Major Histocompatibility Antigens and Antigen-Processing Molecules in Uveal Melanoma1

Home , Antigen, Antigen processing, Histocompatibility, MHC class I

Vol. 9, 4159–4164, September 15, 2003 Clinical Cancer Research 4159

Major Histocompatibility Antigens and Antigen-processing Molecules in Uveal Melanoma1

Subramanian Krishnakumar, Dhiraj Abhyankar, tigens and the APM was seen in nonepithelioid cell melano- Amirtha Lakshmi Sundaram, mas. There was no correlation with largest tumor diameter. Conclusions: Our data suggest decreased expression of Vaijayanthi Pushparaj, HLA, and APM are seen in uveal melanomas with no ex- Mahesh Palanivelu Shanmugam, and trascleral extension and in nonepithelioid cell melanomas. 2 Jyotirmay Biswas Decreased expression of APM may contribute to decreased Department of Ocular Pathology [S. K., A. L. S., V. P., J. B.] and HLA class I antigen expression. Ocular Oncology [M. P. S.], Medical and Vision Research Foundations, Sankara Nethralaya, Chennai, 600 600, India, and Department of Immunology, Roswell Park Cancer Institute, Buffalo, INTRODUCTION New York 142631 [D. A.] Uveal melanoma is the commonest primary intraocular tumor in Caucasian adults, with an incidence rate of 0.7/100,000. The average survival rate after the diagnosis of ABSTRACT metastatic disease, usually in the liver, is between 2 and 7 Purpose: Malignant transformation of cells is fre- months. No effective treatment for metastatic disease is yet quently associated with abnormalities in the human leuko- available (1). The therapy of uveal melanoma remains problem- cyte antigen (HLA) expression. These abnormalities may atic because of the high rate of metastatic dissemination, irre- play a role in the clinical course of the disease, because HLA spective of the success of treatment of the primary tumor. When antigens mediate interactions of tumor cells with T cells and metastatic disease is diagnosed, patient survival is usually Ͻ1 natural killer cells. Uveal melanoma is a highly malignant year (2). Immunotherapy has generated a lot of interest in the tumor of the eye and is characterized by hematogenic spread treatment of metastatic uveal melanomas. MHC and the APM3 to liver. Antigen-processing molecules (APMs) are necessary have generated interest in immunotherapy in many tumors (3). for efficient expression of HLA class I antigens. We studied In uveal melanomas, abnormalities in the HLA class I ␤ the expression of HLA antigens and the APM in uveal antigen, 2-m, and HLA class II antigens have been reported melanomas by immunohistochemistry and correlated clini- (4–6). The pathway of APM involved in the generation of HLA copathologically. class I molecule has been well defined recently. These include ␤ Experimental Design: HLA class I antigen, 2-micro- the low molecular proteasomal complexes, generating antigenic ␤ globulin ( 2-m), HLA class II antigens, and the APM com- peptide fragments comprising of LMP 2, ␤ subunit of LMP2, prising proteasomal subunits low molecular mass polypep- i.e., ⌬, LMP 7, and the recently described multicatalytic en- tide (LMP) 2, ␤-subunit of LMP2-⌬, LMP 10, transporter dopeptidase complex-like-1, also referred to as LMP10. LMP10 associated protein 1 subunit, and chaperone molecules tapa- is not encoded in the MHC locus in contrast to LMP2 and sin and calnexin were studied in 41 primary uveal melanoma LMP7. LMP10 is necessary for the expression of LMP 2. The archival specimens by immunohistochemistry. Immuno- new proteasomal ⌬is highly homologous to LMP2 and ex- analysis was done by a semiquantitative method and corre- pressed in a reciprocal manner with LMP2. lated with extrascleral extension, cell types, and the largest The transporter proteins TAP1, TAP2 translocate peptides tumor diameter. from the cytosol to the ER. Different ER-resident chaperones ␤ Results: HLA class I antigen, 2-m, HLA class II anti- calnexin, calreticulin, the binding protein tapasin, stabilize gen, and the APM were decreased (negative staining in 29 MHC class I molecules during their folding and/or assembly in tumors and dull staining in 3 tumors) in 100% (32 of 32) the ER or assist their loading with peptides. Binding of high -affinity peptides to MHC class I molecules leads to the disso (0.01 ؍ uveal melanomas with no extrascleral extension. (P and positive (bright staining) in 67% (4 of 9) tumors with ciation of this TAP complex and the exit of the ternary MHC ␤ liver metastasis. Decreased immunoexpression of HLA an- class I/ 2-m/peptide complex from the ER through the golgi for presentation to CD8ϩ CTLs. Deficiency of LMP, TAP, and the chaperone proteins reduces the supply and repertoire of peptides available for binding to MHC-1 (7–9). Received 9/19/02; revised 5/12/03; accepted 5/15/03. In contrast to a number of studies (4–6) demonstrating The costs of publication of this article were defrayed in part by the HLA class I antigen down-regulation in primary uveal mela- payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 Supported by a grant from the Vision Research Foundation, Sankara Nethralaya, Chennai, India. 3 The abbreviations used are: APM, antigen-processing molecule; HLA, 2 ␤ ␤ To whom requests for reprints should be addressed, at Medical and human leukocyte antigen; 2-m, 2-microglobulin; LMP, low molec- Vision Research Foundations, Sankara Nethralaya, 18 College Road, ular mass polypeptide; TAP, transporter associated protein; ER, endo- Chennai–600 006, Tamil Nadu, India. Phone: 91-044-28271616; Fax: plasmic reticulum; mAb, monoclonal antibody; LTD, largest tumor 91-044-28254180; E-mail: [email protected]. diameter; NK, natural killer cell.

