[CANO-R RESEARCH53. «I.V-IOT.September15. ITO]

Differential Effect of Human and Mouse ß2-Microglobulinon the Induction and theAntigenic Profile of Endogenous HLA-A and -B Antigens Synthesized byß2-Microg!obulinGene-null FO-1 Melanoma Cells1

Zhigang Wang, Xian-zhen Hu, Revati J. Tatake, Son Y. Yang, Richard A. Zeff, and Soldano Ferrone2

Department of Microbiology and Immunology. New York Medical College, Valhalla, New York /0595 ¡7..W., X-z. H., S. F.]; Department of Pathology. University of ConnecticutHealth Center, Fartninclon, Connecticut 06032 fR. ./. T., R. A. Z.]; and Department of Immunolocv, Memorial Sloan-Kettering Cancer Center, New York, New York 10021

IS. Y. Y./

ABSTRACT

ß2-Microglobulin (ß2-/u)gene-null human melanoma FO-1 cells display

lower reactivity with mili III \ class I monoclonal antibodies (mAb) following transfection with a wild-type mouse /¡r/<gene (referred to asFO-1C cells) than following transfection with a wild-type human |>r/'gene (referred to as FO-1H cells). Furthermore, binding assays with apanel of anti-HLA class I mAb detected higher reactivity of FO-1C cellswith mAb TP25.99 than with mAb CR1-S63, CR10-215, CR11-115, TP67,and W6/32 but similar reactivity of FO-1H cells with all the mAb tested.While mAb TP25.99 recognizes a determinant expressed on ß2-/i-freeandß2-/i-associated HLA class I heavy chains, the remaining mAb recognizedeterminants expressed only on ß2-p-associated HLA class I heavy chains.The differential effects of mouse and human ß2-juon the reactivity withanti-HLA class I mAb of FO-1 cells reflect more than one mechanism.

Besides abnormalities in the processing of HLA class I heavy chainsassociated with mouse (i..-/u. this molecular complex appears to be unstableon the plasma membrane of FO-1 cells.

To analyze the interaction of mouse /{>-/<with HLA-A and -B antigens,the HLA phenotype of FO-1 cells was determined, using a combination ofisoelectric focusing analysis of antigens immunoprecipitated from radio-

labeled cells with mAb to monomorphic determinants of HLA class Iantigens, binding assays with a limited number of mAb recognizing HLAclass I allospecificities, and sequence-specific oligonucleotide probe typing.Although association with mouse ß2-fidoes not cause marked differencesin the expression of HLA-A25 and -B8 antigens on the cell surface of FO-1

cells, it causes a selective reduction in the expression of determinantsrecognized by anti-HLA-A mAb F4/72 and VF19-LL67 and by anti-HLAclass I mAb W6/32 on HLA-A25 allospecificities. The differential effect ofthe association with mouse /!.-// on the antigenic profile of HLA-A25 and-B8 antigens may reflect the different characteristics of the amino acids at

residue 12, which interact with residue 33 of />r/i. The latter residue is theonly one to differ between human and mouse ß2-/iin the stretch of amino

acids interacting with the a, and a2 domains of HLA class I heavy chains.The present results indicate that caution should be exercised in the use ofHLA class I heavy chains associated with mouse ß2-/uto analyze peptide

presentation to cytotoxic T cells and of transgenic mice to establish animalmodels of HLA-linked diseases.

INTRODUCTION

HLA class I antigens consist of a polymorphic heavy chain (a M,41,000-45,000 glycopolypeptide) noncovalently associated withß2-(Li3(an invariant M, 12,000 polypeptide). Association with ß2-ju,is

required for efficient transport and expression of HLA class I heavy

Received 2/26/93; accepted 7/7/93.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely lo indicate this fact.

1This work was supported by USPHS Grant CA39559. awarded by the National

Cancer Institute, DHHS.2 To whom requests for reprints should be addressed, at Department of Microbiology

and Immunology, New York Medical College, Valhalla. NY 105<J5.'The abbreviations used are: ß:-n, fc-microglobulin: FITC-GAM. fluorescein iso-

thyocyanate-conjugated goat anti-mouse immunoglobulin antibodies; 1ER isoelectric focusing; IFN--y. y intcrferon: [IF, indirect immunofluorescence; mAb, monoclonal anti-

body(-ies); PBS. phosphate-buffered saline; PCR. polymcrase chain reaction; SSOP,sequence-specific oligonucleotide probe; TMAC, tetramethylammonium chloride; /(„,affinity constant; SOS. sodium dodecyl sulfate; PAGE, polyacrylamidc gel electrophore-sis; cDNA. complementary DNA.

chains on cell membrane and for peptide binding (1-3). Furthermore,association with ß2-/Mmodulates the antigenic profile of HLA class I

heavy chains. To the best of our knowledge, all but two (4, 5) antibodies elicited to ß2-/j,-associatedHLA class I heavy chains do notreact with ß2-fj.-freeHLA class I heavy chains. Conversely, it is ageneral experience that antibodies elicited to ß2-ju.-freeHLA class Iheavy chains do not react with ß2-/j.-associatedHLA class I heavy

chains (6, 7).The structural characteristics of ß2-/i have been maintained

throughout phylogenetic evolution, because human and mouse ß2-;n

display only 30% amino acid sequence divergence, distributedthroughout the mature polypeptide (8). This structural homology accounts for the ability of mouse ß2-ju.to associate with HLA class I

heavy chains and to support their transport and expression on the cellsurface. However, analysis of HLA class I allospecificities expressedby mouse cells transfected with the corresponding genes has shownthat the cells may display changes in their reactivity patterns withmAb and/or may display a low level of expression (9-12). The latter

findings are likely to be caused by the structural differences betweenhuman and mouse ß2-/x,since in a recent study ß2-ju,gene-null FO-1melanoma cells were found to display reduced reactivity with anti-HLA class I mAb following transfection with a wild-type mouse ß2-ju.gene (13). Furthermore, the results of the experiments with FO-1 cells

argue against the interference of other gene products which regulatethe assembly of allelic HLA class I heavy chains with mouse ß2-ju.(14) and/or competition for endogenous ß2-jain mouse cells as alter

native mechanisms underlying the reduced cell surface expression ofmouse ß2-/j.-associatedHLA class I antigens.

Whether the low reactivity with anti-HLA class I mAb of FO-1 cellstransfected with a wild-type mouse ß2-/u.gene reflects reduced cell

surface expression of HLA class I molecules and/or a change inexpression of the determinants recognized by the anti-HLA class ImAb tested and whether mouse ß2-/nhas different effects on the geneproducts of HLA-A and -B loci is not known. Therefore, in the presentstudy we have analyzed the expression of the gene products of HLA-Aand -B loci and their antigenic profiles in FO-1 melanoma cellstransfected with wild-type mouse and human ß2-jugenes. This infor

mation will also determine the validity of studies performed with HLAclass I heavy chains associated with mouse ß2-/J.to analyze peptide

presentation to cytotoxic T lymphocytes, to generate mAb to HLAclass I allospecificities, and to establish models of HLA-linked dis

eases in transgenic mice (for review, see Ref. 15). Furthermore, toanalyze the interaction of mouse ß2-/nwith distinct HLA class Iallospecificities, we have determined the HLA phenotype of FO-1

cells utilizing a combination of IEF analysis of HLA antigens immunoprecipitated from FO-1 transfectant cells, binding assays with mAb

recognizing HLA allospecificities, and SSOP typing.

