Equivalence of Protein Inventories Obtained from Formalin-fixed Paraffin-embedded and Frozen Tissue in Multidimensional Liquid Chromatography-Tandem Mass Spectrometry Shotgun Proteomic Analysis

Formalin-fixed paraffin-embedded (FFPE) tissue specimens comprise a potentially valuable resource for retrospective biomarker discovery studies, and recent work indicates the feasibility of using shotgun proteomics to characterize FFPE tissue proteins. A critical question in the field is whether proteomes characterized in FFPE specimens are equivalent to proteomes in corresponding fresh or frozen tissue specimens. Here we compared shotgun proteomic analyses of frozen and FFPE specimens prepared from the same colon adenoma tissues. Following deparaffinization, rehydration, and tryptic digestion under mild conditions, FFPE specimens corresponding to 200 µg of protein yielded ∼400 confident protein identifications in a one-dimensional reverse phase liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis. The major difference between frozen and FFPE proteomes was a decrease in the proportions of lysine C-terminal to arginine C-terminal peptides observed, but these differences had little effect on the proteins identified. No covalent peptide modifications attributable to formaldehyde chemistry were detected by analyses of the MS/MS datasets, which suggests that undetected, cross-linked peptides comprise the major class of modifications in FFPE tissues. Fixation of tissue for up to 2 days in neutral buffered formalin did not adversely impact protein identifications. Analysis of archival colon adenoma FFPE specimens indicated equivalent numbers of MS/MS spectral counts and protein group identifications from specimens stored for 1, 3, 5, and 10 years. Combination of peptide isoelectric focusing-based separation with reverse phase LC-MS/MS identified 2554 protein groups in 600 ng of protein from frozen tissue and 2302 protein groups from FFPE tissue with at least two distinct peptide identifications per protein. Analysis of the combined frozen and FFPE data showed a 92% overlap in the protein groups identified. Comparison of gene ontology categories of identified proteins revealed no bias in protein identification based on subcellular localization. Although the status of posttranslational modifications was not examined in this study, archival samples displayed a modest increase in methionine oxidation, from ∼17% after one year of storage to ∼25% after 10 years. These data demonstrate the equivalence of proteome inventories obtained from FFPE and frozen tissue specimens and provide support for retrospective proteomic analysis of FFPE tissues for biomarker discovery. Formalin-fixed paraffin-embedded (FFPE) tissue specimens comprise a potentially valuable resource for retrospective biomarker discovery studies, and recent work indicates the feasibility of using shotgun proteomics to characterize FFPE tissue proteins. A critical question in the field is whether proteomes characterized in FFPE specimens are equivalent to proteomes in corresponding fresh or frozen tissue specimens. Here we compared shotgun proteomic analyses of frozen and FFPE specimens prepared from the same colon adenoma tissues. Following deparaffinization, rehydration, and tryptic digestion under mild conditions, FFPE specimens corresponding to 200 µg of protein yielded ∼400 confident protein identifications in a one-dimensional reverse phase liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis. The major difference between frozen and FFPE proteomes was a decrease in the proportions of lysine C-terminal to arginine C-terminal peptides observed, but these differences had little effect on the proteins identified. No covalent peptide modifications attributable to formaldehyde chemistry were detected by analyses of the MS/MS datasets, which suggests that undetected, cross-linked peptides comprise the major class of modifications in FFPE tissues. Fixation of tissue for up to 2 days in neutral buffered formalin did not adversely impact protein identifications. Analysis of archival colon adenoma FFPE specimens indicated equivalent numbers of MS/MS spectral counts and protein group identifications from specimens stored for 1, 3, 5, and 10 years. Combination of peptide isoelectric focusing-based separation with reverse phase LC-MS/MS identified 2554 protein groups in 600 ng of protein from frozen tissue and 2302 protein groups from FFPE tissue with at least two distinct peptide identifications per protein. Analysis of the combined frozen and FFPE data showed a 92% overlap in the protein groups identified. Comparison of gene ontology categories of identified proteins revealed no bias in protein identification based on subcellular localization. Although the status of posttranslational modifications was not examined in this study, archival samples displayed a modest increase in methionine oxidation, from ∼17% after one year of storage to ∼25% after 10 years. These data demonstrate the