Isolation and characterization of myosin from the adrenal medulla

Myosins II and V in chromaffin cells: myosin V is a chromaffin vesicle molecular motor involved in secretion

The presence of myosin II and V in chromaffin cells and their subcellular distribution is described. Myosin II and V distribution in sucrose density gradients showed only a strong correlation between the distribution of myosin V and secretory vesicle markers. Confocal microscopy images demonstrated colocalization of myosin V with dopamine beta-hydroxylase, a chromaffin vesicle marker, whereas myosin II was present mainly in the cell cortex. Cell depolarization induced, in a Ca2+ and time-dependent manner, the dissociation of myosin V from chromaffin vesicles suggesting that this association was not permanent but determined by secretory cycle requirements. Myosin II was also found in the crude granule fraction, however, its distribution was not affected by cell depolarization. Myosin V head antibodies were able to inhibit secretion whereas myosin II antibodies had no inhibitory effect. The pattern of inhibition indicated that these treatments interfered with the transport of vesicles from the reserve to the release-ready compartment, suggesting the involvement of myosin V and not myosin II in this transport process. The results described here suggest that myosin V is a molecular motor involved in chromaffin vesicle secretion. However, these results do not discard an indirect role for myosin II in secretion through its interaction with F-actin networks.

Molecular motors involved in chromaffin cell secretion

Neurosecretory cells, including chromaffin cells, possess a mesh of filamentous actin underneath the plasma membrane. It has been proposed that filamentous actin network separates the secretory vesicles into two compartments: the reserve pool and the release-ready vesicle pool. Disassembly of chromaffin cell cortical filamentous actin in response to stimulation allows the movement of vesicles from the reserve pool into the release-ready vesicle pool. Electron microscopy of cytoskeletons revealed the presence of polygonal areas almost devoid of actin filaments in stimulated cells. The percentage of stimulated cells showing disrupted cytoskeleton correlates well with the increase in secretion in these cells. Fine filaments also remain in these areas of disassembly, and these reacted with actin antibodies, as demonstrated by immunogold staining. In addition, the movement of vesicles between pools requires Ca(2+) and ATP, a condition for activation of a molecular motor. Confocal microscopy images demonstrated colocalization of myosin Va with dopamine-beta-hydroxylase. Cell depolarization induced the dissociation of myosin Va from chromaffin vesicles. 2,3-Butadione-2-monoxime (BDM), an inhibitor of myosin ATPase, inhibited secretion, suggesting a blockage for chromaffin vesicle transport between the reserve pool and the release-ready vesicle pool. On the other hand, myosin II subcellular distribution was not affected by cell depolarization. Confocal microscopy images show myosin II to be localized in the cell cortex and in some perinuclear structures. Chromaffin vesicles were not stained by myosin II antibody.

Dynamic changes in chromaffin cell cytoskeleton as prelude to exocytosis

Earlier work by us as well as others has demonstrated that filamentous actin is mainly localized in the cortical surface of chromaffin cell. This F-actin network acts as a barrier to the chromaffin granules, impeding their contact with the plasma membrane. Chromaffin granules contain alpha-actinin, an anchorage protein that mediates F-actin association with these vesicles. Consequently, chromaffin granules crosslink and stabilize F-actin networks. Stimulation of chromaffin cell produces disassembly of F-actin and removal of the barrier. This interpretation is based on: (1) Cytochemical experiments with rhodamine-labeled phalloidin indicated that in resting chromaffin cells, the F-actin network is visualized as a strong cortical fluorescent ring; (2) Nicotinic receptor stimulation produced fragmentation of this fluorescent ring, leaving chromaffin cell cortical areas devoid of fluorescence; and (3) These changes are accompanied by a decrease in F-actin, a concomitant increase in G-actin, and a decrease in the F-actin associated with the chromaffin cell cytoskeleton (DNAse I assay). We also have demonstrated the presence in chromaffin cells of gelsolin and scinderin, two Ca(2+)-dependent actin filament-severing proteins, and suggested that chromaffin cell stimulation activates scinderin with the consequent disruption of F-actin networks. Scinderin, a protein recently isolated in our laboratory, is restricted to secretory cells and is present mainly in the cortical chromaffin cell cytoplasm. Scinderin, which is structurally different from gelsolin (different pIs, amino acid composition, peptide maps, and so on), decreases the viscosity of actin gels as a result of its F-actin-severing properties, as demonstrated by electron microscopy. Stimulation of chromaffin cells either by nicotine (10 microM) or high K+ (56 mM) produces a redistribution of subplasmalemmal scinderin and actin disassembly, which preceded exocytosis. The redistribution of scinderin and exocytosis is Ca(2+)-dependent and is not mediated by muscarinic receptors. Furthermore, our cytochemical experiments demonstrate that chromaffin cell stimulation produces a concomitant and similar redistribution of scinderin (fluorescein-labeled antibody) and F-actin (rhodamine phalloidin fluorescence), suggesting a functional interaction between these two proteins. Stimulation-induced redistribution of scinderin and F-actin disassembly would produce subplasmalemmal areas of decreased cytoplasmic viscosity and increased mobility for chromaffin granules. Exocytosis sites, evaluated by antidopamine-beta-hydroxylase (anti-D beta H) surface staining, are preferentially localized in plasma membrane areas devoid of F-actin.

