Expression of the Giant Protein AHNAK (Desmoyokin) in Muscle and Lining Epithelial Cells

Abstract

Here we report a detailed analysis of the expression and localization of the giant protein AHNAK in adult mouse tissues. We show that AHNAK is widely expressed in muscle cells, including cardiomyocytes, smooth muscle cells, skeletal muscle, myoepithelium, and myofibroblasts. AHNAK is also specifically expressed in epithelial cells of most lining epithelium, but is absent in epithelium with more specialized secretory or absorptive functions. In all adult tissues, the main localization of AHNAK is at the plasma membrane. A role for AHNAK in the specific organization and the structural support of the plasma membrane common to muscle and lining epithelium is discussed.

Keywords

muscle AHNAK epithelium tissue localization

The giant protein ahnak has been identified by Shtivelman et al. (1992) as a gene whose transcription is repressed in cell lines derived from neuroblastomas and other tumors. Hieda and Tsukita (1989) identified the mouse homologue of AHNAK (desmoyokin) as a plasma membrane-associated protein localized to the desmosomal attachment plaques in stratified epithelium. AHNAK is unusual in the sense that it is a very large protein of approximately 700 kD whose gene is composed of a single exon located on human chromosome 11q12 (Kudoh et al. 1995). Moreover, AHNAK is characterized by a large internal domain (4300 amino acids) composed of highly conserved repeated elements, most of which are 128 amino acids in length. This internal domain is flanked by a short NH2 terminus and a COOH terminus of 1000 amino acids. The COOH terminus of AHNAK contains nuclear localization sequences and a nuclear export signal (Shtivelman et al. 1992; Sussman et al. 2001). In various cell lines, including NIH 3T3 fibroblasts, MDCK cells, and HeLa cells, AHNAK is located principally (but not exclusively) in the nucleus and is phosphorylated on serine and threonine residues (Shtivelman and Bishop 1993). PKB phosphorylation of AHNAK prevents its nuclear accumulation in HeLa and MDCK cells (Sussman et al. 2001). In contrast, in tissues such as skin and heart, AHNAK has an exclusive plasma membrane localization (Hashimoto et al. 1993; Haase et al. 1999). The targeting of AHNAK to the plasma membrane is dependent on PKC activation in epithelial DJM-1 cells (Hashimoto et al. 1995) and is regulated by cell density in human MDCK cells (Sussman et al. 2001). The cellular functions of AHNAK are unknown, but several studies have implicated AHNAK in signal transduction pathways. In vitro, AHNAK binds and activates PLCγ in the presence of arachidonic acid (Sekiya et al. 1999). In cardiomyocytes, AHNAK associates with the β-subunit of cardiac L-type calcium channels and is phosphorylated by PKA in response to adrenoreceptor stimulation (Haase et al. 1999). Moreover, we have shown that AHNAK interacts with the EF-hand calcium and zinc binding S100B protein (Gentil et al. 2001), which is abundantly expressed in the central nervous system, skin, and muscle (Boni et al. 1997; Shrestha et al. 1998). Although studies on the cellular and sub-cellular localization of AHNAK in cultured cell lines are available in the literature, its tissue localization has been poorly investigated. Recently, a study on the expression of AHNAK in embryonic tissues revealed a tissue-specific expression of AHNAK (Kingsley et al. 2001). At E8.5, AHNAK is expressed in extraembryonic tissue, in migratory mesenchymal cells, in mesenchymal cells lining the branchial artery, and in nasal epithelia and skin. In these cells, AHNAK immunoreactivity was essentially found associated with plasma membrane and cytoplasm. In adult mammals, the expression and function of AHNAK have not yet been investigated. In this study we investigated AHNAK expression in adult mouse tissues.

Materials and Methods

Antibodies and Controls

Rabbit polyclonal antibodies were made against AHNAK using two different epitopes. The first, KIS (CKISMPDVDL HLKGPK), was chosen according to the strategy of Shtivelman and Bishop (1993). The second epitope, CQL (CQLPE VELSVSTKKE), is localized in the C-terminus of the protein (aa 5630–5643). The KIS-AHNAK and CQL-AHNAK antibodies used throughout our study are affinity-purified. Antibodies were purified from immununized rabbit sera by affinity chromatography using KIS or CQL peptide crosslinked to Sepharose beads. The purity of the immunoglobulins was controled by SDS-PAGE in the absence and in the presence of a reducing agent. The purified antibodies were stored in PBS and 50% glycerol at a concentration of 0.15 mg/ml. The specificity of the antibodies was tested in Western blotting on HeLa cell extracts and on the recombinant AHNAK protein. To control for antibody specificity, the KIS-AHNAK antibody and the CQL-AHNAK antibody were incubated with the corresponding KIS or CQL peptides before primary immunoreaction. No immunoreaction was detected using this method. For immunochemistry, the specificity of immunostaining was controlled by preincubation of the affinity-purified KIS-AHNAK antibody with the KIS peptide. For double immunofluorescence experiments, we used a monoclonal anti-PECAM, MEC13.3, from Sigma (St Louis, MO). Cyan3 and Alexa 488 antibodies used as second antibody were provided by Molecular Probes (Eugene, OR).

