Constant Detection of CD2,CD3,CD4,and CD5 in Fixed and Paraffin-embedded Tissue Using the Peroxidase-mediated Deposition of Biotin-Tyramide

Abstract

Immunohistochemical methods are widely used for diagnostic purposes in histopathology. However, the use of most monoclonal anti-leukocyte antibodies is limited to frozen tissues. Initially, it was believed that formalin fixation in particular, which is the gold standard for morphological tissue preservation, destroys most of the antigen binding sites. In recent years, protease digestion and the introduction of microwave techniques have significantly enhanced the sensitivity of immunohistochemical techniques, and a variety of hidden antigen sites in formalin-fixed tissue have been retrieved for initially unreactive antibodies. It therefore became clear that many of the leukocyte antigens are not irreversibly destroyed but are most probably masked during the fixation process. We developed a technique combining optimized pretreatment of formalin-fixed tissue with a dramatic enhancement of the immunohistochemical sensitivity and named it the ImmunoMax method. The ImmunoMax method proves that by optimizing the technique at the following three levels it is possible to detect formalin-sensitive leukocyte antigens: (a) standard fixation of the tissue; (b) sufficient antigen unmasking; and (c) increasing the substrate turnover by multiplication of binding sites with subsequent enhancement of the immunohistochemical reaction. Using this optimized ImmunoMax method, we were able to detect CD2, CD3, CD4, and CD5 with conventional monoclonal antibodies in formalin-fixed, paraffin-embedded tissue specimens of various lymphoid tissues.

Keywords

immunohistochemistry biotin-tyramide ImmunoMax paraffin CD2 CD3 CD4 CD5

Since the introduction of monoclonal antibodies, immunohistochemistry has become an important tool in research and in diagnostic pathology (Sternberger et al. 1970; Poppema et al. 1981,1987; Pallesen et al. 1983; Davey et al. 1990; Lennert and Feller 1992). It enables the detection of defined antigens on cryostat sections and paraffin-embedded tissues. However, the availability of antibodies that detect antigens in formalin-fixed tissues is still limited. Nevertheless, the “gold standard” is still to combine well-preserved morphology with immunohistochemistry. Therefore, many attempts have been made at detection of formalin-sensitive antigens in paraffin-embedded tissues.

The first steps were the use of enzyme digestion and the development of immunohistochemical techniques employing multiple secondary antibodies, both of which led to enhanced sensitivity (Mepham et al. 1979; Hsu et al. 1981; Cordell et al. 1984). The addition of microwave techniques further increased the sensitivity (Gown et al. 1993; Leong et al. 1993; Merz et al. 1993; Peränen et al. 1993; Shi et al. 1993; Kanatsuka et al. 1994; Feller et al. 1995). Various fixatives, such as glutaraldehyde, Bouin's solution, zinc-based solutions, or lysine-phosphate-buffered formalin were also tested after it became clear that it is formalin fixation that is mainly responsible for the unreactivity of many antibodies. This had led to the impression that formalin fixation induces irreversible destruction of antigen binding sites. We developed a new method, the ImmunoMax method, based on a 30–200-fold signal amplification achieved by covalent deposition of biotin molecules (Adams et al. 1992; Bobrow et al. 1992) combined with a modified antigen retrieval technique. This new method ended in a total signal amplification up to 1000–10,000-fold compared with conventional immunohistochemical techniques, which opens new perspectives for detection of other antigens in formalin-fixed tissue, e.g., the T-cell-associated antigen CD7 (Merz et al. 1995). Moreover, it became clear that at least some of the “formalin-sensitive” antigens had not been irreversibly destroyed but were only masked.

Here we describe a combination of antigen retrieval techniques and a highly sensitive immunohistochemical method for detection of the T-cell-associated antigens CD2, CD3, CD4, and CD5 using monoclonal antibodies (MAbs) that were not originally suitable for use in formalin-fixed, paraffin-embedded tissues.

Materials and Methods

Specimens of human tonsils with follicular hyperplasia (n = 10), lymph nodes (n = 10), and spleen (n = 5), and resection specimens from the ileum containing Peyer's patches (n = 5), were investigated. The tissues were routinely fixed in 4% buffered formaldehyde, pH 7.5, and subsequently embedded in paraffin. A portion of each biopsy specimen was snap-frozen and investigated in parallel.

