Neurofilament Protein Subunits Research Articles

Alterations in levels of mRNAs coding for neurofilament protein subunits during regeneration

Animal models of neuronal injury can be used to explore mechanisms that regulate the expression of genes coding for cytoskeletal proteins and transmitter-related markers. In the present study, in situ hybridization was used to measure levels of messenger ribonucleic acid (mRNA) encoding each of the neurofilament subunits and β-tubulin in spinal motor neurons at intervals (4 to 56 days) following a unilateral crush of the sciatic nerve. Levels of β-tubulin mRNA increased (approximately twofold), peaked at 28 days postaxotomy, and returned to control values by 56 days postaxotomy. In contrast, levels of mRNA encoding neurofilament subunits were reduced and returned to control values at 56 days following the lesion. There were significant differences among relative levels of mRNAs coding for each subunit. Other studies have demonstrated that the ratio of pulse-labeled neurofilament subunits in motor axons remained unaltered during regeneration. Therefore, the ratios of neurofilament subunits in axons must be regulated at one of the steps that intervenes between the control of levels of mRNA and the anterograde axonal transport of assembled neurofilaments.

Neurofilament protein-triplet immunoreactivity in distinct subpopulations of peptide-containing neurons in the guinea-pig coeliac ganglion

A battery of polyclonal and monoclonal antibodies raised against the triplet of identified neurofilament protein subunits was used to investigate neurofilament protein immunoreactivity in neurons of the guinea-pig coeliac ganglion. Using optimal conditions of fixation and tissue processing for each antibody we found that only 20% of the postganglionic sympathetic neurons in the guinea-pig coeliac ganglion contain neurofilament protein-triplet immunoreactivity. Double labelling with neurofilament protein-triplet antibodies raised in different species demonstrated that all of these antibodies labelled the same population of neurons. Double labelling using mouse monoclonal antibodies against neurofilament proteins in combination with rabbit polyclonals to neuronal markers showed that neurofilament protein-triplet immunoreactivity is restricted to specific chemically coded subpopulations of noradrenergic neurons. Approximately 52% of neurons in the ganglion contain neuropeptide Y and are presumed vasomotor neurons projecting to blood vessels in the submucosa of the small intestine. Virtually none of the neuropeptide Y-containing neurons were labelled with neurofilament protein-triplet antibodies. Neurons that contain somatostatin (21%) project to the submucous ganglia of the small intestine. Approximately two-thirds of neurons containing somatostatin are immunoreactive for the neurofilament protein-triplet. The other postganglionic neurons in the ganglion (27%) project to the myenteric plexus of the small intestine and do not contain either neuropeptide Y or somatostatin. Approximately a quarter of these neurons were labelled with neurofilament protein-triplet antibodies. These results suggest that the neurofilament protein-triplet may not be an instrinsic component of the cytoskeleton of all neurons. Furthermore the idea of a chemical coding of neurons should be extended to cytoskeletal proteins. The finding that these neurofilament proteins are confined to specific neuronal subpopulations has important implications for the search for a role of the neurofilament protein-triplet in neurons, for the interpretation of classical neurohistological silver impregnation techniques which appear to stain only neurofilament protein-triplet-containing neurons, as well as for neuropathological conditions that may involve these proteins in disease processes.

Monoclonal antibody to neurofilament protein (SMI‐32) labels a subpopulation of pyramidal neurons in the human and monkey neocortex

