Ž .Brain Research 749 1997 358–362

Short communication

Expression of mRNA for brain-derived neurotrophic factor in the dorsal rootganglion following peripheral inflammation

Hee-Jung Cho ) , Sang-Yul Kim, Mae-Ja Park, Dong-Sun Kim, Jeong-Ki Kim, Mi-Young ChuDepartment of Anatomy, School of Medicine, Kyungpook National UniÕersity, 2-101 Dongin Dong, Taegu, Korea

Accepted 23 December 1996

Abstract

Ž .It is well known that the nerve growth factor NGF may serve as a link between inflammation and hyperalgesia. Recent experimentsŽ .showed that systemic injection of NGF dramatically stimulated the expression of brain-derived neurotrophic factor BDNF mRNA in the

Ž .dorsal root ganglion DRG . In the present study, we evaluated the change of BDNF mRNA in the DRG following peripheralinflammation and also observed colocalization of BDNF and trkA mRNAs by means of in situ hybridization histochemistry in rats.Peripheral tissue inflammation produced by an intraplantar injection of Freund’s adjuvant into the paws significantly increased BDNFmRNA levels in the DRG and many neurons expressing trkA mRNA showed increased expession of BDNF mRNA. Intraplantar injectionof antibody to NGF together with Freund’s adjuvant prevented the increase in BDNF mRNA. These findings suggest that peripheralinflammation induces an increased expression of BDNF mRNA which is mediated by NGF in DRG. q 1997 Elsevier Science B.V. Allrights reserved.

Keywords: Brain-derived neurotrophic factor; Nerve growth factor; mRNA; Dorsal root ganglion; Peripheral inflammation; Pain; In situ hybridization

A wealth of evidence have indicated that nerve growthŽ .factor NGF might be a mediator of the persistent pain

states associated particularly with inflammatory conditionw x14 . Experimental inflammation produced by an intraplan-tar injection of formalin or complete Freund’s adjuvant inthe hindpaws of rats was found to increase the level ofNGF many-fold above normal levels in the inflamed tissuew x8,20,23 . An enhanced retrograde transport of NGF by the

Ž . w xsciatic nerve to the dorsal root ganglion DRG occurs 8where it can lead to a rapid and large increase in the

Ž .production of neuropeptide such as substance P SP andŽ .calcitonin gene-related peptide CGRP at the level of gene

w xexpression 15,16 . The increases in the level of thesew xpeptides appear to lead hyperalgesia 13,14 . Furthermore,

it has been reported that administration of a syntheticprotein, trkA-IgG, blocked the hyperalgesia that developed

w xwith peripheral inflammation 18 .w xApfel et al. 3 reported that systemic injection of NGF

dramatically and rapidly stimulated the expression ofŽ .brain-derived neurotrophic factor BDNF mRNA in DRG

neurons and that many of these BDNF-expressing neurons

) Ž .Corresponding author. Fax: q82 53 427-1468.

also expressed trkA, the high-affinity NGF receptor, sug-gesting a paracrine function for BDNF in maturity. How-ever, it is not known whether elevated levels of endoge-nous NGF, such as seen following peripheral tissue inflam-mation, can also stimulate increased BDNF mRNA in theDRG. In the present study, we evaluated the change ofBDNF mRNA in the DRG following peripheral inflamma-tion and also observed colocalization of BDNF and trkAmRNAs.

Male Sprague–Dawley rats weighing 200–250 g werekept under a 12 h lightr12 h dark cycle and given foodand water ad libitum. The animals received a subcutaneous

Ž .injection of 100 ml of complete Freund’s adjuvant Sigmadiluted 1:1 with saline or only saline in the plantar of the

Ž .left ipsilateral hindpaws. Adjuvant injected rats weresacrificed 12 h, 1 day, 2 days, 4 days or 7 days laterŽ .ns23, 4 or 5 at each time point and only saline injected

Ž .ones were killed 1 day later ns4 . To determine whetherNGF mediates the change of BDNF mRNA associatedwith inflammation, a dermal injection of 100 ml of the

ŽFreund’s adjuvant mixed 1:1 with NGF antibody providedby Dr. Robert A. Rush of Flinders University of South

.Australia was adminstered to the plantar of the left hind-Ž .paws of rats ns4 and these animals were sacrificed 1

0006-8993r97r$17.00 Copyright q 1997 Elsevier Science B.V. All rights reserved.Ž .PII S0006-8993 97 00048-6

( )H.-J. Cho et al.rBrain Research 749 1997 358–362 359

day later. Ipsilateral and contralateral L5 DRGs wererapidly removed and immediately frozen on dry ice. Sec-tions of 5 mm were cut on a cryostat and thaw-mountedonto gelatin-chrome alum-subbed microscope slides andstored at y708C until they were processed.

