- Email: [email protected]
Identification of FRAS1 as a human endometrial carcinoma-derived protein in serum of xenograft model
Gynecologic Oncology 127 (2012) 406–411
Contents lists available at SciVerse ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Identification of FRAS1 as a human endometrial carcinoma-derived protein in serum of xenograft model Jing Xu a, Wenjiao Min b, Xinyu Liu c, Chuan Xie a, Jing Tang a, Tao Yi a, Zhengyu Li a,⁎, Xia Zhao a,⁎ a b c
Department of Gynecology and Obstetrics, West China Second University Hospital, Sichuan University, Chengdu, People's Republic of China Department of Psychiatry, The People's Hospital of Sichuan Province, Chengdu, People's Republic of China The State Key Laboratory of Biotherapy, West China University Hospital, Sichuan University, Chengdu, People's Republic of China
H I G H L I G H T S ► A xenograft model system was developed to instantly detect proteins derived from cancer cells. ► FRAS1 was identified as a uniquely human-originated protein by Q-TOF-MS/MS and FT-ICR-MS/MS analyses. ► The xenograft model presented here might be of some values in the discovery of tumor biomarkers.
a r t i c l e
i n f o
Article history: Received 22 June 2012 Available online 26 July 2012 Keywords: Endometrial carcinoma Proteomics Xenograft model Biomarker FRAS1
a b s t r a c t Objective. The value of clinical specimens in identification of tumor biomarker is limited to the great individual variations among patients, the high complexity and dynamic range of protein components, and the possibility that the identified proteins are not produced by cancer cells but are due to secondary body defense mechanisms. Herein we developed a xenograft model system, in which human endometrial carcinoma cells formed a tumor in an immune-deficient nude mouse, to instantly detect proteins derived from cancer cells. Methods. Using one-dimensional electrophoresis, liquid chromatography, and Q-TOF-MS/MS and FT-ICR-MS/ MS analyses, the human-specific proteins in the serum of xenograft mouse could be identified and monitored without the great variations observed in a patient-based approach. Results. We successfully identified 224 proteins, in which 175 (78.1%) were of mouse origin, and 45 (20.1%) were unable to be assigned as human or mouse origin. FRAS1 was identified as a uniquely human-originated protein. Its expression profile was then confirmed by Western blotting in serum samples from xenograft mice, and patients bearing endometrial carcinoma and healthy controls. Conclusions. Although the mechanism of FRAS1 derivation by cancer cells remains to be illustrated, our results suggest that the xenograft model presented here should be a promising tool in the discovery of tumor biomarkers. © 2012 Elsevier Inc. All rights reserved.
Introduction Up to now, the majority of endometrial carcinoma cases are diagnosed in the early stage and its prognosis is usually good, still there are a group of cases with a high risk of recurrence or metastasis [1]. Many markers, such as CD171 [2], PTEN [3], and urokinase-type plasminogen activator (uPA) [4], have been documented and utilized in diagnosis, prognosis prediction, and subtype classification. In previous studies, we adopted a proteomic approach to compare the protein profiles between endometrial carcinoma and normal endometrium tissues [5]. A number of proteins with altered expression were identified, most of which had been reported to be involved in carcinogenesis. By RNA interference and ⁎ Corresponding authors at: Department of Gynecology and Obstetrics, West China Second University Hospital, Sichuan University, Chengdu 610041, People's Republic of China. E-mail addresses: [email protected] (Z. Li), [email protected] (X. Zhao). 0090-8258/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ygyno.2012.07.109
immunohistochemical analyses, we found that some proteins, such as cyclophilin A and calcyphosine, might serve as potential therapeutic targets or prognosis biomarkers for this disease [6]. Up to now, most of the proteomic studies focusing on biomarker discovery utilized dissected tissue sample, serum or other body fluid from clinical patients. However, given the great individual variations among patients, the high complexity, and dynamic range of proteins in serum samples, a considerable sample size was then required to minimize the sample bias and achieve any statistical significance. Also for the identified proteins, still it was difficult to determine whether they were produced by cancer cells or stromal tissue due to secondary body defense mechanisms. Herein we developed a xenograft model system, in which human endometrial carcinoma cells formed a tumor in an immune-deficient nude mouse, to instantly detect proteins derived from cancer cells. By mass spectrometric analyses, the human-specific proteins in the serum of xenograft mouse could be identified and monitored without the great variations observed in a patient-based approach. Theoretically, these proteins
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411
could be viewed as potential biomarkers uniquely derived from human cancer cells.
Materials and methods Endometrial carcinoma xenograft model and clinical serum samples The human endometrial carcinoma cell line HEC-1-B was purchased from the American Type Culture Collection (ATCC) and maintained in DMEM medium supplemented with 10% fetal bovine serum at 37 °C. The Institutional Animal Care and Treatment Committee of Sichuan University approved all studies herein. Healthy female nude mice (BALB/C, 6–8 weeks of age, nonfertile and 18–20 g each) were inoculated subcutaneously with HEC-1-B cells (5×106/100 μl of PBS/mouse) via the right flank. The tumor volume was evaluated according to the following formula: tumor volume (mm3)=0.52×length×width2. The blood samples were collected from the tumor-bearing mice by retro-orbital puncture at indicated time, and the mice were then sacrificed. The blood samples from healthy mice were collected as control samples. For the xenograftbearing or control group, 8 separate serum samples were prepared and stored at −80 °C, respectively. The dissected tumors were fixed in neutral buffered formalin and embedded in paraffin, and sections (5 μm) were stained with hematoxylin and eosin to confirm histology. Clinical serum samples were obtained from 8 patients bearing endometrial carcinoma who underwent surgical treatment at the Gynecologic Department of West China Second University Hospital, Sichuan University, in 2011. Serum samples obtained from 8 healthy volunteers served as controls. All serum samples were prepared and stored at −80 °C. A summary of clinical characteristics of the patients and controls was shown in Table 1. This work was approved by the Institutional Ethics Committee of Sichuan University, and informed consents were obtained from all patients and healthy volunteers prior to blood acquisition.