A monoclonal antibody that recognizes a nonphosphorylated epitope on the 168 kDa and 200 kDa subunits of neurofilament proteins has been used in an immunohistochemical study of cynomolgus monkey (Macaca fascicularis) and human neocortex. This antibody, SMI-32, primarily labels the cell body and dendrites of a subset of pyramidal neurons in both species. A greater proportion of neocortical pyramidal neurons were SMI-32 immunoreactive (ir) in the human than in the monkey. SMI-32-ir neurons exhibited consistent differences in the intensity of their immunoreactivity that correlated with cell size. The cellular specificity of SMI-32 immunoreactivity suggests that a subpopulation of neurons can be distinguished on the basis of differences in the molecular characteristics of basic cytoskeletal elements such as neurofilament proteins. The size, density, and laminar distribution of SMI-32-ir neurons differed substantially across neocortical areas within each species and between species. Differences across cortical areas were particularly striking in the monkey. For example, the anterior parainsular cortex had a substantial population of large SMI-32-ir neurons in layer V and a near absence of any immunoreactive neurons in the supragranular layers. This contrasted with the cortical area located more laterally on the superior temporal gyrus, where layers III and V contained substantial populations of large SMI-32-ir neurons. Both areas differed significantly from the posterior inferior temporal gyrus, which was distinguished by a bimodal distribution of large SMI-32-ir neurons in layer III. Differences across human areas were less obvious because of the increase in the number of SMI-32-ir neurons. Perhaps the most notable differences across human areas resulted from shifts in the density of the larger SMI-32-ir neurons in deep layer III. A comparison between the species revealed that isocortical areas exhibited greater differences in their representation of SMI-32-ir neurons than primary sensory or transitional cortical areas. A comparison of distribution patterns of SMI-32-ir neurons across monkey cortical areas and data available on the laminar organization of cortical efferent neurons suggests that a common anatomic characteristic of this chemically identified subpopulation of neurons is that they have a distant axonal projection. Such correlations of cell biological characteristics with specific elements of cortical circuitry will further our understanding of the molecular and cellular properties that are critically linked to a given neuron's role in cortical structure and function.

Difference in phosphorylation of neurofilament proteins in motor neurons and spinal ganglion cells

The composition of the neurofilament proteins (NFPs) in neuronal perikarya was examined by two-dimensional (2-D) gel electrophoresis of isolated perikarya of bovine spinal motor neurons. The extent of phosphorylation of the high molecular weight subunit of NFP (NFP-H) was compared between motor and sensory neuronal perikarya in spinal cord and spinal ganglion by immunocytochemistry with monoclonal antibodies (MAbs) to NFP. Of the 23 MAbs used in this study, one MAb (82E10) was specific to the highly phosphorylated component of NFP-H examined by 2-D immunoblot whereas another MAb (3A8) was specific to NFP-H irrespective of its level of phosphorylation. Immunocytochemically, 82E10 did not stain the perikarya of bovine and rabbit spinal motor neurons but 3A8 stained the perikarya in both animal species. These findings are consistent with 2-D immunoblot of neuronal perikarya of bovine motor neurons isolated in bulk. As for the spinal ganglia, 82E10 stained many, but not all, perikarya of sensory neurons of both animal species. These results indicate that the extent of phosphorylation of NFP-H in the perikarya of most spinal ganglion cells is higher than that of motor neurons. These findings suggest that the rate of phosphorylation of NFP-H in perikarya or the axonal transport of NFP from perikarya to proximal axons is uniform in spinal motor neurons but variable in spinal ganglion cells.

Immunocytochemical localization of neurofilament protein subunits in the spiral ganglion of the adult rat

Spiral ganglion neurons from adult rats were treated with several monoclonal antibodies that react with neurofilaments (NFs) and NF subunits. An antibody against NFs used with immunocytochemical techniques showed a strong reaction with most neuron processes in the spiral ganglion, whereas only a few neurons presented a reaction. Using monoclonal antibodies against the 3 subunits, we obtained the same results with a small percentage of neurons labelled. From quantitative observations, reacting neurons showed the same percentage as and a smaller size than T II neurons observed with a more conventional method. This shows that reacting neurons are indeed T II neurons and that they can easily be differentiated by an accumulation of NFs in their perikaryon by well characterized commercially available antibodies.

Neurofilament triplet proteins of NB2a/d1 neuroblastoma: posttranslational modification and incorporation into the cytoskeleton during differentiation

Induction of axonal neuritogenesis in NB2a/d1 cells was associated with an increased content of neurofilament proteins (NFPs) by immunoblot analysis. The major NFP subunits in differentiated [NB2a(+)] cells included microheterogenous forms with apparent molecular weights of 200– 190 kDa (NFP-H), 143– 142 kDa (NFP-M) and 70 kDa (NFP-L) on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Only NFP-L was detected in cytoskeletal preparations of undifferentiated [NB2a(−)] cells. All three NFPs of NB2a(+) cells incorporated 32P-orthophosphate in intact cells. A 160/155 kDa NFP-H immunoreactive polypeptide in NB2a(−) and NB2a(+) cells represented a relatively unmodified form of the 200 kDa NFP-H, since dephosphorylation of the 200 kDa NFP-H in vitro with alkaline phosphatase generated the 160/155 kDa forms. Triton-extracted NB2a(+) cells displayed NFP-H immunoreactivity in neurites and occasionally in perikaryal regions at the base of neurites. NFP-M was present throughout the neurites and somata of NB2a(+) cells, and was regularly detected in portions of perikarya in NB2a(−) cells. NFP-L immunoreactivity was distributed throughout the Triton-insoluble cytoskeleton of NB2a(−) and NB2a(+) cells. Immunocytochemical analyses revealed that extensively phosphorylated forms of NFP-H were largely restricted to the neurites of NB2a(+) cells, and less modified forms predominated throughout both perikarya and neurites of NB2a(−) and NB2a(+) cells.