MATERIALS AND METHODS

Cell Lines. The melanoma cell line FO-1 (16) and the FO-1 cell linestransfected with pSV2neo alone (FO-1 neo) (16), mouse ß2-/j.gene andpSV2neo (FO-1C) (16), or human ß,->xgene and pSV2nco (FO-1H) (13) were

4303

Research. on October 18, 2020. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

AND HLA EXPRESSION BY MELANOMA CELLS

cultured in Dulbecco's modified Eagle's medium (GIBCO Laboratories, Grand

Island, NY) supplemented with 10% fetal bovine serum and antibiotics.mAb and Conventional Antisera. mAb CR1-S63, CR10-215, CR11-115,

TP67, and W6/32 to monomorphic determinants expressed on ß2-/x-associatedHLA class I heavy chains (17-19), mAb TP25.99 to a determinant expressedon both ß2-/j,-associated and ß2-;j.-freeHLA class I heavy chains (5), anti-HLA-B mAb H2-89-1 and Q6/64 (20, 21), anti-HLA-Awl9, -B, and -C mAb4E (22), anti-HLA-B8 mAb P8.1 (23), anti-ß^-free HLA class I heavy chainmAb HC-10 (7), anti-human ß2-/xmAb NAMB-1 (17), and anti-mouse ß2-/J.mAb S 19.8 (24) were developed as described. Anti-HLA-A mAb F4/72 andVF19-LL67 were elicited with phytohemagglutinin-activated peripheral blood

T cells and with IFN-y-treated cultured SK-OV-3 human ovarian carcinomacells, respectively. Anti-HLA-B mAb SA24/23 was elicited with cultured

Epstein-Barr virus-transformed B cells. The specificity of the latter three mAb

was characterized with a combination of serological and immunochemicalassays. mAb were used as ascites or culture supernatants. mAb were purifiedfrom ascitic fluid by sequential precipitation with caprylic acid and ammoniumsulfate as described (25). mAb were radiolabeled with I25I using the lodogen

method (26).Purified rabbit anti-mouse immunoglobulin antibodies were purchased from

Jackson ImmunoResearch Laboratories (West Grove, PA). FITC-GAM were

purchased from Boehringer Mannheim Biochemicals (Indianapolis, IN).Cytokine. Recombinant human IFN-y was obtained from Hoffmann La

Roche Inc. (Nutley, NJ).cDNA Probes and Oligonucleotides. The HLA-B7 cDNA probe (27) was

obtained from Dr. S. M. Weissman, Yale University School of Medicine (NewHaven, CT), and the HLA-A- and HLA-B-specific cDNA probes pHLA-2a.land pHLA-1.1 (28) were obtained from Dr. H. T. Orr, University of Minnesota,

School of Medicine (Minneapolis, MN). cDNA probes were radiolabeled with[a-32P]dCTP (>600() Ci/mmol; Amersham Corp., Arlington Heights, IL) byrandom priming (29) to a specific activity of IO8 cpm//j,g. The oligonucleotide5'-TAGAAGACCAGTCCTTGCTGAA-3' (referred to as MH), corresponding

to an exon 2 region identical in human and mouse ß2-/genes, was constructed

on a BioSearch Cyclone DNA synthesizer (MilliGen/Biosearch, Burlington,MA). HLA-A locus-specific 5' primer (5'-GACGCCGCGAGCCAGAGGAT-3') corresponding to codons 39-44 in exon 2, 3' primer (5'-TGCAGC-GTCTCCTTCCCGTT-3') complementary to codons 174-179 in exon 3 of

HLA class I heavy chains, and Oligonucleotides complementary to specificsequences of HLA-A 1, -A25, and/or -A26 alíeles,referred to as SSOP, were

purchased from Oligos Etc. Inc. (Ridgefield, CT). Oligonucleotides were radiolabeled with [y-32P]ATP (>5000 Ci/mmol; Amersham) in the presence of

T4 polynucleotidc kinase (Bethesda Research Laboratories, Gaithersburg,MD).

Serological Assays. IIP was performed by incubating cells (1 x 10'') withan excess of the appropriate mAb for 45 min at 4°C,with occasional shaking.

Following washing with cold PBS containing 0.02% sodium azidc, cells wereincubated with FITC-GAM for 45 min at 4°C.Cells were then washed twice

with PBS-azide, fixed in 3% formaldehyde, and analyzed by cytofluorometry

on a FACS Analyzer (Becton Dickinson, Mountain View, CA). The nonspecificfluorescence was determined with cells incubated with FITC-GAM alone.

Results are expressed as log fluorescence intensity.Binding assays with radiolabeled mAb were performed as described else

where (30). Results of binding assays were converted into Scatchard plots todetermine K¡,and the number of binding sites/cell. K¿was determined from theslope of the straight line plot, and the number of binding sites/cell was obtainedfrom the intercept on the jr-axis.

Immunochemical Methods. Radiolabeling of cells with I25I (Amersham)utilizing the lactoperoxidase method (31) or with [-15S]methionine (>1000

Ci/mmol; Amersham), solubilization with 1% Nonidet P-40 or 0.5% TritonX-114, indirect immunoprecipitation, SDS-PAGE, and two-dimensional gelelectrophoresis (IEF in the first dimension and SDS-PAGE in the second

dimension) were performed as described (16, 32, 33). Ampholines (PharmaciaLKB Biotechnology AB, Bromma, Sweden) of pH 3.5-10.0, pH 5.0-7.0, andpH 6.0-8.0 were used in the ratio of 4:3:3. One-dimensional IEF for HLA class

I typing was performed as described (33), utilizing HLA class I antigens treatedwith neuraminidase type VIII (200 milliunits/sample) and a vertical slab gelapparatus. Gels were processed for fluorography as described (34).

Northern Blot Analysis. Total RNA extraction, electrophoresis, blotting,hybridization, and autoradiography were performed as described elsewhere(16). Blots were reused after the probes had been removed by incubation in a

solution containing 0.1 X standard saline citrate (1 X = 0.15 Msodium cholride,0.015 M sodium citrate, pH 7.0) and 0.1% SDS at 98°Cfor 10 min. The

temperature for hybridization with Oligonucleotides was calculated utilizingthe formula 7d = 4(G + C) + 2(A + C) - 5°C.After prehybridization for 2-3

h, filters were probed with MH oligonucleotide in prehybridization solution for4 h. Following three washes with IX standard saline citrate, 0.1% SDS, for 1h at the hybridization temperature, blots were dried and autoradiographed at-70°C for 3 days using XAR-5 film (Eastman Kodak, Rochester, NY) and an

intensifying screen (DuPont Co., Wilmington, DE).Oligonucleotide Typing by SSOP. Oligonucleotide typing of HLA-A an

tigens was performed as described (35). Genomic DNA was prepared fromFO-1, FO-1C, and FO-1H cells (2 X IO5 cells) and from 50 pii of whole blood

from HLA-typed donors by the cell lysis method using Tween-20 followed by

proteinase K digestion (36). The DNA fragments of HLA-A alíeleswhich

included exon 2, intron 2, and exon 3 were amplified by Taq DNA polymerasein the presence of HLA-A locus-specific 5' and 3' primers, using PCR. Am

plification was performed using a DNA Thermal Cycler (Perkin-Elmer Cetus,Norwalk, CT) for 30 cycles. A cycle profile consisted of 1 min at 96°Cfordenaturation, 20 s at 60°C for annealing, and 1 min at 72°C for primerextension. In the first cycle DNA was denatured at 94°Cfor 5 min. After thelast cycle, all samples were incubated for an additional 5 min at 72°Cto ensure

the completion of the final extension step. An aliquot of each PCR product wasanalyzed by 1% agarose gel electrophoresis, to monitor the extent of amplification and contamination. One (¿Iof each PCR product was applied to a nylonmembrane using an eight-channel syringe. Membranes were soaked for 5 min

in denaturation buffer containing 0.4 N NaOH and for 10 min in neutralizingbuffer containing 10X standard saline-phosphate-EDTA (IX is LSmNaCl, 0.1

