Reduction of High Background Staining by Heating Unfixed Mouse Skeletal Muscle Tissue Sections Allows for Detection of Thermostable Antigens With Murine Monoclonal Antibodies

Abstract

Antigen detection with indirect immunohistochemical methods is hampered by high background staining if the primary antibody is from the same species as the examined tissue. This high background can be eliminated in unfixed cryostat sections of mouse skeletal muscle by boiling sections in PBS, and several proteins including even the low abundant dystrophin protein can then be easily detected with murine monoclonal antibodies. However, not all antigens withstand the boiling procedure. Immunoreactivity of some of these antigens can be restored by subsequent washing in Triton X-100, whereas immunoreactivity of other proteins is not restored by this detergent treatment. When such thermolabile proteins are labeled with polyclonal primary antibodies followed by dichlorotriazinylaminofluorescein–conjugated secondary antibodies and boiled, the fluorescence signal persists, and sections can then be processed with a monoclonal antibody for double immunostaining of a protein unaffected by boiling. This stability of certain fluorochromes on heating can also be exploited for double immunofluorescence labeling of two different thermostable proteins with murine monoclonal antibodies as well as for combination with Y-chromosome fluorescence in situ hybridization. Our method should extend the range of monoclonal antibodies applicable to tissues derived from the same species as the monoclonal antibodies. (J Histochem Cytochem 56:969–975, 2008)

Antigen localization in tissue sections with antibodies (Abs) is hampered by high background staining if Abs and tissue sections are of the same species, probably because of detection of endogenous immunoglobulins (Igs) in tissue samples by secondary Abs (Hierck et al. 1994) or to binding of the Fc-fragment of secondary Abs to tissue components other than the FcRII/III receptor (Lu and Partridge 1998). High levels of background staining are present in skeletal muscle tissue of mdx mice, the mouse homolog of human Duchenne muscular dystrophy (DMD), which lack the 427-kDa subsarcolemmal protein dystrophin. Implantation of stem cells or muscle progenitor cells in mdx mice is an experimental approach in the therapy of DMD (Partridge et al. 1989; Irintchev et al. 1997). In implanted tissue, detection of specific markers of donor cells, such as the Y-chromosome from male donor mice, together with proteins expressed by donor cells, such as dystrophin, is required for the demonstration and quantification of implantation efficiency (Gussoni et al. 1996; Wernig et al. 2005). However, dystrophin contributed by donor cells in host mdx skeletal muscle cell membranes cannot be differentiated from the high intercellular background. There is no background staining in mouse skeletal muscle if dystrophin is detected by rabbit polyclonal Abs; however, analysis of dystrophin expression in cell transplantation studies by polyclonal anti-dystrophin Abs is complicated by the fact that a small minority of mdx skeletal muscle fibers, so-called revertant fibers, express dystrophin through exon skipping. Differentiation between human and murine dystrophin after implantation of human cells (Cooper et al. 2001; Torrente et al. 2004) or detection of the various truncated forms of dystrophin produced by innate (Lu et al. 2000) or experimentally induced exon splicing (Lu et al. 2003) can only be made with exon specific murine MAbs (Lu and Partridge 1998). Fixed-frozen or paraffin sections have a lower background than unfixed cryostat sections, but anti-dystrophin MAbs do not work well on fixed sections (Lu and Partridge 1998). Microwaving of fixed sections for background reduction (Tornehave et al. 2000) does not result in satisfactory dystrophin immunostaining (unpublished data).

Strategies for circumvention of background staining, such as direct labeling of primary Abs (Brown et al. 2004) or application of complexes of primary and secondary Abs (Hierck et al. 1994), may not be sensitive enough for detection of the low-abundant dystrophin protein. Dystrophin can be detected in mouse muscle by blocking of secondary Ab binding with papain-digested unlabeled anti-mouse Igs supplemented with Fc fragments of the same Igs (Lu and Partridge 1998), but we describe an even simpler procedure for reduction of high background in unfixed cryostat sections. Immersion of unfixed sections in boiling PBS abolished background staining, allowing for IHC detection of several proteins. The thermostability of certain fluorescent dyes could be exploited for combination of immunohistochmistry (IHC) and fluorescent in situ hybridization (FISH) or for double immunofluorescence IHC with two Abs from the same species.

