expression pattern of enkephalinergic neurons in the

Fax +41 61 306 12 34E-Mail [email protected]

Original Paper

Cells Tissues Organs 2011;193:404–416 DOI: 10.1159/000321403

Expression Pattern of Enkephalinergic Neurons in the Developing Spinal Cord Revealed by Preproenkephalin-Green Fluorescent Protein Transgenic Mouse and Its Colocalization with GABA Immunoreactivity

Jing Huang a Jing Chen a Wen Wang a Ya Yun Wang a Wei Wang a Yan Yan Wei a Ri Yi Shi b Takeshi Kaneko c Yun Qing Li a Sheng Xi Wu a

a Department of Anatomy, Histology and Embryology, K.K. Leung Brain Research Center, Fourth Military Medical University, Xi’an , PR China; b Department of Basic Medical Sciences, Center for Paralysis Research, School of Veterinary Medicine, Purdue University, West Lafayette, Ind. , USA; c Department of Morphological Brain Science, Graduate School of Medicine, Kyoto University, Kyoto , Japan

ber at postnatal day (P) 7 at the cervical level. The number of GFP-positive neurons reached a peak at P3 at the lumbar lev-el. At P21, the distribution pattern of GFP immunoreactivity was similar to that in the adult spinal cord. Double-labeling results showed that about one-third of the total � -aminobu-tyric acid cell population colocalized with GFP: 34.9 8 3.5% at E16 and 32.4 8 3.7% at P3. Double-labeled neurons ac-counted for nearly half of the GFP-positive neurons: 42.4 8 2.4% at E16 and 44.1 8 2.9% at P3. Taken together, the pres-ent results suggest that ENKergic neurons develop accord-ing to a rostrocaudal and ventrodorsal gradient. These re-sults have broad implications for understanding the func-tional roles of ENKergic neurotransmission in the developing spinal cord.

Copyright © 2010 S. Karger AG, Basel

Key Words Enkephalin � Development � Immunofluorescence labeling � Fluorescent in situ hybridization histochemistry

Abstract To gain better insight into the ontogenic function of enkeph-alin (ENK) in the spinal cord, it is necessary to have a clear picture of the developing pattern of the ENKergic neurons. To address this question, we used transgenic mice which re-veal ENKergic neurons easily by expressing enhanced green fluorescent protein (GFP) under the specific transcriptional control of the preproenkephalin (PPE) gene. GFP-positive neurons first appeared at embryonic day (E) 11.5 in the ven-tromedial part of the cervical ventral gray matter. At E13, they were mainly present in the intermediate zone. Thereafter, GFP-positive neurons increased progressively in number and extended from ventral to dorsal regions. Quantitative analysis showed that GFP-positive neurons peaked in num-

Accepted after revision: September 20, 2010 Published online: December 2, 2010

Dr. Sheng-Xi Wu and Dr. Yun-Qing Li Department of Anatomy, Histology and Embryology, K.K. Leung Brain Research Centre, Fourth Military Medical University , No. 169 West Changle Road Xi’an 710032 (PR China) , Tel. +86 29 8477 3954 or 4501, Fax +86 29 8328 3229E-Mail devneuro @ fmmu.edu.cn or deptanat @ fmmu.edu.cn

© 2010 S. Karger AG, Basel1422–6405/11/1936–0404$38.00/0

Accessible online at:www.karger.com/cto

J.H. and J.C. contributed equally to the study.

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM

http://dx.doi.org/10.1159%2F000321403

Expression Pattern of Spinal ENKergic Neurons

Cells Tissues Organs 2011;193:404–416 405

Introduction

Opioid peptides, which are derived from distinct pre-cursors – preproenkephalin (PPE), preprodynorphin and preproopiomelanocortin – have a wide distribution in the vertebrate central nervous system [Finley et al., 1981]. En-kephalin (ENK), as one of the important opioid peptides, is involved in the inhibitory modulation of the nocicep-tive information transmitted from the peripheral regions of the body to the superficial laminae of the spinal dorsal horn [Basbaum and Fields, 1984]. However, ENK has been proposed to play different roles during individual development stages. In early development, it influences central nervous system developmental events, such as cell proliferation, migration and differentiation [Hammer and Hauser, 1992]. While, during later development and adulthood, it acts on opioid receptors as classical neu-rotransmitters [Basbaum and Fields, 1984]. To better un-derstand the role of ENK, it is necessary to know the evo-lution of the population of ENKergic neurons during the course of development. A wealth of data pointed out that in adulthood, spinal ENKergic neurons exist mainly in the superficial laminae [Bennett et al., 1982; Huang et al., 2008a] and play an important role in the transmission and modulation of nociceptive information [Basbaum and Fields, 1984]. However, developmental studies ofENKergic neurons in the mouse spinal cord are scant.

