655.Full-Text.Pdf

Published OnlineFirst January 25, 2012; DOI: 10.1158/1940-6207.CAPR-11-0412

Cancer Prevention Research Article Research

Increased Plasma Levels of the APC-Interacting Protein MAPRE1, LRG1, and IGFBP2 Preceding a Diagnosis of Colorectal Cancer in Women

Jon J. Ladd1, Tina Busald1, Melissa M. Johnson1, Qing Zhang1, Sharon J. Pitteri1, Hong Wang1, Dean E. Brenner2, Paul D. Lampe1, Raju Kucherlapati3, Ziding Feng1, Ross L. Prentice1, and Samir M. Hanash1

Abstract Longitudinal blood collections from cohort studies provide the means to search for proteins associated with disease before clinical diagnosis. We investigated plasma samples from the Women’s Health Initiative (WHI) cohort to determine quantitative differences in plasma proteins between subjects subsequently diagnosed with colorectal cancer (CRC) and matched controls that remained cancer-free during the period of follow-up. Proteomic analysis of WHI samples collected before diagnosis of CRC resulted in the identification of six proteins with significantly (P < 0.05) elevated concentrations in cases compared with controls. Proteomic analysis of two CRC cell lines showed that five of the six proteins were produced by cancer cells. Microtubule-associated protein RP/EB family member 1 (MAPRE1), insulin-like growth factor– binding protein 2 (IGFBP2), leucine-rich alpha-2-glycoprotein (LRG1), and carcinoembryonic antigen (CEA) were individually assayed by enzyme linked immunosorbent assay (ELISA) in 58 pairs of newly diagnosed CRC samples and controls and yielded significant elevations (P < 0.05) among cases relative to controls. A combination of these four markers resulted in a receiver operating characteristics curve with an area under the curve value of 0.841 and 57% sensitivity at 95% specificity. This combination rule was tested in an independent set of WHI samples collected within 7 months before diagnosis from cases and matched controls resulting in 41% sensitivity at 95% specificity. A panel consisting of CEA, MAPRE1, IGFBP2, and LRG1 has predictive value in prediagnostic CRC plasmas. Cancer Prev Res; 5(4); 655–64. 2012 AACR.

Introduction physician there is only an approximately 50% rate of Current screening methods for colorectal cancer adherence (5). (CRC) have had an impact on mortality associated with Plasma levels of carcinoembryonic antigen (CEA) are this disease (1). There is a 30% to 40% drop in risk of currently used as a preoperative prognostic indicator for developing CRC following a negative result from a CRC, with higher levels of CEA being positively correlated colonoscopy and as much as a 50% reduction in inci- with a poor prognosis (6). CEA also has use for monitoring dence in the portion of the bowel examined by sigmoid- therapy in advanced disease and for patient surveillance oscopy or colonoscopy (2). Even with these decreases in following curative resection (7, 8). However, it lacks the risk and incidence, it is estimated that about 60% of sensitivity and specificity to be used as a diagnostic marker subjects older than 50 years in the United States are not for CRC (6), hence the need for additional markers that screened at recommended intervals (3, 4). In the case of could supplant it or augment its performance. colonoscopies, even when subjects are referred by their The ease of sampling plasma makes it a logical choice for the development of a panel of proteins that inform about risk of developing CRC. However, the plasma proteome is extremely complex and is composed of proteins ranging in Authors' Affiliations: 1Molecular Diagnostics Program, Fred Hutchinson concentration over at least 9 orders of magnitude. Recent Cancer Research Center, Seattle, Washington; 2Department of Internal work has shown that low-abundance plasma proteins may Medicine, University of Michigan Medical Center, Ann Arbor, Michigan; and 3Harvard-Partners Center for Genetics and Genomics and Harvard Medical be identified with high confidence following extensive School, Boston, Massachusetts plasma fractionation (9). High-abundant proteins interfere Note: Supplementary data for this article are available at Cancer Prevention with detection and quantification of less-abundant pro- Research Online (http://cancerprevres.aacrjournals.org/). teins, necessitating their removal before mass spectrometric Corresponding Author: Jon Ladd, Fred Hutchinson Cancer Research (MS) analysis, typically through immunodepletion. After Center, 1100 Fairview Avenue North, Seattle, WA 98109. Phone: 206- removal of the most highly abundant proteins, samples still 667-5778; Fax: 206-667-2537; E-mail: [email protected] require extensive fractionation by anion exchange and/or doi: 10.1158/1940-6207.CAPR-11-0412 reverse-phase chromatography to decomplex the sample to 2012 American Association for Cancer Research. achieve adequate sampling of the plasma proteome.

www.aacrjournals.org 655

Downloaded from cancerpreventionresearch.aacrjournals.org on September 30, 2021. © 2012 American Association for Cancer Research. Published OnlineFirst January 25, 2012; DOI: 10.1158/1940-6207.CAPR-11-0412

Ladd et al.

