Abstract
Epidermal growth factor (EGF) is secreted into sweat from secretory cells of human sweat glands. The function of EGF in sweat is poorly understood. The biological function of EGF is exerted by the binding of EGF to the receptor (EGFR) and its activation. Therefore, we immunohistochemically localized the activated form of EGFR in human eccrine and apocrine sweat glands to assess the functional importance of the EGF–EGFR system in human sweat glands. Frozen sections of human skin were stained with a monoclonal antibody (MAb) specific for tyrosine-phosphorylated (activated) EGFR and with an MAb that stains both activated and non-activated EGFR. In the secretory portion of eccrine sweat glands, nuclei of the secretory cells were stained with the anti-activated EGFR MAb. In coiled and straight portions of eccrine sweat ducts, nuclei of luminal and peripheral cells were stained with the antibody specific for activated EGFR. Luminal cell membranes and luminal cytoplasm of inner ductal cells possessed non-activated EGFR. In the secretory portion of apocrine sweat glands, activated EGFRs were present in cytoplasm and nuclei of secretory cells. These data suggest that EGF, already known to be present in the cytoplasm of secretory cells in eccrine and apocrine sweat glands, activates EGFR in the nuclei of secretory cells themselves in an intracrine manner. Because ductal cells do not express EGF, EGF in the sweat secreted from the secretory cells should activate EGFR in the ductal cells in a paracrine manner. (
Keywords
P
EGF elicits various biological responses by binding and activating EGFR. On ligand binding, EGFR becomes autophosphorylated and it phosphorylates other endogenous proteins (Prigent and Lemoine 1992; Boonstra et al. 1995; Earp et al. 1995). The functional significance of EGF-EGFR system in sweat gland is poorly understood (Nanney et al. 1984; Pierard–Franchimont et al. 1991; Saga and Takahashi 1992). It is still not clear whether EGFR in human sweat glands is activated or non-activated. If EGFR is activated, the EGF–EGFR system must be functioning in human sweat glands. Therefore, we studied the immunohistochemical localization of activated EGFR in human eccrine and apocrine sweat glands using a monoclonal antibody (MAb) that reacts only with the tyrosinephosphorylated (activated) EGFR.
Materials and Methods
Antibodies
EGF was stained with anti-EGF MAb AM 26 from Ohtsuka (Tokushima, Japan) diluted to 45 μg/ml. This MAb was raised against recombinant human EGF (manufacturer's data sheet). Anti-activated EGFR MAb and anti-EGFR MAb were obtained from Transduction Laboratories (Lexington, KY). Anti-activated EGFR MAb (clone 74) was raised against tyrosine-phosphorylated EGFR isolated from EGF-challenged mouse L-cells expressing the human EGFR. This MAb reacts with 180-kD EGF-challenged EGFR. However, it does not react with non-stimulated EGFR (manufacturer's data sheet). Anti-EGFR MAb (clone 13) was raised against a synthetic peptide corresponding to residues 996–1022 of human EGFR. This MAb reacts with both EGF-challenged and EGF-non-challenged mouse L-cells expressing the human EGFR (manufacturer's data sheet). These antibodies were used at the dilution of 10 μg/ml.
The positive site with anti-activated EGFR indicated the presence of activated EGFR. If anti-EGFR showed positive staining and anti-activated EGFR showed negative staining, non-activated EGFR should be present. However it was impossible to determine the presence of non-activated EGFR in the sites at which both anti-activated EGFR and anti-EGFR antibodies showed positive staining because the antibody specific to non-activated EGFR was not available.
Tissue Processing
Small pieces of human skin were obtained from the surgical margin of various non-inflammatory skin lesions. These tissues were mounted in OCT compound and quickly frozen by immersion in liquid nitrogen. Specimens were stored at −70C until use. Five-μm sections were cut with a cryostat and fixed in cold acetone for 10 min, then air-dried.
Immunohistochemistry
Nonspecific immunoglobulin binding sites were inhibited by preincubating the sections with ovalbumin in PBS for 10 min. After blotting the blocking solution, sections were incubated with the first antibody for 30 min at room temperature. After washing with three changes of PBS (5 min each), sections were then incubated with FITC-labeled anti-mouse IgG (Vector Laboratories; Burlingame, CA) for 30 min, followed by washing with PBS. Nuclei were counterstained by incubation with propidium iodide (50 μg/ml in PBS) for 30 min, then washed with PBS. Sections were mounted with Perma Fluor (Shandon; Pittsburgh, PA) aqueous mounting medium then sealed with a coverslip. Antibody binding sites were visualized by observing with a Olympus BX 50 microscope equipped with Olympus Fluoview confocal laser scanning and analysis unit (Olympus; Tokyo, Japan). Color green and color red indicate where FITC and propidium iodide are localized, respectively. Color yellow to orange indicates nuclei are stained with FITC.
