Serological analysis of recombinant cDNA libraries (SEREX) uses high titer IgGs to identify antigens expressed by autologous cancers. In order to identify tumor antigens in soft tissue sarcoma (STS), sera from four patients with malignant fibrous histiocytoma (MFH), gastrointestinal stromal tumor (GIST), pleomorphic liposarcoma, and dedifferentiated liposarcoma were screened against cDNA libraries derived from autologous tumor. We identified 18 antigens encoded by 15 different genes, including DLG7, which is located on chromosome 14q22, a locus previously found to be altered in STS, and the proto-oncogene JUN. Ten of fourteen antigens (71%) do not react with sera from healthy donors, suggesting that antibody recognition takes place during cancer progression. Using oligonucleotide microarray technology, most genes were variably expressed across a panel of 16 benign specimens and 41 STSs of different histologies. DLG7, however, showed restricted expression in testes and cancer, similarly to known germ cell cancer-testis antigens (or germ cell antigens, GCAs). Thus, the current study identified several antigens, including molecules implicated in tumorigenesis that were recognized by high titer IgGs in STS patients. Further studies of these selected novel STS antigens are warranted to identify and characterize potential antigen targets expressed by STS.
This article was published in Cancer Immunity, a Cancer Research Institute journal that ceased publication in 2013 and is now provided online in association with Cancer Immunology Research.
Cancer antigens that are potential targets for immune responses fall into several classes defined by function and/or expression profile: (i) GCAs, for example, MAGE, NY-ESO-1, LAGE-1, SSX-2 (1, 2, 3, 4, 5); (ii) protein differentiation antigens, for example, tyrosinase, DCT, TYRP1, and gp100 (6, 7, 8, 9, 10, 11); (iii) ganglioside differentiation antigens, for example, GD2, GD3, and GM2 (12, 13, 14, 15); (iv) mutated gene products, for example, p53, MUM-1, and CDK4 (16, 17, 18); (v) overexpressed self proteins, for example, p53 and HER2/neu (18, 19); and (vi) viral antigens, for example, EBV and HPV (20, 21).
SEREX is an antigen identification method that allows the screening of cDNA expression libraries using autologous sera derived from patients with cancer (22). Originally described in 1995, SEREX has been widely used to discover multiple novel tumor antigens in diverse types of cancer. The repertoire of SEREX-defined antigens currently deposited in the Cancer Immunome Database (23) includes more than 1000 gene entries, including NY-ESO-1, p53, tyrosinase, MAGE-1, and SSX (2, 22, 24, 25, 26, 27, 28).
The significance of antigens defined in this manner extends to cellular immunity. CD4+ T cell help is required to generate a high titer IgG antibody response (29), thus implicating CD4+ T cells in generating IgG responses against SEREX-defined antigens. Furthermore, several antigens defined in this manner, including epitopes from MAGE-1, MAGE-A4, and tyrosinase, were originally described as CD8+ T cell targets (7, 30, 31). This observation suggests that at least a subset of SEREX-defined antigens is recognized by the cellular immune response, including CTLs. Thus, antigen identification by SEREX may be a surrogate screen for potential targets of a cell-mediated immune response against cancer (32), as shown in the case of NY-ESO-1.
Initially identified by SEREX analysis of squamous cell carcinoma of the esophagus (2), NY-ESO-1 has subsequently emerged as a tumor antigen that elicits both a humoral and cellular immune response in patients with NY-ESO-1-positive cancers (32, 33, 34, 35). In fact, several recent studies have associated NY-ESO-1 seroreactivity with both CD4+ and CD8+ T cell recognition of peptides presented by MHC molecules (32, 36, 37).
STSs define a group of histologically and genetically diverse cancers that arise from mesenchymal tissue. These cancers account for approximately 1% of adult malignancies, with an annual incidence of approximately 8000 cases in the United States (38). STSs represent a unique opportunity for antigen discovery for several reasons: (i) the tumor is in some cases characterized by the presence of tumor-specific gene translocations that may provide unique antigenic epitopes or regulate expression of normally silenced genes; (ii) several subtypes of STS, including synovial sarcoma (39) and myxoid/round cell liposarcoma, express several known GCAs (unpublished data); and (iii) current therapeutic options in the treatment of advanced disease are limited, encouraging a search for new targets.
In the current study, we explored the IgG antibody repertoire in STS patients against four subtypes of STS, including MFH, GIST, dedifferentiated liposarcoma, and pleomorphic liposarcoma. Sera from four patients were screened against cDNA libraries derived from autologous tumor. We identified 18 clones representing 15 distinct genes, including genes implicated in neoplastic transformation. The latter are attractive targets for the future development of immune therapy in STS.
