Clinical Cancer Research CTRC-AACR San Antonio Breast Cancer Symposium Infection and Cancer: Biology, Therapeutics, and Prevention
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sugimura, J.
Right arrow Articles by Teh, B. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sugimura, J.
Right arrow Articles by Teh, B. T.
Clinical Cancer Research Vol. 10, 2368-2378, April 2004
© 2004 American Association for Cancer Research

Molecular Oncology, Markers, Clinical Correlates

Gene Expression Profiling of Early- and Late-Relapse Nonseminomatous Germ Cell Tumor and Primitive Neuroectodermal Tumor of the Testis

Jun Sugimura1,2, Richard S. Foster3, Oscar W. Cummings4, Eric J. Kort1, Masayuki Takahashi1,6, Todd T. Lavery1, Kyle A. Furge1, Lawrence H. Einhorn5 and Bin Tean Teh1

1 Laboratory of Cancer Genetics, Van Andel Research Institute, Grand Rapids, Michigan; 2 Department of Urology, School of Medicine, Iwate Medical University, Morioka, Japan; Departments of 3 Urology, 4 Pathology, and 5 Medicine, School of Medicine, Indiana University, Indianapolis, Indiana; and 6 Department of Urology, School of Medicine, Tokushima University, Tokushima, Japan


   ABSTRACT
Top
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Purpose: To better understand the molecular mechanisms that underlay the development and progression of nonseminomatous germ cell tumor of testis (NSGCTT) as well as malignant transformation of teratoma and primitive neuroectodermal tumor (PNET).

Experimental Design: We studied the gene expression profiles of 17 retroperitoneal NSGCTTs (10 yolk sac tumors, 3 embryonal carcinomas, 4 teratomas) and 2 PNETs obtained from patients with two clinical outcomes. Tissue samples were obtained from the Indiana University. One group of NSGCTT and PNET patients developed metastases within 2 years (early-relapse) of initial successful treatment, and the other group developed metastases after 2 years (late-relapse). Gene expression in these groups of patients was quantified using cDNA microarrays and real-time relative quantitative PCR.

Results: We demonstrate that the gene expression profiles of these tumors correlate with histological type. In addition, we identify type-specific genes that may serve as novel diagnostic markers. We also identify a gene set that can distinguish between early-relapse and late-relapse yolk sac tumors. The expression differences of these genes may underlie the differences in clinical outcome and drug response of these tumors.

Conclusion: This is the first study that used gene expression profiling to examine the molecular characteristics of the NSGCTTs and drug response in early- and late-relapse tumors. These results suggest that two molecularly distinct forms of NSGCTTs exist and that the integration of expression profile data with clinical parameters could enhance the diagnosis and prognosis of NSGCTTs. More importantly, the identified genes provide insight into the molecular mechanisms of aggressive NSGCTTs and suggest intervention strategies.


   INTRODUCTION
Top
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Testicular cancer is a relatively rare malignancy; however, an estimated 7,400 new cases were reported in the United States in 1999, and it is the most common cancer of men between the ages of 15 and 34 (1 , 2) . Based on morphological features, germ cell tumors are divided into seminoma, yolk sac tumor, embryonal carcinoma, choriocarcinoma, and teratoma according to the WHO International Histological Classification of Testicular Tumors (3) . Non-germ-cell malignancies such as primitive neuroectodermal tumors (PNETs) arise in testicular germ cell teratoma and are associated with an increased number of relapses and a decreased relapse-free and overall survival (4) . Teratoma is pluripotential tissue that can differentiate along ectodermal, endodermal, or mesodermal elements. PNET represents malignant transformation of teratoma along mesodermal lines.

Testicular cancer has a 95% cure rate, and most relapses occur within the first 2 years of therapy. Only 2–3% of patients treated for testicular cancer have late-relapse (defined as recurrence >=2 years after initial successful therapy). Metastatic germ cell tumors have remarkably high cure rates. However, those with late relapse are found to be refractory to chemotherapy and carry a poor prognosis (5) . Cytogenetic studies suggest most testicular tumors contain extra copies of the short arm of chromosome 12, and this genomic alteration may be responsible for the development of testicular neoplasms (6) . Unfortunately, many of these additional differences remain unknown. However, differences in chemosensitivity and survival are most likely related to additional differences in underlying genetic and biological characteristics.

Microarray technology has provided new insights into the underlying molecular mechanisms of many types of cancers. The gene expression profiles from microarrays can identify the molecular signatures of cancers and can be used to distinguish histological types and to discover novel types. Classification of tumors by gene expression may reflect heterogeneity in transformation mechanisms, cell types, or aggressiveness among tumors. Furthermore, several studies have identified prognostic sets of genes that may underlie the heterogeneity in tumor aggressiveness (7, 8, 9, 10) . The identification of such genes may lead to the discovery of new potential targets for cancer diagnosis and therapy. Recently, differentially expressed genes on chromosome 17 in 18 testicular samples (15 germ cell tumors; 3 normal testes) were identified using a microarray containing 636 cDNA. Growth factor receptor-bound protein 7 and junction plakoglobin at the 17q11–q21, lethal giant larvae homologue 2, and phosphodiesterase 6G at the 17q24–qter were consistently the most up-regulated in the testicular tumors (11 , 12) . However, no study has identified a prognostic set of genes.

