Candidate Diagnostic Markers and Tumor Suppressor Genes for Adrenocortical Carcinoma by Expression Profile of Genes on Chromosome 11q13
Gustavo G. Fernandez-Ranvier . Julie Weng . Ru-Fang Yeh . Daniel Shibru .
Elham Khafnashar . Ki-Wook Chung . Jimmy Hwang . Quan Yang Duh .
Orlo H. Clark . Electron Kebebew
Published online: 7 March 2008 @ Société Internationale de Chirurgie 2008
Abstract
Background The most common genetic change observed in adrenocortical carcinoma is loss of heterozygozity on chromosome 11q13. As genes on this chromosome may be important in the pathogenesis of adrenocortical carcinoma, we compared their expression profile between benign and malignant adrenocortical tissue.
Methods We used the Affymetrix GeneChip (U133 plus 2.0) array in 54 adrenocortical tumors (11 carcinoma and 43 benign). Differential gene expression was defined as a twofold higher or lower gene expression level (p < 0.05). Differentially expressed genes on microarray analysis were validated by real-time quantitative reverse-transcriptase polymerase chain reaction (RT-PCR). The area under the receiver operating characteristic (ROC) curve (AUC) was used to determined the diagnostic accuracy of the differ- ently expressed genes for distinguishing benign from malignant tumors.
Results We found 25 of the 314 genes on chromosome 11q13 to be differentially expressed between adrenocorti- cal carcinoma and benign adrenocortical tumor. All 25 were downregulated in adrenocortical carcinoma by 2-fold to 4.8-fold; 21 were validated to be differentially expressed by RT-PCR (Pearson’s coefficient > 0.5). Six genes (SERPING1, MRPL48, TM7SF2, DDB1, NDUSF8, PRDX5) validated by RT-PCR were significantly differ- entially expressed between benign and malignant adrenocortical tumors (p < 0.05) with an overall accuracy of 89% for SERPING1, 91% for MRPL48, 87% for TM7SF2, 88% for DDB1, 91% for NDUFS8, and 89% for PRDX5. The AUC was 0.89 for the combination of SER- PING1, MRPL48, TM7SF2, DDB1, and NDUFS8.
Conclusions We have identified 25 genes located on chromosome 11q13 that are downregulated in adrenocor- tical carcinoma and may be candidate tumor suppressor genes. Six of these genes were good diagnostic markers for distinguishing adrenocortical carcinoma from adenoma.
GustavoG. Fernandez-Ranvier . J. Weng . D. Shibru . K .- W. Chung . Q. Y. Duh . O. H. Clark . E. Kebebew Department of Surgery, University of California, San Francisco, School of Medicine, UCSF/Mt. Zion Medical Center, 1600 Divisadero Street, San Francisco, CA 94143-1674, USA e-mail: kebebewe@surgery.ucsf.edu
R .- F. Yeh Department of Epidemiology and Biostatistics, University of California, San Francisco, CA, USA
E. Khafnashar Department of Pathology, University of California, San Francisco, CA, USA
J. Hwang . E. Kebebew Comprehensive Cancer Center, University of California, San Francisco, CA, USA
Introduction
Adrenocortical tumors are relatively common, with a prevalence of 4% in the general population [1]. They most often occur sporadically and, less commonly, in familial syndromes such as multiple endocrine neoplasia type 1 (MEN1), Li-Fraumeni syndrome, Beckwith-Wiedemann syndrome, Carney complex, familial hyperaldosteronism type 1 and 2 (FH1, FH2), McCune-Albright syndrome, and congenital adrenal hyperplasia [2]. Most adrenocortical tumors are benign and are frequently diagnosed inciden- tally by ultrasound, computed tomography scanning, or magnetic resonance imaging.
Adrenocortical carcinomas are infrequent, with an incidence of two cases per million persons per year [3]. They are most often sporadic but may occur in patients with MEN1 and Li-Fraumeni syndrome [4, 5]. Adreno- cortical carcinomas are aggressive, with rapid development of locoregional invasion or distant metastasis. The 5-year survival rate is 50% for patients with resectable tumors, and the median survival rate is less than 1 year for those with metastatic disease [6].
The molecular mechanism of adrenocortical carcino- genesis is poorly understood. Several studies indicate that there is a significantly higher frequency of loss of hetero- zygosity (LOH) on chromosome 11q13 that is associated with sporadic adrenocortical carcinoma in 75%-100% of cases [7-10]. Although germline mutations of the Menin tumor suppressor gene, located on chromosome 11q13, are responsible for the MEN1 syndrome, and although up to 41% of patients with MEN1 syndrome may have an adrenal tumor, Menin gene mutations and/or deletions are absent in sporadic adrenocortical carcinoma [4, 7, 9]. These findings suggest that other genes on chromosome 11q13 may play an important role in the pathogenesis of adre- nocortical carcinoma. We therefore examined the expression profile of genes located on chromosome 11q13 to identify candidate genes that are dysregulated in adre- nocortical carcinoma.
