A yeast two-hybrid system using Sp17 identified Ropporin as a novel cancer–testis antigen in hematologic malignancies

Authors


Abstract

Since most intracellular proteins are expressed with their ligands, ligands of cancer–testis (CT) antigens may also be CT in their distribution. Applying Sperm protein 17 (Sp17) as the bait in a yeast 2-hybrid system of a testicular cDNA library, 17 interacting clones were isolated and all encoded Ropporin, a spermatogenic cell-specific protein that serves as an anchoring protein for the A-kinase anchoring protein, AKAP110. Ropporin showed a very restricted normal tissue gene expression, detected only in testis and fetal liver. Ropporin mRNA could also be detected in tumor cells from patients with multiple myeloma, chronic lymphocytic leukemia and acute myeloid leukemia. Interestingly, expression of Sp17 did not necessarily predict for the expression of Ropporin suggesting that their coexpression in these tumor cells was random rather than coordinated. Ropporin gene expression in tumor cells is associated with the presence of high titer IgG antibodies against Ropporin, suggesting the in vivo translation of the mRNA into protein and the immunogenicity of the protein to the autologous hosts. Using a CT antigen as the bait in a yeast 2-hybrid system may, therefore, identify novel tumor antigen. Our results also suggest that Ropporin is a novel CT antigen in hematologic malignancies. © 2007 Wiley-Liss, Inc.

Although tumor vaccine is an attractive therapeutic option for patients with hematologic malignancies, clinical progresses have not matched laboratory successes. One reason for the lack of successes in clinical tumor immunotherapy is the heterogeneity of tumor antigen expression within individual tumor specimens. Therefore, successful targeting of one specific tumor antigen may be followed by the emergence of antigen-negative variant tumor cells. Furthermore, clinical successes may also be hampered by the relatively high tumor burden, when compared to the antigen-specific effector cells generated by the tumor vaccine in vivo. Identification of tumor antigens that are coexpressed within a tumor specimen may overcome these problems. These antigens could be used in polyvalent vaccines to produce high effector:target ratios of multiple specificity in vivo and reduce the chance for tumor escape by the emergence of variant tumor cells that do not express one particular antigen.

Cancer–testis (CT) antigens are a group of normal testicular proteins aberrantly expressed in cancer cells.1 Although initially thought to be testicular-specific, further sensitive studies involving real-time PCR have found “leaky” expression of many of these antigens in some normal tissues, albeit at a much lower level when compared to that in normal testis and in tumor cells.2, 3 Because of their wide distribution in a number of tumors, including hematologic malignancies, and their primary expression in normal testis, CT antigens represent excellent targets for immunotherapy. The testis is an immune privileged site for two reasons: the apparent lack of HLA class I expression on the surface of germ cells4 and most notably the presence of the blood–testis barrier that inhibits contact between differentiating germline and immune cells.5 As a result, thymic negative selection for CT-reactive T cells is less likely to occur, ensuring the preservation within the immune repertoire of high affinity and frequency of these CT-reactive precursor T cells that may be necessary for successful tumor immunotherapy.

Sperm protein 17 (Sp17) is a testicular protein showing restricted normal tissue expression.6, 7 It is expressed primarily in normal testis but could also be detected at very low levels in ciliated epithelial cells in the bronchi. We have previously identified Sp17 to be aberrantly expressed in tumor cells from patients with hematologic malignancies6 and ovarian cancer.8 Furthermore, we also found that it was possible to generate Sp17-specific cytotoxic T lymphocytes (CTLs) that killed autologous tumor cells expressing Sp17 from cancer-bearing patients,8, 9 suggesting that Sp17 could be a target for the development of tumor vaccines for Sp17-positive tumors.10

Most intracellular proteins are expressed in conjunction with their interacting ligands. We hypothesized that protein molecules interacting with CT antigens may also be testicular restricted and potential CT antigens. Identification of these proteins may provide the opportunity for their application in a polyvalent tumor vaccine to overcome the problems associated with antigen heterogeneity within a tumor specimen. To test our hypothesis, we have applied Sp17 as the bait in a yeast 2-hybrid system of a testicular cDNA library to identify the protein interacting with Sp17 and determine whether the interacting protein is also a CT antigen.

