SIRT1 regulates the phosphorylation and degradation of P27 by deacetylating CDK2 to promote T-cell acute lymphoblastic leukemia progression

RNA-seq data of 264 T-ALL, 267 B-ALL and 337 normal samples were downloaded from the ALL Phase 2 project of the TARGET database (https://​ocg.​cancer.​gov/​) and the TCGA TARGET GTEx cohort of UCSC Xena project (http://​xena.​ucsc.​edu/​). Box plot was drawn with R (http://​www.​r-project.​org/​).

Cell lines (CCRF-CEM, MOLT4, KG-1, THP-1. MV4–11, K562, U937 and 293 T) were purchased from the Type Culture Collection of the Chinese Academy of Sciences, Shanghai, China. The leukemia cell lines MOLT-4, CCRF-CEM, KG-1, THP-1. MV4–11, U937 were maintained in RPMI 1640 medium (Gibco, Cat No. 61870036) supplemented with 10% fetal bovine serum (FBS, Gibco, Cat No. 12483020) and 1% penicillin/streptomycin (Gibco, Cat No. 15140122). K562 cells were maintained in IMDM medium (Gibco, Cat No. 12440046) containing 10% FBS and 1% penicillin/streptomycin. The 293 T cell was maintained in DMEM medium (Gibco, Cat No. 10566016) containing 10% FBS and 1% penicillin/streptomycin. All cell lines were regularly tested for mycoplasma contamination. DAPT dissolved in DMSO was purchased from Selleck (Cat No. S2215). Cells were treated with a final concentration of 10 μM DAPT for 72 h, then washed and refed medium containing DAPT (mock) or medium lacking DAPT (wash) for 24 h. All human samples were conducted with approval from the ethical review committees of biomedical research of Shanghai Tongji Hospital. T-ALL patient sample information is in Supplementary Table 1.

MSCV-IRES-GFP, MSCV-ICN1-IRES-GFP and MSCV-ShRNA-ICN1-IRES-GFP were kindly provided by Professor Hudan Liu, Medical Research Institute, Wuhan University, Wuhan, China. pLVX-ShRNA1, pLVX-ShRNA2 and pLVX-IRES-ZsGreen vectors were purchased from Clontech Laboratories. Hemagglutinin (HA)-Ubiquitin, pMD2.G, psPAX2 and pCL-Eco were purchased from Addgene. pCMV-Blank, pCMV-N-MycTag and pCMV-N-Flag were purchased from Beyotime Biotechnology. A modified pLVX-ShRNA2-mCherry vector in which the ZsGreen fluorescent marker was replaced by an mCherry fluorescent marker was made using HiFi DNA Assembly Master Mix (NEB, Cat No. E2621L). ShSIRT1 and ShScramble were cloned into pLVX-ShRNA2-mCherry. ShScramble, ShMYC and Shp27-human were cloned into pLVX-ShRNA2. The SIRT1 coding region was cloned into pCMV-N-MycTag, pCMV-N-Flag and pLVX-IRES-ZsGreen. Mutant SIRT1-H363Y vector was generated using Q5® site-directed mutagenesis kit (NEB, Cat No. E0554). Sanger sequencing of the SIRT1-H363Y plasmid verified the presence of the desired mutation (Fig. S3a). CDK2, SKP2, MYC and mCherry were cloned into pCMV-Blank. CDK2 and p27 were cloned into pCMV-N-Flag. Shp27-mouse and ShRen were cloned into MSCV-ShRNA-ICN1-IRES-GFP.

The anti-SIRT1 ShRNA sequences used in this study were named ShSIRT1. A scrambled sequence was used as a control for off target effects of ShRNA. The sequences were as follows: ShSIRT1–1: TGGCCATTTCCCTACTTATAA, ShSIRT1–2: GAAGTGCCTCAGATATTAA, ShScramble: GCGCGCTTTGTAGGATTCG, ShMYC-1: CAGTTGAAACACAAACTTGAA, ShMYC-2: CCTGAGACAGATCAGCAACAA, Shp27-human: AGCAATGCGCAGGAATAAGG, Shp27-mouse: CGCAAGTGGAATTTCGACTTT and ShRen: AGGAATTATAATGCTTATCT.

