JOURNAL OF BIOLOGICAL CHEMISTRY 0 1994 by The American Society for Biochemistry and Molecular Biology, Inc. Vol. 269, No. 28, Issue of July 15, pp. 18480-18484, 1994 Printed in U S A . THE Increased Phosphorylationof the Amino-terminal Domainof the Low Molecular Weight Neurofilament Subunit in Okadaic Acid-treated Neurons* (Received for publication, January 21, 1994, and in revised form, April 19, 1994) Michael G. Sacher$, Eric S. AthlanS, and Walter E. Mushynskil From the Department of Biochemistry, McGill University, Montreal, Quebec H3G lY6, Canada Treatment of rat dorsal root ganglion cultures with 1 (Georges and Mushynski,1987) have been shown t o contribute 1.1~okadaic acid leads to a fragmentation of neurofila- to the anomalous migrationof NF subunits on SDS-polyacrylments and a reductionin the electrophoretic mobilities amide gels. of the threesubunits on SDS-polyacrylamide gels Much of the work to dateon NF subunitphosphorylation has (Sacher, M. G., Athlan, E. S., and Mushynski,W. E. (1992) focussed onidentifying theproteinkinases responsible for Biochem. Biophys. Res. Commun. 186,524-530). Based on phosphorylating the subunits. For example, many protein kithe observed response to varying concentrationsof oka- naseshavebeen shown to copurify withNFpreparations daic acid, fragmentation was inferred to be dueinhito (Julien et al., 1983; Toru-Delbauffe and Pierre,1983; Vallano et bition of protein phosphatase-2Aactivity and reduction al., 1985; Caputo et al., 1989; Wible et al., 1989; Dosemeci et al., in electrophoretic mobility to inhibition of protein phos- 1990) and t o phosphorylate NF subunits in vitro. Recently, phatase-1. Okadaic acid treatment led to an increase in cdc2-kinase from starfish oocytes was shown to phosphorylate amino-terminal, relative to carboxyl-terminal, domain the dephosphorylated form of NF-H, returning its mobility on phosphorylation in the lowmolecularweight(NF-L) SDS-gels to that of the native form (Hisanaga et al., 1991). subunit in the Triton X-100-soluble and -insoluble frac- Subsequently, several laboratories have identified and cloned tions. The purified catalytic subunit of protein phospha- cdc2-like protein kinases from nervous tissue (Hellmich et al., tase“dephosphorylated 52P-labeledNF-L and the 1992; Lew et al., 1992; Shetty et al., 1993). These kinases are capable of phosphorylating peptides containing the sequence middlemolecularweightsubunitfromokadaicacidtreated cultures, whereas the catalytic subunit of pro- Lys-Ser-Pro, which occurs in multiplecopies in both NF-M and tein phosphatase-1 had no effect. In the case of NF-L, NF-H. Furthermore, FA-kinase, the component required for phosphate moieties were preferentially removed from protein phosphatase-1 (PP-1) activation, was alsoshown to the amino-terminal domain. These results show that the phosphorylate NFs i n uitro (Guan et al., 1991). NFs were also amino-terminal domain ofNF-L can be phosphorylated shown to be in uitro (Sihag et al., 1988) and in vivo (Georges in situ andimplicateproteinphosphatase-2A in the et al., 1989; Grant and Aunis,1990) substrates for protein turnover of phosphate moieties in this domain. kinase C. Conversely, very little is known about the protein phosphatases involved in maintaining thephosphorylation state of NF Neurofilaments (NFs)’ are components of the cytoskeleton of subunits. Shetty et al. (1992) characterized a protein phosphamost neurons and arecomposed of three subunitsbelonging to tase activity which copurifies with NF preparations from both bovine and rat spinal cord and is inhibited by aluminum, vanathe type IV subclass of intermediate filament (IF) proteins i n vitro study, involving the four date, and fluoride. Another (Steinert and Roop, 1988). The three subunits have apparent molecular weights onSDS-polyacrylamidegels of 68,000 major mammalian serinehhreonine protein phosphatases (see (NF-L), 145,000 (NF-M), and 200,000 (NF-H) (Hoffman and Cohen (1989) for review), showed that none of these enzymes Lasek, 1975), although