BRIEF DEFINITIVE REPORT Genetic manipulation of LKB1 elicits lethal metastatic prostate cancer Ivana Hermanova1*, Patricia Zúñiga-Garcı́a1*, Alfredo Caro-Maldonado1, Sonia Fernandez-Ruiz1,2, Fernando Salvador2,3, Natalia Martı́n-Martı́n1,2, Amaia Zabala-Letona1,2, Marc Nuñez-Olle3, Verónica Torrano1,2,4, Laura Camacho1,4, Jose M. Lizcano5, Ana Talamillo1, Suzanne Carreira6, Bora Gurel6, Ana R. Cortazar1,2, Marc Guiu3, Jose I. López7, Anabel Martinez-Romero2,8, Ianire Astobiza1,2, Lorea Valcarcel-Jimenez1, Mar Lorente9, Amaia Arruabarrena-Aristorena1, Guillermo Velasco9,10, Antonio Gomez-Muñoz4, Cristian Suárez-Cabrera11,12, Iris Lodewijk11,12, Juana M. Flores13, James D. Sutherland1, Rosa Barrio1, Johann S. de Bono6,14, Jesús M. Paramio2,11,12, Jan Trka15,16, Mariona Graupera2,8, Roger R. Gomis2,3,17, and Arkaitz Carracedo1,2,4,18 Gene dosage is a key defining factor to understand cancer pathogenesis and progression, which requires the development of experimental models that aid better deconstruction of the disease. Here, we model an aggressive form of prostate cancer and show the unconventional association of LKB1 dosage to prostate tumorigenesis. Whereas loss of Lkb1 alone in the murine prostate epithelium was inconsequential for tumorigenesis, its combination with an oncogenic insult, illustrated by Pten heterozygosity, elicited lethal metastatic prostate cancer. Despite the low frequency of LKB1 deletion in patients, this event was significantly enriched in lung metastasis. Modeling the role of LKB1 in cellular systems revealed that the residual activity retained in a reported kinase-dead form, LKB1K78I, was sufficient to hamper tumor aggressiveness and metastatic dissemination. Our data suggest that prostate cells can function normally with low activity of LKB1, whereas its complete absence influences prostate cancer pathogenesis and dissemination. Introduction Prostate cancer (PCa) is primarily diagnosed as a localized disease phenomenon in experimental models has been limited (Arriaga that is considered curable with first-line therapies. However, two and Abate-Shen, 2019). major limitations exist in its clinical management (Chang et al., LKB1 (the protein encoded by the gene STK11) is a master kinase 2014). First, a considerable proportion of men will relapse after with a prominent role as a tumor suppressor in various tissue such treatments. These patients will frequently fail to respond to types. This gene is mutated and deleted in lung, cervix, uterine, subsequent lines of treatment (androgen pathway inhibitors and skin, and pancreatic cancers (Momcilovic and Shackelford, 2015). chemotherapy) and succumb to metastatic PCa. Second, a fraction The study of LKB1 in cancer has been boosted by the development of PCa patients will exhibit metastatic disease at diagnosis. of mouse models with varying doses of the tumor suppressor, Whereas the incidence of this phenomenon is estimated as <10% in which result in a reduction of its levels of 50–100% (for hetero- populations heavily prescreened for prostate-specific antigen, it is zygous and homozygous knockout mice, respectively; Pearson thought to account for ≤50% of PCa mortality (James et al., 2015). et al., 2008) or 25–75% (through the combination of hypomor- The clinical management of lethal metastatic disease is the ultimate phic alleles with WT or knockout alleles, respectively; Huang et al., challenge in PCa. However, the success in recapitulating this 2008). ............................................................................................................................................................................. 1 Center for Cooperative Research in Biosciences, Basque Research and Technology Alliance (BRTA), Derio, Spain; 2CIBERONC (Centro de Investigación Biomédica en Red de Cáncer), Madrid, Spain; 3Cancer Science Program, Institute for Research in Biomedicine, The Barcelona Institute of Science and Technology, Barcelona, Spain; 4Biochemistry and Molecular Biology Department, University of the Basque Country, Bilbao, Spain; 5Protein Kinases and Signal Transduction Laboratory, Institut de Neurociències and Departament de Bioquı́mica i Biologia Molecular, Universitat Autònoma de Barcelona, Bellaterra, Barcelona, Spain; 6The Institute of Cancer Research, London, UK; 7Department of Pathology, Cruces University Hospital, Biocruces Institute, University of the Basque Country, Barakaldo, Spain; 8Vascular Signalling Laboratory, Program Against Cancer Therapeutic Resistance (ProCURE), Institut d’Investigació Biomèdica de Bellvitge, Barcelona, Spain; 9Department of Biochemistry and Molecular Biology, School of Biology, Complutense University, Madrid, Spain; 10Instituto de Investigaciones Sanitarias San Carlos, Madrid, Spain; 11Grupo de Oncologı́a Celular y Molecular, Hospital Universitario 12 de Octubre, Madrid, Spain; 12Unidad de Oncologı́a Molecular, Centro de Investigaciones Energéticas, Medioambientales y Tecnológicas, Madrid, Spain; 13Department of Animal Medicine and Surgery, School of Veterinary Medicine, Complutense University of Madrid, Madrid, Spain; 14The Royal Marsden National Health Service Foundation Trust, London, UK; 15Childhood Leukaemia Investigation, Prague, Czech Republic; 16Department of Paediatric Haematology/ Oncology, Second Faculty of Medicine, Charles University and University Hospital Motol, Prague, Czech Republic; 17Institució Catalana de Recerca i Estudis Avançats, Barcelona, Spain; 18Ikerbasque, Basque Foundation for Science, Bilbao, Spain. *I. Hermanova and P. Zúñiga-Garcı́a contributed equally to this paper; Correspondence to Arkaitz Carracedo: acarracedo@cicbiogune.es. © 2020 Hermanova et al. This article is distributed under the terms of an Attribution–Noncommercial–Share Alike–No Mirror Sites license for the first six months after the publication date (see http://www.rupress.org/terms/). After six months it is available under a Creative Commons License (Attribution–Noncommercial–Share Alike 4.0 International license, as described at https://creativecommons.org/licenses/by-nc-sa/4.0/). Rockefeller University Press https://doi.org/10.1084/jem.20191787 1 of 14 J. Exp. Med. 2020 Vol. 217 No. 6 e20191787 Tumor suppressors are clinically and experimentally defined tumor and exhibited positive staining for androgen receptor (AR), based on gene dosage. Knudson (1971) proposed that genes with supporting the notion that the metastases were of prostatic origin such functions would follow the two-hit hypothesis, according (Fig. S1 F). No lesions were observed in heart, liver, spleen, kidney, to which cancer would emerge upon the loss of the second allele or bone (Table S1). in cells with carrying a mutated gene. However, we know that PCa disseminates predominantly to bone. However, dissemi- the relationship between gene dosage and tumor-suppressive nation to lungs is well recognized to occur in 10–20% of metastatic activity is more complex, which led to the introduction of con- disease (Arriaga and Abate-Shen, 2019). We took advantage of cepts such as haploinsufficient tumor suppressors (Berger et al., publicly available PCa transcriptomics and genomics studies to 2011). However, further experimental effort is required for a ascertain the alteration of LKB1 in primary and metastatic lesions. comprehensive classification of tumor suppressors, based on We started by evaluating point mutations in a series of PCa evidence produced through the analysis of mouse models with datasets (Taylor et al., 2010, Cancer Genome Atlas Research hypomorphic alleles (Alimonti et al., 2010). Network, 2015; Kumar et al., 2016; Abida et al., 2017, 2019; In this study, we generate an unprecedented model of lethal Fraser et al., 2017; Armenia et al., 2018; Gerhauser et al., 2018) metastatic PCa by introducing perturbations in Lkb1 in a pre- using cBioPortal (Cerami et al., 2012; Gao et al., 2013). The results cancerous setting and demonstrate the high competence of this revealed a lack of this type of alteration in PCa compared with a kinase to suppress biological features of PCa. compendium of datasets from lung adenocarcinoma (Cancer Genome Atlas Research Network, 2013; Imielinski et al., 2012; Jordan et al., 2017), a tumor type with frequent alterations in LKB1 Results and discussion (Fig. S1 G). We next focused on genomic aberrations using a large- To ascertain the role of LKB1 in PCa, we engineered a series of scale study of PCa (n = 1,013), in which primary tumor biopsies Lkb1 prostate-specific mutant mice based on a conditional mouse and metastatic specimens where sequenced and tropism was an- model (Fig. S1, A–D; see Materials and methods; Nardella et al., notated (Armenia et al., 2018). In this dataset, LKB1 heterozygosity 2010; Pearson et al., 2008). The survival of mice with prostate- was observed in 24.6% of cases (249 out of 1,013), whereas com- specific Lkb1 (Lkb1pc−/−) deficiency was indistinguishable from plete loss was restricted to 3.36% of specimens (34 of 1,013; Table that of control counterparts (Fig. 1 A). Although Lkb1 deletion led S2), in turn suggesting that LKB1 is not a commonly aberrant to increased tissue mass (Fig. S1 E), it did not result in any his- tumor suppressor in the prostate. However, we noted that meta- tological alteration (Fig. 1 B). These data are in contrast to results static tumors had higher rates of LKB1 genomic aberrations, with obtained using other Cre recombinases such as p450 CYP1A1- 4.8% of homozygous deletions (16 of 333) compared with 2.65% in driven Cre, which show a broader tissue distribution in the primary tumors (18 of 680), and 29.1% of heterozygous loss (97 of urogenital track (Pearson et al., 2008). The discrepancies might 333) compared with 22.35% in primary tumors (152 of 680; χ2 P = be associated with the impact of the non-prostate epithelial de- 0.007). We decided to analyze whether perturbations in this gene letion of Lkb1 (Pearson et al., 2008). Also, the original description could be associated with any specific metastatic behavior. 