All animals were housed in specific pathogen-free facilities at the animal experiment center of Tongji University. All the animal experiments were performed in accordance with institutional guidelines for animal care at Tongji University School of Medicine and received ethical approval from the Animal Ethics Committee of Tongji University.

SIRT1^co/co mice (Cat No. 008041) and Mx1-Cre transgenic mice (Cat No. 003556) were purchased from the Jackson Laboratory [21, 22]. Sexually mature SIRT1^co/co and Mx1-Cre mice were crossed to obtain Mx1-Cre SIRT1^+/+ or Mx1-Cre SIRT1^co/co mice. For inducible ablation of SIRT1, 6 weeks old Mx1-Cre SIRT1^co/co and Mx1-Cre SIRT1^+/+ mice were intraperitoneally injected with poly I:C (10 mg/kg, Sigma-Aldrich, Cat No. P1530) every other day for a total of five times. The mice were sacrificed 4 days after the last injection for collection of HSPCs. Genotyping analysis of tail DNA was performed utilizing Mouse SIRT1-co-gDNA primer pair. Genotyping analysis of bone marrow cDNA was performed using a forward primer located in the exon 3 and a reverse primer located in exon 5 (Mouse SIRT1-ko-cDNA primer pair).

Female C57BL/6 J mice (8 weeks old) were purchased from Shanghai SLAC Laboratory Animal Company and used as recipients in all mouse leukemia-model experiments.

B6.SJL (CD45.1) mice were provided by Professor Caiwen Duan at Shanghai Jiao Tong University School of Medicine (Shanghai, China) and were used as transplant recipients in a homing assay.

Cells were plated in 24-well plates in a two-layer soft agar system (0.6% agarose on the bottom and 0.3% agarose on the top) with 200 cells per well in a volume of 400 μl per well as previously described [32].

Cells were lysed with RIPA lysis and extraction buffer (Thermo Scientific, Cat No. 89901) with the phosphatase inhibitor cocktail (EpiZyme, Cat No. GRF102) and the protease inhibitor cocktail (EpiZyme, Cat No. GRF101) and boiled in 5X SDS-PAGE sample loading Buffer (Beyotime, Cat No. P0015). Total cellular proteins were resolved by SDS-PAGE, transferred to nitrocellulose membranes (Amersham, Cat No. 10600001), blocked with 5% fat free milk in TBS/0.05% Tween-20 and incubated with the indicated primary antibody overnight at 4 °C. Then, the membranes were probed with appropriate horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature. The protein bands were visualized with an enhanced chemiluminescent substrate (NCM biotech, Cat No. P10200) on Imager 600 machine (Amersham). For cycloheximide (CHX) assay, cells were treated with 100 μM CHX (Sigma, Cat No. 5087390001) for 2, 4 or 8 h prior to lysis. Antibodies used for immunoblotting are listed in Supplementary Table 2.

Total cellular RNA was extracted using TRIzol (Invitrogen, Cat No. 15596026) and reverse transcribed using a PrimeScript RT reagent kit (Takara, Cat No. RR037A) according to the manufacturer’s instructions. Quantitative PCR (Q-PCR) was performed using SYBR Premix Ex Taq II (Takara, Cat No. RR820A) on a LightCycler 96 PCR system (Roche Diagnostics). Relative mRNA expression was calculated by the comparative Ct method and the results were normalized to the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) results for each sample. The sequences of the PCR primers used are listed in Supplementary Table 3.