Material and methods

Clinical samples

Clinical materials consisted of peripheral blood and bone marrow from patients with hematologic malignancies and healthy donors. All clinical materials were obtained after informed consents and with approval from the local ethics committee. Both presentation and relapsed samples were included. A mouse recombinant Ropporin protein and polyclonal antisera against mouse Ropporin were kindly provided by Professor Shuh Narumiya in the Department of Pharmacology, Kyoto University Faculty of Medicine, Japan.

Screening of a yeast 2-hybrid testicular library using Sp17

Sp17 cDNA was amplified by PCR. The primers were: 5′Sp17: 5′-CGC GGA TCC AGA TGT CGA TTC CAT TCT CC-3′ and 3′Sp17: 5′-ATC TGC AGC TCA CTT GTT TTC CTC TTT TTC-3′. Successful amplification of the Sp17 cDNA was confirmed by sequence analysis and then subcloned into pGBKT7 between BamHI and PstI restriction sites. pGBKT7-Sp17 plasmid was transformed into yeast strain AH109 and selected on SD/-Trp plates. Mating was performed between AH109-pGBKT7-Sp17 and pretransformed human testis cDNA library in yeast strain Y187. Following mating, the culture was first selected on SD/-His/-Leu/-Trp plates and then on SD/-Ade/-His/-Leu/-Trp/X-α-Gal plates. Yeast plasmids were purified from the positive colonies and subjected to nucleotide sequence analysis.

Reverse transcription-polymerase chain reaction

Total RNA was isolated using an RNAEasy kit (Qiagen, Inc., Valencia, CA) according to the manufacturer's recommendation. Reverse transcription-polymerase chain reaction (RT-PCR) was performed. Briefly, all RNA specimens were first treated with DNAse I (Ambion, Inc., Austin, TX) to remove genomic DNA contamination. First strand cDNA was synthesized from 1 μg of total RNA using a random hexamer primer. To amplify Ropporin gene segment, the following pair of oligonucleotides were used—5′Ropporin: 5′-GCG AAT TCA TGG CTC AGA CAG ATA AGC-3′ and 3′Ropporin: 5′-ATG GAT CCG TTA CTC CAG CCA AAC CCT G-3′. They amplify a Ropporin gene segment of 652 bp. PCR was performed using 35 amplification cycles at an annealing temperature of 66°C. Negative controls in all the PCR reactions contained the PCR reaction mixture except for cDNA that was substituted with water. RNA integrity in each sample was checked by amplification of a β-actin gene segment. Successful removal of genomic DNA contamination was confirmed in each sample by amplification of the RNA without prior RT reaction. PCR products were visualized on an ethidium bromide agarose gel. All results were confirmed on 2 independent RT-PCRs.

Generation of Ropporin recombinant protein

Full coding sequence of Ropporin cDNA was isolated and amplified from normal testicular RNA. The PCR products were cloned into the TA-cloning system. The DNA was analyzed for nucleotide sequences bidirectionally and then subcloned into pQE30 vector (QIAGEN Inc, Valencia, CA) between BamHI and HindIII sites to produce a recombinant fusion protein of Ropporin that contained a 6-histidine peptide at the N-terminal of the protein. This strategy allowed affinity purification of the recombinant protein in a Nickel Sephadex column. The recombinant plasmid was transformed into Escherichia coli and selected on agar plates for ampicillin resistance. Recombinant clones were selected by restriction digest for cDNA fragments of the predicted size of 652 bp. To generate the recombinant protein, a recombinant clone was expanded in liquid culture and induced by 1 mM of isopropyl β-D-thiogalactopyranoside (IPTG) for 4 hr. Recombinant Ropporin protein was harvested from E. coli lysate by sonication. Following passage through the Ni-NTA affinity column and numerous rounds of washing, the protein was eluted. Successful generation of recombinant Ropporin protein was confirmed on SDS-PAGE by Coomassie blue staining and Western blotting using an antibody directed at the N-terminal 6-histidine tag (QIAGEN, Inc., Valencia, CA) and polyclonal antisera against a mouse Ropporin that cross-react with human Ropporin.