Results
Localization of EGF
Distribution of EGF in eccrine and apocrine sweat glands was studied by immunohistochemical staining with an MAb specific for EGF. The cytoplasm of secretory cells in eccrine and apocrine sweat glands showed positive staining with the anti-EGF MAb (data not shown), although nuclei of secretory cells and of ductal cells were not stained with the anti-EGF MAb as reported previously (Saga and Takahashi 1992). The lumen of the dermal duct of eccrine sweat glands was occasionally stained with anti-EGF MAb (Figure 1). Anti-EGF positive granules were attached to the luminal membranes of the duct.
Localization of EGFR in Eccrine Sweat Glands
Nuclei of secretory cells and those of ductal cells in eccrine sweat glands (Figure 2) were stained with the MAb that specifically binds to the activated form of EGFR. Luminal cell membranes and luminal cytoplasm of inner ductal cells in eccrine sweat glands were positively stained with anti-EGFR MAb (Figure 3), although they were not stained with the anti-activated EGFR MAb. Other areas of cytoplasm in ductal cells and cytoplasm of eccrine secretory cells were not stained with the anti-activated EGFR antibody. Myoepithelial cells were not stained with the anti-EGFR antibody or with the anti-activated EGFR MAb.
Localization of EGFR in Apocrine Sweat Glands
The anti-activated EGFR antibody stained cytoplasm and nuclei of the apocrine secretory cells (Figure 4) Some nuclei of secretory cells in apocrine sweat glands showed weak staining. The anti-EGFR antibody showed the same pattern of the staining as the antibody against activated EGFR. Myoepithelial cells were not stained with the anti-EGFR antibody nor with the anti-activated EGFR MAb.
Discussion
Confocal laser scanning microscopic observation of the staining with the anti-EGF MAb was consistent with previous reports (Nanney et al. 1984; Saga and Takahashi 1992). Cytoplasm of eccrine and apocrine sweat glands stained positively with the anti-EGF MAb. Ductal cells of eccrine sweat glands were not stained with the anti-EGF MAb. However, the ductal lumen was occasionally stained with the anti-EGF antibody. Substance positively stained with the anti-EGF MAb seemed to be bound to the luminal surface of ductal lumen (Figure 1). These results indicate that the secretory cells of eccrine and apocrine sweat glands secrete EGF and that the EGF goes through the sweat duct, where it binds to the EGF receptors on the luminal cell membranes of ductal cells (Figure 3).
EGFR is a 180-kD tyrosine kinase that mediates signals of EGF and TGF-α to the cells. On binding of ligands, the kinase is activated and phosphorylates the receptor's carboxyl terminus on tyrosine residues. Autophosphorylated tyrosine kinase phosphorylates substrates both bound to and not bound to the receptor. The conformational change induced to the substrates by tyrosine phosphorylation and/or binding to the receptor activates these substrates. These substrates then stimulate a series of downstream signaling pathways (Earp et al. 1995). The activated receptor–ligand complex is endocytosed and degraded within the lysosomes or the receptor is recycled to the cell membrane (Prigent and Lemoine 1992). In this mode of activation of EGFR, EGFR resides on cell membranes and in lysosomes. However, they should not be in the nuclei.
