Influence of tumor environment and host immunity on tumor progression and metastasis


Material Information

Influence of tumor environment and host immunity on tumor progression and metastasis
Physical Description:
xii, 101 leaves : ill. ; 29 cm.
Parratto, Nanette P., 1956-
Publication Date:


Subjects / Keywords:
Neoplasms -- immunology   ( mesh )
Neoplasm Metastasis   ( mesh )
Pathology thesis Ph.D   ( mesh )
Dissertations, Academic -- Pathology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1988.
Bibliography: leaves 91-99.
General Note:
General Note:
Statement of Responsibility:
by Nanette P. Parratto.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 023090788
oclc - 20429303
System ID:

Table of Contents
    Title Page
        Page i
        Page ii
        Page iii
    Table of Contents
        Page iv
        Page v
    List of Tables
        Page vi
    List of Figures
        Page vii
        Page viii
    Key to abbreviations
        Page ix
        Page x
        Page xi
        Page xii
    Introduction and review of literature
        Page 1
        Page 2
        Page 3
    Materials and methods
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
    Biographical sketch
        Page 100
        Page 101
        Page 102
        Page 103
Full Text






This work is dedicated to my father, for his patience and loving support over the many years of its duration; to the memory of my mother who taught me well the power of perseverance; and to the memory my trusting friend, O'Dee, for her unconditional love.


Dr. Arthur Kimura's gentle and persuasive guidance is

most gratefully acknowledged. He has taught me to think like a scientist (we hope). The constructive criticisms of my graduate committee advisors, Dr. Richard T. Smith, Dr. Sigurd Normann, Dr. K. Kendall Pierson, Dr. Maron Calderwood-Mays, and Dr. Chris West, are also greatly appreciated. The invaluable advice of Dr. Byron Croker, Cynthia Bevis, and Norma Houghwout regarding immunohistochemical techniques is much appreciated. The contributions of Roberto Luchetta, Linda Lee-Ambrose, Janice Odelbralski, Xiang-Hong Tian, and Donald Dugger toward the completion of this work is also much appreciated.




ACKNOWLEDGEMENTS ........................................... iii

LIST OF TABLES .............................................. vi

LIST OF FIGURES ............................................ vii

KEY TO ABBREVIATIONS ........................................ ix

ABSTRACT ..................................................... x


MATERIALS AND METHODS ........................................ 4
Mice .................................................... 4
Murine Melanoma Cell Lines .............................. 4
Other Murine Cells and Cell Lines ....................... 6
Monoclonal Antibodies ................................... 7
Affinity Purification of Monoclonal Antibody for
Biotin and Fluorescein Isothiocyanate
Conjugation ........................................ 7
Flow Cytometry .......................................... 9
Subcutaneous Tumor Generation and Spontaneous and
Experimental Metastasis Assays ..................... 9
Immunocytochemistry, Immunohistology and Direct
Immunofluorescence ................................. 10
Immunization Protocols .................................. 12
Radioimmunologic Determination of MoAb and Immune
Sera Binding Indices ............................... 13
Immuno-magnetic Bead Separation of Immunoreactive
Subpopulations .................................... 14
Cell Solubulization, Polyacrylamide Gel
Electrophoresis and Western Blotting .............. 14
Statistical Analyses .................................... 15

RESULTS ...................................................... 16
Anti-Met-72 Monoclonal Antibody Binding to Fixed
Melanoma Cells ..................................... 16
Optimum Biotin Substitution of Monoclonal Antibody ..... 16
Biotin Conjugated Anti-Met-72 Monoclonal Antibody
Specificity ....................................... 17


Fluorescence Activated Cell Sorting Selection of
Met-72 Positive Variants Isolated From a Fresh
Experimental Ovarian Metastasis........................ 17
Bone Marrow Derived Clones of B16 Melanoma Express
High Levels of Met-72.............................. 18
Immunocytology of Fresh Ovarian and Lung Metastases .... 18
Localized Distribution of Met-72 Positive Variants
Within Progressing Subcutaneous Melanoma .......... 19
Met-72 Positive Variants Localized Within B16
Melanoma Metastases................................ 20
Specificity Characteristics of Syngeneic Antibodies
Generated Against B16 Melanoma ....................... 20
Syngeneic Antibody Response Against a C3H Melanoma ..... 23
Metastatic Potential of Melanoma Populations
Negatively Selected By Syngeneic
Anti-Melanoma Antibodies.........................23
Biochemical Characterization of Anti-B16-Fl
Reactive Species................................... 25
In Situ Characterization of Metastatic Variants Within
Experimental Metastases and Subcutaneous Masses of
B16 Melanoma....................................... 25
Immunohistology of Syngeneic Healing Skin Wounds and
Embryos............................................ 28

DISCUSSION................................................... 79

REFERENCES.................................................... 91

BIOGRAPHICAL SKETCH......................................... 100




Table 1. Results of Lung Colonization Assays ...............29

Table 2. Syngeneic Immune Sera Reactivity with C57BL/6
Melanomas .............. .. ...................49

Table 3. Syngeneic Immune Sera Reactivity with C57BL/6
Melanomas......................................... 51

Table 4. Results of Analysis of Variance Comparing Groups
of Mice Injected with Negatively Selected B16
Melanoma Cells.................................... 64

Table 5. Level of Monoclonal Antibody Binding to Cryostat
Sections of Subcutaneous B16 Melanomas ........... 72




Figure 1. Radiolabeled anti-Met-72 monoclonal antibody
binding to fresh or ethanol fixed C57BL/6
melanoma cells .................................. 31

Figure 2. Optimum biotin conjugation of anti-Met-72
monoclonal antibody ............................. 33

Figure 3. Retention of binding specificity of anti-Met-72
MoAb after biotin conjugation ................... 35

Figure 4. Flow cytometric analysis of Met-72 positive
metastatic variants within a fresh ovarian
metastasis of B16-Fl ............................ 37

Figure 5. Clones derived from experimental metastases to
ovaries retain a high expression of Met-72 upon
repeated cycling in vivo ........................ 39

Figure 6. Anti-Met-72 monoclonal antibody binding to clones
of bone marrow derived melanoma metastases ...... 41

Figure 7. Met-72 positive variants of B16 melanoma detected
in cytospin and impression smear preparations of
metastases ...................................... 43

Figure 8. Localization of Met-72 positive variants within
developing B16 subcutaneous melanoma ............ 45

Figure 9. Direct immunofluorescence of Met-72 positive
variants localized within B16 melanoma
metastases ...................................... 47

Figure 10. Flow cytometric analysis of anti-B16-Fl immune
sera (A) and anti-B16-Fl monoclonal antibody (B)
binding to C57BL/6 melanomas .................... 53

Figure 11. Anti-B16-F1 monoclonal antibody binding to a panel
of murine cells ................................. 55

Figure 12. Syngeneic antibody response against K1735
melanoma ........................................ 57


Figure 13. Flow cytometric analysis of B16-Fl cells before
and after immunomagnetic bead selection and
removal of reactive subpopulations .............. 59

Figure 14. Flow cytometric analysis of B16 melanoma before
and after immunomagnetic bead selection and
removal of reactive subpopulations .............. 61

Figure 15. Experimental metastatic activity of negatively
selected B16 melanoma cells ..................... 63

Figure 16. Western blot analysis of anti-B16-Fl immune sera
and MoAb binding to lysates of syngeneic melanoma metastatic variant cell lines ................... 65

Figure 17. Immunocytology of metastatic variants exfoliated
from fresh experimental lymph node metastases of
BL6-10 .......................................... 67

Figure 18. Localization of B16 melanoma metastatic variants
within metastases by immunoperoxidase
cytochemistry .................................... 69
Figure 19. Localization of B16 melanoma variants within
experimentally colonized lungs .................. 71

Figure 20. Localization of B16 melanoma variants within
subcutaneous tumors maintained in normal syngeneic hosts ........................................... 74

Figure 21. Localization of B16 melanoma variants within
subcutaneous tumors maintained in 500 Gy
irradiated syngeneic hosts ...................... 76

Figure 22. Localization of B16 melanoma variants within
subcutaneous tumors maintained in allogeneic
athymic hosts ................................... 78


KEY TO ABBREVIATIONS aminoethylcarbazole (AEC) antibody (Ab) avidin-biotin-horseradish peroxidase complex (ABC) counts per minute (cpm) Eagles Hank's amino acid media (EHAA) fetal bovine serum (FBS) fluorescein isothiocyanate (FITC) fluorescein isothiocyanate conjugated F(ab')2 sheep anti-mouse


flow cytometry (FACS) Gray (Gy)

intraperitoneally (i.p.) intravenous (i.v.) monoclonal antibody (MoAb) newborn calf serum (NCS) normal mouse serum (NMS) phosphate buffered saline (PBS) phosphate buffered saline with 0.7 mM EDTA and 0.6 mM glucose


protein A (pA) radioimmunoassay (RIA) streptavidin (sA) subcutaneous (s.c.) ix

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

Nanette P. Parratto

December 1988

Chairman: Arthur K. Kimura
Major Department: Pathology

Monoclonal antibodies (MoAb) against Met-72, a 72,000dalton cell surface glycoprotein, have been used to characterize B16 melanoma variants expressing high experimental metastatic activity. Freshly isolated bone marrow and ovarian metastases of B16 melanoma were found to express high levels of Met-72 by radioimmunoassay (RIA). Lung and ovarian metastases were shown immunocytochemically to contain melanoma cells with heterogeneous levels of Met-72 antigen. Flow cytometric analysis (FACS) of an ovarian metastasis revealed a discrete subpopulation of positive cells. Clones derived from this ovarian metastasis were found to express high levels of Met-72 and high experimental metastatic potential to ovaries. Finally, Met-72 positive variants were localized by routine immunohistology within developing metastatic and subcutaneous B16 melanomas. The few


cells which showed high intensity Met-72 staining have been consistently observed to be within microscopic foci in colonized organs and positioned within regions of subcutaneous masses.

Melanoma has been shown to be immunogenic in experimental and clinical situations; however, syngeneic antibodies (Ab) against highly metastatic variants have only been obtained by experimental manipulation of the host immune response. Binding of anti-B16 melanoma Ab elicited by syngeneic intravenous immunization was inversely correlated with metastatic activity of melanoma lines and clones by RIA. Flow cytometric analyses similarly revealed that the proportion of reactive cells in melanoma populations was inversely correlated to their experimental metastatic activity. AntiB16 melanoma MoAb derived by hybridoma technology from intravenously challenged mice, showed distinct anti-melanoma specificity in RIA. Monoclonal antibody directed against nonmetastatic melanoma variants defined specific bands in Western blot analyses of Mr 45,000 and 50,000. Finally, negative cellular selection techniques revealed an increased experimental metastatic potential associated with the nonimmunogenic B16 melanoma subpopulation.

The results of these experiments suggest that 1) Met-72 antigen expression may be a common surface phenotype of B16 melanoma metastatic variants; 2) microenvironmental factors may be instrumental in the induction, attraction or


maintenance of metastatic variants, and 3) the host antibody response may be a key factor in the regulation of levels of metastatic variants during tumor progression by selectively targeting poorly metastatic populations.



