Increased sensitivity in peroxidase immunocytochemistry

Histochemistry

Histochemistry (1986) 84:221 230

9 Springer-Verlag 1986

Increased sensitivity in peroxidase immunocytochemistry A comparative study of a number of peroxidase visualization methods employing a model system L. Scopsi and L.-I. Larsson Unit of Histochemistry, University Institute of Pathology, Frederik d. V's vej 11, DK-2100 Copenhagen 0, Denmark Accepted November 16, 1985

Summary. A n u m b e r o f methods for d e m o n s t r a t i o n o f per-

Material and methods

oxidase activity have been tested on immunocytochemical nitrocellulose models. By applying a well-characterized prim a r y a n t i b o d y and Sternberger's peroxidase-antiperoxidase technique, the sensitivity of various protocols has been evaluated. Best results were obtained with diaminobenzidine as chromogen, especially in conjunction with heavy metal salts, either a d d e d directly to the medium or used as " t o n e r s " o f the end-product. Use o f silver intensification o f the diaminobenzidine/metal e n d - p r o d u c t increased sensitivity further.

Immunocytoehemical reagents. Pentagastrin (Peptavlon| was purchased from ICI in injection ampoules (0.25 mg/ml) and further diluted with redistilled water. Rabbit antiserum No. 2717, which recognizes the C-terminal tetrapeptide amide of gastrin/CCK, contained within the sequence of pentagastrin, was kindly given to us by Professor J.F. Rehfeld, Department of Clinical Chemistry, Rigshospitalet, Copenhagen, Denmark. Unlabelled swine antirabbit IgG and rabbit PAP complex were obtained commercially from Dakopatts (Copenhagen, Denmark). Both primary and secondary antibodies as well as PAP complex were diluted in 0.05 M Tris buffer (Sigma7-9, cat. No. T-1378) pH7.4 containing 0.15M NaC1 (TBS) and 0.25 % (w/v) bovine serum albumin (BSA, fraction V, Sigma, cat. No. A-4503).

Introduction M a n y methods for increasing immunocytochemical sensitivity have been p r o p o s e d including different " s a n d w i c h " / multilayer methods, different microscopical techniques, m o r e selective filter settings and, above all, different methods for revealing peroxidase activity (Gibson et al. 1984; G r u b e 1980; Hess a n d Schneider 1981 ; H s u and Ree 1980; Illing and Wfissle 1979; L a n d s d o r p et al. 1984; M a l m g r e n and Olsson 1978; Moller et al. 1984; N a k a n e and Pierce 1966; Olsson and M a l m g r e n 1978; O r d r o n n e a u et al. 1980; Sternberger et al. 1970; Straus 1979; Vacca et al. 1975; Weir et al. 1974). As repeatedly stressed (Petrusz et al. 1975; Petrusz et al. 1980; Scopsi and Larsson 1985; Valnes et al. 1984), detection efficiency is often confused with sensitivity. W e have therefore tested the sensitivity o f several protocols for detection of H R P activity on an I C C model system (Larsson 1981). Our model consisted o f nitrocellulose strips with antigen spots, which were then stained with Sternberger's P A P procedure. H R P activity was demonstrated with a number o f chromogen protocols including diaminobenzidine and its modifications, 3-amino-9-ethylcarbazole, 4 - C l - l - n a p h tol, o-dianisidine, H a n k e r - Y a t e s reagent, tetramethylbenzidine and homovanillic acid. Thus, by keeping b o t h the dilution o f the p r i m a r y a n t i b o d y and sequence of i m m u n o r e a gents (secondary unlabelled a n t i b o d y and P A P complex) constant, it is possible to judge the sensitivity of the system, which in this case will depend entirely u p o n the chromogen step.

