Phosphorylation of the Head Domain of Neurofilament Protein (NF-M)

In neurons the phosphorylation of neurofilament (NF) proteins NF-M and NF-H is topographically regulated. Although kinases and NF subunits are synthesized in cell bodies, extensive phosphorylation of the KSP repeats in tail domains of NF-M and NF-H occurs primarily in axons. The nature of this regulation, however, is not understood. As obligate heteropolymers, NF assembly requires interactions between the core NF-L with NF-M or NF-H subunits, a process inhibited by NF head domain phosphorylation. Phosphorylation of head domains at protein kinase A (PKA)-specific sites seems to occur transiently in cell bodies after NF subunit synthesis. We have proposed that transient phosphorylation of head domains prevents NF assembly in the soma and inhibits tail domain phosphorylation; i.e. assembly and KSP phosphorylation in axons depends on prior dephosphorylation of head domain sites. Deregulation of this process leads to pathological accumulations of phosphorylated NFs in the soma as seen in some neurodegenerative disorders. To test this hypothesis, we studied the effect of PKA phosphorylation of the NF-M head domain on phosphorylation of tail domain KSP sites. In rat cortical neurons we showed that head domain phosphorylation of endogenous NF-M by forskolin-activated PKA inhibits NF-M tail domain phosphorylation. To demonstrate the site specificity of PKA phosphorylation and its effect on tail domain phosphorylation, we transfected NIH3T3 cells with NF-M mutated at PKA-specific head domain serine residues. Epidermal growth factor stimulation of cells with mutant NF-M in the presence of forskolin exhibited no inhibition of NF-tail domain phosphorylation compared with the wild type NF-M-transfected cells. This is consistent with our hypothesis that transient phosphorylation of NF-M head domains inhibits tail domain phosphorylation and suggests this as one of several mechanisms underlying topographic regulation.

Neurofilament (NF) 1 proteins assemble into neuron-specific intermediate filaments, the predominant cytoskeletal component in large myelinated axons. Neurofilaments consist of three protein subunits, high (NF-H), middle (NF-M), and low (NF-L) molecular weight, each composed of three domains, an N-terminal head domain, an a-helix rich central rod domain, and C-terminal tail domains (1). NFs are obligate heteropolymers in vivo, with assembly depending on a core NF-L subunit interacting with NF-M or NF-H subunits (2)(3)(4). Although head and rod domains are essential for assembly, tail domains are dispensable but may regulate filament length (2,5).
NF protein phosphorylation plays a key role in the dynamic remodeling of cytoskeletal architecture during axonal growth, guidance, and synaptogenesis (6 -8). N-terminal head domains are phosphorylated by second messenger-dependent protein kinases, the cyclic AMP-dependent protein kinase (protein kinase A, PKA), and protein kinase C (PKC) (9 -12). The Cterminal tails of NF-M and NF-H, enriched in numerous Lys-Ser-Pro (KSP) repeats, on the other hand, are preferred substrates for the second messenger-independent kinases associated with neurofilaments, such as CKI (13) and GSK3 (14), and the proline-directed kinases Erk1/2 and cdk5 (15)(16)(17).
NF phosphorylation is topographically regulated within neurons. N-terminal head domain phosphorylation by PKA at Ser 55 in NFL or Ser 46 in NF-M inhibits assembly into a heteropolymer in vitro and in vivo (3,9,12,19,20). It has been suggested that this transient head domain phosphorylation in cell bodies inhibits NF polymerization and filament formation (4,21). Extensive phosphorylation of KSP repeats in the tail domain of NF-M and NF-H occurs primarily in axons (7,22,23). These result in sidearm formation, increased inter-neurofilament spacing, radial growth of axons, and increased conduction velocity (24 -26). More recent studies, however, of NF-H null mice, or mice transfected with a NF-H tailless mutant, suggest that neither NF-H nor its phosphorylated tail are essential to the above neuronal properties (27,28). Nevertheless, deregulation of NF-M and NF-H tail domain phosphorylation is correlated with some pathologies seen in neurodegenerative disorders such as amyotrophic lateral sclerosis (ALS) in which tail domain phosphorylation occurs abnormally in perikarya (29 -31).