The dynamics of tumor cell heterogeneity is clearly evident from clinical and experimental studies of tumor progression (1,2). The diagnosis and treatment of solid tumors has been seriously impaired by a poor understanding of the clonal evolution of heterogeneous populations within developing tumors (3,4). Experimental evidence documents the existence of subpopulations of metastatic tumor cell variants, exhibiting a range of metastatic potentials from low to high (3-6). The precise nature and expression of the phenotypic characteristics during the natural history of metastasis remains unique to each tumor bearing host and the particular solid tumor (7). Clearly, classical pathologic descriptions have been inadequate to identify unique metastatic variants within surgical specimens. This remains a hindrance to quantitative classification of metastatic potential and patient prognosis.

Paulus et al. (8) have implicated regional variations of differentiation in B16 melanoma, localized by morphologic heterogeneity. Direct, in situ visualization of metastatic tumor cell variants has not been possible until recently. We have identified a Mr 72,000 cell surface glycoprotein (Met-72) quantitatively associated with highly metastatic tumor cell



variants of the B16 melanoma (5). More recently, a Mr 83,000 native form of this surface molecule (Met-72/83) has been described (9). The experimental metastatic potential of over 30 B16 melanoma clones has been correlated to a quantitative surface expression of Met-72 (5,10,11). In addition, fluorescence activated cell sorting (FACS) has now been used to directly isolate metastatic variants from the heterogeneous parental B16-Fl tumor (12).

The present study was designed to visualize, isolate and histologically localize metastatic variants present in fresh B16 melanoma tumor masses. Anti-Met-72 monoclonal antibodies (MoAb) reveal a unique localization of Met-72 positive cells along the invading front of the progressing tumor. These experiments show that anti-Met-72 MoAb previously used to characterize the expression of Met-72 in vitro, may also be useful for isolation and localization of highly metastatic variants in situ in progressing primary and metastatic B16 melanoma.

Much of the success in defining antigenic systems for

melanoma is related to the immunogenicity of this tumor type. Syngeneic and autochthonous antibody (Ab) responses against melanoma have been demonstrated in a number of experimental (13-20) and clinical situations (21-24), and yet in syngeneic combinations, Ab against highly metastatic variants have only been obtained through experimental manipulation of the host immune response (5). Taken together, these results suggest


that the immunodominant melanoma population may in fact be those cells with a low metastatic activity.

The current view that tumors are comprised of

heterogeneous subpopulations with widely varying and changing metastatic potentials, has been supported in a number of tumor systems (3,4). Selective host responses against the more poorly metastatic subpopulations could thus be expected to dramatically influence the course of tumor growth, progression, and metastasis.

Experiments described here are focused at characterizing the range of Ab specificities elicited by metastatically distinct cell lines and clones of murine melanoma and the potential impact of such responses on tumor progression and metastasis. Antisera and MoAbs derived in these studies have been used to fractionate melanoma populations and deduce the metastatic potentials of the target population. Negatively selected melanoma cells are shown to have enhanced metastatic activity after removal of the anti-melanoma reactive population. These findings strongly support the conclusion that subpopulations of poorly metastatic melanoma are primary targets of host humoral immunity. The relationship of these findings to other anti-melanoma Ab is discussed.



C57BL/6, C3H/HeJ, and Balb/c nu/nu mice were purchased

from the Jackson Laboratory (Bar Harbor, ME) and maintained in the Tumor Biology Unit mouse colony, Department of Pathology, in accordance with the National Institutes of Health Guidelines for the Use of Experimental Animals. Female mice,

8 to 16 weeks of age were used in these studies. C57BL/6 mice were exposed to 500 Gy total body irradiation.

Murine Melanoma Cell Lines

Early passages of the B16-Fl, B16-FlO and B16-BL6

derivatives of the C57BL/6 melanoma, B16, were obtained from the Division of Cancer Treatment Tumor Bank (E. G. and G. Mason Research Institute, Worchester, MA) where they had been deposited by Dr. I. J. Fidler. A recently derived syngeneic melanoma, JB/RH, was provided by Dr. Jane Berklehammer (AMC Cancer Research Center, Denver, CO) (25,26) for comparison in our studies. Clones of the B16 melanoma were derived by limiting dilution and micromanipulation (5) from both the poorly metastatic B16-Fl cell line and the in vitro selected, highly invasive metastatic variant, B16-BL6 (27). Stocks from early passages of all cell lines and clones were frozen at

-70C and restarted every 8 to 12 weeks to limit the


possibilities of functional and phenotypic drift. All cell lines and clones were maintained in vitro at 37*C in a humidified incubator containing 8% CO2, by subculturing every

4 days. Monolayers of cells were detached from the petri dishes (Costar #3100, Cambridge, MA) by a 3-minute room temperature incubation with 136 mM NaCl, 3 mM KC1, 1.5 mM KH2PO4, 8 mM Na2HPO4, 0.7 mM EDTA and 0.6 mM glucose (cPEG)

(28). For routine passage, cells were washed and replated at a concentration of 5 X 105 cells/10 cm dish in 10 ml media.

The poorly metastatic B16-Fl parental line and highly metastatic B16-Fl0 line were maintained in Cellgro MEM (Sybron, Washington, D.C.) supplemented to 10% with heat inactivated fetal bovine serum (Grand Island Biologicals [GIBCO], Grand Island, NY), 100 U/ml penicillin, 100 Ag/ml streptomycin, 1 mM pyruvate, 2 mM glutamine and 27 mM NaHCO3. The highly metastatic clone BL6-10, derived from B16-BL6, clones derived by limiting dilution from extrapulmonary metastases of B16-Fl, and the poorly metastatic line, JB/RH, were maintained in Hanks' balanced salt solution, minimal essential media amino acids, nonessential amino acids and vitamin solution (GIBCO), and nucleosides (Sigma, St. Louis, MO) (EHAA) supplemented to 10% with newborn calf serum (GIBCO), 100 U/ml penicillin, 100 Mg/ml streptomycin, 18 mM NaHCO3 and 0.05 mM P-mercaptoethanol (29).

Other Murine Cells and Cell Lines
A panel of murine cells and cell lines was assembled for specificity analysis of the anti-Bl6-Fl antisera and MoAb secreting hybridomas captured by the immunization procedures. Syngeneic tumor cell lines included PAK17 and PC8 (methylcholanthrene induced fibrosarcomas), Lewis lung carcinoma, EL-4 T-cell lymphoma, and normal syngeneic lung fibroblasts (provided by Dr. Paul Klein, Department of Pathology, University of Florida). An allogeneic B cell lymphoma, 2PK3 (Balb/c origin) was also obtained from the Department of Pathology, Tumor Bank. These cell lines were maintained in Dulbecco's modified minimum essential media (GIBCO) supplemented to 10% with FBS (GIBCO), 100 U/ml penicillin, 100 Ag/ml streptomycin, 1 mM pyruvate and 2 mM glutamine (5). An allogeneic UV induced melanoma, K1735 (C3H/HeN origin) (30) was obtained from the Department of Pathology, Tumor Bank and maintained in 10% newborn calf serum supplemented EHAA. The P815 mastocytoma of the DBA/2 strain (provided by Dr. Sigurd Normann, Department of Pathology, University of Florida), was maintained in RPMI (Gibco) supplemented to 10% with FBS plus 100 Ag/ml streptomycin, 100 U/ml penicillin. Cells were harvested during log phase of growth for use in radioimmunoassays (RIA). Single cell suspensions of normal C57BL/6 and Balb/c spleen cells were prepared by mechanical disaggregation and lysis of RBC in 17 mM Tris, 144 mM NH4Cl, pH 7.2.


Monoclonal Antibodies

The anti-Bl6-F1 monoclonal antibodies (MoAb) derived in these studies were generated as described (5,31). Briefly, spleen cells from mice injected i.v. with 3 X 105 B16-Fl cells were fused with BALB/c myeloma SP2/O-Agl4 cells and dispensed under limiting dilution conditions into 96 well plates. Supernatants from growth positive wells were screened by RIA for binding to the poorly metastatic melanoma line, JB/RH. The highest binding hybridomas were subcloned and screened for specificity using JB/RH and BL6-10 cells. Monoclonal antibody isotypes were determined by enzyme linked immunosorbant assay (Southern Biotechnology Associates, Inc., Birmingham, AL). Monoclonal antibodies derived by syngeneic immunization of C57BL/6 mice with B16-Fl melanoma (anti-Bl6-Fl MoAb; 2D8.3B1 and 2H5.1B1, designating clone and subclone) were both found to IgG3:kappa. Another syngeneic anti-Bl6-Fl MoAb (lFll.1E12) was found to be IgG2a:kappa. Anti-Met 72 MoAb were generated by syngeneic immunization of C57BL/6 mice with selected B16 melanoma clones. The specificity and characteristics of these MoAbs have been reported in detail (5,31). Hybridoma cells secreting an isotype identical, negative control MoAb used in this study (anti-sheep red blood cell, N-S.7, IgG3:kappa) were obtained from the American Type Culture Collection.

Affinity Purification of Monoclonal Antibody for Biotin and Fluorescein Isothiocyanate Conjugation

Monoclonal antibodies were affinity purified by

fractionation through protein A-Sepharose 4B (Pharmacia,


Piscataway, NJ) (32) and purity checked by sodium dodecyl sulfate 10% polyacrylamide gel electrophoresis (33). Affinity purified IgG (1 mg/ml) was dialyzed against 100 mM NaHCO3, 200 mM NaCl pH 8.2 and then reacted with 2 mM sulfosuccinimidyl 6(biotinamido)hexanoate (Pierce Chemical, Rockford, IL) in dimethylformamide (Sigma), at biotin ester:protein ratios of 1:2.5 to 1:10 (weight/weight) for 4 hours at room temperature, in the dark. The reaction was stopped by the addition of 1 M NH4Cl to a final concentration of 100 mM in the reaction mixture. Unreacted biotin was then removed by exhaustive dialysis against PBS. Biotin-conjugated protein concentrations were determined by optical density at 280 nm

(34). Optimal biotin ester/protein ratios used for the conjugation of the various preparations were determined empirically by flow cytometric analysis on cell preparations

(35). Briefly, 2 X 106 melanoma cells were incubated with biotin conjugated antibody at 4*C for 30 min. in the dark. After 3 washes in PBS supplemented to 2% with FBS, either native or biotin conjugated MoAb were incubated with fluorescein isothiocyanate conjugated F(ab')2 sheep anti-mouse IgG (Cooper Biomedical, Malvern, PA) (FITC-SAM) or FITC conjugated avidin (Vector Laboratories, Burlingame,CA). After 3 washes with PBS, 1 X 104 cells were counted and analyzed by flow cytometry (12,35).

Fluorescein isothiocyanate (FITC) conjugation was as described by Goding (36). Briefly, affinity purified MoAb


were dialyzed against 80 mM Na2CO3, 200 mM NaHCO3 pH 9.5 and reacted with 2 mg FITC per ml dimethylsulfoxide (Sigma) at a FITC/protein molar ratio of 4. Unreacted FITC was separated by desalting through Sephadex G-25 (Pharmacia).

