Correspondence to: Yukio Oomori, Division of Anatomy and Physiology, Japanese Red Cross Hokkaido College of Nursing, 664-1 Akebonocho Kitami, Hokkaido 090-0011 Japan. Tel/Fax:+81-157-66-3339. E-mail: firstname.lastname@example.org
Gamma-aminobutyric acid (GABA) exerts its inhibitory actions through GABA receptors in the central and peripheral nervous systems. Previous studies have shown GABA in the nervous systems as a neurotransmitter (Wenthold et al., 1986; Wolff et al., 1986; Kosaka et al.,1987; Furness et al., 1989; Dobó et al., 1990) and in endocrine cells as a hormone (Okada et al., 1976; Taniguchi et al., 1979; Kataoka et al., 1984; Alho et al., 1986; Sakaue et al., 1987; Ahonen et al., 1989; Gilon et al., 1991; Oomori et al., 1992-1994; Iwasa et al., 1998, 1999b). Thus, GABA may be an important bioactive material in both the nervous and endocrine systems.
In the nervous systems, two distinct types of GABA receptors are well known as GABAA and GABAB receptors. GABAA receptors are ionotropic receptors, permeable to chloride ions, where the action of GABA is antagonized by bicuculline (Macdonald and Olsen, 1994). GABAB receptors are metabotropic receptors, blocked by baclofen, which mediates neuronal responses via second messenger systems regulating calcium and potassium channels (Bowery, 1989).
In the adrenal medulla, previous immunohistochemical studies have demonstrated GABA or glutamate decarboxylase (GAD) immunoreactivity in some chromaffin cells as well as in numerous nerve fibers (Kataoka et al., 1984; Alho et al., 1986; Ahonen et al., 1989; Oomori et al., 1992, 1993; Iwasa et al., 1998, 1999b).
Pharmacological, molecular studies have reported the presence of both GABAA and GABAB receptors in adrenal chromaffin cells (Kataoka et al., 1986; Fujimoto et al., 1987; Castro et al., 2003, 1988; Ymer et al., 1989; Droshenko and Neher, 1991; Parramon et al., 1995). However, it is still unknown which types of chromaffin cells, that is, adrenaline (A) cells and noradrenaline (NA) cells or neuronal elements have GABAB receptors in the adrenal medulla.
In order to clarify the issues mentioned earlier, the present immunohistochemical study using a formaldehyde-induced fluorescence (FIF) method examined the localization of GABAB receptors and GABA in each cell type of A and NA cells and in the neuronal elements. Furthermore, we examined the colocalization of choline acetyltransferase (ChAT) or neuronal nitric oxide synthase (nNOS) antibodies with GABAB receptors or GABA in the mouse adrenal medulla, and discuss the possible effects of GABA and GABAB receptor immunoreactive cellular elements on the adrenal functions.
MATERIALS AND METHODS
Thirteen male mice (ddY, 8-weeks old, about 30–40 g in body weight, Japan SLC, Shizuoka, Japan) were used in this study. All experimental procedures were performed according to the Guidelines for Animal Care by the Japanese Red Cross Hokkaido College of Nursing. The animals received commercial food pellets and water ad libitum. They were kept under constant conditions (temperature 22°C, relative humidity 45%, LD12hr light from 0700 to 1900 hr).
