ARTICLE https://doi.org/10.1038/s42003-020-1092-0 OPEN STK11 (LKB1) missense somatic mutant isoforms promote tumor growth, motility and inflammation Paula Granado-Martínez 1,11, Sara Garcia-Ortega 1,11, Elena González-Sánchez1,11, Kimberley McGrail 1, 1234567890():,; Rafael Selgas 1, Judit Grueso1,2, Rosa Gil1, Neia Naldaiz-Gastesi1,9, Ana C. Rhodes1,10, Javier Hernandez-Losa3, Berta Ferrer1,3, Francesc Canals4, Josep Villanueva5, Olga Méndez5, Sergio Espinosa-Gil6, José M. Lizcano 6, Eva Muñoz-Couselo1,7, Vicenç García-Patos 1,8 & Juan A. Recio 1 ✉ Elucidating the contribution of somatic mutations to cancer is essential for personalized medicine. STK11 (LKB1) appears to be inactivated in human cancer. However, somatic mis- sense mutations also occur, and the role/s of these alterations to this disease remain unknown. Here, we investigated the contribution of four missense LKB1 somatic mutations in tumor biology. Three out of the four mutants lost their tumor suppressor capabilities and showed deficient kinase activity. The remaining mutant retained the enzymatic activity of wild type LKB1, but induced increased cell motility. Mechanistically, LKB1 mutants resulted in differential gene expression of genes encoding vesicle trafficking regulating molecules, adhesion molecules and cytokines. The differentially regulated genes correlated with protein networks identified through comparative secretome analysis. Notably, three mutant isoforms promoted tumor growth, and one induced inflammation-like features together with dysre- gulated levels of cytokines. These findings uncover oncogenic roles of LKB1 somatic muta- tions, and will aid in further understanding their contributions to cancer development and progression. 1 Biomedical Research in Melanoma-Animal Models and Cancer Laboratory- Vall d’Hebron Research Institute VHIR-Vall d’Hebron Hospital Barcelona-UAB, Barcelona 08035, Spain. 2 Experimental Therapeutics Group, Vall d’Hebron Institute of Oncology (VHIO), Barcelona 08035, Spain. 3 Anatomy Pathology Department, Vall d’Hebron Hospital Barcelona-UAB, Barcelona 08035, Spain. 4 Proteomics Laboratory, Vall d’Hebron Institute of Oncology (VHIO, Barcelona 08035, Spain. 5 Preclinical Research Program, Vall d’Hebron Institute of Oncology (VHIO, Barcelona 08035, Spain. 6 Protein Kinases and Signal Transduction Laboratory, Neuroscience Institute and Molecular Biology and Biochemistry Department, UAB, Bellaterra, Barcelona 08193, Spain. 7 Clinical Oncology Program, Vall d’Hebron Institute of Oncology (VHIO), Vall d’Hebron Hospital, Barcelona-UAB, Barcelona 08035, Spain. 8 Dermatology Department, Vall d’Hebron Hospital Barcelona-UAB, Barcelona 08035, Spain. 9Present address: Biodonostia, Neurosciences Area, Group of Neuromuscular Diseases, San Sebastian 20014, Spain. 10Present address: Barcelona Clinic Liver Cancer (BCLC) Group, Liver Unit, Hospital Clínic of Barcelona, University of Barcelona, Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Barcelona 08036, Spain. 11These authors contributed equally: Paula Granado-Martínez, Sara García Ortega, Elena González-Sánchez. ✉email: juan.recio@vhir.org COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio 1 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 S TK11 (Liver kinase 1, LKB1) was first identified as a tumor we show the functional consequences of the expression of suppressor gene through its association with the Peutz- LKB1Y49D, LKB1R87K, LKB1G135R, and LKB1D194Y human tumor Jeghers Syndrome (PJS)1. STK11 appears to be inactivated missense mutants compared to the wild-type (LKB1WT) isoform. or mutated in sporadic cancers whose spectrum of tumor types These mutants were selected according to their location in the 3D suggests cooperation with exposure to environmental carcino- LKB1 structure, which affect different parts of the protein and gens. Thus, alterations in LKB1 have been found in non-small- may contribute to its different functions. We investigated their cell lung cancer (NSCLC), malignant melanoma, and cervical tumor suppressor effectiveness and their contributions to cell cancer among others2–4. cycle regulation, motility, and modification of the extracellular The serine/threonine kinase LKB1 belongs to the calcium cal- microenvironment. Finally, we also investigated their contribu- modulin family, which is ubiquitously expressed in several tissues tion to in vivo tumor growth. and highly conserved among eukaryotes. Over the past 15 years, LKB1 has been implicated in a number of essential biological Results processes such as: cell cycle control5,6, cellular energy LKB1Y49D, LKB1G135R, and LKB1D194Y mutants lack LKB1WT metabolism7,8, angiogenesis9,10, cell polarity11, and DNA damage tumor suppressor activity. Initially, mutated residues were response12. The sub-cellular localization and activity of LKB1 is localized within the primary and 3D LKB1 protein structure controlled through its interaction with STRAD and the armadillo (heterotrimeric LKB1-STRADα-MO25α complex (2WTK.pdb)). repeat-containing mouse protein 25 (Mo25)13,14. LKB1 regulates While the Y49 residue was embedded within a β-sheet in the the activity of at least 14 downstream kinases related to the LKB1 N-lobe, two consecutive arginines R86 and R87 in the αB- AMPK family15 and phosphorylates other substrates including helix were completely exposed on the surface of the protein, and STRAD16, PTEN17, and p21CDKN1A12. LKB1 is phosphorylated the G135 and D194 residues were located at the ATP-binding on at least eight residues, and evidence suggests that LKB1 auto- cleft of LKB1 (Fig. 1a and Supplementary Fig. 1a), suggesting phosphorylates itself on at least four of these, whereas the other different functional consequences for these mutations. To study four are phosphorylated by upstream kinases8,16. While these the biological contributions of LKB1 mutant isoforms to tumor post-translational modifications seem not to modify its kinase cells, we initially infected A549 cells, which lack the expression of activity, they are involved in the different biological responses LKB1, with an inducible lentiviral vector containing the different associated with LKB1, and likely in its interactions with other LKB1 isoforms allowing the expression of physiological amounts partners. of LKB1 (Fig. 2b, c and Supplementary Fig. 1b, c). The result of Up to date, more than 400 unique mutations have been colony-formation assays and proliferation curves confirmed the described for the STK11 gene, where ~70% of these mutations tumor-suppressive activity of LKB1WT. LKB1R87K conserved the promote the truncation of the protein and the other 30% repre- tumor suppression activity of LKB1WT, however, LKB1Y49D, sent missense mutations (COSMIC and TCGA-Bioportal). As a LKB1G135R, and LKB1D194Y isoforms lost this function (Fig. 1d, tumor suppressor, a number of studies have shown the con- e). These results were corroborated in HeLa and G361 cell lines, tributions of the genetic loss of LKB1 to tumorigenesis. It has which were also null for LKB1 expression (Supplementary Fig. 1d, been demonstrated that LKB1 controls cell cycle through the e). Furthermore, these data were supported by cell cycle analysis transcriptional regulation of Cyclin D1 and p21CDKN1A56, where studies in which expression of LKB1WT and LKB1R87K but not re-expression of LKB1 leads to G1 cell cycle arrest. The role of the other isoforms led cells to be arrested in G1 phase (Fig. 1f). LKB1 in controlling cell metabolism through AMPK signaling has Metabolic profiling of cells showed a significant increase in been widely documented. We know that the LKB1-AMPK axis mitochondrial proton leak in LKB1G135R expressing cells that controls lipid and glucose metabolism, and acts as a negative might have a have a major impact on mitochondrial coupling regulator of the Warburg effect suppressing tumor growth8,18. efficiency. In relation to this, G135R expressing cells also showed LKB1 is also important in the regulation of catabolic pathways a significant increase of ECAR. Although not significant, D194Y leading to the increase of glucose uptake and modulation of expressing cells also showed a higher acidification rate, suggesting glycolysis19 or the mobilization of lipid stores by stimulating an increase use of glycolysis by cells expressing any of these two lipases, such as adipose triglyceride lipase, to release fatty acids mutants (Fig. 1g). Mitochondrial dysfunction in LKB1G135R from triglyceride stores20. LKB1-AMPK-stimulated pathways also expressing cells correlated with their resistance to metformin, include increased turnover of macromolecules by autophagy, which is known to inhibit mitochondrial respiration (Supple- allowing the turnover of old and damaged molecules, or the mentary Fig. 1f). Thus, although LKB1R87K conserves in vitro replenishment of nutrient stores under starvation20. Additionally, tumor suppression capabilities, LKB1Y49D, LKB1G135R, and several investigations have suggested the role of LKB1 in reg- LKB1D194Y have lost this functional feature. ulation of physiological21 and pathological angiogenesis22 through the regulation of VEGF, MMP-2, MMP-9, bFGF, and NOX1 expression, and its participation in neurophilin-1 degra- Kinase activity, localization, and stability of LKB1 mutant dation23–25. Studies of LKB1 loss of function have also revealed its isoforms. Since the mutations affect different structural domains role in cell polarity and motility through the regulation of PAK115 of LKB1, we next investigated the specific kinase activity of the and the modulation of the phosphorylation status of FAK and different LKB1 