Assessment of the 2-D Gel-Based Proteomics Application of Clinically Archived Formalin-Fixed Paraffin Embedded Tissues

Hospital tissue repositories possess a vast and valuable supply of disease samples with matched retrospective clinical information. Detection and characterization of disease biomarkers in formalin-fixed paraffin-embedded (FFPE) tissues will greatly aid the understanding of the diseases mechanisms and help in the development of diagnostic and prognostic markers. In this study, the possibility of using full-length proteins extracted from clinically archived FFPE tissues in two-dimensional (2-D) gel-based proteomics was evaluated. The evaluation was done based on two types of tumor tissues (breast and prostate) and two extraction protocols. The comparison of the 2-D patterns of FFPE extracts obtained by two extraction protocols with the matching frozen tissue extracts showed that only 7–10 % of proteins from frozen tissues can be matched to proteins from FFPE tissues. Most of the spots in the 2-D FFPE’s maps had pl 4–6, while the percentages of proteins with pl above 6 were 3–5 times lower in comparison to the fresh/frozen tissue. Despite the three-fold lower number of the detected spots in FFPE maps compared to matched fresh/frozen maps, 67–78 % of protein spots in FFPE could not be matched to the corresponding spots in the fresh/frozen tissue maps indicating irreversible protein modifications. In conclusion, the inability to completely reverse the cross-linked complexes and overcome protein fragmentation with the present day FFPE extraction methods stands in the way of effective use of these samples in 2-D gel based proteomics studies.     

Validation of a robust proteomic analysis carried out on formalin‐fixed paraffin‐embedded tissues of the pancreas obtained from mouse and human

A number of reports have recently emerged with focus on extraction of proteins from formalin‐fixed paraffin‐embedded (FFPE) tissues for MS analysis; however, reproducibility and robustness as compared to flash frozen controls is generally overlooked. The goal of this study was to identify and validate a practical and highly robust approach for the proteomics analysis of FFPE tissues. FFPE and matched frozen pancreatic tissues obtained from mice (n = 8) were analyzed using 1D‐nanoLC‐MS(MS)² following work up with commercially available kits. The chosen approach for FFPE tissues was found to be highly comparable to that of frozen. In addition, the total number of unique peptides identified between the two groups was highly similar, with 958 identified for FFPE and 1070 identified for frozen, with protein identifications that corresponded by approximately 80%. This approach was then applied to archived human FFPE pancreatic cancer specimens (n = 11) as compared to uninvolved tissues (n = 8), where 47 potential pancreatic ductal adenocarcinoma markers were identified as significantly increased, of which 28 were previously reported. Further, these proteins share strongly overlapping pathway associations to pancreatic cancer that include estrogen receptor α. Together, these data support the validation of an approach for the proteomic analysis of FFPE tissues that is straightforward and highly robust, which can also be effectively applied toward translational studies of disease.     

Generation of high‐quality protein extracts from formalin‐fixed,paraffin‐embedded tissues

A wealth of information on proteins involved in many aspects of disease is encased within formalin‐fixed paraffin‐embedded (FFPE) tissue repositories stored in hospitals worldwide. Recently, access to this “hidden treasure” is being actively pursued by the application of two main extraction strategies: digestion of the entangled protein matrix with generation of tryptic peptides, or decrosslinking and extraction of full‐length proteins. Here, we describe an optimised method for extraction of full‐length proteins from FFPE tissues. This method builds on the classical “antigen retrieval” technique used for immunohistochemistry, and allows generation of protein extracts with elevated and reproducible yields. In model animal tissues, average yields of 16.3 μg and 86.8 μg of proteins were obtained per 80 mm² tissue slice of formalin‐fixed paraffin‐embedded skeletal muscle and liver, respectively. Protein extracts generated with this method can be used for the reproducible investigation of the proteome with a wide array of techniques. The results obtained by SDS‐PAGE, western immunoblotting, protein arrays, ELISA, and, most importantly, nanoHPLC‐nanoESI‐Q‐TOF MS of FFPE proteins resolved by SDS‐PAGE, are presented and discussed. An evaluation of the extent of modifications introduced on proteins by formalin fixation and crosslink reversal, and their impact on quality of MS results, is also reported.     

