Lats2 kinase potentiates Snail1 activity by promoting nuclear retention upon phosphorylation

Abstract

Snail1 is a central regulator of epithelial cell adhesion and movement in epithelial-to-mesenchymal transitions (EMTs) during embryo development; a process reactivated during cancer metastasis. While induction of Snail1 transcription precedes EMT induction, post-translational regulation of Snail1 is also critical for determining Snail1's protein level, subcellular localization, and capacity to induce EMT. To identify novel post-translational regulators of Snail1, we developed a live cell, bioluminescence-based screen. From a human kinome RNAi screen, we have identified Lats2 kinase as a novel regulator of Snail1 protein level, subcellular localization, and thus, activity. We show that Lats2 interacts with Snail1 and directly phosphorylates Snail1 at residue T203. This occurs in the nucleus and serves to retain Snail1 in the nucleus thereby enhancing its stability. Lats2 was found to positively influence cellular EMT and tumour cell invasion, in a Snail1-dependent manner. Indeed during TGFβ-induced EMT Lats2 is activated and Snail1 phosphorylated at T203. Analysis in mouse and zebrafish embryo development confirms that Lats2 acts as a positive modulator of Snail1 protein level and potentiates its in vivo EMT activity.

Figure 1

A screen for post-translational modifiers of Snail1 protein stability. (A) Stick figure representation of human Snail1-Clic Beetle Green (CBG) bioluminescent plasmid. NES: nuclear export signal region. (B) Bioluminescence and western blot of Snail1–CBG expressing HEK293 (clone #8) or untransfected control HEK293 cells, for the indicated proteins (right). (C) Bioluminescence of HEK293.Sn-CBG clone 8 transfected with Snail1 RNAi, GSK3β RNAi, or treated with the proteosome inhibitor MG132. Results are presented as bioluminescence relative to control untreated cells (set at an arbitrary value of 1.0). ^*Identifies the difference in Snail1 stabilizing effect between inhibition of GSK3β and inhibition of proteasome function. (D) Western blot analysis for the EMT marker E-cadherin (i.e., a Snail1 target gene) in HEK293.Sn-CBG clone 8 cells following transfection with control Luc RNAi, Snail1 RNAi, or GSK3β RNAi. (E) A human kinome RNAi screen (Qiagen) for proteins that stabilize or destabilize Snail1 protein level, as described in Materials and methods. Individual RNAi values are presented as Median Average Deviation (MAD) bioluminescence from the median of the complete library and Luc control RNAi. Control for maximum stabilization is MG1323 treatment (blue diamonds); for maximum destabilization is Snail RNAi (red circles); and for RNAi control is Luciferase RNAi (black triangles). The RNAi library results are in triplicate for each RNAi (orange diamonds). There were two RNAi's per kinase in the library. The blue broken lines identify ±3 MAD.

Figure 2

Presence of Lats2 protein stabilizes Snail1 protein level without affecting Snail1 transcription. Western blots (A) and RT–PCR analysis (B) for indicated proteins or mRNA in Lats2-depleted colon cancer HCT116 or mesenchymal HT1080 cells or Lats1-depleted HT1080 cells using two different shRNAi-containing lentiviruses against each or control luciferase shRNAi (shCTL). (C) Western blots for the indicated proteins in HCT116 and HT1080 cells transfected with wt or kinase-inactive (K765R) Lats2. (D) Western blot (left panel) or RT–PCR (right panel) for indicated protein or mRNA in Lats2+/+, +/−, and −/− MEFs and Lats2^−/− MEFs rescued by Lats2 re-expression. (E) Western blot comparing Snail1 protein level in indicated MEFs treated with the proteasome inhibitor MG132. (F) Snail1 protein stability in Lats2 wt (+/+) and null (−/−) MEFs. Protein translation was inhibited by pretreatment with cycloheximide (CHX). Snail1 protein level relative to control cells (arbitrarily set as 1.0) (A, C) or wt MEFs (D) is listed.