Flow Cytometry

Cultured cell lines were harvested as described for

routine passage and washed twice in ice cold PBS. Aliquots of 106 cells were incubated at 4C for 45 min. with diluted normal or immune serum or with free or conjugated MoAb as described in the results section. After two washes with 100 volume excess of ice cold PBS, labeled cells were counterstained with FITC-SAM for 45 minutes at 4C. Labeled and stained cells were washed, resuspended at 5 X 105 cells per ml and filtered through 41g nylon mesh (Spectrum, Los Angeles, CA). Flow cytometric analysis was performed with the FACSTAR argon laser (Becton/Dickinson, Oxnard, CA) the photomultiplier voltage for green fluorescence set at 400, the amplifier gain set at log scale and the photodiode for forward scatter set at 0 for sample analyses. Data was collected and analyzed by the Becton/Dickinson Consort 30 computer program (Oxnard, CA) (35).

Subcutaneous Tumor Generation and
Spontaneous and Experimental Metastasis Assays

Single cell suspensions of B16-Fl, B16-FlO or BL6-10

metastatic variant lines were injected s.c. (2 X 106 cells per

0.1 ml 136 mM NaCl, 3 mM KCI, 1.5 mM KH2PO4, 8 mM Na2HPO4 [PBS]) into the caudal mid-dorsum of age matched normal or 500


Gy irradiated C57BL/6 mice or into Balb/c nu/nu mice, to generate primary tumor masses. Animals were anesthetized with 50 mg/kg Na pentabarbital i.p. to aseptically excise tumors and associated skin at various intervals of tumor growth; the wounds were then clipped. Spontaneous metastasis formation was assessed by sacrifice and necropsy 21 days postoperatively. Experimental metastasis of all melanoma lines was assessed by sacrifice and necropsy 21 days after lateral tail injection of 3 X 105 cells/0.2 ml PBS into age matched syngeneic female mice.

Immunocytochemistry, Immunohistology and
Direct Immunofluorescence

Cytospin and impression smear preparations of freshly excised B16 melanoma tumor masses were examined for Met-72 expression by immunoperoxidase staining. Single cell suspensions of tumor masses were generated by teasing tissue to 1 mm3 pieces followed by an 8 to 10 hour incubation at 4*C in 10 ml/gm tumor weight, EHAA supplemented to 0.1% with collagenase V (Sigma) (37). Viable cells were separated on Ficoll Hypaque (Pharmacia), washed, and endogenous peroxidase activity blocked by 30 min. incubation in PBS supplemented to 5% with heat inactivated FBS and 0.3% with H202 at 40C. Cells were labeled in the dark with biotin conjugated MoAb for 30 min. at 4*C, washed and incubated with avidin-biotinhorseradish peroxidase complex ([ABC] Vector Laboratories, Burlingame, CA) for 30 min. at 40C. Approximately 3 X 104 labeled cells were cytocentrifuged onto poly-D-lysine coated

microslides, air dried, fixed in acetone for 30 seconds and rehydrated in PBS. The reaction was developed with aminoethylcarbazole ([AEC] Sigma) in the presence of H202. The slides were washed and counterstained with Harris' hematoxylin (Sigma) (38,39).

Freshly excised tumor masses, rinsed in ice cold PBS, were gently touched to coated slides to create impression smears. Slides were air dried, fixed in acetone for 10 min., rehydrated with PBS and either immediately processed by immunohistology or air dried and stored at -70C for future use. Impression smears provide a rapid screening tool for detection of antigen expression in vivo (40,41).

B16 melanoma tumors from subcutaneous or experimental metastatic lesions were excised, snap frozen in isopentane chilled with liquid nitrogen and stored at -70C until sectioned. Subcutaneous masses were examined after 5 to 21 days of growth; metastatic lesions after 14 to 21 days of growth. Skin wounds created surgically on the mid-dorsum were excised at various points during healing, snap frozen and stored at -700C. C57BL/6 embryos were collected at days 11 and 15 of gestation, snap frozen and stored at -70C.

Cryostat sections 2 to 12 Am thick were stained according to a modification of described techniques (42,43). Briefly, air dried, acetone fixed, sections were flushed with 0.01% avidin (Vector Laboratories) for 15 min. and then 0.134 mM biotin (Sigma) for 15 min. with a PBS wash between. Sections


were then blocked for endogenous peroxidase activity as for cytospin preparations. Biotin conjugated MoAb at predetermined optimal concentrations were layered onto sections and allowed to react for 30 min. in a dark, humidified chamber. Slides were washed with 20 volume excess PBS and reacted with the ABC reagent for 30 min. The reaction was developed with AEC, the slides counterstained with hematoxylin and examined under oil immersion.

Frozen sections were similarly air dried, acetone fixed and rehydrated with PBS for direct immunofluorescence using FITC conjugated MoAb. Slides were washed with PBS, coverslipped and examined under high power using a 490 nm excitation filter and 510 nm emission filter for green fluorescence.

Immunization Protocols

Age matched female mice were injected via the tail vein with in vitro maintained melanoma lines after the cells had been lifted with cPEG and washed three times with ice cold PBS. Mice (4 to 6 per experimental group) were immunized with various numbers of cells and immune sera was harvested at various times to determine dose response and kinetics curves. Optimal dosages and response times were determined and used to further characterize the serologic immune response. Normal mouse serum (NMS) harvested from age matched littermates served as background control. Following exsanguination, animals were killed by cervical dislocation and then


necropsied. Systemic experimental metastatic colonization was assessed by counting lung and extrapulmonary tumor nodules after fixation of the organs in buffered formalin (6).

Radioimmunologic Determination of MoAb and
Immune Sera Binding Indices

MoAb binding to the various cell types was measured 1) directly using radiolabeled MoAb (44), or 2) indirectly with 125, labeled protein A as described (45) or with 125, labeled streptavidin (Zymed Laboratories, Inc., San Francisco, CA) for biotin conjugated MoAb using a modification of Philpott, et al. (46). Briefly, 2 X 105 freshly harvested cultured cells were incubated for 45 minutes at 4C with equivalent dilutions of normal or immune mouse serum or MoAb. Washed cells were incubated for an additional 45 minutes with 2 X 105 cpm radiolabeled protein A or streptavidin. Radioactivity remaining bound to the cells after washing was assessed by gamma scintillation counting in an LKB gamma counter (LKB Instruments Inc., Gaithersburg, MD). Specific MoAb or immune sera binding to the different cell types is expressed as mean S.D. cpm bound or as a binding index (47), calculated from duplicate determinations as follows: mean cpm bound with specific

MoAb or immune sera

Binding index

mean cpm bound with

control MoAb or NMS


This normalizes individual cell line differences in background binding and allows comparison between the different cell types.

Immunomagnetic Bead Separation of Immunoreactive Subpopulations

Immunomagnetic beads (Dynabeads M-450, Dynal Inc., Fort Lee, N.J.) coated with sheep anti-mouse IgG were used to separate cells which had been incubated with normal versus anti-B16-Fl immune sera or anti-sheep RBC MoAb versus antiB16-Fl MoAb. Cells were harvested from culture, washed and resuspended in hybridoma exhaustion supernatant or PBS supplemented to 5% with normal or immune serum, and incubated for 45 minutes on ice. Immune labeled cells were washed and then incubated with immuno-magnetic beads at a bead to target cell ratio of 3 to 1 for 10 minutes at 4C. Cells bound to beads were separated from those cells remaining free in suspension by applying a cobalt steel magnetic force (Dynal MPC-l, Dynal Inc., Fort Lee, N. J.) for 1 minute (48). This process was repeated once again with the non-bound cells. The remaining negatively fractionated cells were analyzed by flow cytometry. A portion was injected intravenously at 3 X 105 cells per 0.2 ml PBS to assess experimental metastatic capacity.

Cell Solubilization, Polvacrylamide Gel Electrophoresis and Western Blotting
Cell lines maintained in vitro were lifted with cPEG,

washed in PBS and solubilized in PBS supplemented to 0.5% with Nonidet P-40 [(NP-40) Bethesda Research Laboratories,


Gaithersburg, MD], 1 % epsilon amino caproic acid and 1 mM phenylmethylsulfonylfluoride (Sigma Chemical Co., St. Louis, MO) at 4*C for 30 min. Supernate was recovered by centrifugation of the detergent lysate for 15 min. at 12,000 X g, 4*C (49,50). Lysates and marker proteins (Rainbow Markers, Amersham, Arlington Heights, IL) were electrophoresed on 5 to 15% or 10% acrylamide slab gels under reducing conditions

(33). Proteins were electro-blotted to polyvinylidene difluoride solid membrane (Immobilon, Millipore Corp., Bedford, MA) as described (51,52) and immuno-stained with anti-Bl6-Fl antisera or anti-Bl6-Fl MoAb. Antibody bound protein was visualized by protein A conjugated peroxidase catalysis of H202 in the presence of diaminobenzidene (39).

Statistical Analyses

Significance was determined by analysis of variance, computed using the SAS program (Biostatistics, J. Hillis Miller Health Center, University of Florida).


Anti-Met-72 Monoclonal Antibody Binding to Fixed Melanoma Cells

The highly metastatic B16 melanoma variant, BL6-10, or syngeneic poorly metastatic line, JB/RH (Table 1), were reacted against radiolabeled anti-Met-72 MoAb. Radioactivity remaining bound to fresh or 95% ethanol fixed cells in suspension is shown in Figure 1. The highly metastatic clone, BL6-10, binds equivalently high levels of Ab whether fresh or fixed. Fixation of the poorly metastatic line JB/RH does not artificially increase binding of anti-Met-72 MoAb.

Optimum Biotin Substitution of Monoclonal Antibody

Biotin conjugation to proteins is an empiric reactiion,

the optimum conditions varying uniquely for each protein (34). Affinity purified anti-Met-72 or control MoAb were reacted with the biotin ester at varying weight/weight ratios. Flow cytometric analysis of the binding of three different preparations to BL6-10 is shown in Figure 2. The positive fluorescence under curve C in Figure 2 B is comparable to the binding of affinity purified MoAb in Figure 2 A. This ratio of 8 to 1 (protein to biotin) was also found to be optimal for maintenance of control MoAb negative fluorescence.



Biotin Conjugated Anti-Met-72 Monoclonal Antibody Specificity

Reactivity of biotin conjugated MoAb against the highly

metastatic B16 melanoma clone, BL6-10, is compared in Figure 3 to the poorly metastatic parental B16-Fl melanoma and to a syngeneic, non-metastatic melanoma, JB/RH (Table 1). There is approximately a five-fold lower level of anti-Met-72 MoAb binding by RIA to the poorly metastatic cell lines. This result documents a retention of specificity in binding to the highly metastatic clone after biotin conjugation.

Fluorescence Activated Cell Sorting Selection of Met-72
Positive Variants Isolated From a
Fresh Experimental Ovarian Metastasis

Experimental metastases from the parent B16-Fl melanoma were generated as described. An ovarian metastasis was noted after 21 days and removed for analysis of Met-72 expression by flow cytometry. Biotin conjugated anti-Met-72 MoAb revealed two distinct populations within the freshly excised ovarian metastasis, one highly Met-72 positive (Figure 4). Melanoma cells sorted from this population and maintained in vitro have retained a high binding profile to anti-Met-72 MoAb (data not shown). Cells obtained from the original ovarian metastasis (0-1) were cultured, tested by RIA for Met-72 expression (Figure 5, selection cycle 1), cycled in vivo to generate experimental metastases and cloned by limiting dilution to obtain 0-1.1 (Figure 5). Clone 0-1.1 from the original ovarian metastasis was tested for binding to anti-Met-72 MoAb (Figure 5, selection cycle 2) and 3 X 105 cells were injected


i.v. in the syngeneic host, resulting in the organ selectivity indicated in Figure 5, selection cycle 2. Subsequent ovarian metastases were isolated, cultured, tested for Met-72 expression and cycled in vivo (Figure 5, 0-1.2, cycle 3 and 01.3, cycle 4). Successive cycles of in vivo passage and culturing were seen to enhance organ selectivity for ovaries (53%, 56%, 87% of the mice having ovarian metastases) and to enrich for Met-72 expression (28, 36, 45 times background) (Figure 5).

