Some γ-motoneurons contain γ-aminobutyric acid in the rat cervical spinal cord

BR A IN RE S EA RCH 1 2 01 ( 20 0 8 ) 7 8 –87

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / b r a i n r e s

Research Report

Some γ-motoneurons contain γ-aminobutyric acid in the rat cervical spinal cord Tetsufumi Ito a,⁎, Hiroyuki Hioki b , Kouichi Nakamura b,c , Takeshi Kaneko b,c , Satoshi Iino a , Yoshiaki Nojyo a a

Department of Anatomy, Faculty of Medical Sciences, University of Fukui, Matsuoka-Shimoaizuki, Fukui 910-1193, Japan Department of Morphological Brain Science, Graduate School of Medicine, Kyoto University, Kyoto 606-8501, Japan c CREST of Japan Science and Technology Agency, Kawaguchi 332-0012, Japan b

A R T I C LE I N FO

AB S T R A C T

Article history:

γ-aminobutyric acid (GABA) is utilized in the peripheral as well as central nervous system. In

Accepted 10 January 2008

this study, fibers immunoreactive for 67 kDa isoform of glutamic acid decarboxylase (GAD67),

Available online 2 February 2008

an enzyme which synthesizes GABA, were found to terminate in the intercapsular region of muscle spindles of the upper limb. GABA-containing fibers were also found in the ventral

Keywords:

roots of C5 to T5 spinal segments, brachial plexus, and radial nerve. These fibers were thin

Colocalization

and immunoreactive for choline-acetyl transferase (ChAT). After transection of the brachial

Immunohistochemistry

plexus, GABA immunoreactivity disappeared completely in the ipsilateral triceps brachii

In situ hybridization

muscle (TBM). After the injection of fluorogold into the TBM, some retrogradely labeled

Ventral horn

medium-sized neurons were positive for GAD67, but not VGAT mRNA. All these observations

Schwann cell

clearly indicate that GABA-containing γ-motoneurons in the lower cervical spinal cord send their fibers to muscle spindles in the upper extremities. Since we detected neither GABAA nor GABAB receptors in the TBM by RT-PCR, the function of the GABA-containing γ-motoneurons remains unclear. © 2008 Elsevier B.V. All rights reserved.

1.

Introduction

Somatic motor neurons, which include α-, β-, and γ-motoneurons, use acetylcholine as a neurotransmitter. Once released, acetylcholine excites both extrafusal and intrafusal muscles via nicotinic receptors. Thus, acetylcholine acts as an excitatory neurotransmitter. Years after “Dale's law”, which claims that one neuron uses one transmitter (Dale, 1935),

much is now known about the colocalization of neurotransmitters. Namely, many types of neurons use more than one neurotransmitter (e.g. Johnson et al., 1992; Piehl et al., 1993; Shupliakov et al., 1993; Ito et al., 2005, 2007; Ito and Nojyo, 2008). Nevertheless, only a few studies (Johnson et al., 1992; Piehl et al., 1993; Shupliakov et al., 1993) have examined the colocalization of neurotransmitters in somatic motoneurons. In contrast, there have been many studies about colocalization

⁎ Corresponding author. Fax: +81 776 61 8132. E-mail address: [email protected] (T. Ito). Abbreviations: ir, immunoreactive; ChAT, choline-acetyl transferase; VH, ventral horn; TBM, triceps brachii muscle; DRG, dorsal root ganglion; VR, ventral root; DR, dorsal root; DAPI, 4′,6-diamidino-2-phenylindole dihydrochloride; FG, fluorogold; GAD, glutamic acid decarboxylase; GAD67, 67 kDa isoform of GAD; GABA, γ-aminobutyric acid; PB, phosphate buffer; PBS, 0.05M phosphate-buffered saline; PBS-X, PBS containing 0.3%; Triton X-100, ; PBS-XD, PBS containing 0.3%; Triton X-100, 1%; normal donkey serum, ; RT, room temperature; NLS, N-lauroylsarcosine; DAB, diaminobenzidine; FITC, fluorescein isothiocyanate; VGAT, vesicular GABA transporter; GAPDH, glyceraldehyde-3-phosphate dehydrogenase 0006-8993/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2008.01.056

BR A IN RE S E A RCH 1 2 01 ( 20 0 8 ) 7 8 –8 7

79

Fig. 1 – GABA-containing fibers in the upper extremities. In a muscle spindle (MS) of the triceps brachii muscle (TBM) counterstained with Neutral Red (a), GAD67-ir fibers (white arrows in a) were found in the intercapsular region. Note that a GAD67-negative nerve bundle (arrowhead in a) entered the equatorial region (double arrow in a). At a higher magnification of the boxed region in a (b), we can clearly observe that GAD67-ir fibers were apposed upon the intrafusal muscle fibers (white arrows in b). In sections immunostained for GABA, structures surrounding intrafusal muscle fibers were labeled (c, d). When the optical sections were projected into one Figure (12 optical sections collected from 2.63 μm z-axis range were processed for maximum projection; d), we can observe that these structures were composed of cell bodies (arrows in d) and their processes. Furthermore, GABA-containing fibers (arrowheads in e–g) were found in the ventral roots (VRs; e), brachial plexus (f), and radial nerve (g). All of these fibers were immunoreactive for choline-acetyl transferase (ChAT; red). ChAT-ir fibers which showed immunoreactivity for GABA or GAD67 (arrowheads in e–g) were much thinner than those immunonegative for GABA or GAD67 (arrows in f and g). Note: Inset in a shows a schematic diagram of a, and black arrows indicates sheath of the MS. Bar: 250 μm (a), 100 μm (b), 50 μm (c, d), 20 μm (e, g), and 10 μm (f).