Guidelines for the design of biomarker discovery and S1). In 4 experiments, the pool of cases was labeled with validation studies have been recommended (10). Retro- light acrylamide, and its matched control pool was labeled spective longitudinal repository studies are used to evaluate with heavy 1,2,3-13C-acrylamide isotope. The labeling was biomarkers for their capacity to detect preclinical disease as switched in the other case–control pools. In each experi- a function of time before clinical diagnosis, as well as other ment, the pool of cases and the pool of matched controls sample characteristics that may define clinical applications. were mixed together before further processing and mass This is done through analysis of the most promising mar- spectrometry. kers and developing algorithms for screening positivity Proteins were separated by an automated online 2D- based on a combination of markers. The use of specimens HPLC system controlled by Workstation Class-VP 7.4 (Shi- collected before diagnosis through longitudinal cohort madzu Corporation). Separation consisted of anion studies meets prospective-specimen-collection, retrospec- exchange chromatography followed by reverse-phase tive-blinded-evaluation (PRoBE) design requirements chromatography. (11), reduces bias, and allows identification of proteins In-solution tryptic digestion was conducted with lyoph- that may have value for early detection and risk assessment. ilized aliquots from the reverse-phase (second dimension) Using samples from the Women’s Health Initiative (WHI) fractionation step. Aliquots were subjected to MS shotgun cohort, an intact protein analysis system (IPAS) approach analysis using an LTQ-Orbitrap (Thermo) mass spectrom- that allows quantitative analysis of proteins over 6 to 7 eter coupled with a NanoLC-1D (Eksigent). The acquired orders of magnitude of abundance (9, 12–14) was applied data were automatically processed by the Computational to plasmas from 90 participants who were subsequently Proteomics Analysis System (CPAS; ref. 16). For the iden- diagnosed with colon cancer within 18 months following tification of proteins with false-positive error rate less than blood draw and to 90 matched controls from the same 5%, liquid chromatography/tandem mass spectrometry cohort. Further testing of a protein subset was conducted in (LC/MS-MS) spectra of the samples were subjected to tryptic samples from Early Detection Research Network (EDRN) searches against the human IPI database (v.3.13) using X! collected at the time of diagnosis which included both male Tandem (17). Search results were then analyzed by Pepti- and female subjects. A panel established in the newly deProphet (18) and ProteinProphet (19) programs. Quan- diagnosed cohort was subsequently shown to have predic- titative protein analysis was based on differential labeling of tive value in an independent set of prediagnostic CRC cysteine residues with acrylamide isotopes. Peptide isotopic plasmas from the WHI cohort. ratios were plotted in logarithmic scale in a histogram, and the median of the distribution was centered at zero (Sup- Methods plementary Fig. S2). All normalized peptide ratios for a specific protein were averaged to compute an overall protein Study population ratio. Reported statistical significance of the protein quan- The sample population used in the discovery phase con- titative information was obtained using a one-sample Stu- sisted of plasmas from 90 women who were diagnosed with dent t test. False discovery rates were calculated on the basis CRC within 18 months following a blood draw that of the distribution of P values from permutations of disease occurred in the third year of participation in the WHI labels and the observed P values from the original data. Observational Study. These cases were individually CRC cell line proteomic analysis. HCT116 and SW480 matched on the basis of age ( 2 years), race/ethnicity, and were prepared according to the standard stable isotope baseline blood draw ( 2 months) to a randomly selected labeling with amino acids in cell culture (SILAC) protocol control without a history of cancer diagnosis (Table 1). as previously described (20). Secreted proteins were Plasma from 58 newly diagnosed male and female obtained directly from conditioned media after 48 hours patients with CRC and 58 matched controls were collected of culture. Total cell extract (TCE) was obtained by sonica- through the Community Clinical Oncology Program at the tion of about 2 107 cells followed by centrifugation at University of Michigan, Ann Arbor, MI, following informed 20,000 g. A surface-enriched fraction was obtained by consent. Cases were individually matched on the basis of biotinylating about 2 108 cells in culture plates. Proteins age (within 3 years) and gender. were extracted in a 2% NP-40 solution and subsequently An independent set of plasmas from 32 subjects in the isolated using NeutrAvidin. WHI Observational Study who were diagnosed with CRC Cell line preparations were fractioned by reverse-phase within 7 months following the third year blood draw and 32 chromatography. Reverse-phase fractions from each prep- matched controls was used for validation. Matching was aration were individually digested with trypsin and grouped done based on age ( 4 years), race/ethnicity, baseline into 23 to 27 pools on the basis of chromatographic blood draw ( 4 months), body mass index, hormone features. LC/MS-MS and protein identification were con- therapy use, and a negative history for cancer (Table 1). ducted as described above using v3.57 of the human IPI database. Proteomic analysis IPAS. Nine large-scale proteomic experiments were car- ELISA-based validation ried out on pools of plasmas from 10 cases and 10 controls Human IGFBP2 (R&D Systems), LRG1 (IBL-America), as previously described (refs. 12, 15; Supplementary Fig. CEA (Genway Biotech, Inc.), and MAPRE1 (USCN Life)

656 Cancer Prev Res; 5(4) April 2012 Cancer Prevention Research

Elevation of MAPRE1 Prior to Diagnosis of Colorectal Cancer

Table 1. Prediagnostic WHI cohort sample characteristics

Discovery set Validation set

Case Control Case Control

Number 90 90 32 32 Average age, y 64.9 64.9 68.1 67.8 Stage I 8 (9%) — 4 (13%) — Stage II 29 (32%) — 13 (41%) — Stage III 37 (41%) — 12 (38%) — Stage IV 16 (18%) — 3 (9%) — Caucasian 75 (83%) 75 (83%) 24 (75%) 24 (75%) African-American 10 (11%) 10 (11%) 7 (22%) 7 (22%) Asian 3 (3%) 3 (3%) Hispanic 1 (1%) 1 (1%) Other 1 (1%) 1 (3%) 1 (3%) Missing 1 (1%) Average days to diagnosis 245 — 109 —

