Sarkar et al. SUPPLEMENTAL INFORMATION Figure S1 A 10 20 30 40 50 MADKRKLQGE IDRCLKKVSE GVEQFEDIWQ KLHNAANANQ KEKYEADLKK 60 70 80 90 100 EIKKLQRLRD QIKTWVASNE IKDKRQLIEN RKLIETQMER FKVVERETKT Predicted 110 120 130 140 150 coiled-coil KAYSKEGLGL AQKVDPAQKE KEEVGQWLTN TIDTLNMQVD QFESEVESLS domain 160 170 180 190 200 (1-241) VQTRKKKGDK DKQDRIEGLK RHIEKHRYHV RMLETILRML DNDSILVDAI 210 220 230 240 250 RKIKDDVEYY VDSSQDPDFE ENEFLYDDLD LEDIPQALVA TSPPSHSHME 260 270 280 290 300 DEIFNQSSST PTSTTSSSPI PPSPANCTTE NSEDDKKRGR ST* DSEVSQSP * 310 320 330 340 350 AKNGSKPVHS NQHPQSPAVP PTYPSGPPPT TSALSSTPGN NGASTPAAPT 360 370 380 390 400 SALGPKASPA PSHNSGTPAP YAQAVAPPNA SGPSNAQPRP PSAQPSGGSG 410 420 430 440 450 Intrinsically GGSGGSSSNS NSGTGGGAGK QNGATSYSSV VADSPAEVTL SSSGGSSASS disordered region 460 470 480 490 500 (242-605) QALGPTSGPH NPAPSTSKES STAAPSGAGN VASGSGNNSG GPSLLVPLPV 510 520 530 540 550 NPPSSPTPSF SEAKAAGTLL NGPPQFSTTP EIKAPEPLSS LKSMAERAAI 560 570 580 590 600 SSGIEDPVPT LHLTDRDIIL SSTSAPPTSS QPPLQLSEVN IPLSLGVCPL 610 620 630 640 650 GPVSLTKEQL YQQAMEEAAW HHMPHPSDSE RIRQYLPRNP CPTPPYHHQM 660 670 680 690 700 NAR/CS PPPHSDTVEF YQRLSTETLF FIFYYLEGTK AQYLAAKALK KQSWRFHTKY 710 720 730 740 750 NOT box (654-751) MMWFQRHEEP KTITDEFEQG TYIYFDYEKW GQRKKEGFTF EYRYLEDRDLQ B 2 Supplementary Figure 1: A. Sequence of the mouse CNOT3 protein (UNIPROT Q8K0V4). Vertical blue lines delineate the domains of the protein that are shown in Figure 1B. The green-shaded box indicates the the NOT box region, which is required for the interaction between CNOT3 and Aurora B (Figure 1C). Aurora B consensus phosphorylation sites are underlined. The red box and red text indicate the putative nuclear localization sequence. * indicates residues 292 and 294, which were mutated in this study. B. Phosphorylation sites detected on the CNOT3 protein by in vivo proteomic discovery mass pectrometry. Data was obtained from PhosphoSite® (Hornbeck et al., 2015). The data shown is for the human CNOT3 protein, which is 95% identical to mouse CNOT3. With the exception of one residue (E79), all of the residues that vary between human and mouse are located within the intrinsically disordered region between residues 329 and 604 of the mouse sequence. Blue circles indicate phosphorylated residues. The y- axis shows the number of records where the modification was assigned using proteomic discovery mass spectrometry. The numbered residues show a phosphorylation hotspot that extends from positions 291-299. 3 Figure S2 A B WT Clone 1 110bp ssODN Clone 2 Cnot3 Exon 10 C WT-DMSO WT-AZD1152 WT-AZD1152 24h 48h Counts Counts Counts PI PI PI D E G2/M CNOT3-DM S * ** NS G1 NS ** ** Counts * NS NS Percent of cells PI WT-DMSO F CNOT3-DM WT-AZD1152WT-AZD1152 24h 48h V5-CNOT3-WTV5-CNOT3-T292AV5-CNOT3-S294AV5-CNOT3-DMV5-CNOT3-WTV5-CNOT3-T292AV5-CNOT3-S294AV5-CNOT3-DM V5 (CNOT3) V5 (CNOT3) -TUBULIN LAMIN B Cytoplasm Nucleus HEK 293 cells 4 Supplementary Figure 2: Identification of Aurora B phosphorylation sites in the CNOT3 protein A. MS/MS fragmentation spectrum generated by ion trap collision-induced dissociation (CID) of doubly charged precursor ion with m/z 658.27 identifying CNOT3 tryptic peptide 291-302 (STDSEVSQSPAK). The spectrum presented two fragments which were as follows: Left panel: query number 2609 with assigned