Table 1 HLA antigen and APMs immunoreactivity in Group A uveal melanomas with no extrascleral extensiona Transporter Chaperone MHC class antigens Proteasomal subunits protein molecules LTD Cell HLA HLA Follow-up ␤ ⌬ No. in mm type class I 2-m class II LMP 2 LMP 10 TAP 1 Tapasin Calnexin in months 1 11 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 60 mA 2 11 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 60 mA 3 14 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 46 mAI 4 11 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 56 mA 5 10 S Neg Neg Neg Neg Neg Neg Neg Het Neg 36 mA 6 15.5 S Neg Neg Neg Neg Neg Neg Het Neg Het 31 mA 7 16 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 30 mA 8 8 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 28 mA 9 9.5 S Neg Neg Neg Neg Het Neg Neg Het Neg 24 mA 10 9 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 30 mA 11 7.5 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 24 mA 12 10 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 31 mA 13 7 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 26 mA 14 15.5 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 32 mA 15 16 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 30 mA 16 12.5 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 6 mA 17 9 S Neg Neg Neg Neg Neg Neg Neg Neg Neg 20 mA 18 17 M Neg Neg Neg Neg Neg Neg Neg Neg Neg 18 mA 19 11 M Neg Neg Neg Neg Het Neg Neg Neg Neg 14 mA 20 16.9 M Neg Neg Neg Neg Neg Neg Neg Neg Neg 14 mA 21 9 M Neg Neg Neg Neg Het Neg Neg Neg Neg 18 mA 22 12 M Neg Neg Neg Het Neg Het Het Het Het 13 mA 23 15 M Neg Neg Neg Neg Neg Neg Neg Neg Neg 10 mA 24 23 M Neg Neg Neg Neg Neg Neg Neg Neg Neg 8 mA 25 14 M Neg Neg Neg Neg Het Neg Neg Neg Het 8 mA 26 20 E Het Het Het Het Neg Het Het Het Het 18 mA 27 4 E Het Het Het Het Neg Het Het Het Het 22 mA 28 14 E Neg Neg Neg Neg Neg Neg Neg Neg Neg 16 mA 29 6.5 E Het Het Neg Het Neg Het Het Het Het 24 mA 30 20 E Het Neg Het Het Het Het Het Het Het 14 mA 31 15 E Het Het Het Het Het Het Het Het Het 20 mA 32 12.6 E Het Het Het Het Neg Het Het Het Het 7 mA S, spindle; M, mixed; E, epithelioid; Neg, negative; Het, heterogeneous; Pos, positive; m, months; A, alive; D, dead.