407
In-gel digestion The lanes were excised, and the gel encompassing a range of 8–120 kDa was carefully cut into 1-mm segments using a blade gel slicer, as shown in Fig. 1. Each gel slice was washed and in-gel digested according to the manufacturer's instructions (Promega, Madison, WI) with some slight modifications. In brief, the slices were destained twice with 50% (v/v) acetonitrile (ACN) in 100 mM NH4HCO3 overnight at 4 °C, and incubated with 10 μg/ml Trypsin Gold (Promega, Madison, WI) at room temperature for 1 h and then at 37 °C overnight covered with digestion buffer (40 mM NH4HCO3, 10% ACN). The liquid was then removed and saved, and tryptic peptides were extracted twice with 50% ACN, 5% TFA by sonication for 15 min. Then all extracts were pooled and dried in a SpeedVac at room temperature. Peptides were desalted using C18 ZipTips (Millipore, Bedford, MA) and reconstituted in 70% ACN, 0.1% TFA. Liquid chromatography–mass spectrometry analysis Nano-LC–MS/MS analysis was performed on a nanoflow LC system (CapLC, Waters) coupled to a Q-TOF Ultima mass spectromter (Waters). Peptide mixtures were trapped on a Jupiter™ C18 reversed phase column (Phenomenex; column dimensions, 1.5 cm × 100 μm, packed in-house) at a flow rate of 7 μl/min. Peptides were separated using a Jupiter C18 reversed phase column (Phenomenex; column dimensions, 15 cm × 50 μm, packed in-house) and a gradient of 0–80% acetonitrile in 0.1 M acetic acid in 70 min and at a constant flow rate of 200 nl/min. Fragmentation of the peptides was performed in data-dependent mode, and mass spectra were acquired in continuum mode. Nanoflow LC tandem mass spectrometry was performed by coupling an Agilent 1100 HPLC system (Agilent Technologies), operated as described previously [7]. The mass spectrometer was operated in data-dependent mode, automatically switching between MS and MS/MS acquisition. Full scan MS spectra were acquired by FT-ICR
One-dimensional gel electrophoresis After dilution and filtration, the abundant serum proteins, including albumin, immunoglobulins and transferrin, were removed from the serum samples using multiple affinity removal spin cartridge (Biosciences, St. Louis, MO) according to the manufacturer's instructions. The depleted serum samples were quantified by a Bradford assay kit (Bio-Rad, Hercules, CA) using bovine serum albumin as a standard. After being mixed with Laemmli sample buffer (1:1), the depleted serum samples were subjected to 4–20% linear gradient sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS-PAGE). Prestained high range molecular weight markers were loaded on each gel. The gels were stained using Coomassie Brilliant Blue G-250.
Table 1 Clinical characteristics of the patients and healthy controls. Clinical information Healthy controls (n = 8) Age (range, years) Endometrial carcinoma patients (n = 8) Age (range, years) FIGO grade G1 G2 G3 FIGO staging I II III Lymph node metastasis Yes No
49.1 ± 5.6 (43–58) 49.9 ± 7.0 (43–60) 2 3 3 4 2 2 2 6
Fig. 1. One-dimensional electrophoresis of the xenograft and control serum. The gel encompassing a range of 8–120 kDa was cut into 20 1-mm segments. The human endometrial carcinoma-derived FRAS1 protein was found in segment 3 of the xenograft sample.
408
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411
with a resolution of 20,000 at a target value of 2,000,000. The three most intense ions were then isolated for accurate mass measurements by a FT-ICR selected ion monitoring scan that consisted of 10-Da mass range with a resolution of 50,000 at a target accumulation value of 50,000. These ions were then fragmented in the linear ion trap using collision-induced dissociation at a target value of 15,000. MS/MS data from each LC run were converted to a single file in Mascot generic format. Database searching The resulting raw mass spectra from each pooled fraction were analyzed using Mascot (Matrix Science, London, UK; version 2.1.03) allowing 100-ppm (Q-TOF) or 5-ppm (FT-ICR-MS/MS) mass deviation for the precursor ion. Up to one missed cleavage was allowed, and searches were performed with fixed carbamidomethylation of cysteines and variable oxidation of methionine residues. Individual Mascot scores for each peptide MS/MS spectrum were ≥ 50. Scaffold (Proteome Software Inc., Portland, OR) was used to validate MS/ MS-based peptide and protein identifications. Peptide identifications were accepted if they could be established at greater than 90% probability as specified by the Peptide Prophet algorithm [8]. Protein identifications were accepted if they could be established at greater than 95% probability and contained at least two identified peptides. Protein probabilities were assigned by the Protein Prophet algorithm [9]. Before we annotated a certain peptide derived from the xenograft-bearing mice as human, a strict selection procedure was followed. Peptide mass values were searched against the Swiss-Prot database. The peptides annotated as human in the Swiss-Prot database were re-searched against the non-redundant International Protein Index (IPI) mouse database. Peptides also found in the mouse IPI database are obviously not considered as uniquely human tumor-derived. Peptides that were not identified in the mouse IPI database were also then confirmed as human in the human IPI database (≥62,000 entries). Western blotting The depleted serum samples were subjected to 4–20% linear gradient sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS-PAGE) as described above, and transferred to polyvinylidene difluoride (PVDF) membranes. After blocking with 5% dry milk in TBS-Tween 20, the membranes were incubated with rabbit anti-human FRAS1 antibody (Sigma Aldrich, St. Louis, MO) overnight at 4 °C. The blots were labeled with horseradish peroxidase-conjugated secondary antibodies and visualized by chemiluminescent detection. The equivalent loadings were confirmed using mouse anti-human β-actin antibody (Sigma Aldrich, St. Louis, MO). Results Comparison of serum protein expression between control and xenograftbearing model All eight healthy female BALB/C nude mice inoculated with HEC-1-B cells developed tumor at the right flanks. 5 days after inoculation, when the tumor diameters were about 5 mm, the first serum sample was collected from a tumor-bearing mouse by retro-orbital puncture, and the mouse was sacrificed and the tumor volume was recorded. Then, the other serum samples were collected one by one from the tumor-bearing mice every 5 days, and the tumor volumes were also recorded accordingly. During 40 days, the tumor volume increased, varying between 41.6 and 1048.3 mm 3. All serum samples in one group were pooled into one sample, which was separated on a one-dimensional SDS-PAGE gel after depletion of the high abundant proteins. A total of 224 proteins were identified by Q-TOF-MS/MS. Additionally, the same sample was also analyzed by FT-ICR-MS/MS,
