0022-1514/91/$3.30 The Joumal of Histochemistry and Cytochemistry Copyright 0 1991 by The Histochemical Society, Inc.

HUI Y. L A N , ' W E 1 MU, DAVID J. NIKOLIC-PATERSON, and ROBERT C. ATKINS Department of Nepbdogy, Monad M e d i d Centre, Ckayton, Victoria, Austrdia.

Received for publication May 16, 1994 and in revised form August 23, 1994; accepted September 29, 1994 (4T3389). I

Vol. 43, No. 1, pp. 97-102, 1991 Printed in US.A.

Technical Note I

A Novel, Simple, Reliable, and Sensitive Method for Multiple Immunoenzyme Staining: Use of Microwave Oven Heating to Block Antibody Crossreactivity and Retrieve Antigens

We report a simple and reliable method for detection of two or more antigens within tissue sections by indirect im- munuenzyme staining using mouse monoclonal antibodies (MAbs). This technique involves treating sections with two 5-min miaowave (MW) Oven heatings between sequential rounds of three-layer immunoenzyme staining (mouse MAb, goat anti-mouse IgG, and mouse PAP or mouse APAAP) and color development. Discrete staining of cell surface, cy- toplasmic, and nuclear anti- was evident within individual cells. This technique has a number of advantages Over those currently available. First, MW treatment denatures bound antibody molecules, thereby completely bloddng aosreac- tivity between sequential rounds of staining. This allows the use of primary (and other) antibodies raised in the same

Introduction A technique to detect two or more antigens within the same tissue section has many applications in diagnosis and research. However, one of the main drawbacks to the use of double- or triple-antibody labeling is that the primary antibodies of interest are often raised in the same species. This produces a number of problems caused by antibody crossreactivity in the staining reaction. For example, once a section is labeled with a primary mouse MAb and a second- ary enzyme-conjugated goat anti-mouse antibody, addition of a further primary mouse MAb can be bound by the free arm of the goat anti-mouse antibody and thus produce false-positive stain- ing. Another example is double immunostaining of antigens of low abundance in which peroxidase-conjugated anti-peroxidase an- tibody complexes (PAP) or alkaline phosphatase-conjugated anti- alkaline phosphatase antibody complexes (APAAP) are used. Af- ter the fmt round of staining, addition of a further goat anti-mouse antibody binds to the mouse antibody within the deposited com- plexes, again producing false-positive staining. Therefore, detec-

Correspondence to: Dr. Hui Y. Lan, Dept. of Nephrology, Monash Medical Centre, 246 Clayton Rd., Clayton, Victoria 3168, Australia.

species and the use of a sensitive three-layer staining method. Second, antigen retrieval after MW treatment markedly in- creases the sensitivity of cytoplasmic and nuclear antigen de- tection. Third, inactivation of peroxidase and alkaline phos- phatase enzymes present in PAP and APAAP complexes prevents inappropriate color development. Finally, this method can be used in both paformaldehyde-frxed cryostat sections and formalin-fued p a d m tissue sections. In con- clusion, this is a simple, reliable, and sensitive technique that wil l be useful in many areas of diagnosis and research. (J Histochem Cpochem 4397-102, 1995) KEY WOW: Multiple immunoenzyme staining; Microwave; Anti- body denaturation; Antigen retrieval; Leukocyte; Proliferation.

tion of bound PAP or APMP by an anti-mouse reagent added dur- ing the second round of staining presents a major problem.

