Detection of Apoptosis in Paraffin Embedded Tissues: the Influence of Tissue Type and Fixation

Dubská L. , E. Matalová, I . Mí‰ek: Detection of Apoptosis in Paraffin Embedded Tissues: the Influence of Tissue Type and Fixation. Acta Vet Brno 2002, 71: 529-533. With increasing importance of apoptosis, many assays have been developed to detect apoptosis in cell populations and individual cells. Apoptotic cells are often proved using TUNEL assay based on detection of DNA strand breaks that occur during apoptosis. The method enables quantification of apoptotic cells by flow cytometry, fluorescence as well as by light microscopy when modifications using marker enzymes are applied. Disadvantages of this method are false positive and negative findings. In this study, the influence of different methods of fixation on TUNEL POD results was investigated. Three commonly employed fixative procedures using formaldehyde, methacarn and Bouin’s fluid were applied. Two bovine tissue types of different cell density (thymus and ovary) were used to test this effect. Different pretreatments were tested for an influence upon false negative findings. Results were evaluated under light microscopy and photographed. We observed that, in the absence of a universal molecular marker of programmed cell death, it is essential for reliable apoptosis detection to optimize the chosen method for a given tissue and fixation. It is also strongly recommended to employ at least two methods, measuring different features of the apoptotic process to obtain reproducible results in detection of apoptosis. TUNEL assay, paraffin embedded tissues, thymus, ovary Apoptosis is a cell-suicide mechanism that plays a crucial role in development and homeostasis of metazoans (Jacobson et al. 1997). In spite of the fact that programmed cell death is currently one of the hottest areas of modern biology, reliable apoptosis evaluation in routine histological sections remains problematic. Apoptosis recognition is still commonly based on morphological criteria (Kerr et al. 1972), such as cell shrinkage, blebbing, specific chromatin condensation and loss of cell-cell contacts. The ultrastructural features of apoptosis are relatively easily recognized. A complicating factor is relatively low frequency in which apoptotic cells are observed, especially in lieu of the rapid phagocytosis of these cells in vivo (Bar 1996; Ren et Savill 1998; Holmgren et al. 1999). Associated with the nuclear condensation, the biochemical hallmark of apoptosis is an extensive cleavage of the nuclear DNA into oligonucleosome-sized fragments. These fragments can be visualized as the typical ladder pattern using gel electrophoresis (Aaij et Borst 1972) and DNA breaks can be detected in situ using enzymatic labeling system (Gavrieli et al. 1992). In the absence of a universal molecular marker of programmed cell death, these structural DNA alterations provide the basis for detection systems. TUNEL (terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling) is an often chosen method for rapid identification and quantification of apoptosis in single cells and tissue sections. Two modifications of this assay detecting apoptosis in individual cells by light microscopy have been developed: TUNEL POD and TUNEL AP using peroxidase ACTA VET. BRNO 2002, 71: 529–533 Address for correspondence: Mgr. Lenka Dubská Laboratory of Apoptosis Research Masaryk Memorial Cancer Institute Îlut ̆ kopec 7, 656 53 Brno, Czech Republic Phone: +420 543 136 730 Fax:+420 543 211 169 e-mail: dubska@mou.cz http://www.vfu.cz/acta-vet/actavet.htm

Apoptosis is a cell-suicide mechanism that plays a crucial role in development and homeostasis of metazoans (Jacobson et al. 1997).In spite of the fact that programmed cell death is currently one of the hottest areas of modern biology, reliable apoptosis evaluation in routine histological sections remains problematic.
Apoptosis recognition is still commonly based on morphological criteria (Kerr et al. 1972), such as cell shrinkage, blebbing, specific chromatin condensation and loss of cell-cell contacts.The ultrastructural features of apoptosis are relatively easily recognized.A complicating factor is relatively low frequency in which apoptotic cells are observed, especially in lieu of the rapid phagocytosis of these cells in vivo (Bar 1996;Ren et Savill 1998;Holmgren et al. 1999).Associated with the nuclear condensation, the biochemical hallmark of apoptosis is an extensive cleavage of the nuclear DNA into oligonucleosome-sized fragments.These fragments can be visualized as the typical ladder pattern using gel electrophoresis (Aaij et Borst 1972) and DNA breaks can be detected in situ using enzymatic labeling system (Gavrieli et al. 1992).In the absence of a universal molecular marker of programmed cell death, these structural DNA alterations provide the basis for detection systems.
TUNEL (terminal deoxynucleotidyl transferase-mediated dUTP nick end-labeling) is an often chosen method for rapid identification and quantification of apoptosis in single cells and tissue sections.Two modifications of this assay detecting apoptosis in individual cells by light microscopy have been developed: TUNEL POD and TUNEL AP using peroxidase and alkaline phosphatase, respectively, as marker enzymes.In this article, we report on detection of apoptotic DNA breaks in two mammalian tissues (thymus and ovary) using commercially available TUNEL POD test (TdT-mediated dUTP nick end labeling).Advantages and disadvantages of different fixation procedures of this apoptosis detection method are demonstrated and discussed.