Bone Marrow Derived Clones of B16 Melanoma
Express High Levels of Met-72

Experimentally derived bone marrow metastases of B16-Fl melanoma were aspirated and mechanically disaggregated into a single cell suspension. Aspirates from five animals were cloned by limiting dilution into 96 well plates. Growth positive wells were expanded in culture and tested for binding to anti-Met-72 MoAb. Clones of bone marrow metastases from each of the five animals are shown in Figure 6, compared to BL6-10 and the parental B16-Fl cells. As seen with the ovarian metastases, Met-72 expression of clones derived from bone marrow metastases was increased above the parental B16-Fl level of expression, and in some cases approached that of the highly metastatic BL6-10 clone.

ImmunocytoloQy of Fresh Ovarian and Lung Metastases

B16 melanoma metastatic variants were isolated from fresh lung and ovarian experimental metastases by mechanical and enzymatic disaggregation. Single cell suspensions were

processed for immunocytology using anti-Met-72 MoAb. Cytospin preparations of an experimental ovarian metastasis showed positive immunoperoxidase staining (Figure 7 B). There is cytomorphologic heterogeneity in the level of expression of Met-72 within this mixed cell population. Impression smears of experimental and spontaneous lung metastases generated by injection of the poorly metastatic F1 parental B16 melanoma demonstrated similar levels of individual cellular binding to anti-Met-72 MoAb (Figure 7 D, F). Isotype identical MoAb (anti-sheep RBC, N-S.7) showed no staining in these assays (Figure 7 A, C, E).

Localized Distribution of Met-72 Positive Variants
Within ProQressing Subcutaneous Melanoma

Immunohistologic analyses of cryostat sections from s.c. B16-Fl parental and BL6-10 melanoma were performed at various time points during tumor growth. Predetermined optimal concentrations of biotin conjugated MoAb were incubated with serial sections of snap frozen tumor. Reactive sites were detected as red granules upon development with AEC. A common pattern of reactivity has been noted in all sections examined from tumors as early as 3 days to as late as 15 days of growth. A subcutaneous B16-Fl derived tumor of 7 days growth serially sectioned and stained with anti-Met-72 MoAb is shown in Figure 8. Background levels of peroxidase staining are shown in Figure 8 A and B, using an isotype identical biotinylated control MoAb, N-S.7. In contrast, Met-72 positive cells were observed to be localized within regions of


the tumor mass (Figure 8 C, D, F). The most brightly Met-72 positive cells were seen within satellites which appear to advance from the main tumor mass. Morphologically, these cells are variably melanotic. Early observations showed a red reaction product around vascular areas (Figure 8 E). A definitive association of this Met-72 positive staining with melanoma cells has not been reproducible. Notably, no binding was seen in the necrotic areas of the tumor mass.

Met-72 Positive Variants Localized Within B16 Melanoma Metastases

Direct immunofluorescence utilizing FITC-conjugated antiMet-72 MoAb binding on frozen sections of metastases have revealed Met-72 positive variants to be distributed as single cells and micrometastases within colonized organs. These poorly differentiated cells show a predominantly intense cytoplasmic pattern of staining and weaker cell membrane fluorescence (Figure 9 B, E). Diffusely distributed micrometastatic foci of melanotic cells demonstrate a weaker cytoplasmic fluorescent signal (Figure 9 C). Single cells and micrometastases have been observed in lungs colonized with B16-Fl or B16-FlO metastatic variants and in liver colonized with B16-Fl cells (Figure 9 E, F and G).

Specificity Characteristics of Syngeneic Antibodies
Generated Against B16 Melanoma

The kinetics of an IgG response against the poorly

metastatic B16-Fl or JB/RH lines could be measured as early as

8 days post immunization (Table 2), with moderately higher


binding activity at 16 days (Table 3). In most cases, increasing the immunizing dose above 105 cells had little effect on the levels of Ab produced. Antibody responses generated against either of the poorly metastatic lines (B16F1 and JB/RH) cross-reacted with the other, while the highly metastatic clone was consistently negative in provoking an Ab response or in binding the Ab elicited by the other melanoma lines (Tables 2 and 3).

The ease and reproducibility of this anti-B16-Fl response prompted us to capture these reactivities by hybridoma technology, to further characterize the target cells and surface structures. Using established protocols (5,31), several MoAbs were obtained which demonstrated binding patterns in RIA very similar to anti-B16-Fl antisera. The population distribution of reactive cells was further analyzed by FACS. Specifically, anti-B16-Fl Abs were reacted against selected variants of C57BL/6 melanomas. The results of 6 separate FACS experiments are typified by the example presented in Figure 10 A (antisera) and Figure 10 B (MoAb), where Ab binding (green fluorescence) is plotted versus relative cell number. Antibody binding of anti-B16-Fl antisera or MoAb (stippled) was compared with equivalent dilutions of NMS or control N-S.7 MoAb (solid line). The pattern of reactivity against the immunizing population (B16Fl) is shown in Figure 10 A and 10 B, top left panels. A large proportion of the population shows greater fluorescence


than the NMS. In Figure 10 A, bottom left panel, the high lung colonizing line B16-FlO displayed a qualitatively similar binding pattern, the only difference being a much greater proportion of the cells were coincident with the negative fluorescent population than that seen with the poorly metastatic B16-Fl. These differences were not apparent with anti-B16-Fl MoAb (Figure 10 B, bottom left). When our most highly metastatic clone, BL6-10, was subjected to a similar analysis, only marginal reactivity with the anti-B16-F1 Ab was observed (Figure 10 A and 10 B, top right panels). Finally, a recently derived C57BL/6 melanoma, JB/RH, which shows no experimental metastatic activity displayed the highest reactivity with the syngeneic anti-B16-Fl Ab (Figure 10 A and 10 B, bottom right panels).

The second characteristic of Abs generated by these

immunizations was the apparent melanoma specificity. As shown in Figure 11, anti-B16-Fl MoAbs reacted strongly with syngeneic, poorly metastatic melanomas (BI6-FI and JB/RH), moderately with an allogeneic melanoma K1735, and were nonreactive with the other cells types in our tumor panel (fibrosarcoma, lymphomas, carcinoma, mastocytoma) or nontransformed murine fibroblasts or spleen cells.

The third and most important characteristic of anti-B16F1 Ab was the apparent inverse relationship between Ab binding and experimental metastatic potential of the cell lines and clones tested (Figure 11). This was first suggested by the


reciprocal binding pattern of an anti-Bl6-Fl MoAb (2D8.3B1) versus our anti-metastatic, Met-72 MoAbs (5) (Figure 11).

Syngeneic Antibody Response Against a C3H Melanoma

The binding of anti-B16-Fl MoAb against a moderately

metastatic allogeneic melanoma line, K1735 (30,53), prompted us to examine the specificity of the Ab response to this C3H syngeneic melanoma. Experimental metastases of K1735 were generated and immune sera collected at necropsy. Reactivity against the immunizing line or allogeneic melanoma variants is shown in Figure 12 A. Similar to the Ab response against B16Fl, this response syngeneic to a C3H, moderately metastatic melanoma line was seen to bind higher to the allogeneic, poorly metastatic variants. In addition, the reciprocal binding of anti-B16-Fl versus anti-Met-72 MoAb is apparent on the low versus high metastatic variants, respectively (Figure 12 B).

Metastatic Potential of Melanoma Populations
NeQatively Selected By SynQeneic Anti-Melanoma Antibodies

Strategies and methodology using immunomagnetic bead cellular affinity separation were developed to negatively select for the non-reactive cells in the heterogeneous immunizing B16-Fl population. Flow cytometric analyses of B16-Fl cells reacted against control (solid line) or immune Ab (stippled line) are shown in Figure 13. Anti-B16-Fl antisera (Figure 13 A) or MoAb (Figure 13 B) binding is compared to non-reactive populations, obtained after two cycles of immunomagnetic bead separation, restained with anti-B16-Fl


antisera (Figure 13 C, F) or anti-B16-Fl MoAb (Figure 13 D, E). The efficiency of the immunomagnetic bead separation protocol is apparent in panels C, D, and E.

The FACS profiles of B16-Fl and B16-FlO cells before and after application of the immunomagnetic bead separation procedure are shown in Figure 14. Negatively selected cells obtained by anti-B16-Fl depletion were compared with cells similarly fractionated with NMS or control N-S.7 MoAb. A portion of cells from each of these separations was immediately tested for experimental metastatic activity. As shown in Figure 15, the experimental metastatic potential of control Ab and syngeneic anti-B16-Fl Ab fractionated cells was distinctly different. These results are in agreement with the anticipated doubling in metastatic activity since approximately 60% of the melanoma population was removed by specific Ab + immunomagnetic bead separation. Negligible and significant differences in pulmonary colonization were observed between control and anti-B16-Fl Ab treated B16-Fl and B16-FlO populations, respectively (Table 4). In addition, the fractionation procedure was seen to impact significantly on the extent of extrapulmonary metastases, morbidity, and mortality when compared to controls. As seen in Figure 15, there is a clear trend toward more aggressive extrapulmonary colonization with both the B16-Fl and B16-FlO negatively selected populations. In particular, the involvement of lymph nodes was statistically significant, by analysis of variance,


in the cases of Bl6-F0lO cells negatively selected with antisera (p=0.0196) and MoAb (p=0.0103) (Table 4). Biochemical Characterization of Anti-B16-Fl Reactive Species

Western blot analysis comparing antisera and MoAb

reactivity against detergent lysates of the poorly metastatic lines B16-Fl and JB/RH, versus the highly metastatic B16 clone, BL6-10, is shown in Figure 16. The anti-Bl6-Fl antisera immuno-stains two distinct bands of Mr 26,000 and 45,000 and a diffuse region between Mr 69,000 to 200,000 with both of the poorly metastatic melanoma lines, B16-Fl and JB/RH (Figure 16 lane b). Two of the anti-Bl6-Fl MoAb, 2D8.3B1 and 1F11.1E12, identify a corresponding Mr 45,000 band in lysates from the poorly metastatic melanoma lines (Figure 16 lane c and d). A less intensely stained band of Mr 45,000 has been identified using the same anti-Bl6-Fl MoAb by Western blot analysis of B16-F0lO and K1735 detergent lysates (data not shown). In addition, a Mr 50,000 band was stained with the same anti-Bl6-Fl MoAb in the JB/RH lysate. As anticipated from cell binding assays of the syngeneic antisera, the highly metastatic clone BL6-10 did not reveal any specific bands (Figure 16, lanes b, c, d).