Materials and Methods

Tissue Samples

Tibialis anterior muscle from mdx mice between 10 and 18 months of age (n≥100) and soleus muscle from C57BL/10 mice between 8 and 10 months of age (n=10) were removed and snap-frozen in isopentane cooled in liquid nitrogen. Biopsies from human vastus lateralis muscle (n=4) was obtained according to the protocol of the Ethics Commission of the Medical Faculty of the University of Bonn. Six-μm cryostat sections were collected on SuperFrostPlus slides (Menzel-Gläser; Braunschweig, Germany). Sections were stored at −80C.

Background Reduction

Frozen sections were air-dried for at least 30 min at room temperature. For background reduction, PBS (pH 7.4; Sigma, Dreieich, Germany) was heated in a glass beaker (150 ml) on a hot plate. Single slides were submerged obliquely in the PBS once it started to boil vigorously such that the side with the muscle sections faced the bottom of the beaker. After boiling for times ranging between 15 sec and 5 min, sections were either rinsed briefly in PBS at room temperature or washed in 1% Triton X-100 (Sigma) in PBS at room temperature for 15 min with gentle shaking before IHC.

Antibodies

All Abs used in this study have been previously characterized. The Abs and the dilutions used in this study are shown in Table 1. The polyclonal rabbit anti-dystrophin Ab p6 was a gift from T.A. Partridge (Neuromuscular Research Unit; Hammersmith Hospital, London, UK) and the polyclonal anti-laminin Ab was a gift from R. Timpl (Max Planck Institut; Munich, Germany).

Table 1.

Staining of various antigens in skeletal muscle sections after background reduction^a

Antigen	Dilution	Antibody/isotype	Host	Source	Staining
Desmin	1:50	D33/IgG1	Mouse	Dako; Hamburg, Germany	+
Dystrophin (human)	1:20	Dy8/6C5/IgG1	Mouse	Novocastra; Newcastle upon Tyne, UK	+
Dystrophin	1:40	Dy10/12B2/IgG2a	Mouse	Novocastra	+
Dystrophin	1:1000	p6	Rabbit	Gift from T.A. Partridge	+
Lamin A/C (human)	1:50	636/IgG2b	Mouse	Novocastra	−b
Laminin (fragment P1)	1:1000		Rabbit	Gift from R. Timpl	−b
Laminin (human)	1:1000	LAM-89/IgG1	Mouse	Sigma	−b
Laminin B2	1:1	D18/IgG2a	Mouse	Developmental Studies Hybridoma Bank; Iowa City, IA	−b
M-cadherin	1:50		Rabbit	Irintchev et al. 1994	+
MyoD	1:50	5.8A/IgG1	Mouse	Dako	−b
Myogenin	1:50	F5D/IgG1	Mouse	Dako	−b
Myosin (slow)	1:1	A4-951/IgG1	Mouse	Developmental Studies Hybridoma Bank	+
Neurofilament 68 kDa	1:100	NR4/IgG1	Mouse	Sigma	+
Smooth muscle actin	1:10	ASM-1/IgG2a	Mouse	Progen; Heidelberg, Germany	+
β-tubulin (neuronal Class III)	1:800	Tuj1/IgG2a	Mouse	Covance; Berkeley, CA	+

^aSections were boiled for 5 min followed by washing in Triton X-100 for 15 min.

^bNegative staining even on application of neat antibodies or sera.

Ig, immunoglobulin.

