Application of Immunohistochemical Staining to Detect Antigen Destruction as a Measure of Tissue Damage

Abstract

Electrocautery and directed energy devices (DEDs) such as lasers, which are used in surgery, result in tissue damage that cannot be readily detected by traditional histological methods, such as hematoxylin and eosin staining. Alternative staining methods, including 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) to stain live tissue, have been reported. Despite providing superior detection of damaged tissue relative to the hematoxylin and eosin (H&E) method, the MTT method possesses a number of drawbacks, most notably that it must be carried out on live tissue samples. Herein, we report the development of a novel staining method, “antigen destruction immunohistochemistry” (ADI), which can be carried out on paraffin-embedded tissue. The ADI method takes advantage of epitope loss to define the area of tissue damage and provides many of the benefits of live tissue MTT staining without the drawbacks inherent to that method. In addition, the authors provide data to support the use of antibodies directed at a number of gene products for use in animal tissue for which there are no species-specific antibodies commercially available, as well as an example of a species-specific direct antibody. Data are provided that support the use of this method in many tissue models, as well as evidence that ADI is comparable to the live tissue MTT method.

During the late 19th century, a number of dye types were used in early microscopic investigations of simple plant and tissue materials (Cook 1997). These included (1) hematoxylin and eosin (H&E) used separately, (2) picric acid and carmine used concurrently, and, finally, (3) the combination of H&E (Cook 1997; Gill 2010). The H&E staining method evolved rapidly into a traditional staining standard used for morphological study within histology and cytology settings throughout the world (Cook 1997; Shi et al. 2011; Gill 2010). This standard staining technique is well known within the histopathology and cytopathology communities. The hematoxylin stains nuclei, nucleic acids, and other cellular components (such as keratohyalin granules) blue in color, while eosin normally stains both extracellular and cytoplasmic proteins red. As such, H&E is particularly useful in identifying morphological changes valuable in the diagnosis of cancer and other disease states. There are many H&E protocols available, with selection of the most appropriate being subject to the needs and setting of the investigation (Sullivan-Brown et al. 2011).

While highly effective, H&E does have limitations; of greatest significance is its inability to differentiate between morphologically intact cells and those that are nonviable (Radosevich et al. 1985; Radosevich et al. 1993). This inability is particularly important when tissues are analyzed that have been treated with electrocautery or directed energy devices (DEDs), such as lasers. Most DEDs sources ultimately produce heat (with or without other energetic particles) to alter the tissue. With routine histological stains, it is easy to determine where the gross microscopic damage ends and seemingly normal tissue starts. However, tissue damage can occur that results in cell death without observable morphological changes that are revealed through routine histological methods.

One alternative method to the H&E method that attempts to address this lack of differentiation is the previously reported 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) method (Radosevich et al. 1993). In the MTT assay, the mitochondrial dehydrogenase enzyme in viable cells is able to convert the soluble yellow MTT solution to a purple formazan precipitate. This method allows for better definition of the exact margin where morphologically sound and viable cells border with cells that appear morphologically intact but are not viable (Radosevich et al. 1993; Maker et al. 1995). Utilizing this method for live tissue staining requires (1) the ability to perform sectioning and staining within the time window of tissue viability, typically within 12 hr of surgery, (2) skilled technicians familiar with the process, and (3) an awareness that the methodology limits the amount of live tissue that can be processed immediately after harvesting. Live tissue ex vivo must be used as rapidly as possible after harvesting. Once the tissue is no longer viable (generally about 12 hr postsurgery), it is not suited for this method.

We have attempted to address the shortcomings of both the H&E method (poor differentiation) and the MTT method (live tissue dependence/time constraints) through the development of an alternate treatment method based on immunohistochemistry (IHC). IHC takes advantage of the reactive specificity that exists between an antibody and the corresponding antigen within tissues or cells. Research and clinical laboratories rely on this relationship to rapidly and economically detect the presence or absence of numerous molecules in tissues. Visualization of antigen–antibody complexes can be achieved through a number of colorimetric reactions, typically using a fluorescently labeled antibody, or immunoperoxidase reporting systems.

