Journal of Neuroscience Methods 87 (1999) 195–200

Labelling of peptides with 1.4-nm gold particles to demonstrate theirbinding sites in the rat spinal cord

Klaus Schmidt, Gisela Segond von Banchet 1, Bernd Heppelmann *

Physiologisches Institut der Uni6ersitat Wurzburg, Rontgenring 9, D-97070 Wurzburg, Germany

Received 8 July 1998; received in revised form 9 December 1998; accepted 13 December 1998

Abstract

Recently we presented a method to label the neuropeptide substance P with a 1.4-nm gold particle covalently bound at theN-terminus that can be used for demonstrating its binding sites in histological sections. In this study we examined whether thepeptides neuropeptide Y, somatostatin, calcitonin gene-related peptide and bradykinin can be labelled in the same way.Polyacrylamide gel electrophoresis revealed a reduction in mobility for peptide–gold conjugates over gold particles aloneconsistent with peptide binding. In cryostat sections of the rat lumbar spinal cord, the peptides showed a distinct binding patternin the grey matter corresponding to data of studies using autoradiographic methods. Therefore, we conclude that this simple andfast method can be used for labelling peptides in general to demonstrate their binding sites in histological sections, provided thepeptide binds by its C-terminus. © 1999 Elsevier Science B.V. All rights reserved.

Keywords: Bradykinin; Calcitonin gene-related peptide; Neuropeptide Y; NHS-Nanogold; Peptide receptors; Rat; Somatostatin

1. Introduction

Peptides play an important role in the central andperipheral nervous system. In the innervated tissue,peptides such as bradykinin (Schaible and Grubb,1993), substance P (Heppelmann and Pawlak, 1997b)and somatostatin (Heppelmann and Pawlak, 1997a)affect the sensitivity of the peripheral sensory endings.In the spinal cord, peptides such as calcitonin gene-re-lated peptide, substance P and somatostatin are re-leased upon activation of primary afferents by differentstimuli (e.g. Morton et al., 1989; Schaible et al., 1990,1994).

Their bindings sites have been demonstrated by theuse of radioactively labelled ligands or by the use ofpeptides conjugated with non-radioactive reporter

molecules including colloidal gold (e.g. Mentlein et al.,1989; Anton et al., 1991; Krisch et al., 1993; Bowden etal., 1994). Recently, a method for non-radioactive la-belling was introduced using 1.4-nm gold particles(NHS-Nanogold™) covalently bound at the primaryamino group of the neuropeptide substance P (Segondvon Banchet and Heppelmann, 1995). In histologicalsections, binding sites could be shown in various partsof brain and spinal cord. This method has the advan-tages of a fast and simple labelling of the peptides, avery short procedure time, no need for an additionalimmunohistochemical step, no fading of the signal anda high spatial resolution.

The aim of the present study was to examine whetherthis gold compound can be used to label other peptidesas well. The labelling was performed on neuropeptide Y(NPY), somatostatin (SOM), calcitonin gene-relatedpeptide (CGRP) and bradykinin (BK). The success ofthe labelling was examined by polyacrylamide gel elec-trophoresis (PAGE). For a histological confirmation ofspecific binding, the distribution of binding sites was

* Corresponding author. Tel.: +49-931-31-2726; fax: +49-931-31-2741; e-mail: bernd.heppelmann@mail.uni-wuerzburg.de.

1 Present address: Physiologisches Institut der Universitat Jena,Teichgraben 8, D-07740 Jena, Germany.

0165-0270/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 5 -0270 (99 )00003 -5

K. Schmidt et al. / Journal of Neuroscience Methods 87 (1999) 195–200196

examined in the rat lumbar spinal cord and comparedwith data of previous autoradiographic studies.