IHC

All sera and Abs were diluted in PBS, and all incubations were carried out at 37C unless otherwise stated. After background reduction for application of murine MAbs as described above, sections were blocked with normal goat serum (NGS, 1:20; Jackson ImmunoResearch Laboratories, West Grove, PA) for 15 min and additionally with 1% H₂O₂. (30 min at room temperature) for application of tyramide signal amplification (TSA). Mouse skeletal muscle sections were incubated with primary Abs for 30 min. Primary mouse MAbs were detected by rhodamine-conjugated goat anti-mouse Abs (GAMA, 1:200; Jackson ImmunoResearch Laboratories) and rabbit polyclonal Abs with dichlorotriazinylaminofluorescein (DTAF)-conjugated goat anti-rabbit Abs (1:200; Jackson ImmunoResearch Laboratories) after incubation for 20 min. The human specific MAbs (Table1) were applied to sections of human vastus lateralis muscle.

Protocols for double immunofluorescence with MAbs from the same species on sections fixed with paraformaldehyde (Suzuki et al. 2005) were adapted for unfixed cryostat sections. For double immunofluorescence staining of neurofilament (NF) and smooth muscle actin (SMA) with two murine MAbs, sections were first boiled in PBS for 1 min and incubated with anti-NF Abs followed by peroxidase-conjugated GAMA (1:200; Jackson ImmunoResearch Laboratories) for 15 min and then developed with Cy3-conjugated tyramide (1:100 in amplification solution; PerkinElmer, Boston, MA) for 10 min at room temperature. Sections were boiled again in PBS for 5 min, and anti-SMA Ab was applied, followed by rhodamine-conjugated GAMA (1:200; Jackson ImmunoResearch Laboratories) for 20 min.

Double immunofluorescence staining of laminin and dystrophin was performed by incubation of sections with polyclonal anti-laminin Abs followed by DTAF-conjugated anti-rabbit Abs (1:200; Jackson ImmunoResearch Laboratories) for 20 min. Sections were dried thoroughly, at least for 30 min at room temperature before boiling in PBS for background reduction, and dystrophin was detected with anti-dystrophin MAb as described above.

Negative controls consisted of IHC without primary Abs. MAbs without positive immunostaining (Table 1) after boiling of sections served as isotype controls, and non-immune rabbit serum served as controls for polyclonal Abs.

IHC Combined With FISH

For codetection of dystrophin and Y-chromosome, unfixed skeletal muscle sections from male mdx mice were boiled for background reduction. Dystrophin was detected with TSA by incubation with anti-dystrophin MAb for 1 hr, followed by peroxidase-conjugated GAMA for 15 min and precipitation of Cy3-tyramide (1:50 in amplification solution) for 10 min at room temperature. After dystrophin IHC, sections were fixed in a solution containing 50% isopropanol, 19% formalin, and 4.7% glacial acetic acid in distilled water for 30 min at 4C and washed in PBS for 5 min. Sections were boiled in 0.01 M citrate buffer, pH 6 for 7 min, washed for 5 min in a mixture of 2× SSC and 5 mM EDTA, and incubated with proteinase K (2 μg/ml; Sigma) in PBS for 15 min at 37C. FISH was carried out with Y-chromosome–specific DNA probe Y1 (Nishioka 1988), labeled with digoxigenin with a labeling kit (Roche Diagnostics; Mannheim, Germany) according to the manufacturer's instructions. ISH with 0.8–1.4 ng/μl of DNA probe Y1 was carried out as previously described (Irintchev et al. 1997; Wernig et al. 2005). Hybridized probes were detected by incubation with peroxidase-conjugated anti-digoxigenin Ab (1:100; Roche Diagnostics) for 4 hr at room temperature followed by biotin-conjugated tyramide solution (1:50 in amplification solution) for 5–10 min and avidin-Cy2 (1:1000; Amersham, Poole, UK) for 30 min. Nuclei were stained with Hoechst 33342 (1 μg/ml; Sigma). Controls consisted of omission of Y-chromosome DNA probe or Y-chromosome FISH carried out on skeletal muscle sections from female mdx mice.