Identification of autologous STS antigens
Sera from four patients with STS were tested against recombinant cDNA libraries prepared from autologous tumor, leading to the identification of 18 seroreactive clones. Nine cDNA clones reacted with autologous IgG antibodies in a case of MFH (SEREX-M), two in a case of GIST (SEREX-G), four in a case of pleomorphic liposarcoma (SEREX-P), and three in a case of dedifferentiated liposarcoma (SEREX-D). These clones have been designated sarc-1 through sarc-18 and represent 15 distinct STS antigens. There was no overlap between antigens recognized by autologous antibodies in each of the STS specimens. Of the 15 different antigens, 14 corresponded to proteins of known function, including those implicated in the regulation of transcription and translation, DNA transposition, energy metabolism, and organelle biosynthesis. A single clone, sarc-12, corresponded to a previously uncharacterized EST (Table 1).
The cDNA sequences encoding each of the 15 distinct STS antigens were compared to sequences deposited in the Cancer Immunome Database (23). Seven of the fifteen distinct genes identified in this study had previously been identified in SEREX analyses of diverse tumor types (24, 27, 28, 40, 41, 42, 43) (Table 1).
Expression profile analysis of STS antigens
The mRNA expression profiles of selected antigens were determined using data from oligonucleotide microarray databases published by our group (44, 45), which included transcripts from 41 distinct STS specimens and 16 normal tissues (see the supplementary data for "signal values"). The STS specimens included leiomyosarcoma, GIST, synovial sarcoma, MFH, conventional fibrosarcoma, pleomorphic liposarcoma, dedifferentiated liposarcoma, myxoid/round cell liposarcoma, and clear cell sarcoma/melanoma of soft parts. Benign tissues included testis (pooled from 19 individuals), lung, heart, skeletal muscle, small intestine, colon, stomach, liver, spleen, pancreas, kidney, skin, bone marrow, prostate, adrenal gland, and connective tissue. Genes were chosen according to their representation on the Affymetrix U133A GeneChip®. Of the 11 genes evaluated in this manner, sarc-9 was of particular interest, as it demonstrated the highest levels of expression in testes and cancer. The measured "signal value" for this gene was 5.2 times higher in a pooled testes sample and 2.6 ± 2.3 times higher in STS (n = 41) compared to the maximum "signal value" measured in a panel of normal tissues (n = 15). The remaining antigens did not show restricted expression profiles in testes or in STS specimens (Figure 1).
Seroreactivity profile of STS antigens
To determine whether serological recognition of the STS antigens identified by autologous serum was related to a diagnosis of STS, we tested allogeneic sera from 11 patients with STS (various histologies) and tested 8 healthy donors for seroreactivity against the antigens. Of the 14 antigens screened, 4 were recognized by sera from healthy donors and 8 were recognized by serum from at least one allogeneic STS patient. Three antigens were recognized by sera from both healthy donors and allogeneic STS patients. Five clones reacted solely with serum from the autologous patient (Table 1). Thus, 4/14 antigens (28%) identified in this study were recognized by antibodies produced by the B cell repertoire of healthy individuals, representing autoantibodies. The remaining 10 antigens were recognized by serum from at least one patient with STS, suggesting STS-related IgG responses.
Of the antigens showing STS-related IgG reactivity, JUN and DLG7 were of particular interest. The proto-oncogene JUN, an intronless gene located on chromosome 1p31-32, encodes a 331 amino acid nuclear protein (35,675 Da) (46). JUN is a member of the bZIP family and interacts with FOS to form heterodimers. The protein functions as a transcription factor that recognizes and binds to the enhancer heptamer motif 5'-TGA(C/G)TCA-3'. DLG7, located on chromosome 14q22-23, encodes a 765 amino acid protein (85,668 Da) that is structurally similar to the Drosophila melanogaster discs large-1 (dlg1) tumor suppressor gene and membrane-associated guanylate kinase protein family members. Gene expression of DLG7 has been described as cell cycle-regulated, with maximum expression occurring at the S-phase through the G2- and M-phases of the cell cycle, suggesting a role in checkpoint control and/or DNA repair (47).