Immunohistochemistry and in situ hybridization studies have reported the presence of ENK peptides and PPE messenger RNA (mRNA) in the avian telencephalon [Casini et al., 1995], with a general pattern of distribution that was consistent with their distribution in the telen-cephalon of other vertebrate species [de Lanerolle et al., 1981; Casini et al., 1995]. The ontogeny of the embryonic ENKergic system has only been partially described in the spinal cord of the rat and chick by using indirect immu-

nohistochemistry [Senba et al., 1982; Maderdrut et al., 1986; Du and Dubois, 1988]. Although the distribution of PPE mRNA has been reported in the mouse developing spinal cord [Xu et al., 2008; Huang et al., 2008b], the ap-pearance of the ENKergic system has not been system-atically described. Moreover, even though ENK and � -aminobutyric acid (GABA) have been described as colo-calized in neuronal cells of the adult spinal cord [Todd et al., 1992; Huang et al., 2008a], such information is miss-ing, especially during the course of development.

The identification of ENKergic cell bodies following immunohistochemical labeling is difficult [Du and Du-bois, 1988], which is a big obstacle for the morphological study on spinal ENKergic neurons. To aid in the identi-fication of ENKergic neurons, we used the PPE-green fluorescent protein (GFP) transgenic mouse, in which allENKergic neurons were revealed by the expression of GFP [Koshimizu et al., 2008; Huang et al., 2010].

To address the above-mentioned issues, in the present study, we examined the temporal and spatial develop-mental changes of ENKergic neurons at the cervical and lumbar levels of the spinal cord by using the PPE-GFP transgenic mouse. Immunofluorescent double staining was used to detect the colocalization of ENKergic neu-rons with GABA during spinal cord development.

Materials and Methods

Animals The generation and characterization of the PPE-GFP trans-

genic mice have been described previously [Koshimizu et al., 2008; Huang et al., 2010]. Six adult mice were used for in situ hy-bridization and double immunofluorescence staining experi-ments, and 64 mice (embryos or offspring) at different develop-mental stages (n = 4 for each stage) were used in the present study. Timed pregnancies were established by checking for vaginal smears on the morning following mating. We designated noon on the day sperm was detected as embryonic day (E) 0.5. Postnatal day (P) 0 corresponded to the date of birth. Experimental mice were housed and treated in strict accordance with the Rules for Animal Care and Use for Research and Education of the Fourth Military Medical University. All efforts were made to minimize the number of animals used, as well as to minimize distress to the animals.

Combined Labeling for Immunofluorescence and Fluorescent mRNA Signals The procedure for nonradioactive in situ hybridization has

been described elsewhere [Nakamura et al., 2007]. Complemen-tary DNA fragments corresponding to a portion of mouse PPE complementary DNA (nucleotides 560–846; GenBank accession No. NM_001002927) were cloned into a vector pCRII-TOPO (In-vitrogen). By using linearized plasmids as templates, sense and

Abbreviations used in this paper

ENK enkephalinFISH fluorescent in situ hybridization

histochemistryGABA �-aminobutyric acidGFP green fluorescent proteinNLS N-lauroylsarcosinePBS phosphate-buffered salinePI propidium iodidePPE preproenkephalin

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM

Huang /Chen /Wang /Wang /Wang /Wei /Shi /Kaneko /Li /Wu


antisense single-strand riboprobes with digoxigenin RNA label-ing mix (Roche Diagnostics, Basel, Switzerland) and in vitro tran-scription kit (MEGAscrip T7 and SP6 kits; Ambion, Austin, Tex., USA). Six adult PPE-GFP transgenic mice were perfused trans-cardially with 0.1 M phosphate buffer (pH 7.4) containing 4% paraformaldehyde. Lumbar segments 3–5 of the spinal cord were obtained and postfixed with the same fixative for 24 h, placed in diethylpyrocarbonate-treated 30% (weight/volume, w/v) sucrose solution in 0.05 M phosphate-buffered saline (PBS; pH 7.4) over-night at 4 ° C and cut into 25- � m-thick sections on a freezing mi-crotome. Free-floating sections were acetylated at room tempera-ture (RT) for 10 min by gentle shaking in freshly prepared 0.25% (volume/volume, v/v) acetic anhydride in 0.1 M triethanolamine. After a rinse, the sections were preincubated for 1 h at 60 ° C with a hybridization buffer, which was composed of 5 ! saline-sodi-um citrate (SSC), 2% (w/v) blocking reagent (Roche Diagnostics), 50% (v/v) formamide, 0.1% (w/v) N-lauroylsarcosine (NLS) and 0.1% (w/v) sodium lauryl sulfate. The sections were then hybrid-ized for 18 h at 60 ° C with 1 � g/ml sense or antisense digoxigenin-labeled PPE riboprobe in the hybridization buffer. After two washes at 60 ° C in 50% (v/v) formamide, 2 ! SSC and 0.1% (w/v) NLS for 20 min, hybridized sections were treated for 30 min at 37 ° C with 10 units/ � l RNase H (Toyobo, Osaka, Japan) in 10 m M Tris-HCl (pH 8.0), 1 m M ethylenediaminetetraacetic acid and0.5 M NaCl. The sections were washed at RT twice for 10 min in2 ! SSC and 0.1% (w/v) NLS, and then twice for 10 min in 0.2 ! SSC and 0.1% (w/v) NLS. The following incubations were per-formed at RT with 1% (w/v) blocking reagent (Roche Diagnostics) in 0.1 M Tris-HCl (pH 7.5) and 0.9% (w/v) NaCl. After a wash, the hybridized sections were incubated overnight with a mixture of 1: 3,000-diluted peroxidase-conjugated anti-digoxigenin anti-body Fab fragment (Roche Diagnostics), 1 � g/ml affinity-puri-fied guinea pig antibody to heat-denatured GFP [Nakamura et al., 2008]. The sections were then incubated with TSA (tyramide sig-nal amplification)-biotin (1: 100; NEN Life Science Products, Mass., USA) in 1 ! amplification diluent for 10 min at RT. After washing, sections were incubated with a mixture of 4 � g/ml Alexa Fluor 488-conjugated anti-guinea pig immunoglobulin G (IgG) goat antibody (Invitrogen) and 2 � g/ml Alexa Fluor 594-conju-gated streptavidin (Invitrogen) for 2 h at RT. Finally, the sections were mounted onto aminopropyl-trimethoxysilane (APS)-coated glass slides and cover slipped. Sections hybridized with the sense probe showed no mRNA signals.

