This invention in principle relates to the detection of precipitates of diaminobenzidine (DAB) or other ^' enzidine derivatives on solid substrates and provides an improved procedure for amplifying the density of the DAB related signal to various detection systems, including optical or electron systems.

Diaminobenzidine and its other related benzidine derivatives, have proved to be powerful compounds for the selective detection of horseradish peroxidase and other enzyme markers. These enzymes are used as labels in many systems, for example, immunohistochernistry, "in situ" nucleic acid hybridisation, protein immunoblot studies or as markers in "in vivo" tracer studies. There are many instances however in such studies where the method sensitivity is inadequate, and the ultimate level of diaminobenzidine deposition is not ^' high enough to give rise to a definitive signal. A number of methods therefore has been proposed to increase the signal intensity or change the colour of the diaminobenzidine product: For example, by the simultaneous deposition with DAB of metal ions such as cobalt, nickel and copper. Alternatively a method has been proposed for increasing the intensity of pre-formed

diaminobenzidine precipitate by selective catalytic silver deposition on to it, and it has also been proposed to promote the formation of an intermediate diaminobenzidine gold-sulphide compound to more effectively catalyse the silver precipitation.

A main objective of the present invention is to improve the gold-sulphide-silver amplification (GSSA) procedure by rendering it more simple and economical and therefore amenable to routine application. The GSSA ^" technique was originally designed to detect trace amounts of• diaminobenzidine deposited by immunoperoxidase reagents on thin sections of plastic embedded biological tissue. The current invention extends its application to immunoperoxidase reactions • carried out on thicker paraffin embedded tissue sections and frozen sections; as well as nitrocellulose paper preparations pertaining to in situ nucleic acid hybridisation studies or immunoblot protein analyses. Furthermore, the previously employed three step procedure has been condensed to a two. step procedure.

Thus, in one aspect of the present invention it has been discovered that gold and sulphide can be added simultaneously in the first step, combined in an optimal ratio and at an optimal pH containing altogther much lower amounts of gold and sulphide compounds. According to another aspect of the present invention the final

silver solution is also modified to contain much less tungstosilicic acid but surprisingly giving a more powerful catalysed silver deposition.

Broadly stated from one aspect the invention consists in a procedure for improving the signal densit (e.g. optical density) of precipitates of benzidine derivatives on solid substrates, in which the substrate is a paraffin embedded or frozen tissue section. Conveniently the benzidine derivative is diaminobenzidine polymer, deposited by an immunoperoxidase reagent and reacted with a gold chloride solution and with the sodium sulphide and subsequently with a silver solution containing tungstosilicic acid; preferably the gold-sulphide solution or solutions are acidified, with a pH value between 2.5 and 7. The gold concentration may be between 0.0001 and 0.05%.

From another aspect the invention consists in a procedure for improving the signal density (e.g. optica density) of precipitates of benzidine derivatives on solid substrates, in which an intermediate gold-sulphid compound is used to catalyse the deposition of silver, and the substrate is a non-biological tissue artificial substrate. Preferably the substrate lias a substantially uniform background matrix: for example, it may be nitro

cellulose or the like.

The procedure may be in the form of a nucleic acid hybridisation study or an immunoblot protein analysis.

Conveniently the benzidine derivative is diaminobenzidine polymer, deposited by an immunoperoxidase reagent, and reacted with a gold chloride solution and with sodium sulphide, and subsequently with a silver solution containing tungstosilicic acid.

Preferably the gold-sulphide solution or solutions are acidified, with a pH value between 2.5 and 7. In a particular preferred procedure the gold concentration is between 0.001 and 0.05% and the sulphide concentration is between .005 and 5%. From another aspect the invention consists in a procedure for improving the signal density (e.g.- optical density) of precipitates of benzidine derivatives on solid ^" substrates, in which the gold and sulphide compounds are present in the same solution. ^' Conveniently the sulphide is initially in the form of solid crystals, which are added directly to the gold chloride solution.

The invention also consists in a kit for use in such a procedure comprising two containers or compartments, one containing the gold-chloride solution, and the other the sulphide crystals.