the deduced molecular weights of the were capable of promoting microtubule binding to NF-H in a subunits from rats are 61,000 (NF-L; Chin and Liem, 19891, manner similar to that observed after NF-H is treated with 95,000 (NF-M; Napolitano et al., 19871, and 115,000 (NF-H; either acid or alkaline phosphatase (Hisanaga et al., 1993). protein phosphaChinand Liem,1990). All threesubunits contain multiple Recent studies haveshown that the PP-1 and ( O A Bialojan and Takai, tase-ZA(PP-2A) inhibitor okadaic acid phosphate moieties (Julien and Mushynski,1982) and are rich root in charged amino acid residues. Both the phosphate moieties 1988) causes a disruption of the NF network in rat dorsal (JulienandMushynski, 1982) andglutamic acid residues ganglion (DRG) neurons (Sacheret al., 1992) and an increased deposition of NFsubunitsinnb2aidlneuroblastoma cells (Shea et al., 1993). * This research was supported in part by Grant MT-5159 from the The functional role of NF phosphorylation is unclear and Medical Research Council of Canada. The costs of publication of this may depend on the location of the phosphate moieties within article were defrayed in part by the payment of page charges. This article must thereforebe hereby marked “advertisement”in accordance the subunits (seeNixon and Sihag(1991) for review). Previous work in our laboratory has shown that treatment of rat DRG with 18 U.S.C. Section 1734 solely to indicate this fact. f Recipients of studentships from the Fonds de la Recherche en Sante cultures with 1 PM OA causes a rapid disruption of the NF du Quebec. network (Sacher et al., 1992). Other studies have shown that 5 To whom correspondence and reprint requests should be addressed. phosphorylation of the amino-terminal head domain of NF-L i n Tel.: 514-398-7286; Fax: 514-398-7384. The abbreviations used are: NF, neurofilament; NF-L, NF-M, and vitro causes the subunit t o dissociate from a preexisting netNF-H, low, middle, and high molecular mass neurofilament subunits, work and prevents its assembly (Nakamura et al., 1990; Gonda respectively; DRG, dorsal root ganglia; IF, intermediate filament;NCS, et al.1990). Similarly, head domainphosphorylation of the type N-chlorosuccinimide; PP-1 and PP-2A, protein phosphatase-1and -2A, I11 IF subunits, vimentin, desmin, and glial fibrillary acidic respectively; PAGE, polyacrylamide gel electrophoresis; OA, okadaic protein, has been correlated with filament breakdown i n vitro acid. 18480 of NF-L Amino-terminal Domain Phosphorylation and in vivo (Geisler andWeber, 1988; Geisler et al., 1989; Inagaki et al., 1990; Chou et al., 1991; Matsuoka et al., 1992).We therefore set out to determine the distributionof phosphorylation sites in NF-L subunits after OA treatment by chemical cleavage analysis. Our results show an increase in amino-terminaldomainphosphorylation in NF-L following OA treatment. Inhibition of PP-2A activity was implicated in OA-induced NF fragmentation, and the catalytic subunit of PP-2A was shown to be capable of removing phosphate moieties from the amino-terminal domain of NF-L in vitro. 18481 A 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 NF-H [ NF-M pNF-L NF-L - - EXPERIMENTAL. PROCEDURES B Materials-Okadaic acid and calyculin A were purchased from LC Services(Woburn, MA). Carrier-free 'lP, was from ICN Biomedicals (Mississauga, Ontario, Canada). 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 SDS-Polyacrylamide Gel Electrophoresis (SDS-PAGE)-Gel electrophoresis and Western blotting wereperformed a s described previously NF-H [ (Laemmli, 1970; Sacher et al., 1992). Polypeptides used for chemical NF-M cleavage analysis (seebelow) were obtained from gels run in the presence of 0.025% thioglycolic acid to prevent protein oxidation. Protein on Western blots was quantified usingMicrovision a SCSI camerafollowed by analysis using the Millipore Bio-Image analyzer. Radioactivity was pNF4 = . L quantified with the FujiBAS2000 PhosphorImager. " 0 NF-L Immunoprecipitation-Cells were harvested in cytoskeleton