Invasion of an Artificial Blood Vessel Wall by Human Fibrosarcoma Cells1 · Artificial blood vessel...

[CANCER RESEARCH 41, 4613-4620, November 1981]0008-5472/81 /0041-OOOOS02.00

Invasion of an Artificial Blood Vessel Wall by Human Fibrosarcoma Cells1

Peter A. Jones, Harry B. Neustem, Felicidad Gonzales, and Emil Bogenmann

Departments of Pediatrics and Biochemistry, Division of Hematology-Oncology Â¡P.A. J.. F. G., E. B.Â¡,and Department of Pathology Â¡H.B. N.J. Children* Hospital ofLos Angeles and University of Southern California School of Medicine, Los Angeles, California 90027

ABSTRACT

Artificial blood vessel walls constructed by the addition ofbovine arterial endothelial cells to multilayers of rat smoothmuscle cells were used as substrates for the human fibrosarcoma cell line HT1080. The extracellular matrix proteins elaborated by the smooth muscle cells were prelabeled with [3H]-

proline; therefore, their subsequent digestion could be followedby the appearance of radioactivity in the culture medium. Thefibrosarcoma cells rapidly hydrolyzed smooth muscle multilayers in the absence of endothelial cells, but an endothelial layermarkedly retarded the destructive ability of the tumor cells. Theprotective effect of the endothelium was not due to a lack ofpenetration of this cell layer, since HT1080 cells were observedby light and electron microscopy to be in the subendothelialarea 24 hr after plating. Subsequently, the tumor cells multiplied in the region between the endothelial and smooth musclelayers and, although their degradative ability was retarded,they were ultimately capable of destroying the structure. Endothelial cells also inhibited hydrolysis of the smooth musclelayers if added simultaneously or up to 1 week after HT1080cells, but the degree of inhibition was not as great as that seenwith a preestablished endothelial layer. Measurable inhibitionof tumor cell degradative activity was observed at fibrosar-

coma:endothelial cell ratios of 25:1, demonstrating the potencyof endothelial cells in modulating this aspect of the invasivephenotype. Although the HT1080 cells only slowly degradedthe preexisting matrix proteins in artificial vessel wall cultures,they interfered with the production of new connective tissueproteins which occurred in control cultures. These experimentstherefore suggest that endothelial cells have profound effectson tumor cell proteolytic activity, and the significance of theseobservations to tumor cell extravasation in vivo is discussed.

INTRODUCTION

The exit of malignant cells from the circulation is an importantrequirement for the successful establishment of a secondarytumor during blood-borne metastasis. Most of our knowledge

of extravasation comes from morphological observations ofmetastasis in animal tumor models (4), and little is known ofthe cellular and biochemical properties of tumor cells whichare responsible for this process. This has been due primarilyto the difficulty in obtaining suitable model systems that allowfor simultaneous morphological and biochemical observations.

The process of extravasation in vivo begins after the adherence of circulating tumor cells to vascular endothelium, andthere seem to be several routes by which malignant cells crossthe endothelium and gain access to extravascular tissues.

1This work was supported by Grant CD-18 from the American Cancer Society,

by Grant 83.828.0.80 from the Swiss Science Foundation, and partly by the MaxFactor Family Foundation.

Received May 29, 1981 ; accepted August 5, 1981.

Several studies have indicated that the endothelium isbreached between cell-cell junctions (7, 16, 21), although it isnot clear whether the tumor cells induce endothelial retractionor take advantage of naturally occurring gaps (22, 25). Otherstudies have suggested that metastatic cells might follow thepath set by emigrating leukocytes or lymphocytes (19, 26),migrate directly through endothelial cell cytoplasms (3) orcause the cytotoxic or enzymatic lysis of endothelial cells (5,23). Once tumor cells have penetrated the endothelium, theyoften become sealed off from the circulation by endothelialcells (16, 24, 26) and begin to form secondary tumors.

Recently, several investigators have studied the interactionsof invasive cells with endothelial cell layers (1, 13, 14, 27).These studies have been promising in that some of the morphological features of extravasation in vivo appear to be mimicked in vitro. The adherence of tumor cells to the subendothelial matrix produced by cultured cells (13) and hydrolysis ofthis matrix have also been investigated (10, 15) in order tounderstand more fully extravasation in the animal. This reportextends these studies to include the interaction of a humanfibrosarcoma cell line [HT1080 (17)] with an artificial blood

vessel wall.The artificial blood vessel wall was constructed by the growth

of bovine arterial endothelial cells on a preformed layer of ratsmooth muscle cells to form a structure with several of themorphological and biochemical properties of a vessel wall (8).These structures were then used as substrates for the humanfibrosarcoma cell line which was previously found to be capableof the complete degradation of both endothelial (10) andsmooth muscle cell (9) extracellular matrices. The use of thismodel system allowed us to integrate morphological and biochemical events and demonstrated that normal cells, such asendothelial cells, can modulate tumor cell invasive and degradative ability.