mutant isoforms compared to that of wild type. In CDC42 activation26. Together, these functions contribute to the vitro kinase assays in the presence of STRADα showed that the induction of epithelial mesenchymal transition (EMT) and LKB1R87K mutant exhibited similar kinase activity to the wild- metastasis27,28. In addition to this, in vivo experiments have type isoform. Although LKB1D194A has been described as a shown evidence for the contribution of LKB1 to genotoxic DNA kinase-dead mutant6, LKB1D194Y and LKB1Y49D showed ~30% of damage response and DNA damage repair12,29. the LKB1WT activity (p < 0.001), and LKB1G135R showed a 40% Despite our increased knowledge about contributions of the decrease in the kinase activity, compared to the wild-type isoform loss of LKB1 in the different biological responses, much less is (Fig. 2a). known about the specific biological contributions of the different LKB1 localizes to the nucleus through its nuclear localization STK11 mutants identified in human tumors. Previous studies signal (NLS) and it is translocated to the cytoplasm upon binding have elucidated the tumor suppressor capabilities and the kinase to STRADα13, which is also essential to fully activate LKB1 kinase activity of several mutants found in a group of PJS patients. Here activity. Immunofluorescence experiments showed nuclear and 2 COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 ARTICLE a b Untransfected LKB1WT LKB1Y49D LKB1R87K LKB1G135R LKB1D194Y R86 - Dox LKB1wt LKB1R87K R87 humCMV IE region LKB1Y49D R86 TRE cPPTcts LKB1G135R LKB1D194Y R87 HIV1 RRE HIV1 PBS Woodchuck hepatitis virus + Dox HIV U5 HIV U3 HIV1 R EGFP Hum PGK1 HIV1 R Beta Lactamase Y49 D194 RSV U3 IRES rtTA2 HIV U5 ColE1 origin G135 Y49 R86 c Dox - + - + - + - + - + 70kDa LKB1 50kDa R87 40kDa GAPDH 37kDa LKB1 WT LKB1 Y49D LKB1 R87K LKB1 G135R LKB1 D194Y d e -Dox +Dox 10 10 10 10 Dox - + Cell doublings p<0.001 Cell doublings Cell doublings Cell doublings Cell doublings 100 8 15 8 8 p<0.0001 8 80 6 Parental Parental 6 Parental 6 Parental 6 Parental LKB1WT 10 LKB1 G135R LKB1 D194Y 60 LKB1 WT LKB1Y49D LKB1R87K 4 p<0.01 4 p<0.001 4 4 40 5 2 2 2 2 20 0 0 0 0 0 0 0 24 48 72 96 0 24 48 72 96 0 24 48 72 96 0 24 48 72 96 0 24 48 72 96 160 Time (hours) Time (hours) Time (hours) Time (hours) Time (hours) LKB1Y49D 120 80 f G1 S G2/M 100% Number of colonies 40 80% 0 100 p<0.001 60% LKB1R87K 80 40% 60 20% 40 20 0% -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox 0 p<0.001 180 0h 0h 48h 0h 48h 0h 48h 0h 48h 48h 160 g D194Y LKB1G135R WT Y49D R87K G135R 120 % Baseline ECAR (mpH/min/cell) Oligomycin FCCP Rotenone 80 % Baseline OCR (pmol/min/cell) Oligomycin FCCP Rotenone 150 150 200 40 LKB1WT LKB1Y49D p<0.002 OCR % (Baseline) 0 LKB1WT 150 p<0.001 LKB1R87K 100 LKB1G135R 100 LKB1Y49D 100 LKB1D194Y LKB1R87K 100 LKB1WT LKB1D194Y 80 60 LKB1G135R LKB1Y49D 50 50 p<0.001 LKB1D194Y LKB1R87K 40 50 LKB1G135R p<0.001 20 LKB1D194Y 0 0 0 0 0 50 100 150 0 50 100 150 Baseline Max ATP-linked Proton leak Time (min) respiration respiration Time (min) Fig. 1 LKB1Y49D, LKB1G135R, and LKB1D194Y mutants lack the LKB1WT tumor suppressor activity. a Localization of mutated residues within the 3D structure of the LKB1-STRADα-MO25α complex (2WTK.pdb). Mutated residues are indicated. b Vector map used to clone the different LKB1 isoforms; on the right, expression of GFP in uninduced and induced infected cells with the different constructs is shown. c Western blot, showing the induction of expression of the different LKB1 isoforms in A549 cells, 24 h after doxycycline treatment. d Clonogenic assay with cells expressing the different LKB1 isoforms. Graphs show the quantification of the number of colonies. (n = 3 experiments ±SD). e Representative proliferation curves, showing the effect of LKB1 mutants on cell proliferation (n = 3 experiments ±SD; p-value was calculated by Student’s t test). f Graphs showing cell cycle analysis of cells expressing the different LKB1 isoforms for 48 h. Dark blue bars represent cell cycle phases showing significant changes (n = 3, p < 0.05 calculated by Student’s t test). g Graphs showing oxygen consumption rates (OCR), parameters associated (oxygen consumption baseline; mitochondria proton leak; mitochondrial maximal respiration and ATP-linked respiration), and the extracellular acidification rates (ECAR). The green area and the gray area represent proton leak from LKB1G135R and LKB1WT expressing cells, respectively (n = 5 ± SD; p-value was calculated by Student’s t test). cytoplasmic localization of LKB1WT, LKB1R87K, LKB1G135R, and LKB1R87K and LKB1D194Y confer tumor cells with increased LKB1D194Y. However, besides to the presence of STRADα, in vitro motility. It has been described that LKB1 plays a role in LKB1Y49D showed a mainly nuclear localization in the three metastasis, adhesion, and motility30. Furthermore, LKB1 serves different cell lines tumor models (A549, HeLa, and G361), as a focal adhesion kinase (FAK) repressor, where LKB1 suggesting a lack of binding of this isoform to STRADα (Fig. 2b depletion causes rapid focal adhesion site turnover31. Thus, we and Supplementary Fig. 2a, b). Indeed, analysis of the LKB1 investigated the specific contribution of the different isoforms immunocomplexes in two different cell lines (A459 and G361) to cell motility. As described previously, LKB1WT expression confirmed the lack of binding of LKB1Y49D to STRADα, which is repressed FAKY397 and FAKY861 phosphorylation. However, complimentary to the diminished kinase activity displayed by this LKB1R87K, LKB1G135R, or LKB1D194Y neither repressed FAK mutant (Fig. 2c and Supplementary Fig. 2c). The compromised activation nor increased the phosphorylation of these residues kinase activity of LKB1Y49D, LKB1G135R, and LKB1D194Y mutants (Fig. 3a), including the FAKY576 residue, which is the target of was confirmed through its direct target AMPK upon stress c-SRC (Fig. 3a), suggesting a possible alteration of motility by metabolic conditions (Fig. 2d). In addition, LKB1R87K, LKB1Y49D, LKB1 mutants. Taking advantage of the inducible lentiviral and LKB1D194Y isoforms showed a significant decrease in the construct, we measured the distance traveled by cells upon half-life of the protein compared to LKB1WT (Fig. 2e). Thus, expression of the different mutants by tracking the distance these data show that the investigated LKB1 mutant isoforms have migrated by fluorescent cells upon doxycycline treatment. diminished kinase activity through either different mechanisms LKB1R87K-expressing cells showed significantly increased or decreased protein stability. These data are also in agreement motility compared to wild-type cells or cells expressing the with the tumor suppressor role of the protein and the selection of other mutant isoforms. The LKB1D194Y mutant also promoted these mutations in human cancer. cell motility albeit not as strongly as LKB1R87K (Fig. 3b). COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio 3 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 a LKB1 specific kinase activity b 2.0 p<0.0001 Dapi LKB1 Merged LKB1WT p<0.001 mU/mg of protein 1.5 p<0.0001 1.0 LKB1Y49D 0.5 0.0 LKB1R87K T 9D K 5R 4Y W 87 19 Y4 13 R 1D G 1 B 1 1 1 B LK B B B LK LK LK LK IP: LKB1 c LKB1G135R 5R 4Y 9D K 13 87 19 T Y4 W G R D B1 B1 B1 B1 B1 LK LK LK LK LK T.L. 70kDa LKB1 50kDa 50kDa S.E LKB1D194Y STRADα 40kDa 50kDa L.E. 40kDa d LKB1WT LKB1Y49D LKB1R87K LKB1G135R LKB1D194Y e Chx H.G. H.G. H.G. H.G. H.G. L.G. L.G. L.G. L.G. L.G. Time 0h 1h 2h 4h 6h 70kDa 70kDa p-AMPKα LKB1WT 50kDa 50kDa 70kDa 70kDa Y49D AMPKα LKB1 50kDa 50kDa 70kDa 70kDa LKB1R87K 50kDa LKB1 50kDa 70kDa 50kDa STRADα LKB1G135R 50kDa 40kDa 70kDa 40kDa GAPDH LKB1D194Y 50kDa 37kDa 40000 LKB1 WT p<0.00004 p-AMPK/AMPK LKB1 Y49D 100 LKB1 R87K Arbitrary Units 30000 p<0.00002 H.G. LKB1 G135R % of protein L.G. LKB1D194Y 20000 50 10000 p<0.01 p<0.05 0 0 T 4Y 9D K 5R W 87 0 2 4 6 19 Y4 13 R B1 D G B1 B1 Time (hours) B1 B1 LK LK LK LK LK Moreover, expression of LKB1R87K but not LKB1WT induced degradation32, which was more evident upon LKB1R87K or the formation and reorganization of F-actin fibers (Fig. 3c). LKB1D194Y expression (Fig. 3e). Loss of β-catenin at the cell Matrigel 3D cultures showed that expression of LKB1 promoted junctions upon expression of the LKB1R87K or LKB1D194Y a significant increase in the number of spheres generated; mutants was also observed in the 3D cultures (Fig. 3f). Alto- however, only the expression of the LKB1Y49D, LKB1G135R, and gether, the data suggest that LKB1R87K and to a lesser extent LKB1D194Y mutant isoforms increased the size of the spheres LKB1D194Y, induce cell motility, promoting cytoskeleton reg- (Fig. 3d). As expected, expression of LKB1 promoted β-catenin ulation and β-catenin degradation. 