80

BR A IN RE S EA RCH 1 2 01 ( 20 0 8 ) 7 8 –87

in visceral motor neurons. On that point, we recently reported that a population of cholinergic sympathetic preganglionic neurons, cholinergic visceral motor neurons located in the spinal cord, express glutamic acid decarboxylase (GAD), the enzyme synthesizing γ-aminobutyric acid (GABA), and produce GABA, strongly suggesting the colocalization of both excitatory and inhibitory neurotransmitters (Ito et al., 2005, 2007; Ito and Nojyo, 2008). These GABA-containing sympathetic preganglionic neurons send their axons to the superior cervical ganglion via the thoracic ventral roots (VRs). Surprisingly, in preliminary experiments for these studies, we found that GABA-containing fibers were present not only in the thoracic but also in the cervical VRs (see Fig. 2d), implying that some somatic motor neurons contain GABA. The aim of this study is to elucidate 1) whether somatic motor neurons contain both GABA and acetylcholine, and 2) if so, the type(s) of somatic motor neurons which contain GABA.

2.

Results

2.1. Muscle spindles in the upper extremities receive innervation from GABA-containing fibers In those sections of the upper limb muscles including the triceps brachii muscle (TBM), biceps brachii muscle, flexor and extensor forearm muscles, which were immunostained for 67 kDa isoform of GAD (GAD67), GAD67-immunoreactive (-ir) fibers were found in muscle spindles (MSs) as well as nerve bundles. These fibers were mainly located in the intercapsular region of the MSs (white arrows in Fig. 1a), apposed with intrafusal muscles without coiling (arrows in Fig. 1b), and their terminals showed a very similar morphology of the endings of γ-motor axons (Banks et al., 1985), implying that they were likely to be γ-motor fibers rather than sensory fibers.

In the sections of upper limb muscles immunostained for GABA, GABA-ir structures were found around MSs, and surrounded intrafusal fibers tightly (Fig. 1c). The GABA-ir structures might be composed of cell bodies (arrows in Fig. 1d) and their processes. These cells were flat-shaped (Fig. 4b–b″), and presumably peripheral glia cells which packed nerve fibers innervating intrafusal muscles. We could not discriminate between GABA-ir fibers and GABA-ir thin cellular processes.

2.2. GABA-containing fibers were found in the nerves innervating the upper extremities The GAD67-ir fibers around MSs of the upper limb are likely to be GABA-containing spinal nerve fibers. We then immunostained peripheral nerves which project to the upper limb, including the ventral roots (VRs) of C5 to T1 spinal segments, brachial plexus, and radial nerve. In the VRs, brachial plexus, and radial nerve, both GABA- and GAD67-ir fibers were found. All of these fibers were positive for choline-acetyl transferase (ChAT; arrowheads in Fig. 1e–e′′, f–f′′, g–g′′), and thinner than neighboring ChAT-ir and GAD67-immunonegative fibers, presumably α-motor fibers. These findings imply the presence of GABA-containing γ-motor fibers.

2.3. The distribution of GABA-containing fibers was restricted in the VRs of C5 to T5 spinal segments We further investigated the distribution of GABA- and GAD67ir fibers in the spinal nerve roots of C1 to S4 spinal segments. In the dorsal root ganglia (DRGs; Fig. 2a) or dorsal roots (DRs; Fig. 2b), GAD67 immunoreactivity was not detected at all. In contrast, GAD67-ir fibers were found in the VRs of C5 to T5 segments (Fig. 2c). In the case of immunohistochemistry for GABA, similar results were obtained (not shown). On counting the GAD67-ir fibers in transverse sections of the VRs, we found that the total number was 123.0 ± 6.2 (mean ± S.D; N = 3), and

Fig. 2 – The distribution of GAD67-ir fibers in the VRs. GAD67-immunoreactive (ir) fibers were not found in the dorsal root ganglia (DRGs; a) or dorsal roots (DRs; b), but found in the VRs (c). Bar: 100 μm (a), 20 μm (b, c). GAD67-ir fibers were found in the VRs of C5 to T5 segments (d; N=3). The total number of GAD67-ir fibers was 123 ± 6.2 (mean ± S.D). Transverse sections of the C1-S4 VRs were cut, and immunostained for anti-GAD67 antibody.

BR A IN RE S E A RCH 1 2 01 ( 20 0 8 ) 7 8 –8 7

81

Fig. 3 – Neither GAD65-mRNA (a) nor GAD67-mRNA (b) was expressed in the TBM. Neither muscle fibers nor glia cells around intrafusal muscles in MSs (arrows) expressed GADs. Bar: 100 μm.

GAD67-ir fibers were more numerous in the C6 to C8 VRs (Fig. 2d), corresponding to the level of the brachial plexus. Because there are GABA-containing sympathetic preganglionic neurons located in the rostral part of the intermediolateral nucleus of the spinal cord (Ito et al., 2007), some of the GAD67-ir fibers in the VRs of C8 to T5 segments might be GABA-containing sympathetic preganglionic fibers. However, GAD67-ir fibers in the C5 to C7 VRs were unlikely to be preganglionic fibers, because the rostral end of the intermediolateral nucleus is the C8 spinal segment (Strack et al., 1988).

