Subcellular distribution of melatonin receptors in human parotid glands

Abstract

The hormone melatonin influences oral health through a variety of actions, such as anti-inflammatory, anti-oxidant, immunomodulatory and antitumour. Many of these melatonin functions are mediated by a family of membrane receptors expressed in the oral epithelium and salivary glands. Using immunoblotting and immunohistochemistry, recent studies have shown that the melatonin membrane receptors, MT1 and MT2, are present in rat and human salivary glands. To date, no investigation has dealt with the ultrastructural distribution of the melatonin receptors. This was the aim of the present study, using the immunogold method applied to the human parotid gland. Reactivity to MT1 and, with less intensity, to MT2 appeared in the secretory granules of acinar cells and in the cytoplasmic vesicles of both acinar and ductal cells. Plasma membranes were also stained, albeit slightly. The peculiar intracytoplasmic distribution of these receptors may indicate that there is an uptake/transport system for melatonin from the circulation into the saliva.

Introduction

Melatonin is usually associated with the pinealocytes, where it is synthesised and then discharged into the capillary bed in response to circadian cycles. It is produced prevalently during the night, when its blood levels are higher than during the day. Other tissues, such as the retina (Tosini & Menaker, 1998), bone marrow cells (Conti et al. 2000), skin (Slominski et al. 2005) and gastrointestinal tract, also contain melatonin (Bubenik, 2002, 2008). Locally, melatonin may exert autocrine-paracrine effects. With respect to the gastrointestinal tract, systemic effects may be expected, as it is released into the blood stream in response to food intake (Bubenik, 2002, 2008). Due to its unique lipophilic nature (Hardeland et al. 2011; Chava & Sirisha, 2012), circulating melatonin crosses the cell membranes and distributes easily into tissues and biological fluids; it occurs in milk (Cohen Engler et al. 2012), bile (Messner et al. 2001), cerebrospinal fluid (Tan et al. 2010), urine and saliva (Vakkuri, 1985; de Almeida et al. 2011; Almughrabi et al. 2013).

The free radical scavenging action of melatonin (Cutando et al. 2003, 2007; Galano et al. 2011), as demonstrated in the oral cavity, appears to be the principal non-receptor-mediated role of the hormone. Other actions, such as immunomodulatory and protective functions (Cutando et al. 2003; Carrillo-Vico et al. 2005; Konturek et al. 2007), are regulated by its interaction with retinoid orphan receptors/retinoid Z receptors (RZR/ROR), a family of nuclear receptors (Cutando et al. 2011). Most of the biological functions of melatonin are thought to be mediated by the membrane receptors, melatonin 1 (MT1) and 2 (MT2). In addition to their presence in the central nervous system, melatonin receptors are found in several peripheral structures, such as duodenal enterocytes (Sjöblom & Flemström, 2003, 2004; Slominski et al. 2012), gall-bladder epithelium (Aust et al. 2004), the exocrine pancreas (Aust et al. 2008), the endocrine pancreas (Mulder et al. 2009), the skin (Slominski et al. 2005, 2012) and the parotid gland (Cevik Aras & Ekström, 2008). In humans, using light microscopy, the presence of MT1-receptor immunoreactivity was reported in parotid glands (Aneiros-Fernandez et al. 2013; Arias-Santiago et al. 2012) but not in submandibular glands (Shimozuma et al. 2011). Both MT1 and MT2 receptors have been demonstrated by immunoblotting in rat parotid glands (Cevik Aras & Ekström, 2008). The MT2 type mediates protein synthesis, as well as the secretion of protein/amylase (Cevik Aras & Ekström, 2008; Cevik-Aras et al. 2011). Interestingly, melatonin was shown to induce in vitro ultrastructural changes in a preliminary report, indicating secretory activity, in pieces of human parotid gland (Loy et al. 2010).

Despite the numerous immunohistochemical studies emphasising the wide expression of melatonin receptors (Hardeland et al. 2011), the subcellular localisation of these receptors still remains to be elucidated. In fact, no such investigation has been performed using electron transmission microscopy. This is therefore the aim of our current study using the immunogold post-embedding technique, with the emphasis on human salivary glands. Light microscopy findings were related to those made by electron microscopy.

