B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

ava i l ab l e a t www.sc i enced i r ec t . com

www.e l sev i e r . com/ loca te /b ra i n res

Research Report

Origins of endomorphin-2 immunopositive fibers andterminals in the rat medullary dorsal horn

Chao Zhua, 1, Rui Huia, b, 1, Tao Chena, 1, Zhong-Fu Zuoa, Wei Wanga, Chang-Jun Gaoc,Ting Zhanga, Ya-Yun Wanga, Hui Lia, Sheng-Xi Wua,⁎, Yun-Qing Lia,⁎aDepartment of Anatomy, Histology & Embryology & K. K. Leung Brain Research Centre, The Fourth Military Medical University,Xi'an 710032, PR ChinabDepartment of Neurosurgery, Navy General Hospital of PLA, Beijing 100037, PR ChinacDepartment of Anesthesiology, Tangdu Hospital, The Fourth Military Medical University, Xi'an 710032, PR China

A R T I C L E I N F O

⁎ Corresponding authors at: Department of AnMedical University, No. 169, West Chang-le R

E-mail addresses: shengxi@fmmu.edu.cnAbbreviations: ANOVA, one-way analysis

diaminobenzidine tetrahydrochloride; DMH,endomorphin-2-immunoreactive; FITC, fluorNDS, normal donkey serum; NS, normal steriphosphate-buffered saline; PNS, peripheralhypothalamic nucleus1 These authors contribute equally to this w

0006-8993/$ – see front matter © 2011 Elsevidoi:10.1016/j.brainres.2011.06.067

A B S T R A C T

Article history:Accepted 30 June 2011Available online 14 July 2011

Endomorphin-2-immunoreactive (EM2-IR) fibers and terminals are densely present in themedullary dorsal horn (MDH) and are key factors in regulating central nociceptiveprocessing. However, the origins of these EM2-IR fibers and terminals remain elusive. Itwas hypothesized that there were at least three possible origins of the EM2-IR fibers andterminals in the MDH: intrinsic dorsal horn neurons, primary afferent fibers, and projectionfibers from higher parts of the brain. Different kinds of measures were employed in thecurrent study to elucidate this hypothesis. After intracerebral ventricle administration ofcolchicine, no EM2-IR neuronal cell bodies were detected in the MDH, suggesting that therewas no intrinsic EM2-IR dorsal horn neuron. Disruption of bilateral primary afferents(exposed to the primary afferent neurotoxin, capsaicin) decreased bilateral EM2 expressionbut did not eliminate it. Transection of the trigeminal nerve sensory root significantlydecreased EM2 expression on the ipsilateral but not on the contralateral MDH. Afterinjecting FluoroGold (FG) into the MDH, FG retrogradely labeled some EM2-IR neurons in thebilateral hypothalamus and nucleus tractus solitarii (NTS), and some of the FG retrogradelylabeled neurons in the ipsilateral trigeminal ganglion also showed EM2-immunoreactivities.These results indicate that EM2-IR fibers and terminals in the MDH come not only fromipsilateral primary trigeminal afferents but also from bilateral fibers from the hypothalamusand NTS.

© 2011 Elsevier B.V. All rights reserved.

Keywords:EndomorphinMedullary dorsal hornHypothalamusNucleus tractus solitariiFluoroGoldModulation

atomy, Histology and Embryology & K. K. Leung Brain Research Centre, The Fourth Militaryoad, Xi'an 710032, PR China. Fax: +86 29 8328 3229.(S.-X. Wu), deptanat@fmmu.edu.cn (Y.-Q. Li).of variance; CGRP, calcitonin gene-related peptide; CNS, central nervous system; DAB,

dorsomedial hypothalamic nucleus; DRG, dorsal root ganglion; EM2, endomorphin-2; EM2-IR,escein isothiocyanate; FG, FluoroGold; MDH, medullary dorsal horn; MOR, μ-opioid receptor;le saline; NTS, nucleus tractus solitarii; PAG, periaqueductal gray; PB, phosphate buffer; PBS,nervous system; SDH, spinal dorsal horn; TG, trigeminal ganglion; VMH, ventromedial

ork.

er B.V. All rights reserved.

39B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

1. Introduction

Endomorphin (EM) 1 and EM2 have been shown to beendogenous ligands which are similar to morphine and havelikewise high selectivity and affinity for the μ-opioid receptor(MOR) (Wang, et al., 2003; Zadina, et al., 1997, 1999). Previousreports have described these peptides as being involved inmodulating various functions, e.g. neuroendocrine, cardio-vascular regulation, mood, feeding, sexual behavior and pain(Fichna et al., 2007b; Martin-Schild, et al., 1999; Pierce andWessendorf, 2000; Tseng, et al., 2000). EM1- and EM2-immunoreactive (-IR) structures are distributed in almost thesame regions as those of the MOR-IR structures, indicatingthat EM might exert functions via binding to the MOR (Aicher,et al., 2003; Ding, et al., 1996).

EM1- and EM2-IR structures have different distributionpatterns in the central nervous system (CNS). EM1-IR structuresare widely and densely distributed throughout the brain,whereas EM2-IR structures are more prevalent in the spinalcord than are the EM1-IR structures (Martin-Schild, et al., 1999;Pierce andWessendorf, 2000). Themedullary dorsal horn (MDH,also called the caudal subnucleus of the spinal trigeminalnucleus), which is structurally and functionally identical to thespinal dorsal horn (SDH), is mainly involved in the orofacialnociceptive information transmission and regulation (Bereiter,et al., 2000; Nomura, et al., 2002; Okada-Ogawa, et al., 2009). Inthe superficial laminaeof theMDH,dense aggregation of EM2-IRfibers and terminals is detected (Martin-Schild et al., 1997, 1999).These EM2-IR fibers and terminals in the MDH, which seem toterminate in the region of the trigeminal nucleus where axonsform the trigeminothalamic tract, might exert importantregulatory functions in the central pain pathway subservingthe head (Dubner and Bennett, 1983; Fields and Basbaum, 1978;Martin-Schild, et al., 1997). Previous studies have suggested thatendogenous EM2 released in the dorsal horn acts locally toinhibit pain transmission (Budai and Fields, 1998). In the MDH,EM2 is believed to selectively modulate noxious stimulus-evoked responses (Wang, et al., 2000). All of these observationssuggest that clarification of the origins of EM2-IR fibers andterminals in the MDH would greatly contribute to elucidatingthe mechanism of EM2 action in modulating orofacial paintransmission.