Currently, the main approach to circumventing problems of an- tibody crossreaction has been to conjugate one or more primary antibodies with either an enzyme or a fluorochrome (direct stain- ing) or with a tag such as biotin (indirect staining), or to raise pri- mary antibodies in different species (1-5). This has been successful in many cases but it is limited by the considerable effort required to set up each individual staining system and a lack of sensitivity because antibody-enzyme complexes cannot be used in most cases. An alternative approach to this problem was recently described by Carl et al. (6). In this technique, an irrelevant mouse MAb and an F(ab) fragment of goat anti-mouse IgG was used to block cross- reaction between two rounds of mouse MAb and fluorochrome- conjugated rabbit anti-mouse IgG labeling. In our hands, this technique was time consuming to set up and it sometimes did not completely block antibody crossreactivity in detecting antigens of high abundance. Therefore, we sought to develop a sensitive method of multiple immunostaining that incorporated antibody-enzyme complexes for maximal sensitivity, had general applicability, and that eliminated problems of antibody crossreactivity. The method described here is based on the use of microwave (Mw) oven heat-

97

by guest on July 6, 2015jhc.sagepub.comDownloaded from

LAN, MU, NKOLIC-PATERSON, A " S 98

ing to denature unfixed antibodies bound to tissues and to retrieve fixed tissue antigens.

Materials and Methods Tissue Fixation. Experimental Goodpasture's syndrome (glomerulone-

phritis and pulmonary hemorrhage) was induced in inbred male Sprague- Dawley rats as previously described (7). Briefly, 12 rats were immunized with rabbit IgG and then injected with rabbit anti-rat glomerular base- ment membrane (GBM) serum. Animals were sacrificed at Day 14 after disease induction. In addition, six normal animals were examined. Kidney, lung, and spleen were dissected from each animal, cut into small cubes, and processed according to standard protocols. Tissues fixed in 2% p d o r - maldehyde-lysine-periodate (PLP) for 4 hr were cryoprotected by three changes of 7% sucrose in PBS over a 36-hr period at 4'C and then embed- ded in Tissue-Tek OCT compound (Miles; Elkhart, IN) and stored at - 80°C. Tissues fixed in 10% neutral buffered formalin, Bouin's solution, or methyl Carnoy's solution for 4 hr were dehydrated and embedded in paraffin. In addition, tissues from acutely rejected human renal allografts were fixed in PLP or formalin and processed as above.

Rat peripheral blood mononuclear cells were prepared by Ficoll-Paque (Pharmacia; Pkataway, NJ) gradient centrifugation. Mononuclear cells were recovered from the interface, washed in PBS, re-suspended in a minimal volume of fetal calf serum (FCS), and cell spots applied to gelatin-coated glass slides. Cell spots were dried at room temperature (RT) overnight and stored in a sealed box with desiccant at -80°C.

Animal experiments were performed with the approval of the Monash Medical Centre Animal Experimentation Ethics Committee and according to NH&MRC guidelines. The use of human nephrectomy specimens for research purposes was approved by the Monash Medical Centre Human Re- search and Ethics Committee according to NH&MRC guidelines.

Antibodies. Mouse MAbs used were as follows: OX-1, anti-rat CD45 (leukocyte common antigen) (8); R73, anti-rat up T-cell receptor (non- polymorphic epitope) (9); ED1, anti-rat monocyte and macrophage cyto- plasmic antigen (10). These MAbs were used as hybridoma tissue culture supernatants. PC-10 (code no. M 879, lot no. 121), mouse MAb recogniz- ing the proliferating cell nuclear antigen (PCNA) in all vertebrate species (11); Leu 4, anti-human CD3 recognizing T-cells (12); KP1, anti-human CD68 recognizing macrophages (13); goat anti-mouse IgG mouse PAP and mouse APAAP complexes were purchased from Dako (Glostrup, Denmark). In addition, sheep anti-rabbit IgG, donkey anti-sheeplgoat IgG, and goat PAP were purchased from Silenus (Melbourne, Australia). Control anti- bodies for immunoenzyme staining included 73.5, a mouse anti-human CD45R MAb that does not recognize rat tissues, and non-immune goat serum.