Tissue preparation
Two different bovine tissues -thymus and ovary were obtained from cattle at the abattoir.The tissues were collected immediately after killing and placed on ice for 30 min until fixation.

Paraffin embedded tissue sections
After paraffin processing 3 µm and 4 µm, respectively, thick tissue sections were prepared using microtome SM 2000R (Leica, Germany) and mounted on poly-L-lysine-coated glass slides.The samples were deparaffinized and rehydrated through a graded series of ethanol and distilled water.

TUNEL POD assay -labeling
Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (In Situ Cell Death Detection Kit, POD) was used for proving apoptosis in individual cells.The TUNEL working procedure was carried out following the producer's directions (Roche, Germany).Endogenous peroxidase was blocked by incubation in 1.3% H 2 O 2 in PBS/ 10 min/ RT before enzymatic labeling.During the TUNEL procedure samples were washed in PBS: phosphate-buffer saline (pH 7.4): 137mM NaCl, 2.7mM KCl, 1.5mM KH 2 PO 4 , 8.1mM Na 2 HPO 4 in distilled water.The fluorescent signal conversion using anti-fluorescence antibody conjugated with peroxidase and substrate color reaction applying chromogen DAB (3,3'-diaminobenzidine tetrahydrochloride; Sigma-Aldrich, Germany) were proceeded after enzymatic labeling.

TUNEL POD assay -controls
Positive control of TUNEL POD labeling was prepared using DNase (Sigma-Aldrich, Germany) treatment.After pretreatment, histological sections were incubated with DNase (5 µg/ml) at 37 °C/ humidified chamber/ 10 min to induce DNA strand breaks.
Negative control was obtained by omitting terminal transferase from the labeling procedure (label solution only instead of TUNEL reaction mixture).

TUNEL POD assay -analysis
Counterstaining with alcian blue was performed before final analysis of apoptotic nuclei/bodies.Samples were photographed at a magnification of 400 × using Coolpix camera (Nicon, Japan) after dehydration and mounting in EUKITT.

Cell morphology
For histological analysis tissue sections were stained using hematoxilin-eosin procedure and observed by oil immersion using light microscopy at a magnification of 1000× to evaluate morphological changes characteristic of apoptosis.

Results
Possible exploitation of TUNEL POD assay for proving apoptosis in situ with regards to false positive and negative findings was tested in this study.Positive control was achieved by exogenous DNase application; negative control was without labeling enzyme (terminal deoxynucleotidyl transferase).The fixation procedure (4% formaldehyde, methacarn, Bouin's fluid), the type of tissue (thymus and ovary) and the way of pretreatment were evaluated in respect of TUNEL-positive signal and morphology of cell and tissue.Results of TUNEL POD labeling were compared and documented (Plate XVII and XVIII, Figs.1-6).

Type of tissue
In ovary, which represents the type of tissue with low cell density, the TUNEL-positive cells were scattered throughout in the 4µm sections after proteinase K pretreatment (Figs. 1,3,5).The highest rate of reaction-positive thymocytes was obtained in 3µm samples pretreated by exposure to MW irradiation followed by proteolytic digestion (Fig. 4).

Fixation
No significant difference in false positive or false negative TUNEL POD labeling results was demonstrated after short term fixation procedure in mentioned three types of fixative solution.Tissues fixed in methacarn were more sensitive to pH of the DAB reaction.The false positive findings were avoided when the color reaction was performed in pH=7.4 solution (Figs. 2b,5b).The morphology of cells and tissues was worse in samples fixed in Bouin's fluid, especially in ovary (Fig. 3).Regarding the histology together with appropriate TUNEL labeling, the best results were achieved in ovary and thymus samples after formaldehyde fixation (Fig. 1, 4).

