protocol double immunocytochemistry in pre-embedding ... · protocol double immunocytochemistry in...

Ž .Brain Research Protocols 2 1997 1–8

Protocol

Double immunocytochemistry in pre-embedding electron microscopy for thedetection of neurotensin and tyrosine hydroxylase in the guinea pig, using

two primary antisera raised in the same species

P. Marcos a,b,), M. Corio a, P. Dubourg a, R. Covenas b, G. Tramu a˜a Laboratoire de Neurocytochimie Fonctionnelle, Unite Associee au C.N.R.S., URA 339, UniÕersite de Bordeaux 1, AÕenue des Facultes, 33405 Talence´ ´ ´ ´

cedex, Franceb Departamento de Fisiologıa, Facultad de Medicina, UniÕersidad de Malaga, Campus de Teatinos srn, Malaga, Spain´ ´ ´

Accepted 11 April 1997

Abstract

Ž .In this study we identified for electron microscopy two different antigens neurotensin and tyrosine hydroxylase in the samepre-embedding section of nervous tissue, using two antibodies obtained in the same species. Optimal ultrastructural results were obtained

Ž .without adding to the fixative either glutaraldehyde or acrolein normally used for electron microscopy techniques . The differentŽdeveloping methods used in this study DAB in combination with either 1 nm silver-enhanced colloidal gold or benzidine dihydrochlo-

.ride are perfectly distinguishable at the ultrastructural level, and show some advantages over other previously described developingŽ .procedures. For instance, the use of small gold particles 1 nm reduces the severity of membrane damage caused by tissue penetration of

Ž .the bigger gold particles 5 nm . In addition, the reaction products are stable, so there is no need to stabilize them before osmication, as isnecessary in other developing methods such as the TMB procedure. The immunolabeling results obtained in this study were similar inboth developing methods, although synaptic profiles were more readily visible when the DABrcolloidal gold procedure was used. Usingelectron microscopy, we have detected TH immunoreactivity in dendrites and perikarya receiving synaptic contacts from NT-positiveterminals, as well as TH-immunoreactive inputs on NT-positive neurons, at both the somatic and dendritic levels. q 1997 ElsevierScience B.V.

Themes: Cell biologyTopics: Staining, tracing, and imaging techniquesKeywords: Benzidine dihydrochloride; Developing method; Diaminobenzidine; Double immunocytochemistry; Electron microscopy; Pre-embedding

1. Type of research

The different protocols described in this paper can beused in several types of research:Ø Cellular neurobiology.Ø Electron microscopy.Ø Immunocytochemistry.

2. Time required

For double labeling using DAB and silver-enhancedcolloidal gold as the developing methods, the time esti-

) Corresponding author. Departamento de Biologıa Celular y Patologıa,´ ´Facultad de Medicina, Avenida del Campo Charro srn, 37007 Sala-manca, Spain.

mated is about 85 h. When using the DAB and BDHCprocedures for double staining, the approximate time is 87h.

In both cases, a supplementary period of 3 days isrequired to embed the sections in Epon, obtain ultrathinsections, and prepare these for observation under an elec-tron microscope.

3. Materials

Ø Animals. In this study we used ten adult female guineaŽ .pigs Iffa Credo weighing 400–600 g. Each animal

was kept in a cage and fed ad libitum with food forŽ .guinea pigs Iffa Credo for over 1 week so that the

animals could become adapted to their new environ-ment. Temperature was maintained at 208C, and the

1385-299Xr97r$17.00 q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S1385-299X 97 00018-4

( )P. Marcos et al.rBrain Research Protocols 2 1997 1–82

photoperiod was 12 h lightr12 h darkness. The phaseof the estrous cycle was controlled by taking dailyvaginal smears and post-mortem histological studies ofthe uterus. All animals were in diestrus at the time ofsacrifice.

Ø Special equipment:Ø Animals and food: Iffa Credo, BP 109, 69210 St.

