PAXgene fixation enables comprehensive metabolomic and proteomic analyses of tissue specimens by MALDI MSI

An alcohol-based non-crosslinking tissue fixative, PAXgene Tissue System, has been proposed as alternative fixation method to formalin, providing superior and morphological preservation. To date, metabolites have not been assessed in PAXgene-fixed tissues. The study focuses on a comparison between PAXgene and standard formalin fixation for metabolomic analysis by MALDI mass spectrometry imaging. Therefore, fifty-six samples from seven mice organs were fixed with PAXgene (PFPE) or formalin (FFPE), embedded in paraffin, and processed to a tissue microarray. PAXgene was able to spatially preserve metabolites in organs achieving an overlap of common metabolites ranging from 34 to 78% with FFPE. Highly similar signal intensities and visualization of molecules demonstrated negligible differences for metabolite imaging on PFPE compared to FFPE tissues. In addition, we performed proteomic analysis of intact proteins and peptides derived from enzymatic digestion. An overlap of 33 to 58% was found between FFPE and PFPE tissue samples in peptide analysis with a higher number of PFPE-specific peaks. Analysis of intact proteins achieved an overlap in the range of 0 to 28% owing to the poor detectability of cross-linked proteins in formalin-fixed tissues. Furthermore, metabolite and peptide profiles obtained from PFPE tissues were able to correctly classify organs independent of the fixation method, whereas a distinction of organs by protein profiles was only achieved by PAXgene fixation. Finally, we applied MALDI MSI to human biopsies by sequentially analyzing metabolites and peptides within the same tissue section. Concerning prospective studies, PAXgene can be used as an alternative fixative for multi-omic tissue analysis.     

Microwave stabilization versus chemical fixation. A morphometric study in glycolmethacrylate- and paraffin-embedded tissue

Summary A morphological and morphometrical study was performed on testicular cells after microwave stabilization of the tissue while immersed in phosphate buffered saline (PBS), 0.9 NaCl or Tris-HCl. Fixation in Carnoy's fluid without irradiation was chosen as a control chemical fixation method. After microwave stabilization or chemical fixation, the testes were embedded in paraffin or in plastic (glycolmethacrylate).An excellent morphology, comparable to that after chemical fixation in Carnoy's fluid, was observed in the plastic sections of tissue irradiated in PBS or NaCl, even when the sections were subsequently treated with an aggressive reagent at high temperature, required for the Feulgen reaction. The nuclear area of the microwave-stabilized Sertoli cells was 37–46% smaller in haematoxylin-eosin stained, paraffin sections in comparison with that in the glycolmethacrylate sections. The microwave-stabilized, paraffin-embedded tissue was much more vulnerable to the hot HCl treatment of the Feulgen staining than the chemically fixed tissue, resulting in an additional 10–20% decrease in nuclear size. The latter finding is particularly important for quantitative microscopy, where the Feulgen staining method is often employed.     

Computational methods and evaluation of RNA stabilization reagents for genome-wide expression studies

Gene expression studies require high quality messenger RNA (mRNA) in addition to other factors such as efficient primers and labeling reagents. To prevent RNA degradation and to improve the quality of gene array expression data, several commercial reagents have become available. We examined a conventional hot-phenol lysis method and RNA stabilization reagents, and generated comparative gene expression profiles from Escherichia coli cells grown on minimal medium. Our data indicate that certain RNA stabilization reagents induce stress responses and proper caution must be exercised during their use. We observed that the laboratory reagent (phenol/EtOH, 5:95, v/v) worked efficiently in isolating high quality mRNA and reproducibility was such that reliable gene expression profiles were generated. To assist in the analysis of gene expression data, we wrote a number of macros that use the most recent gene annotation and process data in accordance with gene function. Scripts were also written to examine the occurrence of artifacts, based on GC content, length of the individual open reading frame (ORF), its distribution on plus and minus DNA strands, and the distance from the replication origin.     

Lysozyme antigenicity and tissue fixation

Summary The preservation of lysozyme (LZM) antigenicity was studied in paraffin embedded tissue blocks. The reactivity for LZM varied with the type of tissue studied, the fixative used, the osmolarity and pH of the fixative, fixation time and temperature, and the method of dehydration. In both rat and human tissues equeous fixatives were superior to nonaqueous fixatives in retaining LZM antigenicity. Brief fixation in fixatives of low osmolarity enhanced LZM staining in the parenchymatous tissues but diminished staining in human cartilage; prolonged fixation in fixatives of high osmolarity gave opposite results. Least affected by fixation was the LZM antigenicity in the serous cells of the glands of the respiratory tract. These cells also stained most intensely for LZM of all autopsy material studied.Studies supported by grants from the Sigrid Jusélius Foundation and Finska Läkaresällskapet     

Lysozyme antigenicity and tissue fixation.

