Specificity of Arrest, Survival, and Growth of Selected Metastatic Variant Cell Lines1

[CANCER RESEARCH 38, 4105-4111. November 1978] 0008-5472/78/0038-OOOOS02.00 Specificity of Arrest, Survival, and Growth of Selected Metastatic Variant Cell Lines1

Garth L. Nicolson, ' Kenneth W. Brunson, and Isaiah J. Fidler

Departmental Developmental and Cell Biology. University ol California, Irvine, Calitornia 92717 [G. L.N..K.W. B.Â¡,and Cancer Biology Program, NÃ¬tÃ¬onal Cancer Institute, Frederick Cancer Research Center, Frederick, Maryland 21701 Â¡I.J. F.]

Abstract capillary lodgment (5, 9, 20, 21, 36, 37), suggesting that other factors are involved in specific malignant cell arrest. Animal tumor models for blood-borne metastasis have During transport in the blood, malignant cells interact been developed by in vitro cloning or in vivo selection of with each other as well as with host cells. Common inter malignant tumor cell populations to obtain organ-prefer cellular interactions that affect the subsequent arrest and ring variant tumor cell lines with altered arrest, survival, survival of malignant cells include homotypic adhesion of invasion, and growth properties. Selection and some tumor cells to form multiceli emboli (10, 27, 34), heterotypic tumor cell characteristics of lung-, brain-, and ovary-colo adhesion of tumor cells to platelets (24), lymphocytes (14, nizing metastatic B16 melanoma, liver-colonizing RAW117 17), and noncirculating host cells (33, 34). In addition, lymphosarcoma, and lung-colonizing MSV3T3 vasoforma- malignant cell deformability (40, 48), release of hydrolytic tive sarcoma variant lines will be discussed along with enzymes (4, 41), and interactions with clotting components additional data, suggesting that tumor cells of varying (8, 44, 46) and host immunological defenses (17, 23, 43) malignant potential preexist in the unselected tumor pop can determine whether tumor cells survive circulatory ulation. transport.

Introduction Experimental Models of Blood-borne Metastasis The most life-threatening characteristics of malignant Metastasis via the blood does not necessarily result in tumor cells are their abilities to detach from the primary tumor colonization based on initial malignant cell distribu tumor, invade into surrounding normal tissues, and form tion and circulatory pathways involved. Organ specificity of metastatic growths at distant sites (13, 22, 47). Metastatic implantation, survival and growth of plasmacytoma (37), colonies can be close to or distant from the primary site, histiocytoma (9), melanoma (5, 20, 26), and reticulum cell and the latter may occur when tumor cells penetrate the sarcoma (36) suggests that patterns of malignant tumor lymphatics or circulatory system. Of particular interest are colonization are nonrandom and uniquely dependent on malignant cells that enter the bloodstream, because these specific tumor cell and host properties (22). can be transported to the major organs. However, tumor For studying the various steps in the metastatic process, cell survival during blood-borne transit is not a common it is desirable to have sets of metastatic and equivalent low event, since the overwhelming majority of malignant cells or nonmetastatic tumor cells in order to directly compare expire in the circulation (15, 39). properties important in metastasis. When variants of low The survival of malignant cells after they enter the circu and high metastatic potential are available, comparisons lation depends upon their abilities to successfully implant with dubious 'normal" counterpart cells are unnecessary. in the microcirculation, invade or extravasate the capillary Several of these tumor systems now exist (see below) where endothelium, and establish a microenvironment for subse spontaneously, virally, and chemically transformed cell quent vascularization and growth (13, 22, 47). While in the lines of low metastatic potential have been used to select blood, the distributions of viable malignant cells are not variant lines of high metastatic potential and organ prefer necessarily based on circulatory pathways or patterns of ence of colonization. initial microcirculatory arrest (20, 42). Blood-borne tumor Lung-colonizing Melanoma Variants. In order to obtain cells can be temporarily arrested in the first capillary bed malignant melanoma cell lines of differing metastatic po encountered, but at this point the cells may: (a) die; (b) tential. Fidler (11) used in vivo selection to enrich tumor cell survive and grow at the site of implantation (c) deform and populations with variants of high lung colonization abilities. recirculate to be arrested at other sites; or (d) invade and Starting with an unselected murine B16 melanoma cell line survive at extravascular sites. In a variety of experimental of low in vivo metastatic potential, metastatic variants were blood-borne tumor systems, the distributions of gross mÃ© selected for their abilities to implant, invade, survive, and tastases are nonrandom and do not correlate with initial grow to form gross lung tumor colonies after i.v. injection into syngeneic C57BL/6 mice. When pulmonary tumor nodules formed, these were identified by their melanin 1 Presented at the John E. Fogarty International Center Conference on pigmentation and removed for adaptation to growth in Hormones and Cancer, March 29 to 31, 1978, Bethesda. Md. Supported by National Cancer Institute Contract N01-CO-25423 to Litton Bionetics, Inc. culture. The B16 cells that grew in culture after the first in 2 Supported by Grants BC-211A from the American Cancer Society, vivo selection were designated as line B16-F1, and these Grants RO1-CA-22950 and RO1-CA-15122 from the USPHS. and National Cancer Institute Contract NO1-CB-74153. cells were harvested and injected i.v. into new groups of 3 To whom requests for reprints should be addressed. syngeneic animals to obtain the twice-selected line B16-F2.