When tissue is formalin fixed and paraffin embedded, this process sometimes results in the loss of antigenicity in the tissue. To overcome this problem, a number of methods have been developed to reconstitute antigenicity by gently heating the tissue sections once rehydrated in various mild salt solutions (D’Amico et al. 2009; Leong et al. 2010; Shi et al. 2011). On a molecular level, these “antigen retrieval” methods cause the proteins in the tissue to unfold and then refold with the possibility of having the once “unavailable” epitope to now be exposed and “available” for antibody binding (Fig. 1A).

Figure 1.

Antigen retrieval and antigen destruction immunohistochemistry methods. (A) Antigen retrieval of proteins unfolding and refolding within a histological setting after exposure to a controlled heat source, thereby making antigens available. (B) Antigen destruction upon exposure to heat from a directed energy device where antigens become unavailable.

In this study, we further exploited the concept of protein folding/denaturation upon exposure to a heat source. In contrast to the antigen retrieval method, however, we hypothesized that there would be epitope destruction due to the tissue being exposed to the extreme heat produced during electrocautery. Similar to that previously reported for detecting necrosis after radiofrequency ablation (Itoh et al. 2002), we refer to this new approach as the “antigen destruction immunohistochemistry” (ADI) method (Fig. 1B). Two variants of ADI are presented here. The first makes use of the natural cross reactivity of antibodies to unrelated tissue targets in those cases where there are no known antibodies available for a given model system. This improvement eliminates the need for a specific antibody and relies on the natural reactivity of a readily available assortment of antibodies. The second variant, which employs specific antibodies designed for particular model systems, is also demonstrated for ADI.

Methods and Materials

Porcine Model

Young (4–5 weeks old) live pigs (Sus scrofa Landrace Yorkshire cross, 19–40 lb) were used due to the well-recognized morphological and physiological similarities shared between porcine and human skin (Jacobi et al. 2005; Cilurzo et al. 2007; Jacobi et al. 2007; Yabuki et al. 2007; Drakaki et al. 2008). Detailed protocols were developed for the entire procedure, and these activities were conducted in an AAALAC-approved facility with Institutional Animal Care and Use Committee approval (ACC Protocols 08–206 and 10–002). General anesthesia was administered by a certified large animal veterinarian and monitored by trained anesthesia staff. Once anesthetized, four rectangular sections (approximately 80 × 60 mm) of porcine skin were marked on the hams and shoulders. Two incisions (approximately 50 mm in length) were made lengthwise on each section using a surgical grade (15c) scalpel to induce hemorrhaging. The first incision remained untreated and served as an untreated control. The second incision was cauterized using an electrosurgical unit (Valleylab Force 2 ESU, Covidien Surgical Solutions Group, Boulder, CO) set at 30 W. Following treatment, the tissue sections were immediately harvested and transported to a histology laboratory for processing and analysis.

Tissue Treatment Pathways

In an effort to determine the extent of tissue damage created by exposure to the electrocautery unit, the extracted tissue was treated via one of three pathways: (1) formalin-fixed paraffin-embedded H&E, (2) live tissue sectioning followed by MTT, or (3) the ADI method.