2.4. GABA-containing glia-like cells in muscle spindles of the forelimb do not express GAD, and disappeared after transection of brachial plexus As mentioned, some glia-like cells wrapping around intrafusal fibers showed GABA but not GAD67 immunoreactivity (Figs. 1c,d, 4 c–c′′, and d–d′′). To examine whether these GABA-containing glia-like cells really express GAD or not, we performed in situ hybridization histochemistry for GAD of 65 kDa (GAD65) and GAD67 in the TBM. Neither GAD65 nor GAD67 mRNA-expressing cells were observed in the TBM at all (Fig. 3), indicating that there was no cell which could produce GABA). Then, another question arises: How could the glia-like cells around intrafusal fibers contain GABA without synthesizing GAD? To better understand this issue, we then evaluated GABA-ir structures in the TBM after transection of the hemilateral brachial plexus, through which all fibers innervating the TBM pass. Seven days after the transection, sections of the TBM of both sides were immunostained for GABA and counterstained with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI). On the contralateral side, GABA-ir structures around intrafusal fibers, which were identified by DAPI staining (Fig. 4a–a′′), were found in MSs. GABA-ir fibers were also found (not shown). All GABA-ir cells were flat-shaped (Fig. 4b–b′′). On the ipsilateral side, however, no GABA-ir structure was found in the TBM at all (Fig. 4c–c′′).

2.5. A proportion of γ-motoneurons innervating the triceps brachii muscle express GAD67 Sections of the cervical spinal cord were processed for both fluorescent in situ hybridization for GAD67 and immunohisto-

chemistry for ChAT, to examine GAD67 mRNA expression in cholinergic neurons. Numerous ChAT-ir neurons were found in the ventral horn. It is known that the size of motoneuron shows bimodal distribution; larger α-motoneurons (mean diameter, which was the mean of the major and minor axes of the cell body, larger than 30 µm) and medium-sized γ-motoneurons, of which mean diameter is less than 30 µm (Ichiyama et al., 2006). Indeed, ChAT-ir neurons were successfully divided into two groups by means of their size. Meanwhile, GAD67 mRNAexpressing neurons were small to medium-sized. Some ChAT-ir medium-sized neurons expressed GAD67 mRNA (arrows in Fig. 4d–d′′), while no large neurons expressed GAD67 (arrowheads in Fig. 4d–d′′). We then examined GAD67 mRNA expression in retrogradely labeled motoneurons. Sections of the cervical spinal cord of animals which received an injection of fluorogold (FG), a retrograde tracer, in the TBM were processed for fluorescent in situ hybridization for both GAD67 and VGAT mRNAs, and immunohistochemistry for FG. The number of FG-ir neurons was 429.6± 120.0 (mean ± S.D.; N = 4). FG-ir neurons were divided into two groups, large and medium-sized neurons as before. The number of medium-sized neurons immunoreactive for FG, presumably γ-motoneurons, was 66.2 ± 16.3 (mean ± S.D.). A proportion of medium-sized FG-ir neurons (arrows in Fig. 4e), but no large FG-ir neurons (arrowheads in Fig. 4e), expressed GAD67 mRNA. The number of medium-sized FG-ir neurons expressing GAD67 was 24.4 ± 12.7 (mean± S.D.). Therefore, about 37% of medium-sized FG-ir neurons expressed GAD67 mRNA. Interestingly, the GAD67-expressing medium-sized neurons did not express VGAT mRNA (arrows in Fig. 4e).

2.6. Neither GABAA nor GABAB receptor mRNA was expressed in the triceps brachii muscle Finally, we examined the expression of GABAA and GABAB receptors in the TBM by two-step RT-PCR. Since functional GABAA receptors are known to contain β-subunits (McKernan and Whiting, 1996), we examined subunits β1 to β3 of the GABAA receptor. glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was employed as a positive control. Complementary DNA synthesized from the whole brain and TBM total RNA were used as templates. In the brain, GAPDH, subunits β1 to β3 of the GABAA receptor, and GABAB receptors 1 and 2 were detected by PCR. By contrast, only GAPDH was detected in the TBM (Fig. 5).

82

3.

BR A IN RE S EA RCH 1 2 01 ( 20 0 8 ) 7 8 –87

Discussion

In this study, we demonstrated that 1) GABA-containing fibers and glia-like cells were present in MSs of the upper limb, 2) GABA-containing fibers occurred in peripheral nerves which are pathways to muscles in the upper limb, 3) GABA-containing fibers were present in the VRs of C5 to T5 spinal segments,

but not in the DRs or DRGs, 4) GABA-ir structures in the TBM disappeared after transection of the brachial plexus, 5) a proportion of γ-motoneurons which project to the TBM expressed GAD67, but not VGAT mRNA, and 6) there was no evidence for the expression of mRNA for the GABAA or GABAB receptor in the TBM. These observations clearly indicate that GABAcontaining γ-motoneurons in the lower cervical spinal cord send their fibers to MSs in the upper extremities.

BR A IN RE S E A RCH 1 2 01 ( 20 0 8 ) 7 8 –8 7

83

Fig. 5 – Absence of mRNAs of GABA receptors in the TBM as revealed by RT-PCR. Although GAPDH was detected in extract from both the brain (b) and TBM (m), subunits β1–3 of the GABAA receptor, and GABAB receptors 1 and 2 were detected only in the brain, not in the TBM.

3.1.