Phosphorylation of myosin light chain from adrenomedullary chromaffin cells in culture

The myosin-light-chain (MLC) phosphorylation accompanying catecholamine release in chromaffin cells was investigated with the objective of assessing the possible role of this contractile protein in catecholamine secretion. The electrophoretic characteristics of adrenomedullary MLC were determined by immunochemical techniques using two different specific antibodies. The identified 22 kDa phosphoprotein was mainly present in the cytosol, as demonstrated by ultracentrifugation and immunocytochemical analysis. A part of this protein was located on, or close to, the plasma membrane. Cell stimulation by secretagogues resulted in a Ca2(+)-dependent 32P incorporation into MLC, the time course of this process being related to catecholamine release. These findings were supported by a two-dimensional gel-electrophoretic analysis by which means this protein was resolved into two acidic forms. A role for Ca2(+)-calmodulin and Ca2(+)-phospholipid kinases in adrenomedullary MLC phosphorylation is reported. The results obtained suggest a regulatory role for such a protein in the underlying exocytotic event.

A two-dimensional electrophoresis study of phosphorylation and dephosphorylation of chromaffin cell proteins in response to a secretory stimulus

Phosphorylated proteins of bovine chromaffin cells, radioactively labeled with [32P]orthophosphate, have been analyzed by two-dimensional polyacrylamide gel electrophoresis and autoradiography. Complex two-dimensional electrophoretograms were studied with the aid of computer-assisted image analysis (CAIA). A database map of 32P-labeled proteins was constructed; approximately 500 polypeptides have been detected, numbered, and characterized according to the intensity of labeling, molecular weight, and isoelectric point. The database was constructed from cells kept in resting conditions or stimulated with 59 mM K+ in 2.5 mM Ca2+ or in 0 Ca2+ solution. These manipulations caused statistically significant changes in the degree of phosphorylation of 20 proteins; they were classified as Ca2+-dependent substrates for the phosphorylation or dephosphorylation processes. These changes were also shown in cells stimulated in the presence of the Ca2+ channel activator Bay K 8644. New proteins that show as much as a fivefold increase in their phosphorylation state during cell stimulation have been located with this methodology, as well as many others that had not previously been detected with conventional methods. These experiments provide the first CAIA database of chromaffin cell phosphoproteins; the map constructed with these data will allow the location of specific phosphoproteins and serve as a reference for future ongoing studies. The database will continue to grow to identify more proteins and to facilitate the comparison of complex patterns obtained in different laboratories for normal and transformed pheochromocytoma PC12 cells.

Phosphorylation and dephosphorylation of chromaffin cell proteins in response to stimulation