H&E Staining

Live tissue sections transported from the surgical facility were immediately dissected using a No. 4 surgical scalpel into pieces approximately 5 × 5 mm. These tissue specimens were placed into a 10% formalin solution for 12–24 hrs. The tissue was then removed and processed onto standard tissue cassettes, using a routine paraffin embedding protocol. The paraffin-embedded tissue cassettes were mounted onto a rotary microtome and sectioned to a thickness of approximately 7 µm. These sections were then transferred to a tissue bath set at 48C (approximately 10C below the paraffin’s melting point). Multiple tissue sections were mounted onto a precleaned glass microscope slide coated with 3-aminopropyltriethoxysilane (APES) (Thermo Scientific SuperFrost Excell, Thermo Fisher Scientific, Waltham, MA). A preprinted and serialized bar code label (Z-Xtreme 5000T, Zebra Corporation, Vernon Hills, IL), resistant to staining treatment solutions and chemicals, was applied to each slide for identification. The slides were heat treated at 80C for about 10 min, enough time to melt the wax, adhering the tissue to the glass. After deparaffinization through xylene and graded alcohols, mounted tissue sections were processed using Harris’s H&E (Sigma-Aldrich, St. Louis, MO). The slides were then processed through graded alcohols and xylene, cover-slipped, and allowed to dry at room temperature. An image of each tissue section was captured using a Zeiss Axiovert 40 CFL inverted, digital light microscope (Carl Zeiss AG, Jena, Germany). The microscope was equipped with an InfinityX-32C, 32-megapixal CCD digital video camera (Lumenera Corporation, Ottawa, Canada) with on-board digital image manipulation software.

MTT Staining

Live skin sections transported from the surgical facility were dissected using a No. 4 surgical scalpel into blocks approximately 5 × 5 mm, appropriate for mounting and use with a vibrating microtome (EM Corporation, H1200 Vibrating Microtone, Chestnut Hill, MA). Each tissue block was individually mounted on a tissue-sectioning stage using cyanoacrylate glue and kept submerged in tissue media on the vibrating microtome. The trough was continuously surrounded by an ice/H₂O bath during the sectioning process. Tissue pieces were sectioned into sections approximately 10 µm thick and mounted individually on a precleaned, glass microscope slide coated with APES (Thermo Scientific SuperFrost Excell). The mounted tissue was then treated with an MTT (Sigma-Aldrich, St. Louis, MO) solution (2 mg/mL in PBS) and cover-slipped. The MTT solution reactivity was observed periodically over the course of approximately 3 hr to ensure optimal reactivity time of the dye. At appropriate time intervals, images of each tissue section were captured on an American Optics BioStar microscope (Reichert, Inc., Depew, NY), equipped with an InfinityX-32C, 32-megapixal CCD digital video camera (Lumenera Corporation) with on-board digital image manipulation software.

ADI Staining

Tissue samples were paraffin embedded onto cassettes, sectioned into 7-µm slices, mounted onto glass slides, and bar-coded as described in the H&E section. The slides were heat treated at 80C for about 10 min, enough time to melt the wax, adhering the tissue to the glass. After deparaffinization through xylene and graded alcohols, mounted tissue sections were processed using one of two methods:

Approach 1: “Shotgun approach.”

Untreated scalpel control tissue specimens were incubated with 1 of the 39 well-characterized, commercially available primary antibodies listed in Table 1. Each antibody was tested at a concentration of 5 µg/mL in 1% bovine serum albumin (BSA)/PBS. Each slide was treated with 200 µL of primary antibody solution and incubated for 1 hr at 37C. Following washing with PBS, a VECTASTAIN® Elite ABC kit (Vector Laboratories, Inc., Burlingame, CA) was used per manufacturer protocol to treat each slide with the appropriate secondary antibody (anti-rabbit-IgG or anti-mouse-IgG) and to perform peroxidase staining. Additional specimens were stained with omission of the primary antibody to serve as negative controls; positive controls were stained with 200 µL of a biotinylated goat anti-swine IgG (H+L) antibody (Catalog Number BA-9020, Vector Laboratories, Inc.) at a concentration of 5 µg/mL in 1% BSA/PBS for 1 hr at 37C. After washing with PBS, peroxidase staining was carried out on the positive control specimens using the VECTASTAIN® Elite ABC kit per manufacturer protocol. The slides were then processed through graded alcohols and xylene, cover-slipped, and allowed to dry at room temperature. An image of each tissue section was captured using a Zeiss Axiovert 40 CFL inverted, digital light microscope (Carl Zeiss AG). The microscope was equipped with an InfinityX-32C, 32-megapixal CCD digital video camera (Lumenera Corporation) with on-board digital image manipulation software. All slides were reviewed by the study pathologist (J.A.R.), who recorded the immunohistochemical staining intensities based on the following scale: 0 = no staining, + = weak staining, ++ = moderate staining, +++ = strong staining, and ++++ = very strong staining.

