(67) 1/5(20) 3/6(50) 1/6(17) 1/6(17)

0, 0, 0, 0, 0,

Experiment 3 $2-007:s $2-028:c

2/5(40) 1/4(25)

0, 0, 0, 3, 11 0, 0, 0, 1

Table

0, 0, 0, 0, 0,

1, 0, 0, 0, 0,

6, 0, 3, 0, 0,

16, 30 2 14, 26 0, >200 0, 1

The cells were injected into the lateral tail vein of nude mice [5 x 105 cells/0.1 ml PBS(-)/mouse] as ~'cell clusters' or as b'single cells'. The mice were killed at 8 weeks after injection.

262

Clinical& ExperimentalMetastasisVol 10 No 4

Table 3. Type I collagenolytic activity of SFCM Cell line Native (unita/101° cells.24 h)

Activated (unit/107 cells.24 h)

SUIT-2 $2-007 $2-013 $2-020 $2-028

77.1 67.8 59.7 66.7 2.1

22.4 23.7 18.2 18.1 7.5

_ 0.9 c + 5.3 --- 1.0 + 2.0 + 1.2 b

_+ 8.2 _+ 4.5 + 1.6 d + 17.2 + 3.0 b

al unit = degradation of 1 #g coUagen/min. bLower than the values of other cell lines with statistical significance (P < 0.01) by Student's t-test. CHigher than $2-013 (P < 0.01) and $2-020 (P < 0.05). dLower than SUIT-2 and $2-007 (P < 0.05).

Metastasis and type I collagenolytic activity

a

b

c

d

f

e

g

h

i

j

pro a,1 ( I V ) pro o : 2 ( i v ) . 7 "

Figure 2. S D S - P A G E analysis of type IV collagenolytic activity. One hundred /~g/100/A of type IV collagen from EHS tumor was incubated at 37 ° for 10 h with 100/4 of sample [a: concentrated DME/Ham12, negative control; b: SFCM from SUIT-2 (2 x 107 cells.24 h/ml); c: SFCM from $2-007 (2 x 107 cells.24h/ml); d: SFCM from $2-013 (2 x 107 cells.24 h/ml); e: SFCM from $2-020 (2 × 107 cells.24 h/ml); f: SFCM from $2-028 (2 × 107 cells.24 h/ml)]. Various amounts of type IV collagen (g: 100/zg/100/A; h: 75/~g/100 ~l; i: 50 ~g/100/~l; j: 25 #g/100/A) were mixed with 100//1 of a buffer solution. The reaction products and type IV collagen solutions were electrophoresed in an 8% separating gel. The reduction of two tr chains of type IV collagen by treatments with SFCMs was noticed. $2-028 showed lower activity than the others.

b., 14

(%) 100$2-007 80

(b)

m 0 Q

°~ 60'

..O~oOO~OOoo~..°,.o "''°'°°°°°°°°'°

0

(a)

$2-020 .....-'"'"'""

o Q u Q

u H

0• 0

. 20 Incidence

40 of

60

pulmonary

80

! 100

colonization

(%)

F i g u r e & Correlation of spontaneous metastatic potential with pulmonary colonization in SUIT-2 and sublines. Incidence of spontaneous pulmonary metastasis (16 weeks after s.c injection of 107 cells/mouse) for that of pulmonary colonization (8 weeks after i.v. injection of 5 × 105 cells/mouse) of each line is indicated on the ordinate as (O) following i.v. injection of cell clusters with the correlation line (a) (y = - 6 . 7 + 0.68x, r = 0.87, P = 0.058) and as (O) following i.v. injection of single cells with the correlation line (b) (y = 23.8 + 0.44x, r = 0.32, P = 0.60). These data were ascertained from the results of the present experiments as documented in Tables 1 and 2.

Clinical& ExperimentalMetastasisVol 10 No 4 263

S. T a n i g u c h i et al.

(b)

(%)

i00-

(a)

.H

S2-007

4J

U/

80,

m

4-I 5UIT-2 ~.,.t ~

N

S2-013

$2-020

40 g u @ 'U .,.t u

20]

52-028

M

i0 Type

20

I collagenolytic

activity

30 (unit/l~tcells) (native)

Figure 4. Correlation of spontaneous metastatic potential with type I collagenolytic activity in SUIT-2 and sublines. Incidence of spontaneous pulmonary metastasis (16 weeks after s.c. injection of 107 cells/mouse) or pulmonary colonization (8 weeks after i.v. injection of 5 × 105 cells/mouse in cell clusters) for native type I collagenolytic activity in SFCM of each cell line is indicated on the ordinate as (©) in the former with the correlation line (a) (y = - 3 8 . 6 + 4.4x, r = 0.92, P = 0.027) and as (O) in the latter with the correlation line (b) (y = - 3 4 . 9 + 5.8x, r = 0.95, P = 0.013). These data were ascertained from the results of the present experiments as documented in Tables 1-3. Animal malignant cells, however, have been used in most such research since h u m a n cancer cell lines are hard to metastasize even in experimental animals such as athymic nude mice. Sharkey and Fogh [19] found the metastatic incidence of h u m a n malignant tumors in nude mouse xenograft to be only 1.3% after s.c. inoculation of 106 t u m o r lines, while Kozlowski et al. [20] concluded that the metastatic spread of h u m a n tumors implanted into nude mice is not u n c o m m o n . Although SUIT-2 p e r se is a cell line with high metastatic potential, $2-007, giving m o d e r a t e l y differentiated tabular adenocarcinoma in a nude mouse [12], is the highest metastatic subline and $2-028, giving welldifferentiated papillary adenocarcinoma in a nude mouse [12], is the lowest a m o n g the four present sublines. Liotta et al. [21] stress the significance of t u m o r clumps in metastasis. In the present study, the incidence of p u l m o n a r y colonization after i.v. injection of clusters of each cell line was correlated with that of spontaneous pulmonary metastasis (r = 0.87, P = 0.058) but that of single cells was