Stimulation of bovine chromaffin cell in culture changed (increased or decreased) the phosphorylation state of several proteins as examined by 32P incorporation. Enhanced phosphorylation of 22 protein bands as well as increased dephosphorylation of a 20.4 kilodaltons protein band was observed when extracts of cultured chromaffin cells stimulated by either acetylcholine or high K+ were subjected to mono-dimensional gel electrophoresis. For several protein bands, the degree of phosphorylation was larger in cells stimulated by acetylcholine than in those challenged by a depolarizing concentration of K+. The most affected phosphoproteins have apparent molecular weights of 14,800, 29,000, 33,000, 57,000 (tubulin subunit), 63,000 (tyrosine hydroxylase subunit) and 94,000. The presence of a low extracellular calcium concentration (0.5 mM Ca2+ plus 15 mM Mg2+) in the incubation medium inhibited (38-100%) the acetylcholine-evoked increases in protein phosphorylation observed previously for 18 protein bands. Trifluoperazine at the concentration required for 50% inhibition of acetylcholine-induced catecholamine release decreases (33-100%) the stimulation-induced phosphorylation in all polypeptides, with the exception of the 14.8 kilodaltons and the dephosphorylated 20.4 kilodaltons components which were not affected. Two-dimensional gel electrophoresis analysis revealed that exposure of chromaffin cells to acetylcholine produced two types of effect on protein phosphorylation: activation of protein kinase activities affecting about 30 polypeptides; activation of protein phosphatase activities resulting in the dephosphorylation of about 40 polypeptides, most of them appearing as minor phosphoproteins, with the exception of the alpha-subunit of pyruvate dehydrogenase and the 20.4 kilodaltons polypeptide. On the basis of their molecular properties (molecular weight and pI) and their abundance in chromaffin cells, the 80 kilodaltons phosphoprotein which focused at pI 4.8 and the 117.5 kilodaltons phosphoprotein which focused at pI 5.0 were identified as chromogranins A and B, respectively. The relationship between acetylcholine-induced protein phosphorylation (or dephosphorylation) and catecholamine secretion was also investigated. The time course of protein phosphorylation (or dephosphorylation) paralleled or preceded [3H]noradrenaline release for 16 phosphoproteins.(ABSTRACT TRUNCATED AT 400 WORDS)

Characterization of myosin light-chain kinase from bovine adrenal medulla

Partially purified bovine adrenal medullary myosin light-chain kinase (MLCK) possesses a Stoke's radius of 79 A and a sedimentation coefficient of 3.95 +/- 0.45 S, yielding a native molecular weight of 150,000 +/- 17,000 g/mol and a frictional ratio of 2.24. It catalyzes the phosphorylation of the isolated light chain of skeletal muscle myosin and the light chain of intact adrenal medullary myosin, but not phosphorylase b or histone. The activation of MLCK by calmodulin is specific and dose dependent, yielding a K0.5 value of 9.0 nM; the dose response curve with respect to free Ca2+ is biphasic, exhibiting a stimulatory phase at low free Ca2+ concentrations (K0.5 = 0.17 microM) and an inhibitory phase at higher free Ca2+ concentrations (400-3000 microM). Michaelis-Menten kinetics are observed for ATP, yielding a Km for ATP of 25 microM and a Vmax of 23.2 nmol/min/mg. However, positive cooperative kinetics are observed for the skeletal muscle myosin light chain, yielding a Hill coefficient of 3.57, a K0.5 for light chain of 27 microM and a Vmax of 16.6 nmol/min/mg. A stoichiometry of phosphorylation of approximately 1 mol of phosphate/mol of skeletal muscle myosin light chain was observed. Therefore, adrenal medullary MLCK is similar in most, but not all, of its physical and kinetics properties to MLCKs isolated from other sources and may serve to regulate actin-myosin contractile activity in the adrenal medulla.

Secretory cell actin-binding proteins: identification of a gelsolin-like protein in chromaffin cells

Chromaffin cells, secretory cells of the adrenal medulla, have been shown to contain actin and other contractile proteins, which might be involved in the secretory process. Actin and Ca++-sensitive actin-binding proteins were purified from bovine adrenal medulla on affinity columns using DNase-I as a ligand. Buffers that contained decreasing Ca++ concentrations were used to elute three major proteins of 93, 91, and 85 kD. The bulk of the actin was eluted with guanidine-HCl buffer plus some 93- and 91-kD proteins. These Ca++-sensitive regulatory proteins were shown to inhibit the gelation of actin using the low-shear falling ball viscometer and by electron microscopy. Actin filaments were found to be shortened by fragmentation. Using antibody raised against rabbit lung macrophage gelsolin, proteolytic digestion with Staphylococcus V8 protease and two-dimensional gel electrophoresis, the 91-kD actin-binding protein was shown to be a gelsolin-like protein. The 93-kD actin-binding protein also showed cross-reactivity with anti-gelsolin antibody, similar peptide maps, and a basic-shift in pHi indicating that this 93-kD protein is a brevin-like protein, derived from blood present abundantly in adrenal medulla. Purification from isolated chromaffin cells demonstrated the presence of 91- and 85-kD proteins, whereas the 93-kD protein was hardly detectable. The 85-kD protein is not a breakdown product of brevin-like or gelsolin-like proteins. It did not cross-react with anti-gelsolin antibody and showed a very different peptide map after mild digestion with V8 protease. Antibodies were raised against the 93- and 91-kD actin-binding proteins and the 85-kD actin-binding protein. Antibody against the 85-kD protein did not cross-react with 93- and 91-kD proteins and vice versa. In vivo, the cytoskeleton organization of chromaffin secretory cells is not known, but appears to be under the control of the intracellular concentration of free calcium. The ability of calcium to activate the gelsolin-like protein, and as shown elsewhere to alter fodrin localization, provides a mechanism for gel-sol transition that might be essential for granule movement and membrane-membrane interactions involved in the secretory process.

