Panel of Candidate Biomarkers for Renal Cell Carcinoma Dong Su Kim,†,‡,9 Yoon Pyo Choi,†,§,| Suki Kang,§ Ming Qing Gao,§,| Baekil Kim,§,| Haeng Ran Park,§,| Young Deuk Choi,⊥ Jong Baek Lim,¶ Hyung Jin Na,9 Hye Kyung Kim,9 Young-Pyo Nam,‡ Mi Hyang Moon,‡ Hae Ree Yun,9 Dong Hee Lee,9 Won-Man Park,9 and Nam Hoon Cho*,§,| Research Division, DCD, Inc., Pohang Technopark, Pohang, Kyungbuk, 790-834, Korea, Department of Pathology, Yonsei University College of Medicine and Research Center for Human Natural Defense System, Yonsei University College of Medicine, Seoul, Korea, Brain Korea 21 Projects for Medical Science, Yonsei University College of Medicine, Seoul, Korea, Department of Urology, Yonsei University College of Medicine, Seoul, Korea, Department of Diagnostic Laboratory Medicine, Yonsei University College of Medicine, Seoul, Korea, and Genomine Research Division, Genomine, Inc., Pohang Technopark, Pohang, Kyungbuk, 790-834, Korea Received March 16, 2010
The timely diagnosis and therapeutic monitoring of human renal cell carcinoma (RCC) is limited by the lack of specific biomarkers. To identify candidate RCC biomarkers, we used 2-DE gel electrophoresis with mass spectrometry and 2-DE spot intensity-based ROC analysis to analyze 18 sets of paired normal and RCC tumor tissue including conventional, papillary, and chromophobe subtypes. Validation was performed with RCC patient plasma samples and confirmed by clustergram, shRNA, and immunohistochemistry assays. Cardinal candidates were evaluated by ELISA. The leading candidate biomarker that was upregulated in RCC samples according to the clustergram and validation analysis was nicotinamide N-methyltransferase (NNMT) (13/15, P < 0.0001). Other upregulated candidate biomarkers that were identified by this method include ferritin, hNSE, NM23, secretagogin, and L-plastin. The upregulation of NNMT in RCC was confirmed by immunoblotting and immunohistochemistry. Analysis of fractionated membrane-associated proteins identified CAP-G, mitofillin, tubulin R, RBBP7, and HSP27. Of these, RBBP7 and HSP27 were highly expressed in the chromophobe subtype of RCC (3/3) but were absent from conventional RCC (0/3). The triple combination of the NNMT, FTL, and hNSE biomarkers had the highest predictive capacity of 0.993, while NNMT was the single, most powerful candidate diagnostic biomarker for all types of RCC. Keywords: Renal cell carcinoma • proteome • tumor markers • nicotinamide N-methyltransferase • ferritin light chain • human neuron specific enolase
Introduction Renal cell carcinoma (RCC) is the third most common urological malignancy and accounts for approximately 3% of all malignancies worldwide.1 RCC is also one of the most refractory malignancies due to a metastasis rate of up to 25% at presentation and the lack of a curative therapy.1-4 Generally, RCC is highly resistant to conventional chemotherapy and radiation therapy,3 and the mortality from RCC exceeds 100,000 * To whom correspondence should be addressed. E-mail: cho1988@ yuhs.ac. Tel: +82-2-2228-1767. Fax: +82-2-362-0860. † These authors contributed equally. ‡ Research Division, DCD, Inc., Pohang Technopark. § Department of Pathology, Yonsei University College of Medicine and Research Center for Human Natural Defense SystemYonsei University College of Medicine. | Brain Korea 21 Projects for Medical Science, Yonsei University College of Medicine. ⊥ Department of Urology, Yonsei University College of Medicine. ¶ Department of Diagnostic Laboratory Medicine, Yonsei University College of Medicine. 9 Genomine Research Division, Genomine, Inc., Pohang Technopark.
3710 Journal of Proteome Research 2010, 9, 3710–3719 Published on Web 05/11/2010
per year worldwide.2 Despite significant progress that has been made in the treatment of metastatic RCC, nephrectomy remains the only effective treatment for localized RCC.5 RCC comprises five histologically distinct subtypes classified by morphologic and cytogenetic features including the clear cell (75%), papillary (10-15%), chromophobe (5-10%), collecting duct carcinoma, and unclassified subtypes. While these classifications correlate with tumor progression, the histologic type has been found to be more important for predicting the clinical and biological features associated with the progression of RCC tumors. The loss of the von Hippel-Lindau tumor suppressor gene in chromosomal region 3p is an early event during the evolution of the clear cell type of RCC and is associated with the worst clinical course.3 In contrast, the less common chromophobe subtype of RCC exhibits no change in 3p loss and is associated with a milder clinical course.4 Unlike the more common prostate or bladder cancers, RCC has no useful diagnostic biomarkers or simple diagnostic tools. The development of a screening test to detect early renal cell carcinoma is critical to reduce RCC-related mortality. Despite 10.1021/pr100236r
2010 American Chemical Society
research articles
Panel of Candidate Biomarkers for RCC Table 1. Clinicopathologic Characteristics of Tested Samples pathology no. normal patients age sex
grade
SS04-28358 SS04-30011 SS04-30345 SS04-30446
T2_n T3_n T4_n
T1_t T2_t T3_t T4_t
43 73 55 56
F M F F
4 3 4 benign
SS04-30701 SS04-31538 SS04-31703
T5_n T6_n T7_n
T5_t T6_t T7_t
74 53 35
M F F
nongrade 3 benign
SS04-32390 SS04-32472 SS04-33004 SS04-33261 SS04-33439 SS04-33840 SS04-33849 SS04-34206 SS01-12110 SS01-11651 SS01-13921
T8_n T9_n T10_n T11_n T12_n T13_n T14_n T15_n T34_n T35_n T36_n
T8_t T9_t T10_t T11_t T12_t T13_t T14_t T15_t T34_t T35_t T36_t
55 62 64 33 81 66 57 50 55 45 43
F F F M M M F M F M M
2 4 2 3 1 3 3 3 nongrade nongrade nongrade
diagnosisa RCC_con RCC_con RCC_con angiomyoplipoma_ ang RCC_chr RCC_con nephgrosclerosis_ nep RCC_con RCC_con papillary_pap RCC_con RCC_con RCC_con RCC_con RCC_con papillary_pap RCC_chr RCC_chr
a RCC, renal cell carcinoma; con, conventional; chr, chromophobe; pap, papillary.
many attempts to identify useful diagnostic, prognostic, and treatment monitoring biomarkers with new high-throughput technology techniques such as gene expression profiling6-9 and protein profiling,10-12 no clinically relevant screening assays are currently available to detect asymptomatic RCC. In order to identify clinically useful biomarkers, we used total tissue lysate and differential fractionation of RCC tissue lysate to enrich for membrane-associated proteins in order to detect RCC-specific cell-surface biomarkers that were present at very low levels. In addition, we also sought to identify a panel of biomarkers because previously identified single biomarkers have not had sufficient sensitivity or specificity for clinical applications.13,14 The goal of this study was to identify and verify the feasibility of single or combined panel of candidate RCC diagnostic biomarkers with treatment monitoring potential.