Table 1.

Individual Antibodies Screened for Staining Intensity: Antibodies Tested at 5 µg in 200 µL 1% BSA/PBS

	Antibodies (Anti-Human)	Staining Intensity	Hosts	Manufacturer
1	44-3A6	+	Mouse	Antibodies Incorporated
2	ALB	+	Mouse	Santa Cruz Biotech.
3	ALDH	+++	Rabbit	Santa Cruz Biotech.
4	Ape-1	++++	Mouse	Abcam
5	ASL	++	Mouse	BD Transduction Lab.
6	ASPM	+	Rabbit	Bethyl Lab.
7	ASS	+	Mouse	BD Transduction Lab.
8	Bax	+++	Rabbit	Santa Cruz Biotech.
9	Bcl-2	++	Rabbit	Santa Cruz Biotech.
10	CAIX	++	Rabbit	Santa Cruz Biotech.
11	CD133	+++	Mouse	eBioscience Inc.
12	CD166	+	Mouse	Lifespain Biosciences
13	CD24	+	Mouse	Abcam
14	CD34	+	Mouse	Abcam
15	CD38	+	Mouse	Abcam
16	CD44	+	Mouse	Abcam
17	Cdc20	+	Rabbit	Novus Biologicals
18	Chk1	+	Rabbit	Novus Biologicals
19	Chk2	+++	Mouse	Novus Biologicals
20	Cyclin B2	+++	Mouse	Abcam
21	E2F1	+	Mouse	Abcam
22	EGFr	+	Mouse	Abcam
23	eNOS	++	Rabbit	Cell Signaling Tech.
24	ErbB-2	++++	Mouse	Abcam
25	FXYD2	+++	Mouse	Santa Cruz Biotech.
26	GST- π	++	Rabbit	NeoMarkers
27	H+/K+ATPase alpha	+	Mouse	Fitzgerald Industries
28	H+/K+ ATPase beta	+	Mouse	Fitzgerald Industries
29	HIF-1 alpha	+++	Mouse	Abcam
30	HSP27	0	Mouse	Santa Cruz Biotech.
31	HSP70	++	Mouse	Santa Cruz Biotech.
32	HSP90	++	Mouse	Santa Cruz Biotech.
33	iNOS	+	Rabbit	Abcam
34	KIAA0101	+	Mouse	Abcam
35	Large T SV40	+	Mouse	Lab Vision
36	Nitrotyrosine	++++	Mouse	Millipore
37	p53	+++	Mouse	NeoMarkers
38	PNK	+	Rabbit	Genetex Inc.
39	UBE2C	++	Mouse	Abcam

After the staining intensities of each of the 39 antibodies tested individually were reviewed, the antibodies exhibiting the most intense staining (anti-Ape-1, anti-ErbB-2 and anti-nitrotyrosine) were tested as a mixture in an effort to further increase the observed staining intensity. Equal parts of the three antibodies were mixed to prepare primary antibody solutions. Untreated scalpel control and electrocautery treated tissue specimens were treated at three concentrations of the antibody mixture (2, 5, and 10 µg/mL in 1% BSA/PBS, where concentration refers to the collective concentration of the antibodies together) and two different primary antibody incubation times (30 or 60 minutes). Secondary antibody treatment (utilizing anti-mouse-IgG), peroxidase staining, and all subsequent tissue processing and imaging were performed as described above.

Approach 2: “Host-specific approach.”