NOVEMBER 1978 4105

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1978 American Association for Cancer Research. G. L. Nicolson et al. After 10 selections, a line was obtained (B16-F10) which animals. These results indicated that viable blood-borne was more metastatic when compared to line B16-F1 in lung B16-F1 tumor cells temporarily arrested, recirculated, and colonization (11) and spontaneous pulmonary metastasis traversed the site of circulatory anastomosis within 1 to 3 hr assays (14). postinjection eventually to form pulmonary and extrapul The B16-F melanoma system has proven particularly monary tumors in either parabiotic pair of animals. The useful for studying tumor cell and host properties associ kinetic distributions of viable B16-F cells and the number of ated with metastasis (22). Some advantages of the system experimental mÃ©tastasesthatformed 2 weeks after injection are: (a) the tumor arose spontaneously in the skin and is in the injected parabiont were the same as in the nonpara- maintained in syngeneic hosts; (b) tumors can be seen biotic animals; in contrast, the kinetic distributions in the easily in the skin, lymph nodes, visceral organs, and at uninjected parabionts indicated that circulating tumor cells other sites because of their high melanin pigment content; arrested and survived at lower efficiency in the lungs (B16- and (c) tumor cells can be grown easily in tissue culture or F1) or failed to be arrested and survive in the lungs (BIS- at a variety of sites in vivo. FIO). This indicates that selection for lung colonization Lung specificity of line B16-F10 was assessed by follow leads to a decrease in detachment and recirculation of ing the kinetic distributions of ['25l]-5-iodo-2'-deoxyuridine- tumor cells destined to form mÃ©tastases.Although B16-F10 labeled melanoma cells after i.v. (tail vein) or left ventricle cells were able to arrest, detach, and recirculate from i.e.4 injection (20). By the latter route, extrapulmonary injected to uninjected parabionts, these cells fail to form arrest, detachment, and recirculation are necessary in order lung tumors in the uninjected parabionts (21). to reach the lung microcirculation. Organs and blood from The stability of B16-F lines after prolonged cell growth in groups of animals were analyzed for viable 125l-labeledB16 vitro under nonselecting conditions has recently been ques cells. Although the kinetic distributions of blood-borne tioned (31). Our results suggest that the differences be melanoma cells were different in blood and all organs tween B16-F10 and B16-F1 in ability to form gross lung within 2 min after injection via these 2 routes, after 1 day tumors, while still significant, decrease with long-term the same numbers of viable B16-F10 cells were found in the culture (G. L. Nicolson and I. J. Fidler, unpublished data). It lungs, independent of the route of melanoma cell entry into may prove necessary to start fresh cultures from early the circulation. When animals were sacrificed after 2 weeks passaged cells or to repeat in vivo selection procedures to and the numbers of gross melanoma nodules had been prevent loss of phenotypic properties in tissue culture. counted in the lungs and at other sites, the same numbers However, some metastatic variant systems seem to be quite of pulmonary mÃ©tastaseswere found in mice receiving stable in tissue culture (see description of lymphosarcoma melanoma cells via the tail vein (9 Â±3 and 79 Â±16 nodules system). for B16-F1 and B16-F10, respectively, per 25,000 input Brain-colonizing Melanoma Variants. Brain-colonizing cells) or left ventricle (10 Â±2 and 83 Â±10 for B16-F1 and B16-B melanoma variants have been obtained from line B16-F10, respectively, per 25,000 input cells) (20). B16-F10 B16-F1 by sequential in vivo selection for brain implanta extrapulmonary mÃ©tastaseswere not found, indicating that tion, survival, and growth (Chart 1; Refs. 5, 32). B16-F1 initial blood-borne tumor cell distribution and arrest may cells were injected into the left ventricle of syngeneic mice have little bearing on subsequent metastatic colonization. by i.e. injection, and the resulting rare brain tumors were Significantly more cells from the highly metastatic line B16- recovered and cultured to obtain line B16-B1. The selec F10 implanted, survived, and grew to detectable lung mÃ© tions were repeated 3 more times i.e., and further selections tastases compared to line B16-F1 (11, 20), and B16-F1, but were carried out following injection into the tail vein. After not B16-F10, cells formed a variety of extrapulmonary 7 selections for brain colonization, 2 types of brain tumors mÃ©tastases(Table 1). Although the initial tumor cell distri were found (Chart 1). One type tended to implant and grow butions, rates of vascular entry, and first organ capillary in the meninges of the dorsal cerebrum, and the other was beds encountered were different, the ultimate outcome of located in or near the rhinal fissure between the olfactory metastatic colonization was unchanged. Tumor cells enter bulb and the cerebral cortex. The cell line from the me ing into the circulation via the left ventricle and temporarily ninges was designated B16-B7b, and the rhinal fissure line arresting in the extrapulmonary microcirculation were able was designated B16-B7n. Animals given injections of line B7n, but not B7b, exhibited "neurological symptoms" such to detach, recirculate, and ultimately reach the lungs, implant, and survive to form the same number of tumor as staggering gait, head tilt, and poor balance and orienta colonies as melanoma cells entering the lung microcircula tion. After 3 additional i.v. selections for brain colonization, tion first after tail vein injection. These data indicate that lines B16-B10b and B16-B10n were obtained, and these implantation and survival of blood-borne malignant cells are melanoma lines maintained their brain regionalization not random, and recirculation of once-arrested tumor cells specificities. When line B16-B10n was analyzed for brain can lead to efficient metastatic colonization (20. 32). specificity and neurological disorders, almost every animal Arrest, detachment, recirculation, and rearrest of B16-F given injections had brain mÃ©tastasesexclusively (Table 1), melanoma cells have been studied with the use of para- and animals with B10n tumors exhibited a high degree of biosed pairs of syngeneic mice (21). After injection of 125l-5- neurological symptoms (5). In these latter experiments B16 iodo-2'-deoxyuridine-labeled B16-F cells into one animal of melanoma cells were injected i.v. via the tail vein, requiring each pair, the kinetic distributions of viable melanoma cells the B16 melanoma cells to initially pass through the pul were assessed in injected as well as uninjected parabiotic monary capillary bed and then recirculate in order to reach the brain vascular bed. All brains collected from mice given * The abbreviations used are: i.e., intracardiac; i.t., intratracheal. injections of B16-B10n had melanoma nodules in or near