Oligonucleotide probes were purchased from KoreaŽ .Biotech. Inc. Taejun, Korea . Sequences for probes were

complementary to nucleotides 1198–1245 of trkA mRNAw x w x19,22 and nucleotides 250–298 of BDNF mRNA 12,24 .The probes were 3X end labeled using 35S-labeled deoxy-

w x Ž .adenosine a-thio triphosphate New England Nuclear andŽterminal deoxynucleotidyl transferase Boehringer

.Mannheim . The specific activities of the resultant probeswere 10–15=103 Cirmmol. Several control experimentsfor specificity of hybridization signals were performed;

Ž .they were: 1 competition experiment using 100-fold ex-cess amount of unlabeled probe together with the labeled

Ž .probe; 2 hybridization using labeled sense sequenceŽ . Ž .probe; and 3 RNase A pretreatment 20 mgrml just

before hybridization.

The hybridization procedure was essentially the same asw xthat described by Bloch et al. 5 . Frozen sections were

fixed in 4% paraformaldehyde in 0.1 M phosphate bufferŽ .pH 7.2 for 5 min at room temperature. Sections were

Ž .rinsed 3 times 5 min each in 4=sodium chloride-sodiumŽ . Žcitrate buffer SSC; pH 7.0 1=SSC contained 0.15 M

.sodium chloride and 0.015 M sodium citrate , immersedŽfor 1 h 4=SSC containing 1=Denhardt’s solution 1=

Denhardt’s solution contained 0.02% bovine serum albu-min, 0.02% Ficoll 400 and 0.02% polyvinylpyroridone

.K-30 , and dehydrated with a graded ethanol series. Thesections were treated with chloroform for 5 min to remove

Žfat from the tissue and immersed in 100% ethanol twice 5.min each before hybridization. Hybridization was per-

wformed by incubating the sections with a buffer 4=SSC,Ž50% deionized formamide, 0.12 M phosphate buffer pH

.7.2 , 1=Denhardt’s solution, 2.5% tRNA, 10% dextranx w 35 xsulfate, 50 mM dithiothreitol containing a- S dATP la-

Ž 6 .beled probes 6–9=10 dpmrml for 24–48 h at 378C.After hybridization, the sections were rinsed in 4=SSC

Ž . Ž . Ž . Ž .Fig. 1. Dark-field photomicrographs A and B show BDNF mRNA-positive neurons in the contralateral A and ipsilateral B DRGs L5 1 day afterŽ .peripheral tissue inflammation. An increase in the expression of mRNA for BDNF can be seen on the ipsilateral side. Bright-field micrographs C and D

Ž . Ž . Ž .show a pair of adjacent sections, 5 mm thick, of the DRG L5 hybridized with probe for trkA mRNA C or for BDNF mRNA D 1 day after peripheralŽ . Ž . Ž .inflammation. Many neurons arrows contain either trkA mRNA or BDNF mRNA. Scale bar, 200 A, B and 50 mm C, D .

( )H.-J. Cho et al.rBrain Research 749 1997 358–362360

Ž .pH 7.0 for 10 min and then rinsed 3 times in 1=SSC at558C for 20 min each time. Sections were then dehydrated

Ž .with a graded ethanol series 70–100% and coated withŽ .NTB-2 emulsion dilute 1:1 with water . These sections

were exposed for 6 weeks in a tightly sealed dark box at48C. After being developed in D-19 developer, fixed withphotographic fixer, and washed with tap water, the sec-tions were counterstained with thionine solution. In thepresent study, neurons which had grain density at least 5times higher than the background density were consideredto show positive expression. Background grain densitieswere obtained for each section from fifty 15=15 mm2

areas of emulsion away from the labeled neurons. Forcalculating the percentage of BDNF mRNA-positive neu-ron profiles in the DRG, about 500 neuron profiles fromthe five sections of each DRG of each animal were ana-

Žlyzed. The grain density mean background value wassubtracted from the observed number of grains per 15=15

2 .mm neuronal area showing positive expression. wascalculated over 200 positive cells from five sections fromeach DRG and a mean value was calculated for that DRG.In this case, all positive cells in each field randomlyselected at a magnification of 400= were counted. In

Žcolocalizing studies, montages of photomicrographs mag-.nification, 400= were prepared from pairs of adjacent

sections processed for hybridization with trkA and BDNFprobes. For calculating the percentage of trkA mRNA-positive neurons colocalizing BDNF mRNA, about 100trkA mRNA-positive neurons from each DRG were ana-lyzed. All individual values for each DRG of the ipsilateralŽ .ns4 or 5 or contralateral side at each time point wereaveraged to provide a single mean"S.E.M. Comparisonby paired t-test was made on the ipsilateral and contralat-eral DRGs at each time point.