66% of of this Cre driver reported expression in multiple tissues and cell metastatic cases presented site annotation (220 of 333). Metastatic types beyond the prostate (Ireland et al., 2004). Our results using cases within this subset with one or two intact alleles of LKB1 a well-established prostate epithelium–restricted Cre driver exhibited similar distribution of metastasis, with a predominance support the notion that Lkb1 loss is not an initiating event in of bone and lymph node dissemination and negligible lung me- prostate tumorigenesis. tastasis (Fig. 1 G and Table S3). Among the limited number of Next, we evaluated whether deletion of Lkb1 would contribute metastatic cases with homozygous deletion of LKB1, 50% belonged to PCa pathogenesis in the presence of an oncogenic event. To this to lung metastatic specimens (2 of 4; Fig. 1 G). The incidence of end, we intercrossed prostate-conditional Lkb1 knockout mice LKB1 homozygous deletion was 28.6% (2 of 7, Fisher’s exact test P < with a well-established driver of murine PCa, Pten knockout mice 0.01) in lung metastasis compared with 0.9% (2 of 213) of all other (Fig. S1, C and D). Whereas Ptenpc−/− mice develop nonmetastatic metastatic specimens, whereas we found no enrichment in bone PCa, the phenotype of heterozygous mice is limited to prostate metastasis or lymph node dissemination (Fig. 1 H, Fig. S1 H, and intraepithelial neoplasia (PIN; Nardella et al., 2010). Surprisingly, Table S3). the deletion of Lkb1 resulted in a remarkable reduction in survival As a complementary approach, we analyzed an independent when combined with one allele deletion of Pten (Fig. 1 C). Since the cohort (Kumar et al., 2016) in which we compared the expres- combination of Pten heterozygosity with Lkb1 loss (hereafter re- sion of LKB1 in the two predominant metastatic sites of interest, ferred to as PTLK) resulted in early lethality, we performed his- bone versus lung. In agreement with our previous results, lung tological characterization of all genotypes at 10 mo of age. PTLK metastatic lesions exhibited a significant reduction in the ex- mice exhibited PCa at full penetrance, whereas none of the other pression of the kinase (Fig. 1 I). genotypes (Ptenpc+/− Lkb1pc+/+, Ptenpc+/− Lkb1pc+/−) presented lesions Taking advantage of the aforementioned large genomic da- beyond PIN at the same age (Fig. 1 D). Moreover, the macroscopic taset (Armenia et al., 2018), we explored the genetic events that and histopathological analysis of distant organs unexpectedly re- could co-occur with LKB1 in metastatic specimens. To this end, vealed that PTLK mice developed lung metastasis with an inci- we extracted from this dataset the most frequent copy number dence >80% (justifying their reduced survival rate) and inguinal alterations (Table S4) and evaluated their association to LKB1 lymph node colonization to a lower extent (Fig. 1, E and F). Im- deletion in metastasis. We noted that none of the classic PCa portantly, lung lesions were histologically identical to the prostatic oncogenic mutations were cooccurring with LKB1 deletion Hermanova et al. Journal of Experimental Medicine 2 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Figure 1. Combined Lkb1 loss with Pten heterozygosity in murine prostate epithelia results in PCa progression and dissemination. (A) Cumulative survival curve (Kaplan–Meier survival plot) of Lkb1pc+/+ (n = 27) and Lkb1pc−/− (n = 27) mice. The cumulative survival rate was plotted against age in weeks. (B) Histopathological characterization of the prostate in the indicated genotypes Lkb1pc+/+ (n = 7) and Lkb1pc−/− (n = 19) mice (left). LGPIN, low-grade PIN. Representative image of H&E staining of normal prostate in Lkb1pc+/+ mice (right, upper panel) and Lkb1pc−/− mice (right, lower panel). (C) Cumulative survival curve (Kaplan–Meier survival plot) of Ptenpc+/− Lkb1pc+/+ (n = 9), Ptenpc+/− Lkb1pc+/− (n = 23), and Ptenpc+/− Lkb1pc−/− (n = 21) mice. The cumulative survival rate was plotted against age in weeks. (D) Histopathological characterization of the prostate in the indicated genotypes: Ptenpc+/− Lkb1pc+/+ (n = 4), Ptenpc+/− Lkb1pc+/− (n = 5), and Ptenpc+/− Lkb1pc−/− (n = 18) mice (left). Right panels depict representative H&E micrographs of the histological features observed in each genotype. A macroscopic image of a representative tumor in Ptenpc+/− Lkb1pc−/− mice is included. HGPIN, high-grade PIN. (E) Quantification of the frequency of lymph node colonization in Ptenpc+/− Lkb1pc+/+ (n = 8), Ptenpc+/− Lkb1pc+/− (n = 9), and Ptenpc+/− Lkb1pc−/− (n = 13) mice (left). Representative image of H&E staining of a normal lymph node in Ptenpc+/− Lkb1pc+/+ (right, upper panel) and a lymph node metastasis in Ptenpc+/− Lkb1pc−/− mice (right, lower panel). Red arrows depict colonizing prostate tumor cells. (F) Quantification of the frequency of metastatic lesions in lung of Ptenpc+/− Lkb1pc+/+ (n = 4), Ptenpc+/− Lkb1pc+/− (n = 7), and Ptenpc+/− Lkb1pc−/− (n = 14) mice (left). Representative macroscopic image (right, upper panel) and representative histopathological image (right, lower panel) of metastatic lesions in lung in Ptenpc+/− Lkb1pc−/− mice. (G) Incidence of metastatic lesions in bone, lymph nodes, lung, liver, and other tissues in patients with metastatic PCa with indicated LKB1 genomic status. Numbers below the graph indicate total number of analyzed metastatic samples. Dataset of origin: Armenia et al. (2018). (H) Incidence of LKB1 genomic alterations in metastatic specimens of patients with metastatic PCa (Armenia et al., 2018). Cases without LKB1 homozygous deletion (LKB1 not homdel); cases with LKB1 homozygous deletion (homdel). Not bone, metastasis identified in all analyzed tissues except for bone; not lung, metastasis identified in all analyzed tissues except for lung. Numbers below the graph indicate total number of analyzed metastatic samples. (I) Gene expression levels of LKB1 in bone and lung metastatic samples in patients with PCa (Kumar et al., 2016). pc, prostate-specific allelic changes; +, WT allele; −, deleted allele; n.s., not significant; **, P < 0.01; ***, P < 0.001. Statistic tests: log-rank (Mantel–Cox) test (A and C), Fisher exact test (H), two-tailed unpaired Student’s t test (I). Hermanova et al. Journal of Experimental Medicine 3 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 (Table S5). These data led us to speculate that the aggressive The vast majority of available PCa high-throughput molecular features enabled upon LKB1 loss are not associated with a single studies focus on adenocarcinoma lesions, since this histological oncogenic hit, but rather to the process of transformation itself. cancer type is predominant. However, according to PCa guidelines, In this regard, we envision the perturbation in Pten used in the squamous cell carcinoma of prostate presents an incidence of mouse model as an illustrative oncogenic signal that would un- 0.5–1% and exhibits highly aggressive characteristics (Brunnhoelzl leash prostate cell proliferation and enable the emergence of Lkb1 and Wang, 2018; Munoz et al., 2007). We searched for LKB1- loss–dependent phenotypes. deficient PCa specimens that exhibited squamous pathological LKB1 phosphorylates 13 different effector kinases (Lizcano features. Interestingly, we identified three such cases (Fig. 2 B). et al., 2004). To ascertain which of the potential targets of LKB1 Two of the three cases exhibited positivity for 34βE12, a basal cell could account for its prostate tumor–suppressive activity, we marker, whereas one case showed weak positivity. Interestingly, asked whether any of these targets would phenocopy the pattern two of these three specimens harbored combined LKB1 and PTEN of genomic loss of the master regulator using the same patient alterations, analogous to our murine model (Fig. 2 B). dataset (Armenia et al., 2018). When looking at the loci of the To approach the mechanism by which LKB1 suppresses PCa, aforementioned kinases, we found genomic aberrations in only we characterized a panel of PCa cells. In line with previous re- nine of them. Among these genes, we identified the strongest ports, DU145 cells lack LKB1 protein expression, due to a frame- pattern of lung dissemination in SIK3 (Fig. S1 I). Similar to LKB1, shift homozygous mutation (Ikediobi et al., 2006; Fig. 3 A). We SIK3 homozygous deletion was present in 28.6% (2 of 7, Fisher’s took advantage of this cell line and reintroduced LKB1 at expres- exact test P < 0.01) of all lung metastatic cases (Fig. S1 J and Table sion levels comparable to other prostate cells. As a technical S6). Interestingly, accounting for metastatic specimens with ho- control, we also reexpressed a reported kinase activity-defective mozygous deletion in LKB1 and/or SIK3 increased the incidence in mutant, LKB1K78I (Fig. 3 B and Fig. S2 F; Shaw et al., 2004). lung metastasis in this group to nearly 50% (3 of 7; Fisher’s exact Whereas reintroduction of LKB1 did not affect two-dimensional test P < 0.01; Fig. S1 K and Table S7). It is worth noting that the low cell growth (Fig. S2 G), we observed a remarkable inhibition of number of specimens with alterations in these genes precluded a migration, invasion, invasive growth, and anchorage-independent more conclusive result. The potential involvement of SIK3 in the growth in both WT and K78I mutant LKB1-expressing cells (Fig. 3, prostate tumor–suppressive activity of LKB1 is coherent with re- C–E; and Fig. S2 H). We extended our studies to xenograft assays. cent reports in non–small cell lung cancer (Hollstein et al., 2019; At the level of both subcutaneous tumor growth and metastatic Murray et al., 2019), in turn suggesting that the LKB1-SIK tumor dissemination, LKB1 was tumor suppressive regardless of its ki- suppressive axis might be relevant in different cancer types. nase activity (Fig. 3, F–I; and Fig. S2, I and J). The aggressive nature of PCa in PTLK mice prompted us to The results with LKB1K78I were suggestive of an unprece- characterize the molecular and pathological features of these tu- dented kinase-independent tumor-suppressive activity. We mors. PTLK prostate tissues exhibited reduced AMPK phospho- ruled out this possibility by different means. On the one hand, rylation and elevated activity of mTOR complex 1 (by means of we evaluated the correct and equivalent formation of the ribosomal protein S6 phosphorylation), in line with the reported LKB1 complex in cells (Boudeau et al., 2003; Momcilovic and inhibitory activity of LKB1 on this pathway (Momcilovic and Shackelford, 2015). By both coimmunoprecipitation coupled to Shackelford, 2015; Fig. S2, A and B). The histological analysis of Western blot and mass spectrometry analysis, we could dem- PTLK lesions revealed that these tumors presented squamous onstrate that LKB1WT and LKB1K78I presented equivalent protein features, with elevated p63 positivity and Krt5 expression, in interaction profiles with core components of its functional line with observations in other tumor types (Momcilovic and complex, including MO25, STE20-related adaptor, HSP90 and Shackelford, 2015; Fig. S2, B and C). We then compared the mo- CDC37 (Fig. S3, A and B; and Table S9). On the other hand, the lecular features in PCa with intact Lkb1 (Ptenpc−/−, PIN or adeno- kinase-independent effect could be due to the sequestration of carcinoma as indicated) or Lkb1 complete deletion (PTLK mice, MO25 away from other STK family proteins. MO25A was the squamous cell carcinoma). PIN lesions retained a well-defined predominant isoform in these cells (Fig. S3 C). We hypothesized layer of basal prostate cells, characterized by cytokeratin 5/6 that, if the tumor suppressive activity of both LKB1 constructs (CK5/6) and p63, which were lost in invasive lesions (Fig. 2 A), in relied on the sequestration of MO25, silencing this gene would line with the notion that these tumors emerge from the aberrant mimic LKB1 expression and elicit tumor suppression. However, the proliferation of luminal cells (Arriaga and Abate-Shen, 2019). In silencing of this adaptor protein did not reduce two-dimensional or contrast, prostate tumors and metastases in PTLK mice were anchorage-independent growth (Fig. S3, D–F). predominantly composed of CK5/6- and p63-positive basal pros- The lack of evidence supporting a kinase-independent tumor tate cells and exhibited transcriptional signatures of squamous cell suppressive function of LKB1 prompted us to carefully monitor carcinoma in