Material and Methods Clinical Samples. Frozen sample sets of 15 renal cell carcinoma and adjacent normal tissue were obtained from the Yonsei University College of Medicine. Two cases of benign nephrosclerosis and angiomyolipoma were used as controls. Frozen samples were also prepared from eight ovarian, three liver, three lung, and three cervical cancer tissue samples. The present study was approved by the ethical committee of the university. Each tumor/normal tissue set was isolated by surgical resection and stored in sterile bottles in a -80 °C freezer. Prior to protein extraction, sample cryosections were examined, and compared to the paraffin-embedded tissues. Each frozen tissue sample was selected for protein extraction when the matched cryosection was estimated to contain >98% of the desired cell type (tumor stage or normal). Tumor samples of renal cell carcinoma from several stages are summarized in Table 1. For ELISA assays, plasma was obtained from 88 RCC patients and 80 healthy volunteers. Plasma samples were stored at -80 °C before use. Protein Extraction. For total tissue protein profiling, frozen tissue was directly homogenized with a motor-driven homogenizer (PowerGen125, Fisher Scientific) in 2-DE lysis solution composed of 7 M urea, 2 M thiourea, 4% (w/v) 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate (CHAPS), 1%
(w/v) dithiothreitol (DTT), 2% (v/v) Pharmalyte, and 1 mM benzamidine. Proteins were extracted by vortexing for 30 min at room temperature and centrifugation at 15000g for 30 min at 15 °C. Insoluble material was discarded, and the soluble fraction was saved for 2-DE. For protein profiling of the insoluble tissue fraction containing membrane or membraneassociated proteins, frozen tissue was homogenized with a motor-driven homogenizer in an aqueous buffer solution containing 50 mM Tris-HCl pH 7.5, 150 mM NaCl, and 1 mM benzamidine. After centrifugation at 12000g for 1 h, the supernatant was discarded, and the insoluble material was used for the extraction of membrane proteins. The insoluble material was solubilized in 2-DE lysis solution and protein extracted in the same way as for total tissue protein extraction. The protein concentration of the tissue extract was determined by Bradford assay.15 2-DE and Image Analysis. Two-dimensional gel electrophoresis (2-DE) was basically performed as previously described.16 Briefly, 200 µg protein extract was separated by IEF using an IPG strip with a nonlinear pH gradient of 4 to 10 (Genomine Inc., Kyungbuk, Korea) for the first dimension, and then SDS-PAGE (26 cm ×20 cm format) for the second dimension. Proteins were detected by alkaline silver staining. Image analysis and quantification of protein spots were performed with PDQuest software (BioRad, Hercules, CA). The quantity of protein in each spot was normalized to the total valid spot intensity. Protein Profile Analysis. Clustering of samples and the creation of the protein expression profile were performed with Cluster and Tree View software (http://rana.lbl.gov/EisenSoftware) to analyze and display the differential pattern of protein expression. The statistical significance of the differences in average protein expression was determined with the Student’s t test. Spots with differences of at least (1.5-fold and a P < 0.05 were chosen for further investigation. We applied the median centering for data adjustment and the default options of hierarchical clustering using the uncentered correlation similarity matrix. The distance between the overall expression profiles for normal and tumor samples was determined as described previously by summing the absolute values of the spot intensity ratios as given by formula (a), where n is the number of included proteins and a and b are the normal or tumor samples.17 n
|
∑ log i)1
2
j i,a X j Xi,b
|
(a)
With this formula, we calculated the relative extent of expression changes in differentially expressed protein spots among normal and tumor tissues. The series of pairwise distance calculations of differentially expressed spots among samples generated profile distance-based matrixes. The resulting distance matrix was used to build neighbor joining trees with the MEGA software package (http://www.megasoftware.net). Protein Identification by Mass Spectrometry. For protein identification by peptide mass fingerprinting, protein spots were excised, digested with trypsin (Promega, Madison, WI), mixed with R-cyano-4-hydroxycinnamic acid in 50% acetonitrile/0.1% TFA, and subjected to MALDI-TOF analysis (Ettan MALDI-TOF Pro, Amersham Biosciences, Piscataway, NJ) as described.18 Spectra were collected from 350 shots per specJournal of Proteome Research • Vol. 9, No. 7, 2010 3711
research articles trum over m/z range 600-3000 and calibrated by two-point internal calibration using trypsin autodigestion peaks(m/z 842.5099, 2211.1046). Peak list was generated using Ettan MALDI-TOF Pro Evaluation Module (version 2.0.16). Threshold used for peak-picking was as follows: 5000 for minimum resolution of monoisotopic mass, 2.5 for S/N. The ProFound (http://prowl.rockefeller.edu/prowl-cgi/profound.exe) program was used to search the human NCBInr database (version 2010/ 04/01, 228540 entries) for protein identification. The following parameters were used for the database search: trypsin as the cleaving enzyme, a maximum of one missed cleavage, iodoacetamide (Cys) as a complete modification, oxidation (Met) as a partial modification, monoisotopic masses, and a mass tolerance of (0.1 Da. PMF acceptance criteria is probability scoring. Amino acid sequence-based protein identification by MALDIPSD spectra of selected N-terminal derivatized peptide19 and the derivatization reactions were performed as previously described.20 For protein identification, the fragment masses obtained from MALDI-PSD were used to search database (nr 2010-01-29, 10386837 sequence entries) using Ettan MALDITOF Pro software (Sonar, version 2.0.16) or the protein identification search engine, PepFrag (http://prowl.rockefeller.edu/ prowl/pepfrag.html). Spectrum processing for peak detection was performed using Ettan MALDI-TOF Pro Evaluation Module (version 2.0.16). Smoothing filter was used to generate a peak list with the threshold 3 < S/N. Search inputs included the measured precursors and the fragment ion masses. The mass tolerances used were (1.0 Da for average precursor ions mass and (0.5 Da for average fragment masses. The parameters for database searching were as follows: trypsin as the cleaving enzyme, a maximum of one missed cleavage, iodoacetamide (Cys) as a complete modification, all-taxa as organism type, and only y-type fragmentation ions were allowed as possibilities. Recombinant Protein Production and Generation of Antibodies. Recombinant protein was prepared for use as immunogens in the generation of antibodies. Poly(A)- mRNA was isolated from tumor and corresponding normal tissues for renal cell carcinoma using the