4 COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 ARTICLE Fig. 2 Kinase activity, subcellular localization, and stability of LKB1 mutant isoforms. a Graph showing the specific in vitro kinase activity of LKB1 mutant isoforms (n = 4 assays per isoform ±SD; p-values were calculated by Student’s t test). b Representative immunofluorescence images showing the subcellular localization of LKB1 isoforms in A549 cells. DAPI staining shows nuclear staining (bars represent 250 µm and 50 µm respectively). c Western blot showing the amount of STRADα bound to LKB1 isoforms. Duplicate of immunoprecipitated LKB1 complexes are shown for each isoform in A549 cells. T.L. total lysate, L.E. long exposure, S.E. sort exposure. d Western blot showing the amounts of p-AMPK upon glucose starvation (low glucose (L.G.)) or complete media (high glucose (H.G.)). The graph shows quantification of p-AMPK (n = 3 experiments ±SD). LKB1 show the amounts of mutant isoforms. AMPK and GAPDH are showed as a loading control. e LKB1-isoform protein stability. A549 cells were treated with cycloheximide. (CHX, 5 µg/mL). Then a pulse chase experiment was performed for the indicated time points. The graph shows quantification of LKB1 isoforms (n = 3 experiments ±SD). A representative western blot is shown on the right. Dashed lines show the half-life of proteins. The p-value was calculated by Student’s t test. Vesicle trafficking regulatory molecules, adhesion molecules, belonging to the cytochrome p450 family is dysregulated in and cytokines are differentially regulated by LKB1 mutant LKB1R87K and LKB1G135R mutant cells (Fig. 4d). Albeit the low isoforms. To gain knowledge about the molecular mechanisms number of LKB1D194 mutated samples in the Lung adenocarci- involved in the observed phenotypes, we initially compared the noma Pan Cancer dataset (TCGA), the expression of most of the gene expression profile of A459 cells expressing and not expres- genes above investigated correlated with the observed in the sing LKB1WT. Unsupervised hierarchical clustering of genes (817 mutated human samples (Supplementary Fig. 3b). Altogether, genes 1.2-fold and p > 0.05) showed that LKB1WT promoted a these data show evidence supporting the aberrant regulation by distinct gene expression profile. To strength the relevance of our LKB1 mutant isoforms of several processes contributing to tumor gene dataset, we compared our 817 regulated genes with the 2080 development and progression, including intracellular vesicle unique regulated genes obtained from 15 different datasets (top trafficking, cell adhesion and motility, immunomodulation and 200 regulated genes in each data set) published in Kaufman metabolism. et al.33, comparing the gene expression profiles of human and murine tumors with or without the genetic loss of STK11. One Analysis of LKB1-isoform secretomes identifies differential hundred and fifty-one genes were common to both lists, where alterations in exocytosis regulation, cell adhesion and motility, 25% of the genes were also associated to the functional clusters and cytokine signaling. Since LKB1 expression was shown to described in that study33 (Supplementary Fig. 3a). Gene set regulate vesicle trafficking, cell adhesion and motility, and cyto- enrichment analysis showed that morphogenesis and develop- kine signaling, we decided to analyze the secretome of cells ment, cell signaling, proliferation, adhesion and motility, meta- expressing the different LKB1 isoforms to validate and confirm bolic regulation, and immunomodulation were among the some of the gene expression results and observed phenotypes. processes most significantly affected by LKB1WT expression Relative label-free protein quantification analysis from both (Fig. 4a). The most significantly regulated genes (FDR < 0.1) uninduced and induced cell secretomes expressing each LKB1 included STK11, AP1S3, and RUSC2 (upregulation), which are isoform allowed us to detect 594, 640, 617, 681, and 596 differ- involved in intracellular vesicle trafficking and PLAUR, and entially secreted proteins from cells expressing LKB1WT, TGFB1I1, which regulate cell migration, cell growth, proliferation, LKB1Y49D, LKB1R87K, LKB1G135R, or LKB1D194Y expressing cells, and the transcriptional activity of nuclear receptors. LKB1 also respectively (Supplementary Fig. 4a and Supplementary Data 1 promoted the downregulation of important immunomodulators and 2). In the case of LKB1WT, six proteins identified in the and cytokines such as: SEMA7A, IL8, CXCL1, metabolic enzymes secretome (BMP1 VCL, SOD2, IL8, GREM1 and CXCL1) were (TDO2 and SOD2), and adhesion molecules such as CDH1 also present in the top 31 regulated genes (FDR < 0.1) upon (Fig. 4b). The regulation of these genes upon LKB1WT reintro- LKB1WT expression and were regulated in the same direction duction was validated by RT-PCR in a subset of different samples (Supplementary Fig. 4b). Overall, 60–75% percent of the differ- confirming the regulation of these processes (Fig. 4c). Next, we entially expressed proteins (440 proteins) upon LKB1 expression asked whether the expression of these genes was differentially were common to all isoforms, and only 1–12% of the proteins regulated by the LKB1 mutant isoforms compared to LKB1WT. were identified upon specific LKB1-isoform expression (Fig. 5a All four mutants showed defective regulation of genes involved in and Supplementary Data 3). Pathway and biological process intracellular trafficking of vesicles and endosomes (AP1S3 and enrichment analysis of the list of proteins regulated by all LKB1 RUSC2) and of CDCP1, a molecule involved in cell adhesion and isoforms (Metascape.org) showed that exocytosis regulation, modulated by Src kinases (Fig. 4d). TGFB1I1, a molecular adapter extracellular matrix organization and remodeling, wound-healing linking various intracellular signaling modules to plasma mem- response, signaling by interleukins and response to oxidative brane receptors regulating the Wnt and TGFβ signaling path- stress were the most significantly enriched processes. (Fig. 5a and ways, was also downregulated by LKB1R87K or LKB1D194Y. Supplementary Fig. 5). Most of the proteins that were upregulated PLAUR a protein involved in processes related to cell-surface in the secretome of cells expressing the different LKB1 mutants plasminogen activation and localized degradation of the extra- compared to LKB1WT were either common to the different cellular matrix was upregulated in LKB1R87K or LKB1G135R, and protein lists (Fig. 5b circos plot purple lines) or proteins that DUSP1 which is associated to the negative regulation of cellular belonged to the same enriched ontology term (Fig. 5b circos plot proliferation was downregulated in LKB1Y49D, LKB1G135R, and blue lines). Again, biological process enrichment analysis of all LKB1D194Y-expressing cells (Fig. 4d). One interesting observation upregulated proteins from these four groups (upregulated pro- was the downregulation of IL8 and CXCL1 upon LKB1WT teins in LKB1 mutants vs. LKB1WT) resulted in the identification expression, whereas expression of LKB1Y49D and LKB1D194Y of processes such as protein processing, cell adhesion and motility isoforms either increased or failed to downregulate the expression metabolic regulation and cytokine signaling (Fig. 5b). The results of these immunomodulators (Fig. 4d). TDO2, a rate-limiting from the individual analysis of each LKB1 mutant compared to enzyme in tryptophan metabolism, was differentially regulated in LKB1WT agreed and supported previous observations. The LKB1Y49D, LKB1G135R, and LKB1D194Y-expressing cells, while LKB1Y49D mutant induced the secretion of proangiogenic, the antiapoptotic and stress protective protein SOD2 was mark- inflammatory and immunomodulator molecules (CXCL1, edly upregulated by LKB1D194Y. CYP1B1 a detoxification protein CXCL2, CXCL3, CXCL5, IL8 MMP7 and TIMPs) (Fig. 5c). In COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio 5 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 a - Dox + Dox p=0.0033 p-FAKY576/FAK p-FAKY861/FAK p=0.0158 LKB1G135R LKB1G135R LKB1D194Y LKB1D194Y p-FAKY397/FAK LKB1Y49D LKB1R87K LKB1Y49D LKB1R87K 2.5 LKB1WT LKB1WT 2.5 p=0.0017 2.0 p=0.02 2.0 2.0 Fold Induction 1.5 p=0.0018 150kDa p-FAKY397 1.5 1.5 100kDa 1.0 150kDa 1.0 1.0 p-FAKY576 100kDa 0.5 150kDa 0.5 0.5 p-FAKY861 0.0 0.0 0.0 100kDa -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox 150kDa -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox -Dox +Dox FAK 100kDa T 4Y 9D K 5R W T 4Y 9D K 5R 87 T 4Y 9D K 5R 19 W Y4 13 87 W 87 R 19 Y4 13 19 B1 Y4 13 D G R LKB1 R B1 B1 D G B1 B1 D G B1 B1 B1 B1 50kDa B1 B1 LK B1 B1 B1 B1 LK LK LK LK LK LK LK LK LK LK LK LK LK LK b Cell motility c Dapi LKB1 Ph.+EGFP Merge Dox (-) ≤p<0.001 350 p<0.0001 WT LKB1 300 Dox(+) 250 m/80h 200 150 Dox (-) 100 50 R87K LKB1 0 LKB1WT LKB1Y49D LKB1R87K LKB1G135R LKB1D194Y Dox(+) d 800 p<0.0001 p<0.005 Number of Spheres p<0.0082 p<0.0001 p<0.005 600 p<0.001 p<0.0001 p<0.001 3 105 Sphere size ( m2) 400 2 105 200 1 105 0 Dox - + - + - + - + - + LKB1WT LKB1Y49D LKB1R87K LKB1G135R LKB1D194Y 0 Dox - + - + - + - + - + LKB1WT LKB1Y49D LKB1R87K LKB1G135R LKB1D194Y e LKB1WT LKB1Y49D LKB1R87K LKB1G135R LKB1D194Y f LKB1R87K β-Catenin Dapi Merged 24h 48h 24h 48h 24h 48h 24h 48h 24h 48h Dox (-) 0h 0h 0h 0h 0h Dox β-Catenin 100kDa LKB1R87K LKB1 50kDa 50kDa β-Actin Dox(+) 40kDa 2.0 β-Catenin P=0.0070 P=0.0053 1.5 LKB1D194Y Fold Induction P=0.0091 Dox (-) 1.0 LKB1D194Y 0.5 Dox(+) 0.0 0h 24h 48h 0h 24h 48h 0h 24h 48h 0h 24h 48h 0h 24h 48h LKB1WT LKB1Y49D LKB1R87K LKB1G135R LKB1D194Y Figure 3 agreement with the above results, LKB1R87K promoted the molecules related to extracellular remodeling and adhesion secretion of molecules related to extracellular matrix remodeling, (VCNA, VTN, FLNA, VIM and IQGAP1) (Fig. 5e-f). In the case cytoskeletal maintenance and adhesion, such as MMP2, FN1, of LKB1D194Y, we also identified an increase in the number of IQGAP1 and FLNA (Fig. 5d). Expression of LKB1G135R and molecules involved in the activation of growth factors (cathepsins LKB1D194Y promoted the alteration of common networks of and PAPPA), mitogenic growth factors (PDGFD) and immuno- proteins, including protein-processing components (PSME1, modulators (IL8) (Fig. 5f). Thus, these results support the PSME2 and PSME3), as well as the increased secretion of described phenotypic and molecular observations related to each 6 COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 