Coordinated Expression of Stathmin Family Members by Far Upstream Sequence Element-Binding Protein-1 Increases Motility in Non–Small Cell Lung Cancer Free

Tumor cell growth, survival, and dissemination particularly depend on highly efficient turnover of the microtubule network which contributes to cellular processes such as cell division and migration. Thus, the modulation of microtubule dynamics represents a promising therapeutic strategy (1). Microtubule polymers consist of α-tubulin and β-tubulin heterodimers and are characterized by continuous transition (dynamic instability) between phases of depolymerization (catastrophe) and polymerization (rescue; ref. 2). Several factors have been identified which facilitate dynamic instability in cancer cells, strongly suggesting that specific targeting of dysregulated microtubule-modulating factors affects the biofunctionality of tumor cells but to a lesser extent that of nonmalignant cells.

Stathmin (synonymous with oncoprotein 18/OP18, metablastin, and LAP18) is the prototype member of a phosphoprotein family that also includes stathmin-like 2 (superior cervical ganglion 10; SCG10), stathmin-like 3 (SCG10-like protein; SCLIP), and stathmin-like 4 (RB3 with two splice variants, RB3′ and RB3′′). This family of proteins has been described to induce microtubule depolymerization (3, 4). In contrast to stathmin and SCLIP, the other family members are exclusively expressed in the nervous system under physiologic conditions (5, 6). Stathmin, SCG10, SCLIP, and RB3 share a so-called stathmin-like domain that contains up to four phosphorylation sites (stathmin: Ser¹⁶, Ser²⁵, Ser³⁸, and Ser⁶³) and, at least for stathmin, it has been shown that dephosphorylation by phosphatases promotes microtubule catastrophe by facilitating depolymerization and sequestration of soluble tubulin heterodimers (7, 8).

Besides its role in cell division, stathmin is also involved in other microtubule-dependent processes such as cell motility (9, 10). Thus, this protein family is likely to support cellular processes essential for tumor progression: survival and migration. Indeed, overexpression of stathmin has been reported for different human malignancies such as breast cancer (11), hepatocellular carcinoma (12), sarcoma (9), and lung adenocarcinomas (13). Elevated expression of stathmin supports microtubule-dependent processes and contributes to tumor cell chemoresistance (12, 14, 15). In addition, one recent report implicated that SCG10 is required for maintaining the anchorage-independent growth state of β-catenin/TCF-activated hepatoma cells (16). Although many studies clearly defined the microtubule-destabilizing activity of stathmin in different cell lines, the efficient knockdown of stathmin by gene-specific short interfering RNA (siRNA) only partially influenced microtubule polymerization in malignant cells suggesting functional compensation by other microtubule catastrophe-promoting factors (12, 17). The identification of functionally redundant proteins would significantly improve the understanding of microtubule-mediated effects in tumor cells and would therefore enable the specific adjustment of therapeutic strategies in the treatment of patients with cancer.

In the present study, we show that two stathmin family members (i.e., stathmin and SCLIP) are overexpressed at the invasion front of non–small cell lung cancer (NSCLC). Reduced expression of both factors is associated with a less aggressive phenotype regarding proliferation, morphology, migration, and invasion of NSCLC cells. Finally, the far upstream sequence element (FUSE)-binding protein (FBP)-1 was identified as a novel inducer of stathmin and SCLIP.

Cell culture, transfection, reagents, sequences, and tissue samples. Media for Calu-1 [squamous cell carcinoma (SCC) cell line (ATCC/LGC); MEM + 1% sodium pyruvate (PAA)], Calu-6 (adenocarcinoma cell line, MEM; ATCC/LGC), H157 (SCC cell line, DMEM; ATCC/LGC), A549 (adenocarcinoma cell line, RPMI + 1% l-glutamine; ATCC/LGC), and HT1299 (adenocarcinoma cell line, DMEM; ATCC/LGC), were supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin and cultured at 37°C in a 5% CO₂ atmosphere. SiRNA transfection assays were performed as previously described (18). The sequences of the gene-specific siRNAs are listed in Supplementary Table S1.

