The LKB1 tumor suppressor differentially affects anchorage independent growth of HPV positive cervical cancer cell lines

Virology 446 (2013) 9–16

Contents lists available at ScienceDirect

Virology journal homepage: www.elsevier.com/locate/yviro

The LKB1 tumor suppressor differentially affects anchorage independent growth of HPV positive cervical cancer cell lines Hildegard I.D. Mack 1, Karl Munger n Division of Infectious Diseases, Brigham and Women's Hospital and Department of Medicine, Harvard Medical School, Boston, MA, 02115, USA

art ic l e i nf o

a b s t r a c t

Article history: Received 2 April 2013 Returned to author for revisions 3 May 2013 Accepted 8 July 2013

Infection with high-risk human papillomaviruses is causally linked to cervical carcinogenesis. However, most lesions caused by high-risk HPV infections do not progress to cancer. Host cell mutations contribute to malignant progression but the molecular nature of such mutations is unknown. Based on a previous study that reported an association between liver kinase B1 (LKB1) tumor suppressor loss and poor outcome in cervical cancer, we sought to determine the molecular basis for this observation. LKB1negative cervical and lung cancer cells were reconstituted with wild type or kinase defective LKB1 mutants and we examined the importance of LKB1 catalytic activity in known LKB1-regulated processes including inhibition of cell proliferation and elevated resistance to energy stress. Our studies revealed marked differences in the biological activities of two kinase defective LKB1 mutants in the various cell lines. Thus, our results suggest that LKB1 may be a cell-type specific tumor suppressor. & 2013 Elsevier Inc. All rights reserved.

Keywords: Human Papillomaviruses LKB1 kinase Peutz-Jeghers syndrome Cervical cancer Anoikis Tumor suppressor

Introduction Infection with high-risk alpha genus human papillomaviruses (HPVs) is the cause of almost all cervical cancers and is also linked to the development of other anogenital tract cancers as well as a subset of oral cancers. High-risk HPV associated tumorigenesis is dependent on expression of the E6 and E7 oncoproteins, which antagonize two key cellular tumor suppressors, p53 and pRB, respectively. The HPV E6 and E7 oncoproteins, however, also target additional signaling pathways and cellular processes implicated in tumor development (McLaughlin-Drubin et al., 2012). Of note, even though E6 and E7 expression is rate-limiting for cellular transformation, most high-risk HPV-associated lesions do not progress to cancer (Snijders et al., 2006) and/or spontaneously regress, presumably due to an antiviral immune response (Stanley, 2010). If progression does occur, it is often years or decades after the initial infection. Hence, in addition to persistent HPV infection, mutations in the host genome contribute to cervical cancer development and progression. The HPV E6 and E7 oncoproteins induce chromosomal instability, thereby increasing the mutation rate in infected host cells. Yet, remarkably little is known regarding

n Correspondence to: Division of Infectious Diseases, Brigham and Women's Hospital and Department of Medicine, Harvard Medical School, 181 Longwood Ave, MCP861, Boston, MA, 02115, USA. Fax: +1 617 525 4283. E-mail address: [email protected] (K. Munger). 1 Current address: Department of Biochemistry and Biophysics, University of California, San Francisco, CA, 94158, USA.

0042-6822/$ - see front matter & 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.virol.2013.07.009

specific host cell mutations that contribute to malignant progression of high-risk HPV associated lesions. An increased incidence of high-risk alpha HPV-associated lesions and cancers in Fanconi Anemia patients has been reported by some authors, but other studies have failed to confirm these reports (Kutler et al., 2003; van Zeeburg et al., 2008). Similarly, mutations in EVER1 and EVER2 genes, which predispose patients to develop squamous cell carcinomas in sun exposed areas of the body upon infection with beta HPVs, only play a minor role, if any, in the development of tumors caused by high-risk alpha HPVs (Castro et al., 2012; Orth, 2006; Wang et al., 2010). There is evidence that inactivating mutations in the liver kinase B1 (LKB1) tumor suppressor are associated with a dramatic decrease in progression free survival of patients with cervical cancer, independent of the histologic subtype (Wingo et al., 2009). LKB1 (also called Serine threonine kinase 11, STK11) is an evolutionarily conserved protein kinase that, in mammalian cells, phosphorylates and activates 14 members of the AMPK related kinase family (Jaleel et al., 2005; Lizcano et al., 2004), thereby regulating a variety of cellular processes including energy homeostasis and cell growth, proliferation and polarity (Alessi et al., 2006). Any of these LKB1 activities may contribute to its tumor suppressor activity. The best-studied LKB1-substrates are the catalytic subunits of the AMP-activated protein kinase complex, AMPKα1 and α2, thus linking LKB1 to energy homeostasis (Hawley et al., 2003; Shaw et al., 2004; Woods et al., 2003). However, AMPK has also been implicated in p53-dependent cell cycle arrest, cell migration and cell polarity (Alessi et al., 2006). In addition, the kinases MARK1–4 and, specifically in neuronal cells, BRSK1/2 play an

10

H.I.D. Mack, K. Munger / Virology 446 (2013) 9–16

important role in regulating cell polarity downstream of LKB1 (Alessi et al., 2006). Little is known regarding the other known LKB1substrates, NUAK1/2 (also called ARK5 and SNARK, respectively), SNRK and SIK1–3 (Jaleel et al., 2005; Lizcano et al., 2004). SIK1 may have tumor suppressor activity (Cheng et al., 2009), and oncogenic and/or tumor suppressive activities have been reported for NUAK2 and NUAK1 (Emmanuel et al., 2011; Hou et al., 2011; Humbert et al., 2010; Liu et al., 2012; Namiki et al., 2011; Zagorska et al., 2010). LKB1 itself was first suggested to be a tumor suppressor when gene mutations were detected in over 90% of Peutz-Jeghers syndrome (PJS) patients (Hemminki et al., 1998; Hemminki et al., 1997). PJS is a rare, autosomal-dominantly heritable disease characterized by mucocutaneous hyperpigmentation, development of mostly benign gastrointestinal polyps (hamartomas) and a substantially increased risk for benign and malignant tumors in multiple organs, including but not limited to breast, cervix, colon, lung, stomach, small bowel and testis (Beggs et al., 2010). How to reconcile the apparently benign nature of PJS polyps with the predisposition to cancer, particularly in the gastrointestinal tract, remains controversial (Beggs et al., 2010; Bosman, 1999; Jansen et al., 2006). Additional evidence for LKB1's tumor suppressor activity stems from cell culture and animal studies. For example, ectopic expression of wild type LKB1, but not of a kinasedefective LKB1 mutant suppressed proliferation/survival of G361 melanoma and HeLa S3 cervical cancer cells and inhibited anchorage independent growth of A549 lung cancer cells (Ji et al., 2007; Tiainen et al., 1999). Mice with heterozygous deletion of LKB1 develop gastrointestinal polyps that histologically resemble PJS polyps (Bardeesy et al., 2002; Jishage et al., 2002; Miyoshi et al., 2002).