Immunoeytochemical procedure

Nitrocellulose filters (0.22 lain pore size, cat. No. GSWP 000 10, Millipore S.A., France) were spotted with 2 gl droplets ofpentagastrin solution. In all experiments different pentagastrin dots contained 10,000, 3,000, 1,000, 300, 100, 50, 25 and 5 pg. The filters were subsequently dried with a hairdryer and were then exposed to paraformaldehyde vapours at 80~ C for 1 h in order to immobilize the peptide (Larsson 1981). Thereafter, strips were cut, soaked 3 x 10 rain at room temperature in TBS containing 1% (v/v) triton X-100 (Serva, Heidelberg, FRG, TBS-triton) and then immersed in a 1% solution of human serum albumin (HSA Kabi, Stockholm, Sweden) in TBS for t h at room temperature, in order to block unspecific binding of immunoreagents to the paper. Without any further rinse the strips were then accomodated in glass centrifuge tubes (100 x 16 mm) containing the primary antibody (2717) diluted 1:16,000. The tubes were rotated for 20 h at 4~ plus I h at room temperature. Subsequently, they were rinsed in TBS-triton (3 x 10 rain, with frequent agitation of the dish) and unlabelled antirabbit IgG, diluted 1 : 30 was applied similarly to the primary antiserum for 30 rain at room temperature. The strips were then rinsed 3 x 10 rain in TBS-triton and PAP complex (diluted 1:75) was applied. Subsequently, the strips were rinsed in TBS-triton (2 x 10 rain as above) and then processed according to one of the following protocols. a) Diaminobenzidine at neutral pH (Graham and Karnowsky 1966). Strips were rinsed in TBS for 10 min at room temperature. 3,Y-diaminobenzidine tetrahydrochloride (DAB, Sigma, cat. No. D-5637) was dissolved in, alternatively, TBS, 0.05 M Tris-HCl buffer pH 7.4-7.6 or 0.1 M sodium phosphate buffer, pH 7.3 (PB) at a concentration of 0.5 mg/ml. All solutions were then Millipore-filtered (Millex| pore size 0.45 gm, Millipore S.A., France) and hydrogen peroxide (Perhydrol 30% H202 Merck, Art. No. 7209) was added at a final concentration of 0.0003-0.01%. The reaction

222 was developed for 10 min at room temperature while the strips, immersed in the DAB-H202 solution, were protected from the light by wrapping the Petri dishes with a tin-foil. Subsequently, all the strips were rinsed in redistilled water; some were scored for sensitivity and then further processed according to one of the protocols described in m), whereas others were scored and air dried.

b) Diaminobenzidine-p-cresol at p H 5.0 (Streit and Reubi 1977). Nitrocellulose models were rinsed in 0.05 M citric acid-ammonium acetate buffer pH 5.0 for 10 rain at 37 ~ C. DAB (0.1 mg/ml) was dissolved in the same buffer to which 0.07% p-cresol (Merck, Art. No. 805223, melting interval 32-35 ~ C) and hydrogen peroxide (0.01% final concentration) had been added. The resulting mixture was Millipore-filtered, poured in a Petri dish, the strips were immersed in it and the reaction developed for 10 min at 37 ~ C, in the dark. Strips were subsequently processed as in a). c) Diaminobenzidine in cacodylate buffer p H 5.1 (Malmgren and Olsson 1977). Strips were rinsed for 10 rain at room temperature in 0.1 M cacodylate buffer pH 5.1 (sodium cacodylate, purum, Fluka AG, cat. No. 20840). DAB (0.5-2 mg/ml) was dissolved in the same cacodylate buffer and MiUipore-filtered. Hydrogen peroxide was added to 0.01%. Strips were incubated for 10 min at room temperature, in the dark and subsequently processed as described in a). d) DAB at neutral p H plus imidazole (Straus 1982). Strips were rinsed for 10 min at room temperature in 0.05 M Tris-HC1 buffer pH 7.4. DAB (0.5 mg/ml) was dissolved in the same buffer to which 0.01-0.1 M imidazole (Merck, Art. No. 4716) had been added. The pH of the imidazole solution was adjusted to 7.4 with 1 N HC1 before addition of DAB. Hydrogen peroxide was added to 0.01%. Strips were incubated for 10 rain at room temperature in the dark and subsequently rinsed in redistilled water, scored and air-dried.