Thirty-six fresh-frozen NSCLC and respective tumor-free lung tissue samples were obtained from patients who underwent resection for lung cancer at the Department of Thoracic Surgery, Thoraxklinik Heidelberg-Rohrbach (Germany). Informed consent was obtained from all patients. The study protocol was approved by the local Ethics Committee of the University of Heidelberg (application no. 270/2001).

Each tumor sample was snap-frozen in liquid nitrogen within 30 min after resection and stored at −80°C. Samples were stained with H&E and carefully reviewed by the study pathologist (P.A. Schnabel) to determine the proportion of viable tumor cells, stromal cells, normal lung cell content, infiltrating lymphocytes, and necrotic areas. Only sections with ≥50% viable tumor cells were considered for further analyses. Matched distant lung tissue samples were evaluated by the pathologist to be free of tumor.

Sample preparation, real-time PCR, and Western blot analyses. For sample preparations of total RNA from NSCLC and lung tissues, 10 to 15 cryosections (10 μm each) were homogenized with the TissueLyser mixer-mill disruptor (2 × 30 s, 3.00 Hz; Qiagen). Matched tumor-free tissue pieces were homogenized using the Miccra D-8 with DS-5/K1 (2 × 30 s, 23.500 min⁻¹; Art-Moderne Labortechnik). Total RNA was isolated using the RNeasy Mini Kit (Qiagen). The quantity of RNA was measured with a NanoDrop ND-1000 Spectrophotometer and the quality of total RNA was assessed with Agilent 2100 Bioanalyzer (Agilent Technologies). Samples were considered sufficient for further analysis if the RNA integrity number was ≥8.0. Total RNA isolation from cell lines was performed using the NucleoSpin RNA-II kit (Macherey-Nagel).

Protein extracts were isolated from frozen tissue samples using the mixer mill MM 200 with steel grinding jars and 5 mm steel grinding balls (Retsch). One hundred milligrams of the sample were crushed in cell lysis buffer (New England Biolabs), centrifuged at 18,000 × g (4°C, 15 min), and quantified using the Bradford assay.

For comprehensive semiquantitative real-time PCR analyses of stathmin, SCG10, SCLIP, RB3, FBP-1, and β2-microglobulin from NSCLC and lung tissue samples, reverse transcription (1 μg total RNA) and PCR, reactions (in quadruplicate) were prepared by a liquid handling system robot (Theonyx Liquid Performer, Aviso GmbH). The following cycling program was applied using the Absolute qPCR mix (ABgene): 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 60 s (Prism 7300; Applied Biosystems). Western blot analyses were performed as described previously (12). The sequences of the primers and antibodies used are listed in Supplementary Table S1.

Functional assays: cell viability, proliferation, migration, and invasion. All functional assays were performed 3 days after transient transfection of siRNAs. Cell viability was analyzed in 96-well plates using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetra-zolium-bromide (MTT) assay. Forty-eight hours after seeding, the medium was removed and the MTT solution (0.5 mg/mL in medium) was added. After 2 h, the MTT solution was removed and tetrazolium salt was resolved in 100 μL of DMSO/ethanol solution (1:2). Colorimetric measurement was performed at 570 nm using an ELISA reader. Cell proliferation was examined using a bromodeoxyuridine ELISA (Cell Proliferation ELISA Biotrak, GE Healthcare/Amersham). The number of viable and proliferating cells was calculated based on appropriate standard curves.

For assessing the migratory ability in a two-dimensional assay, tumor cells were treated with mitomycin C (5 μg/mL) for 3 h to repress proliferation. The cell monolayer was scratched using a pipette tip. Cells were stimulated with hepatocyte growth factor (10 ng/mL) to induce migration, and incubated for a further 18 h. Scratches were digitally documented 0 and 18 h after being made, and relative migratory activity was calculated based on the cell-free areas.

For analyzing the invasion capacity in a three-dimensional assay, tumor cell spheroids were formed according to a previously published protocol 24 h after transfection with siRNAs (19). Subsequently, the invasion capacity of the tumor cells was digitally monitored after 48 and 72 h.

