The Neuronal Growth-Associated Protein (GAP)-43 Is Expressed by Corticotrophs in the Rat Anterior Pituitary After Adrenalectomy
http://www.100md.com
《内分泌学杂志》
Department of Cell Biology and Neuroscience (C.M.P., H.J.C.), Montana State University, Bozeman, Montana 59717
Department of Anatomy and Cell Biology (J.A.W.), University of North Dakota, Grand Forks, North Dakota 58203
Department of Anesthesiology (T.H.S.), University of North Carolina, Chapel Hill, North Carolina 27599
WWAMI Medical Education Program (C.M.P., C.L.P.), University of Washington School of Medicine, Seattle, Washington 98195
Abstract
The neuronal growth-associated protein (GAP)-43 has been localized in both long fibers and punctate clusters by immunocytochemistry within the rat anterior pituitary (AP). After adrenalectomy (ADX), GAP-43 immunoreactivity (GAP-43-ir) is greatly increased and is associated with corticotrophs at the light microscopic level. We have undertaken an electron microscopic study to determine the cellular localization of GAP-43 in the post-ADX AP. Using preembedding immunocytochemistry, we found GAP-43-ir localized exclusively to the cytoplasmic surface of the plasmalemma within a subset of endocrine cells with ultrastructure typical of degranulated corticotrophs at 4 d after ADX. We combined preembedding immunoelectron microscopy for GAP-43 with immunogold labeling for ACTH and found that GAP-43-ir was invariably present only in cells containing ACTH-positive granules. The density of GAP-43-ir was highest within extensive processes emanating from the soma, suggesting that these processes are the basis for the punctate clusters of GAP-43 staining seen surrounding corticotrophs in the light microscope. We also observed rare synaptic-like contacts between GAP-43-ir processes and distant cell bodies. GAP-43 mRNA was detected in extracts of the AP 4 d after ADX using RT-PCR, and quantitative PCR confirmed that GAP-43 mRNA was significantly up-regulated in the AP in response to ADX. We postulate that increased expression of GAP-43 may stimulate process outgrowth and intercellular communication by activated corticotrophs.
Introduction
THE GROWTH-ASSOCIATED protein (GAP)-43 (also known as B-50, F1, or neuromodulin) is a membrane-associated phosphoprotein that has been implicated in axonal growth, synaptic remodeling, and secretion of both catecholamines and neuropeptides (1, 2, 3). Expression of GAP-43 is correlated with axonal growth throughout the developing nervous system, whereas expression is substantially reduced and regionally restricted in the adult (reviewed in Ref.3). However, up-regulation of GAP-43 expression is associated with axonal growth by mature neurons in a variety of situations. For example, increased synthesis of GAP-43 accompanies axonal regeneration in the injured optic nerve of nonmammalian vertebrates but is absent from the injured mammalian optic nerve, which fails to regenerate (4). Increased levels of GAP-43 protein and/or mRNA are observed in mature mammalian sensory and motor neurons during axonal regeneration after peripheral nerve injury (5, 6), and up-regulation also occurs in intact brain neurons after induction of collateral axonal sprouting (7, 8).
We previously reported the presence of GAP-43 immunoreactivity (GAP-43-ir) within the rat anterior pituitary (AP) that appears both in the form of long fibers and as dense punctate clusters (9). The extent and intensity of both forms of GAP-43-ir increase after adrenalectomy (ADX) (10, 11), similar to the post-ADX increase observed in the density of nerve fibers in the AP that are immunoreactive for the neuropeptides, substance P and calcitonin gene-related peptide (12). A recent study demonstrated that GAP-43-ir is indeed colocalized within substance P-containing fibers coursing throughout the rat AP and that these axons originate from the nodose ganglia (13). Consistent with these observations, we previously interpreted the post-ADX increase in the intensity and extent of fibrous and punctate GAP-43-ir in the AP as evidence of axonal elongation and sprouting of new axon terminals, respectively. Most interestingly, the putative terminals visualized as punctate clusters of GAP-43-ir were found to be associated almost exclusively with corticotrophs using confocal dual-label immunofluorescence microscopy (11). We now report that, when localized at the electron microscopic level, GAP-43-ir in the post-ADX AP is found within fine cellular processes originating from corticotrophs themselves. These processes can be seen extending to make specialized contacts with other gland cells. In addition, RT-PCR confirmed the presence of GAP-43 mRNA within the AP after ADX. We postulate that this novel nonneuronal expression of GAP-43 may stimulate process outgrowth and formation of cellular contacts by activated corticotrophs.
Materials and Methods
Animals
Male Holtzman albino rats of approximately 250 g body weight were obtained from Harlan Laboratories (Indianapolis, IN) and housed in the Montana State University Animal Resource Center, an Association Assessment and Accreditation of Laboratory Animal Care-accredited facility. All animals were housed under a 12-h light, 12-h dark cycle with ad lib access to lab chow and tap water. Bilateral adrenalectomy was performed under general anesthesia using a ketamine cocktail (26 mg ketamine + 5 mg xylazine + 0.8 mg acepromazine per kilogram body weight, im), incisions were closed with wound clips and treated with wound powder, and 0.05 mg/kg buprenorphine was administered sc to provide postsurgical analgesia. Rats undergoing bilateral adrenalectomy were provided with normal saline to drink instead of tap water throughout the postsurgical survival period. All experimental protocols used in these studies followed the guidelines in the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Montana State University Institutional Animal Care and Use Committee.
Tissue preparation
All animals to be used for immunocytochemistry were anesthetized with isoflurane and perfused intracardially with cold saline for 2 min followed by 500 ml of Nakane’s periodate-lysine-paraformaldehyde (PLP) fixative prepared immediately before use (14) and modified as previously described (9). Pituitaries were removed, postfixed overnight in PLP at 4 C, and then embedded in an egg yolk and gelatin matrix. The matrix block containing the tissue was then submerged for an additional 48 h in PLP at 4 C before sectioning on a Vibratome at 40 μm in the coronal plane.
Animals to be used for analysis of GAP-43 mRNA were decapitated under isoflurane anesthesia and the pituitary quickly removed and frozen on dry ice. A coronal slice of brain containing the piriform cortex and a small piece of liver were also removed and frozen from some animals. Tissues were wrapped in aluminum foil and stored in sealed containers at –80 C until use.
Immunoelectron microscopy
Localization of GAP-43 in free-floating sections of pituitary by peroxidase immunocytochemistry was performed as previously described (9) using sheep antirat GAP-43 at 1:1000 dilution as the primary antisera (generous gift of Dr. Larry Benowitz, Harvard Medical School, Boston, MA), biotinylated goat antisheep IgG at 1:200 dilution (Vector Laboratories, Burlingame, CA) as the secondary, and the Vector Elite ABC kit to produce the avidin-biotin-peroxidase complex. Binding of the ABC reagent was visualized using diaminobenzidine (Sigma, St. Louis, MO) as chromogen with generation of H2O2 by the glucose oxidase method (15). After immunocytochemical staining the tissues were postfixed in 5% glutaraldehyde in 0.1 M PO4 (pH 7.4) for 30 min followed by 4% OsO4 in PBS for 30 min. Sections were then dehydrated through increasing concentrations of ethanol to propylene oxide, infiltrated overnight in Epon/Araldite embedding medium (Ted Pella, Redding, CA), flat mounted between glass microscope slides pretreated with dimethyldichlorosilane (1% in benzene) to facilitate removal, and polymerized at 60 C for a minimum of 72 h. After examination in the light microscope, the top slide was removed and areas of interest in the section were cut out with a razor and glued to the flat surface of a blank Epon/Araldite bullet for ultrathin sectioning. Silver to gold ultrathin sections were counterstained with a combination of aqueous uranyl acetate and Reynolds lead citrate and viewed using an EM10C transmission electron microscope (Zeiss, New York, NY) at 60–80 kV.
For dual-ultrastructural localization of GAP-43 and ACTH, we combined the preembedding peroxidase technique described above with postembedding immunogold labeling of ACTH. After the preembedding procedures outlined above, ultrathin sections of pituitary mounted on 2 x 1 mm formvar/carbon-coated nickel slot grids (EMS Supply, Hatfield, PA) were initially rehydrated by floating on drops of Tris-buffered saline (TBS) (pH 7.4) for 15 min. This was followed by sequential incubation at 4 C on drops of 5% BSA in TBS for 1 h, rabbit antihuman ACTH overnight [Dako, Carpinteria, CA; 1:100 in incubation buffer consisting of 1% normal goat serum + 0.1% Tween 20 + 1% BSA in TBS (pH 8.2)], TBS, and 5 nm gold-conjugated goat antirabbit IgG for 2 h (Ted Pella; 1:100 in incubation buffer). Sections were then washed in TBS and double-distilled sterile filtered H2O, counterstained with a combination of aqueous uranyl acetate and Reynolds lead citrate, and viewed using a Zeiss EM10C transmission electron microscope at 80 kV. Negative controls were prepared in an identical manner but with omission of the 5 nm gold conjugated goat antirabbit IgG.