M NaH2POj, 10 HIMEDTA). After drying on 3 MM filter paper, membraneswere baked at 80°Cfor 1 h in a vacuum oven. Prehybridization was performedat 46°Cfor 1 h in hybridization buffer (50 mM Tris-HCI, pH 8.0, 3 MTMAC,2 mm EDTA, 5X Denhardt's solution, 0.1% SDS, 100 u.g/ml herring spermDNA). All 12P-labeled SSOP were hybridized with membranes at 46°Cfor 2

h, since hybridization was performed in the presence of TMAC (37) utilizingprobes which were all 15 bases long. After hybridization, membranes wererinsed twice in 2X standard saline-phosphate-EDTA, 0.1% SDS, at roomtemperature for 10 min. Also, membranes were washed in 50 mMTris-HCI, pH

8.0, 3 MTMAC, 2 mm EDTA, 0.1% SDS, for 10 min at room temperature andfor 10 min at 50°C.Membranes were then exposed to Kodak X-AR film at-70°C for up to 16 h, with two intensifying screens.

RESULTS

Lower Reactivity of FO-1C Cells than of FO-1H Cells withAnti-HLA Class I mAb. FO-1C cells, which express HLA class Iantigens following transfection with a mouse ßi-ju-gene, were stainedin IIP to a lesser extent by anti-HLA class I mAb TP25.99 and W6/32than were FO-1H cells, which express HLA class I antigens followingtransfection with a human ß2-Mgene (Fig. 1). Following a 48-hincubation at 37°Cwith IFN-y (100 units/ml), the reactivity of mAb

TP25.99 with FO-1C cells was markedly increased, while that of mAbW6/32 was only weakly enhanced. The reactivity of IFN-y-treatedFO-1H cells with mAb TP25.99 and W6/32 was increased to similarextents. To quantitate the differences in the reactivity of FO-1 C andFO-1H cells with the two anti-HLA class I mAb, a Scatchard plotanalysis of the binding of 125I-mAb TP25.99 and 125I-mAb W6/32 to

IFN-y-treated and control FO-1 C and FO-1H cells was performed

(Table 1). This analysis, consistent with the results of IIP staining,demonstrated that the number of mAb TP25.99 molecules bound bycontrol and IFN-y-treated FO-1 C cells was 24.5 and 43.5% of thenumber of mAb TP25.99 molecules bound by control and IFN-y-treated FO-1H cells. The number of mAb W6/32 molecules bound bycontrol and IFN-y-treated FO-1 C cells was about 14% of the numberof mAb W6/32 molecules bound by FO-1H cells. The number ofbinding sites for mAb TP25.99 was 3-fold more than that for mAbW6/32 on control and IFN-y-treated FO-1C cells but was only 1.7-fold more on control FO-1H cells and was similar to that for mAbW6/32 on IFN-y-treated FO-1H cells. The Ka values of mAb TP25.99for control and IFN-y-treated FO-1 C and FO-1H cells were similar;

4304

Research. on October 18, 2020. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

ß,-H AND HLA EXPRESSION BY MELANOMA CELLS

W6/32 TP25.99 NFiMB- 1 SI 9.8 Contro I IFN-y

ALL

AA.

Lag Fluorescence Intensi ty

Fig. 1. Expression and modulation by IFN-y of HLA class I antigens on FO-1 cellstransfcctcd with a wild-type mouse (FO-1C) or human (FO-1H) ßi-/j,gene. Cells wereincubated a! 37°Cwith human recombinant IFN-y (100 units/ml) for 48 h. Control cellswere incubated under the same experimental conditions without IFN-y. After cells weredetached from the flasks with PBS containing 1 HIMEDTA and were washed with Hanks'

balanced salt solution, they were incubated with mAb W6/32 to a determinant expressedby ßi-fi-associated HLA class I heavy chains, with mAb TP25.99 to a determinantexpressed by ß;-(i-associated and ß2-u.-frccHLA class I heavy chains, with anti-humanß^-fi mAb NAMB-1, or with anti-mouse ß2-ßmAb SI9.8. Following washing withPBS-azide, cells were stained with FITC-GAM. The nonspecific fluorescence representscells treated with FITC-GAM alone. Fluorescence was determined on a FACS Analyzer

(Becton Dickinson).

Table 1 Characteristics of Ihe binding of anli-HLA class I mAb TP25.W and W6I32 toFO-IC and FO-IH melanoma cells

FO-1C and FO-1H cells were incubated for 72 h at 37°Cwith IFN-y (final concen

tration, 1000 units/ml). Control cells were incubated under the same experimental conditions without IFN-y. Cells (1 X 10s) were mixed with serial dilutions of I25l-labeled

mAb in a l(X)-(il total volume of PBS containing 0.5% bovine serum albumin, in %-wellV-bottomed microtiter plates. Radioactivity hound to cells was counted in a gamma-counter (LKB-12f>l; LKB-Wallac. Türkin.Finland) after unbound radioactivity had beenremoved by washing. Nonspecific binding was measured in the presence of a > 100-fold

excess of unlabeled mAb. Results were plotted in the form of the linear regressiontransformation of the law of mass equilibrium (38).

mAbTP25.99Cell

lineFO-ICFO-IHNo.

ofmoleculesbound/cell

IFN-y (X10-5)"(2.95r+

13.4012.00+

30.80K.,

X10gM-1)'11.151.280.911.29mAb

W6/32No.

ofmoleculesbound/cell

(X10*5)1.084.207.4029.60Kt(X109M~')0.250.590.540.53

" Number of binding sites/cell was obtained from the intercept on the jt-axis.'' The K„was determined from the slope of the straight line plot.c The data shown are the mean values of triplicates from two experiments. The SD are

the Ka of mAb W6/32 for FO-IH cells was higher than that for FO-1 Ccells, but the values were similar for IFN-y-treated FO-1 C and FO-IH

cells (Table 1).Because of the differential reactivity with mAb TP25.99 and W6/

32, FO-IC and FO-IH cells were tested in IIF with mAb CR1-S63,CR10-215, CR11-115, and TP67, which recognize monomorphic

determinants of HLA class I antigens (Table 2). The four mAb stainedcontrol and IFN-y-treated FO-IH cells strongly. In contrast, the fourmAb did not stain or stained weakly FO-IC cells but stained moderately FO-IC cells which had been incubated with IFN-y (100 units/

ml) for 48 h.The differential reactivity with mAb TP25.99 and W6/32 of FO-IC

and FO-IH cells prompted us to analyze the expression of ß2-;u-free

HLA class I heavy chains by the two transfectants, since mAbTP25.99 recognizes a determinant expressed on ß2-/j,-freeand ß2-/x-associated HLA class I heavy chains (5). FO-IC cells displayed aslightly lower reactivity with anti-ß2-ju,-freeHLA class I heavy chainmAb HC-10 than did FO-IH cells. Following a 48-h incubation withIFN-y (100 units/ml) FO-IC and FO-IH cells displayed similar reactivities with mAb HC-10 (Fig. 2). These findings, in conjunction with

the reduced expression of ß,-/j.-boundHLA class I heavy chains onFO-IC cells, may reflect greater dissociation of mouse ß2-jxfrom the