The preservation of lysozyme (LZM) antigenicity was studied in paraffin embedded tissue blocks. The reactivity for LZM varied with the type of tissue studied, the fixative used, the osmolarity and pH of the fixative, fixation time and temperature, and the method of dehydration. In both rat and human tissues aqueous fixatives were superior to nonaqueous fixatives in retaining LZM antigenicity. Brief fixation in fixatives of low osmolarity enhanced LZM staining in the parenchymatous tissues but diminished staining in human cartilage; prolonged fixation in fixatives of high osmolarity gave opposite results. Least affected by fixation was the LZM antigenicity in the serous cells of the glands of the respiratory tract. These cells also stained most intensely for LZM of all autopsy material studied.     

Thermal stabilization of collagen molecules in bone tissue

Differential thermal calorimetry (DSC) analysis of partially dehydrated bovine bone, demineralized bone and bovine tendon collagen was performed up to 300 °C to determine factors influencing stability of mineralized collagen in bone tissue. Two endothermal regions were recognized. The first, attributed to denaturation of collagen triple helix, was hydration dependent and had a peak at 155–165 °C in bone, 118–137 °C in tendon and 131–136 °C in demineralized bone. The second region extended from 245 to 290 °C in bone and from 200 to 280 °C in tendon and was connected with melting and decomposition of collagen. Differences in thermodynamic parameters between cortical and trabecular bone tissue were stated. Activation energy of collagen unfolding in native bone tissue increased with dehydration of the bone. From the results of the present study we conclude that dehydrated bone collagen is thermally very stable both in native and in demineralized bone. Presence of mineral additionally stabilizes bone tissue.     

Histological assessment of cellular half-life in tissues in vivo

The assessment of cellular half-life is of fundamental importance for cell biology and biomedicine. Here, we show that cellular half-life in tissues can be histologically measured under steady state conditions in vivo by analyzing the loss of 5-bromo-2′-deoxyuridine (BrdU)-labeled cells over time after withdrawal of long-term BrdU labeling. To achieve efficient continuous cell labeling, we implanted BrdU-containing subcutaneous slow-release pellets into 12-month-old male Fischer 344 rats, delivering BrdU at a dose of 75 mg/kg per day over 1 (n = 20) or 3 weeks (n = 20). Four to five rats each were killed directly after the labeling or 1, 3, and 7 weeks post-labeling. Cellular half-life after withdrawal of BrdU was analyzed by nonlinear regression analysis of the labeling index, using a model of one-phase exponential decay. We initially validated our technique in the duodenum, where we determined a half-life of 2.4 days for crypt cells. Next, we applied this method to other tissues, and found a half-life of 2.2 weeks for cardiac endothelial cells, and of 5–6 days for pancreatic duct cells. In conclusion, we believe that this novel approach is an important step forward in the histological assessment of cellular half-life.     

Prevention of non-specific interactions of gold-labeled reagents on tissue sections

Summary The protein A-gold technique is amongst the most useful labeling techniques available for light and electron microscopic immunolabeling. Some electron microscopic studies, however, have suggested that protein A-gold, and other protein-gold complexes as well, may bind non-specifically to certain tissue structures, particularly in skin, creating a specious pattern of labeling.We utilized the protein A-gold technique with antiserum to both involucrin and keratin under a variety of conditions to document the specificity of labeling. When the standard conditions were followed, the protein A-gold technique produces highly specific results. These conditions include: 1. the blocking of unreacted aldehyde groups by amination; 2. the blocking of non-specific binding sites on tissue sections by preincubation with inert proteins; and 3. the use of proper concentration of the protein A-gold complex. However, non-specific labeling could be produced if the three components of the standard protocol were omitted. In particular, the use of too concentrated protein A-gold lead to non-specific labeling.We report here also updated working protocols for antigen detection with protein A-gold on semithin Lowicryl K4M and paraffin sections which provide optimal staining results.Part of this work was presented at the 17th World Congress of Dermatology, Berlin (West), May 24–29, 1987     

Prevention of non-specific interactions of gold-labeled reagents on tissue sections

The protein A-gold technique is amongst the most useful labeling techniques available for light and electron microscopic immunolabeling. Some electron microscopic studies, however, have suggested that protein A-gold, and other protein-gold complexes as well, may bind non-specifically to certain tissue structures, particularly in skin, creating a specious pattern of labeling. We utilized the protein A-gold technique with antiserum to both involucrin and keratin under a variety of conditions to document the specificity of labeling. When the standard conditions were followed, the protein A-gold technique produces highly specific results. These conditions include: 1. the blocking of unreacted aldehyde groups by amination; 2. the blocking of non-specific binding sites on tissue sections by preincubation with inert proteins; and 3. the use of proper concentration of the protein A-gold complex. However, non-specific labeling could be produced if the three components of the standard protocol were omitted. In particular, the use of too concentrated protein A-gold lead to non-specific labeling. We report here also updated working protocols for antigen detection with protein A-gold on semithin Lowicryl K4M and paraffin sections which provide optimal staining results.     

