Identification of genes that are upregulated in tumors, and whose normal expression excludes adult somatic tissues but includes germline and/or embryonic tissues, has resulted in a rich variety of cancer antigens that are attractive targets for cancer vaccine and other therapeutic approaches. In the present study, we extended this approach to include genes strongly and restrictively expressed in the placenta by mining publicly available SAGE and EST databases. We identified a number of genes with high expression in placenta and different cancer types but with relatively restricted expression in normal tissues. The gene with the most distinctive expression pattern was found to be PLAC1, which encodes a putative cell surface protein that is highly expressed in placenta, testis, cancer cell lines and lung tumors. Hence we have designated it CT92. We found by ELISA that PLAC1 is immunogenic in a subset of cancer patients and healthy women. Its physical and expression characteristics render it a potential target for both active and passive cancer immunotherapeutic strategies.
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.
Placenta growth and development is critical to normal embryogenesis. Many similarities exist between embryo implantation and the growth of cancer cells. In a manner analogous to malignant cells, placental trophoblasts, for example, migrate and invade the uterus and its vasculature in order to nourish the developing fetus (1). Likewise, another common characteristic of cancer and the placenta is the ability to downregulate the immune response (2). In contrast to the majority of human somatic tissues, the human placenta displays telomerase activity, as do most cancers (3). Further intriguing parallels between normal placentation and cancer is aneuploidy (4) and even epigenetic changes, such as selective hypermethylation (5).
Because of these similarities between embryo implantation and cancer progression, we reasoned that there might be a selective advantage for tumors to express some otherwise placenta-restricted genes which, due to their absence from normal adult somatic tissues, might also be immunogenic when aberrantly expressed.
High throughput global gene expression profiling techniques such as cDNA microarrays (6), EST sequencing (7), serial analysis of gene expression (SAGE) (8), and massively parallel signature sequencing (MPSS) (9) all generate data that is useful for the investigation of clinical aspects of malignancy through comparative gene expression analysis. In the present study, we used data generated using such large-scale gene expression technologies to identify genes highly expressed in placenta and different cancer types, but with relatively restricted expression in normal tissues. We identified PLAC1, a gene encoding a putative cell surface protein as being highly expressed in placenta, testis, cancer cell lines and lung tumors. Moreover, we show that PLAC1 can be immunogenic in cancer patients and therefore may constitute a target for active and passive cancer immunotherapy.
SAGE data was mined for genes highly expressed in placenta and also expressed in tumor tissues using the Digital Gene Expression Displayer tool. The data mining analysis resulted in the identification of 1075 genes expressed in both placenta and cancer library pools. From this list we selected 74 genes that were expressed at high levels in the placenta SAGE libraries (more than 100 tags) and were present in at least two libraries from cancer tissues. We then examined each SAGE tag and excluded those that were not unique using the SAGE genie tools at the CGAP website (10). Finally, the remaining genes were evaluated using EST data and the Virtual Northern tool in the CGAP website (10) to select those with the most restricted expression patterns in normal tissues. Seven of the 74 genes were found to be expressed in less than 5 somatic tissues (Table 1). Analysis of the predicted amino acid sequence of these genes revealed that two are putative cell surface proteins, of which PLAC1 had the most restricted normal expression pattern.
Analysis of PLAC1 expression in normal tissues
To confirm the normal tissue expression pattern of PLAC1, quantitative RT-PCR was performed and high expression of PLAC1 was found only in placenta. Testis was the tissue with the second highest level of PLAC1 expression, with an expression level corresponding to 13% of the placental expression. All other normal tissues showed very low mRNA levels, ranging from 0.03% to 0.65% (median 0.11%) of that found in placenta (Figure 1).
The expression of PLAC1 protein in placenta was then examined by immunohistochemistry using a rabbit anti-PLAC1 antibody raised against amino acids 125 to 212, a sequence unique to PLAC1. The antibody exhibited strong membranous staining of both syncytiotrophoblasts and cytotrophoblasts in human placental sections, but not of stromal cells (Figure 2A). In the adult human testis, antibody staining was evident on the luminal face of the seminiferous tubules, with the majority of staining appearing to be extracellular, often following a globular contour (Figure 2B).