Immunoblotting

In adult mouse, specimens from kidney, heart, thymus, esophagus, pancreas, liver, and lung were removed. These specimens were cut into small pieces and quickly frozen with liquid nitrogen until used. Specimens were homogenized with a glass homogenizer and were then heated at 90C for 3 min in denaturation buffer composed of 2% SDS, 25 mM Tris-HCl, pH 6.8, 10% glycerol, 0.1% bromophenol blue, 1m M EDTA. Proteins were electrophoretically resolved by one-dimensional SDS-PAGE (6% polyacrylamide), the gel was then equilibrated in transfer buffer containing 0.03% SDS and proteins were transferred onto nitrocellulose membrane. The membrane was first incubated with CQL-AHNAK antibody (1:100 dilution) in TBS containing 0.3% Tween-20 (TTBS) at 4C for 16 hr. After washing, the membrane was incubated with a 1:10,000 dilution of horseradish peroxidase secondary antibody and the immune complex was revealed by chemiluminescence. Nitrocellulose membrane was then stripped in TTBS with 1% SDS and 0.1% sodium azide, and then reprobed with the KIS-AHNAK antibody (1:1000 dilution).

Immunochemistry

Tissues were screened for AHNAK localization using a tissue array produced by ResGen Co. (Invitrogen; Carlsbad, CA). We used paraffin-embedded tissue spots representing several mouse organs: adrenal, brain, heart, large intestine, small intestine, renal cortex, renal medulla, liver, lung, lymph node, skeletal muscle, pancreas, salivary gland, skin, spleen, stomach, thymus, prostate, testis, ovary, uterus, breast, and esophagus. Paraffin was removed and immunohistochemistry was then performed using the DAKO ABC horseradish peroxidase system (DAKO; Carpinteria, CA). After endogenous peroxidase inactivation with H₂O₂ 3% and sodium azide 0.05%, followed by extensive washing, tissues were blocked in 5% normal goat serum in TBS for 30 min and incubated with the KIS-AHNAK antibody at 1:100 dilution in TBS plus 2% normal goat serum overnight at 4C. After rinsing with TBS, tissues were incubated with a biotinylated goat anti-rabbit Ig (DAKO) diluted 1:100 in TBS plus 5% normal goat serum for 30 min at room temperature. After extensive washing with TBS, the slide was incubated with the horseradish peroxidase complex according to the manufacturer′s recommendations. The chromogen used was AEC+ (DAKO). Slides were counterstained with hematoxylin. For indirect immunofluorescence localization of AHNAK, mouse testis, brain, and liver were frozen in isopentane at −80C and cut in 10-μm sections. Slides were fixed with 4% paraformaldehyde for 30 min and washed abundantly before permeabilization for 5 min with 0.5% Triton X-100 in TBS. After extensive washing and blocking with goat serum, primary antibodies were incubated overnight at 4C at the following dilutions: KIS-AHNAK 1:100; PECAM 1:5. Slides were then incubated with secondary antibody for 1 hr and nuclear countstaining was performed with Hoechst 33258 (1 μg/ml). Double immunofluorescence analysis was performed using confocal microscopy (Leica TCS-SP2) or fluorescence microscopy (Zeiss Axiovert 200M). Images were directly captured, saved, and transferred to Adobe Photoshop 5.5.

Results

Immunoblotting Analysis and Tissue Distribution of AHNAK

Extracts from various mouse tissues were analyzed for AHNAK expression by immunoblotting using affinity-purified antibodies raised against the CQL and the KIS peptides (Figure 1). To normalize for total protein content of the tissue preparations, the extracts were also analyzed for α-tubulin content (Figure 1A). Both the CQL and the KIS antibody reacted strongly with a major protein band that migrated with the expected molecular weight for AHNAK, confirming the specificity of our antibodies (Figure 1B). A high level of expression of AHNAK was observed in heart, lung and esophagus, whereas kidney, pancreas, and liver expressed only moderate levels of AHNAK. AHNAK immunoreactivity was absent in the thymus. CQL and KIS antibodies also reacted with lower molecular weight protein bands. These protein bands are present only in tissue extracts that express the full-length AHNAK protein and may therefore represent proteolytic products of AHNAK. In support of that assumption, we found that AHNAK is highly sensitive to metal-dependent proteases. When tissues are lysed in the absence of the metal chelators EDTA and EGTA, the full-length AHNAK immunoreactivity decreases, with a concomitant increase in the lower molecular weight protein bands recognized by both the AHNAK KIS and CQL antibodies (data not shown). A band of approximately 100 kD was also detected with the CQL antibody but not with the KIS antibody. This band might correspond to a C-terminal AHNAK proteolytic fragment. The affinity-purified KIS antibody was next used for immunoperoxidase and indirect immunofluorescence analysis of AHNAK immunoreactivity in mouse tissues (Figures 2–5).