For lentiviral vector production, pLVX-ShRNA1, pLVX-ShRNA2, pLVX-IRES- ZsGreen and pLVX-IRES-mCherry vectors were used for plasmid construction and transfected into 293 T cells with packaging plasmids (pMD2.G and psPAX2). For retroviral vector production, retroviral vectors MSCV-IRES-GFP, MSCV-ICN1-IRES-GFP and a modified MSCV-IRES-GFP vector were used for plasmid construction and transfected into 293 T cells with a helper plasmid (pCL-Eco) [39]. Viral supernatants were generally harvested 2 days after transfection. T-ALL cells were infected by spinoculation (1000 g for 2 h at 32 °C, 4 μg/μl polybrene). After spinoculation, the cells were then supplemented with 3 ml fresh medium and cultured for an additional 48 h.

Cells were fixed with 4% paraformaldehyde, permeabilized with 0.2% Triton X-100 in PBS for 5 min and blocked with 3% bovine serum albumin (BSA) for 1 h. Cells were incubated with an anti-FLAG tag (D6W5B) rabbit monoclonal antibody (mAb) (Cell Signaling Technology, Cat No. #14793, 1:100) and anti-Myc tag (9B11) mouse mAb (Cell Signaling Technology, Cat No. #2276, 1:100) at 4 °C overnight, followed by treatment with a goat anti-rabbit IgG secondary antibody conjugated to Alex Fluor 488 (Invitrogen, Cat No. R37116, 1:200) and donkey anti-mouse IgG secondary antibody conjugated to cyanine cy5 (Jackson ImmunoResearch, Cat No. 715–175-150, 1:200).

To establish Notch1-induced T-ALL leukemia in mice, donor mice (weeks 6–8) were sacrificed and lineage-negative HSPCs were isolated with lineage cell depletion kit (Miltenyi Biotec, Cat No. 130–090-858). Retroviruses expressing MSCV-IRES-GFP (empty vector), MSCV-ICN1-IRES-GFP, or MSCV-Shp27-ICN1-IRES-GFP were transduced into HSPCs by spinoculation (1000 g for 2 h at 32 °C; 4 μg/μl polybrene). The cells were then cultured with RPMI 1640 medium supplemented with 20% FBS, 1% penicillin/streptomycin, IL-3 (PeproTech, Cat No. 213–13, 10 ng/ml), IL-6 (PeproTech, Cat No. 216–16, 10 ng/ml), and SCF (PeproTech, Cat No. 250–03, 100 ng/ml) overnight. A second round of spinoculation was performed 24 h later. After washing with PBS, cells were injected intravenously into lethally irradiated (9 Gy) recipients (1 × 10⁶ cells per mouse). The mice were maintained on antibiotic-treated water. The mice were bled every week to monitor blood counts and evaluate the percentage of GFP⁺ cells by flow cytometry. For the 2nd transplantation, a specific number of GFP⁺ cells sorted from the bone marrow of 1st transplantation recipient mice were injected intravenously into lethally irradiated (9 Gy) recipients together with 1 × 10⁶ unfractionated bone marrow cells for hemogenic support.

For examining homing ability of SIRT1 knock-out (KO) and wild-type (WT) bone marrow (BM) cells, a total of 5 × 10⁶ BM cells from KO and WT mice were transplanted into lethally irradiated recipients (CD45.1). 24 h after transplantation, donor-derived cells (CD45.2) in the BM were detected by FACS. For examining homing ability of SIRT1 KO and WT T-ALL cells, a total of 1 × 10⁶ leukemia cells (GFP⁺) from KO and WT T-ALL were transplanted into lethally irradiated recipients. 24 h after transplantation, GFP⁺ cells in the BM were detected by FACS.