MATERIALS AND METHODS

Production of Substrates. Rat smooth muscle cells of the R22CIDstrain (12) were grown in Eagle's minimum essential medium (Grand

Island Biological Co., Grand Island, N. Y.) containing 10% fetal calfserum (Irvine Scientific, Irvine, Calif.), 2% tryptose phosphate broth(Difco Laboratories, Inc., Detroit, Mich.), penicillin (100 units/ml), andstreptomycin (100 ftg/ml). Stock cultures were maintained in 75-sq cmplastic flasks, and cells derived from them were seeded into 35-mm

dishes (2 x 10Vdish) for the production of tumor cell substrates. Thecultures received ascorbic acid daily (25 fig/ml) and were labeled withL-[3,4-3H]proline (1 jiCi/ml; New England Nuclear, Boston, Mass.).Culture medium was changed twice a week, and the addition of [3H]-

proline was discontinued 6 days after seeding. Medium was changed3 times over the next 6 days to remove as much soluble radioactivityas possible, and the labeled smooth muscle multilayers were used assubstrates for other cell types as described in the text. Artificial vesselwalls were constructed by the addition of 106 bovine arterial endothelial

cells of the A14CL-1 strain to these preestablished multilayers as

NOVEMBER 1981 4613

Research. on August 30, 2021. © 1981 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

P. A. Jones et al.

described previously (8). Ascorbic acid treatment (25 Â¿ig/ml/day) wascontinued throughout all experiments.

Destruction of Substrates. The kinetics of hydrolysis of thesesubstrates was followed by the appearance of radioactivity in thesupernatant medium. Samples (50 Â¿d)of medium were added to 5 ml ofBiofluor (New England Nuclear), and the total radioactivity releasedwas determined before each biweekly medium change. Values ofradioactivity released in control cultures incubated without tumor cellshave been subtracted from all data presented in this paper. In thoseexperiments where nontumorigenic cells such as endothelial cells ornormal fibroblasts were added to the radioactive smooth cell substratesin addition to HT1080 cells, cocultures of the normal cells with smoothmuscle cells provided control values. None of the normal cell typesused induced the release of radioactivity from the substrates in theabsence of HT1080 cells.

Results given in the charts represent the mean values of duplicateor triplicate dishes, and each experiment was repeated at least twice.

Analysis of Extracellular Matrix Proteins. Control cultures or cultures incubated with tumor cells were lysed by the addition of 0.25 MNH4OH for 30 min at room temperature. This treatment dissolved thecells but left the extracellular matrix material firmly anchored to thebottom of culture dishes (11). The composition of the extracellularmatrix was analyzed by sequential enzyme digestion with bovine pancreatic trypsin (Sigma type III) pretreated with elastin to adsorb contaminating elastase; this was followed by porcine pancreatic elastase(Sigma type III) and finally bacterial collagenase (Worthington typeCLSPA). Enzymes were stored at -80Â° at a concentration of 200 fig/

ml in phosphate-buffered saline [NaCI (8 g/liter), KCI (0.2 g/liter),

Na2HPO4, 7H2O (2.16 g/liter), and KH2PO4 (0.2 g/liter)], pH 7.6,containing 1 rriM CaCI2. The matrix was treated sequentially withtrypsin, elastase, and collagenase (1 0 jug/ml) in 1 ml of buffer. Samples(500 /il) of the supernatants obtained after 3 hr incubation at 37Â°with

each enzyme were assayed colorimetrically for protein as describedpreviously (2).

Tumor and Normal Cells. The human fibrosarcoma cell line HT1080(17) was used between passages 38 and 46 and was cultured asdescribed previously (9). Stock cultures of the C3H/10T'/2CL8 line(18) were grown in Eagle's basal medium as recommended (18), butthe medium was changed to Eagle's minimal essential medium after

the cells had been seeded onto smooth muscle multilayers. HFi andHF2 were 2 strains of human fibroblasts obtained from skin biopsies ofnormal individuals. V2CI2 was a cloned strain of bovine endothelial cellsderived from the vena cava of a calf fetus and was verified to be ofendothelial origin by the presence of immunologically reactive FactorVIII antigen.

Electron Microscopy. Cultures were washed with 0.15 M NaCI andthen fixed in situ with 2% glutaraldehyde in 0.1 M phosphate buffer, pH7.3. Postfixation was with 1% osmium tetroxide in 0.1 M phosphatebuffer. The cultures were dehydrated in a graded series of ethanolsand embedded in Epon:Araldite (50:50). Since the cultures wereembedded in situ in plastic dishes, 3 changes of 100% alcohol and nopropylene oxide were used for final dehydration. Polymerization wascarried out in a 37Â°oven for 48 hr followed by a 60" oven for 48 hr.

