• XAGE;
  • cancer/testis antigen;
  • Ewing's sarcoma;
  • melanoma;
  • PAGE;
  • GAGE


  1. Top of page
  2. Abstract
  6. Acknowledgements

The existence of XAGE genes was first reported after database homology searches for PAGE-like sequences identified 3 XAGE EST clusters. One of these clusters, XAGE-1, has in later studies been identified as a cancer/testis-associated gene. Here, we report the expression profiles of all 3 reported XAGE genes, as well as several splice variants of XAGE-1, in normal human tissues, Ewing's sarcoma and melanocytic tumors. We also provide the genetic structure of the corresponding genes. Moreover, by searching the databases for XAGE homologues, we identified 3 additional GAGE-like genes. RT-PCR studies showed frequent expression in melanoma metastases and Ewing's sarcoma for 2 XAGE-1-derived transcripts. XAGE-2 was expressed at lower frequency in these tissues, while XAGE-3 was seen only in normal placenta. Due to a frameshift, the largest XAGE-1 putative protein is far less homologous to GAGE-like proteins than the other XAGEs. Interestingly, all GAGE-like genes contain a large secondary open reading frame, coding for putative proteins homologues to the XAGE-1 primary protein. The XAGE family of cancer/testis-associated genes is located on chromosome Xp11.21-Xp11.22. The data outline a superfamily of GAGE-like cancer/testis antigens, consisting of at least 19 genes. © 2002 Wiley-Liss, Inc.

During the last decade, there has been increased interest in cancer/testis-associated genes (CTAs). Since the detection of the first MAGE (melanoma antigen), many MAGEs1 and MAGE-like genes as well as other CTAs have been found. CTAs, often present as multigene families on the X chromosome, are in general expressed only in normal testis and cancerous tissues.2 This expression pattern qualifies them as potential immunotherapy candidate genes. So far, very little is known about the role of CTAs in gametogenesis and tumor development.3

Based on earlier work, we were interested in the GAGE-like XAGE genes.4, 5 Since the detection of the GAGE family (GAGE-1–GAGE-6) of CTAs,6 the number of GAGE-homologous genes has been steadily growing. PAGE-1, a GAGE-like gene mainly expressed in prostatic tissues, was found by mRNA differential display using prostate cancer cell lines.7 Other members of the PAGE family (PAGE-2, -3 and -4) were found by computerized EST database mining.8 Expression profiling and alignment of the various GAGEs was reported by De Backer et al.9 Except for GAGE-1, which contains an additional exon, the GAGE members vary mainly by amino acid substitutions and, therefore, are highly homologous. The PAGEs are far less homologous with the GAGEs as well as with each other.

Screening EST databases for PAGE-4 homologous genes identified the XAGE family, consisting of 3 homologous EST clusters (XAGE-1, -2 and -3), which are mainly expressed in testis, placenta and sarcoma tissue.10 The XAGE-1 mRNA, which has substantial homology to the GAGE/PAGE genes only at its C terminus, was reported by the same group.11 They demonstrated that XAGE-1 was expressed in testis and in about 50% of the sarcoma samples tested. Previously, we detected a smaller transcript of XAGE-1 in a screen for genes differentially expressed during melanocytic tumor progression.4 This transcript, XAGE-1b, was expressed in a higher percentage and at a higher level in melanoma and Ewing's sarcoma compared to the previously described XAGE-1 transcript (here named XAGE-1a).11 During experiments to discriminate XAGE-1b from XAGE-1a expression and by analyzing the ESTs of the XAGE-1 cluster (Unigene Hs.112208), we found 2 other alternatively spliced XAGE-1 variants, XAGE-1c and XAGE-1d. By HTGS database searches, we identified the corresponding genes of XAGE-1, -2 and -3 on BAC clones. Based on these data, we outlined their genetic structure and aligned the putative proteins with other members of the GAGE/PAGE family. Finally, we determined the expression profile of the XAGE family of CTAs in normal human tissues, Ewing's sarcoma and melanocytic tumors.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Human Tissues

Normal human tissues, Ewing's sarcoma samples and lesions from all stages of melanocytic tumor progression (common nevus, atypical nevus, primary melanoma and melanoma metastasis) were obtained from patients at the University Medical Center (UMC) St. Radboud. As normal human tissues, we used disease-free samples from surgically removed tissues or from autopsies with a postmortem delay shorter than 4 hr. Representative slices from the tissue samples were snap-frozen in liquid nitrogen and stored at –80°C until use.