Tissue Preparation

All tissues were fixed within 1 hr of excision. Variable fixation times were used (minimum 2 hr, maximum 72 hr). Subsequently, the specimens were dehydrated in a graded series of alcohols, cleared in xylene, and embedded in paraffin. Paraffin sections were placed on slides coated with 2% 3-(triethoxysilyl)propylamine (Merck; Darmstadt, Germany). The sections were dewaxed in xylene in a series of three steps (10 min each) and rehydrated in a series of 100%, 96%, and 70% alcohol.

Microwave Oven Treatment

A microwave oven, Miele supratronic M 752 (Miele; Stuttgart, Germany), operating at a frequency of 2.45 GHz with five power level settings, was used. Initially, the maximal power setting of 850 W was used. When the boiling point was reached, the power was reduced to 150 W for different periods of time (Table 1, operating time). Coplin jars were filled with 50 mM Tris-buffered saline (TBS), 10 mM citrate buffer, 4 M guanidine chloride (all from Sigma; Munich, Germany), and 0.1 M formic acid in 2% gelatin (Merck) (Table 2). The microwave operations with formic acid have to be carried out under a fume hood. The jars were covered with loosely fitting glass caps. Every 5 min the fluid level was checked and adjusted (TBS and citrate buffer). The treatment was stopped by intensely rinsing the slides several times in TBS. The number of jars, their position in the microwave oven, and the volume were kept constant because these factors may influence the temperature during microwave treatment.

Table 1

Comparison of CD5 (1:1000) staining after microwave pretreatment using various reagents for different operating times (ImmunoMax method) a , b

Reagent	0.5 min	1 min	2 min	3 min	5 min	10 min	15 min
TBS (50 mM, pH 7.5) c	—	—	—	—	—	—	—
Citrate buffer (10 mM, pH 6.0)	—	—	—	(+)	+	++	(+)
Guanidine chloride (4000 mM, pH 6.0)	(+)	+	++	+ d	− d	− d	− d
Formic acid (100 mM in 2% gelatin)	++	++++	++	+ d	− d	− d	− d

^aOperating at a power level of 150 W from the time when the boiling point was reached (usually after 1.5 min of heating at power level 850 W).

^b-, no specific signal; +, some T-cells slightly positive; + +, all T-cells slightly positive; +++, all T-cells moderately positive; ++++, all T-cells strongly positive.

^cTris-buffered solution.

^dBecause of the inferior morphology no incubation periods longer than 3 min were used.

Enzyme Digestion

Alternatively, the tissue sections were pretreated by protease digestion. Slides were incubated with proteinase K (Boehringer; Mannheim, Germany) 25 μg dissolved in 1 ml TBS at 37C. Protease digestion was stopped after 5 min by rinsing the sections several times in TBS.

Additional immunohistochemical stainings were performed without any procedures for antigen unmasking.

Table 2

Intensity of staining for various leukocyte antigens after microwave pretreatment using TBS, citrate buffer, guanidine chloride, and formic acid (ImmunoMax method) a , b

Reagent	Concentration (mM)	Operating time (min)	CD2 (1:500)	CD3 (1:500)	CD4 (1:500)	CD5 (1:500)
Citrate buffer	10	10	+	+	+	+
Guanidine chloride	4000	2	+	++	+	++
Formic acid	100	1	++++	++++	+++	++++

^aOperating at a power level of 150 W from the time when the boiling point was reached (usually after 1.5 min of heating at power level 850 W).

^b+, some T-cells slightly positive; ++, all T-cells slightly positive; +++, all T-cells moderately positive; ++++, all T-cells strongly positive.

Immunohistochemistry

Before incubation with the primary antibody, slides were pretreated with 0.05% H₂O₂ to block endogenous peroxidase. The primary antibodies (Table 3) were diluted in 10% normal rabbit serum. As detection system we used the ImmunoMax method recently described by Merz et al. (1995) (Table 4). It is based on the streptavidin-biotin complex (ABC) technique with an additional covalent deposition of biotin molecules near the primary antigen. As a secondary antibody, a biotinylated rabbit anti-mouse AB (Dako; Hamburg, Germany) was used, diluted in 20% human serum/RPMI (Seromed; Berlin, Germany) to absorb nonspecific binding. The streptavidin-biotin–horseradish peroxidase complex (Dako) was used as a preadjusted solution. For the next step, biotinylated tyramine was prepared as follows: 100 mg NHS-LC biotin (Pierce, Bender and Hobein; Munich, Germany) and 32 mg tyramine HCl (Sigma) were added to 50 ml boric acid buffer, pH 8.0 (Sigma) and prepared as a stock solution. This stock solution was used at a dilution of 1:50 in 50 mM TBS, 0.05% H₂O₂, pH 8.0. The slides were subsequently incubated with the streptavidin-biotin–alkaline phosphatase complex. As substrate for alkaline phosphatase, naphthol AS-BI phosphate (Sigma) and new fuchsin (Chroma; Köngen, Germany) were used. The slides were counter-stained with hematoxylin and mounted with glycerin-gelatin.

