Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

L552S, an alternatively spliced isoform of XAGE-1, is over-expressed in lung adenocarcinoma


Using a combination of cDNA subtraction and microarray analysis, we report here the identification and characterization of L552S, an over-expressed, alternatively spliced isoform of XAGE-1 in lung adenocarcinoma. Real-time RT–PCR analysis shows that L552S is expressed at levels greater than 10-fold in 12 of 25 lung adenocarcinoma tumors compared with the highest expression level found in all normal tissues tested. L552S is expressed in both early and late stages of lung adenocarcinoma, but it was not detected in large cell carcinoma, small cell carcinoma, or atypical lung neuroendocrine carcinoid. The full-length cDNA for L552S comprises 770 bp and encodes a polypeptide of 160 amino acids. C-terminal 94 amino acids of L552S are identical to a cancer testis antigen, XAGE-1, found in Ewing's sarcoma. Genomic sequence analysis has revealed that L552S and XAGE-1 are alternatively spliced isoforms, and expression of both L552S and XAGE-1 isoforms are present in lung adenocarcinoma. Immunohistochemistry analysis using affinity purified L552S polyclonal antibodies demonstrated specific nuclear staining in 10 of 12 lung adenocarcinoma samples. Furthermore, antibody responses to recombinant L552S protein were observed in seven of 17 lung pleural effusion fluids of lung cancer patients. These results strongly imply that L552S protein is immunogenic and suggest that it might have use as a vaccine target for lung cancer.


Lung cancer causes 28% of all cancer deaths, in large part because conventional therapies are unsatisfactory. Approaches to cancer treatment with adjuvant vaccine therapy is an area being actively pursued, as indicated by several clinical trials using antigens that are specifically expressed or over-expressed in tumors (Brinckerhoff et al., 2000; Robbins and Kawakami, 1996). Cancer vaccines such as Melacine (Corixa, Seattle, WA, USA) have been shown to be effective in preventing recurrence when administrated as adjuvant therapy for melanoma (reported by the Southwest Oncology Group at the Society of Biological Therapy meeting on October 28, 2000 in Seattle, WA, USA). The success of therapeutic cancer vaccines may ultimately rely on the identification of immunogenic proteins that are over-expressed in tumors relative to essential normal tissues.

We previously reported a high throughput methodology for identifying genes over-expressed in lung squamous cell carcinoma (LSCC). Although lung adenocarcinoma and LSCC are generally categorized as non-small cell lung carcinoma (NSCLC), we have found that the majority of genes over-expressed in LSCC were not over-expressed in lung adenocarcinoma (Wang et al., 2000). Using similar approaches, we identified L552S, an alternatively spliced isoform of XAGE-1, which is over-expressed in lung adenocarcinoma tumors. The specificity of L552S expression was demonstrated by real-time PCR, Northern blot, and immunohistochemistry analyses. Furthermore, patient antibody responses to recombinant L552S protein were detected, strongly implying that L552S encodes a protein expressed by lung cancer and that this protein is immunogenic.