In a model of the dynamics of neurofilament phosphorylation, we postulated that the transient phosphorylation of head domains in cell bodies prevented neurofilament assembly and tail domain phosphorylation (32). Only after dephosphorylation of the head domain in the axon hillock, as NFs assembled into filaments for transport in the axon, do tail domain KSP repeats become accessible to phosphorylation. The hypothesis predicts that phosphorylation of head domain sites of NF-M, for example by PKA, should inhibit phosphorylation of the tail domain KSP repeats. To test this hypothesis, we stimulated PKA phosphorylation of NF-M head domain sites with forskolin in primary cortical neurons and in NIH3T3 cells transfected with NF-M mutated in PKA-specific head domain sites to prevent phosphorylation. We showed that activation of PKA in cortical neurons inhibited growth factor-induced tail domain phosphorylation of endogenous NF-M by Erk1/2, despite significant activation of the MAPK cascade. The results of the transfection experiment were more compelling, because they showed that tail domain phosphorylation of transfected wild type NF-M was inhibited by forskolin and EGF stimulation of the MAPK cascade, whereas tail domain phosphorylation was not inhibited in cells transfected with the mutant NF-M. These findings are consistent with the hypothesis that phosphorylation of specific serine sites in the head domain of NF-M by PKA inhibits phosphorylation of KSP sites in the NF-M tail domain and poses a mechanism for topographic regulation of cytoskeletal protein phosphorylation in neurons.
Plasmids and Constructs-The full-length pcNDA-NF-M (wild type) construct was prepared as described previously (17). The head domain phosphorylation sites of NF-M were mutated by site-directed substitution of Ser 1 , Ser 23 , and Ser 46 with alanine according to the methods described previously (33,34). First, NF-M S23A was generated by polymerase chain reaction mutagenesis. 5Ј and 3Ј NF-M primers were used in combination with the mutagenic primers acccggtccgccttcagtcgtggtg and cacacgactgaaggcggcccgggt, respectively. The wild type NF-M was used as a template. Then, the NF-M mutant S1A/S23A was produced by using mutagenic primers S1A, aagatggaattcgcctacacgctg, and mutant NF-M S23A as a template. Finally, the NF-M S1A/S23A/S46A mutant was made by polymerase chain reaction mutagenesis using NF-M S1A/S23A as a template. 5Ј and 3Ј NF-M primers were used in combination with the mutagenic primers S46A, ggctcgcccgccaccgtgtcctcc, and ggaggacacggtggcgggcgagcc, respectively. The final PCR product was digested with EcoRI and NotI and inserted into a pGEM-T vector. After confirming the sequence by DNA sequencing, the triple substitution NF-M mutant was made by cloning the EcoRI/NotI fragment of the pGEM-T NF-M S1A/S23A/S46A mutant into the EcoRI/NotI site of CMV (pCDNA3.1) vector. Mutagenesis was verified by sequencing.
Primary Cultures of Cortical Neurons-Primary cultures of rat cortical neurons were prepared from E18 rat fetuses. Brains were dissected and dissociated with papain and DNase I. For immunohistochemistry, cells were plated at a density of 1.0 ϫ 10 5 /cells per cm 2 on glass coverslips precoated with 20 g/ml poly-L-lysine. For Western blot or immunoprecipitation, cells were plated at 5 ϫ 10 6 per each six wells on Nunc multiwell plates precoated with 20 g/ml poly-L-lysine. Cells were maintained in B27/neurobasal medium containing 2 mM glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin in a humidified atmosphere of 95% air and 5% CO 2 at 37°C for 7 days. Under these conditions the contamination of glial cells, measured by staining with glial fibrillary protein antibody, was less than 10%. After 7 days cells were starved in the presence of 0.2% B27/neurobasal medium overnight (to reduce any background stimulation by serum factors) and treated with 20 M forskolin or 100 M dibutyryl-cAMP, 50 ng/ml NGF, 2.5 M KT5720, or 10 M H89 (PKA inhibitors) and 50 M PD98059 (MEK inhibitor).