Analysis of PLAC1 expression in cancer cell lines and tumor tissues
In order to investigate PLAC1 expression in tumors, we surveyed a panel of 74 tumor cell lines from different origins by quantitative real-time RT-PCR (Figure 3 and Table 2). Expression levels of more than 10% of that of the placenta were found in cell lines from cancers of the breast, cervix, colon, liver, lung, and prostate, as well as melanoma, multiple myeloma, choriocarcinoma and neuroblastoma cell lines.
We then analyzed PLAC1 expression in primary tumors of non-small cell lung cancer (NSCLC) patients. Of the 156 NSCLC tissue samples analyzed using non-quantitative RT-PCR, we detected PLAC1 expression in 119 (76.3%), of which 47 (30.1%) expressed PLAC1 at moderate levels as judged by the intensity of the ethidium bromide staining. Prompted by these results, real-time RT-PCR was performed to determine the relative level of PLAC1 expression in a subset of 85 samples for which information on the histological type of the tumors was available [18 squamous cell carcinomas, 43 non-bronchioloalveolar adenocarcinomas and 24 adenocarcinomas with bronchioloalveolar features (ADC-BAC)], and nine normal lung tissue controls. Expression levels of at least twice that of normal lung were detected in 62 (68.8%) patients, of which 24 (26.6%) exhibited expression levels more than ten times those of the normal lung samples (Figure 4). PLAC1 levels in the adenocarcinoma samples ranged from 0.23 to 319 (median 3.65) relative to normal lung, from 0.31 to 102 (median 3.84) in squamous cell carcinomas, and from 0.17 to 33 (median 3.08) in the ADC-BAC samples. The highest level of PLAC1 expression seen in the NSCLC samples was approximately 8% of that in placenta.
Comparison of PLAC1 and CT antigen expression
CT-X antigens, defined as cancer/testis (CT) antigens that are encoded by genes located on the X chromosome and whose expression is restricted to tumors and testis, represent a group of widely studied targets of cancer immunotherapy (11). Different members of this class of genes are typically co-expressed in tumors (12). Since PLAC1 is located at Xq26, we compared its expression to that of MAGEA1 (CT1.1), MAGEA3 (CT1.3), MAGEA4 (CT1.4), NY-ESO-1 (CT6.1) and LAGE1 (CT6.2) (all CT-X antigens) in the same 156 NSCLC samples tested above. The expression of the CT-X antigens was found to be coordinated, with 53 patients (33.9%) exhibiting high and simultaneous expression of at least two of the five CT-X genes analyzed (marked in red or pink in Figure 5). In this set of 53 patients, 16 (30.2%) presented concomitant high PLAC1 expression, while in the set of patients that expressed no or only one CT-X gene, 24 patients (23.3%) were found to have high levels of PLAC1 expression (P = 0.4595), showing that there is no enrichment of samples with high levels of PLAC1 in the subset of patients with high and simultaneous expression of CT-X genes.
PLAC1 antibody response in cancer patients
Due to the restricted expression profile of PLAC1 in normal tissues and its relatively high expression in some patients with NSCLC, we evaluated the possible spontaneous immunogenicity of PLAC1 by analyzing the plasma of NSCLC patients for the presence of specific anti-PLAC1 antibodies by ELISA (Figure 6). Of the 226 plasma samples from NSCLC patients tested, 14 (6%) were found to have detectable titers to PLAC1, ranging from 1/100 to 1/1200 (mean 1/574). Ninety-three patients whose plasma was tested by ELISA for the presence of anti-PLAC1 antibodies were also analyzed for tumor PLAC1 expression by semi-quantitative RT-PCR; 5 of 8 patients seropositive for PLAC1 showed high levels of PLAC1 mRNA in their tumors. Antibodies to PLAC1 were found in patients with squamous cell carcinomas, adenocarcinomas and bronchioloalveolar carcinomas and were of both genders (6 males and 7 females). Finally, sera from 78 normal healthy adults were analyzed and 4 (5%) were found to react with PLAC1 (Figure 6). Contrary to cancer patients, reactivity from healthy donor sera was restricted to females only, suggesting a possible gender-specific link with the immunogenicity of the antigen.
