Nature Cell Biology 17, (2015)

Transcription

1 Correction notice Nature Cell Biology 17, (215) AMPK and PFKFB3 mediate glycolysis and survival in response to mitophagy during mitotic arrest Elena Doménech, Carolina Maestre, Lorena EstebanMartínez, David Partida, Rosa Pascual, Gonzalo FernándezMiranda, Esther Seco, Ramón CamposOlivas, Manuel Pérez, Diego Megias, Katherine Allen, Miguel López, Asish K. Saha, Guillermo Velasco, Eduardo Rial, Raúl Méndez, Patricia Boya, María SalazarRoa and Marcos Malumbres In the original version of this file posted online, detail was mistakenly omitted from Supplementary Figs 2, 3 and 6. These errors were corrected on 29 September 215.

2 S U P P L E M E N TA R Y I N F O R M AT I O N DOI: 1.138/ncb3231 4OHT Cdc2 sirna Scr. Casp3 Vinculin sirna Bak mrna 1. sirna Scr. Bcl2 Bcl2 Vinculin Casp3 b mrna rel. levels a sirna Scr. BclX BclX Vinculin Bax mrna d 12 h OA ZVAD 24 h Casp3 cleavage.5 sirna Scr. Casp 2 sirna Scr. Mcl1 Casp2 Mcl1 c αtubulin Scr. Bak Scr. Bax Bak/Bax wildtype Bak/Baxnull sirna Scr. Bim Bim e Percentage of cells 6 Annexin Annexin DAPI Annexin Annexin DAPI Apoptosis Necrosis 4 2 4OHT 4OHT Cdc2(/) Cdc2(Δ/Δ) CreERT2 CreERT2 TxpT TxpT MDAMB231 HeLa Supplementary Figure 1 Supplementary Figure 1 Analysis of apoptosis, necrosis and necroptosis. a) Depletion of Cdc2 after treatment of Cdc2(lox/lox); CreERT2 cells with 4hydroxytamoxifen (4OHT) (left top panel). Immunodetection of the indicated proteins after treatment with the indicated sirnas (rest of panels). Vinculin or atubulin were used as a loading control. b) Efficiency of the indicated sirnas by quantification of mrna levels 48 h after sirna nucleofection. Data were normalized against the levels of bactin transcripts. Data are mean ± SD and represent one out of at three independent experiments. c) Immunodetection of caspase 3 cleavage in Cdc2null cells treated with either ZVAD, Okadaic Acid (OA) or both, at the time points indicated. atubulin was used as loading control. d) Representative electron microscopy images of Bak/Baxnull and wildtype cells arrested in mitosis by the nocodazolemg132 treatment. Scale bars, 2 mm. e) Percentage of cells positive for Annexin V or DAPI in the indicated cultures. The left panel represents Cdc2null or control cells expressing CreERT2 in the presence of 4hydroxytamoxifen (4OHT). Right panel represents human cells arrested in mitosis with Taxol and protame (TxpT). Values are represented as mean ± SD (n=3 independent experiments). For (a,c) uncropped scans of the blots can be found in Supplementary Figure 7. Source data can be found in Supplementary Table

