~

Cancer Immunol Immunother (1989) 30:262-268

ancer mmunol9gy mmunotherapy

© Springer-Verlag 1989

In vitro differentiation and antigenic changes in human melanoma cell lines Ludovico Guarini l, Massimo Temponi 3, Gretchen M. Edwalds 2, Joseph R. Vita 2, Paul B. Fisher 2, and Soldano Ferrone 3 Division of Pediatric Hematology/Oncology, and 2Departments of Neurological Surgery, Pathology and Urology, and Comprehensive Cancer Center/Institute of Cancer Research, Columbia University, College of Physicians and Surgeons, New York, NY 10032, and 3Department of Microbiology and Immunology, New York Medical College, Valhalla, NY 10595, USA

Summary. Malignant transformation of melanocytes may be associated with changes in the expression of HLA antigens and melanoma-associated antigens (MAA). To determine whether these changes reflect the differential expression of HLA antigens and MAA by melanocytes at different stages of differentiation, we have studied the effect of the reversible induction of differentiation by fibroblast interferon (interferon 13) and/or 12-O-tetradecanoyl-phorbol 13-acetate (TPA) on the expression of HLA antigens and MAA by the melanoma cell lines DU-2, FO-1 and HO-1. The three melanoma teil lines differed in their sensitivity to the differentiating and antiproliferative activity of these two compounds and displayed an increased growth suppression and induction of differentiation, when incubated with the combination of TPA and interferon [3. Incubation of the three melanoma cell lines with interferon 13, TPA or their combination resulted in a differential modulation of the expression of membrane-bound high-molecular-mass melanoma-associated antigen, l l 5 - k D a MAA, 100-kDa MAA, intercellular adhesion molecule 1, HLA class I antigens and gene products of the HLA-D region. Each melanoma cell line displayed a unique pattern of antigenic modulation when exposed to the two differentiating agents alone or in combination. No direct relationship was found between the effects of interferon 13 and/or TPA on the growth and differentiation of the three melanoma cell lines and the expression of HLA antigens or the MAA evaluated in the present study. These findings argue against a direct role of any of the antigens tested in the reversible induction of human melanoma cell differentiation in the in vitro system.

likely to play a major role in the interaction of melanoma cells with the host's immune system and may represent useful markers to analyze the molecular basis of the malignant transformation of melanocytes. Whether the changes in the antigenic profile of melanocytes associated with their malignant transformation represent an epiphenomenon of the malignant process, or reflect the differential expression of HLA antigens and MAA by melanocytes at different stages of differentiation, remains to be determined. In the present study we have investigated the latter possibility by characterizing the changes of expression of HLA antigens and MAA in melanoma cells after incubation with interferon 13 (IFN-[3) and/or 12-O-tetra-decanoyl-phorbol 13-acetate (TPA). We have utilized these two modulating agents, since they induce both reversible and terminal differentiation of melanoma cells (for review see [1, 4, 5]) and since their combination has a synergistic effect and can induce a dramatic suppression of cell growth and a concomitant induction of terminal differentiation of melanoma cells [6, 9, 10], even in teil lines relatively resistant to either agent alone [9, 10]. We have utilized the melanoma cell lines DU-2, FO-1 and HO-1, since they display differential sensitivity to the differentiating and antiproliferative activity of IFN-13 and TPA, and differ in their expression of HLA class I and class II antigens [9, 10]. Among the MAA defined by mAbs, we have selected the membrane-bound high-molecular-mass melanoma-associated antigen (high-M~ MAA), ll5-kDa MAA, 100-kDa MAA, and intercellular adhesion molecule-1 (ICAM-1) since they have been shown previously to be susceptible to modulation by cytokines [13, 24].

Materials and methods Introduction Malignant transformation of human melanocytes may be associated with changes in their antigenic profile. These changes include the appearance of melanoma-associated antigens (MAA) and HLA class II antigens, and the reduction or loss of HLA class I antigens (for review, see [3, 22, 27, 31]). The biological and clinical significance of these changes and the mechanisms underlying them are the focus of active researeh, since MAA and HLA antigens are Offprint requests to: L. Guarini, Columbia University, College of Physicians & Surgeons, 630 West 168th Street, New York, NY 10032, USA

Cell lines and growth conditions. The melanotic melanoma cell line HO-1 and the amelanotic melanoma cell lines DU-2 and FO-1 [9] were grown in monolayer culture at 37°C in Dulbecco's minimal essential medium (DMEM) supplemented with 5% fetal bovine serum (DMEM-5). The cell lines were used between passages 50 and 150. For each experiment cells were plated at a concentration of 2 x 10s/tal and, 24 h later, the medium was ¢hanged with medium containing the compound to be tested. These conditions allowed the cells to be in a logarithmi¢ phase of growth when exposed to the compounds under investigation. After incubation with the appropriate media for specified periods, cells were harvested by brief treatment with

263 trypsin/versene 0.125%/0.02% (w/v), washed with phosphate-buffered saline (PBS), p H 7.5, and resuspended at a concentration of 5 x 106/ml in cold PBS supplemented with 0.5% bovine serum albumin.

