|Home | About | Journals | Submit | Contact Us | Français|
The aims of this study were to characterize the proteome of normal pancreatic juice, to analyze the effect of secretin on the normal proteome, and to compare these results with published data from patients with pancreatic cancer.
Paired pancreatic fluid specimens (before and after intravenous secretin stimulation) were obtained during endoscopic pancreatography from three patients without significant pancreatic pathology. Proteins were identified and quantified by mass spectrometry-based protein quantification technology. The human RefSeq (NCBI) database was used to compare the data in normal patient samples with published data from three pancreatic cancer patients.
A total of 285 proteins were identified in normal pancreatic juice. Ninety had sufficient amino acid sequences identified to characterize the protein with a high level of confidence. All 90 proteins were present before and after secretin administration but with altered relative concentrations, usually by 1-2 folds, after stimulation. Comparison with 170 published pancreatic cancer proteins yielded an overlap of only 42 proteins.
Normal pancreatic juice contains multiple proteins related to many biological processes. Secretin alters the concentration but not the spectrum of these proteins. The pancreatic juice proteome of normal and pancreatic cancer patients differ markedly.
The proteome is the entire complement of proteins expressed by a genome, cell, tissue or organism. In eukaryotic organisms, the proteome is significantly larger and more complex than the genome secondary to such biological modification processes as gene splicing and post-transcriptional modification.1 Proteomics, the study of the proteome, may prove to be an invaluable tool in the field of oncologic research. Through methods such as gel electrophoresis and mass spectrometry, the proteome of neoplastic tissues and body fluids can be analyzed, leading to the identification of disease specific biomarkers useful in detection, staging, treatment and surveillance of several types of malignancies.2-23
The proteome of human pancreatic juice has been characterized in patients with pancreatic diseases, including pancreatic cancer.24-29 However, to date, the proteome of pancreatic juice from non-diseased humans has not been reported. In addition, the influence of secretin stimulation on the human pancreatic fluid proteome is unknown. Secretin is frequently administered during endoscopic retrograde cholangiopancreatography (ERCP) to increase the yield of pancreas fluid for collection. Thus, an understanding of the effects of secretin on the proteome should be obtained prior to subsequent analysis of such samples for pertinent biomarkers of disease states. Therefore, the aims of this study were 1) to characterize the proteome of normal pancreatic fluid, 2) to analyze the effect of secretin stimulation on the normal proteome and 3) to compare these results with published data from patients with pancreatic cancer.
These studies have been conducted in strict compliance with the Indiana University School of Medicine Institutional Review Board
Three female patients, ages 29-32, underwent ERCP for clinical abdominal symptoms, but no apparent pancreatic pathology was present following diagnostic investigation. Each patient prospectively signed informed consent for collection of pancreatic ductal fluid. Paired fluid specimens were obtained from each patient before and after administration of secretin (one ampule, intravenous administration) to increase the yield of pancreatic fluid. Specimens were placed immediately on ice after procurement and aliquoted for storage at -80°C without protease inhibitors within the Indiana University Pancreas Tissue Fluid Bank. Samples were subsequently analyzed via proteomic methods described below.
One hundred microliters of each pancreatic juice sample was denatured by a buffer containing 8 M urea and 10 mM dithiothreitol (DTT). In order to take the same amount of proteins from each sample for each analysis, protein concentrations were determined by Bradford assay. The same buffer is used as the background reference for protein assay in order to obtain a relatively accurate measurement among all samples (due to the presence of urea in lysis buffer). Resulting protein extracts are subsequently reduced and alkylated with triethylphosphine and iodoethanol to block sulfhydryl groups in proteins. The volatile reagents triethylphosphine and iodoethanol were used to minimize sample preparation variations.30 Protein mixtures were then digested by trypsin, and tryptic digests were filtered with 0.45 μm spin filters before applied to HPLC to avoid column clogging.
All digested samples were randomized for injection order to remove systematic bias from data acquisition. Twenty μg of the tryptic peptides from each sample were injected onto a C18 microbore column (i.d. = 1 mm, length = 5 cm. pore size = 300Å). Peptides were eluted with a linear gradient from 5 to 45% acetonitrile developed over 120 min at a flow rate of 50 μL/min, and effluent was electro-sprayed into a linear trap quadrupole (LTQ) mass spectrometer (ThermoFisher Scientific). The mass spectrometry (MS) data were collected in the data-dependent “Triple-Play” mode (MS scan, Zoom scan, and MS/MS scan). These three important experimental parameters plus chromatographic retention time determine the analytical accuracy of protein identification and protein quantification by ion intensity.31
In this study, both SEQUEST and X! Tandem database search algorithms were used for peptide sequence identification.32-34 Each algorithm compares the observed peptide MS/MS spectra and theoretically derived spectra from the database to assign quality scores. These quality scores and other important predictors are combined in the algorithm that assigns an overall score to each peptide.