Untreated scalpel control and electrocautery tissue specimens were incubated with 200 µL of a biotinylated goat anti-swine IgG (H+L) antibody (Catalog Number BA-9020, Vector Laboratories, Inc.) at a concentration of 5 µg/mL in 1% BSA/PBS for 1 hr at 37C. Peroxidase staining and all subsequent tissue processing and imaging were performed as described above. The slides were then processed through graded alcohols and xylene, cover-slipped, and allowed to dry at room temperature. An image of each tissue section was captured using a Zeiss Axiovert 40 CFL inverted, digital light microscope (Carl Zeiss AG). The microscope was equipped with an InfinityX-32C, 32-megapixal CCD digital video camera (Lumenera) with on-board digital image manipulation software.

In the two approaches outlined in the ADI section, antigen retrieval was not used. Additionally, the antibody concentrations that were used were based on the manufacturer’s recommendations or from empirical use in our laboratory. Additional concentrations were used to test a uniform concentration range across all of the antibodies.

Results

To initially test the ADI method, 39 individual antibodies were screened in untreated scalpel control tissue specimens. The results of the initial antibody screening are show in Table 1. Among the 39 antibodies tested, 3 were found to exhibit very strong staining (anti-Ape-1, anti-ErbB-2 and anti-nitrotyrosine), 8 exhibited strong staining, 8 exhibited moderate staining, and 19 showed weak staining. Only 1 antibody, anti-HSP27, did not reveal any detectable staining.

In an effort to optimize the staining intensities observed in Table 1, mixtures containing equal parts of anti-Ape-1, anti-ErbB-2 and anti-nitrotyrosine were prepared and tested. Figure 2 shows ADI images of this antibody mixture at various concentrations and primary antibody incubation times, along with positive and negative controls, and tissue stained using traditional H&E methods (Fig. 2a). In ADI-treated tissue, if the tissue is viable, the antibody–antigen interaction can be detected using peroxidase staining, resulting in a brown color when 3,3’-diaminobenzidine (DAB) is used as the peroxidase agent (Fig. 3). In contrast, nonviable tissue appears purple because the antigen has been denatured and can no longer interact with the primary antibody. Thus, no antigen–antibody interaction is detected using the peroxidase staining, and the hematoxylin counterstain binds throughout the cell, resulting in a purple color (Fig. 3). As expected, no tissue damage was observed in any of the untreated scalpel control specimens. The intensity of the staining observed in specimens treated with the mixture of three antibodies was markedly greater than any of the specimens treated with individual antibodies. The positive control (biotinylated goat anti-swine IgG [H+L] antibody at 5 µg/mL in 1% BSA/PBS, Fig. 2b) resulted in a uniform brown color throughout the tissue, up to the edge of the scalpel cut. The intensity of the positive control staining was comparable to the antibody mixture stained at 10 µg/mL (Fig. 2f, i) and much more intense than the antibody mixture stained at 2 µg/mL (Fig. 2d, 2g) or 5 µg/mL (Fig. 2e, h). For each concentration tested, staining intensity was found to be similar regardless of the incubation time employed (30 or 60 min). As expected, the negative control (no primary antibody, Fig. 2c) resulted in a uniform purple color throughout the tissue.

Figure 2.

Hematoxylin and eosin with antigen destruction immunohistochemistry staining results of untreated scalpel control tissue. Treatment conditions (antibody mixture refers to equal parts anti-Ape-1, anti-ErbB-2, and anti-nitrotyrosine): (a) Hematoxylin and eosin, (b) positive control (BA-9020, biotinylated, anti-swine IgG [H+L] antibody at 5 µg/mL in 200 µL of 1% BSA/PBS for 60 min), (c) negative control (no primary antibody), (d) antibody mixture at 2 µg/mL for 30 min, (e) antibody mixture at 5 µg/mL for 30 min, (f) antibody mixture at 10 µg/mL for 30 min, (g) antibody mixture at 2 µg/mL for 60 min, (h) antibody mixture at 5 µg/mL for 60 min, (i) antibody mixture at 10 µg/mL for 60 min. The arrow (↓) indicates the surface of the tissue that was exposed to the scalpel blade. Scale bar = 1 mm.

Figure 3.