Tissue microarray analysis, immunohistochemistry, and immunofluorescence. The tissue microarray (TMA) contained two representative areas (0.6 mm diameter) from the tumor center and the invasion front of 117 NSCLCs as well as 117 corresponding nontumorous lung tissues (adenocarcinoma, 58; SCC, 59; grading: 52 grade 2 and 65 grade 3). Cases were all resected at the Department of Thoracic Surgery, Thoraxklinik Heidelberg-Rohrbach and histologically classified according to established criteria by two experienced pathologists (A. Warth and E. Herpel). For assessing the staining for stathmin and SCLIP, the product of cytoplasmic intensity and quantity was calculated based on the previously described scoring system (12). The evaluation was performed independently by two experienced investigators. Immunohistochemistry was performed as previously described with slight modifications for stathmin (citrate buffer) and SCLIP (EDTA–buffer; ref. 20).

For immunofluorescence staining, cells were seeded onto coverslips 1 day after transfection. After 2 days, coverslips were fixed with ice-cold methanol/acetone. The cells were then incubated with anti–α-tubulin antibody and subsequently labeled with a secondary Cy3-linked anti-mouse antibody. After DAPI counterstaining, representative cells were digitally documented using the Leica TCS SL fluorescence microscope system (Leica). The antibodies and respective dilutions used are listed in Supplementary Table S1.

Statistical analysis and software. Data are presented as the mean ± SD. The Spearman rank coefficient was used as a statistical measure of association. The statistical comparison between two groups was accomplished with the nonparametric Mann-Whitney U test, and the significance levels were defined as *, P < 0.05, **, P < 0.01, and ***, P < 0.001. Untreated cells were used for calibration and nonsense siRNA transfected cells were used for statistical evaluation. The AlphaEase FC software was used for densitometric quantification of Western blot signals (V4.0; Alpha Innotech/Biozym). Two-dimensional area determination for migration and invasion assays as well as areas of nuclei and cytoplasm was performed using the digital image analysis software AxiovisionLE (Zeiss).

Stathmin and SCLIP are overexpressed in human NSCLC. In order to define the status of microtubule-destabilizing factors in primary human NSCLCs, the expression of all known stathmin family members (stathmin, SCG10, SCLIP, and all RB3 isoforms) in 36 NSCLCs as compared with respective tumor-free lung tissues was analyzed by real-time PCR. Although no expression of RB isoforms was detected in any sample, elevated transcript levels for stathmin (80%; Fig. 1A), SCLIP (51%; Fig. 1A), and SCG10 (75%; Supplementary Fig. S1A) were observed in NSCLCs compared with corresponding lung tissues. Surprisingly, the mRNA expression profiles of stathmin, SCLIP, and SCG10 were significantly correlated (stathmin/SCLIP, r = 0.40, P < 0.05; stathmin/SCG10, r = 0.42, P < 0.01; SCLIP/SCG10, r = 0.35, P < 0.05).

Accordingly, at the protein level, an increased expression of stathmin (70%; Fig. 1B) and SCLIP (31%; Fig. 1B) were determined in the same NSCLC tissues as compared with respective controls. For stathmin, no down-regulation was detected in NSCLCs whereas reduced expression of SCLIP was observed in 8% of all NSCLC samples. In contrast, an inhomogeneous expression for SCG10 was seen in NSCLCs at the protein level (induced, 8%; reduced, 31%; not regulated, 61%; Supplementary Fig. S1B). To determine whether posttranscriptional mechanisms might influence the amounts of all three microtubule-modulating proteins, we correlated the relative induction on mRNA (real-time PCR) and protein level (densitometric analysis) of stathmin, SCLIP, and SCG10. Although both stathmin and SCLIP transcript and protein profiles significantly correlated with each other (r = 0.446, P < 0.01; r = 0.590, P < 0.01, respectively), no such correlation was detected for SCG10 expression levels suggesting posttranscriptional processing (r < 0.2, P > 0.1). Protein levels of stathmin in nontumorous and tumorous samples significantly correlated with tumor dedifferentiation (r = 0.585, r < 0.01) and to SCLIP expression (r = 0.382, P < 0.01). This suggested that stathmin and SCLIP are jointly induced in NSCLCs at the mRNA and protein levels.

