Regulated Expression of Vascular Cell - Europe PMC

American Journal of Pathology, Vol. 141, No. 6, December 1992 Copyright ) American Association of Pathologists

Regulated Expression of Vascular Cell Adhesion Molecule-1 in Human Malignant Melanoma

Nives Jonjic,* Ines Martin-Padura,* Teresa Pollicino,* Sergio Bernasconi,* Petr Jilek,* Aldo Bigotti,t Roberta Mortarini,4 Andrea Anichini, Giorgio Parmiani, Francesco Colotta,* Elisabetta Dejana,* Alberto Mantovani,* and Pier Giorgio Natalit From the Istituto di Ricerche Farmacologiche Mario Negri,* Milan, Istituto Regina Elena, t Rome, and the Division of Experimental Oncology, Istituto Nazionale Tumori, Milan, Italy

Expression of the endothelial adhesion molecule VCAM-1 was studied in human malignant melanoma lines by flow cytometry. Clones 2/4 and 2/14 (derived from the same lesion) had appreciable levels of VCAM-1 expression, whereas clone 2/21 and the lines A2058, Mel24, and A3 75 were negative. Clone 2/14 was selected for further analysis. Exposure to tumor necrosis factor (TNF) markedly augmented VCAM-I on melanoma cells. Surface VCAM-I was associated with expression of specific transcripts that were augmented by TNF. Analysis by reverse transcriptase and polymerase chain reaction using appropriate primers revealed that TNF-stimulated melanoma cells expressed both 7 and 6 immunoglobulin domain transcripts with predominance of the longer species. Tumor necrosis factor-stimulated melanoma cells bound more VLA-4-expressing cells (melanoma and monocytes) than resting tumor cells and anti-VCAM-1 monoclonal antibodies significantly inhibited binding, thus suggesting that surface VCAM-1 on melanoma is functional. Analysis of melanoma tissue sections demonstrated that VCAM-I is not a marker of transformation of melanocytes because it can be detected in benign nevi. Although, unlike ICAM- 1, VCAM- 1 is not correlated with tumor progression, its expression in a fraction of primary melanomas indicates that it may play a role in regulating host immune response and homotypic interactions in some malignant melanomas. (AmJPathol 1992, 141:1323-1330)

Vascular cell adhesion molecule (VCAM-1, also referred to as INCAM-1 10) is a member of the immunoglobulin gene superfamily originally identified in vascular endothelial cells.'1 Vascular cell adhesion molecule is expressed at low levels in unstimulated endothelial cells and is induced by exposure to inflammatory stimuli, including interleukin 1 (IL-1) and tumor necrosis factor (TNF). The counter-receptor for VCAM-1 is the ,B1 integrin VLA-4 (Oc4p1).4 The VCAM-1NLA-4 adhesion pathway is important in the cell-cell interaction of endothelial cells with lymphocytes,5 monocytes,6 natural killer (NK) cells,7 eosinophils, basophils,8 9 and melanoma cells.10 11 Although in vitro only endothelial cells have been shown to express VCAM-1, there is evidence that certain nonvascular cells have VCAM-1 in vivo.12 In particular, follicular dendritic cells of the germinal centers of lymph nodes are VCAM-1 positive, and this structure is important for the localization of B cells.13'14 Other nonvascular cells expressing VCAM-1 in vivo include certain mononuclear phagocytes of spleen and liver and certain kidney epithelial cells.12 Here we report expression of VCAM-1 by some melanomas in vitro and in vivo.