measurements were conducted on newly diagnosed and protein IDs in the International Protein Index (IPI) data- prediagnostic plasma samples according to the manufac- base (21). Quantitative data based on isotopic ratios for case turer’s protocol. Absorbance was measured using a Spec- versus control was obtained for 1,779 proteins. An overall P traMax Plus 384 and results calculated with SoftMax Pro value and a geometric mean ratio for each protein across all v4.7.1 (Molecular Devices). P values were computed using a 9 experiments were calculated. Six proteins were signiﬁ- paired Mann–Whitney–Wilcoxon test on raw concentration cantly (P < 0.05) elevated in cases compared with controls values. ELISA measurements above and below the detection with a case-to-control ratio >1.2 (Table 2). Microtubule- limit for assays were imputed by the maximum and min- associated protein RP/EB family member 1 (MAPRE1) is a imum computable values for the assay. cytoplasmic protein that binds to adenomatous polyposis coli (APC), a commonly mutated gene in colorectal adenocarcinoma, and functions in mitotic processes (22). Results Leucine-rich alpha-2-glycoprotein (LRG1) is an extracellu- Proteomic analysis of plasma from study subjects and lar protein whose function is largely unknown with varied CRC cell lines expression levels in tissues (23). A role for LRG1 in gran- An in-depth quantitative MS analysis of WHI plasma ulocyte differentiation has been suggested (24). Insulin-like samples in 9 large-scale experiments yielded a total of growth factor–binding protein 2 (IGFBP2) is an extracellu- 1,992,567 mass spectra, resulting in a total of 5,022 unique lar protein that binds IGF2 and has been shown to

Table 2. Proteins with signiﬁcantly elevated levels in plasmas collected before diagnosis of CRC compared with matched controls from the WHI cohort that did not develop CRC during the period of follow-up

Average case– Gene Description control ratio P FDR

MAPRE1 Microtubule-associated protein 4.52 0.019 0.763 RP/EB family member 1 PDIA3 Protein disulﬁde-isomerase a3 1.51 0.027 0.471 IGFBP2 Insulin-like growth factor–binding 1.20 0.027 0.420 protein 2 ENO1 Alpha-enolase 1.96 0.029 0.713 ARMET Mesencephalic astrocyte-derived 1.64 0.043 0.417 neurotrophic factor LRG1 Leucine-rich alpha-2-glycoprotein 1.35 0.046 0.487

Abbreviation: FDR, false discovery rate.

www.aacrjournals.org Cancer Prev Res; 5(4) April 2012 657

Ladd et al.

Figure 1. Tryptic peptides identiﬁed from MS experiments for IPAS experiments for (A) MAPRE1, (B) LRG1, (C) IGFBP2.

potentially have both proliferative and antiproliferative omic analysis of 2 CRC cell lines with different driver roles in cancer (25). Enolase 1 has been identified as a mutations was conducted using SILAC (20). HCT116 and central element in a disease-specific gene network in colon SW480 were analyzed to assess potential differences in cancer (26). Mesencephalic astrocyte-derived neurotrophic protein expression based on APC mutational status. factor (ARMET) and protein disulfide-isomerase A3 MAPRE1, IGFBP2, alpha-enolase (ENO1), PDIA3, and (PDIA3) belong to a family of endoplasmic reticulum ARMET were identified in both HCT116 and the APC- stress–induced proteins which have been found to be upre- mutant SW480 (Supplementary Fig. S3a–S3e), whereas gulated in gastric and hepatocellular carcinomas (27, 28). LRG1 was not identified in either of the 2 cell lines. Mass spectrometric analysis yielded substantial peptide MAPRE1,ENO1,PDIA3,andARMETwereobservedin coverage for all 6 proteins (Fig. 1A–F), indicating a robust TCE, the media, and surface-enriched fractions in both identification of each full-length protein in human plasma. cell lines. The APC-binding domain of MAPRE1 was To determine whether the identified proteins may have enriched in the media and cell surface fractions. PDIA3 originated from tumor cells or from a host response, prote- was enriched in the cell surface compartment with fewer

658 Cancer Prev Res; 5(4) April 2012 Cancer Prevention Research

Elevation of MAPRE1 Prior to Diagnosis of Colorectal Cancer

Figure 1. (Continued ) (D) ENO1, (E) ARMET, and (F) PDIA3. Each line represents peptides identified or quantified in an individual IPAS. The top line in each figure represents theoretical peptides generated by tryptic digestion. Dark grey peptides indicate non-cysteine–containing tryptic peptides, whereas striped peptides are cysteine-containing tryptic peptides that could be quantified from the acrylamide labeling. In the experimental data, dark grey peptides indicate the peptide was identified, but not quantified, whereas striped peptides indicate the peptide had quantification data from the acrylamide labeling.

peptides identified in the conditioned media. ARMET was Assays of IGFBP2, LRG1, and MAPRE1 in plasmas from also enriched on the cell surface of the SW480 cell line but newly diagnosed CRC cases not in HCT116. ENO1 was the most identified protein in Three of these 6 proteins (MAPRE1, LRG1, and IGFBP2) both TCE and conditioned media. IGFBP2 was predom- had ELISA assays available for further validation studies. inantly observed in the conditioned media, with few IGFBP2, LRG1, and MAPRE1 along with CEA were assayed peptides identified in the TCE. These cell findings suggest in plasma from newly diagnosed CRC subjects and controls that tumor cells may contribute to increased levels (Fig. 2). Given that the discovery studies were based on observed in plasma for MAPRE1, IGFBP2, ENO1, PDIA3, pools of cases and controls, ELISA assays of individual and ARMET. samples were relied upon to develop a combination rule

www.aacrjournals.org Cancer Prev Res; 5(4) April 2012 659

Ladd et al.