L552S expression in lung adenocarcinoma

Among 358 cDNAs isolated from two subtracted lung adenocarcinoma cDNA libraries, L552S was recovered twice, indicating a successful subtraction. Both L552S-19107 and L552S-19106 cDNA fragments showed similar microarray expression profiles and they exhibited very restricted expression in normal tissues (Figure 1). L552S was found to be preferentially over-expressed in selected lung adenocarcinoma samples, including LPE100-183 and LPE86-52 tumor probes from which subtracted cDNA libraries were generated (Figure 1). Sequence analysis of these two clones revealed identical cDNA sequences of 470 base pairs excluding the poly A tail, except that clone 19106 had a 16 base pair insertion in the middle of the sequence compared with 19017 (see Figure 4a for detail). Clone 19107 was found later to be a dominant form of L552S during full-length cloning, representing over 95% of L552S cDNA. Initial database analysis failed to reveal any homology with Genbank sequences; however, a dbEST database search revealed 14 EST matches including five cDNAs from testis, two cDNAs from germ cells, three cDNAs from alveolar rhabdomyosarcoma, and four cDNAs from Ewing's Sarcoma. Specific expression of L552S in lung adenocarcinoma was further confirmed by real-time PCR analysis using cDNAs from 13 lung adenocarcinoma samples, two LSCC samples, one small cell lung carcinoma, three normal lung samples, and 17 other normal tissues (Figure 2a). In order to determine the frequency of L552S expression in lung adenocarcinoma patients as well as potential correlation with stages of this lung disease, we examined the L552S expression in an extended cDNA panel using real-time PCR. This panel has 26 lung adenocarcinoma cDNAs derived from tumors of different stages as well as one lung adenocarcinoma cell line (390T), one adenocarcinoma propagated in scid mice (sample 22), and four lung pleural effusions of lung adenocarcinoma. As shown in Figure 2b, L552S is over-expressed in 12 of 25 adenocarcinoma and it is present in both early and late stage samples. Expression of L552S was not observed in a panel of normal tissues. Interestingly, although L552S was not expressed in large cell carcinoma or small cell lung carcinoma cDNA samples (Figure 2b, samples 29–31), expression was observed in two LSCC specimens (Figure 2b, samples 27–28). An extended LSCC cDNA panel, however showed expression of L552S in only three of 24 tumors (data not shown).

Figure 1

Microarray analysis of L552S-19107 and L552S-19106. Pseudo-color images of hybridization intensities (white being the strongest and black being the weakest) were shown between 28 pairs of probes in Cy3 and Cy5 channels as indicated. Notice that L552S-19107 and L552S-19106 exhibit similar expression profiles, showing elevated expression in lung adenocarcinoma tumors of 478A, 373A, LPE100-138, LPE86-52, and 372A

Figure 4

(a) DNA alignment of L552S and XAGE-1. Translation initiation codons for L552S and XAGE-1 were denoted by solid arrows, and the common stop codons were indicated by a solid box. Two in-frame stop codons in front of the L552S translation initiation site were also shown by dotted boxes. A 16 bp insert present in L552S-19106 was also indicated. Three pairs of primers used for real-time PCR analysis were also indicated as L552S-FW and L552S-RV (common cDNA between L552S and XAGE-1), XAGE-1 unique-FW and XAGE unique-RV (XAGE-1 specific cDNA), and L552S unique-FW and L552S unique-RV (L552S specific cDNA). (b) alignment of protein sequences for L552S and XAGE-1. L552S encodes a polypeptide of 160 amino acids. It shares the majority of C-terminal sequences with XAGE-1 protein. Peptides #4, #5, #24, and #29 which are immunoreactive with patient lung pleural effusions are shown directly above or below the amino acids of L552S. (c) Kyte-Doolittle hydrophilicity plot for L552S protein. (d) Localization of L552S cDNA and XAGE-1 cDNA to the same chromosomal X clone RP11-485B17 (180 859 bp). In either case, the genomic DNA base pairs are denoted on top of the exon/intron junctions and cDNA base pairs are denoted at the bottom of the exon/intron junctions. Identical cDNA sequences between L552S and XAGE-1 are in shaded boxes. L552S and XAGE-1 specific probes used for Northern analysis in Figure 7 are denoted

Figure 2

Real-time RT–PCR analysis of L552S using a panel of 36 cDNA samples (a) and an extended panel of 66 cDNA samples for lung adenocarcinoma tumors (b). Over-expression is defined as 10-fold expression compared with highest expression in normal tissues. (a) L552S is over-expressed in five of 13 lung adenocarcinoma. (b) Over-expression of L552S was detected in 12 of 25 lung adenocarcinoma tumors, and it is present in all stages of the disease

Full-length cDNA of L552S

Northern blot analysis was performed to determine the message size and complexity of L552S. A single transcript of 700 base pairs was detected in poly(A)+RNA of two lung adenocarcinoma samples but not in poly(A)+RNA of two normal lung samples or trachea (Figure 3). Full-length cloning was performed using the cDNA library of a lung adenocarcinoma. Due to extreme GC-richness or secondary structures at the 5′ end, the majority of these recovered clones failed to reveal any additional 5′ end sequences beyond L552S-19107 or L552S-19106. However, two clones from independent screens did reveal an additional 300 base pairs with a translation initiation start codon (Figure 4a, solid arrow pointing to L552S sequence) and two upstream in-frame stop codons (Figure 4a, dotted boxes).