Cell Culture and Transfection-NIH3T3 cells were obtained from the American Type Culture Collection (Rockville, MD) and cultured in Dulbecco's modified Eagle's medium with 10% calf serum supplemented with 100 units/ml penicillin and 100 g/ml streptomycin. Cells were maintained at 37°C in a humidified atmosphere of 5% CO 2 . The cells were transiently transfected with wild type NF-M (wtNF-M) or mutant (serines 1, 23, and 46 and alanines 1, 23, and 46) NF-M (mtNF-M) expression constructs using FuGENE 6 transfection reagent (Roche Applied Science, Indianapolis, IN) according to the manufacturer's instructions. Twenty-four hours post-transfection, cells were starved in the presence of 0.2% calf serum overnight (to reduce any background stimulation by serum factors) and then treated with 25 ng/ml EGF, 20 M forskolin, or 100 M dibutyryl-cAMP, 2.5 M KT5720, or 10 M H89 (PKA inhibitors) and 50 M PD98059.
Immunoblotting-Western blot analysis was performed as described previously (35). In brief, cells were harvested by scraping from dishes and lysed in ice-cold lysis buffer (20 mM Tris, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% SDS, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerol phosphate, and 2.5 mM Navo4, supplemented with a mixture of protease inhibitors and 1 mM phenylmethylsulfonyl fluoride) by passing though a 21-gauge needle several times and incubating for 30 min on ice. After centrifugation for 20 min at 13,000 ϫ g at 4°C, the protein concentrations of the supernatants were determined using BCA protein reagent. An equal amount of total protein (25 g of protein/lane) was resolved on a 4 -20% SDS-polyacrylamide gel and blotted onto a polyvinylidene difluoride membrane. This membrane was incubated in blocking buffer containing 20 mM Tris-HCl (pH 7.4), 150 mM NaCl, and 0.1% (v/v) Tween 20 (TTBS) plus 5% dry milk (w/v) for 1 h at room temperature. This was followed by incubation overnight at 4°C in primary antibodies: anti-SMI-31 (phospho-NF-H and phospho-NF-M antibodies, 1:5000), RMO-281, a specific antibody to NF-M phosphoepitopes (1:3000), anti-NF-M C-terminal polyclonal antibody (phosphoindependent, 1:5000), and phospho-or phospho-independent Erk1/2 antibodies (1:2000 and 1:1000). The membranes were then washed four times in TTBS (5 min/each). This was followed by incubation in secondary antibody (goat-anti-mouse or goat anti-rabbit IgG (HϩL)-horseradish peroxidase conjugate at a dilution of 1: 3000) for 2 h at room temperature. Western blots were analyzed using the Enhanced Chemiluminescence (ECL) kit following the manufacturer's instructions (Amersham Biosciences).
Immunofluorescence-After NIH3T3 cells were transfected on glass coverslips coated with poly-L-lysine, cells were washed twice in PBS and fixed for 30 min at room temperature in 4% paraformaldehyde, PBS, and 10 mM EGTA and permeabilized (with 25 mM Tris, pH 7.4, 150 mM NaCl, and 0.2% Triton X-100) for 15 min. The coverslips were incubated overnight at 4°C with primary antibody, RMO-281, a specific antibody to NF-M phospho form (1:250). Antibody was diluted in PBS with 0.1% Triton X-100. After a wash in PBS (three times, 15 min each) the coverslips were incubated with fluorescein isothiocyanate-conjugated goat anti-mouse IgG and rhodamine-labeled goat anti-rabbit IgG secondary antibody for 1 h at room temperature, followed by three PBS washes, and mounted in aqueous medium. Fluorescent images were observed with a Zeiss LSM-410 laser-scanning confocal microscope. Images were processed and merged by using Adobe Photoshop software.