Protein Extraction of Formalin-fixed, Paraffin-embedded Tissue Enables Robust Proteomic Profiles by Mass Spectrometry

Global mass spectrometry (MS) profiling and spectral count quantitation are used to identify unique or differentially expressed proteins and can help identify potential biomarkers. MS has rarely been conducted in retrospective studies, because historically, available samples for protein analyses were limited to formalin-fixed, paraffin-embedded (FFPE) archived tissue specimens. Reliable methods for obtaining proteomic profiles from FFPE samples are needed. Proteomic analysis of these samples has been confounded by formalin-induced protein cross-linking. The performance of extracted proteins in a liquid chromatography tandem MS format from FFPE samples and extracts from whole and laser capture microdissected (LCM) FFPE and frozen/optimal cutting temperature (OCT)–embedded matched control rat liver samples were compared. Extracts from FFPE and frozen/OCT–embedded livers from atorvastatin-treated rats were further compared to assess the performance of FFPE samples in identifying atorvastatin-regulated proteins. Comparable molecular mass representation was found in extracts from FFPE and OCT-frozen tissue sections, whereas protein yields were slightly less for the FFPE sample. The numbers of shared proteins identified indicated that robust proteomic representation from FFPE tissue and LCM did not negatively affect the number of identified proteins from either OCT-frozen or FFPE samples. Subcellular representation in FFPE samples was similar to OCT-frozen, with predominantly cytoplasmic proteins identified. Biologically relevant protein changes were detected in atorvastatin-treated FFPE liver samples, and selected atorvastatin-related proteins identified by MS were confirmed by Western blot analysis. These findings demonstrate that formalin fixation, paraffin processing, and LCM do not negatively impact protein quality and quantity as determined by MS and that FFPE samples are amenable to global proteomic analysis. (J Histochem Cytochem 57:849–860, 2009)     

Integrated and convenient procedure for protein extraction from formalin‐fixed,paraffin‐embedded tissues for LC‐MS/MS analysis

Because fresh‐frozen tissue samples associated with long‐term clinical data and of rare diseases are often unobtainable at the present time, formalin‐fixed paraffin‐embedded (FFPE) tissue samples are considered a highly valuable resource for researchers. However, protein extraction from FFPE tissues faces challenges of deparaffinization and cross‐link reversion. Current procedures for protein extraction from FFPE tissue require separate steps and toxic solvents, resulting in inconvenience in protein extraction. To overcome these limitations, an integrated method was developed using nontoxic solvents in four types of FFPE tissues. The average amount of proteins from three replicates of bladder, kidney, liver, and lung FFPE tissues were 442.6, 728.9, 736.4, and 694.7 μg with CVs of 7.5, 5.8, 2.4, and 4.5%, respectively. Proteomic analysis showed that 348, 417, 607, and 304 unique proteins were identified and quantified without specification of isoform by a least two peptides from bladder, kidney, liver, and lung FFPE tissue samples, respectively. The analysis of individual protein CV demonstrated that 97–99% of the proteins were quantified with a CV ≤ 30%, verifying the reproducibility of the integrated protein extraction method. In summary, the developed method is high‐yield, reproducible, convenient, simple, low cost, nonvolatile, nonflammable, and nontoxic.     

Quantitative Proteomic Analysis of Formalin-fixed and Paraffin-embedded Nasopharyngeal Carcinoma Using iTRAQ Labeling,Two-dimensional Liquid Chromatography,and Tandem Mass Spectrometry

Formalin-fixed, paraffin-embedded (FFPE) tissue specimens represent a potentially valuable resource for protein biomarker investigations. In this study, proteins were extracted by a heat-induced antigen retrieval technique combined with a retrieval solution containing 2% SDS from FFPE tissues of normal nasopharyngeal epithelial tissues (NNET) and three histological types of nasopharyngeal carcinoma (NPC) with diverse differentiation degrees. Then two-dimensional liquid chromatography-tandem mass spectrometry coupled with isobaric tags for relative and absolute quantification (iTRAQ) labeling was employed to quantitatively identify the differentially expressed proteins among the types of NPC FFPE tissues. Our study resulted in the identification of 730 unique proteins, the distributions of subcellular localizations and molecular functions of which were similar to those of the proteomic database of human NPC and NNET that we had set up based on the frozen tissues. Additionally, the relative expression levels of cathepsin D, keratin8, SFN, and stathmin1 identified and quantified in this report were consistent with the immunohistochemistry results acquired in our previous study. In conclusion, we have developed an effective approach to identifying protein changes in FFPE NPC tissues utilizing iTRAQ technology in conjunction with an economical and easily accessible sample preparation method. (J Histochem Cytochem 58:517–527, 2010)     

Protein recovery and identification from the gulf killifish, Fundulus grandis: comparing snap-frozen and RNAlater® preserved tissues