Observation of GFP Fluorescence Animals studied ranged in age from E10.5 to adult. Embryos

at E10.5–E18 were taken from the pregnant mice under diethyl-ether anesthesia. E10.5–E14 embryos were fixed by immersion in 4% paraformaldehyde in 0.1 M phosphate buffer, whereas E15 to adult animals were fixed by vascular perfusion under pentobar-bital anesthesia. After cryoprotection with 30% sucrose in phos-phate buffer and freezing, the cervical and lumbar spinal cords were cut into 16- � m-thick sections with a cryostat (CM1900; Lei-ca, Germany) and mounted on gelatin-coated glass slides. GFP was identified by its native fluorescence or after immunostaining. Immunofluorescence staining was performed by incubating the sections with 0.5 � g/ml rabbit anti-GFP antiserum [Tamamakiet al., 2000] overnight, followed by incubation at RT for 3 h with 4 � g/ml Alexa Fluor 488-conjugated goat anti-rabbit IgG (Mo-lecular Probes). After being rinsed with 0.01 M PBS (pH 7.4), all

sections were treated with 100 � g/ml deoxyribonuclease-free ri-bonuclease A (Sigma) for 30 min at 37 ° C, followed by incubation with 1 � g/ml propidium iodide (Molecular Probes, Carlsbad, Ca-lif., USA) for 5 min at RT. Sections were air dried and cover slipped with a mixture of 50% (v/v) glycerol and 2.5% (w/v) triethylene-diamine (anti-fading agent) in 0.05 M PBS, and were viewed under a confocal laser-scanning microscope (FV-1000, Olympus).

Double Immunofluorescence Staining Sections were processed similarly to the description in ‘Obser-

vation of GFP Fluorescence’. Sections were incubated simultaneously with guinea pig anti-

GFP antiserum [Tamamaki et al., 2000] and rabbit anti-GABA(1: 1,000; Sigma). Secondary antibodies included Alexa Fluor 488-conjugated donkey anti-guinea pig IgG (1: 500; Molecular Probes, Eugene, Oreg., USA) and Alexa Fluor 594-conjugated donkey anti-rabbit IgG (1: 500; Molecular Probes). Sections were air dried and cover slipped with a mixture of 50% (v/v) glycerol and 2.5% (w/v) triethylenediamine (anti-fading agent) in 0.05 M PBS and were viewed under a confocal laser-scanning microscope (FV-1000, Olympus). Control experiments were performed in which the primary or secondary antibody was omitted. No label-ing was observed under these conditions.

Image Analysis and Quantification For fluorescent in situ hybridization histochemistry (FISH)

and double immunofluorescence-labeling results, the total num-ber of positive neurons in laminae I–III of the spinal dorsal horn was counted from 6 sections randomly selected from 3 animals. For relative quantitative analysis of developmental GFP expres-sion, the number of GFP-positive somata in the spinal cord was counted on each section. Six sections from 3 animals at each stage were randomly selected for cell counting. Profile counts were cor-rected by using Abercrombie’s formula. Abercrombie’s formula for the ratio of the ‘real’ number to the observed number is T/T + h, where T = section thickness and h = mean diameter of the GFP-positive somata along the axis perpendicular to the plane of the section [Abercrombie, 1946; Guillery, 2002]. The values were ex-pressed as means 8 SEM.

Results

Specificity of PPE-GFP Transgenic Mice To detect the specificity of GFP fluorescence, we per-

formed the double-labeling experiment by a combination of immunofluorescence staining for GFP protein and FISH for PPE mRNA. GFP immunoreactive productsin the adult spinal dorsal horn were seen in green in cell somata and some axon terminals or dendritic endings ( fig. 1 A). PPE mRNA appeared as a red hybridized reac-tion product in cell cytoplasm ( fig. 1 B). Immunofluores-cence and FISH showed good consistency of the expres-sions of PPE mRNA and GFP immunoreactivities ( fig. 1 C). Quantitative analysis indicated that more than 97.5 8 1.7% of GFP immunoreactive neurons showed

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



signals for PPE mRNA in the adult spinal cord and about 98.1 8 1.1% of PPE-expressing neurons were immunore-active for GFP. Sense RNA displayed no signals in the spinal cord (data not shown). Thus, the above double-la-beling results demonstrated that PPE-GFP transgenic mice showed specific and sufficient expression of GFP at least in the spinal cord.