noma lesions with favorable outcomes, there is little information and liver metastasis; communications received from the rela- about the expression of APM, which are responsible for gener- tives of the patients, and they all had liver metastasis according ating peptides and efficient expression of HLA class I antigens. to their family physicians). The follow-up data available from Therefore, in this present study, we investigated the immuno- the medical records of all of the patients are given in Tables 1 ␤ reactivity of HLA class I antigen, 2-m, HLA class II antigen, and 2. The minimum follow-up was for 6 months, and the and APM comprising proteasomal subunits LMP2, ␤ subunit of maximum follow-up was for 60 months. LMP2, i.e. ⌬, LMP10, transporter protein TAP1 subunit, and Inclusion and Exclusion Criteria. Inclusion criterion chaperone molecules tapasin and calnexin and correlated with was that all of the patients were treated by enucleation. Exclu- extrascleral invasion, cell types, and LTD in uveal melanomas. sion criteria included patients who had received pre or intraop- erative adjunctive treatments, such as radiotherapy and cryo- MATERIALS AND METHODS therapy. Melanomas with extensive necrosis and orbital Patient. We had earlier published a large series on uveal cellulites were excluded (12). Iris melanoma and metastatic melanomas among the Asian-Indians (10). Forty-one uveal mel- melanomas to the uveal tract were excluded. anoma lesions were obtained from 29 male and 12 female Tumor Samples. Neoplastic tissues were obtained from patients with a median age of 45 years. All of the patients were enucleation material of the patients. Each sample was processed evaluated at the ocular oncology clinic of our hospital between for conventional histopathological diagnosis. Histological sec- 1996 and 2000. Thirty-two were choroidal melanomas, 8 in- tions were prepared from tissues fixed in 10% buffered neutral volved both the ciliary body and the choroid, and 1 was a diffuse formalin for 48 h and embedded in paraffin. H&E-stained 6-␮m uveal melanoma, which simulated metastatic tumor (11). sections were prepared through the central region of the tumor The tumors were divided into two groups. Group A (no and reviewed for cell type, LTD, and extension of the tumor. extrascleral extension) had 32 tumors, and group B (extrascleral The blocks were from consecutive patients. extension/liver metastasis) had 9 tumors (5 tumors had extra- Cell Type. The designation of the cell type was based on scleral extension, and 4 tumors had both extrascleral extension the Callenders’ classification (13). There were 17 spindle cell

Table 2 HLA antigen and APMs immunoreactivity in Group B uveal melanomas with extrascleral extension/liver metastasis Transporter Chaperone MHC class antigens Proteasomal subunits protein molecules LTD Cell HLA HLA Follow-up ␤ ⌬ No. in mm type class I 2-m class II LMP 2 LMP 10 TAP 1 Tapasin Calnexin met in months 1 15 M Pos Pos Pos Neg Het Pos Pos Pos Pos Yes 14 mD 2 21.5 E Pos Pos Pos Pos Pos Pos Pos Pos Pos Yes 20 mD 3 16.5 E Pos Pos Pos Pos Neg Pos Pos Pos Pos Yes 18 mD 4 8.5 E Pos Pos Pos Pos Neg Pos Pos Pos Pos Yes 16 mD 5 14 E Het Het Neg Het Neg Het Het Het Het No 20 mA 6 8.5 E Het Het Het Het Neg Het Het Het Het No 20 mA 7 23 M Neg Neg Neg Het Neg Het Het Het Het No 10 mA 8 dif 21 M Neg Neg Neg Het Neg Neg Het Neg Neg No 22 mA 9 9 M Neg Neg Neg Neg Neg Neg Neg Neg Neg No 30 mA S, spindle; M, mixed; E, epithelioid; Neg, negative; Het, heterogeneous; Pos, positive; m, months; A, alive; D, dead; dif, diffuse melanoma; met, metastasis. melanomas, 12 mixed cell melanomas, and 12 epithelioid cell in the previous reading. The staining intensity was scored as Ϫ melanomas. For the analysis, in this report, they were simplified (absent), ϩ/Ϫ (dull), and ϩ (bright). The tumors were graded as into two groups as epithelioid and nonepithelioid melanomas follows: positive (Ͼ75% cells stained and with bright intensity comprising mixed and spindle cell melanomas, and the immuno- of staining), heterogeneous (25–75% of the