Materials and methods
Patients and specimens
We analyzed 54 human adrenocortical tumor specimens from patients who were treated at the University of Cali- fornia San Francisco (UCSF) medical center from 1992 to 2006 (11 adrenocortical carcinomas from 7 patients; 43 benign adrenocortical tumors from 42 patients). Clinical and histopathologic diagnosis of unequivocal adrenocorti- cal carcinoma and benign adrenocortical tumor was confirmed in all cases. Samples of adrenocortical tumors that only produced aldosterone were excluded because they are rarely, if ever, malignant. Histological evaluation was done with hematoxylin & eosin staining to confirm that the specimen to be analyzed contained representative adreno- cortical tissue. The average follow-up time was 26.3 months for patients with benign adrenocortical tumors and 44.4 months for patients with adrenocortical carci- noma. Patient and tumor characteristics are summarized in Table 1. The study was approved by the UCSF Committee on Human Research, and consent was obtained from all patients.
RNA preparation and cDNA microarray hybridization
Adrenal tumor tissue was snap frozen at the time of adre- nalectomy and stored at -80℃. Frozen adrenal tumor tissue was sectioned and confirmed by hematoxylin & eosin histologic examination to be adrenocortical tissue. RNA was isolated using the Trizol reagent (Invitrogen Inc., Carlsbad, CA). RNA quantity was assessed with Nano- Drop®, and quality was assessed in an Agilent 2100 Bioanalyser (Agilent Technologies). Total RNA (1 µg) was amplified and labeled with the Ambion Kit (Ambion Inc, Austin,TX), according to the manufacturer’s protocol. Labeled samples were hybridized to the Affymetrix GeneChip (U133 plus 2.0) microarray.
Microarray data analysis and selection of differentially expressed genes
Probe intensities were extracted from scanned images with GeneChip Operating Software (GCOS, Affymetrix). The expression levels for 314 genes on chromosome 11q13 were determined. The criteria used for differential gene expression between adrenocortical carcinomas and adeno- mas were (1) twofold higher or lower gene expression level, and (2) p < 0.05 by t-test statistics.
Real-time quantitative reverse-transcriptase polymerase chain reaction (RT-PCR)
RT-PCR by the ABI PRISM®7900 Sequence Detection System (Applied Biosystems, Foster City, CA) was used to validate all the differentially expressed genes by cDNA microarray analysis. The same stock of total RNA as for the microarray analysis was reverse transcribed with the RT script cDNA synthesis kit (USB Corporation, Cleve- land, OH). Gene primers and probes were purchased from Applied Biosystems. The RT-PCR was performed in a final volume of 20 ul (9 ul of cDNA and 11 ul of PCR master mix). All experiments were done in triplicate, and the mRNA expression level of each gene was normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA expression levels. Normalized gene expression level = 2-(Ct for gene of interest - Ct for GAPDH) × 100%, where Ct is the PCR cycle threshold.
Gene ontology analysis
Gene ontology information was obtained from the National Center for Biotechnology Information (NCBI), http://www. ncbi.nlm.nih.gov
| Benign adrenocortical tumor | Adrenocortical carcinoma | p Value* | ||||
|---|---|---|---|---|---|---|
| n = 43 | (%) | n =11 | (%) | |||
| Gender | Male | 8 | (19) | 2 | (28.6) | |
| Female | 34 | (81) | 5 | (71.4) | ||
| M:F ratio | 1:4.25 | 1:2.5 | NS | |||
| Age at initial operation, mean ± SD (range) | 48.8 ± 11.9 (18-71) | 49.7 ± 21.8 (18-77) | NS | |||
| Tumor size, cm mean | 3.7 | 9.9 | 0.03 | |||
| Functioning | 23 (Cushing syndrome) | 7 (4 virilizing, 3 Cushing syndrome) | ||||
| Nonfunctioning | 20 | 4 | ||||
* The p values were determined with the chi-square test and the Mann-Whitney U-test for age and tumor size NS not significant
Statistical analysis
Statistical t-tests were used to determine differences in gene expression levels by cDNA microarray and by real- time quantitative RT-PCR. Chi-square tests were used to analyze categorical variables. When appropriate, continu- ous variables were analyzed with the non-parametric Mann-Whitney U-test. The area under the receiver oper- ating characteristic curve (ROC) was used to evaluate the diagnostic accuracy of differentially expressed genes to distinguish benign from malignant adrenocortical tumors.