Germain L’Abresle, France.Ø Stereotaxic apparatus with Hamilton syringe: Kopf.Ø Vibratome: Lancer, St. Louis, MO 63103, USA.Ø Ultramicrotome: Ultracut Reichert, Heidelberg, Ger-

many.Ø Electron microscope: Philips EM 300.Ø Light microscope: Leitz.Ø Perfusion pump: Masterflex, Cole Parmer Instru-

ment Co., Chicago, IL, USA.Ø Chemicals and reagents:

Ø Drugs:Ø Sodium pentobarbital: Sigma, L’Isle d’Abeau

Chesnes, BP 701, 38297 Saint Quentin Fallavier,France.

Ø Colchicine: Sigma.Ø Chemicals:

Ø Sodium veronal: Sigma.Ø Thymerosal: Sigma.

w xØ Antibody against neurotensin: see 11 .Ø Antibody against tyrosine hydroxylase: Institut

Jacques Boy, BP 1430, 51065 Reims cedex,France.

Ø Biotinylated antibody to rabbit IgG: Jackson Lab-oratories, Interchim, 211 Av. J.F. Kennedy, BP1140, 03103 Mont Lucon cedex, France.

Ø Peroxidase-conjugated streptavidin: Jackson Lab-oratories.

Ø 1 nm gold-conjugated goat anti-rabbit IgG: Bio-cell, TEBU, BP 15, 78610 Le Perray en Yve-lines, France.

Ø Silver enhancement kit of colloidal gold probes:Biocell.

X Ž X XØ 3,3 -Diaminobenzidine 3,3 ,4,4 -tetraamino-

.biphenyl : Sigma.Ø Benzidine dihydrochloride: Sigma.Ø Sodium nitroferricyanide: Sigma.

4. Detailed procedure

A. Animals pre-treated with colchicine were anes-thetized by an intraperitoneal injection of sodium pento-

Ž .barbital 50 mgrkg body weight .B. Stereotaxic coordinates for the lateral ventricle were

w xchosen according to the atlas of Tindal 12 for the guineapig. The colchicine solution was prepared by dissolving180 mg of colchicine in 10 ml of 0.9% NaCl, a Hamiltonsyringe coupled to the stereotaxic apparatus being filledwith 10 ml of this solution. Colchicine was slowly injected

Ž .intraventricularly 1 mlrmin in order to avoid a possiblereflux of the solution. In all, 180 mg of the drug wasinjected per 10 min. The treated animals were placed nextto a heat source until they regained consciousness, theirsurvival period being 48 h.

C. Animals whether treated with colchicine or not wereanesthetized as described in step A.

D. Animals were perfused through the ascending aortaŽ .the descending aorta was clamped . A short rinse with

Ž .0.9% NaCl 3–4 min was carried out, followed by perfu-Ž .sion of two fixatives: 200 ml of fixative A 5 min and 400

Ž .ml of fixative B 20 min , both used at room temperature.Fixative A was prepared by mixing 1 vol. of acetate bufferŽ .0.2 M, pH 6.5 with 1 vol. of 8% paraformaldehyde.Fixative B was obtained by mixing 1 vol. of 8% para-

Žformaldehyde and 1 vol. of sodium carbonate buffer 0.2.M, pH 11 . Both fixatives contained 0.2% picric acid. The

final result was 200 ml of 4% paraformaldehyde pH 6.0Ž .fixative A and 400 ml of 4% paraformaldehyde pH 10.5Ž .fixative B .

E. Brains were removed and post-fixed under condi-tions of agitation in fixative B for 4–5 h at room tempera-ture.

F. Brains were cut in a vibratome, obtaining 50 mmcoronal sections of the hypothalamus. Free-floating sec-

Ž .tions were collected in sodium veronal buffer SVB andŽ .rinsed several times 5=10 min . SVB was prepared by

dissolving 4.12 g sodium veronal, 17 g NaCl and 0.2 gthymerosal in 980 ml distilled water, checked and adjustedto pH 7.4 by adding 1 N HCl. As detailed in Section 7, thesaturation, incubation and wash steps were carried out atroom temperature and under conditions of shaking, exceptfor step K, which was performed at 48C.