Identified proteins were classified according to identification (ID) quality (priority). The confidence in protein identification is increased with an increasing number of distinct amino acid sequences identified and increasing peptide ID confidence. Priority 1 proteins have the greatest likelihood of correct identification (multiple unique sequences identified) and priority 4 proteins have the least likelihood of correct identification (but still with <25% false discovery). The “peptide ID confidence” (the quality of the amino acid sequence identification) of the “best peptide” (the peptide with the highest peptide ID confidence) is used to assign the protein to a “high” (between 90% and 100% confidence), “moderate” (between 75% and 89% confidence), or “low” (less than 75% confidence) category. All “low” category proteins < 75% are eliminated from further consideration. The biologic function of priority 1 and 2 proteins was determined by the Panther Classification System.35, 36 The human RefSeq (NCBI) database was used to compare the data in normal patient samples with published data from three pancreatic cancer patients.
The key in our quantitative analysis is the chromatographic peak alignment.31 Due to the fact that large biomarker studies can produce chromatographic shifts as a result of multiple injections of the samples onto the same HPLC column, this peak alignment is critical in order to provide the most accurate comparative data. Un-aligned peak comparison will result in larger variability and inaccuracy in peptide quantification.31 To be qualified for the protein quantification procedure, each aligned peak must match precursor ion (MS data), charge state (zoom scan data), fragment ions (MS/MS data) and retention time (within one-minute window). After alignment, the area-under-the-curves (AUC) for individually aligned peaks from identified peptides from each sample were computed; the AUCs were then compared for relative protein abundance. The data normalization was carried out by Quantile Normalization method.37 Quantile Normalization is a method of normalization that essentially ensures that every sample has a peptide intensity histogram of the same scale, location and shape. This normalization procedure removes trends introduced by sample handling, sample preparation, total protein differences and changes in instrument sensitivity while running multiple samples.
The number of significant changes between groups, the fold changes (FC) and the variability (Coefficient-of-Variation or CV) can be summarized for each priority level. The threshold for significance is set to control the False-Discovery-Rate (FDR) for each comparison at an investigator-desired percentile, normally 5%.38 The FDR is estimated by the q-value which is an adjusted p-value. The FDR is the proportion of significant changes that are false positives. If proteins with a q-value ≤ 0.05 are declared significant, 5% of the declared changes are expected to be false positives. In the method described by Higgs, et al.31, the p-Value to q-Value adjustment is done separately for Priority 1, Priority 2 and the MODERATE confidence categories.
FC is computed from the means on the AUC scale (antilog) as follows:
An FC of 1 means that no change exists.
Also, the median %CV for each priority level is determined. The %CV is the standard deviation / mean on a % scale. The %CV is given both for the replicate variation as well as the combined replicate plus sample variation.
For each protein a separate analysis of variance (ANOVA) model is fit:
Log2(Intensity) is the protein intensity based on the weighted average of the quantile normalized log base 2 peptide intensities with the same protein identification. Group Effect refers to the fixed effects (not random) caused by the experimental conditions or treatments that are to be compared. Patient Effect refers to the fixed effects from individual patients and is NOT assumed to be random. Residual Effect is a mixture of patient by group interaction and pure error due to duplicate samples. This model was necessary due to the experiment only having three patients. This situation resulted in a residual with eight degrees of freedom. Thus, this model implies that significant differences can only be attributed to these three specific patients and NOT generalized to a larger population of patients.
Nine hundred forty-four unique amino acid sequences were identified and quantified resulting in 285 proteins. Of the 285 proteins found in normal pancreatic juice, 90 proteins were detected by the highly confident identification (q < 0.1) of more than one unique peptide sequence within the protein (priority 1 proteins).39 Eighty-two proteins were identified by the highly confident identification of only one unique peptide sequence within the protein (priority 2 proteins). Together, the priority 1 and 2 categories resulted in 172 proteins (Table 1) that will be the focus of this study. The most commonly identified proteins were involved with proteolysis, ribonucleic acid (RNA) and deoxyribonucleic acid (DNA) function or lipid metabolism (Table 2).
Importantly, all 172 proteins were present before and after secretin, suggesting that secretin does not change the spectrum of proteins but rather the relative quantity. Following secretin stimulation, 44% of proteins expressed were increased and 56% decreased, usually by 1-2 folds (Table 2).
Comparison with 170 published pancreatic cancer proteins from Goggins et al40 yielded an overlap of only 42 proteins (Figure 1, Table 3). Putative tumor markers including azurocidin, carcinoembryonic antigen (CEA), insulin-like growth factor (ILGF) binding protein 2, lipocalin 2, mucin 1 (MUC1), pancreatitis associated protein/hepatocarcinoma-intestine-pancreas (PAP/HIP) and tumor rejection antigen were only found in cancer patients.