The anti-EGFR MAb showed positive reaction on the nuclei of sweat glands in our observations. Therefore, a different mode of activation from the one described in the previous paragraph must be working. A number of reports have indicated the presence of EGF (Savion et al. 1981) and/or its receptor in the nucleus of target cells (Boonstra et al. 1995). Extracellular EGF was transported to the nucleus and it bound to chromatin of a human colorectal cell line (Rakowicz–Szulczynska et al. 1986). This binding was inhibited by an antibody against EGFR. The function of nuclear localized EGFR in sweat glands is uncertain. It has been reported that several components known from growth factor-induced signal transduction, such as phosphoinositide kinases, phospholipase C (Payrastre et al. 1992), and protein kinase C (Leach et al. 1992), appear to reside in the nucleus. EGF induced translocation of its phosphorylated receptor in the nucleus of a cell line of squamous cell carcinoma rather than phosphorylation of pre-existent EGFR in the nucleus (Holt et al. 1994). EGF treatment caused a large reduction (80%) in plasma membrane EGFR immunolabeling when compared with untreated cells. This was paralleled by a fourfold increase in the nuclear content of EGFR. Immunoelectron microscopy showed that EGFR was associated with chromatin and, to a lesser extent, with the inner aspect of nuclear membranes. Ichii et al. (1988) reported that the purified nuclear fraction of rat liver showed binding activity of EGF. EGFR translocated to the nucleus provides a mechanism whereby gene activation may be regulated by this growth factor. This may be achieved directly by binding of EGFR to chromatin or indirectly by tyrosine phosphorylation of other chromatin-associated proteins (Holt et al. 1994). It is not limited to the EGF-EGFR system in which binding of an agonist to a cell surface receptor exerts an effect on events in the nucleus (Laduron 1992). In IIC9 cells, a subclone of Chinese hamster fibroblasts, stimulation with α-thrombin caused a massive intranuclear increase of diacylglycerol and protein kinase C (Leach et al. 1992).
Immunohistochemical localization of EGF in ducts of human eccrine sweat glands. Arrows indicate the ducts. Bar = 50 μm.
Secretory portion and ductal portion of eccrine sweat glands were stained with anti-activated EGFR monoclonal antibody. Thin arrows surround the secretory portion of an eccrine sweat gland. Thick arrows indicate the ducts. Bar = 50 μm.
Ductal portions of eccrine sweat glands were stained with the monoclonal antibody that stains both activated and non-activated EGFR. Arrows indicate the ducts. Bar = 50 μm.
Apocrine sweat glands were stained with anti-activated EGFR monoclonal antibody. Bar = 100 μm.
Because most of these studies were carried out using cultured cells, little is known about nuclear EGFR in vivo. An immunohistochemical study showed that EGFR was localized in the nuclei of ductal cells of eccrine sweat glands (Pierard–Franchimont et al. 1991), although it was not certain whether or not the EGFR was activated. Our immunohistochemical study demonstrated that EGFR in the nuclei of eccrine ductal cells were in activated form and therefore were functioning. Because ductal cells were not stained with anti-EGF MAb, EGFR must have been activated by EGF in sweat that was secreted from the secretory cells into the secretory lumen. Supporting this notion, EGF was localized in the lumen of ducts in eccrine sweat gland, as shown in Figure 1. Furthermore, if EGFR was activated by the ligand in blood, EGFR should have been localized on the basolateral cell membranes of ductal cells (Orsini et al. 1993). However, that was not observed in the ducts of eccrine sweat glands. Explanations for activated EGFR in the nucleus of ductal cells could be as follows: EGF in sweat stimulates the EGFR on the luminal cell membranes of ductal cells. Then activated EGFR is transported to the nucleus of ductal cells and it controls gene expression in the nuclei. Yet another possibility is that EGF in the sweat is endocytosed into the cytoplasm of ductal cells and then transported to the nuclei of ductal cells. Then EGF stimulates pre-existing EGFR in the nucleus. We speculate that the former explanation is more likely, because luminal cell membranes and luminal cytoplasm of ductal cells showed the presence of non-activated EGFR (Figure 3) and EGF was not detected in the nucleus or in the cytoplasm of ductal cells. Negative staining with anti-activated EGFR (Figure 2) and positive staining with anti-EGFR MAbs (Figure 3) on the luminal cell membranes and the luminal cytoplasm of eccrine sweat ducts indicate the presence of non-activated EGFR there.
Nuclei of secretory cells in eccrine and apocrine sweat glands stained positively with anti-EGFR MAb specific for the activated form. Because the cytoplasm of eccrine and apocrine secretory cells contains EGF, it would activate EGFR in the nuclei of secretory cells in an intracrine manner (Pittelkow et al. 1991), i.e., cytoplasmic EGF would bind and activate intracellular EGFR without being secreted extracellulary. Therefore, we conclude that at least a part of the function of EGF in secretory cells and sweat is to control the function and the proliferation of secretory cells and ductal cells in human sweat glands by activating EGFR in the nuclei in a intracrine and paracrine manner, respectively.
Footnotes
Acknowledgements
Acknowledgments
Supported in part by a Grant-in-Aid for Scientific Research (08670978) from The Ministry of Education, Science and Culture (KS).