RNeasy plus mini kit (Qiagen, Valencia, CA), and full-length cDNA for NNMT (GeneBank accession number: NM_006169) was generated and amplified by PCR. The PCR product was subcloned into the pBAD/MycHis A vector (Invitrogen, Carlsbad, CA) to construct the recombinant pBAD/Myc-His hNNMT vector. Recombinant hNNMT (rhNNMT) was expressed in Escherichia coli by arabinose induction and purified by affinity chromatography using Ni-NTA Chelating Agarose Resin (Peptron, Korea). For generation of a polyclonal antibody, recombinant hNNMT was used to immunize rabbits or rats. Antisera were collected after 3 months. Immunoblotting. Each 18 µg of sample proteins and controls were resolved with SDS-PAGE. Proteins were then transferred to nitrocellulose membranes using standard methods (Invitrogen). Blots were blocked with 5% nonfat dry milk freshly solubilized in 0.1% Tween-20 in phosphate-buffered saline (PBS-T), rocked on a rotating shaker for 1 h at room temperature, rinsed three times in PBS-T, and probed with 1:1000 antiNNMT IgG (rabbit, polyclonal; Genomine Research Division, Genomine, Inc., Pohang, Korea) or 1:500 anti-GAPDH IgG (goat, polyclonal; Santa Cruz) in PBS-T for 1 h at room temperature. Blots were rinsed three times in PBS-T, probed with an enzymelinked secondary antibody (horseradish peroxidase) in PBS-T (1:1500-10000) for 1 h at room temperature, and rinsed three 3712
Journal of Proteome Research • Vol. 9, No. 7, 2010
Kim et al. times in 100 mL PBS-T for 5 min each. Chemiluminescent detection reagents were used for visualization. Immunoblotting for GAPDH was performed to verify equivalent protein loading. Immunohistochemistry. A tissue microarray (TMA) composed of 795 patients with renal neoplasms diagosed between 1995 and 1997 from 20 hospitals in South Korea was used for validation. Recipient blocks were made using core needles of 2 mm. The paraffin donor blocks were prepared following a thorough evaluation of the hematoxylin/eosin-stained slides. Two adjacent areas of a carcinoma from the matching donor blocks were transplanted to the recipient blocks using a 2-mm core needle. We constructed an array of samples of adjacent normal areas from these patients using paraffin-embedded, formalin-fixed tissue blocks. The histologic subtypes were classified as clear cell, papillary, chromophobe, or collecting duct carcinoma according to the guidelines of the 1997 Union Internationale Contre le Cancer/American Joint Committee on Cancer (UICC/AJCC). The 4-µm sections were placed on silanecoated slides, deparaffinized, immersed in phosphate-buffered saline (PBS) containing 0.3% (v/v) hydrogen peroxide, and processed in a microwave oven (in 10 mM sodium citrate buffer, pH 6.5, for 15 min at 700 W). After blocking with 1% (w/v) bovine serum albumin in PBS containing 0.05% (v/v) Tween-20 for 30 min, they were incubated with diluted (1:2000) biotin-labeled rabbit anti-h6-rhNNMT polyclonal antibody at 4 °C for 16 h. Biotinylation of antibody was done using the antibody biotin-labeling Kit (Genomine Inc., Korea). As a negative control, normal rabbit serum and subtype-matched normal rabbit IgGs were used. The final reaction product was visualized with the addition of 0.03% (w/v) of 3,3′-diaminobenzidine tetrachloride for 5-20 min. Strong cytoplasmic staining was considered to be a positive result. Immunostaining was graded and scored as follows: 0 (no staining), 1+ (weak, diffuse staining), or 2+ (strong, diffuse staining). ELISA for NNMT. ELISA was performed to quantify NNMT protein in plasma. Wells (Corning, NY) were coated with affinity-purified polyclonal rabbit anti-h6-rhNNMT IgG (3 µg/ well) in coating buffer (10 mM Tris HCl pH 8.5) at 4 °C overnight and blocked with 0.1% bovine casein in TBS (50 mM Tris-HCl pH 7.5, 150 mM NaCl). Wells were washed four times with TBS-T (50 mM Tris HCl pH 7.5, 150 mM NaCl, 0.05% Tween-20) and incubated with diluted plasma (1/20, v/v) in incubation buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 0.6% Poly (ethylene glycol)-block-poly (propylene glycol)-block-poly (ethylene glycol) containing 1% polyethylene glycol 3,350, 1.5 mM CaCl2, 0.5 mM EDTA). After a 1-h incubation, wells were washed three times with TBS-T and incubated with biotinlinked polyclonal rabbit anti-h6-rhNNMT IgG (2.5 µg/mL) in incubation buffer containing 0.1% casein and 50 µg/mL bovine IgG for 1 h, and washed as before. Washed plates were incubated with diluted streptavidin-conjugated peroxidase (0.1 µg/mL) for 30 min in incubation buffer, washed again, and incubated with TMB substrate solution (KPL, Gaithersburg, MD, U.S.A.) for 10 min. The reaction was stopped by adding a onehalf volume of 2 N sulfuric acid, and the absorbance was measured at 450 nm. Cell Lines and Culture Conditions. Caki-1, kidney cancer cell line, was obtained from the Korean Cell Line Bank (KCLB, Seoul, Republic of Korea) and cultured in RPMI-1640 medium and DMEM containing 10% fetal bovine serum (FBS), 100 U/mL penicillin, and 100 µg/mL streptomycin (Gibco BRL, Grand Island, NY).
Panel of Candidate Biomarkers for RCC shRNA-Mediated mRNA Depletion. We used the SureSilencing shRNA Plasmid for Human NNMT (KH1607G for the GFP) and a scrambled sequence negative control plasmid (SABiosciences, Frederick, VA). The shRNA target sequences were 5′-CATGATTGGTGAGCAGAAGTT-3′ for NNMT and 5′GGAATCTCATTCGATGCATAC-3′ for the negative control. Cells were seeded overnight and transfected with the SureFECT transfection reagent (SABiosciences) according to the manufacturer’s protocol. At 24, 48, or 72 h after transfection with the shRNA plasmid for human NNMT or the negative control plasmid, cells were harvested, and total RNA was extracted using an RNeasy Protect Mini Kit (Qiagen). The cDNA was synthesized using SuperScript III Reverse Transcriptase kit (Invitrogen). The PCR reactions were then carried out with HotStarTaq DNA polymerase (Qiagen) and the following conditions: denaturation at 95 °C for 15 min, and 28 cycles of 95 °C for 40 s, 55.8 °C (for NNMT) or 53.3 °C (for GAPDH) for 1 min, and 72 °C for 1 min, with a final extension for 10 min at 72 °C. The following primers were used: forward primer for NNMT 5′- CCTTTGACTGGTCCCCAGTG-3′, reverse primer for NNMT 5′- CTGCTTGACCGCCTGTCTC-3′,21 forward primer for GAPDH 5′-ACAACTTTGGTATCGTGGAA-3′, and reverse primer for GAPDH 5′-AAATTCGTTGTCATACC AGG-3′. Cell Proliferation and Viability Analysis. Cells were seeded in a flat 96-well plate and incubated at 37 °C in 5% CO2. Cell proliferation was measured using a Cell Counting Kit-8 (Dojindo Laboratiries, Kumamoto, Japan), according to the manufacturer’s instructions. At 0, 24, 48, and 72 h after transfection with a shRNA plasmid specific for human NNMT or a scrambled sequence negative control, we measured these samples at a wavelength of 450 nm using a Molecular Devices VERSAmax microplate reader (Molecular Devices, Sunnyvale, CA). All experiments were performed in triplicate. (P < 0.05).