Reliable proteomic analysis of biological tissues requires sampling approaches that preserve proteins as close to their in vivo state as possible. In the current study, the patterns of protein abundance in one‐dimensional (1‐D) gels were assessed for five tissues of the gulf killifish, Fundulus grandis, following snap‐freezing tissues in liquid nitrogen or immersion of fresh tissues in RNAlater^®. In liver and heart, the protein profiles in 1‐D gels were better preserved by snap‐freezing, while in gill, the 1‐D protein profile was better preserved by immersion in RNAlater^®. In skeletal muscle and brain, the two approaches yielded similar patterns of protein abundance. LC‐MS/MS analyses and database searching resulted in the identification of 17 proteins in liver and 12 proteins in gill. Identified proteins include enzymes of energy metabolism, structural proteins, and proteins serving other biological functions. These protein identifications for a species without a sequenced genome demonstrate the utility of F. grandis as a model organism for environmental proteomic studies in vertebrates.     

Proteomic analysis of doxorubicin‐induced changes in the proteome of HepG2cells combining 2‐D DIGE and LC‐MS/MS approaches

HepG‐2 cells are widely used as a cell model to investigate hepatocellular carcinomas and the effect of anticancer drugs such as doxorubicin, an effective antineoplastic agent, which has broad antitumoral activity against many solid and hematological malignancies. To investigate the effect of doxorubicin on the protein pattern, we used complementary proteomic workflows including 2‐D gel‐based and gel‐free methods. The analysis of crude HepG2 cell extracts by 2‐D DIGE provided data on 1835 protein spots which was then complemented by MS‐centered analysis of stable isotope labeling by amino acids in cell culture‐labeled cells. The monitoring of more than 1300 distinct proteins, including proteins of the membrane fraction provides the most comprehensive overview on the proteome of the widely used model cell line HepG2. Of the proteins monitored in total, 155 displayed doxorubicin‐induced changes in abundance. Functional analysis revealed major influences of doxorubicin on proteins involved in protein synthesis, DNA damage control, electron transport/mitochondrial function, and tumor growth. The strongest decrease in level was found for proteins involved in DNA replication and protein synthesis, whereas proteins with a function in DNA damage control and oxidative stress management displayed increased levels following treatment with doxorubicin compared with control cells. Furthermore, the doxorubicin‐associated increase in levels of multiple forms of keratins 8, 18, and 19 and other structural proteins revealed an influence on the cytoskeleton network.     

Differential proteomic analysis of lymphatic,venous, and arterial endothelial cells extracted from bovine mesenteric vessels

Differential protein profiling by 2‐D PAGE is generally useful in biomarker discovery, proteome analysis and routine sample preparation prior to analysis by MS. The goal of this study was to compare 2‐D PAGE‐resolved protein profile of lymphatic endothelial cells to those of venous, and arterial endothelial cells isolated from lymphatic and blood vessels of bovine mesentery (bm). Three 2‐D PAGE electrophoretograms were produced for each of the three cell types and quantitatively analyzed. Protein identification by LC‐MS/MS was performed to identify 39 proteins found to be present at statistically significantly different levels in the three cell types (p<0.05). Most of the 39 proteins have not been previously reported in EC proteomic studies of 2‐D PAGE electrophoretograms. Three proteins, HSPA1B (HSP70 family member), HSPB1 (HSP27 family member), and UBE2D3 (a member of E2 ubiquitin‐conjugating enzymes) found to be at highest levels in bm arterial endothelial cells, bm venous endothelial cells, and bm lymphatic endothelial cells, respectively, were validated by immunoblotting with appropriate antibodies. The lack of substantial overlap between our results and those of other groups' comparative studies are discussed. Functional implications of differences in levels of various proteins identified in the three cell types are also discussed.     

mTRAQ‐based quantification of potential endometrial carcinoma biomarkers from archived formalin‐fixed paraffin‐embedded tissues

Formalin‐fixed paraffin‐embedded (FFPE) tissues are the primary and preferred medium for archiving patients' samples. Here we demonstrate relative quantifications of protein biomarkers in extracts of laser microdissected epithelial cells from FFPE endometrial carcinoma tissues versus those from normal proliferative endometria by means of targeted proteomic analyses using LC–multiple reaction monitoring (MRM) MS with MRM Tags for Relative and Absolute Quantitation (mTRAQ) labeling. Comparable results of differential expressions for pyruvate kinase isoform M2 (PK‐M2) and polymeric Ig receptor were observed between analyses on laser microdissected epithelial cells from FFPE tissues and corresponding homogenates from frozen tissues of the same individuals that had previously been analyzed and reported. We also identified PK‐M2 in the normal proliferative phase of the endometrium. Other biomarkers in addition to PK‐M2 and polymeric Ig receptor were also observed but not consistently and/or were at levels below the threshold for quantification.     