Monoclonal antibodies and conventional ant&era. The antiMAA mAb included the anti-(high-M MAA) mAb 225.28, the anti-(115-kDa MAA) mAb 345.134, the anti-(100-kDa MAA) mAb 376.96, and the anti-(96-kDa MAA) mAb CL203.4 [14, 19, 20, 24, 26, 36]. Immunochemical studies have shown that the mAb CL 203.4 recognizes ICAM-1. The anti-HLA mAbs included the mAb W6/32 to a framework determinant of HLA class I antigens, the anti-[32-microglobulin mAb NAMB-1, the anti-HLA-DR mAb CL413, the anti-(HLA-DQwl+DQw3) mAb KS6 and KS7, and the anti-HLA-DP mAb B7/21 [2, 32, 33, 35]. Fluorescein-isothiocyanate-conjugated F(ab')2 fragments of goat anti-(mouse Ig) antibodies (FITC-GaM) were purchased from Jackson Immuno Research Laboratories, Avondale, Pa. Chemicals. TPA was obtained from Consolidated Midland (Brewster, NY). Stock solutions at 1 mg/tal prepared in dimethylsulfoxide were divided into small aliquots and stored at - 20 ° C. IFN-fl. IFN-[3 was produced in E. coli as previously described [25]. IFN-[3 samples were calibrated against the international fibroblast standard, G-023-902-527, from the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, Md. Purified IFN-[3, having an antiviral activity of approximately 1 x 10 7 units/mg, was kindly supplied by Drs. R. Y. Ning and S. J. Tarnowski (Hoffman-LaRoche Inc., Nutley, NJ). The IFN-[3 preparations were divided into aliquots and stored at 10 6 units/ml at - 8 0 ° C , thawed immediately before use, and diluted to the appropriate concentration in DMEM-5.

1. Effects of IFN-[3 and/or TPA on the growth of DU-2, FO-1 and HO-1 human melanoma cell lines"

Table

IFN-[3 (U/ml) -

500 1000 2000 500 1000 2000

TPA (ng/ml) -

-

50 50 50 50

Growth (%) DU-2

FO-1

HO-1

100 96 75 67 5! 47 35 29

100 52 30 ND b 100 22 19 ND

100 93 84

75 80 43 31 27

Cells (2 x 10») were seeded in 35-mm tissue-culture plates; 4 h after plating cells, the indicated compounds in DMEM-5 were added. Following a 72-h incubation at 37°C cell numbers were determined in triplicate samples by Coulter counter and expressed as percentages of the control values. Values in triplicate samples varied by less than 10%., IFN-[3, interferon [3; TPA, 12-O-tetradecanoyl-phorbol 13-acetate b Not determined

were seeded at 2 x 105 cells/35 mm plate. Following a 4-h incubation at 37 ° C, the medium was exchanged with medium containing the indicated compounds and incubation was continued for an additional 72 h. The triplicate cultures were harvested as described above and cell concentrations were determined using a model Zf Coulter counter [9]. Results are expressed as percentages of the control. Replicate samples varied by < 10%. Control cultures containing 0.01% dimethylsulfoxide, the final concentration used in cultures containing TPA, were tested for solvent effect, and no effect on growth or melanin synthesis of the melanoma cell lines used in this study was noted. Results

Serological assays. Indirect immunofluorescence was performed by mixing 50 p~l single-cell suspension (5 x 106/ml) with 50 gl spent medium from a hybridoma secreting the mAb to be tested. This amount was found to be saturating for all of the antibodies used. At the end of a 30-min incubation at 4°C, cells were washed twice with cold PBS, containing 0.001 M sodium azide and incubated with 50 p~l FITC-GaM preparation for 30 min at 4°C. After two washings with PBS/azide, cells were analyzed by flow cytometry using a FACStar (Becton Dickinson, Mountain View, Calif). The Consort 30 software program that runs this instrument was used for the analysis of the data obtained in the list mode. For each sample, 10000 cells were analyzed using a flow rate of 1000 events/s. Dead cells and cell clumps were excluded from the analysis by gating of the data obtained in the list mode. All the fluorescence data were obtained and expressed in a logarithm base-10 scale. The background fluorescence values, obtained by analyzing cells sequentially incubated with nonreactive mAbs and FITC-GaM, were subtracted from the fluorescence values obtained from each experimental point. Growth analysis. The effect of TPA and IFN-13, alone or in combination, on 72 h growth of cultured melanoma cells was determined as described previously [9]. Briefly, cells