In this study, we sought to identify genes that are highly expressed in placenta and in different cancer types but show relatively restricted expression in normal tissues. We envisage that such genes could be involved in processes that are similar between embryo implantation and growth of cancer cells and therefore represent new potential therapeutic targets for cancer. Proteins expressed only in cancer and a restricted number of normal tissues are important vaccine targets, while cell surface exposed antigens expressed on cancer cells and placenta can also be used for the development of therapeutic monoclonal antibodies.
Analysis of PLAC1 expression in normal and cancer tissues suggested that it is a cancer/testis (CT) antigen, but with predominant expression in placenta. Other CT antigens have been shown to be predominantly expressed in placenta, including XAGE2 and XAGE3 (13). CT antigens are found to be expressed in a wide variety of tumors, while their expression in normal tissues is mostly restricted to germ cells of the testis, fetal ovary, and placenta; as such, they are ideal target molecules for cancer vaccines (14, 15). Approximately half of the CT antigens discovered so far map to chromosome X, while the remaining are distributed among the other chromosomes (referred to as CT-X and non-X CTs, respectively) (15). It has been shown that CT-X genes present a much more restricted pattern of expression in normal tissues as compared to non-X CTs (14). PLAC1 can be classified as a CT-X mapping to Xq26, which is in agreement with its pattern of expression in normal tissues. We have therefore given PLAC1 the designation CT92. Interestingly however, PLAC1 expression in testis as evaluated with the polyclonal antibody exhibited a pattern not seen previously for CT-X antigens, which are mainly expressed in the spermatogonia. PLAC1 was present in the luminal face of the seminiferous tubules, raising the possibility that this protein is secreted. In addition, our results show that the expression of PLAC1 is not coordinated with the most frequently expressed CT-X genes in lung cancer (12), which may indicate that its reactivation in cancer tissues is subject to different control mechanisms. Nevertheless, its frequent expression in tumors that are negative or exhibit low frequency of CT-X expression presents an opportunity for developing an effective therapeutic vaccine by using this protein in combination with other CT-X proteins.
The levels of expression of potential targets in cancer tissues, as judged by transcript abundance, are an important criterion for selection. In our evaluation of PLAC1 levels, we showed that while the expression of PLAC1 was undetectable in normal tissues except for placenta and testis, it was found to be expressed in cancer cell lines from different origins and lung tumors at high levels.
The absence of PLAC1 expression in adult somatic cells is consistent with its spontaneous immunogenicity. The presence of antibodies to PLAC1 in the sera of a small subset of healthy women could be linked to self-immunization following pregnancy, and this observation requires further investigation into the immunoglobulin isotype subsets elicited and the kinetics of appearance. Importantly, we were also able to detect a spontaneous serological response to PLAC1 in a number of lung cancer patients, including males. This evidence for tumor-related immunogenicity of PLAC1 in cancer patients provides support for the development of PLAC1-specific cancer vaccines. In a paper published in the Chinese literature, Chen et al. (16) demonstrated PLAC1 expression in gastric cancer and PLAC1 immunogenicity in the serum of gastric cancer patients.
In summary, we have demonstrated that PLAC1 is a novel tumor antigen with very restricted normal tissue expression. We also provide evidence that PLAC1 is expressed in the tumor cells in a subset of lung cancer patients. We suggest that PLAC1 may be a potential target for vaccines and for the development of therapeutic monoclonal antibodies due to its potential cell surface localization.