Materials and Methods Cell Culture and Reagents Human endothelial cells were isolated from umbilical veins and cultured as previously described.1' The following reagents were used for culture of cell lines, separation of effector cells and adhesion assay: pyrogen-free saline for clinical use (S.A.L.F., Bergamo, Italy); pyrogen-free distilled water (S.A.L.F.); medium RPMI 1640 (10x concentrated (Biochrom KG, Berlin, Federal Republic of Germany), and Dulbecco's minimum Supported by PF-CNR-ACRO, Rome, Italy and by a contribution from the Italian Association for Cancer Research. Nives Jonjic was supported by a fellowship of Scientific Community of Croatia. Inbs Martin-Padura was supported by a fellowship of Science Plan of ECC. Accepted for publication May 14, 1992. Address reprint requests to Dr. Alberto Mantovani, Istituto di Ricerche Farmacologiche Mario Negri, Via Eritrea 62, 20157 Milano, Italy.

1323

1324

Jonjic et al

AJP December 1992, Vol. 141, No. 6

essential medium (DMEM) (Gibco, Paisley, Renfewship, United Kingdom); glutamine (Gibco); penicillin and streptomycin for clinical use (Farmitalia, Milan, Italy); gentamicin (Gibco); aseptically collected fetal calf serum (FCS) (Hyclone Lab., Logan, United Kingdom). The routinely employed tissue culture medium was RPMI 1640 with 2 mmol/l glutamine, 50 ,ug/ml gentamicin, 10% or 15% FCS, hereafter referred to as complete medium. All reagents were found negative for endotoxin contamination.

Cytokines and Antibodies Recombinant human TNF (specific activity 8.1 x 106 U/ml) was a gift from BASF/Knoll (Federal Republic of Germany). Monoclonal antibodies (MAb) used in this study were obtained through the courtesy of the following persons: MAb anti VLA-4, clone HP2/1 (IgG1), Dr. F. Sanchez-Madrid (Hospital de la Princesa, Madrid, Spain); MAbs antiVCAM-1, clone 4B9 (IgGl), Dr. J. Harlan (University of Washington, Seattle, WA); and BB16V10 (British Biotechnology, Oxford, United Kingdom). In each experiment, appropriate control irrelevant antibodies were used.

Tumor Cells Tumor lines used in this study included: from a human subcutaneous metastasis (Me 665/2),15 clones 2/4, 2/14, 2/21, melanoma lines A2058, Mel24, and A375 from American-Type Culture Collection (ATCC). Melanoma clones (Me 665/2) were cultured in complete RPMI 164015% FCS medium, and other lines in DMEM-10% FCS. All lines were checked routinely for mycoplasma contam-

ination.

Cytofluorimetric Analysis Phenotyping of tumor cells was performed by indirect immunofluorescence. Briefly, cells were exposed to MAb (5 p.g/ml in saline with 2% human serum) specific for different adhesion structures for 30 minutes at 40C, washed, and incubated with fluoresceinated affinitypurified goat anti-mouse IgG F(ab') 2 (Technogenetics, Turin, Italy) for 30 minutes at 40C. The cells were washed and fixed with phosphate-buffered saline (PBS) containg 1% paraformaldehyde. Fluorescence was measured on a Becton Dickinson FACSstar plus.

Northern Blot Analysis was performed according to standard procedures.16 Briefly, total RNA from melanoma cells was ex-

tracted by guanidine isothiocyanate.17 Ten to fifteen micrograms total RNA were loaded on agaroseformaldehyde gel in the presence of ethidium bromide and transferred to Gene Screen Plus membranes (New England Nuclear, Boston, MA). A 2.2-kb VCAM-1 cDNA fragment (a gift of Dr. L. Osborn, Biogen Inc., Cambridge, MA) was labeled with random hexamers, and aX-32PdCTP (5000Ci/mmol; Amersham, Buckingamshire, UK). Hybridization was in 50% formamide (Merck, Rahway, NJ) and 10% dextran sulfate (Sigma, St. Louis, MO). Membranes were washed twice with 2 x SSC (1 x SSC: 0.15 molA sodium chloride, 0.015 molA sodium citrate) and 1% sodium dodecyl sulfate (SDS) (Merck) _ 2x SSC at 600C per 30 minutes, and finally twice with 0.1 x SSC at room temperature. Membranes were exposed per 12 to 24 hours at -800C with intensifying screens. RNA loading and transfer were checked by UV examination of filters.