Adrenal chromaffin cell calmodulin: its subcellular distribution and binding to chromaffin granule membrane proteins

Bovine adrenal medullae were homogenized in the presence or in the absence of EGTA and different subcellular fractions were prepared by differential and density gradient centrifugations. In the presence of the chelating agent, 69% of the total calmodulin, measured by radioimmunoassay, was present in the cytosol; the rest was bound to different membrane-containing fractions (nuclei, microsomal, and crude granule fraction). When the chelating agent was omitted, 43% of the calmodulin was present in the cytosol, the remaining calmodulin being membrane-bound. Further resolution of the crude granule fraction by sucrose density centrifugation demonstrated that the distribution of calmodulin in the density gradient was similar to the distribution of chromaffin granules rather than to that of mitochondria, Golgi elements, and lysosomes. In this case, there was also more calmodulin bound to chromaffin granules when EGTA was omitted from the density gradient. Experiments with 125I-calmodulin indicated the presence of high-affinity binding sites (KD = 1.3 X 10(-8) M; Bmax = 30 pmol/mg protein) for calmodulin in chromaffin granule membranes. Further, photoaffinity crosslinking experiments with 125I-calmodulin followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography indicated the presence of three calmodulin-binding polypeptide complexes (84,000; 41,000; and 38,000 daltons) in chromaffin granule membranes. These polypeptides were not labelled when either Ca2+ was omitted or an excess of nonradioactive calmodulin was present in the photolysis buffer, indicating the Ca2+ dependency and the specificity of the interaction. On the basis of the results described, it is suggested that the cellular levels of Ca2+ control the cellular distribution of calmodulin and its binding to specific chromaffin granule membrane proteins. Further, it is also suggested that the interactions between calmodulin and granule proteins might play a role in stimulus-secretion coupling.

Trifluoperazine, a calmodulin inhibitor, blocks secretion in cultured chromaffin cells at a step distal from calcium entry

Trifluoperazine, a calmodulin antagonist, inhibited the secretory response of cultured bovine adrenal medullary chromaffin cells to acetylcholine (10(-4) M) or a depolarizing concentration of [K+] (56 mM KCl) in a dose-related fashion. The ID50s of this effect were 2 x 10(-7) M and 2.2 x 10(-6) M for acetylcholine and high [K+], respectively. A decrease in external [Ca2+] concentration of the incubation medium from 4.4 to 0.275 mM resulted in an increase in the percentage of inhibition produced by trifluoperazine on the acetylcholine-evoked secretory response from 20.7 to 96.5%, respectively. However, trifluoperazine inhibited the acetylcholine-evoked catecholamine output by a similar absolute magnitude for all [Ca2+] concentrations tested with the exception of 4.4 mM [Ca2+]. Trifluoperazine, unlike the [Ca2+] channel blocker Ni2+, in concentrations (10(-6)-10(-5) M) that were found to inhibit significantly [K+]-induced amine output did not modify [K+]-induced 45Ca uptake or 45Ca efflux. However, trifluoperazine at a concentration of 2.5 x 10(-5) M was found to produce a small decrease in the 45Ca efflux curve and a decrease in the [K+]-evoked 45Ca uptake of 30 +/- 14% (n = 6). In addition, 2.5 x 10(-6) M trifluoperazine, a concentration which was found to suppress high [K+]-induced amine release by 64 +/- 5%, did not inhibit the 45Ca2+-Ca2+ exchange mechanism. These results demonstrate that trifluoperazine, an antipsychotic agent with anticalmodulin activity, blocks catecholamine release from cultured chromaffin cells at a step distal from calcium entry and, consequently, suggests a role for calmodulin in the secretory process of these cells.