4106 CANCER RESEARCH VOL. 38

Table 1 Locations of experimental mÃ©tastasesfoundafter i.v. injection of lung- or brain-selected melanoma variant lines No. of animals with tumors/total no. of animals

lineB16-F1Melanoma cavity6/9, (lung) 10/10 3/10 6/10 2/10 1/10Lumbar B16-B1 (brain) 9/9, 9/9 4/9, 4/9 4/9, 6/9 2/9, 3/9 spine, B16-F10(lung) 10/10, 10/10 0/10, 0/10 0/10,0/10 0/10,0/10 none B16-B10n (brain)Lung9/9,0/10, 0/10Brain4/9,10/10, 10/10Thoracic0/10, 1/10Ovary2/9,0/10, 1/10OtherAdrenal,none1/10

subcutaneous tumor in C57BL/6 mouse

Chart 1. Experimental design for selec tion in vivo for B16 variant lines with en hanced brain colonization. ^â€”^_^ -O Rhlnal fissure tumors

the rhinal fissure (5); often these tumors extended along ical," and these same 2 clones were also dramatically vascular pathways into the cerebral cortex. Extension into invasive in the brain. Although other clones could be the cortex was not seen with B16-B1 or B16-B5 tumors. classified as grossly invasive due to their extension into the B16-B10n cells also metastasized to brain after s.c. implan brain cortex, not all of the grossly invasive clones caused tation. This has been shown following the s.c. implantation observable neurological symptoms. These data suggest of 100,000 B16-B10n cells into syngeneic mice. After 2 that tumor cell heterogeneity exists even in the highly weeks the resulting tumor was resected, and at this time selected B10n line. regional lymph nodes were palpable. Two additional weeks Ovary-colonizing Melanoma Variants. Ovary-colonizing later mÃ©tastaseswere found in the brain (K. W. Brunson B16 melanoma variants have been selected from B16-F1 for and G. L. Nicolson, unpublished data). enhanced blood-borne implantation, survival, and growth Line B16-B10n has recently been cloned in vitro to deter in ovaries. After 10 selections for ovarian colonization, line mine whether it is composed of a uniform or heterogeneous B16-010 was obtained which forms more gross ovarian tumor cell population. Ten such clones were examined for tumors compared to B16-F1.5 Line B16-010 also forms invasive properties, neurological symptoms, and brain some experimental mÃ©tastasesatsites other than ovary and specificities. Preliminary experiments indicated that all does not appear to be as organ-specific as lines B16-F10 or clones were metastatic to brain, but not all clones were B16-B10n. Cloning B16-010 may yield a highly specific line grossly invasive, or caused observable neurological disor ders. Two of 10 clones were classified as grossly "neurolog 5 K. W. Brunson and G. L. Nicolson. manuscript in preparation.