cells stained, mainly reactiviity of the HLA antigens and the APM was assessed in these. at a dull intensity), with the percentage of cells expressed to Tumor Size. Tumor size was measured as the LTD in nearest 10%), and negative (absent staining; HLA expression in millimeters from histological sections under the microscope cancer. International Histocompatibility Working Group, Pro- (14). Tumors were divided into two groups according to the ject description).4 dimension, LTD Յ 10 mm and LTD Ͼ 10 mm, and immuno- Variations in the percentage of stained cells enumerated by reactivity was assessed in these two groups. the two investigators were within a 10% range. The staining Monoclonal Antibodies. The affinity purified mouse an- intensity of adjacent normal structures (i.e., lymphoid and en- tihuman locus specific mAb HC-10 which recognizes HLA dothelial cells) was used as an internal control to evaluate the ␤ class I antigens, the anti- 2-m mAb L368, the anti-HLA class II staining intensity of malignant cells. For negative control, the mAb LGII-612.14, and the molecules of the APM comprising primary antibody was omitted, and nonimmune serum was used anti-LMP2 mAb SY-1, anti-⌬ mAb SY-4, anti-LMP10 mAb in the immunostaining. TO-6, anti-TAP1 mAb TO-1, anti-Tapasin mAb TO-4, and The study was reviewed and approved by the local ethics anti-Calnexin mAb TO-5, which were used, are target specific committee of our institute, and the committee deemed that it and exhibit no cross-reactivity (15). The antibodies were gifts conformed to the generally accepted principles of research, in from Dr. Soldano Ferrone, Department of Immunology, Roswell accordance with the Helsinki Declaration. Park Cancer Institute (Buffalo, NY). Labeled streptavidin kit Statistical Analysis. For statistical analysis, decreased was purchased from DAKO Laboratories (Glostrup, Denmark). expression (negative and heterogeneous) of HLA antigen and Immunohistochemistry. Immunostaining of tissue sec- APM expression were compared with the positive expression in tions was performed using labeled streptavidin by indirect im- the tumors with no extrascleral extension (group A) and tumors munoperoxidase technique. Briefly, 4-␮m formalin-fixed paraf- with extrascleral extension and liver metastasis (group B). The fin sections were used for the study. Tissue sections were then immunoreactivity was also analyzed in the epithelioid and non- deparaffinised and rehydrated and bleached before immunohis- epithelioid melanomas and in tumors with LTD Յ 10 mm and tochemical procedure. Endogenous peroxidase was blocked LTD Ͼ 10 mm using Fisher’s exact test. with hydrogen peroxide for 10 min at room temperature. No antigen retrieval was performed before antibody incubation. RESULTS Tissue sections were then rinsed in Tris-buffered saline (pH 7.6) The results of immunoperoxidase staining of HLA class I and incubated with respective primary antibody for 1 h. This ␤ antigen, 2-m and HLA class II antigen, and APM in groups A was followed by sequential 40-min incubation with biotinylated (with no extrascleral extension) and B (extrascleral extension) secondary antibody and streptavidin labeled to horseradish per- are shown in Tables 1 and 2. Examples of a negative immuno- oxidase (DAKO). Sections were washed with Tris-buffered staining of HLA class I antigen in spindle melanoma and pos- saline between incubation. The peroxidase reaction was devel- itive immunostaining in melanoma with poor outcome are oped for 5 min using commercially available 3,3 Јdiaminoben- shown in Fig. 1, A and B, respectively. The immunoexpression zidine and counterstained with Harris hematoxylin. of HLA antigens and the APMs were concordant in the majority Assessment of Immunohistochemical Results. Tissue of the tumors. sections were read independently by two ocular pathologists (S. K. and J. B.) without the knowledge of the results obtained by the other investigator. Furthermore, each investigator read all of the slides twice without the knowledge of the results obtained 4 Internet address: http://www.ihwg.org.