e) DAB at neutral p H plus cobalt. Three different protocols were tested: 1) (Adams 1977). Strips were rinsed 10 min in 0.1 M TrisHC1 buffer pH 7.6, transferred to 0.5% cobalt chloride (CoClz-6 H20 Sigma, cat. No. C-2644) in the same Tris buffer for 10 n-fin, rinsed once in Tris and once in PB for 5 rain each. They were then incubated for 10 min in the dark in Millipore-filtered solution of DAB (0.5 mg/ml) in PB containing 0.01% hydrogen peroxide, at room temperature. After development, strips were processed as in a). 2) (Adams 1981). Strips were rinsed in PB for 10 min at room temperature. DAB (0.5 mg/ml) was dissolved in PB and Millipore-filtered. 2.5 ml 1% CoClz and subsequently 2.0 ml 1% Nickel(II)sulphate (NiSO4.6 H20 Merck, Art. No. 6727) were added dropwise to 100 ml DAB solution under stirring. The strips were incubated in this medium for 10 rain at room temperature, hydrogen peroxide was then added to 0.01% and incubation continued for 10 min, at room temperature. Strips were subsequently processed as described in a). 3) (Hsu and Soban 1982) Strips were rinsed 10 min in TBS at room temperature. DAB (50 rag) was dissolved in 100 ml TBS, Millipore-filtered, and 2 ml 1% COC12 in TBS was then added dropwise, under stirring. The strips were incubated in DAB-CoCI2 for 5 min and then hydrogen peroxide 0.0003-0.01% was added and incubation continued for 2-3 rain, whereafter they were processed as described in a).

1.8 ml 5% sodium nitroferricyanide (Merck, Art. No. 6541) in redistilled water was added under stirring. This solution was placed on ice before hydrogen peroxide to 0.006% was added. Strips were incubated at 4 ~ for 10 rain in the dark, thereafter they were transferred through two rinsing-baths of 3% sodium nitroferricyanide at room temperature, rinsed twice more in redistilled water, scored for sensitivity and air-dried.

h) Tetramethylbenzidine (Mesulam 1978). Strips were rinsed for 10 rain at room temperature in redistilled water. 100 ml of incubation medium were freshly prepared by mixing 5 mg of 3,3', 5,5"-tetramethylbenzidine (TMB, Sigma, cat. No. T-2885) dissolved in 2.5 ml 99% ethanol with 92.5 ml redistilled water, containing 100 mg sodium nitroferricyanide and 5 ml 0.2 M acetate buffer, pH 3.3. Strips were incubated for 20 rain at room temperature, in the dark. They were then withdrawn and 2-4 ml of 0.3% hydrogen peroxide (final concentration 0.006-0.012%) were added, after which the strips were reineubated for a further 20-min-period, at room temperature, in the dark (Martin et al. 1984). After completion of development, strips were rinsed 4 x 10 min at 4 ~ C, in 0.01 M acetate buffer, pH 3.3. Some strips were scored, air-dried and photographed, while others were carried through a stabilization procedure (Rye et al. 1984): 50 mg DAB were dissolved in 100 ml PB and Millipore-filtered. 2 ml 1% CoClz in redistilled water was added while stirring. The mixture was chilled to 4 ~ C, hydrogen peroxide to 0.01% was added and the strips, briefly (30 s) rinsed in PB, were incubated for 5 rain, in the dark. They were then rinsed 5 times x 10 rain with cold PB, once more with redistilled water, scored and air-dried.