Immunohistochemical and immunocytochemical localization of myosin, chromogranin A and dopamine-beta-hydroxylase in nerve cells in culture and in adrenal glands

Dopamine-beta-hydroxylase, chromogranin A and myosin were purified from bovine adrenal medulla and antibodies prepared against these proteins. Indirect immunocytochemical methods were used to localize dopamine-beta-hydroxylase, chromogranin A and myosin in bovine adrenal medulla and myosin in rat adrenal glands and cells from rat C.N.S. maintained in primary culture. Dopamine-beta-hydroxylase and chromogranin A were found in chromaffin granules, in agreement with biochemical data and, using electron microscopy, dopamine-beta-hydroxylase was found within the matrix and in the surrounding membrane of the storage granule, whereas chromogranin A was confined to the granule matrix. Myosin was localized in the vascular system irrigating adrenal glands, fibroblasts lining the vessels and chromaffin cells. In chromaffin cells, staining was found at the cell boundaries and electron microscopy showed myosin to be associated with the plasma membrane. Faint immunocytochemical staining by antimyosin antibodies was observed around certain exocytotic profiles but particular association with such structures was not demonstrable. Myosin localization was also studied in bovine adrenal cortex, where it was found in vascular channels and faintly in adrenal cortical cells, as in rat adrenal cortex and medulla, where identical patterns were obtained. In neuronal and glial cells dissociated from 13 day rat embryo cerebral hemispheres and cultured for 48 h, localization of myosin was studied using immunohistochemistry. The neuritic expansions and growth cones of neurons were fluorescent, whereas in glial cells, filamentous networks were visualized enclosing the nucleus and as long fibres traversing the entire cytoplasm.

Immunofluorescent localization of microfilaments in neuronal and glial primary cell cultures with antibody against adrenal medullary myosin

The localization of myosin was studied in rat neuronal and glial cell maintained in primary culture, using the double antibody immunofluorescent method. Antibodies were raised against myosin purified from bovine adrenal medulla. Myosin-specific immunoreactivity was found in the cell body and neurites of neuronal cells and in the cytoplasm of glial cells. In the former no typical substructure was observable, whilst in the latter myosin-rich filaments were found forming either a cage entrapping the nucleus or as long cables in cellular morphogenic expansions.

Some observations on bipolar filaments formed by non-muscle myosins

Adrenal medullary and retinal myosins formed bipolar filaments in vitro. These filaments showed features suggesting flexibility in the rod region of the myosin molecules composing such filaments; in certain cases the myosin heads spread away from the filament backbone, in others the backbone itself was twisted. In addition the bare central backbone showed transverse striations.

Anti-myosin stains chromaffin cells

The presence in fixed chromaffin cells of antigenic sites for a myosin antibody was demonstrated using immunofluorescence techniques. Tests on viable cells showed that at least some of the antigenic sites seem to be localized on or close to the cell surface and explained the cell agglutination that occurred with the addition of the myosin antibody to cells isolated by a method described in this paper.

Purification and characterization of myosin from bovine retina

Myosin has been isolated from bovine retinae and characterised by its ATPase (ATP phosphohydrolase, EC 3.6.1.3) activity, its mobility in sodium dodecyl sulphate polyacrylamide gels and by electron microscopy. The purified myosin shows high ATPase activity in the presence of EDTA or Ca2+ and a low activity in the presence of Mg2+. The Mg2+-dependent ATPase activity is stimulated by rabbit skeletal muscle actin. The presumptive retinal myosin possesses a major component which has a mobility in sodium dodecyl sulphate polyacrylamide gel electrophoresis similar to that of the heavy chain of bovine skeletal muscle myosin. Electron microscopy showed retinal myosin to form bipolar filaments in 0.1 M KCl. It is concluded that the retina possesses a protein with enzymic and structural properties similar to those of muscle myosin.

Distribution of contractile proteins and adrenergic nerves in the adrenal gland of guinea-pig, rat and ox as revealed by immunofluorescence and the glyoxylic acid technique

Myosin and actin were localized in the adrenal gland, using antibodies against these proteins which were isolated from chicken gizzard. Myosin and actin were preferentially located in vascular walls including endothelial cells and in the capsule. In rat and guinea-pig adrenal cortex, the amount of contractile elements in vascular walls corresponded well to the density of adrenergic nerves as revealed with the glyoxylic acid method.