Figure 3

Lats2 phosphorylates Snail1 at T203 in vitro and in cells. (A) Peptide alignment of Snail1 from various organisms. Putative Lats2 phosphorylation sites are underlined. (B) Flag–Lats2 (lanes 1–5) or kinase-inactive Lats2 (K765R) (lane 6) was immunoprecipitated from transfected HEK293 cells, washed, and an in vitro kinase assay performed using purified GST (lane 1), GST-Snail1 (lane 2), GST-T203A.Snail1 (lane 3), GST-T177A.Snail1 (lane 4), GST-T177A; T203A.Snail1 (2TA, lane 5) and GST-Snail1 (lane 6) as exogenous substrate. Left panel: autoradiograph; right panel: Coomassie-stained gel. The left lower panel is a western blot for amount of immunoprecipitated Lats2. (C) A human anti-Snail1 pT203-containing peptide antibody (pT203.Snail1) specifically recognizes T203 phosphorylated Snail. Western blot for indicated proteins in HEK293 cells transfected with empty plasmid (CTL), wt Snail1, or T203A.Snail1. (D) HEK293 cells containing Snail1–CBG (clone #8) were infected with control (CTL: Luciferase siRNA oligos), Lats1, or Lats2 siRNA oligos. Extracts from a confluent plate of cells were immunoblotted (left panel) or RT–PCR performed (right panel) for indicated proteins or mRNA, respectively. (E) Lats2 associates with Snail1 in cells. Endogenous Lats2 was immunoprecipitated from HCT116 cells (upper panel) or HT1080 cells (lower panel) and bound products immunoblotted for Lats2 and Snail1. Input controls (5% of extract used for IP) are in the first lane.

Figure 4

Activation of Lats2, by mitotic stress or oncogenic stress, phosphorylates Snail1 at T203 and stabilizes total cellular Snail1 protein level. (A, B) HCT116 cells were transduced with control (CTL) or Lats2 shRNA lentiviruses and then untreated, treated with nocodazole (noco) or the proteasome inhibitor MG132. Western blot with the indicated antibodies was performed (A) or RT–PCR for the indicated mRNA was performed (B). (C) HCT116 cells were infected with H-RasV12 expressing lentivirus (+) or control empty lentivirus (−) and western blot performed with the indicated antibodies.

Figure 5

Activation of Lats2 by the Hippo pathway phosphorylates Snail1 on T203 and stabilizes total cellular Snail1 protein. (A) HEK293 cells were transfected with epitope-tagged plasmids expressing the indicated cDNAs (HA–Mst2, Flag–WW45, Flag–Lats2, and Flag–Snail1). Western blot (left panel) or RT–PCR (right panel) was performed with the indicated antibodies or for the indicated mRNA, respectively. (B) WT (+/+) or Lats2 null (−/−) MEFs were grown to a low density (L: <50% confluent) or high density (H: confluent) and western blot with the indicated antibodies performed.

Figure 6

Phosphorylation of Snail1 at T203 occurs in the nucleus and affects Snail1 nuclear retention. (A) HCT116 cells were untreated (CTL) or treated with nocodazole and cytosolic (C) and nuclear (N) fractions prepared. Western blot with the indicated antibodies was performed on each fraction. Lamin A/C served as a nuclear fraction control and β-tubulin served as a cytosolic control. The lower two panels are a repeat experiment of the upper two lanes but overexposed. (B) Immunofluorescent analyses of MCF10A cells transfected with the indicated Snail1 mutant protein fused to GFP. DAPI staining was used to identify nuclei. NLS: nuclear export signal. (C) Same cells as in (B) were treated (+) or not (−) with Leptomycin B to inhibit nuclear export. Cells were fractionated into nuclear (N) and cytosolic (C) fractions and each fraction western blotted for the indicated proteins. The relative distribution of Snail1 between the nucleus and cytosol is shown (%).