extraction buffer (CSK buffer) consisting of 1%Triton X-100,150mM NaCI, 50 FIG.1. Western blot analysis of the effects of increasing OAand msr Tris-HC1, pH 7.5, 50 mM NaF, 2 mM EGTA, 2 mM levamisol, 1 mM calyculin A concentrations on NF subunits in DRG neurons. phenylmethylsulfonyl fluoride and centrifuged for 15 min a t 13,000 x g. DRG cultures were fractionated into TritonX-100-insoluble (odd-numbered lanes) and Triton X-100-soluble fractions (even-numhered lanes) SDS was added to the supernatant to a concentration of 0.1% followed by heating in aboiling water bathfor 2 min. The pellet was suspended after a 4-h exposure to eitherOA (A) or calyculin A ( B ) .Samples were in 2% SDS, 50mM Tris-HC1,pH 6.8, heated for 2 min, and then the SDSresolved by SDS-PAGE and the NF subunits detectedby Western blotwas diluted to0.1% with CSK buffer. Immunoprecipitation was carried ting as described under "Experimental Procedures". A, OA concentraout as described previously (Lindenbaum et al., 1987) using either af- tions used were 0 (lanes 1 and 21, 1 nM (lanes 3 and 4 ) , 10 nM (lanes 5 finity-purified antibody or immune serum. In vitro phosphatase treat- and 6 ) , 100 nM (lanes 7 and 8),500 nM (lanes 9 and IO),and 1000 nM ment of the immunoprecipitates was carried out prior to elution from (lanes 11and 12).B , calyculin Aconcentrations used were 0 (lanes 1 and 2), 0.1 nM (lanes 3 and 4 ) , 0.5 nM (lanes 5 and 6 ) , 1 nM (lanes 7 and 8). the protein A-Sepharose beads (seebelow). 10 nM (lanes 9 and lo), and100 nM (lanes 11and 12).NF-H, NF-M, and Cell Culture-Rat DRG were dissected and maintained in defined NF-L refer to the positions of the three NF subunits. pNF-L refers to the medium a s described previously (Sacher et al., 1992). For metabolic slower migrating hyperphosphorylated form of NF-L. labeling cells were incubated with 0.5 mCi of carrier-free "Pjml of Pi-free medium (Flow Laboratories, McLean, VA) for 3 h prior to OA the electrophoretic mobilities of any of the three NF subunits treatments. Phosphatase neatment of Immunoprecipitated Proteins-For phos- (Fig. lA, lanes 7 and 8).As the OA concentration was raised to 500 and 1000 nM, fragmentation wasaccompanied by decreases phatase treatment of '*P-labeled immunoprecipitates, samples were incubated for 5 h at 30 Oc, prior to elutionfrom the protein A-Sepharose in the electrophoretic mobilities of all three subunits (Fig. L 4 , beads (seeabove), in 65mM Tris-HCI, pH 7 , 1 mM MgCI,, 0.5 mM EGTA, compare lanes 7 and 8 with lanes 9 and 10).This would imply 1 mu phenylmethylsulfonyl fluoride= PP-2Ac,PP-1, (1.5 pg/ml) and OA that fragmentation at the lower OA concentration was due to as described in the figure legends. inhibition of PP-2A, whereas theconcomitant decrease in elecChemical Cleavage of Polypeptides-Immunoprecipitated "P-labeled trophoretic mobility of subunits a t the higherOA concentration NF-L was resolved by SDS-PAGE, located by autoradiography, and excised. The gel slice was rehydrated in N-chlorosuccinimide (NCS) (Sacher et al., 1992) was due to the additional inhibition of buffer (1 g of urea, 1ml of water, 1 ml of acetic acid) (Lischwe and Ochs, PP-1. This inference is supported by the results in Fig. 1B 1982) and incubatedat room temperature for 1.5 h in the same buffer showing that both NF fragmentation anddecreases in subunit containing 2 mg/ml NCS. The slices were washed extensively in water, electrophoretic mobility occur at the samecalyculin A concenequilibrated in SDS-PAGE sample buffer, and loaded vertically ontoan SDS-l2% polyacrylamidegel. Digested products were visualizedby au- tration (lanes 9 and 10). Sinceconsensussequences have not been established for toradiographyusinga Du PontLightningPlusintensifyingscreen PP-1 and PP-2A (Ingebristen andCohen, 1983;Agostinis et al., (Swanstrom and Shank, 1978) and quantified as described above. Preparation of Protein Phosphatase-1 and -2A-The catalytic sub19871, it is not possible to predict where the additional phosunits of PP-1 andPP-SA (PP-1, and