Figure 1

Immunoblotting analysis of AHNAK content in mouse organs or tissues. Tissue extracts were analyzed by Western blotting using α-tubulin (A), CQL, or KIS (B) antibody. Lane 1, heart; Lane 2, liver; Lane 3, lung; Lane 4, thymus; Lane 5, kidney; Lane 6, pancreas; Lane 7, esophagus. Asterisk indicates a 100-kD band that may correspond to a proteolylic fragment of the C-terminus of AHNAK.

AHNAK Localization in Muscle Cells

Immunohistochemical analysis of AHNAK immunoreactivity in different tissues revealed a specific localization of AHNAK in muscle cells (Figure 2). As previously reported by Haase et al. (1999), heart sections clearly show a plasma membrane localization of AHNAK in cardiomyocytes (Figure 2A). Figure 2B shows a muscle section incubated with affinity-purified KIS antibody that was preincubated with the KIS peptide and counterstained with biotinylated horseradish peroxidase. In this control, AHNAK staining was completely abolished, confirming our biochemical analysis that the KIS antibody used in our study is specific for AHNAK protein. AHNAK expression is not restricted to cardiomyocytes, but appears to be a general feature of the different muscular cell types. As observed in cardiomyocytes, in the skeletal myocytes forming the striated skeletal muscle fiber, AHNAK immunoreactivity also accumulated at the plasma membrane (Figure 2C). In the small intestine, AHNAK antibodies decorated the layers of smooth muscle cells of the tunica underlining the epithelium (Figure 2D). Both the longitudinal and the circular layer of smooth muscle cells showed strong AHNAK immunoreactivity. A cross-section of the circular layer clearly showed that AHNAK is localized at the plasma membrane. This pattern of AHNAK immunoreactivity in smooth muscle cells was also observed on the muscularis propria of the body of the stomach, which also consists of an inner circular and outer longitudinal layer (data not shown). Other structures composed of smooth muscle cells found in various tissue also express AHNAK (Figure 2E and 2F). In the ovary, theca cells showed strong plasma membrane AHNAK immunoreactivity (Figure 2E). Theca cells, which are important for steroid secretion, also have muscle characteristics (Motta and Familiari 1981; Self et al. 1988). At ovulation, these cells by their contraction may be capable of squeezing the follicular wall or may favor follicle collapse and subsequent involution during atresia. Smooth muscle cells that compose the tunica media and ensheathe large blood vessels in various tissue were also stained with AHNAK antibodies. In testis, smooth muscle cells surrounding small and large arteries were AHNAK-positive (Figure 2F). AHNAK was also expressed by PECAM-positive endothelial cells, as revealed by yellow pixels in merged confocal double-labeled fluorescent image (Figure 2F). We also observed apparently higher AHNAK immunoreactivity in endothelial cells of small arterioles than in large arteries. Slender smooth muscle-like cells surrounding seminiferous tubules also expressed AHNAK (Figure 2F). These cells are responsible for the production of collagen and elastin fibers present in the lamina propria, while their contractile activity helps the movement of spermatogonia along the tubule.

AHNAK Localization in Lining Epithelium

In the stratified epithelium of the skin, AHNAK was expressed at the plasma membrane of keratinocytes of the stratum basale and the stratum spinosum (Figure 3A). AHNAK staining was not observed in the cornified layer. AHNAK is also expressed in cells of papillary dermis and in some cells of the hair follicle (Figure 3A).

Figure 2

AHNAK localization in muscle tissue. (A) Heart section showing strong AHNAK immunoreactivity at the cardiomyocyte plasma membrane. (B) The specificity of AHNAK imunostaining of heart section was controlled by preincubation of the affinity-purified KIS-AHNAK antibody with KIS peptide. Bars = 20 μm. (C) Section of skeletal muscle shows strong membrane localization of AHNAK in the myofiber. (D) The mucosae of the gastrointestinal duct (here the small intestine) is surrounded by a layer of smooth muscle cells, the outer longitudinal (L), the central circular layer (C), and the muscularis mucosae (MM) bordering the mucosae are AHNAK-positive. Note in the circular layer of smooth muscle cells the membranous localization of AHNAK. (E) In the ovarian follicle, an oocyte (O) is seen with the surrounding zona granulosa (ZG) and the theca layer (TC). The layer of theca cells is AHNAK-positive. (F) Double immunofluorescence analysis of AHNAK (red) and PECAM (green) in a seminiferous tubule of testis cut in transverse section. The myoid cells forming a thin layer surrounding the tubule (arrow head), the smooth muscle cells surronding a small arteriol and an artery (arrows), and the endothelial cells are AHNAK-positive. Bars = 50 μm.