Results
All genes significantly differentially expressed by cDNA microarray were downregulated in adrenocortical carci- noma. Of the 314 genes analyzed on chromosome 11q13, 25 were significantly differentially expressed between adrenocortical carcinoma and benign adrenocortical tumors (p < 0.05), and were downregulated in adrenocortical carcinoma by 2-fold to 4.8-fold (Table 2). Six of these genes, SERPING1, TM7SF2, FTH1, HSPC152, C110rf10, and FAU, had > 3-fold lower expression levels in adre- nocortical carcinomas.
Among the 25 significantly differentially expressed genes by cDNA microarray, 21 genes were strongly cor- related with real-time quantitative RT-PCR gene expression levels (Pearson’s coefficient > 0.5). Further- more, when their normalized gene expression levels were compared, 6 of the 25 genes (SERPING1, MRPL48, TM7SF2, DDB1, NDUSF8, PRDX5) were significantly differentially expressed between benign adrenocortical tumors and adrenocortical carcinomas (p < 0.05; Table 3).
The following genes were accurate diagnostic bio- markers: SERPING1, MRPL48, TM7SF2, DDB1, NDUSF8, PRDX5. The AUC, using the normalized gene expression values, was 0.79 for SERPING1, 0.85 for
MRPL48, 0.78 for TM7SF2, 0.86 for DDB1, 0.80 for NDUFS8, and 0.69 for PRDX5 (Fig. 1A and 1B). The sensitivity and specificity of these genes is summarized in Table 3. Five genes in combination (SERPING1, MRPL48, TM7SF2, DDB1, NDUSF8) had the highest AUC value (AUC = 0.89; Fig. 1C). Because tumor size is a common clinical parameter used for the diagnosis of adrenocortical carcinoma, we have also included a comparison of the AUC considering the tumor size (Fig. 1C).
The 25 differentially expressed genes possessed heter- ogeneous molecular function and biological processes, such as RNA processing, protein biosynthesis, cell cycle regulation, cell proliferation, apoptosis, cell adhesion, enzymatic pathways, signal transduction, transcription regulation, hemostasis, and immune response, among oth- ers (Table 4). Several genes (StarD10, FN5, FLJ20625, TncRNA, AYP1, C11orf2, CHCHD8), have no annotated molecular function or biological process yet. Some of the differentially expressed genes have important roles in cel- lular growth, apoptosis, and differentiation, the hallmarks of malignant phenotype (GSTP1, TCIRG1, FTH1, PRKRIR, BANF1, CFL1, and FAU).
Discussion
At present, there are no molecular markers that are accurate enough to help in distinguishing benign from malignant adrenocortical tumors. Because our goal was to identify candidate tumor suppressor genes and to evaluate their diagnostic utility as biomarkers to distinguish carcinoma from benign tumors, we determined the AUC of the nor- malized mRNA gene expression levels of the differentially expressed genes obtained by cDNA microarray analysis and validated by real-time quantitative RT-PCR. We found that DDB1 had the highest AUC when used alone (AUC: 0.86, overall accuracy: 88%). Comparison of several combinations in a multigene analysis showed that five
| Gene symbol | Gene title/definition | UniGene ID | Chromosomal location | p Value* | Fold changeª |
|---|---|---|---|---|---|
| SERPING1 (C1INH) | serpin peptidase inhibitor, clade G (C1 inhibitor), member 1 (angioedema, hereditary) | Hs.384598 | chr11q12- q13.1 | <0.01 | 3.43 |
| NXF1 | nuclear RNA export factor 1 | Hs.523739 | chr11q12-q13 | <0.01 | 2.0 |
| GSTP1 | glutathione S-transferase pi | Hs.523836 | chr11q13 | <0.01 | 2.26 |
| C11orf4 (MRPL49) | mitochondrial ribosomal protein L49 | Hs.75859 | chr11q13 | <0.01 | 2.59 |
| MRPL48 | mitochondrial ribosomal protein L48 | Hs.503239 | chr11q13.4 | <0.01 | 2.0 |
| STARD10 | START domain containing 10 | Hs.188606 | chr11q13 | <0.01 | 2.19 |
| MEN1 | multiple endocrine neoplasia I | Hs.423348 | chr11q13 | <0.01 | 2.0 |
| TCIRG1 | T-cell, immune regulator 1, ATPase, H + transporting, lysosomal V0 protein a isoform 3 | Hs.495985 | chr11q13.2 | <0.01 | 2.81 |
| TM7SF2 | transmembrane 7 superfamily member 2 | Hs.31130 | chr11q13 | <0.01 | 4.28 |
| FN5 | FN5 protein | Hs.438064 | chr11q13.3- q23.3 | <0.01 | 2.64 |
| C11orf59 (FLJ20625) | hypothetical protein FLJ20625 | Hs.530753 | chr11q13.4 | <0.01 | 2.19 |