264

Clinical & Experimental Metastasis Vol 10 No 4

not (r = 0.32, P = 0.60) (Figure 3). $2-007 indicated a high incidence of spontaneous metastasis and pulmonary colonization after i.v. injection of cell clusters, but a much smaller extent of pulmonary colonization after i.v. injection of single cells. In some high metastatic cancer cells the aggregation of the cells may therefore possibly be important to spontaneous metastasis. The contribution of type IV collagenase to cancer metastasis has been considerably emphasized recently [22-27], and Liotta et al. [4] have shown a positive correlation between metastatic potential and type I V collagenolytic activity in culture media or degradation of type IV collagen by living cells. Eisenbach et al. [7] reported both metastatic and non-metastatic clones of the same t u m o r similarly to degrade type IV collagen. According to a S D S - P A G E analysis, low metastatic $2-028 showed lower type I V collagenolytic activity than the others. T y p e I collagen-specific collagenase is considered important mainly for cancer invasion. Yamanishi et al. [28] noted basal cell epitheliomas,

Metastasis and type I collagenolytic activity locally very invasive but not metastt,'2c, to show higher collagenolytic activity than squamous cell carcinomas. Tarin et al. [29], however, found a positive correlation of secretion of type I collagenspecific collagenase into culture medium with metastatic-colonization potential following the i.v. injection of naturally occurring murine m a m m a r y tumor cells in murine m a m m a r y tumor virus infected mice. It is suggested from the present study using human cancer cells that native type I collagenolytic activity in the conditioned medium was correlated not only with the incidence of pulmonary colonization after i.v. injection of cell clusters (r = 0.95, P = 0.013) but also with that of spontaneous pulmonary metastasis (r = 0.92, P = 0.027) (Figure 4). Trypsin-activated type I collagenolytic activity was also high in high metastatic $2-007 and low in low metastatic $2-028. As far as SUIT-2 and its sublines are concerned, the secretion of type I collagen-specific collagenase should be regarded as one of the important factors for spontaneous metastasis.

7.

8.

9.

10.

11.

12.

Acknowledgements This research was supported in part by a Grant-inAid for E n c o u r a g e m e n t of Y o u n g Scientists from the Ministry of Education, Science and Culture of Japan (No. 01770987).

References 1.

2.

3.

4.

5.

6.

Fidler IJ and Kripke ML, 1977, Metastasis results from pre-existing variant cells within a malignant tumor. Science, 197,893-895. Miller FR, Miller BE and Heppner GH, 1983, Characterization of metastatic heterogeneity among subpopulations of a single mouse mammary tumor: heterogeneity in phenotypic stability. Invasion and Metastasis, 3, 22-31. Price JE, Aukerman SL and Fidler IJ, 1986, Evidence that the process of murine melanoma metastasis is sequential and selective and contains stochastic elements. Cancer Research, 46, 5172-5178. Liotta LA, Tryggvason K, Garbisa S, Hart I, Foltz CM and Shafie S, 1980, Metastatic potential correlates with enzymatic degradation of basement membrane collagen. Nature, 284, 67-68. Mullins DE and Rohrlich ST, 1983, The role of proteinases in cellular invasiveness. Biochimica et Biophysica Acta, 695, 177-214. McLaughlin MEH, Liener IE and Wang N, 1983,

Proteolytic and metastatic activities of clones derived from a methylcholanthrene-induced murine fibrosarcoma. Clinical and Experimental Metastasis, 1, 359-371. Eisenbach L, Segal S and Feldman M, 1985, Proteolytic enzymes in tumor metastasis. II. Collagenase type IV activity in subcellular fractions of cloned tumor cell populations. Journal of the National Cancer Institute, 74, 87-93. Liotta LA, Rao CN and Barsky SH, 1983, Tumor invasion and the extracellular matrix. Laboratory Investigation, 49, 636-649. Nabeshima K, Kataoka H and Koono M, 1986, Enhanced migration of tumor cells in response to collagen degradation products and tumor cell collagenolytic activity. Invasion and Metastasis, 6, 270-286. Maslow DE, 1987, Collagenase effects on cancer cell invasiveness and motility. Invasion and Metastasis, 7,297-310. Iwamura T, Katsuki T and Ide K, 1987, Establishment and characterization of a human pancreatic cancer cell line (SUIT-2) producing carcinoembryonic antigen and carbohydrate antigen 19-9. Japanese Journal of Cancer Research, 78, 54-62. Iwamura T, Taniguchi S, Kitamura N, Yamanari H, Kojima A, Hidaka K, Setoguchi T and Katsuki T, 1992, Correlation between CA19-9 production in vitro and histological grades of differentiation in vivo in clones isolated from a human pancreatic cancer cell line (SUIT-2). Journal of Gastroenterology and Hepatology, in press.