Detection of GAP-43 mRNA by RT-PCR and gel electrophoresis
RT-PCR was used to determine whether GAP-43 mRNA was present in extracts of anterior pituitary, brain (piriform cortex), and liver of intact and adrenalectomized rats. Total RNA was extracted from individual tissue samples, which were homogenized in 20 mM sodium acetate containing 4 M guanidinium isothyocyanate, 0.1 mM dithiothreitol, and 0.5% N-lauroylsarcosine (lysis solution; Ambion, Austin, TX) and then diluted with 2 vol 10 mM Tris buffer (pH 7.5) containing 1 mM EDTA. After phenol-chloroform extraction, total RNA was precipitated in isopropanol and resuspended in 20 μl nuclease-free water. Extracts were stored at –80 C until analyzed. RT-PCR was performed with approximately 0.5 μg of total RNA per tube using the Access RT-PCR kit (Promega, Madison, WI) according to the manufacturer’s specifications. We used previously published GAP-43 primer sequences that result in a PCR product of 708 bp that covers the entire coding region of rat GAP-43 (16). These were: upstream (5') tgctgtgctgtatgagaagaacc (3'); and downstream (5') ggcaacgtggaaagccgtttcttaaagt (3'). These primers also span an intron, providing a control for potential amplification of genomic DNA. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primers used as positive controls were: upstream (5') tccaccaccctgttgctgta (3'); and downstream (3') accacagtccatgccatcac (5'). Amplification was performed in a thermal cycler (PerkinElmer, Boston, MA) using a cycle sequence of 94 C for 30 sec, 60 C for 1 min, and 68 C for 2 min. Samples were collected after 30, 35, or 40 cycles and loaded onto 1.5% agarose gels for electrophoreses.
Determination of relative GAP-43 mRNA levels by quantitative PCR
Quantitative PCR was used to measure relative levels of GAP-43 mRNA in extracts of anterior pituitary from intact and adrenalectomized rats. Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol and each sample adjusted to 1.0 μg/μl based on absorbance at 260 nm. Aliquots were run on a 1.7% denaturing agarose gel to verify quality of total RNA before performing reverse transcription of 3.5 μg of total RNA per sample using an oligo T(24) primer and Superscript II reverse transcriptase (Invitrogen) according to the manufacturer’s protocol. Reaction products were diluted to 80 ng/μl and used as templates for quantitative PCR performed in a Roto-Gene 3000 (Corbett Research, Sydney, Australia). Each 20-μl reaction tube contained 10 μl SybrGreen PCR master mix (Applied Biosystems, Foster City, CA), 1 μl template, 1 μl primers (7.5 pmol each), and 8 μl RNase-free water. Left and right primers for GAP-43 mRNA were each 20 bases in length (5'-tttcctctcctgtcctgctc-3', 5'-tggacttgggatctttcctg-3'; ordered from Sigma-Genosys, The Woodlands, TX) and were designed to produce a PCR product 112 bp in length spanning bases 974-1086 in the rat GAP-43 mRNA (accession NM_017195 in the NCBI database). Amplication of GAPDH mRNA was used as a normalization standard. Left and right primers for GAPDH mRNA were each 20 bases in length (5'-gtggacctcatggcctacat-3', 5'-tgtgagggagatgctcagtg-3'; ordered from Sigma-Genosys) and were designed to produce a PCR product 150 bp in length spanning bases 1813–1963 in the rat GAPDH mRNA (accession NM_017195). Both the GAP-43 and GAPDH PCR products were determined to be single bands of the predicted size by isolation on a 1.5% nondenaturing agarose gel.
Results
GAP-43-ir is present in corticotrophs after ADX
We previously reported that GAP-43-ir appears in both punctate clusters and occasional elongated axons scattered throughout the parenchyma of the AP of the intact adult rat (9). Both the extent and intensity of punctate staining are greatly increased within 4 d after ADX, and we observed that punctate clusters of GAP-43-ir visualized in the confocal microscope are colocalized almost exclusively with ACTH-positive corticotrophs after ADX (11). In the present study, we used preembedding peroxidase immunoelectron microscopy to further characterize punctate GAP-43-ir at the ultrastructural level in the post-ADX AP. We found GAP-43-ir localized exclusively to the cytoplasmic surface of the plasmalemma within a subset of endocrine cells (Fig. 1A, arrows). We consistently observed that the reaction product was distributed heterogeneously, with areas of dense reaction product separated by varying distances of nonlabeled plasmalemma. At 4 d after ADX, the GAP-43-ir cells ranged in shape from ovoid to stellate and were characterized by well-developed rough endoplasmic reticulum, numerous mitochondria, and sparse secretory granules arrayed in close juxtaposition to the cytoplasmic surface of the plasmalemma. Many of the secretory granules were of the immature form, having a halo surrounding a dense core. No specific association between secretory granules and GAP-43-ir was observed. The ultrastructure of these GAP-43-ir cells is consistent with that of corticotrophs that are degranulated after ADX (17, 18, 19).
To confirm the presence of GAP-43-ir in corticotrophs, we combined preembedding immunoelectron microscopy for GAP-43 with postembedding immunogold labeling for ACTH on sections prepared from 4-d post-ADX animals. We found that GAP-43-ir was invariably present only in cells containing ACTH-positive granules. Numerous secretory granules labeled with colloidal gold particles were observed along the periphery of each cell that contained GAP-43-ir (Fig. 1B).
GAP-43-ir is concentrated within fine processes that form specialized contacts with other cells
We frequently observed extensive processes of varying caliber emanating from the soma of GAP-43-ir cells. These processes contained numerous secretory granules and often projected for a substantial distance between surrounding cells. Extensive deposits of GAP-43-ir were associated with the cytoplasmic surface of the plasmalemma throughout the length of the processes (Fig. 2, arrows, and Fig. 3A). The density of reaction product was greater within processes than in any other region of the cell, and it is reasonable to conclude that these processes are the basis for the punctate clusters of GAP-43 staining seen in the light microscope. GAP-43-ir was also present in the region of the cell soma immediately adjacent to the base of these processes (Fig. 2), but a much smaller proportion of the plasmalemma was covered by reaction product within the soma than in the processes themselves.
We also observed rare specialized synaptic-like contacts between GAP-43-ir processes and distant cell bodies. These putative terminals contained extensive GAP-43-ir, secretory granules, and mitochondria. Figure 3 (A–C) shows a set of serial sections at increasingly higher magnifications collected through the extent of one such synaptic-like contact. The process can be seen extending beyond this specialized contact in a manner similar to a synapse en passant (Fig. 3C, arrows). Note that the plasmalemma of the recipient cell soma protrudes outward to form a distinct contact surface (Fig. 3, B and C, arrowheads) with the GAP-43-ir process. However, specializations typical of a neuronal postsynaptic membrane were not observed. Whereas the presence of secretory granules in the process immediately adjacent to the synaptic-like contact suggests that exocytosis of granule contents may occur at this site, we did not observe any granules actually captured in the process of exocytosis at the small number of contact sites examined.
Expression of GAP-43 mRNA is up-regulated in the AP after adrenalectomy
RT-PCR and gel electrophoresis.
GAP-43 mRNA was detected in extracts of the AP 4 d after ADX using RT-PCR, and absence of a detectable signal in similar AP extracts from intact rats suggested that expression of GAP-43 mRNA was up-regulated in response to ADX (Fig. 4). GAP-43 mRNA was detected in seven of eight ADX pituitaries examined but in none of six intact pituitaries. Liver and brain extracts were used as negative and positive tissue controls, respectively, for the GAP-43 signal, and GAPDH was used as a positive control for the RT-PCR. The GAP-43 product from both AP and brain was approximately 700 bp in length, as previously reported using identical primers (16), and was confirmed to be authentic rat GAP-43 by extraction from the gel and sequencing.
Quantitative PCR.
A significant increase in the abundance of GAP-43 mRNA in the AP 4 d after ADX was confirmed using quantitative PCR. Templates were prepared from total RNA extracts of the AP of three ADX and three age-matched intact rats. Quantitative PCR analysis was performed twice using triplicate samples from each animal in each run and the mean cycles to threshold calculated for each subject (Table 1). The average decrease of 2.61 in mean cycles to threshold for the ADX group was highly significant (Student’s t test two-tailed; t = 11.76, df = 4, P < 0.001) and corresponds to an approximately 6-fold increase in the level of GAP-43 mRNA in the AP.
