Previous studies demonstrated that stathmin, a microtubule destabilizing protein, promotes cell survival and/or prevents apoptosis, although the underlying molecular mechanism is unknown. Based on a recent work^[1], we are testing the hypothesis that stathmin normally functions to keep the levels of active c-Jun N-terminal kinases (JNK) low. In this model, stathmin, phosphorylated by JNK, functions in a negative feedback loop to inhibit the JNK pathway and limit JNK activation. This model predicts that active JNK will rise in the absence of stathmin; prolonged activation of JNK would then trigger apoptosis. As a first test of whether stathmin regulates JNK activity to control cell survival, we found that treatment with either JNK-IN-8, a JNK inhibitor, or depletion of JNK1,2, prevented cell death in stathmin-depleted HeLa cells. Using a localization-dependent biosensor^[2], we found that active JNK levels were higher in stathmin-depleted cells. Expression of a stathmin phosphomimic restored active JNK level and prevented apoptosis. These data support a model where phosphorylated stathmin, acting independently of the microtubule cytoskeleton, prevents JNK hyperactivation to promote cell survival.

Stathmin/Oncoprotein 18 has been linked to numerous human cancers, where increased stathmin expression is highly correlated with cancer stage progression and chemo-resistance^[3][4]. Previous studies have demonstrated that stathmin depletion is sufficient to slow proliferation and increase cell death in cancer cells^[5][6][7][3], although the mechanism responsible has not been thoroughly described. It is thought that stathmin acts via microtubule destabilization to control cell fate, where stathmin depletion results in greater microtubule stability, a phenotype superficially similar to Taxol treatment^[3]. Stathmin’s microtubule-destabilizing function is inactivated by phosphorylation of up to four serines, where several kinases target one or more sites^[8][9]. Recent work suggested, but did not test, a potential microtubule-independent function for phosphorylated stathmin, where stathmin is both a target of JNK and, in its phosphorylated form, an upstream inhibitor of this kinase^[1]. In this way, phosphorylated stathmin functions as part of a negative feedback loop to prevent hyperactivation of JNK. Importantly, prolonged JNK activation results in apoptosis^[10][11].

To address whether stathmin depletion activates apoptosis via loss of a negative feedback loop and hyperactivation of JNK, we asked whether treatment with a JNK inhibitor, JNK depletion or expression of a stathmin phosphomimic, abrogates apoptosis normally observed in stathmin-depleted cells.

Stathmin depletion increases apoptosis in p53-deficient cancer cells^[5][6]; however, the mechanism by which stathmin regulates cell survival remains unresolved. We hypothesized that stathmin promotes cell survival independently of its regulation of the microtubule cytoskeleton by maintaining low levels of JNK activity. In the absence of stathmin and loss of a negative feedback loop, active JNK levels should rise and eventually activate apoptosis. HeLa cells are a convenient model system to test this hypothesis since they are easily transfected, and previous studies demonstrated that stathmin depletion induces apoptosis in these cells^[7][6][12]. Using a specific JNK inhibitor or siRNA knockdown to abolish JNK activity, we addressed whether JNK is necessary for induction of cell death in stathmin-depleted cells. Treatment with the small-molecule JNK inhibitor, JNK-IN-8, restored cell viability to control levels in stathmin-depleted cells (Fig. 1A). Depletion of JNK 1,2 (Fig. 1B) did not impact cell survival, while co-depletion of stathmin and JNK 1,2 abrogated the cell death observed in cells depleted of stathmin alone (Fig. 1C). These results demonstrate that JNK activity is required for stathmin-depletion-induced cell death.

A model where stathmin normally functions in a negative feedback loop to limit JNK activation predicts that basal levels of JNK activity will rise in the absence of stathmin. We were unable to detect increased phosphorylated (active) JNK in stathmin-depleted cells by western blots. It is possible that increased active JNK occurs asynchronously in a population of cells, much like the asynchrony in the onset of cell death^[12], and, therefore, is undetectable by western blots from cell populations. In order to assess JNK activity in individual cells over time, we used a fluorescently tagged JNK kinase translocation reporter (KTR), a localization-based biosensor of JNK kinase activity^[2]. We confirmed that the JNK KTR localizes appropriately in our system and is translocated from the nucleus to the cytoplasm when cells are treated with a small-molecule JNK activator (Suppl. Fig. S1). Applying the sensor to our experiments, we found that active JNK levels were elevated for at least 72 h post-transfection in stathmin-depleted cells compared to cells treated with non-targeting siRNA, although the increased JNK activity was modest (Fig. 1D,E). We further investigated the subset of stathmin-depleted cells that undergo apoptosis by long-term, live cell imaging. Stathmin-depleted cells expressing the mRuby2-tagged JNK KTR were observed at 24–72 h post-transfection, and JNK activity was quantified on a single cell basis for the 30 min interval prior to cell death. This analysis revealed that JNK activity is significantly elevated in stathmin-depleted cells in the 30 min prior to apoptosis when compared to the neighboring surviving cells (Fig. 1F,G). These data support a role for elevated JNK in stathmin depletion-induced apoptosis, but it is not yet known if JNK activity accumulates over time in all stathmin-depleted cells or if a subset of stathmin-depleted cells is more likely to hyperactivate JNK and undergo apoptosis.