Differential turnover of tubulin and neurofilament proteins in central nervous system neuron terminals

The transport of tubulin and neurofilament protein subunits from the preterminal axons of guinea pig retinal ganglion cells into their presynaptic terminals in the superior colliculus was examined. Newly synthesized tubulin and neurofilament proteins were radiolabeled with tritiated amino acids in the cell bodies and were allowed to be axonally transported through the optic axons and into the terminals in the superior colliculi. Superior colliculi were harvested at appropriate times, synaptosomes were prepared, and radiolabeled proteins were examined by gel electrophoresis and fluoroaphy. Proteins in the radiolabeled synaptosomes were compared with those in the portion of the optic tract immediately proximal to the superior colliculus. Tubulin subunits entered the terminals by 100 days after intraocular labeling, and at least one isoform of tubulin appeared to persist as long as 400 days. Neurofilament proteins, despite the fact that they are axonally transported and delivered to the terminals in concert with the tubulin subunits, disappear rapidly upon entry into the terminals themselves.

Phosphorylation of neurofilament proteins by protein kinase C

The low molecular mass (70 kDa) subunit of neurofilaments (NF-L) contains at least three phosphorylation sites in vivo and is phosphorylated by multiple kinases in a site-specific manner [(1987) J. Neurochem. 48, S101; Sihag, R.K. and Nixon, R.A. submitted]. In this study, we observed that the three subunits of neurofilament proteins from retinal ganglion cell neurons are substrates for purified mouse brain protein kinase C. Two-dimensional α-chymotryptic phosphopeptide map analyses of the NF-L subunit demonstrated that protein kinase C phosphorylates four polypeptide sites, two of which incorporate phosphate when retinal ganglion cells are pulse-radiolabeled with [ 32P]orthophosphate in vivo.

MRNA levels of all three neurofilament proteins decline following nerve transection

The control of neurofilament (NF) protein gene expression was studied by determining and comparing the levels of mRNA to the heavy (NF-H), mid-sized (NF-M) and light (NF-L) NF protein subunits in rat dorsal root ganglia (DRG) following sciatic nerve transection. mRNA to NF-H (4.5 kb), to NF-M (3.4 kb) and to NF-L (2.5 and 4.0 kb) were identified in Northern blots and quantitated in dot blot analyses, using specific cDNA probes for each NF protein. Following transection, there was a coordinated decline in mRNA levels to all 3 NF proteins, beginning as early as 3 days following transection and continuing for at least 28 days. The early and co-terminal fall in mRNAs suggests that the 3 NF genes are regulated by common factor(s) and that the function of these factor(s) is influenced by the state of axonal continuity with the target organ.

Differential distribution of vimentin and neurofilament protein immunoreactivity in NB2a/d1 neuroblastoma cells following neurite retraction distinguishes two separate intermediate filament systems

Mouse NB2a/d1 cells assemble all 3 neurofilament protein subunits (NFPs) into the detergent-insoluble cytoskeleton and segregate phosphorylated forms of the 200-kDa subunit (NFP-H) within neurites when differentiation is induced with dibutyryl cyclic AMP (dbcAMP). Before and after differentiation, these cells also incorporate vimentin into both the perikaryal and neuritic cytoskeleton (Shea et al., 1988, Dev. Brain Res., submitted). To determine whether NFPs and vimentin constitute separate intermediate filament systems or exist as heteropolymers, we perturbed cytoskeletal architecture by inducing the retraction of neurites with colchicine. After cells were exposed to colchicine, vimentin immunoreactivity partitioned into perikarya in the form of fibrous whorls that did not cross-react with antisera to NFPs. By contrast, NFP immunoreactivity remained dispersed throughout the cell body following neurite retraction. We interpret these different responses to colchicine to indicate that NFPs and vimentin are assembled into separate intermediate filaments in NB2a/d1 cells.