and over 300 proteins were identified, including all the proteins found in the Q-TOF-MS/MS analysis. By data searching, 189 of the 224 proteins (84.4%) were observed in both xenograft and control groups, and some proteins were only found in either xenograft (20/ 224, 8.9%) or control (15/224, 6.7%) group. As could be expected, most of the identified proteins (175/224, 78.1%) were of mouse origin, and 45 proteins (20.1%) were unable to be assigned as human or mouse origin due to full homology of the identified peptides. However, using Q-TOF-MS/MS and FT-ICR-MS/MS, we still succeeded in identifying a limited number of uniquely humanoriginated proteins (4/224, 1.8%). These proteins were all observed in xenograft group, and presumed to be derived from the human endometrial carcinoma cells. MS identification of FRAS1 as human-derived protein In gel fragment 3 of the xenograft serum sample, five peptides were identified by Q-TOF-MS/MS (EWASSPCSVC, CGLGFYQAGSLC, LNLVGYCAD, KIHTPSLHV, and PGIQISSFTQAD) that originate from FRAS1 protein. The human (accession: NP_079350.5) and mouse (accession: NP_780682.3) FRAS1 proteins had a sequence identity of 85%, searched by Protein BLAST service in NCBI (http://blast.ncbi.nlm.nih. gov/). As shown in Fig. 2, the peptides CGLGFYQAGSLC, LNLVGYCAD, and PGIQISSFTQAD were all present in human and mouse sequences, making it unable to determine the origin of this protein. However, the EWASSPCSVC and KIHTPSLHV peptides made it possible to assign this protein as human. The mouse FRAS1 presents an alanine instead of serine, and a threonine instead of serine at the fifth and eighth positions of the EWASSPCSVC peptide, respectively. For the KIHTPSLHV peptide, the mouse FRAS1 presents a methionine instead of isoleucine, and a tyosine instead of histidine at the second and third positions, respectively. Proof of human origin of FRAS1 protein was also confirmed by FT-ICR-MS/MS using the same sample, resulting in identification of three of the same peptides (LNLVGYCAD, KIHTPSLHV, and PGIQISSFTQAD) and three additional peptides (HSINITIERKN, STFTMKDIYQ, and SSLYALESGS). Among the three additional peptides, HSINITIERKN and SSLYALESGS were present in both human and mouse sequence. STFTMKDIYQ peptide is specific for human FRAS1 protein, in which the mouse FRAS1 presents a serine instead of threonine, and a glutamic acid instead of lysine at the second and sixth positions, respectively. In the corresponding fragment in control serum sample, no peptide presenting in FRAS1 (either human or mouse origin) was found. Validation of FRAS1 using Western blotting analysis The presence of human FRAS1 in the xenograft serum samples was confirmed by Western blotting using a human-specific monoclonal antibody against FRAS1. A band at ~90 kDa corresponding to FRAS1 was detected in the pooled xenograft serum sample but not in the control sample. With the aim to observe the potential individual variation among samples, the Western blotting analyses were performed using the individual samples for both groups. As expected, expressions of FRAS1 were observed in all 8 xenograft serum samples, and no expression was observed in the 8 control serum samples (shown in Fig. 2). However, we failed to find some correlation between the tumor volume and FRAS1 expression intensity. Additionally, we also examined the expression of FRAS1 in clinical human specimens. Western blotting analyses were performed in serum samples from patients bearing endometrial carcinoma and healthy volunteers. As shown in Fig. 2, clear band corresponding to FRAS1 was observed in each of the 8 endometrial carcinoma samples, but no expression of FRAS1 was observed in healthy samples. Still we failed to find some correlation between the clinical characteristics of endometrial carcinoma patients, including staging, differentiation and metastasis, and the FRAS1 expression intensity.
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411
409
Fig. 2. Identification of human FRAS1 protein. (A) Primary amino acid sequence comparison between human and mouse FRAS1. The eight identified peptides (5 by Q-TOF-MS/MS, and 6 by FT-ICR-MS/MS, including 3 by both) were highlighted in shaded boxes. The human-specific peptides leading to identification of human FRAS1 were shown in larger font style, and the differences in amino acid between human and mouse were underlined. (B) The Mascot score histogram of human FRAS1 in Swiss-Prot database. (C) Validation of FRAS1 expression profiles in serum samples from mouse and patients. FRAS1 expressions were detected in all 8 individual xenograft mice serum samples and all 8 endometrial carcinoma patient serum samples. No expression was detected either in control mice serum samples or in healthy volunteer serum samples.
410
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411
Discussion Up to now, only a few studies have focused on the potential value of xenograft model in the biomarker searching using a proteome-broad approach. Roemer et al. identified mouse haptoglobin in the serum of a human mammary carcinoma xenograft model [9]. Additionally, its altered expression was also observed in serum from xenograft mice of stomach, colon, oral and brain cancer cells [10]. In the serum of xenograft mice with human prostate cancer, van den Bemd et al. identified human nucleoside-diphosphate kinase (NME) and six human enzymes involved in glycolysis [11]. In continued work, they isolated exosomes (small vesicles secreted by cancer cells) from prostate cancer cells, and identified some proteasome subunits and cytoplasmic proteins [12]. In the present study, we attempted to find potential biomarkers for human endometrial carcinoma in a xenograft model system. Immunodeficiency of the host minimized the possibility by definition that these proteins were produced by the host or stromal cells due to body defense mechanisms. Theoretically, the identified proteins were uniquely derived from human endometrial cancer cells. The high homology of amino acid sequences between human and mouse species usually makes it complicated to determine the origin of a protein [13]. Therefore, the ability of distinguishing of human-originated proteins from a complex mixture of mouse proteins should be a critical point to the feasibility of our work. To improve the reliability of protein identification, we adopted two mass spectrometry techniques Q-TOF and FT-ICR in this work. A total of 224 proteins, including 4 human-derived proteins, were successfully identified by Q-TOF-MS/MS; and over 300 proteins were identified by FT-ICR-MS/ MS. We found that FT-ICR-MS/MS could give more distinct peptides and increase sequence coverage for a certain protein, which clearly contributed to the assignment of its origin by data searching. Therefore, we suggested that FT-ICR-MS/MS should be superior to Q-TOF-MS/MS in the identification of protein origin due to greater sensitivity and accuracy. As expected, due to high homology of the amino acid sequences between human and mouse proteins, false positive and/or negative outcomes in the database searching were by definition impossible to be completely avoided. Therefore, we performed a stringent screening procedure in the database searching. The peptides in xenograft group were searched in human database, and the peptides annotated as human origin were re-searched against the non-redundant IPI mouse database. Peptides also found in the IPI mouse database are obviously not considered as uniquely human cancer cell-derived. In our results, over 40 proteins could not be assigned as human or mouse due to positive outcomes in both human and mouse database. Only those proteins identified in human database, but not in mouse database could be considered to be of human origin. In addition, the peptides in control groups were also searched in human database, and over 10 proteins were identified. Of course, these proteins should be regarded as false positive outcomes because by definition human-derived proteins were not presented in control serum sample. FRAS1 is an extracellular matrix protein that appears to function in the regulation of epidermal-basement membrane adhesion and organogenesis during development [14]. It is encoded by the FRAS1 (Fraser syndrome 1) gene, of which mutations are observed to cause Fraser syndrome [15]. Recent studies revealed a functional cooperation between the FRAS1/Frem gene products, in which FRAS1, Frem1 and Frem2 are simultaneously stabilized at the lowermost region of the basement membrane by forming a macromolecular ternary complex [16]. To the best of our knowledge, FRAS1 has rarely been reported in cancer research. Frem1 (FRAS1 related extracellular matrix 1) was observed by Yoon et al. in the conditioned media from starved cancer cells as a secretory factor, which could be induced by NF-κB and STAT3, and promote clonogenic capacities of cancer cells [17]. Still the mechanism of FRAS1 production by cancer cells remains unknown.