After having optimized the pretreatment parameters, we alternatively performed conventional immunohistochemical staining [APAAP and ABC techniques according to Hsu et al. (1981) and Cordell et al. (1984)] with monoclonal antibodies against CD5 on paraffin sections. In addition, cryostat sections of the same tissues were stained for CD2, CD3, CD4, and CD5 with the APAAP and ABC methods using the same monoclonal antibodies.

Table 3

Monoclonal antibodies

Antibody	Reactivity	Clone	Source
CD2 a	T-cells	IOT2a	Dianova; Hamburg, Germany
CD3 a	T-cells	IOT3a	Dianova
CD4 a	T-cells, monocytes, macrophages	IOT4b	Dianova
CD5 a	T-cells, B-cell subset	Leu-1	Becton-Dickinson; San Jose, CA
CD5 b	T-cells, B-cell subset	CD5/54/B4	Novocastra, Medac; Hamburg, Germany

^aAntibodies claimed not to be suitable for use in paraffin sections.

^bAntibody suitable for use in paraffin sections.

Negative Controls

Negative controls were performed, in which each of the reagents (antibodies, biotinylated tyramine) was omitted. In addition, specimens from kidney and liver as control tissues were used, which were expected to be negative for CD2, CD3, and CD5.

Table 4

Flow chart of the ImmunoMax method

Step	ImmunoMax method
1	Antigen retrieval
2	H₂O₂ blocking
3	Primary antibody
4	Secondary biotinylated rabbit anti-mouse antibody
5	ABC-peroxidase
6	Biotin-tyramide + H₂O₂
7	ABC-alkaline phosphatase
8	Color development

Results

The results obtained using different microwave pretreatment parameters are depicted in Tables 1, 2, 5, and 6. Table 5 also shows a comparison of staining results after microwave pretreatment, enzyme digestion, and without any pretreatment of the tissue. Without any pretreatment of the tissue, only for CD5 (Becton-Dickinson, Mountain View, CA; Novocastra, Newcastle, UK) was a weak signal obtained in a primary antibody dilution of 1:50 by use of the ImmunoMax method (Table 5). As Table 2 indicates, the optimal microwave operating time varies from reagent to reagent. Formic acid proved to be superior to citrate buffer and guanidine chloride. Using it for an operating time of 1 min resulted in high sensitivity combined with well-preserved morphology. Moreover, microwave treatment was also superior to enzyme digestion (in the form of pretreatment with 25 μg/ml proteinase K in TBS for 5 min at 37C). Different fixation times, ranging from 2 to 72 hr (2–24 hr for tissue sections with a maximal thickness of 0.5 cm), in PBS-buffered 4% formaldehyde, pH 7.5, did not have any influence on staining intensity (data not shown). Optimized fixation and microwave treatment with formic acid enabled CD5 to be detected with conventional APAAP and ABC techniques, using the primary antibody in a dilution of up to 1:50. These results were obtained equally for the Becton-Dickinson MAb (claimed not to be suitable for use in paraffin sections) and the Novocastra MAb (claimed to be suitable for use in paraffin sections). The combination of microwave pretreatment and the ImmunoMax method resulted in additional enhancement of sensitivity, which allows the primary antibody to be diluted up to 1:10,000 and more (1000-fold and more compared with conventional APAAP or ABC; Table 4). However, only with the ImmunoMax method was a constantly reliable and highly sensitive detection of CD5 possible in formalin-fixed, paraffin-embedded tissue. This was not possible with either the APAAP or the ABC technique. Paradoxically, only a low signal was recognized, or none at all, when the ImmunoMax method was used in combination with high antigen concentrations (e.g., when CD5 was used at a concentration higher than 1:100). This may be due to an effect like that demonstrated by Bigbee et al. (1977) for multistep immunohistochemical reactions in which primary, secondary, and tertiary reagents were not well tuned.