3 a (G2).5 h 2 h 4 h 1 Cdc2( / ) b MDAMB231 8 As h E64d 6 PA h 24 h 28 h 36 h LC Time from RO336 rel. (h) Mitotic cells (%) c f i Time to death in mitosis (h) Merge LC PP242 LC3 GFP Annexin V APC Percentage of cells h in mitosis Annexin LC3 Annexin Scr. Ulk1 Vps34Raptor Beclin1 LC3 sirna n= g n= h *** *** *** *** *** 8 ** *** *** * DMSO ZVAD3MA PP242 DMSO 3MA PP242 HeLa MDAMB231 MDAMB231 As h j Time to death in mitosis (h) Time to death in mitosis (h) d mrna levels (rel.units) 1..5 DMSO ZVAD LY AktIV PD n= ZVAD 3MA Atg5 KO Bak/Bax KO k ZVAD E64d/PA Cdc27 LC3 PARP αtubulin e BEZ235 PP242 Cdc27 LC3 Time to death in mitosis (h) n= * *** *** *** *** DMSO ZVAD PP 242 As. Cdc2( / ) 6 Tem. BEZ h Rap. Cdc2( / ) As h Supplementary Figure 2 Induction of autophagy during mitotic arrest. a) Mitotic entry in Cdc2null cells after release from RO336. Cdc2(lox/ lox) cells were treated with 4OHT to eliminate Cdc2 12 h prior to the addition of the Cdk1 inhibitor RO336 to trigger G2 arrest. This compound was washedout 18 h later allowing the accumulation of mitotic cells over time (left panel). Representative images are shown in the right panel. Scale bars, 5 mm. b) LC3 lipidation was analyzed in MDAMB 231 cells arrested by treatment with taxolprotame, in the presence or absence of E64D/ PA. Two different time exposures for LC3 are shown. c) Colocalization of LC3 and Annexin V (mean ± SD; n= 3 independent experiments) in Cdc2 null cells at the time points of mitotic arrest indicated in the graph. Images are representative for the double staining at 24 h. Scale bar, 1 mm. d) mrna levels (mean ± SD; n=3 independent experiments) of the indicated transcripts 48 h after sirna nucleofection. Data were normalized against the levels of bactin transcripts. e) Duration of mitosis (from metaphase until cell death) in Cdc2null cells treated with ZVAD, the indicated inducers of autophagy (mtor inhibitors PP242, temsirolimus, BEZ235), or transfected with sirnas against Raptor (Rap.). f) Duration of mitosis in Hela and MDAMB 231 cells synchronized by the taxolprotame protocol and treated with the indicated inhibitors. g) Duration of mitosis in Cdc2null cells treated with the indicated PI3K/Akt inhibitors. h) Immunoblot of LC3 after mtor inhibition with PP242 or BEZ235 in Cdc2null cells. Cdc27 bandshift (indicating phosphorylation) was used to confirm mitotic arrest. The samples were processed simultaneously and run in the same blot. i) Immunoblot of LC3 after mtor inhibition with PP242 in MDAMB231 cells synchronized by the taxol protame protocol. j) Duration of mitosis in Atg5null cells treated or not with the caspase inhibitor ZVAD (left) and Bax/Bak double mutant cells treated or not with the autophagy inhibitor 3MA (right), in both cases using the taxolprotame protocol. k) LC3 lipidation and PARP cleavage after treatment with caspase (ZVAD) and autophagy (E64d/PA) inhibitors in Cdc2null cells. Cdc27 phosphorylation (retarded motility) indicates mitotic cells. In (b,h,i,k), atubulin is included as a loading control. In (eg; j) dots represent individual cells and the mean are indicated by red lines. The number of cells analyzed (n) is indicated in each condition, and data are representative of three (e), two (f), six (g) or one (j) separate experiments. Green or red backgrounds indicate a significant delay or premature cell death in mitosis, respectively. *, p<.5; **, p<.1; ***, p<.1; Student s ttest. For (b, h, i, k) uncropped scans of the blots can be found in Supplementary Figure 7, and atubulin is included as a loading control. Source data can be found in Supplementary Table