However, its overexpression in serum of endometrial carcinoma patients makes it feasible to serve as a potential clinical biomarker for endometrial carcinoma. We failed to find any correlation between the FRAS1 level and the clinical characteristics of the patients. We think that this negative outcome should be, at least partly, due to the limited sample size. The most common histological type of endometrial carcinoma is endometrioid carcinoma. We only selected specimens of endometrioid carcinoma in this work, with the aim to minimize the potential bias in histological difference in a limited sample size. As a natural and promising continuation to this work, we are planning to collect a considerable number of clinical specimens, and evaluate the significance of FRAS1 expression in early detection of disease and association with patient prognosis. However, even this approach could minimize the potential influence from body defense response; it is noteworthy that if a protein is originated not from the cancer cells but from stroma, but its expression is a direct consequence from the host response to the cancer, it can also be considered to be a useful biomarker [9]. In addition, we also understand that this approach is unable to by definition identify all the humanderived proteins in xenograft serum samples. For example, CA-125 is a well-known tumor marker for endometrial carcinoma. In our opinion, the failure to detect CA-125 in this work is, at least partly, associated with the following reasons. 1) CA-125 is a transmembrane glycoprotein presenting on the cell surface. The protein extraction procedures we adopted are general for soluble and hydrophilic proteins, of which the ability to extract transmembrane and lipophilic proteins is limited; 2) CA-125 can be secreted into serum and detected by some assays including ELISA. However, the high molecular weight (2358 kDa, UniProtKB Q8WX17) makes it difficult to be detected by SDS-PAGE (common range 5–150 kDa). The mass spectrometric analysis to discriminate the protein origin is based on the difference in peptide mass. Due to full homology of the amino acid sequences for a number of proteins between human and mouse species, false negative outcomes are theoretically inevitable. In further work, we will optimize the procedures in this protocol to facilitate the identification of low abundance proteins, including depletion of other high abundance serum proteins, two-dimensional LC separation, and improving mass spectrometric techniques, etc. Conflict of interest statement The authors declared that they had no conflict of interests.
Acknowledgments This work was supported by the Natural Science Foundation of China (Grant numbers: 30901594 and 81170592) and Program for New Century Excellent Talents in University (Grant number: NECT-10-0594). References [1] Trovik J, Mauland KK, Werner HM, Wik E, Helland H, Salvesen HB. Improved survival related to changes in endometrial cancer treatment, a 30-year population based perspective. Gynecol Oncol 2012;125:381-7. [2] Fogel M, Huszar M, Altevogt P, Ben-Arie A. L1 (CD171) as a novel biomarker for ovarian and endometrial carcinomas. Expert Rev Mol Diagn 2004;4:455-62. [3] Lacey Jr JV, Mutter GL, Ronnett BM, loffe OB, Duggan MA, Rush BB, et al. PTEN expression in endometrial biopsies as a marker of progression to endometrial carcinoma. Cancer Res 2008;68:6014-20. [4] Steiner E, Pollow K, Hasenclever D, Schormann W, Hermes M, Schmidt M, et al. Role of urokinase-type plasminogen activator (uPA) and plasminogen activator inhibitor type 1 (PAI-1) for prognosis in endometrial cancer. Gynecol Oncol 2008;108:569-76. [5] Li Z, Zhao X, Bai S, Wang Z, Chen L, Wei Y, et al. Proteomics identification of cyclophilin A as a potential prognostic factor and therapeutic target in endometrial carcinoma. Mol Cell Proteomics 2008;7:1810-23. [6] Li Z, Huang C, Bai S, Pan X, Zhou R, Wei Y, et al. Prognostic evaluation of epidermal fatty acid-binding protein and calcyphosine, two proteins implicated in endometrial cancer using a proteomic approach. Int J Cancer 2008;123:2377-83. [7] Keller A, Nesvizhskii AI, Kolker E, Aebersold R. Empirical statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search. Anal Chem 2002;74:5383-92.
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411 [8] Nesvizhskii AI, Keller A, Kolker E, Aebersold R. A statistical model for identifying proteins by tandem mass spectrometry. Anal Chem 2003;75:4646-58. [9] Roemer I, Vogel T, Otto A, Fichner I, Klose J. Analysis of mouse beta-haptoglobin chain by lectin affinoblotting detection. Electrophoresis 2001;22:3038-42. [10] Juan HF, Chen JH, Hsu WT, Huang SC, Chen ST, Yi-Chung Lin J, et al. Identification of tumor-associated plasma biomarkers using proteomic techniques: from mouse to human. Proteomics 2004;4:2766-75. [11] van den Bemd GJ, Krijqsveld J, Luider TM, van Rijswijk AL, Demmers JA, Jenster G. Mass spectrometric identification of human prostate cancer-derived proteins in serum of xenograft-bearing mice. Mol Cell Proteomics 2006;5:1830-9. [12] Jansen FH, Krijqsveld J, van Rijswijk A, van den Bemd GJ, van den Berg MS, van Weerden WM, et al. Exosomal secretion of cytoplasmic prostate cancer xenograft-derived proteins. Mol Cell Proteomics 2009;8:1192-205.
411
[13] Sundelin J, Laurent BC, Anundi H, Traqardh L, Larhammar D, Bjorck L, et al. Amino acid sequence homologies between rabbit, rat, and human serum retinol-binding proteins. J Biol Chem 1985;260:6472-80. [14] Pavlakis E, Chiotaki R, Chalepakis G. The role of Fras1/Frem proteins in the structure and function of basement membrane. Int J Biochem Cell Biol 2011;43:487-95. [15] Alazami AM, Shaheen R, Alzahrani F, Snape K, Saggar A, Brinkmann B, et al. FREM1 mutations cause bifid nose, renal agenesis, and anorectal malformations syndrome. Am J Hum Genet 2009;85:414-8. [16] Petrou P, Markrygiannis AK, Chalepakis G. The Fras1/Frem family of extracellular matrix proteins: structure, function, and association with Fraser syndrome and the mouse bleb phenotype. Connect Tissue Res 2008;49:277-82. [17] Yoon S, Woo SU, Kang JH, Kim K, Shin HJ, Gwak HS, et al. NF-κB and STAT3 cooperatively induce IL6 in starved cancer cells. Oncogene 2011, doi:10.1038/onc.2011.517.