The plastic dish was removed from the culture after immersion in liquidnitrogen for a brief period. Thick sections (1 urn) were cut at rightangles to the surface, stained with toluidine blue, and examined bylight microscopy. Thin sections, 600 to 700 A, were mounted onFormvar-carboned one-hole grids, stained with uranyl acetate and lead

citrate, and examined in an electron microscope.

RESULTS

The release by HT1080 cells of radioactivity from cultures ofsmooth muscle cells prelabeled with [3H]proline is shown in

Chart 1. The rate of degradation was dependent on the numberof fibrosarcoma cells added, and cultures seeded with 2 x 10s

or 106 tumor cells were completely destroyed after 18 days of

incubation. The destruction of the smooth muscle cell multilayercould be observed directly with phase-contrast optics, and

holes in the multilayer appeared approximately 11 days afterthe addition of tumor cells. Perforations in multilayers seededwith 4X10" and 8 x 103 HT1080 cells, on the other hand,

did not appear until later in the experiment. The release ofradioactivity from the smooth muscle cell multilayers reflectedthe hydrolysis of the constituent connective tissue proteinssince enzymatic analysis of the residual matrix proteins obtained by NH4OH hydrolysis at the end of the experimentshowed that 95% of the glycoproteins, elastin, and collagenhad been solubilized by the fibrosarcoma cells (data not shown,but see also Fig. 1).

Smooth muscle multilayers containing preformed endotheliallayers (artificial vessel walls) were almost completely refractoryto hydrolysis by fibrosarcoma cells (Chart 2). The protectiveeffect of the endothelial layer was directly visualized when thecultures were lysed by NH4OH treatment at various times afteradding HT1080 cells in order that the residual-insoluble pro

teins could be stained with Coomassie blue (Fig. 1). Culturescontaining only smooth muscle cells showed increases in theamounts of insoluble matrix proteins during the 22 weeks ofthis experiment. The smooth muscle cells and extracellularmatrix were completely destroyed by the added fibrosarcomacells; therefore, virtually no insoluble proteins remained associated with dishes lysed 4 weeks after the beginning of theexperiment. Artificial vessel walls contained more insolublematrix proteins than did smooth muscle cells only, and diges-

30

ÃŒ? 20

f O IO 20 30

DAYS AFTER TUMORCELLS ADOED

Chart 1. Destruction of rat smooth muscle multilayers by HT1080 cells. Tissueculture dishes containing preformed layers of rat smooth muscle cells labeledwith [3H]proline in the presence of daily additions of ascorbic acid were used as

substrates for HT 1080 cells. Medium was changed every 3 to 4 days, and theamount of radioactivity solubilized from dishes seeded with 8x10' (A), 4 x 10'(O), 2 x 105 (A), or 106 (â€¢)HT1080 cells determined on the days indicated.

04 8 10 12 14 16

DAYS AFTER TUMOR CELLS ADDED

Chart 2. Inhibition of the degradative ability of HT1080 cells by a preformedendothelial cell layer. HT1080 cells (2 x 105/35-mm dish) were added to smoothmuscle multilayers prelabeled with [3H]proline in the absence (O) or presence(â€¢)of A1 ,CI-1 cells, 12 days after the addition of endothelial cells. The cumulativerelease of radioactivity into the supernatant was determined every 3 to 4 days atmedium change.

4614 CANCER RESEARCH VOL. 41



Invasion of Blood Vessel Wall by Fibrosarcoma Cells

tion of these structures by HT1080 cells was retarded markedly. Holes appeared only after 17 weeks of incubation (versus4 weeks for smooth muscle cells), but considerable destructionof the artificial vessel wall had occurred after 22 weeks. ViableHT1080 cells were isolated from such cultures on severaloccasions even after these prolonged incubation times.

The interactions of HT1080 cells with these 2 substrateswere also investigated by light and electron microscopy afterfixation of the cultures in situ at various times after the additionof tumor cells. HT1080 cells attached to and occasionallyspread out on the surface of the smooth muscle cells within 24hr but failed to invade the underlying tissue (Fig. 2a). Electronmicroscopy showed the tumor cells spread on top of thesmooth muscle cells and also their typical morphology including many plasma membrane projections, but no defined attachment sites between the 2 cell types were observed (Fig. 3a).

In contrast to this, HT1080 cells had penetrated the layer ofendothelial cells of an artificial blood vessel wall and invadedthe subendothelial space within 24 hr (Fig. 20), although theywere seen also on top of the structure. When viewed in theelectron microscope, the tumor cells had infiltrated betweenthe endothelial cells which then often resulted in a noncontin-

uous endothelium (Fig. 3b). The tumor cells adopted a moreflattened appearance with many membrane projections andwere coated by extracellular material within the structure (Fig.3o).