Cell Lines

We used 4 Ewing's sarcoma cell lines (Ew1, Ew2, Ew3 and SK-ES-1). Ewing's sarcoma cell lines Ew1, Ew2 and Ew3, all grown in DMEM supplemented with 10% FCS and antibiotics, were generously supplied by B. Jansen (Department of Human Genetics, UMC St. Radboud). Total RNA from SK-ES-1 (ATCC, Rockville, MD) was kindly provided by R. Willems (Department of Pathology, UMC St. Radboud).

RNA Isolation

Total RNA was isolated from cultured cells using the RNeasy kit (Qiagen, Hilden, Germany). From tissue samples, total RNA was isolated by disrupting about 25 frozen sections of 20 μm thickness in 1 ml RNAzolB (Campro, Veenendaal, the Netherlands) using a pestle. The RNAzolB method was followed by an additional RNeasy cleaning step. For all methods, we followed the manufacturer's protocols.


Synthesis of cDNA (10′ at 25°C, 59′ at 42°C, 5′ at 95°C) was performed on 1.0 μg of total RNA. The 20 μl reaction mixture contained 250 pmol of random hexadeoxynucleotide primers, 3 mM MgCl2, 200 μM of each dNTP, 50 mM TRIS-HCl (pH 8.3), 75 mM KCl, 10 mM DTT and 200 U MMLV reverse transcriptase (Promega, Madison, WI). For amplification, 1 μl of cDNA was supplemented with 2.5 μl PCR buffer [200 mM (NH4)2SO4, 750 mM TRIS-HCl (pH 9), 0.1% Tween], 5 μl 1 M dNTPs, 10 pmoles of each primer, 2 μl 25 mM MgCl2, 0.15 U of Thermoperfectplus DNA polymerase (Integro, Zaandam, the Netherlands) and water to a final volume of 25 μl. PCR conditions were 30 sec at 94°C, 45 sec at 60°C and 60 sec at 72°C for 30 cycles. These cycles were preceded by a 3 min denaturation step at 94°C and followed by a 5 min elongation step at 72°C. For nested PCRs, comprising an additional 30 PCR cycles, 2 μl of diluted (100×) product of the first PCR were used under the same buffer conditions as described above. DNA m.w. markers were from Roche Diagnostics. For each sample, 5 μl of product was analyzed on an agarose gel using ethidium bromide staining.

To specifically amplify 6 different XAGE transcripts, we used PCR with the primers listed in Table I and shown in Figure 1. Primers, product lengths and positions in the corresponding mRNAs were as follows: XAGE-1a first PCR (467 bp), F1a (92–110) * R1 (540–559) and nested (252 bp), NF1ab (263–283) * NR1 (496–515); XAGE-1b first PCR (356 bp), F1b (47–67) * R1 (384–403) and nested (252 bp), NF1ab (107–127) * NR1 (340–359); XAGE-1c first PCR (657 bp), F1c (21–39) * R1 (659–678) and nested (459 bp), NF1c (176–195) * NR1 (616–635); XAGE-1d first PCR (212 bp), F1d (207–225) * R1 (400–419) and semi-nested (169 bp), NF1d (207–225) * NR1 (357–376); XAGE-2 first PCR (444 bp), F2 (105–124) * R2 (526–549) and nested (251 bp), NF2 (264–284) * NR2 (496–515); XAGE-3 first PCR (468 bp), F3 (30–49) * R3 (467–498) and nested (361 bp), NF3 (111–130) * NR3 (454–472). All first rounds of PCR as well as the nested PCRs consisted of 30 cycles. For XAGE-1d, we used a semi-nested set-up (F1d = NF1d). Only samples positive for both β2-microglobulin and porphobilinogen deaminase control PCRs were included.

Table I. PCR Primers
  • 1

    Semi-nested (F1d = NF1d).

thumbnail image

Figure 1. Alignment of XAGE cDNAs. Exon boundaries are indicated by vertical lines and sharp arrows. Primers used in the first round of PCR are shown with closed arrows; nested primers are represented by arrows with open heads. Start and stop codons are shown in bold. The polyA-signal sequence is underlined.