NOVEMBER 1978 4107

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1978 American Association for Cancer Research. G. L. Nicolson et al. that will exclusively colonize ovary. lines; however, the incidences of experimental pulmonary Liver-colonizing Lymphosarcoma Variants. Brunson and mÃ©tastases after i.v. injection were quite different. The Nicolson (7) have used the RAW117 lymphosarcoma line, lymphocyte-resistant lines B16-F1" 6 and B6-F10'-r 6 formed which forms solid hepatic tumors in BALB/c mice to select significantly fewer lung mÃ©tastaseswhen compared to their for enhanced liver colonization. After 10 sequential selec parental lines B16-F1 and B16-F10, respectively (17). In this tions for liver colonization, a lymphosarcoma line (RAW117- system, host lymphocyte interaction with the B16-F lines H10) was obtained that formed approximately 200 to 250 appears to enhance rather than inhibit the formation of times more gross liver tumor nodules than the parental line mÃ©tastases. Although the mechanism of lymphocyte resist in comparable biological assays (Table 2) and displayed en ance is not known, the lymphocyte-resistant cells show hanced malignancy when assessed by time of host death. In decreased binding of immune lymphocytes in heterotypic the latter assay the approximate times for 50% host death adhesion assays (16). after i.v. injection of 5,000 viable lymphosarcoma cells were 40, 16, and 8 days for lines RAW117 parental, RAW117-H5, Metastatic Cells and the Primary Tumor and RAW117-H10, respectively (7). Although liver coloniza The existence of clonal variability in the properties of in tion potential increased with succeeding in vivo selections, vivo selected tumor cell lines such as B16-B10n and the the numbers of lymphosarcoma nodules in the lungs were ability to select this line from a melanoma line that was low in parental and selected RAW117 lines (Table 2). The selected once for lung colonization (Chart 1) suggest that stability of the unselected and selected lymphosarcoma the original tumor may have been heterogeneous with lines was examined by comparing cell lines passaged for respect to its phenotypic properties. Either these highly meta only a short time in vitro (2 weeks after thawing from liquid static variants preexist in the primary tumor population or, nitrogen) with the same cell lines that had been grown in alternatively, the selected lines could have arisen by the tissue culture for 3 months. Differences in liver colonization process of organ adaptation during the selection process. between RAW lines grown for brief or long periods in tissue Cloning and fluctuation tests have been performed with the culture were not found, indicating that these cell lines are use of B16-F1 to determine whether highly metastatic var stable during long-term growth in vitro. iants preexist in the parent population (19). The cloned B16 Lung-colonizing Sarcoma Variants. Untransformed, den cell lines differed significantly in their metastatic potential. sity-inhibited BALB/c embryo cell lines such as 3T3 have Subcloning experiments ruled out the possibility that the in been virally transformed with oncogenic viruses such as vitro cloning procedures yielded the variations among the SV40 and MSV to SV3T3 and MSV3T3 lines, which are not clones. The existence of phenotypic heterogeneity in the density-inhibited for growth and after in vivo s.c. implanta primary tumor cell population has been hypothesized by tion or i.v. injection form vasoformative sarcomas (1, 3). An Nowell (35). Host selection pressures could allow the emer MSV3T3 line has been used to select lung-colonizing var gence of sublines with enhanced malignant potential, com iants. After 18 in vivo selections, line MSV3T3-LN18 was pared to that of the original tumor. obtained that yields more lung nodules compared to unse Phenotypic heterogeneity of metastatic properties within lected lines. Thus, sarcoma virus transformed cell lines can an unselected tumor cell population has also been found also yield metastatic variants that colonize lung. with vasoformative sarcomas such as MSV3T3-LN (Table 3). Lymphocyte-resistant Melanoma Variants. B16-F mela In this case, groups of BALB/c mice were given injections noma variant lines are susceptible to killing by host immu- i.v. with in vitro cloned tumor cell lines, and the number of nocytes (12). The in vitro killing of B16-F cells by cytotoxic lymphocytes has been used to select B16-F variants resist ant to cell-mediated killing. By sequentially exposing B16-F Table 3 Experimental mÃ©tastases found after i.v. injection of MSV3T3-LN1 lines to syngeneic immune lymphocytes and recovering the clones survivors and repeating the process 6 times (17), lympho Eight-week-old BALB/c mice were given injections i.v. with 2 x cyte-resistant lines (designated B16-F1" 6and B16-F10" 6) 10" single, viable in vitro cloned cells. Mice were sacrificed 18 days were obtained from B16-F1 and B16-F10, respectively. later, and gross pulmonary tumor colonies were visualized by These 4 lines have been injected s.c. or i.v. into nonimmune injections of an India ink-formalin solution, given to the animals i.t. syngeneic hosts. Growth at s.c. sites was similar for all 4 Clone No. of lung tumor colonies Median 39 0,14, 2, 4 Table 2 19 3,3,;20,5,5,6,10, 5 (2-17) Liver colonization of variant lymphosarcoma tumor cell lines 9 15,16, 21,24,27,28 24 (15-29) 37 24,30,34,3325,36,42,46,49,59,6679, 42 (24-66) Five thousand viable lymphosarcoma cells were suspended in (18-70)(12-69) 0.2 ml medium and injected i.v. into groups of 10 mice; after 10 4523 18,12,?015,3428,4952,fin56,5554,6361745970,69 49525556 days the mice were sacrificed, and organs were removed for tumor (16-87)(31-68) counts or weighing. 5566 16,31,?fi42,3650,4?55,7?57,75,64,8768 no. of 14 31,50,,53,57,59,130'.62,689,,28,,73,86145,1 59139(0-4)"(31-86) Lymphosar no. of liver spleenweight(g)0.08(0.06-0.12)lung nod 560,02,106 111,121,2,3,1317141,29 (106-162) comalineRAW117-P nodules0.6 ules0.4(0-1) 1!,55,1,162, (0-4)" RAW117-H1 2.7 (1-7) 0.11 (0.10-0.15) 0.4(0-1) Parent 104,107,110,111,113,125,125, 127(104-198) RAW117-H5 25.1 (9-39) 0.16 (0.09-0.30) 0.6 (0-2) 127,195,198 138, 143, 155, 162, 168, 194, RAW117-H10Av. 237 (200-250)Av. 0.24 (0.17-0.29)Av. 1.0(1-4) " Numbers in parentheses, range. " Numbers in parentheses, range.