Interestingly, LKB1-mutations are much less common in sporadic cancers, with the exception of non-small cell lung cancer and cervical carcinoma, where they are observed in  30% of all cases (Hezel and Bardeesy, 2008; Wingo et al., 2009). To start investigating the molecular basis of the apparent tumorsuppressive activity of LKB1 in cervical cancer cells, we ectopically expressed wild type LKB1 or kinase defective mutants in LKB1negative HeLa and SiHa cervical cancer cell lines as well as in the A549 lung cancer line and examined cell proliferation/viability, anchorage independent growth and energy stress sensitivity. Unexpectedly we found marked differences in how reconstitution with wild type LKB1 or with different kinase defective-mutants affected anchorage independent growth of these cell lines, whereas cell proliferation/ viability and energy–stress responses were similarly modulated by the different LKB1 constructs in all lines. Taken together, our results are consistent with the model that LKB1 is a bona fide tumor suppressor in cervical cancer, yet, our data also indicates that LKB1 tumor suppressor activity may be dependent on the cell type, which may explain why frequent mutation of LKB1 is restricted to certain tumor types. Results Inhibition of proliferation of cervical carcinoma lines by ectopic LKB1 expression Ectopic LKB1 expression was shown to inhibit proliferation of HeLa S3 cervical cancer cells and moreover, this was dependent on

Fig. 1. Effect of LKB1 reconstitution on cell growth of LKB1-negative cervical carcinoma lines. Equal numbers of HPV18 positive HeLa (A–C), HPV16 positive SiHa (D and E) cervical cancer, and A549 lung cancer cells (F) with ectopic expression of wild type LKB1 (WT), the kinase-defective LKB1-K78I mutant (K78I), or infected with parental vector (V) were seeded in triplicate into 24-well plates and cultured for the times indicated. Cell numbers were determined by crystal violet staining. The line graphs show cell numbers relative to the zero hour time point for each cell line. Error bars indicate standard deviations of triplicate measurements. Results shown are representative of at least two independent experiments for each cell population. A Western blot analysis comparing LKB1-WT and LKB-K78I expression in the set of cell lines used is shown on the bottom of each panel. Signal intensities were measured by densitometry and LKB1-signals were normalized to the corresponding actin signal. Numbers indicate the LKB/ Actin ratio relative to LKB-1WT-expressing cells. The sets of HeLa cells used in panels A–C and the sets of SiHa cells used in panels D and E were derived independently.

H.I.D. Mack, K. Munger / Virology 446 (2013) 9–16

its kinase activity (Tiainen et al., 1999). To initially confirm that inhibition of cervical carcinoma cell proliferation by LKB1 was dependent on kinase activity, we generated stable populations of LKB1 deficient HeLa and SiHa cervical carcinoma (Wingo et al., 2009; Yu et al., 2009) cell lines with ectopic expression of wild type LKB1 (LKB1-WT) or the kinase defective LKB1-K78I mutant. Both, HeLa and SiHa cell lines were shown to lack LKB1 expression (Nafz et al., 2007; Wingo et al., 2009; Yu et al., 2009). LKB1negative A549 lung cancer cells were included as a control because they had been previously used to demonstrate that LKB1-WT exerts tumor suppressive activities (Ji et al., 2007). Three independent sets of cell populations were generated for HeLa and two for SiHa cells. Vector expressing cell populations were used as controls and growth curves were recorded (Fig. 1). Proliferation of HeLa, SiHa and A549 cells expressing LKB1-WT was inhibited as compared to vector-expressing control cells. This was observed with each of the independently generated sets of cell populations described above. The results with LKB1-K78I expressing HeLa cells were inconclusive; in one set, the growth rate of LKB1-K78I expressing cells was lower than LKB1-WT expressing cells but higher than vector control cells (Fig. 1A), whereas in the other two independently derived sets of HeLa cell populations, the kinase defective LKB1-K78I mutant suppressed growth almost as effectively as LKB1-WT (Fig. 1B and C). We observed either reduced growth suppression or no growth suppression with the two sets of LKB1-K78I expressing SiHa cell lines (Fig. 1D and E). With A549 cells, we observed growth suppression by ectopic expression of LKB1-WT or the K78I mutant (Fig. 1F).

11

when cellular energy levels drop as a consequence of glucose deprivation (Alessi et al., 2006). To mimic glucose deprivation, we treated HeLa, SiHa and A549 cells with ectopic expression of LKB1WT or LKB1-K78I or empty vector expressing control cells with 2deoxyglucose (2DG) (Ralser et al., 2008; Simons et al., 2009) and examined cell numbers by crystal violet staining and monitored AMPKα-phosphorylation. LKB1-WT expressing HeLa cells were protected from glucose-starvation induced cell death at the 72 and 96 h time points whereas expression of the kinase defective LKB1-K78I mutant provided no survival advantage compared to vector control cells. Similar results were obtained with A549 lung cancer cells although the effects were somewhat less pronounced (Kato et al., 2002; Shaw et al., 2004) (Fig. 2A and C). Consistent with these results, 2DG-treatment caused much stronger AMPK activation in LKB1-WT expressing HeLa and A549 cells than in the corresponding LKB1-K78I expressing cell populations (Fig. 3). Yet, we noted a slight increase in AMPK-phosphorylation in LKB-K78Iexpressing and even in vector control, LKB1 negative cells triggered by 2 DG-treatment. This latter observation has also been made by other authors and suggests that there may be an additional LKB1-independent pathway that triggers AMPKactivation in response to glucose deprivation (Jeon et al., 2012). In contrast, expression of LKB1-WT in SiHa cells had no effect on survival upon glucose starvation (Fig. 2B), even though 2DGtreatment clearly activated AMPK more efficiently in LKB1-WT expressing cells than in vector control or in LKB1-K78I-expressing SiHa cells (Fig. 3).