In preliminary experiments we determined the optimal doses of IFN-[~ and TPA and appropriate in¢ubation times to achieve maximum modulation of the expression of HLA antigens and MAA by the three cultured melanoma ¢ell lines. For those antigens that showed a clear dose-response pattern to TPA (HLA ¢lass I and ¢lass II antigens and high-Mr MAA) or to IFN-[3 (HLA class I and class II antigens, high-Mr MAA and ICAM-1) a plateau was reached at a concentration of 5 ng/ml and 500 U/tal, respectively (data not shown). The time course of the modulating effect of TPA and/or IFN-[~ on the expression of HLA antigens and MAA varied with individual antigens, with the majority showing maximal ¢hange at 72 h. All additional experiments were performed using final concentrations of 2000 U/tal and 50 ng/ml IFN-[3 and TPA, respectively, for the cell lines DU-2 and HO-1, while lower concentrations were used for the cell line FO-1 because of its higher sensitivity to the two compounds, and an incubation time of 72 h. These concentrations were used since they gave the highest growth inhibition (Table 1) and the maximum increase in melanin production (data not shown). To determine whether a common denominator exists between the stage of differentiation of melanocytes and changes of expression of surface antigens, we studied the changes in the antigenic profile of the melanotic melano-

264 Table 2. Effect of IFN-[~ and/or TPA on the expression of HLA antigens and melanoma-associated antigens (MAA) by melanoma cell lines DU-2, FO-1 and HO-1 a Antigen

IFN-[~

TPA

DU-2 Ab

FO-1 B°

A

HO-1 B

A

B

HLA class I

+ +

+ +

27.22 165.69 40.50 158.54

1.00 6.08 1.48 5.82

2.11 2.45 2.09 2.76

1.00 1.16 0.99 1.30

39.15 58.34 50.39 64.79

1.00 1.49 1.28 1.65

HLA-DR

+

2.11 3.15 2.13 22.67

1.00 1.49 1.00 10.74

12.88 43.80 10.44 2.66

1.00 3.40 0.81 0.20

11.02 20.27 8.98 31.22

1.00 1.84 0.81 2.83

+

+ +

High- Mr MAA

+ +

+ +

9.75 6.8 14.9 8.35

1.00 0.69 1.52 0.85

10.58 8.52 8.72 8.74

1.00 0.80 0.82 0.82

10.04 6.46 9.24 3.28

1.00 0.64 0.92 0.32

ll5-kDa MAA

+ +

+ +

23.06 33.34 26.39 27.63

1.00 1.44 1.14 1.19

55.55 95.19 63.58 76.50

1.00 1.71 1.14 1.37

31.11 27.95 47.93 13.87

1.00 0.89 1.54 0.44

100-kDa MAA

+ +

+ +

7.62 7.12 12.46 6.48

1.00 0.93 1.63 0.85

12.78 23.75 13.19 33.47

1.00 1.85 1.03 2.61

10.78 9.04 14.95 8.68

1.00 0.83 1.38 0.80

ICAM-1 d

+ +

+ +

3.15 3.28 3.26 3.43

1.00 1.04 1.03 1.08

10.14 13.67 8.17 18.66

1.00 1.34 0.80 1.84

4.89 6.20 5.80 41.65

1.00 1.26 1.18 8.51

-

Cultured melanoma cells were incubated for 72 h at 37°C with IFN-[3 (2000 U/ml for DU-2 and HO-1 cells and 1000 U/ml for FO-1 cells), TPA (50 ng/ml) and the combination of IFN-[~ (2000 U/tal for DU-2 and HO-1 cells and 1000 U/ml for FO-1 cells) and TPA (50 ng/ml). Control cultures were incubated in parallel with the solvent dimethylsulfoxide (final concentration 0.01%). At the end of the incubation, cells were harvested, washed twice with PBS/azide and incubated for 30 min at 4°C with anti-(HLA class I) mAb W6/32, with anti-HLA-DR mAb CL 413, with anti-(high-Mr MAA) mAb 225.28, with anti-(115-kDa MAA) mAb 345.134, with anti-(100-kDa MAA) mAb 376.96 and with anti-ICAM-1 mAb CL 203.4. Cells were then washed and incubated for 30 min at 4°C with fluoescein-isothiocyanate-labelled goat anti-(mouse Ig) (FITC-GaM). At the end of the incubation, cells were washed and analyzed with a FACStar b Figures indicate the peak fluorescence channel as detected by a FACstar instrument c Figures represent the ratio between values in experimental sample and in the control sample d ICAM-1, intercellular adhesion molecule 1 a