HLA class I molecular complex on the cell membrane.Expression of Mouse and Human (!r¿uand HLA Class I Heavy

Chain mRNA in FO-IC and FO-IH Cells. To exclude the possibility that differences in relative expression of HLA class I antigens bycontrol and IFN-y-treated FO-IC and FO-IH cells reflected differ

ences in the expression of the transfected genes, or differential induction of the ß2-/xgeneexpression in FO-IC and FO-IH cells by IFN-y,the level of mRNA for ß2-jo,was measured using a synthetic oligo-nucleotide which hybridizes to both human and mouse ß2-jxmRNA.As shown in Fig. 3, IFN-y was equally effective at enhancing thelevels of ß2-nmRNA in FO-IC and FO-IH cells, although the baselevel of mouse ß2-jJ.mRNA in FO-IC cells was slightly lower thanthat of human ß2-jamRNA in FO-IH cells. The steady state IFN-y-

induced levels of HLA class I heavy chain mRNA were similar in eachtransfectant (Fig. 3).

Differential Serological Reactivity of FO-IC and FO-IH Cellswith Anti-HLA-A and Anti-HLA-B mAb. To determine whetherthe low reactivity of FO-IC cells with anti-HLA class I mAb reflected

Table 2 Differential reactivity of anti-HLA class I mAb with control and IFN-y-treatedFO-IC and FO-IH cells

FO-IC cells and FO-IH cells were incubated for 48 h at 37°Cwith IFN-y (final

concentration, 100 units/ml) ( + IFN-y). Control cells were incubated under the sameexperimental conditions without IFN-y (-IFN-y). Cells were incubated with an excess ofeach mAb for 45 min at 4"C, with occasional shaking. Cells were then washed with cold

PBS containing 0.02% sodium azide and were incubated with FITC-GAM for 45 min at4°C.Cells were washed twice with PBS-azide and fixed in 3% formaldehyde prior to

analysis.

Positive cells (%)

FO-IC cells FO-IH cells

mAbCR1-S63

CR 10-2 15CR11-115TP67Control''-IFN-y52.2

(4.3)"

63.9 (4.3)44.8 (4.2)18.5(4.1)9.0 (4.2)-HFN-y96.1

(13.1)95.2(19.7)96.5(10.9)96.2(12.4)

6.2 (4.9)-IFN-y98.7

(66.6)99.0 (76.3)98.6 (52.2)95.0(15.9)20.2 (4.1)-HFN-y98.7(191.6)

98.9(231.6)99.5(127.6)97.9 (80.5)11.2 (4.0)

" Values in parentheses represent mean channel fluorescence determined on a FACS

Analyzer (Bccton-Dickinson).'' Control indicates cells which were treated with the secondary antibody (FITC-GAM)

alone.

-IFN-Y +IFN-Y

v0

mfib

HC-10

TP25.QO

Log Fluorescence Intensity

Fig. 2. Expression of ß2-(i-freeHLA class I antigens on FO-IC and FO-IH cells.FO-1C(- •¿�•¿�•¿�)and FO-IH ( ) cells were incubated with lFN-y( 100 units/ml) at 37°C

for 48 h. Control cells were incubated under the same experimental conditions withoutIFN-y. After cells were detached from the tlasks with PBS containing 1 mw EDTA andwere washed with Hanks' balanced salt solution, they were incubated with anti-ß;>-u,-frce

HLA class I heavy chain mAb HC-10 and mAb TP25.99 to a determinant expressed byß2-/i-freeand ß2-u,-associatedHLA class I heavy chains. Following washing with PBS-azide, cells were stained with FITC-GAM. Fluorescence was determined on a FACSAnalyzer (Becton Dickinson). The nonspecific fluorescence represents cells incubatedwith FITC-GAM alone ( ).

4305

Research. on October 18, 2020. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

AND HI.A EXPRESSION BY MELANOMA CELLS

Heavy Cha In »H»Human B2m Gene - + --- + --

Mouse B2m Gene + - - + -

pSV2neo + + +-+ + +-

Fig. 3. Expression and modulation by IFN--y of ß:-n and HLA class I heavy chainmRNA in FO-1C and FO-1H cells. FO-1C. FO-1H. FO-1. and FO-lneo cells wereincubated at 37°Gwith IFN-y ( KM)untis/ml) for 4S h. Control cells were incubated under

the sameexperimental conditions hut without IFN-y. Total cellular RNA (20 u.g) was sizefractionated on a lr/r agarosc gel containing 2.2 M formaldehyde and was transferred tonitrocellulose membranes. 12P-labcled oligonuclcotide probe MH and ':P-labeledH1.A-B7 cf)NA probe were used for hybridization to mouse and human ß:-/imRNA andHLA class I heavy chain mRNA. respectively.

reduced expression of both HLA-A and -B antigens or a selectivereduction in the expression of the gene products of the HLA-A or -Bloci, the reactivity of FO-1 C cells with anti-HI.A-A and anti-HLA-BmAh in IIP was tested and compared with that of FO-1 H cells (Fig. 4).The anti-HLA-B mAh H2-89-1, 06/64, and SA24/23 stained bothIFN-y-treated and control FO-1 C and FO-1 H cells; the intensity ofstaining of FO-1 H cells was brighter than that of FO-1 C cells. Incontrast, the anti-HLA-A mAb F4/72 stained neither IFN-y-treatednor control FO-1 C cells but stained both IFN-y-treated and controlFO-1H cells. The anti-HLA-A mAb VF19-LL67 did not stain FO-1Ccells and stained only 16% of FO-1 H cells with low intensity. Furthermore, the latter mAb stained 55% of IFN-y-treated FO-1 C cellswith low intensity and 95% of IFN-y-treated FO-1 H cells with highintensity. These results indicate that IFN-y-treated FO-1 C cells express the gene products of both HLA-A and -B loci. The associationwith mouse ß;.- changes the antigenic profile of HLA-A molecules.

To determine whether the differential reactivity of FO-1 C cells withanti-HLA-A and anti-HLA-B mAb was caused by differences in thetranscription of HLA-A and -B genes, the steady state levels of mRNAfor HLA-A and HLA-B antigens in FO-1 C cells were compared. Theintensities of the components hybridizing with 12P-labeled probes

pHLA-2a.l and pHLA-1.1 were similar, indicating that the HLA-Aand -B genes are transcribed at similar levels (data not shown).