Protein Kinase A Assay-PKA was assayed by using Promega's SignaTECT TM cAMP-dependent PKA assay system that uses biotinylated Kemptide (LRRASLG), a peptide substrate derived from pyruvate kinase. This substrate is highly specific for PKA. A total reaction volume of 25 l of kinase assay mixture was used, containing 5ϫ PKA assay buffer (200 mM Tris-HCl, pH 7.4; 100 mM MgCl 2 ; 0.5 mg/ml BSA), PKA biotinylated peptide substrate, [␥-32 P]ATP mix, and enzyme sample in the presence or absence of 0.025 mM cAMP. The reaction was initiated by adding 5 l of the enzyme extract and incubated at 30°C for 5min. The reaction was terminated by adding 12.5 l of termination buffer (7.5 M guanidine hydrochloride), and 10 l from each tube was spotted onto a pre-numbered SAM 2TM membrane square. The membranes were washed four times in 2 M NaCl, four times in 2 M NaCl 1% H 3 PO 4 , and one time in de-ionized water then rinsed with 95% ethanol. The radioactivity was measured in a liquid scintillation counter. PKA-specific activity was expressed in picomoles of ATP/min/g of protein and calculated using the formula, enzyme-specific activity in picomoles of ATP/min/g of protein ϭ [(cpm reaction with substrate Ϫ cpm reaction without substrate ) ϫ 37.5]/(10 ϫ time min ϫ amount of protein in reaction pg ϫ specific activity of [␥-32 P]ATP). In this formula 37.5 is the sum of the reaction volume (25 l) plus the termination buffer volume (12.5 l). 10 is the volume in microliters of the sample spotted onto SAM 2TM membranes.
Specific activity of [␥-32 P]ATP ϭ ((37.5/5) X)/2500 (5 is the volume in microliters of the samples spotted onto SAM 2TM membranes without washing; X is the average cpm of the 5-l samples spotted onto SAM 2TM membranes without washing; and 2500 is the number of ATP in the reaction).
In Vitro Phosphorylation of NF-M-The NF-M expression vector was obtained from Dr. R. K. H. Liem (Columbia University, NY). The bacterially expressed full-length NF-M was purified as described previously (36). Phosphorylation of NF-M was performed basically as described previously (16,37). Briefly, phosphorylation was performed in a total volume of 50 l of reaction mixture containing 50 mM Tris-HCl, pH 7.4, 5 mM MgCl 2 , 1 mM each of sodium vanadate, EGTA, EDTA, dithiothreitol, and [␥-32 P]ATP, and 10 g of NF-M, plus 0.1 M okadaic acid or microcystin. 10 l of enzyme was added, and the mixture was incubated for 1 h. The reaction was terminated by the addition of 10% trichloroacetic acid. After washing 2ϫ with 10% trichloroacetic acid and 2ϫ with 90% alcohol, the pellet was dissolved in SDS-PAGE buffer. Phosphorylation was assayed by autoradiography. The phosphorylation of NF-M by PKA was induced in the presence of 10 l of catalytic subunit of PKA (Sigma, St. Louis, MO).

Phosphorylation of the NF-M Head Domain by PKA in Vitro
Inhibits NF-M Tail Domain Phosphorylation-To investigate the effect of PKA phosphorylation of the NF-M head domain on phosphorylation of the KSP repeats in the tail domain, we first studied NF-M phosphorylation by PKA and MAPK; i.e. constitutively active MEK/Erk1/2 in vitro. The kinases were used either alone or together in a standard kinase assay system using bacterial expressed NF-M protein as substrate (36). An autoradiograph of the assay is shown in Fig. 1A  It is important to note that the order of head and tail domain phosphorylation has an important effect on the outcome of tail domain phosphorylation. When the NF-M was phosphorylated first by MAPK then by PKA there was an increase in total NF-M phosphorylation (data not shown), a result consistent with earlier studies (16). On the other hand, when NF-M was first phosphorylated by PKA then by active MAPK, a significant reduction in phosphorylation was observed (Fig. 1A, compare lanes 3 and 4). These results suggest that PKA phosphorylation of the NF-M head domain inhibits subsequent phosphorylation of the tail domain by a prolinedirected kinase.