Figure 3

Northern analysis of L552S. A 700 bp transcript was detected in poly(A)+RNA of two lung adenocarcinoma tumors

The full-length cDNA of L552S has an open-reading frame of 480 base pairs and encodes a putative polypeptide of 160 amino acids. Initial database searches failed to detect any sequence homology with proteins in the database, suggesting that L552S encodes a novel protein of unknown function. However, as this manuscript was in preparation, a cancer-testis antigen, XAGE-1, was found to be over-expressed in Ewing's Sarcoma (Liu et al., 2000). Sequence comparison of L552S and XAGE-1 revealed similarities and differences. The majority of the C-terminal sequences are identical (Figure 4a), but the N-terminal sequences are distinct (Figure 4b). Hydrophilicity analysis of L552S amino acids suggests a very hydrophilic protein with no transmembrane domains (Figure 4c). PSORT analysis of L552S revealed the same result (data not shown). Since L552S is also localized in chromosome X (Figure 5), it is likely to be a new isoform of cancer testis antigen, XAGE-1. The genomic sequence analysis revealed that both genes localized in the same region of the X chromosome and both have four exons (Figure 4d). The last three exons are identical for L552S and XAGE-1. However, the first exon for XAGE-1 is upstream of the first exon for L552S and this results in distinct 5′ nucleotide and amino acid sequences for L552S and XAGE-1. We conclude that L552S and XAGE-1 are alternatively spliced isoforms.

Figure 5

FISH analysis of L552S shown by FISH staining (a), DAPI staining (b), and a diagram of FISH mapping results (c). Each dot in panel c represents the double FISH signals detected on human chromosome X

L552S and XAGE-1 specific expression

In order to delineate the expression of L552S from XAGE-1 in lung cancers, real-time RT–PCR analysis was performed using specific primers localized in the 5′ unique region of L552S, the 5′ unique region of XAGE-1, and 3′ common region (see Figure 4a for detailed primer mapping). Identical expression profiles were observed between 5′ unique region of L552S and the common 3′ sequences (Figure 6a,b). However, the level of abundance for L552S 5′ unique cDNA region was lower. Furthermore, the level of abundance for XAGE-1 cDNA detected in lung tumors using 5′ unique primers of XAGE-1 was much lower compared with L552S. Since the secondary structure posed by L552S and/or XAGE-1 could impede the cDNA synthesis for the 5′ unique sequences of both targets, the real-time RT–PCR analysis may not truly reflect the relative abundance of each isoform expressed in lung cancer. Thus, we assessed the relative abundance of L552S and XAGE-1 using Northern blot analysis. Three identical blots containing two lung tumor samples and one normal lung sample were probed with L552S and XAGE-1 common probe (Figure 7a), the L552S 5′ unique probe (Figure 7b), or XAGE-1 5′ unique probe (Figure 7c). Both L552S and XAGE-1 isoforms were present in lung tumor samples.

Figure 6

Real-time PCR analysis of L552S and XAGE-1 using common primers for L552S and XAGE-1 (a), L552S specific primers (b), or XAGE-1 specific primers (c)

Figure 7

Northern analysis of L552S and XAGE-1 using L552S and XAGE-1 common probe (a), XAGE-1 5′ unique probe (b), or L552S 5′ unique probe (c). L552S and XAGE-1 specific transcripts were detected in RNA from lung adenocarcinoma tumor or tumor cell line, but not from normal lung

L552s protein is localized in the nuclei of lung adenocarcinoma cells

Immunohistochemistry (IHC) analysis using affinity purified L552S polyclonal antibodies revealed that L552S immunoreactivity was observed in the nuclei of lung adenocarcinoma samples (Figure 8a,b). The results of IHC analysis using an extensive panel of lung cancer and normal tissues are summarized in Table 1. Strong ‘dot-like’ nuclear staining was observed in 10 of 12 lung adenocarcinoma samples and this nuclear staining pattern was not observed in any other tissues. Although real-time PCR analysis showed that L552S was also over-expressed in three of 24 LSCC (data not shown), the characteristic nuclear staining was not detected in any of 12 LSCC samples tested by IHC. Similarly, the majority of normal tissues were negative for L552S immunoreactivity, however, some normal tissues such as arterial endothelial cells and liver exhibited light ‘dot-like’ intracytoplasmic staining (see arrows in Figure 8j,k); this staining pattern was not reproducible when multiple samples were tested (Figure 8k,l and Table 1). Thus, L552S protein appears to be specifically expressed in the nuclei of the lung adenocarcinoma cells.