Download figure to PowerPoint

cDNA Sequencing and Homology Searching

All sequence analyses of PCR products were done using the Big Dye terminator kit and ABI-3700 DNA analyzer (Perkin-Elmer, Norwalk, CT). Homology searches and sequence alignments were performed with BLAST12 and other software on various public servers of DNA and protein databases.13


  1. Top of page
  2. Abstract
  6. Acknowledgements

EST Database Query

Database queries show that the XAGE family of CTAs consists of 3 homologous EST clusters.10XAGE-1 transcripts are gathered in Unigene cluster Hs.112208, in which almost all ESTs are located at the 3′ part of the XAGE-1a transcript (AF251237),11 corresponding to the length of the more frequently expressed XAGE-1b transcript described earlier.4 In PCR experiments discriminating XAGE-1b from XAGE-1a, we found a PCR product of unexpected length, possibly representing a splice variant (data not shown). Sequencing revealed that it contained an insert of 220 bp compared to the XAGE-1b mRNA. This transcript was represented in the XAGE-1 Unigene cluster by 2 ESTs (BE875341 and BE881023) and was annotated XAGE-1c (AJ318878). We also identified XAGE-1d (AJ318879), a variant with a 16 bp insert compared to XAGE-1b (represented by ESTs AW386180 and AW386185). The XAGE-2 and XAGE-3 ESTs10 belong to the GAGE-8 similar Unigene cluster Hs.16323 and the GAGE-1 similar Unigene cluster Hs.43879, respectively. In these Unigene clusters, we could not detect any splice variant.

HTGS Database Query

To study the genomic XAGE sequences, we performed a BLAST search for the consensus EST cluster sequences using the HTGS database. For all 3 XAGEs, we found at least 1 BAC clone containing the genetic sequence. The XAGE-1 gene, containing 4 exons (AJ400997),4 was present in clone AC025553.5 (152,000–157,000 bp). The XAGE-2 gene was present on the same BAC clone in 2 distinct regions [exons 1 and 2 at 24,000–25,600 bp (reverse orientation) and exons 3–5 at 50,000–56,000 bp). This separation and reverse orientation of 2 parts of the gene is most likely caused by the unordered subfinal status of the BAC clone. XAGE-3 was present in clones AL445227.7 (89,500–95,000 bp) and AL445236.22 (39,000–44,500 bp). Both XAGE-2 (AJ318891 and AJ318892) and XAGE-3 (AJ318893) genes contain 5 exons.

Using a radiation hybrid panel (Stanford G3RH), we could map XAGE-1 on chromosome Xp11.21-Xp11.22.4 Since AL445236.22 is mapped on chromosome Xp11.21-Xp11.3 and XAGE-2 is present on the same BAC clone as XAGE-1, all 3 XAGEs are present at the same chromosomal region. Next to these 3 XAGE genes, we found several BAC clones containing highly homologous sequences, which will briefly be discussed later.

XAGE Genes, Transcripts and Open Reading Frames

In Figure 1, an alignment of the 4 different transcripts of the XAGE-1 gene and the transcripts of XAGE-2 (AJ318880) and XAGE-3 (AJ318881) is shown. The origin of the XAGE-1a and XAGE-1b transcripts was discussed earlier.4 The alternative transcripts XAGE-1c and XAGE-1d contain an additional 220 bp sequence of intron 1 (exon 1′) and a 16 bp sequence of intron 2 (exon 2′), respectively. These alternatively spliced parts are derived from consensus splice sites.14

Homology between XAGE-1 and the other 2 XAGE genes (XAGE-2 and -3) is mainly present in the part starting at 37 bp in exon 2 of XAGE-1, corresponding to the start of exon 3 in XAGE-2 and -3. XAGE-2 differs from XAGE-3 mainly by an elongated exon 1. Since the start of the open reading frame (ORF) for XAGE-2 and -3 is in exon 2, this does not change the length of the putative proteins.

Schematic representations of the 3 XAGE genes, their transcripts and ORFs are given in Figure 2. The ORF of XAGE-1a, the largest XAGE-1 transcript, starts upstream in exon 1 compared to the other 3 XAGE-1 variants. The XAGE-1b ORF starts further downstream in exon 2 and codes for a smaller putative protein, which is in-frame with the putative XAGE-1a protein. The main ORF of XAGE-1c starts in exon 1′ but shares exons 2–4 with XAGE-1a and -1b.XAGE-1d shares the initiation codon of XAGE-1b, but its reading frame is shifted due to the 16 bp insertion (exon 2′). All 4 XAGE-1 main ORF proteins share a nuclear localization signal (Fig. 3, boxed). Interestingly, all XAGE transcripts contain a relatively large secondary ORF (out of frame), though for XAGE-1 these lack a start codon (dotted bars in Fig. 2). Because of its frame shift, the C-terminal part of the primary XAGE-1d ORF is identical to the C-terminal part of the alternative reading frame of the other 3 putative proteins.

thumbnail image

Figure 2. Schematic representation of XAGE-1, -2 and –3 genes, transcripts and ORFs. Exons are coded with numbered boxes. Primary ORFs are shown by black bars; secondary ORFs are shown by dotted bars. Asterisk indicates a stop codon in a reading frame without a starting methionine. Introns with double diagonal lines are not drawn to scale. The size of intron 2 for XAGE-2 could not be deduced from BAC clone AC025553. Transcription-start sites in exon 1 are indicated in italics.