Numerous BDNF or trkA mRNA-positive neurons werefound in the DRGs on ipsilateral and contralateral sidesafter the injection of complete Freund’s adjuvant into theleft hindpaws. The majority of the neurons showing BDNFor trkA mRNA-positive neurons were among the small andmedium sized cells. In the ipsilateral DRG, inflammationinduced a significant increase in the number and the grain

density of BDNF mRNA-positive neurons. The increasewwas observed at 1 day the number of positive neurons, a

Ž .272% P-0.01 increase; the grain density, a 178%Ž . x ŽP-0.01 increase , 2 days the number of positive neu-

Ž .rons, a 253% P-0.01 increase; the grain density, aŽ . . w187% P-0.01 increase , and 4 days the number of

Ž .positive neurons, a 113% P-0.01 increase; the grainŽ . xdensity, a 86% P-0.01 increase after injection of

adjuvant, and the levels were back to normal at 7 daysŽ .Fig. 1A,B, Table 1 . There was no significant change inthe number and the grain density of trkA mRNA-positiveneurons in the ipsilateral DRG following inflammation and

w xthis result was similar to our previous one 7 . Analysis ofserial sections revealed that the percentage of the numberof trkA mRNA-positive neurons colocalizing BDNF

wmRNA increased significantly at 1 day, contralateral DRG,14.2"1.6% of trkA mRNA-positive neurons; ipsilateral

x Ž .DRG, 67.4"4.2%; P-0.01 Fig. 1C,D . In the controlanimals, no increase of the number and the grain density ofBDNF or trkA mRNA-positive neurons or trkA mRNA-

Žpositive neurons colocalizing BDNF mRNA the number.of these neurons, 12.6"1.5% could be observed. Intra-

plantar injection of antibody to NGF together with Freund’sadjuvant prevented the significant increase in the numberand the grain density of BDNF mRNA-positive neurons.

The present study demonstrated that peripheral tissueinflammation increased BDNF mRNA levels in the DRGand that many neurons expressing trkA mRNA showedincreased expression of BDNF mRNA. Additionally, localinjections of antibody to NGF together with Freund’sadjuvant were found to prevent the increases in BDNFmRNA. These findings suggest that peripheral inflamma-tion induces an increased expression of BDNF mRNAwhich is mediated by NGF in DRG.

It has been shown that NGF administration protectsagainst both cisplatin- and taxol-induced experimental neu-

w xropathies in mice 1,2 . Cisplatin neurotoxicity manifest asw xa neuropathy affecting the large diameter neurons 6,21w xand most large diameter neurons lack trkA receptors 4,25 .

NGF may have paracrine effects on non-trkA mRNA-bearing DRG neurons mediated by local release of BDNF

Table 1The percentage and the grain density of DRG neurons expressing BDNF mRNA following Freund’s adjuvant injection into the hindpaws

Ž . Ž . Ž . Ž . Ž .12 h ns5 1 day ns5 2 days ns5 4 days ns4 7 days ns4

Percentage contra 12.8"0.7 13.8"0.9 15.2"1.1 13.2"0.9 14.1"1.2a a aipsi 14.1"1.3 51.3"3.0 53.6"3.2 28.1"2.0 15.7"1.5

% increase 272% 253% 113%Grain contra 10.9"1.1 12.3"1.2 10.3"1.2 12.1"1.0 14.1"1.0

a a adensity ipsi 8.5"0.9 34.2"2.7 29.6"2.3 23.6"1.9 12.2"1.1% increase 178% 187% 86%

For calculating the percentage of mRNA-positive neuron profiles in the DRG, about 500 neuron profiles from the five sections of each DRG of eachŽ 2 .animal were analyzed. The grain density number of grains per 15=15 mm neuronal area was calculated over 200 cells from five sections from each

Ž . Ž .DRG and a mean value was calculated for that DRG. All individual values for each DRG of the contralateral contra or ipsilateral ipsi side at each timepoint were averaged to provide a single grand mean"S.E.M. Significance between the contralateral and ipsilateral sides was examined by paired t-test.a P-0.01.