i) p-phenylenediamine-pyrocatechol (Hanker et al. 1977). The protocol recommended by Polysciences, Inc. was followed. Briefly, 75 mg of the Hanker-Yates mixture | (two parts, by weight, of purified pyrocatechol and 1 part of purified p-phenylenediamine dihydrochloride, Polysciences Inc., Warrington, PA, USA) were dissolved in 50ml 0.1 M Tris-HC1 buffer pH 7.6, and 0.5 ml 1% hydrogen peroxide (final concentration 0.01%) were added. After a 10 rain-rinse in PB, the strips were incubated in the PPD-PCH202 medium for 10 min at room temperature in the dark, with frequent agitation. They were then rinsed for 10 min in two changes of PB, rinsed once more with redistilled water, scored for sensitivity and air-dried. Alternatively, the procedure recommended by Sprearico et al. (1982) was followed. In this protocol chemicals purchased from Fluka A G were used. Briefly, 25 mg PPD (1,4 phenylenediamine dihydrochloride, cat. No. 78460) and 50 mg PC (1,2 dihydroxy-benzol, Pyrocatechin, cat. No. 15880) were dissolved in 50 ml 0.1 M Tris-HC1 buffer pH 7.6 and 1 ml 1% hydrogen peroxide was added (final concentration 0.02%). The rest of the procedure was carried out exactly as described in the previous protocol. In one set of experiments we attempted to modify the reaction endproduct either by adding metal salts (COC12, NiSO4) directly to the PPD-PC mixture (vide supra: e 2 and f), or by use of a postosmification step (vide infra: m 2). Finally, silver development according to Liposits et al. (1984; vide infra: m 3) was performed after the aforementioned variations as well as after the normal PPD-PC protocol. j ) 3-amino-9-ethylcarbazole (Graham et al. 1965) Strips were rinsed

g) o-dianisidine (Colman et al. 1976). Strips were rinsed for 10 min

for I0 rain at room temperature in 0.05 M acetate buffer pH 5.0. 10-20mg of 3-amino-9-ethylcarbazole (3-A-9-EC, Sigma, cat. No. A-5754) were dissolved in 0.5-2.5 ml N,N-dimethyl-formamide (Merck, Art. No. 2937). Acetate buffer (same as above) was then stirred in, dropwise, to make 50 ml. The resulting solution was Millipore-filtered and hydrogen peroxide (final concentration 0.0015-0.015%) was added just before incubation was started. Development was for 10 min at room temperature and terminated by rinsing the strips in redistilled water. Results were scored and strips air-dried.

at room temperature in redistilled water. 40 mg of o-dianisidine (3,3'-dimethoxybenzidine, Fluka AG, cat. No. 33430) were dissolved in 100 ml redistilled water at room temperature, and then

k) 4-chloro-l-naphtol (Nakane 1968). Three different protocols were used. 1) (Nakane 1968). Strips were rinsed for 10 min at

f ) DAB in acetate buffer p H 6.0 plus nickel (Hancock 1982). Strips were rinsed for 10 min at room temperature in 0.05-0.1 M acetate buffer pH 6.0. DAB (0.2-0.5 mg/ml) was dissolved in the same acetate buffer to which 1.0-2.5% nickel sulphate had been added, the solution was Millipore-filtered and hydrogen peroxide added to 0.007-0.1%. The strips were incubated for 10 min at room temperature and further processed as described in a).

223 room temperature in 0.05 M Tris-HC1 buffer pH 7.6. 20 mg 4-CI-1naphtol (ICN-K & K Lab. Inc., cat. No. 15236) were dissolved in 200 ~tl 99% ethanol and same Tris as above was added to make 50 ml. The resulting solution was Millipore-filtered and pH adjusted to 7.6. Hydrogen peroxide was then added, at a final concentration of 0.00045% and strips were incubated at room temperature, in the dark, for 10 rain. Development was stopped by a rinse in redistilled water, after which the models were scored and airdried. At this point some strips were further processed according to Nakane (1971). They were covered with 1% OsO4 for 1 min at room temperature, rinsed with redistilled water and covered a second time with 1% copper(II)-sulphate (CuSO4" 5 H20, Merck, Art. No. 2790), rinsed again with redistilled water, scored and airdried. 2) (Vandesande and Dierickx 1975). Strips were rinsed 10 min at room temperature in the same Tris as above. 25 mg 4-C1-1-n were dissolved in 2.5 ml 99% ethanol and the same Tris as above was added to make 50 ml. Hydrogen peroxide was then added to a final concentration of 0.005% and strips were incubated for 10 min at room temperature in the dark. Filtration of the medium was deliberately avoided and the solution remained cloudy throughout the whole development step which was terminated by rinsing the strips in redistilled water. Models were then scored and air-dried. 3) (Buffa R. personal communication) Strips were rinsed 10min at room temperature in 0.1 M Tris-HC1 buffer pH 7.6. 15 mg 4-CI-l-n were dissolved in 1 ml 99% ethanol, the same Tris as above was then added to make 50 ml and the resulting mixture was heated to approximatively 50~ under continuous stirring until a clear solution was attained. The medium was then Millipore-filtered and 0.5 m130% H2Oz (final concentration 0.3 %) was added. Strips were incubated in the mixture, while still warm, for 10 rain and were then rinsed in redistilled water, scored and air-dried.