Sections were embedded in Fluoromount G (Southern Biotech; Birmingham, AL) and viewed under a Zeiss Axioskop 2 epifluorescence microscope equipped with a digital Zeiss AxioCam HRc camera (Carl Zeiss; Jena, Germany). Images were processed with Adobe photoshop 5.0 (Adobe Systems; Tucson, AZ).

Results

Dystrophin IHC was carried out on unfixed cryostat serial sections of skeletal muscle from an mdx mouse with MAb before background reduction. The high background on incubation with secondary Ab, in this case rhodamine-conjugated GAMA, precludes identification of dystrophin positive revertant fibers with the MAb (Figure 1A). This background staining was absent if sections were boiled in PBS for 5 min before application of MAb, allowing for clear identification of dystrophin at the sarcolemma of revertant fibers (Figure 1B). Gross morphology did not seem to be altered by boiling. Dry sections heated up to 110C in an oven retained background staining (data not shown). To test whether the epitope recognized by the anti-dystrophin MAb is thermostable when present in the full-length wild-type dystrophin molecule, dystrophin detection was performed with the MAb on an unfixed cryostat section of the soleus muscle from a C57BL/10 mouse after boiling in PBS for 5 min (Figure 1C). Sarcolemmal dystrophin immunoreactivity was detected in all muscle fibers (Figure 1C). Dystrophin can be also be readily detected in human skeletal muscle sections after boiling in PBS with the MAb specific for human dystrophin (Table 1).

Figure 1.

Detection of dystrophin on an unfixed cryostat section of mdx skeletal muscle before (A) and on an adjacent section after (B) boiling in PBS with anti-dystrophin MAb followed by rhodamine-conjugated goat anti-mouse Abs (GAMA) (red). (C) Dystrophin (red) detection in wild-type mouse skeletal muscle with MAb after boiling. (D–H) Effects of boiling and Triton X-100 on patterns of dystrophin expression in serial sections of mdx skeletal muscle. (D) Dystrophin detected with polyclonal Ab before boiling (green). (E,F) Simultaneous dystrophin detection with MAb (red) (E) and polyclonal Ab (green) (F) after boiling. (G,H) Simultaneous dystrophin detection with polyclonal Ab (green) (G) and MAb (red) (H) after boiling and washing in Triton X-100. Dystrophin immunoreactivity with the polyclonal Ab is retrieved by Triton X-100. (I) Codetection of dystrophin with MAb (red) and Y-chromosome (green) by fluorescent ISH in mdx skeletal muscle. Nuclei are blue. (J) Double immunofluorescence labeling of neurofilaments (red) and smooth muscle actin (green) with primary mouse MAbs in mouse skeletal muscle. (K) Laminin detection in serial sections of mdx skeletal muscle with polyclonal primary Ab followed by dichlorotriazinylaminofluorescein (DTAF)-tyramide (green) precipitation before and (L) after boiling in PBS. (M,N) Codetection of laminin and dystrophin. (M) Laminin immunostaining with polyclonal primary Ab and DTAF-conjugated anti-rabbit Abs (green) followed by boiling of section and (N) subsequent detection of dystrophin with MAb (red) on the same section. Bar = 50 μm.

Boiling of sections could interfere with dystrophin detection through extraction of dystrophin or through limited accessibility of the antigen in an environment denatured by boiling. We therefore detected dystrophin with a polyclonal Ab on a section before boiling (Figure 1D) and on a serial section after boiling in PBS for 1 min by simultaneous incubation with the monoclonal (Figure 1E) and polyclonal Abs (Figure 1F). The patterns of dystrophin expression detected by the polyclonal Ab before (Figure 1D) and on the serial section by the MAb after boiling (Figure 1E) were identical; however, dystrophin staining with the polyclonal Ab is barely discernible after boiling (Figure 1F). Dystrophin immunostaining with the polyclonal Ab on a further serial section was restored by washing the section in 1% Triton X-100 for 15 min after boiling in PBS for 1 min (Figure 1G), whereas dystrophin staining with the MAb on the same section (Figure 1H) was unaffected by washing in Triton X-100.