ma cell line HO-1 a n d of the amelanotic m e l a n o m a cell lines DU-2 a n d FO-I. The latter two cell lines display differential sensitivity to the antiproliferative and differentiation-inducing activity of IFN-~ and TPA [9]. The cell line FO-1 was the most sensitive to the antiproliferative activity of IFN-I3, while the cell line DU-2 was the most resistant. At a concentration of 50 n g / m l , TPA did not affect the growth of FO-1 cells, but inhibited that of DU-2 cells slightly and that of HO-1 cells markedly (Table 1). The c o m b i n a t i o n of IFN-[3 and TPA was synergistic in inhibiting the growth of FO-1 and HO-1, while it was additive in DU-2, even in cell lines which were relatively resistant to the individual compounds. The modulating effects of IFN-[3 and TPA, alone and in combination, on the expression of HLA antigens and M A A by the three m e l a n o m a cell lines are summarized in Table 2 and Fig. 1. Like HO-1 cells, DU-2 cells displayed an increase in the expression of H L A class I antigens following incubation with TPA or with IFN-[L The increase induced by IFN-I3 was higher than that induced by TPA, but similar to that induced by the combination of the two compounds. IFN-[3 was more active in enhancing the ex-

pression of HLA class I antigens by DU-2 cells than by HO-1 cells. Unlike HO-1 cells, DU-2 cells incubated with IFN-[3 and TPA displayed a cytofluorographic pattern compatible with the presence of a dual p o p u l a t i o n of cells which differ in their expression of H L A class I antigens (Fig. 2). IFN-[3 alone and in combination with TPA did not induce the expression of HLA class I antigens on FO-1 cells. Unlike FO-1 cells, which displayed an increase of expression of H L A - D R antigens only after incubation with IFN-[~ alone, HO-1 cells increased their expression of H L A - D R antigens following i n c u b a t i o n with IFN-I3 alone and in c o m b i n a t i o n with TPA. The level of expression of H L A - D R antigens on all three m e l a n o m a cell lines was not altered by T P A treatment. DU-2 cells did acquire these antigens following incubation with IFN-[~, alone or in combination with T P A (Fig. 3). H L A - D Q and DP antigens were not detected on three cell lines and they were not induced following i n c u b a t i o n with IFN-[~ a n d / o r TPA (data not shown). M A A were differentially modulated on the three melan o m a cell lines by IFN-[~ a n d / o r TPA (Table 2). TPA did not affect the level of high-Mr M A A on HO-1 cells, en-

265 350

HLA CLASS I /'~ A --Control / I iß -'- lEN-,6'

200

HLA CLASSII

:'"'i

I t."it " TPA

0

fr" LLI ra :Z) Z t--Z LL] > Ld

300

Icl m

2it, HMW-MAA

l--z

115K MAA

>

iI/~~ 1: I;"

300

B

\ \ 0 I0 °

:.

i

IOOK MAA

I

"~

ICAM-I

I0 0

.:"~ \ \.h. \

:..

0

L'I

I01

I0 z

I

103

IO I

:~, IO z

\

, I0 s

FLUORESCENCE

li /

\

ùL)~"'~ ........,. 0

I01

I0 2

i

103

FLUORESCENCE

Fig. 1. Effect of IFN-13 and/or TPA on the expression of HLA antigens and MAA by the melanoma cell line HO-1. Cultured melanoma cells were incubated for 72 h at 37°C with IFN-13 at a final concentration 2000 U/ml (--.--), TPA (50 ng/ml) (...), and the combination of IFN-13(2000 U/ml) and TPA (50 ng/ml) (- - -). Control cultures ( - - ) were incubated in parallel with the solvent dimethylsulfoxide (0.01%). At the end of the incubation cells were washed twice with PBS azide and incubated for 30min at 4°C with anti-(HLA class I) mAb W6/32, with anti-HLA-DR mAb CL 413, with anti-(high-Mr MAA) mAb 225.28, with anti-(ll5-kDa MAA) mAb 345.134, with anti-(100-kDa MAA) mAb 376.96 and with anti-ICAM1 mAb CL 203.4. Cells were then washed and incubated for 30 min at 4°C with FITC-GaM. At the end of the incubation, cells were washed and analyzed with a FACStar. The results are expressed as relative binding of FITC-GaM to melanoma cells coated with mAbs. The graphs from separate analyses were merged with the Consort 30 (Rev. D) software package that runs the Becton-Dickinson FACStar used for fluorescence analysis