Download figure to PowerPoint

thumbnail image

Figure 3. Alignment of XAGE-1 putative proteins. (a) Primary ORFs. Nuclear localization signal is boxed. (b) Alternative ORF (indicated with #). Asterisk marks a stop codon.

Download figure to PowerPoint

Expression Profiles of XAGEs

To determine whether all XAGE transcripts have a CTA expression profile, we tested the expression of XAGE-1, -2 and -3 mRNAs in normal tissues (n = 29), melanocytic lesions (n = 36) and Ewing's sarcoma (n = 9) using the primer sets described in Table I and in Material and Methods. Sixty cycles of RT-PCR were conducted to show a specific CTA expression pattern, even of the low-abundance XAGE variants. Regarding normal tissues, we performed RT-PCR on 4 different samples each of testis and placenta. From other tissues, except brain (n = 3), we tested only 1 sample. XAGE transcripts were found only in testis, placenta and ovary (Table II). In most testis samples, 5 of the 6 different XAGE transcripts could be detected after 60 cycles of PCR. XAGE-3 was undetectable in all 4 testis samples. Only the XAGE-1b PCR was positive in testis after the first round of PCR (30 cycles). XAGE-2 and -3 were expressed at higher levels in placenta compared to testis. After the first PCR rounds, XAGE-3 was seen in 2 of 4 placenta samples, while all 4 samples were positive for XAGE-2. Regarding expression of XAGE-1 transcripts in placenta, we did not find expression of XAGE-1a but XAGE-1b, -1c and -1d could be detected in 1 of 4 samples after nested PCR. Next to testis and placenta, XAGE-2 was also expressed in ovary.

Table II. Expression profile in XAGE-Positive normal human tissues
  • 1

    Other tissues tested (n = 1, but for brain n = 3) negative for all 6 transcripts: appendix, artery, bladder, bone marrow, brain, duodenum, esophagus, gall bladder, heart, ileum, kidney, larynx, liver, lymph node, medulla (spinal), muscle (psoas), pancreas, prostate, rectum, spleen, stomach, thymus, thyroid, uterus and vein.

  • 2

    Total number of PCR cycles.

  • 3

    —, negative; numbers are positive PCR results.


From the origin of the ESTs belonging to the XAGE Unigene clusters and from previous reports, regular expression of XAGEs was expected in Ewing's sarcoma.4, 11 All 4 Ewing's sarcoma cell lines tested were positive for XAGE-1a, -1b and -1d after 30 cycles (Table III). XAGE-1c was detected in 3 of 4 cell lines after the nested PCR. Only 1 cell line showed a XAGE-1c PCR product after the first round of PCR (30 cycles). XAGE-2 was found only in cell line Ew1 after 60 cycles. All Ewing's sarcoma cell lines were negative for XAGE-3. In a number of fresh tumor samples of Ewing's sarcoma, we could also detect all 4 XAGE-1 transcripts. One round of PCR yielded products for XAGE-1a, -1b, -1c and -1d in, respectively 3, 3, 2 and 5 of 9 samples tested. Nested XAGE-1 PCRs were positive except for 1 XAGE-1a PCR and 3 XAGE-1c PCRs. Although detectable only after nested PCR, XAGE-2 expression was seen in a much higher percentage of Ewing's sarcoma lesions (89%, 8 of 9) compared to the cell lines (25%, 1 of 4). A XAGE-3 PCR product was seen after 60 cycles in 1 tumor sample, but sequencing revealed it to be a XAGE-3-like sequence, which we named XAGE-4 (AJ318895). All other sequenced PCR products (several randomly selected products per XAGE-specific PCR) revealed the right sequences.

Table III. XAGE expression profile in Ewing's Sarcoma
  • 1

    This sequenced product was the XAGE-3-like XAGE-4 (AJ318895).

  • 2

    Total number of PCR cycles.

Ewing's sarcoma cell linesEW1+++++++
Ewing's sarcoma lesionsES1+++

Since we were interested in human melanoma, from which XAGE-1b was isolated,4 we tested several samples from all stages of the melanoma progression lineage for XAGE expression (Table IV). In these samples, substantial expression (PCR product after 30 cycles) was seen only in melanoma metastases for XAGE-1b (43%, n = 23) and -1d (35%). Although after nested PCR most XAGE transcripts were detectable in several melanoma metastases, some were also found in a high percentage of nevus samples after 60 cycles. XAGE-2 nested PCR was even positive for all 8 nevi tested, though detectable in only 1 primary melanoma and 3 melanoma metastases. Regarding primary melanoma, this was the only sample with a positive XAGE PCR. All melanocytic tumor samples lacked XAGE-3 expression.