The group of NSCLCs included tumors of various histologic subtypes, especially SCC and adenocarcinoma. In order to further define the expression of both consistently induced stathmin family members in tumors with squamous or glandular differentiation, TMA analyses in a larger group of NSCLC patients were performed. Scattered cells of the basal layer of normal bronchial epithelium and some inflammatory cells stained positive for stathmin, although for SCLIP, a weaker and more diffuse immunoreactivity was observed in nontumorous epithelium (Fig. 2). In the group of adenocarcinomas, a prominent expression of both proteins was detected (stathmin, 42%; SCLIP, 57%). Likewise, in SCCs, both proteins were highly expressed (stathmin, 76% of cases; SCLIP, 52% of cases) with stathmin levels significantly higher than those seen in adenocarcinomas (P < 0.01). In addition, stathmin and SCLIP expression were especially prominent at the invasion front of tumors. The mutual correlation between stathmin family members at the transcript and protein levels and the simultaneous overexpression of stathmin and SCLIP in 37% of the NSCLCs (TMA data) suggests a common higher-order regulation of these microtubule-destabilizing proteins.

Stathmin and SCLIP support tumor cell viability and proliferation of NSCLC cells. The expression data showed up-regulation of stathmin and/or SCLIP in the vast majority of NSCLCs. Therefore, the functional effects of both microtubule-destabilizing factors after siRNA-mediated inhibition were analyzed using different NSCLC cell lines. Based on the TMA analyses, we screened for NSCLC cell lines that reflected the most typical expression characteristics of stathmin and SCLIP with regard to the respective histologic subtype: intensive expression of stathmin in SCC and mutual expression of both factors in adenocarcinoma. Calu-1 (SCC) and Calu-6 (adenocarcinoma) cells met these criteria and were used for further functional assays (Supplementary Fig. S2).

Transient transfection of three independent gene-specific siRNAs for human stathmin and SCLIP were optimized for serum content of culture media, siRNA, and cationic carrier concentration in Calu-1 and Calu-6 cells. Application of chemically labeled siRNA revealed a transfection efficiency of >95% for both analyzed cell lines (siGLO indicators; data not shown). Stathmin and SCLIP were reduced at the respective transcript levels as compared with untransfected and nonsense siRNA-transfected cells (Calu-1, stathmin >95% reduction; Calu-6, stathmin >80%; SCLIP >60%; Supplementary Fig. S3). Efficient inhibition of both stathmin family members in Calu-6 cells was achieved by combining both siRNAs for each target transcript (40 nmol/L for both siRNAs as compared with 20 nmol/L for single gene inhibition). Because the highest efficiency of knockdown was seen 3 days following siRNA treatment, this condition was chosen for further experiments (Fig. 3A and B). No cross-reactivity of the different siRNAs was detected and the combination of two siRNAs targeting stathmin and SCLIP allowed the efficient reduction of both proteins in Calu-6.

Efficient inhibition of stathmin in Calu-1 and stathmin/SCLIP in Calu-6 for 3 days was associated with diminished tumor cell viability (Fig. 3C and D). The combined inhibition of both microtubule-destabilizing proteins in Calu-6 did not lead to a further reduction of viability as compared with single gene inhibition. Similarly, tumor cell proliferation was diminished after single gene inhibition of stathmin and stathmin/SCLIP in Calu-1 and Calu-6, respectively (Fig. 3C and D). Again, no additive effects on proliferation were observed after the combined reduction of both target genes in Calu-6. Only a minor increase of apoptosis was detected in both cell lines after stathmin and stathmin/SCLIP reduction (data not shown).

Stathmin and SCLIP support tumor cell motility and invasion of NSCLC cells. Because the expression of stathmin and SCLIP was typically highest at the invasion front of many NSCLCs and because effects on microtubule dynamics has been described to influence tumor cell migration (9, 10, 12), the effects of both factors on motility and invasiveness were analyzed in NSCLC cells.