In summary, 285 proteins were identified in the normal pancreatic proteome and were most commonly involved with proteolysis, RNA or DNA functions or lipid metabolism. Table 2 lists the 172 priority one and two proteins according to their biological function. Several of the common proteins in the normal pancreatic proteome were those that provide the fundamental basis of the biologic and physiologic function of all cells and tissues, including basic structural cell components, cell signaling molecules and proteins involved in transcription and translation of DNA. Furthermore, as expected, several normal serum components and proteins associated with immunology and detoxification within the body also were present. However, many prominent proteins related to the specific function of the pancreas also were present.
The exocrine pancreas plays a crucial role in digestion, serving as the main source of enzymes responsible for the catabolism of proteins.41 Trypsinogen is an inactive zymogen secreted from the pancreas into the duodenum, where enteropeptidase cleaves it into its active form, trypsin, a serine protease responsible for cleaving peptide bonds and breaking proteins down into amino acids.41 Trypsin and its precursor proteins were commonly present in the normal pancreatic proteome. Other proteases were also represented, including serine proteases, chymotrypsin, elastase and carboxypeptidases.
The exocrine pancreas also contributed to several other components of digestion.41 Our studies detected the presence of pancreatic lipase, colipase and several related proteins all important for the breakdown and utilization of lipids. Amylase and enolase, both important in carbohydrate metabolism, also were well represented. Furthermore, anion exchange proteins and carbonic anhydrase were present, indicating the pancreatic production of bicarbonate ions which serve to neutralize the acidic chime that is produced during the gastric phase of digestion.
All identified proteins of the normal pancreas were present after secretin administration, although 44% of proteins expressed were increased and 56% decreased. Such results suggest that secretin changes the relative quantity of proteins but not the overall spectrum. This observation is extremely important in clinical applications, as secretin is frequently administered during pancreatic procedures to increase the production and flow of pancreatic fluid. Our studies indicate that the use of secretin is unlikely to change the presence of biomarkers within this fluid and lend credence to studies indicating the presence and absence of such molecules. However, future scientists interested in pursuing this important and interesting area of oncologic research should bear in mind the potential for changes in protein concentrations following secretin administration.
Comparison of the 172 proteins established within the normal pancreatic proteome with 170 published pancreatic cancer proteins yielded an overlap of only 42 proteins. Again, the ubiquitous cellular and serum proteins were present. However, also highly represented in this select group were those pancreatic proteins associated with protein and lipid metabolism, as discussed above. Putative tumor markers including azurocidin, carcinoembryonic antigen (CEA), insulin-like growth factor (ILGF) binding protein 2, lipocalin 2, mucin 1 (MUC1), pancreatitis associated protein/hepatocarcinoma-intestine-pancreas (PAP/HIP) and tumor rejection antigen were only found in cancer patients.25, 40 Although we found the lack of these established tumor markers in the normal pancreatic proteome interesting, it must be kept in mind that the samples and proteomics methods employed in our study were different from those used in the pancreatic adenocarcinoma studies, preventing a truly accurate comparison.
In conclusion, this analysis suggests that 1) normal pancreatic juice contains multiple proteins related to many biological processes, 2) secretin alters the concentration but not the spectrum of these proteins and 3) the pancreatic juice proteome of normal and pancreatic cancer patients differ markedly.
Grant Support: This work was supported by the NIH 1R03CA112629-01A1 (CMS)
Disclosure: Portions of this paper were presented at the Pancreas Club Annual Meeting in Washington, DC, May 20, 2007 and the American Pancreatic Association Meeting in Chicago, IL November 1-3, 2007.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Courtney J Doyle, Department of Surgery, Indiana University School of Medicine, Indianapolis, IN.
Kyle Yancey, Department of Surgery, Indiana University School of Medicine, Indianapolis, IN.
Henry A Pitt, Department of Surgery, Indiana University School of Medicine, Indianapolis, IN Indiana University Cancer Center, Indianapolis, IN.
Mu Wang, Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, IN.
Kerry Bemis, Department of Biostatistics, Monarch LifeSciences L.L.C., Indianapolis, IN.
Michele T. Yip-Schneider, Departments of Surgery and Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, IN.
Stuart Sherman, Department of Medicine, Indiana University School of Medicine, Indianapolis, IN Indiana University Cancer Center, Indianapolis, IN.
Keith D. Lillemoe, Department of Surgery, Indiana University School of Medicine, Indianapolis, IN Indiana University Cancer Center, Indianapolis, IN.
Michael D. Goggins, Department of Surgery, Johns Hopkins Medical Institutions, Baltimore, MD.
C. Max Schmidt, Departments of Surgery and Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis, IN Indiana University Cancer Center, Indianapolis, IN.