Immunoreactivity of HLA Class II Antigens in Uveal Melanomas. HLA class II antigen was decreased in 32 tumors (negative in 27 tumors and heterogeneous in 5 tumors with 30–40% cells stained dull), in-group A with no extrascleral extension (Table 1). Among the 9 tumors in group B with extrascleral extension, HLA class II antigen was positive (bright staining in 80% of cells) in 4 tumors with liver metastasis and decreased in 5 tumors (heterogeneous in 1 tumor with 40% cells stained dull and negative in 4 tumors) with no liver metastasis. The decreased expression of HLA class II antigen in the uveal melanoma with no extrascleral extension was significant (P ϭ 0.001). HLA class II antigen was decreased in nonepithelioid cell melanomas. Our results are concordant with another study on HLA class II antigens in uveal melanoma (5). Immunoreactivity of Proteasomal Subunits in Uveal Melanomas. LMP2 and LMP10 were decreased in 32 tumors (negative in 25 tumors and heterogeneous with 30% cells with dull staining in 7 tumors) in group A tumors with no extrascleral extension. LMP2 was positive (bright staining with 80% cells staining) in 3 tumors with liver metastasis and decreased in 5 tumors (heterogeneous with 30% cells staining dull in 4 tumors and negative in 1 tumor). LMP10 was positive in 4 tumors with liver metastasis and decreased (heterogeneous with 30% cells staining dull and negative in 2 tumors) in the remaining 5 tumors. ⌬ was decreased in 32 tumors (negative in 25 and heterogeneous in 7 tumors with 30% cells stained dull) in group A with no extrascleral extension. ⌬ was positive in 1 tumor with liver metastasis and negative in 8 tumors in group B. The decreased expression of LMP2 and LMP10 in the uveal melanoma with no extrascleral extension was significant (P ϭ 0.001), and decreased expression was seen in the nonepithelioid cell melanomas. Immunoreactivity of TAP1 Subunit in Uveal Melano- mas. TAP 1 was decreased in 32 tumors (negative in 24 tumors and heterogeneous with 30% cells stained dull in 8 tumors) with no extrascleral extension. TAP1 was positive (bright staining with 90% cells staining) in all 4 tumors with liver metastasis and decreased in 5 tumors (heterogeneous with Fig. 1 A, negative immunostaining of HLA class I antigen in spindle 40% cells stained dull and negative in 1 tumor) with no liver cell uveal melanoma. B, positive immunostaining of HLA class I antigen metastasis. The decreased expression of TAP1 in the uveal mela- in uveal melanoma with poor outcome (ϫ20 objective). noma with no extrascleral extension was significant (P ϭ 0.001). Decreased TAP1 was seen in nonepithelioid cell melanomas. Immunoreactivity of Chaperone Molecules Tapasin and Calnexin in Uveal Melanomas. Tapasin and calnexin ␤ Immunoreactivity of HLA Class I Antigens and 2-m in were decreased in group A tumors with no extrascleral exten- ␤ Uveal Melanomas. HLA class I antigen and 2-m are de- sion (negative in 23 tumors and heterogeneous with 30% cells creased in 32 tumors (negative staining in 26 tumors and het- stained dull in 9 tumors). Tapasin and calnexin were positive in erogeneous with 30–40% cells stained dull in six tumors) in 4 tumors (bright staining with Ͼ70% cells stained) with liver group A with no extrascleral extension (Table 1). Among the 9 metastasis and decreased in 5 tumors (heterogeneous with 30% tumors in group B with extrascleral extension, HLA class I cells dull stained in 3 and negative in 2 tumors) with no liver ␤ antigen and 2-m are positive (bright staining in 80% cells) in 4 metastasis in group B tumors. The decreased expression of tumors with liver metastasis and decreased in 5 tumors (heter- Tapasin and calnexin in the uveal melanoma with no extras- ogeneous with 30% cells stained dull in 2 tumors and negative cleral extension was significant (P ϭ 0.001). Decreased expres- in 3 tumors) with no liver metastasis shown in Table 2. The sion of chaperone molecules was seen in nonepithelioid cell ␤ decreased expression of HLA class I antigen, 2-m in the uveal melanomas. melanoma with no extrascleral extension was significant (P ϭ Correlation of Immunoreactivity of HLA Class I Anti- 0.001) They were decreased in nonepithelioid cell melanomas. gen and LMP2, LMP10, TAP1, Tapasin, and Calnexin in Our results are concordant with other studies on uveal mela- Uveal Melanomas. Their expression was concordant in most noma using locus-specific antibody for HLA antigens (4–6). of the tumors in groups A and B. HLA class I antigens and