l) Homovanillic acid at p H 6.0 (Papadimitriou et al. 1976) Strips were rinsed in 0.2 M acetate buffer pH 6.0. To 1 vol of freshly prepared 28 m M homovanillic acid (Merck, Art. No. 10177) solution in the same acetate buffer (pH adjusted to 6.0 with 0.2 N NaOH) 1 vol. of a freshly prepared 40 m M lead(II)-nitrate (Merck, Art. No. 7398) solution in the same acetate was added dropwise under constant stirring. To the resulting clear solution, containing 20 m M lead nitrate and 14 m M homovanillic acid, hydrogen peroxide was then added to a final concentration of 0.001-0.002%. Strips were incubated in this mixture for 20 rain to I h, at room temperature, rinsed 2 x 5 min in the above mentioned acetate buffer, carried briefly through 1% ammoniumsulphide (Merck, Art. No. 5442) and finally rinsed in redistilled water. Subsequently, some strips were simply scored and air-dried, whereas others were carried to the procedure of Gallyas et al. (1982) described in m) 3). In one set of experiments 0.05 M acetate buffer pH 5.0 was used and incubation was carried out also at 37 ~ C for 30 min. m) Post-DAB treatments. 1) Copper sulphate intensification (Hanker et al. 1979). Strips were further rinsed (3 x 5 min) in 0.05 M Tris-HC1 buffer pH 7.6 and immersed for 1 to 10 min in 0.5% copper sulphate in the same Tris buffer. They were rinsed first in physiological saline and then in redistilled water, scored and air-dried. 2) Osmium tetroxide intensification (Graham and Karnowsky 1966). Strips were rinsed 3 x 10 rain in chilled Ringer chloride (Johansson and Backman 1983) and then covered with 1% OsO4 in 0.1 M cacodylate buffer pH 7.2 for i to 3 min, rinsed again (3 x 10 min) with cold Ringer and finally with redistilled water. Thereafter, some of them were simply scored and air-dried whereas others were carried directly to the following procedure. 3) Silver-gold intensification (Gallyas et al. 1982). Intensification after Liposits et al. (1984) was performed. Strips were rinsed in redistilled water, transferred to two baths (15 min each) of 2% sodium acetate (Merck, Art. No. 6267) and then placed in Gallyas' physical developer. This last is composed of two stock solutions (A and B) which can be stored in dark bottles at room temperature where they keep for many months. Immediately before use, 10 ml of stock solution B (2 g ammonium nitrate, Merck, Art. No. 1188,