Figure 2. ELISA results in newly diagnosed samples of (A) CEA, (B) MAPRE1, (C) LRG1, and (D) IGFBP2. Each circle represents one of 58 individual cases or matched controls. , P < 0.0001 based on paired Wilcoxon test.

for validation of the marker panel in an independent set of of ELISA responses showed that levels of MAPRE1 and CEA prediagnostic samples. All 4 of the assayed proteins were correlated well (Supplementary Fig. S4). Similarly, IGFBP2 found to be signiﬁcantly (P < 0.05) elevated by more than and LRG1 were also highly correlated, whereas MAPRE1 1.5-fold in cases compared with controls (Table 3) in a set of and LRG1 exhibited an orthogonal relationship. 58 newly diagnosed CRC cases and 58 age-matched controls. Area under the curve values (AUC) for IGFBP2, LRG1, Assays of individual markers in an independent set of MAPRE1, and CEA ranged from 0.712 to 0.782 (Table 3). prediagnostic plasmas Linear regression analyses based on maximum likelihood The linear combination of CEA, IGFBP2, LRG1, and estimation of raw ELISA values were conducted on all MAPRE1 that was constructed on the basis of the newly possible combinations of the 4 markers. A combination of diagnosed samples was evaluated in an independent set of all 4 markers (denoted "Panel") was found to have the prediagnostic WHI plasma samples consisting of 32 CRC highest AUC of 0.841 with 59% sensitivity at 95% speci- cases and 32 matched controls drawn within 7 months ﬁcity, a 23% increase over CEA alone (Fig. 3A). Scatter plots before the diagnosis of CRC. This combination rule resulted

660 Cancer Prev Res; 5(4) April 2012 Cancer Prevention Research

Elevation of MAPRE1 Prior to Diagnosis of Colorectal Cancer

Table 3. ELISA summary for individual proteins in newly diagnosed and prediagnostic CRC samples compared with matched healthy controls

Newly diagnosed samples Prediagnostic samples

Ratio P AUC Ratio P AUC

CEA 45.76 1.9E-06 0.782 (0.706–0.857) 8.42 0.0261 0.602 (0.463–0.740) IGFBP2 1.66 6.2E-05 0.717 (0.624–0.811) 1.27 0.0984 0.586 (0.444–0.728) LRG1 1.60 5.0E-05 0.712 (0.615–0.808) 1.40 8.2E-05 0.723 (0.594–0.851) MAPRE1 10.79 3.9E-07 0.778 (0.691–0.864) 5.30 0.0066 0.701 (0.570–0.831)

NOTE: Newly diagnosed samples are both male and female. Prediagnostic WHI samples are all female. in an AUC of 0.724, with 41% sensitivity at 95% specificity and an equal number of matched controls yielded a set of 6 (Fig. 3B), compared with 19% sensitivity at 95% specificity proteins that were significantly upregulated in cases com- for CEA alone. pared with controls. Three of these proteins, LRG1, IGFBP2, Furthermore, CEA, LRG1, and MAPRE1 were each signif- and ARMET, are known to be secreted whereas MAPRE1, icantly elevated in cases compared with controls (Table 3). PDIA3, and ENO1 are predominantly intracellular. IGFBP2, IGFBP2 was not significantly elevated in the prediagnostic LRG1, and MAPRE1 were selected for further characteriza- samples, with a mean ratio of 1.27 and P < 0.1. Individual tion and validation on the basis of the availability of ELISA markers had AUCs between 0.586 (IGFBP2) and 0.723 assays. Immunologic testing of these 3 proteins along with (LRG1; Table 3). Ratios for each marker were lower in the CEA in plasmas from newly diagnosed subjects showed prediagnostic samples than in the newly diagnosed group. significant (P < 0.05) elevation of each in cases compared with controls. A linear combination of the 4 proteins yielded 59% sensitivity at 95% specificity for plasmas from Discussion newly diagnosed cases relative to controls indicative of the The proteomic analysis of 9 pools from 180 plasma potential of the marker panel for improved monitoring of samples from the WHI cohort collected before diagnosis CRC. Addition of the 3 markers to CEA also improved

Figure 3. A, receiver operating characteristics curve (ROC) analysis of a linear combination of the 4 markers compared with CEA in newly diagnosed samples. B, ROC analysis of the same linear combination of the 4 markers compared with CEA in prediagnostic samples. Coefﬁcients for combination of the 4 markers are: CEA, 3.612e-2; IGFBP2, 7.052e-3; LRG1, 7.263e-5; and MAPRE1, 2.766e-2.

www.aacrjournals.org Cancer Prev Res; 5(4) April 2012 661

Ladd et al.