hanced it on DU-2 cells a n d slightly reduced it on FO-1 cells. On the other hand, IFN-[3 by itsetf and in combination with TPA reduced the expression of high-Mr M A A by the three m e l a n o m a cell lines. The extent of the reduction was greater on HO-1 cells than on the other two cell lines. In the case of the 115-kDa MAA, IFN-I3 slightly reduced its expression by HO-1 cells, whereas it enhanced it on DU-2 and FO-1 cells. In contrast, TPA did not significantly affect the level of l l 5 - k D a M A A on either FO-1 or DU-2 cells, and only slightly increased it on HO-1 cells. The combination of IFN-[3 and TPA reduced the expression of 115-kDa M A A by HO-1 cells, increased it by FO-1 cells and did not affect it on DU-2 cells. The expression of 100-kDa M A A by HO-1 and DU-2 cells was not significantly changed by IFN-[3 a n d / o f by the c o m b i n a t i o n of IFN-13 a n d TPA, while it was increased by TPA alone. On the other hand, the expression of this antigen by FO-1 cells was enhanced by both IFN-13 and T P A individually, and even more by their combination. The two c o m p o u n d s individually did not change the expression of ICAM-1 by the three m e l a n o m a cell lines. The c o m b i n a t i o n of the two

Fig. 2. Effect of IFN-[3 and/or TPA on the expression of HLA class I antigens by the cultured melanoma ceU line DU-2. Cultured melanoma cells were incubated for 72 h at 37°C with IFN-[3 (2000 U/ml) (--.--), TPA (50 ng/ml) (...), or the combination of IFN-13 (2000 U/tal) and TPA (50 ng/ml) (---). Control cultures ( - - ) were incubated in parallel with the solvent dimethylsulfoxide (0.01%). At the end of the incubation cells were washed twice with PBS azide and incubated for 30 min at 4°C with anti-(HLA class I) mAb W6/32. Cells were then washed and incubated for 30 min at 4°C with FITC-GaM. At the end of the incubation, cells were washed and analyzed with a FACStar. The results are expressed as relative binding of FITC-GaM to melanoma cells coated with mAbs. The graphs from separate analyses were merged with the Consort 30 (Rev. D) software package that runs the Becton-Dickinson FACStar used for fluorescence analysis

3OO

w

0 I0 °

I

I01

I0 z

10 3

FLUORESCENCE Fig. 3. Effect of IFN-p and/or TPA on the expression of HLADR antigens by the cultured melanoma ce]l line DU-2. Cu]tured melanoma cel]s were incubated for 72h at 37°C with IFN-13 (2000 U/tal) (--.--), TPA (50 ng/ml) (...), or the combination of IFN-13 (2000 U/tal) and TPA (50 ng/ml) (---). Contro] cultures ( - - ) were incubated in parallel with the solvent dimethylsu]foxide (0.01%). At the end of the incubation ce]ls were washed twice with PBS azide and incubated for 30 min at 4°C with anti-HLADR mAb CL 413. Ce]ls were then washed and incubated for 30 min at 4°(3 with FITC-GaM. At the end of the incubation, ce]]s were washed and ana]yzed with a FACStar. The results are expressed as relative binding of FITC-GaM to melanoma cel]s coated with mAb. The graphs from separate analyses were merged with the Consort 30 (Rev. D) software package that runs the Becton-Dickinson FACStar used for fluorescence ana]ysis

266 agents did not affect it on DU-2 cells, but enhanced it on both FO-1 and HO-1 cells. The extent of the increase was much greater on the latter than on the former cell line. Discussion

Both IFN-[3 and TPA can suppress growth of cultured human melanoma cells. In some cell lines growth suppression is associated with an increase in the synthesis of the melanoma-specific differentiation marker melanin [1, 8-10, 18, 23]. The combination of IFN-[3 and TPA is synergistic in inducing growth suppression and differentiation of melanoma cell lines sensitive to either individual agent and of those which are spontaneously resistant to, or have been selected for resistance to, the individual agents [5, 10, 81. In the present investigation we have utilized this phenomenon to investigate the role of differentiation in the antigenic changes associated with the malignant transformation of melanocytes. We have shown a differential effect of IFN-[~ and/or TPA on the expression of HLA antigens and MAA by the melanoma cell lines DU-2, FO-1 and HO-1, which differ in their sensitivity to the antiproliferative and differentiating activity of the two agents alone and in combination. Under the experimental conditions of reversible differentiation, induced by IFN-[~ and/ or TPA [1, 9, 10] no relationship was found between the effect of the two agents on the antigenic profile of the three melanoma cell lines and the effect on their differentiation. These results indicate that changes in the expression of MAA and HLA antigens associated with the malignant transformation of melanocytes do not represent a differentiation-related phenomenon. Furthermore these results argue against a role for HLA antigens and for the tested MAA in the process of reversible induction of differentiation in the model system we have investigated. The differential susceptibility of the three melanoma cell lines to IFN-[~ a n d / o r TPA is not likely to reflect differënces in the number and/or affinity of the respective receptors, since each agent modulates the expression of the various types of antigens analyzed to a different extent within each cell line. Alternative mechanisms explaining out findings include differences in the ability of the different melanoma cell lines to transmit the necessary transmembrane signals following the binding of IFN-I] and TPA to their receptors and/of in their ability to induce the necessary transcriptional or post-transcriptional changes required for the phenotypic alterations to occur [7, 12]. Furthermore, the differential susceptibility to modulation by IFN-~ and/or TPA of HLA antigens and MAA expressed by the three cell lines suggests that they are controlled by different regulatory mechanisms. This conclusion is in agreement with the results of other investigations that have utilized different approaches to analyze the regulatory mechanisms which eontrol HLA antigens and MAA expression [37, 38]. The effect of IFN-~ or TPA on the antigenic profile of melanoma cells has already been investigated [10, 13, 17, 28]. The following comments emphasize the new information derived from the present investigation. The expression of ICAM-1 by cultured melanoma eells is highly enhanced by immune interferon, but is not modulated by leukocyte interferon or by IFN-[3 [24]. The present study confirmed the inability of IFN-[3 to modulate ICAM-1 significantly, but has shown for the first time that the combination of