In the upper part of the digestive tract, i.e., the esophagus, AHNAK was found in the stratified squamous epithelium, which in rodents may be keratinized (Figure 3B). Here again, AHNAK immunoreactivity was concentrated mainly at the plasma membrane. As shown in Figure 3B, AHNAK was also present in the muscle cells of the striated muscularis layer. In kidney, AHNAK immunochemistry showed that AHNAK was localized in a special type of epithelium of the urinary system that lines the collecting ducts, ducts of Bellini, and the pelvicalyceal system (Figure 3D; and data not shown). This transitional epithelium, called urothelium, is present along the conducting passages of the urinary system and protects the organism from acid and hypertonic urine. In liver, AHNAK immunoreactivity was confined to the portal tract. Figure 3E focuses on a typical portal tract containing three main structures. The largest is a terminal branch of the hepatic portal vein (PV), which has a thin wall lined by endothelial cells labeled with PECAM antibody. These endothelial cells were AHNAK-positive. Smaller-diameter vessels are arterioles, which show lower AHNAK immunostaining. Finally, the bile collecting duct, lined by simple columnar epithelium, was strongly immunostained with AHNAK antibody. In these lining epithelial cells, AHNAK was concentrated at the plasma membrane. There was no AHNAK immunostaining of the liver hepatocytes. Immunohistochemistry with the lung (Figure 3E) confirmed the Western blotting analysis, showing that a high level of AHNAK protein is present in lung (Figure 1, Lane 3). AHNAK was highly expressed in parenchymal cells (epithelium of alveoli) and in the smooth muscle layer surrounding the bronchiolar epithelium. Faint AHNAK immunoreactivity was also present in pseudostratified bronchial epithelium. The labeled epithelial cells were mainly ciliated cells characterized by large nuclei. Their soma was weakly immunostained but concentrated immunoreaction deposit was observed at the apical plasma membrane and in contact with the basement membrane (white arrow).

Figure 3

AHNAK localization in lining epithelium. (A) Skin section. The stratified squamous keratinizing epithelium, but not the cornified layer of the epidermis (e), is AHNAK-positive. The junction between the epidermis and the dermis is characterized by downward folds of the epidermis called rete ridges, which interdigitate with upward projections of the dermis called dermal papillae (DP). Some cells of the dermis and a small layer of cells in the hair follicle complex are AHNAK-positive. (B) The lumen of the esophagus is lined by a thick AHNAK-positive protective stratified squamous epithelium (e) which, in this specimen, is keratinized. The underlying lamina propria is relatively condensed and the underlying muscularis mucosae is barely visible. Note the presence of skeletal muscle (M) in the muscularis, supporting the idea that the specimen is derived from the upper part of the esophagus. Bars = 50 μm. (C) Kidney section. AHNAK staining is located in an epithelium that may correspond to the collecting tubule (arrowhead). This renal specimen contains a glomerulus (arrow) which is not AHNAK labeled. Bar = 20 μm. (D) Liver section. Double immunofluorescence analysis of AHNAK (red) and PECAM (green) by confocal microscopy focuses on a typical portal tract containing a terminal branch of the hepatic portal vein (PV), an arteriole (arrowhead), and the bile ductule (arrow). Bar = 50 μm. (E) In lung, dense immunoreaction deposit is observed in a bronchiole at the apical plasma membrane of epithelial cells and in contact with the basement membrane (white arrow). The underlayer smooth muscle (M) and the lung parenchyma cells are heavily immunostained. Bar = 20 μm.

Figure 4

AHNAK is localized in the ventricular wall of adult brain. Double immunofluorescence analysis of AHNAK and HOECHST (DNA) labeling in the third ventricle (A) and the fourth ventricle (B) shows that AHNAK is localized in the ependymal cells that border the ventricles. AHNAK is also present in the brain parenchymal capillaries (arrowhead). Bar = 20 μm.