LMP2, LMP10, TAP1, Tapasin, and calnexin were all decreased progression in grade of malignancy in cutaneous melanoma in group A tumors with no extrascleral extension and positive in (28, 30). group B tumors with liver metastasis. This reflects the major role played by HLA class I antigen- restricted, tumor antigen-specific CTL in the control tumor growth in cutaneous melanoma. From this, it follows that HLA DISCUSSION class I antigen down-regulation may provide tumor cells with an The results of the present study show that HLA class I escape from CTL recognition and destruction. In uveal mela- ␤ antigen, 2-m, and HLA class II antigens and APMs immuno- noma, the association of low HLA class I antigen expression in expression are decreased in primary uveal melanoma lesions primary uveal melanoma lesions with favorable clinical course with no extrascleral extension (P ϭ 0.001) and are positive in may reflect the susceptibility to NK cell-mediated lysis of low uveal melanomas with liver metastasis. Nonepithelioid cell mel- HLA class I-expressing melanoma cells invading blood vessels. anomas have a decreased immunoexpression of the HLA anti- The molecular basis of the lack of HLA class I antigen gens and APM. However, there was no correlation with the expression on uveal melanoma has not yet been elucidated. ␤ LTD. Our results are similar to the previous studies (5, 6) on Although in cutaneous melanoma, 2-m gene mutations have ␤ uveal melanoma where decreased HLA class I, 2-m, and HLA been found to be responsible for the lack of HLA class I antigen ␤ class II antigens were associated with favorable outcome, and expression (31), 2-m mutations have not been described in there was no correlation with nuclear grade, mitosis, MiB-1, uveal melanoma, although a significantly decreased expression ␤ tumor-infiltrating lymphocytes, and LTD. Our study also shows of 2-m was seen in our study and in an earlier study (5). Thus, that there is a correlation between LMP2 and LMP10 immuno- the difference in expression of HLA class I antigen and APM in reactivity. There is a reciprocal relationship between LMP2 and relation to prognosis in uveal melanoma when compared with ⌬ in mixed and epithelioid melanomas. Immunoexpression of ⌬ cutaneous melanomas could be attributable to the immune priv- subunit of LMP2 was variable in tumors with extrascleral ex- ileged location in the eye where uveal melanomas occur and the tension. role of NK cells in immune surveillance (32). This study shows for the first time that APMs are decreased In this regard, the concordant down-regulation of HLA in primary uveal melanomas with no extrascleral extension. class I antigen and APM in a high percentage of uveal melano- ␤ HLA class I antigen, 2-m, and the APM expressions were mas in our study suggests defects in the regulatory mechanisms concordant in most of the tumors. This suggests the importance that control their expression and not structural defects in the of APM for the expression of HLA class I molecules. Similar corresponding genes. It is unlikely that mutations are present in decreased expression of HLA class I antigen with decreased multiple genes encoding HLA class I antigen and APM (4). IFN expression of the components of the APM has been reported in treatment resulted in significant up-regulation of TAP1/TAP2 cervical cancer (16), cutaneous melanomas (17) in a number of proteins, immunoproteasomal subunits, and the MHC class I other human tumors (18), and in tumor cell lines of distinct heavy chain, indicating that ocular melanomas have functional histology, such as breast cancer (19), lung cancer (20), and HLA class I presentation machinery capable of responding to colon cancer (21). An important role for proteasomes, TAP, as IFN type 2 and that transcriptional defects are responsible for well as chaperones has been postulated using model cell lines, the expression of HLA antigens and APM in uveal melanoma transfected by the respective genes. Tapasin defective cells (4). Similar transcriptional defects in the HLA class I antigens show reduced MHC class I surface expression, which could be and APM have been described in cervical carcinomas (16). corrected by tapasin gene transfer (22). TAP1 knockout mice are Several mechanisms may contribute to the association be- defective in stable assembly of MHC class I molecules and tween disease prognosis and HLA class II expression in uveal show extremely reduced cell surface MHC class I expression melanoma. HLA class II antigen-bearing melanoma cells induce (23). Thus, APMs are necessary for efficient peptide delivery the secretion of immunosuppressive cytokine interleukin-10 by for expression of MHC class I antigen expression. However, T cells, resulting in T-cell anergy (33) that may explain the recently, proteasome-independent, TAP-independent pathways association between high HLA class II antigen expression in supporting class I MHC-mediated presentation of exogenous primary uveal melanoma lesions and poor prognosis. Alterna- antigens, as well as of endogenously synthesizes viral antigens tively, the observed correlation may reflect the resistance to NK has been described (24). cell lysis of hematogenously spreading melanoma cells with Our findings on HLA class I antigen (with mAb HC10) and high HLA class I as well as HLA class II antigen expression, TAP1 expression in uveal melanoma are in contrast to studies because HLA class I and II antigens may share a common on cutaneous melanoma (25–27). Down-regulation of HLA regulatory pathway (34). This may explain why the expression class I antigen in primary cutaneous melanoma is associated of HLA class I and II antigens was found to be correlated in our with increased thickness of the lesion, tumor progression, and study and in an earlier study (5). reduced survival, indicating that HLA class I antigen is an In conclusion, our study supports the previous studies on ␤ important factor in the behavior of this tumor, which first HLA class I, 2-m, and HLA class II antigens in uveal mela- spreads locally, then lymphatically (25–28). In cutaneous mel- noma (5, 6) and highlights the significance of APM in MHC anomas, down-regulation of the expression of HLA class I class I expression in uveal melanoma. Even tumors expressing antigens is associated with a lack of expression of TAP1 and decreased amounts of HLA class I antigen may be highly TAP2 correlating with HLA class I down-regulation (with mAb susceptible to NK cell-mediated lysis (35). The underlying HC10) and reduced survival (17, 29). On the other hand, in- molecular mechanisms of expression of APM in uveal melano- creased expression of HLA class II antigens was correlated with mas have not been analyzed. Additional studies are needed to