4108 CANCER RESEARCH VOL. 38

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 1978 American Association for Cancer Research. Characteristics of Metastatic Variant Cell Lines experimental pulmonary mÃ©tastaseswas determined 18 could lead to arrest in the microcirculation at specific days later. The metastatic potential of the clones differed locations. This might have been due, in part, to homotypic widely (from a median number of 1) of tumors found per (self) adhesion between the circulating blood-borne tumor animal in clone 39 to a median of 139 in clone 56 after i.v. cells, because homotypic adhesion assays for lung-selected administration of 2 x 10" viable cells) compared to that of B16 melanoma (34, 35) and MSV3T3 sarcoma (30) indicated the parental cell line (Table 3), suggesting that mixed cell that the highly metastatic lung-colonizing lines undergo in populations existed in the MSV3T3-LN line. The presence vitro homotypic adhesion at faster rates. The formation of of populations of high and low metastatic potential before multicellular tumor emboli is known to enhance experimen in vivo selection suggests that each in vivo selection results tal pulmonary colonization. This was shown by Fidler (10), in enrichment of more metastatic cells in each subsequent who purposely aggregated B16 melanoma cells and then cell line established in culture. injected the aggregates i.v. and scored for lung tumor formation. In these experiments the cells within the larger Tumor Cell Surface Properties and Metastasis tumor cell emboli implanted and survived at significantly greater rates compared to small emboli or single cells. In A variety of cell surface changes have been correlated addition, heterotypic adhesions of B16-F melanoma cells with transformation or tumor formation (25, 29, 38). Al with platelets (24), lymphocytes (14), and endothelial cell though some of these altered properties may be important lines (34, 45) probably affected rates of tumor cell arrest in in metastasis, little is known about tumor cell surface char the microcirculation. acteristics that determine malignancy and metastatic spread Organ preference in arrest and colonization could be (22, 29). With the use of the metastatic variant lines de explained by adhesive interactions of circulating tumor scribed in preceding sections, it should be possible to cells with noncirculating host cells in the target organ (33). determine tumor cell properties associated with successful In an experimental approach to this problem, B16-F cells blood-borne metastasis (18, 22). have been mixed with partly purified single target or nontar The lactoperoxidase-catalyzed iodination procedure (28) get organ cells in suspension, and heterotypic aggregation has proven to be valuable in detecting cell surface modifi was determined within 5 min. In vivo selected B16-F lines cations on metastatic cells. By this procedure, exposed cell such as B16-F10 and B16-F13 aggregated rapidly with surface proteins are 125l-iodinated, and the labeled proteins suspended lung cells in vitro, but only slowly with non- are separated by sodium dodecyl sulfate-polyacrylamide gel target organ cells (kidney, spleen, liver, etc.) in similar electrophoresis. Autoradiograms made from slab gels of experiments (22, 33). Thus, cell-to-cell recognition and separated B16-F1 and B16-F10 components revealed few adhesion may play an important role in directing tumor cell differences, except in some minor protein bands at low implantation at specific organ sites. molecular weight (31). However, cell surface proteins on Tumor cell invasive properties have been used to charac brain-selected B16-B10n cells had identical autoradi- terize malignant neoplasms, and the selected metastatic ograms, except for 2 prominently labeled components of tumor cell lines that have been placed into in vitro invasion 95,000 and 100,000 molecular weight which incorporated assays have shown an increased ability to invade normal more '25I on the brain-colonizing lines (5). Alternatively, tissue matrix, compared to their unselected parental coun ovary-selected B16-010 cells did not possess modifications terparts. This has been experimentally demonstrated with in components with molecular weights of 95,000 and the use of the chorioallantoic membrane as an invasion 100,000 but instead showed increased exposure of surface substrate in tissue culture. When B16-F1 and B16-F10 were proteins with molecular weights of approximately 140,000 compared for their ability to infiltrate and invade the several and 150,000, compared to those of unselected B16 mela cell-thick chick embryo chorioallantoic membrane, only the noma lines. The degree of exposure to lactoperoxidase- selected B16-F10 cells invaded within 1 day, while the B16- catalyzed iodination of 95K and 100K components on brain- F1 cells were found loosely attached to the chorioallantoic selected and 140K and 150K components on ovary-selected membrane surface (31). Although other in vivo selected B16 melanoma lines correlated with enhanced metastatic tumor cell lines have not been compared with the use of potential. Cell surface changes detected by lactoperoxi these techniques, in vivo metastatic selected lines also dase-catalyzed iodination have also been seen on in vivo appear to be more invasive in syngeneic hosts when in selected RAW lymphosarcoma. Unselected RAW117 parental jected s.c. (Ref. 14; I. J. Fidler and G. L. Nicolson, unpub cells possessed a lactoperoxidase accessible protein with a lished data). molecular weight of 70,000 and this component was barely Invasion of malignant cells into surrounding normal tis detectable on liver-selected RAW117-H10 cells. Other com sues is thought to occur by means of a combination of ponents of higher (-250K) and lower (~30K) molecular mechanical and enzymatic mechanisms (for review, see weight were also modified to a lesser degree in the liver- Refs. 18 and 22). Release or display of degradative enzymes selected RAW lines (6). These results indicated that subtle may be important in invasion, because normal intercellular changes occur on the cell surfaces of the selected meta tissue matrix and similar structures are sensitive to enzy static lines; however, there was no direct proof that the matic destruction (2). In the B16-F system in which degra changes in lactoperoxidase iodinability were directly re dative enzymes have been measured, Bosmann ef al. (4) lated to tumor cell properties required for specific organ found that the more metastatic B16-F10 line in vitro had colonization. slightly higher glycosidase and protease (trypsin- and cath- Preferential in vivo localization of the selected tumor cell epsin-like) activities when compared to those of B16-F1, lines to certain organs suggested that adhesive interactions but only when the cells were replated at low density in