equivalence of proteome inventories obtained from FFPE and frozen tissue specimens and provide support for retrospective proteomic analysis of FFPE tissues for biomarker discovery. Formalin-fixed paraffin-embedded (FFPE) 1The abbreviations used are:FFPEformalin-fixed paraffin-embeddedIHCimmunohistochemistryPMpyridoxamineIDidentificationIEFisoelectric focusingFAformic acidACNacetonitrileFDRfalse discovery ratesANOVAanalysis of variancePTMposttranslational modification. tissue samples are routinely prepared during the pathological characterization of clinical specimens and are abundantly available in pathology archives worldwide. The fixation process yields clinically relevant samples that can be stored at ambient temperature and are suitable for pathological examination by light microscopy even after years in storage. Given the wealth of clinical data associated with specimens collected over a span of decades, such as patient treatment regimens and outcomes, FFPE tissue represents a potentially valuable resource for biomarker discovery through retrospective analysis (1Hood B.L. Conrads T.P. Veenstra T.D. Mass spectrometric analysis of formalin-fixed paraffin-embedded tissue: unlocking the proteome within.Proteomics. 2006; 6: 4106-4114Crossref PubMed Scopus (84) Google Scholar, 2Yamashita S. Heat-induced antigen retrieval: mechanisms and application to histochemistry.Prog. Histochem. Cytochem. 2007; 41: 141-200Crossref PubMed Scopus (125) Google Scholar). formalin-fixed paraffin-embedded immunohistochemistry pyridoxamine identification isoelectric focusing formic acid acetonitrile false discovery rates analysis of variance posttranslational modification. However, fixation of tissue in formalin leads to significant cross-linking among proteins and other biomolecules, rendering the samples incompatible with many biochemical analyses. Immunohistochemical (IHC) analysis of FFPE tissue has been conducted since the 1970s using either proteolysis or protein denaturants to expose antigenic regions of proteins (3Huang S.N. Minassian H. More J.D. Application of immunofluorescent staining on paraffin sections improved by Google Scholar, S. H. S. of in the of and PubMed Scopus Google Scholar). the of in FFPE tissue has been improved through the of antigen in paraffin-embedded for staining based on of tissue Histochem. Cytochem. PubMed Scopus Google Scholar, S. the of formalin fixation with and a antigen Histochem. Cytochem. PubMed Scopus Google Scholar). These application of in the of of a of in the of during the of fixation S. Heat-induced antigen retrieval: mechanisms and application to histochemistry.Prog. Histochem. Cytochem. 2007; 41: 141-200Crossref PubMed Scopus (125) Google Scholar). for FFPE tissue samples been in proteomics studies, to the that tissue fixation proteomic analysis work to this B.L. Conrads T.P. Veenstra T.D. analysis of formalin-fixed PubMed Scopus Google the application of shotgun proteome analysis to FFPE and of proteins were identified from a and the to proteomic analyses of FFPE the same showed that the numbers and of proteins were the to frozen and FFPE support to the of FFPE tissue in biomarker discovery the of FFPE tissues from and been to proteomic analyses H. for proteomic characterization of formalin-fixed and paraffin-embedded PubMed Scopus Google Scholar, of proteins from formalin-fixed paraffin-embedded by PubMed Scopus (125) Google Scholar, tissue a to from formalin-fixed paraffin-embedded archival 2007; PubMed Scopus Google Scholar, S. H. of protein for shotgun proteome analysis of formalin-fixed 2007; 6: PubMed Scopus Google Scholar, S.N. for the proteomic analysis of and PubMed Scopus Google Scholar, from paraffin-embedded tissue by mass Histochem. Cytochem. 2006; PubMed Scopus Google Scholar, analysis of formalin-fixed and paraffin-embedded tissue Histochem. Cytochem. 2007; PubMed Scopus Google Scholar). Although this work suggests the feasibility of biomarker discovery from FFPE of these been on of with one-dimensional reverse phase liquid chromatography-tandem mass spectrometry (LC-MS/MS) The of peptide can the of the LC-MS/MS analyses to proteins. the of as the in a peptide separation for FFPE tissue was H. for proteomic characterization of formalin-fixed and paraffin-embedded PubMed Scopus Google Scholar). this study, of proteins were identified of µg of from FFPE However, the used was not to other of these studies, proteins identified by a peptide were as identifications. of identifications the of false protein and these identifications the of protein identifications for in protein identifications by mass 2007; PubMed Scopus Google Scholar). The equivalence of and FFPE tissue proteomes is a critical in the of FFPE tissues for biomarker discovery by proteomic analyses. B.L. Conrads T.P. Veenstra T.D. analysis of formalin-fixed PubMed Scopus Google and analysis of formalin-fixed and paraffin-embedded tissue Histochem. Cytochem. 2007; PubMed Scopus Google from analyses of fresh and frozen tissue specimens. analysis of formalin-fixed and paraffin-embedded tissue Histochem. Cytochem. 