Technical considerations

We could not determine whether some of the GABA-ir structures were nerve fibers in MSs of the TBM or not, because in our preliminary experiments, immunoreactivity for GAD67 and ChAT especially in muscles decreased with strong fixation, such as following prolonged postfixation or the use of a glutaraldehydecontaining fixative. However, we observed 1) GABA-ir nerve fibers in nerve bundles near MSs, and 2) GAD67-ir nerve fibers both in nerve bundles and around intrafusal muscles. These observations strongly suggest the presence of GABA-containing fibers around MSs.

3.2. GABA immunoreactivity in glia-like cells around intrafusal fibers We observed GABA-ir glia-like cells around intrafusal fibers. In muscle spindles, Schwann cells were shown to wrap both intrafusal muscles and nerve fibers (Banker and Girvin, 1971). Therefore, the GABA-ir cells observed in the MSs were likely to be Schwann cells. We did not obtain any evidence of the production of GABA in GABA-ir glia-like cells in MSs. The disappearance of GABA immunoreactivity in the TBM after transection of the brachial plexus implies that GABA-ir glialike cells do not produce GABA but take up and accumulate GABA released from the neighboring GABAergic terminals. In our recent study (Ito et al., 2007), a similar phenomenon was observed: GABA-ir small intensely fluorescent cells in the superior cervical ganglion did not express GAD65 or GAD67,

and disappeared after transection of the cervical sympathetic trunk.

3.3.

GABA-containing motoneurons in the spinal cord

Previously (Ito et al., 2005, 2007; Ito and Nojyo, 2008), we demonstrated that a proportion of sympathetic preganglionic neurons located in the rostral part (about C8 to T5 spinal segments) of the intermediolateral nucleus contain GABA as well as acetylcholine. In the present study, a part of γ-motoneurons located in the cervical spinal cord contain both GABA and acetylcholine. These observations imply that motoneurons which contain both GABA and acetylcholine are located in the cervical and thoracic spinal cord, corresponding to the distribution of GAD67-ir fibers in the VRs (Fig. 2d). Because motor neuron columnar fate is imposed by transcription factors (Dassen et al., 2003), it may also be probable that the expression of GADs in motoneurons is controlled by transcription factors.

3.4. Possible mechanism for the release of GABA in γ-motoneurons VGAT mRNA was not detected in GAD67 mRNA-expressing γ-motoneurons, suggesting that while these neurons can produce GABA, they cannot release it from their synaptic vesicles. The simplest explanation is that GABA is not released from terminals of γ-motor fibers. However, as discussed in the preceding section, GABA-ir glia-like cells of MSs were likely to store GABA released by other cells. Both immunohistochemistry

Fig. 4 – GABA was produced by the γ-motoneurons in the spinal cord, but not by glia-like cells in muscle spindles. Seven days after transection of the brachial plexus of the right side, GABA-ir cells (arrowhead in b–b′′) surrounding intrafusal muscles were observed on the contralateral side (a, b; b is larger magnification of the boxed region in a), but had disappeared on the ipsilateral side (c). Note that intrafusal muscle is identified by the specific nuclear distribution (e.g. nuclear chain or bag) visualized by DAPI. A combination of immunohistochemistry for ChAT and fluorescent in situ hybridization for GAD67 mRNA (d) revealed that some medium-sized neurons (mean diameter < 30 μm; arrow in d–d′′) but not large neurons (arrowhead in d–d′′) in the ventral horn (VH) of the cervical spinal cord expressed GAD67. Furthermore, in animals injected with fluorogold (FG) into the TBM, some γ-motoneurons innervating the TBM (mean diameter < 30 μm; arrows in e–e′′′) expressed GAD67 but not VGAT mRNA. Indeed, 36.5 ± 15.3% of FG-ir γ-motoneurons expressed GAD67 (N=4). α-motoneurons projecting to the TBM expressed neither GAD67 nor VGAT mRNA (arrowhead in e–e′′′). Note that an interneuron (double arrows in e–e′′′) expressed both GAD67 and VGAT. Photographs were obtained from longitudinal sections. Bar: 50 μm (a, c), 20 μm (b), 100 μm (d), and 40 μm (e).

84

BR A IN RE S EA RCH 1 2 01 ( 20 0 8 ) 7 8 –87

for GAD67 and in situ hybridization for GAD67 mRNA in the TBM clearly suggested that there is no source of GABA other than γ-motor fibers. Therefore, it is probable that nonvesicular release of GABA (Attwell et al., 1993) occurs in γ-motor terminals. In the hippocampus, granule cells showed a modest expression of GAD67 but not VGAT mRNA under normal conditions (Sperk et al., 2003). Nevertheless, granule cells seemed to have the ability to release GABA nonvesicularly via the reversal of membrane GABA transporters (Wu et al., 2001, 2006). The presence or absence of this mechanism in the GABA-containing γ-motoneurons (and also in GABA-containing glia-like cells in the MSs) should be elucidated by conducting electrophysiological studies.

3.5.

Functional considerations

It is difficult to speculate on the function of GABA-containing γ-motoneurons, because 1) the release of GABA has not been confirmed, and 2) the expression of GABAA or GABAB receptors in the TBM has not been observed. The former issue was discussed in the previous section. The latter observation suggests that GABA is not utilized as a neurotransmitter in the TBM. Electrophysiological recording from MSs may answer whether GABA works as a neurotransmitter or not. The function of the storage of GABA in the glia-like cells is also enigmatic. There are reports about GABA in astrocytes (Ochi et al., 1993; Echigo and Moriyama, 2004). Since astrocytes can also release GABA nonvesicularly by reversed uptake (Attwell et al., 1993), it is likely that GABA-containing glia-like cells release GABA. Again, more physiological experiments are needed.