Materials and methods

Fragments of parotid glands were obtained from eight patients, five males and three females, aged 40–65 years, undergoing surgery for the removal of tumours of the neck region at the Otorhinolaryngology Clinic, University of Cagliari. Samples were collected at about 11:00 h. Informed consent was obtained in each case and all procedures were approved by the local institutional committee for human experimentation at the ASL 8 (Azienda Sanitaria Locale 8), Cagliari. The patients were not habitual smokers, alcohol consumers or obese, nor were they affected by autoimmune diseases or subjected to chemotherapy or radiotherapy before surgery.

Tissue preparation

Light microscopy

Samples of human parotid glands were fixed with 4% paraformaldehyde in 0.1 m phosphate-buffered saline, (PBS), for 2 h at 4 °C. They were then rinsed and stored at 4 °C in PBS containing 70 g L⁻¹ sucrose. The samples were frozen and cut into 10-μm-thick sections on a cryostat. The sections were mounted on slides coated with poly-l-lysine and stored at −80 °C. For the demonstration of melatonin receptors using the immunoperoxidase method, slides were pretreated by adding 3% H₂O₂ for 20 min in order to inhibit endogenous peroxidases. The optimal concentration and time of exposure to H₂O₂ were determined after several attempts, before no background staining was obtained. To prevent the non-specific binding of the secondary antibody, 3% normal goat serum (NGS) in PBS with 0.1% Triton X-100 (PBS-T) was added to the sections for 30 min. The sections were incubated with a rabbit polyclonal antibody specific for MT1 or a rabbit polyclonal antibody specific for MT2 (Abcam, Cambridge Science Park, Cambridge, UK) (1 : 125 or 1 : 250), followed for 48 h at 4 °C. Both MT1 and MT2 rabbit polyclonal antibodies are directed against the third cytoplasmic loop of receptors. Labelling was revealed with a goat anti-rabbit secondary antibody conjugated to horseradish peroxidase (1 : 200 in PBS-T with 3% NGS). For contrasting nuclei, sections were stained with Mayer's haematoxylin for 1 min. Slides were studied and photographed with a Leica DMR microscope equipped with a CCD Camera (Leica DC 300). Controls were obtained by incubating with non-immune rabbit serum or by omitting the primary antibody.

Transmission electron microscopy

Samples were cut into small pieces and fixed for 2 h with a mixture of 3% paraformaldehyde and 1% glutaraldehyde in 0.1 m cacodylate buffer. They were then rinsed in cacodylate buffer with the addition of 3.5% sucrose, dehydrated and embedded in Epon Resin (Glycide Ether 100; Merck, Darmstadt, Germany). Osmium tetroxide is omitted in the preparation of immunocytochemical samples in order to preserve the antigenicity of the tissue. Semi-thin sections (2 μm) stained with toluidine blue were examined by light microscopy to check the histological appearance. Ultra-thin sections (80 nm) were collected on nickel grids and processed for the immunohistochemical analysis. The grids were treated with 1% bovine serum albumin (BSA) and 5% NGS in PBS solution to block non-specific binding. They were then incubated overnight at 4 °C with a rabbit polyclonal antibody specific for either MT1 or MT2 (abcam, Cambridge Science Park, Cambridge, UK), diluted 1 : 50 in 1% BSA and 5% NGS. The grids were then incubated for 1 h at room temperature with the secondary antiserum, a goat anti-IgG rabbit antibody conjugated to 15-nm gold particles (GE Healthcare, Milan), diluted 1 : 50 in 1% BSA-PBS. After washing, they were stained with uranyl acetate and bismuth subnitrate and finally observed and photographed in a JEOL 100S transmission electron microscope. Controls were incubated with a non-immune rabbit serum or omitting the primary antibody.