NOVEMBER 1978 4109

tissue culture. Other enzyme activities such as plasminogen host properties probably determine whether metastasis activator, a released serine proteinase that converts blood occurs in each tumor-host system. plasminogen to plasmin, which in turn hydrolyzes fibrin, do not directly correlate with metastatic potential (34). References Some immunological properties of the selected meta static tumor cell lines have been examined. Lung-selected 1. Aaronson, S. A., and Todaro. G. J. Development of 3T3-like Lines from B16-F melanoma cell lines carry a variety of normal and BAL8/C Mouse Embryo Cultures: Transformation Susceptibility to SV40. J. Cellular Physiol., 72: 141-148. 1968 tumor-associated antigens, as well as receptors for cyto- 2. Armstrong, P. B. Cellular Positional Stability and Intercellular Invasion. toxic lymphocytes, but dramatic changes in the expression Bioscience, 27: 803-809, 1977. 3. Boone, C. W.. Takeichi, N., Paranjipe, M., and Gilden, R. Vasoformative of these surface determinants have not been detected on Sarcomas Arising from BALB/3T3 Cells Attached to Solid Substrates. highly selected metastatic B16-F lines (22, 31). Interestingly, Cancer Res., 36. 1626-1633, 1976. liver-colonizing RAW117 lymphosarcoma lines apparently 4. Bosmann, H. B., Bieber, G. F., Brown, A. E., Case, K. R., Gersten, D. M., Kimmerer. T. W., and Lione, A. Biochemical Parameters Correlated have almost lost gp 71, the major RNA tumor virus envelope with Tumour Cell Implantation. Nature, 246: 487-489, 1973. glycoprotein, after 8 to 10 in vivo selections for liver colo 5. Brunson, K. W., Beattie, G., and Nicolson, G. L. Selection and Altered Tumour Cell Properties of Brain-colonising Metastatic Melanoma. Na nization (6). In this case a dramatic reduction in anti-gp 71 ture, 272: 543-545, 1978. binding correlated with the loss in exposure of surface 6. Brunson, K. W., and Nicolson, G. L. Malignant Variants of a Lymphosar protein with a molecular weight of 70,000 detected by coma Tumor Cell Line. J. Cell Biol., 75: 209a, 1977. lactoperoxidase-catalyzed iodination. 7. Brunson, K. W., and Nicolson, G. L. Selection and Biological Properties of Malignant Variantsofa Murine Lymphosarcoma. J. Nati. Cancer Inst., The immunological properties of lymphocyte-resistant in press, November 1978. B16-F10" 6 cells suggested that these lines have lost recep 8. Chew, E. C., Josephson, R. L., and Wallace, A. C. Morphologic Aspects of the Arrest of Circulating Cancer Cells. In: L. Weiss (ed.), Fundamental tors for immune lymphocytes. The resistance of the lympho Aspects of Metastasis, pp. 121-150. Amsterdam: North Holland Publish cyte-selected B16-F" 6 lines to lysis by syngeneic lympho ing Co., 1976. 9. Dunn. T. B. Normal and Pathologic Anatomy of the Reticular Tissue in cytes was not associated with loss of alloantigens such as Laboratory Mice, with a Classification and Discussion of Neoplasms. J. H-2, and these tumor cells remained sensitive to killing by Nati. Cancer Inst., 14: 1281-1433, 1954. allogeneic lymphocytes and syngeneic or allogeneic mac 10. Fidler, I. J. The Relationship of Embolie Homogeneity, Number, Size and Viability to the Incidence of Experimental Metastasis. European J. rophages (22). B16-F10" 6 cells did not protect against Cancer, 9: 223-227, 1973. lymphocyte-mediated cytotoxicity of lymphocyte-sensitive 11. Fidler, I. J. Selection of Successive Tumor Lines for Metastasis. Nature B16-F10 cells in cocultivation experiments, indicating that New Biol.,242: 148-149, 1973. resistance was not due to release of "blocking" factors by 12. Fidler, I. J. Immune Stimulation-inhibition of Experimental Cancer Me tastasis. Cancer Res.,34: 491-498, 1974. the B16-F10" 6 cells (16). Also, immunization of mice with 13. Fidler, I. J. Mechanisms of Cancer Invasion and Metastasis. In: F. F. B16-F10" 6 did not protect against challenge with B16-F10, Becker (ed.), Cancer: A Comprehensive Treatise, Vol. 4, pp. 101-131. New York: Plenum Press, 1975. although immunization with B16-F10 protected against sub 14. Fidler, I. J. Biological Behavior of Malignant Melanoma Cells Correlated sequent B16-F10 challenge. In vitro tumor cell-lymphocyte to Their Survival in Vivo. Cancer Res., 35: 218-224, 1975. 