Although previous studies indicate that primary afferentfibers are important providers for EM2 in the superficiallaminae of SDH and MDH (Martin-Schild, et al., 1998; Pierce,et al., 1998), no relevant reports have systematically revealedorigins of the EM2-IR fibers and terminals in theMDH. As EM2-IR neuronal cell bodies are exclusively located in severalhypothalamic nuclei and nucleus tractus solitarii (NTS) of theCNS, and also in the trigeminal ganglion (TG) and dorsal rootganglion (DRG) of the peripheral nervous system (PNS) (Hui, etal., 2006; Martin-Schild, et al., 1999; Pierce and Wessendorf,2000), it is reasonable to hypothesize that EM2-IR fibers andterminals in the MDH might originate from these primaryafferents and/or brain structures. The intrinsic neurons in theMDH were also considered as a candidate for the possibleorigins and would be checked in the present study.

In order to clarify the sources of the EM2-IR fibers andterminals in the MDH, various techniques were used toelucidate the contribution of the three possible origins. The

trigeminal primary afferent was chemically disrupted (expo-sure to the primary afferent neurotoxin, capsaicin) or me-chanically disrupted (deafferentation by sensory trigeminalrhizotomy) with observation of the consequent expressions ofEM2. Colchicine was used as an axoplasm transport inhibitorto check possible intrinsic EM2-IR neurons in the MDH.Furthermore, FluoroGold (FG) was injected into the MDH,with observation of the retrogradely labeled EM2ergic originsin the supraspinal levels. Besides, the cell types of FGretrogradely labeled EM2-IR neurons in the TG were examinedby immunofluorescent histochemistry multi-staining withcalcitonin gene-related peptide (CGRP) or isolectin IB4 (IB4).

2. Results

2.1. Distribution of EM2-IR fibers and terminals in the MDH

We first observed EM2 expressions in theMDH of the naïve rats.In sections throughout the lower brain stem of the rats, quitedense EM2-IR fibers and terminals were observed in allsubnuclei of the NTS and superficial laminae (lamina I and II)of the MDH (Figs. 1A,B). At the ventral lateral surface of themedulla, an aggregate of fibers coursed into the lateral reticularnucleus. Some of these processes coursed dorsally toward thenucleus ambiguous (Fig. 1C). EM2 immunoreactivities in thesuperficial laminae of rostral open medulla were very weak(Fig. 1D).The distribution patterns of the EM2-IR fibers andterminals in thepresent studywere inaccordancewithpreviousreports (Martin-Schild, et al., 1999; Zadina, et al., 1999).

2.2. Effects of colchicine or capsaicin treatment on EM2expression in the MDH

After treatment with colchicine for 48 h, EM2-IR fibers andterminals located on the bilateral MDH were moderatelyreduced, although no statistical difference was observed com-pared with naïve rats (the optical densities of the naïve group orthe colchicine treated group were respectively 176.92±9.71 or152.35±8.87; n=6, P=0.087; Figs. 1E,K). However, no EM2-IRneuronal cell bodies were detected on the MDH of thesemedullary sections under the light microscope (Fig. 1F).

In order to detect the contribution of the primary afferentin the EM2-IR structures in the MDH, capsaicin was used todisrupt the bilateral primary sensory afferents. After applica-tion of capsaicin for 24 h, the optical density of EM2-IR fibersand terminals in bilateral MDH was 118.47±7.19, which wasstrikingly and symmetrically reduced compared with naïverats (176.92±9.71; Figs. 1G,K; n=6, P=0.009).

2.3. Effects of trigeminal sensory root transection on EM2expression in the MDH

After the unilateral trigeminal root was cut, the intensity ofEM2-IR fibers and terminals was apparently reduced in theMDH ipsilateral to the lesion (Fig. 2C). The optic density ofEM2-IR fibers and terminals of the MDH on the lesioned sidewas 110.35±7.67, which was significantly lower than that ofthe naïve rats (177.32±10.03; n=6, P=0.007) or the contralateralside (179.31±9.52; Fig. 2D; n=6, P=0.007). No apparent changes

Fig. 1 – Endomorphin 2 immunopositive (EM2-IR) fibers and terminals in the medullary dorsal horn (MDH) after treatment ofcolchicine or capsaicin. EM2-IR structures were observed in the NTS and the superficial laminae of the MDH in naïve rats (A–C).EM2-IR structures in the superficial laminae of the rostral openmedullawere veryweak (D). Even after the treatment of colchicine,no intrinsic EM2-IR neuronal cells were observed (E, F). Compared with the naïve group, the intensity of EM2-IR structures wasstrikingly and symmetrically reduced in the bilateralMDH, after application of capsaicin (G, K; n=6, P=0.009). No apparent changeswere observed in the vehicle group (H, K). No immunopositive structureswere detected in the absorption control experiment (I). J isthe schematic illustration of selected areas in the MDH processed for statistical analysis. B–C, F are the magnified images of therectangles respectivelyshowninAandE. *P<0.01comparedwith that ofnaïvegroup. t, spinal trigeminal tract.NTS,nucleus tractussolitarii. pyd, pyramidal decussation. Scale bars=1000 μm in A, D, E, G–I; 100 μm in B, C, F.

40 B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

were observed in the sham surgery group (Fig. 2B). Otherregions in which EM2 immunoreactivities were exhibited, e.g.the NTS, had no apparent changes (data not shown).

2.4. EM2/FG double-labeled neurons in the hypothalamus,NTS and TG after FG injection into the unilateral MDH

In Group 3, FG injections in 6 rats were successfully targeted tothe MDH (Fig. 3A).

2.4.1. EM2/FG double-labeled neurons in the hypothalamusIn the sections through the hypothalamus area, EM2-IRneuronal cell bodies were mainly encountered in the DMH,the ventromedial hypothalamic nucleus (VMH), the centro-medial hypothalamic region and in the arcuate nucleus of thehypothalamus. EM2-IR neurons had fusiform, elliptic or stellar

shaped cell bodies, and usually had several processes. Themajority of these neurons were intermediate in size and a fewwere small (Figs. 3B,C).