Materials and methods
We used the SAGE Digital Gene Expression Displayer (SDGED) tool to select candidate genes based on their high expression in placenta tissues and expression in different tumor types from SAGE data housed at the CGAP website (10). A total of 207,348 tags from two placenta libraries (SAGE_Placenta_normal_B_1 and SAGE_Placenta_first_trimester_normal_B_1) and 7,413,288 tags from 104 cancer tissue libraries (astrocytoma, meningioma, ependymoma, medulloblastoma, colon adenocarcinoma, lung adenocarcinoma, lymphoma, breast carcinoma, and stomach adenocarcinoma) were included in the analysis. All the libraries selected came from bulk tissues. Genes showing high expression in placenta (more than 100 tags) and present in more than one cancer tissue library were further selected based on their restricted expression in normal tissues using EST data assessed using the Virtual Northern tool at the CGAP website (10). The amino acid sequences encoded by the candidate genes were analyzed for subcellular localization using the PSORT prediction algorithm (17).
Tissues and cell lines
All cell lines used in this study were obtained from the cell bank of the New York Branch of the Ludwig Institute for Cancer Research. They were maintained in RPMI medium containing 10% fetal bovine serum (FBS) and non-essential amino acids. Tumor tissue specimens and paired sera, as well as nine normal lung samples from tumor margins from patients with NSCLC who underwent surgery between October 1996 and February 2006 at the Department of Cardio-Thoracic Surgery, Weill Medical College, were included in this study. The study was approved by the Institutional Review Board. All lung cancer patients provided written informed consent prior to our study. Normal tissue RNA preparations from 21 different tissues were purchased from Clontech Laboratories, Inc. (Palo Alto, CA, USA) and Ambion, Inc. (Austin, Texas, USA) and were used to prepare cDNA for RT-PCR and qRT-PCR. Placentas from termed pregnancies were obtained from the Department of Pathology, Weill Cornell Medical College.
Total RNA from cell lines and tissues was prepared with RNeasy (Qiagen, Hilden, Germany). First strand cDNA was synthesized with 1 µg of total RNA using oligo d(T)12-18 primers or random hexamers (Invitrogen, Carlsbad, CA) and Omniscript Reverse Transcription (Qiagen) in a total volume of 20 µl, and 1/5 of the volume of the cDNA was used in the semi-quantitative polymerase chain reaction (PCR) with primers specific to PLAC1 (PLAC1F 5´CAG ACA CAG CAA GTT CCT TC 3´, PLAC1R 5´CCA TGA ACC AGT CTA TGG AG 3´). After preheating at 94˚C for 5 minutes, the PCR cycles (30 repeats) were as follows: 1 minute at 94˚C, annealing at 68˚C for 2 minutes, and extension at 72˚C for 2 minutes; followed by a final extension for 7 minutes at 72˚C. PCR products (expected product 340 bp) were evaluated by agarose gel (1.5%) electrophoresis. Beta-actin was amplified as control. Primers, reaction conditions and annealing temperatures used for CT-X PCR amplification were as described by Gure et al. (12).
Quantitative real-time RT-PC
cDNA samples were run in duplicate for both PLAC1 and the reference gene within the same experiment using the Applied Biosystems apparatus 7500 Fast Real-Time PCR system and Taqman platform (Applied Biosystems). ACTB was amplified as an internal reference gene. The PCR primers and probes were purchased from Applied Biosystems. Primers used for PCR amplification were chosen to encompass introns between exon sequences to avoid amplification of genomic DNA (Applied Biosystems). The gene-specific probes were labeled with the reporter dye 6-FAM at the 5'-end. The TFRC probe was labeled with a reporter dye (VIC) to the 5'-end of the probe and all probes had minor groove binder/nonfluorescent quencher at the 3'-end of the probe (Applied Biosystems). The PCR conditions were 95˚C for 10 minutes followed by 40 cycles at 95˚C for 15 seconds and 60˚C for 1 minute. Duplicate CT values were averaged for each sample. Relative quantification of gene expression (relative amount of target RNA) was determined using the equation 2 exp (-ΔΔCT).