Table IV. Xage expression profile in the Human Melanocytic Tumor progression lineage
  • 1

    Total number of PCR cycles.

  • 2

    Pooled (2–4 lesions/sample).

Normal skinNH1
Nevus nevocellularis2NN1+
Atypical nevus2AN1+
Primary melanomaPM1
Melanoma metastasisMM1

XAGE-like Genes

To discover the gene corresponding to the XAGE-4 transcript found in one of the Ewing's sarcoma samples, we searched the HTGS database for XAGE-like sequences. This search did not yield a XAGE-4 gene, but we found a number of very homologous exons and gene sequences on different BAC clones (data not shown). Most of these clones did not code for XAGE-like proteins due to early stop codons. Only 1 gene, which we named XAGE-5 (AJ318894), had a XAGE-like ORF. XAGE-5 was not represented in the human EST database. Since XAGE-5 was located on the same BAC clone (AL445227) as XAGE-3, it can also be mapped on Xp11.21-Xp11.3.

BLAST searches with mRNA sequences of all known GAGEs, XAGEs and PAGEs rendered 1 additional member. Using the PAGE-2a consensus sequence (AA909141), we found a PAGE transcript (PAGE-5, deposited as AJ344352) represented by ESTs belonging to Unigene cluster Hs.245431.


To deduce consensus sequences for the ORFs of XAGE-like genes, we aligned the putative protein sequences of the 5 XAGE genes described herein with 2 members of the GAGE family (GAGE-1 and GAGE-2), the 4 known members of the PAGE family and PAGE-5 deduced from Unigene cluster Hs.245431 (Fig. 4). To represent XAGE-1, we selected XAGE-1a. Since most GAGE-like genes have a second large ORF, to which the primary ORF of XAGE-1 is partly homologous, we also aligned the secondary ORFs next to the primary ORF (Fig. 4). Previous work indicated that the XAGE-1a and -1b proteins were homologous only to the GAGEs at the C terminus.4 This C-terminal homology is visualized in the alignment of Figure 4a (XAGE-1a, amino acids 119–146). The consensus deduced from the N-terminal and central parts of the GAGE-like proteins is present in the alternative frame of XAGE-1 (XAGE-1a#, amino acids 1–91), indicating that the main ORF of XAGE-1 must have acquired a frameshift mutation compared to the other GAGE-like genes. This frameshift is indeed visible as an insert of 1 nucleotide in exon 3 in XAGE-1 compared to XAGE-2 and -3 (Fig. 1, position 444 in XAGE-1a). As a consequence, part of the putative primary XAGE-1 protein (XAGE-1a, amino acids 66–118) resembles the proteins encoded by the secondary ORF of the other GAGE-like genes (Fig. 4b).

thumbnail image

Figure 4. Alignment of GAGE-like proteins. (a) Putative proteins encoded by the primary ORF of the genes. (b) Alternative ORFs of GAGE-like proteins. Consensus residues (≥8 of 12 identical amino acids) are in bold. Alternative ORFs are marked with #. Stop codons are marked with an asterisk. The partial XAGE-4 protein is flanked by < and >. The XAGE-1a deduced protein is split over both alignments in (a,b).

Download figure to PowerPoint

The alignment of the ORFs of the GAGE-like genes clearly reveals the consensus sequences (Fig. 4, bold) present in all XAGE, GAGE and PAGE families. The primary ORF proteins (aligned in Fig. 4a) have low pI values, ranging from 4.2 to 5.1. The consensus residues (52%, 57 of an average of 109) mainly consist of prolines and charged amino acids. The consensus of the alternative reading frames (Fig. 4b) has a far higher pI value of about 11.5. The putative XAGE-1a has an N-terminal extension of 65 amino acids, which is not present in the other GAGE-like genes, and has no homology to other known proteins in the database. Compared to the other XAGEs, XAGE-4 has an internal deletion of 14 amino acids corresponding to the last part of exon 3 of XAGE-2 and -3. This segment might be lost in the XAGE-4 gene or spliced out in its transcript. The PAGE family members clearly vary more in length and sequence from the consensus than the GAGEs and XAGEs. They also have a much shorter secondary ORF, and most of them do not contain an AUG start codon.