Antigen destruction immunohistochemistry staining method. (Top) Targeted antigen is present on the cell surface/within the cell, allowing primary antibody to bind. Secondary antibody binds to primary antibody. The antibody-antigen interaction is detected using peroxidase staining (DAB). Hematoxylin counterstain binds throughout the cell but is overwhelmed by the intense color of the peroxidase stain, resulting in a brown stain. (Bottom) Targeted antigen is denatured, thereby inhibiting primary antibody. Secondary antibody binds to primary antibody, but no antigen–antibody interaction is detected using peroxidase staining. Hematoxylin counterstain binds throughout the cell, resulting in a purple stain. Scale bar = 1 mm.

Staining results obtained for the electrocautery-treated tissue specimens are shown in Fig. 4. The H&E image (Fig. 4a) revealed a small amount of damaged tissue along the length of the treatment area. In contrast, the ADI-treated samples—the positive control (biotinylated goat anti-swine IgG [H+L] antibody at 5 µg/mL in 1% BSA/PBS, Fig. 4b) and five of the six tissues stained with the antibody mixture (Fig. 4e–i)—revealed a much deeper area of damage, as indicated by the purple color; viable tissue is indicated by the brown color. Among the specimens stained with the antibody mixture, the contrast observed between the viable (brown color) and nonviable (purple color) tissue was dependent on the primary antibody concentration and incubation time. No viable tissue was detected when the tissue was stained at 2 µg/mL for 30 min (Fig. 4d). However, when the concentration of the antibody was increased to 5 µg/mL (Fig. 4e) or 10 µg/mL (Fig. 4f), the 30-min incubation resulted in sharp contrast between the two tissue regions. While all three specimens incubated at 60 min (Fig. 4g–i) resulted in decent contrast between the two tissue regions, little difference was observed among the concentrations tested, and the intensity of the staining in these images was less than that observed for the 30-min incubation time at either 5 µg/mL (Fig. 4e) or 10 µg/mL (Fig. 4f). As expected, the negative control (no primary antibody, Fig. 4c) resulted in a uniform purple color throughout the tissue.

Figure 4.

Hematoxylin & eosin and antigen destruction immunohistochemistry staining results of electrocautery-treated tissue. Treatment conditions (antibody mixture refers to equal parts anti-Ape-1, anti-ErbB-2, and anti-nitrotyrosine): (a) hematoxylin and eosin, (b) positive control (BA-9020, biotinylated, anti-swine IgG [H+L] antibody at 5 µg/mL in 200 µL of 1% BSA/PBS for 60 min), (c) negative control (no primary antibody), (d) antibody mixture at 2 µg/mL for 30 min, (e) antibody mixture at 5 µg/mL for 30 min, (f) antibody mixture at 10 µg/mL for 30 min, (g) antibody mixture at 2 µg/mL for 60 min, (h) antibody mixture at 5 µg/mL for 60 min, (i) antibody mixture at 10 µg/mL for 60 min. The arrow (↓) indicates the surface of the tissue that was exposed to the electrocautery probe. Scale bar = 1 mm.

Figure 5 shows a direct comparison of the MTT, H&E, and IHC staining methods used to visualize untreated scalpel control tissue and the electrocautery-treated tissue. The scalpel tissue specimen stained with MTT (Fig. 5a) exhibits a purple color (indicative of viable tissue) throughout the tissue specimen. The purple color is indicative of cells containing active mitochondrial dehydrogenase that was able to reduce the MTT dye into the purple formazan product (Fig. 6). In contrast, the electrocautery-treated tissue in Fig. 5b displays a tissue zone that lacks the purple MTT stain, indicating nonviable tissue. The cells in this area had been damaged to the point where they no longer had viable mitochondrial dehydrogenase and were therefore unable to reduce the MTT to the purple formazan product (Fig. 6). Optimal MTT reactivity was usually observed between 1.5 and 3 hr; beyond 3 hr, the tissue sections typically became overdeveloped, and distinguishing viable and nonviable sections of tissue became difficult. As with the MTT method, no damage was observed in the untreated scalpel control tissue specimens using the H&E (Fig. 5c) and ADI (Fig. 5e) methods. In contrast, damage was observed in both the H&E- and ADI-treated specimens of the electrocautery-treated tissue. The damage observed in the H&E-stained electrocautery specimen (Fig. 5d) was limited to the exposed edge of the tissue; however, the ADI-stained specimen (Fig. 5f) revealed a much deeper damage area, similar to that observed in the MTT-stained tissue (Fig. 5b).