4.

Experimental procedures

4.1.

Animals

Sixteen adult male Wistar rats (body weight: 200–250 g, Japan SLC, Shizuoka Japan) were used in this study. All animals were maintained and treated according to the Guidelines for Animal Experiments of University of Fukui. All efforts were made to minimize the suffering and number of animals used.

4.2.

Surgery

4.2.1.

Retrograde tracing

Three rats were deeply anesthetized with chloral hydrate (400 mg/kg body weight, i.p.). The right TBM was exposed, and injected with a total of 2 μl of 4% FG at 10 points through a Hamilton syringe. After the injection, the animals were allowed to survive for 3 days.

4.2.2.

Nerve transection

Three rats were deeply anesthetized with chloral hydrate (400 mg/kg body weight, i.p.). The brachial plexus was exposed, and cut with spring scissors. After the surgery, animals were allowed to survive for 7 days.

4.3.

Immunohistochemistry

4.3.1.

Antibodies

We used a mouse monoclonal antibody for GAD67 (MAB5406, Chemicon, Temecula, CA), a goat polyclonal antibody for ChAT (AB144, Chemicon), and rabbit polyclonal antibodies for FG (AB153, Chemicon), and GABA (A2052, Sigma-Aldrich, St Louis, MO) as primary antibodies (Table 1). The specificity of each of these antibodies was adequately checked in previous studies (Table 1; Ligorio et al., 2000; Fong et al., 2005; Ito et al., 2007; Mugnaini and Oertel, 1985).

4.3.2.

Preparation of tissues

Animals were deeply anesthetized with an overdose of chloral hydrate (800 mg/kg body weight, i.p.) and perfused transcardially with saline, followed by a fixative containing 4% paraformaldehyde diluted with 0.1 M phosphate buffer (PB). In the immunohistochemistry for GABA, we instead used 4% paraformaldehyde, 0.05% glutaraldehyde diluted with 0.1 M PB as a fixative. The DRs, VRs, and DRGs from C1 to S4 levels, cervical spinal cord, brachial plexus, and forelimbs were dissected out and postfixed with the same fixative for 24 h and 1 h at room temperature (RT), respectively. The specimens were immersed in 30% sucrose in 0.1 M PB overnight at 4 °C. After embedding in OCT compound (Tissue Tek, Redding, CA), frozen sections of the spinal cord were made longitudinally at a thickness of 35 μm by cryostat, and collected in 0.05 M phosphate-buffered

Table 1 – Details of antibodies used for immunohistochemistry Antibody

Host

Mono or polyclonal

Anti-gamma aminobutyric acid (GABA)

Rabbit

Poly

SigmaAldrich

Anti-glutamic acid decarboxylase of 67 kDa (GAD67)

Mouse

Mono

Chemicon MAB5406 25080061

Anti-fluorogold (FG)

Rabbit

Poly

Chemicon AB153

0509010863 Fluorogold

Poly

Chemicon AB144

24050789

Anti-cholineGoat acetyltransferase (ChAT)

Source

Catalog number A2052

Lot number 101K4837

Antigen GABA conjugated with bovine serum albumin (GABABSA) Recombinant GAD67 protein

Recombinant rat ChAT protein

Specificity Absorbed with GABA; only GABA-BSA, but not glycine-BSA, was detected on dot blot Ligorio et al. (2000). Absorbed with GST-GAD67 fusion protein Ito et al. (2007); single band around 67kDa on the immunoblot Fong et al. (2005); immunoreactivity was well consistent with a previous study Mugnaini and Oeltel (1985). Lack of immunoreactivity in the sections of animals which were not subjected with fluogold. Absorbed with recombinant ChAT protein (AG220, Chemicon).

85

BR A IN RE S E A RCH 1 2 01 ( 20 0 8 ) 7 8 –8 7

Table 2 – Details of riboprobes used for in situ hybridization Riboprobe

GenBank Species Sequence Length accesion number (base)

GAD67 GAD65 VGAT

NM_017007 NM_012563 NM_031782

Rat Rat Rat

226–844 143–766 620–1599

619 624 980

saline (PBS) or DEPC-treated 0.1 M PB. Sections of the DRs, VRs, DRGs, and brachial plexus were made parallel or transverse (for some VRs) to the longitudinal axis at a thickness of 14 μm by cryostat, mounted on APS-coated glass slides, and air-dried.

4.3.3.

Immunohistochemistry for confocal microscopy

Sections were incubated overnight with goat anti-ChAT (1:100), rabbit anti-FG (1:500), mouse anti-GAD67 (1:1000), and rabbit anti-GABA (1:3000) diluted in 0.3% Triton X-100/1% normal donkey serum in 0.05 M PBS (PBS-XD), followed by Cy3 donkey anti-goat IgG (1:200; Rockland Immunochemicals, Gilbertsville, PA) to visualize ChAT-immunoreactivity, fluorescein isothiocyanate (FITC) donkey anti-[mouse IgG] (1:200; Rockland) for GAD67, and FITC donkey anti-[rabbit IgG] (1:200; Chemicon) for GABA diluted in PBS-XD. It is known that GADimmunoreactivity in neuronal somata increases when the sections are incubated without Triton X-100 (Mugnaini and Oertel, 1985). We thus incubated some sections without Triton X-100 to enhance the immunoreactivity of neuronal somata. Some sections were stained with DAPI (1:1000; Dojindo, Kumamoto, Japan) diluted with PBS. Then, sections were mounted on glass slides with VECTORSHIELD (Vector Laboratories, Burlingame, CA).