To explore whether JNK hyperactivation occurs due to loss of phosphorylated stathmin, possibly acting to inhibit JNK activation, or whether stathmin depletion functions by stabilizing the microtubule cytoskeleton, we expressed FLAG-tagged stathmin or phospho-mutant constructs in cells depleted of the endogenous protein (Fig. 1H). Expression of FLAG-tagged stathmin abrogates cell death in these cells. A phosphomimic, with all four serine residues mutated to glutamic acid, also abrogates cell death in the absence of endogenous stathmin (Fig. 1I). This Tetra-E mutant has little microtubule destabilizing activity compared to the unphosphorylated protein^[9][13]. In contrast, a nonphosphorylatable Tetra-A stathmin mutant, with serines mutated to alanine, retains its ability to destabilize microtubules^[9][13][14] and fails to prevent cell death after depletion of endogenous stathmin (Fig. 1I). These results support the idea that stathmin acts independently of the microtubule cytoskeleton in order to mediate cell viability. Additionally, expression of JNK KTR reveals that expression of FLAG-tagged stathmin or the Tetra-E phosphomimic restores JNK activity to levels comparable to control cells (Fig. 1J; Suppl. Fig. S2). In comparison, JNK activity remains elevated in cells expressing the Tetra-A mutant. These data support the hypothesis that phosphorylated stathmin maintains low JNK activity, which is necessary to prevent cell death/promote cell survival.

Earlier studies demonstrated that JNK-dependent phosphorylation of stathmin protects against stress-induced apoptosis^[15], supporting our current hypothesis that stathmin, phosphorylated by active JNK, acts in a negative feedback loop to inhibit the JNK activation pathway. However, the target inhibited by phosphorylated stathmin remains unresolved. Any number of kinases upstream of JNK and its activator MKK-4 in the MAPK cascade may be the critical component^[1], though another hypothesis suggests JNK activation is sustained by the inhibition of MAP kinase phosphatases^[16]. To date, these potential feedback mechanisms for stathmin depletion-stimulated apoptosis remain unexplored; however, our results support the idea that stathmin functions independently of the microtubule cytoskeleton to mediate cell survival, likely acting in a negative feedback loop to maintain basal JNK activity.

Stathmin has been shown to bind to two tubulin dimers^[9], and this has been typically described as stathmin functioning to sequester tubulins and limit their polymerization. Given that we have uncovered a non-microtubule based function for stathmin, it is interesting to speculate that tubulin sequesters stathmin, and not the other way around. By sequestering stathmin, tubulin could prevent its phosphorylation and subsequent inhibition of the JNK activation pathway.

The data presented here provide initial tests of a stathmin-JNK negative feedback loop and demonstrate the potential significance of such a feedback loop in controlling cell survival/apoptosis decisions. But much remains unknown, including where in the JNK activation pathway phosphorylated stathmin acts as an inhibitor. We have not uncovered the link between JNK hyperactivation and apoptosis under our experimental conditions. Initial experiments focused on the E3 ubiquitin ligase, ITCH, which when activated by long-term JNK activation degrades the Caspase 8 inhibitor, c-FLIP^[17][18]. To date we have not detected a role for ITCH in activation of apoptosis. ITCH depletion did not abrogate cell death in stathmin-depleted cells, indicating that ITCH is not sufficient to activate apoptosis after stathmin depletion.

An unsolved question is why so many cancers overexpress stathmin. Indeed, increased stathmin expression has been correlated with cancer progression^[19], and at least in hepatocellular cancer, patient survival is predicted equally well by either stathmin overexpression or by p53 mutations^[20]. It is possible that those cancer cells overexpressing stathmin are selected because the high level of stathmin prevents JNK hyperactivation and subsequent apoptosis.

Cell culture and plasmid transfections

HeLa (from ATCC) cells were grown in DMEM (Sigma) supplemented with 10% fetal bovine serum (FBS; Invitrogen) and 1x antibiotic/antimycotic (Sigma). In some experiments, HeLa cells were transfected with plasmids (1 ug per 35 mm dish) for expression of JNK KTR-mRuby2, STMN-FLAG, STMN-TetraA-FLAG, and/or STMN-TetraE-FLAG using XtremeGene HP DNA Transfection Reagent (version 1.0; Roche Diagnostics) according to the manufacturer’s protocol. Cells were transfected approximately 4 h following siRNA transfection and assessed approximately 24–72 h following initial transfection. In experiments requiring plasmid expression of the mRuby2-tagged JNK KTR (1 ug per 35 mm dish), HeLa cells were transferred to phenol-free DMEM supplemented with 10% FBS and 1x antibiotic/antimycotic after two rinses in PBS approximately 4 h post-transfection. All plasmids were constructed by total gene synthesis, custom cloning, and verified by sequencing (performed by Genewiz).