Discussion
Our results demonstrate that GAP-43 mRNA and protein are expressed by intrinsic cells of the rat AP. Both GAP-43 mRNA abundance and GAP-43-ir are dramatically increased in the AP after ADX, and we found GAP-43-ir to be localized within ACTH-positive corticotrophs. GAP-43-ir is highly concentrated within extensive fine processes extending from the corticotroph soma, and in some cases these processes could be observed forming specialized synaptic-like contacts with other cells. We had previously interpreted the association of GAP-43-ir with corticotrophs at the light microscopic level as representing innervation of corticotrophs by axons entering the AP (11). We observed scattered GAP-43-ir axonal profiles within the AP, and these are presumably the substance P-containing fibers recently found by others to originate from the nodose ganglia (13). However, the present ultrastructural findings show that the extensive punctate GAP-43-ir seen associated with corticotrophs after ADX is not of neuronal origin but rather represents expression of GAP-43 by corticotrophs themselves.
Whereas GAP-43 was earlier considered to be a neuron-specific protein (2, 20), there is more recent evidence of its expression in additional cell types, primarily of nervous origin. These include oligodendroglia, Schwann cells, and some astrocytes (reviewed in Ref.21) as well as noradrenergic chromaffin cells of the adrenal medulla (reviewed in Ref.3). In addition, transient expression in mesoderm of embryonic chick limb (22) and in human embryonic (23) and regenerating muscle fibers (24) has been reported, and GAP-43-ir was observed in regenerating myocytes within biopsies of inflammatory human myopathies (25). However, we are unaware of any previous reports of GAP-43 expression in normal mature tissues outside the nervous system or adrenal medulla. Thus, the finding that GAP-43 is expressed in corticotrophs is novel and unexpected.
Whereas the localization of GAP-43 in corticotrophs was surprising, the increased abundance of GAP-43 mRNA and protein that we observed after ADX is consistent with studies showing that expression of GAP-43 is negatively regulated by glucocorticoids. This effect was first observed in PC12 cells (26) and subsequently demonstrated in the hippocampus by Chao and colleagues (27, 28), who found that GAP-43 mRNA levels increased in pyramidal neurons 7 d after ADX and that this increase could be blocked by corticosterone replacement. It is also possible that ACTH-like peptides could influence the activity of GAP-43, by either altering gene expression (29) or affecting proteolysis (30). Determining whether these or other mechanisms regulate GAP-43 expression in corticotrophs will require additional studies.
GAP-43 is implicated in two principal cell functions, process outgrowth and vesicular release. An extensive literature has correlated GAP-43 expression in neurons with neurite growth during development and regeneration of mature axons, both in vivo and in cell culture (reviewed in Refs.1, 2, 3, 20). Similar observations have been reported in Schwann cells (21). Induced expression of GAP-43 has also been shown to stimulate process extension in a variety of transfected cell lines, including PC12 (31), 3T3 (32), COS (32, 33), CHO (32), and AtT-20 (34) cells. The work of Gamby et al. (34) is especially relevant to the present findings because AtT-20 cells are of pituitary origin and secrete proopiomelanocortin-derived peptide hormones. Whereas GAP-43 was not detectable in the parental AtT-20 mouse pituitary tumor cell line, the D16 subclone was found to express high levels of GAP-43. Consistent with this difference, D16 cells had a flattened morphology and extended processes when attached to a laminin-coated substrate, whereas AtT-20 cells maintained a rounded morphology under similar conditions. In addition, transfection of AtT-20 cells with rat GAP-43 cDNA induced flattening and process extension similar to that seen in D16 cells.
These authors also reported marked effects of GAP-43 expression on secretion of -endorphin from cultured AtT-20 cells. The original AtT-20 line exhibited much less robust K+-induced secretion than did the GAP-43-positive D16 clone, but depolarization induced release of -endorphin was greatly stimulated in transfected AtT-20 cells expressing high levels of GAP-43 (34). A role for GAP-43 in stimulation of vesicular release from AP cells is consistent with numerous studies demonstrating that GAP-43 facilitates release of catecholamines and cholecystokinin-8 from both brain synaptosomes and adrenal chromaffin cells (reviewed in Ref.3). These actions may be mediated by association of GAP-43 with PI(4,5)P(2) plasmalemmal rafts, in which it has been shown to affect actin accumulation and dynamics (reviewed in Ref.35). GAP-43 is a protein kinase C substrate, and it has been proposed that changes in GAP-43 phosphorylation subsequent to modulation of protein kinase C activity may increase the ability of GAP-43 to affect actin and promote high levels of vesicular release under conditions of sustained activation (36). In addition, GAP-43 modulates calcium signaling by binding to calmodulin and reducing activation of calcium-calmodulin-dependent enzymes (reviewed in Refs.3, 37). Interestingly, Gamby et al. (38) found that the ability of GAP-43 to stimulate K+-induced secretion of -endorphin from AtT-20 cells was dependent on both calmodulin binding and membrane association of the protein but that its ability to alter cellular morphology depended on the latter interaction only.
The well-documented effects of GAP-43 expression on cell morphology and hormone secretion suggest that the up-regulation in GAP-43 abundance we observed in activated corticotrophs in vivo may serve to stimulate both process extension and intercellular signaling. The preferential localization of GAP-43-ir within fine processes extending far from the corticotroph somata supports this conclusion. Stellate corticotrophs are known to be highly plastic cells (reviewed in Ref.39). Growth of extensive new processes by activated corticotrophs occurs both in vivo after ADX (17, 40, 41) and cold stress (42) and in vitro in response to CRH (43). Process outgrowth can occur very rapidly, having been observed after only 3–10 min of exposure to CRH in culture (43) or after 30 min of cold stress (42), but extensive processes are also maintained for at least 6 wk and perhaps indefinitely under the sustained activation induced by ADX (41). We previously reported that GAP-43-ir associated with corticotrophs is dramatically increased at both 4 and 14 d after ADX, coinciding with the growth of extensive cell processes (11). GAP-43-ir is also increased in the rat AP in response to 1 h per day of restraint stress applied for 1 wk (Paden, C. M., and J. A. Watt, unpublished observations). Modest GAP-43-ir is present specifically in association with corticotrophs in intact animals (11), and we were able to detect GAP-43 mRNA in the AP of intact rats by quantitative PCR (Table 1). These observations suggest that constitutive expression of the protein at a reduced level could serve to prime corticotrophs for a rapid morphological response to activation. Subsequent up-regulation of GAP-43 expression may then serve to sustain process outgrowth and/or stimulate branching of processes (31) as well as to facilitate cell signaling. Further experiments are necessary to more precisely determine the temporal relation between increased GAP-43 expression and process outgrowth and maintenance.
The occasional synaptic-like contacts that we observed between corticotroph processes containing GAP-43 and other gland cells suggests that these processes are involved in intercellular communication within the AP and are not simply sites of hormone release into the vasculature. Several previous studies have described the extensive processes of activated corticotrophs as being in close contact with or even surrounding neighboring cells as well as extending to capillaries (17, 40, 41). Both the localization of GAP-43-ir processes within clusters of corticotrophs (11) and the ultrastructure of the postsynaptic cells (Fig. 4) suggest that these specialized contacts may be between pairs of corticotrophs. However, we have not been able to confirm the identity of the postsynaptic cells by immunocytochemistry. A previous report of corticotrophs forming synaptic-like contacts in the rat concluded that the morphology of the postsynaptic cells suggested they were lactotrophs (44), but similar synaptic-like contacts have been observed between pairs of corticotrophs in the dog (45). Growth of extensive interdigitated cell processes may serve to facilitate paracrine interactions between corticotrophs, whether or not specialized synaptic-like contacts are formed. This hypothesis is consistent with a variety of studies indicating that intercellular communication between corticotrophs can both stimulate and inhibit ACTH secretion in vivo and in vitro (46, 47, 48). One intriguing possibility is that paracrine interactions between corticotrophs may act to coordinate pulsatile release of ACTH, as suggested by the demonstration of an intrinsic pulse-generating mechanism underlying release of ACTH from perifused human pituitary (49). In this scenario increased expression of GAP-43 could play multiple roles, facilitating both vesicular release of ACTH and intercellular communication between activated corticotrophs.
Footnotes
This work was supported by National Science Foundation Grant IBN-9604670 (to C.M.P.) and National Institutes of Health, National Institute of Neurological Disorders and Stroke Grant 5 R01 NS032507 (to C.M.P.).
The authors have no conflicts of interest.
First Published Online November 3, 2005
Abbreviations: ADX, Adrenalectomy; AP, anterior pituitary; GAP, growth-associated protein; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GAP-43-ir, GAP-43 immunoreactivity; PLP, periodate-lysine-paraformaldehyde; TBS, Tris-buffered saline.
Accepted for publication October 26, 2005.