The LKB1-K78I mutant does not enhance cell survival under conditions of glucose starvation

Ectopic expression of wild type or a kinase-defective LKB1 mutant inhibits anoikis resistance of HeLa but not of SiHa cervical carcinoma cells

Given the unexpected findings with the kinase defective LKB1K78I mutant in our growth inhibition experiments, we subjected the various cell lines to additional assays for LKB1 tumor suppressor activity. The best-studied LKB1-substrate is AMPK, and LKB1dependent phosphorylation of the AMPK activation loop increases

Reconstitution of LKB1-negative A549 lung cancer cells with LKB1-WT has been reported to inhibit anoikis resistance, as measured by anchorage independent growth, and this was shown to be dependent on LKB1 kinase activity (Ji et al., 2007). Hence we determined anoikis resistance of our LKB1-WT and LKB-1K78I

Fig. 2. Effect of LKB1-reconstitution on glucose deprivation of LKB1-negative cervical carcinoma lines. Equal numbers of HeLa (A), SiHa (B) cervical cancer, and A549 lung cancer (C) cells with ectopic expression of wild type (WT) LKB1, the kinase-defective LKB1-K78I mutant (K78I), or infected with parental vector (V) were seeded in triplicate into 24-well plates and cultured in DMEM+10% FBS containing 25 mM 2-Deoxyglucose (2DG) for the times indicated and cell numbers were determined by crystal violet staining. The line graphs show cell numbers relative to the zero hour time point for each cell line. Results are averages and standard deviations of three independent experiments, each done in triplicate.

Fig. 3. AMPKα phosphorylation in response to glucose starvation. HeLa and SiHa cervical carcinoma and A549 lung cancer cells reconstituted with LKB1 wild type (WT) or the K78I kinase defective LKB1 mutant (K78I) or vector infected control cells (V) were treated with 25 mM 2-Deoxyglucose (2DG) for the times indicated and AMPKα phosphorylation at Thr172 (pAMPKα) was monitored by Western Blot analysis. Total AMPKα levels as well as LKB1 expression is also shown, a β-actin blot is shown as a loading control.

12

H.I.D. Mack, K. Munger / Virology 446 (2013) 9–16

expressing cell populations by growing them suspended in agar containing medium. Similar to what was previously reported for A549 lung cancer cells (Ji et al., 2007), LKB1-WT expression inhibited anoikis of HeLa cells. Unexpectedly, however, expression of the kinase-defective LKB1-K78I mutant in HeLa cells inhibited anchorage independent growth to a similar extent as LKB1-WT. In contrast, ectopic expression of LKB1-WT or LKB1-K78I in SiHa cells did not markedly inhibit anoikis resistance. Consistent results were obtained with each of the independently derived sets of cell populations. Moreover, and in contrast to previously published results (Ji et al., 2007), LKB1-K78I expression also clearly reduced anoikis resistance of A549 lung cancer cells (Fig. 4).

SiHa cells do not tolerate long-term ectopic LKB1-WT expression The apparent inability of LKB1-WT to modulate anoikis resistance and survival in response to nutrient deprivation of SiHa cells despite its ability to inhibit proliferation was puzzling. We noted, however, that the growth suppressive effect of LKB1-WT overexpression was no longer observed in higher passage SiHa populations (data not shown). Hence we monitored LKB1 expression in SiHa cells over time. Initially, after retroviral infection and selection of cells, LKB1-WT and the kinase defective mutant LKB1-K78I were expressed at similar levels (see Fig. 1D and E). However, we consistently noted pronounced differences between expression levels of LKB1-WT and kinase-defective mutants (LKB1-K78I and LKB1-D194A, see below) at higher passage numbers. This was specific for SiHa cells and even though LKB1-WT expression also caused growth inhibition in A549 and HeLa cells, LKB1-WT expression did not markedly decrease over time, or it decreased

similarly as in cells expressing the kinase defective mutants. Thus, the inability of LKB1 to inhibit anchorage independent growth of SiHa cells may be a reflection of the loss of expression over time (Fig. 5). Growth suppression by LKB1 is primarily dependent on kinase activity and not expression level The finding that LKB1-WT and the kinase-defective LKB1-K78I mutant inhibit proliferation and anchorage-independent growth to a similar extent in HeLa cells may be explained by several models: (a) the LKB1-K78I mutant has residual kinase activity, which is sufficient to inhibit cell growth and anchorage-independent growth, yet is insufficient to delay starvation-induced cell death of HeLa cells or (b) inhibition of proliferation and anchorage independent growth of HeLa cells are mostly independent of LKB1 kinase activity. To distinguish between these two models, we depleted LKB1 in the LKB1-positive (Nafz et al., 2007; Wingo et al., 2009; Yu et al., 2009) CaSki (HPV16 positive) and C33A (HPV negative) cervical cancer cell lines, and examined the effect on cell growth. We reasoned that if LKB1-mediated inhibition of cell proliferation was largely independent of kinase activity, then LKB1 depletion should enhance proliferation of LKB1-positive CaSki and C33A cells. Conversely, if low-level LKB1 kinase activity (mimicked by low level residual LKB1-expression after depletion) was sufficient for inhibition of proliferation, then LKB1 depletion in these cell lines may not markedly accelerate growth. Indeed, CaSki and C33A cells transduced with different lentiviral shRNAs that diminished LKB1-levels by up to 80% still grew at similar rates than vector control cells. Furthermore, there was no correlation between residual LKB1-expression and resistance to EBSS starvation (Fig. 6). Thus, our results are consistent with a model where, at least in cervical carcinoma lines, low-level LKB1-kinase activity may be sufficient to maintain the growth suppressive activity of the LKB1 tumor suppressor. The kinase-defective LKB1-D194A mutant does not inhibit anoikis resistance of HeLa cells

Fig. 4. Effect of LKB reconstitution on anoikis resistance of LKB1-negative cervical cancer cell lines. HeLa and SiHa cervical cancer and A549 lung cancer cells with ectopic expression of wild type (WT) LKB1, the kinase-defective LKB1-K78I mutant (K78I), or infected with parental vector (V) were seeded in soft agar and colony formation was assessed 3 weeks later. Colonies were imaged and colony size was measured using the ellipse tool in Image J. Each data point represents the size of an individual colony. Horizontal lines and error bars represent the mean colony area with 95% confidence interval. Statistical analysis was performed using unpaired t-tests. nn and nnn indicate nominal P-values o0.01 and o 0.001, respectively. Similar results were obtained in two additional independent experiments, although the differences between the various SiHa-cell lines were not consistently statistically different as in the experiment shown here. The sets of cell lines used in the experiment shown are HeLa (II), SiHa (II) and the single A549-set as defined in Fig. 1. See Fig. 1B, E and F for Western Blot analysis of LKB1-WT and LKB-K78I expression levels in the different cell lines.