Reverse Transcriptase-Polymerase Chain Reaction (RT-PCR) One microgram total RNA was reverse transcribed and then amplified by polymerase chain reaction (Perkin Elmer Cetus, Norwalk, CT). RNA was mixed with 2.5 mmol/l random hexamers in 5 mmol/l MgCI2, 25 mmol/l KCI, 10 mmol/l TRIS-HCI pH 8.3, 1 mmol/l each dNTP and 2.5 U/,l Moloney Murine Leukemia Virus reverse transcriptase in a final volume of 20 ,lI. Reverse transcription was carried out at 42°C for 15 minutes. This reaction then was amplified by adding in a final volume of 100 IlI 2 mmol/l MgCI2, 50 mmol/l KCI, 10 mmol/l TRIS-HCI pH 8.3, 2.5 U Taq DNA Polymerase, and 0.15 mol/l each specific primers. Primers were as described. Oligomer 370-y is 5'-GGAACCTTGCAGCTTACAGTGACAGAGCTCCC-3' and oligomer VC-16 is 5' CAAGTCTACATATCACCCAAG-3'. Samples were amplified in a Perkin-Elmer Cetus thermal cycler using 30 cycles at 950C (1'), 550C (2'), 720C (3'). Ten microliters of each RT-PCR then was electrophoresed through a 1.2% agarose gel, which was stained, photographed, and blotted onto Zeta Probe Blotting membranes. These then were hybridized to the VCAM-1 probe described above according to standard procedures for Southern

blotting."6

Tissue and Cell Specimens Surgical biopsies of benign (n = 26), malignant primary (n = 26), and metastatic (31 cases) melanocytic lesions were obtained from the Surgical Pathology Section of the Regina Elena Cancer Institute, Roma, Italy. Tissue samples were immediately snap-frozen in liquid nitrogen.

VCAM-1 in Melanoma

1325


From each specimen, 4-,. cryostat sections were obtained that were fixed in absolute acetone for 10 minutes. Fixed sections were either immediately used in immunohistochemical assays or kept frozen at - 30°C with no loss of serologic activity. Fixed sections stained with 1% toluidine blue were used to evaluate the histologic features of the lesions. Histologic diagnosis was evaluated according to Breslow.i8 Culture of normal human melanocytes were purchased from Clonetics (San Diego, CA). Acetone-fixed cytospins were prepared using a Shandon (Rouncorn, Cheshire, United Kingdom) cytocentrifuge.

Immunohistochemical Studies The immunoperoxidase stain was performed with commercially available reagents (Immunocolor, Sorin Biomedica, Saluggia, Italy). Slides were incubated overnight with MAb at 4°C in a moist chamber. The enzymatic activity was developed using 3-amino-9 ethylcarbazole (AEC) as chromogenic substrate for 8 minutes. Slides then were rinsed with PBS and counterstained with Mayer's hematoxylin.

Monocytes Monocytes were obtained from the Ficoll-Hypaque separated mononuclear cells by centrifugation on a discontinuous (46%) gradient of isoosmotic (285 mOsm) Percoll (Pharmacia Fine Chemicals, Uppsala, Sweden).19

Adhesion Assay Tumor cells were grown to confluence in flat-bottomed 96-well trays (Falcon 3072, Becton Dickinson, Lincoln Park, NJ). Tumor monolayer was activated with 500 U/ml of human recombinant TNF for 24 hours. Monoclonal antibodies directed to adhesive receptors were preincubated with tumor monolayer for 30 minutes at 37C before addition of cell suspension and then allowed to stay during the adhesion assay. Adhesion of monocytes to tumor cells was studied as described previously.20 Monocytes (1 07/mI in complete medium) were incubated at 370C for 1 hour with 100 mCi 51Cr (sodium chromate, Amersham, UK). At the end of incubation, the cells were washed three times with medium and were resuspended at 106/ml. 5 x 104 monocytes in 0.2 ml medium were added to tumor monolayers and incubated for 30 minutes at 370C. Adhesion of tumor cells has been previously described in detail.11 1251lododeoxyuridine (1 mCi/mI for 18 hours, Amersham International, United Kingdom) labeled melanoma cells were detached, washed twice, and fi-

