Both fixation and storage time are considered to be critical steps that determine block quality and its suitability for subsequent genomic and proteomic analysis. Sample handling, including surgical procedures, anesthesia, and tissue excision, together with fixation time significantly affects the degree of sample degradation and autolysis. This preanalytical phase is critical, and should be considered independently, regardless of the preservation method (formalin fixed or snap frozen). Therefore, the duration of tissue excision or biopsies should be kept as short as possible, and exact fixation time should be used to prevent significant changes in DNA, RNA transcripts, and proteins. It is highly unlikely that any process subsequent to fixation and embedding (other than tryptic digestion and probably storage time) can alter the secondary structure of the tissue . It has been reported that the duration of fixation affects the results of FFPE proteome shotgun analysis. For instance, when the samples from the same specimen were frozen or fixed in 10% buffered formalin for 1, 2, and 4 days, proteome analysis revealed comparable results between frozen tissue and samples preserved in formaldehyde for 1 and 2 days. In contrast, samples preserved for 4 days yielded less protein (Fig. 6) . The spectral count derived from this sample was significantly lower compared to the others. In agreement with this observation, an earlier study that investigated fixation time/protein binding patterns using [14C] formaldehyde to fix rat kidney sections under different conditions showed that the amount of covalent binding between protein and formaldehyde is proportional to fixation time until approximately 37 h . That study also suggested that, since covalent binding of formaldehyde forming cross-links is a fundamental event in fixation, the appropriate duration of formaldehyde fixation is 24 h at room temperature (25°C) or 18 h at 37°C, and recommended not exceeding 48 h for fixation [1, 13, 32, 33]. Moreover, it clearly depicted the probable relevance of formalin fixation time to the degree of protein cross-linking, and consequently to any subsequent extraction of protein [1, 17]. In fact, this does not just affect FFPE proteins, fixation time would be critical in determining nucleic acid (DNA and RNA) integrity as well [34, 35]. Compared with the amount of protein required for protein profiling, only a small fragment of mRNA is required to generate an expression profile of a given gene. Therefore, degradation might not always be clearly visible when quantifying mRNA, and we would expect that genes with low expression and a short half-life might be affected dramatically . This technical point needs to be investigated more closely. Although the preservation of architecture by the formaldehyde has long been known, whether formaldehyde can preserve macromolecules (DNA, RNA, and protein) is still questionable. A recent trial aimed at investigating the changes in protein identification in FFPE blocks as a result of long-term storage. The trial revealed that there might be an archival effect on low abundance proteins quantified by having less than ten spectral counts: These proteins were more difficult to be retrieved from tissue blocks that were older than 10 years . That study utilized K-means cluster analysis to investigate the possible effect of archival time on tissue proteome analysis of FFPE samples that had been stored from 9 to 21 years. The findings showed that, in terms of distinct peptide and protein identification, slightly fewer were identified in 21-year-old blocks compared to ones stored for 9 years. The authors attributed this finding to the difficulty in retrieving proteins from aged blocks.
Figure 6. Assessment of FFPE sample fixation time, storage time, and extraction buffer variability on protein identification and yield by shotgun LC-MS/MS. (A) Effect of fixation time on the number of identified proteins. (B) Effect of storage time on the number of identified proteins. Error bars represent SD, (n = 9) for both (A) and (B). *** indicates significant different from all groups (p<0.001). (C) Electrophoretic pattern of 20-μg FFPE mouse heart tissue extracted with different extraction buffers and stained with Coomassie. (a) Tris buffer containing 1% β-octylglucoside; (b) 2% CHAPS; (c) Laemmli buffer containing 2% SDS; (d) RIPA buffer containing 2% SDS and 1% NP40; (e, h) acidic Tris buffer containing glycine and 2% SDS (e) or 0.2% Tween 20 (h); (f, g) neutral Tris buffer containing glycine, 2% SDS, and 1% NP40 (f) or 0.2% Tween 20 (g); (i, j) basic Tris buffer containing 2% SDS, 0.2% Tween 20, glycine, with DTT (i) or without DTT (j); (k) commercially FFPE extraction buffer (Qproteome FFPE tissue kit, Qiagen, Germany); (l) basic Tris buffer containing 2% SDS, 1% ß-octylglucoside, DTT, and glycine (reproduced with permission from Refs.  and ).
Download figure to PowerPoint