We also found AHNAK immunoreactivity associated with the plasma membrane of several other lining epithelia, such as the stratified epithelium of the tunica albuginea of the testis (data not shown), and also in the epithelial cells lining the ventricles of the central nervous system (Figure 4). In the adult mouse brain, AHNAK immunoreactivity was abundant in epithelial cells that bordered all the ventricles, including the lateral ventricles (not shown), the third ventricle (Figure 4A), and the fourth ventricle (Figure 4B). There was no apparent AHNAK immunostaining of neuronal and glial cell soma. Brain capillary endothelial cells were also immunostained (see Figure 4A).

In contrast to muscle cells, not all epithelial cells synthesize AHNAK. As shown in Figure 5A, the different secretory epithelial cells that constitute the stomach mucosa, including mucus-secreting cells, acid-secreting cells, and pepsin-secreting cells, were not AHNAK-immunoreactive. AHNAK staining appeared as a thin strand between the gastric gland, which corresponds to the muscularis mucosae that extends from the base of a gastric gland, to the lumen, whose contraction expels gastric secretion into the stomach lumen. The absence of AHNAK in secretory epithelium was confirmed in the salivary gland (Figure 5B), the mammary gland (Figure 5C), and the prostate gland (Figure 5D). In the salivary gland, the epithelium of serous acini, with nuclei flattened against the basement membrane, and the epithelium of mucous acini were not AHNAK-immunoreactive. The thin AHNAK immunoreactivity surrounding serous acini corresponds to the processes of contractile myoepithelial cells. In mammary glands, myoepithelial cells surrounding the gland and endothelial cells of capillaries were also labeled by the AHNAK antibody (Figure 5C). The prostate glandular epithelium was not positive for AHNAK, whereas the supporting tissue, which contains numerous smooth muscle cells and capillaries, was labeled by the AHNAK antibody (Figure 5D). Figure 5E shows the immunolabeling of AHNAK in the pancreas, a large gland composed of glandular epithelium with exocrine and endocrine functions. The endocrine tissue formed by the islets of Langerhans was not stained, nor were the glandular acini, which constitute the exocrine part of the pancreas. Only the endothelial cells of the capillaries were immunolabeled. AHNAK immunostaining was also not observed in epithelium with absorptive functions, such as the epithelium of the small intestine (data not shown; see also Figure 2D) and the absorptive epithelium lining the colon, which also has protective functions (Figure 5F). The mucosa of the colon, formed of straight tubular glands consisting of cells specialized for water absorption and of mucus-secreting globlet cells, was devoid of AHNAK immunoreactivity. However, significant cytoplasmic immunoreactivity was present in the cells of the luminal portion of the epithelium. These cells, which are derived from precursor cells present in the basal portion of the epithelium, are more mature and are subjected to the stretching forces of the luminal content. The muscularis mucosae, which extends into the lamina propria between the folds of the gland, also expressed AHNAK.

Discussion

Table 1 summarizes the tissue distribution and the cellular localization of AHNAK in mouse tissues. The protein AHNAK is a hallmark of muscular and muscle-like cells with contractile properties organized in contractile tissue or scattered in the tissue as fibromyocytes or myoepithelial cells. AHNAK is also specifically expressed in epithelial cells of the skin, the esophagus, the tunica albugina of the testis, the bile ducts, and the brain ventricles. In contrast to muscle cells, not all epithelial cells express AHNAK. We did not find significant AHNAK immunoreactivity in the glandular epithelium of the mammary gland, the salivary gland, the stomach, the prostate, or the exocrine and endocrine pancreas, all of which have secretory functions. We were also unable to detect AHNAK in the absorptive epithelium of the proximal part of the nephron and of the intestine. In addition to muscle cells and lining epithelial cells, AHNAK is also present in endothelial cells of several capillary and blood vessels (see Figures 2F, 3D, 4A, and 5C–5E). However, in fenestrated endothelium of the kidney glomeruli (Figure 3C), the hepatic sinusoid (data not shown), and the continuous capillaries of lung (data not shown), AHNAK immunoreactivity was below the detection limit. This suggests that the level of AHNAK expression in endothelial cells might depend on the blood vessel properties.

Figure 5

AHNAK localization is restricted to lining epithelium. (A) The mucosae of the stomach is thrown into prominent folds consisting of gastric glands (GG). AHNAK is observed in a thin strand that lies immediately beneath the base of the gastric glands and extends from the base of the glands in the lamina propria between the glands (arrowhead). This strand corresponds to the muscu-laris mucosae and supporting tissue (lamina propria and blood vessels). (B) Salivary glands are composed of mucous acini (MA) and serous acini. Both are embraced by processes of contractile cells called myoepithelial cells (arrowhead) which, on contraction, force the secretion from the acinar lumen into the duct system. (C) In the alveolar duct of the breast, myoepithelial cells surrounding the ductal lining cells are AHNAK-positive, whereas cuboidal epithelial cells are not immunostained. (D) The prostate glandular epithelium (GE) is AHNAK-negative, whereas the supporting tissue, which contains many smooth muscle cells, is AHNAK-positive. (E) The exocrine pancreas, which forms the bulk of the gland, is composed of glandular acini (a), whereas the endocrine tissue of the pancreas forms the islets of Langerhans (I) scattered throughout the exocrine tissue. In this organ, AHNAK immunoreactivity is restricted to vessels (arrows). (F) The simple tubular glands (LP) of the colon extend from lumen into the muscularis mucosae (MM). Lamina propria, composed of supporting tissue and a thin strand of muscularis mucosae, is AHNAK-positive. Note also the cytoplasmic AHNAK immunoreactivity in luminal enterocytes. Arrows in C-E point to blood vessels. Bars = 50 μm.