Level of HLA-A and -B Heavy Chain Expression by FO-1 C andFO-1H Cells. To determine whether the differential reactivity ofFO-1 C cells with anti-HLA-A and anti-HLA-B mAb reflected differences in the expression of HLA-A and HLA-B heavy chains or of thedeterminants recognized by the mAb used, HLA-A and HLA-B heavychains were immunoprecipitated from [vsS|methionine- and '-5I-la-

beled FO-1C melanoma cells with the anti-mouse ß;>-jumAb S19.8and analyzed by two-dimensional gel electrophoresis. For comparisonpurposes. HLA-A and HLA-B heavy chains were also immunoprecipitated with the anti-human ß^-jtxmAb NAMB-1 from [^S]methio-nine- and I2sl-labeled FO-1 H cells and analyzed by two-dimensional

gel electrophoresis. Representative results are shown in Fig. 5. Theintensities of the basic spots which represent the gene products of theHLA-A loci were higher than those of the acidic spots which representthe gene products of the HLA-B loci, especially when FO-1 C cells

were used as an antigen source (Fig. 5, a and c). but intensities weresimilar when both IFN-y-treated FO-1 C and FO-1 H cells were used as

an antigen source (Fig. 5, h and d). These results indicate that thedifferential reactivity of FO-1 C cells with anti-HLA-A and anti-HLA-B mAb reflects a lack or reduction of expression of the determinants recognized by the anti-HLA-A mAb tested.

HLA Class I Phenotype of FO-1 Cells. To identify the HLA class

I allospecificity with lacking or low expression of the determinantsdefined by anti-HLA-A mAb F4/72 and VF19-LL67, the HLA class Iphenotype of FO-1 cells was determined. Since the complement-

dependent cytotoxic assay is not a reliable procedure for HLA typingof cells in long term culture, the HLA class I phenotype of FO-1 cellswas determined by IFF. To facilitate the analysis, IFN-y-treatedFO-1 C and FO-1 H cells, which express an increased level of HLA

class I antigens, were used as an antigen source. The electrophoreticpattern of HLA class I heavy chains isolated from FO-1 C and FO-1 Hcells is compatible with the presence of HLA-A1, HLA-A25, orHLA-A26 alloantigens. The electrophoretic pattern of HLA-B heavychains was compatible with the presence of HLA-B8 antigens (Fig. 6).

It is noteworthy that the intensity of the components corresponding tothe gene products of HLA-A loci immunoprecipitated by mAb W6/32from FO-1 C cells was markedly lower than that of the moleculesimmunoprecipitated from FO-1 H cells. In contrast, the intensity of theHLA-B alloantigens immunoprecipitated from FO-1 C cells by mAbW6/32 was only slightly lower than that of those from FO-1 H cells;the intensities of HLA-B antigens immunoprecipitated from FO-1 Cand FO-1 H cells by mAb 4E were similar (Fig. 5). The expression ofHLA-B8 antigens by FO-1 C and FO-1 H cells was confirmed by IIFstaining with the anti-HLA-B8 mAb P8.1. mAb with the appropriatespecificity to discriminate between HLA-A 1, -A25, and -A26 antigenswere not available to us. Therefore, FO-1 cells were typed by SSOPanalysis. All the probes complementary to HLA-A25 alíele,to bothHLA-A25 and -A26 alíeles,and to both HLA-A1 and -A25 alíeleshybridized to DNA amplified by PCR from FO-1, FO-1C, and FO-1 Hcells, while probes complementary to HLA-A 1 and/or -A26 alíelesdidnot (Table 3). Therefore, FO-1, FO-1C, and FO-1H cells expressHLA-A25 and -B8 specificities.

- IFN-y + IFN-Y mflb

F4/72

VFI9-LL67

H2-88-I

08/64

SR24/23

Log Fluorescance Intensity

Fig. 4. Differential reactivities of FO-1C and FO-1H cells with anti-HLA-A and anti-HLA-B mAb. FO-1 («).FO-lC(h). and FO-1H (c) cells were incubated with IFN-y (100units/ml) al 37°Cfor 48 h. Control cells were incubated under the same experimental

conditions without IFN-y. After cells were detached from the flasks with PBS containingI mM EDTA and were washed with Hanks" balanced salt solution, they were incubated

with anti-HI.A-A mAb F4/72 and VF19-LL67 and anti-HLA-B mAb H2-X9-1. Ofi/64.and SA24/23. Following washing with PBS-azide. cells were stained with FITC-GAM.Fluorescence was determined on a FACS Analyzer (Becton Dickinson).

4306

Research. on October 18, 2020. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

ß:-H AND HLA EXPRESSION BY MELANOMA CELLS

lEF

Fig. 5. Two-dimensional gel electrophoresis

analysis of HLA class I antigens expressed byFO-1C and FO-1H cells. FO-1C (a and fc) andFO-IH (r and d) cells were incubated with IFN--y(100 units/ml) at 37°Cfor 48 h (b and d). Control

cells (a and c) were incubated under the same experimental conditions but without IFN-y, and cellswere then surface labeled with '-5I using the lac-toperoxidase method (31). Lysates from FO-1Ccells were immunoprecipitated with anti-mouseß2-fimAb S 19.8 (a and h). Lysates from FO-IHcells were immunoprecipitated with anti-humanßi-p.mAb NAMB-1 (c and d). 1EF was from right

to left (acidic to basic) in the first dimension, andSDS-PAGE was from lop to bottom in the seconddimension on 10% polyacrylamide gels. A, HLA-A-related spots; B. HLA-B-related spots. Arrow

heads, human

a

DISCUSSION

Transfection with a wild-type human or mouse ß2-/xgene inducesHLA class I antigen expression by FO-1 melanoma cells, which do not

express these antigens because of a structural abnormality of theirendogenous ß2-|ugene(s) (16). The level of HLA class I antigens, as

indicated by the extent of binding of mAb recognizing monomorphicdeterminants of HLA class I antigens and by the intensity of components corresponding to HLA class I heavy chains immunoprecipitatedby anti-mouse and anti-human ß2-ju,mAb, is markedly higher in FO-1cells transfected with a human ß2-/Ligene than in FO-1 cells trans-fected with a mouse ß2-/xgene. This difference is likely to reflect

more than one mechanism. Besides abnormalities in the processing ofHLA class I heavy chains by mouse ß2-ju,transfectant FO-1 cells (13),the recently described dissociation of ß2-fj,from major histocompat-ibility complex class I heavy chains on mouse EL4 and RAM-S cells(39) suggests that mouse ß2-/xdissociates from the HLA class I

molecular complex inserted in the cell membrane because of the lowstability of the complex formed by HLA class I heavy chains bound tomouse ß2-pi-This possibility is supported by the following lines of

evidence. The cell surface expression of HLA class I antigens bymouse ß2-fAtransfectant FO-1 cells is increased by incubation at30°C,which has been shown to enhance the stability on the cell

membrane of mouse ß2-/ji-associatedHLA class I heavy chains (13).Furthermore, FO-1C and FO-IH cells react to similar extents withanti-ß,-/x-free HLA class I heavy chain mAb HC-10, although FO-IHcells display higher reactivity with mAb reacting with ß2-/j,-associatedHLA class I heavy chains. The expression of ß2-/x-freeHLA class Iheavy chains can also account for the higher reactivity of FO-1 C cellswith mAb TP25.99 than with mAb CR1-S63, CR10-215, CR11-115,