In this study, we characterized the molecular signature of 17 metastatic regions of nonseminomatous germ cell tumors of testes (NSGCTTs) and two PNETs by using 19,968 cDNA microarrays to elucidate their underlying molecular mechanisms. Furthermore, we sought to identify a gene set that can distinguish between early-relapse and late-relapse tumors and correlate with differences in survival and chemosensitivity.


   MATERIALS AND METHODS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Patient Information and Tumor Samples.
Tissue samples were obtained from 19 patients with metastatic lesions (NSGCTTs and PNETs) of the testis. The patients had undergone retroperitoneal lymph node dissection after chemotherapy at Indiana University. Nine patients had early relapse (within 2 years of initial management), whereas 10 had late relapse (after 2 years of initial management). Informed consent was obtained from patients to use their removed tissue specimens and clinicopathological data for research purposes. Approval was obtained by the Indiana University Internal Review Boards. The samples were rendered anonymous before the study. Part of each tumor sample was frozen immediately after the operation and stored at –80°C. Total RNA was isolated from the frozen tissue by using TRIzol reagent (Invitrogen, Carlsbad, CA). For all samples, total RNA was purified with 2.5 M final concentration of LiCl. The remaining parts of tumors were fixed with 10% buffered formalin, and the paraffin sections were stained with H&E. The WHO International Histological Classification of Tumors was used for histological evaluation of the specimens (3) . Union Internationale Contre le Cancer Tumor-Node-Metastasis classification and stage groupings were also used (13) .

cDNA Microarray Fabrication and Procedures.
Microarray production was performed as described previously (14 , 15) with slight modification. Briefly, 19,968 cDNA clones were PCR-amplified directly from bacterial stocks purchased from Research Genetics (Huntsville, AL). Following ethanol precipitation and transfer to 384-well plates, clones were printed onto aminosilane-coated glass slides using a custom-built robotic microarrayer. Normal human testicle total RNA (Ambion, TX) purified with 2.5 M final concentration of LiCl was used as a reference. For all samples, 50 µg of total RNA from tumors and reference were reverse-transcribed with oligodeoxythymidylic acid primer and Superscript II (Invitrogen) in the presence of Cy5-dCTP and Cy3-dCTP (Amersham Pharmacia Biotech, Peapack, NJ). The Cy5- and Cy3-labeled cDNA probes were mixed with probe hybridization solution containing formamide and hybridized to pre-warmed (50°C) slides for 20 h at 50°C. Following hybridization, slides were washed, dried by snap centrifugation, and scanned immediately using Scan Array Lite operating at 532 and 635 nm (GSI Lumonics, Billerica, CA).

Data Analysis.
Images were analyzed by using the software GENEPIX PRO 3.0 (Axon Instruments, Foster City, CA). Spots showing no signal or obvious defects were excluded from the analysis. The local background was subtracted from the remaining spots, and the ratios of net fluorescence from the Cy5-specific channel to the net fluorescence from the Cy3-specific channel were calculated for each spot, representing tumor RNA expression relative to the normal testicular total RNA.

Microarray Analysis.
Two sources of systematic bias were reduced in the data by normalization. Loess regression was performed on the log-transformed data to eliminate the intensity-dependent variation in the ratio of the spot intensities that has been observed in gene expression data (16 , 17) . This normalization was performed in a pin-dependent fashion to further reduce systematic variation related to the physical properties of each pin of the slide-spotting robot. Following loess normalization, the data were median-centered and rescaled using the median absolute deviation as described elsewhere (18 , 19) . These normalizations were carried out using the Bioconductor package7 for the R statistical analysis framework (20) . Unsupervised clustering was performed on the normalized data using CLUSTER and visualized using TREEVIEW.8 The significance threshold was set at P < 0.05, and only those genes for which data were available in 75% of all 19 cases (i.e., at least 15 in all) were included in the analysis (4,569 genes). Because the cases initially clustered according to histological type, indicating unique genetic signatures among the types, it was decided that the histological types should not be combined for the purpose of determining possible genetic predictors of outcome (early-relapse versus late-relapse). However, with the exception of yolk sac tumors, there were not enough cases within each histological type to perform this analysis. Therefore, the analysis of genes related to outcome was restricted to the yolk sac tumors.

Differentially expressed genes were identified using the Cluster Identification Tool application (21) , which compares the mean expression in each group and uses random permutation to estimate P. Because of the possibility of false discovery, we selected the two differentially expressed genes in late-relapse NSGCTTs and four differentially expressed genes in late-relapse yolk sac tumors identified by this method and confirmed their expression level using real-time PCR.

Comparative Genomic Microarray Analysis Algorithm.
To identify regional gene expression biases, gene expression values that map to a given chromosomal arm are collected, and a binomial test is used to determine if a significant upward or downward bias is present. First, sequence comparisons are used to map microarray probe sequences (the sequences that are placed on the microarray) to predicted Ensembl transcripts (Ensembl version 10.2; Ref. 22 ). Included in the Ensembl transcript annotations are chromosomal mapping locations at base-pair resolution. Expression values from multiple probes that map to the same gene are condensed to a single value by averaging. To apply the binomial test to expression data, of n non-zero expression values that map to a given chromosomal arm, r gene expression values are scored as "up" if the log2(R/G) value is positive and (n – r) "down" if the log-transformed ratio is negative. This binomial test assumes that in a cytogenetically normal genomic region, the probability of a gene expression value being positive (p) is equal to the probability of the expression value being negative (q) such that p = q = 0.5. For the binomial test, the probability of obtaining r "up" observations with a probability p and n – r "down" observations with a probability q is

In the cases where n > 25 and n*p*q > 10, the binomial probability density function is approximated using the normal probability density function, and a z-statistic for the chromosomal arm is computed such that z = (2r – n)/sqrt(n). Therefore, a large positive z-statistic indicates a significant positive expression bias (i.e., genomic gain), and a large negative z-statistic indicates the presence of a negative expression bias (i.e., genomic loss). A set of chromosomal arm z-statistics can be plotted as a heat map to identify and summarize predicted cytogenetic features (23) .