Given our model that low level LKB1 kinase activity may be sufficient for tumor suppressor activity in cervical carcinoma lines, and that the kinase defective LKB1-K78I mutant inhibits anoikis resistance in HeLa cells as efficiently as LKB1-WT, raises the possibility that the LKB1-K78I mutant may retain some kinase activity. Thus, we investigated the effect of expression of a different kinase defective mutant, LKB1-D194A on anoikis resistance of A549 lung cancer and the HeLa and SiHa cervical carcinoma line. As seen previously for the K78I mutant, SiHa cells could be reconstituted with the D194A mutant, but in LKB1-WT reconstituted cells generated in parallel, expression was rapidly silenced (see Fig. 1D and E and 4) and both LKB1-reconstituted cell lines behaved similar to vector control-expressing SiHa cells in our anoikis resistance experiments (data not shown).

Fig. 5. SiHa cervical cancer cells do not tolerate high levels of LKB1-WT expression. LKB1 levels were determined by western blotting in SiHa, HeLa and A549 cells with ectopic expression of wild type LKB1 (WT) or the kinase defective LKB1-D194A mutant (D194A) or infected with the parental vector and expression at passage 2 and passage 8 after derivation of stable populations. Signal intensities were measured by densitometry and LKB1-signals were normalized to the corresponding actin signal. Numbers indicate the LKB/Actin ratio relative to passage 2 cells for each cell line.

H.I.D. Mack, K. Munger / Virology 446 (2013) 9–16

13

Fig. 6. Effect of LKB1-depletion on growth and starvation resistance of LKB1-positive cervical cancer cell lines. HPV16 positive CaSki (A) and HPV negative C33A (B) cervical carcinoma cells were infected with a panel of LKB1 specific lentiviral shRNAs expression vectors. After selection, residual LKB1 expression was analyzed by the Western blotting. β-actin and GAPDH served as loading controls. Signal intensities were measured by densitometry and LKB1-signals were normalized to the corresponding signal of the loading control. Numbers indicate the LKB/loading control ratio relative to cells infected with a non-targeting control shRNA (left panels). Equal numbers of cells with stable expression of the indicated LKB1 specific shRNAs or a control shRNA were seeded in triplicate into 24-well plates and cultured in normal growth medium DMEM+10% FBS (middle panels) or starved in EBSS (right panels) for the times indicated and cell numbers were determined by crystal violet staining. The line graphs show cell numbers relative to the zero hour time point for each cell line. Error bars indicate standard deviations of triplicate measurements. Results shown are representative of 3 independent experiments each.

was defective in suppressing anoikis resistance in this cell line (Fig. 7). These results suggest that the kinase-defective LKB1-K78I and LKB1-D194A mutants may be functionally distinct, and that these functional differences only become detectable in particular cellular backgrounds such as HeLa cells.

Discussion

Fig. 7. The kinase-defective LKB1-D194A mutant fails to inhibit anoikis resistance of HeLa cells. (A) Western blot analysis to determine expression levels of LKB1-WT and LKB1-D194A in the cell lines indicated. Signal intensities were measured by densitometry and LKB1-signals were normalized to the corresponding tubulin signal. Numbers indicate the LKB/tubulin ratio relative to LKB1-WT-expressing cells. (B) HeLa cervical cancer and A549 lung cancer cells with ectopic expression of LKB1-WT (WT), the kinase defective LKB1-D194A mutant (D194A) or infected with the empty vector were seeded in soft agar and colony formation was assessed three weeks later. Colonies were imaged and colony size was measured using the ellipse tool in Image J. Each data point represents the size of an individual colony. Horizontal lines and error bars represent the mean colony area with a 95% confidence interval. Statistical analysis was performed using unpaired t-tests nn and nnn indicate nominal P-values o 0.01 and o 0.001, respectively. Similar results were obtained in two additional independent experiments.

In contrast to LKB1-K78I, which in our hands partially inhibited anoikis resistance in A549 cells, the LKB1-D194A mutant did not inhibit anoikis resistance of these cells. Moreover, the LKB1-D194A mutant differed from the LKB1-K78I mutant in HeLa cells in that it did not inhibit anoikis resistance similar to LKB1-WT but instead

In this study, we aimed to elucidate the molecular basis for the published observation that loss of the LKB1-tumor suppressor, which is detected in approximately 30% of cervical cancers, dramatically decreases progression free survival of patients (Wingo et al., 2009). Increased tumor aggressiveness as a consequence to LKB1 loss may be due to anoikis resistance as LKB1's tumor-suppressor activities have been linked to its ability to inhibit anchorage independent growth of A549 lung cancer cells (Ji et al., 2007). Although our results with A549 cells are consistent with the Ji et al. (2007) study and indicate a clear dependence on LKB1 kinase activity, our results with the SiHa and HeLa cervical carcinoma cell lines were quite different. In particular, out of the two different kinase-defective LKB1-mutants that we examined, and which did not efficiently suppress A549 cell growth in soft agar, only one, D194A, was defective in suppressing HeLa cell growth in soft agar. In contrast, the LKB1-K78I mutant inhibited anchorage independent growth of HeLa cells to a similar extent as the wild type protein, while it has at least a partial suppressive effect in A549 cells. In contrast, SiHa cells appeared not to tolerate ectopic LKB1 expression, confounding our efforts to study a potential tumor suppressor activity of LKB1 in this cell line. We note that HeLa cells are HPV18 positive whereas SiHa cells contain HPV16, suggesting the interesting possibility that HPV genotype may contribute to the marked differences we observed between these two cervical carcinoma cell lines. These results further support the notion that the LKB1 tumorsuppressive activity may depend on the cellular context (Contreras et al., 2008). Indeed, the increase in tumor incidence in PJS patients compared to the general population varies widely across different organs. Specifically, risk is elevated 1.8- and 4.5-fold for cervical and testicular cancers in female and male PJS patients,