  1. Top of page
  2. Abstract
  6. Acknowledgements

In this report, we characterize the XAGE family of CTAs. Next to the identification of several new transcripts and genes, an exploration of their expression pattern in normal tissues, Ewing's sarcoma and melanocytic tumors is given. Alignment of the XAGEs with other homologous CTAs clearly outlines a GAGE superfamily, consisting of at least 9 GAGEs, 5 PAGEs and 5 XAGEs.

Database mining is an increasingly powerful and reliable tool to explore the expression profiles of genes.15, 16 Although libraries of several tissues and tumor types are underrepresented, insight into the distribution of the transcript can be obtained by the origin of ESTs present in its Unigene cluster. A clear example is the placental specificity of the XAGE-3 transcript (based on the ESTs in Unigene cluster Hs.43879), which was confirmed by our PCR results. Furthermore, we were able to prove the existence of splice variants XAGE-1c and -1d, encountered in the database, by RT-PCR and sequencing. For other CTAs, splice variants have also been described. The GAGE-1 transcript has an additional exon, which resembles the intron of the other GAGEs.9 Intron-including transcripts have also been found for the NY-ESO-1 family member LAGE-1, leading to a putative protein with an altered C-terminus.17 For PAGE-2, 2 variants have their own specific cluster (PAGE-2a, Hs.293317; PAGE-2b, Hs.245431). PAGE-2b has a 17–amino acid deletion compared to PAGE-2a.10 Since little is known about the function of CTAs, the role of these variant-encoded proteins is also unclear.

Like most CTAs, XAGE-1, -2, -3 and -5 are localized on chromosome X (Xp11.21-Xp11.3). The coexistence of several XAGE genes on the same BAC clones makes it very likely that the unidentified XAGE-4 gene also resides on that chromosomal region. In the same region, the GAGE,9SSX18 and MAGE-D, -H, -I and -J1 families have been mapped, marking this region as CTA-rich. Regarding the PAGEs, only for PAGE-1 was a subchromosomal localization (Xp21.3-Xp11.3) reported in the Unigene mapping information, locating it near the other GAGE-like genes. The genetic organization of XAGE-2 and -3 is much more similar to the GAGEs than that of XAGE-1. Like XAGE-2 and –3, the GAGEs consist of 5 exons and the primary ORF starts at the beginning of the second exon.9

From the PCR results, it is evident that XAGE-1 and -2 transcripts share a cancer/testis-characteristic expression pattern. Expression in female reproductive tissues is common among CTAs.2, 7, 19–21 In tumor tissues, XAGE-1 transcripts are more abundantly expressed than XAGE-2 transcripts. In melanocytic tumors, the highest percentages of expression are found for XAGE-1b, expressed in 43% (10 of 23) of melanoma metastases after 30 cycles of PCR. This level of expression compares well with the percentages of 38% (23 of 61 after 30 cycles)4 and 43% (20 of 47 after 35 cycles)5 reported before. XAGE-3 might not be a real CTA since we could detect it only in normal placenta and not in normal testis or any tumor lesion.

More frequent expression of XAGEs in melanoma metastases compared to primary melanoma might be induced by loss of regulation of gene expression. This deregulation may be a result of phenomena like altered methylation status or cumulative genetic damage in later stages of tumor progression. With regard to an eventual use of XAGEs in immunotherapy, immunohistochemistry must clarify their potential. This will be of special interest for Ewing's sarcoma since XAGEs are the first CTAs described in this tumor type. Interestingly, several CTL-recognized epitopes are derived from nonprimary ORFs of melanoma differentiation antigen TRP-122 and from CTA NY-ESO-1, called CAMEL.23 Later studies have shown that efficient generation of CTL epitopes, capable of inducing antitumor responses, can occur simultaneously from both ORFs of NY-ESO-1.24 A possible explanation for the production of both proteins from 2 overlapping ORFs on 1 mRNA could be the proposed leaky-scanning mechanism regularly passing the first AUG.22 CTL antitumor epitopes can even be derived from alternative ORFs of widely expressed genes like N-acetylglucosaminyltransferase V25 and intestinal carboxyl esterase.26 The latter epitope was processed from a non-AUG-defined reading frame.

The protein alignment in Figure 4 clearly outlines the consensus present in the putative proteins of the GAGE-like genes. The short AUG start codon missing secondary ORFs of the PAGEs makes it unlikely that the corresponding proteins will be expressed. However, alternative translation starts from CUG codons (coding for leucine) have been described before in humans.27, 28 Most of the non-AUG secondary ORFs shown in Figures 3 and 4 indeed have a CUG codon at the beginning of their ORF. The bipartite nuclear localization signal, described in the XAGE-1 transcripts, is also present in the putative secondary ORF-defined proteins of most other GAGE-like genes. The striking difference in pI values of the GAGE-like proteins derived from both ORFs indicates that, if produced, these proteins will most probably have different biologic functions.