13. Wexler H, 1966, Accurate identification of experimental pulmonary metastases. Journal of the National Cancer Institute, 36,641-645. 14. Terato K, Nagai Y, Kawanishi K and Yamamoto S, 1976, A rapid assay method of collagenase activity using 14C-labeled soluble collagen as substrate. Biochimica et Biophysica Acta, 445,753-762. 15. Terato K, Hashida R, Miyamoto K, Morimoto T, Kato Y, Kobayashi S, Tajima T, Otake S, Hori H and Nagai ¥, 1982, Histological, immunological and biochemical studies on type II collagen-induced arthritis in rats. Biomedical Research, 3,495-505. 16. Yamori T, Tsuruo T, Naganuma K, Tsukagoshi S and Sakurai Y, 1984, Isolation and characterization of highly and rarely metastatic clones from murine colon adenocarcinoma 26. Invasion and Metastasis, 4, 84-97. 17. Grimstad IA, Thorsrud AK and Jellum E, 1988, Marker polypeptides distinguishing between cancer cell clones with high and low potential for spontaneous metastasis in murine fibrosarcoma cells. Cancer Research, 48,572-577. 18. Albini A, Aukerman SL, Ogle RC, Noonan DM, Fridman R, Martin GR and Fidler IJ, 1989, The in vitro invasiveness and interactions with laminin of K-1735 melanoma cells. Evidence for different laminin-binding affinities in high and low metastatic variants. Clinical and Experimental Metastasis, 7, 437-451.

Clinical & Experimental Metastasis Vol 10 No 4

265

S. Taniguchi et al. 19. Sharkey FE and Fogh J, 1979, Metastasis of human tumors in athymic nude mice. International Journal of Cancer, 24, 733-738. 20. Kozlowski JM, Fidler IJ, Campbell D, Xu Z, Kaighn ME and Hart IR, 1984, Metastatic behavior of human tumor cell lines grown in the nu6e mouse. Cancer Research, 44, 3522-3529. 21. Liotta LA, Kleinerman J and Saidel GM, 1976, The significance of hematogenous tumor cell clumps in the metastatic process. Cancer Research, 36, 889-894. 22. Liotta LA, Abe S, Robey PG and Martin GR, 1979, Preferential digestion of basement membrane collagen by an enzyme derived from a metastatic murine tumor. Proceedings of the National Academy of Sciences, USA, 76, 2268-2272. 23. Salo T, Liotta LA, Keski-Oja J, TurpeenniemiHujaen T and Tryggvason K, 1982, Secretion of basement membrane collagen degrading enzyme and plasminogen activator by transformed cells -- Role in metastasis. International Journal of Cancer, 30, 669-673. 24. Turpeenniemi-Hujanen T, Thorgeirsson UP, Hart IR, Grant SS and Liotta LA, 1985, Expression of collagenase IV (basement membrane collagenase) activity in murine tumor cell hybrids that differ in metastatic potential. Journal of the National Cancer Institute, 75, 99-103.

266 Clinical& Experimental Metastasis Vol 10 No 4

25. Garbisa S, Pozzatti R, Muschel ILl, Saffiotti U, Ballin M, Goldfarb RH, Khoury G and Liotta LA, 1987, Secretion of type IV collagenolytic protease and metastatic phenotype: induction by transfection with c-Ha-ras but not c-Ha-ras plus Ad2-Ela. Cancer Research, 47, 1523-1528. 26. Garbisa S, De Giovanni C, Biagini G, Vasi V, Grigioni WF, D'Errico A, Mancini AM, Del Re B, Lollini P-L, Nanni P, Nicoletti G and Prodi G, 1988, Different metastatic aggressiveness by murine TS/A clones: ultrastructure, extracellular glycoproteins and type IV collagenolytic activity. Invasion and Metastasis, 8, 177-192. 27. Ura H, Bonfil RD, Reich R, Reddel R, Pfeifer A, Harris CC and Klein-Szanto AJP, 1989, Expression of type IV collagenase and procollagen genes and its correlation with the tumorigenic, invasive, and metastatic abilities of oncogene-transformed human bronchial epithelial cells. Cancer Research, 49, 4615-4621. 28. Yamanishi Y, Dabbous MK and Hashimoto K, 1972, Effect of collagenolytic activity in basal cell epithelioma of the skin on reconstituted collagen and physical properties and kinetics of the crude enzyme. Cancer Research, 32, 2551-2560. 29. Tarin D, Hoyt BJ and Evans D J, 1982, Correlation of collagenase secretion with metastatic colonization potential in naturally occurring murine mammary tumours. British Journal of Cancer, 46,266-278.