Diffusion artefacts and tissue fixation in thyroperoxidase cytochemistry

Summary The distribution of endogenous peroxidase activity in rat, mouse and human thyroid follicle cells was studied with electron microscopic cytochemistry after incubation in 3-3-diaminobenzidine (DAB).In all three species enzyme activity was found at the apical plasma membrane (facing the follicle lumen) as well as in intracellular compartments. The enzyme activity in the apical plasma membrane was more sensitive to changes in fixation conditions than the activity in intracellular compartments. Under optimal conditions more than 90% of the follicle cells in normal rat thyroids displayed a cytochemical reaction at the apical plasma membrane.In all three species the reaction product at the apical plasma membrane formed a gradient which extended into the colloid which otherwise was unreactive. Evidence obtained indicated that this gradient was not due to the presence of soluble peroxidase in the lumen but most likely signified the diffusion of the reaction product from the membrane-bound enzyme.This study was supported by Grant No. 12X-537 from the Swedish Medical Research Council     

Special symposium: fixation and tissue processing models

Abstract

Fixation and processing of tissue to paraffin blocks permit thin (4-5 µm) sections of tissues to be cut. Tissues and their subcellular components and surrounding stroma are visualized by cutting thin sections and staining them histochemically or immunohistochemically and viewing the sections using a bright field microscope. During the last century, anatomists and pathologists have used fixation with 10% neutral buffered formalin (10% NBF) as the fixative of choice. Also, both human and veterinary pathologists have trained to use fixation with 10% NBF, so these professionals are reluctant to change the familiar microscopic appearance of diagnostic tissues by using different fixatives. In addition, the effects of tissue processing on the microscopic appearance of tissue essentially has been ignored in most studies. Archives of paraffin blocks of pathological tissue contain essentially paraffin blocks fixed in 10% NBF. Therefore, if retrospective studies use archival paraffin blocks to correlate the molecular features of diseases with their outcomes, the studies must be based on tissue fixed in 10% NBF. Studies of how fixation in 10% NBF interacts with histochemical and immunohistochemical staining are limited in number and most are based on relatively long fixation times (≥36 h). Currently, fixation times in 10% NBF have been reduced to <24 h. Little is known about fixation in 10% NBF and its interaction with tissue processing for any period of fixation, especially short times. Less is known about how fixation of tissues with 10% NBF interacts with more modern assays using immunohistochemistry, real time quantitative polymerise chain reaction (PCR), and techniques that depend on analysis of proteins extracted from paraffin blocks including multiplex immunoassays or mass spectrometry. In general, multiple antibody–antigen combinations are reported not to work in tissues fixed in 10% NBF, i.e., loss of immunorecognition is nearly complete for such antibody–antigen combinations as Ki67/MIB, estrogen receptor alpha (ERα) and Progesterone receptor (PR), and partial for Bcl-2. Several models have been developed to study the interactions of tissue fixation and immunorecognition, but most have viewed the problem with immunorecognition as completely caused by fixation. Also, some of the models discussed in this special symposium do not predict the effects of fixation on frozen tissues fixed in 10% NBF and not processed to paraffin blocks. This article is a brief review of issues attending the use of 10% NBF combined with tissue processing as an interrelated process to study biomarkers identified by immunohistochemistry.     

Cell and tissue fixation, 1972-1982

Conclusions There have been changes in the practice of fixation over the past 10 years. There seem to be at least two different pressures working. On the one hand, there is increasing diversity in cell biological techniques which in turn demands more variable fixation procedures. Some of these have been outlined. Some of this change in practice has percolated through to pathology where it has been found to be diagnostically useful. In surgical pathology on the other hand, there is the continuing financial pressure for more rapid through-put of specimens which includes more rapid fixation, often with the loss of potential for subsequent chemical investigations. These are the horns of the dilemma; both are wanted at the same time.It seems that there is no magical fixative in sight which will permit all investigations on all tissues. Rather, it seems that the future will hold increasing diversity in fixation procedures which will demand that practitioners be well informed as to possibilities which, hopefully, may lend to better understanding of the problems and mechanisms of fixation.     