RNA interference and transient transfection

HeLa cells were grown in 35 mm dishes and transfected with siRNAs 1–2 days after plating using GeneSilencer (Genlantis) according to the manufacturer’s protocol. Cells were serum-starved from the time of transfection to 4 h post-transfection. siRNA oligonucleotides were purchased from GE Dharmacon and included: STMN1, 5’- CGUUUGCGAGAGAAGGAUADTDT -3’; 5’UTR targeted-STMN1, 5’- CCCAGUUGAUUGUGCAGAAUU -3’; SMARTpool targeting JNK2. SiGenome non-targeting siRNA sequence was used as control siRNA sequences for these experiments.

Drugs and reagents

Chemical inhibitor to JNK (JNK-IN-8) was purchased from Selleckchem. Activator of JNK, anisomycin, was purchased from Sigma. Drugs were prepared as stocks in DMSO and stored at -20°C.

Indirect immunofluorescence and microscopy

HeLa cells were grown on glass coverslips and treated as described above. Cells were fixed with 4% paraformaldehyde (Electron Microscopy Sciences)/ 20% glycerol in PEM (100 mM PIPES, 1 mM MgSO4, 2 mM ethylene glycol tetraacetic acid, pH 6.9) for 15 min at room temperature. Cells were permeabilized with methanol at -20°C for 5 min. Fixed cells were incubated with blocking reagent (10% FBS in PBS) for 30 min at 37°C. Cells were then incubated with primary antibody for 45 min at 37°C. Cells were next washed with PBS and incubated with secondary antibody for 45 min at 37°C. Antibodies used were: anti-FLAG (1:100; Sigma cat#F7425) and goat anti-rabbit Alexa Fluor 568 (1:50; Invitrogen cat#A-11036). Coverslips were then washed with PBS and mounted on slides with Vectashield (Vector Laboratories). Cells were imaged by wide-field microscopy using a 40x/1.4 numerical aperture planapo objective on an inverted microscope (TE300; Nikon).

Live cell imaging

To follow cell fates over several days, HeLa cells were plated on glass bottom dishes (CellVis) and imaged using a Nikon Biostation IM. Cells were imaged with phase-contrast and fluorescence (eg., 510–560, em:590; G-2A part#96306, Nikon) using a 20x objective, and images were collected at 5 min intervals for 72 h. Cell fates were tracked from image series; cell death was determined based on changes in cell morphology characterized by cell retraction and blebbing of the plasma membrane.

Western blotting

Soluble cell extracts were prepared and protein concentrations were measured by Bradford assay. Lysates were diluted in PAGE sample buffer; 10–20 ug total protein per lane was typically loaded and resolved in 10% polyacrylamide gels and transferred to PVDF membranes (Trans-Blot Turbo Mini; Bio Rad). Membranes were blocked in 5% milk or 5% BSA (for membranes probed with anti-JNK antibody) in Tris-buffered saline with 0.1% Tween and then probed with primary antibodies, including anti-STMN1 (1:2000; EMD Millipore cat#AB2967), anti-JNK (1:1000; Cell Signaling cat#9252), and anti-FLAG (1:1000; Sigma cat#F7425). Membranes were then probed with horseradish peroxidase-linked secondary antibodies, including anti-mouse (1:10,000; Sigma cat#A4416) or anti-rabbit (1:5000; Sigma cat#A0545) IgG. Immunoreactive bands were developed using enhanced chemiluminescence (GE Amersham). Membranes were reprobed with anti-α-tubulin (1:10,000; Sigma cat#T5168) or anti-GAPDH (1:1000; Abcam cat#ab9483) as a loading control.

Cell viability measurements

Cells were allowed to grow for 2 days following treatment with the JNK inhibitor, JNK-IN-8. On the day of measurement, cells were trypsinized and resuspended in PBS with 0.2% Trypan Blue and counted using a hemocytometer to determine viability via Trypan Blue exclusion. Alternatively, cell viability was assessed by the CellTox Green Cytotoxicity Assay (Promega) by the addition of CellTox Green Dye (prepared according to manufacturer’s protocol) approximately 8 h following initial transfection. Cells were allowed to grow for up to 3 days following transfection with siRNA and/or plasmids. Wide-field phase contrast and fluorescence images were acquired at 24 h time points following initial transfection using a 20x/1.4 numerical aperture planapo objective as described above. Approximately 200 cells were counted per condition per experiment. Each experiment was repeated at least three times.

Image analysis

Images were acquired and stored as 12-bit files using MetaMorph or NIS Elements AR 3.2 (Nikon). JNK activity was assessed by the measurement of the integrated intensity ratio of cytoplasmic to nuclear regions (10×10 px) representing active and inactive JNK, respectively. Cytoplasmic regions were measured just outside the nuclear envelope to ensure measurements of consistent cellular volumes.

Data analysis

Statistical analysis of integrated fluorescence intensity and cell viability were performed using unpaired t-tests with either GraphPad Software or Kaleidagraph.

Supported by NIH GM100381.

The authors are indebted to undergraduates Kyle Peters and Arianna Caruso for performing the ITCH experiments.

Not applicable.

No fraudulence is committed in performing these experiments or during processing of the data. We understand that in the case of fraudulence, the study can be retracted by ScienceMatters.