G. Sections were immersed in an SVB solution contain-ing 0.5% casein for 30 min. In order to avoid non-specificstaining, this solution was used to dilute all the antibodies.

H. Some sections were incubated in rabbit anti-neuro-Ž .tensin anti-NT diluted 1:5000 and others in rabbit anti-TH

diluted 1:20 000. Both incubations were carried out overŽ .16 h. Then sections were rinsed in SVB 4=10 min and

incubated in biotinylated secondary antibody to rabbit IgGdiluted 1:2000.

Ž .I. After exhaustive rinses in SVB 4=15 min , sectionswere incubated for 90 min in peroxidase-conjugatedstreptavidin diluted 1:2000.

ŽJ. Free-floating sections were washed in SVB 4=15. Ž .min and in 0.05 M Tris buffer, pH 7.6 10 min . Then

peroxidase product was demonstrated by incubation of theX Žsections for 4–6 min in 3,3 -diaminobenzidine 0.025 g in

.100 ml Tris buffer and 0.3% hydrogen peroxide. Reactionwas verified under the microscope and stopped by washing

Ž .in SVB 4=15 min .Sections were then saturated as described in step G.K. In the second labeling, sections first treated with

Ž .anti-NT diluted 1:5000 were now incubated in antibodyŽ . Žagainst TH diluted 1:10 000 and vice versa sections

( )P. Marcos et al.rBrain Research Protocols 2 1997 1–8 3

incubated with anti-TH diluted 1:20 000 in the first reac-tion were treated with anti-NT diluted 1:5000 in the sec-

.ond reaction . Sections were incubated at 48C under gentleshaking for 48 h.

The procedures to which the different sections weresubjected thus differed according to the developing methodused.

4.1. Colloidal gold-silÕer intensified method

L. Sections were washed in SVB for 1 h. Next theywere incubated in 1 nm gold-conjugated goat IgG anti-rab-bit diluted 1:400 in SVB containing 0.5% casein and 0.1%bovine serum albumin for 4 h. Then they were rinsed over

Ž .1 h in 0.1 M phosphate buffer PB , pH 7.4.M. Sections were post-fixed for 10 min in 0.1 M PB,

pH 7.4 containing 2.5% glutaraldehyde. Then they wereŽ .rinsed in PB and quickly in distilled water 3=1 min .

The reaction was enhanced over 5 min with a silverenhancement kit of colloidal gold probes. Enhancementwas stopped by quick rinse in distilled water and in 0.1 MPB, pH 7.4, for 15 min.

4.2. BDHC method

ŽN. After step K, sections were rinsed in SVB 3=10.min and incubated for 90 min in biotinylated secondary

antibody to rabbit IgG diluted 1:2000. Step I was thenperformed as previously described. After rinses in SVBŽ . Ž .3=10 min and in 0.01 M PB, pH 6.8 5=10 min ,sections were pre-incubated for 10 min in a BDHC solu-tion prepared as follows: 10 mg of BDHC were added to95 ml distilled water and stirred for 30 min before filter-ing. Next, 25 mg of sodium nitroferricyanide and 5 ml of0.2 M sodium phosphate buffer were added just beforeuse. Reaction was started by adding 200 ml of 3% hydro-gen peroxide in fresh BDHC solution. The progress of thereaction was checked under a microscope. After 5–10 min,dark blue crystals were detected, and the reaction wasstopped by exhaustive rinsing in 0.01 M PB, pH 6.8Ž .5=10 min .