performance in prediagnostic samples. Sensitivity was expression data from BioGPS indicated that MAPRE1 was increased from 19% for CEA alone to 41% at 95% specificity strongly expressed in colorectal adenocarcinoma compared for the panel in blood drawn within 7 months before the with most other tissues, including normal colon (37). diagnosis of CRC. In addition, CEA, LRG1, and MAPRE1 Immunohistochemistry for MAPRE1 in colorectal tumor were each significantly elevated in the prediagnostic plas- tissues from Human Protein Atlas (38) shows an increase in mas. IGFBP2 yielded a case-to-control ratio of 1.27 before cytoplasmic staining compared with normal tissues. Our diagnosis but was not statistically significant. Prediagnostic study has revealed for the first time an association between samples separated by stage showed increased levels of CEA circulating levels of MAPRE1 and CRC. and MAPRE1 in stage III/IV cases compared with stage I/II Elevated plasma levels of LRG1 have previously been cases (P ¼ 0.068 and 0.120, respectively; Supplementary reported for pancreatic and ovarian cancers (39–41) but Fig. S5). For both proteins, only stage III/IV cases not for CRC. LRG1 is primarily expressed in the liver (37) were significantly higher than controls. The elevation of and has been associated with acute-phase response, being LRG1 in cases compared to controls was more statistically induced by proinflammatory cytokines, such as interleukin significant in stage III/IV cases than in stage I/II cases. 6 (IL-6; ref. 23). LRG1 was not observed in proteomic Our findings suggest that circulating plasma levels of analysis of conditioned media from 2 CRC cell lines, sug- CEA, LRG1, and MAPRE1 may all increase with tumor gesting that increased circulating levels are a response to progression. tumor development. Elevated circulating levels have previ- Extensive mass spectrometric evaluations of IGFBP2, ously been associated with GVHD (14), as well as autoim- LRG1, and MAPRE1 in other cancers and inflammatory mune diseases (42). LRG1 may be released from neutro- diseases have been carried out by our group. Protein levels phils (24, 43). LRG1 has also been associated with TGFb were on average unchanged across multiple experiments in signaling (44), specifically through interaction with TGFb both breast and lung cancer for each of the 3 proteins. In receptor type II (45). patients who developed coronary heart disease, MAPRE1 CEA has long been established as a marker for CRC (46). and IGFBP2 were decreased or unchanged in diseased It is a member of the immunoglobulin superfamily and has individuals compared with matched controls whereas LRG1 been associated with cancer dissemination (47). Because of was not quantified. CEA was not quantified in any of the its low sensitivity and specificity, CEA has limited use in mass spectrometric experiments likely due to its high degree screening or diagnosis of early-stage CRC and has primarily of glycosylation. been assayed to determine preoperative prognosis and for A comprehensive proteomic analysis of an Apc D580 disease monitoring (6, 48). Plasma levels of CEA are mouse model was previously conducted by our group reduced following surgical removal of cancerous polyps (29). From that analysis, it was observed that circulating (49–51). Our data suggest that CEA may have use for early levels of both LRG1 and IGFBP2 were significantly (P < detection of CRC as part of a panel of markers. 0.05) elevated in tumor-bearing mice compared with con- IGFBP2 has been previously investigated as a potential trols. ENO1 and PDIA3 were also identified in the analysis plasma marker for CRC and other cancers (52–54) with of mouse plasma samples based on non-cysteine–contain- mostly negative findings (55, 56). In this study, plasma ing peptides, thus lacking quantification, whereas no pep- levels were significantly elevated in newly diagnosed tides from MAPRE1 or ARMET were identified, likely due to patients, but not in preclinical samples, suggesting that their very low abundance in plasma. circulating levels of IGFBP2 increase with progressive tumor Mutation of the APC gene is considered to be one of the development (57). Given the occurrence of IGFBP2 in the initiating events in the development of colorectal adeno- conditioned media of CRC cell lines, it is likely that an carcinoma (30). The mutated form of APC is commonly increase in tumor cell mass contributes to the observed truncated, retaining only the N-terminus, resulting in increase in plasma levels with advanced CRC. increased protein mobility and altered function (31). ENO1, PDIA3, and ARMET have all previously been MAPRE1 is known to bind to APC and participate in the investigated in various cancers but only ENO1 has been stabilization of microtubules through interactions with the associated with colorectal tumor development (26–28). formin mDia (32). Overexpression of MAPRE1 has been Levels of PDIA3 in human plasma were found to be elevated found to induce nuclear accumulation of b-catenin and in hepatocellular carcinoma on the basis of an immunoas- activate the b-catenin/T-cell factor pathway leading to a say (27) and in gastric cancer on the basis of proteomic mass promotion of cell growth and increase in colony formation spectrometric analysis (28). Our findings suggest that these (33, 34). Our study shows a significant elevation of circu- 3 proteins may be elevated in the plasma of patients with lating MAPRE1 protein in newly diagnosed and prediag- CRC before the clinical diagnosis of the disease. These nostic CRC plasma samples. Expression of MAPRE1 has markers may further improve the performance of the 4- been reported to be elevated in tissue from head and neck marker panel presented in this study. cancer (35) and to be correlated with tumor size and Multiple steps were used to facilitate the in-depth, quan- associated with poor differentiation in hepatocellular car- titative plasma proteomic profiling in this study: depletion cinoma tissue (36). Extensive proteomic analysis of 2 CRC of the 6 most highly abundant plasma proteins, extensive cell lines, as well as Western blot analysis, resulted in the protein fractionation using reverse-phase and anion identification of MAPRE1 in conditioned media. Gene exchange chromatography, and the use of heavy and light