Table 1

Tissue and cell localization of AHNAK

Tissue and organ	AHNAK-positive structure or cell type	Type of epithelium in the organ	AHNAK in epithelium
Epidermis	Keratinocytes	Lining	++
Esophagus	Stratified epithelium of mucosae, muscular layer (smooth muscle and skeletal muscle)	Lining	++
Liver	Bile duct epithelium	Lining	++
Brain	Ependymal cells	Lining	++
Testis	Tunica albuginea, myofibroblasts	Lining	++
Prostate	Supporting tissue (smooth muscle cells)	Secretory	−
Breast	Myoepithelial cells	Secretory	−
Salivary gland	Myoepithelial cells	Secretory	−
Small intestine	Smooth muscle layers	Secretory	−
Stomach	Muscularis mucosae, smooth muscle layers	Secretory	−
Adrenal	Negative	Secretory	−
Pancreas	Negative	Secretory	−
Heart	Cardiomyocytes
Skeletal muscle	Muscle fiber, myocytes
Uterus	Myometrium
Ovary	Theca cells
Kidney	Epithelium of collecting tubule, duct of Bellini, pelvicalyceal system	Lining/absorptive	+ +/−
Large intestine	Smooth muscle layers	Absorptive and protective	+
Lung	Epithelium of bronchiole,	Lining/secretory	+
	epithelium of alveoli (pneumocytes),^a smooth muscle	Lining	++
Spleen	Negative
Thymus	Negative
Lymph node	Negative
Blood vessels	Endothelium and smooth muscle cells

^aConfirmed in human lung.

In adult tissue, the main subcellular localization of AHNAK is at the plasma membrane. Cytoplasmic staining was also observed in fibroblast or certain epithelial cells, such as those in the colon (Figure 5F). Exclusive membrane location of AHNAK has also been recently reported in rat cardiomyocytes by Hohaus et al. (2002), using two region-specific AHNAK antibodies raised against either amino-terminal or carboxyl-terminal epitopes. These findings contrast with the predominant nuclear localization of AHNAK observed in cultured cells lines of nonepithelial origin (Shtivelman and Bishop 1993). However, recent data reported that in MDCK cells derived from canine kidney epithelium, membrane targeting of AHNAK is regulated by cell density and the formation of cell-cell contacts (Sussman et al. 2001). The authors proposed that the plasma membrane-anchored AHNAK may be involved in the growth arrest of normal epithelial cells (Sussman et al. 2001). In adult mouse tissues, AH-NAK-expressing cells are terminally differentiated cells and therefore do not proliferate. This status might account for the predominant membrane localization of AHNAK in most adult tissues. Alternatively, recent studies indicate that AHNAK may be proteolytically cleaved in vivo (Skoldberg et al. 2002) and that AHNAK fragments may have functional roles intracellularly (Sekiya et al. 1999). It is therefore possible that, in cell culture, cytoplasmic AHNAK is more susceptible to proteolysis, and that AHNAK proteolysis could contribute to the apparent nuclear localization of the protein. In normal keratinocytes, endogenous full-length AHNAK accumulates at the cell membrane, but the transfected C-terminus of AHNAK accumulates within the cell nuclei (Nie et al. 2000).