When such cultures were kept for another 36 days, fixed insitu, and analyzed microscopically, it was found that the artificial vessel structure remained essentially intact although onlyoccasionally a continuous layer of endothelial cells could beobserved. The invading tumor cells had proliferated within thevessel wall and appeared to be viable at the end of the experiment (Figs. 2d and 4b). The HT1080 cells had also infiltratedinto the smooth muscle tissue of the blood vessel leaving thesecells intact (Fig. 4b). In contrast to these observations, thetumor cells had completely destroyed the smooth muscle cultures after the same period of time (Fig. 2c). Very few smoothmuscle cells could be found with electron microscopy (Fig. 4a).The remaining extracellular matrix consisted mainly of collagenfibers and some glycoprotein material. The fibrosarcoma cells,however, showed a viable appearance and could be regrownfrom such cultures (data not shown).

Further experiments showed that endothelial cells weremarkedly inhibitory to tissue destruction, even if they wereadded simultaneously or after the tumor cells (Chart 3). Aninhibition of tumor cell-induced hydrolysis was detectable if the

endothelial cells were added 7 days after the HT1080 cells,but the extent of protection was greater if the time betweenadding the 2 cell types was less. No measurable digestion ofsmooth muscle cell matrix proteins occurred if the fibrosarcomacells were added 1 day after the endothelial cells. Thus, although the protective effect of the endothelial cells did notrequire a preexisting layer, the maximum resistance of thestructure to tumor cell-induced destruction was seen when

such a layer was present.Two strains of human skin fibroblasts and the mouse embryo

cell line C3H/10TV2CL8 were also able to retard degradationby HT1080 cells if they were plated on top of smooth musclemultilayers 1 week prior to fibrosarcoma cells (Chart 4). Thesenontumorigenic cells formed monolayers before the tumor cellswere added, however, the degree of protection afforded was

less than that seen with a strain of bovine venous endothelialcells (V2CL2), which behaved identically to the arterial endothelial cells in this experiment. A second addition of nonradioactiverat smooth muscle cells to the labeled smooth muscle layersonly weakly retarded the destructive activity of the tumor cells.None of the normal cell types used induced the release ofradioactivity from the smooth muscle substrates.

The inhibitory effect of the endothelial cells was proportionalto the number of cells seeded (Chart 5). In this experiment, 2x 10s HT1080 cells were added to smooth muscle multilayers

simultaneously with different numbers of endothelial cells. Thepresence of 8 x 103 endothelial cells caused a measurabledecrease in the rate of destruction, and 4x10* endothelial

cells (tumor celhendothelial cell ratio, 5:1) markedly inhibitedthe hydrolytic activity of the fibrosarcoma cells.

The effect of the fibrosarcoma cells on the total amounts ofextracellular matrix proteins present in smooth muscle or arti-

OAYSChart 3. Degradation of rat smooth muscle cultures by HT1080 cells in the

presence or absence of bovine endothelial cells. HT1080 cells (2 x 10s cells/35-mm dish) were added to smooth muscle layers prelabeled with [3H]proline 1

day after (A), simultaneously with (A), 1 day before (O), 7 days before (x), or inthe absence (â€¢)of 106 bovine endothelial cells. The total radioactivity present in

the supernatant medium was determined every 3 to 4 days at medium change,and values for control cultures incubated without tumor cells have been subtracted from the values given. Results are expressed as the cumulative releaseof radioactivity, and control cultures contained 4 x 105 cpm in matrix proteins at

the end of the experiment. Arrows indicate when endothelial cells were added.

30

E 20

ilIO

4 8 I2 I6 20 24 28 32 36

DAYS AFTER HTIO80 CELLS ADDED

Chart 4. Inhibition of the degradative ability of HT1080 cells by other celltypes. HT1080 cells (2 x 10V35-mm dish) were added to smooth musclemultilayers prelabeled with [3H]proline 7 days after the addition of other nontu

morigenic cells. Nontumorigenic cells used were: â€¢¿�,no additional cells added;O, unlabeled R22CID; A, HF,; A, C3H/10TVÃŒCL8; â€¢¿�HF2; x, V2CL2.

NOVEMBER 1981 4615



P. A. Jones et al.

ficial vessel cultures was determined directly by protein determination (Chart 6). The HT1080 cells caused a decrease in thequantities of all the extracellular matrix proteins present insmooth muscle cultures (Chart 6/4). The amount of glycoproteinpresent in artificial vessel walls was not affected by the presence of HT1080 cells, but the cultures contained less elastinand collagen than did control cultures (Chart 66). This resultimplied that the HT1080 cells interfered with the continueddeposition of these proteins, since the earlier experiments hadshown that there was no degradation of preexisting proteins insuch cultures (Charts 2 and 3).