14

H.I.D. Mack, K. Munger / Virology 446 (2013) 9–16

respectively, but is increased 520-fold for small bowel cancers in patients of both genders (Beggs et al., 2010; Giardiello et al., 1987). The LKB1-K78I and LKB1-D194A mutants have both been described as “kinase inactive”. K78 is the lysine residue involved in binding and correctly orienting the phosphate donor, ATP, and D194 is involved in binding the cofactor Mg2+. Unexpectedly, our experiments revealed clear functional differences between these two mutations. One possibility is that the LKB1-K78I, but not the D194A mutant, may retain some low level of kinase activity that is sufficient to exert anoikis resistance in HeLa but not in A549 cells. At least 14 LKB1 phosphorylation substrates with partially overlapping activities have been described (Alessi et al., 2006). It is conceivable that anoikis resistance is regulated via different LKB-1 downstream substrates in different cell types and it will be interesting to determine which substrates are particularly important for LKB1 tumor suppression in different cell types and whether the LKB1-K78I and D194A mutants differ in their residual activity to phosphorylate certain substrates. Last but not least, our results also are consistent with the hypothesis that some of the LKB1 tumor suppressor activities may be independent of its kinase activity. Experiments to distinguish between these two possibilities will be the focus of future studies. Interestingly, the loss of LKB1 kinase activity due to the K78I mutation, while it did not affect anoikis resistance in HeLa cells, was sufficient to cause some sensitivity to cell death upon glucose starvation. This is interesting since cancer cells generally acquire resistance to anoikis as well as to cell death in response to nutrient deprivation. Unlike what would be expected from a prototypical tumor suppressor, loss of LKB1 activity can have both tumor promoting (anoikis resistance) and tumor suppressive (increased sensitivity to cell death caused by glucose starvation) effects, depending on the stage of tumorigenesis when LKB1 is lost. A similar stage-dependent tumor-modulatory role has been proposed recently for the LKB1-substrate AMPK (Jeon et al., 2012). Greater than 99% of cervical cancers are causally linked to infection with high-risk HPVs. This raises the interesting question whether in the subset of cervical cancers that are negative for LKB1-expression ( 20% of cases), HPV infection or LKB1-loss occurs earlier in the course of tumorigenesis. The fact that cervical cancer risk is elevated in female PJS patients at the first glance suggests that LKB1 loss may be the first hit. However, one has to be extremely cautious before drawing this conclusion. First, there are no studies reporting HPV-status of PJS associated cervical cancers (Beggs et al., 2010; Giardiello et al., 1987). Second, a substantial fraction of PJS-associated cervical cancers, approximately 78% (Giardiello and Trimbath, 2006), has been classified as “adenoma malignum” (also called “minimal deviation adenocarcinoma”), a subtype that is very rare in the general population, and which frequently appears to be HPV-negative (Ferguson et al., 1998; Mikami et al., 2004). Thus, it is possible that HPV infection does not play a major role in those cervical cancers that occur at a higher frequency in PJS patients. On the other hand, studies with LKB-/- MEFs support the hypothesis that HPV oncoprotein expression after HPV-infection is the initiating event in PJS-associated cervical cancers. LKB-/- MEFs were reported to be resistant to transformation by RAS, SV40 large T antigen and adenovirus E1A (Bardeesy et al., 2002), and, given the mechanistic similarities between SV40 large T antigen, adenovirus E1A and HPV E6/E7 driven cell transformation, LKB-/- MEFs may also be resistant to HPV E6/E7 transformation. Hence high-risk HPV infections would be predicted to have a lower risk for cancer formation in PJS patients than in the general population. Clearly, additional studies are needed to determine the sequence of oncogenic hits and mechanism of cervical carcinogenesis in PJS patients. In summary, our studies support the model that the tumorsuppressive activity of LKB1 is context dependent and may be

modulated by host cell and/or viral factors. HPV E6 and E7 are prime candidates for such factors in cervical cancer cells. Since HeLa cells but not SiHa cells tolerate ectopic LKB1 expression it is possible that modulation of LKB1-function may be different for HPV18 and HPV16 E6/E7 proteins. Although it remains to be determined whether and how HPV E6/E7 and LKB1 functionally interact, our study clearly provides further evidence that LKB1 is a bona fide tumor suppressor in cervical cancer and suggests that LKB1-dependent inhibition of anoikis resistance represents an important underlying molecular mechanism.

Material and methods Cell culture and treatments C33A, CaSki, HeLa and SiHa cervical cancer cell lines were obtained from ATCC. A549 lung cancer cells were a gift from Kwok-Kin Wong. One independently derived set of LKB1-WT, -K78I and vector-control expressing HeLa cells was a gift from Dr. Bin Zheng (Columbia University, New York, NY); (Zheng et al., 2009). All cell lines were maintained in DMEM supplemented with 10% fetal bovine serum, 50 U/ml penicillin and 50 mg/ml streptomycin at 37 1C, 5% CO2. For all treatments, cells were changed to fresh growth media two hours before treatment started. To mimic glucose starvation, 25 mM 2-deoxyglucose (Sigma) was added directly to the growth media. For complete starvation treatments, cells were washed twice with phosphate-buffered saline (PBS) and then incubated in Earl's balanced salt solution (EBSS, Invitrogen, E2888). Plasmids and viral infections The pBabe-FLAG-LKB1-WT and pBabe-FLAG-LKB1- K78I expression vectors were obtained from Addgene (plasmids #8592 and 8593; (Shaw et al., 2004)). pBabe-FLAG-LKB1-D194A was a generous gift from Dr. Bin Zheng (Columbia University, New York, NY; (Zheng and Cantley, 2007)). Lentiviral pLKO-based shRNA expression vectors targeting LKB1 (TRCN0000000407, TRCN0000000408, TRCN0000000409, TRCN0000000410, TRCN0000000411) were obtained from Thermo Fisher Scientific and the pLKO-based MISSION Non-Target shRNA vector (SHC002, Sigma) was used as a control. For production of recombinant viruses, HEK293T cells were seeded at 60% confluence and cotransfected with the following plasmids using polyethylenimine (PEI; Polysciences Inc., 23966-2) at a 3:1 ratio to DNA: for shRNAcontaining lentiviruses, cells were cotransfected with shRNA-construct and packaging plasmids psPax2 and pMD2.G. For retroviruses, cells were cotransfected with the retroviral plasmid of interest and packaging plasmids pMD.MLV and pMD.G. 16 h later, the media was changed and virus-containing supernatant was collected 24, 36 and 48 h post transfection and centrifuged to eliminate cells. For infections, cells were seeded at 30% confluence and incubated in viral supernatant containing 8 mg/ml polybrene for 16 h. Infected cells were selected with 2.5 mg/ml puromycin. Antibodies and western blotting For Western blot analysis, cells were lysed in 1% SDS, 60 mM Tris, pH 7.4. Lysates were boiled at 95 1C for 10 min and cleared by centrifugation for 10 min at room temperature. Protein concentration was determined using a detergent compatible (Dc) protein assay (Bio-Rad). Fifty to hundred mg of total protein were separated by SDS-PAGE and transferred to PVDF membranes (ImmunBlot PVDF, Bio-Rad). Membranes were blocked in 5% nonfat dry milk in TBST, followed by incubation in primary antibody for 16 h at 4 1C. After washing, membranes were incubated in the appropriate