GFP Distribution in the Developing MouseSpinal Cord Next, we took advantage of the PPE-GFP transgenic

mice and analyzed the distribution of ENKergic neurons from E10.5 to adult spinal cord at cervical and lumbar levels.

By E10.5, no GFP-positive cells were detected regard-less of the spinal level considered ( fig. 2 A). GFP-positive cells started to appear at around E11.5 at cervical level ( fig. 2 B). These stained somata were located in the ven-tromedial part of the ventral gray matter and lie pref-erentially in the vicinity of the ventricular zone ( fig.2 B). Most GFP-positive neurons were spindle-shaped with fairly long processes ( fig. 2 C, D). At E12.5, GFP-positive neurons were detected at all considered levels of the spinal cord, but were more numerous at the cervical than at the lumbar level ( fig. 2 E, I; fig. 7 ). Numerous ENKergic neurons were found in ventromedial areas of the gray matter close to the ventricular zone ( fig. 2 E). At this stage, the characteristics of GFP-positive cells were similar to those at E11.5, and these aspects were con-firmed by confocal laser scanning microscopy ( fig. 2 F, G). However, as by E12.5 stage, the whole dorsal part of the spinal cord was devoid of ENK immunoreactivity at both the cervical ( fig. 2 E) and the lumbar levels ( fig. 2 I).

Sections were counterstained with propidium iodide ( fig. 2 F, H).

At E13, GFP-positive neurons were mainly present in the intermediate zone at the cervical and lumbar levels ( fig. 3 A, B, F). At this period of development, the inter-mediate zone contained spindle-shaped or triangular GFP-positive cells with extended processes in various di-rections ( fig. 3 B–E, G). However, by E13 stage, the whole dorsal part of the spinal cord was still devoid of GFP im-munoreactivity at both the cervical and lumbar levels ( fig. 3 A, F). The first remarkable evolution within the dis-tribution of ENKergic population occurred at around E13.5 stage. A few GFP-positive cells were first observed at the cervical level of the dorsal gray matter ( fig. 3 H, I, M). In the intermediate region at E13.5, GFP-positive cells increased and appeared to form streams extending long processes ( fig. 3 H–M, fig. 7 ). Several GFP-positive cells were spindle-shaped with processes and oriented horizontally between the ventricular and pial surface ( fig. 3 J–L).

As development proceeds, at E14, GFP-positive cells were detected at all considered levels of the dorsal spinal cord but were more numerous at the cervical than at the lumbar level ( fig. 4 A, E, F). However, ENKergic neurons were primarily located in the intermediate zone ( fig. 4 A, F) and some of them were spindle-shaped and had ver-tical or horizontal processes ( fig. 4 B–D). At the lum-bar level, the intermediate lamina exhibited numerous ENKergic cell bodies while fewer ENKergic cells in the dorsal horn were detected ( fig. 4 F). Thereafter, ENKer-gic neurons increased in extent and intensity. At E15, rostral and caudal spinal levels exhibited numerous GFP somata, as illustrated in figure 4 G, H, whereas cells

A B C

Fig. 1. Colocalization of GFP ( A , green) and PPE mRNA ( B , red) in the spinal dorsal horn of the adult PPE-GFP transgenic mouse. GFP expression was revealed by fluorescence immunohistochemistry and PPE mRNA expres-sion was revealed by FISH. C Merged picture of A and B . Arrows indicate GFP and PPE mRNA coexpressing neu-rons. The dashed line delimitates lamina I and II of the spinal dorsal horn. Scale bar = 20 � m in C (applies to A–C ).

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



A B

C

D

E

F

G

H

I

Fig. 2. E10.5, E11.5 and E12.5 stages of development. A At E10.5, no GFP-positive cell was detected at the cervical level. B First evi-dence of GFP staining at E11.5. At the cervical level, some ENKer-gic neurons were clearly detected in the ventromedial part ofthe gray matter at E11.5. C Higher magnification of the spindle-shaped GFP-positive cell (arrow) in B . D Higher magnification of the spindle-shaped GFP-positive cell in the left ventral part of the spinal cord from a section at E11.5. E Distribution of GFP-positive neurons at the cervical level at E12.5. F, G Enlarged view of the area (asterisk) in E . Note the colocalization of GFP-positive neurons with propidium iodide (red). I Distribution of GFP-positive neu-rons at the lumbar level at E12.5. H Confocal images of the lumbar spinal cord at E12.5 stained with propidium iodide (PI). Arrows indicate some GFP-positive neurons with long processes. dh = Dorsal horn; vh = ventral horn. Scale bars = 100 � m in A , B , E , H , I ; 10 � m in C , D , F , G .