understand the precise biological role of APM in uveal mela- 18. Seliger, B., Maeurer, M. J., and Ferrone, S. Antigen processing noma. This will add to our understanding in the development of machinery breakdown and tumor growth. Immunol. Today, 21: 455– immunotherapy in uveal melanoma. 464, 2000. 19. Kaklamanis, L., Leek, R., Koukourakis, M., Gatter, K. C, and Harris, A. L. loss of transporter in antigen processing I transport protein ACKNOWLEDGMENTS and MHC class I molecules in metastatic versus primary breast cancer. Cancer Res., 55: 5191–5194, 1995. We thank Sukumar for assistance in statistical analysis and Dr. R 20. Korkolopoulou, P., Kaklamanis, L., Pezzella, F., Harris, A. L., and Rajagopal for the linguistic revision. Gatter, K. C. Loss of antigen-processing molecules (MHC class I and TAP1) in lung cancer. Br. J. Cancer, 73: 148–153, 1996. REFERENCES 21. Kaklamanis, L., Townsend, A., Doussis-Ananostopoulou, L. A., Mortensen, N., Harris, A. L., and Gatter, K. C. Loss of MHC-encoded 1. Mooy, C. M., and De Jong, P. T. Prognostic parameters in uveal transporter associated with antigen processing (TAP) in colorectal can- melanoma: a review. Surv. Ophthalmol., 41: 215–228, 1996. cer. Am. J. Pathol., 145: 505–509, 1994. 2. Kath, R., Hayungs, J., Bornfeld, N., Sauerwein, W., Hoffken, K., and 22. Dick, T. P., Ruppert, T., Groettrup, M., Kloetzel, P. M., Kuehn, L., Seeber, S. Prognosis and treatment of disseminated uveal melanoma. Koszinowski, U. H., Stevanovic, S., Schild, H., and Rammensee, H. G. Cancer, 72: 2219–2223, 1993. Coordinated dual cleavage induced by the proteasome regulator PA28 3. Hicklin, D. J., Marincola, F. M., and Ferrone, S. HLA class I antigen alpha lead to dominant MHC ligands. Cell, 86: 253–262, 1996. down regulation in human cancers. T cell immunotherapy revives an old 23. Ortmann, B., Androlewicz, M. J., and Cresswell, P. MHC class story. Mod. Med. Today, 5: 178–188, 1999. I/beta 2 microglobulin complexes with TAP transporters before peptide 4. Jager, M. J., Hurks, H. M., Levitskaya, J., and Kiessling, R. HLA binding. Nature, 368: 864–867, 1994. expression in uveal melanoma: there is no rule without some exception. 24. Fromm, S. V., Duady-Ben Yaakov, S., Schechter, C., and Ehrlich, Hum. Immunol., 63: 444–451, 2002. R. Assembly and cell surface expression of TAP-independent, chloro- 5. Ericsson, C., Seregard, S., Bartolazzi, A., Levitskaya, E., Ferrone, S., quine-sensitive and interferon-gamma-inducible class I MHC com- Kiessling, R., and Larsson, O. Association of HLA class I and Class II plexes in transformed fibroblast cell lines are regulated by tapasin. Cell. antigen expression and mortality in uveal melanoma. Invest. Ophthal- Immunol., 215: 207–218, 2002. mol. Vis. Sci., 42: 2153–2156, 2001. 25. Bro¨cker, E. B., Suter, L., Bru¨ggen, J., Ruiter, D. J., Macher, E., and 6. Blom, D. J., Luyten, G. P., Mooy, C., Kerkvliet, S., Zwinderman, Sorg, C. Phenotypic dynamics of tumor progression in human malignant A. H., and Jager, M. J. Human leukocyte antigen class I expression: melanoma. Int. J. Cancer, 36: 29–35, 1985. marker of poor prognosis in uveal melanoma. Invest. Ophthalmol. Vis. 26. Ruiter, D. J., Bhan, A. K., Harrist, T. J., Sober, A. J., and Mihm, Sci., 38: 1865–1872, 1997. M. C. Major histocompatibility antigens and mononuclear inflammatory 7. Belich, M. P., and Trowsdale, J. Proteasome and class I antigen infiltrate in benign nevomelanocytic proliferations and malignant mel- processing and presentation. Mol. Biol. Rep., 21: 53–56, 1995. anoma. J. Immunol., 129: 2808–2815, 1982. 8. Koopmann, J. O., Hammerling, G. J., and Momburg, F. Generation, 27. Ruiter, D. J., Bergman, W., Welvaart, K., Scheffer, E.,. Van Vloten, intracellular transport and loading of peptides associated with MHC W. A., Russo, C., and Ferrone, S. Immunohistochemical analysis of class I molecules. Curr. Opin. Immunol., 9: 80–88, 1997. malignant melanomas and nevocellular nevi with monoclonal antibodies 9. Lankat-Buttgereit, B., and Tampe, R. The transporter associated with to distinct monomorphic determinants of HLA antigens. Cancer Res., antigen processing: function and implications in human diseases. 44: 3930–3935, 1984. Physiol. Rev., 82: 187–204, 2002. 