nally resuspended in RPMI-1 0% FCS at 6 x 1 05 cells/ml. Then an aliquot (50 ml) of radiolabeled tumor cell suspension was added to each well and incubated for 30 minutes at 370C. Nonadherent cells then were removed by washing the plates three times with PBS-2% FCS. The content of each well was solubilized with 100 ml of 0.025 mol/l NaOH - 1% SDS and counted in a gamma counter (Beckman, Fullerton, CA). Results are presented as mean (± standard deviation) percentage of adherent cells, with six replicates/group.

Results Figure 1 shows a representative flow cytometric analysis of expression of VCAM-1 in the 2/14 melanoma clone. Endothelial cells were used as reference VCAM-1expressing population. It can be seen that a fraction of 2/14 cells (10%) were VCAM-1 positive in the absence of stimulation. In a series of four experiments, the percentage of VCAM-1-positive cells in clone 2/14 was from 6% to 43% (23.5 ± 15.8%). On exposure to TNF (500 U/ml for 8 or 24 hours), a substantial increase of VCAM-1 expression was observed, both in terms of per cent positive (10% versus 62% in Figure 1), and fluorescence intensity (mean channel of fluorescence 320 versus 575, Figure 1). Table 1 summarizes results obtained with a limited series of melanoma clonesAines. Among three clones derived from the same melanoma lesion (Me665/2), only two of 14 and two of four showed some constitutive expression of VCAM-1 and on exposure to TNF this expression augmented, especially in clone 2/14. The clone 2/21, and the A2058, Mel24, and A375 lines were almost VCAM-1 negative under these conditions. After initial submission of this manuscript, five additional melanoma lines were examined (B. Azzarone, personal communication). Of these, three (ME1477, GLL19, JUSO) expressed VCAM-1 constitutively, and exposure to TNF or interferon-y augmented or induced it. In an effort to define the molecular basis for VCAM-1 on melanoma, we examined mRNA transcripts by Northern blot analysis and PCR. As shown in Figure 2, the melanoma clones 2/14 and 2/4 expressed appreciable levels of VCAM-1-related mRNA transcripts. In agreement with surface expression, clone 2/21 and A 2058 cells did not have appreciable levels of VCAM-1 mRNA transcripts. Exposure to TNF augmented mRNA levels in clone 2/14 cells. Alternative splicing yields a 7- and 6-Ig domain containing mRNA in endothelial cells.21.22 To identify the forms of VCAM-1 expressed in melanoma cells, we used RT-PCR to amplify fragments of different sizes. Figure 3A shows the ethidium-bromide-stained gel and Figure 3B shows the Southern transfer and hybridization with a

1326

Jonjic et al


A

I

2

3

I

I

I 0 .0

E II

B

1.

I 200

9

I

400

I

---k.

.

k

'I

see

999

id I

1414

I''1999-r

F.

c1

290

400

C999r

9G

-Li-.

2

II..

209

.. ....-r

9O9

agoX

]

logo

3 44

II

I

;_I .. ". 409

1 a>>

0~

.. -1! a

.

.

.VT'lo 299

490

.I

ClOO

I

1

I 1.

I

21D0

I

1

I OGI

_.

!-1 *

'Ia

'I i qI i3q-7311 22 rd

40aa

Eai

c

IB

1.r

X T1 PIr l § r Y XWIML-A.lj X

la

299

409

co9

299

F I u o r e s c e n c e H e i g h t one Figure 1. Expression of VCAM-1 on 2/14 melanoma cells and endothelial cells. Melanoma cells (A) and endothelial cells (B) with 500 U/ml of TNFfor 0 (1), 8 (2), and 24 (3) hours before staining with anti VCAM-1 MAb.