Insulin growth factors regulate the mitotic cycle in cultured rat sympathetic neuroblasts.

While neuronal mitosis is uniquely restricted to early development, the underlying regulation remains to be defined. We have now developed a dissociated, embryonic sympathetic neuron culture system that uses fully defined medium in which cells enter the mitotic cycle. The cultured cells expressed two neuronal traits, tyrosine hydroxylase [L-tyrosine, tetrahydropteridine:oxygen oxidoreductase (3-hydroxylating); EC 1.14.16.2] and the neuron-specific 160-kDa neurofilament subunit protein, but were devoid of glial fibrillary acidic protein, a marker for non-myelin-forming Schwann cells in ganglia. Approximately one-third of the tyrosine hydroxylase-positive cells synthesized DNA in culture, specifically incorporating [3H]thymidine into their nuclei. We used this system to define factors regulating the mitotic cycle in sympathetic neuroblasts. Members of the insulin family of growth factors, including insulin and insulin-like growth factors I and II, regulated DNA synthesis in the presumptive neuroblasts. Insulin more than doubled the proportion of tyrosine hydroxylase-positive cells entering the mitotic cycle, as indicated by autoradiography of [3H]thymidine incorporation into nuclei. Scintillation spectrometry was an even more sensitive index of DNA synthesis, revealing a 4-fold insulin stimulation with an ED50 of 100 ng/ml. Insulin-like growth factor I was 100-fold more potent than insulin, whereas insulin-like growth factor II was less potent, suggesting that insulin growth factor type I receptors mediated the mitogenic responses. In contrast, the trophic protein nerve growth factor exhibited no mitogenic effect, suggesting that the mitogenic action of insulin growth factors is highly specific. Our observations are discussed in the context of the detection of insulin growth factors and receptors in the developing brain.

Binding of brain spectrin to the 70‐kDa neurofilament subunit protein

Brain spectrin, or fodrin, a major protein of the subaxolemmal cytoskeleton, associates specifically in in vitro assays with the 70-kDa neurofilament subunit (NF-L) and with glial filaments from pig spinal cord. As an initial approach to the identification of the fodrin-binding proteins, a crude preparation of neurofilaments was resolved by electrophoresis on SDS/polyacrylamide gels and then transferred to nitrocellulose paper, which was 'blotted' with 125I-fodrin. A significant binding of fodrin was observed on polypeptides of 70 kDa, 52 kDa and 20 kDa. These polypeptides were further purified and identified respectively as the NF-L subunit of neurofilaments, the glial fibrillary acidic protein (GFP) and the myelin basic protein. The binding of fodrin to NF-L was reversible and concentration-dependent. The ability of the pure NF-L and GFP to form filaments was used to quantify their association with fodrin. a) The binding of fodrin to reassembled NF-L was saturable with a stoichiometry of 1 mol fodrin bound/50 +/- 10 mol NF-L and an apparent dissociation constant Kd = 4.3 x 10(-7) M. b) The binding involved the N-terminal domain of the polypeptide chain derived from the [2-(2-nitrophenylsulfenyl)-3-methyl-3'-bromoindolenine] cleavage of NF-L. c) Binding occurred optimally at physiological pH (6.8-7.2) and salt concentrations (50 mM). d) Interestingly, calmodulin, a Ca2+-binding protein, which has been shown to bind to fodrin, was found to reinforce the binding of fodrin to the NF-L, at Ca2+ physiological concentrations. The binding of fodrin to pure neurofilaments was not affected by the presence of the 200-kDa (NF-H) and the 160-kDa (NF-M) subunits. The apparent dissociation constant for the binding of fodrin to NF-L in the pure NF was 1.0 x 10(-6) M with 1 mol fodrin bound/80 +/- 10 mol NF-L. Moreover, the binding of fodrin to GFP, demonstrated in blot assays, was confirmed by cosedimentation experiments. The apparent dissociation constant Kd for the fodrin binding was 2.8 x 10(-7) M and the maximum binding was 1 mol fodrin/55 +/- 10 mol GFP.

Cell-cell interactions modulate the responsiveness of PC12 cells to nerve growth factor.