ARTICLE Fig. 3 Differentially regulated cell motility by LKB1 mutant isoforms. a Western blot showing the regulation of FAK phosphorylation after the expression of LKB1 mutant isoforms for 48 h. Graphs show quantification of every phosphorylation site. (n = 3, p-value was calculated by Student’s t test). b A graph showing the quantification of A549 cell motility induced by LKB1 isoforms. The movement of 35–50 cells expressing each different isoform (colored circles) was quantified for 80 h and plotted. The p-value was calculated by Student’s t test. c Representative images of LKB1 immunofluorescence (red) and F-actin (phalloidin, green). Cells appear filled with green upon doxycycline treatment because of IRES-GFP expression linked to LKB1 expression. Bars represent 100 µm and 20 µm. d Graphs showing the quantification of the number and size of spheres grown in Matrigel upon expression of the indicated LKB1 isoform. For number of spheres experiments were performed in triplicate. For determining size, 200 spheres per group (from the three experiments) were analyzed. p-value was calculated by Student’s t test. e Representative western blot showing the amount of β-Catenin upon LKB1 isoform induction. β- Actin is shown as loading control. Graphs shows normalized quantification. (n = 3, p-value was calculated by Student’s t test). f Immunoflurescence of Matrigel spheres showing the amounts of β-catenin upon the expression of the indicated LKB1 isoforms. DAPI staining shows nuclear staining. Bars represent 500 µm. LKB1 mutant in regulating cell proliferation, cell motility and that of the wild-type isoform, the enzymatic activity in LKB1Y49D, cytokine production. LKB1G135R, and LKB1D194Y was almost extinct, even in the presence of STRADα14. In this regard, mutations at residue Asp194 affect the conserved DLG triplet lying in the activation Distinct contributions of LKB1 mutant isoforms to in vivo loop that helps to orientate the γ phosphate of ATP for transfer34; tumor growth. Next, we studied the specific contribution/s of in fact LKB1D194A is considered a dead kinase. However, each LKB1 mutant to in vivo tumor growth. As expected (Fig. 1), LKB1D194Y still has ~20% of the in vitro LKB1WT kinase activity, LKB1WT and LKB1R87K behaved as tumor suppressors. The possibly because the deprotonated OH group of the Tyr residue reduction in tumor growth was more evident for the wild-type still has a negative charge that might help to conserve some isoform, and in both cases (LKB1WT and LKB1R87K), some of the amount of activity. Residue Gly135 is also located in the ATP tumors showed more differentiated areas with no proliferating binding cleft; thus, it is tempting to speculate that substitution of (Ki67-negative) nested polarized epithelial cells (Fig. 6a, b). a Gly for a charged amino acid might affect the ATP binding to LKB1Y49D promoted tumor growth and, in agreement with the LKB1 and, consequently its catalytic activity. Importantly, the above results, these tumors presented signs of hemorrhage and described kinase activity of the different isoforms correlated with inflammation that actively contributed to the tumor size (Fig. 6c). the metabolic stress response showed by the different cell lines Indeed, these tumors showed increased angiogenesis (ERG posi- expressing the different isoforms, according to the amounts of the tive), increased proliferation (Ki67 positive), and increased direct LKB1 target AMPK. amounts of cytokines (CXCL1, 2 and 3) and immunomodulatory The subcellular localization and activity of LKB1 is controlled molecules (IL8). Again, LKB1Y49D showed nuclear localization through its interaction with STRADα and the armadillo repeat- within tumor cells (Fig. 6c). Loss of tumor suppressor function in containing mouse protein 25 (Mo25)13,14. All mutants except the LKB1G135R and LKB1D194Y mutants was also reflected in vivo. LKB1Y49D showed nuclear-cytoplasmatic localization. LKB1Y49D Tumor growth was particularly promoted by the LKB1G135R localized in the nucleus and had impaired activity, which is in mutant which also showed increased production of vimentin. agreement with its diminished binding capability to STRADα. According to the Ki67 proliferation marker both mutants pro- Interestingly, recent studies support that LKB1Y49D mutation moted tumor cell proliferation (Fig. 6d, e). We did not observe promotes variations in the binding energy pertaining to spatial metastasis in any of the models tested. Thus, the in vitro tumor conformation and flexibility, impairing the binding to STRADα suppressor capabilities of the investigated LKB1 mutants were and MO2535. Thus, the tumor suppressor activity linked to LKB1 reflected in vivo. Additionally, tumors harboring the LKB1Y49D kinase activity could be acquired through STRADα-dependent mutation showed phenotypic features compatible with the (LKB1G135R and LKB1D194Y) or STRADα-independent mechan- molecular profile induced by this isoform. isms (LKB1Y49D); the later also affect the subcellular localization and most likely other processes, such as transcriptional regula- Discussion tion6. In addition, LKB1Y49D, LKB1R87K, and LKB1D194Y showed Loss of function of the tumor suppressor STK11 (LKB1) has been a significantly shorter half-life than LKB1WT, and LKB1R87K was observed in many types of cancer. However, LKB1 also plays a the isoform showing the shortest half-life. In this case, the sub- role in a number of pathways involved in controlling metabolism, stitution of the Arg residue by a Lys residue could promote post- cell growth, angiogenesis, adhesion, and motility; even recently, translational modifications that might affect the protein stability. LKB1 has been related to immunotherapy responses. Therefore, it Beyond its tumor suppressor activity, it is known that is likely that the tumorigenic potential of LKB1 mutations is LKB1 serves as a FAK repressor to stabilize focal adhesion sites, mediated through alternative mechanism/s. Here, we studied the contributing to cell adhesion and motility31. Both LKB1R87K and contribution of four LKB1 somatic mutations to tumorigenesis LKB1D194Y failed to repress FAK phosphorylation compared to in vitro and in vivo, confirming the role of LKB1 mutants in pro- the wild-type isoform. Interestingly, the LKB1R87K mutant pro- tumorigenic processes such as adhesion, motility, angiogenesis, moted cell motility together with cytoskeletal reorganization. This and inflammation. effect on cell motility was also induced by LKB1D194Y, although it As a tumor suppressor, LKB1 is usually lost in human cancer. was less significant than the effects seen in LKB1R87K cells. In Thus, selected missense LKB1 mutations should contribute with relation to this, it is known that LKB1 induces the degradation of either an additional useful feature/s for the tumor or loss of the β-catenin, a molecule involved in cell–cell contact in epithelial tumor suppressor function of the protein. Three out of the four cells36. Both LKB1R87K and LKB1D194Y promoted more β-catenin LKB1 mutants studied, failed to function as tumor suppressors in degradation than the wild-type counterpart and induced the three different cell lines from three different tumor types (lung downregulation of TGFB1I1 and CDCP1, two molecules also cancer, melanoma, and cervical cancer). However, LKB1R87K involved in adhesion and motility37,38. behaved as the wild-type isoform in controlling cell proliferation. It is clear that reconstitution of LKB1 implies changes in the In fact, while the kinase activity of LKB1R87K was comparable to transcriptional profile of cells that were supported by previous COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio 7 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 a b -Dox +Dox Adhesion and motility ZNF655 2.11 TNC GSEA MSIgDB GLIPR1 STK11 Cell signaling AP1S3 BMP1 SOD2 GREM1 Apoptosis OXTR PLAUR Other DUSP1 HDAC3 TGFB1I1 SORL1 GEM DUSP1 KRT80 OPTN VCL TGFB1I1 Morphogenesis and development STK11 IL8 PLAUR SEMA7A Cell proliferation and cell cycle SEMA7A CYP1B1 0.00 PDZK1IP1 BMP1 Immune system regulation IL8 CXCL1 CXCL1 RARRES3 Metabolism regulation RARRES3 OXTR SOD2 TNC HDAC3 KRT80 PPAPDC1B VCL HSD17B14 GEM SORL1 AP1S3 GREM1 ATP9A OPTN HSD17B14 CYP1B1 PPAPDC1B TDO2 CABLES1 -2.11 0 1 E-3 2 E-3 3 E-3 4 E-3 5 E-3 6 E-3 7 E-3 CABLES1 Downregulated in cells expressing LKB1 FDR q-value Upregulated in cells expressing LKB1 c STK11 PLAUR RUSC2 AP1S3 GLIPR1 TGFB1IP1 SEMA7A IL8 CXCL1 TDO2 RARRES3 SOD2 CDH1 2.5 p=0.00012 1.5 2.5 3.0 2.5 1.5 1.5 p=0.0052 1.5 1.5 1.5 1.5 1.5 1.5 mRNA Fold Induction p<0.001 p=0.0001 p=0.002 p<0.001 p=0.0003 p<0.001 p<0.0001 p<0.0001 2.0 2.0 2.0 1.0 2.