28. Ruiter, D. J., Mattijssen, V., Bro¨cker, E. B., and. Ferrone, S. MHC 10. Biswas, J., Krishnakumar, S., and Shanmugam, M. P. Uveal mel- antigens in human melanomas. Semin. Cancer Biol., 2: 35–45, 1991. anoma in Asian Indians. A clinicopathological study. Arch. Ophthal- 29. Hicklin, D. J., Kageshita, T., Dellaratta, D., and Ferrone, S. Trans- mol., 120: 522–523, 2002. porter associated with antigen processing (TAP) downregulation in 11. Biswas, J., Raghavendran, R., Ratra, V., Krishnakumar, S., Gopal, human melanoma cells. In: HLA. Genetic Diversity of HLA. Functional L., and Shanmugam, M. P. Diffuse malignant melanoma of the choroid and Medical Implication. Proceedings of the Twelfth International His- simulating metastatic tumor to the choroid. Indian J. Ophthalmol., 48: tocompatibility Workshop and Conference. Vol. II. Edited by D Char- 137–140, 2000. ron. Sevres, France, EDK, pp. 742–744, 1997. 12. Biswas, J., Ahuja, V. K., Shanmugam, M. P., Kurian, R., and 30. D’Alessandro, G., Zardawi, I., Grace, J., McCarthy, W. H., and Fernandez, T. Malignant melanoma of the choroid presented as orbital Hersey, P. Immunohistological evaluation of MHC Class I and II cellulitis: report of 2 cases with review of the literature. Orbit, 18: antigen expression on nevi and melanoma: relation to biology of mel- 123–130, 1999. anoma. Pathology, 19: 339–346, 1987. 13. Callender, G. R. Malignant melanotic tumors of the eye: a study of 31. Hicklin, D. J., Dellaratta, D. V., Kishore, R., Liang, B., Kageshita, histologic type in 111 cases. Trans. Am. Acad. Ophthamol. Otolaryn- T., and Ferrone, S., Beta 2-microglobulin gene mutations in human gol., 36: 131–142, 1931. melanoma cells: molecular characterization and implications for im- 14. Folberg, R., Game, I. J. W., Greenberg, R. A., Donoso, L. A., and mune surveillance. Melanoma Res., 2: S67–S74, 1997. Naids, R. M. Comparison of direct and microslide pathology measure- 32. Ma, D., and Niederkorn, J. Y. Transforming growth factor-beta ments of uveal melanomas. Investig. Ophthalmol. Vis. Sci., 26: 1788– down regulates major histocompatibility complex class I antigen ex- 1791, 1985. pression and increases susceptibility of uveal melanoma cells to natural 15. Kishore. R., Hicklin, D. J., Dellaratta, D. V., Golde, U., Kageshita. killer cell-mediated cytolysis. Immunology, 86: 263–269, 1995. T., Seliger, B and Ferrone, S. Development and characterization of 33. Brady, M. S., Lee, F. Petrie, H.,. Eckels, D. D., and. Lee, J. S., mouse anti-human LMP2, LMP7, TAP1 and TAP2 monoclonal anti- CD4ϩ T cells kill HLA-class-II-antigen postive melanoma cells present- bodies. Tissue Antigens, 51: 129–140, 1998. ing peptide in vivo. Cancer Immunol. Immunother., 48: 621–626, 2000. 16. Ritz, U., Momburg, F., Pilch, H., Huber, C., Maeurer, M. J., and 34. Van den Elsen, P. J., Gobin S. J., van Eggermond, M. C., and Seliger, B. Deficient expression of components of the MHC class I Peijnenburg, A. Regulation of MHC class I and II gene transcription: antigen processing machinery in human cervical carcinoma. Int. J. differences and similarities. Immunogenetics, 48: 208–221, 1998. Oncol., 19: 1211–1220, 2001. 35. Pende, D., Accame, L., Pareti, L., Mazzocchi, A., Moretta, A., 17. Kageshita, T., Hirai, S., Ono, T., Hicklin, D. J., and Ferrone, S. Parmiani, G., and Moretta, L. The susceptibility to natural killer cell- Downregulation of HLA class I antigen-processing molecules in malig- mediated lysis of HLA class 1-positive melanoma reflects the expres- nant melanoma: association with disease progression. Am. J. Pathol., sion of insufficient amounts of different HLA class I alleles. Eur. 154: 745–754, 1999. J. Immunol., 28: 2384–2394, 1998.

Downloaded from clincancerres.aacrjournals.org on September 24, 2021. © 2003 American Association for Cancer Research. Major Histocompatibility Antigens and Antigen-processing Molecules in Uveal Melanoma

Subramanian Krishnakumar, Dhiraj Abhyankar, Amirtha Lakshmi Sundaram, et al.

Clin Cancer Res 2003;9:4159-4164.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/9/11/4159

Cited articles This article cites 34 articles, 6 of which you can access for free at: http://clincancerres.aacrjournals.org/content/9/11/4159.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/9/11/4159.full#related-urls

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

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

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/9/11/4159. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.