In summary, we studied the XAGE family of CTAs. The high expression of several members of the XAGE family in melanoma and Ewing's sarcoma could be of interest in immunotherapeutic strategies. Homology searching identified several new members of the XAGE and PAGE families of GAGE-like CTAs and their genetic structure. Alignment of the XAGE, PAGE and GAGE putative proteins clearly outlines a GAGE-like superfamily of CTAs. Future studies are needed to address their biologic function in testis and cancers and their putative use in cancer immunotherapy.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Note Added in Proof

Shortly after acceptance of our manuscript, a PubMed search for “XAGE” revealed a report on L552S (Wang et al., Oncogene 2001;20:7699–709). L552S, an alternative splice variant of the XAGE-1 gene, codes for a 160–amino acid nuclear protein and was discovered in lung adenocarcinoma by differential gene expression analysis. L552S, like XAGE-1, has a CTA expression pattern. The mRNA and putative protein are identical to the XAGE-1c splice variant described herein.

The authors also described a low-expression L552S variant in lung cancer (clone 19106), which contains both an insert corresponding to our exon 1′ as well as a 16 bp insert corresponding to our exon 2′. We suggest naming this variant XAGE-1e. Although our RT-PCR set-up for XAGE-1d does not distinguish it from the suggested XAGE-1e, database analysis proves the existence of XAGE-1d: ESTs AW386180 and AW386185 do not contain alternative exon 1′, while in both ESTs exon 2′ is present.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Chomez P, de Backer O, Bertrand M, de Plaen E, Boon T, Lucas S. An overview of the MAGE gene family with the identification of all human members of the family. Cancer Res 2001;61: 554451.
  • 2
    Chen YT, Gure AO, Tsang S, Stockert E, Jager E, Knuth A, Old LJ. Identification of multiple cancer/testis antigens by allogeneic antibody screening of a melanoma cell line library. Proc Natl Acad Sci USA 1998;95: 691923.
  • 3
    Old LJ. Cancer/testis (CT) antigens—a new link between gametogenesis and cancer. Cancer Immunity 2001;1: 17.
  • 4
    Zendman AJ, van Kraats AA, den Hollander AI, Weidle UH, Ruiter DJ, van Muijen GN. Characterization of XAGE-1b, a short major transcript of cancer/testis-associated gene XAGE-1, induced in melanoma metastases. Int J Cancer 2002;97: 195204.
  • 5
    Zendman AJ, de Wit NJ, van Kraats AA, Weidle UH, Ruiter DJ, van Muijen GN. Expression profile of melanoma differentiation antigens and cancer/testis antigens in metastatic lesions of human cutaneous melanoma. Melanoma Res 2001;11: 4519.
  • 6
    Van den Eynde B, Peeters O, de Backer O, Gaugler B, Lucas S, Boon T. A new family of genes coding for an antigen recognized by autologous cytolytic T lymphocytes on a human melanoma. J Exp Med 1995;182: 68998.
  • 7
    Chen ME, Lin SH, Chung LW, Sikes RA. Isolation and characterization of PAGE-1 and GAGE-7. New genes expressed in the LNCaP prostate cancer progression model that share homology with melanoma-associated antigens. J Biol Chem 1998;273: 1761825.
  • 8
    Brinkmann U, Vasmatzis G, Lee B, Yerushalmi N, Essand M, Pastan I. PAGE-1, an X chromosome-linked GAGE-like gene that is expressed in normal and neoplastic prostate, testis, and uterus. Proc Natl Acad Sci USA 1998;95: 1075762.
  • 9
    De Backer O, Arden KC, Boretti M, Vantomme V, de Smet C, Czekay S, Viars CS, de Plaen E, Brasseur F, Chomez P, van den Eynde B, et al. Characterization of the GAGE genes that are expressed in various human cancers and in normal testis. Cancer Res 1999;59: 315765.
  • 10
    Brinkmann U, Vasmatzis G, Lee B, Pastan I. Novel genes in the PAGE and GAGE family of tumor antigens found by homology walking in the dbEST database. Cancer Res 1999;59: 14458.
  • 11