References
Benowitz LI, Routtenberg A 1997 GAP-43: an intrinsic determinant of neuronal development and plasticity. Trends Neurosci 20:84–91
Skene JHP 1989 Axonal growth-associated proteins. Annu Rev Neurosci 12:127–156
Oestreicher AB, De Graan PNE, Gispen WH, Verhaagen J, Schrama LH 1997 B-50, the growth associated protein-43: modulation of cell morphology and communication in the nervous system. Prog Neurobiol 53:627–686
Skene JH 1984 Growth-associated proteins and the curious dichotomies of nerve regeneration. Cell 37:697–700
Schreyer DJ, Pate Skene JH 1991 Fate of GAP-43 in ascending spinal axons of DRG neurons after peripheral nerve injury: delayed accumulation and correlation with regenerative potential. J Neurosci 11:3738–3751
Palacios G, Mengod G, Sarasa M, Baudier J, Palacios JM 1984 De novo synthesis of GAP-43: in situ hybridization histochemistry and light and electron microscopy immunocytochemical studies in regenerating motor neurons of cranial nerve nuclei in the rat brain. Brain Res Mol Brain Res 24:107–117
Benowitz LI, Rodriguez WR, Neve RL 1990 The pattern of GAP-43 immunostaining changes in rat hippocampal formation during reactive synaptogenesis. Brain Res Mol Brain Res 8:17–23
Bendotti C, Pende M, Samanin R 1994 Expression of GAP-43 in the granule cells of rat hippocampus after seizure-induced sprouting of mossy fibres: in situ hybridization and immunocytochemical studies. Eur J Neurosci 6:509–515
Paden CM, Moffett CW, Benowitz LI 1994 Innervation of the rat anterior and neurointermediate pituitary visualized by immunocytochemistry for the growth-associated protein GAP-43. Endocrinology 134:503–506
Lu CR, Meng FT, Benowitz LI, Ju G 1995 Evidence for axonal sprouting in the anterior pituitary following adrenalectomy in the rat. J Endocrinol 147:161–166
Paden CM, Babcock C, Conner KA, Duong DKT, Kuhl JM 1998 Axons containing the growth associated protein GAP-43 specifically target rat corticotrophs following adrenalectomy. J Neuroendocrinol 10:693–699
Ju G, Ma D, Fan S-C 1994 Response of calcitonin-gene-related peptide-like immunoreactive nerve fibers of the anterior pituitary to adrenalectomy in the rat. Neuroendocrinology 59:505–510
Vanhatalo S, Soinila S 2001 Evidence for nodose ganglion as the source of innervation to the anterior lobe of the pituitary gland. Neuroscience 107:491–497
McLean IW, Nakane PK 1974 Periodate-lysine-paraformaldehyde fixative—a new fixative for immunoelectron microscopy. J Histochem Cytochem 22:1077–1083
Itoh K, Konishi A, Nomura S, Mizuno N, Nakamura Y, Sugimoto T 1979 Application of coupled oxidase reaction to electron microscopic demonstration of horseradish peroxidase: cobalt-glucose oxidase method. Brain Res 175:341–346
Baetge EE, Hammang JP 1991 Neurite outgrowth in PC12 cells deficient in GAP-43. Neuron 6:21–30
Siperstein ER, Miller KG 1973 Hypertrophy of the ACTH-producing cell following adrenalectomy: a quantitative electron microscopic study. Endocrinology 93:1257–1268
Taniguchi Y, Shiino M 1990 Immunoelectron-microscopic and morphometric study of the rat corticotrophs after adrenalectomy. Acta Anat 137:45–48
Moriarty GC, Halmi NS 1972 Electron microscopic study of the adrenocorticotropin producing cell with the use of unlabeled antibody and the soluble peroxidase-antiperoxidase complex. J Histochem Cytochem 20:590–603
Benowitz LI, Perrone-Bizzozero NI 1991 The expression of GAP-43 in relation to neuronal growth and plasticity: when, where, how and why In: Gispen WH, Routtenberg A, eds. Progress in brain research. Amsterdam: Elsevier Science Publishers B.V.; 69–87
Sensenbrenner M, Lucas M, Deloulme J-C 1997 Expression of two neuronal markers, growth-associated protein 43 and neuron-specific enolase, in rat glial cells. J Mol Med 75:653–663
Stocker KM, Baizer L, Ciment G 1992 Transient expression of GAP-43 in nonneuronal cells of the embryonic chicken limb. Dev Biol 149:406–414
Moos T, Christensen LR 1993 GAP43 identifies developing muscle cells in human embryos. Neuroreport 4:1299–1302
Huess D, Schluep M 1998 Subcellular localization of phosphoprotein B-50 in regenerating muscle. An immuno-electron microscopic study. Neurol Res 20:360–364
Huess D, Engelhardt A, Gobel H, Neundorfer B 1995 Light-microscopic study of phosphoprotein B-50 in myopathies. Virchows Arch 426:69–76
Federoff HJ, Grabczyk E, Fishman MC 1988 Dual regulation of GAP-43 gene expression by nerve growth factor and glucocorticoids. J Biol Chem 263:19290–19295
Chao HM, McEwen BS 1994 Glucocorticoids and the expression of mRNAs for neurotrophins, their receptors and GAP-43 in the rat hippocampus. Brain Res Mol Brain Res 26:271–276
Chao HM, Sakai RS, Ma L-Y, McEwen BS 1998 Adrenal steroid regulation of neurotrophic factor expression in the rat hippocampus. Endocrinology 139:3112–3118
Zwiers H, Veldhuis HD, Schotman P, Gispen W-H 1976 ACTH, cyclic nucleotides, and brain protein phosphorylation in vitro. Neurochem Res 1:669–677
Coggins PJ, Zwiers H 1994 Detergents and peptides alter proteolysis and calmodulin binding of B-50/GAP-43 in vitro. J Neurochem 63:1491–1498
Jap Tjoen San ERA, Van Rozen AJ, Nielander HB, Oestreicher AB, Gispen W-H, Schotman P 1995 Expression levels of B-50/GAP-43 in PC12 cells are decisive for the complexity of their neurites and growth cones. J Mol Neurosci 6:185–200
Zuber MX, Goodman DW, Karns LR, Fishman MC 1989 The neuronal growth-associated protein GAP-43 induces filopodia in non-neuronal cells. Science 244:1193–1195
Strittmatter SM, Valenzuela D, Fishman MC 1994 An amino-terminal domain of the growth-associated protein GAP-43 mediates its effects on filopodial formation and cell spreading. J Cell Sci 107:195–204
Gamby C, Waage MC, Allen RG, Baizer L 1996 Growth-associated protein-43 (GAP-43) facilitates peptide hormone secretion in mouse anterior pituitary AtT-20 cells. J Biol Chem 271:10023–10028
Caroni P 2001 Actin cytoskeleton regulation through modulation of PI(4,5)P(2) rafts. EMBO J 20:4332–4336
Iannazzo L 2001 Involvement of B-50 (GAP-43) phosphorylation in the modulation of transmitter release by protein kinase C. Clin Exp Pharmacol Physiol 28:901–904
Slemmon JR, Feng B, Erhardt JA 2000 Small proteins that modulate calmodulin-dependent signal transduction: effects of PEP-19, neuromodulin, and neurogranin on enzyme activation and cellular homeostasis. Mol Neurobiol 22:99–113
Gamby C, Waage MC, Allen RG, Baizer L 1996 Analysis of the role of calmodulin binding and sequestration in neuromodulin (GAP-43) function. J Biol Chem 271:26698–26705
Childs GV 1992 Structure-function correlates in the corticotropes of the anterior pituitary. Front Neuroendocrinol 13:271–317
Westlund KN, Aguilera G, Childs GV 1985 Quantification of morphological changes in pituitary corticotropes produced by in vivo corticotropin-releasing factor stimulation and adrenalectomy. Endocrinology 116:439–445
Kubba MA, McNicol AM 1987 Stimulation of pituitary corticotrophs in the rat—ultrastructural studies. Virchows Arch B Cell Pathol 54:119–126
Sasaki F, Wu P, Rougeau D, Unabia G, Childs GV 1990 Cytochemical studies of responses of corticotropes and thyrotropes to cold and novel environmental stress. Endocrinology 127:285–297
Childs GV, Lloyd J, Unabia G, Rougeau D 1989 Growth and secretory responses of enriched populations of corticotropes. Endocrinology 125:2540–2549
Liu YY, Ju G 1998 Synaptoid contacts between gland cells of the anterior pituitary of the rat. Anat Rec 251:181–184
Ju G, Zhang X 1990 An electron microscopic study on substance P-like immunoreactive nerve fibers in the pars distalis of the adenohypophysis in the dog. Neuroscience 38:503–513
Schwartz J 1990 Evidence for intrapituitary intercellular control of adrenocorticotropin secretion. Mol Cell Endocrinol 68:77–83
Jia L-G, Canny BJ, Leong DA 1992 Paracrine communication regulates adrenocorticotropin secretion. Endocrinology 130:534–539
Giraldi FP, Cavagnini F 1998 Corticotropin-releasing hormone is produced by rat corticotropes and modulates ACTH secretion in a paracrine/autocrine fashion. J Clin Invest 101:2478–2484
Gambacciani M, Liu JH, Swartz WH, Tueros VS, Rasmussen DD, Yen SS 1987 Intrinsic pulsatility of ACTH release from the human pituitary in vitro. Clin Endocrinol (Oxf) 26:557–563(Charles M. Paden, John A. Watt, Tiffany )
Department of Anatomy and Cell Biology (J.A.W.), University of North Dakota, Grand Forks, North Dakota 58203
Department of Anesthesiology (T.H.S.), University of North Carolina, Chapel Hill, North Carolina 27599
WWAMI Medical Education Program (C.M.P., C.L.P.), University of Washington School of Medicine, Seattle, Washington 98195
Abstract
The neuronal growth-associated protein (GAP)-43 has been localized in both long fibers and punctate clusters by immunocytochemistry within the rat anterior pituitary (AP). After adrenalectomy (ADX), GAP-43 immunoreactivity (GAP-43-ir) is greatly increased and is associated with corticotrophs at the light microscopic level. We have undertaken an electron microscopic study to determine the cellular localization of GAP-43 in the post-ADX AP. Using preembedding immunocytochemistry, we found GAP-43-ir localized exclusively to the cytoplasmic surface of the plasmalemma within a subset of endocrine cells with ultrastructure typical of degranulated corticotrophs at 4 d after ADX. We combined preembedding immunoelectron microscopy for GAP-43 with immunogold labeling for ACTH and found that GAP-43-ir was invariably present only in cells containing ACTH-positive granules. The density of GAP-43-ir was highest within extensive processes emanating from the soma, suggesting that these processes are the basis for the punctate clusters of GAP-43 staining seen surrounding corticotrophs in the light microscope. We also observed rare synaptic-like contacts between GAP-43-ir processes and distant cell bodies. GAP-43 mRNA was detected in extracts of the AP 4 d after ADX using RT-PCR, and quantitative PCR confirmed that GAP-43 mRNA was significantly up-regulated in the AP in response to ADX. We postulate that increased expression of GAP-43 may stimulate process outgrowth and intercellular communication by activated corticotrophs.