In the current study, serological analysis of recombinant cDNA expression libraries has led to the identification of 15 new antigens expressed by STS. Fourteen of the fifteen antigens are the products of known genes, and one, sarc-12, represents a previously uncharacterized EST. The identified gene products represent diverse cellular processes and functions, including the regulation of transcription and translation, DNA transposition, energy metabolism, and organelle biosynthesis. In particular, serum IgG antibodies in a patient with MFH recognized gene products with a known or suspected etiological role in cancer. The first, sarc-7/JUN, dimerizes to form AP-1 via Jun/Jun homodimers or Jun/Fos heterodimers. AP-1 regulates the transcription of a number of genes, some of which may mediate neoplastic transformation (48, 49, 50). The second, sarc-9/DLG7, is cell cycle-regulated and possibly implicated in cell cycle/DNA checkpoint control, a functional characteristic of certain oncogenes. It is noteworthy that DLG7 was mapped to 14q22-23. This chromosome region has previously been described as altered in STS, including fibrosarcoma and GIST (51, 52).
In order to determine the significance of SEREX-defined antigens in the context of STS immunity, we sought to associate immune recognition with the diagnosis of STS using sera from a panel of STS patients and healthy donors, and we investigated antigen expression profiles in a panel of benign tissues and STS using microarray technology.
We found that 10/14 antigens, including both sarc-7/JUN and sarc-9/DLG7, reacted with sera from patients with STS and not with sera from healthy donors, demonstrating STS-restricted reactivity. The remaining antigens presumably represented natural autoantigens. Most of the genes were variably expressed in a panel of normal tissue specimens and STS tumors. This observation, however, does not exclude the potential relevance of these antigens to either cancer progression or to cancer immunity, in particular when expression is restricted in normal tissues and is not associated with natural autoimmunity. It is noteworthy that the expression profile of sarc-9/DLG7 is similar to that of known cancer-testis GCAs with restricted expression in testes and cancer. The observations that sarc-9/DLG7 is recognized by the host immune system, displays limited expression in normal adult tissues, and is potentially implicated in tumorigenesis, including STS, make it an attractive candidate for further studies.
Several antigens identified here had been reported in SEREX analyses by other groups. A recent analysis of synovial sarcoma and testes, using sera from 2/54 patients with fibrosarcoma and MFH selected for antibody responses against NY-ESO-1, identified several antigens in STS, including sarc-13/RBPSUH and sarc-16/RBPSUH (28). None of the other STS antigens identified here had been recognized by patients with STS. Several antigens, however, had been identified in prior SEREX analyses, including diverse types of cancer. These include breast carcinoma, colorectal carcinoma, gastric carcinoma, melanoma, ovarian carcinoma, and renal cell carcinoma, showing that immune recognition of these antigens is not limited to patients with STS (24, 27, 28, 40, 41, 42, 43).
SEREX-defined antigens may be surrogate gene products recognized by CD4+ T cells, because T-cell help is implicated in generating high titer IgG antibody responses. Therefore, the identification of autologous IgG antibodies against our panel of tumor antigens suggests that in the corresponding patient there could be active cellular, as well as humoral, immune responses against STS. In summary, the current study has identified antigens recognized by high-titer IgGs in patients with STS, including antigens encoded by genes implicated in tumorigenesis; that is, DLG7 located on chromosome 14q22, a region previously found to be altered in STS, and the proto-oncogene JUN. Further studies defining the cell-mediated immune response against these two antigens are warranted in order to identify candidate tumor vaccines in STS.
Materials and methods
Tumor specimens that had been described in tumor classification studies reported by our group (45, 53) were obtained as surgical specimens and frozen at -80°C. Specimens were obtained from previously untreated patients, with the exception of SEREX-P, which was obtained subsequent to chemotherapy. SEREX-M was a primary tumor. The remaining tumors were recurrent. All specimens, including sera and tumors, were collected under procurement protocols reviewed and approved by the Memorial Sloan-Kettering Cancer Center (MSKCC, NY, USA) Institutional Review Board. Representative tumor tissue was embedded in OCT compound and frozen as tissue blocks using liquid nitrogen. Tumor specimens were selected for analysis following validation of histologic diagnosis.
RNA extraction and construction of cDNA libraries
Cryopreserved tumor sections were homogenized under liquid nitrogen by mortar and pestle. Total RNA was extracted by the guanidium thiocyanate method, and RNA integrity was assessed by ethidium bromide agarose gel electrophoresis. Poly(A)+ RNA was isolated using the Fast Track mRNA purification kit (Invitrogen, Life Technologies, Carlsbad, CA, USA). Poly(A)+ RNA from normal tissues was purchased from Clontech (Palo Alto, CA, USA) and from Ambion (Austin, TX, USA). Separate cDNA libraries were constructed in the ZAP Expression vector (Stratagene, La Jolla, CA, USA), according to the manufacturers instructions, using 2-5 µg Poly(A)+ RNA. The resultant libraries contained between 1.9 x 105 and 1.2 x 106 recombinants and were not amplified prior to immunoscreening. Approximately 1 x 105 recombinants were screened per library.