Real-Time Relative Quantitative PCR (RT-PCR).
RT-PCR was performed in triplicate using the ABI PRISM 7700 Sequence Detection System according to the manufacturer’s instructions, and these data were averaged. The two primers and the TaqMan probe were specifically designed for the following six genes: glutathione S-transferase {theta}1 (GSTT1); fatty acid synthase (FASN); phospholipase A2 group IIA (PA2IIA); trinucleotide repeat containing 3 (TNRC3); glypican 3 (GPC3); and glutaredoxin (GLRX) using Primer Express v1.5a (Applied Biosystems, Foster City, CA).

One hundred ng of each cDNA were amplified using PCR Master Mix according to the following PCR conditions: 50°C for 2 min; 95°C for 10 min; followed by 40 cycles of 95°C for 15 s; and 60°C for 1 min. Because 18S rRNA resulted in the least variation throughout the samples among a total of 11 housekeeping genes using Taqman Human Endogenous Control Plate (Applied Biosystems), this gene was used as the endogenous control. Each threshold cycle (CT), which indicates the cycle at which an increase in reporter fluorescence just goes over the optimal value line, was determined. The CT value of 18S rRNA was subtracted from each CT value of tumor or normal for normalization, and the ratio of tumor to normal testicular RNA expression was calculated to compare with microarray data.


   RESULTS
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Gene Expression Profiling.
We analyzed the gene expression profiling data in two ways. First, we compared the gene expression of each tumor sample with normal testicular RNA to identify gene expression alterations that occur in NSGCTT and PNET. In addition, we identified particular subsets of genes that most strongly distinguished between late- and early-relapse patients.

Differentially expressed cDNAs in each type of tumors are listed in Tables 1Citation 2Citation 3Citation 4Citation . They are significantly up-regulated or down-regulated in each type of tumors compared to other types of tumors. A patient dendrogram based on a set of differentially expressed genes in each of the tumors compared with all other types of NSGCTTs (or PNETs) studied was shown in Fig. 1ACitation . All differentially expressed cDNA in late-relapse tumors compared to early-relapse tumors and all differentially expressed cDNA in late-relapse yolk sac tumors compared to early-relapse yolk sac tumors are listed in Tables 5Citation and 6Citation . A patient dendrogram based on the whole differentially expressed 13 cDNA set that might distinguish the late-relapse group from the early-relapse group is shown in Fig. 1BCitation .


View this table:
[in this window]
[in a new window]
  		Table 1 Differentially expressed genes in yolk sac tumor

 

View this table:
[in this window]
[in a new window]
  		Table 2 Differentially expressed genes in embryonal carcinoma

 

View this table:
[in this window]
[in a new window]
  		Table 3 Differentially expressed genes in teratoma

 

View this table:
[in this window]
[in a new window]
  		Table 4 Differentially expressed genes in primitive neuroectodermal tumor

 


View larger version (50K):
[in this window]
[in a new window]
  		Fig. 1. Panel A, hierarchical clustering of 19 testicular tumors. Hierarchical clustering of the gene expression profiles of testicular tumors. Rows represent individual cDNAs, and columns represent individual tumor samples. The color of each square represents the median-polished, normalized, log-transformed gene expression ratio (tumor versus reference). Expression levels greater than the median are shaded in red, those below the median, green; equal to the median, black; and inadequate or missing data, gray. The color saturation indicates the degree of divergence from the median. Clustering is based on a combined set of genes that are unique to each subtype (total 543 cDNA). Basically, there are two main groups; one group consists of embryonal carcinomas and teratomas, and the other consists of yolk sac tumors and primitive neuroectodermal tumor. The highly-expressed genes from each tumor type are shown using sidebars with different colors (A, yolk sac tumor; B, teratoma; C, embryonal carcinoma; D, primitive neuroectodermal tumor). Panel B, clustering of 10 yolk sac tumors. Clustering based on 13 differentially expressed cDNAs that discriminate between late-relapse and early-relapse samples is shown in Fig. 1BCitation . A, genes mostly over-expressed in tumors with the late relapse yolk sac tumors. B, genes mostly under-expressed in tumors with the late relapse yolk sac tumors.