VCAM-1 cDNA probe. As expected, IL-1-stimulated endothelial cells (EC) gave a predominant 'long' 631-bp fragment and a faint 355-bp band. Unstimulated two of 14 cells had only the 631 -bp band. On exposure to TNF for 4 or 24 hours, the intensity of the 631 band increased and a 355-bp fragment became visible. Having defined that at least certain human melanomas have the capacity to express VCAM-1 mRNA and surface protein, we wanted to establish the biologic significance of this adhesion molecule. We therefore studied binding of VLA-4-expressing cells (clone 2/14) on monolayers of resting or TNF-stimulated 2/14 cells. As shown in Figure 4, VLA-4-expressing cells were captured more efficiently by TNF stimulated than by resting 2/14 cells. Similar results were obtained with blood monocytes. Anti-VLA-4 and anti-VCAM-1 MAbs decreased the bindTable 1. Expression of VCAM-1 on Melanoma Clones/Lines Melanoma TNF 8 hr cells Control TNF 24 hr 2/4 10% (370) 43% (574) 33% (560) 2/14 10% (320) 62% (575) 59% (646) 2/21 1% 1% 4% 0% 4% A2058 2% A 375 1% 2% 3% Mel 24 0% 6% 3% Cell surface staining by flow cytometric analysis. Melanoma cells were treated with 500 U/ml TNF for 8 or 24 h at 370C and cell surface staining for VCAM-1 determined by FACS. Results are expressed as % of positive cells with the mean of channel of fluorescence in parenthesis.

were

1999

treated

ing to TNF-stimulated 2/14 cells with substantially greater inhibition with anti-VLA-4 than with anti-VCAM-1 (Figure 4). That ant-VCAM-1 was a less effective inhibitor of binding than antKVLA-4 is hardly surprising. We made C%M

-T-mam

20 24 205 .NF

214

1W

2114

2121 2114 2/21 End. -

TNF TNW IL-1

VCAI-

Figure 2. Northern blot ana{vszs of VCAM-1 expression in melanoma cells. Melanoma cells were either untreated or incubated with TNF (500 U/ml)for 4 hours. Endothelial cells (last lane) were incubated with IL-1 (10 ng/ml) for 4 hours. The lower part of the figure shows the ethidium bromide stained RiVA bloted on thefilter.

VCAM-1 in Melanoma

1327


A

355.fi

6

831631 355

Figure 3. RT-PCR analysis of VCAM-1 transcript in melanoma cells. Melanoma cells were either untreated or incubated with TNF (500 U/mI) for 2, 4, and 12 hours. Total RNA was extracted and subjected to RT-PCR with specific primers; (A) shows the ethidium bromide stained gel electrophoresis ofRT-PCR; (B) shows the bybridization of the same gel ofpanel A with a VCAM-1 probe blotted onto nylon filter. Lanes are as follows: ST: molecular weight marker (pBR328 Bgl I + Hinf I). 1: 2/4. 2: 2/4 + TNF (4 hr). 3: 2/14. 4: 2/14 + TNF (12 hr). 5: 2/14. 6: 2/14 + (4 hr). 7: Endothelial cells treated with IL-1 for 4 hr. 8: Negative control (no RNA).

similar observations when we studied the interaction of melanoma clones with endothelium.11 VLA-4 may interact with fibronectin or other as yet unrecognized ligands on the surface of melanoma cells. The results discussed so far suggested that some human melanomas cultured in vitro have the potential to express VCAM-1. It was important to investigate whether melanoma lesions in vivo express VCAM-1. Immunohistochemical studies using MAb BB1 6-V10 provided the following information (Table 2 and Figure 5). Positive and negative lesions did not differ morphologically in an appreciable way. None of the blue nevi and one of two congenital nevi reacted homogeneously with MAb BB1 6V10. Testing of primary melanomas disclosed no major differences in reactivity among the various histotypes. The expression of VCAM-1 did not appear to be correlated with increasing tumor thickness (Table 3). Four of five lesions with thickness equal to or less than 1 mm and four of six lesions thicker than 4 mm were in fact stained by MAb BBIG-V1 0. Detectable levels of VCAM-1 were detected in fewer than one third of metastatic foci.