Contents lists available at SciVerse ScienceDirect
Gynecologic Oncology journal homepage: www.elsevier.com/locate/ygyno
Identification of FRAS1 as a human endometrial carcinoma-derived protein in serum of xenograft model Jing Xu a, Wenjiao Min b, Xinyu Liu c, Chuan Xie a, Jing Tang a, Tao Yi a, Zhengyu Li a,⁎, Xia Zhao a,⁎ a b c
Department of Gynecology and Obstetrics, West China Second University Hospital, Sichuan University, Chengdu, People's Republic of China Department of Psychiatry, The People's Hospital of Sichuan Province, Chengdu, People's Republic of China The State Key Laboratory of Biotherapy, West China University Hospital, Sichuan University, Chengdu, People's Republic of China
H I G H L I G H T S ► A xenograft model system was developed to instantly detect proteins derived from cancer cells. ► FRAS1 was identified as a uniquely human-originated protein by Q-TOF-MS/MS and FT-ICR-MS/MS analyses. ► The xenograft model presented here might be of some values in the discovery of tumor biomarkers.
a r t i c l e
i n f o
Article history: Received 22 June 2012 Available online 26 July 2012 Keywords: Endometrial carcinoma Proteomics Xenograft model Biomarker FRAS1
a b s t r a c t Objective. The value of clinical specimens in identification of tumor biomarker is limited to the great individual variations among patients, the high complexity and dynamic range of protein components, and the possibility that the identified proteins are not produced by cancer cells but are due to secondary body defense mechanisms. Herein we developed a xenograft model system, in which human endometrial carcinoma cells formed a tumor in an immune-deficient nude mouse, to instantly detect proteins derived from cancer cells. Methods. Using one-dimensional electrophoresis, liquid chromatography, and Q-TOF-MS/MS and FT-ICR-MS/ MS analyses, the human-specific proteins in the serum of xenograft mouse could be identified and monitored without the great variations observed in a patient-based approach. Results. We successfully identified 224 proteins, in which 175 (78.1%) were of mouse origin, and 45 (20.1%) were unable to be assigned as human or mouse origin. FRAS1 was identified as a uniquely human-originated protein. Its expression profile was then confirmed by Western blotting in serum samples from xenograft mice, and patients bearing endometrial carcinoma and healthy controls. Conclusions. Although the mechanism of FRAS1 derivation by cancer cells remains to be illustrated, our results suggest that the xenograft model presented here should be a promising tool in the discovery of tumor biomarkers. © 2012 Elsevier Inc. All rights reserved.
Introduction Up to now, the majority of endometrial carcinoma cases are diagnosed in the early stage and its prognosis is usually good, still there are a group of cases with a high risk of recurrence or metastasis [1]. Many markers, such as CD171 [2], PTEN [3], and urokinase-type plasminogen activator (uPA) [4], have been documented and utilized in diagnosis, prognosis prediction, and subtype classification. In previous studies, we adopted a proteomic approach to compare the protein profiles between endometrial carcinoma and normal endometrium tissues [5]. A number of proteins with altered expression were identified, most of which had been reported to be involved in carcinogenesis. By RNA interference and ⁎ Corresponding authors at: Department of Gynecology and Obstetrics, West China Second University Hospital, Sichuan University, Chengdu 610041, People's Republic of China. E-mail addresses: [email protected] (Z. Li), [email protected] (X. Zhao). 0090-8258/$ – see front matter © 2012 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ygyno.2012.07.109
immunohistochemical analyses, we found that some proteins, such as cyclophilin A and calcyphosine, might serve as potential therapeutic targets or prognosis biomarkers for this disease [6]. Up to now, most of the proteomic studies focusing on biomarker discovery utilized dissected tissue sample, serum or other body fluid from clinical patients. However, given the great individual variations among patients, the high complexity, and dynamic range of proteins in serum samples, a considerable sample size was then required to minimize the sample bias and achieve any statistical significance. Also for the identified proteins, still it was difficult to determine whether they were produced by cancer cells or stromal tissue due to secondary body defense mechanisms. Herein we developed a xenograft model system, in which human endometrial carcinoma cells formed a tumor in an immune-deficient nude mouse, to instantly detect proteins derived from cancer cells. By mass spectrometric analyses, the human-specific proteins in the serum of xenograft mouse could be identified and monitored without the great variations observed in a patient-based approach. Theoretically, these proteins
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411
could be viewed as potential biomarkers uniquely derived from human cancer cells.
Materials and methods Endometrial carcinoma xenograft model and clinical serum samples The human endometrial carcinoma cell line HEC-1-B was purchased from the American Type Culture Collection (ATCC) and maintained in DMEM medium supplemented with 10% fetal bovine serum at 37 °C. The Institutional Animal Care and Treatment Committee of Sichuan University approved all studies herein. Healthy female nude mice (BALB/C, 6–8 weeks of age, nonfertile and 18–20 g each) were inoculated subcutaneously with HEC-1-B cells (5×106/100 μl of PBS/mouse) via the right flank. The tumor volume was evaluated according to the following formula: tumor volume (mm3)=0.52×length×width2. The blood samples were collected from the tumor-bearing mice by retro-orbital puncture at indicated time, and the mice were then sacrificed. The blood samples from healthy mice were collected as control samples. For the xenograftbearing or control group, 8 separate serum samples were prepared and stored at −80 °C, respectively. The dissected tumors were fixed in neutral buffered formalin and embedded in paraffin, and sections (5 μm) were stained with hematoxylin and eosin to confirm histology. Clinical serum samples were obtained from 8 patients bearing endometrial carcinoma who underwent surgical treatment at the Gynecologic Department of West China Second University Hospital, Sichuan University, in 2011. Serum samples obtained from 8 healthy volunteers served as controls. All serum samples were prepared and stored at −80 °C. A summary of clinical characteristics of the patients and controls was shown in Table 1. This work was approved by the Institutional Ethics Committee of Sichuan University, and informed consents were obtained from all patients and healthy volunteers prior to blood acquisition.