H.I.D. Mack, K. Munger / Virology 446 (2013) 9–16

HRP-coupled secondary antibody (anti-mouse IgG, NA931V, GE Healthcare or anti-rabbit IgG, NA934V, GE Healthcare), for one hour at room temperature. Signals were detected using enhanced chemiluminescence (Millipore, WBKLS0500) on X-ray films (Denville, Kodak) or on a molecular imaging system (Carestream, GL 4000 Pro). The following primary antibodies were used at the indicated dilutions: AMPK (Cell Signaling #2603, 1:1000), pAMPK (Cell Signaling #2535, 1:1000), β-Actin (Sigma #A1978, 1:5000), GAPDH (Millipore MAB374, 1:1000), LKB1 (Abcam Ab15095, 1: 1000), α-tubulin (Sigma #T6199, 1:2000). Cell growth analysis and soft agar assays To measure cell growth, 10,000–15,000 cells/well (for measurements of cell death upon 2DG- or EBSS-treatment: 25,000–30,000 cells/well) depending on cell line and experiment, were seeded in triplicate for each time point in 24-well plates. Cells were fixed with 3.7% paraformaldehyde in PBS for 10 min at room temperature and stored in PBS at 4 1C until fixation of the last time point. Then, cells were stained with 0.1% crystal violet in 10% ethanol for one hour at room temperature, washed 5 times with H2O and dried. Crystal violet was extracted with 0.5 ml/well 10% acetic acid and absorption at 590 nm was read in a plate reader (Perkin Elmer Victor or Bio-Tek PowerWave HT). For soft agar growth assays, 6 well plates were coated with 2 ml 0.6% noble agar (Difco) in growth media. 24 h later, cells were trypsinized and 20,000 cells/well were resuspended in 2 ml 0.3% noble agar and plated on top of the 0.6% agar layer (2 wells/ cell line). Colonies were imaged after 3 weeks and colony size was measured using the ellipse tool in Image J.

Acknowledgments We thank Drs. Kwok-Kin Wong (Dana-Farber Cancer Institute, Boston, MA) and Bin Zheng (Columbia University, New York, NY) for providing cell lines and reagents, and all members of the Munger Lab for helpful discussion. This work was supported by PHS grant R01 CA081135 from the National Cancer Institute (K.M.). H.I.D.M. was the recipient of graduate fellowships from the Bavarian State Ministry of Sciences, Research and the Arts, the German Academic Exchange Service and the German National Academic Foundation. References Alessi, D.R., Sakamoto, K., Bayascas, J.R., 2006. LKB1-dependent signaling pathways. Annu. Rev. Biochem. 75, 137–163. Bardeesy, N., Sinha, M., Hezel, A.F., Signoretti, S., Hathaway, N.A., Sharpless, N.E., Loda, M., Carrasco, D.R., DePinho, R.A., 2002. Loss of the Lkb1 tumour suppressor provokes intestinal polyposis but resistance to transformation. Nature 419 (6903), 162–167. Beggs, A.D., Latchford, A.R., Vasen, H.F., Moslein, G., Alonso, A., Aretz, S., Bertario, L., Blanco, I., Bulow, S., Burn, J., Capella, G., Colas, C., Friedl, W., Moller, P., Hes, F.J., Jarvinen, H., Mecklin, J.P., Nagengast, F.M., Parc, Y., Phillips, R.K., Hyer, W., Ponz de Leon, M., Renkonen-Sinisalo, L., Sampson, J.R., Stormorken, A., Tejpar, S., Thomas, H.J., Wijnen, J.T., Clark, S.K., Hodgson, S.V., 2010. Peutz-Jeghers syndrome: a systematic review and recommendations for management. Gut 59 (7), 975–986. Bosman, F.T., 1999. The hamartoma-adenoma-carcinoma sequence. J. Pathol. 188 (1), 1–2. Castro, F.A., Ivansson, E.L., Schmitt, M., Juko-Pecirep, I., Kjellberg, L., Hildesheim, A., Gyllensten, U.B., Pawlita, M., 2012. Contribution of TMC6 and TMC8 (EVER1 and EVER2) variants to cervical cancer susceptibility. Int. J. Cancer 130 (2), 349–355. Cheng, H., Liu, P., Wang, Z.C., Zou, L., Santiago, S., Garbitt, V., Gjoerup, O.V., Iglehart, J.D., Miron, A., Richardson, A.L., Hahn, W.C., Zhao, J.J., 2009. SIK1 couples LKB1 to p53dependent anoikis and suppresses metastasis. Sci. Signal 2 (80), ra35. Contreras, C.M., Gurumurthy, S., Haynie, J.M., Shirley, L.J., Akbay, E.A., Wingo, S.N., Schorge, J.O., Broaddus, R.R., Wong, K.K., Bardeesy, N., Castrillon, D.H., 2008. Loss of Lkb1 provokes highly invasive endometrial adenocarcinomas. Cancer Res. 68 (3), 759–766. Emmanuel, C., Gava, N., Kennedy, C., Balleine, R.L., Sharma, R., Wain, G., Brand, A., Hogg, R., Etemadmoghadam, D., George, J., Birrer, M.J., Clarke, C.L., Chenevix-

