Abstract
We studied monoamine oxidase (MAO) A and B localization in human pancreas, thyroid gland, and adrenal gland by immunohistochemistry. The primary antibodies used were mouse monoclonal anti-human MAO-A (6G11/E1) and anti-human MAO-B (3F12/ G10/2E3). Samples were obtained from six routine autopsy cases and fixed in 2% paraformaldehyde. Exocrine pancreas showed a widespread distribution of MAO-A, whereas MAO-B was present only in centroacinar cells and epithelial cells of pancreatic ducts. In endocrine pancreas, MAO-A was observed in around 50% of islet cells, whereas MAO-B was less abundant and was restricted to the periphery of islets. Thyroid gland showed strong MAO-A immunoreactivity in all cell types and was MAO-B-negative. In adrenal gland, the capsule displayed MAO-A but not MAO-B immunoreactivity, whereas the cortex showed widespread MAO-A staining but was MAO-B-negative in interstitial cells. Finally, in the medulla only a few scattered cells showed either MAO-A or MAO-B immunoreactivity. To our knowledge, these data represent the first study of the cellular distribution of MAO-A and MAO-B in the three human tissues included.
M
In an attempt to understand the role of MAO in the brain, many studies have investigated the localization of MAO-A and MAO-B in the human brain. In contrast, little is known about their distribution in peripheral tissues, especially in humans. In a recent paper we described the distribution of MAO-A and MAO-B in six human peripheral tissues by means of quantitative autoradiography (Saura et al. 1996). In the present study, specific anti-human MAO-A and MAO-B monoclonal antibodies (MAbs) were used to immunostain cryosections of pancreas, thyroid, and adrenal gland to describe the distribution and, most importantly, the cellular location of both isoenzymes in these glands for the first time.
Samples of pancreas, adrenal gland, and thyroid gland were generously provided by Dr. J.A. Bombi (Departament de Biologia Cel·lular i Anatomia Patològica, Facultat de Medicina, Universitat de Barcelona, Spain) and were obtained from six routine autopsy cases (three men and three women) and taken in accordance with European guidelines. Tissues affected by the pathological state of the patient were discarded. The age range was 50-84 years (mean ± SEM 73.6 ± 4.9) and the postmortem delay ranged from 4 to 22 hr (mean ± SEM 12.3 ± 2.5).
Immediately after autopsy, tissues were fixed by immersion in 2% (w/v) paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) at 4C for 4 hr. After fixation, tissues were immersed for 72 hr in 15% (w/v) sucrose in 0.1 M phosphate buffer for cryoprotection, frozen with dry ice, and stored at −30C.
Cryostat sections (12 μm) were obtained and mounted on gelatinized slides. The immunohistochemical peroxidase-anti-peroxidase (PAP) technique was used. Endogenous peroxidase activity was inhibited by a 30-min preincubation in H2O2-methanol-PBS (0.01 M phosphate buffered saline) (0.3/9.7/90) followed by a 10-min wash in PBS. As primary antibodies for detection of MAO-A and MAO-B protein, the mouse monoclonals anti-human MAO-A (6G11/E1) (see Figure 1) and anti-human MAO-B (3F12/G10/2E3) (Yeomanson and Billett 1992) were used diluted 1:50 and 1:75, respectively, in PBS-5% normal swine serum. Incubation was performed at 4C for 24 hr. The antibody 9H7 (an anti-γ-gliadin, which has been typed as subclass IgG1) was diluted 1:50 in normal swine serum and used to define the nonspecific staining. After washing, sections were incubated for 30 min in rabbit anti-mouse IgG (1:100), rinsed and washed in PBS, and incubated for 30 min in mouse IgG-PAP (1:100). Sections were developed for 15 min in a 0.05 M Tris solution containing 0.03% diaminobenzidine and 0.006% H2O2. Some sections were counter-stained with Mayer's hematoxylin. All antibodies except the primary ones were purchased from Sigma (St Louis, MO).
In the exocrine pancreas, strong MAO-A immunoreactivity (ir) was seen in all acinar and centroacinar cells, in the intercalated ducts, and in the epithelial cells of pancreatic ducts. The walls of pancreatic ducts and the connective tissue showed no MAO-A ir (Figures 2A and 2B). In contrast, acinar cells were devoid of MAO-B ir. Centroacinar cells showed punctate MAO-B staining, and both intercalated ducts and epithelial cells of pancreatic ducts were enriched in MAO-B ir. Connective tissue was devoid of MAO-B ir (Figure 2C).
In the endocrine pancreas, both MAO-A and MAO-B ir was present but in different cell populations (Figures 2D–2F). MAO-A-positive cells were homogeneously distributed in the islets. They accounted for approximately 50% of the islet cells and probably correspond to β-cells. MAO-B-positive cells were less abundant, appeared in the periphery of the islets, and may correspond to α-cells.
In the thyroid gland, all follicular and parafollicular cells showed strong MAO-A immunoreactivity but were MAO-B-negative (Figures 2G–2I).
Specificity of anti-MAO-A antibody: Western blot analysis. Human liver mitochondrial membranes (Billett and Mayer 1996) were fractionated by SDS-PAGE (10% w/v) and electroblotted onto nitrocellulose. One lane was stained with amido black (Lane M). Other lanes were blocked with 3% (w/v) Marvel milk powder in Tris-buffered saline (TBS), washed, and incubated overnight at 25C with culture supernatants of hybridomas raised to human MAOs 5A11 (Lane 1) and 6G11/E1 (Lane 2) in the presence of 0.1% (w/v) Tween-20 (Yeomanson 1990). Unbound antibody was revealed by extensive washing in TBS/Tween and bound antibody was revealed with an alkaline phosphatase-conjugated goat anti-mouse antibody (Dako, Carpinteria, CA; diluted X1000 in blocking buffer), using 5-bromo-4-chloro-indolyl phosphate (enhanced with nitroblue tetrazolium) as substrate. Molecular weight markers (standards) were also included (stained with amido black; Lane S, ranging from 29 kD to 116 kD). It is clear that 5A11 bound to two bands (apparent Mr 63,000 and 59,000, MAO-A and MAO-B, respectively), whereas 6G11/E1 bound to MAO-A only.