In the hypothalamic area, FG retrogradely labeled neuronswere bilaterally present with a strong ipsilateral predomi-nance in the hypothalamic regions, such as the paraventri-cular nucleus, dorsomedial hypothalamic nucleus (DMH), andthe arcuate nucleus of the hypothalamus. The size and shapeof FG retrogradely labeled neurons were variable in thehypothalamus. They were usually round, elliptical, triangularor fusiform in shape, and most of them were intermediatewhile a few had small diameters (Fig. 3D).

Some of the FG-labeled neurons also showed EM2-immu-noreactivities in the hypothalamus. EM2/FG double-labeledneurons were mainly found in the DMH (Figs. 3B–D). A few ofthem were also scattered in the VMH and the periventricular

Fig. 2 – Densities of EM2-IR fibers and terminals in the MDH decreased after transection of the sensory root of the trigeminalnerve. The optical density of EM-IR structures of the spinal trigeminal nucleus on the lesioned sidewas significantly lower thanthat of the naïve rats and the contralateral side (A, C, D; n=6, P<0.01 respectively). There were no significant changes observedin both sides of the sham group (B). *, #P<0.01 compared with that of naïve group and contralateral side, respectively. Scalebar=1000 μm.

41B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

hypothalamic nucleus. In the whole DMH region, the totalnumber of EM2-IR neurons and FG labeled neurons in each ratwas 78±5 and 112±9 respectively (Table 1). The percentage ofthe EM2/FG double-labeled neurons to the total number ofEM2-IR neurons and FG labeled neurons was respectively37.5%±3.2% and 25.3%±2.1% (Table 1).

2.4.2. EM2/FG double-labeled neurons in the NTSIn the sections through the NTS area, EM2-IR neuronal cellbodies exhibited small and intermediate neuronal somata,many of which were round and elliptical (Fig. 3E). FG-labeledneurons were found in all parts of the NTS, including themedial part, commissural part, and the lateral part of the NTS.

Fig. 3 – Fluorescence photomicrographs of sections showing fluophotomicrographs through the dorsomedial hypothalamic nucleuEM2-immunoreactivities (B, C, E) or labeledwith FG (D, F). Light grasmallest FG injection areas. The region demarcated with rectangidentical fields were taken under different filters. Double arrowharrowheads indicate EM2-IR neurons (C, E) and arrows show FG-la4th ventricle; cc, the central canal. Scale bars=500 μm in A; 200 μ

Neurons labeled with FG varied in size and shape, from smalland round or elliptical to intermediate and fusiform ortriangular (Fig. 3F). In the medial part and commissural partof the NTS, EM2/FG double-labeled neurons were detected(Figs. 3E,F). Statistical analysis showed that 92±9 neuronsshowed EM2-IR and 238±21 neurons were labeled with FG.Thus, EM2/FG double-labeled neurons were 16.8%±1.4% of thetotal number of EM2-IR neurons and were 6.9%±0.8% of thetotal number of FG retrogradely labeled neurons (Table 1).

2.4.3. EM2/FG double-labeled neurons in the TGA number of EM2-IR neurons were symmetrically distributedin the TG. Most of EM2-IR cell bodies (63%) were intermediate

ro-gold (FG) injection sites in the MDH (A) and fluorescences (DMH) (B–D) and the NTS (E–F) showing neurons exhibitingy and dark gray in (a) respectively indicate the largest and theles in B is shown at higher magnification in C and D. Theeads point to EM2/FG double-labeled neurons (C–F),beled neurons (D, F), respectively. 3v, the 3rd ventricle; 4v, them in B; 50 μm in C–F.

Table 1 – Quantity of EM2, FG, EM2/FG, EM2/FG/CGRP or EM2/FG/IB4 labeled neurons in the DMH, NTS and TG for FGinjections into the MDHa.

EM2 FG EM2/FG Triple-labeled

Mean±SD Mean±SD Mean±SD Rate1 (%) Rate2 (%) Mean±SD Rate3 (%)

DMH 78±5 112±9 31±4 37.5±3.2 25.3±2.1 – –NTS 92±9 238±21 16±2 16.8±1.4 6.9±0.8 – –TGCGRP 142±11 601±42 114±12 79.8±8.5 20.9±1.6 108±8 95.2±3.1TGIB4 152±13 623±38 121±13 77.4±8.1 21.6±1.8 8±1 5.8±0.5

a Counts were made on 5 randomly selected sections of each of six rats. SD, standard deviation. Rate1 and Rate2 are the percentages of thequantity of EM2/FG double labeled neurons relative to that of EM2 immunopositive or FG labeled neurons respectively. Rate3 is the percentage ofthe quantity of triple-labeled neurons with CGRP or IB4 to that of EM2/FG double labeled neurons.

42 B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

in size, some others (33%) small, and only a small percentage(4%) of EM2-IR neurons were large (Fig. 4A). In the ipsilateralganglion, a large number of FG-labeled neuronswere detected,while no FG-labeled neurons were observed in the contralat-eral ganglion. FG-positive neurons were also symmetricallydistributed in the TG and no regional preferencewas observed.The shape of FG-labeled neurons was round or elliptical(Fig. 4B). Counting results showed that 72% of the FG-labeledneurons in the TG exhibited intermediate neuronal somata,while the other 21% exhibited small neuronal somata, withonly a few (7%) exhibiting large neuronal somata.

A large number of EM2/FG double-labeled neurons wereevenly and diffusely distributed in the TG (Fig. 4D). Thesedouble-labeled cell bodies, which were mainly small tomedium in size, accounted for 79.8%±8.5% of EM2-IR neuronsand 20.9%±1.6% of FG-labeled neurons in the CGRP stained TGsections (in the IB4 stained TG sections, the number is 77.4%±8.1% or 21.6%±1.8% respectively; Table 1). All of the resultsshowed that themajority of EM2-IR neurons in TG projected tothe MDH and these EM2-IR neurons could be the main sourceof EM-2 positive fibers and terminals in the MDH. Besides,EM2/FG/CGRP or EM2/FG/IB4 triple-labeled neurons accountedfor 95.2%±3.1% or 5.8%±0.5% of EM2/FG double-labeledneurons respectively, which indicated that almost all EM2/FG double-labeled neurons were labeled by CGRP.

