Histochemical Journal 27, 204-209 (1995)

An easy method for the from semi-thin sections. avidin-biotin technique

removal of Epon resin Application of the

S E R G I O V I D A L , M A T I L D E L O M B A R D E R O , P A B L O S A N C H E Z , A L B I N A R O M A N and L U C A S M O Y A

Department of Anatomy and Histology, University of Santiago, Faculty of Veterinary Sciences, 27002-Lugo, Spain

Received 26 April 1994 and in revised form 20 September 1994

Summary

A simple procedure is described for removing Epon resin from semi-thin 1 ~m sections, which permits excellent postembedding immrmohistochemical staining (avidin-biotin complex technique). The procedure was developed for the detection of growth hormone and prolactin in bovine adenohypophysis fixed with 2% paraformaldehyde and 0.5% glutaraldehyde in 0.1 M sodium cacodylate buffer pH 7.4-7.6. The results indicate that the removal of the epoxy embedding medium prior to the application of the immunohistochemical reagents was essential for the successful localization of the antigenic determinants of the two hormones. The immrmocytochemical reactivity was obtained only after treating the sections with a solution of potassium hydroxide in a mixture of absolute methyl alcohol and propylene oxide (Maxwell's solution). An enhanced immunoreactivity was obtained when this treatment was followed by an additional treatment with either 4% hydrogen peroxide or a saturated aqueous solution of sodium metaperiodate. Because of the easy preparation of the Epon removal solution and the good structural preservation without damage to the antigenic determinants, MaxwelFs solution is suggested as a good etching agent which can be used in immrmohistochemical studies on semi-thin sections with excellent results.

Introduction

The avidin-biotin-peroxidase complex (ABC) is a popular immunohistochemical method because it is often more sensitive and produces lower background staining compared with other methods (Hsu et aI., 1981). In 1982, Child and Unabia applied for the first time the ABC technique for the localization of pituitary hormones on semi-thin (0.5-1 p~m) plastic sections. Postembedding immunohistochemical staining of semi-thin plastic sections combines the advantages of high morphological resolution with the possibility of staining adjacent sections with different antibodies (Romijn et al., 1993). An additional application of this procedure is the utilization of semi-thin sections for immunohistochemical superimposition techniques (Beauvilliain et al., 1975). The superimposition techni- que consists of examining ultrathin sections adjacent to immunostained semi-thin sections.

The main problem with the application of immuno- histochemical techniques on semi-thin sections is the difficulty of removing plastic from the sections with-

out damage to tissue and antigenic determinants. Therefore, potential etching agents (for removing plastic) should be solutions capable of enhancing antigen-antibody interactions, facilitating infiltration of the immunological reagents and the partial to complete removal of plastic (Rodning et al., 1980). Saturated solutions of sodium hydroxide in absolute methanol or ethanol (sodium methoxide or ethoxide) (Lane & Europa, 1965; Erlandsen et al., 1979) have been widely used for removing the plastic from semi-thin sections allowing subsequent immunohistochemical techniques to be applied. Other etching agents have been described by Luft (1961), Imai et al. (1968), Snodgress et al. (1972) and Iwadare et al. (1990) for the removal of Epon resin from semi-thin sections without causing tissue damage, and giving excellent results using standard histological dyes (Haematoxylin and Eosin, PAS, and Giemsa's solutions). However, these agents have not been used hitherto in immunohisto- chemical studies on semi-thin sections.

We describe here the use of postembedding immunohistochemical staining (ABC) on Epon

0018-2214 �9 1995 Chapman & Hall

Immunohi s tochemis t ry on deplasticized Epon sections 205

sections after r emova l of the Epon according to the technique suggested b y Maxwell (1978) (modified f rom Snodgress et al., 1972). Excellent results were obtained for localizing bovine g rowth ho rmone (GH) and prolactin (PRL) in semi-thin plastic sections. The etching solution was easy and rapid to use compared wi th other etching agents.

Materials and methods

Bovine pituitary glands were obtained from a commercial slaughterhouse. The glands were quickly removed after sacrifice and the anterior lobes disected to improve fixation. They were fixed in 2% paraformaldehyde + 0.5% glutaralde- hyde in 0.1 M sodium cacodylate buffer pH 7.4-7.6 for three hours at 4~ After primary fixation, the specimens were cut into small pieces with a razor blade and postfixed in osmium tetroxide. The samples were dehydrated in graded ethanol solutions, including one bath of ethanol (70%) with uranyl acetate, and embedded in Epon 812. The embedded tissues were cut with a glass knife using a RMC ultramicrotome. Semi-thin sections (1 ~m-thick) were mounted on gelatin- coated glass slides.