In the adrenal gland, the capsule showed MAO-A but not MAO-B ir (Figures 2J–2L). Positive staining was seen for both isoenzymes in the secretory cells of the cortex, but only interstitial cells displayed MAO-A ir. The only MAO-A immunoreactivity in the medulla was observed in the walls of the central vein and in scattered cells. The pattern of MAO-B ir in the medulla was similar to that of MAO-A (Figure 2L).
In this article we report the cellular distribution of MAO-A and MAO-B in human pancreas, thyroid, and adrenal glands by immunohistochemistry. To our knowledge, this is the first study that characterizes the cellular distribution of MAO isoenzymes in these tissues.
Acinar cells and centroacinar/duct cells participate in the production of pancreatic juice from the exocrine pancreas. We found MAO-A in both cell populations but MAO-B only in the latter. Because the sympathetic innervation of the exocrine pancreas is mainly if not entirely vascular, sympathetic norepinephrine is unlikely to be a MAO substrate in these cells. On the basis of rat experimental data, circulating mono-amines, especially epinephrine, dopamine and 5-HT are the most likely MAO substrates in these cells.
MAO-A, 9H7, and MAO-B staining in adjacent sections of pancreas, thyroid gland, and adrenal gland. (
In the endocrine pancreas, we have shown a different cellular distribution for the two isoenzymes. MAO-A is localized mainly in β-cells, whereas MAO-B appears to be restricted to α-cells. Of the three tissues included in this study, this is the only example of a cell type expressing MAO-B but not MAO-A, a feature also shared by platelets.
Both circulating epinephrine and sympathetic norepinephrine stimulate glucagon secretion and inhibit insulin secretion, and both are putative substrates for islet MAOs. However the pattern of distribution of MAO-A does not match that of norepinephrine found in all islet cells (Lloyd et al. 1986) 5-HT is another possible substrate for islet MAO. Because of multiple species-related differences regarding 5-HT and the endocrine pancreas (e.g., Cetin 1992), more precise studies of the cellular localization of 5-HT in human islets are needed to ascertain the role of MAOs. If found in human β-cells, 5-HT could be a physiological substrate for MAO-A. If found in α-cells, the role of MAO-B may be to prevent the accumulation of other amines that might interfere with 5-HT, as suggested for MAO-B in serotonergic neurons.
There are two types of parenchymal cells in the thyroid gland: the follicular cell, which synthesizes and secretes thyroxin (T4) and triiodothyronine (T3), and the parafollicular cell, which synthesizes and secretes calcitonin. Previous studies have shown that MAO-A is the predominant MAO isoform in the human thyroid gland (more than 90% of total MAO activity) (Cabanillas et al. 1989). Our study confirms these observations but, most significantly, it shows that both follicular and parafollicular cells express MAO-A.
The role of MAO-A in follicular cells is unclear. It has been suggested that it may supply the H2O2 required in thyroid hormone synthesis (Masini-Repiso and Coleoni 1981). However, studies with cultured human thyroid cells have raised doubts about this hypothesis.
The detection of MAO-A but not MAO-B in parafollicular cells is a new finding. It is consistent with the reported presence of MAO-A, but not MAO-B, in human medullary thyroid carcinoma (MTC) cells, a cell line derived from a carcinoma of parafollicular cells. 5-HT has also been proposed as an MAO-A substrate in parafollicular cells, in which it may stimulate calcitonin secretion. However, an inhibitory effect has been proposed for intracellular 5-HT (Zabel 1985).
The adrenal gland is the main producer of hormonal catecholamines. However, we found neither MAO-A nor MAO-B ir in adrenal medullary cells. Nevertheless, we found MAO-A ir in the central vein and ganglion cells, which agrees with the findings of Carmichael and Pfeiffer (1987). These authors showed no evidence of MAO activity in human adrenal medullary cells. These combined data suggest that there are no uptake mechanisms for epinephrine and norepinephrine and that their inactivation takes place in the target organs.
In the adrenal cortex we found a more widespread distribution, with both MAO-A and -B ir present in all secretory cells. This is also the case in rat tissue, in which both enzymes have specific functions. Thus, the autocrine effect of dopamine synthesized from circulating L-dopa could be regulated by MAO-B, whereas MAO-A may metabolize the neuronal norepinephrine that modulates cortical endocrine release (Gilchrist and Charlton 1993).
In conclusion, this study offers the first comprehensive results showing that MAO-A and MAO-B are localized at the cellular level in these human peripheral tissues. They show a widespread localization of both enzymes, especially MAO-A, and frequent co-localization. These observations are in marked contrast to the pattern in brain, in which most cells are devoid of both isoenzymes (e.g., oligodendrocytes, microglia, and most neurons) and co-localization of MAO-A and MAO-B is extremely rare. The physiological role of the enzymes in pancreas, thyroid gland, and adrenal gland is far from being understood, but these results suggest that specific uptake or synthesis of monoamines occurs in cell types in which MAO is present. Further experiments are required to determine the physiological substrates for MAO-A and MAO-B in the different cell types.
Footnotes
Acknowledgements
Supported by the Wellcome Trust and FISss 94/1461.