prostate tumor microdissected specimens (Fig. 2 A, the impact of K78I mutation on the activity of the kinase. Sur- Fig. S2 D, and Table S8). In agreement, prostate squamous cell prisingly, LKB1K78I retained 10% of the WT activity (Fig. 4 A). We carcinoma lesions from PTLK mice exhibited a 10-fold increase in evaluated the impact of LKB1 expression on the activity of its basal cell proliferation, shown by Ki67, compared with adeno- targets. In line with our clinical observations, the phosphoryl- carcinoma lesions (Fig. S2 E). Our results support the notion that ation of SIK1-3 was increased by LKB1WT and to a lower extent PCa in PTLK mice develops predominantly from the aberrant by the K78I mutant (Fig. 4 B), and this effect was retained, albeit proliferation of basal cells, leading to highly metastatic squamous with lower amplitude, when monitoring AMPK activity (Fig. S3, cell carcinoma. G and H). To ascertain whether the residual activity of LKB1K78I Hermanova et al. Journal of Experimental Medicine 4 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Figure 2. Histological characterization of prostate tumors with LKB1 deficiency. (A) Immunostaining images of H&E, CK5/6, p63, and Ki67 in squamous cell carcinoma of prostate (Ptenpc+/− Lkb1pc−/− mice), lung metastasis (Ptenpc+/− Lkb1pc−/− mice), PIN lesions (Ptenpc−/− Lkb1p+/+ mice), and adenocarcinoma (Ptenpc−/− Lkb1p+/+ mice). (B) Immunohistochemical images of H&E (left, upper panels) and 34βE12 (left, lower panels) in PCa specimens with squamous features. Right panels depict the status of LKB1 (genomic) and PTEN (genomic and protein). is sufficient to exert tumor suppression, we generated a second PCa is a highly prevalent disease in men. Although it is mutant, harboring a nucleotidic change in the magnesium- generally considered curable with first-line therapies, a fraction binding domain (D194A; Lizcano et al., 2004) that we con- of patients will exhibit biochemical recurrence associated with firmed to be fully devoid of kinase activity (Fig. 4 C and Fig. S3 I). the failure on subsequent therapeutic alternatives (Chang et al., Importantly, compared with the deleted parental cells, the ex- 2014). These men will eventually develop metastatic disease, pression of LKB1D194A neither activated SIK1-3 nor influenced the thus accounting for the mortality of the disease, together with signaling pattern or the biological response in vitro and in vivo, those individuals that exhibit metastatic PCa at presentation as opposed to LKB1WT or LKB1K78I mutant (Fig. 4, D–G; and Fig. (James et al., 2015). Bone is the predominant site of metastasis in S3, J–O). Altogether, our results demonstrate that loss of LKB1 is this disease, although a subset of highly aggressive PCa dis- associated with the emergence of squamous features in murine seminates to the lung and soft tissues (Arriaga and Abate-Shen, PCa and that the kinase requires minimal activity to suppress 2019). The biology of lethal metastatic PCa is poorly understood, aggressive PCa features in cellular systems. and there are few murine models to date that recapitulate this Hermanova et al. Journal of Experimental Medicine 5 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Figure 3. WT and mutant LKB1K78I exert significant tumor suppressive activity. (A) Western blot analysis of LKB1 protein expression in a normal prostate epithelial cell line (PWR1E) and in a panel of PCa cell lines (technical dupli- cates are shown). Representative image from three independent experiments is shown. (B) Western blot analysis of LKB1 protein ex- pression in DU145 cell line with ectopic expression of WT Flag-LKB1 (LKB1WT), the Flag-LKB1-K78I mutant (LKB1K78I), or transduced with mock vector (pBABE) using 22RV1 cells as a positive control (n = 3). Representative image from three independent experiments is shown. (C–E) Quantification (left) and representative image (right) of anchorage-independent growth (C; n = 8, independent experiments), migration (D; n = 6, independent experiments), and invasive growth (E; n = 5, independent experiments). Black circle highlights the original size of the spheroid and dashed line highlights the size of the spheroid after 48 h of incubation in DU145 cells with ectopic expression of LKB1WT or LKB1K78I or transduced with mock (pBABE). Data are normalized to con- trol (pBABE; dashed line). (F and G) Effect of LKB1 reconstitution in DU145 cells with ectopic ex- pression of LKB1WT or LKB1K78I or transduced with mock (pBABE) on tumor formation (F) and tumor weight (G; left). n = 13–14 tumors per condition. (H and I) Evaluation of metastatic capacity of control (pBABE) and LKB1WT- and LKB1K78I-ex- pressing DU145 cells using intracardiac xenotrans- plant assays. Luciferase-based signal intensity of DU145 cells in limbs was monitored in vivo (H; n as indicated) and ex vivo (I; mean signal of both tibia was calculated and presented; n as indicated in the dot plot) for ≤38 d. Representative luciferase im- ages are presented referred to the quantification plots. Statistic tests: one-sample t test (C–E), one- tailed unpaired Student’s t test (F and G), one- tailed Mann–Whitney U test (H and I). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Error bars represent SEM. pattern of aggressiveness (Arriaga and Abate-Shen, 2019). Our we find an association of this event with metastasis in the lung. study provides the field with an invaluable model to study the It is important to note that our data in murine models and hu- process of PCa dissemination in an immune-competent and man specimens is insufficient to postulate a role for LKB1 in tissue-specific context. Importantly, our patient dataset analysis metastatic tropism. However, together with the in vitro assays, using the largest cohort available (Armenia et al., 2018) shows our data indicate that LKB1 deletion results in the acquisition of that homozygous deletion in LKB1 is rare in PCa (2.6% in primary aggressiveness features, which would enable prostate tumor tumors and 4.8% in metastasis), and point mutations are negli- cells to gain access to the lung instead of the long metastatic gible. Despite this low frequency of LKB1 homozygous deletion, process that underlies bone colonization. Hermanova et al. Journal of Experimental Medicine 6 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Figure 4. LKB1 is a high-competence prostate tumor suppressor. (A) Evaluation of LKB1 kinase activity of DU145 cell line with ectopic expression of LKB1WT or LKB1K78I or transduced with mock (pBABE). Representative data from three independent experiments. (B) Western blot analysis for LKB1, SIK1-3 phospho- rylation, and HSP90 protein expression in control and LKB1WT- and LKB1K78I-expressing cells. The asterisk indicates a nonspecific band. Representative image from three independent experiments is shown. (C) Evaluation of LKB1 kinase activity of DU145 cell line with ectopic expression of LKB1WT, LKB1K78I, or LKB1D194A or transduced with mock (pBABE). Representative data from two independent experiments. (D) Western blot analysis for LKB1, SIK1-3 phosphorylation, and β-actin protein expression in control, LKB1WT-, LKB1K78I-, and LKB1D194A-expressing cells. Representative image from three independent experiments is shown. (E) Effect of LKB1WT, LKB1K78I, and LKB1D194A expression on anchorage-independent growth (n = 3, independent experiments). Data are normalized to control (pBABE; dashed line). (F) Effect of LKB1 reconstitution in DU145 cells with ectopic expression LKB1WT, the kinase-defective mutants (LKB1K78I, LKB1D194A) or infected with mock vector (pBABE) on tumor formation in xenotransplantation experiments (n = 20 injections per condition). (G) Metastatic capacity of control (pBABE), LKB1WT-, LKB1K78I-, and LKB1D194A-expressing DU145 cells using intracardiac xenotransplant assays. Quantification (left) and representative tibia view (right) of luciferase-dependent signal intensity of DU145 cells in ex vivo tibia analysis on day 42 after injection (mean signal of both tibias was calculated and presented; n as indicated in the dot plot). (H) Schematic representation of the proposed classification of tumor suppressors according to their competence. Flox, results based on tissue-specific conditional mouse models. Red filling, oncogenic features associated to dose reduction. Statistic tests: one-tailed unpaired Student’s t test (A, C, and F); one-sample t test (E), one-tailed Mann–Whitney U test (G). n.s., not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001. Lkb1 deletion is associated with the onset of metastatic disease features once oncogenic signaling is present. Following this line in other tumor types when combined with perturbation of dif- of reasoning, and in line with our patient analysis, it is tempting ferent cancer genes (Momcilovic and Shackelford, 2015). Over- to speculate that the oncogenic signaling provides a proliferative all, experimental evidence suggests that loss of LKB1 is not an advantage to cells, regardless of the cancer gene perturbed, that initiating event in cancer, but rather an enhancer of aggressive will unleash the metastasis-suppressor potential of LKB1. Hermanova et al. Journal of Experimental Medicine 7 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Squamous cell carcinoma is included in the clinical guidelines et al., 2000; Li et al., 2002). The integration of our results with the as a highly aggressive subtype of PCa, with an incidence study of the group of Alessi (Huang et al., 2008) justifies the def- of 0.5–1% and average survival of 14 mo after diagnosis inition of LKB1 as a prostate tumor suppressor with high compe- (Brunnhoelzl and Wang, 2018; Munoz et al., 2007). The combi- tence. In this context, reduction of the gene dosage of 50% nation of Pten and Lkb1 deletion in the prostate epithelium per- (heterozygous) or 70% (hypomorphic) is inconsequential for PCa formed in this study results in aggressive squamous cell prostate development. Moreover, retention of LKB1 activity as low as 10% is carcinoma. In line with the notion that basal prostate cells have sufficient for the kinase to exert tumor suppression in vitro and reduced androgen requirements (Xie et al., 2017), PTLK tumors in vivo. Only complete deletion of this gene is associated with the exhibit delocalized or negative AR expression. It is worth pathogenesis of PCa, a genetic condition that turns a premalignant highlighting the different cellular compartments that contribute disease (PIN, typical of Pten+/− mice) into lethal metastatic PCa. It is to PCa aggressiveness upon loss of tumor suppressors. Lkb1 loss worth noting that the introduction of competence in the definition affects basal cell proliferation. In contrast, deletion in other tu- of tumor suppressors remains a conceptual implementation, based mor suppressors such as Tp53 predominantly favors luminal cell on the integration of results derived from conditional and germline renewal and expansion (Agarwal et al., 2015). Importantly, mouse models of two exemplifying genes, PTEN and LKB1. Further whereas the incidence is low in human PCa, mouse models of research is warranted to explore the relevance of this concept for PCa are more prone to exhibit squamous features (Ittmann et al., other tumor suppressors, but it could harbor therapeutic relevance, 2013; Martin et al., 2011). In addition, several studies report that since it would help establish the extent of pathway reactivation Lkb1 deletion in different tissues predisposes to squamous lesions required for targeted therapies to show clinical effectiveness. (Momcilovic and Shackelford, 2015). Specifically, combined de- letion of Pten and Lkb1 in the lung was recently reported as the first bona fide mouse model of lung squamous cell carcinoma Materials and methods (Xu et al., 2014). Interestingly, the existence of compound LKB1 Genetically engineered mouse models and PTEN deletions in human prostate tumors with squamous The conditional tissue-specific phosphatase and tensin homologue pathological and molecular features opens an avenue of research (Pten) knockout (C57/BL6/129sv; Pb-Cre4; Pten lox/lox) model toward better understanding of this disease. was kindly provided by Pier Paolo Pandolfi (Beth Israel Deaconess The introduction of hypomorphic alleles in mouse models has Cancer Center, Boston, MA). The conditional tissue-specific Lkb1 enabled the fine-tuning of dosage beyond the use of gene null homogeneous background model (FVB;Lkb1lox/lox) was from knockouts. Taking advantage of this technology, we previously Mouse Models of Human Cancer Consortium in a pure FVB reported that Pten is a weak breast tumor suppressor, since the background. First, we backcrossed this line into C57BL/6J for four introduction of a hypomorphic allele in heterozygosity (70–75% generations to obtain animals with genetic background enriched of WT Pten expression) resulted in breast cancer (Alimonti et al., in C57/BL6 (>90% C57/BL6). Next we crossed Lkb1lox/lox mice 2010). The combined use of conditional mouse models and LKB1 with Pb-cre4 and Pten lox/lox mice for at least four generations to mutant in cells in this study provides strong support for the obtain a founder colony with mixed homogeneous background, notion that this kinase is a potent prostate tumor suppressor that and experimental animals were generated. Littermates where can operate at very low doses. Moreover, our analysis of the analyzed when possible. However, due to the complexity of the mutant LKB1K78I provides a feasible explanation for the discor- allele combination, the different genotypes of interest with Pten dance of our results with the elegant study performed by Huang and/or Lkb1 alteration were obtained from parallel contemporary et al. (2008). Using a whole-body hypomorphic Lkb1 mutant breeding pairs. Probasin Cre was always retained in male mice, mouse combined with systemic Pten heterozygous deletion, they since in females Pb-Cre4 expression in utero can lead to recom- observed that there was Lkb1 dose-dependent acceleration of bination in embryos during pregnancy. The Cre recombinase various cancerous lesions. However, Pten+/−;Lkb1Hy/Hy mice did expression, under the control of androgen-dependent ARR2B not exhibit alterations in prostate pathology compared with probasin promoter (Pb-Cre4), allowed the deletion of Pten in the Pten+/−;Lkb1+/+. Our report, together with past results from our prostate epithelium at puberty. Prostate Pten/Lkb1-deleted male group and others, suggests that classic genetic dosage dis- mice were termed pc−/+ (heterozygous) or pc−/− (homozygous tributions or definitions such as haploinsufficiency are unable to knockout) for each gene. correctly define the extent of their role in cancer. We put forward the concept of tumor suppressive “compe- Patient cohort and ethics approval tence” as a defining feature of tumor suppressors, taking PTEN Patients were identified from a cohort of men with advanced and LKB1 as illustrative genes (Fig. 4 H). We define standard castrate-resistant PCa treated at the Royal Marsden National competence as a continuum of reduction in tumor suppression Health Service Foundation Trust. All patients had given written capacity along with the dosage, which is recapitulated in the informed consent and were enrolled in institutional protocols prostate phenotype of Pten mutant mice (Nardella et al., 2010; Chen approved by the Royal Marsden Hospital (London, UK) ethics et al., 2005; Di Cristofano et al., 1998; Trotman et al., 2003). We review committee (reference no. 04/Q0801/60). define low competence as a reduction of <50% when a tumor suppressor gene dose results in cancerous lesions, as reported in Generation of stable cell lines the mammary tissue with Pten mutants expressing 75% of Pten DU145 cells expressing WT LKB1 (LKB1WT), LKB1 K78I (LKB1K78I), (Alimonti et al., 2010) and at lower gene doses by others (Stambolic and LKB1 D194A (LKB1D194A) mutants were generated by retroviral Hermanova et al. Journal of Experimental Medicine 8 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 gene transfer. pBABE-puro (Addgene plasmid 1764; RRID:Add- Anchorage-independent growth gene_1764) was a gift from Hartmut Land (Imperial Cancer Re- A 6-well tissue culture dish was coated with the bottom agar search Fund, Lincoln’s Inn Fields, London, UK), Jay Morgenstern, layer (complete DMEM containing 0.6% agar; SeaKem LE aga- and Bob Weinberg (Whitehead Institute for Biomedical Re- rose, Lonza) and stored at 4°C for ≥30 min to let the agar solidify. search, Cambridge, MA; Morgenstern and Land, 1990). pBABE- Cells (2,500 cells/well) were resuspended in a 0.3% low- FLAG-LKB1 (Addgene plasmid 8592; RRID:Addgene_8592) and melting-point agar (Agarose LM, Pronadisa, Conda) and plated pBABE-FLAG-KD LKB1 (Addgene plasmid 8593; RRID:Add- on top of the bottom agar. The plates were incubated (37°C and gene_8593) were a gift from Lewis Cantley (Meyer Cancer Center, 5% CO2) for 4 wk for growth of colonies. Colonies were counted New York, NY; Shaw et al., 2004). pBabe-FLAG-LKB1-D194A was using Fiji software. a gift from Dr. Bin Zheng (Columbia University, New York, NY). Lentiviral shRNA constructs targeting MO25α (TRCN0000044254; Cell migration assay TRCN0000044256) were purchased from Sigma-Aldrich, and a Cells (50,000 cells/well) were resuspended in final volume of scramble shRNA (hairpin sequence: 59-CCGGCAACAAGATGAAGA 500 µl 0.1% FBS DMEM and seeded in the upper part of the GCACCAACTCGAGTTGGTGCTCTTCATCTTGTTG-39) was used as chamber. The bottom part was filled with 1.4 ml of complete control. pLKO.1 puro was a gift from Bob Weinberg (Addgene DMEM. Plates were maintained at 37°C and 5% CO2 for 24 h. plasmid 8453; RRID:Addgene_8453; Stewart et al., 2003). Nonmigrated cells from the upper surface of the membrane HEK293FT cells were transfected with retroviral or lentiviral were carefully removed by rinsing with PBS and using a cotton vectors following standard procedures, and viral supernatant swab. The membrane was fixed with 10% formalin (15 min at was used to infect cells. Selection was done using puromycin 4°C) and stained with crystal violet. Transwell invasion was (2 µg/ml; Sigma-Aldrich). performed using Matrigel-coated chambers (Corning BioCoat Matrigel Invasion Chamber). After an incubation period of 48 h Cell culture (37°C and 5% CO2), membrane washing and fixation were per- Human prostate carcinoma cell lines PC3, DU145, and LNCaP formed under the same conditions as in the migration assay. were purchased from Leibniz-Institut Deutsche Sammlung von Both Transwell assays were evaluated by photomicrographs Mikroorganismen und Zellkulturen GmbH, who provided an from five randomly chosen fields (100×) per membrane. The authentication certificate. Human prostate cell line PWR1E and number of migrated or invaded cells was counted using Fiji. human prostate carcinoma cell lines 22RV1 and VCaP were purchased from ATCC. HEK293FT and C4-2 were generously Invasive growth assay provided by the laboratories of Rosa Barrio (CIC bioGUNE, Derio, 28,000 cells were resuspended in a final volume of 500 µl Spain) and Pier Paolo Pandolfi, respectively. Cell lines were DMEM with 6% methylcellulose (Sigma-Aldrich, M0387). Drops periodically subjected to microsatellite-based identity valida- (25 µl/drop) were pipetted on the cover of a 100-mm culture tion. The cell lines used in this study were not found in the plate. The inverted plates were incubated at 37°C and 5% CO2 for database of commonly misidentified cell lines maintained by the 48 h. Once formed, spheroids were collected, resuspended in International Cell Line Authentication Committee and National collagen I solution (1.7 mg/ml in DMEM), and seeded into 24- Center for Biotechnology Information Biosample. All cell lines well plates that were maintained at 37°C and 5% CO2 for 48 h. were routinely monitored for mycoplasma contamination and The area of the invading spheroids was measured using Fiji. quarantined while treated if positive. DU145, PC3, VCaP, and Relative invasive growth was quantified as area difference on 293FT cell lines were maintained in DMEM supplemented with day 2 minus day 0. 10% (vol/vol) FBS (Gibco) and 1% (vol/vol) penicillin-streptomycin (Gibco). LNCap, C4-2, and 22RV1 cell lines were maintained in Xenotransplant assays RPMI supplemented with 10% (vol/vol) FBS and 1% (vol/vol) All mouse experiments were performed following the ethical penicillin-streptomycin. PWR1E cell line was maintained in Ke- guidelines established by the Biosafety and Welfare Committee at ratinocyte Serum Free Medium (Gibco) supplemented with the Center for Cooperative Research in Biosciences as previously 0.05 mg/ml bovine pituitary extract (Gibco) and 5 ng/ml epider- described (Martı́n-Martı́n et al., 2018; Torrano et al., 2016). The mal growth factor (Gibco). procedures employed were performed following the recom- mendations from Association for Assessment and Accreditation Cellular assays of Laboratory Animal Care. DU145 cells (pBABE, LKB1WT, Cell proliferation LKB1K78I, and LKB1D194A; 4 × 106 cells per site) were resuspended 1 × 104 cells were plated in triplicate in 12-well dishes. in a final volume of 100 µl cold PBS supplemented with glucose Quantification was performed as reported (Torrano et al., (5 µM) mixed with 50 µl of Matrigel (Corning) and injected 2016). Anchorage-independent growth was tested by the subcutaneously in two flanks per 6–12-wk-old male athymic nude ability to grow in soft agar as described previously (Torrano mouse. Measurement of tumor size was performed three times a et al., 2016). The cell migration and invasion assay was per- week, and tumor volume was estimated using the following for- formed in a 24-well-size Transwell plate of 8-µm pores (BD mula: volume = length × width2 × 0.526. After euthanasia, tumors Falcon 351185). The invasive growth assay was performed as were weighed, and tissue was fresh frozen or fixed overnight referenced (Martı́n-Martı́n et al., 2018) using the hanging in 10% neutral buffered formalin, embedded in paraffin, sec- drop method. tioned 3-µm thick, and dried. Proliferation was assessed in Hermanova et al. Journal of Experimental Medicine 9 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 dewaxed and rehydrated slides using Ki67 antibody (MA5- CK5/6 (Ventana, 790-4554, ready to use, cytoplasmic staining), 14520, Thermo Fisher Scientific). p63 (Ventana, 790-4509, ready to use, nuclear staining), and Ki67 (Ventana, 790-4286, ready to use, nuclear staining) stains were Metastasis assays performed in automated immunostainers (BenchMark Ultra, All animal studies were performed according with the institu- Ventana Medical Systems) following routine methods. Tris-EDTA tional animal care and use committee of Institute for Research in was used for antigen retrieval. Negative controls were slides not Biomedicine, Barcelona, Spain. For these studies, male BALB/c exposed to the primary antibody, and these were incubated in PBS nude mice (Harlan) were used (12 wk old). Before cell injection, and then processed under the same conditions as the test slides. mice were anaesthetized as previously described (Torrano et al., The analysis was performed using a Nikon Eclipse 80i microscope. 