In Situ Characterization of Metastatic Variants Within
Experimental Metastases and Subcutaneous Masses of B16 Melanoma

The cellular morphology of metastatic variants expressing Met-72 or anti-Bl6-Fl antigen was evaluated by immunocytochemical analysis of touch preparations of freshly


excised metastases. Impression smears of liver, lymph node and lung metastases derived experimentally from B16-Fl and BL6-10 cells were examined. A lymph node metastasis derived experimentally by injection of BL6-10 is shown in Figure 17. Variably melanotic cells stained most positively with antiMet-72 MoAb (Figure 17 B) with individual cellular heterogeneity evident for anti-Met-72 MoAb and anti-B16-Fl MoAb (Figure 17 C) binding. Most cells exfoliated from BL6-10 derived metastases demonstrated background levels of anti-B16F1 MoAb binding with an occasional (less than 1%) strongly positive cell.

Previously, analyses of cytospin preparations of B16 melanoma ovarian metastases had revealed cellular heterogeneity in Met-72 expression (54). Analysis of frozen sections of experimental metastases revealed similar levels of cytoplasmic and membrane heterogeneity within metastases to the ovary (Figure 18 B) and greater omentum (not shown) when stained using anti-Met-72 MoAb. Anti-B6-Fl MoAb generally showed less intense staining, with an occasional brightly positive cell (less than 1%). Frozen sections of lungs were analyzed using the immunoperoxidase technique for comparison with the earlier direct immunofluorescence studies. Again, Met-72 positive cells were variably melanotic and stained brightly as individual cells or microscopic foci within the parenchyma. The majority of melanotic cells within larger metastatic foci showed weak cytoplasmic staining (Figure 19 C,


D). Interestingly, anti-B16-Fl MoAb revealed strong cytoplasmic and membrane staining within the larger pulmonary metastatic foci (Figure 19 E, F).

To address the effect of the host immune response on the emergence of metastatic variants, B16 melanoma variants were localized within developing s.c. tumors maintained in normal or sublethally irradiated hosts. B16-Fl or BL6-10 cells were injected s.c. into normal or 500 Gy irradiated syngeneic mice and into Balb/c nu/nu mice (2 mice per group). Resulting masses were removed after 12 days, measured in three dimensions to calculate a mean tumor volume (Table 5) and snap frozen. Serial cryostat sections (6 to 12 Am) of each mass from each animal were examined by ABC immunoperoxidase analysis as described. Monoclonal antibody binding was assessed on a scale of negative (lowest) to 5+ (highest) and is summarized in Table 5. Anti-Met-72 MoAb binding was clearly evident within regions of tumors and on variably melanotic cells within normal syngeneic hosts (Figure 20 C and D) as had been noted previously (54). Anti-B16-Fl MoAb binding to serial sections of the same masses clearly showed no reaction within the regions which were most brightly Met-72 positive (Figure 20 E and F). The most striking observation was the increase in the number of Met-72 positive cells and the increase in the level of anti-Met-72 MoAb binding per cell throughout the masses in 500 Gy irradiated syngeneic hosts (Figure 21 C, D and G). This increase was much more apparent


in tumors induced by BL6-10 cells than in those induced by B16-Fl cells. This difference was also more apparent in sublethally irradiated, syngeneic hosts than in allogeneic athymic hosts (Figure 22 C and D). There was a slight increase in the level of anti-Bl6-Fl MoAb binding per cell within tumors induced by BL6-10 cells maintained in irradiated syngeneic or allogeneic athymic hosts (Table 5, Figures 21 and 22, E and F).

Immunohistology of Syngeneic Healing Skin Wounds and Embryos

Surgically created skin wounds of C57BL/6 mice were

examined by ABC immunoperoxidase analysis of serial cryostat sections (6 Am) during early, mid, and late phases of healing. Anti-Met-72 MoAb reactivity was not detected in 24, 48, 72, or 96 hour healing wounds (data not shown).

Embryos of C57BL/6 mice were removed at day 11 and 15 of gestation and snap frozen. Immunohistology of serial cryostat sections (12 Am) were examined from 2 embryos at each time point. Neither anti-Met-72 MoAb nor anti-B16-Fl MoAb (2D8.3B1) binding was detected at either time point during gestation (data not shown).


H 0 0 C-4 0 0 0

H r-4 0 0 0 0 0


lw m V-4 m m C14 0 1!

m 0 U') 0


r, cl r-I LO tD LO 0


a) cj Lo Ln r, r, ui C-4 N co N
+1 +1 +1 +1 +1 +1 +1
LO 0 .4, cn %D co 0 r-i H M r-q tD cn =w f-i V-4 N cv


co cn 00 U') ko U) ko

( 4
00 H r-I 00

0 4.)
9 (o 4-)
0 0) 9 4J
0 4-1 :3 04 M

Lf) Im 0 Ch (44
44 (L) ul
M r.
0 04-) +1 0 N
4-) 4
r. 44 4J
0 co Wra a) r-4 z 4) 91 41 -rq
tQ it 4 4) r-4 4)
9i 0 rq qr C) >4 V r-4 pq
to 4 4-) (0
0 4.) m (o U 9:
4J M r. 4 EO
r. -r-i 4) r-A
(0 4J LO U) r
0 C) 4 tO 4J
9 rc$ 0 (0 0 -H Q) 4-4 g () 0 rq
9 r-q r-4
0 r-4 P It (d 04
M 0 ;: 4)
z 0 p ci r-4
(n r. N 4J N 4) 0 r, -H -H 41 0 -rq 1 10 to
4) 0 -4 M V '0
z 9 r-i 9 0 010 I fli (1) z z
-H r-i r-4 134 1 (n to V 0) r-4 (n -r-I r. 0 4) 0 4-) P rA
(0 u M 9 4)
%D (0 4J 4J
(0 0 44 9
91 --4 ro to 4) rq m 0 4) 0
r- r. r-I FO -4 LO to z (1) r. p
-rq 4 0 (1) 0 9V 0 04 Ord Ol
N 0 tri
-rq -rq r r
C4 44 4) MN -H -,i
9 or-I 9
W 44 -ri W
ON r. (a r-A
4J 0 r-4
CO "Cr t 4 R



.. . . .

...... ....E
.............~~ ............ .......... ........ .........

N 0


... ... ... ..
~ 2


,-x LU

V- x qyoyj Z L-ISYW-flU8 peleqelojpei punoq wudo ueen

Figure 2. Optimum biotin conjugation of anti-Met-72
monoclonal antibody.

Affinity purified anti-Met-72 MoAb binding to BL610 cells analyzed by FACS indirectly using FITC-SAM
(panel A). Biotin was reacted with affinity
purified anti-Met-72 MoAb at three biotin ester to protein ratios (A = 1 to 2.5, B = 1 to 5, C = 1 to 8). The binding of these preparations, indirectly
indicated by FITC-avidin, is shown in panel B.



media affinity purified
anti-Met-72 MoAb



blotin conjugated anti-Met-72 MoAb B


16,4 0 i 41 2 a
0c e
Gre Uoezec n n

Figure 3. Retention of binding specificity of anti-Met72 MoAb after biotin conjugation.

Various amounts of biotin conjugated anti-Met-72 MoAb were reacted with the highly metastatic B16 melanoma clone, BL6/10, the poorly metastatic B16
melanoma parent, B16-Fl, and a recently derived, poorly metastatic C57BL/6 melanoma, JB/RH. The
extent of specific binding was measured after the addition of 125I labeled streptavidin. Values are expressed as the mean of triplicate determinations
of cpm bound S.D.


1316- 10 0
_~ B16-F 1
I' 1 JB/RH 4n

w Ixr10 CO,


0.375 0.75 1.5 3.0

(/Lg per well' per 2X105 cells)

4-) ro -ri
10 (1) >
9: 4.)

>i E-4
it %4

u r m
.rq 0 r-4
4-) 4)
(d M

0 $4
P4 (1) I 4J
> ko 9:
.rq rq 0
41 m 04 0
.H to U
tn 44 :5
0 0 m ro
04 9
rn r-4 Id
Cj 9-4 -rq r-4 r- rT4 M 4) 10 I I tv 0 (L)
4-) tD 4-)

0) t
44 44 0) r.
0 0 -rq
to to tt
.rq -r-4 (0 0
tn tn -,i >1 to m 0
r-4 41 (o 4j C-4 tv 0) > r. rr. IV 0 -4 1
(0 41 4-)
W Oro (1)
u 01 0) 4) z
.H > 41 1
4 -r-I (0 -rA
4J $4 4J
(1) -rq (D 9
0 p 10 0 to
0 to to
41 > >IMV w
>i 0 r-i -,-I (D l
r-4 ro 4-) m (0 0 >4
M 4-) >4 0) "q
0 4) Z r-4 0 (d
r-4 p Q) r-I -m 9
44 44 R (v z (d
i U 0
4 -H 0 tO
4) 4J U

4) W 4J
4 0 p
:% 9 9 -el 0
UN 4).044


0 (D


CY 0 N
2 0 E
cm 0 :3
'D N C4
(D co 03 cl) ;,% ca
CO < .2 *CIO
co 0 w
cl) 0
< z
w w cy 0
w 0 o < V- D
> -0 -J
co LL


(spun OAIIUIOJ) -13NNVHO U3d Nnoo

Figure 5. Clones derived from experimental metastases to
ovaries retain a high expression of Met-72
upon repeated cycling in vivo.

a. in vivo cycle number for the selection of
metastatic variants with ovary specificity.

b. number of mice per group.

c. number of mice with metastases to organ site.

d. Binding of anti-Met-72 MoAb to ovarian
metastatic variants as detected by RIA. Results
are expressed as a binding index, which is
calculated by dividing the mean cpm 125I pA
bound with anti-Met-72 MoAb divided by the mean
cpm 125I pA bound with N-S.7 MoAb (background

e. Other = kidney, mesentery.


a c % of Mice with Metastases Met-72d

Selection Metastases Expression
Group cycle nb' to 20 40 60 80
LungsZ 40

Primary Ovary
Metastasis 1 8 Bone
Marrow // -20
01 Lymph

Lungs 40

Clone 0-1.1 Ovary
of Ovarian 2 1 9 Bone
Metastasis Marrow 20 X
Lymph U

Lungs Z
var /////////////// 40 n Oary

0-1.2 3 23 Bone
Marrow 4077 7Lymph
Node 77= / 20

Lungs 40

0-1.3 4 8
Marrow 20

> v w
.rq Z FA
$4 0 0 9 9
4) (1) 0 0 (d
4 p 0 -H.0
4) 44 H V 4.)
9 I co
0 4) (L) %D 0 to
p (n > 9 4
p r-4 -,A 0 m
to 44
0 W
-H M 4.) 4J >4
r. 0 4) 4) r-4 (d k r. ro 10 4)
0 44 0 (D 0 0
A v r-4 p 0 -r4 v
r. PC 0 (d 4-) 4-) r. 44 ow 04000
0 4-) 4) 0 0 0 10
0 > 0 H k
k rq 'I
rq 04 m
a 0
0 0 0 ra x
r-I -rq
w 44 A
4J M 0 W 0 < 0 0 W-1 4) 0 >
M 4J P r. Z -H (0 0 04 (d 4J
0 4J r-4 X 0 r- 0 -H tO -H 4) 4J
.H to ra U) r.
W p C,4 4) 1 4)
k 0 tPr- r. Z M (D 0 r. 1 V 0)
4 V 44 P
4J 4) W 0 04
0 -ri z (d 4)
CO $4 0 0) k
0 m k .-I p 10 9
A 44 to rq 0 W -4 r4
.r4 44 tO PO 4)
r-f >1 >
0 q 0
H th
0 ro g a
M 4) X 10 44
0 9 >4 4) 9: 0
0 Q :3
0 rn p r-I (L) 0 4)
(1) w u 10 k p 9
w 044) 4) (d tyl 0
0 0 4-) 0 4J 9
r. 4.) Ln lid p V) to () (n 0 m 0 r-I 0 w w 0 -H
0 M H g 4J :3 M
4-) Im r-4 0
4) x >4 M (L) (a : : 0 M r-4 p > 0
m 10 It (1) 41 0 04
Ol -rq
0 C,4 r.
r (d r-4 rA +1 4.)
to 10 r-4 (d
0 to 4) 0
r-4 04 k r4 0
4. o a) r-4 A s;
k 44 1 0
0 44 u %D 0 H
(1) -rq 04





-ME= 75 .........