4 Supplementary Figure 3 Autophagosomes and ROS accumulate during mitotic arrest. a) Representative profiles of mitotracker (left) and Dihydroethidium (DHE; right) in Cdc2null (green) or control (grey) cells expressing CreERT, at the indicated time points after the release from RO336 (all the cells were previously treated 18 hours with 4OHT). b) Immunodetection of the indicated mitochondrial proteins at 24 h after the release from RO336 (all the cells were previously treated 18 hours with 4OHT) in controlcreert2 (left) and Cdc2null cells treated with vehicle or the autophagy inhibitor 3MA. Uncropped scans of the blots can be found in Supplementary Figure 7. bactin was used as a loading control. c) Effect of 3MA and Cyclosporin A (CsA) on mitotracker levels in controlcreert2 (left panels) and Cdc2null cells 24 h after the release from RO336 (all the cells were previously treated 18 hours with vehicle or 4OHT). Representative FACS profiles are shown and quantification of mitotracker signal is represented in the bottom graphs. Data are mean ± SEM (n=3 independent experiments). d) Quantification of ROS (mean ± SEM; n=3 independent experiments) in Cdc2null cells as determined by the OxiSelect TM Intracellular ROS assay kit. The effect of NAC treatment in the accumulation of ROS (mean ± SEM; n=3 independent experiments) is also shown at the indicated times. e) Quantification of ROS (mean ± SEM; n=3 independent experiments) in Cdc2null cells or Cdc2(lox/lox) cells arrested in mitosis after 24 h with taxol and protame, as determined by the OxiSelect TM Intracellular ROS assay kit. f) Duration of mitosis until cell death in Cdc2null and HeLa cells treated with the antioxidant NAC or the caspase inhibitor ZVAD after synchronization with nocodazole and release in MG132. The mean is indicated by red bars and data represent two separate experiments.. In (d,e,f) the number of cells analyzed (n) is indicated in each condition. *, p<.5; **, p<.1; ***, p<.1 (Student s ttest). Source data can be found in Supplementary Table

5 a c AMPK activity (ratio CFP/YFP; rel units) e RO336 Release in taxol protame mrna levels (rel.units) G DG Scr. siα1 siα2 AMPK 36 h b Time (h) f OCR (re. levels) Cdc27 pampk AMPK Caspase 3 Scr. siα1 siα2 siα1 siα2 AMPK HeLa MDAMB231 As G As G d siscr. 2DG (n=23) siampk 2DG (n=22) siscr. (n=16) siampk (n=22) ECAR (rell levels) Scr. siα1 siα2 siα1 siα2 AMPK AMPK activity (ratio CFP/YFP; rel units) g Events 2 1 2DG Gluc. Mitosis Time (min) Cdc2(Δ/Δ) 24 h in mitosis Mitotracker siscr. siα1 siα2 siα1,2 Mitotracker (rel. units) ns Scr. siα1 siα2 n=38 n= n=31 * siα1,2 Supplementary Figure 4 Supplementary Figure 4 AMPK is reactivated during mitotic arrest. a) Schematic representation of the protocol followed for synchronization in human cells. Cells were treated with RO336 and this compound was washedout 18 h later allowing mitotic entry in the presence of taxol and pro TAME (an APC/C inhibitor) which mimic the arrest in metaphase imposed by Cdc2 ablation. b) The levels of the indicated antigens were analyzed by immunoblot in HeLa or MDAMB231 cells. atubulin was included as a loading control and uncropped scans of the blots can be found in Supplementary Figure 7. c) Changes in AMPK activity were monitored by using a FRET biosensor in Cdc2(lox/lox) cells in the absence of 4OHT. Upon treatment of these control cells with 2DG, a glucose analog, AMPK is activated giving an increase of about 2fold in biosensor signal (indicated in red in the images after 2DG addition; bottom panel). This increase is not observed in cells treated with sirnas against AMPK. Data are mean ± SD (n=4 independent experiments). d) AMPK activity was also monitored by the FRET biosensor in mitotic cells treated with vehicle (blue), 2DG (green) or Glucose (red). Data are represented as mean ± SD (n=3 independent experiments) after scoring the indicated number of mitotic cells in each condition. e) mrna levels of APMKa1 and AMPKa2 48 h after nucleofection of specific sirnas. Data are mean ± SD (n=3 independent experiments) normalized against bactin mrna levels. f) Levels of OCR (left) and ECAR (right) in Cdc2null cells arrested in mitosis for 24 h and transfected with sirnas against siampka1 (sia1), siampka2 (sia2) or the combination of both relative to Scrambled (Scr.) sequences. Data are mean ± SD (one experiment with 12 replicates). g) Effect of sirnas against AMPKa1 and/or AMPKa2 on mitotracker levels in Cdc2null cells 24 h after the release from RO336. Representative FACS profiles are shown on the left, and the quantification of mitotracker signal mean ± SEM (n=3 independent experiments) is represented in the histogram. In (g), ns, no significant; *, p<.5 (Student s ttest). Source data for all the figures can be found in Supplementary Table