| FTH1 | ferritin, heavy polypeptide 1 | Hs.558804 | chr11q13 | <0.01 | 4.84 |
| TncRNA | trophoblast-derived noncoding RNA | Hs.556258 | chr11q13.1 | <0.01 | 2.06 |
| AYP1 | AYP1 protein | Hs.397010 | chr11q13.1 | 0.01 | 2.67 |
| C11orf2 (ANG2) | chromosome 11 open reading frame2 | Hs.277517 | chr11q13 | 0.011 | 2.0 |
| DDB1 | damage-specific DNA binding protein 1, 127kDa | Hs.290758 | chr11q12-q13 | 0.013 | 2.85 |
| PRKRIR | protein-kinase, interferon-inducible double stranded RNA dependent inhibitor, repressor of (P58 repressor) | Hs.503315 | chr11q13.5 | 0.017 | 2.76 |
| NDUFS8 | NADH dehydrogenase (ubiquinone) Fe-S protein 8, 23kDa (NADH-coenzyme Q reductase) | Hs.90443 | chr11q13 | 0.018 | 2.0 |
| BANF1 | barrier to autointegration factor 1 | Hs.433759 | chr11q13.1 | 0.019 | 2.10 |
| HSPC152 | hypothetical protein HSPC152 | Hs.333579 | chr11q13.1 | 0.020 | 3.22 |
| CHCHD8 | coiled-coil-helix-coiled-coil-helix domain containing 8 | Hs.475387 | chr11q13.4 | 0.022 | 2.09 |
| CFL1 | cofilin 1 (non-muscle) | Hs.170622 | chr11q13 | 0.022 | 2.08 |
| C11orf10 | chromosome 11 open reading frame 10 | Hs.437779 | chr11q12- q13.1 | 0.026 | 3.25 |
| FAU | Finkel-Biskis-Reilly murine sarcoma virus (FBR-MuSV) ubiquitously expressed (fox derived); ribosomal protein S30 | Hs.387208 | chr11q13 | 0.029 | 3.30 |
| PRDX5 | peroxiredoxin 5 | Hs.502823 | chr11q13 | 0.031 | 2.40 |
* The p values were determined with the t-test
a All genes were downregulated in adrenocortical carcinoma
genes (SERPING1, MRPL48, TM7SF2, DDB1, NDUFS8) improved diagnostic accuracy (AUC: 0.89). We therefore propose a pool of six candidate genes (SERPING1, MRPL48, TM7SF2, DDB1, NDUFS8, PRDX5) as diag- nostic biomarkers, with an overall accuracy of 87%-91%, to distinguish benign from malignant adrenocortical tumors. Among this group of genes, SERPING1 encodes a protein with a serine-type endopeptidase inhibitor activity and peptidase activity, and it regulates activation of the complement, the intrinsic coagulation system, and the contact coagulation system. Alteration in SERPING1 is associated with hereditary angioneurotic edema [11]. The
TM7SF2 gene encodes the endoplasmic reticulum 3-ß hydroxysterol Delta (14)-reductase (C14SR), an enzyme that operates during the conversion of lanosterol to cho- lesterol in mammalian cells. Mutations in this gene have been implicated in lipid disorders [12]. The mammalian mitochondrial ribosomal proteins (MRPL48) are structural constituents of ribosomes and are involved in protein bio- synthesis and in translation, in addition to their function as initiation factors, but their molecular significance remains unknown. The DDB1 gene encodes a subunit of DNA damage-binding protein. This protein functions in nucleo- tide-excision repair of UV-damaged DNA and initiates the
| Gene symbol | Pearson's coefficientª | p Valueb | Sensitivityb (%) | Specificityb (%) | Overall accuracyb (%) | Area under ROC* |
|---|---|---|---|---|---|---|
| SERPING1 | 0.71 | <0.01 | 82 | 70 | 89 | 0.79 |
| MRPL48 | 0.72 | <0.01 | 93 | 82 | 91 | 0.85 |
| TM7SF2 | 0.73 | <0.01 | 73 | 84 | 87 | 0.78 |
| DDB1 | 0.57 | <0.01 | 73 | 93 | 88 | 0.86 |
| NDUFS8 | 0.69 | <0.01 | 73 | 93 | 91 | 0.80 |
| PRDX5 | 0.86 | 0.03 | 98 | 55 | 89 | 0.69 |
a The log2 microarray gene expression level versus the Ct on RT-PCR were compared using the Pearson’s correlation test
b Analysis based on RT-PCR normalized values
ROC receiver operating curve
nucleotide excision repair process. Its defective activity is associated with a repair defect in patients with xeroderma pigmentosum complementation group E (XPE) [13]. The NADH dehydrogenase (ubiquinone) Fe-S protein 8 enco- ded by the NDUFS8 gene is a subunit of one of the enzymes that constitute the mitochondrial respiratory chain, which is important in electron transport [14]. Mutations of the NDUFS8 gene result in Leigh syndrome [15]. The PRDX5 gene encodes a member of the perox- iredoxin family of antioxidant enzymes, which reduce hydrogen peroxide and alkyl hydroperoxides, thereby protecting the genome in response to oxidative stress [16]. The encoded protein may play an antioxidant protective role in different tissues under normal conditions and during inflammatory processes.