Figure 8

Immunohistochemistry analysis of L552S using affinity purified L552S polyclonal antibodies. Illustrated here are results of immunohistochemical staining on tissue sections of two different lung adenocarcinoma samples (a,b), lung squamous cell carcinoma sample (c), normal lung alveoli (d), normal trachea (e), normal bronchus (f), normal pancreas (g), normal ovary (h), normal cerebellum (i), normal artery endothelial cells (j), and two normal liver samples (k,l). Strong nuclear staining was observed in lung adenocarcinoma samples (a,b). Arrows denote light cytoplasmic staining in artery endothelial cells (j) and very light cytoplasmic ‘dot-like’ staining in one of the liver samples (k) (see Table 1 for the summary)

Table 1 IHC analysis of L552S protein

Antibody responses to L552S and L552S peptides in lung cancer patients

To test whether lung cancer patients can develop antibody responses to L552S protein, ELISA was performed testing recombinant L552S protein on lung pleural effusions from 17 patients (Table 2). Seven pleural effusions from lung cancer patients gave positive signals (more than twofold signals in at least two different titers compared with normal sera of 48 donors). These positive signals are significant as the immunoreactivity could not be abolished using E. coli protein lysates as competitors. Strong antibody titers were observed in two patients, and the specificity of the antibody response in patient lung pleural effusion #3 and #16 was demonstrated by both peptide epitope mapping and Western blot analysis. Using 29 overlapping peptides across L552S protein sequences, the peptide epitope for lung pleural effusion #3 was mapped to peptides 29 and 24, and the peptide epitope was mapped to peptides 4 and 5 for lung pleural effusion #16 (see Materials and methods and Figure 4b). Since peptides 4 and 5 localized to the unique N-terminal amino acid sequences of L552S (different from XAGE-1) and peptides 24 and 29 localized to the C-terminal amino acid sequences of L552S. These data suggest that L552S unique region may be immunogenic, further substantiating its specific expression in lung cancer patients. Western blot analysis of pleural effusion antibody responses in a positive and a negative patient was shown in Figure 9, recombinant L552S protein showed specific antibody reactivity was present in patient pleural effusion #3 but not #13 (Table 2 and Figure 9b).

Table 2 Antibody responses to L552S in lung cancer patients
Figure 9

Western blot analysis of L552S recombinant protein. (a) Silver staining of recombinant L552S protein purified through nickel chromatography (lane 2). The arrow points to L552S protein confirmed by N-terminal sequencing analysis. B, L552S recombinant protein was probed with pleural effusion from lung cancer patient #3 (lane 1′) and #13 (lane 2′). Notice that L552S protein as well as L552S aggregates or oligomers were detected by patient fluid #3 (lane 1′)


In this report, we have identified and characterized a new cancer testis antigen, L552S, through a combination of cDNA subtraction and microarray analysis. L552S is located on chromosome X and is over-expressed in lung adenocarcinoma as determined by cDNA microarray, real-time PCR, and immunohistochemistry analysis. Furthermore, L552S encodes an immunogenic protein and antibody responses to recombinant L552S protein were detected in seven of 17 pleural effusion fluids of lung cancer patients.