The invention also resides in a light-insensitive silver developing solution, comprising:

Sodium carbonate 1 - 10% Ammonium nitrate .01% - 1% Silver nitrate .01% - 1%

Formalin .01% - 2%

(balance water) and including tungstosilicic acid in a proportion below 0.5%. Broadly stated the silver developing solution comprises a buffer, a silver salt, a reducing agent, and a light stabiliser. In a particular preferred form of the solution the buffer is sodium carbonate, the stabiliser is a nitrate, the silver salt is silver nitrate, the light stabiliser contains a • tungsten compound, and the reducing agent is formalin.

Basis for application to semi-thin plastic sections

1) Section preparation

Biological, tissue under study is fixed in a choice of aldehyde fixatives and embedded in LR White resin (London Resin Co. - Registered Trademark). Semi- thin, sections (0.25 u to 0.5 u in thickness) are cut -from a selected, plastic embedded tissue block using an ultra-microtome apparatus, (Reichert Ultracut E) , and transferred to a droplet (100 ul) of double- distilled water kept on a clean glass slide. ,The water is allowed to evaporate gradually at 50°C overnight leading to flattening and very firm adherence of the section to the glass slide.

The dried section is then equilibrated in a droplet (200 ul) of phosphate-buffered saline (PBS - sodium phosphate buffer 0.01M pH 7.2 containing 0.15M sodium. chloride) for.5 to 15 minutes. Thereafter immunoperoxidase labelling procedure is performed on the sections bearing the following conditions in mind.

(a) The primary antibody reagent and the subsequent secondary reagents are applied to the sections for a minimum of 15 minutes and maximum of 2 hours. The upper time limit relates to prevention of deterioration of the section quality due to overexposure to aqueous media.

,(b) The immunoperoxidase labelling technique used should be ensured to have a high level of method specificity since the silver amplification is unable to

discriminate' between specifically and non-specifically deposited diaminobenzidine product. Antibody bridge techniques (e.g. PAP and DHSS ^" techniques) with intrinsically high level of method specificity are therefore preferred over the indirect labelling techniques (e.g. indirect peroxidase conjugate). The

DHSS technique which has been shown to be significantly more sensitive than the PAP has been employed by us and is the preferred method in conjunction with the GSSA procedure. The more recent avidin-biotin complex approach (ABC kit Vectastain Laboratories) and the

Fab ¹ -peroxidase conjugate system described by Ishikawa and co-workers which are not antibody bridge techniques have high levels of method specificity and may be used with this system.

(c) The diaminobenzidine product once deposited should be subjected 'to the GSSA procedure without much time delay as ^' there is tendency for it to deteriorate on long-term storage in the aqueous medium.'

2) DHSS Procedure ^' -

The DHSS system depends ^" on the use of dinitrophenyl (DNP) hapten labelled primary antibody directed against the molecular species to be detected in the tissue section' or the nitrocellulose or other solid preparation. The DNP groups on the primary antibodies binding to ^' he target elements are then specifically linked to DNP-labelled peroxidase conjugate molecules using a multivalent bridge antibody system consisting of

monoclonal anti-DNP IgM (i.e. decavalent) molecules.

The bound peroxidase conjugate is then visualised by adding .diam.inobenzidine and H„0 containing solution which results in deposition of the precipitate of peroxidatively polymerised diaminobenzidine.

3) Improved gold-sulphide silver amplification procedure

The immunoperoxidase labelled plastic sections are first rinsed in double. distilled water (DDH-O) for at least 5 minutes with at least 2 changes of washing medium. They are then exposed for 5 m'inutes to freshly made gold-sulphide solution : 5 ml. of 0.03% gold chloride solution containing 0.3% (w/v) of sodium sulphide to which 84 ul of IN hydrochloric acid have been added just prior to use. The sections are then rinsed as before in DDH-0 and then exposed to freshly constituted silver developer reagent, but containing only .4% of tungstosilicic acid. The preferred silver developer reagent is prepared by a rapid addition of 1 volume of 5% (w/v) sodium carbonate stock solution to 1 volume of a stock solution containing ammonium nitrate (0.2% w/v), silver nitrate (0.2% w/v),. tungstosilicic acid (0.4% w/v) and formalin (0.5% v/v) .