To detect apoptosis, cells were washed with PBS and resuspended with Annexin V binding buffer at 1 × 10⁶ cells/ml. 100ul cell suspension was stained with 5 ul APC-labeled Annexin V (BioLegend, Cat No. 640932) at room temperature for 1 h, and 10ul propidium iodide (PI) solution was added 10 min before flow cytometric analysis. Bromodeoxyuridine (BrdU, MedChemExpress, Cat No. HY-D0184) and PI staining was performed to detect the cell cycle distribution. Cells (1 × 10⁶) were plated in 100 mm dishes with 10 μM BrdU for 1 h before harvesting. For detection of the cell cycle distribution in the T-ALL model, KO and WT mice received an initial intraperitoneal injection of BrdU (1 mg/6 g mouse weight) and were then maintained on 1.0 mg/ml BrdU in the drinking water for 24 h prior to sacrifice. Cells treated with BrdU were fixed in 75% ethanol at − 20 °C for at least 2 h and processed according to manufacturer’s instructions. An APC-labeled anti-BrdU antibody (BioLegend, Cat No. 364114) was used to detect BrdU, and a PI solution was added 10 min before flow cytometric analysis.

Cells for co-immunoprecipitation assay were lysed in Pierce IP lysis buffer (Thermo Scientific, Cat No. 87787) containing a protease inhibitor cocktail. For immunoprecipitation of FLAG or MycTag, the whole cell lysates were incubated with anti-Flag magnetic beads (Bimake, Cat No. B26101), anti-Myc magnetic beads (Bimake, Cat No. B26301) or Mouse IgG Magnetic beads (Cell Signaling Technology, Cat No. #5873) respectively overnight at 4 °C with rotation. After washing three times with PBST buffer, the immune complexes were boiled in SDS sample loading buffer and subjected to western blot analysis.

293 T cells were transduced with vectors as indicated with HA-tagged ubiquitin constructs for 48 h. Before harvesting, cells were treated with 10 μM MG132 for 24 h (MedChemExpress, Cat No. HY-13259). The cells were then lysed in Pierce IP lysis buffer containing a protease inhibitor cocktail and immunoprecipitated with anti-Flag magnetic beads. Ubiquitination was detected by using an anti-HA (F-7) antibody (Santa Cruz, Cat No. sc-7392).

SIRT1-GST, SIRT1-HIS, CDK2-GST, CDK2-HIS fusion protein was expressed in E coli. and purified using the GST-tagged fusion protein purification kit and BeyoGold™ His-tag Purification Resin and BeyoGold™ GST-tag Purification Resin according to the manufacturer’s instructions. Recombinant GST fusion proteins were incubated with HIS-tagged protein in vitro. The protein complexes were then pulled down with GST beads, eluted with SDS sample buffer, and resolved by SDS/PAGE.

We first studied SIRT1 expression in public data base and found that the level of SIRT1 mRNA expression was higher in T-ALL than in normal samples or B-ALL (Fig. 1a). Then we measured the protein levels of SIRT1 in different types of human cell lines by western blot and Q-PCR analysis. SIRT1 was highly expressed at both protein and mRNA levels in T-ALL cell lines MOLT-4 and CCRF-CEM (Fig. 1b-c). We then transduced murine HSPCs with the intracellular form of Notch1 (MSCV-ICN1-IRES-GFP) or the Vector control (MSCV-IRES-GFP) and transplanted into lethally irradiated mice (Fig. 1d). Recipients of ICN1 transduced HSPCs showed accumulation of GFP⁺ T cells with CD4⁺ CD8⁺ immunophenotype in peripheral blood (PB), organ infiltration and succumb to leukemia (Fig. 1e, S1a-c). As anticipated, ICN1 overexpression resulted in activation of the expression of the known Notch targets MYC and SKP2 at the mRNA and protein levels. Although the SIRT1 mRNA level was modestly elevated, the SIRT1 protein level was significantly increased in the T-ALL mouse model (Fig. 1f-h). This would suggest that the increased SIRT1 level was also due to a posttranscriptional mechanism.