Cultures of smooth muscle cells or artificial vessels incubatedwith HT1080 cells showed decreased levels of incorporationof [3H]proline into elastin and collagen, whereas incorporation

into glycoprotein(s) was relatively unaffected by the tumor cells(result not shown). Thus, although the tumor cells were ineffective in degrading preexisting proteins in the artificial vesselwalls, they did inhibit the increases in the levels of theseproteins which were occurring in the control cultures.

DISCUSSION

The complexities of whole-animal experiments necessitate

that the enzymology and cell biology of tumor cell extravasationbe investigated in isolated systems. The fidelity with whichculture systems reflect extravasation as it occurs in vivo can atpresent only be assessed by morphological comparisons of in

30

20

- IO

: O IO 20 30


Chart 5. Inhibition of degradatlve abilities of HT1080 cells by different numbers of endothelial cells. HT1080 cells (2 X 105/35-mm dish) were added tosmooth muscle multilayers prelabeled with [3H]proline simultaneously with 8 x103 (O), 4 X 10* (A). 2 x 105 (A). 106 (x), or no (â€¢)A1.CI-1 endothelial cells.

The cumulative release of radioactivity in the supernatant medium was determinedevery 3 to 4 days at medium change.

1.5

IO

A

D.

B.

0 34 0 34


Chart 6. Effect of HT1080 cells on the total amounts of protein present insmooth muscle or artificial vessel wall cultures. HT1080 cells (2 x 105/35-mm

dish) were added to rat smooth muscle cultures alone (A) or artificial vessel wallcultures (B) 4 days after the addition of 106 AUCI-1 cells. The total amounts of

glycoproteins (GP), elastin (E), and collagen (C) in cultures incubated without(a) or with (o) HT1080 cells were determined 0 or 34 days after the addition offibrosarcoma cells.

vitro and in vivo extravasation. In this sense, the morphologicaldata presented here provide an essential frame of referenceby which the significance of the biochemical data can bejudged.

The substrates had several of the morphological characteristics of vessel walls and contained an endothelial layer anchored firmly to the underlying smooth muscle multilayer (8).Electron micrographs also showed the presence of an irregular,dense basal lamina-like structure between the 2 cell types.

Although detailed studies on the intactness of the endotheliallayers have not yet been performed, the existence of somegaps in the endothelium would not invalidate our approach,since such gaps do occur in vivo (20) and invasive cellspenetrate "tight" endothelia both in vivo and in vitro (1, 22,

25). The structures also contained biochemically and morphologically identifiable glycoproteins, elastin, and collagen (8,12), which are 3 of the major protein constituents of vesselwalls. They therefore appeared to be reasonable substrates forthe tumor cells and resembled tissues from animals becausethe producer cells were embedded in the extracellular matrixproteins which they had synthesized. One disadvantage of thesystem was that the amounts of protein present in controlcultures increased throughout the experiments (Chart 6);hence, the substrates represented dynamic rather than staticsystems. However, our ability to measure these increasesmeant that quantitative data could still be obtained.

The fact that the fibrosarcoma cells rapidly established themselves in the subendothelial zone suggested that they mightfollow an adhesive gradient as suggested by Kramer andNicolson (14). It was also interesting that they assumed a muchflatter shape when growing in this area and often appeared tobecome coated by extracellular matrix material. The walling offof the tumor cells by endothelial cells following penetrationappeared to mimic what has been described for extravasationin vivo (16, 24, 26).

Although the number of HT1080 cells in the subendothelialspace appeared to be less than that seen in smooth musclecultures, it was unlikely that the inhibitory effects of the endothelial cells on vessel destruction were due primarily to growthinhibition. Examination of the electron micrographs showedthat considerable fibrosarcoma cell division had occurred after36 days, and no signs of cell atrophy due to anoxia or nutrientdeprivation were observed. Also, the facts that endothelial cellsinhibited HT1080-induced destruction at a celiiceli ratio of

1:25 or when added up to 1 week after tumor cells arguedagainst growth inhibition or toxicity as being completely responsible for the effects.

Nevertheless, the microscopic observations appeared todemonstrate a decreased proliferation of HT1080 cells in thesubendothelial area, and these observations may have somebearing on the phenomenon of tumor cell dormancy. In thisregard, it is interesting to note that Warren (24) has proposedthat the subendothelium may indeed be a site where tumorcells remain dormant in the intact animal. Indeed, the connective tissue substrates appeared to increase the viability of theHT1080 cells since viable tumor cells were regrown from thecocultures on several occasions. It should also be noted thattumor cell turnover appeared to occur during the course of theexperiments since floating cells, which were assumed to be ofHT1080 origin, were always present in the medium beforerefeeding.