2016). Using a 26-G needle, 500,000 cells were injected into the High molecular weight CK immunohistochemistry was per- left cardiac ventricle of mice resuspended in 1× PBS. Luciferase formed on the Bond RX automated platform (Leica Micro- activity from the bone legs was detected by IVIS imaging. Data systems) using a mouse monoclonal anti–high molecular weight were processed using Living Image software. Photon Flux was CK antibody (Abcam, ab776, clone 34βE12). Heat-induced epitope measured every week for 38–42 d and normalized to values retrieval was achieved with Bond Epitope Retrieval Solution obtained on day 0. Ex vivo tibia analysis was performed in those 2 (Leica Microsystems, AR9640) for 20 min at 100°C before in- mice with evidence of bioluminescence signal above background cubation with primary antibody (1:100 dilution) for 15 min at and that survived for the entire experiment. room temperature. The reaction was visualized using the Bond Polymer Refine Detection system (DS9800, Leica Microsystems). Histopathologic analysis, immunohistochemistry, PTEN immunostaining and scoring criteria was done fol- and immunofluorescence lowing the procedure in Ferraldeschi et al. (2015). Proliferative Tissue sample collection (prostate gland, lymph nodes, long bones basal epithelium cells were identified by staining with Ki67 from lower limbs, lungs, and liver) was performed at 18 mo of (Thermo Fisher Scientific, 1/200, 14-5698-82) and CK5 (Bio- age (LKB1pc+/+ and LKB1pc−/− mice) or 10 mo of age (Ptenpc+/− Legend, 1/500, PRB-160P) primary antibodies. For quantifica- LKB1pc+/+, Ptenpc+/− LKB1pc+/−, and Ptenpc+/− LKB1pc−/−). Tissue tion of double-positive cells, five independent sections from five samples were fixed overnight in 10% neutral buffered formalin, independent mice were counted using ImageJ 1.52a (National embedded in paraffin, sectioned 3-µm thick, and dried. Slides Institutes of Health). were dewaxed and rehydrated through a series of graded ethanol to water and stained with the required antibody and/or H&E. LKB1 kinase activity Histological observations on H&E-stained tissues were per- LKB1 kinase activity was measured as described previously formed using an Olympus DP73 digital camera. Prostatic lesions (Lizcano et al., 2004). Cells stably overexpressing LKB1 WT, were histologically classified according to the criteria of the K78I, or D194A mutants (FLAG-tagged), were used to immuno- Consortium Prostate Pathology Committee (Shappell et al., precipitate anti-FLAG-agarose resin, in triplicate. Reactions 2004) and scored as 0, no lesion observed; 1, focal or multifo- were performed in FLAG-agarose immunoprecipitates from cal low-grade PIN; 2, focal or multifocal high-grade PIN; 3, 1.5 mg protein, using 200 µM NuaKtide as peptide and 100 µM 32 carcinoma affecting <50% of tissue; 4, carcinoma affecting >50% P-ATP as substrates in a 60-min reaction. 1 mU of LKB1 kinase of tissue. SPCs originate from squamous metaplasia of neoplastic activity is the amount that catalyzed the phosphorylation of prostatic epithelial cells. Lungs metastasis were described as 1 pmol of NuaKtide in 1 min in the standard assay at 30°C. micrometastasis when neoplastic cell growth in a small focus or multifocal or severe metastasis when neoplastic cells progressed Quantitative real-time PCR (qRT-PCR) involving multiple or large lung regions. Lymph node lesions RNA was extracted using NucleoSpin RNA isolation kit were scored as follows: 0, no lesion observed; or 1, metastasis. (Macherey-Nagel, 740955.240C). For animal tissues, a Trizol-based A standard immunohistochemistry protocol was applied. implementation of the NucleoSpin RNA isolation kit protocol Briefly, slides were deparaffinized, and antigen retrieval was was used as reference (Ugalde-Olano et al., 2015). For all cases, 1 µg performed in a pressure cooker in citric acid buffer, pH 6 of total RNA was used for cDNA synthesis using Maxima H Minus (phospho-S6, p63) or EDTA, pH 8 (phospho-AMPK). Sections cDNA Synthesis Master Mix (Thermo Fisher Scientific, M1682). were then immersed in 3% hydrogen peroxide to block endog- qRT-PCR was performed as previously described (Carracedo et al., enous peroxidase activity and washed in PBS. Incubation with 2012). Universal Probe Library primers (Roche) and probes (Roche specific blocking serum for 30 min was followed by incubation and Thermo Fisher Scientific) are detailed in Table S10. All qRT- with primary antibodies at 4°C overnight using the following PCR data presented were normalized using GAPDH/Gapdh (Ap- concentrations: phospho-AMPK (CST, 2535, 1:50), phospho-S6 plied Biosystems, Hs02758991_g1, Mm99999915_g1). (CST, 4858, 1:500), p63 (Abcam, ab735, 1:100), and AR (Santa Cruz Biotechnology, N-20, 1:200). The next day, after washing in Western blot PBS, signal was amplified with avidin/biotin technology (Vec- Western blot was performed as previously described (Carracedo tastain Elite ABC system) or ImmPress HRP IgG (Peroxidase) et al., 2012). Briefly, cells were lysed in radioimmunoprecipitation Polymer Detection Kit (Vector Laboratories) and visualized with assay buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA, ImmPact DABSubstrate (Vector Laboratories). Finally, tissues 0.1% SDS, 1% Nonidet P40, 1% sodium deoxycholate, 1 mM sodium were counterstained with H&E. fluoride, 1 mM sodium orthovanadate, 1 mM β-glycerophosphate, Hermanova et al. Journal of Experimental Medicine 10 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 and protease inhibitor cocktail; Roche). Antibodies used were 30 min (5–60% ACN) was used with a flow rate of 300 nl/min. phospho-AMPK substrate motif (CST 5759), phospho-ACC (CST Mobile phase A was water containing 0.1% vol/vol formic acid, 11818), phospho-AMPK (CST 2535), LKB1 (CST 3050), MO25α/ and mobile phase B was ACN containing 0.1% vol/vol formic CAB39 (CST 2716s), PTEN (CST 9559), phospho-Raptor (CST acid. A lock mass compound [Glu1]-fibrinopeptide B (100 fmol/ 2083), Raptor (CST 2280), β-actin (CST 3700S), and HSP90 (CST µl) was delivered by an auxiliary pump of the LC system at 500 4874). All antibodies were used at a 1:1,000 dilution except β-actin nl/min to the reference sprayer of the NanoLockSpray (Waters) at 1:2,000. Mouse and rabbit secondary antibodies were pur- source of the mass spectrometer. chased from Jackson Immunoresearch. After standard SDS-PAGE LC was connected online to a Synapt G2Si electrospray ioni- and Western blotting techniques, proteins were visualized using zation Q-Mobility-TOF mass spectrometer (Waters) equipped the enhanced chemiluminescence. with an ion mobility chamber (T-Wave-IMS) for high-definition data acquisition analyses. All analyses were performed in positive- Proteomics mode electrospray ionization. Data were postacquisition lock mass Immunoprecipitation corrected using the double-charged monoisotopic ion of [Glu1]- Cells were lysed in immunoprecipitation buffer (40 mM Tris- fibrinopeptide B. Accurate mass LC-MS data were collected in HCl, pH 7.5, 1 mM EDTA, 150 mM NaCl, 0.5% Triton X-100, hexadehydro-Diels–Alder mode, which enhances signal intensities 150 mM NaCl, 1 mM MgCl2, 1 mM sodium fluoride, 1 mM sodium using the ion mobility separation step. orthovanadate, 1 mM β-glycerophosphate, and protease inhibi- Database searching was performed using Mascot 2.2.07 tor cocktail; Roche). An equal amount of each protein lysate was (Matrixscience) against a Uniprot-Swissprot database filled only incubated with anti-FLAG M2 Affinity Gel (Sigma-Aldrich) for with entries corresponding to Homo sapiens (without isoforms). 1 h at 4°C. The immune complexes and protein lysates were For protein identification, the following parameters were adopted: analyzed by Western blot with anti-LKB1 and anti-MO25 anti- carbamidomethylation of cysteines (C) as fixed modification, ox- body. The immune complexes were in parallel subjected to liq- idation of methionines (M) as variable modification, 15 ppm of uid chromatography/tandem mass spectrometry (LC–MS/MS) peptide mass tolerance, 0.2 D fragment mass tolerance, ≤3 missed analysis. cleavage points, and peptide charges of +2 and +3. SDS-PAGE Progenesis LC-MS software analysis For LC–MS/MS analysis, protein samples were boiled for 5 min Progenesis LC-MS (v2.0.5556.29015, Nonlinear Dynamics) was and resolved in 12.5% acrylamide gels, using a Mini-Protean II used for the label-free differential protein expression analysis. electrophoresis cell (Bio-Rad). A constant voltage of 150 V was One of the runs was used as the reference to which the precursor applied for 45 min. Gels were fixed in a solution containing 10% masses in all other samples were aligned. Only features com- acetic acid and 40% ethanol for 30 min and stained overnight in prising charges of 2+ and 3+ were selected. The raw abundances Sypro Ruby (Bio-Rad). Gels were then washed in a solution of each feature were automatically normalized and logarith- containing 10% ethanol and 7% acetic acid for 30 min, and the mized against the reference run. Samples were grouped in ac- image was acquired using a Typhoon Trio scanner (GE Health- cordance to the comparison being performed, and ANOVA was care). Each lane was cut in slices and subjected to tryptic di- performed. A peak list containing the information of all the gestion, followed by LC–MS/MS analysis. features was generated and exported to the Mascot search en- gine (Matrix Science). This file was searched against a Uniprot/ Tryptic digestion Swissprot database under the conditions stated in the previous Gel bands were sliced into three small pieces, which were section, and the list of identified peptides was imported back to washed in milli-Q water. Reduction and alkylation were per- Progenesis LC-MS. Protein quantitation was performed based on formed using 10 mM dithiothreitol in 50 mM ammonium bi- the three most intense nonconflicting peptides (peptides oc- carbonate at 56°C for 20 min, followed by iodoacetamide curring in only one protein), except for proteins with only two (50 mM iodoacetamide in 50 mM ammonium bicarbonate) for nonconflicting peptides. The significance of expression changes another 20 min in the dark. Gel pieces were dried and incubated was tested at the protein level, and proteins identified with at with trypsin (12.5 µg/ml in 50 mM ammonium bicarbonate) for least two peptides, an ANOVA P value ≤0.05, and a ratio >1.5 in 20 min on ice. After rehydration, the trypsin supernatant was either direction were selected for further analyses. discarded. Gel pieces were hydrated with 50 mM ammonium bicarbonate and incubated overnight at 37°C. After digestion, Gene expression microarrays acidic peptides were cleaned with trifluoroacetic acid 0.1% and Gene expression analysis of tumors from different mouse models dried in an RVC2 25 speedvac concentrator (Martin Christ). (the original Ptenpc−/− mouse model [Chen et al., 2005] versus the Peptides were resuspended in 10 µl of 0.1% fluoroacetic acid and Ptenpc+/− Lkb1pc−/− generated in this study) was performed using sonicated for 5 min before analysis. GeneChip Mouse Transcriptome Array 1.0 (Affymetrix). Differ- ential expression was obtained using Transcriptome Analysis LC-MS analysis Console v4.0. Expression data were normalized and background- LC was performed using an NanoAcquity nano-HPLC (Waters), and batch-corrected using the Signal Space Transformation- equipped with a Waters BEH C18 nano-column (200 mm × Robust Multi-Chip Analysis implemented in that software. 