2 g silver nitrate, Merck, Art. No. 1512, 10 g tungstosilicic acid, Merck, Art. No. 659 and 5 ml 37% formaldehyde, Merck, Art. No. 4002, dissolved in 1,000 ml redistilled water, in the order given) were added dropwise to 10 ml of stock solution A (50 g of anhydrous sodium carbonate, Merck, Art. No. 6329, dissolved in 1,000 ml of redistilled water) kept under vigorous stirring. Strips were reacted in the developer, at room temperature, in the daylight and development was stopped after 4 to 6 min by immersing the strips into a 1% solution of acetic acid (Merck, Art. No. 63) for 5 rain. Subsequently, strips were rinsed in 2% sodium acetate for 5 min and "gold-toning" was accomplished by covering the strips with a 0.05% solution of gold chloride (HAuC14, Fluka AG, cat. No. 50800) in redistilled water, for 10 min at 4 ~ C. Further rinses in 2% sodium acetate (5 min), 3% sodium thiosulphate 5 - H 2 0 (Merck, Art. No. 6516) (10 rain twice), 2% sodium acetate (10 min twice) and redistilled water (10 rain twice) were carried out at room temperature, whereafter strips were scored and air-dried. 4) Goldsilver intensification (Newman et al. 1983a). Strips were covered with 0.1% HAuC14 for 5 rain at room temperature, rinsed in redistilled water (2 x 10 min) and then either carried through Gallyas' procedure (cf. m) 3)) or treated for 5 min at room temperature with neutralized sodium sulphide (11.7 g Na2S'7-9 H 2 0 , Merck, Art. No. 6638 and 11.9 g NaH2PO4.H20, Merck, Art. No. 6346 were dissolved in 1,000 ml of redistiUed water and the pH adjusted to 7.4), rinsed 2 x 10 min in redistilled water and exposed to Gallyas' developer. In both the cases the developing time did not exceed 30 s, thereafter strips were treated as described in m) 3). 5) Silver methenamine intensification (Rodriguez et al. 1984). Strips were reacted at 60 ~ C with a silver methenamine solution freshly prepared by adding 2.5 ml of a 5% AgNO3 solution and 5.0 ml of a 2.5% di-sodium tetraborate. 10 H 2 0 (Merck, Art. No. 6308) to 50 ml of a 3% hexamethylenetetramine (Merck, Art. No. 4339) solution. Strips were then rinsed 3 x 5 rain in redistilled water and "gold-toning" was accomplished by covering them with 0.05% HAuC14 for 1 min at room temperature. Subsequently, three rinses of 5 min each with redistilled water plus one rinse of 1 rain with 2.5% sodium thiosulphate and one final rinse of 10 min in redistilled water were carried out. After which strips were scored and air-dried.

Con~o& Routine immunocytochemical controls were carried out, including sequential deletion of the various layers and substitution of the primary antibody with non-immune rabbit serum and with antiserum 2717 preabsorbed for 24 h with 200 gg per ml of pentagastrin.

Results T h e nitrocellulose m o d e l s w o r k e d well w i t h all t e c h n i q u e s tested, i n c l u d e d silver intensification, w i t h o u t giving b a c k g r o u n d p r o b l e m s or unspecific b i n d i n g o f c h r o m o g e n s . A p p l i c a t i o n o f 2 gl droplets o f a n t i g e n s o l u t i o n o n t o nitrocellulose filters p r o d u c e d fairly h o m o g e n e o u s spots o f 2-2.5 m m in diameter. A direct c o r r e l a t i o n b e t w e e n a n t i g e n c o n c e n t r a t i o n a n d staining intensity was observed. Results were j u d g e d by visual i n s p e c t i o n o f the strips b o t h i m m e d i a t e l y after c o m p l e t i o n o f every single p r o c e d u r e , w h e n nitrocellulose was still wet, a n d after it h a d been dried. A f a d i n g in c o l o u r intensity was o b s e r v e d with all c h r o m o g e n s u p o n drying. All c o n t r o l strips were c o n s t a n t l y negative. Results are p r e s e n t e d b e l o w in three g r o u p s : D A B g r o u p (I), n o n - D A B g r o u p (II), D A B - i n t e n s i f i c a t i o n g r o u p OID.