4616 CANCER RESEARCH VOL. 41



Invasion of Blood Vessel Wall by Fibrosarcoma Cells

The nature of the inhibitory effect of the endothelial cells onthe HT1080 degradative activity is presently unknown. However, preliminary experiments have shown that medium conditioned by endothelial cells can inhibit the plasminogen activatoractivity in the HT1080 supernatant without the need for cell-

cell contact (6). Since plasmin production has been foundpreviously to play a significant role in the degradation ofextracellular matrix by HT1080 cells (9), the lowering of plasminogen activator activity may well influence the rate of tissuedestruction. The use of these artificial vessel walls may therefore increase our knowledge of tumor cell extravasation andprovide more information on interactions between normal andtumor cells in the animal.

REFERENCES

1. Bogenmann, E , and Sordat, 6. Interactions of tumor cells with endothelialcell monolayers grown on collagen gels. J. Cell Biol., 83 (Part 2). C93.1979.

2. DeClerck, Y. A., and Jones, P. A. The effect of ascorbic acid on the natureand production of collagen and elastin by rat smooth muscle cells. Biochem.J.. f86.21 7-225, 1980.

3. Dingemans, K. P., Roos. E.. van der Bergh Weerman, M. A., and van dePavert, I. V. Invasion of liver tissue by tumor cells and leukocytes: comparative ultrastrudture. J. Nati. Cancer Inst., 60.583-598, 1978.

4. Fidler, I. J., Gersten, D. M., and Hart, l. R. The biology of cancer invasionand metastasis. Adv. Cancer Res., 28. 149-250, 1978.

5. Fonck-Cussac, Y., Delage, J., and Petit, J. Observations ultrastructures surle mode d'implantation endovasculaire des mÃ©tastasesd'un carbronchique.

Poumon Colur, 25. 231-234, 1969.6. Heisel. M. A.. Jones, P. A., and Laug, W. E. Modulation of the degradative

properties of human fibrosarcoma cells by endothelial cells. Proc. Am.Assoc. Cancer Res.. 22. 248, 1981.

7. Jones, D. S.. Wallace, A. C., and Fraser, E. E. Sequence of events inexperimental mÃ©tastasesof Walker 256 tumor: light, Â¡mmunofluorescent andelectron microscopic observations. J. Nati. Cancer Inst., 46: 493-504.1971.

8. Jones, P. A. Construction of an artificial blood vessel wall from culturedendothelial and smooth muscle cells. Proc. Nati. Acad. Sei. U. S. A., 76.1882-1886, 1979.

9. Jones, P. A., and DeClerck, Y. Destruction of extracellular matrices containing glycoproteins, elastin, and collagen by metastatic human tumor cells.Cancer Res., 40: 3222-3227, 1980.

10. Jones, P. A., and Laug, W. E. Destruction by human tumor cells of basementmembranes produced by cultured endothelial cells. Proc. Am. Assoc. CancerRes., 2). 210, 1980.

11. Jones, P. A., and Scott-Burden, T. Activated macrophages digest the extra

cellular matrix proteins produced by cultured cells. Biochem. Res. Commun.,86. 71-77, 1979.

12. Jones, P. A., Scott-Burden, T., and Gevers, W. Glycoprotein, elastin andcollagen secretion by rat smooth muscle cells. Proc. Nati. Acad. Sei. U. S.A., 76. 353-357, 1979.

13. Kramer, R. H.. Gonzalez. R., and Nicolson, G. L. Metastatic tumor cellsadhere preferentially to the extracellular matrix underlying vascular endothelial cells. Int. J. Cancer, 26. 639-645, 1980.

14. Kramer, R. H . and Nicolson, G. L. Interactions of tumor cells with vascularendothelial cell monolayers: a model for metastic invasion. Proc. Nati. Acad.Sei. U. S. A., 76. 5704-5708. 1979.

15. Kramer, R. H., and Nicolson, G. L. Invasion of vascular endothelial cellmonolayers and underlying matrix by metastatic human cancer cells. In: H.G. Schweiger (ed.). International Cell Biology, pp. 794-799. Heidelberg:Springer-Verlag, 1981.

16. Ludatscher, R. M., Luse, S. A., and Suntzeff. V. An electron microscopicstudy of pulmonary tumor emboli from transplantable Morris hepatoma5123. Cancer Res., 27. 1939-1952, 1967.

17. Rasheed, S., Nelson-Rees, W A., Toth, E. M.. Arnstein, P., and Gardner, M.B. Characterization of a newly derived human sarcoma cell line (HT-1080).Cancer (Phila.), 33. 1027-1033, 1974.