Following the observation that T cells derived from human cancer patients are capable of recognizing tumor antigens in an MHC class I-restricted manner, the first cancer-testis antigen, MAGE-1 was isolated in 1991 (van der Bruggen et al., 1991). Over the last 10 years, many cancer-testis antigens including MAGE, GAGE, BAGE, LAGE, SCP-1, SSX2, and NY-ESO-1 etc. have been identified by using immunological approaches such as T cell expression cloning (Chen and Old, 1999; Van den Eynde et al., 1995; Boel et al., 1995; Tureci et al., 1998) and serological screening (SEREX) (Tureci et al., 1996; Chen et al., 1998). Recently, bioinformatic and molecular approaches brought additional members such as XAGE, PAGE, and SAGE (Brinkmann et al., 1999; Martelange et al., 2000). Although antigens within each family such as GAGE-1, GAGE-2, GAGE-3, and GAGE-4 share sequence homologies, the overall sequence similarity between genes of different families are extremely limited (Brinkmann et al., 1999). Activation of these genes is believed to be associated with malignant transformation.

L552S share significant sequence identity with XAGE-1 and they are alternatively spliced isoforms (Figure 4). The predicted L552S protein is hydrophilic whereas N-terminal sequence of XAGE-1 protein is hydrophobic (Liu et al., 2000). Although real-time PCR analysis suggests that L552S isoform is expressed at a higher level compared with XAGE-1 (Figure 6), the Northern blot data suggests that both isoforms are present in lung cancers (Figure 7). Since strong secondary structure is present at the 5′ end of L552S and XAGE-1 messages, neither real-time nor Northern analysis can truly reflect the relative abundances between L552S and XAGE-1. Nevertheless, the L552S isoform is specifically over-expressed in lung adenocarcinoma. Peptide epitope mapping studies using positive patient lung pleural effusions suggest that the unique N-terminal amino acid sequence for L552S is immunogenic and is consistent with the expression of L552S unique transcripts. Since the immunohistochemistry analysis was performed using affinity purified L552S polyclonal antibodies which recognize both L552S 5′ unique and L552S/XAGE-1 common epitopes (data not shown), we cannot determine which epitope(s) is contributing to the nuclear staining of lung adenocarcinoma samples. Generation of monoclonal antibodies to L552S is in progress and epitope specific monoclonal antibody will be carried out to delineate L552S and XAGE-1 specific expression in future immunohistochemistry analyses. However, based on mRNA analysis, it is likely that protein from both spliced isoforms is expressed in lung adenocarcinoma.

L552S is the first cancer testis antigen shown to be specifically over-expressed in lung adenocarcinoma by both mRNA and protein analyses. A previous report has suggested that 32% of lung cancer specimens were positive to MAGE-1 antibody, however, the majority of the positive specimens showed low and heterogeneous staining (Jungbluth et al., 2000). Similarly, RT–PCR analysis indicated that MAGE-3 gene was present in 45% lung carcinoma but the level of MAGE-3 expression was not quantitative (Yoshimatsu et al., 1998). Because of its high prevalence of expression and apparent immunogenicity, L552S represents an attractive candidate for therapeutic cancer vaccines in patients who over-express this protein. At present, we do not know whether the antibody response in lung cancer patients represents a successful or unsuccessful immune response, or whether the level of L552S expression in tumors bears a direct or inverse correlation with serum antibody responses. Further studies are underway to address this relationship. In addition, studies of L552S immunogenicity using in vitro T cell priming are also underway to evaluate L552S as a potential therapeutic target for lung cancer.

Materials and methods

Tissue and RNA sources

Tumor and normal tissues used in this study were obtained from Cooperative Human Tissue Network (CHTN), National Disease Research Interchange (NDRI), and Roswell Park Cancer Center. Other normal tissues were purchased from Clontech (Palo Alto, CA, USA) and Invitrogen (Carlsbad, CA, USA).

Construction of a cDNA library for metastatic lung adenocarcinoma

A human lung adenocarcinoma cDNA library was constructed from a pool of two poly(A)+RNA extracted from two patient lung pleural effusions (LPE100-183 and LPE86-52) as described (Wang et al., 2000). The primary sites for both LPE tumors are lung adenocarcinoma. A total of 1.7×106 independent colonies were obtained, with 100% of clones having inserts and the average insert size being 1200 base pairs.

Construction of cDNA libraries using normal tissues

Using the same procedure, three human normal tissue cDNA libraries were prepared with poly(A)+RNA extracted from a pool of four human normal lung tissues, a pool of normal human liver and heart specimens, and a pool of normal tissues of lung, liver, pancreas, PBMC, skin, kidney, and brain. The number of independent colonies in each library range from 1.4–1.7×106 with more than 90% of clones having inserts and the average insert size being 1500–1800 base pairs.