15. Fidler, I. J. Patterns of Tumor Cell Arrest and Development. In: L. Weiss aggregation experiments have demonstrated that resist (ed.), Fundamental Aspects of Metastasis, pp. 275-289. Amsterdam: ance to lymphocyte cytotoxicity is associated with de North Holland Publishing Co., 1976. 16. Fidler, I. J., and Bucana, C. Mechanism of Tumor Cell Resistance to creased lymphocyte binding (16, 17, 22). Collectively, these Lysis by Syngeneic Lymphocytes. Cancer Res., 37: 3945-3956, 1977. experiments suggest that the lymphocyte binding sites 17. Fidler, l. J., Gersten, D. M., and Budmen, M. B. Characterization in Vivo could be tumor-associated transplantation antigens and and in Vitro of Tumor Cells Selected for Resistance to Syngeneic that these are altered or missing on the B16-F1MÃŽcelllines Lymphocyte-mediated Cytotoxicity. Cancer Res.. 36: 3160-3165, 1976. 18. Fidler, I. J., Gersten, D. M., and Hart, l. R. The Biology of Cancer (22). Invasion and Metastasis. Advan. Cancer Res., in press, 1978. 19. Fidler, I. J., and Kripke, M. L. Metastasis Results from Pre-existing Variant Cells within a Malignant Tumor. Science, 797: 893-895, 1977. 20. Fidler, I. J., and Nicolson, G. L. Organ Selectivity for Implantation, Conclusion Survival and Growth of B16 Melanoma Variant Tumor Lines. J. Nati. Cancer Inst., 57: 1199-1202, 1976. The spread of solid tumors via blood-borne metastasis 21. Fidler, I. J., and Nicolson. G. L. Fate of Recirculating B16 Melanoma Metastatic Variant Cells in Parabiotic Syngeneic Recipients. J. Nati. usually marks the final, terminal stage of the pathogenic Cancer Inst., 58: 1867-1872, 1977. sequence of malignant neoplasms, and the complex series 22. Fidler, I. J., and Nicolson, G. L. The Immunobiology of Experimental of interactions between tumor cells and their host is of Metastatic Melanoma. In: M. Hanna, Jr. (ed.). Basic Immunologie Mech anisms in Cancer. New York: Marcel Dekker, Inc., in press, 1978. paramount importance in this process. The unique proper 23. Fidler, I. J., and Nicolson, G. L. Tumor Cell and Host Properties Affecting ties of malignant tumor cells determine, in part, whether the Implantation and Survival of Blood-borne Metastatic Variants of B16 they are able to invade, enter the blood, detach, circulate, Melanoma. Israel J. Med. Sci., 14: 38-50, 1978. 24. Gasic, G. J., Gasic, T. B., Galanti, N., Johnson, T., and Murphy, S. arrest, survive, and grow to form secondary tumor colonies. Platelet-tumor Cell Interaction in Mice. The Role of Platelets in the Few tumor cells in the initial population are probably able Spread of Malignant Disease. Intern. J. Cancer, 11: 704-718, 1973. 25. Hynes, R. O. Cell Surface Proteins and Malignant Transformation. to successfully continue at each step in the metastatic Biochim. Biophys. Acta, 458: 73-107, 1976. process, and this selective pressure undoubtedly results in 26. Kinsey, D. L. An Experimental Study of Preferential Metastasis. Cancer, enrichment of cells that possess unique properties in the Ã•3:674-676, 1960. 27. Liotta, L. A., Kleinerman, J., and Saidel, G. M. The Significance of highly metastatic cell lines. In addition to unique tumor cell Hematogenous Tumor Cell Clumps in the Metastatic Process. Cancer characteristics, host properties such as immune response Res.,36: 889-894, 1976. can influence metastasis. The role of host immune antitu- 28. Marchalonis, J. J., Cone, R. E., and Santer, V. Enzymatic Iodination: A Probe for Accessible Surface Proteins of Normal and Neoplastic Lym mor response in metastasis has been reviewed elsewhere phocytes. Biochem. J., 124: 912-927, 1971. (18, 22). It suffices to mention here that both tumor cell and 29. Nicolson, G. L. Transmembrane Control of the Receptors on Normal and