662 Cancer Prev Res; 5(4) April 2012 Cancer Prevention Research

Elevation of MAPRE1 Prior to Diagnosis of Colorectal Cancer

acrylamide labels for comparison of cases and controls. (10). Because discovery of markers was also done in preclin- Using these steps, proteins across 7 orders of magnitude and ical samples, phase I and II were not applicable, which is an with concentrations in the picogram per milliliter range advantage of this study. The primary aims of phase III, to have been identified. MS-based discovery of this nature, evaluate the capacity of the biomarkers to detect preclinical while quantitative, does not recognize posttranslational disease and to define criteria for a positive screening test, were modifications, such as glycosylation that may be cancer- addressed. CEA, MAPRE1, and LRG1 were shown to signif- related (58). icantly differentiate preclinical cases from matched controls, Most prior discovery studies of blood-based biomarkers whereas IGFBP2 was elevated in preclinical cases, but not for early detection have been based on analysis of specimens significantly so. Furthermore, a linear combination of these 4 collected at the time of diagnosis. In contrast, our study markers was established that differentiates preclinical cases relied on plasma samples collected before the clinical man- from controls with 41% sensitivity at 95% specificity. ifestation of CRC to identify and validate a panel of markers Mass spectrometric analysis of preclinical CRC compared that could be useful for early detection and identification of with matched controls yielded a set of elevated proteins that subjects at increased risk of developing CRC. The WHI were further validated by ELISA. Three of these proteins in cohort samples used in discovery and validation sets consist conjunction with CEA show promise as a preclinical test for entirely of postmenopausal women and may not be repre- CRC. Further improvements in sensitivity and specificity sentative of the general population as a whole. Hormone based on inclusion of additional markers may ultimately therapy use, which may alter the circulating levels of some lead to a blood-based test to aid in screening for CRC. proteins, was not a factor used in matching case and control samples in this study and could have impacted the plasma Disclosure of Potential Conflicts of Interest proteome. However, there was no bias in this regard No potential conflicts of interests were disclosed. between cases and controls of which we are aware. How- The costs of publication of this article were defrayed in part by the ever, validation data from newly diagnosed patients suggest payment of page charges. This article must therefore be hereby marked that levels of the assayed markers are not confounded by advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate gender (Supplementary Fig. S6) or hormone therapy. this fact. The WHI cohort meets the requirements of phase III of Received August 29, 2011; revised January 9, 2012; accepted January 11, biomarker development as outlined by Pepe and colleagues 2012; published OnlineFirst January 25, 2012.

References 1. Hawk ET, Levin B. Colorectal cancer prevention. J Clin Oncol 12. Faca V, Coram M, Phanstiel D, Glukhova V, Zhang Q, Fitzgibbon M, 2005;23:378–91. et al. Quantitative analysis of acrylamide labeled serum proteins by LC- 2. Singh H, Turner D, Xue L, Targownik LE, Bernstein CN. Risk of MS/MS. J Proteome Res 2006;5:2009–18. developing colorectal cancer following a negative colonoscopy exam- 13. Misek DE, Kuick R, Wang H, Galchev V, Deng B, Zhao R, et al. A wide ination - Evidence for a 10-year interval between colonoscopies. JAMA range of protein isoforms in serum and plasma uncovered by a 2006;295:2366–73. quantitative intact protein analysis system. Proteomics 2005;5: 3. Seeff LC, Manninen DL, Dong FB, Chattopadhyay SK, Nadel MR, 3343–52. Tangka FKL, et al. Is there endoscopic capacity to provide colorectal 14. Wang H, Clouthier SG, Galchev V, Misek DE, Duffner U, Min CK, et al. cancer screening to the unscreened population in the United States? Intact-protein-based high-resolution three-dimensional quantitative Gastroenterology 2004;127:1661–9. analysis system for proteome profiling of biological fluids. Mol Cell 4. Wee CC, McCarthy EP, Phillips RS. Factors associated with colon Proteomics 2005;4:618–25. cancer screening: the role of patient factors and physician counseling. 15. Wang H, Hanash S. Intact-protein analysis system for discovery of Prev Med 2005;41:23–9. serum-based disease biomarkers. In: Simpson RJ, Greening DW, 5. Denberg TD, Melhado TV, Coombes JM, Beaty BL, Berman K, Byers editors. Serum/plasma proteomics: methods and protocols. New TE, et al. Predictors of nonadherence to screening colonoscopy. J Gen York: Humana Press; 2011. p. 69–85. Intern Med 2005;20:989–95. 16. Rauch A, Bellew M, Eng J, Fitzgibbon M, Holzman T, Hussey P, et al. 6. Duffy MJ. Carcinoembryonic antigen as a marker for colorectal cancer: Computational Proteomics Analysis System (CPAS): an extensible, is it clinically useful? Clin Chem 2001;47:624–30. open-source analytic system for evaluating and publishing proteomic 7. Duffy MJ, van Dalen A, Haglund C, Hansson L, Holinski-Feder E, data and high throughput biological experiments. J Proteome Res Klapdor R, et al. Tumour markers in colorectal cancer: European Group 2006;5:112–21. on Tumour Markers (EGTM) guidelines for clinical use. Eur J Cancer 17. MacLean B, Eng JK, Beavis RC, McIntosh M. General framework for 2007;43:1348–60. developing and evaluating database scoring algorithms using the 8. Ludwig JA, Weinstein JN. Biomarkers in cancer staging, prognosis and TANDEM search engine. Bioinformatics 2006;22:2830–2. treatment selection. Nat Rev Cancer 2005;5:845–56. 18. Keller A, Nesvizhskii AI, Kolker E, Aebersold R. Empirical statistical 9. Faca V, Pitteri SJ, Newcomb L, Glukhova V, Phanstiel D, Krasnoselsky model to estimate the accuracy of peptide identifications made by MS/ A, et al. Contribution of protein fractionation to depth of analysis of the MS and database search. Anal Chem 2002;74:5383–92. serum and plasma proteomes. J Proteome Res 2007;6:3558–65. 19. Nesvizhskii AI, Keller A, Kolker E, Aebersold R. A statistical model for 10. Pepe MS, Etzioni R, Feng ZD, Potter JD, Thompson ML, Thornquist M, identifying proteins by tandem mass spectrometry. Anal Chem et al. Phases of biomarker development for early detection of cancer. J 2003;75:4646–58. Natl Cancer Inst 2001;93:1054–61. 20. Faca VM, Ventura AP, Fitzgibbon MP, Pereira-Faca SR, Pitteri SJ, 11. Pepe MS, Feng ZD, Janes H, Bossuyt PM, Potter JD. Pivotal evaluation Green AE, et al. Proteomic analysis of ovarian cancer cells reveals of the accuracy of a biomarker used for classification or prediction: dynamic processes of protein secretion and shedding of extra-cellular standards for study design. J Natl Cancer Inst 2008;100:1432–8. domains. PLoS One 2008;3:e2425.

www.aacrjournals.org Cancer Prev Res; 5(4) April 2012 663

Ladd et al.