Enzyme-linked immunosorbent assay

Antibody directed to Ropporin was determined in the sera of patients with hematological malignancies by enzyme-linked immunosorbent assay (ELISA). Briefly, 96-well flat-bottom microtiter plates were coated with the purified recombinant Ropporin protein for 4 hr at room temperature. The plates were washed and blocked with 3% bovine serum albumin (BSA) in phosphate-buffered saline (PBS) at 37°C for 3 hr. Patient sera were diluted to 1:1000 in blocking buffer and added to the wells in triplicate and allowed to bind at 4°C overnight. Following extensive washing with PBS/Tween solution, alkaline phosphatase-conjugated goat antihuman IgG (Sigma, St. Louis, MO) at 1:1000 dilutions in blocking buffer was added to each well and incubated at room temperature for 2 hr. Negative control consisted of wells coated with PBS/BSA and wells with another E. coli-derived recombinant protein. All experiments were performed in triplicates. After 2 hr of incubation at room temperature, p-nitrophenylphosphate solution was added to each well and incubated at room temperature for color development. Twenty five microliters of 2 N NaOH was added to stop the reaction. Color intensity was measured on a microplate reader (Molecular Devices, Sunnyvate, CA) and analyzed using the Softmax data analysis program.

Western blot analysis

Purified recombinant Ropporin protein was fractionated in 12% SDS-PAGE electrophoresis and then transferred onto a nitrocellulose membrane. Successful generation of recombinant Ropproin protein was confirmed by an antibody directed to the histidine tag as well as polyclonal antisera, directed against mouse Ropporin, that cross-react with human Ropporin. Antibody binding was visualized by reaction with the Western Blue® stabilized substrate (Promega, Madison, WI).

Results

Sp17 interacts with human AKAP-binding sperm protein Ropporin

Following plating on selection plates, a total of 17 positive colonies were isolated. These colonies were expanded and the plasmids screened by restriction digest to confirm the presence of DNA fragments. All 17 interacting clones were subjected to nucleotide sequence analysis that identified all the 17 DNA fragments to encode human Ropporin. The interaction between Sp17 and Ropporin was further confirmed by a second round of yeast 2-hybrid interaction.

Ropporin is a novel CT antigen in hematologic malignancies

Using a pair of sequence-specific primers in RT-PCR on total RNA derived from a panel of normal tissues, we next determined the normal tissue expression of human Ropporin gene. Ropporin shows a very restricted normal tissue expression; transcripts encoding Ropporin could not be detected in any of the normal tissues except in testis and fetal liver (Fig. 1). Although not designed primarily for quantification, the signals obtained from fetal spleen were consistently much lower than that from normal testis, suggesting the presence of much lower transcript numbers in the fetal spleen.

Figure 1.

Expression of Ropporin mRNA, as detected by RT-PCR, in a panel of normal tissues (M = molecular marker; 1 = negative control; 2 = positive control; 3 = liver; 4 = lung; 5 = prostate; 6 = spleen; 7 = brain; 8 = pancreas; 9 = prostate; 10 = placenta; 11 = skeletal muscle;12 = thyroid; 13 = stomach; 14 = small intestine; 15 = uterus; 16 = kidney; 17 = heart; 18 = fetal liver; 19 = testis).

Having demonstrated the very restricted normal tissue expression of human Ropporin gene, we next determined whether or not Ropporin was being expressed by tumor cells from patients with hematologic malignancies. Transcripts encoding Ropporin could be detected in the tumor cells derived from the bone marrow in 6 of 16 (37.5%) cases of multiple myeloma, 6 of 14 (43%) cases of chronic lymphocytic leukemia and 2 of 11 (18%) cases of acute myeloid leukemia (Fig. 2). In contrast, Ropporin transcripts were not detected in the PBMC from any of 17 healthy donors (p < 0.02, χ2 test) and bone marrow/peripheral blood stem cells from 11 healthy donors (p < 0.02, χ2 test).

Figure 2.