( )H.-J. Cho et al.rBrain Research 749 1997 358–362 361

from DRG neurons because NGF treatment in vivo in-w xcreases the level of BDNF mRNA in the DRG 3 . Consid-

ering these results, peripheral tissue inflammation, whichcauses an increase in the level of NGF, may have sometrophic effects on non-peptidergic, large diameter neuronsmediated by BDNF release in the DRG. However, on theother hand, it has been demonstrated that BDNF is trans-ported not only retrogradely to, but also anterogradelyfrom, the DRG to terminals in the periphery and spinal

w xcord 27 . Increased anterograde transport of BDNF incentral nerve processes of the DRG following peripheralinflammation may act directly on postsynaptic neuronswithin the dorsal horn because this area contains neurons

w xexpressing trkB mRNA and trkB immunoreactivity 11,26 .Alternatively, increased release of BDNF from the axonterminals in the dorsal horn may modulate synaptic trans-

w xmission 10,17 . Furthermore, because the peripheral in-flammation-stimulated BDNF synthesis takes place in manytrkA bearing DRG neurons which are known to contain

w xCGRP or SP 4,9,25 , it is possible that BDNF may coexistwith these peptides in the DRG neurons and that BDNFmay have effects on the functions of these peptides relatedwith nociception.

Acknowledgements

We thank Dr. Robert A. Rush, Department of Physiol-ogy and Centre of Neuroscience, Flinders University ofSouth Australia, for his helpful comments on themanuscript. This study was supported by a grant of ’96Good Health R&D Project, Ministry of Health and Well-

Ž .fare R.O.K. .

References

w x1 Apfel, S.C., Arezzo, J.C., Lipson, L. and Kessler, J.A., Nervegrowth factor prevents experimental cisplatin neuropathy, Ann. Neu-

Ž .rol., 31 1991 76–80.w x2 Apfel, S.C., Lipton, R.B., Arezzo, J.C. and Kessler, J.A., Nerve

growth factor prevents toxic neuropathy in mice, Ann. Neurol., 29Ž .1991 87–90.

w x3 Apfel, S.C., Wright, D.E., Wildeman, A.M., Dormia, C., Snider,W.D. and Kessler, J.A., Nerve growth factor regulates the expres-sion of brain-derived neurotrophic factor mRNA in the peripheral

Ž .nervous system, Mol. Cell. Neurosci., 7 1996 134–142.w x4 Averill, S., McMahon, S.B., Clary, D.O., Reichardt, L.F. and Priest-

ley, J.V.P., Immunocytochemical localization of trkA receptors inchemically identified subgroups of adult rat sensory neurons, Eur. J.

Ž .Neurosci., 7 1995 1484–1494.w x5 Bloch, E.A., Popovici, T., Le Guellec, D., Normand, E., Chouham,

S., Guitteny, A.F. and Bohlen, P., In situ hybridization histochem-istry for analysis of gene expression in the endocrine and centralnervous system tissue: a three year experience, J. Neurosci. Res., 16Ž .1986 183–200.

w x6 Boogered, W., Bokkel Huinink, W.W., Dalesio, O., Hoppenbrouw-

ers, W.J.F. and Van der Sande, J., Cisplatin induced neuropathy:central, peripheral and autonomic nerve involvement, J. Neuroon-

Ž .col., 9 1990 255–263.w x7 Cho, H.J., Park, E.W., Bae, M.A. and Kim, J.K., Expression of

mRNAs for preprotachykinin and nerve growth factor receptors inthe dorsal root ganglion following peripheral inflammation, Brain

Ž .Res., 716 1996 197–201.w x8 Donnerer, J., Schuligoi, R. and Stein, C., Increased content and

transport of substance P and calcitonin gene-related peptide insensory nerve innervating inflamed tissue: evidence for a regulatory

Ž .function of growth facror in vivo, Neuroscience, 49 1992 693–698.w x9 Kashiba, H., Ueda, Y. and Senba, E., Coexpression of prepro-

tachykinin-A, a-calcitonin gene-related peptide, somatostatin, andneurotrophin receptor family messenger RNAs in rat dorsal root

Ž .ganglion neurons, Neuroscience, 70 1996 179–189.w x10 Kim, H.G., Wang, T., Olafsson, P. and Lu, B., Neurotrophin-3

potentiates neuronal activity and inhibits gamma-aminobutyratergicsynaptic transmission in cortical neurons, Proc. Natl. Acad. Sci.( ) Ž .USA , 91 1994 12341–12345.