Generation of recombinant PLAC1 protein
Using the BLASTP tool at the NCBI website (18) , we selected a region of PLAC1 that does not share sequence similarity with other human proteins. The region specific to PLAC1 used to produce recombinant protein was from amino acids 125 to 212 (NP_068568). For the construction of the PLAC1 prokaryotic expression vector, the template vector pBluescriptR containing the cDNA clone BC022335 was PCR amplified with the primers 5'-TTTT GGATCC ACC AAG CCC TGC TCC ATC AGC-3' and 5'-TTTT CTCGAG TCA CAT GGA CCC AAT CAT ATC-3' (the BamHI and XhoI restriction sites, respectively, are underlined). The PCR-amplified products were inserted between the BamHI and XhoI restriction sites of the E. coli expression vector PGEX-4T. The fusion protein (amino acids 125 to 212) was translated in-frame from the vector's start codon. After sequence verification, the resulting prokaryotic expression vector pGEX-4T-PLAC1 was introduced into the bacterial host E. coli following a standard protocol and expression of the fusion protein was induced by adding isopropyl-1-thio-D-galactopyranoside (IPTG). The fusion protein was purified using the GST tag. After 10% SDS-PAGE analysis, a 35 kDa band was detected in the sample of fusion proteins purified by glutathione-sepharose beads.
Polyclonal antibody production
For antibody production, New Zealand rabbits were injected subcutaneously with 50 µg of the PLAC1 recombinant protein obtained as described above. For the first immunization the antigen was admixed 1:1 with complete Freund's adjuvant; for the next four boosts (on days 28, 42, 60 and 78) incomplete Freund's adjuvant was used. Anti-PLAC1 antibody production was determined by enzyme-linked immunosorbent assay (ELISA).
To detect PLAC1 protein, five micron paraffin sections were prepared and dried overnight at 37˚C. These were deparaffinized in xylene and rehydrated through alcohol solutions. Antigen retrieval was performed by microwave boiling in 100 mM citrate buffer pH 6.0 for 15 minutes. Endogenous peroxidase activity was quenched with 0.3% hydrogen peroxide for 5 minutes. Sections were then incubated with PLAC1-specific rabbit polyclonal antibody or pre-immune serum diluted 1 in 1000 in Tris-buffered saline with 10% BSA for 1 hour at room temperature. The slides were then processed using Dako Envision+™ HRP (DAB)- kit (DakoCytomation, Glostrup, Denmark) following the manufacturer's protocol, counterstained briefly with Mayer's hematoxylin (Amber Scientific, Belmont, WA) and cover slipped.
Enzyme-linked immunosorbent assays
Patient plasma samples were analyzed by ELISA for seroreactivity against recombinant PLAC1 protein (amino acids 125 to 212) and against control DHFR or NY-ESO-1 proteins. Briefly, plasma was serially diluted from 1/100 to 1/100,000 and added to low-volume 96-well plates (Corning, NY) coated with 1 µg/ml antigen and blocked with PBS containing 5% non-fat milk. After incubation, the plates were washed automatically (Bio-Tek, Winooski, VT) with PBS containing 0.2% Tween and rinsed with PBS. Plasma IgG (total or subclasses) bound to antigens was detected with alkaline phosphatase-conjugated specific monoclonal antibodies (Southern Biotech, Birmingham, AL). Following addition of ATTOPHOS substrate (Fisher Scientific, Waltham, MA), absorbance was measured using a fluorescence plate reader Cytofluor Series 4000 (PerSeptive Biosystems, Framingham, MA). A reciprocal titer was calculated for each plasma sample as the maximal dilution still significantly reacting to a specific antigen. This value was extrapolated by determining the intersection of a linear trend regression with a cutoff value. The cutoff was defined as 10x the average of the OD values from the first 4 dilutions of a negative control pool of 5 healthy donor sera. Sera with reciprocal titers >100 were considered significantly reactive and specificity was determined by comparing seroreactivity across antigens. Each assay was validated using patient sera of known reactivity.
This work was conducted as part of the Atlantic Philanthropies/Ludwig Institute for Cancer Research Clinical Discovery Program. WASJr. was supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq) (Proc. Num. 201158/2005-1).
- Received October 4, 2007.
- Accepted October 31, 2007.
- Copyright © 2007 by Otavia L. Caballero