The animals were anesthetized with ether and perfused with 50 mL of physiological saline and 50 mL of 4% paraformaldehyde or 0.3% glutaraldehyde plus 4% paraformaldehyde in 0.1 M phosphate buffered (PB) pH 7.4 through the heart. The adrenal gland was removed and immersed in the same fixative for 2 hr at 4°C. After rinsing in PB saline (PBS), the adrenal gland was left overnight in PBS containing 30% sucrose at 4°C. The adrenal gland was cut about 12 µm thick using a cryostat, and mounted on glass slides coated with poly-l-lysine (Sigma, St. Louis, Mo). To identify NA cells in the medulla, the cryostat sections were examined and photographed using a Zeiss fluorescent microscope equipped with a filter No. 9 for NA fluorescence. Fixation containing 4% paraformaldehyde is suitable for demonstrating NA fluorescence in these tissues (Falck and Torp, 1961). In order to confirm the distribution of NA cells and A cells in the mouse adrenal medulla, we used both the FIF for NA cells and phenylethanolamine N-methyltransferase (PNMT) immunohistochemistry for A cells as a control in the same sections of the mouse adrenal medulla. For immunohistochemistry and FIF, the cryostat sections were photographed by fluorescence microscope and then immunostained by the primary antibodies. For immunofluorescent microscopy, the tissues after overnight at 4°C with primary antiserum (Table 1) were washed in PBS three times, 10 min each, and incubated for 2 hr with a single secondary antibody or a mixture of secondary antibodies conjugated with indocarbocyanine (Cy3), cyanine (Cy2) (Table 1).
Table 1. List of primary antisera and secondary fluorescence conjugated antisera used for immunohistochemistry in the present study
The specificity of the immunohistochemical staining was confirmed by replacing the primary antibody with normal rabbit serum, and by using diluted antiserum pretreated with adequate antigen (5.7–10 µg/mL) for 24 hr at 4°C. No immunostaining was observed under these conditions.
At the microscopic level, we randomly selected at least 10 fields of view of the adrenal medulla from at least 10 sections obtained from each of 5 animals. The selected fields were photographed using a digital camera (EOS KISS 5: Canon, Tokyo, Japan). Adobe Photoshop Elements (version 2.0; Adobe Systems, San Jose, CA) and Win ROOF Professional Ver.5.7 (MITANI CORPORATION, Fukui, Japan) were used to view the photographed images and measure the unit area (1 mm2). The number of varicosities of ChAT-immunoreactive nerve fibers in the areas of A cells and NA cells was counted and calculated and shown as mean ± S.D. in the 1 mm2. The cell size of adrenal chromaffin and ganglion cells was also measured using the aforementioned software.
GABAB receptor immunoreactivity was seen in numerous chromaffin cells (ca. 15 µm in diameter) in the medulla, and was also visible as fine dots on chromaffin cells. By using immunohistochemical techniques combined with a FIF method, GABAB receptor immunoreactivity was found predominantly in A cells, but not in the NA cells showing blue-white fluorescence (Fig. 1a,b). GABAB receptor immunoreactivity was also seen in a few ganglion cells (20–40 µm in diameter) with a large nucleus in the medulla (Fig. 2a). In such cases, it was denser than that of the chromaffin cells. nNOS immunoreactivity was seen in ganglion cells, and nerve fibers and nerve bundles, but not in numerous chromaffin cells of the medulla. By using a double immunostaining method, GABAB receptor immunoreactive ganglion cells were shown to be nNOS immunopositive (Fig. 2a,b).
Weak GABA immunoreactivity was found only in some clusters of chromaffin cells, but not in the cortical cells. Using the FIF method, these immunoreactive cells showed blue-white NA fluorescence (Fig. 3a,b). However, no GABA immunoreactivity was observed in ganglion cells of the mouse adrenal medulla. Thick GABA-immunoreactive nerve bundles ran through the cortex directly into the medulla and divided into thinner nerve fibers in the medulla. However, no GABA-immunoreactive nerve fibers ran along blood vessels in the cortex. GABA-immunoreactive varicose nerve fibers were frequently observed around the A cells without the fluorescence, but not around the NA cells with the fluorescence. These immunoreactive nerve fibers were dense and in close contact with A cells, but not with NA cells. GABA-immunoreactive nerve fibers were in close contact with a few ganglion cells.
By using a double immunofluorescence method with GABA and ChAT antibodies, intra-adrenal GABA-immunoreactive nerve fibers in the area of the A cells exhibited all ChAT immunoreactivity (Fig. 3e,f). However, no GABA-immunoreactive nerve fibers were seen in NA cell areas showing blue-white fluorescence. ChAT-only immunoreactive nerve fibers in NA cell area showing the fluorescence were more numerous than those in A cell areas not showing the fluorescence (Fig. 3c,d). The number of varicosities of ChAT-immunoreactive nerve fibers in the NA cell area and in the A cell area were 5,667 ± 962, 2,707 ± 227 (average varicosities/mm2 ± S.D.), respectively.