75 µm; internal diameter, 1.8 µm), A chromatographic ramp of Data are deposited in GEO (GSE133837). Hermanova et al. Journal of Experimental Medicine 11 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Total RNA was extracted from formalin-fixed and paraffin- David Barneda, Len Stephens, Ana M. Aransay, and Felix Elortza embedded (FFPE) tissue using RNAstorm FFPE RNA extrac- for technical advice. tion kit (Cell Data Sciences) according to the manufacturer’s I. Hermanova and F. Salvador were funded by the Juan de la instructions. cDNAs from a total of 12 ng RNA were generated, Cierva program from the Spanish Ministry of Research. V. fragmented, biotinylated, and hybridized to the GeneChip Torrano is funded by Fundación Vasca de Innovación e Inves- Mouse Transcriptome Array 1.0. The arrays were washed and tigación Sanitarias, Bioef (BIO15/CA/052), the Ministerio de stained on a GeneChip Fluidics Station 450 (Affymetrix); Ciencia, Innovación y Universidades (RTI2018-097267-B-100), scanning was performed with the GeneChip Scanner 3000 7G the Fundación Cientı́fica Asociación Española Contra el Cáncer and image analysis with the Affymetrix GeneChip Command J.P. Bizkaia, and the Basque Department of Health (2016111109). Console Scan Control. Gene set enrichment analysis was The work of A. Carracedo is supported by the Basque Depart- performed using the MSigDB database (Subramanian et al., ment of Industry, Tourism and Trade (Elkartek), the Department 2005). of Education (IKERTALDE IT1106-16, with A. Gomez-Muñoz), the Department of Health (2019222031), the Fundación BBVA, the Next-generation sequencing and copy number variation (CNV) Ministerio de Ciencia, Innovación y Universidades (SAF2016- calling 79381-R, FEDER [European Regional Development Fund, EU]; DNA was extracted from FFPE tumor blocks positive for tumor Severo Ochoa Excellence Accreditation SEV-2016-0644; Excel- content using the FFPE Tissue DNA kit (Qiagen). DNA was lence Networks SAF2016-81975-REDT), European Training Net- quantified with the Quant-iT high-sensitivity PicoGreen double- works Project (H2020-MSCA-ITN-308 2016 721532, with G. stranded DNA Assay Kit (Invitrogen), and quality control was Velasco), the Fundación Cientı́fica Asociación Española Contra el performed with the Illumina FFPE quality control kit (WG-321- Cáncer (IDEAS175CARR; GCTRA18006CARR, with M. Graupera 1001). Libraries for next-generation targeted sequencing were and R.R. Gomis), “La Caixa” Foundation (HR17-00094), and the constructed from 40 ng of DNA using a customized panel European Research Council (Starting Grant 336343, PoC 754627, (Generead DNaseq Mix-n-Match Panel v2; Qiagen) as previously and Consolidator Grant 819242). We are grateful for the support described (Mateo et al., 2015). CNVs were assessed from the of Mondravember and Movembergara. J.M. Paramio was funded aligned sequence files (.bam) using CNVkit v0.7.3 (Seed et al., by Ministerio de Ciencia, Innovación y Universidades (SAF2015- 2017; Talevich et al., 2016; Turajlic et al., 2018). 66015-R and CB16/12/00228). J.I. López was funded by the Ministerio de Ciencia, Innovación y Universidades (SAF2016- Statistics and reproducibility 79847-R). R. Barrio and J.D. Sutherland acknowledge Ministerio No statistical method was used to predetermine sample size. de Ciencia, Innovación y Universidades/European Regional De- The experiments were not randomized. The investigators were velopment Fund (BFU2017-84653-P, SEV-2016-0644). G. Ve- not blinded to allocation during experiments and outcome assess- lasco’s research is supported by grants integrated into the State ment. n values represent the number of independent experiments Plan for R & D + I2013-2016 and funded by the Instituto de Salud performed, the number of individual mice, or the number of patient Carlos III (PI18/00442 and PI15/00339), the European Regional specimens. Unless otherwise stated, data analyzed by parametric tests Development Fund, and Fundació la Marató de TV3 (20134031). are represented by the mean ± SEM of pooled experiments and the J.M. Lizcano was supported by the Ministerio de Ciencia, In- median ± interquartile range for experiments analyzed by nonpara- novación y Universidades (SAF2015-64237-R) and cofunded by metric tests. For each independent in vitro experiment, at least three the European Regional Development Fund. R.R. Gomis is sup- technical replicates were used, and a minimum number of three ported by the Ministerio de Ciencia, Innovación y Universidades, experiments was performed to ensure adequate statistical power. European Regional Development Fund (CIBERONC and SAF016- In the in vitro experiments, normal distribution was assumed, and 76008R), and “La Caixa” Foundation (HR17-00092). CIBERONC one-sample t tests were applied for one-component comparisons was cofunded with European Regional Development funds and with control and Student’s t test for two-component comparisons. funded by Instituto de Salud Carlos III. L. Valcarcel-Jimenez was For in vivo experiments, D’Agostino–Pearson, Shapiro–Wilk, and funded by a Basque Government predoctoral grant. Kolmogorov–Smirnov normality tests were applied. Student’s Author contributions: In vitro studies: I. Hermanova, P. t test was used to compare data with normal distribution, and Zúñiga-Garcı́a, and A. Caro-Maldonado, with support from S. nonparametric Mann–Whitney U exact test was used to analyze Fernandez-Rui, N. Martı́n-Martı́n, A. Arruabarrena-Aristorena, samples not following a normal distribution. The confidence L. Valcarcel-Jimenez, and A. Talamillo. In vivo studies: I. Her- level used for all the statistical analyses was 95% (α = 0.05). manova, P. Zúñiga-Garcı́a, A. Caro-Maldonado, and M. Nuñez- Two-tailed statistical analysis was applied for experimental Olle, with support from N. Martı́n-Martı́n, A. Zabala-Letona, A. design without predicted result, and one-tailed for validation or Carracedo, M. Lorente, M. Guiu, and V. Torrano. In vivo biolu- hypothesis-driven experiments. GraphPad Prism 8.0.2 soft- minescence assays: F. Salvador and R.R. Gomis. Genotyping and ware was used for statistical calculations. cloning: I. Astobiza and S. Fernandez-Rui, with support from J.D. Sutherland and R. Barrio. Pathologic and immunohistochemical analysis of murine tissues: J.M. Flores and J.I. López, with par- Acknowledgments ticipation of I. Hermanova, I. Lodewijk, C. Suárez-Cabrera, S. We are grateful to the Carracedo laboratory for valuable input Fernandez-Rui, and L. Camacho for immunohistochemistry. and to Diego Barriales, Elisabet Megias-Roda, Juan Anguita, Pathologic, genetic, and immunohistochemical analysis of Hermanova et al. Journal of Experimental Medicine 12 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 human PCa specimens: B. Gurel and S. Carreira, supervised by Cerami, E., J. Gao, U. Dogrusoz, B.E. Gross, S.O. Sumer, B.A. Aksoy, A. Ja- cobsen, C.J. Byrne, M.L. Heuer, E. Larsson, et al. 2012. The cBio cancer J.S. de Bono. Immunofluorescence analysis of tissues: A. Martinez- genomics portal: an open platform for exploring multidimensional Romero, supervised by M. Graupera. LKB1 kinase radiochemical cancer genomics data. Cancer Discov. 2:401–404. https://doi.org/10.1158/ assay: J.M. Lizcano. Bioinformatics and statistical analy- 2159-8290.CD-12-0095 sis: A.R. Cortazar. Murine tissue microdissection and gene Chang, A.J., K.A. Autio, M. Roach III, and H.I. Scher. 2014. High-risk prostate cancer-classification and therapy. Nat. Rev. Clin. Oncol. 11:308–323. expression analysis: C. Suárez-Cabrera, I. Lodewijk, and https://doi.org/10.1038/nrclinonc.2014.68 J.M. Paramio. Other contributions: G. Velasco (supervi- Chen, Z., L.C. Trotman, D. Shaffer, H.K. Lin, Z.A. Dotan, M. Niki, J.A. sion in xenograft experiments), A. Gomez-Muñoz (shared Koutcher, H.I. Scher, T. Ludwig, W. Gerald, et al. 2005. Crucial role of p53-dependent cellular senescence in suppression of Pten- supervision of L. Camacho), J. Trka (shared supervision of deficient tumorigenesis. Nature. 436:725–730. https://doi.org/10 I. Hermanova), V. Torrano (support with human cancer .1038/nature03918 dataset analysis). General supervision of data analysis, Di Cristofano, A., B. Pesce, C. Cordon-Cardo, and P.P. Pandolfi. 1998. Pten is essential for embryonic development and tumour suppression. Nat. project direction, and manuscript writing: A. Carracedo. Genet. 19:348–355. https://doi.org/10.1038/1235 Ferraldeschi, R., D. Nava Rodrigues, R. Riisnaes, S. Miranda, I. Figueiredo, P. Disclosures: Dr. Graupera reported, "the Graupera laboratory Rescigno, P. Ravi, C. Pezaro, A. Omlin, D. Lorente, et al. 2015. PTEN protein loss and clinical outcome from castration-resistant prostate has research agreements with ArQule and Venthera. None of cancer treated with abiraterone acetate. Eur. Urol. 67:795–802. https:// those have a relationship with cancer or LKB1." Dr. Gomis re- doi.org/10.1016/j.eururo.2014.10.027 ported shares of Inbiomotion (<$10,000). No other disclosures Fraser, M., V.Y. Sabelnykova, T.N. Yamaguchi, L.E. Heisler, J. Livingstone, V. Huang, Y.J. Shiah, F. Yousif, X. Lin, A.P. Masella, et al. 2017. Genomic were reported. hallmarks of localized, non-indolent prostate cancer. Nature. 541: 359–364. https://doi.org/10.1038/nature20788 Submitted: 23 September 2019 Gao, J., B.A. Aksoy, U. Dogrusoz, G. Dresdner, B. Gross, S.O. Sumer, Y. Sun, A. Jacobsen, R. Sinha, E. Larsson, et al. 2013. Integrative analysis of com- Revised: 16 December 2019 plex cancer genomics and clinical profiles using the cBioPortal. Sci. Accepted: 6 February 2020 Signal. 6:pl1. https://doi.org/10.1126/scisignal.2004088 Gerhauser, C., F. Favero, T. Risch, R. Simon, L. Feuerbach, Y. Assenov, D. Heckmann, N. Sidiropoulos, S.M. Waszak, D. Hübschmann, et al. 2018. Molecular Evolution of Early-Onset Prostate Cancer Identifies Molec- References ular Risk Markers and Clinical Trajectories. Cancer Cell. 34:996–1011.e8. Abida, W., J. Armenia, A. Gopalan, R. Brennan, M. Walsh, D. Barron, D. https://doi.org/10.1016/j.ccell.2018.10.016 Danila, D. Rathkopf, M. Morris, S. Slovin, et al. 2017. Prospective Ge- Hollstein, P.E., L.J. Eichner, S.N. Brun, A. Kamireddy, R.U. Svensson, L.I. nomic Profiling of Prostate Cancer Across Disease States Reveals Vera, D.S. Ross, T.J. Rymoff, A. Hutchins, H.M. Galvez, et al. 2019. The Germline and Somatic Alterations That May Affect Clinical Decision AMPK-related kinases SIK1 and SIK3 mediate key tumor-suppressive Making. JCO Precis. Oncol. 