 

View this table:
[in this window]
[in a new window]
  		Table 5 Differentially expressed genes in late relapse nonseminomatous germ cell tumor of testes (NSGCTT)s

 

View this table:
[in this window]
[in a new window]
  		Table 6 Differentially expressed genes in late relapse yolk sac tumors

 
Real-Time Relative Quantitative PCR.
Six genes were selected to confirm the microarray results with RT-PCR. To compare the gene expression ratios obtained by microarray to the RT-PCR data, the expression ratio of tumor to noncancerous testis was calculated by RT-PCR (Fig. 2)Citation . RT-PCR showed a significantly higher expression ratio of GSTT1 (P = 0.016) in late-relapse yolk sac tumors (Fig. 2A)Citation but not FASN (P = 0.222, Fig. 2BCitation ). PA2IIA and TNRC3 were significantly under expressed in late-relapse yolk sac tumors (P = 0.032 and 0.016, respectively; Fig. 2, C and DCitation ), and PA2IIA was more highly expressed in early-relapse yolk sac tumors. GPC3 and GLRX did not have significant changes (P = 0.310 and 0.548, respectively; Fig. 2, E and FCitation ).



View larger version (32K):
[in this window]
[in a new window]
  		Fig. 2. Real-time relative quantitative PCR analysis of GSTT1 (A), FASN (B), phospholipase A2 group IIA (C), TNRC3 (D), glypican 3 (E), and glutaredoxin (F). The left side bar in gray of each sample indicates the expression ratio of tumor to normal testis by real-time relative quantitative PCR and the right side bar in white indicates the expression ratio by microarray experiments. Data grouped based on clinical parameters are given as the mean and the SE. They are evaluated statistically using the Mann-Whitney U test. All real-time relative quantitative PCR data were consistent with microarray data. GSTT1 was significantly highly expressed in late-relapse yolk sac tumors (P = 0.016; A), but FASN was not significantly different between late- and early-relapse yolk sac tumors (P = 0.222; B). PA2IIA and TNRC3 were significantly under-expressed in late-relapse yolk sac tumors (P = 0.032 and 0.016, respectively; C and D). However, glypican 3 and glutaredoxin also were not significantly different between the two groups. P = 0.310, E; P = 0.548, F.

 
Chromosomal Changes.
We also used the gene expression profiles to infer chromosomal abnormalities in these samples. Chromosomal changes identified in our study are shown in Fig. 3Citation . Frequent chromosomal changes of 6 and 12, as reported previously in NSGCTTs were confirmed. In addition, a novel chromosome change of 14q was frequently indicated by comparative genomic microarray analysis among late-relapse tumor in this study.



View larger version (19K):
[in this window]
[in a new window]
  		Fig. 3. Chromosomal changes in nonseminomatous germ cell tumor of testis (or primitive neuroectodermal tumors) are shown. Red corresponds to a gain of chromosome, and green corresponds to a loss of chromosome. Frequent changes of chromosome 6 and 12 as reported previously in nonseminomatous germ cell tumor of testes were confirmed.

 

   DISCUSSION
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
Differentially Expressed Genes in Each Type of NSGCTT.
Identifying differentially expressed genes based on histological diagnosis may lead to the discovery of new markers for differential diagnosis. Because global clustering analysis using 4,569 cDNAs showed that each type of NSGCTT had distinct molecular signatures, we sought to identify the differentially expressed genes contributing to these distinctions. However, the combination of teratoma and yolk sac tumor is quite common as is the combination of embryonal carcinoma and yolk sac tumor. These combinations may affect gene expression, therefore we will require a more detailed pathological review of those specimens that cluster with other tumor types.

Generally, while some immunoglobulin genes and immunological-related genes are differentially expressed in yolk sac tumors (Table 1)Citation , a large number of immunoglobulin genes, including IFN-related genes, are found in embryonal carcinomas (Table 2)Citation . The possibility exists that immunological response to the tumor may be important or that embryonal carcinoma is more immunogenic than yolk sac tumors. Differentially expressed genes in teratoma (Table 3)Citation include proteins such as mucin, elastin, matrix protein, and fibulin, all of which are structural proteins and would be consistent with the type of histological characterization of well-differentiated teratoma. Similarly, the differentially expressed genes in PNET (Table 4)Citation do not display any patterns other than relatedness to neuronal tissue (see below).

If we concentrate on individual genes, gap junction protein ß 1 (GJB1, connexin 32) is the most over-expressed gene in yolk sac tumors (Table 1)Citation . Mutations in the GJB1 gene are responsible for the majority of cases of X-linked Charcot-Marie-Tooth disease. Connexin expression is frequently decreased in neoplasia and may contribute to defective growth control and loss of differentiated functions (24) .

The genes of acyl-CoA dehydrogenase (C-2 to C-3 short chain; ACADS) are the most differentially under-expressed in yolk sac tumors. The acyl-CoA dehydrogenases (ACDs) are a family of mitochondrial flavoenzymes required for fatty acid ß-oxidation and branched-chain amino acid degradation. The defects of ACDs in isoleucine and valine catabolism have been proposed in clinically diverse patients with an abnormal pattern of metabolites in their urine. In ACDs deficiency, the maternal serum and amniotic fluid concentrations of {alpha}-fetoprotein were elevated (25) . This is a very interesting observation because there is a similar elevation of serum {alpha}-fetoprotein in yolk sac tumors.

ACADS and immunoglobulin heavy constant {gamma}3 are the most over-expressed in embryonal carcinomas. Interestingly, ACADS is the most under-expressed gene in yolk sac tumors but is the most up-regulated gene in embryonal carcinomas. Immunoglobulin heavy chain has a relationship with multiple myeloma and plasma cell leukemia (26) . Immunoglobulin genes including immunoglobulin heavy constant {gamma}3 may play important roles in embryonal carcinoma.