Introduction
THE GROWTH-ASSOCIATED protein (GAP)-43 (also known as B-50, F1, or neuromodulin) is a membrane-associated phosphoprotein that has been implicated in axonal growth, synaptic remodeling, and secretion of both catecholamines and neuropeptides (1, 2, 3). Expression of GAP-43 is correlated with axonal growth throughout the developing nervous system, whereas expression is substantially reduced and regionally restricted in the adult (reviewed in Ref.3). However, up-regulation of GAP-43 expression is associated with axonal growth by mature neurons in a variety of situations. For example, increased synthesis of GAP-43 accompanies axonal regeneration in the injured optic nerve of nonmammalian vertebrates but is absent from the injured mammalian optic nerve, which fails to regenerate (4). Increased levels of GAP-43 protein and/or mRNA are observed in mature mammalian sensory and motor neurons during axonal regeneration after peripheral nerve injury (5, 6), and up-regulation also occurs in intact brain neurons after induction of collateral axonal sprouting (7, 8).
We previously reported the presence of GAP-43 immunoreactivity (GAP-43-ir) within the rat anterior pituitary (AP) that appears both in the form of long fibers and as dense punctate clusters (9). The extent and intensity of both forms of GAP-43-ir increase after adrenalectomy (ADX) (10, 11), similar to the post-ADX increase observed in the density of nerve fibers in the AP that are immunoreactive for the neuropeptides, substance P and calcitonin gene-related peptide (12). A recent study demonstrated that GAP-43-ir is indeed colocalized within substance P-containing fibers coursing throughout the rat AP and that these axons originate from the nodose ganglia (13). Consistent with these observations, we previously interpreted the post-ADX increase in the intensity and extent of fibrous and punctate GAP-43-ir in the AP as evidence of axonal elongation and sprouting of new axon terminals, respectively. Most interestingly, the putative terminals visualized as punctate clusters of GAP-43-ir were found to be associated almost exclusively with corticotrophs using confocal dual-label immunofluorescence microscopy (11). We now report that, when localized at the electron microscopic level, GAP-43-ir in the post-ADX AP is found within fine cellular processes originating from corticotrophs themselves. These processes can be seen extending to make specialized contacts with other gland cells. In addition, RT-PCR confirmed the presence of GAP-43 mRNA within the AP after ADX. We postulate that this novel nonneuronal expression of GAP-43 may stimulate process outgrowth and formation of cellular contacts by activated corticotrophs.
Materials and Methods
Animals
Male Holtzman albino rats of approximately 250 g body weight were obtained from Harlan Laboratories (Indianapolis, IN) and housed in the Montana State University Animal Resource Center, an Association Assessment and Accreditation of Laboratory Animal Care-accredited facility. All animals were housed under a 12-h light, 12-h dark cycle with ad lib access to lab chow and tap water. Bilateral adrenalectomy was performed under general anesthesia using a ketamine cocktail (26 mg ketamine + 5 mg xylazine + 0.8 mg acepromazine per kilogram body weight, im), incisions were closed with wound clips and treated with wound powder, and 0.05 mg/kg buprenorphine was administered sc to provide postsurgical analgesia. Rats undergoing bilateral adrenalectomy were provided with normal saline to drink instead of tap water throughout the postsurgical survival period. All experimental protocols used in these studies followed the guidelines in the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Montana State University Institutional Animal Care and Use Committee.
Tissue preparation
All animals to be used for immunocytochemistry were anesthetized with isoflurane and perfused intracardially with cold saline for 2 min followed by 500 ml of Nakane’s periodate-lysine-paraformaldehyde (PLP) fixative prepared immediately before use (14) and modified as previously described (9). Pituitaries were removed, postfixed overnight in PLP at 4 C, and then embedded in an egg yolk and gelatin matrix. The matrix block containing the tissue was then submerged for an additional 48 h in PLP at 4 C before sectioning on a Vibratome at 40 μm in the coronal plane.
Animals to be used for analysis of GAP-43 mRNA were decapitated under isoflurane anesthesia and the pituitary quickly removed and frozen on dry ice. A coronal slice of brain containing the piriform cortex and a small piece of liver were also removed and frozen from some animals. Tissues were wrapped in aluminum foil and stored in sealed containers at –80 C until use.
Immunoelectron microscopy
Localization of GAP-43 in free-floating sections of pituitary by peroxidase immunocytochemistry was performed as previously described (9) using sheep antirat GAP-43 at 1:1000 dilution as the primary antisera (generous gift of Dr. Larry Benowitz, Harvard Medical School, Boston, MA), biotinylated goat antisheep IgG at 1:200 dilution (Vector Laboratories, Burlingame, CA) as the secondary, and the Vector Elite ABC kit to produce the avidin-biotin-peroxidase complex. Binding of the ABC reagent was visualized using diaminobenzidine (Sigma, St. Louis, MO) as chromogen with generation of H2O2 by the glucose oxidase method (15). After immunocytochemical staining the tissues were postfixed in 5% glutaraldehyde in 0.1 M PO4 (pH 7.4) for 30 min followed by 4% OsO4 in PBS for 30 min. Sections were then dehydrated through increasing concentrations of ethanol to propylene oxide, infiltrated overnight in Epon/Araldite embedding medium (Ted Pella, Redding, CA), flat mounted between glass microscope slides pretreated with dimethyldichlorosilane (1% in benzene) to facilitate removal, and polymerized at 60 C for a minimum of 72 h. After examination in the light microscope, the top slide was removed and areas of interest in the section were cut out with a razor and glued to the flat surface of a blank Epon/Araldite bullet for ultrathin sectioning. Silver to gold ultrathin sections were counterstained with a combination of aqueous uranyl acetate and Reynolds lead citrate and viewed using an EM10C transmission electron microscope (Zeiss, New York, NY) at 60–80 kV.
For dual-ultrastructural localization of GAP-43 and ACTH, we combined the preembedding peroxidase technique described above with postembedding immunogold labeling of ACTH. After the preembedding procedures outlined above, ultrathin sections of pituitary mounted on 2 x 1 mm formvar/carbon-coated nickel slot grids (EMS Supply, Hatfield, PA) were initially rehydrated by floating on drops of Tris-buffered saline (TBS) (pH 7.4) for 15 min. This was followed by sequential incubation at 4 C on drops of 5% BSA in TBS for 1 h, rabbit antihuman ACTH overnight [Dako, Carpinteria, CA; 1:100 in incubation buffer consisting of 1% normal goat serum + 0.1% Tween 20 + 1% BSA in TBS (pH 8.2)], TBS, and 5 nm gold-conjugated goat antirabbit IgG for 2 h (Ted Pella; 1:100 in incubation buffer). Sections were then washed in TBS and double-distilled sterile filtered H2O, counterstained with a combination of aqueous uranyl acetate and Reynolds lead citrate, and viewed using a Zeiss EM10C transmission electron microscope at 80 kV. Negative controls were prepared in an identical manner but with omission of the 5 nm gold conjugated goat antirabbit IgG.