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.5 1.5 1.5 1.0 1.0 1.0 0.5 1.0 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 +D x ox +D x ox +D x ox +D x ox +D x ox +D x ox +D x ox +D x ox +D x ox ox ox +D x ox +D x ox +D x ox o o o o o o o o o o o o -D -D -D -D -D -D -D -D -D -D -D +D -D -D d RUSC2 AP1S3 TGFB1IP1 CDCP1 3 4 4 4 LKB1WT p<0.001 LKB1WT LKB1WT LKB1WT mRNA Fold Induction p<0.001 LKB1Y49D p<0.0001 3 LKB1Y49D 3 LKB1Y49D LKB1Y49D p<0.0001 3 2 LKB1R87K LKB1R87K LKB1R87K LKB1R87K p<0.0001 LKB1G135R LKB1G135R LKB1G135R p<0.001 p<0.001 2 2 2 LKB1G135R p<0.0001 LKB1D194Y LKB1D194Y LKB1D194Y LKB1D194Y 1 1 p<0.001 1 1 0 0 0 0 - Dox + Dox - Dox + Dox - Dox + Dox - Dox + Dox PLAUR DUSP1 IL8 CXCL1 4 4 4 4 LKB1WT mRNA Fold induction LKB1WT LKB1WT LKB1WT 3 LKB1Y49D 3 LKB1Y49D 3 LKB1Y49D p<0.0001 3 LKB1Y49D p<0.0001 LKB1R87K p<0.05 p<0.05 LKB1R87K p<0.0001 p<0.001 p<0.05 LKB1R87K LKB1R87K 2 LKB1G135R 2 LKB1G135R 2 LKB1G135R LKB1G135R 2 LKB1D194Y LKB1D194Y LKB1D194Y LKB1D194Y 1 1 1 1 0 0 0 0 - Dox + Dox - Dox +Dox - Dox + Dox - Dox + Dox TDO2 SOD2 CYP1B1 4 3 4 LKB1WT mRNA Fold induction LKB1WT LKB1WT LKB1Y49D LKB1Y49D 3 3 LKB1Y49D LKB1R87K 2 LKB1R87K LKB1R87K LKB1G135R LKB1G135R LKB1G135R 2 p<0.001 2 p<0.001 p<0.001 LKB1D194Y LKB1D194Y LKB1D194Y p<0.0001 p<0.001 p<0.001 1 1 1 0 0 0 - Dox + Dox - Dox + Dox - Dox +Dox Fig. 4 Vesicle trafficking regulating molecules, adhesion molecules, and cytokines are differentially regulated by LKB1 mutant isoform. a Gene set enrichment analysis of genes showing >1.2-fold difference (p < 0.05) in A549 cells expressing LKB1 compared to non-LKB1 expressing cells. The most significant ontology terms are shown in different colors. b Heat map of the unsupervised hierarchical clustering of the top regulated genes (FDR < 0.1) upon LKB1 expression. On the right the classification of the top regulated genes within the ontology terms (same code color as in a) is shown. c Validation by RT- PCR of genes regulated upon LKB1WT expression. The graphs show the mean of three different experiments. The p-values were calculated by Student’s t test. d RT-PCR of the indicated genes upon the expression of the different LKB1 isoforms (+doxycycline). The graphs show the mean of three different experiments. p-values were calculated by Student’s t test. 8 COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 ARTICLE a -Dox 48h. Regulated exocytosis LKB1 Isoform Secretome Extracellular matrix organization +Dox MS/MS Axon guidance Regulation of IGF transport and uptake by IGF Bi Platelet degranulation Cellular responses to stress Degradation of the extracellular matrix ECM proteoglycans Response to wounding Supramolecular fiber organization Cofactor metabolic process Signaling by interleukins Protein folding Infectious disease Cell morphorgenesis involved in differentiation Response to oxydative stress Response to inorganic substance Regulation of peptidase activity Lysosome Negative regulation of cellular component organization Number of identified proteins in LKB1expressing cells Protein Folding Metabolic Regulation Ribosomal and mRNA processing b MCODE_1 MCODE_2 MCODE_3 MCODE_4 MCODE_5 Proteasome Cytoskeleton, Cell Adhesion, Motility Cytokine Signaling MCODE_7 MCODE_9 MCODE_6 MCODE_8 MCODE_12 MCODE_11 c d WT vs. R87K e f WT vs. D194Y WT vs. Y49D WT vs. G135R -10 -5 -20 -15 ACTB -8 -4 CTSD -15 Log p-Value Log p-Value VCAN Log p-Value VIM Log p-Value -10 CTSL1 -6 -3 CTSB IGFBP1 VCAN FN1 MMP2 -10 -4 CXCL5 -2 TIMP2 CXCL3 MMP7 CXCL2 TPM4 -5 TIMP1 PLEC FLNA VCAN PAPPA VTN -2 SOD1 IL8 -1 CORO1C FLNA IQGAP1 CDH6 -5 PSMA6 MMP2 IL8 SPTAN1 PSMB6 ACTN4 SPTBN1 PSMA1 CDC37 CTTN NRP2 ACTN1 PDGFD 0 CXCL1 0 IQGAP1 0 -2 -1 0 1 2 0 -2 -1 0 1 2 -2 -1 0 1 2 -2 -1 0 1 2 Log FC Log FC Log FC Log FC WT Y49D R87K G135R D194Y WT Y49D R87K G135R D194Y Dox - + - + - + - + - + Dox - + - + - + - + - + MMP9 IL8 10kDa MMP2 70kDa 3 IL8 2.0 MMP2 activity 1.5 Fold Induction Fold Induction 2 1.0 1 0.5 0 0.0 Dox - + - + - + - + - + Dox - + - + - + - + - + WT Y49D R87K G135R D194Y LKB1 LKB1 LKB1 LKB1 LKB1 LKB1WT LKB1Y49D LKB1R87K LKB1G135RLKB1D194Y LKB1Y49D 24h 48h 0h 14kDa GRO α,β,γ 10kDa Ponceau S studies33. Some genes involved in vesicle trafficking, regulation of of regulated genes, given to their clinical relevance, is cytokines autophagy, and inflammation (RUSC2 and AP1S3) failed to be (i.e., IL8 and CXCL1) which become downregulated upon STK11 upregulated by the mutant isoforms, which might have some expression. Our data suggest that loss of LKB1 and/or somatic effects in autophagosome maturation and distribution, as well as mutation of this protein in tumor cells might lead to the upre- proinflammatory consequences39–42. The latter, could be more gulation of these cytokines contributing to the deregulation of the prominent in LKB1Y49D expressing cells that showed a stronger immune response and tumor angiogenesis. In particular, both dysregulation of AP1S341,42. Another particularly interesting set LKB1Y49D and LKB1D194Y not only failed also repress IL8 and COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio 9 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 Fig. 5 Secretome analysis induced by LKB1 isoforms identified differential alterations in exocytosis regulation, cell adhesion and motility, and cytokine signaling. a The recovered conditioned media from A549 parental and LKB1-isoform induced cells (after 48 h) was analyzed by mass spectrometry. Venn diagram showing the common proteins identified in the secretomes obtained from cells expressing the different LKB1 isoforms. On the right, a functional enrichment and interactome analysis of all the protein identified after LKB1-isoform expression is shown. The intracluster and intercluster similarities of enriched terms are shown. Cluster annotations are shown in color coded. (metascape.org). b Circos plot showing the overlap between protein lists (increased proteins in the LKB1 mutant isoforms vs. LKB1WT). The purple curves indicate overlap at the protein level and the blue curves link proteins that belong to the same enriched ontology term. On the right the protein–protein interaction network and the MCODE component48 identified in the gene lists are shown. c Volcano plot depicting protein data p-values vs. fold change after comparison of proteomics data from LKB1WT vs. LKB1Y49D cells. The most representative protein–protein interaction network is shown. Below, western blot showing the levels of IL8 and CXCL1, CXCL2, and CXCL3 in A549 cell’s conditioned media after 48 h of LKB1 induction (IL8) or at the indicated time points. The graphs show quantification of the bands. Ponceau S was used as a loading control. d Same as in c, showing data from LKB1WT vs. LKB1R87K cells. Below, zymogram showing the MMP9 and MMP2 activity in conditioned media upon induction of the different LKB1 isoforms. The graphs show quantification of the bands. e, f Same as in c showing data from LKB1WT vs. LKB1G135R and LKB1WT vs. LKB1D194Y cells, respectively. In the protein network, pink circles depict shared proteins between two sets of samples and red circles represent proteins detected only in each set. CXCL1 transcription but induced their transcriptional regulation. LKB1 mutants that are catalytically deficient enhance cyclin D1 Vascular abnormalities in part due to deregulation of VEGF have expression6, which contributes to tumor growth. This finding also been described in LKB1-deficient mice. Our observations suggest supports the ability of these mutants to differentially regulate the that loss of STK11 and/or LKB1Y49D and LKB1D194Y somatic expression of specific genes (i.e., LKB1Y49D regulation of IL8 and mutations in tumor cells could also contribute to angiogenesis CXCL1) that might contribute to tumor development and and immune response through upregulation of cytokines. This progression. piece of data could be especially relevant in relation to the clinical In summary, we show that beyond the role of the non-mutated responses to immunotherapy observed in LKB1-deficient KRAS- protein as a tumor suppressor, missense LKB1 somatic mutations mutated lung tumors43. TDO2 an enzyme catalyzing the pro- could contribute to tumor development and/or progression by duction of kynurenine, which promotes immune-tolerant den- modifying not only intrinsic cell capabilities such as proliferation, dritic cells (DCs) and regulatory T cells and thus contributes to an motility, or adhesion but also the tumor microenvironment, immunosuppressive environment, is also transcriptionally upre- affecting inflammatory responses and likely the immune system. gulated through the loss of STK11 or mutated isoforms. The Interestingly, these effects can be both kinase dependent and contributions of these molecules to immunotherapy responses in kinase independent, unveiling possible roles for LKB1 indepen- a STK11-deficient or STK11-mutated context are currently under dent of its enzymatic activity. These results will contribute to investigation. clarify the unknown significance of missense somatic LKB1 Comparative secretome analysis of samples not only confirmed mutations in human cancer, assisting with the diagnosis of dis- the role of LKB1 in regulating processes involved in motility and eases to help guide optimal treatment. cell adhesion such as extracellular matrix organization and morphogenesis, but also revealed its participation in processes Methods such as vesicular transportation and cytokine production. In fact, Reagents. Doxycycline, cycloheximide, and Ponceau S solution were obtained these results support both the possible role of LKB1Y49D in reg- from Sigma-Aldrich Quimica (Madrid, Spain). Horseradish peroxidase and sec- ulating IL8 and CXCL1, 2, 3, and 5, and promoting inflammation ondary fluorescent antibodies were obtained from GE Healthcare (Little Calfont, and angiogenesis, and the contribution of LKB1R87K to cell UK) and Thermo Scientific (Fremont, CA, USA), respectively. FITC-phalloidin was purchased from Abcam (Cambridge, UK); antibodies against LKB1 (ley37D/G6), motility and adhesion through the regulation of extracellular IL-8, and GROα,β,γ (CXCL1, CXCL2, and CXCL3) were obtained from Santa Cruz matrix-remodeling molecules such as MMP2 and protein net- Biotechnology (Heidelberg, Germany); p-AMPKα, AMPKα, Anti-LKB1 works related to this phenotype. LKB1G135R and LKB1D194Y (D60C5F10), and Anti-β-catenin antibodies were obtained from Cell Signaling mutants showed certain similarities: both mutants lost the tumor (Leiden, The Netherlands); anti-Ki67 antibodies were purchased from Abcam (Cambridge, UK); and anti-vimentin