4.3.4.

Immunohistochemistry for bright field microscopy

The sections were incubated with mouse anti-GAD67 (1:3000), and rabbit anti-GABA (1:10000) diluted in PBS-XD, followed by biotinylated donkey secondary antibodies (biotinylated anti[mouse IgG], or biotinylated anti-[rabbit IgG]; 1:200; Rockland) diluted in PBS-XD, and further incubated with avidin-biotinylated peroxidase complex (1:50; ABC-Elite, Vector) in PBS containing 0.3% Triton X-100 (PBS-X). To reveal GAD67 immunoreactivity in the neuronal somata, we processed some sections without Triton

X-100. These sections were used for the Nickel-DAB reaction. Some sections were counterstained with Neutral Red (Merck, Germany). The sections were dehydrated through graded alcohol, cleared with xylene, and mounted on glass slides with Entellan (Merck).

4.3.5.

Bright field in situ hybridization histochemistry

Complementary DNA fragments corresponding to a region of the rat GAD65 (nucleotides of 69–693, GenBank accession number NM_012563), GAD67 (nucleotides of 226–844, GenBank accession number NM_017007), and VGAT (nucleotides of 620– 1599, GenBank accession number NM_031782) cDNAs (Table 2) were cloned into pBluescript II SK (+) (Stratagene, La Jolla, CA). Using this plasmid as a template, sense and antisense singlestranded RNA probes were synthesized with a digoxigenin or fluorescein labeling kit (Roche Diagnostics, Mannheim, Germany). The specificity of the riboprobe for GAD67 was established in our previous study (Ito et al., 2007). The specificity of the riboprobe for VGAT was checked by in situ hybridization as described below and by comparing the distribution of VGATexpressing neurons in a previous study (Sagne et al., 1997), and the distribution was confirmed to be very similar. The procedure for non-radioactive in situ hybridization was described elsewhere (Liang et al., 2000; Ito et al., 2007). The free-floating sections of brain and spinal cord were washed in 0.1 M PB (pH 7.0) for 5 min twice, immersed in PB containing 0.3% Triton X100, and washed in 0.1 M PB. Then, sections were treated with acetylation buffer (0.003% acetic acid anhydrate/1.3% (v/v) triethanolamine/6.5% (w/v) HCl in DEPC-treated water) for 10 min at RT. After being washed in 0.1 M PB twice, the sections were incubated in a prehybridization buffer containing 50% (v/v) formamide (Nacalai Tesque, Kyoto, Japan)/5× SSC/2% blocking reagents (Roche)/0.1% N-lauroylsarcosine (NLS)/0.1% SDS for 1 h at 60 °C. Then, the sections were hybridized with 1 μg/ml digoxigenin-labeled sense or antisense RNA probe for GAD67 and VGAT in prehybridization buffer for 20 h at 60 °C and 70 °C, respectively. After two washes in 2× SSC/50% formamide/0.1% NLS for 20 min at 60 °C or 70 °C, the sections were incubated with 20 μg/ml RNase A (Nacalai) for 30 min at 37 °C, washed in 2× SSC/ 0.1% NLS for 20 min twice at 37 °C, and washed in 0.2× SSC/0.1% NLS for 20 min twice at 37 °C. These sections were incubated with 1% blocking reagent (Roche) diluted in Tris–HCl (pH 7.5),

Table 3 – Details of primers used for two-step RT-PCR Target GABAA receptor β1 subunit a GABAA receptor β2 subunit a GABAA receptor β3 subunit a GABAB receptor 1 GABAB receptor 2 GAPDH

a

Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse

Primer Sequence

Starting at base

Product size (bp)

5′-ACAGCTCCAATGAACCCAGCAA-3′ 5′-TGCTCCCTCTCCTCCATTCCA-3′ 5′-GGAGTGACAAAGATTGAGCTTCCT-3′ 5′-GTCTCCAAGTCCCATTACTGCTTC-3′ 5′-CCGTCTGGTCTCCAGGAATGTTGTC-3′ 5′-CGATCATTCTTGGCCTTGGCTGT-3′ 5′-AAGTGGGGCTGGAAGAAGAT-3′ 5′-CCAGGCTGGAGAGAACTGAG-3′ 5′-GGCTGACACACTGGAGATCA-3′ 5′-GATGAGCTTTGGCACAAACA-3′ 5′-CCTGGAGAAACCTGCCAAGTAT-3′ 5′-GGTCCTCAGTGTAGCCCGAGAT-3′

154 674 677 1240 718 1128 1085 1982 1137 2041 814 908

521

Liu and Burt, J Neurosci Methods 85 (1998) 89–98.

564 411 898 905 95

86

BR A IN RE S EA RCH 1 2 01 ( 20 0 8 ) 7 8 –87

0.15 M NaCl (TS 7.5) for 1 h at RT, and then with alkaline phosphatase-conjugated anti-digoxigenin antibody Fab fragment (1:2000; Roche) in 1% blocking reagent (Roche) diluted in TS7.5 at RT overnight. The bound phosphatase was visualized by a reaction with NBT/BCIP for 4 h at 37 °C in TS9.5. Sections were mounted on glass slides, dehydrated, cleared with xylene, and coverslipped.

DNA as a template, and confirmed that there was no amplification. After denaturation at 94 °C for 5 min, PCR was carried out for 30 cycles: 94 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s. The final extension time was 7 min. The PCR products were separated by electrophoresis in 2% agarose gels containing 0.5 µg/ml ethidium bromide.

