QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

First published online 3 February 2004
doi: 10.1242/jcs.00906

This Article Summary Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Eriksson, J. E. Articles by Goldman, R. D. Search for Related Content PubMed PubMed Citation Articles by Eriksson, J. E. Articles by Goldman, R. D. Social Bookmarking                         What's this?

Specific in vivo phosphorylation sites determine the assembly dynamics of vimentin intermediate filaments

John E. Eriksson^1,2,*, Tao He^2,3,4,, Amy V. Trejo-Skalli^5,, Ann-Sofi Härmälä-Braskén^2,3, Jukka Hellman², Ying-Hao Chou⁵ and Robert D. Goldman⁵

¹ Department of Biology, Laboratory of Animal Physiology, University of Turku, Science Building 1, FIN-20014 Turku, Finland
² Turku Centre for Biotechnology, University of Turku and Åbo Akademi University, POB 123, FIN-20521 Turku, Finland
³ Department of Biochemistry, Åbo Akademi University, FIN-20521 Turku, Finland
⁴ Turku Graduate School of Biomedical Sciences, Kiinanmyllynkatu 13, FIN-20520, Turku, Finland
⁵ Department of Cell, Molecular, and Structural Biology, Northwestern University Medical School, 303 E. Chicago Avenue, Chicago, IL-60611-3008, USA

View larger version (40K): [in a new window]   Fig. 1. Inhibition of vimentin dephosphorylation in vivo induces hyperphosphorylation and disassembly of vimentin polymers. (A) 32P in vivo labeled BHK-21 cells were incubated without or with 20 nM of the protein phosphatase inhibitor cl-A for 10 and 20 minutes and with 50 nM cl-A for 30 minutes. The inhibition of constitutive vimentin phosphatase activities causes a dose and time-dependent elevation of vimentin phosphorylation, as shown in the autoradiography of the 32P in vivo labeled proteins separated by 10% SDS-PAGE. (B) The phosphorylation causes a disassembly of the vimentin IF filaments, which is reflected by their increase in both the low and high speed centrifugation supernatants, as measured by western blotting of the respective supernatant fractions. While the IF pool in the low speed supernatant is primarily composed of both large filament fragments and soluble subunits, the high speed supernatant contains only truly soluble subunits. The vimentin levels are elevated both in the low and high speed supernatants but desmin concentrations are only elevated in the low speed supernatants. This implies that desmin is not disassembled under these conditions into soluble subunits but merely fragmented. This is consistent with the phosphorylation of desmin, which is not elevated by inhibition of dephosphorylation. (C) The specific phosphorylation (32P labeling/protein units) of polymer-associated and depolymerized vimentin subunits was analyzed by double in vivo labeling with [32P]orthophosphate and [35S]methionine followed by phosphorimager analysis (32P + 35S is the signal collected from both isotopes on the gel; the 32P signal was obtained by exposure through four layers of aluminum foil). The results indicate that phosphate incorporation takes place both on polymer-associated vimentin and on the dissociated subunits, as shown by (D) quantification of the same specific levels of 32P per protein unit (32P/35S), with a certain preference for phosphate incorporation into the soluble subunits, as indicated by the ratios between the different protein pools, as compared to the filamentous pool (=1), shown above the bars.

View larger version (26K): [in a new window]   Fig. 2. Phosphorylation-induced disassembly of vimentin IFs in vitro and in vivo results in disassembled subunits with the same molecular mass categories. (A) Bacterially expressed human vimentin IF polymers were phosphorylated in vitro by PKA and the resultant disassembly of depolymerized subunits was assayed by glutaraldehyde crosslinking (GA; concentration range: 0-0.06%) of the vimentin supernatant fractions after 30 minutes centrifugation at 200,000 g. (B) The in vitro disassembly of vimentin was compared to that in vivo, by incubating BHK-21 cells in the presence of 50 nM cl-A for 30 minutes. The cells were treated with 1% Triton X-100 and the particulate material was pelleted by centrifugation at 200,000 g for 30 minutes. The supernatants were then subjected to glutaraldehyde crosslinking (same concentration range as above) followed by immunoblottting. The results indicate that the phosphorylation-mediated release of vimentin subunits also occurs as tetramers in vivo.