Head Domain Phosphorylation of NF-M Inhibits the Tail Domain Phosphorylation of NF-M in Primary Cortical Neurons-To determine whether PKA activation and phosphorylation of the NF-M head domain in neuronal cells affects tail domain phosphorylation, we cultured E18 primary cortical neurons in the presence and absence of forskolin. After 7 days of culture, cells were starved and some were then treated with forskolin (20 M), whereas others were treated with forskolin plus KT5720 (2.5 M), a PKA inhibitor. The cell lysates were assayed for PKA kinase activity and showed that forskolin stimulated activity 6-fold ( Fig. 2A, lanes 1 and 2), whereas the addition of the specific inhibitor restored activity to the control level (lane 3). The same cell lysates were subjected to Western blot analysis using antibody specific for NF-M phosphoepitopes in the tail domain (RMO-281) and anti NF-M antibody (for total NF-M). We found that activation of PKA decreases the tail domain phosphorylation of NF-M (Fig. 2B, lane 2). The PKA inhibitor (Fig. 2B, lane 3) reversed this inhibitory effect. Because it is well established that forskolin activation of PKA activity affects the MAPK pathway in most cells, including neurons (up-regulation or down-regulation, depending on specific neuron type) (38,39), it was essential to determine how the MAPK pathway was affected by forskolin in these cortical neurons. In Fig. 2C the data show that forskolin significantly activates Erk1/2 activity more than 2-fold (lane 2), which is inhibited by the PKA inhibitor (lane 3). This suggests that, despite forskolin activation of the MAPKs, known to phospho-  (40 -42). In a previous study we showed that the activated MAPK cascade induced tail domain phosphorylation of NF-M in transfected NIH3T3 cells (17). We assumed that, if the MAPK cascade can be induced in cortical neurons as in PC12 cells, it may result in elevated NF-M tail domain phosphorylation. This activated system could then be used to study the effect of forskolin stimulation of PKA phosphorylation on NF-M tail domain phosphorylation. We first treated neurons with NGF and showed a significant activation of Erk1/2 phosphorylation, suggesting that the MAPK cascade had indeed been stimulated (Fig. 3A, lane 2). Use of the MEK1 inhibitor PD98059 eliminated this activation (lane 3), suggesting that NGF could induce MAPK activity in primary cortical neurons. Moreover, activation of Erk1/2 by NGF resulted in increased phosphorylation of NF-M as seen in the Western blots of the same lysates using NF-M phospho-specific antibody RMO-281 (Fig. 3B, lane 2), and this also could be inhibited by the MEK1 kinase inhibitor PD98059 (Fig. 3B, lane 3). These results confirmed that the Ras/Raf/MAPK signaling pathway was involved in mediating endogenous NF-M tail domain phosphorylation in primary cortical neurons (16,17).
To determine whether the activation of PKA inhibits tail domain phosphorylation of NF-M induced by NGF stimulation of MAPK, we treated neurons with NGF only, forskolin plus NGF, or forskolin, NGF, and KT5720. With this regimen, it was evident that forskolin plus NGF significantly enhanced Erk1/2  2, Fig. 3C), which was restored to basal levels by the PKA inhibitor KT5720 (Fig. 3C, lane 3). Western blot analysis of these same lysates was performed using RMO-281, SMI31, SMI32, and anti-NF-M antibodies to detect the tail domain phospho-NF-M, dephospho-NF-M and total NF-M, respectively. Results showed that the induced tail domain phosphorylation of NF-M by NGF (Fig. 3D, lane 1) was markedly inhibited by forskolin (Fig. 3D, lane 2, compared with lane 1), but it was reversed by the PKA inhibitor, KT5720 (Fig. 3D, lane  3, compared with lane 2). In vitro PKA kinase assay was performed in same cell lysates to show the activity of PKA (Fig.  3E). Here again, as in the previous experiment, despite the stimulated activity of Erk1/2, tail domain phosphorylation was decreased in the presence of activated PKA. These results are consistent with the hypothesis that head domain phosphorylation of NF-M by PKA inhibits the tail domain phosphorylation of NF-M induced by NGF via the MAPK cascade. Although the activity of MAPK (Erk1/2) was elevated by PKA activation in these experiments, NF-M tail domain phosphorylation was reduced suggesting that a conformational change induced by PKA phosphorylation limits KSP domain phosphorylation.  Fig. 4 (A-C). The phosphorylation of tail domain NF-M was detected by Western blot analysis and immunocytochemistry using phospho-specific antibodies, RMO-281 and SMI31 (Fig. 4, A and D). In wtNF-M-transfected cells, the results showed that induced tail domain phosphorylation of NF-M by EGF stimulation of the MAPK cascade (Fig. 4, A and B, lane 4) was inhibited more than 6-fold by forskolin treatment (Fig. 4, A and B, lane 2). PKA inhibitor KT5720 (Fig. 4, A and B, lane 3) reversed this inhibitory effect. These results are in agreement with the experiments on cortical neurons (Fig. 3). It should be noted that, contrary to previous reports that forskolin inhibits the EGFinduced MAPK cascade in NIH3T3 cells (45,46), we found that in all cases of forskolin treatment Erk1/2 was equally activated by EGF (data not shown). What is most significant, however, is that in cells transfected with the head domain mutant (mtNF-M), the induced tail domain phosphorylation of NF-M by EGF was not inhibited by forskolin (Fig. 4, A and B, lane 5). These results suggest that activation of PKA (Fig. 4C, lane 5) does not inhibit tail domain phosphorylation of NF-M in the absence of phosphorylation at PKA-specific serine residues in the NF-M head domain.