Analysis for Ropporin mRNA in tumor cells derived from patients with hematologic malignancies. (a) RT-PCR showing the expression of Ropporin gene in patients with hematologic malignancies (1 = bone marrow; 2 = peripheral blood leukocytes; 3 = MM; 4 = MM; 5 = AML; 6 = CLL; 7 = testis; 8 = positive control using a Ropporin plasmid; 9 = negative control; M = molecular marker; a = PCR of DNase-treated RNA; b = RT-PCR for Ropporin gene segment; c = RT-PCR for β-actin gene segment). (b) Restriction digest with BclI of Ropporin PCR products produced two DNA fragments of 397 and 256 bp, showing the specificity of the PCR products (1 = normal testis; 2 = MM; 3 = MM; 4 = AML; 5 = CLL; M = molecular marker; a = BclI digest; b = mock digest).

The identity of the PCR products was confirmed either by sequence analysis or by restriction digest. There is a BclI internal restriction site within the Ropporin cDNA. Restriction digest analysis with BclI produced two Ropporin gene fragments of 397 and 256 bp (Fig. 2b). These results, therefore, indicate that Ropporin is aberrant expressed in tumor cells from patients with hematologic malignancies and suggest that Ropporin is a novel CT antigen.

Since Ropporin interacts with Sp17 in the yeast 2-hybrid system, we next determined whether or not the expression of Ropporin in tumor cells was random or coordinated to the expression of Sp17. The frequencies of Sp17 expression in this cohort of samples were higher than previously reported6 because these were not consecutive samples and were selected for high frequencies of Sp17 expression. We found that, although Ropporin was coexpressed with Sp17 in some samples, expression of Sp17 did not necessarily predict for the expression of Ropporin in individual tumor specimens (Table I). Only 40% of Sp17+ tumor specimens were also positive for Ropporin expression, suggesting that the expression of Ropporin gene by the tumor cells in most patients is likely a random process rather than being coordinated to the expression of Sp17.

Table I. Expression of Ropporin and its Relationship to Sp17 Expression in Hematologic Malignancies
DiagnosisRop+/Sp17+Rop+/Sp17−Rop−/Sp17+Rop−/Sp17−
MM (n = 16)5173
CLL (n = 14)5153
AML (n = 11)2063

Successful generation of recombinant Ropporin protein

The coding sequence of human Ropporin was sub-cloned into pQE30 plasmid and expressed in E. coli (Top10) as a fusion protein with a 6-histidine tag at NH2-terminal. The protein was purified in a Ni+ Sepharose column, then fractionated by SDS-PAGE gel electrophoresis and detected by Coomassie blue staining (Fig. 3). Successful generation of recombinant human Ropporin protein was confirmed by Western blot analysis (Fig. 3) using an antibody directed to His-tag antibody as well as antisera directed at mouse Ropporin that cross-react with human Ropporin.

Figure 3.

Successful generation of human Ropporin recombinant protein from E.coli. (a) Coomassie blue staining of a 12% SDS–polyacrylamide gel showing purified human Ropporin recombinant protein (1 = purified recombinant human Ropporin protein; 2 = mouse Ropporin recombinant protein). (b) Western blot analysis of the generated recombinant human Ropporin protein with an antibody directed at the 6-His tag (1 = cell lysate of E. coli Top10 used as negative control; 2 = recombinant human Ropporin protein fused to 6-His). (c) Western blot analysis of generated recombinant human Ropporin protein with polyclonal rabbit antisera directed to mouse Ropporin protein but cross-react with human Ropporin (1 = recombinant human Ropporin protein; 2 = negative control using another recombinant protein; 3 = positive control using the recombinant mouse Ropporin protein). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

High titer IgG antibodies against Ropporin were detected in sera from patients with hematologic malignancies

To determine the in vivo immunogenicity of Ropporin, antibodies against Ropporin in sera of patients with hematologic malignancies were detected by ELISA using the purified recombinant human Ropporin protein. To do so, we first established the background signals in the ELISA systems using sera from 31 healthy donors. With a serum dilution of 1:1000, the background signals among these healthy donors were consistently low (OD492 nm = 0.0736) (SD = 0.0147). Using mean + 3SD from the 31 sera as the cut-off signal intensity at OD492 nm, high titer IgG antibodies against Ropporin were detected at a serum dilution of 1:1000 in 8 of 30 MM (26.7%), 7 of 24 AML (29.2%), 18 of 31 CLL (58.1%), 9 of 27 CML (33.3%) and 1 of 3 ALL (33.3%) (Fig. 4). The OD492 nm in the two sets of negative control were consistently less than 0.010, indicating the specificity of the binding of the antibodies from these sera to the recombinant human Ropporin protein.