Differentially under-expressed genes in the embryonal carcinoma are GPC3 and gap junction protein ß1. GPC3 is a membrane-bound heparin sulfate proteoglycan. The GPC3 gene is located at Xq26, frequently deleted in advanced ovarian cancer cell lines (27) . Expression of GPC3 is also silenced in human breast cancer (28) and decreased in human gastric cancer (29) . In contrast, the expression of GPC3 in Wilms’ tumor, hepatoblastoma, and hepatocellular carcinoma were increased (30 , 31) . In our study, expression of GPC3 and Gap junction protein ß1 in yolk sac tumors was increased. These results show some of the molecular genetic differences between yolk sac tumors and embryonal carcinomas.

Matrix Gla protein (MGP) is the most over-expressed gene in teratomas (Table 3)Citation . MGP is a vitamin-K-dependent protein and is synthesized in a variety of tissues such as lung, heart, kidney, cartilage, and bone. MGP mRNA levels have been found to be elevated in a breast cancer cell lines, 600 polyethylenimine, primary renal-cell carcinomas, prostate carcinomas, and testicular germ-cell tumors (32 , 33) .

PNET is a malignant, small, round cell tumor that exhibits neuroepithelial differentiation. The histological designation of PNETs in the testis (neuroblastoma or medulloepithelioma) did not predict which tumor metastasized. Extratesticular PNETs in patients with testicular germ-cell tumors are usually fatal, but patients with neuroblastomatous metastases may have a more prolonged course (4) .

Neuronal pentraxin II and ubiquitin COOH-terminal esterase L1 (UCHL1) are the most over-expressed genes in the PNETs (Table 4)Citation . The existence of a family of pentraxin proteins that are expressed in the brain, and other tissues may play important roles in the uptake of extracellular material (34) . Protein gene product 9.5 (PGP 9.5), most likely identical to UCHL1, is a major constituent of cytoplasmic polypeptides in neurons. There is immunoreactive PGP 9.5 in many neuroendocrine cells, in spermatogonia, Leydig cells of the testis, ova, and some cells of the corpus luteum (35) . PGP 9.5 is also highly expressed in peripheral PNET cell lines and embryonic tumors (36) . These findings are also consistent with our gene expression profiling results.

Serine protease inhibitor Kunitz type 2 (SPINT2) and RNase inhibitor-related are the most differentially under-expressed genes in the PNETs. SPINT2 is down-regulated in normal human fibroblast MRC-5 cells expressing an activated H-ras oncogene and in the human fibrosarcoma cell line HT1080 (37) . The human RNase inhibitor 2 is expressly limited to testis and may play critical roles in human spermatogenesis (38) .

Differentially Expressed Genes in Each Prognosis: Late-Relapse and Early-Relapse NSGCTTs.
To identify differentially expressed genes in late-relapse NSGCTTs and PNETs compared to early-relapse NSGCTTs and PNETs, we first identified genes that were at least 2-fold up- or down-regulated in at least 75% of the tumors. Next, we identified genes that could distinguish between early- and late-relapse samples. The six up-regulated and five down-regulated genes that met the above criteria are summarized in Table 5Citation . Both previously known and unknown genes were found to be significantly up- or down-regulated in 19 tumors, but the change values, both up and down, were small. The most over-expressed gene in the late-relapse group is small nuclear ribonucleoprotein 70kD polypeptide. The U1–70kD small nuclear ribonucleoprotein auto antigen is a major target of B-cell responses in patients with connective tissue diseases (39) .

Glutaredoxin (GRX) and aldehyde dehydrogenase (ALDH)-1 family member A2 are the most differentially under-expressed genes in the late-relapse group. GRX, also known as thioltransferase, is a small dithiol protein that has been shown to be involved in regulating various cellular functions (40) . ALDHs are a group of enzymes that catalyze the conversion of aldehydes to the corresponding carboxylic acids. The expression of ALDH-1 is higher in radioresistant cervical squamous cell carcinomas cells (41) . Generally, genes related to the cell function may play some role between the two categories.

Differentially expressed genes in late-relapse yolk sac tumors are shown in Table 6Citation . The seven up-regulated and six down-regulated genes that met the criteria are summarized in Table 6Citation . The most over-expressed gene in late-relapse yolk sac tumors is GSTT1. The glutathione S-transferases (GSTs) have been implicated as susceptibility genes for a number of cancers. The homozygous deletions of GSTT1 [GSTT1(–)] group had a worse prognosis than either the homozygous or heterozygous for GSTT1 [GSTT1(+)] group in acute myeloid leukemia (42) . On the other hand, the GSTT1(–) genotype conferred the reduction in risk of relapse compared to the GSTT1(+) in children with acute lymphoblastic leukemia (43 , 44) . Furthermore, GSTT1(–) was associated with decreased time to the next primary tumor presentation in the basal cell carcinoma (45) . Our results suggest that GSTT1 may play an important role for the relapse frequencies of yolk sac tumors.