Microwave Oven Heating. MW oven heating of tissue sections was per- formed as previously described (14,15). A Panasonic MW oven (model NN- 5453) with operating frequency of 2450 mHz and 800 W power output was employed. Slides were placed in a glass Coplin jar filled with 0.01 M sodium citrate buffer, pH 6.0. The Coplin jar was situated in the center of a square plastic box containing 300 ml water, which was always placed in the center of the rotating plate of the MW oven. Slides were heated for periods of 5 min at the maximal power setting (800 W). The citrate buffer reached boiling point within 140-145 sec of heating and the fluid level in the Coplin jar was topped up with water between heating periods to prevent drying of sections. In developing the staining procedure, multiple steps of 5-min MW oven heating were used, which resulted in total heating times of 5 , 10, 15, and 20 min. After MW treatment, sections were washed in PBS and labeled with MAbs as described below.

the enzymatic activity of peroxidase and alkaline phosphatase conjugates was assessed. During a single round of three-layer immunoenzyme stain- ing, MW oven heating (0, 5 , 10, 15. or 20 min) was performed (a) after labeling with the first mouse MAb only, (b) after labeling with the second- layer goat anti-mouse IgG only, or (c) after labeling with the third-layer mouse PAP or mouse APAAP only. When MW treatment followed the mouse MAb or goat anti-mouse IgG step, three-layer antibody labeling was com- pleted and developed with either DAB or Fast Blue substrate as appropri- ate. In the case of MW treatment after the third-layer antibody, two ap- proaches were used. First. sections were labeled with goat anti-mouse IgG and then mouse PAP or mouse APAAP and developed with Fast Blue or DAB substrate. Second, sections were developed with substrates with no further antibody addition to assess the effect of MW treatment on the en- zymatic activity of peroxidase and alkaline phosphatase.

To check that there was no detectable antibody crossreactivity in triple- and quadruple-stained sections, the localization (cell membrane, cytoplas- mic, GBM, or nuclear) and color of reaction products were examined under high power. This assessment was made relative to serial tissue sections that were labeled with single mouse MAbs using the same reagent for color de- velopment and by comparing sections with and without MW oven heating.

Multiple Immunoenzyme Staining of PLPfued Cryosections. Serial cryostat sections (6 pm) of PLFfixed tissue were adhered to gelatin-coated microscope slides and air-dried. Sections were pre-incubated with 10% fe- tal calf serum (FCS) and 10% normal goat serum for 10 min. Sections were drained and incubated with the first mouse MAb OX-1 (1:lOO dilution) or R73 (1:4) for 60 min, washed twice in PBS, and incubated with goat anti-mouse IgG (150) for 30 min. After two washes in PBS. sections were incubated with mouse APAAP (130) for 30 min and then developed with the Fast Blue BB base (Sigma; St Louis, MO) substrate in the presence of 2 mM levamisole. Positive cells exhibited a blue stain on the cell mem- brane. To denature all antibodies bound to the tissue, sections were micro- waved twice for 5 min as described above. Sections were washed in PBS, pre-incubated with 10% FCS and 10% goat serum for 10 min, and then incubated with the second mouse MAb ED1 (1:4). After washing in PBS, endogenous peroxidase activity was blocked by incubation in 0.3% Hz02 in methanol for 20 min. Sections were then labeled sequentially with goat anti-mouse IgG (1:50) and mouse PAP (1:50) for 30 min each and devel- oped with 3-amino-9-ethylcarbazole (AEC; Sigma) substrate as previously described (3). Positive cells exhibited a red color within the cytoplasm. The two 5-min MW oven heating treatments were repeated, sections pre- incubated as before, and then labeled with the third mouse MAb PC-10 (1:300). After washing, sections were labeled sequentially with goat anti- mouse IgG (1:50) and mouse PAP (1:50) and developed with 3,3- diaminobenzidine (DAB Sigma) in the presence of 1% nickel ammonium sulfate. Positive cells exhibited black-stained nuclei. Slides were counter- stained with methyl green or PAS, or not at all, and then coverslipped with an aqueous mounting medium. The intensity of the reaction product was scored as: - , no staining; + , weak; + + , moderate; and + + +, strong staining.