Detection of GAP-43 mRNA by RT-PCR and gel electrophoresis
RT-PCR was used to determine whether GAP-43 mRNA was present in extracts of anterior pituitary, brain (piriform cortex), and liver of intact and adrenalectomized rats. Total RNA was extracted from individual tissue samples, which were homogenized in 20 mM sodium acetate containing 4 M guanidinium isothyocyanate, 0.1 mM dithiothreitol, and 0.5% N-lauroylsarcosine (lysis solution; Ambion, Austin, TX) and then diluted with 2 vol 10 mM Tris buffer (pH 7.5) containing 1 mM EDTA. After phenol-chloroform extraction, total RNA was precipitated in isopropanol and resuspended in 20 μl nuclease-free water. Extracts were stored at –80 C until analyzed. RT-PCR was performed with approximately 0.5 μg of total RNA per tube using the Access RT-PCR kit (Promega, Madison, WI) according to the manufacturer’s specifications. We used previously published GAP-43 primer sequences that result in a PCR product of 708 bp that covers the entire coding region of rat GAP-43 (16). These were: upstream (5') tgctgtgctgtatgagaagaacc (3'); and downstream (5') ggcaacgtggaaagccgtttcttaaagt (3'). These primers also span an intron, providing a control for potential amplification of genomic DNA. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) primers used as positive controls were: upstream (5') tccaccaccctgttgctgta (3'); and downstream (3') accacagtccatgccatcac (5'). Amplification was performed in a thermal cycler (PerkinElmer, Boston, MA) using a cycle sequence of 94 C for 30 sec, 60 C for 1 min, and 68 C for 2 min. Samples were collected after 30, 35, or 40 cycles and loaded onto 1.5% agarose gels for electrophoreses.
Determination of relative GAP-43 mRNA levels by quantitative PCR
Quantitative PCR was used to measure relative levels of GAP-43 mRNA in extracts of anterior pituitary from intact and adrenalectomized rats. Total RNA was extracted using TRIzol reagent (Invitrogen, Carlsbad, CA) according to the manufacturer’s protocol and each sample adjusted to 1.0 μg/μl based on absorbance at 260 nm. Aliquots were run on a 1.7% denaturing agarose gel to verify quality of total RNA before performing reverse transcription of 3.5 μg of total RNA per sample using an oligo T(24) primer and Superscript II reverse transcriptase (Invitrogen) according to the manufacturer’s protocol. Reaction products were diluted to 80 ng/μl and used as templates for quantitative PCR performed in a Roto-Gene 3000 (Corbett Research, Sydney, Australia). Each 20-μl reaction tube contained 10 μl SybrGreen PCR master mix (Applied Biosystems, Foster City, CA), 1 μl template, 1 μl primers (7.5 pmol each), and 8 μl RNase-free water. Left and right primers for GAP-43 mRNA were each 20 bases in length (5'-tttcctctcctgtcctgctc-3', 5'-tggacttgggatctttcctg-3'; ordered from Sigma-Genosys, The Woodlands, TX) and were designed to produce a PCR product 112 bp in length spanning bases 974-1086 in the rat GAP-43 mRNA (accession NM_017195 in the NCBI database). Amplication of GAPDH mRNA was used as a normalization standard. Left and right primers for GAPDH mRNA were each 20 bases in length (5'-gtggacctcatggcctacat-3', 5'-tgtgagggagatgctcagtg-3'; ordered from Sigma-Genosys) and were designed to produce a PCR product 150 bp in length spanning bases 1813–1963 in the rat GAPDH mRNA (accession NM_017195). Both the GAP-43 and GAPDH PCR products were determined to be single bands of the predicted size by isolation on a 1.5% nondenaturing agarose gel.
Results
GAP-43-ir is present in corticotrophs after ADX
We previously reported that GAP-43-ir appears in both punctate clusters and occasional elongated axons scattered throughout the parenchyma of the AP of the intact adult rat (9). Both the extent and intensity of punctate staining are greatly increased within 4 d after ADX, and we observed that punctate clusters of GAP-43-ir visualized in the confocal microscope are colocalized almost exclusively with ACTH-positive corticotrophs after ADX (11). In the present study, we used preembedding peroxidase immunoelectron microscopy to further characterize punctate GAP-43-ir at the ultrastructural level in the post-ADX AP. We found GAP-43-ir localized exclusively to the cytoplasmic surface of the plasmalemma within a subset of endocrine cells (Fig. 1A, arrows). We consistently observed that the reaction product was distributed heterogeneously, with areas of dense reaction product separated by varying distances of nonlabeled plasmalemma. At 4 d after ADX, the GAP-43-ir cells ranged in shape from ovoid to stellate and were characterized by well-developed rough endoplasmic reticulum, numerous mitochondria, and sparse secretory granules arrayed in close juxtaposition to the cytoplasmic surface of the plasmalemma. Many of the secretory granules were of the immature form, having a halo surrounding a dense core. No specific association between secretory granules and GAP-43-ir was observed. The ultrastructure of these GAP-43-ir cells is consistent with that of corticotrophs that are degranulated after ADX (17, 18, 19).
To confirm the presence of GAP-43-ir in corticotrophs, we combined preembedding immunoelectron microscopy for GAP-43 with postembedding immunogold labeling for ACTH on sections prepared from 4-d post-ADX animals. We found that GAP-43-ir was invariably present only in cells containing ACTH-positive granules. Numerous secretory granules labeled with colloidal gold particles were observed along the periphery of each cell that contained GAP-43-ir (Fig. 1B).
GAP-43-ir is concentrated within fine processes that form specialized contacts with other cells
We frequently observed extensive processes of varying caliber emanating from the soma of GAP-43-ir cells. These processes contained numerous secretory granules and often projected for a substantial distance between surrounding cells. Extensive deposits of GAP-43-ir were associated with the cytoplasmic surface of the plasmalemma throughout the length of the processes (Fig. 2, arrows, and Fig. 3A). The density of reaction product was greater within processes than in any other region of the cell, and it is reasonable to conclude that these processes are the basis for the punctate clusters of GAP-43 staining seen in the light microscope. GAP-43-ir was also present in the region of the cell soma immediately adjacent to the base of these processes (Fig. 2), but a much smaller proportion of the plasmalemma was covered by reaction product within the soma than in the processes themselves.
We also observed rare specialized synaptic-like contacts between GAP-43-ir processes and distant cell bodies. These putative terminals contained extensive GAP-43-ir, secretory granules, and mitochondria. Figure 3 (A–C) shows a set of serial sections at increasingly higher magnifications collected through the extent of one such synaptic-like contact. The process can be seen extending beyond this specialized contact in a manner similar to a synapse en passant (Fig. 3C, arrows). Note that the plasmalemma of the recipient cell soma protrudes outward to form a distinct contact surface (Fig. 3, B and C, arrowheads) with the GAP-43-ir process. However, specializations typical of a neuronal postsynaptic membrane were not observed. Whereas the presence of secretory granules in the process immediately adjacent to the synaptic-like contact suggests that exocytosis of granule contents may occur at this site, we did not observe any granules actually captured in the process of exocytosis at the small number of contact sites examined.
Expression of GAP-43 mRNA is up-regulated in the AP after adrenalectomy
RT-PCR and gel electrophoresis.
GAP-43 mRNA was detected in extracts of the AP 4 d after ADX using RT-PCR, and absence of a detectable signal in similar AP extracts from intact rats suggested that expression of GAP-43 mRNA was up-regulated in response to ADX (Fig. 4). GAP-43 mRNA was detected in seven of eight ADX pituitaries examined but in none of six intact pituitaries. Liver and brain extracts were used as negative and positive tissue controls, respectively, for the GAP-43 signal, and GAPDH was used as a positive control for the RT-PCR. The GAP-43 product from both AP and brain was approximately 700 bp in length, as previously reported using identical primers (16), and was confirmed to be authentic rat GAP-43 by extraction from the gel and sequencing.
Quantitative PCR.
A significant increase in the abundance of GAP-43 mRNA in the AP 4 d after ADX was confirmed using quantitative PCR. Templates were prepared from total RNA extracts of the AP of three ADX and three age-matched intact rats. Quantitative PCR analysis was performed twice using triplicate samples from each animal in each run and the mean cycles to threshold calculated for each subject (Table 1). The average decrease of 2.61 in mean cycles to threshold for the ADX group was highly significant (Student’s t test two-tailed; t = 11.76, df = 4, P < 0.001) and corresponds to an approximately 6-fold increase in the level of GAP-43 mRNA in the AP.
Discussion
Our results demonstrate that GAP-43 mRNA and protein are expressed by intrinsic cells of the rat AP. Both GAP-43 mRNA abundance and GAP-43-ir are dramatically increased in the AP after ADX, and we found GAP-43-ir to be localized within ACTH-positive corticotrophs. GAP-43-ir is highly concentrated within extensive fine processes extending from the corticotroph soma, and in some cases these processes could be observed forming specialized synaptic-like contacts with other cells. We had previously interpreted the association of GAP-43-ir with corticotrophs at the light microscopic level as representing innervation of corticotrophs by axons entering the AP (11). We observed scattered GAP-43-ir axonal profiles within the AP, and these are presumably the substance P-containing fibers recently found by others to originate from the nodose ganglia (13). However, the present ultrastructural findings show that the extensive punctate GAP-43-ir seen associated with corticotrophs after ADX is not of neuronal origin but rather represents expression of GAP-43 by corticotrophs themselves.