6 a 13CGlc Cdc2(lox/lox) Asynchronous RO336 Release in 13CGlc Cdc2(lox/lox) G2 4OHT RO336 Release in 13CGlc Rel. from RO336 G Cdc2(Δ/Δ) Rel. from RO h c f Glucose uptake (mm) Cdc2(lox/lox) Cdc2(Δ/Δ) As. G h Release from G2 Alive cells (%) h Time from G2 release d Lactate (mm) Increase in metabolite concentration (24 h h; µm) Ile Cdc2(Δ/Δ) h Time from G2 release b Leu Val His Trp Phe Gln Amino acids g RO336 Release MDAMB231 G2 RO336 Release in taxol protame Tyr Thr Ala Cys Lys Pfkfb1/4 h Gapdh Pfkfb1 G Glu Gly Pyr Formate αketobutarate 2OH butyrate Hydroxysobutyrate Ketosolvalerate Ketonic bodies Alive cells (%) As. G2 Mitotic arrest Veh. Cord. Chx. Veh. Cord. Chx. Veh. Cord. Chx. siscr. sipfkfb1 Pfkfb3 siscr. sipfkfb3 e i 36 h Time to death in mitosis (h) Asynchronous Rel. from RO336 Rel. from RO336 (arrested in mitosis) MDAMB231 (cycling,mitotic) As. G h Release from G2 n=5 5 siscr. sipfkfb1 Supplementary Figure 5 Metabolic profile of mitotic arrested cells. a) Protocol for synchronized mitotic entry. Cdc2 (lox/lox) cells were untreated or treated with 4OHT to eliminate Cdc2 expression, 12 h prior the addition of the Cdk1 inhibitor RO336 to trigger G2 arrest. This compound was washedout 18 h later, allowing cells to progress through the cell cycle (Cdc2(lox/lox); blue) or arrest in mitosis in the absence of Cdc2 (Cdc2(D/D); red). Green lines indicate asynchronous cultures and these color codes are maintained throughout the figure to indicate the three different cell cultures. b) Similar protocol followed for synchronization in human MDAMB231 cells. c) Cell viability (TOPRO3 staining) assessed by flow cytometry in the different cultures represented in (a). d) Differential extracellular concentration of the indicated metabolites after 24 h of adding new medium, as monitored by NMR. 34 metabolites were analyzed and only those exhibiting significant differences are represented (23 in this figure glucose and lactate). No differences were found in carnosine, adenine, glyoxylate, cholesterol, EtOH, niacinamide, acetate, asparagine and aspartic acid. e) Cell viability (TOPRO3 staining) assessed by flow cytometry in the cultures represented in (b). f) Glucose uptake and concentration of extracellular lactate in asynchronous (green), normally release from G2 arrest (blue) or mitotic arrested (red) MDAMB231 cells, in the presence of 13 Clabelled glucose. Data in d) and f) are not normalized for death cells and the differences in mitoticarrested cells are likely underestimated. g) PFKFB1/4 protein levels in asynchronous (As.), G2arrested or mitotic Cdc2deficient cells in the presence or absence or cordycepin (Cord.) or cycloheximide (Chx.). GAPDH was used as a loading control. Images are representative of 2 independent experiments. h) Immunodetection of PFKFB1 and PFKFB3 48 hours after transfection. atubulin was included as a loading control. For panels (g,h), uncropped scans of the blots can be found in Supplementary Figure 7. i) Duration of mitosis (from mitotic entry until cell death) in Cdc2null cells transfected with sirnas against PFKFB1 or Scrambled sequences. Each dot indicates a single cells and red lines indicate mean (n=5 cells per condition; one independent experiment). ns, not significant differences. In (c,f) data are mean ± SD (n=3 independent experiments. Source data can be found in Supplementary Table