Pretreatment
Detergent pretreatment or its combination with proteinase K was not efficient in improving permeability of cell, uniformly throughout the section.As noted above, proteolytic digestion provided low accessibility in thymic tissue (highest cell density).Only thymocytes at the margin of the lobules ("pseudolobules") were TUNEL-positive.MW irradiation (5 min/ 400 W) was sufficient for adequate TUNEL POD labeling throughout thymic sample.No increase in signal intensity and amount of TUNEL-positive cells were demonstrated in ovary sections when specimens were MW irradiated before proteinase K digestion.In addition, morphology of samples pretreated by this way was artifactually altered.

TUNEL POD assay
Different concentrations of DNase were tested (5, 50, and 500 µg/ml) for positive controls with the best results when using 5 µg/ml.
Strong nonspecific background was present when the blocking of endogenous peroxidase activity was omitted.However, application of higher peroxide concentration (e.g. higher than 1.3 %) had no discernable benefit for reducing false positives.
Throughout our experiments, performing the TUNEL reaction under parafilm resulted in decreased levels of labeling in the center of the sections.This uneven labeling was eliminated and drying was avoided when specimens were incubated, without parafilm, in a humidified chamber.

Discussion
The aim of this study was to observe the influence of tissue type and its fixation on recognition of cells with DNA fragmentation after TUNEL POD labeling in histological sections.Effect of TUNEL POD assay in two thymic and ovarian tissues under three commonly used fixatives was investigated under light microscopy at a magnification of × 400.False positive and negative findings were evaluated, along with morphology of tissue, background staining and recognition of apoptotic cells were the criteria employed.
The TUNEL technique is widely used to demonstrate apoptosis in individual cells.This method enables enzymatic labeling and detection of genomic DNA breaks and thus quantification of apoptotic cells by FACS, fluorescence or light microscopy.However, the TUNEL assay is prone to false positive and negative results.Extensive TUNEL-positive signals can be explained by nonapoptotic DNA fragmentation, which occurs in numerous biological events, including: necrosis (Ohno et al. 1998); RNA synthesis and splicing (Kockx et al. 1998); DNA repair (Eastman and Barry 1992).Nonspecific enzymatic DNA disruption and labeling might be induced by sample fixation and paraffin-embedding (Ansari et al.1993;Negoescu et al. 1998); sectioning (Sloop et al. 1999); postmortem proteolysis (W ijsman et al. 1993); the staining kinetics and the reagent concentration (Migheli et al. 1994).A goal of this experimental study was to improve the accessibility of all cells within the tissue and while avoiding the nonspecific staining.
No significant influence of fixation on the accessibility of the tissues for the TUNELstaining was noted.Our results demonstrated that short-term fixation in 4% formaldehyde is suitable for application of TUNEL method.In addition to the disadvantages of Bouin's fluid and methacarn, there is also their inadequacy for long-term fixation with respect to the histologic structures of the tissue.We recommend formaldehyde fixation for TUNEL POD labeling of tissues with low cell density (ovary) where no false positive or negative findings were detected.The susceptibility to DNA breaks for enzymatic labeling is reduced by the nuclear protein environment (Kerrigan et al. 1987) and impaired by cell fixation (Gold et al. 1994).Therefore, chemical pretreatments of the biological samples have been devised to improve TUNEL sensitivity (Negoescu et al. 1997).Nevertheless, in this study the TUNEL sensitivity was dependent mainly on the type of tissue and the adequate pretreatment.In the absence of proteolytic digestion, nuclear labeling was faint in both thymus and ovary.False negative findings especially in the high-cell-density sections were avoided by adding a procedural step of MW irradiation.Detergent permeabilisation was found inefficient in the enhancing accessibility of the cells in tissues.Even the most intensive positive signal was lost when the DNase at low concentration was applied.High concentration of the enzyme probably results in extensive DNA cleavage, which alters incorporation of the labeled deoxynucleotides.
According to our results, the effect of enzymatic labeling using TUNEL POD test is tissue and fixation type dependent.Combining other methods with TUNEL, such as microscopic evaluation of morphological changes or assessment of the apoptotic DNA fragmentation pattern can substantiate the specificity of results.To obtain reproducible results, it is highly recommended to standardize the TUNEL test using positive control of apoptosis as well as of the labeling procedure and to evaluate enough microscopic fields.It is also essential to optimize the pretreatment process, the color reaction and the counterstaining for specific fixation and also tissue type, whereas TUNEL method can be applied without an extensive modification.