A previous study suggested that MOLT-4 and CCRF-CEM cells have mutations within the Notch1 gene [42]. We then observed that the γ-secretase inhibitor (GSI) DAPT blocked Notch signaling induced SIRT1 downregulation at protein level. Removal of DAPT reversed the GSI-mediated inhibition of SIRT1 protein level but had no effect on mRNA level (Fig. 2a-d, S2 a-b). MYC has been proven to be a direct downstream target of Notch1 that contributes to the growth of T-ALL cells [43]. Previous studies have shown an association between MYC and increased SIRT1 protein expression [23, 30]. To evaluate the role of MYC in regulating SIRT1 expression, we inhibited MYC expression using MYC-specific ShRNAs (ShMYC-1 and ShMYC-2). MYC knockdown only reduced SIRT1 protein expression but not mRNA in MOLT-4 and CCRF-CEM cells (Fig. 2e-f, S2 c-d). Inhibition of MYC expression led to accelerated SIRT1 protein decreases in MOLT-4 and CC RF-CEM cells following CHX treatment (Fig. 2g-h). Furthermore, the SIRT1 protein decrease resulting from GSI treatment can be rescued by MYC overexpression, supporting that NOTCH regulates the protein levels of SIRT1 via MYC in T-ALL (Fig. S2e-f). 293 T cells were co-transfected with HA-ubiquitin, Flag-SIRT1, and MYC to evaluate the role of MYC in regulating SIRT1 ubiquitination and the result showed that overexpression of MYC inhibited the polyubiquitination of SIRT1 (Fig. 2i).

To investigate the effects of SIRT1 on T-ALL cells, two ShRNAs (ShSIRT1–1 and ShSIRT1–2) were designed to inhibit SIRT1 expression (Fig. 3a). MOLT-4 and CCRF-CEM cells with SIRT1 knockdown proliferated less than their parental counterparts and exhibited growth inhibition in a colony formation assay (Fig. 3b-c). Flow cytometry analysis of BrdU (S-phase) and PI (DNA content) showed arrested cell cycle entry (increased G0/G1-phase distribution, and decreased S-phase distribution) in ShSIRT1 MOLT-4 and CCRF-CEM cells compared to cells transduced with the ShRNA control (Fig. 3d). However, ShSIRT1 did not increase tumor cell apoptosis compared to the scramble control (Fig. 3e). ShSIRT1–1 was designed to target the untranslated region (UTR) region of the SIRT1 mRNA transcript, and therefore, it did not have any effect on a construct containing the coding region of SIRT1. It was reported that SIRT1-H363Y was a deacetylase-inactive point mutant of SIRT1 [8]. To exclude off-target effects of ShRNA, we overexpressed a SIRT1 coding region construct (SIRT1-CDS) or a SIRT1 deacetylase-mutant construct (SIRT1-H363Y) in T-ALL cell lines expressing ShSIRT1–1. The expression of SIRT1-CDS and SIRT1-H363Y maintained SIRT1 protein levels after ShSIRT1–1 transduction (Fig. 3f). Expression of SIRT1-CDS abrogated the ability of ShSIRT1 to inhibit T-ALL cell growth, while expression of SIRT1-H363Y did not produce this effect (Fig. 3g). SIRT1 is nicotinamide adenine dinucleotide (NAD+) dependent and is inhibited by nicotinamide. T-ALL cell viability was reduced by nicotinamide in a dose-dependent manner (Fig. S3b-c).