w x11 Klein, R., Martin-Zanca, D., Barbacid, M. and Parada, L.F., Expres-sion of the tyrosine kinase receptor gene trkB is confined to themurine embryonic and adult nervous system, DeÕelopment, 109Ž .1990 845–850.

w x12 Leibrock, J., Lottspeich, F., Hohn, A., Hofer, M., Hengerer, B.,Masiakowski, P., Thoenen, H. and Barde, Y.A., Molecular cloningand expression of brain-derived neurotrophic factor, Nature, 341Ž .1989 149–152.

w x13 Levine, J.D., Fields, H.L. and Basbaum, A.I., Peptides and theŽ .primary afferent nociceptor, J. Neurosci., 13 1993 2273–2286.

w x14 Lewin, G.R. and Mendell, L.M., Nerve growth factor and nocicep-Ž .tion, Trends Neurosci., 16 1993 323–359.

w x15 Lindsay, R.M., Lockett, C., Sternberg, J. and Winter, J., Neuropep-tide expression in cultures of adult sensory neurons: modulation ofsubstance P and calcitonin gene-related peptide levels by nerve

Ž .growth factor, Neuroscience, 33 1989 53–65.w x16 Lindsay, R.M. and Harmar, A.J., Nerve growth factor regulates

expression of neuropeptide genes in adult sensory neurons, Nature,Ž .337 1989 362–364.

w x17 Lohof, A.M., Ip, N.Y. and Poo, M., Potentiation of developingneuromuscular synapses by the neurotropins NT-3 and BDNF, Na-

Ž .ture, 363 1993 350–353.w x18 McMahon, S.B., Bennett, D.L.H., Priestley, J.V. and Shelton, D.L.,

The biological effects of endogenous nerve growth factor on adultsensory neurons revealed by a trkA-IgG fusion molecule, Nature

Ž .Med., 1 1995 774–780.w x19 Martin-Zanca, D., Oskam, R., Mitra, G., Copeland, T. and Barbacid,

M., Molecular and biochemical characterization of the human trkŽ .protooncogene, Mol. Cell. Biol., 9 1989 24–33.

w x20 Otten, U., Nerve growth factor: a signalling protein between thenervous and the immune system. In A.I. Basbaum and J.M. BessonŽ .Eds. , Towards a New Pharmacology of Pain, John Wiley, Chich-ester, 1991, pp. 353–363.

w x21 Thompson, S.W., Davis, L.E., Kornfeld, M., Hilgers, R.D. andStandefer, J.C., Cisplatin neuropathy: clinical, electrophysiologic,

Ž .morphologic and toxicologic studies, Cancer, 54 1984 1269–1275.w x22 Verge, V.M.K., Merlio, J.P., Grondin, J., Ernfors, P., Persson, H.,

Riopelle, R.J., Hokfelt, T. and Richardson, P.M., Colocalization of¨NGF binding sites, trk mRNA, and low-affinity NGF receptormRNA in primary sensory neurons: responses to injury and infusion

Ž .of NGF, J. Neurosci., 12 1992 4011–4022.w x23 Weskamp, G. and Otten, U., An enzyme-linked immunoassay for

Ž .nerve growth factor NGF : a tool for studying regulatory mecha-nisms involved in NGF production in brain and in peripheral tissues,

Ž .J. Neurochem., 48 1987 1779–1786.w x24 Wetmore, C. and Olson, L., Neuronal and nonneuronal expression of

neurotrophins and their receptors in sensory and sympathetic ganglia

( )H.-J. Cho et al.rBrain Research 749 1997 358–362362

suggest new intercellular trophic interactions, J. Comp. Neurol., 353Ž .1995 143–159.

w x25 Wright, D.E. and Snider, W.D., Neurotrophin receptor mRNA ex-pression defines distinct populations of neurons in rat dorsal root

Ž .ganglia, J. Comp. Neurol., 351 1995 329–338.w x26 Zhou, Z.F., Parada, L.F., Soppet, D. and Rush, R.A., Distribution of

trkB tyrosine kinase immunoreactivity in the rat central nervousŽ .system, Brain Res., 622 1993 63–70.

w x27 Zhou, X.F. and Rush, R.A., Endogenous brain-derived neurotrophicfactor is anterogradely transported in primary sensory neurons,

Ž .Neuroscience, 74 1996 945–951.