2017:PO.17.00029. effects of LKB1 in NSCLC. Cancer Discov. 9:1606–1627. https://doi.org/10 Abida, W., J. Cyrta, G. Heller, D. Prandi, J. Armenia, I. Coleman, M. Cieslik, M. .1158/2159-8290.CD-18-1261 Benelli, D. Robinson, E.M. Van Allen, et al. 2019. Genomic correlates of Huang, X., S. Wullschleger, N. Shpiro, V.A. McGuire, K. Sakamoto, Y.L. clinical outcome in advanced prostate cancer. Proc. Natl. Acad. Sci. USA. Woods, W. McBurnie, S. Fleming, and D.R. Alessi. 2008. Important 116:11428–11436. https://doi.org/10.1073/pnas.1902651116 role of the LKB1-AMPK pathway in suppressing tumorigenesis in Agarwal, S., P.G. Hynes, H.S. Tillman, R. Lake, W.G. Abou-Kheir, L. Fang, O.M. PTEN-deficient mice. Biochem. J. 412:211–221. https://doi.org/10 Casey, A.H. Ameri, P.L. Martin, J.J. Yin, et al. 2015. Identification of .1042/BJ20080557 Different Classes of Luminal Progenitor Cells within Prostate Tumors. Ikediobi, O.N., H. Davies, G. Bignell, S. Edkins, C. Stevens, S. O’Meara, T. Cell Rep. 13:2147–2158. https://doi.org/10.1016/j.celrep.2015.10.077 Santarius, T. Avis, S. Barthorpe, L. Brackenbury, et al. 2006. Mutation Alimonti, A., A. Carracedo, J.G. Clohessy, L.C. Trotman, C. Nardella, A. Egia, L. analysis of 24 known cancer genes in the NCI-60 cell line set. Mol. Salmena, K. Sampieri, W.J. Haveman, E. Brogi, et al. 2010. Subtle var- Cancer Ther. 5:2606–2612. https://doi.org/10.1158/1535-7163.MCT-06 iations in Pten dose determine cancer susceptibility. Nat. Genet. 42: -0433 454–458. https://doi.org/10.1038/ng.556 Imielinski, M., A.H. Berger, P.S. Hammerman, B. Hernandez, T.J. Pugh, E. Armenia, J., S.A.M. Wankowicz, D. Liu, J. Gao, R. Kundra, E. Reznik, W.K. Hodis, J. Cho, J. Suh, M. Capelletti, A. Sivachenko, et al. 2012. Mapping Chatila, D. Chakravarty, G.C. Han, I. Coleman, et al. PCF/SU2C Inter- the hallmarks of lung adenocarcinoma with massively parallel se- national Prostate Cancer Dream Team. 2018. The long tail of oncogenic quencing. Cell. 150:1107–1120. https://doi.org/10.1016/j.cell.2012.08.029 drivers in prostate cancer. Nat. Genet. 50:645–651. https://doi.org/10 Ireland, H., R. Kemp, C. Houghton, L. Howard, A.R. Clarke, O.J. Sansom, and .1038/s41588-018-0078-z D.J. Winton. 2004. Inducible Cre-mediated control of gene expression Arriaga, J.M., and C. Abate-Shen. 2019. Genetically Engineered Mouse Models in the murine gastrointestinal tract: effect of loss of beta-catenin. of Prostate Cancer in the Postgenomic Era. Cold Spring Harb. Perspect. Gastroenterology. 126:1236–1246. https://doi.org/10.1053/j.gastro.2004 Med. 9:a030528. https://doi.org/10.1101/cshperspect.a030528 .03.020 Berger, A.H., A.G. Knudson, and P.P. Pandolfi. 2011. A continuum model for Ittmann, M., J. Huang, E. Radaelli, P. Martin, S. Signoretti, R. Sullivan, B.W. tumour suppression. Nature. 476:163–169. https://doi.org/10.1038/ Simons, J.M. Ward, B.D. Robinson, G.C. Chu, et al. 2013. Animal models nature10275 of human prostate cancer: the consensus report of the New York Boudeau, J., M. Deak, M.A. Lawlor, N.A. Morrice, and D.R. Alessi. 2003. Heat- meeting of the Mouse Models of Human Cancers Consortium Prostate shock protein 90 and Cdc37 interact with LKB1 and regulate its stability. Pathology Committee. Cancer Res. 73:2718–2736. https://doi.org/10.1158/ Biochem. J. 370:849–857. https://doi.org/10.1042/bj20021813 0008-5472.CAN-12-4213 Brunnhoelzl, D., and J. Wang. 2018. Clinical features, treatment, prognosis, James, N.D., M.R. Spears, N.W. Clarke, D.P. Dearnaley, J.S. De Bono, J. Gale, J. and outcome of 47 patients with pure squamous cell carcinoma of the Hetherington, P.J. Hoskin, R.J. Jones, R. Laing, et al. 2015. Survival with prostate. J. Clin. Oncol. 36(6 suppl):7. https://doi.org/10.1200/JCO.2018 Newly Diagnosed Metastatic Prostate Cancer in the “Docetaxel Era”: .36.6_suppl.7 Data from 917 Patients in the Control Arm of the STAMPEDE Trial (MRC Cancer Genome Atlas Research Network. 2015. The Molecular Taxonomy of PR08, CRUK/06/019). Eur. Urol. 67:1028–1038. https://doi.org/10.1016/j Primary Prostate Cancer. Cell. 163:1011–1025. https://doi.org/10.1016/j .eururo.2014.09.032 .cell.2015.10.025 Jordan, E.J., H.R. Kim, M.E. Arcila, D. Barron, D. Chakravarty, J. Gao, M.T. Carracedo, A., D. Weiss, A.K. Leliaert, M. Bhasin, V.C. de Boer, G. Laurent, Chang, A. Ni, R. Kundra, P. Jonsson, et al. 2017. Prospective Compre- A.C. Adams, M. Sundvall, S.J. Song, K. Ito, et al. 2012. A metabolic hensive Molecular Characterization of Lung Adenocarcinomas for Ef- prosurvival role for PML in breast cancer. J. Clin. Invest. 122:3088–3100. ficient Patient Matching to Approved and Emerging Therapies. Cancer https://doi.org/10.1172/JCI62129 Discov. 7:596–609. https://doi.org/10.1158/2159-8290.CD-16-1337 Hermanova et al. Journal of Experimental Medicine 13 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Knudson, A.G. Jr. 1971. Mutation and cancer: statistical study of retinoblas- and classification. The consensus report from the Bar Harbor meeting toma. Proc. Natl. Acad. Sci. USA. 68:820–823. https://doi.org/10.1073/ of the Mouse Models of Human Cancer Consortium Prostate Pathology pnas.68.4.820 Committee. Cancer Res. 64:2270–2305. https://doi.org/10.1158/0008 Kumar, A., I. Coleman, C. Morrissey, X. Zhang, L.D. True, R. Gulati, R. Etzioni, -5472.CAN-03-0946 H. Bolouri, B. Montgomery, T. White, et al. 2016. Substantial interin- Shaw, R.J., N. Bardeesy, B.D. Manning, L. Lopez, M. Kosmatka, R.A. DePinho, dividual and limited intraindividual genomic diversity among tumors and L.C. Cantley. 2004. The LKB1 tumor suppressor negatively regu- from men with metastatic prostate cancer. Nat. Med. 22:369–378. lates mTOR signaling. Cancer Cell. 6:91–99. https://doi.org/10.1016/j.ccr https://doi.org/10.1038/nm.4053 .2004.06.007 Li, G., G.W. Robinson, R. Lesche, H. Martinez-Diaz, Z. Jiang, N. Rozengurt, Stambolic, V., M.S. Tsao, D. Macpherson, A. Suzuki, W.B. Chapman, and T.W. K.U. Wagner, D.C. Wu, T.F. Lane, X. Liu, et al. 2002. Conditional loss of Mak. 2000. High incidence of breast and endometrial neoplasia re- PTEN leads to precocious development and neoplasia in the mammary sembling human Cowden syndrome in pten+/- mice. Cancer Res. 60: gland. Development. 129:4159–4170. 3605–3611. Lizcano, J.M., O. Göransson, R. Toth, M. Deak, N.A. Morrice, J. Boudeau, S.A. Stewart, S.A., D.M. Dykxhoorn, D. Palliser, H. Mizuno, E.Y. Yu, D.S. An, D.M. Hawley, L. Udd, T.P. Mäkelä, D.G. Hardie, and D.R. Alessi. 2004. LKB1 is Sabatini, I.S. Chen, W.C. Hahn, P.A. Sharp, et al. 2003. Lentivirus- a master kinase that activates 13 kinases of the AMPK subfamily, in- delivered stable gene silencing by RNAi in primary cells. RNA. 9: cluding MARK/PAR-1. EMBO J. 23:833–843. https://doi.org/10.1038/sj 493–501. https://doi.org/10.1261/rna.2192803 .emboj.7600110 Subramanian, A., P. Tamayo, V.K. Mootha, S. Mukherjee, B.L. Ebert, M.A. Martin, P., Y.N. Liu, R. Pierce, W. Abou-Kheir, O. Casey, V. Seng, D. Camacho, Gillette, A. Paulovich, S.L. Pomeroy, T.R. Golub, E.S. Lander, and J.P. R.M. Simpson, and K. Kelly. 2011. Prostate epithelial Pten/TP53 loss Mesirov. 2005. Gene set enrichment analysis: a knowledge-based leads to transformation of multipotential progenitors and epithelial to approach for interpreting genome-wide expression profiles. Proc. mesenchymal transition. Am. J. Pathol. 179:422–435. https://doi.org/10 Natl. Acad. Sci. USA. 102:15545–15550. https://doi.org/10.1073/pnas .1016/j.ajpath.2011.03.035 .0506580102 Martı́n-Martı́n, N., A. Zabala-Letona, S. Fernández-Ruiz, L. Arreal, L. Ca- Talevich, E., A.H. Shain, T. Botton, and B.C. Bastian. 2016. CNVkit: Genome- macho, M. Castillo-Martin, A.R. Cortazar, V. Torrano, I. Astobiza, P. Wide Copy Number Detection and Visualization from Targeted DNA Zúñiga-Garcı́a, et al. 2018. PPARδ Elicits Ligand-Independent Repres- Sequencing. PLOS Comput. Biol. 12:e1004873. https://doi.org/10.1371/ sion of Trefoil Factor Family to Limit Prostate Cancer Growth. Cancer journal.pcbi.1004873 Res. 78:399–409. https://doi.org/10.1158/0008-5472.CAN-17-0908 Taylor, B.S., N. Schultz, H. Hieronymus, A. Gopalan, Y. Xiao, B.S. Carver, V.K. Mateo, J., S. Carreira, S. Sandhu, S. Miranda, H. Mossop, R. Perez-Lopez, D. Arora, P. Kaushik, E. Cerami, B. Reva, et al. 2010. Integrative genomic Nava Rodrigues, D. Robinson, A. Omlin, N. Tunariu, et al. 2015. DNA- profiling of human prostate cancer. Cancer Cell. 18:11–22. https://doi Repair Defects and Olaparib in Metastatic Prostate Cancer. N. Engl. .org/10.1016/j.ccr.2010.05.026 J. Med. 373:1697–1708. https://doi.org/10.1056/NEJMoa1506859 Torrano, V., L. Valcarcel-Jimenez, A.R. Cortazar, X. Liu, J. Urosevic, M. Cas- Momcilovic, M., and D.B. Shackelford. 2015. Targeting LKB1 in cancer - ex- tillo-Martin, S. Fernández-Ruiz, G. Morciano, A. Caro-Maldonado, M. posing and exploiting vulnerabilities. Br. J. Cancer. 113:574–584. https:// Guiu, et al. 2016. The metabolic co-regulator PGC1α suppresses prostate doi.org/10.1038/bjc.2015.261 cancer metastasis. Nat. Cell Biol. 18:645–656. https://doi.org/10.1038/ Morgenstern, J.P., and H. Land. 1990. Advanced mammalian gene transfer: ncb3357 high titre retroviral vectors with multiple drug selection markers and a Trotman, L.C., M. Niki, Z.A. Dotan, J.A. Koutcher, A. Di Cristofano, A. Xiao, complementary helper-free packaging cell line. Nucleic Acids Res. 18: A.S. Khoo, P. Roy-Burman, N.M. Greenberg, T. Van Dyke, et al. 2003. 3587–3596. https://doi.org/10.1093/nar/18.12.3587 Pten dose dictates cancer progression in the prostate. PLoS Biol. 1:E59. Munoz, F., P. Franco, P. Ciammella, M. Clerico, M. Giudici, A.R. Filippi, and U. https://doi.org/10.1371/journal.pbio.0000059 Ricardi. 2007. Squamous cell carcinoma of the prostate: long-term Turajlic, S., H. Xu, K. Litchfield, A. Rowan, T. Chambers, J.I. Lopez, D. Nicol, T. survival after combined chemo-radiation. Radiat. Oncol. 2:15. https:// O’Brien, J. Larkin, S. Horswell, et al. TRACERx Renal Consortium. 2018. doi.org/10.1186/1748-717X-2-15 Tracking Cancer Evolution Reveals Constrained Routes to Metastases: Murray, C.W., J.J. Brady, M.K. Tsai, C. Li, I.P. Winters, R. Tang, L. Andrejka, TRACERx Renal. Cell. 173:581–594.e12. https://doi.org/10.1016/j.cell R.K. Ma, C.A. Kunder, P. Chu, and M.M. Winslow. 2019. An Lkb1-Sik .2018.03.057 axis suppresses lung tumor growth and controls differentiation. Cancer Ugalde-Olano, A., A. Egia, S. Fernández-Ruiz, A. Loizaga-Iriarte, P. Zuñiga- Discov. 9:1590–1605. https://doi.org/10.1158/2159-8290.CD-18-1237 Garcı́a, S. Garcia, F. Royo, I. Lacasa-Viscasillas, E. Castro, A.R. Cortazar, Nardella, C., A. Carracedo, L. Salmena, and P.P. Pandolfi. 2010. Faithfull et al. 2015. Methodological aspects of the molecular and histological modeling of PTEN loss driven diseases in the mouse. Curr. Top. Micro- study of prostate cancer: focus on PTEN. Methods. 77-78:25–30. https:// biol. Immunol. 