Phospholipase A2 group IIA (PLA2IIA) is the most differentially under-expressed gene in late-relapse yolk sac tumors. On the other hand, it is the most differentially over-expressed gene in early-relapse yolk sac tumors. This enzyme releases free fatty acids through catalysis of membrane in mammalian cells. The product arachidonic acid is metabolized to produce prostaglandins and leukotrienes that mediate a diverse array of biological activities including inflammation, mitogenesis, and tumor cell invasion. PA2IIA can generate arachidonate from cellular phospholipids. The expression of PA2IIA is elevated in the prostatic cancer, and dysregulation of this enzyme may play a role in prostatic carcinogenesis and progression (46 , 47) . Furthermore, breast cancer patients with high PLA2 expression have significantly shorter disease-free and overall survival than those patients with low PLA2 expression (48) . Therefore, PLA2IIA may play some roles in the relapse of yolk sac tumors.

Real-Time Relative Quantitative PCR.
All real-time PCR data were consistent with microarray data (Fig. 2)Citation ; however, in some cases the expression ratios calculated by real-time PCR were different from those obtained by microarray. Spot saturation and competitive hybridization during the microarray experiments and the amplification methods of PCR may result in these discrepancies.

Initial cisplatin-based combination chemotherapy is curative for many patients with metastatic testicular cancer. Subsequent second-line chemotherapy with high-dose carboplatin plus etoposide will cure approximately half of the relapsing patients. However, patients who experience a late relapse are rarely cured with any type of chemotherapy (49) . Our results suggest that yolk sac tumor, with over-expressions of GSTT1 and under-expressed PLA2IIA, will have a risk of late-relapse and drug resistance. For this group of patients, we need careful follow-up on the patients after 2 years of initial management. In contrast, yolk sac tumors with GSTT1 under-expression, and PLA2IIA over-expression will have relapsed early, but cisplatin-based combination chemotherapy has proven more effective.

In conclusion, we have identified gene expression alterations that are specific for each type of NSGCT and PNET. The identified genes may give insight into tumorigenesis and progression of these tumors. We have also identified gene expression alterations that distinguish between early- and late-relapse yolk sac tumors. This may have important clinical implications because the late relapse patients may require a different approach. In addition, some of these discriminating genes may give potential insights into new drug targets.


   ACKNOWLEDGMENTS
 
We thank the Laboratory of DNA and Protein Microarray Technology at the Van Andel Research Institute. We also thank Dr. David Petillo, Dr. Sok Kean Khoo, and David E. Nadziejka for critically reviewing this manuscript. We are grateful to the Lance Armstrong foundation for supporting this work.


   FOOTNOTES
 
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Bin Tean Teh, Laboratory of Cancer Genetics, Van Andel Research Institute, 333 Bostwick Avenue, North East, Grand Rapids, MI 49503. Phone: (616) 234-5296; Fax: (616) 234-5297; E-mail: bin.teh{at}vai.org

7 http://www.bioconductor.org. Back

8 MB Eisen; http://rana.lbl.gov. Back

Received 10/ 3/03; revised 12/17/03; accepted 12/31/03.