PP-2Ac, respectively) were purified phate moieties would be found in the NF subunits when the from rabbit skeletal muscle following the procedure of Cohen et al. phosphatases are inhibited. However, phosphoamino acid anal(1988). ysis shows a similar lack of phosphothreonine in NF-L and no PhosphorylatedAmino Acid Analysis-Immunoprecipitated NF-L and NF-M were resolved by SDS-PAGE, located by autoradiography, change in the proportions of phosphoserine to phosphothreoexcised, and processed a s described previously (Julien and Mushynski, nine in NF-M after OA treatment (data not shown). These 1982). Unlabeled phosphoamino acid standards were detected by nin- phosphoamino acid profiles correspond to those of in vivo 32Phydrin staining (Cooper et al., 1983). labeled NF-L and NF-M (Julien and Mushynski,1982). --- RESULTS To determine which phosphatase(s) playsa role in maintaining the integrity of NFs, DRG cultures were treated with increasing concentrations ofOA or calyculin A.OA is a more potent inhibitor of PP-SA than of PP-1, whereas calyculin A inhibits both PP-1 and PP-2A with similarpotencies (Ishihara et al., 1989). As shown in Fig. l A , treatment of cultures with 100 nM OA for 4 h led to NF fragmentation with no changes in We examined the distribution of phosphorylation sites in NF-L by chemicalcleavage with NCS (Lischwe and Ochs, 1982). Rat NF-L contains 1 tryptophan residue a t amino acid 280 (Chin and Liem, 19891, and the two halves of the protein migrate differently on SDS-PAGE (Mahboub et al., 1986). Determining the ratioof carboxyl-terminal t o amino-terminal "*P phosphorylation (C/N ratios) provides a means for assessing the location of phosphorylation sites on NF-L following OA treatment. DRG cultures were labeled with n*Pifollowed by a Amino-terminal Domain Phosphorylation 18482 1 2 3 M, ( W 2051169766- of NF-L A - OA TREATED CELLS PP-PA PP-1 10 nM OA 1 DM OA " " 1 NF-M - NF-L - CIN 10 1.7 1.6 FIG.2. Gel electrophoretic analysis of the amino- and carboxyl-terminal halves of chemically cleaved NF-L fromuntreated and okadaic acid-treated DRG cultures. DRG cultures were labeled with 32P,for 3 h and either untreated (lane 1 ) or treated with 1 p . OA ~ for 1 h (lanes 2 and 3 ) . NF-L was immunoprecipitated from the Triton X-100-insoluble (lane 2 ) or Triton X-100-soluble (lane 3 ) fractions. Protein was resolved by SDS-PAGE and digested with NCS (see "Experimental Procedures"). C and N refer to the carboxyl-terminal and amino-terminal portions, respectively, of NF-L after NCS treatment. NF-L refers to the migration of uncleaved NF-L subunit and the M , (x standards are shown to the left of the autoradiograph. CIN ratios are shown at the bottom of each lane. + + " 4529- + - 2 3 + + + - + - - + " - + 4 5 6 + + B 1-h treatment with 1 p~ OA. Samples were fractionated into Triton X-100-soluble and -insoluble fractions, and immunoprecipitated NF-L was treated with NCS. As shown in Fig. 2 the C/N ratio of Triton X-100-soluble and -insoluble NF-L from OA-treated cells was reduced by about 6-fold in comparison with that of the untreatedcontrol. I t is interesting to note that the CM ratiosof both the Triton X-100-soluble and -insoluble NF-L from OA-treated cultures were the same. These results indicate that althoughOA treatment causes a preferential increase in amino-terminaldomain phosphorylation in NF-L, the Triton X-100-soluble subunit shows nofurther increase in phos1 2 3 4 5 6 phorylation of this domain. FIG.3. Gel electrophoretic analysis of NF subunits from OAAlthough enhanced phosphorylation of the amino-terminal treated DRG cultures after in vitro treatment with PP-1, and domain in NF-L from OA-treated cultures was seen both in the PP-PA,. "P-Labeled NF-L and NF-M were immunoprecipitated from Triton X-100-soluble and -insoluble subunits, theTriton X-100- control or OA-treated (1 h) DRG cultures (OA). Protein phosphatase soluble subunits were phosphorylated to a higher degree than treatment (PP-1, and PP-BA,) was carried out prior to elution of the from