Mutation of Serine 1, 23, and 46 Residues to Alanine in the Head Domain of NF-M Prevents PKA Phosphorylation and Fails to Inhibit Tail Domain Phosphorylation Induced by Erk1/2 Activation in Transfected NIH3T3 Cells-Previous
In Fig. 4C assays of PKA activity corresponding to each lysate in Fig. 4 (A and B) are shown to indicate the levels of forskolin activation. Fig. 4D represents the results of an im-  Fig. 4 (A and B). DISCUSSION Based on our studies of the squid giant axon and mammalian brains, we proposed that phosphorylation of cytoskeletal com- The cells were treated as follows: EGF only (a); forskolin followed by EGF (b and d); KT5720, then forskolin was added followed by EGF (c). ponents in neurons is compartmentalized and topographically regulated (47)(48)(49). Although the NF subunits and their relevant kinases are synthesized in cell bodies, stable phosphorylation of most KSP repeats in the NF-M and NF-H C-terminal tail domains by proline-directed kinases occurs primarily in the axon during slow transport (7, 16, 50 -53). On the other hand, phosphorylation of N-terminal head domain motifs by PKA and PKC occurs transiently in cell bodies and prevents NF assembly (3,4,9,10,12,19,21,55). If this occurs immediately after synthesis, premature cell body assembly is inhibited thereby protecting the neuron from abnormal accumulations of phosphorylated cytoskeletal aggregates, as seen in such neurodegenerative disorders as ALS (56). This would suggest that dephosphorylation of head domains must occur before NF assembly and extensive tail domain phosphorylation can begin. In fact, few NFs are seen in cell bodies; only small oligomers can be detected, which suggests that consequential neurofilament assembly is restricted to axons during transport (57). Results of a study in which dephosphorylation of NF-L Ser 55 was prevented (by mutation to Asp to mimic permanent phosphorylation), resulted in pathological accumulation of NF aggregates in brain neuronal cell bodies, are certainly consistent with this model (58).
The mechanisms underlying compartment-specific patterns of phosphorylation are not understood. We have proposed that transient head domain phosphorylation of NF-L and presumably of NF-M by PKA occurs within neuronal cell bodies to prevent NF assembly and tail domain phosphorylation by proline-directed kinases (8,32). As partially assembled subunits enter the axon hillock and proximal axon, the hypothesis poses dephosphorylation of head domain sites, followed by NF assembly within the axon, where KSP repeats in C-terminal tail domains of NF-M and NF-H are extensively phosphorylated (6,7).
The experimental results to test this model are, indeed, consistent with a key prediction of the hypothesis, that PKA phosphorylation of head domain sites in NF-M should inhibit phosphorylation of KSP repeats in the NF-M C-terminal tail domain. In vitro, the addition of PKA to an activated MAPK assay with NF-M as substrate resulted in a 50% reduction in tail domain phosphorylation, a result that could be attributed to simultaneous head domain phosphorylation by PKA. Instead of an additive effect, as one might expect if the kinases acted synergistically, an inhibition of phosphorylation was seen instead.