IFN-I3 and TPA can markedly enhance the expression of ICAM-1 by HO-1 cells. The effect of IFN-[3 on the expression of the high-Mr MAA, 115-kDa MAA and 100-kDa MAA has been previously investigated using only the melanoma cell line Colo 38 [13]. In this cell line, IFN-~ did not affect the expression of high-Mr MAA, but increased its shedding. Furthermore IFN-~ increased the expression and shedding of the ll5-kDa MAA and 100-kDa MAA by Colo 38 cells. By contrast, in the present investigation, IFN-[~ decreased the expression of the high-Mr MAA by DU-2, FO-1 and HO-1 cells and increased its shedding by the three cell lines (data not shown). Furthermore, IFN-[~ enhanced the expression of the 115-kDa MAA in both FO-1 and HO-1 cells but of the 100-kDa MAA only in HO-1 cells. TPA has been reported to have a differential effect on the expression of high-Mr MAA by four melanoma cell lines [28]. Similarly, in out study TPA also varied in its ability to alter the expression by DU-2, FO-1 and HO-1 cells of high-M~ MAA, HLA antigens and the other types of MAA analyzed. The meehanism(s) underlying the differential susceptibility to modulation of HLA antigens and MAA by the various melanoma cell lines is not known. These findings may account for the antigenic heterogeneity that has been found among melanoma cells within a surgically removed lesion, among autologous lesions removed from different anatomic sites, and among lesions removed from different patients [29, 30, 34]. IFN-I3 enhances the expression of HLA class I antigens by melanoma cells, but is unable to induce the expression of HLA class II antigens on all the melanoma cell lines tested, except the MEM 50-10 cell line. This cell line, as the DU-2 cell line we describe hefe, after incubation with IFN-[~, acquires reactivity with anti-HLA-DR mAb, but not with anti-HLA-DQ or anti-HLA-DP mAb. The higher susceptibility of HLA-DR antigens compared to HLA-DQ antigens to modulation by IFN-[~ is not a general phenomenon, since IFN-[3 has been shown to be more effective in enhancing the expression of HLA-DQ than that of HLADR antigens by melanoma cells Colo 38 [13, 15]. IFN-[3 enhanced the surface expression of HLA class I antigens on DU-2 and HO-1 cells, but did not induce their expression on the essentially HLA-class-I-antigen-negative melanoma cell line FO-1. TPA enhanced HLA class I antigen expression only by HO-1 cells. These results suggest that the mechanism(s) by which IFN-[~ and TPA modulate HLA class I antigen expression on melanoma cells may be different. It has recently been suggested that up-regulation of tumor-associated antigen expression by immunomodulators may provide an approach to enhancing the sensitivity of immunodiagnostie and immunotherapeutic uses of mAb. The described differential susceptibility to modulation of human melanoma cells by biological response modifiers, such as IFN-[3, stresses the need to test the antigenic changes induced by these agents in melanoma lesions removed from patients individually before using these agents for clinical applications [11, 16, 21].

Acknowledgements: We thank Liana Apelis for the preparation of the manuscript. These studies were supported by grants from the National Institutes of Health AI21384, CA35675, CA37959 and CA39559. Dr. Guarini is supported in part by Clinical Investigator Award 1 KO8 A100750-01A1 from the National Institutes of Health.