the protein A-Sepharose beads as described under "Extheir insoluble counterparts. The increase ranged from 2-fold subunits perimental Procedures". Addition of OA during phosphatase treatment for NF-L to 3-fold for NF-M (data not shown). at either 10 nM or 1PM is indicated above each lane. Equal amounts of Since inhibition of PP-2A had alreadybeen implicated in the the proteins were fractionatedby SDS-PAGE, and the positions of the to the left of the autoradiograph(A). B shows a fragmentation of NFs (cf. Fig. 1, A and B ) , we examined two subunits are shown whether PP-2Acwas capable of acting on NF-M and NF-L from histogram obtained by quantifying the autoradiograph from A. Lanes 1-6 in B correspond to the six lanes of the autoradiograph inA. Data OA-treated cultures. 32P-LabeledNF-L and NF-M were immu- pertaining to NF-L are indicated by stippled bars, whereas those for noprecipitated from control and OA-treated DRG cultures. NF-M are represented by solid black bars. Prior to their elutionfrom the proteinA-Sepharose beads, the subunits were treated either with purified PP-2Ac or purified Treatment of NF-L from OA-treated cultures withPP-2Acled to PP-1, in theabsence or presence of OA as indicated above each an increase in the CM ratio compared as with untreated sublane. Equal amounts of protein were fractionated by SDS- unit, to almost the same level a s was seen in NF-L from culPAGE. Theautoradiographin Fig. 3A shows that PP-2Ac tures not treated withOA (Fig. 4, compare lanes l , 2,and 3 ) . caused a reduction of about 50% in the amount of 32Pin both OA at low concentrations prevented thisPP-2Ac-dependent reNF-M and NF-L relative to non-phosphatase-treated subunits versal of the C/N ratio (Fig. 4, lane 4 ) , and PP-1, had no effect PP-2ACpreferentially removed the from OA-treated cultures, whereasOA, at 10 nM, inhibited this on the ratio, indicating that PP-24-dependent decrease (Fig. 3B, compare lanes 3 and 4 ) . OA-induced phosphate moieties from the amino-terminal porPurified PP-1, had no effect on the 32Pcontent of either subunit tion of NF-L. (compare lanes 2 and 5). DISCUSSION Since PP-PA had been implicated in the maintenanceof NF integrity (Fig. l ) , we wished to determine whether PP-2A, could Thefragmentation of NFs that occurs in DRG cultures remove the phosphatemoieties seen in the amino-terminaldo- treated with OA (Sacher et al., 1992) may represent theamplimain of NF-L after OA treatment. NF-L bands from the gel in fication of a mechanism normally involved in modulating the Fig. 3A were excised, subjected to NCS digestion, and the CM local reorganization of NFs ina variety of situations, including ratios were determined (Fig. 4).Again it can be seen thatOA axonal transport. In this respect it is interesting to note that treatment led to an increase in amino-terminal phosphoryla- Okabe et al. (1993) showed a recovery in NF-L fluorescence tion of NF-L, as indicated by the 4.5-fold reduction in the CM following photobleaching with no movement of the photoratio following OA treatment (Fig. 4, compare lanes 1 and 2). bleached zone. Additionally they showed that biotin-labeled - Amino-terminal Domain Phosphorylation of NF-L + + - + + + +- ++ " " + 18483 + cells (Eriksson et al., 1992). We also showed that PP-PA,, but not PP-l,, preferentially removed the OA-induced amino-terminal domain phosphorylation sites in vitro (Fig. 4). further suggesting that inhibitionof this enzymeis responsible for the fragmentation of NFs that accompanies OA treatment. However, this result does not exclude the involvement of PP-2A in maintaining the phosphorylation state of the tail domain of NF subunits, since the heterotrimeric form of PP-2A was shown to be the mosteffective form for acting on microtubule-associated protein taufollowing its phosphorylation by p42 mitogen-associated protein kinase (Goedert et al., 1992). Hisanaga et al. (1993) showed that PP-PA can dephosphorylate NF-M which had been phosphorylated in oitro with protein kinase A. This implies that PP-2A can