TP67, and W6/32. While the latter five mAb recognize determinantsexpressed only on ß2-ju.-associatedHLA class I heavy chains (17-19,

40), mAb TP25.99 recognizes a determinant which is expressed onboth ß2-/j,-associatedHLA-A, -B, and -C heavy chains and ß2-/j,-freeHLA-B and -C heavy chains (5). The reduced stability of the complex

CMn

LU CD LU

CVJnco

human

R25-

B8-

Fig. 6. Analysis by one-dimensional IEF gel electrophoresis of HLA class I allospeci-ficities immunoprecipitated from FO-IC and FO-IH cells. Following a 48-h incubationwith IFN-y (1(K) units/ml), FO-IC and FO-IH cells (5 X 10ft) were metabolically labeledwith [^Sjmethionine. Labeled cells were lysed with 0.5% Triton X-114 and cellularlysates were mixed with 0.5% Nonidet P-40. HLA class I antigens were immunoprecipitated with anti-HLA-B and -C mAb 4E and with anti-HLA class I mAb W6/32. IsolatedHLA class I antigens were treated with neuraminidase and analyzed by one-dimensional

IEF gel electrophoresis using a vertical slab gel apparatus. The gel was then processed forfluorography.

4307

Research. on October 18, 2020. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

ß2-nAND HLA EXPRESSION BY MELANOMA CELLS

Table 3 Sequence-specific oligonucleotide probes for HLA-A1, -A25, and -A26 typing

Probe-aï

domain62RN76AN77S78R«2

domain11401I4R131

R144Q149T150V152EI5AWI63R166DGAmino

acidsequenceDRNTRDRANLESLRILRIALYQQDAGYRQDRSWTAQITQRWETAHEAVHAAHEAEEQWRAEGRCVVDGLRHLA''

Niirlentirle sequenceFfl-1(5'-3r)Al A25 A26cellsrGACCGGAACACACGG

+ ++GACCGAGCGAACCTG

++GAGAGCCTGCGGATC

++TGCGGATCGCGCTCC

++TACCAGCAGGACGCC

+ ++GGGTACCGGCAGGAC

+CGCTCTTGGACCGCG

+ + ++AGATCACCCAGCGCA

+ ++TGGGAGACGGCCCAT

+ ++GAGGCGGTCCATGCG

+GCCCATGAGGCGGAG+ -f+GAGCAGTGGAGAGCC

+ ++GAGGGCCGGTGCGTG+ + ++GTGGACGGGCTCCGC

+

" The probe designation indicates the position number in the amino acid sequence and the symbol of the amino acid residue that discriminates distinct HLA-A alíeles.h Complementarity of the nucleotide sequence of the probe to that of the indicated HLA-A alíelesis shown.r Hybridization in the dot blot assay of the indicated probes with DNA amplified from FO-1 cells is shown.

formed by HLA class I heavy chains bound to mouse ß2-/xmay reflectthe difference between human and mouse ß2-/xin the polarity of one

of the amino acids interacting with the a, and a2 domains of HLAclass I heavy chains (41). Residue 33 is the hydrophobic (nonpolar)amino acid proline in mouse ß2-/xbut the hydrophilic (polar) aminoacid serine in human ß2-M(8). An alternative but not exclusive

mechanism is represented by the reduced efficiency of the binding ofpeptides to the mouse ß2-pi-HLAclass I heavy chain complex, since

expression of this complex was increased by incubation of cells at lowtemperature (13). Incubation of cells at low temperature has beenshown to increase the expression of empty major histocompatibilitycomplex class I molecules (42).

The present study has also shown a differential effect of the association with mouse ß2-fion the antigenic profile of HLA-A25 and -B8allospecificities. HLA-B8 heavy chains associated with mouse ß2-/Mdisplay only moderately lower reactivity, compared with HLA-B8heavy chains associated with human ß2-/x,with anti-HLA-B mAbH2-89-1, Q6/64, and SA24/23 and with anti-HLA class I mAb W6/32.In contrast, HLA-A25 heavy chains associated with mouse ß2-p.display no or very low reactivity with anti-HLA-A mAb F4/72 andVF19-LL67 and with anti-HLA class I mAb W6/32. The latter resultsparallel the lacking or markedly low reactivity of anti-HLA-A3 mAbX1.23 with HLA-A3 antigens and of mAb W6/32 with HLA-A3, -B7,and -Cw3 antigens expressed by mouse cells transfected with the

corresponding HLA class I genes (10, 11). The differential ability ofHLA-A25 and -B8 allospecificities to tolerate the structural differences between mouse and human ß2-/xmay reflect the different char

acteristics of valine and methionine, which are present at residue 12 ofthe heavy chains of these two allospecificities (43). Residue 12 interacts with residue 33 of ß2-|t(41). The latter residue is the only onethat differs between human and mouse ß2-/xin the stretch of amino

acids which interact with the a, and a2 domains of HLA class I heavychains (8, 41). This mechanism is supported by the reduction in theexpression of the determinant defined by mAb W6/32 on the HLA-A3, -B7 and -Cw3 allospecificities when they are associated withmouse ß2-/i.Like HLA-A25 allospecificities, HLA-A3, -B7, and-Cw3 allospecificities express valine at position 12 (43). Furthermore,

the recent information about the mapping of the determinants recognized by mAb CR1-S63, CR10-215, CR11-115, TP67, and W6/32(20) indicates that association with mouse ß2-;u.may change the

expression of determinants located on a2 and a3 domains of HLAclass I heavy chains. Whether these changes have any effect onpeptide presentation and/or CDS binding is not known at the presenttime.

Interest in the characterization of HLA allospecificities expressedby malignant cells has been stimulated by at least two experimentalfindings. Malignant transformation of cells may be frequently associated with changes in their HLA phenotypes (for review, see Ref. 44).These changes may affect the interaction of tumor cells with the hostimmune system and the clinical course of the disease (for review, seeRef. 45). Additional impetus to analyze the HLA phenotype of tumorcells has been provided by the growing interest in the use of cytotoxicT cells recognizing tumor-associated antigens to develop immuno-

therapy of malignant diseases, since the success of these therapeuticapproaches requires expression of the restricting HLA class I allospecificities by tumor cells (for review, see Refs. 46). However, characterization of the HLA phenotype of cultured cells as well as ofmalignant cells isolated from surgically removed lesions has beenhampered by their abnormal susceptibility to complement-dependentlysis, which reduces the usefulness of the widely used complement-

dependent cytotoxic assay to identify HLA allospecificities on tumorcells. Furthermore, HLA phenotyping by absorption is restricted to afew allospecificities, since most of the available anti-HLA alloantisera

are not suitable for this technique because of the low titer and/oraffinity of anti-HLA alloantibodies. The present study has shown that

a combination of binding assays with the limited number of availablemAb recognizing HLA allospecificities, IEF of HLA molecules im-munoprecipitated from 12SI-labeled tumor cells with a mAb recogniz

ing a monomorphic determinant shared by the gene products ofHLA-A, -B, and -C loci, and oligonucleotide hybridization may over

come the limitations imposed by the previously used techniques.Furthermore, the three methods we have used complement each other,so that their combination provides reliable information about the HLAphenotype of tumor cells. In contrast, each of the three methodsindividually is not suitable to identify the HLA allospecificities expressed on the membranes of tumor cells. The usefulness of thebinding assay with mAb for HLA typing of tumor cells is restricted toa limited number of HLA allospecificities, since the number of mAbspecific for HLA alloantigens is still small. Furthermore, the IEFtyping technique does not have the discriminatory power to identifyall the serologically defined HLA allospecificities, since some of themhave very similar if not identical electrophoretic patterns, as shown inthis study. The typing techniques based on the nucleotide sequencesencoding the polymorphic polypeptides of HLA antigens have beenused to only a limited extent, to assess their practical usefulness (35,47). At any rate, a major limitation of DNA-based typing techniques

is represented by their inability to determine whether the HLA allospecificities encoded within a tumor cell are expressed on its plasma

4308

Research. on October 18, 2020. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

ß2-li AND HLA EXPRESSION BY MELANOMA CELLS

membrane. The latter information contributes to determining whetherHLA class I allospecificities fulfill their function to bind peptidesintracellularly and transport them to the cell surface, where antigenicpeptide presentation to CD8+ T lymphocytes can occur.