18. Reznikoff, C. A., Brankow, D. W., and Heidelberger, C. Establishment andcharacterization of a cloned line of C3H mouse embryo cells sensitive topostconfluence inhibition of cell division. Cancer Res., 33. 3231-3238,

1973.19. Sherwin, R. P., and Richters, A. Pathobiologic nature of lymphocyte inter

actions with human breast cancer. J. Nati. Cancer Inst., 48: 1111-1115,1972.

20. Simionescu, N., Simionescu, M., and Palade G. E. Structural basis ofpermeability in sequential segments of the microvasculature of the diaphragm. II Pathways followed by microperoxidase across the endothelium.Microvasc. Res., 15: 17-36, 1978.

21. Sindelar, W. F., Tralka, T. S., and Ketcham, A. S. Electron microscopicobservations on formation of pulmonary mÃ©tastases.J. Surg. Res.. Ã•8.137-161, 1975.

22. Sugarbaker, E. V.. and Ketcham, A. S. Mechanisms and prevention ofcancer dissemination: an overview. Semin. Oncol., 4: 19-32, 1977.

23. Vlaeminck, M. N., Adenis, L., Mouton. Y., and DÃ©maille,A. Etude expÃ©rimentale de la diffusion metastatique chez l'oeuf de poule embryonne repar

tition, microscopie et ultrastructure des foyers tumoraux. Int. J. Cancer, 10:619-631, 1972.

24. Warren, B. A. Some aspects of blood borne tumour emboli associated withthrombosis. Z. Krebsforsch, 87. 1-15, 1976.

25. Warren, B. A., Chauvin, W. J., and Phillips. J. Blood-borne tumor emboliand their adherence to vessel walls. In Cancer Invasion and Metastasis:BiologicS. B. Day et al. (eds), Mechanismsand Therapy, pp. 185-197. NewYork: Raven Press, 1977.

26. Wood, S. Pathogenesis of metastasis formation observed in vivo in the rabbitear chamber. Arch. Pathol., 66. 550-568, 1958.

27. Zamora, P. O., Danielson, K. G., and Hosick, H. L. Invasion of endothelialcell monolayers on collagen gels by cells from mammary tumor spheroids.Cancer Res., 40: 4631-4639, 1980.

NOVEMBER 1981 4617



146 17 22WEEKS AFTER FIBROSARCOMA CELLS ADDED

2a 2b

2c 2d

Fig. 1. Effect of human fibrosarcoma cells on the total amounts of extracellular matrix material present in smooth muscle or artificial vessel walls. HT1080 cells(2 x 10!/35-mm dish) were seeded onto rat smooth muscle layers or artificial vessel walls, and the cells were lysed by the addition of 0.25 M NH<OH at the indicated

times after seeding. The residual matrix was then stained with Coomassie blue before photography. Top row, cultures containing rat smooth muscle cells only;second row, as above but seeded with HT1080 cells; third row, cultures containing artificial blood vessel walls; bottom row, as above but incubated with HT1080cells.

Fig. 2. Human fibrosarcoma cells were seeded on cultures of smooth muscle cells (a. c) or artificial blood vessel walls (6, tfl, fixed in sifu, and processed forelectron microscopy. Within 24 hr after addition, tumor cells (closed arrows) attached to the surface of smooth muscle cells and some were spread out (a), whereason the artificial blood vessel some of the HT1080 cells (closed arrows) penetrated the endothelium and were seen to invade the subendothelial space (b) When suchcultures were fixed 36 days later, the smooth muscle cultures had been destroyed by overgrowing tumor cells (c). In contrast, the structure of the artificial bloodvessel was more intact with the endothelial cells on top (open arrows), while the tumor cells had invaded and proliferated within the vessel wall, x 1.250.

Fig. 3. a, HT1080 cells were seeded on smooth muscle cell cultures (SMC) and fixed in situ 24 hr later. This micrograph shows 2 tumor cells, which developeda primitive intercellular junction (arrow), attached to the upper surface of smooth muscle cells. The fibrosarcoma cells showed their characteristic morphologyincluding numerous plasma membrane projections. No invasion of the smooth muscle cell tissue by the tumor cells occurred within this period of time. Note theextracellular material between the smooth muscle cells, x 10,400. b, tumor cells (TC) were seeded on the artificial blood vessel and fixed in sifu as indicated in a.The fibrosarcoma cells had penetrated the endothelial cell layer which shows an area of noncontinuity. Beneath the endothelial cells (EC) is a basement membrane-like structure of extracellular matrix material (open arrows). The tumor cells, which contain many structurally normal mitochondria, are surrounded by extracellularmaterial. Smooth muscle cells show their characteristic thin and flat morphology, x 10.600.