Fractionation and identification of proteins by 2-DE and MS: towards a proteomic analysis of Plasmodium falciparum

Since completion of genome sequencing of the malarial parasite Plasmodium falciparum, proteomic tools for the identification of parasite proteins have become particularly attractive as they allow a more thorough interpretation of these data. Recent advances in 2-D PAGE, MS, and bioinformatics have created great opportunities for mapping and characterization of protein populations. We employed these improvements in a proteomic approach for the analysis of proteins detected in two blood stages of P. falciparum, (i) in the schizont stage and (ii) in the merozoite stage. For the isolation of merozoites, we introduced a new protocol based on the preparation of clustered structures of merozoites upon treatment of cultures with the common cysteine proteinase inhibitor E64. Peptide mass fingerprints of excised and trypsinated protein spots, acquired by MALDI-TOF MS were generated to identify a variety of proteins. Moreover, prefractionation procedures were used to enrich and map low-abundance proteins in protein samples. The data demonstrate that classic proteomic analyses using 2-D PAGE are now feasible for P. falciparum and represent the first step in the direction of creating 2-D reference maps for this medically most relevant protozoon.     

Two methods for proteomic analysis of formalin‐fixed,paraffin embedded tissue result in differential protein identification,data quality,and cost

Formalin‐fixed paraffin‐embedded (FFPE) tissue is a rich source of clinically relevant material that can yield important translational biomarker discovery using proteomic analysis. Protocols for analyzing FFPE tissue by LC‐MS/MS exist, but standardization of procedures and critical analysis of data quality is limited. This study compared and characterized data obtained from FFPE tissue using two methods: a urea in‐solution digestion method (UISD) versus a commercially available Qproteome FFPE Tissue Kit method (Qkit). Each method was performed independently three times on serial sections of homogenous FFPE tissue to minimize pre‐analytical variations and analyzed with three technical replicates by LC‐MS/MS. Data were evaluated for reproducibility and physiochemical distribution, which highlighted differences in the ability of each method to identify proteins of different molecular weights and isoelectric points. Each method replicate resulted in a significant number of new protein identifications, and both methods identified significantly more proteins using three technical replicates as compared to only two. UISD was cheaper, required less time, and introduced significant protein modifications as compared to the Qkit method, which provided more precise and higher protein yields. These data highlight significant variability among method replicates and type of method used, despite minimizing pre‐analytical variability. Utilization of only one method or too few replicates (both method and technical) may limit the subset of proteomic information obtained.     

Analysis of the budding yeast pH 4–7 proteome in meiosis

Meiosis, the developmental programme generating haploid gametes from diploid precursors, requires two cell divisions and many innovations. In budding yeast, a large number of genes are expressed exclusively during meiosis while others are repressed compared to vegetative growth. Microarray analysis has shown that gene expression during meiosis is highly regulated, and has been used to classify yeast genes according to meiotic temporal expression pattern. In this study, we have begun to investigate the kinetics of meiotic protein expression using a proteomics approach. 2‐D DIGE was used to characterise the temporal protein expression patterns of the budding yeast pH 4–7 proteome in meiosis. More than 1400 meiotic protein spots were visualised and at least 63 spots were temporally regulated during meiosis in a statistically significant manner. Gel spots with significant expression changes were excised and 26 unique proteins were identified using LC‐MS/MS. The identified proteins could be classified into functional categories and the genes encoding a number of these were previously shown to be involved in yeast sporulation and meiosis. This data set was used to assemble the first differential 2‐D PAGE map of budding yeast meiosis, which can be accessed through a web server. This work represents one of the first quantitative proteomic analyses of meiosis in yeast and will provide a valuable resource for future investigations.     

Altered protein expression pattern in colon tissue of mice upon supplementation with distinct selenium compounds

The essential trace element selenium (Se) is controversially discussed concerning its role in health and disease. Its various physiological functions are largely mediated by Se incorporation in the catalytic center of selenoproteins. In order to gain insights into the impact of Se deficiency and of supplementation with different Se compounds (selenite, selenate, selenomethionine) at defined concentrations (recommended, 150 μg/kg diet; excessive, 750 μg/kg diet) in murine colon tissues, a 20‐week feeding experiment was performed followed by analysis of the protein expression pattern of colon tissue specimens by 2D‐DIGE and MALDI‐TOF MS. Using this approach, 24 protein spots were identified to be significantly regulated by the different Se compounds. These included the antioxidant enzyme peroxiredoxin‐5 (PRDX5), proteins with binding capabilities, such as cofilin‐1 (COF1), calmodulin, and annexin A2 (ANXA2), and proteins involved in catalytic processes, such as 6‐phosphogluconate dehydrogenase (6PGD). Furthermore, the Se compounds demonstrated a differential impact on the expression of the identified proteins. Selected target structures were validated by qPCR and Western blot which mainly confirmed the proteomic profiling data. Thus, novel Se‐regulated proteins in colon tissues have been identified, which expand our understanding of the physiologic role of Se in colon tissue.     