Table 5

Comparison of no pretreatment, optional enzyme digestion (proteinase K), and optimal microwave pretreatment (formic acid) using the ImmunoMax method a

Pretreatment	CD2 (1:1000)	CD3 (1:1000)	CD4 (1:1000)	CD5 (1:1000)
None	—	—	—	− b
Microwave (formic acid) c	++++	++++	++++	++++
Enzyme digestion (proteinase k) d	++	++	++	+++

^a+, some T-cells slightly positive; ++, all T-cells slightly positive; +++, all T-cells moderately positive; ++++, all T-cells strongly positive.

^bCD5 was used in a concentration of 1:50.

^c0.1 M formic acid in 2% gelatin, 1 min, operating at a power level of 150 W from the time when the boiling point was reached.

^d25 mg/ml proteinase K in TBS for 5 min at 37C.

Table 6

Comparison of CD5 staining after optimized microwave pretreatment (formic acid) by the ImmunoMax method and by conventional immunohistochemical techniques (APAAP, ABC) a , b

	Dilution
Method	1:10	1:20	1:100	1:1000	1:5000	1:10000
ImmunoMax	—	+	++++	++++	++++	+++
APAAP	+	(+)	—	—	—	—
ABC	+	(+)	—	—	—	—

^a0.1 M formic acid in 2% gelatin 1 min operating at a power level of 150 W from the time when the boiling point was reached.

^b-, no specific signal; +, some T-cells slightly positive; ++, all T-cells slightly positive; +++, all T-cells moderately positive; ++++, all T-cells strongly positive.

All snap-frozen and paraffin-embedded tissues were tested in parallel. They showed identical reaction patterns with regard to the quantity and topography of cells stained by defined antigens. This was true for lymph nodes, tonsils, and Peyer's patches of the ileum. In formalin-fixed, paraffin-embedded tissues the cytology was well preserved, enabling a detailed study of the T-cells in the different compartments. In the T-zone and in interfollicular areas, the cells that stained positive for CD2, CD3, CD4, and CD5 consisted mostly of small lymphocytes (Figure 1). In exceptional cases a blast cell was positive. In the mantle zone and in the border between mantle zone and germinal center, the CD5 MAb stained two different populations of cells (Figure 2). Some cells were strongly stained and had the morphology of small lymphocytes or pleomorphic T-cells. Other cells with almost the same nuclear irregularity showed less intense staining and were interpreted as precursor cells of the mantle zone lymphocytes. Pleomorphic T-cells were detected particularly in germinal centers by CD2 and CD3 staining (Figures 3 and Figures 4). Macrophages in the germinal centers showed weak staining for CD4 (Figure 5).

In the spleen, the staining pattern was also identical in both frozen and formalin-fixed tissues. Kidney and liver specimens as control tissues showed the expected negative reaction for all antibodies tested. No staining and enhancement was obtained in lymphatic tissue when one of the reagents (antibodies, ABC complex, or biotin-tyramide) was omitted.

Figure 1

Paraffin section of tonsil tissue stained for CD5 using the ImmunoMax technique after microwave-based pretreatment with formic acid. The distribution of the T-cells is identical to the staining pattern by ABC or APAAP on cryostat sections. Mostly small lymphocytes are stained in the interfollicular areas (i) as well as in the follicle mantle (f) and in the germinal center (g) (A, B). Bars: A = 180 μm; B = 60 μm.

Figure 2

Paraffin section of tonsil tissue stained for CD5 using the ImmunoMax technique after microwave-based pretreatment with formic acid. In the germinal center, in addition to the strongly stained small lymphocytes or pleomorphic T-cells, some cells with almost the same nuclear irregularity show less intense staining (arrows). Bar = 10 μm.

Figure 3

Paraffin section of tonsil tissue stained for CD2 using the ImmunoMax technique after microwave-based pretreatment with formic acid. In the border between mantle zone and germinal center, in addition to small T-lymphocytes some pleomorphic cells (arrows) are stained. Bar = 10 μm.

Figure 4

Paraffin section of tonsil tissue stained for CD3 using the ImmunoMax technique after microwave-based pretreatment with formic acid. In addition to small T-lymphocytes, staining shows pleomorphic cells (arrows) positive for CD3 in the border between mantle zone and germinal center (A) and in the germinal center (B). Bars: A = 12 μm; B = 10 μm.