Both the siRNA-mediated knockdown of stathmin in Calu-1 as well as the reduction of stathmin or SCLIP in Calu-6 drastically diminished cell migration in a two-dimensional scratch assay as compared with nonsense siRNA-transfected cells (Calu-1, stathmin >52% reduction; Calu-6, stathmin >66% reduction, SCLIP >37% reduction; Fig. 4A and B). The combined inhibition of stathmin and SCLIP in Calu-6 further significantly reduced the ability of tumor cells to close the scratch by active migration compared with single gene inhibition (stathmin siRNA no. 1 versus combi-1: Calu-6, > 85% reduction; P < 0.05; Fig. 4B).

In order to analyze the effect of microtubule-destabilizing factors with regard to invasiveness of NSCLC cells, we performed matrix invasion assays after inhibition of stathmin and SCLIP. Blocking of stathmin expression in Calu-1 cells significantly reduced tumor cell invasion as compared with appropriate controls (>55% reduction; Fig. 4C). Equally, in Calu-6, the single gene inhibition of stathmin and SCLIP was associated with a diminished capacity to penetrate the matrix (stathmin, >40% reduction; SCLIP, >10% reduction; Fig. 4D). Surprisingly, the invasion capacity was not further reduced after combined efficient inhibition of both microtubule-destabilizing factors in Calu-6 in contrast to the scratch assay (Fig. 4D).

Because the elevated expression of microtubule-destabilizing proteins should result in a reduction of stabilized/polymerized tubulin (4, 12), we reasoned that changes in tumor cell morphology may correlate with the observed changes in microtubule-dependent processes such as migration and invasion. Therefore, the nuclear and cytoplasmic two-dimensional extension of NSCLC cells was measured with and without inhibition of stathmin family members. Indeed, the cytoplasmic area significantly increased after stathmin reduction in both cell lines with clear effects on cell morphology, including a decreased nuclear to cytoplasmic ratio (Fig. 5A and B). The cytoplasm of tumor cells with reduced stathmin and SCLIP expression was enlarged with reduced cytoplasmic integrity as compared with controls. The observed effects were even stronger after SCLIP inhibition and combined reduction of stathmin and SCLIP in Calu-6 cells (Fig. 5B). It is worth mentioning that efficient single and combined inhibition in NSCLC cells only partially affected the stability of microtubule polymers with the strongest effects following SCLIP inhibition (20–30% increased polymer stability as measured by tubulin assay; data not shown, ref. 12), again suggesting that additional redundant destabilizing factors provide residual microtubule dynamics.

Stathmin and SCLIP are induced by the single-strand DNA binding factor FBP-1. Correlated expression of stathmin and SCLIP in NSCLCs at the transcript levels suggested potential coregulation by a common upstream factor (Fig. 1A). Because EZH2 and wild-type/mutated p53 have recently been described to modulate stathmin expression (12, 21), we speculated that these factors might also influence stathmin and SCLIP expression in NSCLC cells. However, siRNA-mediated reduction of EZH2 or wild-type/mutated p53 did not significantly affect stathmin and SCLIP expression in NSCLC cells (data not shown).

We therefore asked whether other factors might regulate the expression of both stathmin family members. Based on comprehensive NSCLC microarray expression data (22), the FUSE-binding protein was identified as coregulated with SCLIP (r = 0.3). Indeed, FBP-1 was strongly expressed in many tumor cell lines including NSCLC cells (data not shown). FBP-1 has been described to affect target gene protein levels through different molecular mechanisms such as direct interaction and subsequent regulation of transcript availability as well as transcriptional regulation through promoter binding.