Results The Majority of the Differentially Expressed Proteins Were Down-Regulated in Renal Cell Carcinoma Samples. To identify differentially expressed proteins in renal cell carcinoma samples, 15 cases of RCC and two cases of non-neoplastic disease were examined. The 15 cases of RCC included 11 cases of clear cell carcinoma, two of chromophobic RCC, and two of papillary RCC. Tissues from the RCC lesion and adjacent normal tissue were subjected to protein profiling using 2-DE of unfractionated whole protein extract. Differential expression patterns were analyzed by quantification of the spot volume of expressed proteins and classification using clustering software. For a more comprehensive view of the similarity relationships between the samples, distance calculations were performed as described in the Experimental Procedures. Of the 1450 detectable protein spots, 108 spots were differentially expressed in RCC (P < 0.05) samples. Of these, 52 spots have a P < 0.02 and were subjected to cluster analysis. 83% of the protein spots were lower (Figure 1A) and nine spots were higher in RCC samples than adjacent normal tissue (see Supporting Information 1, Figure S1). The Expression pattern was more heterogeneous in RCC tissue than in the corresponding normal tissue (Figure 1B) as indicated by scattering on the extended distance in tree. Membrane Fraction Biomarkers Were Differentially Expressed in RCC Subtypes. Membrane or membrane associated proteins were barely detectable in the whole cell extract 2-DE gels; therefore, we enriched for membrane proteins using fractionation of tissue lysate and performed 2-DE analysis
research articles again. Using this approach, twenty three differentially expressed proteins for RCC had a significance of P < 0.05, and were subjected to cluster analysis (Figure 1C) and profile distance calculation (Figure 1D). In contrast to the total protein profile, half of the differentially expressed proteins in the membrane fraction were higher in RCC or in subtypes of RCC than in normal tissue. This result was reflected in the tree of clustergram which was constructed with 23 differentially expressed proteins, and in the profile distance relationship which was constructed with a neighbor-joining tree based on the protein profile distance calculation with 13 up-regulated proteins (Figure 1C and 1D). In the distance tree and clustergram, chromophobe cases were clustered out from conventional or papillary RCC cases. This relatedness was the result of the upregulated proteins identified in the membrane associated protein fraction (Figure 1D). In contrast, there was no distinct pattern related to the subtypes of RCC based on the downregulated proteins (data not shown). Identification of Up-Regulated Proteins in Renal Cell Carcinoma. Proteins up-regulated in RCC were identified by peptide mass fingerprinting using MALDI-TOF (Table 2) and confirmed by database searching with chemically assisted MALDI-PSD spectra (see Supporting Information 2, Table S1). The proteins up-regulated in RCC were nicotinamide-Nmethyl-transferase, secretagogin, L-plastin, human neuron specific enolase, nonmetastatic cell 1, ferritin light chain, and thioredoxin peroxidase. Some protein spots, for example spot No. 1011 identified as ferritin light chain (data not shown), were found to be redundant proteins and presumed to be modified forms of the parental polypeptide. Nicotinamide-N-methyltransferase (NNMT) was found to be the most commonly upregulated protein for all types of RCC. It was up-regulated in 13 out of 15 RCC cases with an average increase of 16.5-fold over normal tissue levels (P < 0.0001) (Table 2). Secretagogin, was also elevated (P < 0.023) and was detected in about half of the conventional RCC samples, but not in normal tissue, papillary RCC, or chromophobe RCC. NM23A (1.8-fold, P < 0.008) and L-plastin (3.9-fold, 0.002 < P) were also elevated, although the magnitude of changes was not as much as that for NNMT or secretagogin, the difference of changes was significant (P < 0.008). hNSE (P < 0.0001) and ferritin (P < 0.003), which was detected in serum using an enzyme immunoassay as previously described (28-29), were also elevated in RCC tissue (Figure 2A). Other proteins that were identified in the membrane fraction and found to be upregulated in all three subtypes of RCC included tubulin R (0.00002 < P), mitofilin (P ) 0.002), ECGF-1 (P < 0.003), and CAP-G (P < 0.007) (Table 2 and Figure 2B). ECGF-1 was found to have a heterogeneous expression pattern (P ) 0.01, 16 normal vs 7 RCCs) (data not shown). Proteins specifically elevated in chromophobe RCC samples were the retinoblastoma binding protein 7 (P ) 0.00004, 8 normal vs 3 chromophobe RCC; P ) 0.026, 3 conventional and 2 papillary vs 3 chromophobe), and the heat shock 27 kDa protein 1 (P ) 0.00001, 8 normal vs 3 chromophobe RCC; P < 0.0002, 3 conventional and 2 papillary vs 3 chromophobe) (Figure 2B). Both proteins were up-regulated in all three chromophobe RCCs and one of two papillary RCC, but did not have significantly different expression levels in conventional RCC and normal tissue. Expression of NNMT was Highly Elevated in RCC. Elevated NNMT expression was restricted to three distinct subtypes of RCC (Figure 2A and Table 2). The elevation of NNMT was Journal of Proteome Research • Vol. 9, No. 7, 2010 3713
research articles
Kim et al.