Results

The experiments carried out using light and electron microcopy showed unequivocally that all the examined parotid glands were highly reactive to melatonin receptor MT1 and weakly reactive to MT2. No appreciable differences between male and female samples were observed.

MT 1 reactivity

Using light microscopy, a diffuse cytoplasmic reactivity to MT1 was observed in acinar cells of the parotid gland. However, the intensity varied between the cells. The striated duct cells exhibited intense intracytoplasmic staining for MT1. It was difficult directly to associate the reactivity with the cytoplasmic membranes (Fig. 1B). Controls did not display any staining, which meant that both endogenous peroxidase reaction and non-specific antibody binding were absent (Fig. 1A).

Fig. 1.

MT1-receptor immunohistochemistry in the human parotid gland. (A) Staining was completely absent in control samples. (B) The diffuse localisation of MT1-receptor staining occurred in acinar cells (a) and in striated ducts (sd). The diffuse staining made it difficult specifically to locate the reactivity to the plasma membrane of acinar and duct cells. As can be seen from the image, the intensity of reactivity varied between the acinar cells. Scale bar: 10 μm.

Transmission electron microscopy enabled us to define the intracellular distribution of MT1 reactivity, as well as its relationship to the cytoplasmic membrane. In the acinar cells, it was principally associated with the secretory granules (Fig. 2A). Gold particles were also observed in small vesicles diffused throughout the cytoplasm and surrounding the secretory granules (Figs. 2A-C). The plasma membranes of acinar cells showed an evident reactivity, localised in particular in their basal and lateral portions (Fig. 2A,B). In intercalated duct cells, immunoreactivity was found in vesicles near the luminal and lateral plasma membranes. No reactivity was observed in the granules of this ductal portion (Fig. 3A). In striated duct cells, reactivity was detected in a few apical vesicles, as well as in basal folds of the cytoplasmic membrane and nearby vesicles (Fig. 3B,C). Occasionally, gold particles were seen on nuclei and close to mitochondria of both acinar and duct cells. None of the control samples exhibited positivity (Fig. 4).

Fig. 2.

Immunogold staining of MT1 in the human parotid gland. (A) In acinar cells, gold particles were principally associated with secretory granules (g). Basal membranes exhibited reactivity (asterisks), as did (B) lateral membranes (asterisks). (B,C) MT1-positive small vesicles were localised throughout the cytoplasm, particularly surrounding the secretory granules (arrows). As osmium tetroxide is omitted in the preparation of immunocytochemical samples, the membranes of positive vesicles are not always well contrasted. The variation in immunoreactivity intensity between the cells observed in light microscopy (Fig. 1) was not apparent at the electron microscopy level. Importantly, the reactivity represented by the gold particles (‘black dots’) differs by its uniformity with respect to shape, size and electron density from structures such as ribosomes and glycogen granules. Scale bar: 1 μm.

Fig. 3.

Immunogold staining of MT1 in ducts of the human parotid gland. (A) An intercalated duct cell (id) adjacent to an acinar cell: MT1 staining was observed in vesicles near the luminal membrane (arrows). In striated duct cells, apical membrane reactivity was fairly low, as was the number of positive vesicles (B), whereas marked labelling was observed in basolateral membranes and in some small vesicles (C). m, Mitochondria. Scale bar: 1 μm.

Fig. 4.

In control samples, MT1 staining was completely absent, as was the case for MT2 (not shown). The image shows a portion of acinar cells. Scale bar: 1 μm.

MT2 reactivity

The distribution of MT2 reflected that of MT1. Using light microscopy, staining for MT2 was weaker than for MT1 (data not shown). Transmission electron microscopy showed the MT2 antibody decorating the same subcellular structures as the MT1 antibody (Fig. 5A). With respect to the cytoplasmic cell membranes, MT2 reactivity occurred only occasionally (Fig. 5B).

Fig. 5.

Immunogold staining of MT2 in the human parotid gland. On the whole, the distribution pattern was the same for MT2 as for MT1, but the number of gold particles for MT2 appeared much smaller than for MT1. (A) The reactivity in a portion of an acinar cell and (B) in the plasma membrane. Scale bar: 1 μm.