View larger version (52K): [in a new window]   Fig. 3. Phosphopeptide mapping of the major interphase-specific in vivo phosphorylation sites on vimentin. (A) The relatively low level of 32P-labeling of vimentin in interphase cells shown in Fig. 1, reflects constitutive phosphorylation on a few major sites, as indicated by the presence of a few more prominently labeled tryptic peptides (1, 4, 5, and 8) from vimentin isolated from untreated 32P in vivo labeled BHK-21 cells. (B) When dephosphorylation is inhibited with 20 nM cl-A for 20 minutes, all (except peptide 8) of these constitutively labeled phosphopeptides displayed significant increases in labeling and, in addition, many new peptides (2, 3, 6, 7) showed marked elevations in 32P labeling. These results indicate that there are some sites that maintain a certain level of constitutive phosphorylation and, in addition, some sites that are subjected to a high constitutive phosphate turnover. (C) The interphase-specific sites do not correspond to the mitosis-specific in vivo phosphorylation sites, as shown on a vimentin phosphopeptide map derived from in vivo labeled cells, blocked in metaphase by treatment with 2 µg/ml nocodazole for 3 hours. (D-H) The in vivo phosphopeptide maps were compared to those obtained by in vitro phosphorylation using a number of potential vimentin kinases. (D) PKA, (E) PKC, (F) CaMKII, (G) p37 K, and (H) cdc2 K, all resulted in characteristic phosphopeptide maps. Some of the major interphase-specific in vivo phosphopeptides showed a similar migration as the major phosphopeptides generated by PKA and PKC but not CaMK. The phosphopeptides generated by the mitotic kinases cdc2 K and p37 K did not correspond to any of the interphase-specific phosphopeptides but co-migrated with the major mitosis-specific phosphopeptides.

View larger version (31K): [in a new window]   Fig. 4. Chromatographic separation of the major tryptic phosphopeptides from 32P in vivo labeled vimentin. Four 20 cm plates of semi-confluent BHK-21 cells were preincubated for 4 hours with 0.5 mCi/ml [32P]orthophosphate. In order to maximize the in vivo labeling, vimentin dephosphorylation was inhibited by addition of 50 nM cl-A to the cells. Vimentin was isolated by preparative SDS-PAGE as described in Materials and Methods and then subjected to tryptic cleavage. (A) The generated phosphopeptides were first isolated by reversed phase chromatography on a microbore C-18 column. The UV chromatogram (left panel) indicates the presence of a high number of tryptic peptides, among which seven 32P labeled fractions were isolated (right panel; fractions a-g). (B) Peaks a-g obtained on the C18 chromatography were then separated a second time on a reversed phase microbore C-8 column, yielding one single purified peptide from fractions a, b and d-g, and two peptides from fraction c. The inserts show the elution profile of the 32P label, which in all cases corresponded to a peak on the UV chromatogram. The isolated peptides were numbered according to their hydrophobicity on the C8 reversed phase HPLC. The obtained peaks were subjected to automatic sequencing and manual Edman degradation. Peptide masses were confirmed by mass-spectrometry (data not shown). The results are presented in Table 1.

View larger version (30K): [in a new window]   Fig. 5. Identified in vivo phosphorylation sites on vimentin. (A) A summary of all results obtained by phosphopeptide mapping, sequencing and manual Edman degradation. P, identified phosphorylation sites; *, sites that have not been previously described as in vivo phosphorylation sites. (B) The schematic figure shows the relative location of the phosphorylation sites in the N and C termini of vimentin.

View larger version (15K): [in a new window]   Fig. 6. Mutation of the two major PKA sites results in a reduction of vimentin phosphorylation and alters the phosphate distribution on vimentin phosphopeptides. When two-dimensional tryptic phosphopeptide mapping of PKA-phosphorylated (A) wild-type and (B) S38:A/S72:A vimentin was performed, peptide 1 showed negligible phosphorylation in the mutant vimentin, whereas peptide 2 showed reduced phosphorylation. Each map was loaded with peptides generated from 5 µg of 32P-labeled vimentin. The direction of electrophoresis (+,-) and ascending chromatography (arrow) are indicated. (C) An overall reduction in 32P incorporation as a result of the phosphate-site mutations could be seen when 7.5 µg of wild-type (lane 1) and S38:A/S72:A (lane 2) vimentin, respectively, were phosphorylated in vitro by PKA and subjected to one-dimensional SDS-PAGE followed by autoradiography.

View larger version (88K): [in a new window]   Fig. 7. Presence of phosphate groups on Ser-38 and Ser-72 affect filament forming ability following microinjection into BHK-21 fibroblasts. (A) Wild-type (a-c), PKA-phosphorylated wild-type supernatant fraction (d-f), and S38:A/S72:A (g-i) myc-tagged vimentin were microinjected at a concentration of 1 mg/ml into BHK-21 fibroblasts. At 10 minutes (a, d, g), 30 minutes (b, e, h) and 3 hours (c, f, i) post-injection, cells were fixed with methanol and indirect immunofluorescence was performed using a monoclonal antibody directed against the myc-tag to trace the microinjected protein. Filament formation by the injected protein was assessed, and confocal micrographs were obtained, demonstrating the degree of filament formation observed for the majority of cells at a given time point. (B) Co-localization of the injected protein with the endogenous BHK-21 IF network was determined at 3 hours post-injection by double-label immunofluorescence using a polyclonal antibody specific for the endogenous vimentin (a, b, c). The images show the behavior of the S38:A/S72:A mutant pelletable protein but similar results were obtained with all three injected proteins.

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

Twitter