15

Trench, G., Bowtell, D.D., Harnett, P.R., deFazio, A. Comparison of expression profiles in ovarian epithelium in vivo and ovarian cancer identifies novel candidate genes involved in disease pathogenesis. PLoS One 6(3), 2011, e17617. Ferguson, A.W., Svoboda-Newman, S.M., Frank, T.S., 1998. Analysis of human papillomavirus infection and molecular alterations in adenocarcinoma of the cervix. Mod. Pathol. 11 (1), 11–18. Giardiello, F.M., Trimbath, J.D., 2006. Peutz-Jeghers syndrome and management recommendations. Clin. Gastroenterol. Hepatol. 4 (4), 408–415. Giardiello, F.M., Welsh, S.B., Hamilton, S.R., Offerhaus, G.J., Gittelsohn, A.M., Booker, S.V., Krush, A.J., Yardley, J.H., Luk, G.D., 1987. Increased risk of cancer in the Peutz-Jeghers syndrome. N. Engl. J. Med. 316 (24), 1511–1514. Hawley, S.A., Boudeau, J., Reid, J.L., Mustard, K.J., Udd, L., Makela, T.P., Alessi, D.R., Hardie, D.G., 2003. Complexes between the LKB1 tumor suppressor, STRAD alpha/beta and MO25 alpha/beta are upstream kinases in the AMP-activated protein kinase cascade. J. Biol. 2 (4), 28. Hemminki, A., Markie, D., Tomlinson, I., Avizienyte, E., Roth, S., Loukola, A., Bignell, G., Warren, W., Aminoff, M., Hoglund, P., Jarvinen, H., Kristo, P., Pelin, K., Ridanpaa, M., Salovaara, R., Toro, T., Bodmer, W., Olschwang, S., Olsen, A.S., Stratton, M.R., de la Chapelle, A., Aaltonen, L.A., 1998. A serine/threonine kinase gene defective in Peutz-Jeghers syndrome. Nature 391 (6663), 184–187. Hemminki, A., Tomlinson, I., Markie, D., Jarvinen, H., Sistonen, P., Bjorkqvist, A.M., Knuutila, S., Salovaara, R., Bodmer, W., Shibata, D., de la Chapelle, A., Aaltonen, L.A., 1997. Localization of a susceptibility locus for Peutz-Jeghers syndrome to 19p using comparative genomic hybridization and targeted linkage analysis. Nat. Genet. 15 (1), 87–90. Hezel, A.F., Bardeesy, N., 2008. LKB1; linking cell structure and tumor suppression. Oncogene 27 (55), 6908–6919. Hou, X., Liu, J.E., Liu, W., Liu, C.Y., Liu, Z.Y., Sun, Z.Y., 2011. A new role of NUAK1: directly phosphorylating p53 and regulating cell proliferation. Oncogene 30 (26), 2933–2942. Humbert, N., Navaratnam, N., Augert, A., Da Costa, M., Martien, S., Wang, J., Martinez, D., Abbadie, C., Carling, D., de Launoit, Y., Gil, J., Bernard, D., 2010. Regulation of ploidy and senescence by the AMPK-related kinase NUAK1. EMBO J. 29 (2), 376–386. Jaleel, M., McBride, A., Lizcano, J.M., Deak, M., Toth, R., Morrice, N.A., Alessi, D.R., 2005. Identification of the sucrose non-fermenting related kinase SNRK, as a novel LKB1 substrate. FEBS Lett. 579 (6), 1417–1423. Jansen, M., de Leng, W.W., Baas, A.F., Myoshi, H., Mathus-Vliegen, L., Taketo, M.M., Clevers, H., Giardiello, F.M., Offerhaus, G.J., 2006. Mucosal prolapse in the pathogenesis of Peutz-Jeghers polyposis. Gut 55 (1), 1–5. Jeon, S.M., Chandel, N.S., Hay, N., 2012. AMPK regulates NADPH homeostasis to promote tumour cell survival during energy stress. Nature 485 (7400), 661–665. Ji, H., Ramsey, M.R., Hayes, D.N., Fan, C., McNamara, K., Kozlowski, P., Torrice, C., Wu, M.C., Shimamura, T., Perera, S.A., Liang, M.C., Cai, D., Naumov, G.N., Bao, L., Contreras, C.M., Li, D., Chen, L., Krishnamurthy, J., Koivunen, J., Chirieac, L.R., Padera, R.F., Bronson, R.T., Lindeman, N.I., Christiani, D.C., Lin, X., Shapiro, G.I., Janne, P.A., Johnson, B.E., Meyerson, M., Kwiatkowski, D.J., Castrillon, D.H., Bardeesy, N., Sharpless, N.E., Wong, K.K., 2007. LKB1 modulates lung cancer differentiation and metastasis. Nature 448 (7155), 807–810. Jishage, K., Nezu, J., Kawase, Y., Iwata, T., Watanabe, M., Miyoshi, A., Ose, A., Habu, K., Kake, T., Kamada, N., Ueda, O., Kinoshita, M., Jenne, D.E., Shimane, M., Suzuki, H., 2002. Role of Lkb1, the causative gene of Peutz-Jegher's syndrome, in embryogenesis and polyposis. Proc. Natl. Acad. Sci. USA 99 (13), 8903–8908. Kato, K., Ogura, T., Kishimoto, A., Minegishi, Y., Nakajima, N., Miyazaki, M., Esumi, H., 2002. Critical roles of AMP-activated protein kinase in constitutive tolerance of cancer cells to nutrient deprivation and tumor formation. Oncogene 21 (39), 6082–6090. Kutler, D.I., Wreesmann, V.B., Goberdhan, A., Ben-Porat, L., Satagopan, J., Ngai, I., Huvos, A.G., Giampietro, P., Levran, O., Pujara, K., Diotti, R., Carlson, D., Huryn, L. A., Auerbach, A.D., Singh, B., 2003. Human papillomavirus DNA and p53 polymorphisms in squamous cell carcinomas from fanconi anemia patients. J. Natl. Cancer Inst. 95 (22), 1718–1721. Liu, L., Ulbrich, J., Muller, J., Wustefeld, T., Aeberhard, L., Kress, T.R., Muthalagu, N., Rycak, L., Rudalska, R., Moll, R., Kempa, S., Zender, L., Eilers, M., Murphy, D.J., 2012. Deregulated MYC expression induces dependence upon AMPK-related kinase 5. Nature 483 (7391), 608–612. Lizcano, J.M., Goransson, O., Toth, R., Deak, M., Morrice, N.A., Boudeau, J., Hawley, S. A., Udd, L., Makela, T.P., Hardie, D.G., Alessi, D.R., 2004. LKB1 is a master kinase that activates 13 kinases of the AMPK subfamily, including MARK/PAR-1. EMBO J. 23 (4), 833–843. McLaughlin-Drubin, M.E., Meyers, J., Munger, K., 2012. Cancer associated human papillomaviruses. Curr. Opin. Virol. 2 (4), 459–466. Mikami, Y., Kiyokawa, T., Hata, S., Fujiwara, K., Moriya, T., Sasano, H., Manabe, T., Akahira, J., Ito, K., Tase, T., Yaegashi, N., Sato, I., Tateno, H., Naganuma, H., 2004. Gastrointestinal immunophenotype in adenocarcinomas of the uterine cervix and related glandular lesions: a possible link between lobular endocervical glandular hyperplasia/pyloric gland metaplasia and adenoma malignum. Mod. Pathol. 17 (8), 962–972. Miyoshi, H., Nakau, M., Ishikawa, T.O., Seldin, M.F., Oshima, M., Taketo, M.M., 2002. Gastrointestinal hamartomatous polyposis in Lkb1 heterozygous knockout mice. Cancer Res. 62 (8), 2261–2266. Nafz, J., De-Castro Arce, J., Fleig, V., Patzelt, A., Mazurek, S., Rosl, F., 2007. Interference with energy metabolism by 5-aminoimidazole-4-carboxamide-1-beta-D-ribofuranoside induces HPV suppression in cervical carcinoma cells and apoptosis in the absence of LKB1. Biochem. J. 403 (3), 501–510.

