The following commercial antibodies were used: Mouse monoclonal anti-vimentin (V9, Dako, Tokyo, Japan) and anti-β-actin (AC-15, Sigma, Tokyo, Japan); rabbit polyclonal anti-vimentin (SC-5565, Santa Cruz Biotechnology, Dallas, USA) and anti-GST (60–021, BioAcademia, Osaka, Japan); horseradish peroxidase (HRP)-conjugated goat F(ab')₂ anti-mouse (710–1332, Rockland Immunochemicals, Limerick, USA) and goat anti-rabbit IgG (111–035-003, Jackson ImmunoResearch Laboratories, West Grove, USA).

The MIA PaCa-2 human pancreatic cancer cell line JCRB0070 and the mouse embryonic fibroblast NIH3T3 cell line JCRB1503 were purchased from the Japanese Collection of Research Bioresources Cell Bank (Osaka, Japan). The mouse fibroblast L cell line CRL-2648 was obtained from the American Tissue Culture Center (Manassas, USA). To maintain authenticity of the cell lines, multiple aliquots of frozen stocks were prepared from initial stocks, and every 3 months, a new frozen stock was used for the experiments. The cells were routinely inspected for identity by morphology and growth curve analysis and were validated to be mycoplasma free. The cells were cultured in DMEM medium containing 10% fetal bovine serum and were maintained at 37°C in a humidified atmosphere with 5% CO₂.

Full-length and truncated mutants of the vimentin protein were prepared by polymerase chain reaction (PCR) amplification using the Pfu polymerase (BioAcademia) and HeLa-derived cDNA (full-length) or the full-length human vimentin cDNA (truncated mutants) as the template. The primers used, containing BamHI and EcoRI restriction sites as indicated (underlines), were 5′-GGGGATCCTCCACCAGGTCCGTGTCCTCG-3′ and 5′-GGGAATTCTTATTCAAGGTCATCGTGATGCTG-3′ for full-length vimentin (amino acids 2 to 466); 5′-GGGGATCCTCCACCAGGTCCGTGTCCTCG-3′ and 5′-GGGAATTCTTAGGCAGCCGTGAGGTCAGGCTT-3′ for vim-NT (amino acids 2 to 268); 5′-GGGGATCCGCAGCCATGGACAAAGCCCGCGTCGAGGT-3′ and 5′-GGGAATTCTTATTCAAGGTCATCGTGATGCTG-3′ for vim-CT (amino acids 167 to 466); 5′-GGGGATCCAACACCCTGCAATCTTTCAG-3′ and 5′-GGGAATTCTTAGCGCAGGGCGTCATTGTTCCG-3′ for vim-CT1 (amino acids 201 to 310); 5′-GGGGATCCGCTGCCAACCGGAACAATGAC-3′ and 5′-GGGAATTCTTACAGCAGCTTCCTGTAGGTGG-3′ for vim-CT2 (amino acids 301 to 404); and 5′-GGGGATCCGCCCTTGACATTGAGATTGCC-3′ and 5′-GGGAATTCTTATTCAAGGTCATCGTGATGCTG-3′ for vim-CT3 (amino acids 392 to 466). The amplified fragments were digested with BamHI and EcoRI and cloned into the pGEX-4T2 vector (GE Healthcare, Buckinghamshire, England) digested with the same two enzymes. BL21 bacteria were transformed with the plasmids, and protein expression was induced with 1 mM isopropyl-β-D-1-thiogalactopyranoside (IPTG). The final constructs were sequenced to ensure that no mutations had occurred during the PCR and cloning processes.

The point mutation construct (N417T) was generated using the PCR overlap extension method. Two reactions were performed in the first round of PCR using full-length vimentin as a template and the following primer pairs, containing BamHI and EcoRI restriction sites as indicated (underlines) 5′-GGGGATCCTCCACCAGGTCCGTGTCCTGGG0-3′ and 5′-GGTTCAGGGAGGAAAAGGTTGGAAGAGGGT0GGT1GGT2-3′; and 5′-TCTCTGCCTCTTCCAACCTTTTCCTCCCTGAAC-3′ and 5′-GGGAATTCTTATTCAAGGTCATCGTGATGCTG-3′. The two purified PCR products were mixed with the primer pair 5′-GGGGATCCTCCACCAGGTCCGTGTCCTGGG0-3′ and 5′-GGGAATTCTTATTCAAGGTCATCGTGATGCTG-3′ for the second round of PCR. The PCR product was cloned into the BamHI/EcoRI sites of the pGEX-4T2 vector for expression in E. coli. The target mutation was identified by sequence analysis of the region of the mutation.

All short vimentin fragments were generated by annealing complementary oligos. The oligos used to synthesize the indicated fragments were: 5′-GATCCATTTCTCTGCCTCTTCCAAACTTTTCCTCCCTGAACCTGTAAG-3′ and 5′-AATTCTTACAGGTTCAGGGAGGAAAAGTTTGGAAGAGGCAGAGAAATG-3′ (vim 411–423), 5′-GATCCATTTCTCTGCCTCTTCCAACATTTTCCTCCCTGAACCTGTAAG-3′ and 5′-AATTCTTACAGGTTCAGGGAGGAAAATGTTGGAAGAGGCAGAGAAATG-3′ (vim 411–423_N417T), 5′-GATCCTCTCTGCCTCTTCCAAACTTTTCCTCCCTGAACTAAG-3′ and 5′-AATTCTTAGTTCAGGGAGGAAAAGTTTGGAAGAGGCAGAGAG−3′ (vim 412–422), 5′-GATCCCTGCCTCTTCCAAACTTTTCCTGAT0GAT1GAT2-3′ and 5′-AATTCTTACAGGGAGGAAAAGTTTGGAAAT0AAT1AAT2-3′ (vim 413–421). Annealing reactions containing equal molar amounts of each oligo were performed in the thermal cycler as follows: Initial denaturation at 98°C for 3 min followed by a temperature decrease of 2°C every 20 s down to 4°C to allow the oligos to anneal slowly. The oligos contained the BamHI/EcoRI restriction sites for cloning of the vimentin fragments into these restriction sites within the pGEX-4T2 vector.