typic cells. Finally, and most importantly, in vivo the expression of VCAM-1, although not restricted to transformed melanocytes, was demonstrated in a substantial fraction of human melanomas. Vascular cell adhesion molecule is not the first example of antigenic structure common to melanoma and endothelial cells that is susceptible to modulation by cytokines, because this property is shared by ICAM- .23,24 At variance from the latter molecule, VCAM-1 does not display a differential distribution in primary and metastatic melanoma lesions. Furthermore, its expression does not seem to be modulated during tumor progression, at least in the number of cases analyzed in the current study. There was obvious gross correlation between VCAM-1 expression on tumor cells and extent of the lymphoreticular infiltrate or expression in the tumor vascular bed. This 40

co 0

Discussion The results presented here show that at least certain human malignant melanomas have the potential to express the endothelial adhesion molecule VCAM-1. Tumor necrosis factor-inducible cell surface, MAb-defined VCAM-1 was associated with expression of specific mRNA transcripts. The TNF-inducible VCAM-1 on 2/14 melanoma cells was biologically functional in that it was recognized by VLA-4-expressing homotypic and hetero-

0

-

10

none anti VLA-4 anti VCAM-1 Figure 4. Adhesion of VLA-4 positive cells on TNF treated and non-treated 2/14 melanoma cells. 2/14 melanoma cells were labeled and subsequently added on non-treated (O) 2/14 or TNF treated (F]) 2/14 melanoma cells. Effect of anti-VLA4 and antVCAM-1 mAb was investigated

1328

Jonjic et al


Figure 5. Immunohistochemical detection of VCAM-1 by indirect avidinebiotin immunoperoxidase stain on 4 pLm acetonefixed cryostat sections of an intradermal nevus (a) of a melanoma from superficial spreading (b) and ofa metastatic melanoma (c). VCAM is distributed with a crescent-like pattern in the nevic and primary melanoma cells while is homogenously expressed on the plasma membrane of the metastatic cells. Mayer's haematoxylin counterstain. (a, b x 139; c X200) ep: epidermis.

point should be examined more in depth by studying cytokine expression in vivo as well as 'activation' of the VCAM-1 counter receptor VLA-4. Melanoma cells express a panoply of adhesion molecules that can engage in cell-to-cell and cell-to-matrix interactions. Previous studies using clones derived from the same melanoma lesion have demonstrated intratumoral heterogeneity in the expression of ICAM-1 and VLA-4.25 The current study extends this observation to VCAM-1. The relevance of VCAM-1-regulated expression in malignant melanoma remains to be established. One would predict that VCAM-1 on melanoma cells should facilitate retention of, and interaction with, infiltrating mononuclear cells. Also, this adhesion molecule has the potential to engage in homotypic interactions given the expression of VLA-4 on many melanomas. The significance of cytokine-regulated VCAM-1 in this neoplasia

remains a matter of speculation and should probably be viewed in the context of the ensemble of adhesive molecules expressed on this tumor. Table 2. Expression of VCAM-1 Molecule in Benign and Malignant Lesions of the Melanocyte Lineage Benign Junctional nevi 2/6* (weak stain) Intradermal nevi 5/13 (weak stain of "A" type cells) 1/2 Congenital nevi Blue nevi 0/5 Malignant 2/3 (< 1 mm) Superficial spreading (SS) 6/12 (>1 mm) Melanoma from SS Acral lentiginous melanoma 3/5 (>1 mm) Nodular melanoma 3/6 (>1 mm) 9/31 Metastatic melanoma * Fraction positive.