Fig. 3. Distribution of GFP-positive ENKergic neurons at E13 and E13.5. A, F Overview of GFP-positive neurons and propidium io-dide at E13 at cervical ( A ) and lumbar ( F ) levels. B–E, G Note the spindle-shaped or triangular GFP-positive cells which extended processes in various directions. H Overview of GFP-positive neu-rons and propidium iodide at E13.5 at the cervical level. I–M High-er magnification of some spindle-shaped GFP-positive cells with long processes, oriented horizontally between the ventricular and pial surface. In this figure, as in the following figures, boxed areas indicate enlargements, with the relative lowercase letters in the boxed areas. L Enlarged view of the area (asterisk) in K . Arrows indicate some GFP-positive neurons with long processes. Arrow-heads indicate the processes of the GFP-positive neurons. Scale bars = 100 � m in F , H ; 50 � m in A , B , I; and 10 � m in C–E , G , J–M .

2

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



A B

C

D

E F G

H I J

K L M 3

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



A

B

C D E

F G H

I J K L

Fig. 4. Expression of GFP at E14, E15 and E16. A, F Distribution of GFP-positive cells in the spinal cord at E14 at the cervical ( A ) and lumbar ( F ) levels. B–E Higher magnification views of the boxed areas in A . G, H At E15, the total number of GFP-positive cells in-creased in the dorsal horn and the intermediate zone both at the cervical ( G ) and lumbar ( H ) levels. I , L At E16, the GFP-positive cells were localized within the whole spinal cord in the future gray matter. J , K Higher magnification views of the boxed areas in I . Some GFP-positive neurons around the ventricle extended long processes ( K ). Arrows indicate some GFP-positive neurons with long processes. Arrowheads indicate the processes of the GFP-positive neurons. Scale bars = 100 � m in A , F , G , I , L ; 50 � m in E , H ; 10 � m in B , C , D , J , K . Fig. 5. Overview of ENKergic immunoreactivity at E17 ( A ), E18 ( B–E ), P1 ( F–H ), P3 ( I–K ) and P7 ( L, M ). A, B At E17 and E18, the cervical ENKergic immunoreactivity evolved toward dorsal parts

of the spinal cord. C Enlarged view of the box in B . D Distribution of GFP-positive cells in the section of spinal cord parallel to the longitudinal direction at E18. E Enlarged view of the box in D . F At P1, GFP labeling increased in the neuronal population of the dorsal horn. G Enlarged view of the box in F showing a few mul-tipolar cells in the intermediate zone which gave rise to dendrites radiating in all directions. H Compared to the cervical level, GFP-positive neurons were less at the lumbar level. I At the cervical level, GFP-positive neurons were primarily located in the super-ficial layers of the dorsal horn at P3. J Enlarged view of the boxin I . K At the lumbar level, more GFP-positive neurons were dis-tributed in the intermediate zone. L , M View at low magnification of GFP-positive neurons at P7. Arrows indicate some GFP-posi-tive neurons with long processes. Scale bars = 100 � m in A , B , D , F , H , I , K– M ; 50 � m in C , E , J ; 10 � m in G .

4

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



A B C

D E

F G H

I J K

L M 5

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



immunoreactive for GFP were less numerous at the lumbar level ( fig. 4 G, H, fig. 7 ). A similar pattern was observed at E14 and the total number of GFP-positive cells increased in the dorsal horn and in the intermedi-ate laminae (compare fig. 4 G, H with 3 A, F). At this stage of development, GFP-positive cells occupied almost the entire area of the section, invading the previously GFP-

negative region in the spinal dorsal horn. GFP-positive neurons in the intermediate lamina appeared to be more intensely labeled ( fig. 4 G, H). At E16, the GFP-positive cells were localized within the whole spinal cord ( fig. 4 I, L). A few spindle-shaped GFP-positive cells were also detected in the intermediate area ( fig. 4 J). At this stage, we found some GFP-positive cells surrounding the cen-

A B C

D E F

Fig. 6. Distribution pattern of ENKergic neurons at P14, P21 and adult spinal cord. A At P14, the majority of immunoreactive cell bodies were found in the superficial layers of the dorsal horn, al-though some GFP-positive neurons were occasionally seen in the deeper layers of the dorsal horn. B At P21, GFP immunoreactivity appeared as a dense band within lamina II of the dorsal horn with moderate intensities within lamina I and III. The distribution pat-

tern at this stage is similar to that of the adult ( D ). C Enlarged view of the box in B . Note the presence of some ENKergic neurons with very long processes in the deep layers of the dorsal horn at P21. E , F Higher magnification of the boxed areas in D . Arrows indicate some GFP-positive neurons with long processes. Arrowheads in-dicate the processes of the GFP-positive neurons. Scale bars = 100 � m in A , B , D ; 10 � m in C , E , F .

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



tral canal and extending processes into the cavity of the central canal ( fig. 4 K). At E17 and E18, the GFP immu-noreactivity evolved toward the dorsal part of the spinal cord. The dorsal horn exhibited numerous GFP-positive cell bodies ( fig. 5 A–C). A large amount of GFP-positive perikarya was found in the dorsal horn at E18 ( fig. 5 C, D). As illustrated in the sections of spinal cord parallel to the longitudinal direction, the ventral part of the spi-nal cord exhibited sparse GFP-positive cell bodies ( fig. 5 D, E).