Figure 1. Clustergram and distance tree representing the expression profile created from 2-DE gel image analysis, expression profile similarity analysis, and clustering and profile distance calculations. (A) Clustergram. (B) Distance tree from total protein extract of tissues. Fifty-two spots that were up-regulated or down-regulated in RCC samples and had P < 0.02 were used for this analysis. (C) Clustergram. (D) Distance tree for the membrane associated protein fraction. Twenty three spots which were up-regulated or downregulated in RCC samples and had P < 0.05 were used for this analysis. Panels A and B: green color represents lower expression, red represents high expression.
confirmed by immunoblotting analysis (Figure 2C). NNMT was up-regulated in RCC, whereas in normal tissue it was appeared in relatively low level. We also examined NNMT expression levels in cervical, liver, lung, and ovarian cancer tissues using two-dimensional gel electrophoresis. The expression level of NNMT was calculated by comparing the spot quantity of NNMT in 200 µg of total protein from each sample, and normalizing to the total valid spot density of each sample (Figure 3A, B). In normal tissue, the level of NNMT was higher in the liver. The expression of NNMT in the normal liver tissue was 3.3-fold higher than in the normal kidney tissue. In addition, NNMT expression was elevated in most of the tumor tissue except the liver cancer tissue. In RCC, NNMT was markedly up-regulated by 16.5-fold, but was only slightly elevated in the cervical, lung, and ovarian cancer tissues (Figure 3B). The identity of NNMT was confirmed by immunoblotting (Figure 3C). Expression of NNMT Using Immunohistochemistry. NNMT expression was primarily restricted to RCC tumor tissue (Figure 4). All renal cell carcinoma cells were intensely immu3714
Journal of Proteome Research • Vol. 9, No. 7, 2010
noreactive for NNMT, irrespective of grading or histology. Lowgrade clear cell subtype RCC with anastomosing sinusoidal patterns had the strongest staining (Figure 4C), while highgrade RCC with sarcomatous differentiation showed less intense immunoreactivity (Figure 4D). Papillary RCC demonstrated staining restricted to tumor cells covering papillary cores with grade 1+ intensity (Figure 4E). Chromophobe RCC was intensely stained, grade 2+. (Figure 4F). Effects of the Depletion of NNMT on Cell Proliferation in Vitro. To determine if down-regulation of NNMT could suppress cell proliferation in vitro, we depleted NNMT expression with an NNMT-specific shRNA plasmid. By flow cytometry and immunofluorescence microscopy, transfection efficiency for Caki-1 cell line was over 90% (data not shown). NNMTspecific shRNA transfected Caki-1 cells showed a marked decrease in the NNMT mRNA and protein levels 48 h after transfection (Figure 5A, B). The depletion of NNMT expression decreased cell proliferation by 8 to 13% for Caki-1 (Figure 5C). NNMT as a Diagnostic Biomarker for RCC. To examine plasma NNMT levels and evaluate the potential NNMT as a
research articles
Panel of Candidate Biomarkers for RCC Table 2. Protein Identification of Up-Regulated Proteins in Renal Cell Carcinoma
sample origin
tissue
ratio
8/2
2.34/1.0E+000
16.5
53 32 45
17/8 20/11 17/7
2.36/1.0E+000 2.36/1.0E+000 2.41/1.0E+000
24.4 3.9 6.1
17.3/16.9
38
6/4
1.64/1.0E+000
1.8
5.5/5.2
20.0/19.7
53
9/5
2.22/1.0E+000
2.9
AAX37099 NP_006073.2
5.4/5.9 5.2/4.9
27.0/30.74 39.1/50.8
37 29
12/6 8/2
2.38/1.0E+000 2.26/1.0E+000
2.3 3.8
Q16891
6.5/6.1
82.4/84.1
19
11/3
2.32/1.0E+000
2.5
AAH18160.1
5.4/4.9
50.4/53.3
36
17/6
2.39/1.0E+000
2.9
P40121
6.3/5.9
40.6/38.7
32
13/3
2.26/1.0E+000
3.3
CAI41284.1
4.1/4.9
51.8/47.6
18
9/1
2.35/1.0E+000
2.6
NP001531.1
5.5/6.0
27.6/22.8
32
7/1
2.41/1.0E+000
5.8
protein name
accession
pIa/pIb
Mwa/ Mwb
2107
NNMT (nicotinamide-Nmethyltransferase) SCGN (secretagogin) L-plastin hNSE (human neuron specific enolase) NM23A (nonmetastatic cell1) FTL (ferritin light chain) PRDX4 tubulin R
NP_006160.1
5.6/5.1
30.0/28.8
30
NP_008929.2 NP_002289.1 1TE6|A
5.2/4.7 5.2/4.8 5.2/4.2
32.2/31.3 70.8/68.4 48.2/48.6
NP_000260.1
5.8/5.7
CAG32996.1
mitofilin proliferationinducing gene 4 protein ECGF-1 (endothelial-l cell growth factor 1) CAP-G (actin-regulatory protein) RBBP7 (retinoblastoma binding protein7) HSP27 (heat shock 27 kDa protein 1)
4010 2014 3113 2411 5810 1613 5412 618 3213 a
Est’Z score/ probability
spot no.
1212 1614 409
membrane fraction
sequence coverage (%)
Theoretically calculated.
b
matched peptides/ unmatched peptides
Observed.
Figure 2. Expression pattern of up-regulated proteins in kidney tumor and adjacent normal tissue. (A) 2-DE gel of total kidney tumor tissue extract, (b ) benign). (B) 2-DE gel of the membrane a fraction of kidney tumor tissue extract. (C) Immunoblot of NNMT in kidney tumor and normal tissue. Refer to Table. 1 for abbreviations.
diagnostic biomarker for RCC, a highly sensitive sandwich-type immunoassay was established and used to analyze NNMT levels in plasma from healthy volunteers (n ) 80) and RCC patients (n ) 88). The plasma levels of NNMT were elevated in RCC patients (median, 5231 pg/mL) in comparison with
healthy volunteers (median, 321 pg/mL) and the difference was significant (t test, P < 0.0001) (Figure 6A). Receiver operating characteristic (ROC) curve analysis was performed to assess the diagnostic accuracy of the test using statistical software, MedCalc (http://www.medcalc.be).32 When choosing a 90% Journal of Proteome Research • Vol. 9, No. 7, 2010 3715
research articles
Kim et al.
Figure 3. Expression pattern of NNMT in various tumor and normal tissues. (A) NNMT spots in 2-DE gels of tumor tissue extracts. Abbreviations: I∼III, the stage of cervical cancers; TC, typical carcinoma; SCLC, small-cell lung carcinoma; LCNEC, large-cell neuroendocrine carcinoma in lung cancer; ST, serous tumor; ET, endometrioid tumor; MT, mucinous tumor. (B) Graph of the relative NNMT intensity, normalized intensity value of 2-DE spots in various tumor (T) and normal (N) tissues. (C) Immunoblot of NNMT.
Figure 5. Depletion of NNMT in Caki-1 cell line. (A) Quantitative RT-PCR for NNMT mRNA expression at 24, 48, and 72 h after the transfection with NNMT-specific shRNA plasmid or negative control plasmid. (B) Immunoblot of NNMT protein expression at 48 h after transfection. (C) Effect of NNMT depletion on cell proliferation for Caki-1 at 24, 48, 72 h after transfection. The decrease of cell proliferation for Caki-1 at 24 h was 12.88%, at 48 h was 8.22%, and at 72 h was 8.42% as compared with control levels. The assay was performed in triplicate. (P < 0.05) GAPDH was used to normalize NNMT expression levels in quantitative RT-PCR and immunoblotting.