To our knowledge, there have been no previous cDNA microarray studies focusing on chromosome 11q13, a locus that has been found to have loss of heterozygosity (LOH) in 75%-100% of adrenocortical carcinomas.
Previous studies have identified several genes-insulin- like growth factor genes, among them-that appear to be important in adrenocortical tumorigenesis. Dysregulation of the IGF2 gene located at 11p15 seems to be the most common finding among different microarray studies in sporadic adrenocortical tumors, and it was overexpressed in carcinomas when compared with adenomas and normal adrenal cortex [17-19]. However, these studies have not provided information on how accurate IGF2 or other IGF- related genes might be as biomarkers to distinguish cancer from normal or benign disease. Another study identified two distinctive genetic clusters that differed between the adrenocortical carcinomas and adenomas: the IGF2 cluster and the steroideogenic cluster, which contain different genes encoding growth factors and steroideogenic and other enzymes, respectively [20]. High expression of IGF2 cluster genes was correlated with adrenocortical carcinoma in 75% of the cases, whereas a low level of expression of the steroideogenic cluster correlated with adrenocortical
carcinoma in 81% of the cases. This study also identified a cluster of genes that was helpful in discriminating between recurring and nonrecurring tumors. However, once again, although the authors identify a pool of genes that could help distinguish cancer from benign tumors, no further analysis has been done to determine their accuracy as clinical biomarkers.
The gene ontology analyses of the differentially expressed genes suggest many different molecular and biologic functions, as well as possible unknown functions. The six most underexpressed genes were SERPING1, TM7SF2, FTH1, HSPC152, C11orf10, and FAU (3- to 4.8- fold lower expression levels in adrenocortical carcinomas). To our knowledge, these genes have not been reported in previous adrenocortical cDNA microarray analyses. Of these genes, FTH1 seems to be the best characterized; it encodes the heavy subunit of ferritin, the major intracel- lular iron storage protein in prokaryotes and eukaryotes, and it is responsible for mammalian iron homeostasis. Alterations in FTH1 are associated with iron overload disorders [21, 22]. This protein plays also an important role in cell proliferation, actin filament polymerization, and protein complex assembly. The HSPC152 and C11orf10 genes are not well characterized. It is possible that C11orf10 may interact with the FEN1 gene, which plays an important role in DNA replication and repair in eukaryotic cells [23]; it has also been implicated in cell adhesion. The FAU gene is the cellular homolog of the fox sequence in the Finkel-Biskis-Reilly murine sarcoma virus (FBR- MuSV). It encodes a fusion ribosomal protein. Overex- pression of FAU induces cell death, hence, the proposed role of FAU in apoptosis and its importance during onco- genesis [24].
In our study, genes that regulate apoptosis and cellular growth (GSTP1, TCIRG1, PRKRIR, BANF1, CFL1, and FAU) were also underexpressed in adrenocortical carci- nomas. The GSTP1 gene has been extensively reported to be overexpressed in several cancers (breast, bladder, colon,
A
1.0
0.9
0.8
Sensitivity
0.7
0.6
0.5
SERPING1: 0.79
0.4
MRPL 48:
0.85
TM7SF2:
0.78
0.3
0.2
0.1
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1-Specificity
B
1.0
0.9
0.8
Sensitivity
0.7
0.6
0.5
0.4
0.3
0.2
DDB1:
NDUSF8:
0.86
0.1
PRDX5:
0.80
0.69
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1-Specificity
C
1.0
0.9
0.89
0.8
Sensitivity
0.7
0.71
0.6
0.5
0.4
0.3
multigene
0.2
analysis: 0.89
0.1
tumor
size: 0.71
0.0
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.0
1-Specificity
ovary, cervix) [25, 26]; however, hypermethylation of GSTP1, which results in gene silencing, is one of the most frequent epigenetic modifications reported in more than 90% of prostate carcinomas [27, 28]. The GSTP1 gene also encodes an important detoxifying enzyme that inactivates
carcinogens and products of oxidative stress, and it has been implicated in the regulation of cell proliferation and apoptosis [29]. It is possible that a similar epigenetic change may be a mechanism of tumorigenesis in adreno- cortical carcinomas.