In some cases a fourth antigen (rabbit IgG bound to the GBM) was also detected within the same section. The MW denatures fixed IgG to a degree intermediate between that of cell membrane and cytoplasmiclnu- clear antigens. Therefore, the following adjustment was made to the stain- ing protocol described above. Once sections were developed with Fast Blue after completion of the first round of three-layer staining (OX-1 or R73), sections were MW oven heated twice for 5 min. pre-incubated with 10% FCS and 10% normal donkey serum, and then labeled sequentially with sheep anti-rabbit IgG (1:50), donkey anti-sheep/goat IgG (1:50), and goat PAP (1:50), and developed with DAB to produce a brown reaction product. Thereafter, sections were labeled with ED1 (red) and then PC-10 (black)

Blodcing of Anubcdy CtossreaaiVity by MW Oven Heating. The ability of MW oven heating to block antibody-antibody binding and to inactivate

MAbs as above. Triple-immunoenzyme staining of human cryostat tissue sections with

by guest on July 6, 2015jhc.sagepub.comDownloaded from

MICROWAVES IN MULTIPLE IMMUNOENZYME STAINING 99

Table 1. Effect of microwave oven heating on antibody crossreactiui2y, enzymatic activity, and antigen retrieval in PLPfied ctyostat and formalin-jived parafin tissue sections"

PLP-hCd tissue F~rmnlin-fired tissue (miaoprvc min) (miaowave min)

0 5 10 0 5 10

Antibody crossreactivity) + + + +

- MAb + + + c + + + + - MAb + GAMC + + + + + + + MAb + GAM + PAP + + + + + + + MAb + GAM + APAAP + + + + + + +

- - - - - -

Enzymatic activity Endogenous P + + + + + + + + + ND Endogenous AP + + + PAP complex + + + + + + + - APAAP complex + + + + + + + -

Cytoplasmic + + + + + + + + + + + + + + +

+ + +

- ND - -

Detection of tissue antigens Membranous + + + + - - f *

Nuclear + + + + + + + + + + + +

GAM, goat anti mouse IgG; P, peroxidase; AP, alkaline phosphatax; ND, not determined. Antibody crossreactivity was assessed as described in Materials and Methods. Intensity of staining reaction (graded - to + + + ).

Leu 4, KP1, and PC-10 MAbs followed the same protocol. Cell spots, once warmed to RT, were fmed in PLP for 15 min at 4'C, washed in PBS, and then stained as above.

A number of controls were employed to check the specificity of i"u- nostaining. First, three-layer staining of serial sections with just one pri- mary MAb, with or without MW Oven heating, was used as a comparison for labeling patterns produced in multiple immunoenzyme staining. Sec- ond, serial sections had multiple immunoenzyme staining with one or more MW treatments omitted. Third, irrelevant primary mouse MAbs and/or secondary normal goat IgG were used as negative controls.

Double Immunomzyme Staining 0f-m Tissue Sections. Most cell membrane antigens are modified by fmtion and the subsequent process- ing steps involved in preparation of paraffin tissue sections. This includes antigen epitopes recognized by OX-1, R73, and Leu 4 MAbs. Therefore, a double immunoenzyme staining method was developed for labeling cy- toplasmic and nudear antigens in paraffin tissue sections. This method was bascd on the finding that cytoplasmic and nuclear antigens could be retrieved by MW m n heating. Briefly, p d i sections (4 pm) were mounted on gelatin-coated slides, deparaffinized, and rehydrated to water. Slides were MW oven heated twice for 5 min and washed in PBS. Sections were pre-incubated and then labeled with mouse MAb PC-10 (1:lOOO) as de- scribed for cryostat tissue sections to produce black-stained nuclei. Slides had a second round of two 5-min MW Oven heatings, followed by PBS wash- ing, pre-incubation, and labeling with a second mouse MAb ED1 (1:4) as described above to produce a red cytoplasmic stain. Sections were counter- stained with standard PAS or PAS with hematoxylin omitted and were ei- ther mounted in an aqueous medium or dehydrated and mounted in De- pex (BDH Chemicals; Kilsyth, Australia).