    Liu XF, Helman LJ, Yeung C, Bera TK, Lee B, Pastan I. XAGE-1, a new gene that is frequently expressed in Ewing's sarcoma. Cancer Res 2000;60: 47525.
  • 12
    Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 1997;25: 3389402.
  • 13
    Zendman AJ, Cornelissen IM, Weidle UH, Ruiter DJ, van Muijen GN. TM7XN1, a novel human EGF-TM7-like cDNA, detected with mRNA differential display using human melanoma cell lines with different metastatic potential. FEBS Lett 1999;46: 2928.
  • 14
    Breatnach R, Chambon P. Organisation and expression of eucaryotic split genes coding for proteins. Annu Rev Biochem 1981;50: 34983.
  • 15
    Krizman DB, Wagner L, Lash A, Strausberg RL, Emmert-Buck MR. The Cancer Genome Anatomy Project: EST sequencing and genetics of cancer progression. Neoplasia 1999;1: 1016.
  • 16
    Strausberg RL. The Cancer Genome Anatomy Project: new resources for reading the molecular signatures of cancer. J Pathol 2001;195: 3140.
  • 17
    Lethé B, Lucas S, Michaux L, de Smet C, Godelaine D, Serrano A, de Plaen E, Boon T. LAGE-1, a new gene with tumor specificity. Int J Cancer 1998;76: 9038.
  • 18
    Clark J, Rocques PJ, Crew AJ, Gill S, Shipley J, Chan AM, Gusterson BA, Cooper CS. Identification of novel genes, SYT and SSX, involved in the t(X;18)(p11.2;q11.2) translocation found in human synovial sarcoma. Nat Genet 1994;7: 5028.
  • 19
    Chen YT, Scanlan MJ, Sahin U, Tureci O, Gure AO, Tsang S, Williamson B, Stockert E, Pfreundschuh M, Old LJ. A testicular antigen aberrantly expressed in human cancers detected by autologous antibody screening. Proc Natl Acad Sci USA 1997;94: 19148.
  • 20
    Scanlan MJ, Altorki NK, Gure AO, Williamson B, Jungbluth A, Chen YT, Old LJ. Expression of cancer-testis antigens in lung cancer: definition of bromodomain testis-specific gene (BRDT) as a new CT gene, CT9. Cancer Lett 2000;150: 15564.
  • 21
    De Smet C, Courtois SJ, Faraoni I, Lurquin C, Szikora JP, de Backer O, Boon T. Involvement of two Ets binding sites in the transcriptional activation of the MAGE1 gene. Immunogenetics 1995;42: 28290.
  • 22
    Wang RF, Parkhurst MR, Kawakami Y, Robbins PF, Rosenberg SA. Utilization of an alternative open reading frame of a normal gene in generating a novel human cancer antigen. J Exp Med 1996;183: 1131– 40.
  • 23
    Aarnoudse CA, van den Doel PB, Heemskerk B, Schrier PI. Interleukin-2–induced, melanoma-specific T cells recognize CAMEL, an unexpected translation product of LAGE-1. Int J Cancer 1999;82: 4428.
  • 24
    Rimoldi D, Rubio GV, Dutoit V, Lienard D, Salvi S, Guillaume P, Speiser D, Stockert E, Spagnoli G, Servis C, Cerrotini J-C, Lejeune F, et al. Efficient simultaneous presentation of NY-ESO-1/LAGE-1 primary and nonprimary open reading frame-derived CTL epitopes in melanoma. J Immunol 2000;165: 725361.
  • 25
    Guilloux Y, Lucas S, Brichard VG, van Pel A, Viret C, de Plaen E, Brasseur F, Lethé B, Jotereau F, Boon T. 1996. A peptide recognized by human cytolytic T lymphocytes on HLA-A2 melanomas is encoded by an intron sequence of the N-acetylglucosaminyltransferase V gene. J Exp Med 1996;183: 117383.
  • 26
    Ronsin C, Chung-Scott V, Poullion I, Aknouche N, Gaudin C, Triebel F. A non-AUG-defined alternative open reading frame of the intestinal carboxyl esterase mRNA generates an epitope recognized by renal cell carcinoma-reactive tumor-infiltrating lymphocytes in situ. J Immunol 1999;163: 48390.
  • 27
    Arnaud E, Touriol C, Boutonnet C, Gensac M, Vagner S, Prats H, Prats A-C. A new 34-kilodalton isoform of human fibroblast growth factor 2 is cap dependently synthesized by using a non-AUG start codon and behaves as a survival factor. Mol Cell Biol 1999;19: 50514.
  • 28
    Meiron M, Anunu R, Scheinmann EJ, Hashmueli S, Levi BZ. 2001. New isoforms of VEGF are translated from alternative initiation CUG codons located in its 5′UTR. Biochem Biophys Res Commun 2001;282: 105360.