The growth of PC12 cells on a collagen substratum or on monolayers of several non-neuronal cell types was studied by measuring nerve growth factor (NGF)-dependent increases in the expression of a 150 X 10(3) (Mr) neurofilament protein subunit and the membrane glycoprotein Thy-1. Both responses were found to be greatly suppressed in cultures of fibroblasts as compared to the C2 and G8-1 muscle cell lines and the C6 glioma cell line. This suppression was associated with an inhibition of NGF-dependent neuritic outgrowth from PC12 cells grown on fibroblast monolayers. There was no evidence that fibroblasts secrete soluble molecules that directly inhibit these responses or neutralize NGF. In addition, there was no difference in the neurofilament protein response from PC12 cells that had been treated with NGF prior to coculture, and the now primed PC12 cells readily extended axons over fibroblast monolayers. These data demonstrate that cell-cell and/or cell-matrix interactions can modulate biochemical responses to NGF and suggest that responsiveness of neuronal cells to environmental cues is not immutable. Control of the latter may be at the level of expression of receptor molecules for cell-surface- or matrix-associated macromolecules and a similar mechanism operating during development could play a role in growth cone guidance.

Phosphorylation of neurofilament proteins and chromatolysis following transection of rat sciatic nerve

States of phosphorylation of neurofilament proteins were examined in the perikarya of rat sensory and motor neurons between 3 and 28 d following either a distal transection [6-7 cm from the L4-L5 dorsal root ganglia (DRG)] or a proximal transection (1-2 cm from the L4-L5 DRG) of the sciatic nerve. Paraffin sections of the right (experimental) and left (control) L4 and L5 DRG from animals with unilateral transection of the right distal sciatic nerve were stained immunocytochemically with monoclonal antibodies to phosphorylation-dependent (NF-P), dephosphorylation-dependent (NF-dP), or phosphorylation-independent (NF-ind) epitopes on the largest (NF200), mid-sized (NF150), or smallest (NF68) neurofilament protein subunits. Increased immunoreactivity to NF-P on NF200 and NF150 was detected in experimental DRC at 10 d, peaking by 20 d, and declining to near control levels by 28 d. Conversely, immunoreactivity to NF-dP declined in experimental DRG beginning at 6 d, reaching a maximum decline at 10-16 d, and returning to near control levels by 28 d. Immunocytochemical changes were confirmed with biochemical studies on tissue homogenates that demonstrated an increase of immunoreactivity to NF-P and a decrease of reactivity to NF-dP in the experimental DRG. Changes in immunoreactivities to NF-P and NF-dP were observed only in the perikarya of large neurons and were closely associated with chromatolytic changes in these neurons. Marked enhancement of chromatolysis, as well as the immunoreactivities to NF-P and NF-dP, occurred following a proximal (left side) versus distal (right side) transection in the same animal.(ABSTRACT TRUNCATED AT 250 WORDS)

Posttranslational modification of neurofilament proteins by phosphate during axoplasmic transport in retinal ganglion cell neurons

The progressive modification of newly synthesized neurofilament proteins (NFPs) during axoplasmic transport in mouse retinal ganglion cell (RGC) neurons was studied after RGC perikarya were pulse-labeled with 32P-orthophosphate or radiolabeled amino acids. The 3 NFP subunits, H(igh), M(iddle), and L(ow), were among a group of axonally transported proteins that incorporated high levels of 32P. Covalent addition of phosphate slowed the electrophoretic mobility of H and M on SDS polyacrylamide gels and shifted the charge of all 3 subunits toward more acidic pH values, thereby providing an index of the phosphorylation state of this radiolabeled population of NFPs. NFPs were extensively phosphorylated before they entered axons at the optic nerve level, and continued to be modified during transport along RGC axons at the optic nerve and tract level. H and M exhibited charge shifts of 0.2-0.6 units toward a more acidic pH during axoplasmic transport. The charge modifications became more prominent when NFPs reached distal axonal levels, which may indicate regional differences in the activity of this modification process along axons. By contrast, the L subunit became more basic in charge, consistent with decreases in the phosphorylation state during transport. Additional observations (Nixon and Lewis, 1986) that a considerable proportion of phosphate groups initially added to L and M were later removed as neurofilaments advanced along RGC axons support the notion that the changing phosphorylation state of NFP subunits during axoplasmic transport reflects a dynamic equilibrium between phosphorylation and dephosphorylation events. Topographical remodeling of phosphate groups on NFPs during axoplasmic transport is proposed as a possible mechanism for coordinating interactions between neurofilaments and other constituents, as these elements are transported and integrated into the axonal cytoskeleton.