Figure 5

Paraffin section of tonsil tissue stained for CD4 using the ImmunoMax method after microwave-based pretreatment with formic acid. In the mantle zone, mostly small lymphocytes are stained (A). Bar = 12 μm. In the germinal center, macrophages also show weak staining (B). Bar = 10 μm.

Discussion

Since the introduction of MAbs that can detect leukocyte differentiation antigens, we have gained a better understanding of the biology of lymphoid tissues (Sternberger et al. 1970; Poppema et al. 1981,1987; Norton et al. 1989; Davey et al. 1990). Nevertheless, the number of antibodies available for staining in for-malin-fixed material is still limited. Originally, it was assumed that formalin fixation destroys the vast majority of antigens. Since various new pretreatment techniques have been tested, however, it has become evident that at least some antigens are more likely masked by formalin fixation than completely destroyed (Merz et al. 1993; Peränen et al. 1993; Feller et al. 1995).

In the past, various approaches have been tested for detecting T-cell antigens in formalin-fixed, paraffin-embedded tissues. One strategy was to synthesize new MAbs by using synthetic peptide sequences, as has been described for CD8 (Mason et al. 1992).

The fact that T-cell antigens are, in principle, available in formalin-fixed tissues stimulated us to develop procedures for unmasking such antigens, combined with highly sensitive staining techniques. Our staining technique, termed the ImmunoMax method, increases the sensitivity up to about 100-fold compared to conventional ABC and APAAP methods (Merz et al. 1995). In combination with an equivalent pretreatment, the sensitivity can be increased up to 10,000-fold. Without any pretreatment of the tissue, only for CD5 was a weak signal obtained in a primary antibody concentration of 1:50 by use of the ImmunoMax method. This fact underscores the significance of well-performed antigen unmasking. Various pretreatment procedures employing chaotropic substances and different detergents have been described (Merz et al. 1993). Here, we introduce a new rapid and simple pretreatment method, in which the slides are incubated in formic acid during the microwave procedure. The advantage of this pretreatment lies in a short operating time (1 min) that does not destroy the morphology of the tissue after formalin fixation. Moreover, the sensitivity achieved in combination with the ImmunoMax method is higher than that obtained with citrate buffer or guanidine chloride. The same is true when formic acid is compared with proteinase K. The results clearly document that a number of T-cell antigens are detectable in formalin-fixed material with well-preserved morphology. The sensitivity of this technique is evident in the antibody dilution, which was as much as 10,000-fold, compared with conventional APAAP and ABC techniques.

Monoclonal and polyclonal antibodies have been described that were prepared by immunization with synthetic peptide sequences and are available for detection of CD3 and CD8 in paraffin-embedded tissue (Mason et al. 1988, 1989). Therefore, the synthesis of new antibodies using synthetic peptide sequences is a potential approach. Moreover, the polyclonal CD3 antibody also stains plasma cells and nonlymphoid cells, especially epithelial cells (Mason et al. 1988). However, the production and specificity testing of such antibodies are much more cost-intensive and time-consuming than the use of an adequate immunohistochemical technique with well-defined antibodies.

Another strategy is to improve the immunohistochemical technique by starting with optimized fixation of the tissue, in the hope of being able to detect formalin-sensitive antigens in paraffin-embedded material. Many different fixatives, such as Bouin's solution, glutaraldehyde, lysine-phosphate-buffered formalin, acetic acid, or zinc-based solutions, have been tested (Feller et al. 1995). The use of zinc-based solutions has been described to have several advantages. The morphology was quite well preserved, and the formalin-sensitive antigens CD1, CD4, CD7, CD8, and CD19 could be detected using a 10- to 20-fold dilution of the primary antibodies in a modified ABC procedure (Beckstead et al. 1994). However, the morphology is less well preserved than in tissues fixed in PBS-buffered formaldehyde. On the other hand, this method cannot be used on archival material. In recent studies, we showed that fixation in PBS-buffered 4–6% formaldehyde, pH 7.5, for 12–24 hr leads to optimal morphological preservation of the tissue and allows very high antibody dilutions to be used (Merz et al. 1993; Feller et al. 1995). This range of fixation times is an optimum for tissue blocks of maximal 0.5-cm thickness (data not shown). Furthermore, formaldehyde is a widely established fixative that is frequently used in pathological laboratories. Therefore, formalin-fixed archival material can also be used, keeping in mind that the fixation utilized may be critical. In working with such archival material, it is usually necessary to perform titration experiments with the different antibodies. After optimized antigen unmasking, we were able to detect CD5 by conventional APAAP and ABC techniques with an MAb that was not originally intended for use on paraffin-embedded tissue. However, we had to use a dilution of 1:10, which is at least 50 times higher than when it is used in frozen tissues. The combination with the ImmunoMax method dramatically increased the sensitivity, allowing a final antibody dilution of 1:1000 and greater.