b- and y-ions; MASCOT score: -5 54; expectation value: 1.8x10 ; neutral loss of H3PO4 (-97.98 Da) from precursor ion detected at m/z 609.49; localisation probability for phosphorylation of Ser291: 3.0%, Thr292: 3.0%, Ser294 is 93.9%. Right panel: query number 2608 with assigned b- and y-ions; MASCOT score: 53; expectation value: 3.6x10-5; localisation probability for phosphorylation of Ser291: 25.6%, Thr292: 25.6%, Ser294: 48.4%. B. Left panel: Schematic representation showing use of CRISPR/Cas9 targeting of Cnot3 in exon 10 to mutate Threonine 292 and Serine 294 to alanine, using the guide RNA described in Methods and a 110 bp single-stranded donor oligonucleotides (ssODN) carrying the mutations. Right panel: Sequence images showing the wild-type (WT) sequence and the mutated sequences obtained from two CRISPR generated clones. C. FACS analysis of the cell cycle profiles of ESCs incubated for 24 and 48 hours with the Aurora B inhibitor AZD1152 or with vehicle (DMSO) and stained with propidium iodide (PI). D. FACS analysis of the cell cycle profile of Cnot3-DM ESCs. E. Histograms showing the percentage of cells measured at different phases of the cell cycle in (c) and (d). Mean ± SEM; P values (with respect to WT-DMSO) calculated by unpaired t-test, ns = non-significant, n=3. F. Representative immunoblot analysis of V5-tagged CNOT3 carried out on cytoplasmic and nuclear extracts from HEK 293 cells transfected with WT Cnot3, Cnot3-T292A, Cnot3-S294A, or Cnot3-DM expression constructs. Cells were harvested 48 hours post- transfection. a-Tubulin and Lamin B were used as loading controls for the cytoplasm and nucleus respectively. 5 Figure S3 A B DAPI OTX2 WT CNOT3-DM WT *** *** SURVIVAL RATIO SURVIVAL * BMP4+FGF2 + - - CNOT3-DM ACTIVIN A - + - WNT3 - - + C Mesoderm Endoderm Late endoderm and Ectoderm Pluripotency trophectoderm *** ns *** *** * *** * * ** ** *** (RELATIVE TO UNDIFF) TO (RELATIVE FOLD CHANGE IN mRNA (RELATIVE TO UNDIFF) TO (RELATIVE *** *** *** FOLD CHANGE IN mRNA (RELATIVE TO UNDIFF) TO (RELATIVE FOLD CHANGE IN mRNA Cdx2 Fgf5 Mixl1Gata4Gata6Sox17 PdgfrCxcr4 Tead4 Sox1 Oct-4 Nanog Sox2 Brachury Lineage marker expression in differentiated wild-type (WT) mesendodermal cells 4 days mesendoderm differentiation Supplementary Figure 3: Differentiation and survival of Cnot3-DM mesendoderm A. Representative confocal images of 8-day embryoid bodies (EBs) derived from wild- type (WT) and Cnot3-DM ESCs and subjected to immunostaining for the ectodermal marker OTX2 (green). Nuclei were counterstained with DAPI; scale bar: 100 µm. B. Survival of WT and Cnot3-DM ESCs over 4 days of differentiation induced in the presence of BMP4+FGF2 or Activin A or Wnt3 and the cell survival was determined using WST-1 reagent on day 4 of the differentiation. Histograms show survival ratios calculated relative to the values obtained for wild-type (WT) cells treated with BMP4+FGF2, which was assigned a value of 1. Mean ± SEM, significance calculated using unpaired t-test *P<0.05, ***P<0.001, n=3. C. Analysis of mRNA levels of lineage marker genes after BMP4+FGF2 induced differentiation of WT ESCs for 4 days. Values for the undifferentiated (undiff) ESCs were set at 1 as indicated by a horizontal line in each histogram. Cdx2 is a marker for late endoderm and trophectoderm. Tead4 is a marker for early trophectoderm. Mean ± SEM; n=3. P values are calculated by unpaired t-test, *P<0.05, **P<0.01, ***P<0.001, ns = non significant. 