Fig. 4. a. smooth muscle cell cultures (SMC) were fixed 36 days after tumor cell addition. These cultures were overgrown and completely destroyed by thefibrosarcoma cells. The remaining smooth muscle cells showed many lysosomal bodies and the majority of the extracellular material had been degraded by theHT1080 cells leaving a few collagen fibers and some glycoprotein material, x 10,480. b, this micrograph shows an artificial blood vessel fixed 36 days aftercocultivation with HT1080 cells. These cultures occasionally showed endothelial cells (EC) on top of the structure, but most of the surface was covered by tumor cells(TC). HT1080 cells appeared viable with well-organized cytoplasm and nuclei; however, some of them showed a few lysosomal bodies. The fibrosarcoma cells wereseen to have invaded into the smooth muscle tissue of the artificial blood vessel; the surrounding smooth muscle cells, however, showed their typical flat appearanceand were coated by extracellular material, x 6.500.

4618



3a

3bEC

' â€¢¿�â€¢â€¢¿�^v>.-

t

..r-

>V" ^

-Â»*%*-xC-M&fcfeÃ

-... **â€¢**/; â€¢¿�-â€¢¿�-.. ' > .: - ;... â€¢¿�".

--o v, 'â€¢' , â€¢¿�fe , v. â€¢¿�â€¢;;- -â€¢' .'- , ; *â€¢ ' Â»â€¢¿�- C3

- ..9- ;- - . SÂ£ .,;>i ." ' " - Â¡ -. ,-.-..:

M

EG

-riâ€¢¿�'

^->/rf..._â€” L- --'-^ ^â€”-

-*, i .,--.- ''^-T*^^^*aÂ«-. â€¢¿� tfÂ»/Tr"-'r ^ â€¢¿�"':Â¿aiâ€”^-^- -j- ;1!;^ ^ft -â€¢^â€¢*â€¢^ â€¢¿�C '-jo* gC^

Â¿^* '^. - \f ; ^ *ÃÂ«ic,j#ÃŸ *Â¿?'L^.-

- *f *A ^T^ â€¢¿�*-^iV..Ã¶fTAâ€¢¿�'j.''*v 53^- â€¢¿� Â£*&.*>'â€¢¿�Â¿.TC;â€”yâ€”>-r>--â€¢¿�*â€¢Â£*Â£

SMC

4619



* I' -tf*.dfc

-:'v^Â«kâ€¢¿�;*v"<â€¢.-*.**'â€¢::. ^ -,. â€¢¿�â€¢â€¢â€¢

â€¢¿�-' -.-*â€¢*'â€¢'- â€¢¿�

^^v^^*^^Â«^^^^^/. Â«^^Ã•Ã^^V^A''Â¿'Â¿*'' %^ '- 4'* '--i!*' '-ÃÃÃ-j '^-^

- i, i. "â€¢'"- .;* 'Ã¬tMÃŒfVV?* I*-'' JtÃ³:.Â¿Â¿#?'-.i Â¿ i' ? â€¢¿�""-*â€¢":>-â€¢¿�â€¢¿�'v './-i *x :.

**'" '/'":F- - ;

"~~*Ã;^^t: :':".--;i^^Ã„^f^SSS^'â€¢^r.-/.*<â€¢ .â€¢:'"v-'<Ã„^*r*"';."

TtÃ¼

4b Ã>ÃlC n->

\---'^*Ã•^-*â€¢¿�â€¢aK^1^--v-:-.:;--.-*1'"''

eT^

^-i^X**. ^wsgÂ»*^*â€¢ Â«<* .

TC

â€”¿� (^ , Â«Jâ€”¿�.â€”, <* >* â€¢¿�*.-" ^^-x.* -'^^ 'â€¢â€¢* Â£^Nâ€¢¿�.*'if' "Â» c t:**"1-/~::s: .i|r. -.'V--., â€¢¿�Â«fe ÃŸfÃ„Â® -â€¢ â€¢¿�

; \'â€¢ T*â€¢\ i>*- -e^--^- -*--?--.-.-;^.;,; ^^-^-^^.i-J . . c, â€¢¿�â€¢;â€¢â€¢â€¢;'â€¢¿�Â»- â€¢¿�'-Â«J.-

.Â». '*-â€¢ . r* * - - SMC,._. ^* â€¢¿�-* -r*' *?u

â€¢¿�

r.-Ã®;:

4620



1981;41:4613-4620. Cancer Res Peter A. Jones, Harry B. Neustein, Felicidad Gonzales, et al. Fibrosarcoma CellsInvasion of an Artificial Blood Vessel Wall by Human

Updated version

http://cancerres.aacrjournals.org/content/41/11_Part_1/4613

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/41/11_Part_1/4613To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



Invasion of an Artificial Blood Vessel Wall by Human Fibrosarcoma Cells1 · Artificial blood vessel...

Documents

Transcript of Invasion of an Artificial Blood Vessel Wall by Human Fibrosarcoma Cells1 · Artificial blood vessel...