MERKEL CELL TUMOR COEXPRESSING CYTOKERATIN AND NEUROFILAMENT PROTEINS

Whorled filaments 10 nm in width were identified by anti-intermediate filaments antibodies in a Merkel cell tumor from a 52-year-old man. Immunohistochemical tests revealed that the tumor was stained with anti-keratin antibody and antibodies against the 68-kd and 200-kd subunits of neurofilament proteins but not antibody against the 150-kd subunit. This is the first reported case of Merkel cell tumor expressing a 200-kd subunit of neurofilament proteins.

Differential turnover of phosphate groups on neurofilament subunits in mammalian neurons in vivo.

The phosphorylation and dephosphorylation of specific proteins was demonstrated directly in the intact vertebrate nervous system in vivo. By exploiting the neurons' ability to segregate a select group of cytoskeletal proteins from most other phosphorylated constituents of the cell by axoplasmic transport, we were able to examine the dynamics of phosphate turnover on neurofilament proteins in mouse retinal ganglion cell neurons simultaneously labeled with [32P]orthophosphate and [3H]proline in vivo. Three [3H]proline-labeled neurofilament protein (NFP) subunits, designated H (160-200 kDa), M (135-145 kDa), and L (68-70 kDa), entered optic axons in a mole:mole ratio similar to that of isolated axonal neurofilaments, supporting the notion that newly synthesized NFPs are transported along axons as assembled neurofilaments. NFP subunits incorporated high levels of 32P before reaching axonal sites at the level of the optic nerve. As neurofilaments were transported along axons, however, many initially incorporated [32P]phosphate groups were removed. Loss of these phosphate groups occurred to a different extent on each subunit. A minimum of 50-60 and 35-40% of the labeled phosphate groups was removed in a 5-day period from the L and M subunits, respectively. By contrast, the H subunit exhibited relatively little or no phosphate turnover during the same period. Dephosphorylation of L in axons is accompanied by a decrease in its net state of phosphorylation; changes in the phosphorylation state of H and M, however, also reflect ongoing addition of phosphates to these polypeptides during axonal transport (Nixon, R.A., Lewis, S.E., and Marotta, C.A. (1986) J. Neurosci., in press). The possibility is raised that dynamic rearrangements of phosphate topography on NFPs represent a mechanism to coordinate interactions of neurofilaments with other proteins as these elements are transported and incorporated into the stationary cytoskeleton along retinal ganglion cell axons.

Calcium-activated neutral proteinase of human brain: subunit structure and enzymatic properties of multiple molecular forms.

Calcium-activated neutral proteinase (CANP) was purified 2,625-fold from postmortem human cerebral cortex by a procedure involving chromatography on diethylaminoethyl (DEAE)-cellulose, phenyl-Sepharose, Ultrogel AcA-44, and DEAE-Biogel A. The major active form of CANP exhibited a molecular weight of 94-100 kilodaltons (Kd) by gel filtration on Sephacryl 300 and consisted of 78-Kd and 27-Kd subunits. Two-dimensional gel electrophoresis resolved the small subunit into two molecular species with different isoelectric points. CANP degraded most human cytoskeletal proteins but was particularly active toward fodrin and the neurofilament protein subunits (145 Kd greater than 200 Kd greater than 70 Kd). The enzyme required 175 microM Ca2+ for half-maximal activation and 2 mM Ca2+ for optimal activity toward [methyl-14C]azocasein. Other divalent metal ions were poor activators of the enzyme, and some, including copper, lead, and zinc, strongly inhibited the enzyme. Aluminum, a neurotoxic ion that induces neurofilament accumulations in mammalian brain, inhibited the enzyme 47% at 1 mM and 100% at 5 mM. A second CANP form lacking the 27-Kd subunit was partially resolved from the 100-Kd heterodimer during DEAE-Biogel A chromatography. The 78-Kd monomer exhibited the same specific activity, calcium ion requirement, pH optimum, and specificity for cytoskeletal proteins as the 100-Kd heterodimer, suggesting that the 27-Kd subunit is not essential for the major catalytic properties of the enzyme. The rapid autolysis of the 27-Kd subunit to a 18-Kd intermediate when CANP is exposed to calcium may explain differences between our results and previous reports, which describe brain mCANP in other species as a 76-80-Kd monomer or a heterodimer containing 76-80-Kd and 17-20-Kd subunits. The similarity of the 100-Kd human brain CANP to CANPs in nonneural tissues indicates that the heterodimeric form is relatively conserved among various tissues and species.