2. Materials and methods

2.1. Preparation of the peptide–gold conjugates

In principle, the peptide–gold conjugates (CGRP–,SOM–, NPY– and BK–gold) were prepared as de-scribed earlier for substance P (Segond von Banchetand Heppelmann, 1995). In brief, 1 mmol of peptide(Sigma, Deisenhofen, Germany) was dissolved in 500 mlof HEPES (20 mmol, pH 7.5), added to 500 ml of 6nmol Nanogold reagent (NHS-Nanogold™; BioTrend,Koln, Germany) dissolved in double-distilled H2O(ddH2O), and incubated for 1 h at room temperatureby rotating the micro test tubes. Peptide–gold conju-gates were separated from excess unbound peptides bymembrane centrifugation (Microcon-10 system; Ami-con, Germany) with a membrane molecular cut-off of10 kDa. After separation, the peptide–gold conjugateswere diluted in 20 mmol sodium phosphate buffersaline (PBS plus 150 mmol NaCl, pH 7.4) containing 10mmol sodium azide, 1% (w/v) bovine serum albumin(BSA; Fluka, Deisenhofen, Germany), 4 mg/ml leu-peptin (Sigma, Deisenhofen, Germany) and 0.2 molsucrose, aliquoted and stored at −20°C for not longerthan 3 months.

2.2. Control of the peptide–gold preparation

To control the preparation and the purity of thepeptide–gold conjugate, we used sodium dodecyl sul-phate (SDS)–disk-PAGE separation with high concen-trations of acrylamide. Gels with a polyacrylamideconcentration of 20% were prepared and the NHS-Nanogold reagent and the peptide–gold conjugateswere separated by the method described originally byLaemmli (1970). All buffer solutions were made with-out b-mercaptoethanol, because the peptide–gold con-jugate degrades upon exposure to thiols.

The bands were visualised by a silver stain(Heukeshoven and Dernick, 1985). To compare therelative mobility, the lanes of the NHS-Nanogold andthe peptide–gold conjugate were scanned and analysedby an image analysing system (OPTIMAS; Scheer &Wabbel, Wurzburg, Germany).

2.3. Binding experiments on tissue sections

Four male Wistar rats (240–400 g) were killed by anoverdose of thiopentone (Trapanal). Segments of thelumbar spinal cord were removed and quickly frozen.Cryostat sections (30 mm) of the spinal cord werethaw-mounted on Vectabond-coated slides (Vector

Laboratories, Peterborough, UK) and stored at −70°C. Before the binding experiments, sections werefixed with 0.5% glutaraldehyde in PBS for 30 min. After5 min washing with PBS, they were incubated with 50mmol glycine in PBS for 30 min to block free aldehydegroups and thereafter with 5% BSA in PBS for 30 minto block unspecified binding sites within the tissue. Theslides were then covered with 0.6 nmol peptide–gold inPBS plus 0.1% acetylated bovine serum albumin (BSA-C; BioTrend, Koln, Germany), 40 mg/ml bacitracin, 4mg/ml leupeptin and 2 mg/ml chymostatin (Sigma,Deisenhofen, Germany). After an overnight incubationat 4°C in a moist chamber, sections were extensivelyrinsed in PBS plus 0.1% BSA-C and thereafter in PBS.The sections were postfixed with 2% glutaraldehyde inPBS for 2 min. After extensive washing with PBS andddH2O, the gold particle was intensified with silverenhancer (R-Gent, pH 5.5; BioTrend, Koln, Germany)for 25 min at 23°C. The reaction was stopped bywashing in ddH2O. Thereafter, sections were embeddedin DePeX and examined by darkfield microscopy. Thistechnique was used because the silver-enhanced goldparticles are still small and can be examined moreclearly with darkfield illumination.

2.4. Control incubations

Two controls were performed to determine the spe-cificity of the peptide–gold complexes used in the bind-ing studies. Unlabeled peptide (1 mmol) was added tothe peptide–gold incubation of the sections. Further-more, to determine any non-specific binding of theNanogold particles, some slides were incubated with 0.6nmol of NHS-Nanogold only.

3. Results

3.1. Demonstration of the peptide labelling

To demonstrate the labelling of the peptides with thegold particle, relative mobilities of the peptide–goldconjugates and of the gold reagent were compared byelectrophoresis. In each case, the relative mobility ofthe peptides incubated with the gold reagent was lowerthan the NHS-Nanogold, indicating a successful la-belling of the peptides (Fig. 1, left side). This findingwas further supported by a computerised analysis (Fig.1, right side).

3.2. Distribution of peptide binding sites in the dorsalhorn of the rat spinal cord

Incubations of spinal cord sections resulted in adistinct binding pattern of gold-labelled peptides in thegrey matter (Figs. 2–5). The addition of an excess of

K. Schmidt et al. / Journal of Neuroscience Methods 87 (1999) 195–200 197

Fig. 1. Comparison of the relative mobility of the peptides incubatedwith the gold reagent (lower bands) with that of the NHS-Nanogold(upper bands) by SDS–disk-PAGE (20%). All bands were alsoscanned and analysed by an image analysing system (right side).