Ž .O. Sections showing double labeling NT and THobtained with either method described above were pro-cessed for electron microscopy. Sections developed by theBDHC method were post-fixed for 45 min at room temper-ature in 1% osmium tetroxide diluted in 0.1 M PB, pH 6.8,whereas post-fixation of the silver-enhanced colloidal goldtreated sections was carried out in 1% osmium tetroxidediluted in 0.1 M PB, pH 7.4, for 30 min. After dehydrationin several graded alcohols and propylene oxide, sectionswere flat-embedded in Epon 812 and placed between twoslides for 14–16 h at 378C for them to polymerize.

Fig. 1. Ultrastructural appearance of the A14 catecholaminergic cell group. A cell body labeled for TH with the gold-silver technique receives a synapticŽ . Ž .contact arrow from a NT-immunoreactive terminal showing dense-cored vesicles arrowhead . =32 000.


ŽP. Samples of both arcuate nucleus A12 catecholamin-.ergic cell group and A14 hypothalamic dopaminergic cell

groups were removed and attached to Epon supports forfurther sectioning. Ultrathin sections were counterstainedwith lead citrate and observed under an electron micro-scope.

5. Results

In this study, two hypothalamic regions were examined:the A14 dopaminergic cell group, in the periventriculararea of the anterior hypothalamus, and the A12 arcuatecatecholaminergic cell group.

At the level of electron microscopy, the followingproducts were observed: DAB staining resulted in a diffuseprecipitate throughout the structures containing NT or THŽ .Figs. 1–3 . Those stained by BDHC procedure showed acharacteristic precipitate in the form of crystals which

Ž .were not located on any specific organelle Fig. 3 andŽwere equally dense whatever the stained structure den-

drites, dendritic spines, cell bodies, axons, and axon termi-.nals . The product of the silver-enhanced colloidal gold

method appeared as electrodense round grains distributed

Ž . Ž .throughout the cell bodies Fig. 1 , dendrites Fig. 2 ,Ž . Ž .axons Fig. 2 and axon terminals not shown .

Synaptic profiles were more easily seen when theDABrcolloidal gold procedure was used, although identi-cal results were found using one or the other double-label-

Ž w x .ing method see Ref. 8 for details .Injection of colchicine did not affect the number or the

distribution of TH-immunoreactive perikarya. However,observation of NT-positive cells is possible only after thistreatment, and exclusively in the A12 arcuate cell group.

Ž . Ž .Axo-somatic Fig. 1 and axo-dendritic Fig. 2 synapseswere detected between NT-positive fibers and all the TH-immunoreactive neurons observed in the A14 region, al-though in the arcuate nucleus some of the individualprofiles of TH-immunoreactive neurons did not receiveany NT-positive input.

In this region, but not in the A14 dopaminergic cellgroup, synaptic contacts between TH-containing boutons

Ž .and NT-immunoreactive perikarya Fig. 3 were frequentlyobserved.

This ultrastructural analysis also disclosed the presenceof axo-dendritic and axo-somatic synaptic contacts be-tween NT-positive terminals and NT-containing cell bod-ies, as well as between TH-immunoreactive boutons and

Ž .Fig. 2. NT-positive terminal labeled by DAB establishes a synaptic contact arrow with a TH-immunoreactive dendrite from the A14 periventricularŽ .hypothalamic nucleus. The arrowhead points to dense-cored vesicles in the NT-immunoreactive fiber. An axon asterisk is also labeled for TH by the

gold-silver method but receives no synaptic input from NT-positive terminals. =19 300.


Ž .Fig. 3. An arcuate NT-immunoperoxidase-BDHC-labeled neuron receives a synaptic contact from a TH-immunoperoxidase-DAB-labeled fiber arrow . AsŽ .in all structures stained for NT, dense-cored vesicles are visible arrowhead . =31 600.

TH-positive neurons. NT-immunoreactive structures con-tained some dense-cored vesicles, whereas TH-positiveprofiles did not. In general, TH-immunoreactive boutonswere smaller than NT-positive terminals.