To directly assess the role of SIRT1 loss as a leukemia-initiating event in T-ALL transformation, we crossed a mouse line containing floxed alleles for SIRT1 exon 4 with a Mx1-Cre transgenic mouse line to delete SIRT1 exon 4 (Fig. S4a-d). HSPCs from SIRT1 KO (SIRT1^−/− Mx1-Cre) and WT (SIRT1^+/+ Mx1-Cre) mice were infected with retroviruses driving the expression of ICN1 and transplanted into lethally irradiated mice (Fig. 4a). SIRT1 KO T-ALL recipients exhibited lower GFP⁺ cell percentages and white blood cell counts than SIRT1 WT T-ALL recipients by peripheral blood monitoring (Fig. 4b, c). The frequencies of GFP⁺ cells in the bone marrow and spleen were also lower in the SIRT1 KO T-ALL group (Fig. 4d, e). Knocking out SIRT1 significantly extended the survival of T-ALL mouse model and reduced spleen and liver size (Fig. 4f-i). Similarly, hematoxylin and eosin (HE) staining indicated reduced infiltration of leukemia cells into the liver and spleen (Fig. 4j). Since the extent of leukemia cell maturation often correlates with prognosis, peripheral blood GFP⁺ cells were evaluated for surface markers [44]. Knocking out SIRT1 did not change the CD4⁺ CD8⁺ phenotype of leukemia cells (Fig. S4e). Consistently, SIRT1 KO T-ALL mice prolonged survival during the secondly transplantation (Fig. 4k). To quantify the leukemia- initiating cell (LIC) frequencies in SIRT1 KO or WT T-ALL, equivalent numbers of GFP⁺ cells sorted from primary leukemia model mice were transplanted into secondary recipients at a range of doses from 5 × 10³–2 × 10⁵ cells per recipient. Limiting dilution analysis showed that the deletion of SIRT1 resulted in decrease in the LIC frequency compared with the WT counterparts (1 in 43,997 GFP⁺ cells vs. 1 in 4753 GFP⁺ cells, Fig. 4l, m). To rule out the possibility that a failure of cell homing caused the reduction of T-ALL, we injected an equal number of SIRT1 WT or SIRT1 KO BM cells into lethally irradiated CD45.1 mice. FACS analysis showed that the percentages of SIRT1 WT and SIRT1 KO donor-derived cells identified by CD45.2 expression were similar 24 h after bone marrow transplantation (Fig. S4f). We transplanted equal numbers of SIRT1 WT T-ALL and SIRT1 KO T-ALL into recipients and the percentages of GFP⁺ cells in the bone marrow of SIRT1 WT T-ALL and SIRT1 KO T-ALL recipients were similar after 24 h (Fig. S2g). Consistently, BrdU-PI FACS analysis showed that SIRT1 loss blocked cell cycle entry (Fig. 4n, o).

To determine the potential target responsible for G1/S arrest induced by loss of SIRT1, multiple proteins associated with cell cycle entry, including MYC, SKP2, p16, p21, p27, cyclin E1, CDK4, CDK6, AKT and p-AKT were evaluated. The protein levels of p27 were markedly increased in ShSIRT1 transfected MOLT-4 and CCRF-CEM cells compared with control cells (Fig. 5a). Moreover, Q-PCR analysis demonstrated similar p27 mRNA levels in control and SIRT1 knockdown cells, suggesting that SIRT1 might decrease p27 protein levels but not the corresponding mRNA levels (Fig. 5b). As expected, the protein half-life of p27 was prolonged in SIRT1-specific ShRNA-transfected MOLT-4 and CCRF-CEM cells (Fig. 5c-d). Conversely, increased SIRT1 protein expression but not SIRT1-H363Y mutant reduced p27 protein levels (Fig. S5a-b). Western blot analysis with anti-SIRT1 (D1D7) rabbit monoclonal antibody (CST, #9475) and anti-SIRT1 (07–131) rabbit polyclonal antibody (Sigma-Aldrich, #07–131) revealed that SIRT1 KO T-ALL cells expressed the expected SIRT1 mutant protein with the in-frame deletion of exon 4 and that SIRT1 deficiency upregulated p27 protein expression but had little effect on p27 mRNA levels (Fig. 5e-f).