   REFERENCES
Top
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1999. CA Cancer J Clin, 49: 8-31, 1999.[Abstract/Free Full Text]
  2. Silverberg E. Cancer in young adults (ages 15 to 34). CA Cancer J Clin, 32: 32-42, 1982.[Free Full Text]
  3. Mostofi FK, Sobin LH, Sesterhenn IA. . Histological typing of testis tumours (International Histological Classification of Tumours), Springer-Verlag Berlin 1998.
  4. Michael H, Hull MT, Ulbright TM, Foster RS, Miller KD. Primitive neuroectodermal tumors arising in testicular germ cell neoplasms. Am J Surg Pathol, 21: 896-904, 1997.[CrossRef][Medline]
  5. Baniel J, Foster RS, Gonin R, Messemer JE, Donohue JP, Einhorn LH. Late relapse of testicular cancer. J Clin Oncol, 13: 1170-6, 1995.[Abstract]
  6. Atkin NB, Baker MC. Specific chromosome change, i(12p), in testicular tumours?. Lancet, 2: 1349 1982.
  7. Takahashi M, Rhodes DR, Furge KA, et al Gene expression profiling of clear cell renal cell carcinoma: gene identification and prognostic classification. Proc Natl Acad Sci USA, 98: 9754-9, 2001.[Abstract/Free Full Text]
  8. Alizadeh AA, Eisen MB, Davis RE, et al Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature (Lond), 403: 503-11, 2000.[CrossRef][Medline]
  9. Sorlie T, Perou CM, Tibshirani R, et al Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc Natl Acad Sci USA, 98: 10869-74, 2001.[Abstract/Free Full Text]
  10. Bhattacharjee A, Richards WG, Staunton J, et al Classification of human lung carcinomas by mRNA expression profiling reveals distinct adenocarcinoma subclasses. Proc Natl Acad Sci USA, 98: 13790-5, 2001.[Abstract/Free Full Text]
  11. Rodriguez S, Jafer O, Goker H, et al Expression profile of genes from 12p in testicular germ cell tumors of adolescents and adults associated with i(12p) and amplification at 12p11.2-p12.1. Oncogene, 22: 1880-91, 2003.[CrossRef][Medline]
  12. Skotheim RI, Monni O, Mousses S, et al New insights into testicular germ cell tumorigenesis from gene expression profiling. Cancer Res, 62: 2359-64, 2002.[Abstract/Free Full Text]
  13. Sobin LH Wittekind C eds. . International Union Against Cancer. TNM classification of malignant tumours, 5th ed. John Wiley & Sons New York 1997.
  14. Hegde P, Qi R, Abernathy K, et al A concise guide to cDNA microarray analysis. Biotechniques, 29: 548-50, 552–4, 556 passim 2000.[Medline]
  15. Eisen MB, Brown PO. DNA arrays for analysis of gene expression. Methods Enzymol, 303: 179-205, 1999.[Medline]
  16. Cleveland WS. Robust locally weighted regression and smooth-ing scatterplots. J Am Statistic Assoc, 74: 829-836, 1979.[CrossRef]
  17. Cleveland WS, Devlin SJ. Locally-weighted regression: an approach to regression analysis by local fitting. J Am Stat Assoc, 83: 596-610, 1988.[CrossRef]
  18. Yang YH, Dudoit S, Luu P, et al Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res, 30: e15 2002.[Abstract/Free Full Text]
  19. Yang YH, Dudoit S, Luu P, Speed TP. Normalization for cDNA microarray data Bittner ML Chen Y Dorsel AN Dougherty ER eds. . Microarrays: optical technologies and informatics. Vol. 4266 of Proceedings of SPIE, The International Society for Optical Engineering Bellingham, WA 2001.
  20. Ihaka R, Gentleman RR. A language for data analysis and graphics. J Comp Graph Stat, 5: 299-314, 1996.[CrossRef]
  21. Rhodes DR, Miller JC, Haab BB, Furge KA. CIT: identification of differentially expressed clusters of genes from microarray data. Bioinformatics, 18: 205-6, 2002.[Abstract/Free Full Text]
  22. Clamp M, Andrews D, Barker D, et al Ensembl 2002: accommodating comparative genomics. Nucleic Acids Res, 31: 38-42, 2003.[Abstract/Free Full Text]
  23. Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci USA, 95: 14863-8, 1998.[Abstract/Free Full Text]
  24. Schwarz M, Wanke I, Wulbrand U, Moennikes O, Buchmann A. Role of connexin32 and ß-catenin in tumor promotion in mouse liver. Toxicol Pathol, 31: 99-102, 2003.[CrossRef][Medline]
  25. Chisholm CA, Vavelidis F, Lovell MA, et al Prenatal diagnosis of multiple acyl-CoA dehydrogenase deficiency: association with elevated {alpha}-fetoprotein and cystic renal changes. Prenat Diagn, 21: 856-9, 2001.[CrossRef][Medline]
  26. Nishida K, Tamura A, Nakazawa N, et al The Ig heavy chain gene is frequently involved in chromosomal translocations in multiple myeloma and plasma cell leukemia as detected by in situ hybridization. Blood, 90: 526-34, 1997.[Abstract/Free Full Text]
  27. Lin H, Huber R, Schlessinger D, Morin PJ. Frequent silencing of the GPC3 gene in ovarian cancer cell lines. Cancer Res, 59: 807-10, 1999.[Abstract/Free Full Text]
  28. Xiang YY, Ladeda V, Filmus J. Glypican-3 expression is silenced in human breast cancer. Oncogene, 20: 7408-12, 2001.[CrossRef][Medline]
  29. Zhu Z, Friess H, Kleeff J, et al Glypican-3 expression is markedly decreased in human gastric cancer but not in esophageal cancer. Am J Surg, 184: 78-83, 2002.[CrossRef][Medline]
  30. Toretsky JA, Zitomersky NL, Eskenazi AE, et al Glypican-3 expression in Wilms tumor and hepatoblastoma. J Pediatr Hematol Oncol, 23: 496-9, 2001.[CrossRef][Medline]
  31. Zhu ZW, Friess H, Wang L, et al Enhanced glypican-3 expression differentiates the majority of hepatocellular carcinomas from benign hepatic disorders. Gut, 48: 558-64, 2001.[Abstract/Free Full Text]
  32. Chen L, O’Bryan JP, Smith HS, Liu E. Overexpression of matrix Gla protein mRNA in malignant human breast cells: isolation by differential cDNA hybridization. Oncogene, 5: 1391-5, 1990.[Medline]
  33. Levedakou EN, Strohmeyer TG, Effert PJ, Liu ET. Expression of the matrix Gla protein in urogenital malignancies. Int J Cancer, 52: 534-7, 1992.[Medline]
  34. Hsu YC, Perin MS. Human neuronal pentraxin II (NPTX2): conservation, genomic structure, and chromosomal localization. Genomics, 28: 220-7, 1995.[CrossRef][Medline]
  35. Wilson PO, Barber PC, Hamid QA, et al The immunolocalization of protein gene product 9.5 using rabbit polyclonal and mouse monoclonal antibodies. Br J Exp Pathol, 69: 91-104, 1988.[Medline]
  36. Wang Y, Einhorn P, Triche TJ, Seeger RC, Reynolds CP. Expression of protein gene product 9.5 and tyrosine hydroxylase in childhood small round cell tumors. Clin Cancer Res, 6: 551-8, 2000.[Abstract/Free Full Text]
  37. Izumi H, Takahashi C, Oh J, Noda M. Tissue factor pathway inhibitor-2 suppresses the production of active matrix metalloproteinase-2 and is down-regulated in cells harboring activated ras oncogenes. FEBS Lett, 481: 31-6, 2000.[CrossRef][Medline]
  38. Miyamoto T, Hasuike S. Isolation and expression analysis of the human RNH2 gene encoding ribonuclease inhibitor 2. J Assist Reprod Genet, 19: 394-7, 2002.[CrossRef][Medline]
  39. Talken BL, Lee DR, Caldwell CW, Quinn TP, Schafermeyer KR, Hoffman RW. Analysis of T cell receptors specific for U1–70kD small nuclear ribonucleoprotein autoantigen: the {alpha} chain complementarity determining region three is highly conserved among connective tissue disease patients. Hum Immunol, 60: 200-8, 1999.[CrossRef][Medline]
  40. Hirota K, Matsui M, Murata M, et al Nucleoredoxin, glutaredoxin, and thioredoxin differentially regulate NF-{kappa}B, AP-1, and CREB activation in HEK293 cells. Biochem Biophys Res Commun, 274: 177-82, 2000.[CrossRef][Medline]
  41. Kitahara O, Katagiri T, Tsunoda T, Harima Y, Nakamura Y. Classification of sensitivity or resistance of cervical cancers to ionizing radiation according to expression profiles of 62 genes selected by cDNA microarray analysis. Neoplasia, 4: 295-303, 2002.[CrossRef][Medline]
  42. Naoe T, Tagawa Y, Kiyoi H, et al Prognostic significance of the null genotype of glutathione S-transferase-T1 in patients with acute myeloid leukemia: increased early death after chemotherapy. Leukemia (Baltimore), 16: 203-8, 2002.
  43. Anderer G, Schrappe M, Brechlin AM, et al Polymorphisms within glutathione S-transferase genes and initial response to glucocorticoids in childhood acute lymphoblastic leukaemia. Pharmacogenetics, 10: 715-26, 2000.[CrossRef][Medline]
  44. Stanulla M, Schrappe M, Brechlin AM, Zimmermann M, Welte K. Polymorphisms within glutathione S-transferase genes (GSTM1, GSTT1, GSTP1) and risk of relapse in childhood B-cell precursor acute lymphoblastic leukemia: a case-control study. Blood, 95: 1222-8, 2000.[Abstract/Free Full Text]
  45. Lear JT, Smith AG, Heagerty AH, et al Truncal site and detoxifying enzyme polymorphisms significantly reduce time to presentation of further primary cutaneous basal cell carcinoma. Carcinogenesis (Lond), 18: 1499-1503, 1997.[Abstract/Free Full Text]
  46. Jiang J, Neubauer BL, Graff JR, et al Expression of group IIA secretory phospholipase A2 is elevated in prostatic intraepithelial neoplasia and adenocarcinoma. Am J Pathol, 160: 667-71, 2002.[Abstract/Free Full Text]
  47. Graff JR, Konicek BW, Deddens JA, et al Expression of group IIa secretory phospholipase A2 increases with prostate tumor grade. Clin Cancer Res, 7: 3857-61, 2001.[Abstract/Free Full Text]
  48. Yamashita S, Yamashita J, Ogawa M. Overexpression of group II phospholipase A2 in human breast cancer tissues is closely associated with their malignant potency. Br J Cancer, 69: 1166-70, 1994.[Medline]
  49. Bhatia S, Abonour R, Porcu P, et al High-dose chemotherapy as initial salvage chemotherapy in patients with relapsed testicular cancer. J Clin Oncol, 18: 3346-51, 2000.[Abstract/Free Full Text]