6. Discussion

Ž .The main outcomes of this work are: 1 the identifica-tion of two different antigens in the same pre-embeddingsection for electron microscopy by means of two antibod-

Ž .ies obtained in the same species and 2 the preservation ofoptimal tissue ultrastructure using a fixative without glu-taraldehyde.

6.1. Troubleshooting

The main problem was to find an adequate fixative foruse in this double-labeling method. The fixatives mostcommonly used in electron microscopy contain both para-formaldehyde and glutaraldehyde. The problem was that

Ž .even with a very small amount of glutaraldehyde 0.01% ,NT immunoreactivity completely disappeared from thetissue. We therefore attempted to perfuse animals with

only 4% paraformaldehyde containing 0.2% picric acid,pH 7.4, as for light microscopy; however, the ultrastruc-tural appearance of the tissue pointed to considerabledamage. When the animals were perfused with two fixa-tives, both made up with 4% paraformaldehyde but at

Ždifferent pH first fixative pH 6.0 and second fixative pH.10.5 , the NT immunostaining was preserved and the

ultrastructure of the tissue was optimal. This fixation pro-w xtocol has been previously described by Berod et al. 1 for

light microscopic studies, but to our knowledge this is thefirst time that it has been used for electron microscopy.

Another problem was the fragility of the median emi-nence. This structure was usually broken, as were parts ofthe arcuate nucleus. It was necessary to search for asuitable cutting speed on the vibratome in order to avoidsuch breakage. In general, no broken sections resultedwhen the block was slowly cut.

The appropriate dilution for each antibody must bepreviously tested, especially when two different antibodiesŽ .against NT and TH obtained in the same species areused. The immunological properties of both NT and TH

w xantibodies have been described previously 8,11 . The dilu-tions used in this study were identical to those previously

w xreported 8 . Staining specificity was assured by means of


separate immunocytochemical procedures for each anti-body. The results obtained here were in agreement with

w xthose reported in a previous work 7 at the level of lightmicroscopy, in which we carried out double staining withantibodies obtained in different species. The results ofincubations carried out by omitting one or the other of theprimary antibodies were identical to those obtained withseparate immunocytochemical procedures. In sum, the ab-sence of cross-reactions was guaranteed.

The DAB developing method was always used first. Asw xpreviously described 10 , this reaction enables a second

immunocytochemical reaction on the same section withoutantibody removal, even when both primary antibodies areobtained in the same species. This is due to the DABreaction product, which masks the antigen and catalyticsites of the first sequence of immunoreagents and thusprevents interaction with the reagents of the second se-quence. However, there are also technical reasons in favorof using the DAB developing procedure first. When thecolloidal gold method was carried out first, the seconddeveloping with DAB masked the gold particles and it wasimpossible to differentiate the two methods. The samephenomenon was observed when the BDHC procedurewas used first, since crystals of BDHC changed to agranular deposit similar in appearance to the final product

w xobtained with the DAB method 4 . No double-labeledstructures were detected.

ŽThe products of the different developing methods DAB.and BDHC were readily distinguishable at the light micro-

scopic level, allowing the reactions to be checked. Thesilver-enhanced colloidal gold method involved someproblems because its product is extremely weak for obser-vation at the light microscopic level. Accordingly, theappropriate times of incubation with both 1 nm gold-con-

Ž .jugated goat anti-rabbit IgG 4 h and the silver enhance-Ž .ment kit of colloidal gold probes 5 min were established

after several attempts with different periods. Worse resultsŽ .weak immunoreactivity or high background were ob-served when the times used were different.

The ultrathin sections obtained from the most superfi-cial portions of the tissue showed the best differentiallabeling, because this region has maximal access to anti-bodies and gold and silver enhancers. However, ultrastruc-tural morphology was less optimal than that observed insections taken from greater depth. In any case, the ultra-structural appearance of the tissue obtained from superfi-cial regions was acceptable even when glutaraldehydeŽ .normally used in electron microscopic techniques wasnot added. NT immunoreactivity disappeared when glu-taraldehyde was added to the fixative. Moreover, the mor-phology of the tissue was not adequate when animals wereonly perfused with neutral 4% paraformaldehyde. How-ever, using the fixation method described here we solvedboth problems: the ultrastructural appearance of the tissuewas optimal and NT immunostaining was preserved.