7 Supplementary Figure 6 Cooperation between taxol and inhibitors of mitotic survival pathways. Cell death was quantified by scoring ToPro3 levels by highthroughput microscopy after the addition of the indicated compounds in the presence or absence of taxol in MDAMB468, EVSAT or MCF7 breast cancer cells (left panels) or BJTERT2 and RPE nontransformed cells (right panels). Pink columns indicate significant cooperation when compared to both single inhibitors (black dotted comparisons) and taxol (red dotted comparisons). The number of cells analyzed (n) is indicated in each condition. Data are mean ± SEM (n=6 independent experiments). ***, p<.1; Student s ttest, and represent one out of 6 independent experiments. Source data can be found in Supplementary Table

8 Figure 2b Figure 4d 2 LC3bI/II 2 15 Survivin 5 αtub 2 Figure 3d 15 1 ph Tom2 1 Cdc LC3I LC3II Tim23 5 pampk 5 2 Mcl1 15 Tom4 1 PARP 5 5 Actin Supplementary Figure 7 Uncropped images of immunoblots. 7

9 Figure 5d 15 1 Figure 6a Pfkfb3 pser 5 Actin Figure 6d Pfkfb Gapdh Pfkfb3 Figure S1a 5 5 Cdc2 Casp3 Casp2 Bcl2 BclX Mcl1 Bim Bcl2 BclX Mcl1 Bim 15 1 Casp 2 Cdc2 Casp3 Vinculin Supplementary Figure 7 continued 8

10 Figure S1c Figure S2k 15 1 Cdc27 2 Caspase 3 1 LC3bI/II αtub 1 5 PARP αtub Figure S2b Figure S3b LC3bI/II LC3bI/II 15 5 Cdc2( / ) CreERT2 4OHT Tim23 αtub Figure S2h 15 Cdc2( / ) CreERT2 4OHT Tom Cdc Cdc2( / ) CreERT2 4OHT Actin 1 LC3bI/II Cdc2( / ) CreERT2 4OHT Tim23 αtub 2 Figure S2i LC3bI/II 2 5 Cdc2( / ) CreERT2 4OHT Tom2 5 αtub Cdc2( / ) CreERT2 4OHT Actin Supplementary Figure 7 continued 9

11 Figure S4b HeLa MDAMB Cdc27 pampk AMPK Caspase 3 Figure S5g Figure S5h Pfkfb1/4 5 Pfkfb1 Mw (KDa) Gapdh 5 63 tub. 48 Pfkfb tub. Supplementary Figure 7 continued 1

12 Inhibitor Target Concentration Source 2DG Hexokinase 52 mm Sigma 3MA Vps34 (PI3KCIII) 5 mm Sigma 3PO PFKFB3 11 µm Millipore ABT263 Bcl2 family Mcl1 1 nm ActiveBiochem ABT7 Bcl2 family 1 nm Selleckbio AKTIV Akt 2 µm Calbiochem Actinomycin A mitochondrial electron transport 1 µm Sigma BEZ235 mtorc1 and PI3K 2 nm Selleck CCCP H ionophore 1 µm Sigma Compound C AMPK 1 µm Calbiochem Cordycepin Polyadenylation 2 µg/ml Sigma Cycloheximide Protein synthesis 1 µg/ml Sigma Cyclosporin A Cyclophilin 5 µm Sigma E64d Cystein proteases 1 µg/ml Sigma LY2942 PI3K 2 µm Calbiochem MG132 Proteasome inhibitor 1 µm Sigma Necrostatine1 RIPK1 2 µm Enzo Nocodazol Microtubules 2 nm Sigma Oligomycin A F part of the HATPsynthase 21 µm Sigma Oxamate LDHA 1 mm Sigma PD9859 MEK1 and MEK2 2 µm Calbiochem Pepstatin A Aspartyl proteases 5 µg/ml Sigma protame APC/cyclosome 1 µm Boston Biochem PP242 mtorc1 1 nm Sigma Rapamycin mtorc1 5 nm Sigma RO336 Cdk1 11 µm Sigma Roscovitine Cdk1 1 µm Sigma SB2358 pp38mapk 2 µm Enzo Taxol Microtubules 27 nm Sigma Temsirolimus mtorc1 1 nm Sigma ZVADfmk Pan caspase inhibitor 2 µm Calbiochem Supplementary Table 1 Smallmolecule inhibitors and other chemicals used in this work. 11