407
In-gel digestion The lanes were excised, and the gel encompassing a range of 8–120 kDa was carefully cut into 1-mm segments using a blade gel slicer, as shown in Fig. 1. Each gel slice was washed and in-gel digested according to the manufacturer's instructions (Promega, Madison, WI) with some slight modifications. In brief, the slices were destained twice with 50% (v/v) acetonitrile (ACN) in 100 mM NH4HCO3 overnight at 4 °C, and incubated with 10 μg/ml Trypsin Gold (Promega, Madison, WI) at room temperature for 1 h and then at 37 °C overnight covered with digestion buffer (40 mM NH4HCO3, 10% ACN). The liquid was then removed and saved, and tryptic peptides were extracted twice with 50% ACN, 5% TFA by sonication for 15 min. Then all extracts were pooled and dried in a SpeedVac at room temperature. Peptides were desalted using C18 ZipTips (Millipore, Bedford, MA) and reconstituted in 70% ACN, 0.1% TFA. Liquid chromatography–mass spectrometry analysis Nano-LC–MS/MS analysis was performed on a nanoflow LC system (CapLC, Waters) coupled to a Q-TOF Ultima mass spectromter (Waters). Peptide mixtures were trapped on a Jupiter™ C18 reversed phase column (Phenomenex; column dimensions, 1.5 cm × 100 μm, packed in-house) at a flow rate of 7 μl/min. Peptides were separated using a Jupiter C18 reversed phase column (Phenomenex; column dimensions, 15 cm × 50 μm, packed in-house) and a gradient of 0–80% acetonitrile in 0.1 M acetic acid in 70 min and at a constant flow rate of 200 nl/min. Fragmentation of the peptides was performed in data-dependent mode, and mass spectra were acquired in continuum mode. Nanoflow LC tandem mass spectrometry was performed by coupling an Agilent 1100 HPLC system (Agilent Technologies), operated as described previously [7]. The mass spectrometer was operated in data-dependent mode, automatically switching between MS and MS/MS acquisition. Full scan MS spectra were acquired by FT-ICR
One-dimensional gel electrophoresis After dilution and filtration, the abundant serum proteins, including albumin, immunoglobulins and transferrin, were removed from the serum samples using multiple affinity removal spin cartridge (Biosciences, St. Louis, MO) according to the manufacturer's instructions. The depleted serum samples were quantified by a Bradford assay kit (Bio-Rad, Hercules, CA) using bovine serum albumin as a standard. After being mixed with Laemmli sample buffer (1:1), the depleted serum samples were subjected to 4–20% linear gradient sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS-PAGE). Prestained high range molecular weight markers were loaded on each gel. The gels were stained using Coomassie Brilliant Blue G-250.
Table 1 Clinical characteristics of the patients and healthy controls. Clinical information Healthy controls (n = 8) Age (range, years) Endometrial carcinoma patients (n = 8) Age (range, years) FIGO grade G1 G2 G3 FIGO staging I II III Lymph node metastasis Yes No
49.1 ± 5.6 (43–58) 49.9 ± 7.0 (43–60) 2 3 3 4 2 2 2 6
Fig. 1. One-dimensional electrophoresis of the xenograft and control serum. The gel encompassing a range of 8–120 kDa was cut into 20 1-mm segments. The human endometrial carcinoma-derived FRAS1 protein was found in segment 3 of the xenograft sample.
408
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411
with a resolution of 20,000 at a target value of 2,000,000. The three most intense ions were then isolated for accurate mass measurements by a FT-ICR selected ion monitoring scan that consisted of 10-Da mass range with a resolution of 50,000 at a target accumulation value of 50,000. These ions were then fragmented in the linear ion trap using collision-induced dissociation at a target value of 15,000. MS/MS data from each LC run were converted to a single file in Mascot generic format. Database searching The resulting raw mass spectra from each pooled fraction were analyzed using Mascot (Matrix Science, London, UK; version 2.1.03) allowing 100-ppm (Q-TOF) or 5-ppm (FT-ICR-MS/MS) mass deviation for the precursor ion. Up to one missed cleavage was allowed, and searches were performed with fixed carbamidomethylation of cysteines and variable oxidation of methionine residues. Individual Mascot scores for each peptide MS/MS spectrum were ≥ 50. Scaffold (Proteome Software Inc., Portland, OR) was used to validate MS/ MS-based peptide and protein identifications. Peptide identifications were accepted if they could be established at greater than 90% probability as specified by the Peptide Prophet algorithm [8]. Protein identifications were accepted if they could be established at greater than 95% probability and contained at least two identified peptides. Protein probabilities were assigned by the Protein Prophet algorithm [9]. Before we annotated a certain peptide derived from the xenograft-bearing mice as human, a strict selection procedure was followed. Peptide mass values were searched against the Swiss-Prot database. The peptides annotated as human in the Swiss-Prot database were re-searched against the non-redundant International Protein Index (IPI) mouse database. Peptides also found in the mouse IPI database are obviously not considered as uniquely human tumor-derived. Peptides that were not identified in the mouse IPI database were also then confirmed as human in the human IPI database (≥62,000 entries). Western blotting The depleted serum samples were subjected to 4–20% linear gradient sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS-PAGE) as described above, and transferred to polyvinylidene difluoride (PVDF) membranes. After blocking with 5% dry milk in TBS-Tween 20, the membranes were incubated with rabbit anti-human FRAS1 antibody (Sigma Aldrich, St. Louis, MO) overnight at 4 °C. The blots were labeled with horseradish peroxidase-conjugated secondary antibodies and visualized by chemiluminescent detection. The equivalent loadings were confirmed using mouse anti-human β-actin antibody (Sigma Aldrich, St. Louis, MO). Results Comparison of serum protein expression between control and xenograftbearing model All eight healthy female BALB/C nude mice inoculated with HEC-1-B cells developed tumor at the right flanks. 5 days after inoculation, when the tumor diameters were about 5 mm, the first serum sample was collected from a tumor-bearing mouse by retro-orbital puncture, and the mouse was sacrificed and the tumor volume was recorded. Then, the other serum samples were collected one by one from the tumor-bearing mice every 5 days, and the tumor volumes were also recorded accordingly. During 40 days, the tumor volume increased, varying between 41.6 and 1048.3 mm 3. All serum samples in one group were pooled into one sample, which was separated on a one-dimensional SDS-PAGE gel after depletion of the high abundant proteins. A total of 224 proteins were identified by Q-TOF-MS/MS. Additionally, the same sample was also analyzed by FT-ICR-MS/MS,