21. Kersey PJ, Duarte J, Williams A, Karavidopoulou Y, Birney E, Apweiler (2D-DIGE): up-regulation of leucine-rich alpha-2-glycoprotein in pan- R. The International Protein Index: an integrated database for prote- creatic cancer. J Chromatogr B Analyt Technol Biomed Life Sci omics experiments. Proteomics 2004;4:1985–8. 2007;852:257–67. 22. Green RA, Wollman R, Kaplan KB. APC and EB1 function together in 41. Okano T, Kondo T, Kakisaka T, Fujii K, Yamada M, Kato H, et al. Plasma mitosis to regulate spindle dynamics and chromosome alignment. Mol proteomics of lung cancer by a linkage of multi-dimensional liquid Biol Cell 2005;16:4609–22. chromatography and two-dimensional difference gel electrophoresis. 23. Shirai R, Hirano F, Ohkura N, Ikeda K, Inoue S. Up-regulation of the Proteomics 2006;6:3938–48. expression of leucine-rich alpha(2)-glycoprotein in hepatocytes by the 42. Serada S, Fujimoto M, Ogata A, Terabe F, Hirano T, Iijima H, et al. mediators of acute-phase response. Biochem Biophys Res Commun iTRAQ-based proteomic identification of leucine-rich alpha-2 glyco- 2009;382:776–9. protein as a novel inflammatory biomarker in autoimmune diseases. 24. O'Donnell LC, Druhan LJ, Avalos BR. Molecular characterization Ann Rheum Dis 2010;69:770–4. and expression analysis of leucine-rich alpha 2-glycoprotein, a 43. Hofman PM. Pathobiology of the neutrophil-intestinal epithelial cell novel marker of granulocytic differentiation. J Leukoc Biol 2002; interaction: role in carcinogenesis. World J Gastroenterol 2010;16: 72:478–85. 5790–800. 25. Wolf E, Lahm H, Wu MY, Wanke R, Hoeflich A. Effects of IGFBP-2 44. Sun DT, Kar S, Carr BI. Differentially expressed genes in TGF-beta 1 overexpression in vitro and in vivo. Pediatr Nephrol 2000;14:572–8. sensitive and resistant human hepatoma-cells. Cancer Lett 1995;89: 26. Jiang W, Li X, Rao SQ, Wang LH, Du L, Li CX, et al. Constructing 73–9. disease-specific gene networks using pair-wise relevance metric: 45. Munoz NM, Upton M, Rojas A, Washington MK, Lin L, Chytil A, et al. application to colon cancer identifies interleukin 8, desmin and enolase Transforming growth factor beta receptor type II inactivation induces 1 as the central elements. BMC Syst Biol 2008;2:72. the malignant transformation of intestinal neoplasms initiated by Apc 27. Chignard N, Shang SF, Wang H, Marrero J, Brechot C, Hanash S, et al. mutation. Cancer Res 2006;66:9837–44. Cleavage of endoplasmic reticulum proteins in hepatocellular carci- 46. Ballesta AM, Molina R, Filella X, Jo J, Gimenez N. Carcinoembryonic noma: detection of generated fragments in patient sera. Gastroenter- antigen in staging and follow-up of patients with solid tumors. Tumor ology 2006;130:2010–22. Biol 1995;16:32–41. 28. Ren H, Du N, Liu G, Hu HT, Tian W, Deng ZP, et al. Analysis of 47. Hostetter RB, Augustus LB, Mankarious R, Chi KF, Fan D, Toth C, et al. variabilities of serum proteomic spectra in patients with gastric cancer Carcinoembryonic antigen as a selective enhancer of colorectal-can- before and after operation. World J Gastroenterol 2006;12:2789–92. cer metastasis. J Natl Cancer Inst 1990;82:380–5. 29. Hung KE, Faca V, Song K, Sarracino DA, Richard LG, Krastins B, et al. 48. Fletcher RH. Carcinoembryonic antigen. Ann Internal Med 1986;104: Comprehensive proteome analysis of an Apc mouse model uncovers 66–73. proteins associated with intestinal tumorigenesis. Cancer Prev Res 49. McCall JL, Black RB, Rich CA, Harvey JR, Baker RA, Watts JM, et al. (Phila) 2009;2:224–33. The value of serum carcinoembryonic antigen in predicting recurrent 30. Fodde R, Smits R, Clevers H. APC, signal transduction and genetic disease following curative resection of colorectal-cancer. Dis Colon instability in colorectal cancer. Nat Rev Cancer 2001;1:55–67. Rectum 1994;37:875–81. 31. Brocardo M, Henderson BR. APC shuttling to the membrane, nucleus 50. Wang JY, Tang RP, Chiang JM. Value of carcinoembryonic antigen in and beyond. Trends Cell Biol 2008;18:587–96. the management of colorectal-cancer. Dis Colon Rectum 1994;37: 32. Wen Y, Eng CH, Schmoranzer J, Cabrera-Poch N, Morris EJS, Chen M, 272–7. et al. EB1 and APC bind to mDia to stabilize microtubules downstream 51. Bannura G, Cumsille MA, Contreras J, Barrera A, Melo C, Soto D. of Rho and promote cell migration. Nat Cell Biol 2004;6:820–30. [Carcinoembryonic antigen (CEA) as an independent prognostic 33. Liu M, Yang SB, Wang YH, Zhu HX, Yan S, Zhang W, et al. EB1 acts as factor in colorectal carcinoma]. Rev Med Chile 2004;132: an oncogene via activating beta-catenin/TCF pathway to promote 691–700. cellular growth and inhibit apoptosis. Mol Carcinog 2009;48:212–9. 