The T-cell associated-antigens CD2, CD3, CD4, and CD5 were detected in paraffin-embedded specimens of tonsils, lymph nodes, spleen, and ileum with Peyer's patches. The staining patterns were identical to those seen in cryostat sections tested in parallel and to staining patterns described previously (Stein et al. 1980; Poppema et al. 1981,1987; Norton et al. 1989). The staining procedure can be used not only in lymphoma research and diagnosis but also for studying cytological details. This is best illustrated by the results of staining for CD5 and CD3. On slides stained for CD5, morphologically different lymphoid cells can be detected. In addition to small lymphocytes with round nuclei, which are most probably T-lymphocytes, slightly and more strongly stained cells can be detected in the follicular mantle and in the outer zone of the germinal center. The slightly stained CD5-positive cells exhibit irregularly shaped nuclei similar to centrocytes. It can be speculated that they are the precursor cells of mantle cell lymphoma. This view is supported by the staining for CD3. In addition to small lymphocytes with round or only slightly irregular nuclei within the germinal center, cells with irregularly shaped nuclei (pleomorphic cells) can be detected. Undoubtedly, these are pleomorphic T-cells. This has not yet been described for normal T-cells. Pleomorphic nuclei in T-cell lymphomas were believed to be more probably a sign of anaplasia.

In the future, the use of CD4 and CD8 may help us to better describe T-cell lymphomas when immunohistochemistry can be combined with well-preserved morphology. Our approach will enable other leukocyte differentiation antigens to be detected in formalin-fixed material. This study demonstrates that it is possible to detect a number of T-cell antigens with highly sensitive but conventional immunohistochemical techniques, and it can be assumed that other B-cell or myelomonocytic antigens can also be detected by the use of equivalent techniques. They can then also be studied in combination with the morphology.

It must be stressed that formalin fixation is still a critical issue in immunohistochemistry. Our experiments also show that it is necessary to standardize formalin fixation by using PBS-buffered formaldehyde with defined fixation times. Such a standardization would also increase the availability of intact DNA and, to some extent, RNA for molecular genetic studies.

In summary, we have described a three-step system of standardized fixation, optimal antigen unmasking, and highly sensitive immunohistochemistry, which is a reliable and feasible technique that can open new perspectives in diagnostic and scientific pathology.

Footnotes

Acknowledgements

Supported by Deutsche Forschungsgemeinschaft, grants Me 1037/4–1 and Me 1037/5–1.

References

Adams

(1992) Biotin amplification of biotin and horseradish peroxidase signals in histochemical stains. J Histochem Cytochem 40:1457–1463

Beckstead

(1994) A simple technique for preservation of fixation-sensitive antigens in paraffin-embedded tissues. J Histochem Cytochem 42:1127–1134

Bigbee

Kosek

Eng

(1977) Effects of primary antiserum dilution on staining of antigen rich tissues with peroxidase-anti-peroxidase techniques. J Histochem Cytochem 25:443–447

Bobrow

Litt

Shaugnessy

Mayer

Colon

(1992) The use of catalyzed reporter deposition as a means of signal amplification in a varity of formats. J Immunol Methods 150:45–49

Cordell

Falini

Erber

Ghosh

Abdulaziz

MacDonald

Pulford

KAF

Stein

Mason

(1984) Immunoenzymatic labelling of monoclonal antibodies using immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP-complexes). J Histochem Cytochem 32: 219–229

Davey

Elghetany

Kurec

(1990) Immunophenotyping of hematologic neoplasms in paraffin-embedded tissue sections. Am J Clin Pathol 93(suppl 1):17–26

Feller

Malisius

Venske

Merz

(1995) New immunohistochemical methods for the visualization of formalin-sensitive antigens in routinely processed paraffin-embedded material. Med J Kagoshima Univ 47(suppl 2): 33–38

Gown

deWever

Battifora

(1993) Microwave based antigenic unmasking. A revolutionary new technique for routine immunohistochemistry. Appl Immunohistochem 1:256–266

Hsu

Raine

Fanger

(1981) The use of avidinbiotin peroxidase complex (ABC) in immunoperoxidase techniques: comparison between ABC and unlabelled antibody (PAP) procedure. J Histochem Cytochem 29:577–583

10.