Figure 4. Immunohistochemistry of NNMT in tissue microarray. (A) Tissue microarray composed of each 2 cm cores from tumor sites. (B) No staining for NNMT in normal kidney. (C and D) Clear cell carcinoma, low grade and high grade (sarcomatous differentiation), revealed NNMT cytoplasmic staining. (E) Papillary RCC demonstrated weak staining along the tumor cells covering papillary cores. (F) Chromophobe RCC was stained diffusely and intensely.
specificity parameter, the sensitivity of NNMT for RCC was 81.8% and the corresponding area under curve (AUC) was 0.868 (Figure 6B). 3716
Journal of Proteome Research • Vol. 9, No. 7, 2010
A Potential Diagnostic Biomarker Panel for RCC. Using the protein expression profile, we performed spot intensity-based receiver operating characteristic (ROC) curve analysis. The potential accuracy, specificity, and sensitivity of each panel were assessed. Of the proteins up-regulated in RCC, NNMT, ferritin light chain, hNSE, and L-plastin were selected and surveyed in lung cancer (n ) 9), cervical cancer (n ) 9), liver cancer (n ) 9), ovarian cancer (n ) 14), RCC (n ) 15), and benign nephritic disease (n ) 2) tissue samples. Corresponding normal tissues (n ) 16) were utilized to verify the tumorspecificity of the biomarker panels. For RCC, NNMT was the most accurate with an area under curve (AUC) value of 0.962, which was higher than hNSE (0.889), ferritin (0.885), and L-plastin (0.748) (Figure 7A). NNMT had a higher specificity for RCC than lung cancer (AUC; 0.911), cervical cancer (0.923),
Panel of Candidate Biomarkers for RCC
research articles tumor tissue are seldom detectable in proteome analysis of the whole plasma due to the low plasma concentration in general. We approached this problem by digging the tumor tissue directly and by enriching tumor cell extracts for membrane proteins by fractionation. We considered the hypothesis that shedding of tumor cell membrane proteins release biomarkers into patient plasma.27 The candidate biomarkers were assessed for prognostic potential using plasma prospectively obtained from RCC patients before and after chemotherapy.
Figure 6. Receiver operating characteristic (ROC) curve for NNMT. Plasma levels of NNMT were measured by ELISA for RCC patients (n ) 88) and healthy donors (n ) 80). The concentration of NNMT in the plasma was calculated using protein standards and used as input for ROC curve analysis. (A) Interactive dot diagram. (B) ROC curve. The AUC for NNMT was 0.868.
and ovarian cancer (0.913) (see Supporting Information 2, Table S2). In contrast, hNSE had the lowest value (0.77) for RCC when compared to that for lung cancer (Figure 7B). The AUC value from the combined panel of NNMT, ferritin, and hNSE had the highest value for RCC (0.993) (Figure 7A and see Supporting Information 2, Table S2).
Discussion The identification of diagnostic and therapeutic biomarker candidates for RCC is a major goal of RCC research, and has led to identification of pro-MMP7, HIG-2, adipose-differentiation-related protein (ADRP), hsp27, and NMP22 (22-26). Yet these early candidate diagnostic biomarkers are did not have sufficient utility for clinical applications. To identify more useful RCC biomarkers, we preformed protein profiling of the RCC plasma, tissue and membrane protein enriched samples using 2-DE analysis. However, tumor biomarkers originated from
We found that the candidate biomarker levels varied according to the histologic RCC subtype, with the majority showing lower expression in RCC tumors. Observed heterogeneity of tumor compared to normal (shown in Figure 1B) comes in part of heterogeneity of tumor sample histology. We focused our study on the candidate biomarkers that were elevated in RCC tumors which included nicotinamide-N-methyltransferase (NNMT), secretagogin, L-plastin, neuron specific enolase (NSE), NM23, ferritin light chain, and thioredoxin peroxidase. While the ferritin light chain and NSE have been previously identified as candidate biomarkers, secretagogin and L-plastin are novel RCC biomarkers.28,29 We found that secretagogin was expressed mainly in the clear cell subtype of RCC, whereas L-plastin and NM23 were elevated in all types of RCC. Despite its cytoplasmic localization and presumed low levels in patient plasma, NNMT was initially selected for further evaluation because of the typeand grade-independent specificity. NNMT is a cytoplasmic enzyme which is mainly expressed in the liver where it metabolizes xenobiotics. NNMT has been previously identified as a candidate biomarker for colorectal, thyroid, gastric, and bladder cancer.21,30-32 Furthermore, a recent study demonstrated that elevated NNMT expression is correlated with lower cell migration rates.21 Depletion of NNMT with shRNA caused decreased Caki-1 cell proliferation. In the present study, NNMT was very strikingly elevated up to16.5-
Figure 7. Receiver operating characteristic (ROC) curve for proteins and panels. Normalized spot intensities from 2-DE gels were used. To assess the RCC-specificity of panels, RCC expression levels were designated as positive and expression levels in normal or benign tissue as negative. Regressed expression levels of each panel protein were used as the input data for the ROC of the panel. Logistic regression was used as a regression model, and the calculation was performed with Medcalc statistical software. The area under curve (AUC) expresses the diagnostic accuracy of the test. Complete separation between two groups would have an AUC value of 1. Journal of Proteome Research • Vol. 9, No. 7, 2010 3717
research articles fold in all subtypes of RCC (clear cell, papillary, and chromophobe). The clear cell and chromophobe subtypes of RCC had the highest NNMT expression, whereas the papillary type was only weakly positive. NNMT expression was inversely related to histologic grading, grade 1 or 2 clear cell subtypes had higher staining than RCC samples with higher graded RCC samples exhibiting sarcomatoid differentiation. In addition, tubulin R, proliferation-inducing gene 4 protein, endothelial cell growth factor 1, actin regulatory protein, retinoblastoma binding protein 7, and heat shock 27 kDa protein 1 were identified as candidate biomarkers using the membrane-associated fractions of RCC. The identified proteins could have originated either from the plasma membrane or the nuclear fraction because the membrane associated fraction was not prefractionated with low-speed centrifugation for removing the nuclear fraction. However, we found that the majority of cytosolic proteins was removed and have had a chance to reveal many low abundant proteins on gels which could not be detected in whole tissue protein profiling using 2-DE. Although it was crude fraction, unlike the candidate biomarkers from total tissue extract, membrane associated candidate biomarkers had differential expression in RCC subtypes. For example, retinoblastoma binding protein 7 and heat shock 27 kDa protein 1 were mainly up-regulated in chromophobe RCC. Besides tissue and membrane associated protein profiling, we performed plasma protein profiling of preand postsurgery RCC patients. (Data not shown in Results.) During plasma protein profiling, the proteins originated from tumor tissue could not be detected, but some complement pathway proteins were found to be higher in preoperative RCC patient plasma than those from postoperative RCC patients. One of them was a fragment of complement C4a containing the anaphylatoxin region (see Supporting Information 2, Figure S3A). This protein was also detected in the plasma from patients with other tumor types including ovarian and cervical cancer (see Supporting Information 2, Figure S3B). To evaluate the potential of the candidate biomarkers as diagnostic or prognostic markers, the plasma levels of selected proteins were measured for RCC patients and healthy volunteers. ROC curves were also produced using the tissue profiling intensity data. Using our highly sensitive and accurate sandwich immunoassay for NNMT, we found that the median NNMT level in RCC patient plasma was significantly higher than in healthy volunteer plasma. This indicates that NNMT has potential as a new diagnostic marker for RCC. However, the NNMT plasma levels from apparently healthy donors had a wide range from 101 pg/mL to 18,950 pg/mL. The origin of the observed high plasma level of NNMT (although it is a minor proportion) in apparently healthy donors needed to be deciphered by examination with a further expanded cohort of healthy donor and/or RCC patients. In particular, to increase the specificity of a plasma test for RCC, we suggest using a panel of biomarker sets and ROC analysis. ROC curves were also produced using the tissue profiling intensity data. The panels tested for this approach included combinations of NNMT, ferritin light chain, hNSE, and L-plastin. Of the tested panels, the combined panel of NNMT, ferritin, and hNSE had the highest AUC of up to 0.993. In conclusion, NNMT is a powerful diagnostic plasma biomarker for patients with all types of RCC, and can be combined with ferritin and hNSE to create a biomarker panel with diagnostic potential for RCC. 3718
Journal of Proteome Research • Vol. 9, No. 7, 2010
Kim et al. Abbreviations: 2-DE, Two-dimensional gel electrophoresis; ROC, Receiver operating characteristic; AUC, Area under curve; GAPDH, Glyceraldehyde 3-phosphate dehydrogenase.