Fig. 4 – Fluorescence photomicrographs of rat trigeminal gangliomulti-labeled neurons after FG injections into the MDH. Double aArrowheads indicate EM2/FG double-labeled neurons without IB

3. Discussion

The present study sought to determine the origins of EM2-IRfibers and terminals within the MDH. There were possiblythree origins for EM2-IR fibers and terminals in the MDH:intrinsic dorsal horn neurons, primary central afferent termi-nals, and projection fibers from superior structures. Previouswork had demonstrated the possibility of peripheral originsfor EM2-IR fibers and terminals (Martin-Schild, et al., 1998), butthe other two possible origins remained unconfirmed. Theresults of our present study indicate that the remnant EM2-IRfibers and terminals in the MDH originate from superiorstructures rather than from intrinsic dorsal horn neurons.

After disruption of the primary sensory afferent by admin-istration of capsaicin, EM2-IR structures in the bilateral super-ficial laminae of the MDH were markedly reduced. The resultsentirely corresponded to the previous study (Martin-Schild,et al., 1998). Capsaicin selectively causes nociceptive primaryafferent C- and Aδ-fibers to be eliminated with a certain dose(Holzer, 1991), suggesting that primary sensory afferent is animportant contributor of the EM-2 in the superficial MDH.

Inorder to further clarify the contributionof primary sensoryafferent to EM2-IR fibers and terminals in the MDH, transectionof the sensory root of the trigeminal nerve was performed. As

n (TG) sections showing EM2/FG/CGRP or EM2/FG/IB4rrowheads point to EM2/FG/CGRP triple-labeled neurons (D).4 staining (H). Scale bar=50 μm.

Fig. 5 – Schematic summary of the projection relationship ofEM2-IR structures in the central nervous system. EM2-IRneurons in the hypothalamus (Hp) project to theperiaqueductal gray (PAG, a), parabrachial nucleus (PB, b),NTS(c) and the MDH (d). EM2-IR neurons in NTS sendprojections to the PAG (e), PB (f), MDH (g) and to the spinaldorsal horn (SDH, h). EM2-IR neurons in the TG and the dorsalroot ganglion (DRG) are respectively the main sources ofEM2-IR fibers and terminals in the MDH (i) and SDH (j).

43B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

predicted, the densities of EM2-IR fibers and terminals in thesuperficial dorsal horn of the lesioned side were inordinatelydecreased. In addition, after FGwas injected into theMDH, largenumbersof EM2-IRneuronswere retrogradely labeledwithFG inthe ipsilateral TG,while no FG labeledneuronswere observed inthe contralateral ganglion. All of these results support the viewthat most of EM2-IR fibers and terminals originate from theipsilateral primary afferent, probably the TG (Martin-Schild,et al., 1997). Interestingly, the EM2-IR structure in the lesionedside was not extinguished although it was reduced, suggestingthat theremay be other sources of EM-IR fibers and terminals inthe superficial MDH. As there was no FG retrogradely labeledneurons observed in any of the contralateral ganglions, thepossibility of contralateral primary afferent origin may beexcluded. We further checked the possible origins of theremnant EM2-IR fibers and terminals in the higher parts of thebrain.

It has been demonstrated by many reports that thehypothalamus and the NTS project to the superficial laminaeof not only the SDH but also the MDH (Gamboa-Esteves, et al.,2001; Kausz, 1990; Swanson and Kuypers, 1980). In the presentstudy, EM2/FG double-labeled neurons were detected in boththe hypothalamus and the NTS, indicating that the EM2-IRneurons in these regions send their axons to the superficiallaminae of the MDH. In another study, we observed that EM2-IR neurons in the NTS were retrogradely labeled aftertetramethyl rhodamine dextran-amine injection into thelumbar SDH, which indicated that EM2-IR neurons in theNTS not only sent axons to the MDH, but also to the SDH (Hui,et al., 2010). The hypothalamus is the center for controllingnumerous homeostatic and analgesic processes whereas theNTS is described as a main target of primary sensoryinformation termination for visceral and taste inputs in thebrainstem (Almeida, et al., 2000; Hardy, 2001; Monnikes, et al.,2003; Silva-Carvalho, et al., 1995; Ter, et al., 1989). Projectionfibers from these two regions might regulate nociceptiveinformation transmission by releasing EM2 and binding to theMOR-expressing neurons within the superficial laminae of theMDH (Aicher, et al., 2003). Given that the proportion of EM2/FGdouble-labeled neurons to the FG labeled or EM2 immunopo-sitive neurons as well as the number of that in thehypothalamus is much higher than that of the NTS respec-tively, it is assumed that the EM2-IR neurons in the hypothal-amus play more important roles in modulating analgesicfunctions than the neurons in the NTS. However, morephysiological data is needed to verify this assumption.

Our previous studies have shown that in addition tosending EM2-IR fibers to the MDH, EM2-IR neurons in thehypothalamus also send projections to the periaqueductalgray and the parabrachial nucleus (Chen, et al., 2004, 2008).Furthermore, there exist reciprocal EMergic connectionsbetween the hypothalamus and NTS (Hui, et al., 2006).Regarding the NTS, it also sends EM2-IR axons to PAG (Lü, etal., 2010) and the parabrachial nuclei (Lü, et al., 2009) (Fig. 5).These regions, which are part of the well-characterizednociceptive pathways (Fields and Basbaum, 1994; Guilbaud,et al., 1994), are known to be involved in the transmission ofthe nociceptive information by direct input from the primaryafferents and/or as relay nuclei to other pain-processingcircuits (Basbaum and Fields, 1984; Fields and Basbaum,

1978; Przewlocki, et al., 1999; Przewlocki and Przewlocka,2001). The EM2-containing neuronal elements and theirintercommunication were found in most regions of thespino (trigemino)-ponto-amygdaloid pathway, reflectingtheir potential endogenous involvement in many majorbiological processes, including perception, responses relatedto stress, and such complex functions as reward, arousal, andvigilance, as well as autonomic, cognitive, neuroendocrine,and limbic homeostasis (Fichna et al., 2007a).