The reaction is allowed to proceed until the first signs of non-specific silver deposition at the section edges (due to mechanical trapping) are visible. This is seen usually between 1 to 4 minutes from the time of the application of the silver reagent to the

section. Although wherever possible one can and should extend this time limit to obtain the most intense colour intensification ' in the desired components of the tissue section. The reaction is promptly quenched by immersing the slide into 1% (w/v) acetic acid solution. The sections are allowed to dry gradually at room temperature (overnight) and then counter-stained, if required, with Harris's haemalum for 2 minutes keeping the glass slides heated (on a hot plate at 60 C) before mounting them under coverslips using Gurr's medium (Registered Trademark).

Basis for application of GSSΛ procedure to immunoperoxidase labelled paraffin-embedded and frozen tissue sections

The procedural recommendations for silver amplification of diaminobenzidine product using the GSSA method are essentially the same as those for the semi- thin ^' plastic sections. However, there is less restriction on the time these types of sections may be exposed to the im unological reagents. For example, the sections are able to tolerate 24 to 48 hours of immuno¬ peroxidase labelling time without undue loss in section quality (i.e. section lifting and/or disintegration).

Basis for application to nitrocellulose studies

Immunoperoxidase staining of nitrocellulose preparations relating to either in situ nucleic acid hybridisation studies (e.g. Southern and Northern blots

and Hybri-dots) or protein analysis (e.g. Western blots or Immunoblots) on the whole presents less stringent demands in terms of method specificity _* This is because, unlike biological tissues, the background matrix provided by nitrocellulose is relatively uniform and therefore much more predictable in terms of its non¬ specific interaction with the secondary immunoperoxidase reagents. Thus, indirect peroxidase conjugate techniques are likely to prove less problematic and therefore on balance as effective as the antibody bridge techniques for nitrocellulose studies.

The advantage of nitrocellulose- in terms of its relatively uniform composition also results in a more predictable handling of any non-specific interaction of gold-sulphide-silver amplification reagents, in particular the silver developing reagent. As long as 10 minutes silver developing time produces no undue non-specific staining of the background on the nitrocellulose paper, provided the previously added reagents are not in any way contributory. The inclusion of 0.05% Tween 20 (Registered Trademark) in- all the immunoperoxidase reagents applied to the nitrocellulose prior to the gold-sulphide-silver reagents application, is very effective in minimising any non-specific ^• adherence of such reagents . The result is that it is possible to achieve much higher signal to noise amplification ratios ending in a markedly significant increase in the detection sensitivity of the overall technique.

Finally,. the stability of the nitrocellulose preparations is sufficient enough to allow,, if necessary, much longer .incubation times for the immuno¬ peroxidase reagents.

The following examples illustrate the effectiveness of the improved GSSA procedure

Example 1 - Immunoblot staining amplification

Nitrocellulose paper was pre-treated with distilled H_0 and allowed to dry. Square pieces

2 (0.5 cm ) were then cut out and arranged in a duplicate series of 10 such pieces, designated numbers 1 to 10 starting from the left-hand side. These were spotted with 3 ul amounts of each of the ten corresponding rabbit IgG solutions containing 3125, 625,

-15 125, 25, 5 and 1 femtograms (i.e. 10 g) respectively. The spots were allowed to dry for at least 1 hour at room temperature, and. then subjected to immunoperoxidase staining with the DHSS system in the following manner.

DNP-labelled sheep anti-rabbit IgG antibody was applied in 100 ul amounts to> each of the nitrocellulose

• pieces, at a concentration of 1.5 ug/ l sheep IgG. The incubation was allowed to proceed for 1.5 hours at room temperature. This was followed by washing off excess

. reagent using phosphate-buffered saline containing Tween

20 (0.05% v/v) in changes each lasting 5 minutes. Next anti-DNP IgM monoclonal antibody (1:100 dilution of a

hybridoma ascites preparation containing 0.1% Tween 20) was applied for a period of 1.5 hours at room temperature. The excess reagent was once again washed off as before. The DNP-peroxidase conjugate (1:800 -dilution of a standard preparation) was- ^' then applied for 1.5= hours at room temperature followed by washing as before. The peroxidase conjugate bound to the nitrocellulose pieces was reacted for 5 minutes at room temperature with diaminobenzidine in the presence of hydrogen peroxide. This was followed by the standard wash but using DDH_0. At this stage the first series of 10 pieces were allowed to dry gradually at room temperature, whilst the second series of 10 pieces were all treated with the _' GSSA reagents in the following manner.