13 Gene Orientation Sequence Bak Forward CATCTTGGGTCAGGTGGGTC Reverse TGGAACTCTGTGTCGTAGCG Bax Forward TCTCCGGCGAATTGGAGATG Reverse CGTGTCCACGTCAGCAATCA Beclin1(BECN1) Forward TTGAAACTCGCCAGGATGGT Reverse ATCAGATGCCTCCCCGATCA Drp1 Forward TCAGAAATGCTACTGGCCCC Reverse TCACGGGCAACCTTTTACGA PFKFB3 Forward CCGCGTACCATCTACCTGTG Reverse CAGAGCACTGGCAAACTTCTTG Raptor Forward CTGCTCGTGGCAAGTTTGTT Reverse CTGCTTACTGGGGTGCAGTT RIPK1 Forward ACATCAATGCAAAGCCCACG Reverse TGCAGATCACGAACTGCTCA RIPK3 Forward GTACGTTCGAAAGAGCCGGG Reverse ATCTTGTCACCAGAGCCTGC ULK1 Forward CCTGACTTCCTACAGCGGAG Reverse CTAGCCAACAGGGTCAGCAA Vps34 (PI3KCIII) Forward AGACTTCAGGCCTTGCTTGG Reverse CTTGACGCAAGTCGTCTCCA Supplementary Table 2 Oligonucleotides used in this work. 12

14 Antigen Catalogue number Clone number Source Ig Dilution Source αtubulin T926 DM1A Mouse 5 Sigma AMPK 32S Rabbit 1 Cell Signalling bactin A5441 Mouse 1 Sigma Bcl Rabbit 1 Cell Signalling BclX H12 Mouse 1 BD Pharmigen Bim Rabbit 1 Abcam Caspase 2 21 Rabbit 1 Abcam Caspase Rabbit 1 Cell Signalling Caspase 3, Active MAB835 Rabbit 1 RYD Systems Cdc Rabbit 5 Santa Cruz Cdc Mouse 1 Abcam CEPB AP Rabbit ProteinTech CEPB4 839 Rabbit 1 Abcam LC3b L43 Rabbit 1 Sigma Mcl Rabbit 1 Rockland Parkin Rabbit 1 Abcam PARP 9542 Rabbit 1 Cell Signalling PFKFB1/4 196 E16 Goat 5 Santa Cruz PFKFB Rabbit 1 Cell Signalling PhosphoAMPK(Thr172) 31 Rabbit 1 Cell Signalling PhosphoHistone 3 (ser1) 657 Rabbit 1 Millipore PhosphoSer 51 4A4 Rabbit 1 Millipore Survivin 521 Rabbit 1 Novus Biological Tim Rabbit 1 BD Bioscience Tom Rabbit 1 Santa Cruz Tom Rabbit 1 Santa Cruz Vinculin V9131 hvin1 Mouse 1 Sigma Supplementary Table 3 Antibodies used in this work for immunodetection of proteins. 13

15 Supplementary Table 4 Source data used for statistical analysis. 14