. . . . . . . . . . .*. =-- - . . . . . . . . . .
.. . .. . . .. . .. . . .. . .. . . .. . .. . .
.. .. .. .. .. .. .
..... .. .. ..
... .. .. ... .. .. ...
.. .. .. ...
... .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. -.
.. .. .. .. .. .. .. .. .. .. .. .. .
.. .. .. .. .. .. .. .. .. .. .. 0.


0 40 0mc 0

CO0L x qyy ZL-Ion-llue punoq wda ueeW

Figure 7. Met-72 positive variants of B16 melanoma
detected in cytospin and impression smear
preparations of metastases.

Cell suspensions from a fresh ovarian metastasis
were stained using biotinylated control N-S.7 MoAb (A) or anti-Met-72 MoAb (B). Impression smears of experimental lung metastases (C, D) or spontaneous
lung metastases (E, F) were stained using biotin conjugated N-S.7 MoAb (C, E) or anti-Met-72 MoAb
(D, F). (A Fx 270).


i!, l,

Figure 8. Localization of Met-72 positive variants
within developing B16 subcutaneous melanoma.

Biotin conjugated N-S.7 MoAb bound to a cryostat
section of a B16-Fl subcutaneous melanoma (A) x 55 and (B) x 135. Biotin conjugated anti-Met-72 MoAb
bound to a serial section of the same B16-Fl s.c.
melanoma x 55 (C), x 135 (D), x 1350 (E), x 550

45 JW

N (V 4.) E-4
.rj 41 :3 H
r-i (A ()

0 4) $4 4J
4) -r-I
R :it C)
to LO
41 LO
r. 0 M 0
m :I x
.r4 I
4) p r. x io
> 4) 0
.r-l > -H 0 FZ4
41 rq 41
r-q 010
4) 9 ()
0 rA En (0
I r-A 4 0 to (o -rl H "I N
41 4) x
4) 10 to 0
z pq
0 9
U 4) U) to 4) 4 1
z m
4) 4) 4 Oro u to rh ol
W 0 4) 4J
4) 4.) ON 10 (o 0 m til r. ON z
0 0 0 w 0
0 4J r-I (0
r-I 4)
4-4 14 0 9 0 1 0 r-I 4) () 4J
r. 0 44 tQ 1 (1)
9 1 ouz
0 %D k E-4 I
9 ri 44 H -H
.rq (0 M 44 4J
r-i ro 9
41 4) H r. .0 (0
() co co 41
4) 4.)
4) 4) 41 0 4J V 0
(v 4) UN
4) 4) -H
X 4) 4J 0 4) N ON W


47 LL ul co


I 0+1

*x 0

g0 +1


.1 I4-iIVIw


> i;Fi4


0 r rl r- m m co 0
f-A C14 m m co %0 m
0% 10 0

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
C19 0 tn m r- tn m cn r, c4 o% m a%
I Ln 0% r- co co V le co r- Lo v
Ln m eq eq lw C*4 2 Ln q,
r, ON C19 00 m r- 0 Ln

ON r-4 N %D r-4 %D in Ole OD c%3
Cj .1 %D cv m w r- H

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
Ln r-q N N cn m 0 %0%0 cri H N 0
Lo qr C14 0 90 m Ln m m w r, 0 0
q:r lqr co m %0 o w Lo Ln co m m lw Lo Lo

0% q Ln w v tn co H cj H
m H OD cn cn H H CN le lw m
v -W N m cli

+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
R41 0 a% 0 Ln r, 0 H r- 0 N r, C14 0
%00 %0 CD M M r-4 m T 0 %o 0
IV Lo H m C14 v Ln in H r, r,
r, r, OD t%- V-4 r-4 H H Lo co

H ri
x N cli lw C14 10


1 +1


4.1 10


.'4 +1
0 44

r, r-4 -rl A4

ri to w

(4-4 so CHO 44 Lo 0
44 In 0

44 j

Ln r-4 -0,



tA %D m %D C4 (n OLA O % 0 H lr r O %0c m N-co Ln N UH co H O0 qw H H0 C-D H
r- HH CV H N N,
+1 +1 11+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
4O NO% H 0 n 0 N NH LAn Itc %0 wH enco m~ HH HH HH w rI GH H1 HN

WLO HN Cq N r H qw Cn Nn '0
N1 LO H m~ rm N

+1+1 +1+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1
H co~ l co 00 % %C1 c
C% LO mmA4N N N

m~ cnlN m H O H NN NC m 0 co

%D +1 +1 +1+ +1+ +1+ +1+ +1+ +1 1 SNC 00-qH LA oiA O H N cn N m %0 O m O H H H H H N m~ H- r H N N1

I- H
ij H_ _Nv1 HH NN 1 q 1

Figure 10. Flow cytometric analysis of anti-Bl6-FI
immune sera (A) and anti-Bl6-Fl MoAb
(2D8.3Bl) (B) binding to C57BL/6 melanomas.

The solid line indicates control Ab binding; the
stippled line indicates anti-Bl6-Fl Ab binding.
The antisera is from the same individual mouse
(challenged i.v. with 3 X 105 B16-Fl cells) which
was used to generate the anti-Bl6-Fl hybridoma, 2D8. Log green fluorescence is on the x axis; relative cell number per channel on the y axis.



450 B16-Fl 45BL6-1O

2 0

~~01 101 1o io oo0~

40B16-FlO 45 J/RH

0 00 12 130 0
12 101 0 2 10 101 10- 1o


450 -450
B16-Fl BL6- 10

0 C3


to0 to1 to0- '1 0 10~- toi t02 c

H 00 -*d

H~ 4.4

0 .

CO 41)4 ) 0 g
44-40 0 (1)M 0 cq Ha4U.0
r- rj- rd 4H1 ImH r. owfl

4) 04-)
4- ) 0-4 -4
mO W1 -r-4 0 (12 H
b~ I o 0) -4 -I 9 H+ 4--4 0-4 C 0 m~ :1 4g t r
04-) .
>1 H to4)0 9 o ru4 o0 AQ I M (0 4J .I 4J 01(0 4-i wi 4-) 0) 0 4 0 4A-P

*r 4 ) 0 W) X Hq : H- 1 jV J' 4) 4) -Hq

o n xi I) X 04 H 4-0 r.U)-H4 0 ) (1)V
o a) .4A ia 0)
(0 W toH~ r- 4).n)4
H CDa)r.H0 (140. 2.U(1 > El 0 4- .1)3 +1 04-)

V '0r 0J 1 (1 ) 0 0-10 M/ to) H 4OH RlVtA 0
X 0) 9 0) 01444 0) 01 wp4rou0p



0 CD cc 0:

O!n o o-) I Ln

r r co
M (N 0 00 (C)

Ln C114


E 0



C4 Vr, U. o I

c c cm
fu a 0


> 0
> m
> E

> CL
0 C

> to


L 0

Ln t r- op rq r- Ln w w 0 0
ro E cn (0
0 0- CO Q) Q)
u U LLo m UJ y Q- CL
0 0 _j
In co
0 m

0) 0


4 r-I r.
4) r-4 m

0 co 4-)
H 0 4) 9 Id
H 0 ON 4J 00
4) -4 014 to .4
0 > r- E-4 0 4J
"A >4 r4
tn 0 *r-4 m a;
rq +I
41 -H
44 44 4) 0
in IV 4J M 1 0 $40 0 M 4-) %D 4) co
0 -rq (d
x -,q to-,q
0 4.) ON M -rq
C: Ln (0 It (0 4-) 04.)
to -I > 9:
H -r to r. 41 CA
W 41 tn
0 $4 0 -14 -,A $4 4)
0 V 4j 0 to r-) r-4
Lo r. tv tv x
m LC) -H 4J C-4 (L) >4
r- 0 to r- ro r-I
H H 4) M 1 4) 4)
4-1 $4 4-) 4-) W 0 r
X 44 0 W 4) -ri
4J (U 4-4 01 Z (n 4J
U) 1 $4 0
9 -,4 r-I 0
.rq H >4 H 0 V r-I $4
(v fu z 0 rz 4) 4
ON PC r. 05 w
M 4J RJ 4) (0 4J
H -rq r-i 0 m C-4 4) 4J M
t3, to V; I .rq 0 >w4jurw om
04 4-) tQ -4 -rj :1 +1 Z (1)
U (L) 4) V r-4 m
W 4J 9 r. to rd r- X $4 H 4) -H > 9 4)
r-I 4) MA 0 U)
>4 4-) 0 $4 0 0 1 0
'0 0 0 4) r-I $4 Q Z R
0 r-i 0- V
4 (0 44 CQ 0 44
.r4 04 0 4-4
qo 10 u 0
w 0)
>4 44 9 4)
$4 It 10 0 -H >
4) 4) 04 to v q
R in to 41
0 9 fo
W lz r-4 0 0 10 41
U - -H r.
M -ri cl) t3l (1) 41 4)
r 0 M to 9 m to tn
>j C14 (1) to 9 :3 4)
-ri 0 4
9 m 04
0 0 z X: ON 0)
040 v 9 m
04 Ea 9 4J Q
F4 r. (1) 0 to


0 0
2 2 T- I -- ---- x

4.0 40


lu U.
ui w LU LLI V) 0 U) ca m 0 3 m
0 0 0



4c -3
uj lu uj w
CO 0 (A Cb U.

0 0 0 0
............... .....
......... .....