ACKNOWLEDGMENTS

The authors wish to acknowledge the secretarial assistance of Edwina L.Jones, Harriett V. Harrison, Donna D. James, and Gail D. Price.

REFERENCES

1. Poulik, M. D., Ferrone, S., Pellegrino, M. A., Sevier, D. E., Oh, S. K., and Reisfeld,R. A. Association of HL-A antigens and ßi-microglobulin: concepts and questions.Transplant. Rev., 2.Õ:106-125, 1974.

2. Ploegh, H. L., Orr, H. T., and Strominger, J. L. Major histocompatibility antigens: thehuman (HLA-A.-B.-C) and murine (H-2K, H-2D) class I molecules. Cell, 24: 287-

299, 1981.3. Silver, M. L., Parker, K. C., and Wiley, D. C. Reconstitution by MHC-restricted

peptides of HLA-A2 heavy chain with ßi-microglobulin, in vitro. Nature (Lond.),350: 619-622, 1991.

4. Quaranta, V, Walker, L. E., Ruberto, O., Pellegrino, M. A., and Ferrone, S. The freeand the ß2-microglobulin-associated heavy chains of HLA-A, B alloantigens share theantigenic determinant recognized by the monoclonal antibody 01/28. Immunogenet-¡cs,13: 285-295, 1981.

5. Temponi, M., Kekish, U. M., Wang, Z., and Ferrone, S. Reactivity of monoclonalantibody (mAb) TP25.99 with a determinant shared by ß2-microglobulin (ß2-M)associated HLA-A, B and C heavy chains and ß2-nfree HLA-B and C heavy chains.In: K. Tsuji, M. Aizawa, and T. Sasaznki (eds.), HLA 1991, Proceedings of theEleventh International Histocompatibility Workshop and Conference (Japan), Vol. 2,pp. 395-393. New York: Oxford University Press, Inc., 1992.

6. Nakamuro, K., Tanigaki, N., and Pressman, D. Common antigenic structures of HL-Aantigens. VI. Common antigenic determinants located on [he 33,000-Dalton alloan-tigenic fragment portion of papain-solubilized HL-A molecules. Immunology, 29:1119-1132, 1975.

7. Stam, N. J., Spits, H., and Ploegh, H. L. Monoclonal antibodies raised againstdenatured HLA-B locus heavy chains permit biochemical characterization of certainHLA-C locus products. J. Immunol., 757: 2299-2306, 1986.

8. Gates, F. T., Ill, Coligan, J. E., and Kindt, T. J. Complete amino acid sequence ofmurine ß2-microglobulin: structural evidence for strain-related polymorphism. Proc.Nati. Acad. Sci. USA, 78: 554-558, 1981.

9. Rein, R. S., Seemann, G. H. A., Neefjes, J. J., Hochstenbach, F. M. H., Stam, N. J.,and Ploegh, H. L. Association with ß^-microglobulin controls the expression oftransfected human class 1 genes. J. Immunol., 138: 1178-1183, 1987.

10. Perarnau, B. M., Gillet, A. C., Hakem, R., Barad, M., and Lemonnier, F. A. Humanß2-microglobulin specifically enhances cell-surface expression of HLA class I molecules in transfected murine cells. J. Immunol., 141: 1383-1389, 1988.

11. Ferrier, P., Fontecilla-Camps, J. C., Bucchini, D., Caillol, D. H., Jordan, B. R., andLemonnier, F. A. Altered structure of HLA class I heavy chains associated with mouseß2-microglobulin. Immunogenetics, 2/: 321-331, 1985.

12. Ferrier, P., Layet, C., Caillol, D. H., Jordan, B. R., and Lemonnier, F. A. Theassociation between murine ß2-microglobulin and HLA class I heavy chains results in

serologically detectable conformational changes of both chains. J. Immunol., 135:1281-1287, 1985.

13. Tatake, R. J., Ferrone, S., and Zeff, R. A. The role of ß2-microglobulin in temperature-sensitive and interferon-y-induced exocytosis of HLA class I molecules. Transplantation (Baltimore), 54: 395-403, 1992.

14. Nickerson-Nutter, C. L., Hogen, K. L., and David, C. S. Expression of HLA-B27 intransgenic mice is controlled by gene(s) mapping between H-2D and H-2L loci.Immunogenetics, 35: 199-204, 1992.

15. Arnold, B., and Hämmerling, G. J. MHC class-I transgenic mice. Annu. Rev. Immunol., 9: 297-322, 1991.

16. D'Urso, C. M., Wang, Z., Cao, Y., Tatake, R., Zeff, R. A., and Ferrone, S. Lack of

HLA class I antigen expression by cultured melanoma cells FO-1 due to a defect inB2m gene expression. J. Clin. Invest., 87: 284-292, 1991.

17. Pellegrino, M. A., Ng. A-K.. Russo, C., and Ferrone, S. Heterogeneous distribution ofthe determinants defined by monoclonal antibodies on HLA-A and B antigens bearingmolecules. Transplantation (Baltimore), 34: 18-23, 1982.

18. Turco, M. C., De Felice, M., Corbo, L., Morrone, G., Mertelsmann, R., Ferrane, S.,and Venuta, S. Regulatory role of a monomorphic determinant of HLA class I antigensin T cell proliferation. J. Immunol., /.?5: 2268-2278, 1985.

19. Barnstable, C. J., Bodmer, W. F, Brown, G., Galfre, G., Milstein, C.. Williams, A. F,and Ziegler, A. Production of monoclonal antibodies to group A erythrocytes, HLAand the human cell surface antigens: new tools for genetic analysis. Cell, 14: 9-20,1978.

20. Tanabe, M., Sekimata, M., Ferrone, S., and Takiguchi, M. Structural and functionalanalysis of monomorphic determinants recognized by monoclonal antibodies reacting

with the HLA class I a, domain. J. Immunol., 148: 3202-3209, 1992.

21. Quaranta, V., Pellegrino, M. A., and Ferrone, S. The monoclonal xenoantibody 06/64recognizes a determinant expressed by certain gene products of the A and B loci ofthe HLA region. Immunogenetics, 14: 403—413,1981.