and over 300 proteins were identified, including all the proteins found in the Q-TOF-MS/MS analysis. By data searching, 189 of the 224 proteins (84.4%) were observed in both xenograft and control groups, and some proteins were only found in either xenograft (20/ 224, 8.9%) or control (15/224, 6.7%) group. As could be expected, most of the identified proteins (175/224, 78.1%) were of mouse origin, and 45 proteins (20.1%) were unable to be assigned as human or mouse origin due to full homology of the identified peptides. However, using Q-TOF-MS/MS and FT-ICR-MS/MS, we still succeeded in identifying a limited number of uniquely humanoriginated proteins (4/224, 1.8%). These proteins were all observed in xenograft group, and presumed to be derived from the human endometrial carcinoma cells. MS identification of FRAS1 as human-derived protein In gel fragment 3 of the xenograft serum sample, five peptides were identified by Q-TOF-MS/MS (EWASSPCSVC, CGLGFYQAGSLC, LNLVGYCAD, KIHTPSLHV, and PGIQISSFTQAD) that originate from FRAS1 protein. The human (accession: NP_079350.5) and mouse (accession: NP_780682.3) FRAS1 proteins had a sequence identity of 85%, searched by Protein BLAST service in NCBI (http://blast.ncbi.nlm.nih. gov/). As shown in Fig. 2, the peptides CGLGFYQAGSLC, LNLVGYCAD, and PGIQISSFTQAD were all present in human and mouse sequences, making it unable to determine the origin of this protein. However, the EWASSPCSVC and KIHTPSLHV peptides made it possible to assign this protein as human. The mouse FRAS1 presents an alanine instead of serine, and a threonine instead of serine at the fifth and eighth positions of the EWASSPCSVC peptide, respectively. For the KIHTPSLHV peptide, the mouse FRAS1 presents a methionine instead of isoleucine, and a tyosine instead of histidine at the second and third positions, respectively. Proof of human origin of FRAS1 protein was also confirmed by FT-ICR-MS/MS using the same sample, resulting in identification of three of the same peptides (LNLVGYCAD, KIHTPSLHV, and PGIQISSFTQAD) and three additional peptides (HSINITIERKN, STFTMKDIYQ, and SSLYALESGS). Among the three additional peptides, HSINITIERKN and SSLYALESGS were present in both human and mouse sequence. STFTMKDIYQ peptide is specific for human FRAS1 protein, in which the mouse FRAS1 presents a serine instead of threonine, and a glutamic acid instead of lysine at the second and sixth positions, respectively. In the corresponding fragment in control serum sample, no peptide presenting in FRAS1 (either human or mouse origin) was found. Validation of FRAS1 using Western blotting analysis The presence of human FRAS1 in the xenograft serum samples was confirmed by Western blotting using a human-specific monoclonal antibody against FRAS1. A band at ~90 kDa corresponding to FRAS1 was detected in the pooled xenograft serum sample but not in the control sample. With the aim to observe the potential individual variation among samples, the Western blotting analyses were performed using the individual samples for both groups. As expected, expressions of FRAS1 were observed in all 8 xenograft serum samples, and no expression was observed in the 8 control serum samples (shown in Fig. 2). However, we failed to find some correlation between the tumor volume and FRAS1 expression intensity. Additionally, we also examined the expression of FRAS1 in clinical human specimens. Western blotting analyses were performed in serum samples from patients bearing endometrial carcinoma and healthy volunteers. As shown in Fig. 2, clear band corresponding to FRAS1 was observed in each of the 8 endometrial carcinoma samples, but no expression of FRAS1 was observed in healthy samples. Still we failed to find some correlation between the clinical characteristics of endometrial carcinoma patients, including staging, differentiation and metastasis, and the FRAS1 expression intensity.
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411
409
Fig. 2. Identification of human FRAS1 protein. (A) Primary amino acid sequence comparison between human and mouse FRAS1. The eight identified peptides (5 by Q-TOF-MS/MS, and 6 by FT-ICR-MS/MS, including 3 by both) were highlighted in shaded boxes. The human-specific peptides leading to identification of human FRAS1 were shown in larger font style, and the differences in amino acid between human and mouse were underlined. (B) The Mascot score histogram of human FRAS1 in Swiss-Prot database. (C) Validation of FRAS1 expression profiles in serum samples from mouse and patients. FRAS1 expressions were detected in all 8 individual xenograft mice serum samples and all 8 endometrial carcinoma patient serum samples. No expression was detected either in control mice serum samples or in healthy volunteer serum samples.
410
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411
Discussion Up to now, only a few studies have focused on the potential value of xenograft model in the biomarker searching using a proteome-broad approach. Roemer et al. identified mouse haptoglobin in the serum of a human mammary carcinoma xenograft model [9]. Additionally, its altered expression was also observed in serum from xenograft mice of stomach, colon, oral and brain cancer cells [10]. In the serum of xenograft mice with human prostate cancer, van den Bemd et al. identified human nucleoside-diphosphate kinase (NME) and six human enzymes involved in glycolysis [11]. In continued work, they isolated exosomes (small vesicles secreted by cancer cells) from prostate cancer cells, and identified some proteasome subunits and cytoplasmic proteins [12]. In the present study, we attempted to find potential biomarkers for human endometrial carcinoma in a xenograft model system. Immunodeficiency of the host minimized the possibility by definition that these proteins were produced by the host or stromal cells due to body defense mechanisms. Theoretically, the identified proteins were uniquely derived from human endometrial cancer cells. The high homology of amino acid sequences between human and mouse species usually makes it complicated to determine the origin of a protein [13]. Therefore, the ability of distinguishing of human-originated proteins from a complex mixture of mouse proteins should be a critical point to the feasibility of our work. To improve the reliability of protein identification, we adopted two mass spectrometry techniques Q-TOF and FT-ICR in this work. A total of 224 proteins, including 4 human-derived proteins, were successfully identified by Q-TOF-MS/MS; and over 300 proteins were identified by FT-ICR-MS/ MS. We found that FT-ICR-MS/MS could give more distinct peptides and increase sequence coverage for a certain protein, which clearly contributed to the assignment of its origin by data searching. Therefore, we suggested that FT-ICR-MS/MS should be superior to Q-TOF-MS/MS in the identification of protein origin due to greater sensitivity and accuracy. As expected, due to high homology of the amino acid sequences between human and mouse proteins, false positive and/or negative outcomes in the database searching were by definition impossible to be completely avoided. Therefore, we performed a stringent screening procedure in the database searching. The peptides in xenograft group were searched in human database, and the peptides annotated as human origin were re-searched against the non-redundant IPI mouse database. Peptides also found in the IPI mouse database are obviously not considered as uniquely human cancer cell-derived. In our results, over 40 proteins could not be assigned as human or mouse due to positive outcomes in both human and mouse database. Only those proteins identified in human database, but not in mouse database could be considered to be of human origin. In addition, the peptides in control groups were also searched in human database, and over 10 proteins were identified. Of course, these proteins should be regarded as false positive outcomes because by definition human-derived proteins were not presented in control serum sample. FRAS1 is an extracellular matrix protein that appears to function in the regulation of epidermal-basement membrane adhesion and organogenesis during development [14]. It is encoded by the FRAS1 (Fraser syndrome 1) gene, of which mutations are observed to cause Fraser syndrome [15]. Recent studies revealed a functional cooperation between the FRAS1/Frem gene products, in which FRAS1, Frem1 and Frem2 are simultaneously stabilized at the lowermost region of the basement membrane by forming a macromolecular ternary complex [16]. To the best of our knowledge, FRAS1 has rarely been reported in cancer research. Frem1 (FRAS1 related extracellular matrix 1) was observed by Yoon et al. in the conditioned media from starved cancer cells as a secretory factor, which could be induced by NF-κB and STAT3, and promote clonogenic capacities of cancer cells [17]. Still the mechanism of FRAS1 production by cancer cells remains unknown.