0 co C14

C-OLxqVoWJOBJO$14uogeL )1-1;uwpunoqLudouvorI

4) 0
9 r-4 CA $4
ON r-I 4)
to fO M rO 04
0 (1) it
0 r-i 4) 1 $4
04-ri co o 4)
ID4 >
M 4J ri I
0 -4-) -4 r.
(d (1) 41
14 (1) Q 9 9 H
a) $4 0 01 M H
4J tv 4)
44 0 $4 0
id to (1) 0 H 0 0
9 9 _rq N 9 4)
fo 0 0 4-) 1 :3 >
r. rq to %D I PW -r-I
to 41 r-I -1 41
(0 0 M-H -to
4) r-q I U r_4
M :3 4 0 -1 44 4)
0 M 0 04 4J 0 $4
44 0 4 9 ed
4) 04 :3 (0 M $4
to a) 4)
r-I W _r4
to _rq m () _rq x
r" 4 4j (a
4) > 10 r. r. to x
0 -r-4 "A "A (d 0 4-) r-i V : rl 4)
H () 0 9: 4) P4 4 44 to (A it I 4J
$4 >i W
(1) rA 9 ta m 0 _rq M 44 0 _rq 10 1 x
0 M (a V (1) -rq W (0
44 4-) 1 9 > 41 _rq
0 r-I z u 0 0 9 >4
03 0 E-4 (0 (1)
> H r-I () 4)
0 PW 44 m 0) 9 4
M 41 1 9 4) 4-)
(1) W $4 %D 4) -4 0
4) 0 (1) 0
(a 44 1 R p
(C $ r. U) III _rq 10 0M (v 41 w 4) :1 0
u cn r. r-i 9 r-i Ln
.H U (0 r-I -ri 44 M
m 0 0) 4 4) co
4j rq 4-) N 0 V r. 11
41 U 4 0 (1)
0 0 >4 10 0 w 0
0 (1) MA W 4) m m 0
V m H rq ON M
(L) (D ro -14 r. 4 0 t3m x 0 fu to
c ro r-4 >4 04 r-4 3 0
0 M 0) rq
r-4 4) 0 to r. >1 0 FZ4 A 0 _H 10 _rq
0 4J Q)

r_4 %0 M 0 4J Cd
4 9: 04



F 0 Cl


im cg


Figure 14. Flow cytometric analysis of B16 melanoma
before and after immunomagnetic bead
selection and removal of reactive

B16-Fl and B16-FlO0 cells reactive with NMS (solid
line) or anti-Bl6-Fl antisera (stippled line)
before and after immunomagnetic bead depletion were analyzed by FACS after FITC-SAM binding. Log green
fluorescence is on the x axis; relative cell
number per channel (maximum = 350) on the y axis.



816-Fl 816-Fl


10 1/0 10


ON 0 ON FA it >
r. 0 M M 4) (1) 0 -H -H r-4 M 0 0 M 0 r-I r-I
W co tp 0 -rq
0 0 0 $4
9 4-) 0 FA 0 Wro (d
tic r. co 4) (d 44 91 > r.
r-q 0 In 04 u 4) 4) 0 0 r-4 (o 9
4) rq N rq A V 0 f-i -rq
41 0 41 m v m > tr
4) u go 0 4) 0 9 10
%D -4 A -4 V 44 44 H -rq r-q 4)
V-4 04 < 0 w 0 V It 4J
Q) 0 tv >1 1 '10 f-q 9: ed
v z 0 43 41 9 V r-4 0) 4)
H 4) -rq W (0 M 04
0 m 0 0 ro 0
4P 0 P4 0 0 4 41 m 0 m
I 4J > 4) V 0 0
ko Q),4 0 44
H W W E-1 >i 4-) 4)
4) u m 9 .0 w p 4)
--I 1 0 044 4) IV A
4J --I *J 0 0 -'U (d M 4) 0 W V r. H W r. 0 4 W Z
9 C-H 0 0 010 0 u tv
4) UN to 04 p M -ri 0, to 10
> (d 0 4) > q 9
.f-I to p u > -ri 0 r-i 0 go
V 0 0 M V tA 4) rd 04 to M
(v r 04X 0 Mumom
ON 0 w rq m m -ri (04 m
r-4 (d V 0 4J M 0 W W
0 0 0 0 w g x 0 tA
z .0-1-1 0 4) 0 0 0 -H
44 N E-4 M 0 :1 9 4J 4.)
0 W r-4 44 r. 0 (0 r,
0 04 11 (d 0 A 0 go
>1 41 cn :3 0 4)
4J 44 1 9 LO -H X r, Q)
.rq tv z w w r. 4) V W 14
> 04 4j it (d ra 4) X fo
.rq tP r--4 10 0 9 44 -H 0) W
4J 9 0 01-4 0 m 0 -H 0 m ro
u -H M r-I 0 w 0 IV r. 4)
(v r. 41 r-4 W > > 4) 0, > -ri r.
-H Z 0 >I.H 0 9.0 0 0 -P
u 0 0 0 0 tT W -H 4J M r-4
.rq 0 ri rd r. M UN
V W Lo w 41 V w w -P 9
co p -0 00 tv W -ri r r w 0 -ri
V 0 P rO 4J W
m 0 0 0
co H (D x v r-I r-4 r. r 4)
V r-A w 4) tv tv .0 4J
0) 4) -H M -14 0 11 UN (0 .4 V r.
0: U 4J W rj M r. 04 4)
Z 1340 9-4 0 9 0 r-4 14
r-q 0 it 0 cu- rq 0) tv -rq
fu H > m v tv M r-4 & m
V WH 04 >4 4) 4J 0 4)
9 1 W w u to >A to 4) 04
4) ko 1 (0 04 r-I 0 U 4) x
0 ko (D 0 0 -4 rd 0 4)
r-A M > 0 44 0 4J
w 4J 0 0 0 -P -H r-I V
r-I -rq w -rq 0 u w 9
04 r-4 r. r 0 04 W 0 W
X 4) M V 4J :3 04 "1 V
9: 9: tO -H 4J U 0 M 4) r
rq t 0 to m w 43 4)
it 4) ON (d 0 :3 04
Lo 1 4) Oro 4) 0 0
W M M -H (0 V M M $4 ON V V H I w c 0-H 0 0 0 m 0 0
r-4 tA 44 $2,44 0 IV -H


..... .. ...

.. .. . .. .. . *. .. . .. . 0

L 0
z .4 .. .. .. .. .. dA

...... *(U***


................ -- --
.......... ...... .. .. .. .. .. .. .. .. .
.. . . . .. . . . .. . . . .. . . . .
..... .... .... .... ....

peol .iowni olelselow Aieuowindejlx, 50 x~pul


zc IL


soinpou Bunt, uee.y




00 0 0 0

00 w 0 0

0 0 0


r-4 %D~


B16-F1 JB/RH BL6-10

a bc d e f ab cdef abcdef


9 2 .5 .. .....



30 .


Figure 16. Western blot analysis of anti-Bl6-Fl immune sera
and MoAb binding to lysates of syngeneic melanoma
metastatic variant cell lines.

A 10% polyacrylamide gel was western blotted as described.
Apparent molecular weights of marker proteins (Mr X 10-3)
are indicated on the abcissa.

a. Normal mouse serum.

b. Anti-Bl6-Fl antisera.

c. Anti-Bl6-F1 MoAb, 1F11.1E12.

d. Anti-Bl6-Fl MoAb, 2D8.3B1.

e. Anti-Bl6-Fl MoAb, 2H5.1B1.

f. Anti-sheep RBC MoAb, N-S.7.

Figure 17. Immunocytology of metastatic variants
exfoliated from fresh experimental lymph node
metastases of BL6-10.

Impression smears of expermental lymph node
metastases of BL6-10 were processed by the ABC
immunoperoxidase method as described. The level of
staining generated by N-S.7 MoAB (A) is compared
with anti-Met-72 MoAb (B) and anti-Bl6-Fl MoAb,
2D8.3B1 (C) (A, B and C x 1350).


0 C-4

H 4

0 0
to 44


4J 4)
W U)

to Ea r
.H It -r-f
p 4.) rd
M 4) r.
> 14-1-10
A In
.rq H
41 4J r,
to r. X
4J co
to 1
M z
0 04 d 10
x 9
4-) 0
(1) 0
4-4 M 0 tp
to 4)
(D 0 0 tV
0 zi co
4-) ON
ko >4 N r.
r-4 r-I -H

44 M M -H
0 tic r. --1 4 ro 0 p
-rq -f-4 () 0 x 4-1 W
.rq 0 u (1) 0
V P (1) PO Z
tio (L) m N 04 M N
.H 0 V M rH r (t% I
M 0 V '0 4J
u tri W (D
0 0 r. z
p 0
U (d 4J co x r.
r-q (1) As
4 V
ON co r 3:

69 21.

Figure 19. Localization of B16 melanoma variants within
experimentally colonized lungs.

Mice were injected i.v. with 3 X105 B16-Fl cells per 0.2 ml PBS; colonized lungs were removed and
snap frozen at necropsy 21 days later. Serial
cryostat sections (12 Am) were analyzed by the ABC
immunoperoxidase technique as described. Control
(N-S.7) binding is shown in A and B; anti-Met-72 MoAb binding in C, D and G; and anti-B16-Fl MoAb
(2D8.3B1) binding in E and F. (A, C and E x 135; B,
D, F and G x 1350).







+ + + I + +4

1 44

0+ +

+ + + +4

4-i- 0 0O CV H- CD 4-4 $)
H %D H mCi N N 06 4-)

+1 +1 +1 +1 +1 +1 +1 0 >

Hi Li H

0 4

0 0

A ~Lo -Lon_ j~


I _0


Figure 20. Localization of B16 melanoma variants within
subcutaneous tumors maintained in normal
syngeneic hosts.

A s.c. mass induced by injection of 2 X 106 BL6-10
cells per 0.1 ml PBS into a normal C57BL/6 mouse
was excised after 12 days and snap frozen. Serial cryostat sections (12 Am) were examined by the ABC
immunoperoxidase technique as described. Control
N-S.7 MoAb binding is shown in A and B; anti-Met-72 MoAb in C and D; and anti-Bl6-Fl MoAb (2D8.3Bl) in
E and F. (A, C and E x 135; B, D and F x 1350).


v iv a LI.'

;W -low F

L ~

Figure 21. Localization of B16 melanoma variants within
subcutaneous tumors maintained in 500 Gy
irradiated syngeneic hosts.

A C57BL/6 mouse was exposed to 500 Gy irradiation and injected with 2 X 106 BL6-10 cells per 0.1 ml
PBS on the same day. A s.c. mass was excised after
12 days, snap frozen and serial cryostat sections
(12 gm) were examined by ABC immunoperoxidase
analysis as described. Control N-S.7 MoAb binding
is shown in A and B; anti-Met-72 MoAb in C, D and G; and anti-Bl6-F1 MoAb (2D8.3B1) in E and F. (A,
C and E x 135; B, D, F and G x 1350).



IF$L D Imp






Figure 22. Localization of B16 melanoma variants within
subcutaneous tumors maintained in allogeneic
athymic hosts.

A s.c. mass induced by injection of 2 X106 BL6-10
cells per 0.1 ml PBS into a Balb/c nu/nu mouse was
excised after 12 days and snap frozen. Serial
cryostat sections (12 Am) were examined by the ABC
immunoperoxidase technique as described. Control
N-S.7 MoAb binding is shown in A and B; anti-Met-72 MoAb in C and D; and anti-Bl6-Fl MoAb (2D8.3B1) in
E and F. (A, C and E x 135; B, D and F x 1350).



k4M lk IV

I vi.




These studies have focused on 1) the localization of B16 melanoma metastatic variants within developing s.c. tumors and metastases and 2) the effect of the tumor bearer's humoral immune response on the regulation of the levels of metastatic variants within developing B16 melanomas.

The significant observations were first, that Met-72

expression is discretely localized in situ in the primary B16 melanoma mass. Met-72 positive variants are found to be variably melanotic cells regionally located within developing s.c. tumors grown in normal syngeneic hosts. Secondly, Met-72 expression is enriched on experimentally derived metastases. As increased ovarian colonizing potential is generated, Met-72 expression of melanoma cells derived from metastases to ovaries is enriched. There is individual cellular heterogeneity in the level of expression of Met-72, as visualized by immunocytology of metastases, with small foci of cells being the most positive within colonized organs. Thirdly, melanoma lines with poor experimental metastatic activity consistently elicit strong syngeneic Ab responses, while highly metastatic clonal variants do not. This Ab response can be used to subdivide metastatic from nonmetastatic cells present within the immunizing population.