I) D A B group The standard DAB-H202 (S-DAB) protocol of Graham a n d K a r n o w s k y (1966), w i t h o u t osmification, d e t e c t e d pen-

224

Fig. 1. Nitrocellulose immunocytochemical models containing 2 gl spots of pentagastrm solutions (final content per spot - in pg - is showed on the left side). All strips were immunocytochemically stained as described in material and methods. Peroxidase activity was then revealed either by the S-DAB procedure (A), by the DAB-p-cresol procedure at pH 5.0 (B), by the DAB-cacodylate procedure at pH 5.1 (C), by the DAB-imidazole procedure (D), by the DAB-CoClzNiSO4 procedure of Adams, 1981 (E), by the DAB-CoClz procedure of Hsu and ,oban, 1982 (F), or by the DAB-NiSO4 ,rocedure of Hancock, 1982 (G). Strip A erves as reference also for Figs. 2 and 3

Fig. 2. Nitrocellulose models as described in Fig. 1. Peroxidase activity was revealed by either the S-DAB procedure (A), by the 3-A-9-EC procedure (B), by the 4-C1-1-n procedure (C), by the PPDPC procedure (D), by the o-dianisidine procedure (E), by the TMB procedure as recommended by Mesulam et al. 1980 (F) or by the TMB procedure as modified by Rye et al. t984 (G)

tagastrin spots down to 25 pg (Fig. 1). Use of different buffers at neutral pH (TBS, Tris-HC1 or PB) and of different amounts of hydrogen peroxide did not influence sensitivity, but at low concentrations (0.0003%) of HzO2 the staining intensity decreased. Use of buffers at low p H (around 5) resulted in a lower sensitivity as 100 pg/spot were detected with the DAB-pcresol method and 300 pg/spot with DAB-cacodylate (Fig. 1). Addition of imidazole to a D A B medium at neutral p H increased sensitivity to 5 pg/spot. Even low concentrations (0.01 M) of imidazole increased sensitivity to this level

and a general increase in staining intensity was observed for all the spots (Fig. 1). Addition of metals to D A B medium increased the sensitivity but slight differences among the various protocols were noticed. Addition, for instance, of CoCla to D A B at neutral p H and use of 0.01% H202 (cf. procedure of Hsu and Soban 1982) produced a clear staining of the 5 pg spot but the colour intensity of the spots was not increased (Fig. 1), whereas addition of NiSO4 to DAB at p H 6.0 and use of 0.01% H202 (Hancock 1982) did stain the 5 pg spot, but an overall increase in the staining intensity was also observed (see Fig. 1). In an intermediate position between

225

Fig. 3. Nitrocellulose models as described in Fig. 1. Peroxidase activity was revealed by either the S-DAB procedure (A), by the DAB + CuSO4 procedure (B), by the DAB + OsO4 procedure (C), by the S-DAB procedure plus Gallyas development (D), by the DAB + OsO4 procedure plus Gallyas development (E), by the DAB-NiSO~ procedure plus Gallyas development (F), by the S-DAB + HAuC14-Na2S procedure plus Gallyas development (G) or by the S-DAB procedure plus silver methenamine intensification (H)

these two methods stand the results of the second of the two Adams' methods (Adams 1981) in which a DAB medium at neutral pH is supplemented with both COC12 and NiSO4 and 0.01% H202 is added. Here the 5 pg spot was stained and a slight increase in staining intensity was obtained (Fig. 1). The other Adams' method (1977) displayed, on the contrary, a lower sensitivity (25 pg spot was faintly stained) and the colour attained was not dark-blue as it happened with all the other methods employing metals, but brown as seen with the S-DAB procedure (results not shown).