Boiling periods as low as 15 sec also effectively reduced background levels; however, background levels vary between individual animals and between muscle groups in an animal, and we therefore routinely boiled sections of 6 μm thickness for 5 min. Sections were washed in Triton X-100 for 15 min after background reduction for Abs showing no or equivocal immunostaining after boiling in PBS. These conditions yielded consistent abolition of background staining even in denervated mdx muscles with intrinsically high background levels (data not shown). Secondary Ab dilutions as low as 1:40 could be used without detectable background staining. Immunostaining results for several antigens after boiling of unfixed skeletal muscle sections followed by washing in Triton X-100 are shown in Table 1.

For detection of dystrophin and the Y-chromosome by FISH on the same section, an unfixed cryostat section of a tibial muscle from a male mdx mouse was boiled, and dystrophin was detected with MAb followed by peroxidase-catalyzed precipitation of Cy3-conjugated tyramide. The section was fixed, and FISH carried out as described above. Fluorescent conjugates of tyramide are resistant to boiling, and the sequence of staining allows for the demonstration of dystrophin together with the nuclear Y-chromosome FISH signal (Figure 1I) in a revertant fiber.

The thermostability of certain fluorescent dyes can also be exploited for immunofluorescence double labeling with MAbs from the same species by multiple boiling steps. For double immunofluorescence labeling of neurofilaments and SMA, an unfixed frozen skeletal muscle section was first boiled for background reduction and sequentially incubated with mouse monoclonal anti-neurofilament Ab (subclass IgG1), peroxidase-conjugated GAMA (reacts with all mouse IgGs), and Cy3-conjugated tyramide in amplification solution. After precipitation of tyramide, sections were once again boiled for denaturation of primary and secondary Ab complexes. Sections were incubated with mouse monoclonal anti-SMA (subclass IgG2a), followed by DTAF-conjugated GAMA (reacts with all mouse IgGs). Staining patterns for neurofilaments and SMA (Figure 1J) are restricted to peripheral nerves and blood vessels walls, respectively. Repeated cycles of boiling did not alter SMA or neurofilament immunoreactivity (data not shown). M-cadherin, a marker for satellite cells and regenerating muscle fibers, is thermostable (Table 1) and can be conveniently codetected with dystrophin after just one boiling step for background reduction (data not shown).

Certain antigens, however, do not withstand boiling in PBS. Human lamin A/C, useful for tracing implanted human cells in mouse skeletal muscle, or laminin, useful for the spatial localization of antigens such as dystrophin or satellite cells in skeletal muscle fibers, showed no immunostaining after boiling. Immunoreactivity of such proteins was not restored if sections were washed in Triton X-100 after boiling in PBS (Table 1). Fixation of sections with acetone or methanol had no effect on background reduction by boiling of sections but did not result in preservation of the immunoreactivity of such thermolabile proteins. Laminin could not be detected by any of the monoclonal or polyclonal Abs listed in Table 1, even with amplification methods. Laminin immunostaining with polyclonal rabbit anti-laminin Ab and DTAF-conjugated tyramide on cryostat sections before boiling (Figure 1K) is not possible if IHC is carried out after boiling sections (Figure 1L). However, if laminin is first detected by rabbit polyclonal anti-laminin Abs followed by DTAF-conjugated anti-rabbit Abs and the sections are then boiled, the fluorescence DTAF signal of the denatured secondary Ab persists (Figure 1M), and the background reduction on boiling then allows for detection of dystrophin with anti-dystrophin MAb on the same section (Figure 1N).