4.5. 4.3.6.

For fluorescent in situ hybridization for GAD67 and VGAT, the fluorescein-labeled riboprobe for GAD67 and digoxigeninlabeled riboprobe for VGAT were diluted in hybridization buffer at a final concentration of 1 µg/ml. The hybridization temperature was 70 °C. After the incubation and washing described above, these sections were incubated with alkaline phosphatase-conjugated anti-fluorescein antibody Fab fragment (1:2000; Roche), peroxidase-conjugated anti-digoxigenin antibody Fab fragment (1:200; Roche), rabbit anti-FG (1:500), and goat antiChAT (1:100) in 1% blocking reagent (Roche) diluted in TS7.5 at RT overnight. The sections were washed 3 times, then the bound peroxidase was reacted with biotinylated-tyramide signal amplification (Perkin-Elmer, Waltham, MA). After 3 more washes, sections were incubated with Cy5-conjugated streptavidin (1:100; GE healthcare UK, Buckinghamshire, UK) and Alexa488conjugated anti-[goat or rabbit] IgG (1:400; Invitrogen, Carlsbad, CA). To visualize alkaline phosphatase by fluorescent microscopy, sections were developed with 0.005% (w/v) FastRed (Roche), 1% (v/v) HNPP (Roche) diluted in Tris–HCl (pH 8.0), 0.15 M NaCl for 30 min at RT. The sections were mounted on glass slides with Permafluor (Beckman Coulter, Fullerton, CA).

4.4.

RT-PCR

4.4.1.

Total RNA extraction and reverse transcription

Three rats were deeply anesthetized with an overdose of chloral hydrate (800 mg/kg body weight, i.p.), and killed by cutting the common carotid artery. The cerebral cortex and TBM were quickly dissected out, and frozen with liquid nitrogen. Total RNA was extracted with Trizol (Invitrogen), precipitated with iso-propanol, washed with 80% EtOH, air-dried, and dissolved in 50 μl of DEPC-treated water. The yield of the RNA was determined by measuring the absorbance of an aliquot at 260 and 280 nm. Some RNA samples were treated with DNase I (Takara Bio, Otsu, Japan) to remove genomic DNA. Then complementary DNA was synthesized from the total RNA by using reverse transcriptase (SuperScript III RT, Invitrogen). The template RNA was digested by RNase H (Invitrogen).

4.4.2.

Micrographs and counting

Fluorescent in situ hybridization histochemistry

Polymerase chain reaction

We used primers for subunits β1-3 of the GABAA receptor, GABAB receptors 1 and 2, and GAPDH (Table 3). Sequences of the primers for GABAA receptors were obtained from a previous study (Liu and Burt, 1998). Sequences of the other primers were determined with Primer3 software (http://primer3.sourceforge.net/). One microliter of each RT product was added to 50 µl of PCR solution (Blend Taq, Toyobo, Osaka, Japan), containing 25 mM TAPS (pH 9.3), 50 mM KCl, 10 mM MgCl2, 200 µM dNTPs, 100 µM [α-32P]-dCTP, 20 mg of activated salmon sperm DNA, 500 pM of each primer, and 0.6 U of Blend Taq DNA polymerase. In some experiments, we used total RNA instead of complementary

All micrographs were collected digitally. Bright field micrographs were taken with a CCD-camera (HC2500, Fujifilm, Japan). Fluorescent micrographs were taken under a confocal laser scanning microscope (TCS-SP2-AOBS, Leica Microsystems, Germany). FITC and Alexa488 were excited with a 488 nm laser beam and observed through a 510–560 nm AOBS emission filter. FastRed and Cy3 were excited with a 543 nm laser beam, and observed through a 570–630 nm AOBS emission filter. Cy5 was excited with a 633 nm laser beam, and observed through a ≥650 nm AOBS emission filter. We used an ×63 oil-immersive objective (N.A.= 1.4; Leica) for the determination of axonal colocalization in double immunostained sections. Minimal adjustments of brightness and contrast were done with Photoshop CS3(Adobe Systems, San Jose, CA) for both bright field and confocal micrographs. When counting the number of GAD67-immunoreactive fibers in the VRs, we used transverse sections of the VRs of C1 to S4 spinal segments obtained from 3 animals. On counting the number of neurons, serially collected longitudinal sections of the spinal cord of C3 to T2 spinal segments obtained from 4 animals were used. To avoid double-counting, we only counted the number of neurons of which nucleoi could be observed. We also calculated the mean diameter, which was the mean of the major and minor axes of the cell body.

Acknowledgments This work was supported by Grants-in-Aid from the Ministry of Education, Science and Culture of Japan (T. Ito, No. 18700339, Y. Nojyo, No. 14580726, 19500292, T. Kaneko, No. 16200025, 16500217, 17022020, 17022024, 15-5638, H. Hioki, No. 18700341).

REFERENCES

Attwell, D., Barbour, B., Szatkowski, M., 1993. Nonvesicular release of neurotransmitter. Neuron 11, 401–407. Banker, B.Q., Girvin, J.P., 1971. The ultrastructural features of the mammalian muscle spindle. J. Exp. Neurol. 30, 155–195. Banks, R.W., Barker, D., Stacey, M.J., 1985. On the classification of motor endings in mammalian muscle spindles. In: Boid, I.A., Gladden, M.H. (Eds.), The muscle spindle. Stockton Press, New York. Dale, H., 1935. Pharmacology and nerve endings. Proc. R. Soc. Med. 28, 319–332. Dassen, J.S., Liu, J.P., Jessell, T.M., 2003. Motor neuron columnar fate imposed by sequential phases of Hox-c activity. Nature 425, 926–933. Echigo, N., Moriyama, Y., 2004. Vesicular inhibitory amino acid transporter is expressed in gamma-aminobutyric acid (GABA)-containing astrocytes in rat pineal glands. Neurosci. Lett. 367 (1), 79–84.