4110 CANCER RESEARCH VOL. 38

Tumor Cells. II. Surface Changes Associated with Transformation and 39. Salsbury, A. J. The Significance of the Circulating Cancer Cell. Cancer Malignancy. Biochim. Biophys. Acta. 458: 1-72. 1976. Treatment Rev.. 2: 55-72, 1975. 30. Nicolson, G. L. Cell and Tissue Interactions Leading to Malignant Tumor 40. Sato, H., and Suzuki, M. Deformability and Viability of Tumor Cells by Spread (Metastasis). Am. Zool., 78: 77-86, 1978. Transcapillary Passage, with Reference to Organ Affinity of Metastasis 31. Nicolson, G. L.. Birdwell.C. R., Brunson, K. W.. Bobbins, J. C., Beattie, in Cancer. In: L. Weiss (ed.). Fundamental Aspects of Metastasis, pp. G.. and Fidler, l. J. Cell Interactions in the Metastatic Process: Some 311-317. Amsterdam: North Holland Publishing Co., 1976. Cell Surface Properties Associated with Successful Blood-borne Tumor 41. Strauch, L. The Role of Collagenases in Tumor Invasion. In: D. Tarin Spread. In: J. Lash and M. M. Burger (eds.), Cell and Tissue Interactions, (ed.), Tissue Interactions in Carcinogenesis. pp. 399-434. New York: pp. 225-241. New York: Raven Press. 1977. Academic Press. Inc., 1972. 32. Nicolson, G. L., and Brunson, K. W. Organ Specificity of Malignant B16 42. Sugarbaker, E. D. The Organ Selectivity of Experimentally Induced Melanomas: In Vivo Selection for Organ Preference of Blood-borne Metastasis in Rats. Cancer. 5: 606-612. 1952. Metastasis. Gann Monogr. Cancer Res.. 20. 15-24. 1977. 43. Vaage, J., Chen. K., and Merrick, S. Effect of Immune Status on the 33. Nicolson, G. L., and Winkelhake, J. L. Organ Specificity of Blood-borne Development of Artificially Induced MÃ©tastasesin Different Anatomical Tumour Metastasis Determined by Cell Adhesion? Nature, 255: 230-232. Locations. Cancer Res., 37: 496-500, 1971. 1975. 44. Warren, B. A. Tumor Metastasis and Thrombosis. In: E. Deutsch. K. M. 34. Nicolson, G. L., Winkelhake, J. L., and Nussey, A. C. An Approach to Brinkhous, K. Lechner, and E. Hinnom (eds), Thrombosis: Pathogenesis Studying the Cellular Properties Associated with Metastasis: Some in and Clinical Trials, pp. 139-156. Stuttgart: F. K. Schattauer-Verlag. 1973. Vitro Properties of Tumor Variants Selected in Vivo for Enhanced 45. Winkelhake, J. L., and Nicolson, G. L. Determination of Adhesive Metastasis. In: L. Weiss (ed.), Fundamental Aspects of Metastasis, pp. Properties of Variant Metastatic Melanoma Cells to BALB/3T3 Cells and 291-303. Amsterdam: North Holland Publishing Co., 1976. Their Virus-transformed Derivatives by a Monolayer Attachment Assay. 35. Nowell, P. C. The Clonal Evolution of Tumor Cell Populations. Science, J. Nati. Cancer Inst., 56: 285-291, 1976. 794: 23-28, 1976. 46. Wood, S., Jr. Experimental Studies of the Intravascular Dissemination of 36. Parks, R. C. Organ-specific Metastasis of a Transplantable Reticulum Ascitic V2 Carcinoma Cells in Rabbit, with Special Reference to Fibrin- Cell Sarcoma. J. Nati. Cancer Inst.. 52: 971-973, 1974. ogen and Fibrinolytic Agents. Bull. Schweiz. Akad. Med. Wiss.. 20: 92- 37. Potter, M., Rahey, J. L., and Pilgrim, H. I. Abnormal Serum Protein and 121,1964. Bone Destruction in Transmissible Mouse Plasma Cell Neoplasm (Multi 47. Zeldman. I. Metastasis: A Review of Recent Advances. Cancer Res., 77: ple Myeloma). Proc. Soc. Exptl. Biol. Med., 94: 327-333, 1957. 157-162, 1957. 38. Roblin, R., Chou, I-N., and Black. P. H. Proteolytic Enzymes, Cell 48. Zeidman, I. The Fate of Circulating Tumor Cells. I. Passage of Cells Surface Changes and Viral Transformation. Advan. Cancer Res., 1422: through Capillaries. Cancer Res..27: 38-39. 1961. 203-259. 1975.

NOVEMBER 1978 4111

Garth L. Nicolson, Kenneth W. Brunson and Isaiah J. Fidler

Cancer Res 1978;38:4105-4111.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/38/11_Part_2/4105

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/38/11_Part_2/4105. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.