In the control, clusters of NA cells showing blue-white fluorescence using the FIF method and A cells showing PNMT immunoreactivity by immunohistochemistry were identified in the same sections of the mouse adrenal medulla (Fig. 4a,b). Immunohistochemical staining was confirmed by replacing the primary antibody with normal rabbit serum. No immunoreactivity was observed in the control sections of the mouse adrenal medulla (Fig. 4c,d).
The present immunohistochemical study combined with histofluorescence revealed that GABAB receptor immunoreactive A cells but not NA cells were innervated by GABA-immunoreactive nerve fibers in the mouse adrenal medulla. This suggests the possibility that GABA released from these nerve fibers may activate GABAB receptor in A cells and consequently influence the release of catecholamines from A cells of the mouse adrenal medulla. Previous physiological studies showed that GABA and the metabotropic GABAB receptor system increased basal secretion and inhibited the evoked catecholamine secretion from chromaffin cells (Oset-Gasque et al., 1993; Parramón et al., 1995). In addition, the present study showed colocalization of GABA and ChAT in numerous intra-adrenal nerve fibers of the mouse adrenal medulla. This supports a previous study which reported intra-adrenal GABA-immunoreactive nerve fibers in the mouse exhibited AChE active (Iwasa et al., 1999b). From these results, it can be inferred that GABA and acetylcholine may be released from the same presynaptic nerve fibers and may have secretory effects on the A cells of the mouse adrenal medulla. Previous physiological studies have actually reported that GABA inhibited the secretion of catecholamines from the chromaffin cells by cholinergic stimulation (Kataoka et al., 1986; Oset-Gasque et al., 1993). Our previous immunoelectron microscopic study showed membrane specialization in the chromaffin cells at the contact between the GABA-immunoreactive nerve fiber and chromaffin cell in the mouse adrenal medulla (Oomori et al., 1993). GABAB receptors were present at presynaptic terminals where it serves as an autoreceptor to influence transmitter release. They were also present on the postsynaptic neurons where activation produces an increase in membrane K+ conductance and associated neuronal hyperpolarization (Bowery et al., 2002). This suggests the possibility that GABA may be released from the nerve fibers and have similar inhibitory effects on the A cells via pre and postsynaptic GABAB receptors.
In the present immunohistochemical study, GABAB receptor immunoreactivity was observed in the A cells of the mouse adrenal medulla. On the other hand, previous physiological studies demonstrated only GABAA, only GABAB, both GABAA and GABAB receptors in the A and NA cells of the bovine chromaffin cells (Castro et al., 2003), GABAA receptor in canine (Kataoka et al., 1986) and rat (Busik et al., 1996) chromaffin cells. These results suggest that the expression of GABAA and GABAB receptors on adrenal chromaffin cells may be species difference and may be different in A and NA cells. In the present study, GABA- and ChAT-immunoreactive nerve fibers were in contact with A cells in the mouse adrenal medulla. ChAT-only immunoreactive nerve fibers were in contact with NA cells and the number of varicosities of ChAT-immunoreactive nerve fibers in NA cell areas was almost more than twice that of the nerve fibers in A cell areas. Previous immunohistochemical studies have demonstrated that neurotensin and substance P immunoreactive nerve fibers only innervated NA cells in the hamster and rat adrenal medulla (Pelto-Huikko et al., 1985b; Murabayashi et al., 2007), whereas enkephalin and GABA-immunoreactive nerve fibers were only numerous in A cells (Pelto-Huikko et al., 1985a; Oomori et al., 1993; Iwasa et al., 1998). Furthermore, AChE positive nerve fibers, ChAT, vesicular acetylcholine transporter and neurocalcin immunoreactive nerve fibers were in dense contact with NA cell groups compared to A cell groups (Allen et al., 1958; Eränkö 1959; Lewis and Shute, 1968; Iino et al., 1997; Iwasa et al., 1999a; Murabayashi et al., 2009). These findings have demonstrated that A and NA cells differ in degree of innervation and type of nerves. It has been demonstrated that distinct preganglionic neurons innervate NA cells and A cells in the adrenal medulla (Edwards et al., 1996). In fact, differential NA and A secretions evoked by secretagogue are known. The release of A in effluent blood from the canine adrenal gland is greater than that of NA after administration of a GABA receptor agonist (Kataoka et al., 1986). Nicotine and high K+ cause greater secretion of NA than A (Douglas and Poisner, 1965; Marley and Livett, 1987). Thus, these differential secretions of A and NA may represent differential GABA and acetylcholine innervation between A and NA cells.