Fig. 3. Binding sites of gold-labelled SOM in the dorsal horn of therat lumbar spinal cord shown by darkfield microscopy. Cells indeeper laminae with a dense binding at the soma are marked byarrows. Magnification, see Fig. 5.

horn. A moderate binding was present in the outer partof lamina II and in laminae III, IV and X (Fig. 2).Some larger cells in lamina V and some motor neuronesin the ventral horn also showed NPY binding sites.

A binding of gold-labelled SOM was found in thewhole grey matter. The highest density was seen inlaminae I–II (Fig. 3). A moderate density occurred inlaminae III and X. Some larger cells in lamina V andmotor neurones in the ventral horn also showed a clearbinding reaction.

CGRP binding sites were predominantly found inlamina III of the dorsal horn and in lamina X (Figs. 4

unlabelled peptide completely suppressed the binding ofgold-labelled peptides. An example is presented in Fig.6 showing a control incubation of gold-labelled SOM inthe presence of an excess of unlabelled SOM. Incuba-tions with the Nanogold substrate also obtained nospecific staining of the tissue.

Gold-labelled NPY bound in the whole grey matterof the spinal cord. A very dense binding occurred inlamina I and the inner part of lamina II of the dorsal

Fig. 4. Binding sites of gold-labelled CGRP in the dorsal horn of therat lumbar spinal cord shown by darkfield microscopy. Cells indeeper laminae with a dense binding at the soma are marked byarrows. Magnification, see Fig. 5.

Fig. 2. Binding sites of gold-labelled NPY in the dorsal horn of therat lumbar spinal cord shown by darkfield microscopy. Magnifica-tion, see Fig. 5.

K. Schmidt et al. / Journal of Neuroscience Methods 87 (1999) 195–200198

Fig. 5. Binding sites of gold-labelled BK in the dorsal horn of the ratlumbar spinal cord shown by darkfield microscopy. Bar: 0.5 mm.

Fig. 7. Higher magnification of cells with a dense binding of gold-la-belled CGRP (see Fig. 4). Bar: 100 mm.

possible to produce specific antibodies against the dif-ferent subtypes of peptide receptors. As this immuno-histochemical detection can be performed at sections offixed tissue, the structural preservation of the tissue isvery good. However, the production of the antisera isvery laborious and time-consuming.

In most studies labelled peptides have been used todemonstrate their binding sites. These methods areperformed on unfixed tissue in order to preserve thereceptor–ligand affinity binding. The advantage of us-ing radioactive isotopes is that almost every ligandmolecule can be labelled. However, it requires radio-chemical facilities and working with radioactively la-belled ligands has the disadvantage of commonlyrequiring exposure times of up to several weeks.

Therefore, non-radioactive methods have been devel-oped using peptides conjugated with biotin orfluorochromes (e.g. Anton et al., 1991; Bowden et al.,1994). However, for detection of biotin an additionalimmunohistochemical step is necessary. Fluorochromescan be visualised directly. The disadvantages of thismethod are that a fluorescence microscope must beused and that fluorochromes tend to fade very quickly.

Another non-radioactive labelling was introduced us-ing colloidal gold (e.g. Mentlein et al., 1989; Krisch etal., 1993). However, in this method the gold particle isnot bound covalently to the peptide at a specific part ofthe molecule. In addition, the labelling technique needssome experience.

The present study shows that a variety of peptidescan be labelled with NHS-Nanogold to demonstratetheir binding sites in neuronal tissue. As the molecularweight of the NHS-Nanogold complex is increased by

and 8). In the other parts of the spinal cord binding wasmoderate. Some cells in lamina V and motor neuroneswere also stained (Fig. 7).

BK showed a less-dense binding compared to theother peptides examined, but it also bound to all partsof the spinal cord. The highest densities of binding werein laminae I and II of the dorsal horn (Fig. 5). Bindingwas also seen at some motor neurones.

4. Discussion

Various methods can be used to demonstrate thedistribution of peptide receptors in histological sections.With the development of molecular biology it became

Fig. 6. Control incubation using gold-labelled SOM in the presence ofan excess of unlabelled SOM. Bar: 0.5 mm. Fig. 8. Binding of gold-labelled CGRP in lamina X. Bar: 100 mm.