As the inhibitor of the axoplasm transport, colchicinemarkedly reduces the volumeof axoplasmand thus accumulatesacertainsubstance in theperikarya (Hughes, et al., 1983). Previousstudies have shown that it is more likely for neurons to showEM2-immunoreactivity after experimentally utilizing colchicine(Chen, et al., 2004, 2008; Hui, et al., 2006). In order to inhibitaxoplasmtransport ofpossible intrinsicascendingprojectingEM2neurons in the MDH, the unilateral lateral ventricle was alsotargeted in the present study. However, no EM2-IR neuronal cellbody was found in the superficial laminae of the medullaoblongata, even after colchicine administration in the cerebello-medullary cistern and unilateral lateral ventricle, partiallyindicating that there are no intrinsic EM2-IR neurons in thisregion, which further confirms previous report (Martin-Schild, etal., 1999). What should be noted is that a colchicine blockade oftheprimary trigeminal afferent didn't induce a significant changecomparedwith naïve rats. Two possible reasonsmay account forthis phenomenon. Firstly, the drug concentration may not havebeen sufficient to inhibit axoplasm transport to induce asignificant change. Secondly, the survival time of the colchicine

44 B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

treated rats (48 h) may have been too short. However, it was notdetermined if larger dose of colchicine or if a longer survival timemight induce a statistically significant change.

Previous studies showed that almost all EM2-labeled cellbodies were CGRP positive and few were labeled with IB4 inspinalDRGneurons (Sanderson, etal., 2004).Our results indicatethat high fraction of EM2/FG double-labeled cell bodies immu-nostained with CGRP antibody, and low percentage of EM2/FGdouble-labeled neurons were labeled with IB4. High degree ofco-localization with peptide-containing neurons suggests thatthese EM2-IR neurons send projections to the MDH and maymodulate peptide neurotransmitter release via presynapticautoreceptors (Aicher, et al., 2003; Sanderson, et al., 2004).

In conclusion, the present results have provided morpho-logical evidence for originsofEM2-IR fibers and terminalswithintheMDH.No intrinsic EM2-IRneuronwasdetected in this study.In addition to the primary afferent from the ipsilateral sensorialganglia, the EM2-IR fibers and terminals in the superficiallaminae of the MDH originate from the bilateral superiorstructures, i.e. the hypothalamus and the NTS.

4. Experimental procedures

Forty-eight male Sprague–Dawley rats weighing 250–300 g wereused in the present study. All experimental protocols wereapproved by the Committee of Animal Use for Research andEducation of the Fourth Military Medical University (Xi'an,China), and all efforts were made to minimize the number ofanimals used and their suffering, in accordancewith the ethicalguidelines for animal use in pain studies (Zimmermann, 1983).All surgical procedures were performed under anesthesia withsodium pentobarbital (40–60 mg/kg, i.p.).

4.1. Surgical procedures

4.1.1. Application of colchicine or capsaicin (Group 1)Twenty-four rats were randomly divided into four subgroups(n=6 each) to receive the following treatment. In the colchicinesubgroup, 5–8 μl of 1% (w/v) colchicine (Sigma, St. Louis, MO,USA; dissolved innormal sterile saline,NS)was injected into theunilateral lateral ventricle and cerebellomedullary cistern(Martin-Schild, et al., 1999). In the capsaicin subgroup, 5–8 μl of0.15% (w/v) capsaicin solution (Sigma; vehicle: 10% ethanol, 10%Tween 80 and 80%NS) was injected into the cerebellomedullarycistern (Hui, et al., 2010). In the sham group, 5–8 μl of vehicle forcapsaicinwas injected,while thenaïvegroupwas leftuntreated.All rats were allowed to survive for 24 to 48 h and were thensubjected to the immunohistochemical staining for EM2.

4.1.2. Transection of the sensory root of the trigeminal nerve(Group 2)Eighteen rats were randomly divided into three subgroups (n=6each). In the transection subgroup, rats were anesthetized andthen the sensory root of the trigeminal nervewas transected onthe left side under sterile conditions. The surgical approach tothe root of the trigeminal nerve was through a ventromedianincision and a hole drilled in the base of the skull. At the pointwhere the trigeminal roots left theprotective bony ridge to enterthe ganglion, the motor root was carefully separated from the

sensory root.All these stepswere carriedoutunderanoperationmicroscope. In the sham surgery subgroup, the sensory root ofthe trigeminal nerve was separated but not transected. The ratswere allowed to survive for 10 days after the operation. All ofthese rats, subsequently, were processed for EM2 immunohis-tochemical staining in the lower medullary oblongata.

4.1.3. Injections of FluoroGold (Group 3)After anesthesia, 0.1 μl of 4% (w/v) FG (Fluorochrome, Denver,CO, USA) dissolved in distilled water was stereotaxicallyinjected into the MDH of 6 rats on the left side according tothe atlas of the rat brain (−14.28 mm from Bregma, 2.6 mmlateral to the midline and 8 mm deep from the brain surface)(Paxinos and Watson, 2005). Each injection was performedslowly and lasted 15–20 min and the injection needle was keptin place for another 20 min. Four to 5 days after the injection,5–8 μl of 1% (w/v) colchicine (Sigma) dissolved in NS wasinjected into the unilateral lateral ventricle and cerebellome-dullary cistern. Thirty-six to 48 h later, all rats were subjectedto immunohistochemical staining for EM2.

4.2. Immunohistochemistry

All rats were re-anesthetized with an overdose of sodiumpentobarbital and perfused through the ascending aorta with100ml of NS, followed by 500 ml of 4% (w/v) paraformaldehydein 0.1 M phosphate buffer (PB, pH 7.4). The brains of all rats andtheTGof FG-treated ratswere immediately removed (ipsilateralside was labeled by piercing a needle into the left ventralmedulla and the left temporal cortex of rats) and placed into thesame fresh fixative for 2 h (4 °C), and then saturated with 30%(w/v) sucrose in 0.1 M PB (pH 7.4) overnight at 4 °C. All of thelower brainstems of Groups 1 and 2 as well as brains and TGfrom Group 3 were then serially cut into 30 μm-thick frontalsections on a frozen microtome (Kryostat 1720; Leitz, Mann-heim, Germany). All sections were serially collected into 5dishes containing 0.01 M phosphate-buffered saline (PBS, pH7.4) as 5 sets of every fifth serial sections. Each dish contained aone-fifth set of serial sections.