remove phosphate moieties from the head domainof NF-M, since it is this domain CIN 6.0 1.3 4.2 1.4 1.6 1.4 which is phosphorylatedby protein kinaseA in vitro (Sihag and FIG.4. Gel electrophoretic analysis of chemically cleaved NF-L Nixon, 1990). Our results showing that PP-ZA, removes phosfollowing in vitro treatment with PP-1, and PP-2AC.The NF-L subunit from the gel shown in Fig. 3 was excised and digested with NCS phate from OA-treated NF-M (Fig. 3) indicate that OA treat(see "Experimental Procedures"). Treatmentsare the same as in Fig. 3 ment also leads to an increase in phosphorylation of the head and are shown above each lane. C, N , and NF-L are as indicated in the domain of NF-M, although this remains to be determined dilegend to Fig. 2. C I N ratios are shown at the bottom and the M,( x 10-7 rectly. We also show a similar dephosphorylation pattern for standards are shown at the left. NF-L. The removal by P P - 2 4 of -50% of the "P from OANF-L was incorporated into numerous discrete sites along the treated NF-M implies that the enzyme may also dephosphoaxon, indicating thatsoluble oligomers of NF subunits can be rylate sites in the carboxyl-terminal tail domain. The demontransported along the axon faster than the bulk movement of stration by Sola et al. (1991) that PP-PA removes phosphate from Lys-Ser-Pro sequences in histone H1 is consistent with NFs. Itis possible that amino-terminalphosphorylation/ dephosphorylation allows a transient release of oligomers to this deduction, because similar tripeptide repeats arefound in occur during the transport of NFs down the axon. The abun- the tail domain of NF-M (Napolitano et al., 1987). The fact that PP-1, had no effect on the nzP levels in both dance of protein phosphatases in the cell (Hardie et al., 1991) would ensure that any local disruption of NFs be of a short NF-L and NF-M is consistent with our studies implicating inof electrophoretic variantsof duration andfully reversible. Treatment withOA may allow for hibition of PP-1 in the appearance the accumulation of phosphorylated species which normally all three subunits (Fig. 1).These varianb were not present in the PP-1,-treated subunits which wereobtained from cells have a transient existence. This model implies that the fragmented NFs shouldbe in a treated with OA for 1 h (Fig. 4 and Sacher et al., 1992). Alterform which isreadily re-incorporated into 'hiton X-lOO-in- natively, PP-1, may require additional subunits to be active soluble structures upon restoration of protein phosphatase ac- against NF subunits. Any attempt to explain the OA-induced fragmentation of tivity. Indeed, we have found that the early reversible stages of OA-induced NF fragmentation give rise to a heterogeneous NFs in DRG cultures must take into account the similar inin the Triton population of very large oligomers,2 thus allowing for a rapid crease in amino-terminal domain phosphorylation recovery of the NF network after OA is removed (Sacher et al., X-100-insoluble and -soluble subunits (Fig. 3) and the higher specific radioactivity of the latter as compared with the former 1992). A major reason for speculating that increased phosphoryla- (data not shown). We suggest that hyperphosphorylation of to a relatively tion was responsiblefor the OA-induced fragmentation of NFs subunits in their amino-terminal domains leads for the disruptionof was because recentin vitro studies showed that IFs composed slow fragmentation of NFs. The time frame of vimentin (Geisler et al., 1989), glial fibrillary acidic protein the NF network after OA treatment contrasts with that of the (Inagaki et al., 1990), desmin (Geisler and Weber, 1988) or glial fibrillary acidic protein network, found in the Schwann NF-L (Nakamura et al., 1990;Gonda et al., 1990) were all cells of the DRG cultures, which was shown to be complete within 1h.3 The dynamic natureof NFs has been demonstrated disrupted upon phosphorylation of the amino-terminal head by several investigators in recent years(Angelides et al., 1989; domain. In