Preservation of tissue mucins by freeze-drying and vapour fixation

Summary Histochemical studies previously undertaken showed that tissue mucins (glycoproteins) of the rat submaxillary salivary gland and other organs are preserved very satisfactorily by formaldehyde vapour treatment applied after freeze-drying of the tissue.It was thought desirable to confirm these histochemical findings by quantitative chemical data. This was performed by studying the effect of formaldehyde vapour treatment on the solubility of proteins in the freeze-dried rat submaxillary gland.Large quantities of protein (about 30 to 60 per cent of the dry weight) could be removed by aqueous extraction from the freeze-dried control samples, which had not received any formaldehyde vapour treatment, but very little protein (about 0.5 to 4 per cent of the dry weight) could be extracted from those samples which had been exposed to this vapour at 50°C for 3 hours.Each of the experiments performed confirmed this overall picture, but there were differences in the amount of protein extracted among the control samples, as well as among the formaldehyde vapour treated ones; it has been suggested that these differences were due to variations in the proportion and/or type of protein present, most probably caused by fluctuations in the content of secretory mucins.In part fulfilment for the Doctorate of Philosophy, University of London.     

Diffusion artefacts and tissue fixation in thyroperoxidase cytochemistry

The distribution of endogenous peroxidase activity in rat, mouse and human thyroid follicle cells was studied with electron microscopic cytochemistry after incubation in 3-3'-diaminobenzidine (DAB). In all three species enzyme activity was found at the apical plasma membrane (facing the follicle lumen) as well as in intracellular compartments. The enzyme activity in the apical plasma membrane was more sensitive to changes in fixation conditions than the activity in intracellular compartments. Under optimal conditions more than 90% of the follicle cells in normal rat thyroids displayed a cytochemical reaction at the apical plasma membrane. In all three species the reaction product at the apical plasma membrane formed a gradient which extended into the colloid which otherwise was unreactive. Evidence obtained indicated that this gradient was not due to the presence of soluble peroxidase in the lumen but most likely signified the diffusion of the reaction product from the membrane-bound enzyme.     

On the stabilization by fixation of configurational states in beef heart mitochondria

The energized configuration of the cristal membrane of beef heart mitochondria can be maintained only as long as oxygen is available for electron transfer. When the oxygen supply is exhausted, the membrane undergoes a transition to the nonenergized configuration. Since the exhaustion of the available oxygen supply is complete in 5–20 sec, it is impossible to apply the method of sedimenting the mitochondria prior to fixation for studying the energized configurational states of mitochondria. The direct addition of glutaraldehyde followed by osmium tetroxide to the mitochondrial suspension is the most effective way of freezing the configurational state of the cristal membrane. Fixation with glutaraldehyde appears to be complete within 1–2 sec even at 0°. Osmium tetroxide alone can also freeze the energized configuration by fixation but the concentration of the fixative is critical. The problem of capturing the configurational state applies not only to energized transitions (nonenergized to energized) but also to nonenergized transitions (orthodox to aggregated). The freezing by fixation of the cristal membrane in the aggregated configuration is best accomplished by the sequential use of glutaraldehyde and osmium tetroxide. When the levels of glutaraldehyde and osmium tetroxide are respectively too low or too high, the mitochondrion will undergo a transition from the aggregated to the orthodox configuration before fixation is complete. Light-scattering studies provide an independent method for monitoring configurational changes in mitochondria; these light-scattering measurements confirm that the conditions for fixation which lead to stabilization of the energized state as judged by electron microscopy, also show maintenance of configuration as judged by absence of light-scattering changes after the fixatives are introduced. Reagents used in negative staining will induce the geometrical form of the energized configuration of the mitochondrion even under nonenergizing conditions. These reagents are thus unsuitable for use in studies of configurational transitions in mitochondria.     

Histological assessment of intermediate- and long-term creatine monohydrate supplementation in mice and rats

Creatine monohydrate (CrM) supplementation appears to be relatively safe based on data from short-term and intermediate-term human studies and results from several therapeutic trials. The purpose of the current study was to characterize pathological changes after intermediate-term and long-term CrM supplementation in mice [healthy control and SOD1 (G93A) transgenic] and rats (prednisolone and nonprednisolone treated). Histological assessment (18-20 organs/tissues) was performed on G93A mice after 159 days, and in Sprague-Dawley rats after 365 days, of CrM supplementation (2% wt/wt) compared with control feed. Liver histology was also evaluated in CD-1 mice after 300 days of low-dose CrM supplementation (0.025 and 0.05 g x kg-1x day-1) and in Sprague-Dawley rats after 52 days of CrM supplementation (2% wt/wt) with and without prednisolone. Areas of hepatitis were observed in the livers of the CrM-supplemented G93A mice (P < 0.05), with no significant inflammatory lesions in any of the other 18-20 tissues/organs that were evaluated. The CD-1 mice also showed significant hepatic inflammatory lesions (P < 0.05), yet there was no negative effect of CrM on liver histology in the Sprague-Dawley rats after intermediate-term or long-term supplementation nor was inflammation seen in any other tissues/organs (P = not significant). Dietary CrM supplementation can induce inflammatory changes in the liver of mice, but not rats. The observed inflammatory changes in the murine liver must be considered in the evaluation of hepatic metabolism in CrM-supplemented mice. Species differences must be considered in the evaluation of toxicological and physiological studies.