The procedures for immunohistochemical staining of EM2-IR neurons, fibers and terminals were described in our previousreports (Chen, et al., 2004, 2008; Hui, et al., 2006, 2010; Li, et al.,1996). In Groups 1 and 2, the sections in the first dish weremounted onto gelatin-coated glass slides and processed forNissl staining. The sections in the second dish were incubatedwith rabbit antiserumagainst EM2 (1:200; Chemicon, Temecula,CA, USA) in 0.01 M PBS containing 5% (v/v) normal donkeyserum(NDS), 0.3% (v/v)TritonX-100, 0.05% (w/v)NaN3and0.25%(w/v) carrageenan (PBS-NDS, pH 7.4) at 4 °C. 48 h later, allsections of the second dish were incubated with biotinylatedgoat anti-rabbit IgG (1:200; Vector, Burlingame, CA, USA) dilutedin PBS-NDS for another 4 h, and then with avidin–biotin–peroxidase complex (ABC) Elite Kit (1:100; Vector) in 0.01 M PBS(pH7.4) for 1 h. Betweeneachstep, the sectionswere completelywashed with 0.01 M PBS (pH 7.4). Finally, the sections werereacted with 0.05 M Tris–HCl buffer (pH 7.6) containing 0.04%diaminobenzidine tetrahydrochloride (DAB) (Dojin, Kumamoto,Japan) and 0.003% H2O2 for visualizing EM2-IR structures.

In Group 3, the first set of brain sections was mounted ontogelatin-coated glass slides and processed for Nissl staining. The

45B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

second set of sections wasmounted onto clean glass slides andused to examine the FG injection sites in the MDH and thedistributions of FG retrogradely labeled neurons in the hypo-thalamus, NTS and TG. The third set of the sections through thehypothalamus, lower brainstem (containing NTS) was used forimmunofluorescence histochemical staining of EM2. The sec-tions were sequentially incubated at 4 °C in 3 steps: (1) rabbitantiserum against EM2 (1:200; Chemicon) in PBS-NDS for 72 h;(2) biotinylated goat anti-rabbit IgG (1:200; Vector) in PBS-NDSfor 4 h; (3) Texas Red-labeled avidin D (1:500; Vector) in 0.01 MPBS (pH 7.4) for 4 h. The third set and the fourth set of TGsections were used for immunofluorescence histochemicalmulti-staining of EM2/CGRP or EM2/IB4 respectively. Briefly,the third set of TG sectionswas sequentially incubated at 4 °C in3 steps: (1) rabbit antiserum against EM2 (1:200; Chemicon) andgoat antiserumagainst CGRP (1:3000; Abcam, San Francisco, CA,USA) in PBS-NDS for 72 h; (2) biotinylated goat anti-rabbit IgG(1:200; Vector) and Alexa488-conjugated donkey antibody togoat IgG (1:500; Invitrogen, Carlsbad, CA, USA) in PBS-NDS for4 h; (3) Texas Red-labeled avidin D (1:500; Vector) in 0.01 M PBS(pH 7.4) for 4 h. The fourth set of TG sections was sequentiallyincubated at 4 °C in 3 steps: (1) rabbit antiserum against EM2(1:200; Chemicon) in PBS-NDS for 72 h; (2) biotinylated goat anti-rabbit IgG (1:200; Vector) and fluorescein isothiocyanate (FITC)-labeled rabbit antiserumto IB4 (1:200;Vector) inPBS-NDS for 4 h;(3) Texas Red-labeled avidin D (1:500; Vector) in 0.01 M PBS (pH7.4) for 4 h. Between the steps, the sections were carefullywashed with 0.01 M PBS (pH 7.4).

Our previous experimentation certified the specificity ofthe EM2 antibody used in this study (Chen, et al., 2008; Hui, etal., 2006, 2010). In the present study, some sections in the fifthdish of all groups were used as an absorption control ornegative control. Absorption control tests were conductedaccording to our previous report (Hui, et al., 2006). In thenegative control tests, the primary antibodies were replacedwith normal rabbit serum. Other procedures were the same asthose for normal immunohistochemistry or immunofluores-cent histochemical staining. Immunoreactivities werecompletely eliminated by preabsorption or replacement ofprimary antibodies with normal rabbit serum.

4.3. Digital photomicroscopy and image analysis

The sections were mounted onto gelatin-coated glass slides,air-dried, cover-slipped and observed under a light micro-scope (VANOX; Olympus, Tokyo, Japan). Five nonadjacentsections from theMDHof each rat were randomly selected. 16-bit grayscale images with a resolution of 300 pixels/in. werecaptured and digitized by a high-resolution CCD video camerasystem (Nikon DXM1200). Leica Q500MC Image Processing andAnalysis software (Leica, Wetzlar, Germany) was used toanalyze the grayscale images. Image data were collectedsimilarly as previously performed (Hui, et al., 2010). Astandardized field area of superficial laminae of MDH wassampled and the optical densities were automatically trans-ferred to Excel software for the subsequent statistical analysis.The configuration remained consistent during image dataacquisition in all experimental groups. Densities of thenegative control experiment (primary antibody replacement)were set as 100.

The immunofluorescence stained sections were mountedonto cleanglass slides, air-dried, cover-slippedwithamixtureof50% (v/v) glycerin and 2.5% (w/v) triethylene diamine in 0.01 MPBS, and observed with an epifluoresence microscope (BX-60;Olympus, Tokyo, Japan) under appropriate filters for gold-emitting FG (excitation 360 nm; emission 450 nm), green-emitting Alexa488/FITC (excitation 488 nm; emission 530 nm),and for red-emitting Texas Red (excitation 550 nm; emission615 nm) labeled neurons. The number of FG labeled neurons,EM2-IR neurons and EM2/FG double-labeled neurons in thehypothalamus, NTS and TG was counted in five randomlyselected sections fromeach rat (n=6 rats; total 30 sections). Onlycellswith a clear nucleuswere counted. The longest diameter ofthe cell body through the nucleus was regarded as the majordiameter of the cell. The neurons counted in the present studywere divided into small (≤15 μm), intermediate (16–34 μm) andlarge (≥35 μm)cells according to theirmajor somadiameter. Thenumbers of the counted cells were corrected according to themethod of Abercrombie(1946)(Guillery, 2002): 20 small, inter-mediate or large cells were randomly selected and the majordiameter of the nuclei was measured. The mean nucleardiameter of the small, intermediate or large cells was respec-tively 5.23±0.62 μm, 11.32±1.16 μm, or 18.71±1.44 μm (mean±SD). The numbers of counted cells were corrected by usingAbercrombie's equation: number of cell=number of cellscounted×T/(T+h), where T = thickness of the sections and h =the mean nuclear diameter of the small, intermediate or largecells.