Discussion

Our electron transmission images of the ultrastructural localisation of melatonin receptors in the human parotid gland appear to be the first on record. Immunoreactivity to the receptors was found in the parenchymal cells of this salivary gland both intracellularly and at the level of the cell membrane, although the reactivity was less intense at the latter site. In acinar cells, the electron transmission images enabled us to detect small positive vesicles diffused among the secretory granules, which themselves exhibited reactivity. Labelled vesicles were also observed in intercalated and striated duct elements, especially near the lumen. In all the examined samples, we found the same distribution pattern for both MT1 and MT2 receptors, but the reactivity for MT1 was higher than for MT2. Notably, tissues do not always express both MT1 and MT2 to the same extent. MT1 is preferentially expressed in the ovary, testis, mammary gland, retina, kidney and skin, whereas MT2 is preferentially expressed in the lung, heart and intestines, as well as in adipocytes and immune cells (Konturek et al. 2007; Pandi-Perumal et al. 2008). As melatonin receptors are thought to be activated by circulating melatonin (Vakkuri, 1985; Vakkuri et al. 1985), their staining might be expected primarily on cell membranes. In fact, both receptors are members of the seven transmembrane G protein-coupled receptor family (Dubocovich et al. 1998; Dubocovich & Markowska, 2005). Instead, our experiments revealed a chiefly cytoplasmic staining, in agreement with what was recently described by others (Aneiros-Fernandez et al. 2013; Arias-Santiago et al. 2012) at the light microscopy level. Although the fate of melatonin receptors remains to be elucidated in salivary glands, we hypothesise that the reactivity within vesicles reflects the immediate internalisation of melatonin upon its binding to the membrane receptors. The scarce membrane reactivity may therefore reflect the high turnover of receptor internalisation (Gerdin et al. 2003; Miller et al. 2007; Sethi et al. 2008). Previous studies propose that the membrane-binding sites for melatonin not only may act as an uptake system for the hormone but also might control the quantity of melatonin that is allowed to reach the intracellular compartments, such as nuclear, mitochondrial and cytosolic fractions (Lee & Pang, 1993; Acuñ-Castroviejo et al. 1994; Venegas et al. 2012). In this respect, the reactivity occasionally observed in the nuclei of both acinar and ductal cells might suggest that MT receptors play an active role in driving melatonin inside the nucleus. Interestingly, the work by Withyachumnarnkul et al. (1987) focuses on the possibility of the active accumulation of a melatonin metabolite in the nuclei fraction of the rat salivary gland.

The pattern of distribution of the MT1 receptor in the acinar cells differed from that in the intercalated duct cells. In the acinar cells, MT1-receptor staining was identified in the granules, whereas in the intercalated ducts, the granules lacked staining for the receptor. This difference in distribution may imply that the receptor plays various roles in the different parts of the gland.

The abundant staining of the vesicles in both acinar and ductal cells could indicate that, at these locations, the vesicular trafficking of the melatonin receptor is involved in the transport of the hormone, internalised at the cell base and released from the apex. This phenomenon of transcytosis has been described for both IgA and transferrin in acinar cells (Nashida et al. 2009; Xu et al. 2013). The transcytosis of melatonin bound to its receptor might parallel the synthesis of melatonin in ductal cells, as reported by Shimozuma et al. (2011). In addition, in the acinar cells, the vesicles may fuse with the secretory granules to discharge melatonin by the major regulated secretion into the lumen of the glands (Castle & Castle, 1996). The two pathways may co-exist (Xu et al. 2013). We assume that, once secreted apically, melatonin is detached from its receptor and becomes a component of the saliva (Kennaway & Voultsios, 1998; Cutando et al. 2007), while the receptor is immediately re-absorbed. In keeping with this, the apical vesicles in ductal cells should be both absorptive and secretory; however, the two kinds cannot be distinguished using a morphological approach. To conclude, melatonin originating from the circulation may reach the gland lumina through active transport involving the formation of a complex with its receptors.