and anti-ERG antibodies were obtained from suppressor capability of LKB1 and both mutations affect the ROCHE, Ventana (Basel, Switzerland). Anti-E-cadherin antibodies were purchased LKB1 ATP binding cleft. In addition, the secretome analysis also from R&D Systems (Minneapolis, MN, USA). Anti-GAPDH antibodies were identified similar protein–protein networks for these two iso- purchased from Trevigen (Gaithersburg, MD, USA). forms, including processes networks involved in antigen proces- sing, which could be relevant for immunotherapy. Construct generation. The pLenti-rtTA2-IRES-H2B-GFP doxycycline-inducible Notably, the tumor suppressor activity of LKB1WT and plasmid was obtained from S. Tenbaum, HG Palmer’s Lab (Vall d´Hebron Institute LKB1R87K was also observed in vivo. Unfortunately, we did not of Oncology, VHIO). The human LKB1 sequence was subcloned from pCMV5- Flag-LKB1WT to obtain pLenti-rtTA2-LKB1WT-IRES-GFP. The different mutant observe any increased metastasis promoted by LKB1R87K. Inter- isoforms were generated via site-directed mutagenesis using the Quick-Change II estingly, all tumors expressing LKB1Y49D showed signs swelling Kit (Stratagene, Cedar Creek, TX, USA) and the following primers: LKB1Y49D: supporting a role of LKB1Y49D in regulating cytokine production ATCGGCAAGGACCTGATGGGG, LKB1R87K: AAGTTGCGAAAGATCCCCA and inflammation-related processes. These data were braced by AC, LKB1G135R: GCGTGTGTCGCATGCAGGAAA and LKB1D194Y: AAAATCT an increased expression of IL8 and GROα, β, and γ (CXCL1, 2, CCTACCTGGGCGTG. and 3) in these tumors compared to the expression of these in parental cells or LKB1WT−expressing tumors. LKB1G135R and Cell culture. A549 lung cancer cells, G361 melanoma cells and HeLa cells, all of LKB1D194Y not only did not function as tumor suppressors but which were null for STK11 mutations, were obtained from ATCC. Cells were grown in Dulbecco’s modified Eagle medium (DMEM) (Biowest, Riverside, MO, also the promoted tumor growth. Since parental cells lack the USA) supplemented with 10% fetal bovine serum (FBS) (Biowest) and 100 µg/mL tumor suppressor activity of LKB1, these results suggest that penicillin/streptomycin (Thermo Scientific, Waltham, MA, USA) and maintained LKB1G135R and LKB1D194Y mutants promote in vivo tumor at 37 °C and 5% CO2. Cells were infected with the doxycycline-inducible construct growth by a mechanism independent of their kinase activity, a rtTA2-H2B-GFP containing the different isoforms of LKB1 (LKB1WT, LKB1Y49D, LKB1R87K, LKB1G135R, or LKB1D194Y). Cells were induced with 1 µg/mL doxycy- result that is reflected in the proliferation index of the tumors cline (Sigma) for 48 h, and green fluorescent cells were sorted via a FacsAria Digital (Ki67 staining). In agreement with this, it has been described that Cell Sorter (BD Biosciences, San Jose, CA, USA). For glucose starvation 10 COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 ARTICLE a DOX - + LKB1WT b DOX 200µm LKB1R87K Dox - + Dox + - + - 200µm 100µm Ki67 Tumor growth (Fold induction) 8 Tumor growth (Fold induction) 10 - Dox - Dox 100µm + Dox 6 + Dox p<0.0178 4 5 VIMENTIN Ki67 2 n.s. 0 0 0 10 20 30 40 50 GRO α,β,γ 0 10 20 30 40 Time (days) Time (days) LKB1 LKB1 IL8 DOX c LKB1Y49D DOX DOX d - + - + - + G135R - Dox + LKB1 Dox + 200µm - 200µm ERG Tumor growth (Fold induction) 50 Tumor growth (Fold induction) 40 - Dox 100µm 150 100µm p<0.05 - Dox + Dox Ki67 Ki67 30 + Dox 100 20 p<0.0001 10 GRO α,β,γ 50 VIMENTIN 0 0 10 20 30 40 Time (days) 0 0 10 20 30 40 Time (days) LKB1 LKB1 IL8 DOX e D194Y - + LKB1 Dox + - 200µm 50 - Dox 40 + Dox 100µm 30 p<0.01 Ki67 20 10 0 LKB1 0 10 20 30 40 Time (days) experiments cells were culture in DMEM without glucose for 3 h before total Quantitative-reverse transcriptase polymerase chain reaction (qRT-PCR). protein was recollected. Two hundred micrograms of RNA per sample was used to generate cDNA using the SuperScript III First-Strand Synthesis System for RT-PCR Kit (Invitrogen, Carlsbad, CA, USA). Quantitative PCR analysis was performed using the SYBR Green PCR Master Mix Kit (Applied Biosystems Inc., Foster City, CA, USA) and Immunoblots. Cells were lysed in RIPA lysis buffer, equal amounts of protein were the ABI Prism 7900HT Fast Real-Time PCR System (Applied Biosystems Inc.). The subjected to SDS-PAGE and transferred to a PVDF membrane. Immunoblots were primers used are shown in Table 1. The measurements were calculated by performed as previously described44,45. employing the ΔΔCt method using SDS 2.3 Software (Applied Biosystems, Inc.). COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio 11 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 Fig. 6 Distinct contributions of LKB1 mutant isoforms to in vivo tumor growth. a In vivo tumor growth upon expression of LKB1WT (n = 5 per group). IVIS imaging of tumor growth at 45 days is shown. H&E staining of representative tumors, immunohistochemistry images from Ki67 staining and vimentin staining, and immunofluorescence images of LKB1, IL8, and GROα,β,γ (CXCL1, CXCL2, and CXCL3) are shown. b In vivo tumor growth upon expression of LKB1R87K (n = 5 per group). The representative images of H&E staining, immunostaining for Ki67 and immunofluorescence staining for LKB1 are shown c. In vivo tumor growth upon expression of the LKB1Y49D mutant (n = 5 per group). Pictures of fresh tumors are shown. Arrows point to hemorrhagic areas. The representative images of immunostaining and immunofluorescence for ERG and Ki67, and LKB1, IL8 and GROα,β,γ (CXCL1, CXCL2, and CXCL3), respectively, are shown. d In vivo tumor growth upon expression of LKB1G135R (n = 4 per group). The representative images of H&E staining, immunostaining for Ki67 and vimentin, and immunofluorescence staining for LKB1 are shown. e In vivo tumor growth upon expression of LKB1D194Y (n = 5 per group). The representative images of H&E staining, immunostaining for Ki67 and immunofluorescence staining for LKB1 are shown. The p-value was calculated by a two-tailed binomial test. Bars represent 500 µm. (0 h.). For colony-formation assays, 300 cells per well were seeded in triplicate for Table 1 List of primers used for qPCR. every condition tested. Cells were maintained at 37 °C in 5% CO2 for ~20 days. Then, the cells were washed twice with PBS and fixed for 10 min with 4% paraf- Gene FWR/REV Sequence-5′-3′ ormaldehyde in PBS at RT. Cells were dyed for 10 min with a crystal violet staining solution. After distaining, colonies were photographed and counted. Colony STK11 FWR TCTACACTCAGGACTTCACG quantification was performed manually and by using ImageJ software. At least two REV GTTCATACACACGGCCTT biological replicates with three technical replicates each were performed for every PLAUR FWR CATGAATCAATGTCTGGTAGCC cell line. REV GCCTCTTACCATATAGCTTTG RUSC2 FWR ATTTCCATTGACCTGCTTCAG Cell cycle analysis. Cells were grown in complete media and treated for 48 h with REV CTTGTGCCAAAATGAGCC doxycycline 1 µg/mL to determine LKB1-isoform expression. Time point treat- AP1S3 FWR CTGTGCAATAGAAAATCAGGAC ments were performed in triplicate. Then, the medium and cells were collected, and REV AGCTCACAGACATTTCCA after centrifugation, the cells were fixed and stained with the Cell Cycle Analysis GLIPR1 FWR GCCCCAATAATGACAAGTG Guava reagent (Guava Technologies, Hayward, CA, USA). Samples were analyzed REV TTTGACTTGGTCTCGCTG with the Guava PCA cytometer (Guava Technologies Hayward, CA, USA). TGFB1I1 FWR GCCACTCAGTTCAACATCAC REV TCCTCCTTCTGCTCTCCT Metabolic profiling. Mitochondrial function and glycolytic function were assessed SEMA7A FWR TCTCAATGTGTCCCGTGT using Seahorse technology (Seahorse XF Cell Mito stress kit, Agilent Technologies, REV TGAACTTTCCCCACCCTG Wilmington, DE, USA). Briefly A459 cells harboring the different STK11 isoform IL8 FWR CTGAGAGTGATTGAGAGTGGAC constructs were cultured on Seahorse XF-24 plates at a density of 75,000 cells per well. Cells were grown in the presence of doxycycline for 4 days before experiment. REV TTTTCCTTGGGGTCCAGA On the day of metabolic flux analysis, cells were changed to unbuffered DMEM CXCL1 FWR TCAATCCTGCATCCCCCATA (DMEM base medium supplemented with 10 mM glucose, 1 mM sodium pyruvate, REV TTCCTCCTCCCTTCTGGT 2 mM Glutamine, pH 7.4) and incubated at 37 °C in a non-CO2 incubator for 1 h. TDO2 FWR ACTTCTGGGGAAAGCTTG All medium and injection reagents were adjusted to pH 7.4 on the day of assay. REV GTTCCTCTTTTTCTTCAGACTC Four baseline measurements of OCR and ECAR were taken before sequential RARRES3 FWR TGAGCACTTTGTCACCCA injection of mitochondrial inhibitors. Three readings were taken after each addition REV CACACCGACTTCAACCTT of mitochondrial inhibitor before injection of the subsequent inhibitors. The SOD2 FWR GACAAACCTCAGCCCTAA mitochondrial inhibitors used were oligomycin (1 µM), FCCP (0.5 µM), and rotenone (0.5 µM). OCR and ECAR were automatically calculated and recorded by REV CAGCTTCTCCTTAAACTTGTC the Seahorse XF-24 software. After the assays, plates were saved and protein CDH1 FWR AGAACGCATTGCCACATACACTC readings were measured for each well to confirm equal cell numbers per well. The REV CATTCTGATCGGTTACCGTGATC percentage of change compared with the basal rates was calculated as the value of CDCP1 FWR GACGGTGTCCTTTATACC change divided by the average value of baseline readings. REV GACCTTGCTTTTTGTGTCAG DUSP1 FWR AGGACAACCACAAGGCAG Motility assay. Cells were seeded at low confluence in 24-well plates in duplicate. REV TGGACAAACACCCTTCCT After 24 h, cells were induced with doxycycline and placed in the IncuCyte Live PLAUR FWR CATGAATCAATGTCTGGTAGCC Cell Analysis Platform (Essen, Ann Arbor, MI, USA). Pictures were