Whereas GAP-43 was earlier considered to be a neuron-specific protein (2, 20), there is more recent evidence of its expression in additional cell types, primarily of nervous origin. These include oligodendroglia, Schwann cells, and some astrocytes (reviewed in Ref.21) as well as noradrenergic chromaffin cells of the adrenal medulla (reviewed in Ref.3). In addition, transient expression in mesoderm of embryonic chick limb (22) and in human embryonic (23) and regenerating muscle fibers (24) has been reported, and GAP-43-ir was observed in regenerating myocytes within biopsies of inflammatory human myopathies (25). However, we are unaware of any previous reports of GAP-43 expression in normal mature tissues outside the nervous system or adrenal medulla. Thus, the finding that GAP-43 is expressed in corticotrophs is novel and unexpected.
Whereas the localization of GAP-43 in corticotrophs was surprising, the increased abundance of GAP-43 mRNA and protein that we observed after ADX is consistent with studies showing that expression of GAP-43 is negatively regulated by glucocorticoids. This effect was first observed in PC12 cells (26) and subsequently demonstrated in the hippocampus by Chao and colleagues (27, 28), who found that GAP-43 mRNA levels increased in pyramidal neurons 7 d after ADX and that this increase could be blocked by corticosterone replacement. It is also possible that ACTH-like peptides could influence the activity of GAP-43, by either altering gene expression (29) or affecting proteolysis (30). Determining whether these or other mechanisms regulate GAP-43 expression in corticotrophs will require additional studies.
GAP-43 is implicated in two principal cell functions, process outgrowth and vesicular release. An extensive literature has correlated GAP-43 expression in neurons with neurite growth during development and regeneration of mature axons, both in vivo and in cell culture (reviewed in Refs.1, 2, 3, 20). Similar observations have been reported in Schwann cells (21). Induced expression of GAP-43 has also been shown to stimulate process extension in a variety of transfected cell lines, including PC12 (31), 3T3 (32), COS (32, 33), CHO (32), and AtT-20 (34) cells. The work of Gamby et al. (34) is especially relevant to the present findings because AtT-20 cells are of pituitary origin and secrete proopiomelanocortin-derived peptide hormones. Whereas GAP-43 was not detectable in the parental AtT-20 mouse pituitary tumor cell line, the D16 subclone was found to express high levels of GAP-43. Consistent with this difference, D16 cells had a flattened morphology and extended processes when attached to a laminin-coated substrate, whereas AtT-20 cells maintained a rounded morphology under similar conditions. In addition, transfection of AtT-20 cells with rat GAP-43 cDNA induced flattening and process extension similar to that seen in D16 cells.
These authors also reported marked effects of GAP-43 expression on secretion of -endorphin from cultured AtT-20 cells. The original AtT-20 line exhibited much less robust K+-induced secretion than did the GAP-43-positive D16 clone, but depolarization induced release of -endorphin was greatly stimulated in transfected AtT-20 cells expressing high levels of GAP-43 (34). A role for GAP-43 in stimulation of vesicular release from AP cells is consistent with numerous studies demonstrating that GAP-43 facilitates release of catecholamines and cholecystokinin-8 from both brain synaptosomes and adrenal chromaffin cells (reviewed in Ref.3). These actions may be mediated by association of GAP-43 with PI(4,5)P(2) plasmalemmal rafts, in which it has been shown to affect actin accumulation and dynamics (reviewed in Ref.35). GAP-43 is a protein kinase C substrate, and it has been proposed that changes in GAP-43 phosphorylation subsequent to modulation of protein kinase C activity may increase the ability of GAP-43 to affect actin and promote high levels of vesicular release under conditions of sustained activation (36). In addition, GAP-43 modulates calcium signaling by binding to calmodulin and reducing activation of calcium-calmodulin-dependent enzymes (reviewed in Refs.3, 37). Interestingly, Gamby et al. (38) found that the ability of GAP-43 to stimulate K+-induced secretion of -endorphin from AtT-20 cells was dependent on both calmodulin binding and membrane association of the protein but that its ability to alter cellular morphology depended on the latter interaction only.
The well-documented effects of GAP-43 expression on cell morphology and hormone secretion suggest that the up-regulation in GAP-43 abundance we observed in activated corticotrophs in vivo may serve to stimulate both process extension and intercellular signaling. The preferential localization of GAP-43-ir within fine processes extending far from the corticotroph somata supports this conclusion. Stellate corticotrophs are known to be highly plastic cells (reviewed in Ref.39). Growth of extensive new processes by activated corticotrophs occurs both in vivo after ADX (17, 40, 41) and cold stress (42) and in vitro in response to CRH (43). Process outgrowth can occur very rapidly, having been observed after only 3–10 min of exposure to CRH in culture (43) or after 30 min of cold stress (42), but extensive processes are also maintained for at least 6 wk and perhaps indefinitely under the sustained activation induced by ADX (41). We previously reported that GAP-43-ir associated with corticotrophs is dramatically increased at both 4 and 14 d after ADX, coinciding with the growth of extensive cell processes (11). GAP-43-ir is also increased in the rat AP in response to 1 h per day of restraint stress applied for 1 wk (Paden, C. M., and J. A. Watt, unpublished observations). Modest GAP-43-ir is present specifically in association with corticotrophs in intact animals (11), and we were able to detect GAP-43 mRNA in the AP of intact rats by quantitative PCR (Table 1). These observations suggest that constitutive expression of the protein at a reduced level could serve to prime corticotrophs for a rapid morphological response to activation. Subsequent up-regulation of GAP-43 expression may then serve to sustain process outgrowth and/or stimulate branching of processes (31) as well as to facilitate cell signaling. Further experiments are necessary to more precisely determine the temporal relation between increased GAP-43 expression and process outgrowth and maintenance.
The occasional synaptic-like contacts that we observed between corticotroph processes containing GAP-43 and other gland cells suggests that these processes are involved in intercellular communication within the AP and are not simply sites of hormone release into the vasculature. Several previous studies have described the extensive processes of activated corticotrophs as being in close contact with or even surrounding neighboring cells as well as extending to capillaries (17, 40, 41). Both the localization of GAP-43-ir processes within clusters of corticotrophs (11) and the ultrastructure of the postsynaptic cells (Fig. 4) suggest that these specialized contacts may be between pairs of corticotrophs. However, we have not been able to confirm the identity of the postsynaptic cells by immunocytochemistry. A previous report of corticotrophs forming synaptic-like contacts in the rat concluded that the morphology of the postsynaptic cells suggested they were lactotrophs (44), but similar synaptic-like contacts have been observed between pairs of corticotrophs in the dog (45). Growth of extensive interdigitated cell processes may serve to facilitate paracrine interactions between corticotrophs, whether or not specialized synaptic-like contacts are formed. This hypothesis is consistent with a variety of studies indicating that intercellular communication between corticotrophs can both stimulate and inhibit ACTH secretion in vivo and in vitro (46, 47, 48). One intriguing possibility is that paracrine interactions between corticotrophs may act to coordinate pulsatile release of ACTH, as suggested by the demonstration of an intrinsic pulse-generating mechanism underlying release of ACTH from perifused human pituitary (49). In this scenario increased expression of GAP-43 could play multiple roles, facilitating both vesicular release of ACTH and intercellular communication between activated corticotrophs.
Footnotes
This work was supported by National Science Foundation Grant IBN-9604670 (to C.M.P.) and National Institutes of Health, National Institute of Neurological Disorders and Stroke Grant 5 R01 NS032507 (to C.M.P.).
The authors have no conflicts of interest.
First Published Online November 3, 2005
Abbreviations: ADX, Adrenalectomy; AP, anterior pituitary; GAP, growth-associated protein; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GAP-43-ir, GAP-43 immunoreactivity; PLP, periodate-lysine-paraformaldehyde; TBS, Tris-buffered saline.
Accepted for publication October 26, 2005.