Figure 7

Lats2 influences Snail1-dependent cellular EMT. (A, B) MCF10A cells were transfected with an empty control vector (CTL), Snail1, Lats2, or Lats2 and an shRNAi against Snail1. Cells were grown to confluence and analysed for EMT changes. (A) Upper set of panels is phase images of cell morphology and lower panels are immunofluorescent analysis for E-cadherin and DAPI staining. (B) Western blot for the indicated markers of EMT: epithelial E-cadherin and mesenchymal Vimentin. (C) MCF10A cells were transduced with control (CTL: luciferase shRNAi), Snail1 or Lats2 shRNA lentiviruses. Following selection in puromycin, cells were grown to confluence and then treated with TGFβ1 (2 ng/ml) for 8 h, washed, fresh media without TGFβ added and cells cultured for 4 days. Western blot of extracts from cells at each of 4 days was performed using the indicated antibodies. (D) Immunofluorescence analysis of Smad2 subcellular distribution on the same set of cells as in (C) at day 0 (−) and day 4 after 8 h treatment with TGFβ (+). DAPI staining was used to identify nuclei. (E) Western blot analysis, with the indicated antibodies, of untreated confluent parental MCF10A cells (−) or parental MCF10A cells 4 days after treatment with TGFβ (+) to induce EMT.

Figure 8

Metastatic breast cancer cells contain increased amounts of Lats2 and in these cells Lats2 affects their invasive capacity. (A) Western blot analysis with the indicated antibodies of cell lysates from the non-transformed and non-tumourigenic human breast epithelial cell line MCF10A and two tumourigenic and metastatic human breast cancer cell lines MDA-MB-231 and BT549. (B) A Snail1 antibody western blot analysis of MDA-MB-231 cells transduced with lentiviruses expressing both a Snail1 shRNAi and various YFP-tagged, RNAi-resistant Snail1 mutants, as indicated. TA: T203A mutation; TE: T203E mutation; N or NLS: nuclear localization signal. (C, D) MDA-MB-231 cells as described in (B) were aggregated and placed in a 3D collagen I gel (2 mg/ml). Phase image from a representative imbedded cell aggregate shows the distance of invasion/migration of cells at 48 h. Magnified boxes show presence of cells leaving the aggregate. Results from multiple aggregates (@10) per well from multiple experiments (3) are quantified in (E) and depict the distance cells migrated centrifugally from the aggregates, relative to control cells arbitrarily set at 100%. Data are represented as mean±s.d. ^**Indicates P<0.01.

Figure 9

Snail1 rescues the Last2 morpholino defect in migration of the mesoderm during zebrafish embryo development. Zebrafish embryos at the one cell stage were injected with indicated morpholinos, mRNA, or combination of morpholino and mRNA as indicated above each panel. All were analysed at the end of gastrulation (tailbud stage). Prechordal plate (p.p.) migration was visualized by expression of hatching gland gene hgg1. The arrows indicate the distance occupied by the p.p. along the anteroposterior axis. (A) A diagram showing the shapes of the p.p. under different conditions is shown: WT: wild type; LOF: loss of function; GOF: gain of function. The phenotype associated with the specified injections is indicated below each panel (C–L). Human Snail1 is equivalent to zebrafish snail1b. Human Snail1 T203 is equivalent to zebrafish snail1b T196. MOC is a control, scrambled morpholino. TA: T196A; TE: T196E; NLS: nuclear localization signal. (M) Snail1b rescues the lats2MO phenotype. The percentage of phenotypes (LOF: no rescue; partial rescue; WT: rescue; GOF: gain-of-function) obtained after injection of the corresponding mRNAs with lats2MO, relative to embryos injected with lats2MO alone. Pictures (I, J, L) show representative embryos of each condition. Conditions: (I) lats2MO+snail1b WT mRNA n=85 (GOF: 7%; rescue: 31%; partial rescue 37%; no rescue: 25%). (J) lats2MO+TA-snail1b n=91 (GOF: 4%; rescue: 24%; partial rescue: 22%; no rescue: 50%). (L) lats2MO+NLS-TE-snail1b n=88 (GOF: 18%; rescue: 38%; partial rescue: 23%; no rescue: 21%). GOF phenotype: similar to snail1b overexpression (E). Rescue: similar to the wild-type condition (B). Partial rescue: intermediate phenotype between lats2MO (D) and the wild-type condition (B). No rescue: similar to lats2MO (D).