However, its overexpression in serum of endometrial carcinoma patients makes it feasible to serve as a potential clinical biomarker for endometrial carcinoma. We failed to find any correlation between the FRAS1 level and the clinical characteristics of the patients. We think that this negative outcome should be, at least partly, due to the limited sample size. The most common histological type of endometrial carcinoma is endometrioid carcinoma. We only selected specimens of endometrioid carcinoma in this work, with the aim to minimize the potential bias in histological difference in a limited sample size. As a natural and promising continuation to this work, we are planning to collect a considerable number of clinical specimens, and evaluate the significance of FRAS1 expression in early detection of disease and association with patient prognosis. However, even this approach could minimize the potential influence from body defense response; it is noteworthy that if a protein is originated not from the cancer cells but from stroma, but its expression is a direct consequence from the host response to the cancer, it can also be considered to be a useful biomarker [9]. In addition, we also understand that this approach is unable to by definition identify all the humanderived proteins in xenograft serum samples. For example, CA-125 is a well-known tumor marker for endometrial carcinoma. In our opinion, the failure to detect CA-125 in this work is, at least partly, associated with the following reasons. 1) CA-125 is a transmembrane glycoprotein presenting on the cell surface. The protein extraction procedures we adopted are general for soluble and hydrophilic proteins, of which the ability to extract transmembrane and lipophilic proteins is limited; 2) CA-125 can be secreted into serum and detected by some assays including ELISA. However, the high molecular weight (2358 kDa, UniProtKB Q8WX17) makes it difficult to be detected by SDS-PAGE (common range 5–150 kDa). The mass spectrometric analysis to discriminate the protein origin is based on the difference in peptide mass. Due to full homology of the amino acid sequences for a number of proteins between human and mouse species, false negative outcomes are theoretically inevitable. In further work, we will optimize the procedures in this protocol to facilitate the identification of low abundance proteins, including depletion of other high abundance serum proteins, two-dimensional LC separation, and improving mass spectrometric techniques, etc. Conflict of interest statement The authors declared that they had no conflict of interests.
Acknowledgments This work was supported by the Natural Science Foundation of China (Grant numbers: 30901594 and 81170592) and Program for New Century Excellent Talents in University (Grant number: NECT-10-0594). References [1] Trovik J, Mauland KK, Werner HM, Wik E, Helland H, Salvesen HB. Improved survival related to changes in endometrial cancer treatment, a 30-year population based perspective. Gynecol Oncol 2012;125:381-7. [2] Fogel M, Huszar M, Altevogt P, Ben-Arie A. L1 (CD171) as a novel biomarker for ovarian and endometrial carcinomas. Expert Rev Mol Diagn 2004;4:455-62. [3] Lacey Jr JV, Mutter GL, Ronnett BM, loffe OB, Duggan MA, Rush BB, et al. PTEN expression in endometrial biopsies as a marker of progression to endometrial carcinoma. Cancer Res 2008;68:6014-20. [4] Steiner E, Pollow K, Hasenclever D, Schormann W, Hermes M, Schmidt M, et al. Role of urokinase-type plasminogen activator (uPA) and plasminogen activator inhibitor type 1 (PAI-1) for prognosis in endometrial cancer. Gynecol Oncol 2008;108:569-76. [5] Li Z, Zhao X, Bai S, Wang Z, Chen L, Wei Y, et al. Proteomics identification of cyclophilin A as a potential prognostic factor and therapeutic target in endometrial carcinoma. Mol Cell Proteomics 2008;7:1810-23. [6] Li Z, Huang C, Bai S, Pan X, Zhou R, Wei Y, et al. Prognostic evaluation of epidermal fatty acid-binding protein and calcyphosine, two proteins implicated in endometrial cancer using a proteomic approach. Int J Cancer 2008;123:2377-83. [7] Keller A, Nesvizhskii AI, Kolker E, Aebersold R. Empirical statistical model to estimate the accuracy of peptide identifications made by MS/MS and database search. Anal Chem 2002;74:5383-92.
J. Xu et al. / Gynecologic Oncology 127 (2012) 406–411 [8] Nesvizhskii AI, Keller A, Kolker E, Aebersold R. A statistical model for identifying proteins by tandem mass spectrometry. Anal Chem 2003;75:4646-58. [9] Roemer I, Vogel T, Otto A, Fichner I, Klose J. Analysis of mouse beta-haptoglobin chain by lectin affinoblotting detection. Electrophoresis 2001;22:3038-42. [10] Juan HF, Chen JH, Hsu WT, Huang SC, Chen ST, Yi-Chung Lin J, et al. Identification of tumor-associated plasma biomarkers using proteomic techniques: from mouse to human. Proteomics 2004;4:2766-75. [11] van den Bemd GJ, Krijqsveld J, Luider TM, van Rijswijk AL, Demmers JA, Jenster G. Mass spectrometric identification of human prostate cancer-derived proteins in serum of xenograft-bearing mice. Mol Cell Proteomics 2006;5:1830-9. [12] Jansen FH, Krijqsveld J, van Rijswijk A, van den Bemd GJ, van den Berg MS, van Weerden WM, et al. Exosomal secretion of cytoplasmic prostate cancer xenograft-derived proteins. Mol Cell Proteomics 2009;8:1192-205.
411
[13] Sundelin J, Laurent BC, Anundi H, Traqardh L, Larhammar D, Bjorck L, et al. Amino acid sequence homologies between rabbit, rat, and human serum retinol-binding proteins. J Biol Chem 1985;260:6472-80. [14] Pavlakis E, Chiotaki R, Chalepakis G. The role of Fras1/Frem proteins in the structure and function of basement membrane. Int J Biochem Cell Biol 2011;43:487-95. [15] Alazami AM, Shaheen R, Alzahrani F, Snape K, Saggar A, Brinkmann B, et al. FREM1 mutations cause bifid nose, renal agenesis, and anorectal malformations syndrome. Am J Hum Genet 2009;85:414-8. [16] Petrou P, Markrygiannis AK, Chalepakis G. The Fras1/Frem family of extracellular matrix proteins: structure, function, and association with Fraser syndrome and the mouse bleb phenotype. Connect Tissue Res 2008;49:277-82. [17] Yoon S, Woo SU, Kang JH, Kim K, Shin HJ, Gwak HS, et al. NF-κB and STAT3 cooperatively induce IL6 in starved cancer cells. Oncogene 2011, doi:10.1038/onc.2011.517.