Pretreatment of sections Before labelling, the semi-thin sections of adenohypophysis were pretreated in five different ways: (a) incubation for 15 min in a 10% solution of hydrogen peroxide; (b) incubation for 60 min in a saturated aqueous solution of sodium metaperiodate; (c) incubation for 5 min in a Maxwell's solution (2 g KoH in 10 ml absolute methyl alcohol and 5 ml propylene oxide) to remove Epon resin (Maxwell, 1978); (d) incubation for 5 min in a Maxwell's solution, followed by incubation for 5 min in 4% hydrogen peroxide, or (e) for 60 min in a saturated aqueous solution of sodium metaper- iodate. After the treatment with Maxwell's solution, all sections were rinsed in absolute methyl alcohol to remove any residual resin. Subsequently the sections were washed in running water prior to incubation in 4% H202 or saturated aqueous solution of sodium metaperiodate. Some semi-thin sections were not pretreated in order to confirm the need for etching.

Immunohistochemical labelling Before the immunohistochemical procedure, the sections were washed in Tris-buffer-saline (TBS). Subsequently, they were immunolabelled by the avidin-biotin-peroxidase com- plex (ABC) technique (Hsu et al., 1981), with minor

modifications. The sections were incubated with antiserum diluted in TBS with 0.5% bovine serum albumin (BSA) at 4~ for 18 h. Ovine anti-GH raised in rabbit (NIDDK-AFP-Anti- oGH-2) diluted 1:2000 and human anti-PRL raised in rabbit (NIDDK-AFP-55781789) diluted 1:1000 were used.

Following the incubation in primary antiserum solution, the tissue was washed in TBS and incubated for 1 h in secondary antibody biotin (Sigma), diluted 1:10, and 1 h in streptavidin peroxidase complex (Boehringer) diluted 1:1000. The final reaction was achieved by incubating the sections for 5 min in a solution of 5 mg DAB and 1 ml 1% H202 in 100 ml TRIS-buffer (pH 7.6). Subsequently, some sections were counterstained with Haematoxylin at 60~ for 30 min.

Immunocytochemical controls Controls for the immunocytochemical reactions included the following. (a) Omission of the specific antiserum. The labelling procedures were the same as those described above. (b) Pre-absorption of the specific antiserum with homologous and heterologous antigens. Absorption tests were carried out on semi-thin sections of bovine pituitary gland: 0.1 ml GH antibody (working dilution) was incubated with 5 ~ g ovine GH and 5 g human PRL; 0.1ml PRL antibody (working dilution) with I ~g porcine PRL and 1 ~g ovine GH. Pre-incubation was performed for 24 h at 4~

Results

The omission of the specific ant iserum abolished the immunolabell ing. In the pre-absorpt ion process, the addit ion of homologous antigens inhibited each immunolabel l ing in all instances and the absorpt ion with heterologous antigens did not affect the reaction.

The removal of the epoxy e m b e d m e n t resin prior to the applicat ion of the immunohis tochemical reagents was essential for the successful localization of the antigenic determinants of G H and PRL. Thus, the immunocytochemica l reactivity to G H and PRL anti- sera was obtained only after the sections were treated with Maxwel l ' s solution. No immunohis tochemical staining for G H or PRL was detectable in the sections wi th the epoxy resin intact or in sections treated only wi th either saturated aqueous sod ium metaper ioda te or 10% hydrogen peroxide.

The best results were obtained with semi-thin sections pret reated with Maxwel l ' s solution followed

Figs 1-3. Semi-thin sections (1 ~m) of bovine adenohypophysis embedded in Epon and subjected to different pretreatments before immunolabelling with antibody against growth hormone. Fig. 1. Pretreatment with Maxwell's solution followed by hydrogen peroxide. Intense immunostaining of the cytoplasm and no tissue damage were observed in the sections deplasticized by this method. Bar = 5 p~m. Fig. 2. Pretreatment only with Maxwell's solution. A weak immunostaining appeared in the sections deplasticized by this method. Bar = 5 ~m. Fig. 3. Pretreatment with Maxwell's solution followed by saturated aqueous sodium metaperiodate. An excellent preservation of the tissue and a good immunostaining (although less intense than in the sections pretreated with Maxwell's solution followed by hydrogen peroxide) were observed in the semi-thin sections deplasticized by this method. Bar = 5 p~m.