All data was collected by researchers completely unawareof the surgery and reagents used. Statistical analysis wasperformed using SPSS software (version 16.0). Measurementfor every group was expressed as mean±SD. Statisticalsignificance was determined as P<0.05 using one-way anal-ysis of variance (ANOVA), or using multi-factor ANOVA,followed by the least significant difference test.

Acknowledgments

This work was supported by grants from the National NaturalScience Foundation of China (nos. 30771133, 30971123), theNational Program of Basic Research of China (G2006CB500808)and the Program for Innovation Research Team Program ofMinistry of Education of China (IRT0560). The authorwould alsolike to thank Professor Tom Kellie of the Graduate University ofthe Chinese Academy of Sciences and Peking University foreditorial assistance.

R E F E R E N C E S

Abercrombie, M., 1946. Estimation of nuclear population frommicrotome sections. Anat. Rec. 94, 239–247.

Aicher, S.A., Mitchell, J.L., Swanson, K.C., Zadina, J.E., 2003.Endomorphin-2 axon terminals contact mu-opioidreceptor-containing dendrites in trigeminal dorsal horn. BrainRes. 977, 190–198.

Almeida, R.S., Ferrari, M.F., Fior-Chadi, D.R., 2000. Quantitativeautoradiography of adrenergic, neuropeptide Y andangiotensin II receptors in the nucleus tractus solitarii andhypothalamus of rats with experimental hypertension. Gen.Pharmacol. 34, 343–348.

46 B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

Basbaum, A.I., Fields, H.L., 1984. Endogenous pain control systems:brainstem spinal pathways and endorphin circuitry. Annu.Rev. Neurosci. 7, 309–338.

Bereiter, D.A., Hirata, H., Hu, J.W., 2000. Trigeminal subnucleuscaudalis: beyond homologies with the spinal dorsal horn. Pain88, 221–224.

Budai, D., Fields, H.L., 1998. Endogenous opioid peptides acting atmu-opioid receptors in the dorsal horn contribute to midbrainmodulation of spinal nociceptive neurons. J. Neurophysiol. 79,677–687.

Chen, T., Hui, R., Dong, Y.X., Li, Y.Q., Mizuno, N., 2004.Endomorphin 1- and endomorphin 2-like immunoreactiveneurons in the hypothalamus send axons to the parabrachialnucleus in the rat. Neurosci. Lett. 357, 139–142.

Chen, T., Hui, R., Wang, X.L., Zhang, T., Dong, Y.X., Li, Y.Q., 2008.Origins of endomorphin-immunoreactive fibers and terminalsin different columns of the periaqueductal gray in the rat.J. Comp. Neurol. 509, 72–87.

Ding, Y.Q., Kaneko, T., Nomura, S., Mizuno, N., 1996.Immunohistochemical localization of mu-opioid receptors inthe central nervous system of the rat. J. Comp. Neurol. 367,375–402.

Dubner, R., Bennett, G.J., 1983. Spinal and trigeminal mechanismsof nociception. Annu. Rev. Neurosci. 6, 381–418.

Fichna, J., Janecka, A., Costentin, J., Do, R.J.C., 2007a. Theendomorphin system and its evolving neurophysiological role.Pharmacol. Rev. 59, 88–123.

Fichna, J., Janecka, A., Piestrzeniewicz, M., Costentin, J., do, R.J.C.,2007b. Antidepressant-like effect of endomorphin-1 andendomorphin-2 in mice. Neuropsychopharmacology 32,813–821.

Fields, H.L., Basbaum, A.I., 1978. Brainstem control of spinalpain-transmission neurons. Annu. Rev. Physiol. 40, 217–248.

Fields, H.L., Basbaum, A.I., 1994. Central nervous systemmechanisms of pain modulation. In: Wall, P.D., Melzack, R.(Eds.), Textbook of Pain. Churchill Livingstone, New York,pp. 243–257.

Gamboa-Esteves, F.O., Tavares, I., Almeida, A., Batten, T.F.,McWilliam, P.N., Lima, D., 2001. Projection sites of superficialand deep spinal dorsal horn cells in the nucleus tractus solitariiof the rat. Brain Res. 921, 195–205.

Guilbaud, G., Bernard, J.F., Besson, J.M., 1994. Brain areas involvedin nociception and pain. In: Wall, P.D., Melzack, R. (Eds.),Textbook of Pain. Churchill Livingstone, NewYork, pp. 113–128.

Guillery, R.W., 2002. On counting and counting errors. J. Comp.Neurol. 447, 1–7.

Hardy, S.G., 2001. Hypothalamic projections to cardiovascularcenters of the medulla. Brain Res. 894, 233–240.

Holzer, P., 1991. Capsaicin: cellular targets, mechanisms of action,and selectivity for thin sensory neurons. Pharmacol. Rev. 43,143–201.

Hughes, S.E., Sloan, H.E., Jones, L.B., Oakley, B., 1983. Colchicinereduces myelin thickness and axoplasm volume. Neurosci.Lett. 37, 181–186.

Hui, R., Chen, T., Li, Y.Q., 2006. The reciprocal connections ofendomorphin 1- and endomorphin 2-containing neuronsbetween the hypothalamus and nucleus tractus solitarii in therat. Neuroscience 138, 171–181.

Hui, R., Wang, W., Chen, T., Lu, B.C., Li, H., Zhang, T., Wu, S.X., Li,Y.Q., 2010. Origins of endomorphin-2 immunopositive fibersand terminals in the spinal dorsal horn of the rat. Neuroscience169, 422–430.

Kausz, M., 1990. Distribution of hypothalamic neurons projectingto the thoracic and sacral spinal segments in the cat. J.Hirnforsch. 31, 697–703.

Li, Y.Q., Wang, Z.M., Zheng, H.X., Shi, J.W., 1996. Central origins ofsubstance P-like immunoreactive fibers and terminals in thespinal trigeminal caudal subnucleus in the rat. Brain Res. 719,219–224.