Droplets of freshly made combined 'gold-sulphide ' reagent .containing 140 ul of IN HCl- (see text above) were first applied to- the pieces for a period of 5 minutes at room temperature, followed by a three-changes .wash in DDH _p O each lasting 1 minute. Freshly made ■combined silver solution was finally applied to the

^• pieces ^; and reaction allowed to proceed for full 10

^• minutes _' before quenching the reaction with 0.1% acetic ^■ .-' ^■ acid 'in DDH-0. The -nitrocellulose pieces were allowed

^: to dry gradually at room temperature. The results of

• this study show dramatic improvement in the intensity of staining resulting from the application of the GSSA procedure.

Example .2 - Enhancement of immunoperoxidase staining in paraffin sections

Fresh pancreas was cut ^• into small blocks

3 measuring approximately 1cm in size. These were fixed in a standard formol-saline fixative and embedded into paraffin in a routine manner. Serial sections (5 u thick) -taken 'from .a block containing islet tissue were taken onto chrome-gel' coated glass slides ^; and allowed to dry and become firmly adherent by incubating them overnight in a '37°C incubator. The sections were deparaffinised in a routine manner using serial xylene and alcohol washes , and treated for 30 minutes at room temperature with a mixture of methanol (47.2 ml) with

0.8 ml of a stock solution concentrated H-0- (30% v/v) in order to inhibit any endogenous peroxidatic activity.

An optimal dilution (1:800) DNP-labelled guinea pig anti-bovine insulin antibody preparation was ^' applied for 30 minutes to two pairs of selected serial sections.

This 'was followed by a wash in PBS (3 changes, one minute each)> to remove the .excess reagent. DNP-specific

IgM- monoclonal and DNP-peroxidase conjugate reagents as used ^: in the Example 1 were applied in sequence for 30

^• ■ minutes each, with the standard washes in between. The diaminobenzidine/H _p O ^• under conditions identical to those-described for Exmaple 1, was applied' to develop the peroxidase reaction. ^■ One section from each pair was set aside for counterstaining with hae atoxylin and subsequent mounting under Canada Balsam (Registered

Trademark) , whilst the remaining sections were treated . ith the GSSA reagents .as per Example 1 but. with combined gold-sulphide .reagent containing 84 ul of IN HC1 and giving 2 to 5 minutes instead of 10. minutes developing time with the silver solution.

The sections were counterstained with ■haematoxylin, and .mounted as before. The results of -.this study again showed dramatic improvement in the intensity of staining of the islet B cells in the GSSA treated sections.

Example 3 - Amplification of immunoperoxidase staining reaction on frozen sections

Fresh lymphoid tissue was dissected into 0.5

3 cm blocks and frozen in isopentane cooled with liquid nitrogen.. The frozen blocks were stored in containers packed with crushed ice at -70 C until used. A block was. serially sectioned at 4 u setting using an automated cryostat unit.

^• The sections were taken onto dry alcohol cleaned glass -slides and air-dried for 5 minutes. They were then treated for 5 minutes with yery dry acetone (<0.02%

H-0 content) obtained from BDH as a chromatography grade reagent. The sections were then equilibrated with

,PBS. over ^" a, 5 ,to 10 minute, period and exposed to the DHSS procedure reagents as follows:

DNP-labelled monoclonal antibody (RF - B4) to

B-cells was applied at dilution of 1.15 to 4 serial sections for 15 minutes followed by three 1 minute

changes wash in PBS. The DNP-specific IgM monoclonal antibody and DNP-peroxidase conjugate reagents as used in Example 2, were applied in sequence for 30 minutes each with the washing procedure included in between. Diaminobenzidine/H _p O solution was applied to develop the colour reaction as described in Example 1. A pair of sections was counterstained with Harris's haemalu and mounted permanently under Canada Balsam. The other pair of sections was treated with GSSA reagents followed by haematoxylin counterstaining as described for the paraffin sections in Example 2. The results of this study showed a dramatic increase in the intensity of staining of the B lymphocytes in the GSS treated sections.