Therefore, the expression of FBP-1 protein levels was reduced using gene-specific siRNA-mediated knockdown in Calu-1 and Calu-6 cells. Indeed, efficient inhibition of FBP-1 in both cell lines reduced stathmin (Calu-1) as well as stathmin and SCLIP expression (Calu-6; Fig. 6A). Vice versa, elevated expression of FBP-1 was associated with increased stathmin and SCLIP expression in NSCLC cells (data not shown). In order to verify the coexpression of FBP-1 and all elevated stathmin family members in NSCLC tissues, the transcript levels of stathmin, SCG10, SCLIP, and FBP-1 were correlated. A significant correlation with elevated FBP-1 mRNA levels was detected for all three stathmin family members (Fig. 6B; Supplementary Fig. S4). Moreover, the expression of FBP-1 in NSCLCs and respective nontumorous lung tissues was analyzed by Western immunoblotting. Indeed, expression of FBP-1 was strongly increased in >70% of NSCLCs as compared with nontumorous lung tissues and significantly correlated with stathmin (r = 0.81; P < 0.01) and SCLIP (r = 0.4; P < 0.01) at the protein levels (Figs. 1 and 6C).

As the most common type of lung cancer, NSCLC accounts for 80% of all cases including the major subtypes, adenocarcinoma and SCC with distinct clinical, histologic, and molecular characteristics. Identification and dissection of the molecular mechanisms underlying the development and progression of NSCLC is essential for the improvement of current treatment regimens and may provide novel therapeutic strategies (23). Regulating the dynamic instability of the microtubule apparatus represents a promising therapeutic strategy because rapid changes in the microtubule network are critically involved in many tumor cell–relevant processes such as increased mitosis and motility while showing only little effects in normal resting cells.

Here, we identified two microtubule-destabilizing proteins, i.e., stathmin and SCLIP, as highly overexpressed in 70% to 80% (stathmin) and 31% to 51% (SCLIP) of NSCLC tissues as compared with respective normal lung tissues. Both factors independently contributed to tumor cell viability, proliferation, motility, and elevated nuclear to cytoplasmic ratio, suggesting that reduced availability of stathmin and SCLIP is functionally and morphologically associated with a “less” malignant phenotype (24). Because inhibition of proliferation was incomplete after efficient protein reduction and combined reduction of stathmin/SCLIP did not further minimize the observed biological effects, the existence of additional, functionally redundant microtubule-destabilizing factors (non-stathmin family members) are likely to be active. This is supported by the fact that >15% of all NSCLCs samples did not exhibit detectable stathmin or SCLIP expression (TMA data).

The effects of stathmin and SCLIP reduction on different cellular processes suggests that selective inhibition of microtubule-interacting factors might sensitize cancer cells for further antineoplastic treatment. Indeed, sensitization of tumor cells after inhibition of stathmin levels has been reported for breast cancer (17), osteosarcoma (15), and HCC cells (12). This is also valid for NSCLC cells because inhibition of stathmin and SCLIP in both analyzed cell lines significantly increased responsiveness to irradiation (Supplementary Fig. S5). Moreover, overexpression of stathmin was shown to sensitize lung cancer cells to treatment with antimitotic Vinca alkaloids, suggesting that any imbalance of the microtubule network affected tumor cell maintenance and sensitivity (25). Very recently, it was shown that chemical modification of stathmin by nitrosoureas inhibits microtubule depolymerization, further supporting the notion that direct targeting of these family members represents a promising therapeutic strategy (26, 27).

Surprisingly, stathmin and SCLIP did not equally affect tumor cell properties after combined inhibition. Although additive effects on cell migration were detected as compared with single gene inhibition, no such effects were observed for other functions. Indeed, each stathmin family member exhibits structural characteristics which are associated with distinct functional specifications (4). All factors contain a COOH-terminal stathmin domain including subdomains with different degrees of amino acid homology (e.g., SCLIP/stathmin, 36–73%). Particularly, stathmin and SCLIP differ in the number of phosphorylation sites (stathmin, four sites; SCLIP, three sites) and the extent of the NH₂-terminal palmitoylation sites. These sites are essential for the association of SCG10, SCLIP, and the splice variants RB3/RB3′/RB3′′ with Golgi and vesicular membranes, whereas stathmin is exclusively found in the cytoplasm (28, 29). Therefore, differential posttranslational modifications (phosphorylation and palmitoylation) might account for a distinct subcellular localization and partly specific biological effects of stathmin family members (28). The structural specificities of microtubule-destabilizing factors might therefore explain the observed differences between additive (migration) and nonadditive effects (proliferation) after combined inhibition of stathmin and SCLIP in NSCLC cells.