Finally, in order to obtain the best contrast when DAB

labeling was used, ultrathin sections were stained with leadcitrate, contrast being reduced when ultrathin sections werestained with uranyl acetate.

6.2. AlternatiÕe and support protocols

There are several developing procedures for the electronw xmicroscopic level 2–4,9,13 which use 5 nm gold particles

w x Ž .13 or the tetramethylbenzidine TMB developing methodw x9 . We did not choose these procedures for the followingreasons.

Ž .1 Colloidal gold particles smaller than 5 nm showed amore optimal labeling due to their greater penetration

w xcapacity 2 . In addition, the use of the smaller goldŽ .particles 1 nm reduces the severity of membrane damage

w xcaused by tissue penetration of these particles 13 . Ultra-structural appearance was also improved by post-fixing in2.5% glutaraldehyde for 10 min before silver enhance-ment. This step was not carried out in the study by Van

w xden Pol 13 . This procedure minimized the damage at-tributed to disruption of extracellular spaces during expo-sure to unbuffered water rinses before and after silverenhancement.

Ž .2 Other developing methods, such as TMB, requireincubation of the sections in a DAB-cobalt solution inorder to stabilize the TMB reaction product during osmica-

w xtion 9 , because this reaction is extremely unstable andeventually disappears. Moreover, it is difficult to controlthis step: the different labeled structures were not distin-guishable under the light microscope because the DAB-cobalt solution rapidly turned both antigens blue-black. Inaddition, the TMB procedure sometimes discloses non-

w xspecific aggregates in the nucleoplasm 5,6 . Following thetechniques described in this paper, we found no suchartifacts, and hence the stabilization step could be avoided.Thus, our protocol is shorter and easier than the TMBprocedure. Our method is also more advantageous, since

Ž .the pH of the BDHC reaction pH 6.8 or the gold-silverŽ .enhanced procedure pH 7.4 is not very different from pH

Ž .used for incubations and washes in our protocol pH 7.6 .By contrast, in the TMB developing method the pH usedwas only 6.0. This acid pH is detrimental not only forultrastructural preservation but also for optimal retentionof antigen-antibody interactions in immunohistochemistryw x3 .

A comparison of our results with other double-labelingmethods in which DAB and colloidal gold-silver enhanced

w xwere used as developing procedures 2 will show that ourmethod has several advantages.Ø The perfusion of the animals is easier, because acrolein

is not necessary to obtain good ultrastructural preserva-tion of the tissue, even when fixation is not carried outwith glutaraldehyde. In our case, both antigens areperfectly detectable without acrolein and glutaralde-hyde.

Ø There is no need to treat the sections with sodium


borohydride, since the brains have not been perfusedwith acrolein-containing fixative.

Ø The different buffers used in our study are easier tow xprepare than those in the protocol of Chan et al. 2 .

Those authors used 0.1% gelatin-PBS, BSA for incuba-tions with goat anti-rabbit IgG bound to 1 nm colloidalgold. We show that this incubation can be carried outwith only SVB containing 0.5% casein and 0.1% BSA,immunocytochemical detection being optimal withoutthe use of gelatin.

w xØ In comparison with the protocol used by Chan et al. 2 ,the technique described here allows a higher dilution of

Žthe antibodies primary, biotinylated, IgG bound to.colloidal gold, etc. and shorter incubation periods with

silver enhancer.Moreover, in this study we offer several details to

improve the DABrsilver-enhanced colloidal gold methodw x w x2,13 and DABrBDHC procedure 3,4 for double im-munocytochemical staining in pre-embedding sections forelectron microscopy. The main advantages of our proce-dure are:Ø The location at cellular and subcellular levels of two

Ž .different substances in this case, NT and TH , althoughthe primary antibodies were obtained in the same speciesand the antigenic sites were fragile.