BR A IN RE S E A RCH 1 2 01 ( 20 0 8 ) 7 8 –8 7

Fong, A.Y., Stornetta, R.L., Foley, C.M., Potts, J.T., 2005. Immunohistochemical localization of GAD67-expressing neurons and processes in the rat brainstem: subregional distribution in the nucleus tractus solitarius. J. Comp. Neurol. 493 (2), 274–290. Ichiyama, R.M., Broman, J., Edgerton, V.R., Havton, L.A., 2006. Ultrastructural synaptic features differ between alpha- and gamma-motoneurons innervating the tibialis anterior muscle in the rat. J. Comp. Neurol. 499 (2), 306–315. Ito, T., Nojyo, Y., 2008. GABAergic fibers in the rostral paravertebral sympathetic ganglia: do they originate from intrinsic neurons or preganglionic neurons? In: Kaneko, M. (Ed.), Sympathetic Nervous System Research Developments. Nova Publishers, New York. Ito, T., Iino, S., Nojyo, Y., 2005. A part of cholinergic fibers in mouse superior cervical ganglia contain GABA or glutamate. Brain Res. 1046 (1–2), 234–238. Ito, T., Hioki, H., Nakamura, K., Tanaka, Y., Nakade, H., Kaneko, T., Iino, S., Nojyo, Y., 2007. Gamma-aminobutyric acid-containing sympathetic preganglionic neurons in rat thoracic spinal cord send their axons to the superior cervical ganglion. J. Comp. Neurol. 502 (1), 113–125. Johnson, H., Hokfelt, T., Ulfhake, B., 1992. Galanin- and CGRP-like immunoreactivity coexist in rat spinal motoneurons. Neuroreport 3 (4), 303–306. Liang, F., Hatanaka, Y., Saito, H., Yamamori, T., Hashikawa, T., 2000. Differential expression of gamma-aminobutyric acid type B receptor-1a and -1b mRNA variants in GABA and non-GABAergic neurons of the rat brain. J. Comp. Neurol. 416 (4), 475–495. Ligorio, M.A., Akmentin, W., Gallery, F., Cabot, J.B., 2000. Ultrastructural localization of the binding fragment of tetanus toxin in putative gamma-aminobutyric acidergic terminals in the intermediolateral cell column: a potential basis for sympathetic dysfunction in generalized tetanus. J. Comp. Neurol. 419 (4), 471–484. Liu, Z.F., Burt, D.R., 1998. A synthetic standard for competitive RT-PCR quantitation of 13 GABA receptor type A subunit mRNAs in rats and mice. J. Neurosci. Methods 85, 89–98.

87

McKernan, R.M., Whiting, P.J., 1996. Which GABAA-receptor subtypes really occur in the brain? Trends Neurosci. 19, 139–143. Mugnaini, E., Oertel, W.H., 1985. An atlas of the distribution of GABAergic neurons and terminals in the rat CNS as revealed by GAD immunohistochemistry. In: Bjorklund, A., Hokfelt, T. (Eds.), Handbook of chemical neuroanatomy, vol. 4. Elsevier, New York. Ochi, S., Lim, J.Y., Rand, M.N., During, M.J., Sakatani, K., Kocsis, J.D., 1993. Transient presence of GABA in astrocytes of the developing optic nerve. Glia 9 (3), 188–198. Piehl, F., Arvidsson, U., Hokfelt, T., Cullheim, S., 1993. Calcitonin gene-related peptide-like immunoreactivity in motoneuron pools innervating different hind limb muscles in the rat. Exp. Brain Res. 96 (2), 291–303. Sagne, C., El Mestikawy, S., Isambert, M.F., Hamon, M., Henry, J.P., Giros, B., Gasnier, B., 1997. Cloning of a functional vesicular GABA and glycine transporter by screening of genome databases. FEBS Lett. 417 (2), 177–183. Shupliakov, O., Ornung, G., Brodin, L., Ulfhake, B., Ottersen, O.P., Storm-Mathisen, J., Cullheim, S., 1993. Immunocytochemical localization of amino acid neurotransmitter candidates in the ventral horn of the cat spinal cord: a light microscopic study. Exp. Brain Res. 96 (3), 404–418. Sperk, G., Schwarzer, C., Heilman, J., Furtinger, S., Reimer, R.J., Edwards, R.H., Nelson, N., 2003. Expression of plasma membrane GABA transporters but not of the vesicular GABA transporter in dentate granule cells after kainic acid seizures. Hippocampus 13, 806–815. Strack, A.M., Sawyer, W.B., Marubio, L.M., Loewy, A.D., 1988. Spinal origin of sympathetic preganglionic neurons in the rat. Brain Res. 455, 187–191. Wu, Y., Wang, W., Richerson, G.B., 2001. GABA transaminase inhibition induces spontaneous and enhances depolarization-evoked GABA efflux via reversal of the GABA transporter. J. Neurosci. 21 (8), 2630–2639. Wu, Y., Wang, W., Richerson, G.B., 2006. The transmembrane sodium gradient influences ambient GABA concentration by altering the equilibrium of GABA transporters. J. Neurophysiol. 96 (5), 2425–2436.