2007; PubMed Scopus Google overlap of in protein identifications between FFPE and frozen tissue B.L. Conrads T.P. Veenstra T.D. analysis of formalin-fixed PubMed Scopus Google did not the of but that FFPE tissue yielded of the identifications for frozen The of protein identifications in were based on peptide These did not the effect of modifications on the inventories of identified question with FFPE tissue specimens is the to which in fixation process and storage the proteomes The of tissue fixation is not and from to of the of FFPE specimens is the for retrospective biomarker but the of storage for many years on tissue proteomes we these through of the analysis of fresh frozen and FFPE tissues by shotgun used the same fresh tissue specimens to frozen and FFPE samples for for tissue and digestion and the of fixation and storage on the of protein obtained during shotgun proteomic analysis of FFPE characterized the differences in peptides between and frozen specimens in to the effect of fixation from a biomarker discovery we compared analyses of fresh frozen and FFPE colon adenoma tissue by LC-MS/MS using isoelectric focusing of peptides The demonstrate a overlap in the and of proteins between the and frozen tissue and that in of fixation and storage a effect on protein inventories obtained by shotgun proteomic analysis. The data equivalence between protein inventories obtained from fresh frozen and FFPE tissue specimens by shotgun proteomics and the of FFPE tissue specimens for biomarker discovery. was obtained from and buffered formalin was from and pyridoxamine were from was from was from and and were from and were from and was from and frozen colon adenoma were obtained from the The samples and were to of tissue were in was with in of and was with in of and were in of either or with or pyridoxamine as in the were at for 2 digestion was by of the of H. and of a and proteomic PubMed Scopus Google Scholar). was and the samples were for by on The was The was for at by a of as The was with and at for by with in the at ambient temperature for The and was to with by of at The was at by at and were in of liquid over and in with a focusing of tryptic peptides was from the of isoelectric focusing as a separation in shotgun Google Scholar). tryptic peptides were in of and in were over the samples and to at ambient The were at on using the at for to for to for to for The were and in of a were from the as 200 of formic acid for 200 of acetonitrile for 200 of for of peptides were in in of acid and over a were in in of and in for LC-MS/MS analysis. LC-MS/MS analyses were on mass with and or were on a with was at ambient temperature at a of using a of formic acid in and formic acid in from the were the in with a of A was obtained for peptides in the of by MS/MS MS/MS were using of for with a of and a of MS/MS were by of the peptide at of to a of and as major The mass stored as from and to MS/MS were to the were of the of a mass at a that of the a was the was under and mass were to peptides from the 2007; by the mass spectral peptide identification by 2007; 6: PubMed Scopus Google Scholar). The was to in and false discovery was to to modifications and to for the of on peptides were to or protein at of was A of was but were to The through analysis and 2007; 6: PubMed Scopus Google the identifications for reverse phase liquid to the for which a identification false discovery be as by of peptide and protein identifications from mass spectrometry and the PubMed Scopus Google Scholar). proteins were and and were to a of proteins that of the peptides that through analysis and 2007; 6: PubMed Scopus Google Scholar). analysis to a of protein identifications with of These identifications were for were to at least two peptide discovery rates for peptide identifications were by the for in protein identifications by mass 2007; PubMed Scopus Google The the of and of distinct for protein and protein group in of at the tissue and the for are in the as of protein group obtained and lysine to arginine were compared using a with in of and were that spectral a of protein in protein for proteins H. A for and of protein in shotgun PubMed Scopus Google Scholar). analysis and using a based on used for data in the of PubMed Scopus Google with a for the false discovery a and to Scholar). compared protein group and peptide identifications between the FFPE and frozen tissue groups and and of to proteins and peptides that had between these two analyses of proteins identified between is of protein from FFPE tissue for the acid is not that to the of are to with the protein from FFPE tissue be the by for this of a frozen colon adenoma were The adenoma was in in protein to differences in tissue were to a were and to fixation in formalin for at ambient The formalin was and the and was to the and frozen tissue in the was using the acid The sections a protein that was of the for frozen tissue equivalent