206

%

4 ~ Z

Figs 1-3. Caption on p. 205.

Immunohistochemistry on deplasticized Epon sections 207

Figs 4, 5. Semi-thin sections (1 ~m) of bovine adenohypophysis embedded in Epon and Subjected to different pretreatments before immunolabelling with antibody against prolactin. Fig. 4. Pretreatment with Maxwell's solution followed by hydrogen peroxide. As in the sections immunolabelled with anti- growth hormone, the sections pretreated with this method showed an intense and specific immunolabelling of cytoplasm in the immunopositive cells. Bar = 5 I~m. Fig. 5, Pretreatment with Maxwell's solution followed by saturated aqueous sodium metaperiodate. Specific staining in the cytoplasm of prolactin cells grouped in clusters. Bar = 5 ~m.

by treatment with 4% solution of hydrogen peroxide (Figs 1 and 4) or with a saturated aqueous solution of sodium metaperiodate (Figs 3 and 5). In both cases an excellent structural preservation and a lack of back-

ground staining were observed. However, slight differences were found between both cases concerning the intensity of the immunolabelling. A more intense immunolabelling appeared when the sections were

208 VIDAL et al.

treated with Maxwell's solution followed by 4% hydrogen peroxide.

Although the sections pretreated only with Max- well's solution and without any additional treatment appeared immunolabelled (Fig. 2), they were weaker than those in previous cases.

Discussion

Epoxy resins are very useful embedding media for electron microscopy due to their low shrinkage during polymerization and good stability under the electron beam (Hayat, 1989). However, the utilization of these resins in immunohistochemical studies raises the question of Epon removal before examination of the semi-thin sections. Sternberger (1967) has suggested that in plastic resins, the antigenic determinants of the large molecular weight antigens may not be destroyed but are trapped by the embedment matrix. Accord- ingly, an etching of the embedment should be an agent capable of rendering the antigenic determinants more accessible to the immunohistochemical reagents with- out causing morphological damage.

Our results indicate that Maxwell's solution is a useful etching agent since it permits the immunohis- tochemical localization of GH and PRL in the adenohypophysis without producing damage to the tissue. Previously, Snodgress et al. (1972) and Maxwell (1978) had demonstrated, by conventional histological techniques (Haematoxylin-Eosin, PAS, Giemsa), that this solution did not cause deterioration of the tissue. On the other hand, we also observed that hydrogen peroxide or sodium metaperiodate were unsuccessful as etching agents, because epoxy sections treated by them still contained unaffected embedding resin. Erlandsen et al. (1979) considered hydrogen peroxide as an 'etching' agent because it does not remove plastic. Our results conflict with those obtained by Baskin et al. (1979), who observed the existence of faint but detectable immunolabelling of GH and PRL cells in semi-thin sections exclusively treated with 10% H202. However, the lack of immunolabelling in semi- thin sections subjected to the same treatment has also been noted by Rodning et al. (1980). These authors attempted to localize IgA antigens in rat ileum but found that H202 was an unsuccessful etching agent. Rodning et al. (1980) concluded that the removal of the epoxy embedment was essential for the successful localization of the antigenic determinants.

We have observed that the application of either hydrogen peroxide or sodium metaperiodate to sec- tions previously treated with Maxwell's solution produced a marked increase in the immunolabelling. The effects of H202 and sodium metaperiodate on immunocytochemical staining are unclear at this stage. Their effect is probably related to the removal of osmium, as strong oxidizing agents serve to re-oxidize

reduced osmium molecules. This modification makes these molecules soluble and leads to their removal, unmasking antigens in osmium postfixed tissues (Bendayan & Zollinger, 1983). Osmium tetroxide has been universally used as a fixation procedure for ultrastructural preservation. Although osmium fixa- tion does not destroy protein antigenicity, it causes a cross-linking mechanism that inactivates the immuno- logical reativity of the antigenic determinants on the GH and PRL molecules (Baskin et al., 1979).

The application of Maxwell's solution as an etching solution for epoxy-based semi-thin sections offers two advantages in comparison to the popular etching technique based on sodium ethoxide or methoxide solutions (Lane & Europa, 1965; Erlandsen et al., 1979). First, Maxwell's solution is easy to prepare because of the rapid solubilization of its components and sec- ondly, it can be used immediately after preparation. In contrast, the sodium ethoxide and methoxide solutions of Lane & Europa (1965) and Erlandsen et al. (1979) must be left to mature for several days. The sodium methoxide method proposed by Casanova (1974) for removing Araldite from semi-thin sections uses a strong exothermic reaction when metallic sodium is added to absolute methanol. Currently this etching agent is unpopular because of its hazardous nature, despite the advantage of it being useable immediately.