Unrestricted modification search reveals lysine methylation as major modification induced by tissue formalin fixation and paraffin embedding

Formalin‐fixed paraffin‐embedded (FFPE) tissue is considered as an appropriate alternative to frozen/fresh tissue for proteomic analysis. Here we study formalin‐induced alternations on a proteome‐wide level. We compared LC‐MS/MS data of FFPE and frozen human kidney tissues by two methods. First, clustering analysis revealed that the biological variation is higher than the variation introduced by the two sample processing techniques and clusters formed in accordance with the biological tissue origin and not with the sample preservation method. Second, we combined open modification search and spectral counting to find modifications that are more abundant in FFPE samples compared to frozen samples. This analysis revealed lysine methylation (+14 Da) as the most frequent modification induced by FFPE preservation. We also detected a slight increase in methylene (+12 Da) and methylol (+30 Da) adducts as well as a putative modification of +58 Da, but they contribute less to the overall modification count. Subsequent SEQUEST analysis and X!Tandem searches of different datasets confirmed these trends. However, the modifications due to FFPE sample processing are a minor disturbance affecting 2–6% of all peptide‐spectrum matches and the peptides lists identified in FFPE and frozen tissues are still highly similar.     

Efficient method for the proteomic analysis of fixed and embedded tissues

Formalin-fixed and paraffin-embedded (FFPE) tissues present a particular challenge for proteomic analysis. Yet, most of the archived tissues in hospitals and tissue banks worldwide are only available in this form. We have developed conditions for removal of the embedding medium and protein digestion, such that informative tryptic peptides are released from fixed proteins which are suitable for analysis by liquid chromatography-mass spectrometry (LC-MS). We demonstrate that the peptide identifications made by this approach compare favorably to those made from matched fresh frozen tissue. Moreover, we demonstrate that a high level of sequence coverage can be observed for proteins of interest.     

Stopping the clock on proteomic degradation by heat treatment at the point of tissue excision

The effectiveness of rapid and controlled heating of intact tissue to inactivate native enzymatic activity and prevent proteome degradation has been evaluated. Mouse brains were bisected immediately following excision, with one hemisphere being heat treated followed by snap freezing in liquid nitrogen while the other hemisphere was snap frozen immediately. Sections were cut by cryostatic microtome and analyzed by MALDI‐MS imaging and minimal label 2‐D DIGE, to monitor time‐dependent relative changes in intensities of protein and peptide signals. Analysis by MALDI‐MS imaging demonstrated that the relative intensities of markers varied across a time course (0–5 min) when the tissues were not stabilized by heat treatment. However, the same markers were seen to be stabilized when the tissues were heat treated before snap freezing. Intensity profiles for proteins indicative of both degradation and stabilization were generated when samples of treated and nontreated tissues were analyzed by 2‐D DIGE, with protein extracted before and after a 10‐min warming of samples. Thus, heat treatment of tissues at the time of excision is shown to prevent subsequent uncontrolled degradation of tissues at the proteomic level before any quantitative analysis, and to be compatible with downstream proteomic analysis.     

LC‐MS/MS as an alternative for SDS‐PAGE in blue native analysis of protein complexes

Two‐dimensional blue native/SDS‐PAGE is widely applied to investigate native protein–protein interactions, particularly those within membrane multi‐protein complexes. MS has enabled the application of this approach at the proteome scale, typically by analysis of picked protein spots. Here, we investigated the potential of using LC‐MS/MS as an alternative for SDS‐PAGE in blue native (BN) analysis of protein complexes. By subjecting equal slices from BN gel lanes to label‐free semi‐quantitative LC‐MS/MS, we determined an abundance profile for each protein across the BN gel, and used these profiles to identify potentially interacting proteins by protein correlation profiling. We demonstrate the feasibility of this approach by considering the oxidative phosphorylation complexes I–V in the native human embryonic kidney 293 mitochondrial fraction, showing that the method is capable of detecting both the fully assembled complexes as well as assembly/turnover intermediates of complex I (NADH:ubiquinone oxidoreductase). Using protein correlation profiling with a profile for subunits NDUFS2, 3, 7 and 8 we identified multiple proteins possibly involved in the biogenesis of complex I, including the recently implicated chaperone C6ORF66 and a novel candidate, C3ORF60.