Acknowledgment. This work was supported by the National Research Foundation of Korea (NRF) Grant funded by the Korea government (MEST) (No. 20090079165; CNH) and by a Faculty Research Grant of Yonsei University College of Medicine for 2008-0273. Supporting Information Available: Figure S1, 2-DE gels of total protein extract and membrane associated protein fractions from RCC and adjacent normal tissue; Figure S2, expression pattern of up-regulated proteins in RCC and comparison with the expression pattern in various solid tumors; Figure S3, 2-DE protein profile of RCC patient plasma; Table S1, protein identification of up-regulated proteins in renal cell carcinoma by MALDI-PSD; Table S2, multipanel protein ROC analysis. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Jemal, A.; Siegel, R.; Ward, E.; Murray, T.; Xu, J.; Smigal, C.; Thun, M. J. Cancer statistics. 2006. CA Cancer. J. Clin. 2006, 56, 196– 130. (2) Patel, P. H.; Changanti, R. S.; Motzer, R. J. Targeted therapy for metastatic renal cell carcinoma. Br. J. Cancer 2006, 94, 614–619. (3) Patard, J. J.; Leray, E.; Rioux-Leckercq, N.; Cindolo, L.; Ficarra, V.; Zisman, A.; De La Taille, A.; Tostain, J.; Artibani, W.; Abbou, C. C.; Lobel, B.; Guille´, F.; Chopin, D. K.; Mulders, P. F.; Wood, C. G.; Swanson, D. A.; Figlin, R. A.; Belldegrun, A. S.; Pantuck, A. J. Prognostic value of histologic subtypes in renal cell carcinoma: a multicenter experience. J. Clin. Oncol. 2005, 23, 2763–2771. (4) Godley, P.; Kim, S. W. Renal cell carcinoma. Curr. Opinion Oncol. 2002, 14, 280–285. (5) Hafez, K. S.; Fergany, A. F.; Novick, A. C. Nephron sparing surgery for localized renal cell carcinoma: impact of tumor size on patient survival, tumor recurrence and TNM staging. J. Urol. 1999, 162, 1930–3. (6) Takahashi, M.; Rhodes, D. R.; Furge, K. A.; Westphal, M.; Zhang, Z.; Kort, E. J.; The, B. T. Gene expression profiling of clear cell renal cell carcinoma: gene identification and prognostic classification. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 9754–9759. (7) Young, A. N.; Amin, M. B.; Moreno, C. S.; Lim, S. D.; Cohen, C.; Petros, J. A.; Marshall, F. F.; Neish, A. S. Expression profiling of renal epithelial neoplasms. A method for tumor classification and discovery of diagnostic molecular markers. Am. J. Pathol. 2001, 158, 1639–1651. (8) Takahashi, M.; Yang, X. J.; Sugimura, J.; Backdahl, J.; Tretiakova, M.; Qian, C. N.; Gray, S. G.; Knapp, R.; Anema, J.; Kahnoski, R.; Nicol, D.; Vogelzang, N. J.; Furge, K. A.; Kanayama, H.; Kagawa, S.; The, B. T. Molecular subclassification of kidney tumors and the discovery of new diagnostic markers. Oncogene 2003, 22, 6810– 6818. (9) Higgins, J. P.; Shinghal, R.; Gill, H.; Reese, J. H.; Terris, M.; Cohen, R. J.; Fero, M.; Pollack, J. R.; van de Rijn, M.; Brooks, J. D. Gene expression patterns in renal cell carcinoma assessed by complementary DNA microarray. Am. J. Pathol. 2003, 162, 925–932. (10) Bui, M. H.; Visapaa, H.; Seligson, D.; Kim, H.; Han, K. R.; Huang, Y.; Horvath, S.; Stanbridge, E. J.; Palotie, A.; Figlin, R. A.; Belldegrun, A. S. Prognostic value of carbonic anhydrase IX and KI67 as predictors of survival for renal clear cell carcinoma. J. Urol. 2004, 171, 2461–2466. (11) Seliger, B.; Fedorushchenko, A.; Brenner, W.; Ackermann, A.; Atkins, D.; Hanash, S.; Lichtenfels, R. Ubiquitin COOH-terminal hydrolase 1: a biomarker of renal cell carcinoma associated with enhanced tumor cell proliferation and migration. Clin. Cancer Res. 2007, 13, 27–37. (12) Seligson, D. B.; Pantuck, A. J.; Liu, X.; Huang, Y.; Horvath, S.; Bui, M. H.; Han, K. R.; Correa, A. J.; Eeva, M.; Tze, S.; Belldegrun, A. S.; Figlin, R. A. Epithelial cell adhesion molecule (KSA) expression: pathobiology and its role as an independent predictor of survival in renal cell carcinoma. Clin. Cancer Res. 2004, 10, 2659–2669.