captured at 30 REV GCCTCTTACCATATAGCTTTG min intervals from five separate 950 × 760 µm2 regions per well using a ×20 CYP1B1 FWR TGCTCCTCCTCTTCACCA objective for 5 days. Motility rates were measured for 30 to 50 GFP-positive cells REV GGTCACCCATACAAGGCA individually with ImageJ software and graphed. Kinase assay. A specific LKB1 kinase assay was performed as previously described We applied geNorm algorithms to select TATA-binding protein (TBP) and human in Lizcano et al.15. Briefly, different combinations of His-tagged LKB1 isoforms and peptidyl-prolyl cis-trans isomerase A (HPPIA). to select TATA-binding protein FLAG-tagged STRADα were expressed in 293 cells and the complexes purified on (TBP) and peptidylprolyl isomerase A (cyclophilin A, PPIA) as the most stable cobalt binding resin. Protein complexes were washed twice with 1 ml of lysis buffer reference transcripts. The geometric means of the expression values for both (50 mM Tris/HCl pH 7.5, 1 mM EGTA, 1 mM EDTA, 1% (w/v), Triton-X 100, 1 housekeeping genes were used to normalize the expression and to calculate the mM sodium orthovanadate, 50 mM sodium fluoride, 5 mM sodium pyropho- normalized SD of all transcripts analyzed. Relative expression levels were calculated sphate, 0.27 M sucrose, 0.1% (v/v) 2-mercaptoethanol and ‘complete’ proteinase after normalization. Data were represented as mean ± SD of triplicates from three inhibitor cocktail (one tablet/50 ml) containing 0.5 M NaCl, and twice with 1 ml of independent experiments (biological replicates). Buffer A (50 mM Tris/HCl pH 7.5, 0.1 mM EGTA, and 0.1% (v/v) 2-mercap- toethanol) and eluted from column with elution buffer containing 50 mM sodium phosphate, 300 mM NaCl and 150 mM imidazole. Phosphotransferase activity Gene expression analysis. For microarray analyses, we used a genome-wide towards the NUAKtide peptide (SNLYHQGKFLQTFCGSPLYRRR residues Human Gene 1.0 ST Array (Affymetrix, Santa Clara, CA, USA). Genes were 241–260 of human NUAK2 with three additional Arg residues added to the C- considered differentially expressed in A549-LKB1-WT cells if the fold change was terminal to enable binding to P81 paper), was then measured in a total assay >1.2 and the p was <0.05 (noninduced cells versus cells treated with doxycycline for volume of 50 μl consisting of 50 mM Tris/HCl pH 7.5, 0.1 mM 48 h) using a two-tailed one-way ANOVA test. EGTA, 0.1% (by vol) 2-mercaptoethanol, 10 mM magnesium acetate, 0.1 mM [γ32P]ATP (~200 cpm/pmol), and 200 μM NUAKtide peptide. The assays were Proliferation and colony-formation assays. We seeded 0.16 × 106 cells per well in carried out at 30 °C with continuous shaking, to keep the immunoprecipitates in 6-well plates. The data were collected in triplicate at 24, 48, 72, and 96 h after the suspension, and were terminated after 10 min by applying 40 μl of the reaction initial seeding. For every time point, viable cells were counted (using a Neubauer mixture onto p81 membranes. The p81 membranes were washed in phosphoric chamber). Data are presented as the fold change with respect to the first measure acid, and the incorporated radioactivity was measured by scintillation counting. 12 COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 ARTICLE Secretome proteomics and statistical analysis. The secretomes were prepared as Data availability previously described46. In brief, 4 × 106 cells were seeded in 150 cc tissue culture The datasets generated during and/or analyzed during the current study are available in: plates and allowed to grow for 48 in the presence or absence of doxycycline 1 µg/ ArrayExpress public repository with accession number E-MTAB-8863 (https://www.ebi. ml. After that, media was aspirated and cells were washed twice with PBS and then ac.uk/arrayexpress/experiments/E-MTAB-8863/) and the proteomeXchange public three times with serum-free media. Then, cells at 60–70% confluency were database PRIDE (https://www.ebi.ac.uk/pride/archive) with accession number: maintained in serum-free media for 24 h before the collection of the conditioned PXD018041. media (secretome). Secretomes were spun down, and filtered through a 0.22-μm pore filter. Then, secretomes were concentrated using a 10,000 MWCO Millipore Amicon Ultra filter (Millipore) until a final volume of 50 μL was reached. The Received: 2 November 2019; Accepted: 19 June 2020; protein concentration was determined using the Pierce BCA Protein Assay kit (Thermo Scientific). All samples were digested with trypsin in-solution prior to analysis by liquid chromatography−mass spectrometry (LC − MS) as previously described (1). Tryptic digests were analyzed by shotgun proteomics using an LTQ Velos-Orbitrap mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). The RAW files of each MS run were processed using Proteome Discoverer References (Thermo Fisher Scientific), and MS/MS spectra were searched against the human 1. Hemminki, A. et al. Localization of a susceptibility locus for Peutz-Jeghers database of Swiss-Prot using the MASCOT (Matrix Science, London, U.K) algo- syndrome to 19p using comparative genomic hybridization and targeted rithm. The results files generated from MASCOT (.DAT files) were then loaded linkage analysis. Nat. Genet. 15, 87–90 (1997). into Scaffold (Proteome Software, Portland, OR), resulting in a nonredundant list 2. Guldberg, P. et al. Somatic mutation of the Peutz-Jeghers syndrome gene, of identified proteins per sample achieving a protein false discovery rate (FDR) LKB1/STK11, in malignant melanoma. Oncogene 18, 1777–1780 (1999). under 1.0%, as estimated by a search against a decoy database. 3. McCabe, M. T., Powell, D. R., Zhou, W. & Vertino, P. M. Homozygous Relative spectral counting-based protein quantification analysis was performed deletion of the STK11/LKB1 locus and the generation of novel fusion on the different samples analyzed using Scaffold. Files containing all spectral counts for each sample and its replicates were generated and then exported to R transcripts in cervical cancer cells. Cancer Genet. Cytogenet. 197, 130–141 software for normalization and statistical analysis47. All statistical computations (2010). were done using the open-source statistical package R. The data were assembled in 4. Sanchez-Cespedes, M. et al. Inactivation of LKB1/STK11 is a common event in a matrix of spectral counts, where columns represent the biological conditions and adenocarcinomas of the lung. Cancer Res. 62, 3659–3662 (2002). rows represent the identified proteins. An unsupervised exploratory data analysis 5. Tiainen, M., Vaahtomeri, K., Ylikorkala, A. & Makela, T. P. Growth arrest by (EDA) by means of principal components analysis and hierarchical clustering of the LKB1 tumor suppressor: induction of p21(WAF1/CIP1). Hum. Mol. the samples on the SpC matrix was first performed. Then, the GLM model based Genet. 11, 1497–1504 (2002). on the Poisson distribution was used as a significance test47. Finally, the Benjamini- 6. Scott, K. D., Nath-Sain, S., Agnew, M. D. & Marignani, P. A. LKB1 Hochberg multiple test correction was used to adjust the p-values with control on catalytically deficient mutants enhance cyclin D1 expression. Cancer Res. 67, the false discovery rate (FDR). 5622–5627 (2007). 7. Spicer, J. & Ashworth, A. LKB1 kinase: master and commander of metabolism and polarity. Curr. Biol. 14, R383–385 (2004). Animal study. Animal experiments were conducted and designed according to 8. Esteve-Puig, R., Canals, F., Colome, N., Merlino, G. & Recio, J. A. Uncoupling protocols approved by the Institutional Animal Care and Use Committee of Vall d of the LKB1-AMPKalpha energy sensor pathway by growth factors and ´Hebron Institute of Research. Nude mice (athymic nu/nu, female 4-6weeks old, oncogenic BRAF. PLoS ONE 4, e4771 (2009). Harlan Laboratories) were used for xenograft studies. 9. Londesborough, A. et al. LKB1 in endothelial cells is required for angiogenesis and TGFbeta-mediated vascular smooth muscle cell recruitment. Development 135, 2331–2338 (2008). In vivo tumor growth experiments. For xenograft animal models, 5 × 105 10. Bonanno L, et al. LKB1 and tumor metabolism: the interplay of immune and luciferase-expressing A549-LKB1 cells were subcutaneously implanted into 8- angiogenic microenvironment in lung cancer. Int. J. Mol. Sci. 20, 1874 (2019). week-old female athymic nude-Foxn1nu mice (n = 5 per group) (Envigo, India- 11. Williams, T. & Brenman, J. E. LKB1 and AMPK in cell polarity and division. napolis, IN, USA). Tumors were measured with a digital Vernier caliper, and the Trends Cell Biol. 18, 193–198 (2008). mice were weighed twice a week. Tumor volume was calculated as D × d2/2, where 12. Esteve-Puig, R. et al. A mouse model uncovers LKB1 as an UVB-induced D was the major diameter and d was the minor diameter. Bioluminescence imaging DNA damage sensor mediating CDKN1A (p21WAF1/CIP1) degradation. (IVIS Spectrum In Vivo Imaging System), PerkinElmer Life Science, Waltham, PLoS Genet. 10, e1004721 (2014). MA, USA) was performed at the end of the experiment for metastasis screening, 13. Dorfman, J. & Macara, I. G. STRADalpha regulates LKB1 localization by both in vivo and ex vivo (viscera). Once the treatment was started, the animal blocking access to importin-alpha, and by association with Crm1 and weights and tumor sizes were monitored every 2 days. Representative tumor pic- exportin-7. Mol. Biol. Cell 19, 1614–1626 (2008). tures were taken and tumor samples were obtained for further analysis. 14. Boudeau, J. et al. Analysis of the LKB1-STRAD-MO25 complex. J. Cell Sci. 117, 6365–6375 (2004). 