Another interesting finding from our study is that the expression of the MEN1 gene, Menin, is downregulated. No MEN1 somatic mutations have been previously found in sporadic adrenocortical carcinomas, although a high frequency of LOH present at 11q13 and its known tumor suppressor function have been reported [7-9]. It remains to be determined whether epigenetic changes are respon- sible for the downregulation of MEN1 gene expression. Alterations in DNA methylation and histone acetylation status (epigenetic changes) are an important mechanism of gene silencing. Epigenetic changes or modifications that affect phenotype without altering DNA sequences or amount are increasingly recognized as being among the most common molecular alterations in human cancers. Tumor suppressor genes may be inactivated as a result of epigenetic changes (promoter CpG hypermethylation), as well as genetic alterations (deletions, chromosomal translocation, and point mutation). Proto-oncogenes are activated by epigenetic changes (promoter CpG demeth- ylation), as well as genetic alterations (gene amplification, chromosomal translocation, and point mutation). Genetic or epigenetic alterations in a variety of genes are funda- mental to the processes of growth, cell proliferation, differentiation, and programmed cell death and removal. This has been evidenced by many studies on hyperme- thylation of normally unmethylated promoter region (CpG islands) demonstrating transcriptional inactivation of sev- eral tumor suppressor genes (hMLH1, VHL, BRCA1, and RASSF1A) in a variety of human cancers [30-33]. It is possible that both genetic and epigenetic mechanisms may act together to inactivate gene expression, and these mechanisms need to be studied in the 25 differentially expressed genes if their role in adrenocortical carcinoma is to be appreciated.
There are several limitations to our study. We used a 2- fold or higher threshold to identify differentially expressed genes between benign and malignant adrenocortical tumors, and this criterion may be too stringent. Also, multiple comparison testing of the microarray results for the 314 genes may result in false-positive differentially expressed genes, but to address this concern we validated individually all 25 differentially expressed genes by microarray performing quantitative RT-PCR, an approach that is more accurate than any statistical inference (false discovery rate, etc.), giving to the data a unique power, something that has not been done in similar previously published studies. Lastly, because of our sample size, it would be important to validate the diagnostic accuracy of
| Gene symbol | Gene title | Molecular function | Biological process | Pathway |
|---|---|---|---|---|
| SERPING1 (C1INH) | serpin peptidase inhibitor, clade G (C1 inhibitor), member 1 (angioedema, hereditary) | Serine-type endopeptidase Inhibitor, peptidase activity | complement activation, classical pathway, immune response | – |
| NXF1 | nuclear RNA export factor 1 | Nucleotide and protein binding | mRNA export from nucleus | mRNA processing reactome |
| GSTP1 | glutathione S-transferase pi | Glutathione transferase activity | anti-apoptosis | circadian exercise, glutathione metabolism, metabolism of xenobiotics by cytochrome P450 |
| MRPL48, 49 (C11orf4) | mitochondrial ribosomal protein L48-L49 | Structural constituent of ribosome | protein biosynthesis | – |
| STARD10 | START domain containing 10 | – | – | – |
| MEN1 | multiple endocrine neoplasia I | Protein binding | regulation of transcription | – |
| TCIRG1 | T-cell, immune regulator 1, ATPase, H+ transporting, lysosomal V0 protein a isoform 3 | Transporter activity | cell proliferation | – |
| TM7SF2 | transmembrane 7 superfamily member 2 | Oxidoreductase activity, delta14-sterol reductase activity | sterol biosynthesis | – |
| FN5 | FN5 protein | – | – | – |
| FLJ20625 | hypothetical protein FLJ20625 | – | – | – |
| FTH1 | ferritin, heavy polypeptide 1 | Protein, kinase and ion binding | iron homeostasis, cell proliferation, actin filament polymerization, protein assembly | – |
| TncRNA | trophoblast-derived noncoding RNA | – | – | – |
| AYP1 | AYP1 protein | – | – | – |
| C11orf2 (ANG2) | chromosome 11 open reading frame2 | – | – | – |
| DDB1 | damage-specific DNA binding protein 1, 127kDa | Damaged DNA binding | nucleotide-excision repair, response to DNA damage stimulus | – |
| PRKRIR | protein-kinase, interferon-inducible double stranded RNA dependent inhibitor, repressor of (P58 repressor) | DNA binding, kinase activity, protein binding | regulation of translation, response to signal transduction, regulation of cell proliferation | stress, – |