research articles
Panel of Candidate Biomarkers for RCC (13) Mor, G.; Visintin, I.; Lai, Y.; Zhao, H.; Schwartz, P.; Rutherford, T.; Yue, L.; Bray-Ward, P.; Ward, D. C. Serum protein markers for early detection of ovarian cancer. Proc. Natl.Acad. Sci. U.S.A. 2005, 102, 7677–7682. (14) Zhang, Z.; Barnhill, S. D.; Zhang, H.; Xu, F.; Yu, Y.; Jacobs, I.; Woolas, R. P.; Berchuck, A.; Madyastha, K. R.; Bast, R. C. Combination of multiple serum markers using an artificial neural network to improve specificity in discriminating malignant from benign pelvic masses. Gynecol. Oncol. 1999, 73, 56–61. (15) Bradford, M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 1976, 72, 248–254. (16) Rabilloud, T.; Kieffer, S.; Procaccio, V.; Louwagie, M.; Louwagie, M.; Courchesne, P. L.; Patterson, S. D.; Martinez, P.; Garin, J.; Lunardi, J. Two-dimensional electrophoresis of human placental mitochondria and protein identification by mass spectrometry: toward a human mitochondrial proteome. Electrophoresis 1998, 9, 1006–1014. (17) Enard, W.; Khaitovich, P.; Klose, J.; Zo¨llner, S.; Heissig, F.; Giavalisco, P.; Nieselt-Struwe, K.; Muchmore, E.; Varki, A.; Ravid, R.; Doxiadis, G. M.; Bontrop, R. E.; Pa¨a¨bo, S. Intra- and interspecific variation in primate gene expression patterns. Science 2002, 296, 340–343. (18) Fernandez, J.; Gharahdaghi, F.; Mische, S. M. (1998) Routine identification of proteins from sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels or polyvinyl difluoride membranes using matrix assisted laser desorption/ionization-time of flight-mass spectrometry (MALDI-TOF-MS). Electrophoresis 1998, 19, 1036–1045. (19) Flensburg, J.; Belew, M. Characterization of recombinant human serum albumin using matrix-assisted laser desorption ionization time-of-flight mass spectrometry. J. Chromatogr. A 2003, 1009, 111–117. (20) Wang, D.; Kalb, S. R.; Cotter, R. J. Improved procedures for N-terminal sulfonation of peptides for matrix-assisted laser desorption/ionization post-source decay peptide sequencing. Rapid Commun. Mass Spectrom. 2004, 18, 96–102. (21) Wu, Y.; Siadaty, M. S.; Berens, M. E.; Hampton, G. M.; Theodorescu, D. Overlapping gene expression profiles of cell migration and tumor invasion in human bladder cancer identify metallothionein 1E and nicotinamide N-methyltransferase as novel regulators of cell migration. Oncogene 2008, 27, 6679–6689. (22) Sarkissian, G.; Fergelot, P.; Lamy, P.; Patard, J.; Culine, S.; Jouin, P.; Rioux-Leclercq, N.; Darbouret, B. Identification of pro-MMP-7
(23) (24) (25)
(26)
(27)
(28) (29) (30)
(31)
(32)
as a serum marker for renal cell carcinoma by use of proteomic analysis. Clin. Chem. 2008, 54, 574–581. Ozer, G.; Altinel, M.; Kocak, B.; Yazicioglu, A.; Goneme, F. Value of urinary NMP-22 in patients with renal cell carcinoma. Urology 2002, 60, 593–597. Seliger, B.; Dressler, S. P.; Lichtenfels, R.; Kellner, R. Candidate biomarkers in renal cell carcinoma. Proteomics 2007, 7, 4601–4612. Yao, M.; Tabuchi, H.; Nagashima, Y.; Nakaigawa, N.; Ishiguro, H.; Hamada, K.; Inayama, Y.; Kishida, T.; Hatton, K.; Yamada-Okabe, H.; Kubota, Y. Gene expression analysis of renal carcinoma: adipose differentiation-related protein as a potential diagnostic and prognostic biomarker for clear-cell renal carcinoma. J. Pathol. 2005, 205, 377–387. Togashi, A.; Katagiri, T.; Ashida, S.; Fujioka, T.; Maruyama, O.; Wakumoto, Y.; Sakamoto, Y.; Fujime, M.; Kawachi, Y.; Shuin, T.; Nakamura, Y. Hypoxia-inducible protein 2 (HIG2), a novel diagnostic marker for renal cell carcinoma and potential target for molecular therapy. Cancer Res. 2005, 65, 4817–4826. Black, R. A.; Rauch, C. T.; Kozlosky, C. J.; Peschon, J. J.; Slack, J. L.; Wolfson, M. F.; Castner, B. J.; Stocking, K. L.; Reddy, P.; Srinivasan, S.; Nelson, N.; Boiani, N.; Schooley, K. A.; Gerhart, M.; Davis, R.; Fitzner, J. N.; Johnson, R. S.; Paxton, R. J.; March, C. J.; Cerretti, D. P. A metalloproteinase disintegrin that releases tumour-necrosis factor-alpha from cells. Nature 1997, 385, 729–733. Singh, K. J.; Singh, S. K.; Suri, A.; Vijian, V.; Goswami, A. K.; Khullar, M. Serum ferritin in renal cell carcinoma: effect of tumor size, volume grade, and stage. Indian J. Cancer 2005, 42, 197–200. Yaman, O.; Baltaci, S.; Arikan, N.; Ozdiler, E.; Go¨g˘u ¨s¸, O.; Mu ¨ftu ¨og˘lu, Y. Z. Serum neuron specific enolase: can it be a tumour marker for renal cell carcinoma? Int. Urol.Nephrol. 1996, 28, 207–210. Xu, J.; Moatamed, F.; Caldwell, J. S.; Walker, J. R.; Kraiem, Z.; Taki, K.; Brent, G. A.; Hershman, J. M. Enhanced expression of nicotinamide N-methyltransferase in human papillary thyroid carcinoma cells. J. Clin. Endocrinol Metab. 2003, 88, 4990–4996. Roessler, M.; Rollinger, W.; Palme, S.; Hagmann, M. L.; Berndt, P.; Engel, A. M.; Schneidinger, B.; Pfeffer, M.; Andres, H.; Karl, J.; Bodenmu ¨ ller, H.; Ru ¨ schoff, J.; Henkel, T.; Rohr, G.; Rossol, S.; Ru ¨ sch, W.; Langen, H.; Zolg, W.; Tacke, M. Identification of nicotinamide N-methyltransferase as a novel serum tumor marker for colorectal cancer. Clin. Cancer Res. 2005, 11, 6550–6557. Lim, B. H.; Cho, B. I.; Kim, Y. N.; Kim, J. W.; Park, S. T.; Lee, C. W. Overexpression of nicotinamide N-methyltransferase in gastric cancer tissues and its potential post-translational modification. Exp. Mol. Med. 2006, 38, 455–465.
PR100236R
Journal of Proteome Research • Vol. 9, No. 7, 2010 3719