347:135–168. doi.org/10.1016/j.ymeth.2015.02.005 Pearson, H.B., A. McCarthy, C.M. Collins, A. Ashworth, and A.R. Clarke. Weinstein, J.N., E.A. Collisson, G.B. Mills, K.R. Shaw, B.A. Ozenberger, K. 2008. Lkb1 deficiency causes prostate neoplasia in the mouse. Cancer Ellrott, I. Shmulevich, C. Sander, and J.M. Stuart. Cancer Genome Atlas Res. 68:2223–2232. https://doi.org/10.1158/0008-5472.CAN-07-5169 Research Network. 2013. The Cancer Genome Atlas Pan-Cancer analysis Seed, G., W. Yuan, J. Mateo, S. Carreira, C. Bertan, M. Lambros, G. Boysen, R. project. Nat. Genet. 45:1113–1120. https://doi.org/10.1038/ng.2764 Ferraldeschi, S. Miranda, I. Figueiredo, et al. 2017. Gene Copy Number Xie, Q., Y. Liu, T. Cai, C. Horton, J. Stefanson, and Z.A. Wang. 2017. Dissecting Estimation from Targeted Next-Generation Sequencing of Prostate cell-type-specific roles of androgen receptor in prostate homeostasis Cancer Biopsies: Analytic Validation and Clinical Qualification. Clin. and regeneration through lineage tracing. Nat. Commun. 8:14284. Cancer Res. 23:6070–6077. https://doi.org/10.1158/1078-0432.CCR-17 https://doi.org/10.1038/ncomms14284 -0972 Xu, C., C.M. Fillmore, S. Koyama, H. Wu, Y. Zhao, Z. Chen, G.S. Herter-Sprie, Shappell, S.B., G.V. Thomas, R.L. Roberts, R. Herbert, M.M. Ittmann, M.A. E.A. Akbay, J.H. Tchaicha, A. Altabef, et al. 2014. Loss of Lkb1 and Pten Rubin, P.A. Humphrey, J.P. Sundberg, N. Rozengurt, R. Barrios, et al. leads to lung squamous cell carcinoma with elevated PD-L1 expression. 2004. Prostate pathology of genetically engineered mice: definitions Cancer Cell. 25:590–604. https://doi.org/10.1016/j.ccr.2014.03.033 Hermanova et al. Journal of Experimental Medicine 14 of 14 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Supplemental material Hermanova et al. Journal of Experimental Medicine S1 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Figure S1. Consequences of aberrations in Lkb1 or its targets in murine and human prostate tissue. (A) Analysis of Lkb1 gene expression by qRT-PCR in Lkb1pc+/+ (n = 5) and Lkb1pc−/− (n = 5) mice. Data were normalized to Gapdh expression. (B) Western blot analysis for Lkb1 protein expression in Lkb1pc+/+ (n = 5) and Lkb1pc−/− (n = 5) mice. (C) Analysis of Lkb1 gene expression by qRT-PCR in Ptenpc+/− Lkb1pc+/+ (n = 3), Ptenpc+/− Lkb1pc+/− (n = 3), and Ptenpc+/− Lkb1pc−/− (n = 3) mice. Data are normalized to Gapdh expression. (D) Western blot analysis of Lkb1 and Pten protein expression in Ptenpc+/− Lkb1pc+/+ (n = 3), Ptenpc+/− Lkb1pc+/− (n = 3), and Ptenpc+/− Lkb1pc−/− (n = 3) mice. (E) Quantification of ventral (VP), dorsolateral (DLP) and anterior (AP) prostate weight in Lkb1pc+/+ and Lkb1pc−/− mice (n = 27; two lobes per mouse). (F) Immunohistochemical images of AR staining in prostate tissue in Ptenpc+/− Lkb1pc+/+ mice (upper panels) and prostate tumors and lung metastatic lesions in Ptenpc+/− Lkb1pc−/− mice (lower panels). (G) Lollipop representation of LKB1 point mutations in PCa (Abida et al., 2017; Abida et al., 2019; Armenia et al., 2018; Cancer Genome Atlas Research Network, 2015; Fraser et al., 2017; Gerhauser et al., 2018; Kumar et al., 2016; Taylor et al., 2010) and lung adenocarcinoma (Imielinski et al., 2012; Weinstein et al., 2013; Jordan et al., 2017) datasets. (H) Incidence of LKB1 genomic alterations in metastatic specimens of patients with metastatic PCa (Armenia et al., 2018). Cases without LKB1 homozygous deletion (LKB1 not homdel); cases with LKB1 homozygous deletion (homdel). Not LN, metastasis identified in all analyzed tissues except for lymph node. Numbers below the graph indicate total number of analyzed metastatic samples. (I) Incidence of genomic aberrations in metastatic lesions in lung, lymph nodes, bone, liver, and other tissues in patients with metastatic PCa for the indicated genes (LKB1 targets; Armenia et al., 2018). (J) Incidence of LKB1 genomic alterations in metastatic specimens of patients with metastatic PCa (Armenia et al., 2018). Cases without SIK3 homozygous deletion (SIK3 not homdel); cases with SIK3 homozygous deletion (homdel). Not bone, metastasis identified in all analyzed tissues except for bone; not lung, metastasis identified in all analyzed tissues except for lung. Numbers below the graph indicate total number of analyzed metastatic samples. (K) Incidence of LKB1 and/or SIK3 genomic alterations in metastatic specimens of patients with metastatic PCa (Armenia et al., 2018). Cases without LKB1 and/or SIK3 homozygous deletion (Rest); cases with LKB1 and/or SIK3 homozygous deletion (homdel). Not bone, metastasis identified in all analyzed tissues except for bone; not lung, metastasis identified in all analyzed tissues except for lung. Numbers below the graph indicate total number of analyzed metastatic samples. Statistic tests: one-tailed Mann–Whitney test (A, C, and E); Fisher exact test (J and K). n.s., not sig- nificant; *, P < 0.05; **, P < 0.01. Error bars represent SEM. Hermanova et al. Journal of Experimental Medicine S2 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Figure S2. Pathological and molecular characterization of Lkb1 perturbations in murine and human prostate cells. (A) Immunohistochemical images of prostate tissue in Ptenpc+/− Lkb1pc+/+ and Ptenpc+/− Lkb1pc−/− mice. Staining as indicated: p-AMPK (Thr172). (B) Representative immunohistochemical images of prostate tissue in Ptenpc+/− Lkb1pc+/+ and Ptenpc+/− Lkb1pc−/− mice. Staining as indicated: H&E, pS6 (Ser235/236), and p63. (C) Analysis of Krt5 gene expression by qRT-PCR in Ptenpc+/− Lkb1pc+/+ (n = 3), Ptenpc+/− Lkb1pc+/− (n = 3), and Ptenpc+/− Lkb1pc−/− (n = 3) mice. Data are normalized to Gapdh expression. (D) Gene set enrichment analysis of the squamous cell carcinoma signature in Ptenpc+/− Lkb1pc−/− (SCC-squamous cell carcinoma) and Ptenpc−/− Lkb1pc+/+ (PADC-prostate adenocarcinoma) mice. (E) Representative immunofluorescence images (left) of Ki67 (green) and CK5 (red), with quantification of double positive cells (right) in prostate tissue of Ptenpc−/− Lkb1pc+/+ and Ptenpc+/− Lkb1pc−/− mice. (F) Analysis of LKB1 gene expression by qRT-PCR in DU145 cells with ectopic expression of WT LKB1 (LKB1WT), the kinase-defective LKB1-K78I mutant (LKB1K78I), or transduced with mock vector (pBABE; n = 6; independent experiments). Data are normalized to control (pBABE; dashed line). (G) Effect of LKB1WTand LKB1K78I expression on cellular growth (n = 3; independent experiments). (H) Quanti- fication of invasion (C; n = 4, independent experiments) in LKB1WT, the LKB1K78I, or pBABE-transduced DU145 cells. Data are normalized to control (pBABE; dashed line). (I) Western blot analysis for LKB1 protein expression in harvested xenograft tumors (n = 3 samples per condition). (J) Metastatic capacity of control (pBABE) and LKB1WT- and LKB1K78I-expressing DU145 cells using intracardiac xenotransplant assays. Quantification (left) and representative ventral view (right) of luciferase-dependent signal intensity of DU145 cells captured on day 23. Statistic tests: one-tailed Mann–Whitney test (C, E, and J); one-sample t test (F and H); *, P < 0.05; **, P < 0.01; ****, P < 0.0001. Error bars represent SEM. Hermanova et al. Journal of Experimental Medicine S3 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Figure S3. Molecular analysis of LKB1 reexpression in human PCa cells. (A) Western blot analysis for LKB1 and MO25 protein expression of input and flag immunoprecipitates in DU145 cell line with ectopic expression of Flag-tagged LKB1WT and LKB1K78I. (B) Picture of SDS-PAGE gel with separation of im- munoprecipitated proteins. Blue squares indicate areas selected for subsequent LC-MS–based identification and quantitation. (C) Representative analysis of relative MO25α and MO25β mRNA abundance by qRT-PCR in DU145 cells (n = 3). (D) Representative analysis of MO25α gene expression by qRT-PCR in DU145 cells transduced with scramble shRNA (shCtrl) or two independent shRNA (shMO25α 02; shMO25α 04; n = 3). (E and F) Effect of MO25α gene silencing on cellular growth (E; n = 3; independent experiments) and anchorage independent growth (F; n = 5; independent experiments; data are normalized to control; shCtrl, dashed line). (G) Representative Western blot analysis using AMPK motif antibody in control and LKB1WT- and LKB1K78I-expressing cells (n = 3). (H) Representative Western blot analysis of LKB1, p-AMPK, p-ACC, p-Raptor, Raptor, and β-actin protein expression in control and LKB1WT- and LKB1K78I- expressing cells (n = 5). (I) Analysis of LKB1 gene expression by qRT-PCR in DU145 cells with ectopic expression of LKB1WT, LKB1K78I, and LKB1D194A or transduced with mock vector (pBABE; n = 3; independent experiments). Data are normalized to control (pBABE; dashed line). (J) Effect of LKB1WT, LKB1K78I, and LKB1D194A expression on cell growth (n = 3; independent experiments). (K) Western blot analysis using AMPK motif antibody in control and LKB1WT-, LKB1K78I-, and LKB1D194A-expressing DU145 cells. (L) Representative Western blot analysis for LKB1, p-AMPK, p-ACC, p-Raptor, Raptor, and HSP90 protein expression in control and LKB1WT-, LKB1K78I-, and LKB1D194A-expressing DU145 cells (n = 5). (M) Western blot analysis of LKB1 protein expression in harvested xenograft tumors (n = 5 samples per condition). (N) Effect of LKB1 reconstitution in DU145 cells with ectopic expression LKB1WT, the kinase-defective mutants (LKB1K78I, LKB1D194A) or infected with mock vector (pBABE) on tumor weight (left). Representative picture of harvested tumors (right) in xenotransplantation experiments (n = 20 injections per condition). (O) Quantification (left) and representative image (right) of Ki67 staining in harvested xenograft tumors in control and LKB1WT-, LKB1K78I-, and LKB1D194A-expressing DU145 cells. Statistic test: one-sample t test (I); one-tailed Mann–Whitney test (N and O). *, P < 0.05; ****, P < 0.0001. Error bars represent SD (C, E, and F) and SEM (I, J, N, and O). Hermanova et al. Journal of Experimental Medicine S4 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787 Tables S1–S10 are provided online as separate Excel files. Table S1 shows the incidence of metastasis in indicated tissues in Ptenpc+/− Lkb1pc−/− mice. Table S2 shows STK11 in primary tumors and metastasis (Armenia et al., 2018). Table S3 shows the incidence of primary prostate tumors and metastasis in patients with indicated LKB1 genomic status (Armenia et al., 2018). Table S4 shows the frequency of copy number alterations (Armenia et al., 2018). Table S5 shows the frequency of coexisting mutations with LKB1 deletion among the top genes with copy number aberrations in the metastatic specimens (Armenia et al., 2018). Table S6 shows the incidence of primary prostate tumors and metastasis in patients with indicated SIK3 genomic status (Armenia et al., 2018). Table S7 shows the incidence of primary prostate tumors and metastasis in patients with indicated LKB1 and SIK3 genomic status (Armenia et al., 2018). Table S8 lists differentially expressed genes in microdissected specimens from Ptenpc+/− Lkb1pc−/− (G2) versus Ptenpc−/− Lkb1pc+/+ mice (G5). Table S9 shows interactome analysis of LKB1 using Flag immunoprecipitation. Table S10 shows the Universal Probe Library (Roche) primers and probes used in this study. Hermanova et al. Journal of Experimental Medicine S5 Modeling lethal metastatic prostate cancer https://doi.org/10.1084/jem.20191787