6 Figure S4 A DAPI BRACHURY GATA4 MERGED DAPI SMA WT WT CNOT3-DM CNOT3-DM 4 days mesendoderm differentiation B DAPI BRACHURY FOXA2 MERGED DAPI SMA WT CNOT3-DM 8 days mesendoderm differentiation DAPI BRACHURY GATA4 MERGED WT CNOT3-DM 8 days mesendoderm differentiation Supplementary Figure 4: CNOT3-T292/S294 phosphorylation promotes survival of differentiating mesendodermal cells. A. Staining for ME markers shows that differentiation of Cnot3-DM ESCs with BMP4 + FGF2 for 4 days generates reduced numbers of ME cells. Left panel: Representative IF analyses of staining of differentiated cells for the mesodermal marker Brachury (red) and endodermal marker GATA4 (green) carried out on ME cells differentiated from WT and Cnot3-DM ESCs by incubating them with BMP4 + FGF2. Right panel: staining for the additional mesodermal marker smooth muscle actin (SMA) (green). Nuclei were 7 stained with DAPI. Merged red and green images show the ME cells. Scale bars in (C) = 100 µm; scale bar in (D) = 60 µm. For IF analysis at 8 days of differentiation, see Figure S4A. B. Representative immunofluorescence analysis images of ME cells generated by inducing differentiation of WT and Cnot3-DM ESCs with BMP4 + FGF2 for 8 days (left panel: scale bar: 100 µm). Top left panel: staining for the mesodermal marker Brachury (red) and endodermal marker FOXA2 (green). Bottom panel: staining for Brachyury (red) and the endodermal marker GATA4. Merged images show dual staining of the mesendodermal cells. Top right panel: staining for the mesodermal marker a-SMA (scale bar: 60 µm). Nuclei were counterstained with DAPI. 8 Figure S5 A DAPI NESTIN DAPI -SMA DAPI FOXA2 + WT WT WT CNOT3-DM CNOT3-DM CNOT3-DM Ectoderm Mesoderm Endoderm ns ns *** SURVIVAL RATIO SURVIVAL SURVIVAL RATIO SURVIVAL SURVIVAL RATIO SURVIVAL WT WT WT CNOT3 CNOT3 CNOT3 -DM -DM -DM B CNOT3-DM C WT CNOT3-DM G2/M WT Clone 2 Clone 3 ns S G1 Counts Counts 8 days mesendoderm differentiation Percent of cells WT PI PI CNOT3-DM Supplementary Figure 5: Differentiation of Cnot3-DM ESCs into the three germ layers. A. Top panels: Representative confocal images of differentiation of wild-type (WT) and Cnot3-DM ESCs into ectoderm (3 days; immunostained for Nestin (green)), Activin A induced mesoderm (3 days; immunostained for a-SMA (green)) and FGF2+retinoic acid induced endoderm (3 days; immunostained for FOXA2 (green). Nuclei were counterstained with DAPI; scale bar: 12 µm. (See Methods). Bottom panels: Graphs represent the ratio of suriving Cnot3-DM cells versus WT cells after 3 days of differentiation into each lineage. Cell survival was measured using WST-1 reagent (see Methods). Mean ± SEM; (P values calculated by unpaired t-test, ***P<0.001. ns = non- significant, n=3). B. Crystal violet staining showing the efficiency of BMP4+FGF2 induced mesendoderm (ME) differentiation of wild-type (WT) ESCs and two additional clones of Cnot3-DM ESCs after 4 days of differentiation. 9 C. Cell cycle analysis of ME cells obtained by inducing differentiation of WT and Cnot3- DM ESCs with BMP4 + FGF2 for 4 days. Histograms represent the percentage of cells at different phases of the cell cycle.