addition, second messenger-dependent protein kinases were shown be to involved in phosphorylationof the head Okabe et al., 1993). These reports have suggested that NFs are domain of NF-L (Sihag andNixon, 1989) andNF-M (Sihag and in equilibrium with a poolof soluble subunits or oligomers. Nixon, 1990), leading to the suggestion that head domain phos- Phosphorylation of NFs following OA treatment of DRG neuphorylation of NFs influences filament assembly states (Nixon rons maylead to a slow accumulation of these soluble oligomers due to amino-terminal domain phosphorylation. and Sihag, 1991). There have been some recent reports suggesting that doNCS cleavage of NF-L (Fig. 2) showed a correlation between OA treatment and increased amino-terminal domainphospho- main-specific phosphorylation may not be involved in modulatof IFs. Chou et al. (1993) recently rylation. Since phosphorylation in the a-helical rod domain of ing the assembly states the phosNF subunits has not been reported (Steinert et al., 1982), we showed thatthere were no differencesbetween suspect that the phosphate moieties in the amino-terminal seg- phopeptide maps of the soluble and insoluble forms of keratin 8 and keratin18. One importantdifference between the soluble ment of NF-L are located in the head domain. X-100-soluble NFs in this The OA-induced fragmentation of NFs was attributedto in- keratins in their study and the Triton report is that the keratins were found in a naturally occumng hibition of PP-PA based on the concentration response data for OA and calyculin A (Fig. 1).A similar approach was used to soluble pool made up of tetramers, whereas the OA-induced implicate PP-1 in the disruptionof the IF network inBHK-21 soluble pool in this study was composed of large oligomers.2 OA TREATED CELLS PP-PA PP-1 10 nM OA 1 pM OA " I . " -+ * M. G.Sacher, E. S. Athlan, and W.E. Mushynski, manuscript in preparation. M. G. Sacher, E. S. Athlan, and W. E. Mushynski. unpublished observation. 18484 Amino-terminal Domain Phosphorylation Nevertheless, these authorsproposed a mechanism similar to the one we invoked above, suggesting that phosphorylation may shift thesolubility equilibrium constant. Klymkowsky et al. (1991) suggested that fragmentation of the cytokeratin network observed duringxenopus oocyte maturation maybe due toa severing activity, similar to theone seen for microtubules (McNally and Vale, 1993). They speculated a severing activity to that phosphorylation may serve to target the filament network giving rise t o a heterogeneous population of relatively large oligomers similar t o those seen inDRG neurons following OA treatment. Inview of the diversityof potential mechanisms that could render IFs Triton X-100-soluble, further studies are required t o define the role of domain-specific phosphorylation in NF dynamics. Acknowledgment-The expert technical assistance of Sylvia Levine is gratefully acknowledged. REFERENCES Agostinis, P., G o r i s , J., Waelkens,E., Pinna, L.A,, Marchiori, F., and Merlevede, W. (1987) J. Biol. Chem. 262, 1060-1064 Angelides, K. J., Smith, K. E., and Takeda,M. (1989) J . Cell Biol. 108, 1495-1506 Bialojan, C., and Takai, A. (1988) Biochem. J. 256, 283-290 Caputo, C. B., Sygowski, L. A,, Brunner, W. F., Scott, C. W., and Salama,A. (1989) Biochim. Biophys. Acta 1012,299307 Chin, S. S., and Liem, R. K. H. (1989) Eur. J. Cell Biol. 50, 475-490 Chin, S.S. M., and Liem, R. K. H. (1990) J. Neurosci. 10, 37143726 Chou, Y-H., Ngai, K.-L., and Goldman, R. (1991)J . B i d . Chem. 266, 7325-7328 Chou, C.-F., Riopel, C. L., Rott, L. S., and Omary, M. B. (1993) J. Cell Sci. 105, 433444 Cohen, P. (1989) Annu. Reu. Biochem. 68,453-508 Cohen, P., Alemany, S., Hemmings, B. A., Resink, T. J., StrAlfors, P., and Tung, H. Y. L. (1988) Methods Enzymol. 159,390-408 Cooper, J. A,, Sefton, B. M., and Hunter, T.(1983) Methods Enzymol. 99,387401 Dosemeci,A., Floyd, C. C., and Pant,H. C. (1990) Cell. Mol. Neurobiol.10,369382 Eriksson, J. E., Brautigan, D. L., Vallee, R., Olmsted, J., Fujiki,H., and Goldman, R. D. (1992) Proc. Natl. Acad. Sci. U. S. A . 