Example 4 - Use of separate Gold and Sulphide Reagents ■ in GSSA

In the previous example the GSSA procedure was performed using a combined gold-sulphide reagent. In certain applications however it may be preferred to use separately compounded gold and sulphide reagents.

In such case it is preferred to wash the labelled sections, after addition of diaminobenzidine, in an acid wash as mentioned under (3) above, before exposure to each ^' of the solutions. (When using a combined gold-sulphide reagent it may also be preferred to use an acid wash before and after exposure to the gold-sulphide reagent. )

Basis for Commercial Kit Forms

Although the invention relates specifically to the improved GSSA procedure, applicable to diaminobenzidine precipitate deposited by any means, it is important to note that the DHSS system of choice has formed the back-bone of its application to various types of immunoperoxidase analysis. For this reason, the invention contemplates two kit forms: Kit 'A' relating to GSSA system alone for amplification of diaminobenzidine precipitate in general, including immunoperoxidase analysis; and Kit 'B' relating to GSSA system applied in conjunction with the DHSS procedure for high grade immunoperoxidase analysis under specified conditions .

The Kit A for example, may consist of two reagent bottles each consisting of a two compartment reagent bottle, 'A* having a readily breakable partition between the two compartments containing in one compartment 5 ml of gold chloride solution (0.03% w/v) plus 140 ul of IN HCl, and the other compartment containing solid sodium sulphide reagent (15mg); and the reagent bottle •B*, again having a readily breakable portion, and solution I and solution II of Gallyas's silver regime in separate compartments.

On breaking the partition between the two compartments in each reagent bottle there should then result the ready-to-use reagents for each of the two steps in the GSSA procedure.

The silver developer reagent of the invention essentially comprises a buffer, a silver salt, a stabiliser for the silver salt, and a light stabiliser, optionally with a reducing agent, and an overall stabiliser such as a surface active detergent. It has been found of advantage to establish the developer in acid conditions, in which case the constituents need to be selected accordingly. One particular developer formulation of this nature is as follows:

Sodium acetate (anhydrous) in 13.57 g. amount is dissolved in 300 ml of water (analytical grade) and the solution treated with 3.75 ml of glacial acetic acid (analytical grade) to make buffer solution with a pH of between 5.1 and 5.2. To this is added silver nitrate (analytical grade) in 0.5 g. amount and thoroughly dissolved. Finally, 15 millilitres of a one per cent solution of a charged detergent such as Triton X-100 Registered Trade Mark (in analytical grade water) is gently added and allowed to dissolve and the resulting solution made up to 400 ml with analytical grade water. This solution is ready to use after an overnight storage in an analytically clean brown glass-ware bottle. It has been found to be stable for at least 6 months at ambient temperatures of between 17°C to 20°C.

The overall stability of the solution can be extended by inclusion of a charged detergent such as 15 ml of one per cent solution cetylpyridinium chloride (analytical grade) in analytical grade water.

The developer reagent also requires the presence of a light stabiliser containing a tungsten compound, and a reducing agent. These may be incorporated in the developer or applied to the tissue sections beforehand. Thus in one procedure for performing the invention the following steps -are taken.

1. Treat the sections bearing the DAB-gold sulphide complex with a freshly made solution of silver developer reagent (I or II) , sodium tungstate solution (5% in distilled water) and freshly made hydroquinone (1% in distilled water) in 8:1:1 parts, respectively, and added together in that order. The reaction with frozen and paraffin sections may be usefully continued for between 5-10 minutes and that with plastic sections between 1-2 minutes.

2. The sections at the end of the treatment should be rinsed in distilled water and exposed to the fixer in form of sodium thiosulphate (5% in distilled water) for 1 to 2 minutes, and then rinsed again with distilled water before counter-staining and mounting.

Great improvements in silver amplification are obtained by these silver developing procedures:-

1. Background non-specific silver staining of connective tissue and nuclei almost eliminated.

2. Signal intensification more reproducible and particularly high for trace amounts of virtually invisible diaminobenzidine product deposits.

3. The overall reagent conditions less deleterious to the frozen sections.