267

References 1. Ahmed MA, Guarini L, Ferrone S, Fisher PB (1988) Induction of differentiation in human melanoma cells by the combination of different classes of interferons or interferon plus mezerein. In: Viral oncogenes and cell differentiation. NY Acad Sci (in press) 2. Barnstable CJ, Bodmer WF, Brown G, Galpe G, Milstein C, Williams AF, Ziegler A (1978) Production of monoclonal antibodies to group A erythrocytes, HLA and other human cell surface antigens - new tools for genetic analysis. Cell 14:9 3. Ferrone S, Imberti L, Natali PG, Schrier PI, Ruiter D (1986) HLA antigens on tumor cells. In: Monoclonal antibodies in cancer: advances in diagnosis and treatment. Futura Publishing Co., New York, p 111 4. Fisher PB (1984) Enhancement of viral transformation and expression of the transformed phenotype by tumor promoters. In: Tumor promotion and cocarcinogenesis in vitro: mechanisms of tumor promotion. CRC Press, Boca Raton, p 57 5. Fisher PB, Grant S (1985) Effects of interferon on differentiation of normal and tumor cells. Pharmacol Ther 27:143 6. Fisher PB, Mufson RA, Weinstein Iß (1981) Interferon inhibits melanogenesis in B-16 mouse melanoma cells. Biochem Biophys Res Commun 100:823 7. Fisher PB, Schachter D, Abbott RE, Callaham MF, Huberman E (1984) Membrane lipid dynamics in human promyelocytic leukemia cells sensitive and resistant to 12-O-tetradecanoyl-phorbol-13-acetate induction of differentiation. Cancer Res 44:5550 8. Fisher PB, Hermo H Jr, Pestka S, Weinstein Iß (1985) Modulation of differentiation in murine and human melanoma cells by interferon and phorbol ester tumor promoters. In: Biological, Molecular and Clinical Aspects of Pigmentation. Univ of Tokyo Press, Tokyo, Japan, p 325 9. Fisher PB, Prignoli DR, Hermo H Jr, Weinstein Iß, Pestka S (1985) Effects of combined treatment with interferon and mezerein on melanogenesis and growth in human melanoma cells. J Interferon Res 5:11 10. Fisher PB, Hermo H Jr, Solowey WE, Dietrich MC, Edwalds GM, Weinstein Iß, Langer JA, Pestka S, Giacomini P, Kusama M, Ferrone S (1986) Effect of recombinant human fibroblast interferon and mezerein on growth differentiation, immune interferon binding and tumor associated antigen expression in human melanoma cells. Anticancer Res 6:765 11. Fisher PB, Griener JW, Mufson RA, Schlom J (1988) Molecules for tumor diagnosis and therapy. In: Application of genetic engineering. Marcel Dekker, New York, NY (in press) 12. Fisher PB, Schachter D, Mufson RA, Huberman E (1988) The role of membrane lipid dynamics and translocation of protein kinase C in the induction of differentiation in human promyelocytic leukemia cells. In: Pharmacological effects of lipids (in press) 13. Giacomini P, Aguzzi A, Pestka S, Fisher PB, Ferrone S (1984) Modulation by recombinant DNA leukocyte (~t) and fibroblast (1~)interferons of the expression and shedding of HLAand tumor associated antigens by human melanoma cells. J Immunol 133:1649 14. Giacomini P, Veglia F, Cordiali Fei P, Rehle T, Natali PG, Ferrone S (1984) Level of a membrane-bound high-molecularweight melanoma-associated antigen and of a cytoplasmic melanoma-associated antigen in surgically removed tissues and in sera from patients with melanoma. Cancer Res 44: 1281 15. Giacomini P, Gambieri R, Barbieri R, Nistico P, Tecce R, Pestka S, Gustafsson K, Natali PG, Fisher PB (1986) Regulation of the antigenic phenotype of human melanoma cells by recombinant interferon. Anticancer Res 6: 877 16. Greiner JW, Guadagni F, Noguchi P, Pestka S, Colcher D, Fisher PB, Schlom J (1987) Recombinant interferon enchances monoclonal antibody-targeting of carcinoma lesions in vivo. Science 235:895