II) non-DAB group Among the "alternative" chromogens best results were obtained with PPD-PC. Both the protocol established by Hanker et al. (1977) and recommended by Polysciences Inc. and the protocol proposed by Spreafico et al. (1982) resulted in the staining of the 5 pg spot. The colour intensity of all the spots was by far higher than that given by S-DAB (Fig. 2). Moreover, addition of metal salts to the medium produced a sharper definition of the spot contours along with a change towards a darker colour (results not shown). The o-dianisidine protocol (Colman et al. 1976) displayed a somewhat lower sensitivity (the 25 pg spot was well stained) but the staining intensity was still considerable (Fig. 2). Best results with 3-A-9-EC were obtained when 0.04% 3-A-9-EC and 0.075% H202 were used, when a wellstained 50 pg spot was obtained (Fig. 2). Among the 4-C1-1n protocols, the one suggested by Vandesande and Dierickx (1975), with omission of the filtration step, resulted in the staining of the 100 pg spot (Fig. 2). The least sensitive of the three protocols tested was that in which 4-C1-1-n solution had to be warmed and then filtered (300 pg spot was

faintly stained). The TMB procedure of Mesulam (1978) was found to possess a sensitivity comparable to that of S-DAB as a 25 pg spot was attained with all three concentrations of H202 (Fig. 2). The intensity of the staining was, however, somewhat lower and the colour faded with time. Use of the stabilization step (Rye et al. 1984) resulted in a stable end-product and, at the same time, in an increase both of sensitivity (the 5 pg spot was clearly stained) and staining intensity (Fig. 2). The method employing homovanillic acid and lead nitrate (Papadimitriou et al. 1976) was found to be the least sensitive. In fact, employing low concentrations of H202 (0.001-0.002%) only the I0,000 pg spot was stained (results not shown).

III) DAB-intensification group All these post-treatments of the DAB end-product resulted in an overall increase in the staining intensity and most of them produced also a remarkable rise in sensitivity (Fig. 3). Simple post-treatment of the strips with CuSO4 (Hanker et al. 1979) was superior to the single osmium treatment in terms of sensitivity (Fig. 3). Both produced, however, an increase of the staining intensity. The Gallyas' method when applied to strips developed with S-DAB produced a strong increase in the staining intensity but only a marginal increase in sensitivity (Fig. 3). The same method, when applied to strips which had been developed either with DAB-metals (especially nickel) or first developed with SDAB and t h e n " toned" with gold chloride (with or without NazS step; Newman et al. 1983 a) produced both a strong increase in the staining intensity and a remarkable increase in sensitivity, without affecting background level. The Gallyas' method was also able to increase the sensitivity of

226 the method based on homovanillic acid and lead nitrate plus a reduction with ammonium sulphide. In this case, after a short (30 s to 1 min) "physical development" the 1,000 pg spot was stained (results not shown). The silver methenamine enhancement recently proposed by Rodriguez et al. (1984) was found to give an increase both of the sensitivity and of the colour intensity of strips treated with SDAB, which was by far superior to that obtained with the Gallyas' method. Discussion

A great number of reports deal exclusively with comparisons among various protocols for demonstration of H R P activity in the central or peripheral nervous system (Carson and Mesulam 1982; Deschenes et al. 1979; Mesulam and Rosene 1979; MorreU et al. 1981). The results obtained by these investigators can not directly be transferred to ICC. The "sensitivity" of a given chromogen may hence differ according to its use. Among chromogens used in ICC for detecting H R P activity DAB is, by far, the most popular and its use has since long been based on a well-working standard protocol, employing Tris or phosphate buffers at near neutral p H (Graham and Karnowsky 1966). Under these conditions the method works well and has a good sensitivity. Attempts have been made to enhance its sensitivity, by using buffers at low pH and/or by use of additives (p-cresol, imidazole, heavy metals) and silver development (Adams 1981; Gallyas et al. 1982; Straus 1980; Vacca et al. 1978; Weir et al. 1974). By testing two protocols with DAB in low p H buffers (Streit and Reubi 1977; Malmgren and Olsson 1977) we have repeatedly found a lower sensitivity when compared with S-DAB. This is evidently in conflict with the findings obtained at tissue level both in HRP-neurohistochemistry (Malmgren and Olsson 1978; Streit and Reubi 1977) and in ICC (Weir et al. 1974; Vacca et al. 1978; Straus 1980). Low pH buffers were introduced in DAB staining because it was thought that carrying out the enzymatic reaction at a pH close to the optimal p H of H R P (