Ø The optimal preservation of the ultrastructure, althoughthe perfusion was not carried out with glutaraldehyde oracrolein, two substances normally used in electron mi-croscopic immunocytochemistry.Although the results on immunolabeling were similar

with both developing procedures, the synaptic profileswere more easily visible with the DABrcolloidal goldmethod. In any case, the results obtained usingDABrBDHC were of such good resolution that we havealso included them in our methods for double-stainingprocedures at both light and electron microscopic levels.Our procedure provides a very good resolution with anacceptable ultrastructural morphology of the tissue.

Under the conditions reported here, the DABrBDHCdeveloping method is not adequate to perform co-existencestudies; both DAB and BDHC deposits were so intensethat it would be difficult to resolve both chromogens in thesame structure at the electron microscopic level. However,the method might be adapted for co-localization studies,and a combination of both chromogens could be used tostudy the localization of a pair of antigens, despite thecombination of peroxidase methods employed: direct, indi-rect, PAP, avidin-biotin-HRP, streptavidin coupled to per-oxidase, or even axonally transported HRP, as has been

w xpreviously reported 3,4 . Our procedure may also be usedto identify antigens concentrated in vesicles when vesiclemembranes are rendered permeable after using methodssuch as sonication. Finally, two antigens can also beidentified by an adapted method for pre-embedding peroxi-

Ž .dase or silver-intensified colloidal gold staining followedby post-embedding labeling.

7. Quick procedure

Note: Steps G–N were carried out with sections beinggently agitated. Incubations and washes were carried out atroom temperature, except for step K, which was done at48C. In order to avoid non-specific staining, we used a

Ž .solution of sodium veronal buffer SVB containing 0.5%casein to dilute all antibodies, as well as the sections instep G. Some animals did not receive intracerebroventricu-lar injections of colchicine, manipulation being initiatedinstead at step C.

A. Anesthetize the animals with sodium pentobarbitalby means of an intraperitoneal injection.

B. Inject colchicine and let the animal survive 48 h.C. Anesthetize with sodium pentobarbital as in step A.D. Perfuse with rinse and fixation fluids through the

ascending aorta.Ž .E. Post-fix in the second fixative pH 10.5 for 4–5 h.

F. Cut 50 mm coronal sections with a vibratome andrinse extensively.

G. Saturate sections in a solution of SVB containing0.5% casein.

H. Incubate in primary and biotinylated secondary anti-sera.

I. Incubate in peroxidase-conjugated streptavidin.J. Develop peroxidase according to the 3,3X-diamino-

benzidine method.K. Second labeling: process sections containing neu-

Ž .rotensin NT by immunostaining with primary antibodiesŽ .against tyrosine hydroxylase TH .

After step K the protocols differed, depending onwhether the sections were developed by the colloidal goldmethod or by the BDHC method.

7.1. Colloidal gold-silÕer enhanced method

L. Wash in SVB and incubate some sections in 1 nmgold-conjugated goat IgG anti-rabbit.

M. Post-fix in buffer containing 2.5% glutaraldehydeand intensify with a silver enhancement kit of colloidalgold probes.

7.2. BDHC procedure

N. Incubate remaining sections with biotinylated sec-ondary antibody and peroxidase-conjugated streptavidin,rinse in 0.01 M PB, pH 6.8, and pre-incubate in a solutionof BDHC. Start reaction by adding hydrogen peroxide infresh BDHC solution.

O. Post-fix all double-immunostained sections with os-mium tetroxide. Dehydrate in graded alcohols and propy-lene oxide, and then flat-embed in Epon 812.

P. Remove samples of hypothalamic region and attachthem to Epon supports for sectioning. Stain ultrathin sec-tions with lead citrate and observe under electron micro-scope.


8. Essential literature references

w xOriginal papers: Refs. 1–4,6–10,13 .