The predominant cellular expression of AHNAK in muscles and in lining epithelium that we report in this study might help to elucidate the membrane function of this giant protein. First, it allows restriction of the search for AHNAK function to essentially two specialized cell types. Moreover, several structural and functional features are shared between muscle and lining epithelium. These tissues are subjected to stretching force and have elastic properties. For example, the epithelium of the esophagus is subjected to the stretching force of the bolus, and muscle cells to stretching forces during the contraction-relaxation cycles. This implies that common cellular organizations are involved in both tissues to respond and control the stretching forces. The first is the cell-cell interactions necessary for maintaining tissue integrity. Several molecular structures are known to participate in cell-cell interactions, such as desmosomes, tight junctions, and adherens junctions. Although initially AHNAK was co-purified with the desmosomal component during cell fractionation of keratinocytes (Hieda and Tsukita 1989), and therefore was been named desmoyokin, later studies clearly demonstrated that AHNAK (desmoyokin) is not directly associated with desmosomes (Masunaga et al. 1995). We have investigated the possible association of AHNAK with tight junctions in endothelial cells of brain capillary and ependymal cells (data not shown). Double immunofluorescence labeling and confocal microscopy analysis for AHNAK and the tight junction-associated protein ZO1 revealed a distinct localization of the two proteins (unpublished data). Finally, the observation that AHNAK can migrate from the cytoplasm to the plasma membrane in cell culture without any effect on cell-cell adhesion (Hashimoto et al. 1995) does not support a direct role for AHNAK in cell-cell interactions. We envision that AHNAK may be more likely associated with the structural support of the membrane during changes in membrane shape, the reinforcement of the cortical framework, and signaling pathways. It is significant that AHNAK shares structural similarity and subcellular localization with two other giant proteins abundantly expressed in muscle cells, titin and dystrophin. Both of these giant proteins lie close beneath the plasma membrane (Shimada et al. 1993; Mussini et al. 1995). Dystrophin is part of a multimeric protein complex that links actin on the inside of the cells to extracellular matrix protein through transmembrane proteins (Samitt and Bonilla 1990; Wakayama et al. 1993; Norwood et al. 2000). Titin is a giant muscle protein. Its size and location within the sarcomere structure suggests that it plays an important role in the mechanism of muscle elasticity (Freiburg et al. 2000). It will be very interesting to investigate if AHNAK may also act as a scaffold protein, recruiting and maintaining the organization of cytoskeletal proteins at the plasma membrane. This will obviously require the identification and characterization of AHNAK-associ-ated proteins at the plasma membrane. Two membrane-associated signaling pathways regulated by AHNAK have already been characterized. In the cardiomyocyte, AHNAK modulates membrane-bound L-type calcium channel activity (Haase et al. 1999). Intriguingly, organization of the cortical actin network has been implicated in modulation of voltage-gated ion channel activity (Maguire et al. 1998), and in vitro interaction between the carboxyl-terminal region of AHNAK and actin has recently been described (Hohaus et al. 2002). Peptides from the central domain of AHNAK also bind and activate PLCγ in the presence of arachidonic acid (Sekiya et al. 1999). The activation of this enzyme mediates several cellular responses, including cytoskeletal rearrangements that lead to protection of the plasma membrane in response to mechanical stress (Ryan et al. 2000; Ruwhof et al. 2001). Finally, taking into account that AHNAK is an abundant muscular protein with possible important functions associated with structural support of the plasma membrane, it will be now of primary interest to investigate if its function is altered in muscular dystrophies that are characterized by cell membrane defects.

Footnotes

Acknowledgements

Supported in part by a grant from Association pour la Recherche contre le Cancer (Christian Delphin) and a Ligue Nationale contre le Cancer fellowship (Benoit Gentil).

We thank Dr Peoc'h for valuable advice and Dr La Marre Jonathan for critical reading of the manuscript.

References

Boni

Burg

Doguoglu

Ilg

Schafer

Muller

Heizmann

(1997) Immunohistochemical localization of the Ca2+ binding S100 proteins in normal human skin and melanocytic lesions. Br J Dermatol 137:39–43

Freiburg

Trombitas

Hell

Cazorla

Fougerousse

Centner

Kolmerer

. (2000) Series of exon-skipping events in the elastic spring region of titin as the structural basis for myofibrillar elastic diversity. Circ Res 86:1114–1121

Gentil

Delphin

Mbele

Deloulme

Ferro

Garin

Baudier

(2001) The giant protein AHNAK is a specific target for the calcium- and zinc-binding S100B protein: potential implications for Ca2+ homeostasis regulation by S100B. J Biol Chem 276:23253–23261

Haase

Podzuweit

Lutsch

Hohaus

Kostka

Lindschau

Kott

. (1999) Signaling from beta-adrenoceptor to L-type calcium channel: identification of a novel cardiac protein kinase A target possessing similarities to AHNAK. FASEB J 13:2161–2172

Hashimoto

Amagai

Parry

Dixon

Tsukita

Miki

Sakai

. (1993) Desmoyokin, a 680 kDa keratinocyte plasma membrane-associated protein, is homologous to the protein encoded by human gene AHNAK. J Cell Sci 105:275–286

Hashimoto

Gamou

Shimizu

Kitajima

Nishikawa

(1995) Regulation of translocation of the desmoyokin/AHNAK protein to the plasma membrane in keratinocytes by protein kinase C. Exp Cell Res 217:258–266

Hieda

Tsukita

(1989) A new high molecular mass protein showing unique localization in desmosomal plaque. J Cell Biol 109:1511–1518

Hohaus

Person

Behlke

Schaper

Morano

Haase

(2002) The carboxyl-terminal region of ahnak provides a link between cardiac L-type Ca2+ channels and the actin-based cytoskeleton. FASEB J 16:1205–1216

Kingsley

McGrath

Maltby

Koniski

Ramchandran

Palis

(2001) Subtractive hybridization reveals tissue-specific expression of ahnak during embryonic development. Dev Growth Differ 43:133–143

10.