Next to differential expression and posttranslational alterations of microtubule-destabilizing factors, there is increasing evidence for a large number of reversible posttranslational modifications which generate a diversity of α/β-tubulin heterodimers with molecular and functional peculiarities (30). In particular, acetylation of the microtubule network has been described to affect cell motility, whereas polyglycylation is critically involved in cytokinesis (31, 32). Because the stathmin family members significantly differ in cellular localization and structure, distinct microtubule-destabilizing factors are likely to affect different posttranslationally modified microtubule network components and might display only partial functional redundancy with other family members.

Interestingly, we detected differences between two-dimensional migration and three-dimensional invasion after combined inhibition of stathmin and SCLIP. Accordingly, controversial results regarding the influence of stathmin on cell migration in a two-dimensional setup have been reported in the literature. Although no effects after stathmin overexpression (wild-type and a mutated hyperactive isoform) were observed in fibrosarcoma cells, significant inhibitory effects were detected in human HCC cells, glioma cells, and murine embryonic fibroblasts after stathmin reduction (9, 12, 26, 33). In our cell lines representing pulmonary SCC and adenosarcoma cells, significant migratory effects in a two-dimensional and three-dimensional setup were detected, suggesting the cell type–specific effects of both processes after manipulation of stathmin bioavailability. Because additive biological effects after combined inhibition of both microtubule-destabilizing factors were observed for migration but not invasive behavior of NSCLC cells, we conclude that both processes impose different strains on tumor cells. This hypothesis is supported by recently published data showing that proteins differentially affect cell migratory and invasive behavior (34).

We were able to show that overexpression of FBP-1 induces different microtubule-destabilizing proteins in NSCLC cells. FBP-1 and its relatives, FBP-2 and FBP-3, belong to a multifunctional family of factors that have been shown to bind and modify RNA stability, editing, and trafficking (35, 36). In addition, FBPs bind the previously described sequence of single-stranded FUSEs in target gene promoters (37). Torsional stress, which induces melting and transient perturbation of susceptible sequences, are induced by counter-rotation of DNA, enabling FBP to recognize and bind appropriate single-stranded promoter regions. Therefore, the FBP network, together with its partner protein FBP-interacting repressor, represent a molecular mechanism for induction and fine-tune regulation of target gene expression (38, 39). To date, different modes of regulation have been described for stathmin such as induction by specific mutations of the tumor suppressor gene p53 (12), modulation of the chromatin structure by enhancer of zeste homologue 2 (EZH2; ref. 21), and small noncoding RNAs (miRNA-223; ref. 40). Induction of stathmin family members by FBP-1 represents an additional regulatory level likely based on the adjustment of transcription initiation as well as controlling mRNA processing and stability. However, it is worth mentioning that we did not find FUSE elements in the promoter of STMN1 (stathmin gene) and STMN3 (SCLIP gene) as was previously described for c-MYC (37).

In addition, induction of stathmin and SCLIP by FBP-1 at the transcript level represents a mechanism for the regulation of microtubule dynamics. Subsequently, posttranscriptional and reversible posttranslational modifications of protein levels/bioactivity lead to the required adjustment of microtubule dynamics in tumor cells which are required for progression through the cell cycle and motility.

Targeting common regulators or distinct effectors responsible for microtubule dynamics is of special therapeutic significance because it may enable us to focus on selected cell functions (proliferation or migration), and thus, progression steps during carcinogenesis.

No potential conflicts of interest were disclosed.

Grant support: Initiative and Networking Fund of the Helmholtz Association within the Helmholtz Alliance on Immunotherapy of Cancer (P. Schirmacher).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Sarah Messnard for excellent technical assistance and Anthony Barsotti for critical reading of the manuscript.