References
Benowitz LI, Routtenberg A 1997 GAP-43: an intrinsic determinant of neuronal development and plasticity. Trends Neurosci 20:84–91
Skene JHP 1989 Axonal growth-associated proteins. Annu Rev Neurosci 12:127–156
Oestreicher AB, De Graan PNE, Gispen WH, Verhaagen J, Schrama LH 1997 B-50, the growth associated protein-43: modulation of cell morphology and communication in the nervous system. Prog Neurobiol 53:627–686
Skene JH 1984 Growth-associated proteins and the curious dichotomies of nerve regeneration. Cell 37:697–700
Schreyer DJ, Pate Skene JH 1991 Fate of GAP-43 in ascending spinal axons of DRG neurons after peripheral nerve injury: delayed accumulation and correlation with regenerative potential. J Neurosci 11:3738–3751
Palacios G, Mengod G, Sarasa M, Baudier J, Palacios JM 1984 De novo synthesis of GAP-43: in situ hybridization histochemistry and light and electron microscopy immunocytochemical studies in regenerating motor neurons of cranial nerve nuclei in the rat brain. Brain Res Mol Brain Res 24:107–117
Benowitz LI, Rodriguez WR, Neve RL 1990 The pattern of GAP-43 immunostaining changes in rat hippocampal formation during reactive synaptogenesis. Brain Res Mol Brain Res 8:17–23
Bendotti C, Pende M, Samanin R 1994 Expression of GAP-43 in the granule cells of rat hippocampus after seizure-induced sprouting of mossy fibres: in situ hybridization and immunocytochemical studies. Eur J Neurosci 6:509–515
Paden CM, Moffett CW, Benowitz LI 1994 Innervation of the rat anterior and neurointermediate pituitary visualized by immunocytochemistry for the growth-associated protein GAP-43. Endocrinology 134:503–506
Lu CR, Meng FT, Benowitz LI, Ju G 1995 Evidence for axonal sprouting in the anterior pituitary following adrenalectomy in the rat. J Endocrinol 147:161–166
Paden CM, Babcock C, Conner KA, Duong DKT, Kuhl JM 1998 Axons containing the growth associated protein GAP-43 specifically target rat corticotrophs following adrenalectomy. J Neuroendocrinol 10:693–699
Ju G, Ma D, Fan S-C 1994 Response of calcitonin-gene-related peptide-like immunoreactive nerve fibers of the anterior pituitary to adrenalectomy in the rat. Neuroendocrinology 59:505–510
Vanhatalo S, Soinila S 2001 Evidence for nodose ganglion as the source of innervation to the anterior lobe of the pituitary gland. Neuroscience 107:491–497
McLean IW, Nakane PK 1974 Periodate-lysine-paraformaldehyde fixative—a new fixative for immunoelectron microscopy. J Histochem Cytochem 22:1077–1083
Itoh K, Konishi A, Nomura S, Mizuno N, Nakamura Y, Sugimoto T 1979 Application of coupled oxidase reaction to electron microscopic demonstration of horseradish peroxidase: cobalt-glucose oxidase method. Brain Res 175:341–346
Baetge EE, Hammang JP 1991 Neurite outgrowth in PC12 cells deficient in GAP-43. Neuron 6:21–30
Siperstein ER, Miller KG 1973 Hypertrophy of the ACTH-producing cell following adrenalectomy: a quantitative electron microscopic study. Endocrinology 93:1257–1268
Taniguchi Y, Shiino M 1990 Immunoelectron-microscopic and morphometric study of the rat corticotrophs after adrenalectomy. Acta Anat 137:45–48
Moriarty GC, Halmi NS 1972 Electron microscopic study of the adrenocorticotropin producing cell with the use of unlabeled antibody and the soluble peroxidase-antiperoxidase complex. J Histochem Cytochem 20:590–603
Benowitz LI, Perrone-Bizzozero NI 1991 The expression of GAP-43 in relation to neuronal growth and plasticity: when, where, how and why In: Gispen WH, Routtenberg A, eds. Progress in brain research. Amsterdam: Elsevier Science Publishers B.V.; 69–87
Sensenbrenner M, Lucas M, Deloulme J-C 1997 Expression of two neuronal markers, growth-associated protein 43 and neuron-specific enolase, in rat glial cells. J Mol Med 75:653–663
Stocker KM, Baizer L, Ciment G 1992 Transient expression of GAP-43 in nonneuronal cells of the embryonic chicken limb. Dev Biol 149:406–414
Moos T, Christensen LR 1993 GAP43 identifies developing muscle cells in human embryos. Neuroreport 4:1299–1302
Huess D, Schluep M 1998 Subcellular localization of phosphoprotein B-50 in regenerating muscle. An immuno-electron microscopic study. Neurol Res 20:360–364
Huess D, Engelhardt A, Gobel H, Neundorfer B 1995 Light-microscopic study of phosphoprotein B-50 in myopathies. Virchows Arch 426:69–76
Federoff HJ, Grabczyk E, Fishman MC 1988 Dual regulation of GAP-43 gene expression by nerve growth factor and glucocorticoids. J Biol Chem 263:19290–19295
Chao HM, McEwen BS 1994 Glucocorticoids and the expression of mRNAs for neurotrophins, their receptors and GAP-43 in the rat hippocampus. Brain Res Mol Brain Res 26:271–276
Chao HM, Sakai RS, Ma L-Y, McEwen BS 1998 Adrenal steroid regulation of neurotrophic factor expression in the rat hippocampus. Endocrinology 139:3112–3118
Zwiers H, Veldhuis HD, Schotman P, Gispen W-H 1976 ACTH, cyclic nucleotides, and brain protein phosphorylation in vitro. Neurochem Res 1:669–677
Coggins PJ, Zwiers H 1994 Detergents and peptides alter proteolysis and calmodulin binding of B-50/GAP-43 in vitro. J Neurochem 63:1491–1498
Jap Tjoen San ERA, Van Rozen AJ, Nielander HB, Oestreicher AB, Gispen W-H, Schotman P 1995 Expression levels of B-50/GAP-43 in PC12 cells are decisive for the complexity of their neurites and growth cones. J Mol Neurosci 6:185–200
Zuber MX, Goodman DW, Karns LR, Fishman MC 1989 The neuronal growth-associated protein GAP-43 induces filopodia in non-neuronal cells. Science 244:1193–1195
Strittmatter SM, Valenzuela D, Fishman MC 1994 An amino-terminal domain of the growth-associated protein GAP-43 mediates its effects on filopodial formation and cell spreading. J Cell Sci 107:195–204
Gamby C, Waage MC, Allen RG, Baizer L 1996 Growth-associated protein-43 (GAP-43) facilitates peptide hormone secretion in mouse anterior pituitary AtT-20 cells. J Biol Chem 271:10023–10028
Caroni P 2001 Actin cytoskeleton regulation through modulation of PI(4,5)P(2) rafts. EMBO J 20:4332–4336
Iannazzo L 2001 Involvement of B-50 (GAP-43) phosphorylation in the modulation of transmitter release by protein kinase C. Clin Exp Pharmacol Physiol 28:901–904
Slemmon JR, Feng B, Erhardt JA 2000 Small proteins that modulate calmodulin-dependent signal transduction: effects of PEP-19, neuromodulin, and neurogranin on enzyme activation and cellular homeostasis. Mol Neurobiol 22:99–113
Gamby C, Waage MC, Allen RG, Baizer L 1996 Analysis of the role of calmodulin binding and sequestration in neuromodulin (GAP-43) function. J Biol Chem 271:26698–26705
Childs GV 1992 Structure-function correlates in the corticotropes of the anterior pituitary. Front Neuroendocrinol 13:271–317
Westlund KN, Aguilera G, Childs GV 1985 Quantification of morphological changes in pituitary corticotropes produced by in vivo corticotropin-releasing factor stimulation and adrenalectomy. Endocrinology 116:439–445
Kubba MA, McNicol AM 1987 Stimulation of pituitary corticotrophs in the rat—ultrastructural studies. Virchows Arch B Cell Pathol 54:119–126
Sasaki F, Wu P, Rougeau D, Unabia G, Childs GV 1990 Cytochemical studies of responses of corticotropes and thyrotropes to cold and novel environmental stress. Endocrinology 127:285–297
Childs GV, Lloyd J, Unabia G, Rougeau D 1989 Growth and secretory responses of enriched populations of corticotropes. Endocrinology 125:2540–2549
Liu YY, Ju G 1998 Synaptoid contacts between gland cells of the anterior pituitary of the rat. Anat Rec 251:181–184
Ju G, Zhang X 1990 An electron microscopic study on substance P-like immunoreactive nerve fibers in the pars distalis of the adenohypophysis in the dog. Neuroscience 38:503–513
Schwartz J 1990 Evidence for intrapituitary intercellular control of adrenocorticotropin secretion. Mol Cell Endocrinol 68:77–83
Jia L-G, Canny BJ, Leong DA 1992 Paracrine communication regulates adrenocorticotropin secretion. Endocrinology 130:534–539
Giraldi FP, Cavagnini F 1998 Corticotropin-releasing hormone is produced by rat corticotropes and modulates ACTH secretion in a paracrine/autocrine fashion. J Clin Invest 101:2478–2484
Gambacciani M, Liu JH, Swartz WH, Tueros VS, Rasmussen DD, Yen SS 1987 Intrinsic pulsatility of ACTH release from the human pituitary in vitro. Clin Endocrinol (Oxf) 26:557–563(Charles M. Paden, John A. Watt, Tiffany )