| NDUFS8 | NADH dehydrogenase (ubiquinone) Fe-S protein 8, 23kDa (NADH-coenzyme Q reductase) | NADH Dehydrogenase activity, electron carrier activity, | mitochondrial electron transport | electron transport chain |
| BANF1 | barrier to autointegration factor 1 | DNA binding | response to virus | – |
| HSPC152 | hypothetical protein HSPC152 | Protein binding | – | – |
| CHCHD8 | coiled-coil-helix-coiled-coil-helix domain containing 8 | – |
| methane | coumarine and | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| metabolism, | |||||||||
| stilbene, | |||||||||
| proteins | biosynthesis | ||||||||
| Pathway | metabolism, | ||||||||
| – | ribosomal | phenylalanine | lignin | ||||||
| process | biosynthesis | response, | oxidative | ||||||
| cell | to | ||||||||
| Biological | homophilic | adhesion | protein | inflammatory | response | stress | |||
| function | activity | binding | of | ||||||
| Molecular | Transmembrane | receptor G13 signaling pathway | Calcium | constituent Structural | Peroxiredoxin ribosome | activity, | antioxidant activity | ||
| blood actin | ribosomal protein (FBR-MuSV) | ||||||||
| biogenesis, transduction, | 10 | derived); virus | |||||||
| frame | |||||||||
| and signal | sarcoma | ||||||||
| (fox | |||||||||
| organization protein adhesion | reading | murine | |||||||
| expressed | |||||||||
| Rho cell | open | ||||||||
| 11 | 5 | ||||||||
| title | (non-muscle) | cytoskeleton anti-apoptosis, coagulation, | chromosome | ubiquitously Finkel-Biskis-Reilly | peroxiredoxin | ||||
| 1 | |||||||||
| Gene | cofilin | S30 | |||||||
| matrix ion | |||||||||
| calcium | |||||||||
| continued | Extracellular and constituent. | ||||||||
| symbol | protein | ||||||||
| 4 | binding. structural | C11orf10 | |||||||
| Table | Gene | CFL1 | Actin, | FAU | PRDX5 | ||||
A dash indicates unknown molecular function, biological process, or pathway
the six genes in an independent dataset to determine their clinical diagnostic utility.
In summary, the results of our study show that the expression profiles of six genes located on chromosome 11q13 have the potential to be used as diagnostic markers of adrenocortical carcinoma, thereby distinguishing it from benign adrenocortical tumors. Our results also provide new evidence to support the possible inactivation of genes on chromosome 11q13 that was suggested by LOH. The mechanisms involved in the pathogenesis of adrenocortical tumors need to be further investigated. The absence of chromosomal and/or microsatellite instability may reflect the presence of genetic or epigenetic changes that could be the mechanism responsible for inactivation of gene expression, a critical step in the tumorigenic process. All these events might play an important role in early clinical diagnosis and, possibly, in the management and treatment of adrenocortical carcinoma.
Acknowledgments This work was supported in part by grants from the University of California San Francisco (UCSF) Comprehensive Cancer Center, the Mount Zion Health Fund, and the American Cancer Society.
The authors thank Dr. Christopher Barker of the UCSF Genomic Core for helpful technical assistance, and Pamela Derish, for helping to edit this manuscript.
References
1. Kjellman M, Kallioniemi OP, Karhu R et al (1996) Genetic aberrations in adrenocortical tumors detected using comparative genomic hybridization correlate with tumor size and malignancy. Cancer Res 56:4219-4223
2. Koch CA, Pacak K, Chrousos GP (2002) The molecular patho- genesis of hereditary and sporadic adrenocortical and adrenomedullary tumors. J Clin Endocrinol Metab 87:5367-5384
3. Fehaily MA, Duh QY (2003) Adrenocortical carcinoma. In Clark OH, Duh QY, Perrier N et al (eds) Endocrine Tumors: American Cancer Society, Atlas of Clinical Oncology. BC Decker, Ham- ilton, Ontario, pp 123-130
4. Skogseid B, Larsson C, Lindgren PG et al (1992) Clinical and genetic features of adrenocortical lesions in multiple endocrine neoplasia type 1. J Clin Endocrinol Metab 75:76-81
5. Houdelette P, Chagnon A, Dumotier J et al (1989) Malignant adrenocortical tumor as a part of Wermer’s syndrome. Apropos of a case. J Chir (Paris) 126:385-387
6. Sidhu S, Sywak M, Robinson B et al (2004) Adrenocortical cancer: recent clinical and molecular advances. Curr Opin Oncol 16:13-18
7. Heppner C, Reincke M, Agarwal SK et al (1999) MEN1 gene analysis in sporadic adrenocortical neoplasms. J Clin Endocrinol Metab 84:216-219