In conclusion, Maxwell's solution is a good etching agent which can be used with excellent results for immunohistochemical studies on semi-thin sections. It is easy to prepare and provides good structural preservation without damage to antigenic determi- nants.

Acknowledgements

The authors thank NIDDK for the generous donation of ovine-GH and human-PRL antisera.

References

BASKIN, D. G., ERLANDSEN, S. L. & PARSONS, J. A. (1979) Influence of hydrogen peroxide or alcoholic sodium hydroxide on the immunocytochemical detection of growth hormone and prolacfin after osmium fixation. J. Histochem. Cytochem. 27, 1290-2.

BENDAYAN, M. & ZOLLINGER, M. (1983) Ultrastructural localization of antigenic sites on osmium-fixed tissues applying the protein A-gold technique. J. Histochem. Cytochem. 31, 101-9.

BEAUVILLIAIN J. C., TRAMU G. & DUBOIS M. P. (1975) Characterization by different techniques of adrenocor- ticotropin and gonadotropin producing cells in Lerot pituitary (Eliomys quercinus). A superimposition techni- que and an immunocytochemical technique. Cell Tissue Res. 158, 301-17.

Immunohis tochemistry on deplasticized Epon sections 209

CASANOVA, P. (1974) Techniques de coloration des tissus osmids et inclus dans l'araldite ou l'epon. An. Anat. Pathol. 19, 231-43.

CHILD, G. & UNABIA, G. (1982) Appl ica t ion of the a v i d i n -

biotin-peroxidase complex (ABC) method to the light microscopic localization of pituitary hormones. J. Histochem. Cytochem. 30, 713~.

ERLANDSEN, S. L., PARSONS, J. A. & RODNING C. B. (1979) Technical parameters of immunostaining of osinicated tissue in epoxy sections. J. Histochem. Cytochem. 27, 1286-9.

HAYAT, M. A. (Ed.) (1989) In Principles and Techniques of Electron Microscopy, Biological Applications, 3rd edn. p. 93. London: MacMillan.

HSU, S. M., RAINE, L. & FANGER, H. (1981) Use of a v i d i n -

biotin-peroxidase complex (ABC) immunoperoxidase techniques. A comparison between ABC and immuno- labeled antibody (PAP) procedures. J. Histochem. Cytochem. 29, 577-80.

IMAI, Y., SUE, A. & YAMAGUCHI, A. (1968) A removing method of the resin epoxy-embedded sections for light microscopy. J. Electron Microsc. 17, 84-5.

IWADARE, T., HARADA, E., YOSHINO, S. & ARAI, T. (1990) A solution for removal of resin from epoxy sections. Stain Technol. 65, 205-9.

LANE, B. P. & EUROPA, D. L. (1965) Differential staining of

ultrathin sections of epon-embedded tissues for light microscopy. J. Histochem. Cytochem. 13, 579-82.

LUFT, J. H. (1961) Improvements in epoxy resin embedding methods. J. Biophys. Biochem. Cytol. 9, 409-14.

MAXWELL, M. H. (1978) Two rapid and simple methods used for the removal of resins from 1.0 p,m thick epoxy sections. J. Microsc. 112, 253-5.

RODNINGp CH. B. r ERLANDSEN, S. L., COULTER, H. D. & WILSON, I. D. (1980) Immunohistochemical localization of IgA antigens in sections embedded in epoxy resin. J. Histochem. Cytochem. 28, 199-205.

ROMIJN, H. J., JANSZEN, A. W. J. W., POOL, CH. W. & BUIJS, R. M. (1993) An improved immunocytochemical stain- ing method for large semi-thin plastic epon sections: applications to GABA in rat cerebral cortex. L Histochem. Cytochem. 41, 1259-65.

SNODGRESS, A. B., DORSEY, C. H., BAILEY, G. W. H. & DICKSON, L. G. (1972) Conventional histopathologic staining method compatible with epon-embedded, osmicated tissue. Lab. Invest. 26, 329-37.

STERNBERGER, L. A. (1967) Electron microscopic immuno- cytochemistry: a review. J. Histochem. Cytochem. 15, 139-59.