Figure 4.

ELISA analysis of diluted serum samples from patients with hematologic malignancies and health donors, showing the presence of high titer IgG antibodies directed at human Ropporin in patients with hematological malignancies.

Western blot analysis using the recombinant Ropporin protein was then used to further confirm the specificity of the ELISA results indicating the presence of antibodies against Ropporin in the sera of the patients (Fig. 5). Not surprising, because ELISA system is generally more sensitive than the Western blot analysis, only 50% of the ELISA+ specimens produced positive signals in Western blot analysis. Importantly, none of the specimens that were negative by ELISA produced positive signals in Western blot analysis (Table II). These results, therefore, support the specificity of the antibodies against Ropporin detected in the sera of patients with hematologic malignancies.

Figure 5.

Western blot analysis of serum samples from patients with hematological malignancies using the recombinant human Ropporin protein, showing the specificity of the antibodies detected by ELISA (1, 2 and 3 = samples showing positive results with ELISA; 4 = samples showing a negative result with ELISA). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]

Table II. Western Blot (WB) Analysis of Serum Samples
DiagnosisWB+/ELISA+WB+/ELISA−WB−/ELISA+WB−/ELISA−
MM (n = 4)1021
CLL (n = 4)2011
Donor (n = 3)0003

Correlation between gene expression and immune response

Finally, we determined the correlation between Ropporin gene expression (GE) and immune response (IR) against Ropporin. Paired tumor samples and sera were available from 14 patients (4 MM, 8 CLL and 2 AML). There was a good correlation between Ropporin GE and IR. Ropporin GE predicted for the development of IR s against human Ropporin (Table III), except in one MM patient in whom Ropporin antibodies were detected without Ropporin GE.

Table III. Correlation between Gene Expression (GE) and Immune Response (IR)
DiagnosisGE+/IR+GE+/IR−GE−/IR+GE−/IR−
MM (n = 4)1012
CLL (n = 8)3005
AML (n = 2)1001

Discussion

In this study we hypothesized that, because most intracellular proteins are expressed in conjunction with their interacting ligands, identification of the proteins interacting with a CT antigen may not only provide the opportunity to determine whether the interacting proteins are also CT antigens suitable for immune targeting but also give insight into the mechanisms regulating antigen coexpression. By applying a testicular cDNA library to a yeast 2-hybrid system involving Sp17 as the bait, we have identified Ropporin as the protein interacting with Sp17.

Ropporin is a rhophilin-associated protein11 of 212 amino acids. The gene encoding Ropporin is localized to chromosome 3q21.1. It is an important component of the inner fibrous sheath of sperm flagella11 and has previously been found to interact with the A-kinase anchoring protein, AKAP110.12 The expression profile by analysis of EST counts in UniGene suggested that it is expressed primarily in normal testis but at much lower levels in skin, mouth, eye, brain and mammary gland. Using RT-PCR on total RNA derived from a panel of normal tissues, we have demonstrated that the transcripts for Ropporin was only detected in normal testis and fetal liver and not in any other normal tissues examined. Furthermore, we also demonstrated that Ropporin gene is expressed aberrantly by tumor cells from patients with a variety of hematologic malignancies, including MM, CLL and AML. The function of Ropporin in tumor cells is not known although it is tempting to speculate that it could be involved in the motility of these tumor cells. Provided the protein is immunogenic in vivo, Ropporin could also be a potential target for immunotherapy of these hematologic malignancies. Our results, therefore, support our hypothesis that the interacting proteins of a CT antigen may also show a very restricted normal tissue expression and may also be suitable targets for tumor vaccine design.