15. Lizcano, J. M. et al. LKB1 is a master kinase that activates 13 kinases of the Immunohistochemistry and immunofluorescence. Formalin-fixed paraffin- AMPK subfamily, including MARK/PAR-1. EMBO J. 23, 833–843 (2004). embedded tumor samples were subjected to immunocytochemistry according to 16. Alessi, D. R., Sakamoto, K. & Bayascas, J. R. LKB1-dependent signaling the manufacturer’s antibody protocol. Samples were developed either by using pathways. Annu Rev. Biochem. 75, 137–163 (2006). either secondary antibodies linked to horseradish peroxidase or secondary anti- 17. Mehenni, H. et al. LKB1 interacts with and phosphorylates PTEN: a functional bodies linked to a fluorophore. Immunostaining was performed on 4 µm sections link between two proteins involved in cancer predisposing syndromes. Hum. from formalin-fixed paraffin-embedded tissues. Staining was performed either Mol. Genet. 14, 2209–2219 (2005). manually or on the automated immunostainer Beckmarck XT (Ventana Medical 18. Shackelford, D. B. & Shaw, R. J. The LKB1-AMPK pathway: metabolism and Systems, Roche, Tucson, AZ, USA). Antibodies were visualized by the Ultra- growth control in tumour suppression. Nat. Rev. Cancer 9, 563–575 (2009). ViewTM Universal DAB Detection Kit (Ventana Medical Systems). For samples 19. Chavez, J. A., Roach, W. G., Keller, S. R., Lane, W. S. & Lienhard, G. E. processed manualy antigen retrieval was performed using target retrieval solution pH 6.0 (Dako,Agilent, Santa Clara CA, USA)) Samples were scanned (panoramic Inhibition of GLUT4 translocation by Tbc1d1, a Rab GTPase-activating slide digital scanner) and evaluated by two independent pathologists (using protein abundant in skeletal muscle, is partially relieved by AMP-activated 3DHistech software). protein kinase activation. J. Biol. Chem. 283, 9187–9195 (2008). 20. Tamargo-Gomez I, Marino G. AMPK: regulation of metabolic dynamics in the context of autophagy. Int. J. Mol. Sci. 19, 3812 (2018). Statistics and reproducibility. Statistical tests used are reported in the figure 21. Leprivier, G. et al. The eEF2 kinase confers resistance to nutrient deprivation legends. In summary, statistical analyses were performed in GraphPad Prism 6 by blocking translation elongation. Cell 153, 1064–1079 (2013). (GraphPad Software Inc.) using a two-tailed Student’s t test to compare differences 22. Zhuang, Z. G., Di, G. H., Shen, Z. Z., Ding, J. & Shao, Z. M. Enhanced between two groups or one-way ANOVA with multiple-comparisons tests to expression of LKB1 in breast cancer cells attenuates angiogenesis, invasion, compare 3 or more groups. Statistical tests used for the analyses of transcriptomes and metastatic potential. Mol. Cancer Res. 4, 843–849 (2006). (microarrays and gene set enrichment analysis) were performed in Partek Genomic 23. Zulato, E. et al. Involvement of NADPH oxidase 1 in liver kinase B1-mediated Suite software (Partek Inc.). Number of biological replicates was ≥3 and is indi- effects on tumor angiogenesis and growth. Front. Oncol. 8, 195 (2018). cated in the figure legends. 24. Xia, C. et al. Reactive oxygen species regulate angiogenesis and tumor growth through vascular endothelial growth factor. Cancer Res. 67, 10823–10830 (2007). Reporting summary. Further information on research design is available in 25. Okon, I. S. et al. Protein kinase LKB1 promotes RAB7-mediated neuropilin-1 the Nature Research Reporting Summary linked to this article. degradation to inhibit angiogenesis. J. Clin. Invest. 124, 4590–4602 (2014). COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio 13 ARTICLE COMMUNICATIONS BIOLOGY | https://doi.org/10.1038/s42003-020-1092-0 26. Zhang, S. et al. The tumor suppressor LKB1 regulates lung cancer cell polarity 46. Villarreal, L. et al. Unconventional secretion is a major contributor of cancer by mediating cdc42 recruitment and activity. Cancer Res. 68, 740–748 (2008). cell line secretomes. Mol. Cell. Proteom. 12, 1046–1060 (2013). 27. Roy, B. C. et al. Involvement of LKB1 in epithelial-mesenchymal transition 47. Gregori, J. et al. Batch effects correction improves the sensitivity of (EMT) of human lung cancer cells. Lung Cancer 70, 136–145 (2010). significance tests in spectral counting-based comparative discovery 28. Lin, R. et al. 6-Phosphogluconate dehydrogenase links oxidative PPP, proteomics. J. Proteom. 75, 3938–3951 (2012). lipogenesis and tumour growth by inhibiting LKB1-AMPK signalling. Nat. 48. Bader, G. D. & Hogue, C. W. An automated method for finding molecular Cell Biol. 17, 1484–1496 (2015). complexes in large protein interaction networks. BMC Bioinform. 4, 2 (2003). 29. Wang, Y. S. et al. LKB1 is a DNA damage response protein that regulates cellular sensitivity to PARP inhibitors. Oncotarget 7, 73389–73401 (2016). 30. Konen, J. et al. LKB1 kinase-dependent and -independent defects disrupt Acknowledgements polarity and adhesion signaling to drive collagen remodeling during invasion. This work was supported by funds from the Spanish Health Ministry (Fondo de Mol. Biol. Cell 27, 1069–1084 (2016). Investigaciones Sanitarias-FIS) PI1400375-Fondos FEDER, PI17/00043-Fondos FEDER, 31. Kline, E. R., Shupe, J., Gilbert-Ross, M., Zhou, W. & Marcus, A. I. LKB1 Euronanomed2-ISCIII (AC16/00019)-Fondos FEDER, Asociación Española Contra el represses focal adhesion kinase (FAK) signaling via a FAK-LKB1 complex to Cancer (AECC-GCB15152978SOEN) supported PGM, KM., Ramón Areces Foundation regulate FAK site maturation and directional persistence. J. Biol. Chem. 288, supported KM and research. 17663–17674 (2013). 32. Wang, S. et al. LKB1 and YAP phosphorylation play important roles in Author contributions Celastrol-induced beta-catenin degradation in colorectal cancer. Ther. Adv. Conceptualization: J.A.R., P.G.M., and E.G.S. Investigation: P.G.M., E.G.S., S.G.O., K.M., Med. Oncol. 11, 1758835919843736 (2019). N.N.G., A.C.R., R.S., J.H.L., B.F., J.V., O.M., S.E.G., J.M.L. and J.A.R. Resources: E.M.C. 33. Kaufman, J. M. et al. LKB1 Loss induces characteristic patterns of gene and V.G.P. Methodology: J.G. and R.G. Formal analysis: J.A.R., J.M.L., F.C., J.V. Writing- expression in human tumors associated with NRF2 activation and attenuation review and editing: J.A.R., P.G.M., S.G.O., and K.M. Supervision: J.A.R. of PI3K-AKT. J. Thorac. Oncol. 9, 794–804 (2014). 34. Schumacher, V. et al. STK11 genotyping and cancer risk in Peutz-Jeghers syndrome. J. Med. Genet. 42, 428–435 (2005). Competing interests 35. Lopus, M., Paul, D. M. & Rajasekaran, R. Unraveling the deleterious effects of The authors declare no competing interests. cancer-driven STK11 mutants through conformational sampling approach. Cancer Inf. 15, 35–44 (2016). 36. Ossipova, O., Bardeesy, N., DePinho, R. A. & Green, J. B. LKB1 (XEEK1) Additional information Supplementary information is available for this paper at https://doi.org/10.1038/s42003- regulates Wnt signalling in vertebrate development. Nat. Cell Biol. 5, 889–894 020-1092-0. (2003). 37. Hetey, S. E., Lalonde, D. P. & Turner, C. E. Tyrosine-phosphorylated Hic-5 Correspondence and requests for materials should be addressed to J.A.R. inhibits epidermal growth factor-induced lamellipodia formation. Exp. Cell Res. 311, 147–156 (2005). Reprints and permission information is available at http://www.nature.com/reprints 38. Benes, C. H., Poulogiannis, G., Cantley, L. C. & Soltoff, S. P. The SRC- associated protein CUB Domain-Containing Protein-1 regulates adhesion and Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in motility. Oncogene 31, 653–663 (2012). published maps and institutional affiliations. 39. Ivankovic, D. et al. Axonal autophagosome maturation defect through failure of ATG9A sorting underpins pathology in AP-4 deficiency syndrome. Autophagy 16, 391–407 (2020). 40. Davies, A. K. et al. AP-4 vesicles contribute to spatial control of autophagy via Open Access This article is licensed under a Creative Commons RUSC-dependent peripheral delivery of ATG9A. Nat. Commun. 9, 3958 Attribution 4.0 International License, which permits use, sharing, (2018). adaptation, distribution and reproduction in any medium or format, as long as you give 41. Mahil, S. K. et al. AP1S3 mutations cause skin autoinflammation by disrupting appropriate credit to the original author(s) and the source, provide a link to the Creative keratinocyte autophagy and up-regulating IL-36 production. J. Invest. Commons license, and indicate if changes were made. The images or other third party Dermatol. 136, 2251–2259 (2016). material in this article are included in the article’s Creative Commons license, unless 42. Park, S. Y & Guo, X. Adaptor protein complexes and intracellular transport. indicated otherwise in a credit line to the material. If material is not included in the Biosci. Rep. 34, 00123 (2014). article’s Creative Commons license and your intended use is not permitted by statutory 43. Skoulidis, F. et al. STK11/LKB1 mutations and PD-1 inhibitor resistance in regulation or exceeds the permitted use, you will need to obtain permission directly from KRAS-mutant lung adenocarcinoma. Cancer Discov. 8, 822–835 (2018). the copyright holder. To view a copy of this license, visit http://creativecommons.org/ 44. Andreu-Perez, P. et al. Methylthioadenosine (MTA) inhibits melanoma cell licenses/by/4.0/. proliferation and in vivo tumor growth. BMC Cancer 10, 265 (2010). 45. Lopez-Fauqued, M. et al. The dual PI3K/mTOR inhibitor PI-103 promotes immunosuppression, in vivo tumor growth and increases survival of © The Author(s) 2020 sorafenib-treated melanoma cells. Int. J. Cancer 126, 1549–1561 (2010). 14 COMMUNICATIONS BIOLOGY | (2020)3:366 | https://doi.org/10.1038/s42003-020-1092-0 | www.nature.com/commsbio