17. Greiner JW, Schlom J, Pestka S, Langer JA, Giacomini P, Kusama M, Ferrone S, Fisher PB (1987) Modulation of tumor associated antigen expression and shedding by recombinant human leukocyte and fibroblast interferons. Pharmacol Ther 31 : 209 18. Huberman E, Heckman C, Langenbach R (1979) Stimulation of differentiated functions in human melanoma cells by tumor-promoting agents and dimethylsulfoxide. Cancer Res 39:2618 19. Imai K, Natali PG, Kay NE, Wilson BS, Ferrone S (1982) Tissue distribution and molecular profile of a differentiation antigen detected by a monoclonal antibody (345.134S) produced against human melanoma cells. Cancer Immunol Immunother 12:159 20. Imai K, Wilson BS, Bigotti A, Natali PG, Ferrone SA (1982) 94,000-Dalton glycoprotein expressed by human melanoma and carcinoma cells. J Natl Cancer Inst 68:761 21. Leon JA, Mesa-Tejada R, Gutierrez MC, Vita JR, Greiner JW, Schlom J, Fisher PB (1989) Increased surface expression and shedding of tumor associated antigens by human breast carcinoma cells treated with recombinant human interferons on phorbol ester tumor promoters. Cancer Res (in press) 22. Lloyd KO (1983) Human tumor antigens: detection and characterization with monoclonal antibodies. In: Basic and clinical tumor Immunology. Martinus Nijhoff, Boston, Mass, p 159 23. Loms Ziegler-Heitbrock HW, Munker R, Johnson J, Petersmann I, Schmoeckel C, Riethmuller G (1985) In vitro differentiation of human melanoma cells analyzed with monoclonal antibodies. Cancer Res 45:1344 24. Matsui M, Temponi M, Ferrone S (1987) Characterization of a monoclonal antibody-defined human melanoma-associated antigen susceptible to induction by immune interferon. J Immunol 139:2088 25. Moschera JA, Woehle, D, Tsai KP, Chen CH, Tarnowski SJ (1986) Purification of recombinant human fibroblast interferon produced in Escherichia coli. Methods Enzymol 119:177 26. Natali PG, Wilson BS, Imai K, Bigotti A, Ferrone S (1982) Tissue distribution, molecular profile and shedding of a cytoplasmic antigen identified by the monoclonal antibody 465.12S to human melanoma cells. Cancer Res 42:583 27. Natali PG, Aguzzi A, Veglia F, Imai K, Burlage RS, Giacomini P, Ferrone S (1983) The impact of monoclonal antibodies on the study of human malignant melanoma. J Cutaneous Pathol 10:514 28. Natali PG, Cavaliere R, Bigotti A, Nicotra MR, Russo C, Ng AK, Giacomini P, Ferrone S (1983) Antigenic heterogeneity of surgically removed primary and autologous metastatic human melanoma lesions. J Immunol 130:1462 29. Natali PG, Bigotti A, Cavaliere R, Nicotra MR, Ferrone S (1984) Phenotyping of lesions of melanocyte origin with monoclonal antibodies to melanoma-associated antigens and to HLA antigens. J Natl Cancer lnst 73:13 30. Natali PG, Bigotti A, Cavaliere R, Liao S-K, Taniguchi M, Matsui M, Ferrone S (1985) Heterogeneous expression of melanoma-associated antigens and HLA antigens by primary and multiple metastatic lesions removed from patients with melanoma. Cancer Res 45:2883 31. Natali PG, Bigotti A, Cavaliere D, Ruiter DJ, Ferrone S (1988) HLA class II antigens synthesized by melanoma cells. Cancer Rev 9:34 32. Pellegrino MA, Ng AK, Russo C, Ferrone S (1982) Heterogeneous distribution of determinants defined by monoclonal antibodies on HLA-A and B antigens bearing molecules. Transplantation 34:18 33. Quaranta v, Pellegrino MA, Ferrone S (1981) Serological and immunochemical characterization of the specificity of four monoclonal antibodies to distinct antigenic determinants expressed on subpopulations of human Ia-like antigens. J Immunol 126:548 34. Real FX, Houghton AN, Albino AP, Cordon-Cardo C, Melamed MR, Oettgen HF, Old LJ (1985) Surface antigens of

268 melanomas and melanocytes defined by mouse monoclonal antibodies: specificity analysis and comparison of antigen expression in cultured cells and tissues. Cancer Res 45:4401 35. Watson AJ, DeMars R, Trowbridge IS, Bach FH (1983) Detection of a novel human class II HLA antigen. Nature 304: 358 36. Wilson BS, Imai K, Natali PG, Ferrone S (1981) Distribution and molecular characterization of a cell-surface and a cytoplasmic antigen detectable in human melanoma cells with monoclonal antibodies. Int J Cancer 28: 293 37. Ziai RM, Imberti L, Tongson A, Ferrone S (1985) Differential modulation by recombinant immune interferon of the expression and shedding of HLA antigens and melanoma asso-

ciated antigens by a melanoma cell line resistant to the antiproliferative activity of immune interferon. Cancer Res 45: 5877 38. Ziai MR, Imberti L, Kobayashi M, Perussia B, Trinchieri G, Ferrone S (1986) Distinct functional domains on the recombinant human immune interferon molecule. Cancer Res 46: 6187

Received 26 April 1989/Accepted 13 July 1989