VCAM-1 in Melanoma

1329


Table 3. Relationship Between of VCAM-1 Molecule and Degree of Dermal Invasiveness of Cutaneous Melanoma

Patient

Histodiagnosis

Ri Ba Di Bar 5 Mo 6 Gu 7 Ga 8 Po 9 Da 10 Fr 11 Fo 12 Gi 13 Ca 14 Ta 15 Can 16 Ri 17 Pa 18 Al 19 Po 20 Be 21 Op 22 Tar 23 Bor 24 By 25 Fl 26 Fi

SS SS SS SSM SSM SSM ALM SSM SSM SSM SSM ALM SSM NM SSM SSM SSM SSM ALM NM NM ALM NM NM ALM NM

1 2 3 4

Thickness (mm)

VCAM-1

0.5 0.5 0.5 1.0

h h h

1.0

is h

1.5 1.5 1.5

1.5 1.8 1.8 2.0 2.0 2.2

is + +

h

2.7

3.0 3.0 3.2 3.5 4.0 4.5 4.5 4.5 5.0 5.0 6.0

h -

-

h + h + -

-: no stain, h: heterogeneous stain with negative areas, is: stain of isolated cell nests, +: homogenous stain of various intensity. SS: superficial spreading; SSM: superficial spreading melanoma; ALM: acral lentiginous melanoma; NM: nodular melanoma.

Acknowledgment The authors thank Miss Maria Rita Nicotra from First Biomedical Technology, CNR, Rome, for technical help.

References 1. Osborn LC, Hession R, Tizard C, Vassallo S, Luhowskyj S, Chirosso G, Lobb RR: Direct expression cloning of vascular cell adhesion molecule 1, a cytokine-induced endothelial protein that binds to lymphocytes. Cell 1989, 59:1203-1211 2. Rice GE, Bevilacqua MP: An inducible endothelial cell sur-

face glycoprotein mediates melanoma adhesion. Science 1989, 246:1303-1306 3. Rice GE, Munro JM, Bevilacqua MP: Inducible cell adhesion molecule 110 (INCAM-1 10) is an endothelial receptor for lymphocytes: A CD1 1/CD1 8-independent adhesion mechanism. J Exp Med 1990,171:1369-1374 4. Elices MJ, Osborn LC, Takada Y, Crouse C, Luhowskyj S, Hemler ME, Lobb RR: VCAM-1 on activated endothelium interacts with the leukocyte integrin VLA-4/fibronectin binding site. Cell 1990, 60:577,584 5. Carlos TM, Schwartz BR, Kovach NL, Yee E, Rosso M, Osbom L, Chi-Roso G, Newman B, Lobb RR, Harlan JM: Vascular cell adhesion molecule-1 mediates lymphocyte adherence to cytokine-activated cultured human endothelial cells. Blood 1990, 76:965-970