Figure 5.

Comparison of the MTT, hematoxylin and eosin (H&E), and antigen destruction immuno- histochemistry (ADI) staining methods in untreated scalpel control tissue and electrocautery-treated tissue specimens. (a) Untreated scalpel control tissue stained with the MTT method, (b) electrocautery-treated tissue stained with the MTT method, (c) untreated scalpel control tissue stained with the H&E method, (d) electrocautery-treated tissue stained with the H&E method, (e) untreated scalpel control tissue stained with the ADI method (BA-9020, biotinylated, anti-swine IgG [H+L] antibody at 5 µg/mL in 200 µL of 1% BSA/PBS for 60 min), (f) electrocautery-treated tissue stained with the ADI method (BA-9020, biotinylated, anti-swine IgG [H+L] antibody at 5 µg/mL in 200 µL of 1% BSA/PBS for 60 min). The arrow (↓) indicates the surface of the tissue that was exposed to the scalpel blade or electrocautery probe. Scale bar = 1 mm.

Figure 6.

Diagram of the MTT staining method. Scale bar = 1 mm.

Discussion

H&E staining is widely used to evaluate tissue morphology. It is the most popular tissue-staining method among histologists and has been used for decades to train clinicians, researchers, veterinary practitioners, students, and so on. While extremely popular, H&E lacks the ability to distinguish morphologically intact but metabolically nonviable tissue. Conversely, the MTT method can clearly define metabolically active tissue from non-metabolically active tissue. But this method also has a number of inherent limitations, including the requirements of live tissue, specialized equipment, and trained personnel.

Given the drawbacks of these methods, considerable effort has been made in immunohistrochemistry to enhance the antigen availability in formalin-fixed embedded tissues. During the use of these “antigen retrieval methods,” rehydrated paraffin-embedded tissues are boiled or steamed in various buffers that attempt to replicate physiological conditions. The energy produced by the heating allows proteins and other molecular components to relax their structures and unfold. When the heat source is removed, these components have the chance to refold in such a way that epitope-binding sites made inaccessible during the paraffin-embedding processing now become available again (Fig. 1a). (Proteins and other components are heated by the paraffin, thereby causing a relaxing/refolding but in a very hydrophobic environment.)

In this study, we report the development of a novel IHC method called antigen destruction immunohistochemistry (ADI) that is, in essence, the reverse of the antigen retrieval method commonly employed to correct tissue damage induced by the paraffin-embedding process. We predicted that epitopes would be destroyed by devices such as electrocautery and directed energy devices, which generate significant heat. When applied to tissue, we hypothesized that these devices would create a zone within the tissue that would be heated to a point where the morphology of the cells would still be intact, but the proteins and other molecular components that had undergone a transient folding/refolding phase due to the in situ heating may have resulted in lost epitopes (Fig. 1b).

In the present case, we employed antibodies directed at human proteins that cross-reacted to swine proteins. To reduce this idea to practice, we took two approaches: the first being a universal approach that could be used for any combination of animal and tissue damaging device (shotgun approach), and the second being specific for our porcine model system (host-specific approach).

In the shotgun approach, we tested a broad spectrum of commercially available, off-the-shelf antibodies (Table 1) to select the most reactive types against the specific porcine tissue under study. All of the antibodies tested in the current study were available because they were being used in various other investigations in the laboratory. This approach takes advantage of the natural cross-reactivity of antibodies (which may be highly selective for a given epitope in a specific species) with tissue from different species and allows one to identify the antibodies that a given laboratory may already have on hand that might provide maximum staining intensity for a given tissue type. While not all antibodies exhibit this cross-reactivity, the findings in this study suggest the ADI method has broad applicability, given that only 1 of the 39 different antibodies tested failed to produce visible staining. This finding suggests that researchers can use almost any available antibody on hand to develop other species model systems beyond the porcine model and specific antibodies studied herein.