8. Kjellman M, Roshani L, Teh BT et al (1999) Genotyping of adrenocortical tumors: very frequent deletions of the MEN1 locus in 11q13 and of a 1-centimorgan region in 2p16. J Clin Endo- crinol Metab 84:730-735
9. Schulte KM, Mengel M, Heinze M et al (2000) Complete sequencing and messenger ribonucleic acid expression analysis of the MEN I gene in adrenal cancer. J Clin Endocrinol Metab 85: 441-448
10. Wachenfeld C, Beuschlein F, Zwermann O et al (2001) Dis- cerning malignancy in adrenocortical tumors: are molecular markers useful? Eur J Endocrinol 145:335-341
11. Davis AE (2005) The pathophysiology of hereditary angioedema. Clin Immunol 114:3-9
12. Bennati AM, Castelli M, Della Fazia MA et al (2006) Sterol dependent regulation of human TM7SF2 gene expression: role of the encoded 3beta-hydroxysterol Delta14-reductase in human cholesterol biosynthesis. Biochim Biophys Acta 1761:677-685
13. Cang Y, Zhang J, Nicholas SA et al (2007) DDB1 is essential for genomic stability in developing epidermis. Proc Natl Acad Sci USA 104:2733-2737
14. Lescuyer P, Martinez P, Lunardi J (2002) YY1 and Sp1 activate transcription of the human NDUFS8 gene encoding the mito- chondrial complex I TYKY subunit. Biochim Biophys Acta 1574: 164-174
15. Procaccio V, Wallace DC (2004) Late-onset Leigh syndrome in a patient with mitochondrial complex I NDUFS8 mutations. Neu- rology 62:1899-1901
16. Kropotov A, Serikov V, Suh J et al (2006) Constitutive expres- sion of the human peroxiredoxin V gene contributes to protection of the genome from oxidative DNA lesions and to suppression of transcription of noncoding DNA. FEBS J 273:2607-2617
17. Velázquez-Fernández D, Laurell C, Geli J et al (2005) Expression profiling of adrenocortical neoplasms suggests a molecular sig- nature of malignancy. Surgery 138:1087-1094
18. Giordano TJ, Thomas DG, Kuick R et al (2003) Distinct tran- scriptional profiles of adrenocortical tumors uncovered by DNA microarray analysis. Am J Pathol 162:521-531
19. Slater EP, Diehl SM, Langer P et al (2006) Analysis by cDNA microarrays of gene expression patterns of human adrenocortical tumors. Eur J Endocrinol 154:587-598
20. de Fraipont F, El Atifi M, Cherradi N et al (2005) Gene expression profiling of human adrenocortical tumors using com- plementary deoxyribonucleic acid microarrays identifies several candidate genes as markers of malignancy. J Clin Endocrinol Metab 90:1819-1829
21. Sanchez M, Galy B, Dandekar T et al (2006) Iron regulation and the cell cycle: identification of an iron-responsive element in the
3’-untranslated region of human cell division cycle 14A mRNA by a refined microarray-based screening strategy. J Biol Chem 281:22865-22874
22. Hayashi H, Wakusawa S, Motonishi S et al (2006) Genetic background of primary iron overload syndromes in Japan. Intern Med 45:1107-1111
23. Adachi N, Karanjawala ZE, Matsuzaki Y et al (2002) Two overlapping divergent transcription units in the human genome: the FEN1/C11orf10 locus. Omics 6:273-279
24. Mourtada-Maarabouni M, Kirkham L, Farzaneh F et al (2004) Regulation of apoptosis by fau revealed by functional expression cloning and antisense expression. Oncogene 23:9419-9426
25. Toffoli G, Viel A, Tumiotto L et al (1992) Expression of gluta- thione-S-transferase-pi in human tumours. Eur J Cancer 28A: 1441-1446
26. Meiers I, Shanks JH, Bostwick DG (2007) Glutathione S-trans- ferase pi (GSTP1) hypermethylation in prostate cancer: review 2007. Pathology 39:299-304
27. Bastian PJ, Ellinger J, Wellmann A et al (2005) Diagnostic and prognostic information in prostate cancer with the help of a small set of hypermethylated gene loci. Clin Cancer Res 11:4097-4106
28. Dobosy JR, Roberts JL, Fu VX et al (2007) The expanding role of epigenetics in the development, diagnosis and treatment of pros- tate cancer and benign prostatic hyperplasia. J Urol 177:822-831
29. Holley SL, Fryer AA, Haycock JW et al (2007) Differential effects of glutathione S-transferase pi (GSTP1) haplotypes on cell proliferation and apoptosis. Carcinogenesis 28:2268-2273
30. Herman JG, Umar A, Polyak K et al (1998) Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma. Proc Natl Acad Sci USA 95:6870-6875
31. Herman JG, Latif F, Weng Y et al (1994) Silencing of the VHL- suppressor gene by DNA methylation in renal carcinoma. Proc Natl Acad Sci USA 91:9700-9704
32. Baldwin RL, Nemeth E, Tran H et al (2000) BRCA1 promoter region hypermethylation in ovarian carcinoma: a population- based study. Cancer Res 60:5329-5333
33. Yoon JH, Dammann R, Pfeifer GP (2001) Hypermethylation of the CpG islands of the RASSF1A gene in ovarian and renal cell carcinomas. Int J Cancer 94:212-217