89, 11093-11097 Geisler, N., and Weber, K. (1988) EMBO J. 7,15-20 Geisler, N., Hatzfeld, M., and Weber, K. (1989) Eur. J. Biochem. 183, 441447 Georges, E., and Mushynski, W. E. (1987) Eur. J. Biochem. 166,281-287 Georges, E., Lindenbaum, M. L., Sacher, M. G., Trifar6, J.-M., and Mushynski, W. E. (1989) J. Neurochem. 52, 1156-1161 Goedert, M., Cohen, E. S., Jakes, R., and Cohen, P. (1992) FEBS Lett. 312,95-99 Gonda, Y., Nishizawa, K., Ando, S., Kitamura, S., Minoura, Y., Nishi, Y., and Inagaki, M. (1990) Biochem. Biophys. Res. Commun. 167,131&1325 Grant, N. J., and Aunis, D. (1990) Eur. J. Cell B i d . 52, 3 6 4 6 Guan, R. J., Khatra, B. S., and Cohlberg, J.A.(1991)J. Biol. Chem. 266,8262-8267 Hardie, D. G., Haystead, T. A. J., and Sim,A. T.R. (1991) Methods Enzymol. 201, 469-476 View publication stats of NF-L Hellmich, M. R., Pant, H. C., Wada, E., and Battey, J. F. (1992) h o c . Natl. Acad. Sci. U. S. A. 89, 10867-10871 Hisanaga, S., Kusubata, M., Okumura, E., and Kishimoto, T. (1991) J. Biol. Chem. 266,21798-21803 Hisanaga, S.,Yasugawa,S., Yamakawa, T.,Miyamoto, E., Ikebe, M., Uchiyama, M., and Kishimoto, T. (1993) J. Biochem. (Tokyo)113, 705-709 Hoffman, P. N., and Lasek, R. J. (1975) J. Cell B i d . 66, 351-366 Inagaki, M., Gonda, Y., Nishizawa, K., Kitamura,S., Sato, C., Ando, S., Tanabe, K., Kikuchi, K., Tsuiki, S., and Nishi, Y. (1990) J. Biol. Chem. 265,47224729 Ingebristen, T. S., and Cohen, P. (1983) E m J. Biochem. 132,255-261 Ishihara, H., Martin, B. L., Brautigan, D. L., Karaki, H., Ozaki, H., Kato, Y., Fusetani, N., Watabe, S., Hashimoto, K., Uemura, D., and Hartshorne, D. J. (1989) Biochem. Biophys. Res. Commun. 159, 871-877 Julien, J.-P., and Mushynski,W. E. (1982) J. B i d . Chem. 257, 10467-10470 Julien, J . 2 , Smoluk, G. D., and Mushynski, W. E. (1983) Biochim. Biophys. Acta 755,2531 Klymkowsky, M. W., Maynell, L. A., and Nislow, C. (1991)J. Cell Biol. 114,787-797 Laemmli, U. K. (1970) Nature 227,680-685 Lew, J., Winkfein, R. J., Paudel, H. K., and Wang, J. H. (1992) J. Biol. Chem. 267, 25922-25926 Lindenbaum, M. H., Carbonetto, S., and Mushynski, W. E. (1987) J . Biol. Chem. 262,605-610 Lischwe, M. A., and Ochs, D. (1982) Anal. Biochem. 127,453457 Mahboub, S., Richard, C., Delacourte, A,, and Han, K.-K. (1986) Anal. Biochem. 154, 171-182 Matsuoka, Y.,Nishizawa, K., Yano, T., Shibata, M., Ando, S., Takahashi, T., and Inagaki, M. (1992) EMBO J. 11,2895-2902 McNally, E J., and Vale, R. D. (1993) Cell 75, 419-429 Nakamura, Y., Takeda, M., Angelides, K. J., Tanaka, T., Tada, K., and Nishimura, T. (1990) Biochem. Biophys. Res. Commun. 169, 744-750 Napolitano, E. W., Chin, S. S. M., Colman, D. R., and Liem, R. K. H. (1987) J. Neurosci. 7, 2590-2599 Nixon, R. A., and Sihag, R. K. (1991) lFends Neurosci. 14, 501-506 Okabe, S., Miyasaka, H., and Hirakawa, N. (1993) J . Cell Biol. 121,375486 Sacher, M. G., Athlan, E. S., and Mushynski, W. E. (1992) Biochem. Biophys. Res. Commun. 186,524-530 Shea, T. B., Paskevich, P. A,, and Beermann, M. L. (1993) J. Neurosci. Res. 35, 507-521 Shetty, K. T., Veeranna, and Guru, S. C. (1992) Neurosci. Lett. 137, 83-86 Shetty, K. T.,Link, W. T., and Pant, H.C. (1993) Proc. Natl. Acad. Sci.U. S. A . 90, 6844-6848 Sihag, R. K., and Nixon, R. A. (1989) J. Biol. Chem. 264,457464 Sihag, R. K., and Nixon, R. A. (1990) J. B i d . Chem. 265,41664171 Sihag, R. K., Jeng, A. Y., and Nixon, R. A. (1988) FEBS Lett. 233, 181-185 Sola, M. M., Langan, T., and Cohen, P. (1991) Biochim. Biophys. Acta 1094, 211216 Steinert, P. M., and Roap, D. R. (1988) Annu. Reu. Biochem. 57,593-625 Steinert, P. M., Wantz, M. L., and Idler, W.W. (1982)Biochemistry 21, 177-183 Swanstrom, R., and Shank, P. R. (1978)Anal. Biochem. 86, 184-192 lbru-Delbauffe, D., and Pierre, M. (1983) FEBS Lett. 162,23&234 Vallano, M. L., Buckholz, T.M., and DeLorenzo, R. J. (1985)Biochem.Biophys. Res. Commun. 130,957-963 Wible, B.A., Smith, K. E., andAngelides, K. J. (1989)Proc.Natl. Acad. Sci.U. S. A . 86, 720-724