Multiple calcium-activated neutral proteinases (CANP) in mouse retinal ganglion cell neurons: specificities for endogenous neuronal substrates and comparison to purified brain CANP

Calcium-activated neutral proteinases (CANPs) and their specificities for axonally transported proteins were studied within intact axons of mouse retinal ganglion cell (RGC) neurons in vitro. Two CANP activities with markedly different properties were identified. CANP B, at endogenous calcium levels, selectively cleaved the 145,000 Da (145 kDa) neurofilament protein subunit to yield 143 and 140 kDa neurofilament proteins that are also major constituents of the axonal cytoskeleton. This process represents a posttranslational modification of the neurofilament protein subunit rather than the initial step in its degradation (Nixon et al., 1982, 1983). A second calcium-activated neutral proteinase activity, CANP A, appeared only when calcium levels in the incubating medium were 100 microM or higher. CANP A degraded most proteins in RGC axons but acted considerably more rapidly on high-molecular-weight species. In particular, a 290-320 kDa protein in the Group IV (SCb) phase of axoplasmic transport was degraded 3 X faster than other major axonal proteins, including neurofilament proteins and fodrin. When maximally expressed, CANP A activity represented an enormous proteolytic potential in RGC axons--more than 50% of the total axonal content of proteins larger than 60 kDa could be hydrolyzed within 5 min. The calcium requirements, inhibitor profile, and substrate specificity of CANP A were similar to those of mCANP, the major CANP of mouse brain purified to homogeneity, suggesting that these enzymes may be the same or highly related proteins. The existence in a single neuron type of two CANP activities with markedly different substrate specificities and enzymatic properties emphasizes the possible functional diversity of calcium-activated neutral proteinases in neurons. These functions include the posttranslational modification, as well as degradation of neuronal proteins.

Multiple fates of newly synthesized neurofilament proteins: evidence for a stationary neurofilament network distributed nonuniformly along axons of retinal ganglion cell neurons.

We have studied the fate of neurofilament proteins (NFPs) in mouse retinal ganglion cell (RGC) neurons from 1 to 180 d after synthesis and examined the proximal-to-distal distribution of the newly synthesized 70-, 140-, and 200-kD subunits along RGC axons relative to the distribution of neurofilaments. Improved methodology for intravitreal delivery of [3H]proline enabled us to quantitate changes in the accumulation and subsequent decline of radiolabeled NFP subunits at various postinjection intervals and, for the first time, to estimate the steady state levels of NFPs in different pools within axons. Two pools of newly synthesized triplet NFPs were distinguished based on their kinetics of disappearance from a 9-mm "axonal window" comprising the optic nerve and tract and their temporal-spatial distribution pattern along axons. The first pool disappeared exponentially between 17 and 45 d after injection with a half-life of 20 d. Its radiolabeled wavefront advanced along axons at 0.5-0.7 mm/d before reaching the distal end of the axonal window at 17 d, indicating that this loss represented the exit of neurofilament proteins composing the slowest phase of axoplasmic transport (SCa or group V) from axons. About 32% of the total pool of radiolabeled neurofilament proteins, however, remained in axons after 45 d and disappeared exponentially at a much slower rate (t 1/2 = 55 d). This second NFP pool assumed a nonuniform distribution along axons that was characterized proximally to distally by a 2.5-fold gradient of increasing radioactivity. This distribution pattern did not change between 45 and 180 d indicating that neurofilament proteins in the second pool constitute a relatively stationary structure in axons. Based on the relative radioactivities and residence time (or turnover) of each neurofilament pool in axons, we estimate that, in the steady state, more neurofilament proteins in mouse RGC axons may be stationary than are undergoing continuous slow axoplasmic transport. This conclusion was supported by biochemical analyses of total NFP content and by electron microscopic morphometric studies of neurofilament distribution along RGC axons. The 70-, 140-, and 200-kD subunits displayed a 2.5-fold proximal to distal gradient of increasing content along RGC axons. Neurofilaments were more numerous at distal axonal levels, paralleling the increased content of NFP.(ABSTRACT TRUNCATED AT 400 WORDS)