52. Cohen P, Peehl DM, Stamey TA, Wilson KF, Clemmons DR, Rosenfeld 34. Wang YH, Zhou XB, Zhu HX, Liu S, Zhou CQ, Zhang G, et al. Over- RG. Elevated levels of insulin-like growth factor-binding protein-2 in expression of EB1 in human esophageal squamous cell carcinoma the serum of prostate-cancer patients. J Clin Endocrinol Metab (ESCC) may promote cellular growth by activating beta-catenin/TCF 1993;76:1031–5. pathway. Oncogene 2005;24:6637–45. 53. Flyvbjerg A, Mogensen O, Mogensen B, Nielsen OS. Elevated serum 35. Ralhan R, Desouza LV, Matta A, Tripathi SC, Ghanny S, Gupta SD, et al. insulin-like growth factor-binding protein 2 (IGFBP-2) and decreased Discovery and verification of head-and-neck cancer biomarkers by IGFBP-3 in epithelial ovarian cancer: correlation with cancer antigen differential protein expression analysis using iTRAQ labeling, multidi- 125 and tumor-associated trypsin inhibitor. J Clin Endocrinol Metab mensional liquid chromatography, and tandem mass spectrometry. 1997;82:2308–13. Mol Cell Proteomics 2008;7:1162–73. 54. Liou JM, Shun CT, Liang JT, Chiu HM, Chen MJ, Chen CC, et al. 36. Orimo T, Ojima H, Hiraoka N, Saito S, Kosuge T, Kakisaka T, et al. Plasma insulin-like growth factor-binding protein-2 levels as diagnos- Proteomic profiling reveals the prognostic value of adenomatous tic and prognostic biomarker of colorectal cancer. J Clin Endocrinol polyposis coli-end-binding protein 1 in hepatocellular carcinoma. Metab 2010;95:1717–25. Hepatology 2008;48:1851–63. 55. Jenab M, Riboli E, Cleveland RJ, Norat T, Rinaldi S, Nieters A, et al. 37. Wu C, Orozco C, Boyer J, Leglise M, Goodale J, Batalov S, et al. Serum C-peptide, IGFBP-1 and IGFBP-2 and risk of colon and rectal BioGPS: an extensible and customizable portal for querying and cancers in the European Prospective Investigation into Cancer and organizing gene annotation resources. Genome Biol 2009;10: Nutrition. Int J Cancer 2007;121:368–76. R130. 56. Kaaks R, Lundin E, Manjer J, Rinaldi S, Biessy C, Soderberg S, et al. 38. Berglund L, Bjorling€ E, Oksvold P, Fagerberg L, Asplund A, Prospective study of IGF-I, IGF-binding proteins, and breast cancer Szigyarto C, et al. A genecentric Human Protein Atlas for expres- risk, in Northern and Southern Sweden. Cancer Causes Control sion profiles based on antibodies. Mol Cell Proteomics 2008;7: 2002;13:307–16. 2019–27. 57. Diehl D, Hessel E, Oesterle D, Renner-Muller I, Elmlinger M, Langham- 39. Heo SH, Lee SJ, Ryoo HM, Park JY, Cho JY. Identification of putative mer M, et al. IGFBP-2 overexpression reduces the appearance of serum glycoprotein biomarkers for human lung adenocarcinoma by dysplastic aberrant crypt foci and inhibits growth of adenomas in multilectin affinity chromatography and LC-MS/MS. Proteomics chemically induced colorectal carcinogenesis. Int J Cancer 2009;124: 2007;7:4292–302. 2220–5. 40. Kakisaka T, Kondo T, Okano T, Fujii K, Honda K, Endo M, et al. Plasma 58. Zhang Q, Faca V, Hanash S. Mining the plasma proteome for disease proteomics of pancreatic cancer patients by multi-dimensional liquid applications across seven logs of protein abundance. J Proteome Res chromatography and two-dimensional difference gel electrophoresis 2011;10:46–50.

664 Cancer Prev Res; 5(4) April 2012 Cancer Prevention Research

Increased Plasma Levels of the APC-Interacting Protein MAPRE1, LRG1, and IGFBP2 Preceding a Diagnosis of Colorectal Cancer in Women

Jon J. Ladd, Tina Busald, Melissa M. Johnson, et al.

Cancer Prev Res 2012;5:655-664. Published OnlineFirst January 25, 2012.

Updated version Access the most recent version of this article at: doi:10.1158/1940-6207.CAPR-11-0412

Supplementary Access the most recent supplemental material at: Material http://cancerpreventionresearch.aacrjournals.org/content/suppl/2012/03/28/1940-6207.CAPR-11-0412.DC 1

Cited articles This article cites 57 articles, 9 of which you can access for free at: http://cancerpreventionresearch.aacrjournals.org/content/5/4/655.full#ref-list-1

Citing articles This article has been cited by 9 HighWire-hosted articles. Access the articles at: http://cancerpreventionresearch.aacrjournals.org/content/5/4/655.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerpreventionresearch.aacrjournals.org/content/5/4/655. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.