of protein in tissue this a a of protein in FFPE colon adenoma of the FFPE tissue were in to the of confident protein obtained during shotgun analysis of tryptic from tissue The were on the of for mass spectrometry a phase and in and pyridoxamine in was has been that can between groups that are as between cross-linked proteins are during S. of protein by is the Histochem. Cytochem. PubMed Scopus Google Scholar, H. of in antigen PubMed Scopus Google Scholar). of and to was to which through the of was to to to a of of the numbers of protein groups identified from shotgun analysis of the indicated a significant difference between the groups a for revealed a significant in numbers of protein groups identified in the and pyridoxamine FFPE the frozen No significant difference in the of protein groups identified was frozen FFPE tissue with the FFPE tissue to one These are with that the and little effect on protein but that at is critical S. Heat-induced antigen retrieval: mechanisms and application to histochemistry.Prog. Histochem. Cytochem. 2007; 41: 141-200Crossref PubMed Scopus (125) Google Scholar). and cross-linking by formaldehyde are to the of lysine H. of modifications in with 2006; PubMed Scopus Google Scholar). one C-terminal peptides to be shotgun proteomic analysis of a tryptic of FFPE Comparison of the of C-terminal lysine to arginine during shotgun analysis of FFPE and frozen tissue this The of to peptides revealed a significant of this in the FFPE a of peptides in the samples 3, is with the chemistry of formaldehyde and with that peptides are in proteomic analyses of FFPE tryptic B.L. Conrads T.P. Veenstra T.D. analysis of formalin-fixed PubMed Scopus Google Scholar). for this in to be that of acid with formaldehyde identification during modifications on and lysine and and on and arginine H. of modifications in with 2006; PubMed Scopus Google Scholar). However, the the and for these modifications did not peptide that be through spectral conducted analysis using the which discovery of peptide modifications among proteins to be in a the to the mass of and to modifications to peptide using PubMed Scopus Google Scholar). mass and modifications and the bias by of with acid mass as of the of the proteins on spectral counts from the the MS/MS for frozen and FFPE tissues to modifications to formaldehyde The that acid modifications by formalin fixation of tissue in or that these modifications the and of the The used for peptide from MS/MS are to for the of from the of cross-linked for this been S. for protein cross-linking and mass 2006; PubMed Scopus Google Scholar, of cross-linked peptides from PubMed Scopus Google but of the proteins in the a the of FFPE tissue samples identification of cross-linked at this of in peptide and protein identifications from FFPE samples is fixation Fixation can among pathology and specimens be in formalin for of the of fixation and protein this we a frozen colon of was and the other were in a buffered formalin for 1, or tissue was in paraffin and of tissue from treatment were and were for indicated a significant difference in protein groups identified among the fixation between the groups revealed no significant difference in the numbers of protein groups identified among the and 2 fixation samples However, the fixation yielded protein group with a of that was from the other groups The numbers of spectral counts from these samples were compared and to significant differences The indicated a significant in the of for the 2 and fixation compared with the frozen and suggests that of fixation has a impact on the of the samples with proteomic peptide spectral counts were at the fixation compared with the frozen with the from the analysis of the effect of on protein identifications These to be in with a recent in which numbers of were obtained between frozen tissue and tissue for up to days H. for proteomic characterization of formalin-fixed and paraffin-embedded PubMed Scopus Google Scholar). is that in the fixation was in a using with no paraffin The of fixation to is to be in a frozen was and to in as the and of fixation are to impact on for shotgun proteomic data represents a FFPE tissue specimens be for of in analyses of archival FFPE tissue is the of of storage. this we conducted shotgun proteomic analysis on FFPE colon adenoma tissue samples that had been in storage for 1, 3, 5, or 10 years. The samples were by on the of tissue and from of FFPE colon at were as under No significant difference was in the of protein obtained from tissues that had been stored for up to a and we compared the spectral counts to the identification of a peptide among the as a of the of obtained from no significant difference in the of identified among of the was no difference in the of or peptide spectral counts for the storage that the of storage has no effect on the of cross-linking and impact other modifications among such that one modifications as methionine to be in did a modest increase in the of methionine among