The present study showed that GABA-immunoreactive nerve fibers innervated GABAB receptor immunoreactive ganglion cells. Our previous electron microscopic study revealed that the GABA-immunoreactive nerve fibers were in close apposition to the intra-adrenal ganglion cells and the postsynaptic membrane specialization of the ganglion cells was seen at the contact point between the GABA-immunoreactive nerve fiber and the ganglion cell in the mouse adrenal gland (Oomori et al., 1993). Previous physiological studies have shown that GABA-produced depolarization depended on the resting membrane potential, and GABA-induced depolarization inhibited synaptic transmission in the ganglion cells of the autonomic ganglia (De Groat, 1970; Adams and Brown, 1975). The inhibition of GABAB receptors in neurons was mainly achieved via modulation of neurotransmitters release from presynaptic terminals and hyperpolarization of postsynaptic membranes (Bowery et al., 2002). It is probable that GABA may exert inhibitory effects on the secretory activity of the intra-adrenal ganglion cells via pre and postsynaptic GABAB receptor. Furthermore, the present study revealed that nNOS immunoreactivity was found in ganglion cells and nerve fibers, but not in chromaffin cells of the mouse adrenal medulla, and that GABAB receptor immunoreactive ganglion cells were also NOS immunopositive. NOS immunoreactivity was seen in the ganglion cells and nerve fibers (Holgert et al., 1995; Afework et al., 1996), and in the chromaffin cells (Hyme et al., 1994; Oset-Gasque et al., 1998) of the adrenal medulla. Previous studies have reported that there are two types of ganglion cells, large ganglion cells showing NPY, tyrosine hydroxylase and dopamine β-hydroxylase immunoreactivities, and small ganglion cells exhibiting vasoactive intestinal polypeptide (VIP) and NOS immunoreactivities in the rat adrenal gland (Oomori et al., 1994; Holgert et al., 1996). These findings infer that GABA from the nerve fibers may also inhibit the release of NO and VIP from the terminals of the immunoreactive ganglion cells via GABAB receptor.
The present study showed weak GABA, but not GABAB receptor immunoreactivity in the NA cells of the mouse adrenal medulla. This infers that GABA and NA may be synthesized, stored, and released from the NA cells by adequate stimuli. In fact, previous studies reported GABA or GAD immunoreactivities in some chromaffin cells in canine, mouse, and cow (Kataoka et al., 1986; Oomori et al., 1993; Iwasa et al., 1998; Castro et al., 2003).
Probably, GABA in the adrenal medulla may be released partly from the NA cells as a hormone and partly from the neuronal elements as a neurotransmitter.
As no GABA immunoreactive ganglion cells could be observed in the mouse adrenal medulla, GABA-immunoreactive nerve fibers in the present study may be considered as extrinsic in origin. The present study revealed that intra-adrenal GABA-immunoreactive nerve fibers were all ChAT immunopositive. It is well known from tracer experiments (Kesse et al., 1988) that the neurons innervated in the adrenal medulla are located in the intermediolateral nucleus of the spinal cord. Thus, the GABA nerve fibers in the mouse adrenal gland may come mainly from the neurons in the intermediolateral nucleus of the spinal cord.