Lung adenocarcinoma specific subtracted cDNA libraries

To enrich for genes preferentially expressed in lung adenocarcinoma, we performed cDNA library subtractions using the above lung adenocarcinoma cDNA library as the tester and normal tissue cDNA libraries as drivers, as previously described (Wang et al., 2000). Two subtracted cDNA libraries were generated and they are referred as METS-S2 and METS-S3. To analyse each subtracted library, 48–96 clones were randomly picked and plasmid DNA was prepared for sequence analyses with a Perkin Elmer/Applied Biosystems Division Automated Sequencer Model 373A and/or Model 377 (Foster City, CA, USA). These sequences were compared to sequences in the GenBank and human EST databases.

cDNA microarray analysis of L552S

Fifteen cDNA of METS-S2 and 343 cDNA from METS-S3 subtracted cDNA libraries were PCR amplified and arrayed in high density as multiple replicas also known as chips (Incyte, Palo Alto, CA, USA). Expression profiles of these cDNAs including L552S were evaluated using 28 pairs of cDNA probes synthesized from mRNA of lung tumor, normal lung, and other tumor and normal tissues as described previously (Wang et al., 2000). The expression profile for L552S was illustrated in pseudo colors representing pairs of Cy3 and Cy5 hybridization signals, white being the highest and black being the lowest (see Figure 1 for detail).

Northern analysis of L552S

Northern blot analysis was performed using 1.0 μg or mRNA or 10 μg of total RNA from lung tumors and normal tissues of lung and trachea. Blot was stained with 0.02% methylene blue to reveal the heterogeneous population of mRNA before hybridization was proceeded with. Pictures and autoradiographs were scanned and processed through Photoshop 4.0. Probes used for Northern analysis of L552S or XAGE-1 specific isoforms are generated from L552S or XAGE-1 unique PCR fragments, respectively (Figure 4d).

Quantitative real-time RT–PCR analysis

To compare the relative level of gene expression in multiple tissue samples, a panel of 36 to 66 cDNA samples was constructed using total RNA extracted from tissues and/or cell lines, and real-time PCR was performed using gene specific primers to quantify the copy number in each cDNA sample. Each cDNA sample was done in duplicate and each reaction repeated in duplicated plates. The final Real-time PCR result was reported as an average of copy number of a gene of interest normalized against internal actin number in each cDNA sample. Four cDNA panels were constructed and used for real-time PCR analysis. One panel focused on selected normal and tumor tissues and includes cDNAs of lung adenocarcinoma samples, other types of lung tumors, and a variety of normal tissues (Figure 2a). The second panel includes 66 cDNA samples done on two separate plates using six over-lapping cDNAs for normalizing two plates. This panel has 26 lung adenocarcinoma cDNAs involving different stages of the disease including one lung adenocarcinoma cell line (390T), one adenocarcinoma propagated in scid mice (sample 22), and four lung pleural effusions of lung adenocarcinoma (Figure 2b). The third panel is similar to the second except that 26 cDNAs of squamous tumors were used including 20 LSCC, two lung adeno-squamous tumors, two head & neck squamous tumors, and two LSCC propagated in scid mice (data not shown). The fourth panel is similar to the first panel with majority of cDNAs replaced by newer samples as old ones became exhausted, and a testis cDNA was also included (Figure 6). L552S-FW (5′-CTGAAAGTCGGGATCCTACA-3′) and L552S-RV (5′-AATCCAGATTTATCCCG-3′) primers were used for real-time PCR analysis reported in Figures 2 and 6a, whereas L552S unique-FW (5′-GTTGTGTGGTCAGTGACTCAGAG-3′) and L552S unique-RV (5′-CATGAGAGGGACGACGACTTC-3′) primers were used for L552S specific expression (Figure 6b) and XAGE-1 unique-FW (5′-CAGGGCAAGGCGGGATAAG-3′) and XAGE-1 unique-RV (5′-AGACGCCCAGTGAACATGC-3′) primers were used for XAGE-1 specific expression (Figure 6c). Known copy number of L552S plasmid DNA or XAGE-1 PCR product (amplified from human testis cDNA and confirmed by sequence analysis) was used as standards for L552S or XAGE-1. Real-time PCR reactions were performed on a GeneAmp 5700 Detector using SYBR Green I dye (Perkin Elmer/Applied Biosystems Division, Foster City, CA, USA).