which with storage and for the 1, 3, 5, and 10 year storage peptides were detected this was as a in analyses with were to corresponding to peptides from proteins and in and FFPE These analyses detected but did not acid These that a modest of protein during FFPE tissue on methionine These are that storage of up to 10 years not be a to retrospective proteomic analysis. since is that these samples were in the same been collected and over the span of a these that in fixation can be during shotgun proteomic analysis. separation been at the protein and peptide to proteomic by peptide for LC-MS/MS analysis H. A for and of protein in shotgun PubMed Scopus Google Scholar). However, the of on for analysis of FFPE samples has to be this separation 200 µg of tryptic from of FFPE and frozen colon adenoma tissue were using a peptide separation in A separation tryptic peptides from a of proteins and was based on isoelectric focusing as a separation in shotgun Google Scholar, of isoelectric focusing of peptides for liquid chromatography-tandem mass PubMed Scopus Google Scholar). yielded a increase in protein group identifications compared with reverse phase LC-MS/MS analysis. the of protein groups identified in the FFPE samples was of that identified in frozen tissue and a of identifications over the was for and The of peptides identified was from FFPE samples and from the frozen the of spectral counts was and The protein were using the gene which of protein S. for gene in PubMed Scopus Google Scholar). analysis of protein numbers by analyses to gene in differences in the numbers of proteins in this analysis compared with the The overlap in protein identifications between the two was differences in the of proteins between frozen and FFPE samples of is the FFPE and frozen were from the same The of proteins among subcellular was between the FFPE and frozen tissue that the differences in proteins in the two were not a subcellular of proteins identified from frozen and FFPE were using the gene analysis A for the in proteins identified in the frozen and FFPE is that and cross-linking of peptides lysine the of peptides to in this we examined the of to peptides for the the frozen this to the of lysine and arginine among proteins and Scholar). The was for the FFPE and the difference between the frozen and FFPE was significant using a did that the peptide was in the FFPE samples in the FFPE samples 3, that the peptides in are to and cross-linking by with the difference in for frozen and FFPE the of proteins identified by peptides was in the frozen samples in the FFPE samples the identified by peptides was in the frozen FFPE these differences in peptide of the proteins in these categories were identified in the FFPE lysine arginine of the to the FFPE samples The for the in protein groups examination of the numbers of peptides identified by with the combined FFPE and frozen in this using and a of protein groups were of which were to the FFPE and were to the overlap of 92% and analysis at least two distinct peptides to a protein proteins with a peptide in the FFPE were not identified the frozen and FFPE the are peptide identifications from to confident identifications. The combined from and the overlap in identifications is compared numbers of spectral counts from proteins identified in the combined proteins were for which was a confident difference in spectral counts of or between the frozen and FFPE These proteins were and a protein associated protein. of associated with the of FFPE we no of a difference in the proteins through shotgun proteomic analysis of FFPE and frozen tissue These that the of proteins using peptide separation on FFPE tissue samples the proteome in frozen The data obtained from LC-MS/MS analysis of the FFPE and frozen colon adenoma samples were combined and to analysis at and peptide identifications to a protein A in identified proteins between frozen and FFPE proteomes was proteins identified by peptides were as and and the FFPE and frozen data in to protein identification and to the FFPE and frozen specimens were from the same we overlap in the of proteins. However, this is not the and in to the 92% overlap obtained the and at least two peptides per protein Given the of the of peptides in LC-MS/MS and the that a significant of the proteins were identified by a either a or the proteome through such analysis be as of peptide identifications can for as many as of protein identifications in as be the of such analyses. work the feasibility of FFPE tissue specimens for shotgun proteome analyses the of retrospective biomarker discovery B.L. Conrads T.P. Veenstra T.D. analysis of formalin-fixed PubMed Scopus Google Scholar, H. for proteomic characterization of formalin-fixed and paraffin-embedded PubMed Scopus Google Scholar, of proteins from formalin-fixed paraffin-embedded by PubMed Scopus (125) Google Scholar, tissue a to from formalin-fixed paraffin-embedded archival 2007; PubMed Scopus Google Scholar, S. H. of protein for shotgun proteome analysis of formalin-fixed 2007; 6: PubMed Scopus Google Scholar, S.N. for the proteomic analysis of and PubMed Scopus Google Scholar, from paraffin-embedded tissue by mass Histochem. Cytochem. 