Several factors determine the suitability of a molecule as an antigen for tumor vaccine development.13 In addition to expression in tumor cells, restricted normal tissue expression will provide specificity and hence reduced toxicity from a tumor vaccine based on the molecule. Furthermore, the antigen has to be immunogenic in vivo in the cancer-bearing patients that are usually immunocompromised. Given that previous works by us14, 15 and others16 demonstrated the ability of tumor-derived molecules to elicit B-cell IRs in the autologous host, we proceeded to determine the in vivo immunogenicity of Ropporin in these patients. To do this, we first generated a Ropporin recombinant protein from E. coli. This was achieved through the cloning of the Ropporin cDNA as a fusion gene to produce a recombinant Ropporin protein that contained a 6-histidine tag at the N-terminal of the fusion protein. We successfully generated the recombinant fusion protein from E. coli. Since the Ropporin recombinant protein is bacteria derived, to exclude IRs due to contaminating bacterial antigens that may be present in the recombinant protein preparation, we included in all experiments a control recombinant 6-histidine fusion protein that has also been prepared from E. coli lysate.

Using the purified recombinant protein in ELISA and Western blot analysis, we demonstrated that B-cell responses, in the form of high titer IgG, against Ropporin protein occur frequently in patients with hematologic malignancies and not in healthy donors. This finding suggests that the Ropporin mRNA in the tumor cells is likely being translated to Ropporin protein and that the protein is immunogenicity in these patients. Since there was a good correlation between IRs against Ropporin and Ropporin GE, the IRs against Ropporin are likely to be specific and elicited by the Ropporin expressed in these tumor cells. Because B-cell IRs to a molecule are usually generated only with cognitive help from T cells, the presence of high titer IgG responses in these patients suggests that Ropporin is likely also immunogenic to T cells.

Although there was generally a good correlation between IRs against Ropporin and Ropporin GE, antibodies against Ropporin were also detected in the serum of a MM patient despite failure to detect Ropporin mRNA by RT-PCR. The underlying cause of this is unclear. The patient had not received any blood or plasma transfusion; therefore, the origin of the antibodies could not be attributed to exogenously derived antibodies from the transfused blood products. Even though mRNA level generally correlates with protein level, exceptions frequently occur, ranging from total lack of translation, despite abundance of mRNA, to high levels of protein without significant up-regulation of mRNA.13 Such discrepancy may, therefore, explain the results we have observed in this patient, although the possibility that the antibodies in this patient were generated as a result of immune dysregulation cannot be excluded.

The successful identification of Ropporin as another CT antigen in hematologic malignancies, using Sp17 as the bait in a yeast 2-hybrid system, suggests that this approach may be applied to other molecules to identify novel CT antigens. However, the lack of a good correlation between the expressions of Sp17 and Ropporin in the tumor cells from patients with hematologic malignancies suggests four important points. First, Ropporin and Sp17 are unlikely to be suitable for use in pair in a polyvalent vaccine for most patients with Sp17+ tumor. Second, the aberrant expression of these interacting pair of molecules is not a result of coordinated intracellular regulatory mechanisms but likely the result of some random molecular processes such as DNA demethylation that frequently occurs as a result of tumorigenesis and tumor progression. Third, if the function of one protein (either Sp17 or Ropporin) is dependent on the presence of its ligand, then these individual molecules expressed within the tumor cells are unlikely to be of any functional significance in the tumor cells from most patients. Finally, although the yeast two-hybrid system could be applied using known CT antigens as the baits to identify other novel CT antigens, unlike in normal testis, the lack of coordinated expression of the protein pairs within tumor cells precludes the ability to confidently conclude that, just because one CT antigen is being expressed by the tumor cells, the interacting CT antigen is also expressed within individual tumor cells/specimens.

In conclusion, we have provided the first evidence that Ropporin is another novel CT antigen in hematologic malignancies. It is capable of eliciting IRs in the autologous host. Furthermore, we have also demonstrated that it is possible to isolate a novel CT antigen by applying a known CT antigen as the bait to a yeast 2-hybrid system of a testicular cDNA library.

Ancillary