To confirm the importance of p27 in the inhibitory effects of ShSIRT1, Shp27 was used to knock down p27 expression in T-ALL cell lines expressing ShSIRT1–1 (Fig. 6a). The inhibitory effects of ShSIRT1 on proliferation were significantly reduced following p27 knockdown in MOLT-4 and CCRF-CEM cells (Fig. 6b). ShRNA-mediated knockdown of p27 significantly reduced SIRT1 knockdown-induced cell cycle arrest, as determined by BrdU-PI FACS analysis (Fig. 6c-d). We next evaluated the role of p27 in the SIRT1 KO T-ALL mouse model. A retroviral vector co-expressing Shp27 and ICN1 in one construct was used to transduce SIRT1 KO HSPCs (Fig. 5e). The efficiency of p27 knockdown was confirmed by western blot analysis (Fig. S6a). Moreover, SIRT1 KO did not decrease the peripheral blood or bone marrow leukemic burden in mice engrafted with p27-specific ShRNA-expressing T-ALL cells (Fig. 5f-h and S6b). Loss of SIRT1 did not induce cell cycle arrest in shp27 expressing T-ALL mouse model (Fig. 5i-j).

The Notch target SKP2 is a major ubiquitin ligase that controls the abundance of cell cycle regulatory proteins, such as p21, p27 and p57 [11]. The interaction between p27 and SKP2 was confirmed by co-immunoprecipitation experiments in CCRF-CEM cells (Fig. S7a-b). As shown in Fig. 7a, overexpression of SIRT1 increased the ubiquitination of p27 and the interaction between p27 and SKP2 compared to treatment with the empty vector or SIRT1HY vector. It has been reported that the degradation of p27 is promoted by its phosphorylation [31, 38]. We assessed the phosphorylation of p27 in MOLT-4 and CCRF-CEM cells and found that silencing SIRT1 with ShRNA decreased the phosphorylation of p27-Thr187 but did not affect p27-S10 (Fig. 7b-e, Fig. S7c-d). Expression of wild-type SIRT1 decreased Flag-p27 protein expression and increased the phosphorylation of p27-Thr187 but did not affect p27-Ser10 (Fig. S7c). Expression of wild-type SIRT1 did not affect the p27T187A mutant protein levels (Fig. S7e). However, SIRT1H363Y did not affect the p27 and p27T187A mutant protein levels (Fig. S7f). CHX experiments indicated that SIRT1 accelerated the Flag-p27 protein degradation but did not affect the p27T187A mutant protein degradation (Fig. 7f, g).

A previous study proved that CDK2 can directly phosphorylate p27 at Thr187 [38]. We co-transfected 293 T cells with a Flag-CDK2 vector, a Myctag-SIRT1 vector and a pCMV-mCherry vector. Immunofluorescence showed that both SIRT1 and CDK2 mainly co-localized in the nucleus (Fig. 8a). Co-transfection of vectors expressing Myctag-SIRT1 and Flag-CDK2 in 293 T cells revealed that SIRT1 directly interacts with CDK2 (Fig. S7g-h). Endogenous protein interaction between SIRT1 and CDK2 is detected in CCRF-CEM cells (Fig. 8b-c). In vitro GST pulldown assays also showed interaction between SIRT1 and CDK2 (Fig. 8d-e). Acetylation has been reported to downregulate CDK2-mediated phosphorylation, while deacetylation upregulates it [17, 28]. We found that SIRT1 but not the SIRT1-H363Y mutant decreased the acetylation level of CDK2 (Fig. 8f-g). The ubiquitination assay showed that overexpression of CDK2 increased the polyubiquitination of p27 (Fig. S7i). However, the polyubiquitination of Flag-p27 was only slightly increased in 293 T cells transfected with SIRT1 (Fig. S7j). This may be due to a relatively low endogenous expression of SKP2 in 293 T cells. We performed a western blot assay in blood cells from 3 healthy donor and 3 primary T-ALL patient samples. The result revealed that primary T-ALL cells harboring NOTCH1 mutations showed higher SIRT1 protein expression and lower p27 protein expression than normal blood cells from healthy donor (Fig. 8h).