K. Schmidt et al. / Journal of Neuroscience Methods 87 (1999) 195–200 199

adding a peptide, polyacrylamide gel electrophoresiswas used to examine whether the peptides have beenlabelled successfully. In each case a clear reduction ofthe relative mobility was observed after the incubationof the NHS-Nanogold with a peptide.

In a second set of experiments we examined whetherthe gold-labelled peptides can be used to demonstratetheir binding sites in the rat lumbar spinal cord. Thisneuronal tissue was used because several studies havealready shown peptide binding sites byautoradiography.

Using radioactively labelled NPY, binding sites weredetected in all parts of the dorsal horn of the rat spinalcord with highest densities in laminae I and II (Kar andQuirion, 1992). Similar results were obtained by Zhanget al. (1995), who found NPY binding sites mainlylocated in laminae I–IV and X with the highest densityin lamina I–II. The results of the present study corre-spond with those data, as the main binding of gold-la-belled NPY was found in lamina I–II and X. Thisbinding, however, only represents the distribution ofthe Y2-receptor, as the Y1-receptor requires intact N-and C-termini of the NPY. The Y2-receptor, on theother hand, makes strong demands on the C-terminusonly (Grundemar and Hakanson, 1993). The binding ofthe gold-labelled NPY indicates that labelling at theN-terminus does not prevent receptor recognition.

Using gold-labelled SOM a distinct pattern of bind-ing sites was found in the grey matter of the rat lumbarspinal cord. This pattern corresponds to previous au-toradiographic studies, which showed a high density ofSOM binding sites in laminae I and II of the dorsalhorn (Reubi and Maurer, 1985; Sato et al., 1991). Inaddition, moderate amounts of labelling were found inlaminae III and X (Kar and Quirion, 1995). Althoughreceptor autoradiography did not show binding sites inthe ventral horn of the spinal cord (Reubi and Maurer,1985), mRNA of one SOM receptor subtype (SSTR3)was demonstrated in motor neurones of the rat (Senariset al., 1995).

The pattern of CGRP binding differed from those ofthe other peptides, as the main binding occurred inlaminae III and X. This finding corresponds to datafrom studies using a radioactively labelled peptide.They found binding sites mainly localised in laminaeIII–VI and in lamina X, whereas only moderate toweak CGRP binding was observed in the superficiallayers (Kar et al., 1993; Rossler et al., 1993).

Radioactively labelled BK bound in all parts of thespinal cord with highest density in lamina II (Lopes etal., 1995). A moderate density was seen in lamina III,and low but still significant densities were found inlamina I and deeper in the dorsal horn (laminae IV andV), the ventral horn and in lamina X. A few motorneurones also showed a binding of radioactively la-belled BK. In the present study a comparable distribu-tion of binding was obtained.

5. Conclusion

This study shows that various peptides can be la-belled with NHS-Nanogold covalently bound at theirN-terminus. Those peptides can then be used to demon-strate their binding sites in histological sections, pro-vided the receptor–ligand interaction is based on anintact C-terminus only.

Acknowledgements

The authors thank B. Beckmann for skilful assistancewith the laboratory work, and Prof. R.F. Schmidt forcritically reading the manuscript. This work was sup-ported by the Deutsche Forschungsgemeinschaft (He1919/2-3).

References

Anton PA, Reeve JR Jr., Vidrich A, Mayer E, Shanahan F. Develop-ment of a biotinylated analogue of substance P for use as areceptor probe. Lab Invest 1991;64:703–8.

Bowden JJ, Garland AM, Baluk P, Lefevre P, Grady EF, Vigna SR,Bunnett NW, McDonald DM. Direct observation of substanceP-induced internalization of neurokinin 1 (NK1) receptors at sitesof inflammation. Proc Natl Acad Sci USA 1994;91:8964–8.

Grundemar L, Hakanson R. Multiple neuropeptide Y receptors areinvolved in cardiovascular regulation. Peripheral and centralmechanisms. Gen Pharmacol 1993;24:785–96.

Heppelmann B, Pawlak M. Inhibitory effect of somatostatin on themechanosensitivity of articular afferents in normal and inflamedknee joints of the rat. Pain 1997a;73:377–82.

Heppelmann B, Pawlak M. Sensitisation of articular afferents innormal and inflamed knee joints by substance P in the rat.Neurosci Lett 1997b;223:97–100.