The general distribution of GFP immunoreactivityat P1 was not fundamentally changed as compared with E18, although the number and intensity of immunofluo-rescent cells increased (compare fig. 5 F and B). A few multipolar cells in the intermediate zone gave rise to den-drites radiating in all directions ( fig. 5 F, G). The staining intensity of GFP-positive cells became strong and their dendrites showed relatively developed arborization ( fig. 5 G). The distribution at the lumbar level ( fig. 5 H) was similar to that observed at the cervical level ( fig. 5 F), while with a lower density.

At P3 and P7 and at both levels, we detected (1) some GFP-positive cells in the intermediate zone and around the central canal, (2) a high number of stained cells in the dorsal horn and (3) a few GFP immunoreactive cells in the ventral horn of the spinal cord ( fig. 5 I–M). In the in-termediate zone, several GFP-positive cells displayed long processes ( fig. 5 J). Quantitative analysis showed that

GFP-positive neurons peaked in number at P7 at the cer-vical level ( fig. 7 A), while the number of GFP-positive neurons reached a peak at P3 at the lumbar level ( fig. 7 B). Thereafter, there was a gradual decrease in the number of GFP-positive cells ( fig. 7 ).

Two weeks after birth, the majority of GFP immuno-reactive cell bodies and puncta were found in the super-ficial layers of the dorsal horn, although some GFP-posi-tive neurons were occasionally seen in the deeper layers of the dorsal horn ( fig. 6 A). By P21, GFP immunoreactiv-ity appeared as a dense band within Rexed’s lamina II of the dorsal horn with moderate intensities within laminae I and III ( fig. 6 B). In the deep layers of the dorsal horn, we found some ENKergic neurons with very long processes. These processes increased in length up to this date, and were around 2–5 times the size of the cell body ( fig. 6 C). The general appearance of GFP immunoreactivity at this stage showed an adult-like distribution ( fig. 6 B, D). In the superficial layer of the adult spinal dorsal horn, ENKergic neurons were rounded or oval ( fig. 6 E). Some GFP-posi-tive neurons in the deep layer extended long processes ( fig. 6 F).

The relative quantitative analysis of GFP-positive neu-rons in the mouse spinal cord during the development is shown in figure 7 . Finally, the quantitative analysis indi-cated that at both spinal levels, a rostrocaudal delay in the appearance of the ENK immunoreactivity exists. ENKer-gic neurons in the spinal cord appear to undergo develop-

0

50

100

150

200

250

300

E12 E13 E14 E15 E16 E17 E18 P1 P3 P7 P14 P21 Adult

EN

Ke

rgic

ne

uro

ns/

sec

tio

n(n

)

Developmental stagesA

0

50

100

150

200

E13 E14 E15 E16 E17 E18 P1 P3 P7 P14 P21 Adult

EN

Ke

rgic

ne

uro

ns/

sec

tio

n(n

)

Developmental stagesB

Fig. 7. Relative quantitative analysis of GFP-positive cells in the PPE-GFP transgenic mouse spinal cord during development at the cervical ( A ) and lumbar ( B ) levels. The values were expressed as mean 8 SEM. Each histo-gram corresponds to six measurements.

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



A B C

D E F

G H I

Fig. 8. Double labeling of GFP with GABA in the spinal cord of PPE-GFP transgenic mice at E16 ( A–F ) and P3 ( G–I ). In each case, GFP (green) is shown on the left, GABA (red) in the middle, and a merged image on the right. D–F Higher magnification views of the boxed area in C . Arrows indicate some double-labeled neurons. Scale bars = 50 � m in C (applies to A–C ); 20 � m in F (applies to D–I ).

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



ment in the form of upregulation, downregulation or dis-tribution changes ( fig. 7 ).

Analysis of Colocalization of GFP Expression and GABA Immunoreactivity At E16, colocalization of GFP/GABA was mainly ob-

served in the dorsal horn ( fig. 8 A–F). At P3, the localiza-tion pattern of double-stained perikarya remained quite similar to the pattern observed at E16 ( fig. 8 G–I). Final-ly, to improve our understanding of the dynamic of the GFP/GABA colocalization evolution during the develop-ment period, we analyzed the ratio of GABA immunore-activity, GFP, and GFP/GABA somata in the spinal cord. Double-stained cells represented about one-third of the total GABA cell population (34.9 8 3.5% at E16; 32.4 8 3.7% at P3) and corresponded to nearly half of the ENKergic neurons (42.4 8 2.4% at E16; 44.1 8 2.9% at P3).

Discussion

Specificity of GFP Expression in ENKergic Neurons The validity of GFP in the PPE-GFP transgenic mouse

as a marker for ENKergic neurons has been confirmed [Koshimizu et al., 2008; Huang et al., 2010]. In this study, by using FISH and immunofluorescent histochemistry double staining, we observed that PPE mRNA signals in the spinal cord showed GFP immunoreactivity with a co-localization rate of 1 95%. The above double-staining ex-periments in the spinal cord and previous study in other brain regions in the PPE-GFP transgenic mice further confirmed that GFP-positive neurons were ENKergic and thus confirmed the reliability of the PPE-GFP trans-genic mouse in our studies on the developmental pattern and characteristics of spinal ENKergic neurons.