Discussion

The method described above for background reduction in mouse skeletal muscle tissue for application of murine MAbs is a simple alternative to other methods (Lu and Partridge 1998; Tornehave et al. 2000; Brown et al. 2004), is quick to perform, and does not require additional reagents. Heating of sections resulted in denaturation and fixation of several proteins. Antibodies capable of reacting with their denatured target proteins, such as those suitable for Western blotting, obviously work best on boiled sections. Fixation of sections by boiling, however, also causes loss of antigenicity of several proteins, as is often the case with chemical fixatives. Solubilization of proteins from unfixed cryostat sections by heating, including those involved in the binding of anti-mouse Abs, is a major factor in loss of antigenicity. In contrast, microwaving of formalin-fixed sections for antigen retrieval results in negligible elution of proteins (Ikeda et al. 1998), and the proteins responsible for high background seem to be denatured rather than extracted by microwaving (Tornehave et al. 2000). Heat-induced masking of epitopes through alteration of conformation (Linsenmayer et al. 1986) or steric hindrance by denaturation of neighboring proteins could lead to loss of antigenicity in unfixed sections. Immunostaining of some antigens can be retrieved in unfixed sections by Triton X-100 treatment after boiling. Triton treatment may result in antigen retrieval through tissue permeabilization or through alteration of conformation of denatured proteins. Still remaining to be investigated are the effects of other detergents such as SDS (Nguyen et al. 1992; Brown et al. 1996) — and of the pH, ionic strength, and buffer composition of the solutions unfixed sections are boiled in (Shi et al. 1997) or exposed to after boiling (Boenisch 2006) — on the reduction of background staining with maximal preservation of antigenicity.

Microwave treatment of formalin-fixed sections conveniently combines background reduction with antigen retrieval (Tornehave et al. 2000), but boiling of unfixed sections in PBS is even simpler, and antigens can be rapidly screened for preservation of antigenicity after boiling. This is obviously the method of choice for formalin-sensitive heat-stable antigens such as dystrophin, and because morphology is not adversely affected by boiling, formalin fixation is unnecessary for the IHC detection of several other proteins in skeletal muscle tissue. Application of this method to unfixed tissue sections from other organs would depend on whether tissue morphology is significantly affected by boiling.

Antigenicity of thermostable proteins is not affected by multiple short boiling periods, resulting in simple protocols for double indirect immunofluorescence with secondary antibodies, coupled with thermostable fluorochromes, from the same species. Various methods of heating fixed sections for inactivation of the first pair of primary–secondary antibody complexes have been previously described (Wang and Larsson 1985; Kolodziejczyk and Baertschi 1986; Lan et al. 1995; Suzuki et al. 2005). For unfixed mouse skeletal muscle sections, the simplest approach consists of boiling sections in PBS once for background reduction and boiling sections again after application of the first round of primary and secondary antibodies in PBS (Suzuki et al. 2005).

In our experience, the thermolabile proteins are extremely sensitive to heating in PBS, because reduction of boiling times (down to 15–30 sec) or reduction in temperature (down to 90C) still abolishes their antigenicity. However, we have not been able to selectively preserve the antigenicity of a thermolabile protein with a concomitant reduction in background staining by manipulating boiling times or temperatures. Heat labile antigens can be “salvaged” if detected with polyclonal antibodies before boiling and can be combined with MAbs to heat-stable antigens for double immunofluorescence staining.

For combination of dystrophin IHC and FISH, IHC with TSA followed by fixation (Donadoni et al. 2004) resulted in dystrophin immunostaining stable enough to withstand the lengthy FISH procedure. A method for combination of dystrophin IHC with FISH using monoclonal anti-dystrophin Abs in mouse skeletal muscle tissue briefly fixed in paraformaldehyde has been previously described (Donadoni et al. 2004); however, reduction of background staining was apparently not needed in tissues from the immunodeficient mouse strain (CD-1 nude) used in that study. Our method for combining dystrophin IHC with FISH has been developed for mice strains with high background staining.

In conclusion, we described a simple, rapid, inexpensive method for reduction of background staining in unfixed mouse tissue sections with preservation of antigenicity of several proteins. Our method should facilitate the application of murine MAbs to mouse tissue sections in a variety of experimental settings (Wernig et al. 2005).