Heukeshoven J, Dernick R. Simplified method for silver staining ofproteins in polyacrylamide gels and the mechanism of silverstaining. Electrophoresis 1985;6:103–12.

Kar S, Quirion R. Quantitative autoradiographic localization of[125I]neuropeptide Y receptor binding sites in rat spinal cord andthe effect of neonatal capsaicin, dorsal rhizotomy and peripheralaxotomy. Brain Res 1992;574:333–7.

Kar S, Quirion R. Neuropeptide receptors in developing and adultrat spinal cord: an in vitro quantitative autoradiography study ofcalcitonin gene-related peptide, neurokinins, m-opioid, galanin,somatostatin, neurotensin and vasoactive intestinal peptide recep-tors. J Comp Neurol 1995;354:253–81.

Kar S, Rees RG, Quirion R. Altered calcitonin gene-related peptide,substance P and enkephalin immunoreactivities and receptorbinding sites in the dorsal spinal cord of the polyarthritic rat. EurJ Neurosci 1993;6:345–54.

Krisch B, Buchholz C, Mentlein R, Turzynski A. Visualization ofneuropeptide-binding sites on individual telencephalic neurons ofthe rat. Cell Tissue Res 1993;272:523–31.

Laemmli UK. Cleavage of structural proteins during the assembly ofthe head of bacteriophage T4. Nature 1970;227:680–5.

Lopes P, Kar S, Chretien L, Regoli D, Quirion R, Couture R.Quantitative autoradiographic localization of [125I-TYR8]bradykinin receptor binding sites in the rat spinal cord:effects of neonatal capsaicin, noradrenergic deafferentation, dor-sal rhizotomy and peripheral axotomy. Neuroscience1995;68:867–81.

K. Schmidt et al. / Journal of Neuroscience Methods 87 (1999) 195–200200

Mentlein R, Buchholz C, Krisch B. Binding and internalization ofgold-conjugated somatostatin and growth hormone-releasinghormone in cultured rat somatotropes. Cell Tissue Res1989;258:309–17.

Morton CR, Hutchison WD, Hendry IA, Duggan AW. Somato-statin: evidence for a role in thermal nociception. Brain Res1989;488:89–96.

Reubi JC, Maurer R. Autoradiographic mapping of somatostatinreceptors in the rat central nervous system and pituitary. Neuro-science 1985;15:1183–93.

Rossler W, Gerstberger R, Sann H, Pierau F-K. Distribution andbinding sites of substance P and calcitonin gene-related peptideand their capsaicin-sensitivity in the spinal cord of rats andchicken: a comparative study. Neuropeptides 1993;25:241–53.

Sato H, Ota Z, Ogawa N. Somatostatin receptors in the senescentrat brain: a quantitative autoradiographic study. Regul Pept1991;33:81–92.

Schaible H-G, Grubb BD. Afferent and spinal mechanisms of jointpain. Pain 1993;55:5–54.

Schaible H-G, Jarrott B, Hope PJ, Duggan AW. Release of im-munoreactive substance P in the spinal cord during development

of acute arthritis in the knee joint of the cat: a study withantibody microprobes. Brain Res 1990;529:214–23.

Schaible H-G, Freudenberger U, Neugebauer V, Stiller RU. In-traspinal release of immunoreactive calcitonin gene-related pep-tide during development of inflammation in the joint in vivo-astudy with antibody microprobes in cat and rat. Neuroscience1994;62:1293–305.

Segond von Banchet G, Heppelmann B. Non-radioactive localiza-tion of substance P binding sites in rat brain and spinal cordusing peptides labeled with 1.4-nm gold particles. J HistochemCytochem 1995;43:821–7.

Senaris RM, Schindler M, Humphrey PPA, Emson PC. Expressionof somatostatin receptor 3 mRNA in the motorneurones of therat spinal cord, and the sensory neurones of the spinal cord.Mol Brain Res 1995;29:185–90.

Zhang X, Ji RR, Nilsson S, Villar M, Ubink R, Ju G, Wiesenfeld-Hallin Z, Hokfelt T. Neuropeptide Y and galanin binding sitesin rat and monkey lumbar dorsal root ganglia and spinal cordand effect of peripheral axotomy. Eur J Neurosci 1995;7:367–80.

.