To reduce nonspecific reactivity, serum antibodies reactive against vector or bacterial transformation related antigens were absorbed prior to immunoscreening in the following manner. Sera (1:10 dilution) was preabsorbed by repeated passage through columns of Sepharose 4B coupled to lysates of E. coli Y1090 and bacteriophage-infected E. coli BNN97 (5 prime 3 prime, Boulder, CO, USA). Final serum dilutions (1:200) were prepared in 0.2% nonfat dried milk preserved with 0.02% sodium azide and stored at 4°C. For further absorption of nonspecific antibodies, serum was incubated for 12-20 h at room temperature in the presence of nitrocellulose membranes prepared using a negative control phage, as described below.
Library screenings were performed as previously described (24, 27). Briefly, recombinant phage were plated at a concentration of 5 x 103/15 cm plate and amplified for 4 h, then transferred to nitrocellulose membranes over an additional 15 h at 37°C. Membranes were blocked with 5% nonfat dried milk and prescreened with a 1:2000 dilution of peroxidase-conjugated Fc fragment-specific goat antihuman IgG (Jackson Immunoresearch, West Grove, PA, USA) for 1 h at room temperature. Color was developed with 3,3-diaminobenzidine tetrahydrochloride, and clones encoding IgG were then marked for exclusion from further analysis.
Membranes were incubated in autologous diluted sera for 12-20 h at room temperature and subsequently incubated in a 1:3000 dilution of alkaline phosphatase-conjugated Fc fragment-specific goat antihuman IgG (Jackson Immunoresearch, West Grove, PA) for 1 h at room temperature. Color was developed with 4-nitro blue tetrazolium chloride/5-bromo-4-chloro-3-indolyl-phosphate. Seroreactive clones were subcloned as previously described (24, 27).
DNA sequence analysis
Recombinant phage were converted to pBK-CMV phagemid by in vivo excision. Plasmid DNA was purified using a Qiaprep spin column (Qiagen, Chatsworth, CA, USA) and digested by EcoRI and XbaI to confirm the presence of insert. Individual clones were sequenced at the Cornell University DNA sequencing facility (Ithaca, NY, USA) and at the Memorial Sloan-Kettering Cancer Center (MSKCC, NY, USA) sequencing facility using ABIprism automated DNA sequencers (Perkin Elmer, Norwalk, CT, USA).
These microarray data were generated from published work by our group and included mRNA isolated from tumors used in this study (44, 45). As was previously described, cryopreserved tumor sections were homogenized under liquid nitrogen by mortar and pestle. Total RNA was extracted in Trizol reagent (Invitrogen, Carlsbad, CA, USA) and purified using the Qiagen RNeasy kit (Qiagen, Chatsworth, CA, USA). RNA quality was assessed by ethidium bromide agarose gel electrophoresis. Complementary DNA was then synthesized in the presence of oligo(dT)24-T7 from Genset Corporation (La Jolla, CA, USA). Complementary RNA was prepared using biotinylated UTP and CTP and hybridized to HG_U133A oligonucleotide arrays (Affymetrix, Santa Clara, CA, USA). Fluorescence was measured by a laser confocal scanner (Agilent, Palo Alto, CA, USA) and converted to signal intensity using Affymetrix Microarray Suite v5.0 software. Selected gene signal intensities were compared in a set of 41 STS specimens and 16 normal tissues.
We are grateful to the late Dr. Matthew Scanlan of the Ludwig Institute for Cancer Research for instruction in the SEREX technique and for his advice, and we dedicate this manuscript to him and to his family. We also wish to thank Barbara Kaye-Injeian, Alwyn Maynard, Raul Meliton, and Cora Mariano of the Memorial Sloan-Kettering Cancer Center (MSKCC) tumor procurement service for their work collecting and cryopreserving specimens, and to Dr. Rodica Stan for helping us to prepare this manuscript. This work has been supported by NIH grant CA-47179, the Etta Weinheim Memorial Fund, Swim Across America, and the Kennedy Family Fund.
- Received November 24, 2004.
- Accepted November 24, 2004.
- Copyright © 2005 by Neil H. Segal