Kudoh

Wang

Minoshima

Hashimoto

Amagai

Nishikawa

Shtivelman

. (1995) Localization of the human AHNAK/desmoyokin gene (AHNAK) to chromosome band 11q12 by somatic cell hybrid analysis and fluorescence in situ hybridization. Cytogenet Cell Genet 70:218–220

11.

Maguire

Connaughton

Prat

Jackson

Jr Cantiello

(1998) Actin cytoskeleton regulates ion channel activity in retinal neurons. Neuroreport 9:665–670

12.

Masunaga

Shimizu

Ishiko

Fujiwara

Hashimoto

Nishikawa

(1995) Desmoyokin/AHNAK protein localizes to the non-desmosomal keratinocyte cell surface of human epidermis. J Invest Dermatol 104:941–945

13.

Motta

Familiari

(1981) Occurrence of a contractile tissue in the theca externa of atretic follicles in the mouse ovary. Acta Anat (Basel) 109:103–114

14.

Mussini

Sogos

Della Barbera

Ennas

Gremo

(1995) Immunolocalisation of dystrophin in the immature human neurons and muscles. Ital J Anat Embryol 100:155–163

15.

Nie

Ning

Amagai

Hashimoto

(2000) C-Terminus of desmoyokin/AHNAK protein is responsible for its translocation between the nucleus and cytoplasm. J Invest Dermatol 114: 1044–1049

16.

Norwood

Sutherland-Smith

Keep

Kendrick-Jones

(2000) The structure of the N-terminal actin-binding domain of human dystrophin and how mutations in this domain may cause Duchenne or Becker muscular dystrophy. Structure Fold Des 8:481–491

17.

Ruwhof

van Wamel

Noordzij

Aydin

Harper

van der Laarse

(2001) Mechanical stress stimulates phospholipase C activity and intracellular calcium ion levels in neonatal rat car-diomyocytes. Cell Calcium 29:73–83

18.

Ryan

Gross

Hajduczok

(2000) Calcium-dependent activation of phospholipase C by mechanical distension in renin-expressing As4.1 cells. Am J Physiol 279:E823–829

19.

Samitt

Bonilla

(1990) Immunocytochemical study of dystrophin at the myotendinous junction. Muscle Nerve 13:493–500

20.

Sekiya

Bae

Jhon

Hwang

Rhee

(1999) AHNAK, a protein that binds and activates phospholipase C-gamma 1 in the presence of arachidonic acid. J Biol Chem 274:13900–13907

21.

Self

Schroeder

Gown

(1988) Hamster thecal cells express muscle characteristics. Biol Reprod 39:119–130

22.

Shimada

Atsuta

Sonoda

Shiozaki

Maruyama

(1993) Distribution of connectin (titin) and transverse tubules at myotendinous junctions. Scann Microsc 7:157–163

23.

Shrestha

Muramatsu

Kudeken

Mori

Takai

Ilg

Schafer

. (1998) Localization of Ca(2+)-binding S100 proteins in epithelial tumours of the skin. Virchows Arch 432:53–59

24.

Shtivelman

Bishop

(1993) The human gene AHNAK encodes a large phosphoprotein located primarily in the nucleus. J Cell Biol 120:625–630

25.

Shtivelman

Cohen

Bishop

(1992) A human gene (AHNAK) encoding an unusually large protein with a 1.2-microns polyionic rod structure. Proc Natl Acad Sci USA 89:5472–5476

26.

Skoldberg

Ronnblom

Thornemo

Lindahl

Bird

Rorsman

Kampe

. (2002) Identification of AHNAK as a novel autoantigen in systemic lupus erythematosus. Biochem Biophys Res Commun 291:951–958

27.

Sussman

Stokoe

Ossina

Shtivelman

(2001) Protein kinase B phosphorylates AHNAK and regulates its subcellular localization. J Cell Biol 154:1019–1030

28.

Wakayama

Shibuya

Jimi

Takeda

Oniki

(1993) Size and localization of dystrophin molecule: immunoelectron microscopic and freeze etching studies of muscle plasma membranes of murine skeletal myofibers. Acta Neuropathol (Berl) 86:567–577