Finally, the Ab response to poorly metastatic variants detects uniformly distributed variants within larger metastatic foci in experimentally colonized lungs.

The existence of subpopulations of cells exhibiting a

range of metastatic potential within heterogeneous tumors has been substantiated in a number of systems (3,55). In the original studies leading to the present work, a strong correlation between the quantitative expression of a Mr 72,000 glycoprotein (Met-72) and experimental metastatic activity of over 30 in vitro grown B16 melanoma clones was demonstrated (5,11). Flow cytometric analysis and cell sorting procedures using anti-Met-72 MoAb have directly shown that high levels of Met-72 expression is characteristic of cells with a high experimental metastatic potential (12). Our current studies were designed to determine the potential utility of anti-Met72 MoAb to visualize and localize Met-72 positive metastatic variants within progressing and metastatic B16 melanoma masses.

Results of the experiments reported here greatly expand our previous knowledge of Met-72 antigen expression and its correlation with metastatic potential in vivo. Localization of its expression in primary s.c. tumors is notably discrete, and not randomly distributed throughout the developing tumor mass (Figures 8 and 20) in immunologically normal syngeneic hosts. Immunohistologic examination of progressing tumors, excised at sequential times during s.c. growth, shows a

recurrent pattern of localization of Met-72 positive variants. These variably melanotic cells are seen within regions of the masses, the most brightly positive cells within smaller satellites. The implications of this selective localization are twofold. First, induction of Met-72 positive variants may be controlled by microenvironmental factors. Secondly, even if random somatic mutational events yield single metastatic variant cells within a solid tumor mass, their outgrowth into colonies may be directed by chemotactic factors which are microenvironmentally determined. These observations suggest that microenvironmental influences may function regionally to influence metastatic potential. Our observations in the B16 melanoma model of metastasis are consistent with those of Gabbert, et al. (56). They suggest that morphologic transitions at the invading front of rat malignant carcinoma may signify a localized process of tumor de-differentiation. Tumor cell locomotion may be specifically enhanced in regions observed to have a loss of basement membrane and decreased numbers of desmosomes between tumor cells. Clearly, the interactions in tumor microregions greatly influence tumor heterogeneity, as recently reviewed by Sutherland (57).

An important aspect of these studies focused on isolation and visualization of Met-72 positive variants within metastatic foci of B16 melanoma. The sophisticated capabilities of the fluorescence activated cell sorter have provided evidence that Met-72 expression may be a common


surface phenotype of B16 melanoma metastatic variants, irrespective of their organ colonization after i.v. inoculation. Experimentally induced ovarian metastases were directly shown to express Met-72 (54), as has been reported for experimental lung metastases of the B16 melanoma (12). As previously demonstrated with lung colonizing melanoma cells (58,59), ovarian colonizing variants were selected by repeated in vivo cycling. These variants showed increased levels of Met-72 positivity (Figure 5). Clones of bone marrow derived metastases also demonstrated increased levels of Met-72 expression (Figure 6). Clones derived from B16-Fl metastases to the liver, heart, lymph node and stomach wall have all demonstrated increased Met-72 antigen expression which is maintained after repeated in vitro passage.

Rapid visualization of individual cellular expression of Met-72 was achieved by two immunocytologic methodologies: 1) cytospin preparations of experimental ovarian metastases and 2) impression smears of experimental and spontaneous lung metastases. The ease of impression smear immunocytochemistry, especially, has permitted rapid characterization of the surface phenotype of cells dislodged from colonized lungs. Primary melanomas and other well encapsulated masses do not present suitable specimens for impression smears.

The significant findings of these studies are that (1) Met-72 expression is discretely localized in situ in the primary B16 melanoma mass, and (2) cell surface expression of


Met-72 may prove to be a generalized phenomenon of melanoma metastatic variant populations, regardless of organ selectivity. The study of metastatic tumor cell evolution influenced by 1) interactions with tumor derived elements and 2) host selection pressures during tumor progression has eluded tumor biologists. The ability to isolate metastatic variant cells from primary tumor tissue may enable us to quantitate their presence and evaluate their differences during tumor progression and metastatic outgrowth.

With few exceptions, humoral immunity towards syngeneic tumors has been generally regarded as beneficial for host survival (4,6,55,57,60). This belief, and the availability of therapeutic quantities of "tumor reactive" MoAb has resulted in a number of clinical trials for the treatment of human cancer.

The notion that poorly immunogenic variants exist within tumors has been supported primarily by the isolation of subpopulations of tumor cells which are resistant to cellular effector systems in vitro (61-64). A major question left unanswered by these studies is the extent to which humoral immunity may contribute to the changing proportion of metastatic variants within tumor populations. Previous studies have not attempted to correlate syngeneic Ab responses against B16 melanoma-associated antigens with spontaneous or experimental metastatic activity of melanoma subpopulations (14-16,20,65,66). Some studies have indirectly correlated


syngeneic anti-B16 melanoma responses with metastatic activity (13,17,18) or have found no correlation between metastatic potential of clones derived from B16 melanoma and syngeneic Ab responses induced by s.c. immunization with those clones (19). McDonald, et al. (67) reported syngeneic Ab detected in sera of mice challenged with the highly invasive variant of B16 melanoma, B16-BL6. However the subpopulation and functional specificity of these Ab was not determined.

The ease and frequency with which host anti-melanoma

humoral immunity can be demonstrated in both experimental and clinical situations has prompted us to examine whether Ab responses in these situations may in fact be targeted against functionally distinct subsets of melanoma. While studies directly correlating metastatic activity and immunogenicity of melanoma subpopulations have not been described previously, several lines of investigation have suggested the possibility that poorly metastatic subpopulations are the primary Ab targets. First, syngeneic and autochthonous Ab responses against melanoma are frequently observed (13-20,65-67). However, detection of syngeneic Ab reagents defining highly metastatic melanoma variants often require manipulation of host immunity to override these primary responses (5). Second, Ab against human melanoma, obtained by xenogeneic immunization, are most reactive with the more differentiated, benign melanoma forms (68-71). Recently, Herd reported the production and characteristics of MoAb from syngeneic mice

bearing B16 melanoma masses (72). Some of the MoAb were reported to enhance or inhibit B16-Fl lung colonization and at the same time to decrease lung colonization of the high lung colonizing variant, B16-FlO. Recent studies by others within our laboratory now suggest that in addition to the primary host responses against subpopulations of poorly metastatic B16 melanoma, the isotype of these antibodies has a dramatic biologic consequence (Kimura et al., unpublished results). Mice challenged i.v. with syngeneic B16-Fl or BL6-10 cells after i.p. treatment with the IgG3 isotype switch variant of anti-Met-72 MoAb were observed to have increased numbers of lung nodules.

While it is true that syngeneic anti-melanoma MoAbs which interfere with experimental metastasis can be isolated (5,65,72), these Ab responses are rare. In these cases multiple immunizations, with or without adjuvant, coupled to highly specific screening procedures are needed to isolate these reactivities. The major finding to emerge from our study is that melanoma lines with poor metastatic activity consistently elicit strong syngeneic Ab responses, while highly metastatic clonal variants do not. These antinonmetastatic Ab activities are readily demonstrated by analyzing the serum from melanoma immunized mice. Poorly metastatic melanoma lines such as B16-Fl and JB/RH elicit Ab responses which can be used to further subdivide metastatic from non-metastatic cells present within the immunizing

population. This conclusion is based on 1) Ab binding to well characterized melanoma lines and clones shown by RIA, FACS, and western blot analysis, and 2) experimental metastatic capabilities of negatively selected cells.

The results of our immunomagnetic bead depletion assays

show a total increase in the extent of experimental metastases generated after the removal of anti-B16-Fl reactive subpopulations. This increase is subtle with the heterogeneous parental B16-Fl line, but significant with the B16-FlO line. It is not unexpected that a subpopulation of anti-B16-Fl antisera reactive cells remains after removal of the anti-B16-Fl MoAb reactive cells (Figure 13 F). Our results support the suggestion that B16-FlO was selected for increased lung colonization potential, but that organ selectivity is not absolutely equivalent to increased metastatic potential. B16-Fl and B16-FlO are heterogeneous lines with respect to many phenotypes (6). Metastases generated from B16-Fl and B16-FlO may maintain heterogeneity with regard to organ colonization potential. Our findings are consistent with the interpretation that non-metastatic cells present within the parental B16-Fl line (and at lower levels in the selected B16-FlO line) elicit a strong and selective Ab response; the negative binding cells representing the more highly metastatic variants. Removal of the poorly metastatic variants from the total population allows the extrapulmonary colonization potential to become more evident. This is more

so with B16-FlO, because on a cell to cell basis this heterogeneous population contains more cells with a more efficient metastatic potential.

The mechanism of this extrapulmonary colonization may be suggested by the results of experiments by Nakajima, et al.

(73). Metastatic variants of B16 melanoma, including B16-Fl and B16-FlO, were shown to differ from each other in their ability to degrade heparan sulfate which may be related to their ability to adhere and grow in specific organs (73). Specificity of adhesion between murine tumor cells and microvascular endothelium from different organs has been demonstrated by Auerbach et al. (74). Thus, variants remaining after depletion of anti-B16-Fl reactive subpopulations may mainfest their adhesive phenotypes more efficiently.

The influence of host immunity on the induction or dissemination of metastatic variants appears variable depending upon which of the heterogeneous phenotypes required for tumor progression one is attempting to study (1-4). The studies reported here strongly suggest the relevance of comparing metastatic variant clonal lines with heterogeneous lines. Caution must be exercised when interpreting results obtained using heterogeneous lines which appear homogeneous with regard to one phenotype, such as B16-FlO cells.

Anti-B16-Fl antisera and MoAb have recently been used to immunoprecipitate from detergent lysates of melanoma cells


biosynthetically labeled with [3H]-leucine, bands which correspond to those identified by western blot analyses (Mr 32,000, 45,000 and 55,000 on 5 to 15% or 10% acrylamide slab gels under reducing conditions, Parratto & Kimura, unpublished results). Molecular weight characteristics of anti-B16-Fl reactive species bear some resemblance to those reported by others (14-16,20,75-79). However, surface expression of those markers versus metastatic behavior of the cells was not determined in those studies. The Mr 45,000 band identified by the anti-B16-Fl antisera or corresponding MoAb cannot be against normal Class I antigens of the H-2 complex, as there is no measurable binding to C57BL/6 spleen cells or other syngeneic tumor cell lines.

Additional immunohistologic analyses of cells isolated from and localized within developing B16 melanoma metastases revealed several interesting binding profiles. First, impression smears revealed an increased level of anti-Met-72 MoAb binding to individual cells exfoliated from the parietal lung surface compared to cells exfoliated from cut surfaces of well encapsulated extrapulmonary metastases. On the other hand, anti-B16-Fl MoAb binding was weak to cells exfoliated from these same metastases. In addition, cryostat sections of colonized organs showed intensely Met-72 positive variants to be variably melanotic individual cells or cells within small metastatic foci. However, anti-B16-Fl MoAb showed a diffuse pattern of reactivity with melanotic cells within larger B16-