Full-length cloning of L552S

Full-length cloning of L552S was performed using colony hybridization according to standard procedure. In general, 100 000 to 500 000 colonies from the above lung adenocarcinoma cDNA library were screened for each attempt, and 12 to 20 clones were picked from the primary screening and proceeded for the secondary screening. Two clones were then picked from each primary pick on the secondary screening, and DNA was prepared and sequenced on an Automated Sequencer Model 373 or 377 (Perkin/Elmer Applied Biosystems, Foster City, CA, USA). The analysis, assembly, and alignment of sequences were performed using Lasergene software (DNASTAR, Inc. Madison, WI, USA). PSORT software was used for prediction of localization and topology of the protein.

Fluorescence in situ hybridization (FISH) analysis of L552S

The FISH analysis was performed using full-length cDNA of L552S as a probe. Preparation of slides, biotinylation of DNA probes, hybridization, and analysis were performed by SeeDNA Biotech as previously described (Heng et al., 1992; Heng and Tsui, 1993).

L552S protein expression and polyclonal antibody generation

E. coli recombinant L552S protein was expressed in pPDM His, a modified pET28 vector (Novagen Inc., Madison, WI, USA). Amino acid 2–160 of L552S was expressed in frame with an N-terminal His tag (MQHHHHHH). L552S protein was purified through nickel chromatography according to standard protocol. L552S protein was subjected to N-terminal sequencing analysis to confirm the identity. The protein was used for immunization of rabbits for generating polyclonal antibodies.

Immunohistochemistry analysis

An affinity purified anti-L552S polyclonal antibody was subjected to immunohistochemistry analysis by QualTek Molecular Laboratories (Santa Barbara, CA, USA). In all cases, four micron sections of formalin fixed, paraffin-embedded tissues were used. Tissues were subjected to enzyme and steam heat induced epitope retrieval before being stained with 2.5 μg/ml of anti-L552S rabbit polyclonal antibody. Tissues were then incubated with a biotinylated anti-rabbit secondary antibody. After endogenous peroxidase blocking, the Avidin-Biotin Complex/HRP (ABC/HRP) was used along with DAB chromogen to visualize protein expression.

ELISA and Western blot analysis

ELISA was performed using recombinant L552S protein with lung pleural effusion fluids from lung cancer patients and sera from normal donors. Titers of fluids from 17 patients as well as 48 sera from normal donors were tested at 1:30, 1:100, 1:300, and 1:1000 dilutions. 1:100 dilutions of patient pleural effusion fluids #3 and #13 were used for Western blot analysis. For epitope mapping, a total of 29 overlapping peptides representing L552S polypeptide were synthesized and tested for their immunoreactivities with lung pleural effusions. The peptides are: peptide 1, MRCHAHGPSCLVTAITREEG; peptide 2, HGPSCLVTAITREEGGPRSG; peptide 3, LVTAITREEGGPRSGGAQAK; peptide 4, TREEGGPRSGGAQAKLGCCW; peptide 5, GPRSGGAQAKLGCCWGYPSP; peptide 6, GAQAKLGCCWGYPSPRSTWN; peptide 7, LGCCWGYPSPRSTWNPDRRF; peptide 8, GYPSPRSTWNPDRRFWTPQT; peptide 9, RSTWNPDRRFWTPQTGPGEG; peptide 10, PDRRFWTPQTGPGEGRHERH; peptide 11, WTPQTGPGEGRHERHTQTQN; peptide 12, GPGEGRHERHTQTQNHTASP; peptide 13, RHERHTQTQNHTASPRSPVM; peptide 14, TQTQNHTASPRSPVMESPKK; peptide 15, HTASPRSPVMESPKKKNQQL; peptide 16, RSPVMESPKKKNQQLKVGIL; peptide 17, ESPKKKNQQLKVGILHLGSR; peptide 18, KNQQLKVGILHLGSRQKKIR; peptide 19, KVGILHLGSRQKKIRIQLRS; peptide 20, HLGSRQKKIRIQLRSQCATWK; peptide 21, QKKIRIQLRSQCATWKVICK; peptide 22, IQLRSQCATWKVICKSCISQ; peptide 23, QCATWKVICKSCISQTPGIN; peptide 24, KVICKSCISQTPGINLDLGS; peptide 25, SCISQTPGINLDLGSGVKVK; peptide 26, TPGINLDLGSGVKVKIIPKE; peptide 27, LDLGSGVKVKIIPKEEHCKM; peptide 28, GVKVKIIPKEEHCKMPEAGE; peptide 29, IIPKEEHCKMPEAGEEQPQV.