The total protein samples from cell lysates after IPTG induction were mixed with SDS buffer (125 mM Tris-HCl, pH 6.8, 5% glycerol, 4% SDS, 0.005% bromophenol blue, and 10% 2-mercaptoethanol). Extracts were subjected to 12% SDS-polyacrylamide gel electrophoresis and transferred onto a polyvinylidene difluoride membrane (Merck Millipore, Darmstadt, Germany). After blocking with 5% nonfat dry milk for 30 min, the membranes were incubated with the indicated antibodies for 1 h at room temperature followed by incubation with secondary antibodies (HRP-conjugated anti-mouse or anti-rabbit IgG at the appropriate dilutions) for 45 min at room temperature. Antibody binding was detected with an enhanced chemiluminescence detection system (PerkinElmer, Waltham, USA).

Sequences were aligned with Clustal Omega (http://www.ebi.ac.uk/Tools/msa/clustalo/).

To map the epitope recognized by the V9 antibody, we first generated recombinant N-terminal (aa 2–268) or C-terminal (aa 167–466) vimentin protein fragments fused with GST (Fig. 1A). Western blot analysis showed that the V9 mAb bound to the C-terminal but not to the N-terminal vimentin fragment (Fig. 1B), confirming previous reports (16,17). To further determine the precise location of the V9 mAb epitope, we constructed three GST-tagged C-terminal truncated mutants of vimentin, vim-CT1 (aa 201–310), vim-CT2 (aa 301–404) and vim-CT3 (aa 392–466) (Fig. 1C). These mutants were expressed in E. coli and their reactivity with the V9 mAb was examined by Western blotting. Of these mutants only vim-CT3 was recognized by the V9 mAb; neither vim-CT1 nor vim-CT2 reacted with the V9 mAb (Fig. 1D). These results suggested that the epitope recognized by the V9 mAb is located within amino acid residues 392–466 of the vimentin protein.

The antibody V9 was shown to react with the vimentin of various mammalian species, including human, hamster, rat, and pig (18), but it did not detect vimentin in mouse cells or chicken fibroblasts (18). To confirm this observation, we first examined whether the V9 mAb could recognize vimentin in the mouse fibroblast cell line, NIH 3T3, which normally expresses endogenous vimentin (19). Western blot analysis of vimentin in human (MIA PaCa-2) and mouse (NIH 3T3) cell lines confirmed that the V9 mAb recognizes the human but not the mouse vimentin protein (Fig. 2A). Similar results were obtained using another mouse L fibroblast cell line, in which expression of mouse vimentin of the cell line was detected with a rabbit anti-vimentin polyclonal antibody (Fig. 2B). To investigate the mechanism behind the inability of the V9 mAb to recognize vimentin in mouse cells, we compared the amino acid residues (residues 392–466) flanking the epitope among different species (Fig. 2C). This sequence comparison showed that most of the amino acid residues were conserved among the species analyzed. However, one amino acid that was unique to mouse was observed at position 417; mouse vimentin contains threonine at position 417, whereas human, hamster, rat, and pig have an asparagine residue at the corresponding position. We therefore hypothesized that the inability of the V9 mAb to recognize mouse vimentin was probably attributable to the variation at position 417. To test this hypothesis, we generated an expression construct in which a single amino acid substitution, that of threonine for asparagine at position 417, was introduced into the full-length GST-tagged human vimentin protein. The plasmid construct was verified by DNA sequence analysis and the recombinant protein was expressed in E. coli BL21 cells. Western blot analysis showed that the mutation N417T completely inhibited V9 antibody binding (Fig. 2D). These results indicate that the asparagine residue at position 417 is a critical residue for V9 antibody binding and that its replacement with threonine at position 417 effectively inhibited this interaction.

To precisely define the amino acid residues that are involved in V9 binding, we constructed a set of GST-tagged vimentin fragments with amino acid sequences containing the asparagine at position 417 that was implicated as being essential for binding based on our above results. These truncated peptide fragments were expressed and analyzed for their reactivity with the V9 mAb by Western blotting. The blotting showed that the V9 mAb could only recognize the peptide vim 411–423 (⁴¹¹ISLPLPNFSSLNL⁴²³) (Fig. 3). As expected, asparagine to threonine substitution at position 417 (vim 411–423_N417T) abolished V9 mAb binding. Neither of the peptides vim 412–422 (⁴¹²SLPLPNFSSLN⁴²²) or vim 413–421 (⁴¹³LPLPNFSSL⁴²¹) were recognized by the V9 mAb. We concluded that the thirteen amino acid residues ⁴¹¹ISLPLPNFSSLNL⁴²³ in vimentin represent the minimal requirement for recognition of the epitope by the V9 mAb.