2006; PubMed Scopus Google Scholar). Although this of work indicates that FFPE tissue samples protein these not whether shotgun proteome inventories the of proteins identified in analyses of fresh frozen tissues. data that shotgun proteome analyses of FFPE tissues proteome inventories that are equivalent to from frozen in numbers of proteins identified and in of the protein in FFPE tissues modifications to formaldehyde with lysine groups and protein after and rehydration, these cross-linked proteins are with under mild used for fresh tissues and for that proteome inventories of fresh and FFPE tissues the same a of C-terminal proteins Comparison of identifications based on gene ontology categories and indicates no or bias in identifications from FFPE tissues. the of work is that proteome analyses of FFPE and frozen tissues the same proteomic of the of the under is this to protein identifications that the of proteomic analysis of FFPE tissues for biomarker discovery The of formaldehyde with proteins be to protein lysine and the of fixation are in FFPE peptide inventories a of C-terminal lysine peptides and FFPE samples yielded spectral counts proteins identified by or peptides in the for for and peptides with frozen These differences in in protein group identifications the FFPE and frozen are in we did not peptide modifications with no MS/MS for modifications in the FFPE for modifications in a modifications as modifications to lysine in and in of data for proteins identified in the that that to of peptides as the of from with formaldehyde on to cross-linked that are not to identification using or However, we protein was between FFPE and frozen the of peptides the was not A of during the of peptides from the FFPE samples be but analyses demonstrate that proteome inventories for frozen and FFPE samples are equivalent at the of protein However, we did not to the posttranslational status of proteins in FFPE specimens using for modifications among the proteins. These analyses that modifications were not analysis of for the of which we did not in analyses of frozen been detected to the of peptides to Although we the analyses of in FFPE tissues from to that analyses in FFPE be by during the fixation process and to of and the of using with FFPE in is the impact of for data analysis on the of overlap between frozen and FFPE protein identifications. analysis of the data of the data from frozen and FFPE Comparison of the identification indicated a significant in protein group identifications which that application of a analysis to the two yielded proteome However, this was by the FFPE and frozen tissue and the identified protein A to either FFPE or frozen tissue a on protein identifications to the for two peptide identifications per protein in of the FFPE and frozen tissue a protein is identified by two peptides in the frozen tissue but by one peptide in the FFPE tissue that protein is to the frozen tissue the protein are the combined are peptide identifications of proteins from one can be by identification of one or peptides in the other The combined for the same protein in and represents the overlap of protein identifications between frozen and FFPE tissues at over is distinct from the of whether to protein identifications. the for protein identification by at least two peptide identifications is the overlap between FFPE and frozen tissues even the for peptide identification is of work with FFPE tissues is the of protein a for protein analysis of FFPE tissues using the acid and used this to the of protein a colon adenoma FFPE of and yielded µg of protein. FFPE samples corresponding to 200 µg of which is equivalent to of fresh or frozen These samples yielded protein group identifications in reverse phase LC-MS/MS analyses and protein group identifications in phase LC-MS/MS analyses. work with B.L. Conrads T.P. Veenstra T.D. analysis of formalin-fixed PubMed Scopus Google identified proteins with at least two peptide identifications from of gene ontology categories by from FFPE and frozen tissue specimens little or no bias in analysis of FFPE we that the equivalence of FFPE and frozen tissues to specimens. of with FFPE tissues is the of these specimens with clinical to or to these after the proteins in archival FFPE samples be that analyses provide of the proteome at the the tissue was that storage of FFPE colon adenoma samples did not the identification of proteins in these specimens. the other fixation for 2 days did protein identifications. These provide that of archival FFPE specimens for retrospective is these we that the and of archival FFPE specimens and with to work the work on shotgun proteomics with FFPE tissue specimens and the of FFPE samples with corresponding fresh frozen is a critical in in retrospective biomarker discovery in the for with