Comparative Analysis of the Spinal ENKergicNeuron Development Although it is well established that ENK plays a crucial

role during the development of neuronal tissues [Ham-mer and Hauser, 1992], the ontogenic establishment of ENKergic neurons in the mouse spinal cord has never been fully documented. Previous studies showed that Met-ENK immunoreactivity was first detected in the chick spinal cord at embryonic stages 29–30 (incubation day 6). From stage 40 (incubation day 13 or 14), much more Met-ENK immunoreactivity was observed in sev-eral spinal areas. Such a distribution of Met-ENK immu-noreactivity remained visible at later embryonic stages,

but labeled cells gradually decreased in number and dis-appeared after hatching [Du and Dubois, 1988]. High-performance liquid chromatography showed that the lev-el of total radioimmunoassayable Met-ENK in the lum-bar spinal cord of the chicken embryo increased more than 1,000-fold between day 4.5 and day 18 [Maderdrut et al., 1986]. By using in situ hybridization, two recent studies observed the spatial and temporal expression pat-terns of some peptide genes, including Penk1 . Their data indicated that Penk1 -expressing cells emerged in the ven-tral spinal cord at E14.5 and then expanded to the dorsal horn from E16.5 to P0. From P0 and on, Penk1 -expressing cells were scattered throughout the spinal cord [Xu et al., 2008; Huang et al., 2008b]. However, our data showed that GFP-positive ENKergic neurons first appeared in the ventral region of the spinal cord at E11.5. The first labeled GFP-positive cells were observed in the cervical dorsal gray matter at E13.5, though ENKergic neurons were pri-marily located in the intermediate lamina. Thereafter, GFP-positive neurons increased progressively in number and extended from the ventral to the dorsal regions. Be-tween E11.5 and E18, a rostrocaudal and ventrodorsal evolution of the ENKergic population occurs, leading to a redistribution of ENK immunoreactivity, from ventral-ly early in development to dorsally. When compared with embryonic stages, GFP-positive neurons became primar-ily restricted to the dorsal horn during the postnatal pe-riod. We suspect that the difference in these two sets of results might be related to the different research methods, particularly the use of the PPE-GFP transgenic mouse, which most likely accounts for the discrepancy.

Parallels between ENK and GABA Immunoreactivity during Development Our previous study examined the temporal and spatial

developmental changes of GABAergic neurons at the cer-vical level of the spinal cord by using glutamic acid decar-boxylase-67-GFP knock-in mouse [Huang et al., 2007]. GFP-positive GABAergic neurons appeared at E11.5 in the ventral region of the spinal cord and became abundant in the future gray matter at E12. Although there was a vast difference in the expression pattern of ENKergic and GABAergic neurons at early developmental stages, the evolution was similar at late developmental and postnatal stages. The majority of ENKergic and GABAergic neu-rons were found in the superficial layers of the dorsal horn, and the ventral part of the spinal cord exhibited only sparse cell bodies at late developmental and postnatal stages. In view of our data indicating that ENKergic and GABAergic neurons were detected in the same dorsal

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM



References

Abercrombie, M. (1946) Estimation of nuclear population from microtome sections. Anat Rec 94: 239–247.

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

Bennett, G.J., M.A. Ruda, S. Gobel, R. Dubner (1982) Enkephalin immunoreactive stalked cells and lamina IIb islet cells in cat substan-tia gelatinosa. Brain Res 240: 162–166.

Casini, G., M. Molnar, B.M. Davis, P. Bagnoli (1995) Posthatching development of prepro-enkephalin mRNA-expressing cell popula-tions in the pigeon telencephalon. Brain Res Dev Brain Res 84: 233–244.

de Lanerolle, N.C., R.P. Elde, S.B. Sparber, M. Frick (1981) Distribution of methionine-enkephalin immunoreactivity in the chick brain: an immunohistochemical study. J Comp Neurol 199: 513–533.

Du, F., P.M. Dubois (1988) Development and dis-tribution of enkephalin-immunoreactive el-ements in the chicken spinal cord. Neurosci-ence 27: 251–266.

Finley, J.C., J.L. Maderdrut, P. Petrusz (1981) The immunocytochemical localization of en-kephalin in the central nervous system of the rat. J Comp Neurol 198: 541–565.

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

Hammer, J.R., K.F. Hauser (1992) Consequences of early exposure to opioids on cell prolifera-tion and neuronal morphogenesis; in Miller M (ed): Development of the Central Nervous System: Effects of Alcohol and Opiates. New York, Wiley Liss, pp 319–339.

Huang, J., J. Chen, W. Wang, W. Wang, Y. Koshi-mizu, Y.Y. Wei, T. Kaneko, Y.Q. Li, S.X. Wu (2010) Neurochemical properties of enkeph-alinergic neurons in lumbar spinal dorsal horn revealed by preproenkephalin-green fluorescent protein transgenic mice. J Neu-rochem 113: 1555–1564.