Three antibodies tested (anti-Ape-1, anti-Erb-2, and anti-nitrotyrosine; see Table 1) resulted in extremely strong staining, and this staining intensity was further increased when these three antibodies were run in a single mixture. While the anti-nitrotyrosine reactivity is not surprising because it is not species specific, it is surprising that the anti-Ape-1 (DNA repair enzyme) and anti-ErbB-2 (proto-oncogene receptor and tyrosine kinase) were so cross-reactive. While the perceived amount of abundant staining cannot be explained (cross-reactive vs. specific binding), it would be unlikely that these two antibodies were detecting highly conserved (evolutionarily) epitopes that were also upregulated in expression.

While this initial shotgun approach allows existing antibodies to be used to develop a system to use the ADI approach for any tissue/model system, we have shown via the host-specific approach that a specific antibody directed at the animal species being studied can be used to obtain comparable results. In this particular study, we found the biotinylated goat anti-swine IgG (H+L) antibody provided comparable staining results to our best expression of cross-reactive antibodies, and determined that it was also less expensive to use this single reagent. The concept of using cross-reactive antibodies is important in any research setting where specific species-reactive agents are not available. It should be noted that the use of some of these antibodies may work through other mechanisms of antibody binding, other than the specific antigen–antibody binding relationship. The carbohydrate components expressed on these antibodies may be selectively but non-specifically causing this binding. However, for this study, the nature of the binding is not important. What is of relevance is that antibodies can be used to detect damaged tissues, regardless of how they interact with the tissue.

While Itoh et al. (2002) used antibodies directed at DNA and mitochondrial antigens, we, herein, have used a wide range of cellular antigens as targets to detect the difference between denatured and non-denatured tissue. The methods are similar in that they both use formalin-fixed, paraffin-embedded tissues without the use of antigen-retrieval protocols. The methods differ in that we have directly compared our method with the MTT method of staining live viable tissue, which is the most accurate method reported to date, and documented that the MTT method and our immunostaining are comparable. Itoh et al. provide no direct comparison and only imply that their method is valid for detecting denatured tissue. The methods also differ in the fact that we have demonstrated a universal staining method based on antibody cross-reactivity (which may be specific or not) with tissue, making this approach useful for model systems in which specific antiantibodies are not available for the animal system being studied. We also provide evidence that when tissue-specific antibodies are available, they too can be used and provide comparable results to the MTT method.

We hypothesized that the epitope region identified via the ADI method would appear similar to the loss seen in MTT-stained live viable tissue. Indeed, this is the case (Fig. 5), and we now have a method that can be used on formalin-fixed, paraffin-embedded tissue that does not have a restrictive processing time but has the same sensitivity as the MTT tissue-staining method. It is important to note that there are slight differences in the amount of damage perceived by the MTT and ADI methods due to the fact that the tissue changes size (during dehydration when processed to be paraffin embedded). Therefore, quantitative measurements of damage obtained via the ADI method will not be directly comparable to those obtained via the MTT method.

Conclusion

In conclusion, we have demonstrated the ability to visualize epitope destruction using an immunological staining method that is comparable to live tissue staining via the MTT method. The ADI method provides several advantages over the MTT and/or H&E methods: (1) the tissue is fixed in the same manner as that used in routine H&E staining, thus eliminating the need for specialized equipment; (2) the tissue can be stored almost indefinitely, later sectioned, and stained as needed; and (3) no new skills are needed to carry out ADI. The results presented herein support the use of antibodies that are directed against a number of gene products as well as species-specific antibodies, to better define the tissue margins that have been spared morphologically but are in fact tissues that have been damaged by a directed energy device. This new improved method of assessing tissue damage should have a broad applicability within research and medical communities.