The in vivo experiments are much more compelling, however. It is well established that there is considerable cross-talk between the PKA and MAPK signaling pathways in a variety of cell types, including neurons (38). The response is specific to cell type; in some cases forskolin activation of PKA can downregulate the MAPK cascade as in NIH3T3 cells (59), whereas in primary neurons, depending on developmental age, MAPKs are either unaffected (1 day in culture) or are significantly activated (7 days in culture) (39). In agreement, the cortical neurons in our experiments, cultured for 7 days, did in fact respond with activation of the MAPK pathway after forskolin treatment (Fig. 2C). Nevertheless, the endogenous level of NF-M tail domain phosphorylation was reduced, instead of increased, as one might expect, because MAPKs are the principal prolinedirected kinases that phosphorylate KSP repeats in rat NF-M and NF-H (16).
Experimental stimulation of the MAPK pathway by the addition of NGF to these primary cortical neurons resulted in a 3-fold activation of Erk1/2 kinase activity, with a concomitant increase in C-terminal tail domain phosphorylation of NF-M. Surprisingly, NGF is primarily a sympathetic and sensory neu-ron trophic factor with few reported instances of cortical neuron effects. Nevertheless, despite almost 2-fold further activation of the Erk1/2 pathway after the addition of forskolin and PKA stimulation, NF-M tail domain phosphorylation was inhibited, rather than activated (Fig. 3). The reduced phosphorylation of NF-M tail domain phosphorylation after forskolin treatment in both of these in vivo experiments may be interpreted as an intramolecular inhibition of tail domain phosphorylation by phosphorylation of PKA motifs in the head domain.
This conclusion is strongly supported by the demonstration that NF-M mutants with three alanine-blocked and PKA-specific serine head domain sites, when transfected into EGFstimulated NIH3T3 cells, exhibited normal tail domain phosphorylation in the presence of forskolin-activated PKA. Tail domain phosphorylation of wild type NF-M, on the other hand, was markedly inhibited by forskolin. Here, too, the level of EGF-induced MAPK activity after forskolin treatment was unchanged compared with controls, so that the reduction in tail domain phosphorylation could not be attributed to reduced Erk1/2 activation. One hypothesis to explain these data is that phosphorylation of at least one (possibly all three head domain sites) is sufficient to change the conformation of the subunit protein so that tail domain KSP sites are less accessible to proline-directed kinases Erk1/2. Additional experiments are necessary to determine whether phosphorylation at one specific PKA site, or all sites, is essential in this intramolecular modulation.
Sequential phosphorylation of large molecules with many potential sites depends on the action of several different mutually dependent kinases that usually interact synergistically (18). In these instances, phosphorylation of one site by one kinase often activates phosphorylation of another site by an unrelated kinase. Presumably, this results from intramolecular conformational changes that facilitate kinase-substrate interactions. It is not too much of a speculative leap to postulate that a similar mechanism can inhibit phosphorylation in one domain (C-terminal tail?) by prior phosphorylation at specific N-terminal domain sites.
An alternative explanation, and one probably more likely, considering the large size of the NF-M substrate with its rigid coiled-coil rod, is an intermolecular model of inhibition, based on the role of NF-head domains in filament assembly (5,60). According to models of filament formation, dimers interact longitudinally head to tail to form protofilaments, and subsequently, these lead to staggered interactions to form larger protofibrils (60). This polymerization is inhibited by phosphorylation of the head domain sites by PKA. Hence, according to our model, inhibition of assembly would prevent proline-directed kinase phosphorylation of NF-M and NF-H tail domains. Tail domains are not integral to assembly of the core filament but are collapsed and extend laterally as cross-bridges as they are slowly phosphorylated in the axon. Head domain phosphorylation, by preventing assembly of oligomers, could result in conformational changes of the tail domains such that they are inaccessible to the kinase. The tail domains, in the absence of assembly, may either retain or assume a globular configuration, which prevents association with the kinase.
In a highly asymmetrical cell such as the neuron, where compartments are morphologically quite distinct, topographic regulation of phosphorylation of cytoskeletal proteins is tightly regulated. NF proteins are localized in soma, dendrites, and axons, but their states of assembly and phosphorylation differ and are compartment-specific. We suggest that transient phosphorylation of the head domain by PKA and PKC in the soma prevents assembly and tail domain phosphorylation until NFs enter the axon. If this tight topographic regulation is compro-mised by a variety of exogenous factors, then abnormal filament assembly and tail domain phosphorylation are induced in cell bodies with severe pathological consequences.