Results

Bloching of Antibody Crossreactivity by MW Oven Heating The ability of MW oven heating to block antibody crossreactivity

was assessed in two ways. First, antibodies were bound to tissue sec- tions, microwaved for different periods of time, and then detected by the addition of further antibody-enzyme conjugates and sub- strate development. Results from the systematic evaluation of MW oven heating are shown in Table 1. The main finding was that all antibody molecules bound to tissue sections (cryostat or paraffin) during a three-layer antibody labeling procedure were completely denatured by two 5-min MW treatments so that they were no longer detected by the subsequent addition of goat anti-mouse IgG and mouse PAP/mouse APAAP and color development. This procedure has been repeated many times and found to be a reliable method for completely blocking the detection of bound antibody mole- cules during a subsequent round of immunoenzyme staining. In addition, two 5-min MW treatments completely inactivated per- oxidase and alkaline phosphatase enzymes present within the mouse PAP and mouse APAAP complexes. This was demonstrated by the ability of MW treatment, applied after the completion of three- layer antibody labeling, to block the development of a colored reac- tion product (lkble 1). We also found that two 5-min MW treat- ments inactivated endogenous alkaline phosphatase in cryostat sec- tions (Tible 1). However, MW treatment of up to 20 min had no effect on endogenous peroxidase activity.

The second means by which antibody crossreactivity was evalu- ated was the ability to detect three antigens of distinct cellular localization -cell membrane, cytoplasmic, and nuclear-within in- dividual cells in tissue sections. The effect of MW treatment on double immunoenzyme staining of a formalin-fixed diseased kid- ney is shown in Figures la and b. In the absence of MW treatment (Figure la), an indistinct weak (+ ) brownlblue color was observed in nuclei (PC-10 plus ED1 labeling) and a weak ( + ) blue color within the cytoplasm of some cells (ED1 labeling). The indistinct brown/ blue nuclear stain is probably due to (a) the binding of ED1 MAb and mouse APAAP used in the second round of staining by free arms of the goat anti-mouse IgG already bound to the PC-10 MAb

by guest on July 6, 2015jhc.sagepub.comDownloaded from

100 LAN, MU, NJKOLIC-PATERSON, ATKINS

by guest on July 6, 2015jhc.sagepub.comDownloaded from

MICROWAVES IN MULTIPLE IMMUNOENZYME STAINING 101

used in the first round ofstaining, and (b) detection of bound mouse PAP complexes by goat anti-mouse IgG used in the second round of staining. The use of two 5-min MW treatments between rounds of three-layer immunoenzyme staining produced a dramatic re- sult (Figure 1b). Not only were a greater number of PC-10' and ED1' cells detected (described below) but strong (+ + +) nuclear staining with the PC-10 MAb gave the distinctive brown color nor- mally produced by cleavage of the DAB substrate and had no over- lap with the strong ( + + +) blue cytoplasmic stain produced by ED1 MAb labeling. Indeed, many doubly positive cells can be clearly identified that have brown nuclei and blue cytoplasm, demonstrat- ing no crossreactivity between the two rounds of three-layer anti- body staining.

Further examples of two-, three-, and four-color immunoenzyme staining of cryostat tissue sections using MW treatment are shown in Figures IC-lf. In spleen from a diseased rat (Figure IC), three distinct cell populations were labeled with mouse MAbs: T-cells (blue cell membrane staining) within the periateriolar sheath (PALS) area and occasionally within an adjacent germinal center; occasional macrophages (red cytoplasmic stain) within the densely packed PAIS area; and proliferating cells (black nuclear staining) prominent within the active germinal center, and a few T-cells were also PC- 10' within the PALS area. An example of triple labeling (blue cell membrane, red cytoplasm, and black nucleus) of individual prolifer- ating macrophages within in a section of diseased lung is shown in Figure Id. Four-color immunoenzyme staining was also performed with the use of MW oven heating. In this case, blue cell surface (R73), red cytoplasmic (EDl), brown glomerular basement mem- brane (sheep anti-rabbit IgG), and black nuclei (PCNA) can be clearly distinguished in a section of diseased kidney (Figure le). In addition, the MW technique was successful in detecting T-cells (blue cell membrane) and macrophages (red cytoplasm) in acutely rejected human kidney allografts (Figure If), demonstrating the applicability of this technique to clinical specimens. Multiple im- munoenzyme staining was also successful using cell spots.