  1. Boel P, Wildmann C, Sensi ML, Brasseur R, Renauld JC, Coulie P, Boon T, van der Bruggen P . 1995 Immunity 2: 167–175

  2. Brinckerhoff LH, Thompson LW, Slingluff Jr CL . 2000 Curr. Opin. Oncol. 12: 163–173

  3. Brinkmann U, Vasmatzis G, Lee B, Pastan I . 1999 Cancer Res. 59: 1445–1448

  4. Chen YT, Gure AO, Tsang S, Stockert E, Jager E, Knuth A, Old LJ . 1998 Proc. Natl. Acad. Sci. USA 95: 6919–6923

  5. Chen YT, Old LJ . 1999 Cancer. J. Sci. Am. 5: 16–17

  6. Heng HH, Squire J, Tsui LC . 1992 Proc. Natl. Acad. Sci. USA 89: 9509–9513

  7. Heng HH, Tsui LC . 1993 Chromosoma 102: 325–332

  8. Jungbluth AA, Stockert E, Chen YT, Kolb D, Iversen K, Coplan K, Williamson B, Altorki N, Busam KJ, Old LJ . 2000 Br. J. Cancer 83: 493–497

  9. Liu XF, Helman LJ, Yeung C, Bera TK, Lee B, Pastan I . 2000 Cancer Res. 60: 4752–4755

  10. Martelange V, De Smet C, De Plaen E, Lurquin C, Boon T . 2000 Cancer Res. 60: 3848–3855

  11. Robbins PF, Kawakami Y . 1996 Curr. Opin. Immunol. 8: 628–636

  12. Tureci O, Sahin U, Schobert I, Koslowski M, Scmitt H, Schild HJ, Stenner F, Seitz G, Rammensee HG, Pfreundschuh M . 1996 Cancer Res. 56: 4766–4772

  13. Tureci O, Sahin U, Zwick C, Koslowski M, Seitz G, Pfreundschuh M . 1998 Proc. Natl. Acad. Sci. USA 95: 5211–5216

  14. Van den Eynde B, Peeters O, De Backer O, Gaugler B, Lucas S, Boon T . 1995 J. Exp. Med. 182: 689–698

  15. van der Bruggen P, Traversari C, Chomez P, Lurquin C, De Plaen E, Van den Eynde B, Knuth A, Boon T . 1991 Science 254: 1643–1647

  16. Wang T, Hopkins D, Schmidt C, Silva S, Houghton R, Takita H, Repasky E, Reed SG . 2000 Oncogene 19: 1519–1528

  17. Yoshimatsu T, Yoshino I, Ohgami A, Takenoyama M, Hanagiri T, Nomoto K, Yasumoto K . 1998 J. Surg. Oncol. 67: 126–129

Download references


We would like to thank Qualtek for their expert IHC analysis. We also thank Dr Martin A Cheever for his critical reading of this manuscript.

Author information



Corresponding author

Correspondence to Tongtong Wang.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Wang, T., Fan, L., Watanabe, Y. et al. L552S, an alternatively spliced isoform of XAGE-1, is over-expressed in lung adenocarcinoma. Oncogene 20, 7699–7709 (2001).

Download citation


  • tumor antigen
  • lung adenocarcinoma
  • lung squamous cell carcinoma (LSCC)
  • XAGE-1
  • cancer testis antigen

Further reading


Quick links