16

H.I.D. Mack, K. Munger / Virology 446 (2013) 9–16

Namiki, T., Tanemura, A., Valencia, J.C., Coelho, S.G., Passeron, T., Kawaguchi, M., Vieira, W.D., Ishikawa, M., Nishijima, W., Izumo, T., Kaneko, Y., Katayama, I., Yamaguchi, Y., Yin, L., Polley, E.C., Liu, H., Kawakami, Y., Eishi, Y., Takahashi, E., Yokozeki, H., Hearing, V.J., 2011. AMP kinase-related kinase NUAK2 affects tumor growth, migration, and clinical outcome of human melanoma. Proc. Natl. Acad. Sci. USA 108 (16), 6597–6602. Orth, G., 2006. Genetics of epidermodysplasia verruciformis: insights into host defense against papillomaviruses. Semin. Immunol. 18 (6), 362–374. Ralser, M., Wamelink, M.M., Struys, E.A., Joppich, C., Krobitsch, S., Jakobs, C., Lehrach, H., 2008. A catabolic block does not sufficiently explain how 2deoxy-D-glucose inhibits cell growth. Proc. Natl. Acad. Sci. USA 105 (46), 17807–17811. Shaw, R.J., Kosmatka, M., Bardeesy, N., Hurley, R.L., Witters, L.A., DePinho, R.A., Cantley, L.C., 2004. The tumor suppressor LKB1 kinase directly activates AMPactivated kinase and regulates apoptosis in response to energy stress. Proc. Natl. Acad. Sci. USA 101 (10), 3329–3335. Simons, A.L., Mattson, D.M., Dornfeld, K., Spitz, D.R., 2009. Glucose deprivationinduced metabolic oxidative stress and cancer therapy. J. Cancer Res. Ther. 5 (Suppl. 1), S2–S6. Snijders, P.J., Steenbergen, R.D., Heideman, D.A., Meijer, C.J., 2006. HPV-mediated cervical carcinogenesis: concepts and clinical implications. J. Pathol. 208 (2), 152–164. Stanley, M., 2010. HPV—immune response to infection and vaccination. Infect Agent Cancer 5, 19. Tiainen, M., Ylikorkala, A., Makela, T.P., 1999. Growth suppression by Lkb1 is mediated by a G(1) cell cycle arrest. Proc. Natl. Acad. Sci. USA 96 (16), 9248–9251. van Zeeburg, H.J., Snijders, P.J., Wu, T., Gluckman, E., Soulier, J., Surralles, J., Castella, M., van der Wal, J.E., Wennerberg, J., Califano, J., Velleuer, E., Dietrich, R., Ebell,

W., Bloemena, E., Joenje, H., Leemans, C.R., Brakenhoff, R.H., 2008. Clinical and molecular characteristics of squamous cell carcinomas from fanconi anemia patients. J. Natl. Cancer Inst. 100 (22), 1649–1653. Wang, S.S., Gonzalez, P., Yu, K., Porras, C., Li, Q., Safaeian, M., Rodriguez, A.C., Sherman, M.E., Bratti, C., Schiffman, M., Wacholder, S., Burk, R.D., Herrero, R., Chanock, S.J., Hildesheim, A., 2010. Common genetic variants and risk for HPV persistence and progression to cervical cancer. PLoS One 5 (1), e8667. Wingo, S.N., Gallardo, T.D., Akbay, E.A., Liang, M.C., Contreras, C.M., Boren, T., Shimamura, T., Miller, D.S., Sharpless, N.E., Bardeesy, N., Kwiatkowski, D.J., Schorge, J.O., Wong, K.K., Castrillon, D.H., 2009. Somatic LKB1 mutations promote cervical cancer progression. PLoS One 4 (4), e5137. Woods, A., Johnstone, S.R., Dickerson, K., Leiper, F.C., Fryer, L.G., Neumann, D., Schlattner, U., Wallimann, T., Carlson, M., Carling, D., 2003. LKB1 is the upstream kinase in the AMP-activated protein kinase cascade. Curr. Biol. 13 (22), 2004–2008. Yu, S.Y., Chan, D.W., Liu, V.W., Ngan, H.Y., 2009. Inhibition of cervical cancer cell growth through activation of upstream kinases of AMP-activated protein kinase. Tumour Biol. 30 (2), 80–85. Zagorska, A., Deak, M., Campbell, D.G., Banerjee, S., Hirano, M., Aizawa, S., Prescott, A.R., Alessi, D.R., 2010. New roles for the LKB1-NUAK pathway in controlling myosin phosphatase complexes and cell adhesion. Sci. Signal 3 (115), ra25. Zheng, B., Cantley, L.C., 2007. Regulation of epithelial tight junction assembly and disassembly by AMP-activated protein kinase. Proc. Natl. Acad. Sci. USA 104 (3), 819–822. Zheng, B., Jeong, J.H., Asara, J.M., Yuan, Y.Y., Granter, S.R., Chin, L., Cantley, L.C., 2009. Oncogenic B-RAF negatively regulates the tumor suppressor LKB1 to promote melanoma cell proliferation. Mol. Cell 33 (2), 237–247.