Antigen Retrieval by M W Oven Hegting The effect of MW treatment on detection of cellular antigens is shown in Eble 1. In PLFfixed cryostat tissue sections, MW treat- ment abolished recognition of cell membrane antigens (CD45 and up T-cell receptor) but had no discernible effect on detection of the cytoplasmic antigens (ED1 antigen and CD68), even after 20 min of treatment. Of particular interest was a marked increase in the detection of PCNA by two 5-min microwave treatments. The situation was somewhat different in paraffin tissue sections, in that cell membrane antigens could not be detected and detection of cytoplasmic antigens and PCNA was suboptimal. MW treatment of paraffin sections - whether formalin- or Bouin's-fixed tissue - demonstrated a marked increase in detection of both cytoplasmic

antigens and PCNA (Eble 1). An example of MW treatment block- ing antibody crossreactivity and facilitating antigen retrieval is shown by comparing Figures la and 1b. In the absence of MW treatment, few cells were labeled in formalin-fixed paraffin sections with PC- 10 and ED1 MAbs (Figure la). However, after MW treatment, many ED1+ and P C " cells, including distinct double-labeled cells, could be identified (Figure Ib). In non-microwaved sections, there were 3.8 & 0.5 ED1' cellslglomerular cross-section (gcs) in anti-GBM disease. MW treatment of serial sections from the same animal gave a score of 11.2 2 1.8 ED1' cells/gcs, which is comparable with that obtained by scoring ED1+ cells in PLP-fixed cryostat sections in the absence of MW treatment (12.0 k 1.2 ED1' cells/gcs from a group of six animals). In addition, the distribution of ED1' macrophages throughout the kidney, lung, and spleen was very similar in micro- waved paraffin sections compared with PLP-fixed cryostat sections.

Detection of the PCNA in cryostat and paraffin tissue sections by PC-10 MAb staining was also markedly enhanced by MW treat- ment. For example, in formalin-fixed paraffin sections of diseased rat kidney, the number of PC-10+ cells increased from 5.5 k 0.6 cellslgcs to 15.3 k 2.3 cellslgcs with MW treatment. In addition, the pattem of cell proliferation in kidney, lung, spleen, and pe- ripheral blood mononuclear cells in normal rats and in rats with anti-GBM disease was not altered by the MW treatment. The in- creased sensitivity of PC-10 MAb staining was reflected by the fmding that significantly lower concentrations of primary MAb were re- quired for maximal staining intensity (1:loOO vs 1:300 dilution). Indeed, good staining was achieved with a 1:4000 dilution of the PC-10 MAb. Titration of the PC-10 MAb after MW treatment was critical, since the use of a 1300 antibody dilution resulted in non- specific nuclear staining in both normal and diseased tissues,

Discussion This study describes the development of a simple, reliable, and sen- sitive method for detection of multiple antigens within cryostat and paraffin tissue sections. The advantages of this novel technique over current methods used for detection of two or more antigens within tissue sections are discussed below.

First, MW treatment was used to completely block antibody crossreactivity. This means that primary antibodies raised in the same species and the same secondary antibodies can be used in each round of immunostaining performed in between MW treat- ments. This is much simpler than other methods used to avoid an- tibody crossreactivity, which involve conjugation of purified anti- bodies with enzymes, fluorochromes (direct staining), or tags such as biotin, raising of primary or secondary antibodies in different species, or the use of irrelevant mouse MAb and F(ab) fragments of unconjugated goat anti-mouse antibody (indirect staining) (1-6).