6. Carlos TM, Harlan JM: Membrane proteins involved in phagocyte adherence to endothelium. Immunol Rev 1990, 114:5 7. Allavena P, Paganin C, Martin-Padura I, Peri G, Gaboli M, Dejana E, Marchisio PC, Mantovani A: Molecules and structures involved in the adhesion of natural killer cells to vascular endothelium. J Exp Med 1991, 173:439-448 8. Bochner BS, Luscinskas FW, Gimbrone MA Jr, Newman W, Sterbinsky SA, Derse-Anthony CP, Klunk D, Schleimer RP: Adhesion of human basophils, eosinophils and neutrophils to interleukin-1 activated human vascular endothelial cells: Contribution of endothelial cell adhesion molecules. J Exp Med 1991,173:1553-1556 9. Walsh GM, Mermond JJ, Hartnell A, Kay AB, Wardlaw AJ: Human eosinophil, but not neutrophil, adherence to IL-1stimulated human umbilical vascular endothelial cells is alfa4 betal (very late antigen-4) dependent. J Immunol 1991, 146:3419-3423 10. Rice GE, Gimbrone MA Jr, Bevilacqua MP: Tumor cellendothelial cell interactions: Increased adhesion of human melanoma cells to activated vascular endothelium. Am J Pathol 1989, 133:204-210 11. Martin-Padura I, Mortarini R, Lauri D, Bernasconi S, Sanchez-Madrid F, Parmiani G, Mantovani A, Anichini A, Dejana E: Heterogeneity in human melanoma cell adhesion to cytokine activated endothelial cells correlates with VLA-4 expression. Cancer Res 1991, 51:2239-2241 12. Rice GE, Munro JM, Corless C, Bevilacqua P: Vascular and nonvascular expression of INCAM-1 10. A target for mononuclear leukocytes adhesion in normal and inflamed human tissues. Am J Pathol 1991, 138:385-393 13. Freedman AS, Munro JM, Rice GE, Bevilacqua MP, Morimoto C, Mclntrye BW, Rhynhart K, Pober JS, Nadler LM: Adhesion of human B cells to germinal centres in vitro involves VLA-4 and INCAM-1 10. Science 1990, 249:1030-1033 14. Koopman G, Parmentier HK, Schuurman HJ, Newman W, Meijel CJLM, Pals ST: Adhesion of human B cells to follicular dendritic cells involves both the lymphocyte functionassociated antigen 1 /vascular cell adhesion molecule pathways. J Exp Med 1991, 173:1297-1304 15. Anichini A, Mazzochi A, Fossati G, Parmiani G: Cytotoxic T lymphocyte clones from peripheral blood and from tumor site detect intratumor heterogeneity of melanoma cells. J Immunol 1989,142:3692-3701 16. Maniatis T, Fritsch EF, Sambrook J: Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY, Cold Spring Laboratory, 1982, p 545 17. Chirgwin JM, Przybyla AE, MacDonald RJ, Rutter WJ: Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry 1979, 18:52945295 18. Breslow A: Tumor thickness, level of invasion and node dissection in stage cutaneous melanoma. Ann Surg 1975,

182:572-575 19. Colotta F, Peri G, Villa A, Mantovani A: Rapid killing of actinomycin D-treated tumor cells by human mononuclear cells: I. Effectors belong to the monocyte-macrophage lineage. J Immunol 1984, 132:936-943 20. Bemasconi S, Peri G, Sironi M, Mantovani A: Involvement of

1330

Jonjic et al


leukocyte (beta2) integrins (CD18/CD1 1) in human monocyte tumoricidal activity. Int J Cancer 1991, 49:267-273 21. Cybulsky MI, Fries JWU, Williams AJ, Sultan P, Davis VM, Gimbrone MA Jr, Collins T: Alternative splicing of human VCAM-1 in activated vascular endothelium. Am J Pathol 1991, 138:815-820 22. Hession C, Tizard R, Vassallo C, Schiffer BS, Goff D, Moy P, Chi-Rosso G, Luhowskyj S, Lobb RR, Osborn LC: Cloning of an alternate form of vascular cell adhesion molecule-1 (VCAM1). J Biol Chem 1991, 266:6682-6685 23. Johnson JP, Stade BG, Holzmann B, Schwable W, Rieth-

muller G: De novo expression of intercellular-adhesion molecule 1 in melanoma correlates with increased risk of metastasis. Proc Natl Acad Sci USA 1989, 86:641-644 24. Natali PG, Nicotra MR, Cavaliere R, Bigotti A, Romano G, Temponi M, Ferrone S: Differential expression of intercellular adhesion molecule 1 in primary and metastatic melanoma lesions. Cancer Res 1990, 50:1271-1278 25. Anichini A, Mortarini R, Supino R, Parmiani G: Human melanoma cells with high susceptibility to cell-mediated lysis can be identified on the basis of ICAM-1 phenotyte, VLA profile and invasive ability. Int J Cancer 1990, 46:508-515