Huang, J., N. Tamamaki, K. Obata, Y.Q. Li, S.X. Wu (2007) Prenatal and postnatal develop-ment of GABAergic neurons in the mouse spinal cord revealed by GFP expression in GAD67-GFP knock-in mouse. Neuroem-bryol Aging 4: 147–154.

Huang, J., Y. Wang, W. Wang, Y. Wei, Y. Li, S.Wu (2008a) Preproenkephalin mRNA is ex-pressed in a subpopulation of GABAergic neurons in the spinal dorsal horn of the GAD67-GFP knock-in mouse. Anat Rec (Hoboken) 291: 1334–1341.

Huang, M., T. Huang, Y. Xiang, Z. Xie, Y. Chen, R. Yan, J. Xu, L. Cheng (2008b) Ptf1a, Lbx1 and Pax2 coordinate glycinergic and pepti-dergic transmitter phenotypes in dorsal spi-nal inhibitory neurons. Dev Biol 322: 394–405.

Koshimizu, Y., S.X. Wu, T. Unzai, H. Hioki, T. Sonomura, K.C. Nakamura, F. Fujiyama, T. Kaneko (2008) Paucity of enkephalin pro-duction in neostriatal striosomal neurons: analysis with preproenkephalin-green fluo-rescent protein transgenic mice. Eur J Neu-rosci 28: 2053–2064.

Maderdrut, J.L., I. Merchenthaler, D.K. Sundberg, N. Okado, R.W. Oppenheim (1986) Distribu-tion and development of proenkephalin-like immunoreactivity in the lumbar spinal cord of the chicken. Brain Res 377: 29–40.

Nakamura, K., A. Watakabe, H. Hioki, F. Fuji-yama, Y. Tanaka, T. Yamamori, T. Kaneko (2007) Transiently increased colocalization of vesicular glutamate transporters 1 and 2 at single axon terminals during postnatal de-velopment of mouse neocortex: a quantita-tive analysis with correlation coefficient. Eur J Neurosci 26: 3054–3067.

Nakamura, K.C., H. Kameda, Y. Koshimizu, Y. Yanagawa, T. Kaneko (2008) Production and histological application of affinity-purified antibodies to heat-denatured green fluores-cent protein. J Histochem Cytochem 56: 647–657.

Senba, E., S. Shiosaka, Y. Hara, S. Inagaki, M. Sakanaka, K. Takatsuki, Y. Kawai, M. Tohya-ma (1982) Ontogeny of the peptidergic sys-tem in the rat spinal cord: immunohisto-chemical analysis. J Comp Neurol 208: 54–66.

Tamamaki, N., K. Nakamura, T. Furuta, K. Asa-moto, T. Kaneko (2000) Neurons in Golgi-stain-like images revealed by GFP-adeno-virus infection in vivo. Neurosci Res 38: 231–236.

Todd, A.J., R.C. Spike, G. Russell, H.M. Johnston (1992) Immunohistochemical evidence that Met-enkephalin and GABA coexist in some neurones in rat dorsal horn. Brain Res 584: 149–156.

Xu, Y., C. Lopes, Y. Qian, Y. Liu, L. Cheng, M. Goulding, E.E. Turner, D. Lima, Q. Ma (2008) Tlx1 and Tlx3 coordinate specifica-tion of dorsal horn pain-modulatory pepti-dergic neurons. J Neurosci 28: 4037–4046.

area, we performed double staining to reveal possible co-localization of GABA and ENK in the developing spinal cord. Double GFP/GABA staining revealed a significant population of neurons expressing both of them. Interest-ingly, this proportion remained stable during the course of development. From a functional point of view, ENK to-gether with GABA may exert a trophic influence on sec-ondary events of axonal growth and synapse formation.

Moreover, as studies of the development or electrophys-iological characteristics of ENKergic neurons pro gress, the necessity for visualizing ENKergic neurons in living tissues becomes more urgent. PPE-GFP transgenic mice presented a major advantage in that ENKergic neurons could be easily visualized by GFP. Compared with those previously observed with ENK antibody labeling, this newly generated mouse line can be a reliable tool for the study of specific spinal ENKergic neuronal populations.

In summary, the present study examined the expres-sion pattern of ENKergic neurons and its colocalization with GABA in the developing spinal cord of the PPE-GFP transgenic mouse. The generation timetable and migra-tion of ENKergic neurons still remain to be elucidated and we will study further to clarify the molecular mech-anisms that control the fate of the ENKergic neurons. Identifying core transcription factors that coordinate a large set of transmitter phenotypes may help to generate specific types of neurons in vitro for future functional and therapeutic studies.

Acknowledgements

This study was supported by a grant from the National Natural Science Foundation of China (30770694, 30970946) and Shaanxi Provincial Natural Science Foundation (2010JQ4003).

Dow

nloa

ded

by:

Pur

due

Uni

v.Li

b T

SS

128.

210.

106.

129

- 9/

11/2

019

7:36

:23

PM

expression pattern of enkephalinergic neurons in the

Documents