A second important aspect of this new technique is that the complete blocking of antibody crossreactivity produced by two 5 -min

4

Figure 1. Examples of multiple immunoenzymatic staining with MW treatment (b-f) on (a,b) formalin-fixed paraffin sections and (c-f) PLP-fixed cryostat sections. Serial sections of diseased rat kidney (a) without or (b) with MW oven heating demonstrating detection of PCNA (PC-10 MAb, brown) and macrophages (ED1, blue) and counterstained with PAS minus hematoxylin. (C) Spleen from a diseased rat with detection of T-cells (R73, blue), macrophages (ED1. red; arrows), and PCNA (PC-10, black). (d) Diseased rat lung with detection Of total leukocytes (0x4, blue), macrophages (ED1, red), and PCNA (PC-10, black). Triple-labeled cells indicated by arrows. (e) Diseased rat kidney with detection of T-cells (R73, blue), rabbit IgG on the GEM (sheep anti-rabbit, brown), macrophages (ED1, red), and PCNA (PC-10, black). (f) Acute human kidney transplant rejection with detection of T-cells (Leu 4, blue) and macrophages (KP1, red). Bars - 15 pm.

by guest on July 6, 2015jhc.sagepub.comDownloaded from

102 LAN, MU, NMOLIC-PATERSON, ATKINS

MW treatments allows the use of sensitive three-layer immunoen- zyme staining for each primary mouse MAb. This is particularly important for detecting antigens of low abundance. Previous methods for multiple immunoenzyme staining have suffered from a lack of sensitivity because they could not employ PAP or APAAP complexes owing to problems of antibody crossreactivity.

Third, the use of MW treatment had a marked positive effect on retrieval of cytoplasmic and nuclear antigens in both cryostat and paraffin tissue sections, consistent with previous studies (14J5). Indeed, although no ED1’ or PC-10* cells could be detected in tis- sues fixed in formalin for over 1 month, these antigens were op- timally recognized &er MW oven heating twice for 5 min. The specificity of PC-10 MAb staining after MW treatment was con- firmed in three ways: (a) the pattern of PCNA expression in nor- mal and diseased tissues was the same in MW-treated and untreated sections; (b) dilution of the PC-IO MAb was consistent with increased antigen availability; and (c) PCNA expression in MW-treated sec- tions was very similar to that detected in methyl Carnoy’s-fixed par- affin tissue sections. The latter point is important, since excellent PC-10 staining can be achieved in methyl Camoy’s-fixed tissue that is not altered by MW treatment.

One further advantage of MW treatment is the ability to per- manently destroy enzymatic activity within PAP and APAAP com- plexes. This allows the use of more than one round of PAP or APAAP complex labeling during the same staining protocol. In addition, inactivation of endogenous alkaline phosphatase within the kid- ney proved very useful, since this activity can be a problem.

One drawback of this method of multiple immunostaining is that it cannot be used to detect more than one cell membrane an- tigen, since two 5-min MW oven heatings denatured cell mem- brane antigens. Although reducing the time of microwave treat- ment to a single 5-min heating allowed some antibody detection of cell membrane antigens, this period was insufficient to com- pletely block antibody crossreactivity.

The development of MW-based immunohistochemistry protocols has relied on empirical observation, with little clear understand- ing ofthe mechanisms involved. It has been shown that heat denatu- ration of native protein secondary structure can be prevented by cross-linking with formaldehyde (16). Therefore, formalin fixation may protect tissue antigens from the effects of MW treatment, whereas native antibodies bound to tissue sections during the im- munostaining procedure are denatured by this treatment. How- ever, this is clearly a complicated issue, as there are differential ef- fects of MW treatment in terms of denaturing fixed cell membrane antigens but enhancing recognition of fixed cytoplasmic and nu- clear antigens. Antigen retrieval is presumably a combination of improving antibody access to the antigen and possibly a partial breakdown of protein cross-linking that allows proteins sufficient freedom to assume conformations optimal for antibody binding. The immediate microenvironment of individual proteins (lipid vs macromolecular protein and nucleic acid) and the nature of sec- ondary and tertiary protein structure may also be important fac- tors in the differential effects of MW oven heating on antigen rec- ognition.