This article has been cited by other articles:


Home page
 Cancer Res.Home page
R. D. Palmer, N. L. Barbosa-Morais, E. L. Gooding, B. Muralidhar, C. M. Thornton, M. R. Pett, I. Roberts, D. T. Schneider, N. Thorne, S. Tavare, et al.
Pediatric Malignant Germ Cell Tumors Show Characteristic Transcriptome Profiles
Cancer Res., June 1, 2008; 68(11): 4239 - 4247.
[Abstract] [Full Text] [PDF]

Home page
 JAMAHome page
D. R. Feldman, G. J. Bosl, J. Sheinfeld, and R. J. Motzer
Medical Treatment of Advanced Testicular Cancer
JAMA, February 13, 2008; 299(6): 672 - 684.
[Abstract] [Full Text] [PDF]

Home page
 JCOHome page
J. Houldsworth, J. E. Korkola, G. J. Bosl, and R. S. K. Chaganti
Biology and Genetics of Adult Male Germ Cell Tumors
J. Clin. Oncol., December 10, 2006; 24(35): 5512 - 5518.
[Abstract] [Full Text] [PDF]

Home page
 Proc. Natl. Acad. Sci. USAHome page
D. Juric, S. Sale, R. A. Hromas, R. Yu, Y. Wang, G. E. Duran, R. Tibshirani, L. H. Einhorn, and B. I. Sikic
Gene expression profiling differentiates germ cell tumors from other cancers and defines subtype-specific signatures
PNAS, December 6, 2005; 102(49): 17763 - 17768.
[Abstract] [Full Text] [PDF]

Home page
 JCOHome page
E. A. Ronnen, G. V. Kondagunta, J. Bacik, S. Marion, D. F. Bajorin, J. Sheinfeld, G. J. Bosl, and R. J. Motzer
Incidence of Late-Relapse Germ Cell Tumor and Outcome to Salvage Chemotherapy
J. Clin. Oncol., October 1, 2005; 23(28): 6999 - 7004.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sugimura, J.
Right arrow Articles by Teh, B. T.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sugimura, J.
Right arrow Articles by Teh, B. T.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online