22. Yang, S. Y, Morishima, Y, Collins, N. H., Alton, T., Pollack, M. S., Yunis, E. J., andDupont, B. Comparison of one dimensional IEF patterns for serologically detectableHLA-A and HLA-B allotypes. Immunogenetics, 19: 217-231, 1984.

23. Antonelli, P., Nísperos,B., Braun, M., and Hansen, J. A. Recognition by a murinemonoclonal antibody of a unique epitope specific for the human alloantigen HLA-B8.Hum. Immunol., 11: 11-18, 1984.

24. Tada, N., Kimura, S., Hatzfeld, A., and Hämmerling, U. Ly-mll: the H-3 region ofmouse chromosome 2 controls a new surface alloantigen. Immunogenetics, //: 441-

449, 1980.25. Temponi, M., Kageshita, T, Perosa, F., Ono, R., Okada. H., and Ferrone, S. Purifi

cation of murine IgG monoclonal antibodies by precipitation with caprylic acid:comparison with other methods of purification. Hybridoma, 8: 85-95, 1989.

26. Fraker, P. J., and Speck, J. C., Jr. Protein and cell membrane iodinations with asparingly soluble chloramide, l,3,4,6-tetrachloro-3a,6a-diphenylglycoluril. Bio-chem. Biophys. Res. Commun., 80: 849-857, 1978.

27. Sood, A. K., Pereira, D., and Weissmann, S. M. Isolation and partial nucleotidesequence of a cDNA clone for human histocompatibility antigen HLA-B by use of anoligodeoxynucleotide primer. Proc. Nati. Acad. Sci. USA, 78: 616-620, 1981.

28. Koller, B. H., Sidwell, B., DeMars, R., and Orr, H. T. Isolation of HLA locus-specificDNA probes from the 3'-untranslated region. Proc. Nati. Acad. Sci. USA, Sì:5175-

5178, 1984.29. Feinberg, A. P., and Vogelstein, B. Addendum: a technique for radiolabeling DNA

restriction endonuclease fragments to high specific activity. Anal. Biochem., 137:266-267, 1984.

30. Temponi, M., Romano, G., D'Urso, C. M., Wang, Z., Kekish, U., and Ferrone, S.

Profile of intercellular adhesion molecule-1 (ICAM-1) synthesized by human melanoma cells. Semin. Oncol., 15: 595-607, 1988.

31. Zweig, S. E., and Shevach, E. M. Production and properties of monoclonal antibodiesto guinea pig la antigens. Methods Enzymol., 92: 66-85, 1983.

32. Russo, C., Ng, A-K., Pellegrino, M. A., and Ferrone, S. The monoclonal antibodyCR11-351 discriminates HLA-A2 variants identified by T cells. Immunogenetics, 18:23-35, 1983.

33. Yang, S. Y. Population analysis of class I HLA antigens by one-dimensional isoelec-

tric focusing gel electrophoresis: workshop summary report. In: B. Dupont (éd.),Immunobiology of HLA, Vol. 1, pp. 309-331. New York: Springer Verlag Inc., 1989.

34. Bonner, W. M., and Laskey. R. A. A film detection method of tritium labelled proteinsand nucleic acids in polyacrylamide gels. Eur. J. Biochem., 46: 83-88, 1974.

35. Oh, S-H., Fleischhauer, K., and Yang. S. Y. Isoelectric focusing subtypes of HLA-Acan be defined by oligonucleotide typing. Tissue Antigens, 41: 135-142, 1993.

36. Kawasaki, E. S. Sample preparation from blood, cells and other fluids. In: M. A. Innis,D. H. Gelfand, J. J. Sninsky, and T. J. White (eds.), PCR Protocols: A Guide toMethods and Applications, pp. 146-158. San Diego, CA: Academic Press, Inc., 1990.

37. Wood, W. I., Gitschier, J., Lasky, L. A., and Lawn, R. M. Base composition-inde

pendent hybridization in tetramethylammonium chloride: a method for oligonucleotide screening of highly complex gene libraries. Proc. Nati. Acad. Sci. USA, 82:1585-1588, 1985.

38. Scatchard, G. The attractions of proteins for small molecules and ions. Ann. NY Acad.Sci., 51: 660-672, 1949.

39. Rock, K. L., Gamble, S., Rothstein, L.. Gramm, C., and Benacerraf, B. Dissociationof ß2-microglobulin leads to the accumulation of a substantial pool of inactive classI MHC heavy chains on the cell surface. Cell, 65: 611-620, 1991.

40. Parham, P., Barnstable, C. J., and Bodmer, W. F. Use of a monoclonal antibody(W6/32) in structural studies of HLA-A, B, C antigens. J. Immunol., 123: 342-349,1979.

41. Bjorkman, P. J., Saper, M. A., Samraoui, B., Bennett, W. S., Strominger, J. L., andWiley, D. C. Structure of the human class I histocompatibility antigen, HLA-A2.Nature (Lond.), 329: 506-512, 1987.

42. Liunggren, H-G., Stam, N. J., Ohlen. C, Neefjes, J. J., Höglund, P., Heemels, M-T.,Bastin, J., Schumacher, T. N. M., Townsend, A., Karre, K., and Ploegh, H. L. EmptyMHC class I molecules come out in the cold. Nature (Lond.), 346: 476-480, 1990.

43. Bjorkman, P. J., and Parham, P. Structure, function and diversity of class I majorhistocompatibility complex molecules. Annu. Rev. Biochem., 59: 253-288, 1990.

44. Ferrone, S., Imberti, L., Natali, P. G., Schrier, P. I., and Ruiter, D. HLA antigens ontumor cells. In: J. A. Roth (ed.). Monoclonal Antibodies in Cancer: Advances inDiagnosis and Treatment, pp. 111-139. New York: Futura Publishing Company. Inc.,1986.

45. van Duinen, S. G., Ruiter, D. J., Broecker, E. B., van der Velde, E. A., Sorg, C.,Welvaart, K., and Ferrone, S. Level of HLA antigens in locoregional métastasesandclinical course of the disease in patients with melanoma. Cancer Res., 48: 1019-1025,1988.

46. Knuth, A., Wölfel,T, and Meyer zum Büschenfelde,K-H. T cell response to humanmalignant tumors. Cancer Surv., 13: 39—52,1992.

47. Browning, M. J., Krausa, P., Rowan, A., Bicknell, D. C., Bodmer, J. G., and Bodmer,W. F. Tissue typing the HLA-A locus from genomic DNA by sequence-specific PCR:comparison of HLA genotype and surface expression on colorectal tumor cell lines.Proc. Nati. Acad. Sci. USA, 90: 2842-2845, 1993.

4309

Research. on October 18, 2020. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from

1993;53:4303-4309. Cancer Res   Zhigang Wang, Xian-zhen Hu, Revati J. Tatake, et al.   Melanoma Cells

-Microglobulin Gene-null FO-12β-B Antigens Synthesized by Induction and the Antigenic Profile of Endogenous HLA-A and

-Microglobulin on the2βDifferential Effect of Human and Mouse

  Updated version

  http://cancerres.aacrjournals.org/content/53/18/4303

Access the most recent version of this article at:

   

   

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

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

  Permissions

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

.http://cancerres.aacrjournals.org/content/53/18/4303To request permission to re-use all or part of this article, use this link

Research. on October 18, 2020. © 1993 American Association for Cancercancerres.aacrjournals.org Downloaded from