In summary, the combination of microwave oven heating and sequential rounds of three-layer immunoenzyme staining has pro- duced a simple, reliable, and sensitive method for detection of two

or more antigens within paraffin and cryostat tissue sections. This technique has a number of advantages over existing methods and has many possible applications in diagnosis and research.

Literature Cited 1.

2.

3.

4.

5 .

Boorsma DM. Direct immunoenzyme double staining applicable for monoclonal antibodies. Histochemistry 1984;80:103 Van der Loos CM, Das PK, Houthoff H-J. An immunoenzyme triple- staining method using both polydonal and monoclonal antibcdies from the same species. Application of combined direct, indirect, and avidin-biotin complex (ABC) technique. J Histochem Cytochem 1987;35:1199 Claassen E, Boorsma DM, Kors N, Van Rooijen N. Double-enzyme conjugates, producing an intermediate color, for simultaneous and di- rect detection of three different intracellular immunoglobulin deter- minants with only two enzymes. J Histochem Cytochem 1986;34:423 Tidman N, Janossy G, Bodger M. Delineation of human thymocyte differentiation pathway utilizing double staining techniques with mono- clonal antibodies. Clin Exp Immunol 1981;45:457 Van Rooijen N, Kors N. Double immunocytochemical staining in the study of antibody-producing cells in vivo. Combined detection of an- tigen specificity (anti-3”) and (sub)class of intracellular antibodies. J Histochem Cytochem 1985;33:175

6. Carl SAL, Gillete-Ferguson I, Ferguson DG. An indirect immunoflu- orescence procedure for staining the same cryosection with two mouse monoclonal primary antibodies. J Histochem Cytochem 1993;41:1273

7. Lan HY, Nikolic-Paterson DJ, Zarama M, Vannice JL, Atkins RC. Sup- pression of experimental crescentic glomerulonephritis by the inter- leukin-1 receptor antagonist. Kidney Int 1993;43:479

8. Sunderland CA, McMaster WR, Williams AF. Purification with mono- clonal antibody of a predominant leukocyte-common antigen and gly- coprotein from rat thymocytes. Eur J Immunol 1979;9:155

9. Hiiing T, Wallny HJ, HartleyJK, Lawetzky A, Tiefenthaler G. A mono- clonal antibody to a constant determinant ofthe rat T cell antigen recep- tor that induces T cell activation. Differential reactivity with subsets of immature and mature T lymphocytes. J Exp Med 1989;169:73

10. Dijkstra CD, Dopp EA, Joling P, Kraal G. The heterogeneity of mononuclear phagocytes in lymphoid organs: distinct macrophage sub- populations in the rat recognized by monoclonal antibodies ED1, ED2, ED3. Immunology 1985;54:589

11. Waseem NH, Lane DP. Mondonal antibody analysis of the prolifer- ating cell nuclear antigen (PCNA). J Cell Sci 1990;96:121

12. Kung PC, Goldstein G, Rineherz EL, Schlossman SF. Monoclonal an- tibodies defining human T cell surface antigens. Science 1979;206:347

13. Pulford KA, Rigney EM, Micklem KJ, Jones M, Stross WP, Gatter KC, Mason DY. KP1: a new monoclonal antibody that detects a mono- cytelmacrophage associated antigen in routinely processed tissue sec- tions. J Clin Path01 1989;42:414

14. Shi S-R, Key ME, Kalra KL. Antigen retrieval in formalin-fxed, panffin- embedded tissues: an enhancement method for immunohistochemi- cal staining based on microwave oven heating of tissue sections. J Histochem Cytochem 1991;39:741

15. Shi S-R, Chaiwun B, Young L, Cote RJ, Taylor CR. Antigen retrieval technique utilizing citrate b&r or mea solution for immunohistochem- ical demonstration of androgen receptor in formalin-fixed paraffin sec- tions. J Histochem Cytochem 1993;41:1599

16. Mason JT, O’Leary TG. Effects of formaldehyde fixation on protein sec- ondary structure: a calorimetric and infrared spectroscopic investiga- tion. J Histochem Cytochem 1991;39:225

by guest on July 6, 2015jhc.sagepub.comDownloaded from