Kanatsuka

Makino

Yagui

Huang

Taira

Mikata

Yoshida

(1994) Islet amyloid polypeptide and its N-terminal and C-terminal flanking peptides' immunoreactivity in islet amyloid of diabetic patients. Diabet Res Clin Pract 26:101–107

11.

Lennert

Feller

(1992) Histopathology of Non-Hodgkin's Lymphomas. New York, Springer-Verlag

12.

Leong

Milos

(1993) An assessment of the efficacy of the microwave antigen-retrieval procedure on a range of tissue Ag. Appl Immunohistochem 1:267–274

13.

Mason

Cordell

Brown

Pallesen

Ralfkiaer

Rothbard

Crumpton

Gatter

(1989) Detection of T cells in paraffin wax embedded tissue using antibodies against a peptide sequence from the CD3 antigen. J Clin Pathol 42:1194–1200

14.

Mason

Cordell

Gaulard

Tse

AGD

Brown

(1992) Immunohistological detection of human cytotoxic/suppressor T cells using antibodies to a CD8 peptide sequence. J Clin Pathol 45:1084–1088

15.

Mason

Krissansen

Davey

Crumpton

Gatter

(1988) Antisera against epitopes resistant to denaturation on T3 (CD3) antigen can detect reactive and neoplastic T cells in paraffin embedded tissue biopsy specimens. J Clin Pathol 41:121–127

16.

Mepham

Frater

Mitchell

(1979) The Use of Proteolytic Enzymes to Improve Immunoglobulin Staining by the PAP-Technique. London, UK, Chapman and Hall

17.

Merz

Malisius

Mannweiler

Zhou

Hartmann

Orscheschek

Moubayed

Feller

(1995) ImmunoMax. A maximized immunhistochemical method for the retrieval and enhancement of hidden antigens. Lab Invest 73:149–156

18.

Merz

Rickers

Schrimel

Orscheschek

Feller

(1993) Constant detection of surface and cytoplasmatic immunglobulin heavy and light chain expression in formalin-fixed and paraffin-embedded material. J Pathol 170:257–264

19.

Norton

Isaacson

(1989) Lymphoma phenotyping in formalin-fixed and paraffin wax-embedded tissues. I. Range of antibodies and staining patterns. Histopathology 14:437–446

20.

Pallesen

Mason

Schifter

(1983) Immune marker expression in 53 lymphomas of high grade malignancy. Histopathology 7:841–857

21.

Peränen

Rikkonen

Kääriäinen

(1993) A method for exposing hidden antigenic sites in paraformaldehyde-fixed cultured cells, applied to initially unreactive antibodies. J Histochem Cytochem 41:447–454

22.

Poppema

Bhan

Reinherz

McCluskey

(1981) Distribution of T-cell subsets in human lymph nodes. J Exp Med 153:30–41

23.

Poppema

Hollema

Visser

Vos

(1987) Monoclonal antibodies (MT1, MT2, MB1, MB2, MB3) reactive with leukocyte subsets in paraffin-embedded tissue sections. Am J Pathol 127:418–429

24.

Shi

Chaiwun

Young

Cote

Taylor

(1993) Antigen retrieval technique utilizing citrate buffer or urea solution for immunohistochemical demonstration of androgen receptor in formalin-fixed paraffin sections. J Histochem Cytochem 41:1599–1604

25.

Stein

Bonk

Tolksdorf

Lennert

Rodt

Gerdes

(1980) Immunohistologic analysis of the organization of normal lymphoid tissue and non-Hodgkin's lymphomas. J Histochem Cytochem 28:746–760

26.

Sternberger

Hardy

Jr Cuculis

Meyer

(1970) The unlabeled antibody enzyme method of immunohistochemistry. Preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochetes. J Histochem Cytochem 18:315