Acknowledgements

The authors gratefully acknowledge N. Skinner andProf. Cooper R. Mackin for manuscript review and thetechnical assistance of M. Chaigniau. P.M. is supported by

Ža grant from the Spanish Government D.G.I.C.Y.T.; Pro-.grama F.P.U. .

References

w x1 A. Berod, B.K. Hartman, J.F. Pujol, Importance of fixation inimmunohistochemistry: use of formaldehyde solutions at variablepH for the localization of tyrosine hydroxylase, J. Histochem.

Ž .Cytochem. 29 1981 844–850.w x2 J. Chan, C. Aoki, V.M. Pickel, Optimization of different immuno-

gold-silver and peroxidase labeling with maintenance of ultrastruc-ture in brain sections before plastic embedding, J. Neurosci. Meth-

Ž .ods 33 1990 113–127.w x3 S. Lakos, A.I. Basbaum, Benzidine dihydrochloride as a chromogen

for single- and double-label light and electron microscopic immuno-Ž .cytochemical studies, J. Histochem. Cytochem. 34 1986 1047–

1056.w x4 A.I. Levey, J.P. Bolam, D.B. Rye, A.E. Hallanger, R.M. Demuth,

M.-M. Mesulam, B.H. Wainer, A light and electron microscopicprocedure for sequential double antigen localization using di-

aminobenzidine and benzidine dihydrochloride, J. Histochem. Cy-Ž .tochem. 34 1986 1449–1457.

w x5 R. Magoul, P. Dubourg, W. Benjelloun, G. Tramu, Synaptic inputsof tachykinin-containing nerve terminals to target tyrosine hydroxyl-ase-, b-endorphin- and neuropeptide Y-producing neurons of thearcuate nucleus. Double pre-embedding immunocytochemical study

Ž .in the rat, J. Chem. Neuroanat. 6 1993 419–429.w x6 R. Magoul, P. Dubourg, O. Kah, G. Tramu, Ultrastructural evidence

for synaptic inputs of enkephalinergic nerve terminals to targetŽ .neurons in the rat arcuate nucleus, Peptides 15 1994 883–892.

w x7 P. Marcos, M. Corio, R. Covenas, G. Tramu, Neurotensin and˜tyrosine hydroxylase in the tuberoinfundibular system of the guineapig hypothalamus with special emphasis to lactation status, Peptides

Ž .17 1996 139–146.w x8 P. Marcos, M. Corio, P. Dubourg, G. Tramu, Reciprocal synaptic

connections between neurotensin- and tyrosine hydroxylase-im-munoreactive neurons in the mediobasal hypothalamus of the guinea

Ž .pig, Brain Res. 715 1996 63–70.w x9 R.B. Norgren Jr., M.N. Lehman, A double-label pre-embedding

immunoperoxidase technique for electron microscopy using di-aminobenzidine and tetramethylbenzidine as markers, J. Histochem.

Ž .Cytochem. 37 1989 1283–1289.w x10 L.A. Sternberger, S.A. Joseph, The unlabeled antibody method:

contrasting colour staining of paired pituitary hormones withoutŽ .antibody removal, J. Histochem. Cytochem. 27 1979 1424–1429.

w x11 J.M. Studler, P. Kitabgi, G. Tramu, D. Herve, J. Glowinski, J.P.Tassin, Extensive colocalization of neurotensin with dopamine in ratmeso-cortical-frontal dopaminergic neurons, Neuropeptides 11Ž .1988 95–100.

w x12 J.S. Tindal, The forebrain of the guinea pig in stereotaxic coordi-Ž .nates, J. Comp. Neurol. 124 1965 259–266.

w x13 A.N. Van den Pol, Silver-intensified gold and peroxidase as dualultrastructural immunolabels for pre- and post-synaptic neurotrans-

Ž .mitters, Science 228 1985 332–335.

protocol double immunocytochemistry in pre-embedding ... · protocol double immunocytochemistry in...

Documents