Lü, B.C., Ji, G.L., Huo, F.Q., Chen, T., Li, H., Li, Y.Q., 2010.Topographical distributions of endomorphinergic pathwaysfrom nucleus tractus solitarii to periaqueductal gray in the rat.J. Chem. Neuroanat. 39, 166–174.

Lü, B.C., Li, H., Chen, T., Huo, F.Q., Zhang, T., Li, Y.Q., 2009.Endomorphin 1- and endomorphin 2-containing neurons innucleus tractus solitarii send axons to the parabrachial nucleiin the rat. Anat. Rec. (Hoboken) 292, 488–497.

Martin-Schild, S., Gerall, A.A., Kastin, A.J., Zadina, J.E., 1998.Endomorphin-2 is an endogenous opioid in primary sensoryafferent fibers. Peptides 19, 1783–1789.

Martin-Schild, S., Gerall, A.A., Kastin, A.J., Zadina, J.E., 1999.Differential distribution of endomorphin 1- and endomorphin2-like immunoreactivities in the CNS of the rodent. J. Comp.Neurol. 405, 450–471.

Martin-Schild, S., Zadina, J.E., Gerall, A.A., Vigh, S., Kastin,A.J., 1997. Localization of endomorphin-2-likeimmunoreactivity in the rat medulla and spinal cord. Peptides18, 1641–1649.

Monnikes, H., Ruter, J., Konig, M., Grote, C., Kobelt, P., Klapp, B.F.,Arnold, R., Wiedenmann, B., Tebbe, J.J., 2003. Differentialinduction of c-fos expression in brain nuclei by noxious andnon-noxious colonic distension: role of afferent C-fibers and5-HT3 receptors. Brain Res. 966, 253–264.

Nomura, H., Ogawa, A., Tashiro, A., Morimoto, T., Hu, J.W., Iwata,K., 2002. Induction of Fos protein-like immunoreactivity in thetrigeminal spinal nucleus caudalis and upper cervical cordfollowing noxious and non-noxious mechanical stimulation ofthe whisker pad of the rat with an inferior alveolar nervetransection. Pain 95, 225–238.

Okada-Ogawa, A., Suzuki, I., Sessle, B.J., Chiang, C.Y., Salter, M.W.,Dostrovsky, J.O., Tsuboi, Y., Kondo, M., Kitagawa, J., Kobayashi,A., Noma, N., Imamura, Y., Iwata, K., 2009. Astroglia inmedullary dorsal horn (trigeminal spinal subnucleus caudalis)are involved in trigeminal neuropathic pain mechanisms.J. Neurosci. 29, 11161–11171.

Paxinos, G., Watson, C., 2005. The Rat Brain in StereotaxicCoordinates, five ed. Academic Press, New York.

Pierce, T.L., Grahek, M.D., Wessendorf, M.W., 1998. Immunoreactivityfor endomorphin-2 occurs in primary afferents in rats andmonkey. Neuroreport 9, 385–389.

Pierce, T.L., Wessendorf, M.W., 2000. Immunocytochemicalmapping of endomorphin-2-immunoreactivity in rat brainJ. Chem. Neuroanat. 18, 181–207.

Przewlocki, R., Labuz, D., Mika, J., Przewlocka, B., Tomboly, C.,Toth, G., 1999. Pain inhibition by endomorphins. Ann. N. Y.Acad. Sci. 897, 154–164.

Przewlocki, R., Przewlocka, B., 2001. Opioids in chronic pain. Eur.J. Pharmacol. 429, 79–91.

Sanderson, N.K., Skinner, K., Julius, D., Basbaum, A.I., 2004.Co-localization of endomorphin-2 and substance P in primaryafferent nociceptors and effects of injury: a light and electronmicroscopic study in the rat. Eur. J. Neurosci. 19, 1789–1799.

Silva-Carvalho, L., Dawid-Milner, M.S., Goldsmith, G.E., Spyer, K.M.,1995. Hypothalamic modulation of the arterialchemoreceptor reflex in the anaesthetized cat: role of thenucleus tractus solitarii. J. Physiol. 487 (Pt 3), 751–760.

Swanson, L.W., Kuypers, H.G., 1980. A direct projection from theventromedial nucleus and retrochiasmatic area of thehypothalamus to the medulla and spinal cord of the rat.Neurosci. Lett. 17, 307–312.

Ter, H.G.J., de Boer, P., Luiten, P.G., van, W.J.D., 1989. Ascendingprojections from the solitary tract nucleus to the hypothalamus.A Phaseolus vulgaris lectin tracing study in the rat. Neuroscience31, 785–797.

Tseng, L.F., Narita, M., Suganuma, C., Mizoguchi, H., Ohsawa, M.,Nagase, H., Kampine, J.P., 2000. Differential antinociceptiveeffects of endomorphin-1 and endomorphin-2 in the mouse.J. Pharmacol. Exp. Ther. 292, 576–583.

47B R A I N R E S E A R C H 1 4 1 0 ( 2 0 1 1 ) 3 8 – 4 7

Wang, Q.P., Zadina, J.E., Guan, J.L., Shioda, S., 2003.Morphological evidence of endomorphin as an agonist for the mu-opioid receptor in the rat spinal cord. Neurosci. Lett.341, 107–110.

Wang, X.M., Zhang, K.M., Long, L.O., Flores, C.A., Mokha, S.S.,2000. Endomorphin-1 and endomorphin-2 modulateresponses of trigeminal neurons evoked by N-methyl-D-asparticacid and somatosensory stimuli. J. Neurophysiol. 83,3570–3574.

Zadina, J.E., Hackler, L., Ge, L.J., Kastin, A.J., 1997. A potent andselective endogenous agonist for the mu-opiate receptor.Nature 386, 499–502.

Zadina, J.E., Martin-Schild, S., Gerall, A.A., Kastin, A.J., Hackler, L.,Ge, L.J., Zhang, X., 1999. Endomorphins: novel endogenousmu-opiate receptor agonists in regions of high mu-opiatereceptor density. Ann. N. Y. Acad. Sci. 897, 136–144.

Zimmermann, M., 1983. Ethical guidelines for investigations ofexperimental pain in conscious animals. Pain 16, 109–110.