Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Nov;7(11):1336-1353.
doi: 10.1158/2159-8290.CD-17-0267. Epub 2017 Oct 3.

TOX Regulates Growth, DNA Repair, and Genomic Instability in T-cell Acute Lymphoblastic Leukemia

Free PMC article

TOX Regulates Growth, DNA Repair, and Genomic Instability in T-cell Acute Lymphoblastic Leukemia

Riadh Lobbardi et al. Cancer Discov. .
Free PMC article


T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive malignancy of thymocytes. Using a transgenic screen in zebrafish, thymocyte selection-associated high mobility group box protein (TOX) was uncovered as a collaborating oncogenic driver that accelerated T-ALL onset by expanding the initiating pool of transformed clones and elevating genomic instability. TOX is highly expressed in a majority of human T-ALL and is required for proliferation and continued xenograft growth in mice. Using a wide array of functional analyses, we uncovered that TOX binds directly to KU70/80 and suppresses recruitment of this complex to DNA breaks to inhibit nonhomologous end joining (NHEJ) repair. Impaired NHEJ is well known to cause genomic instability, including development of T-cell malignancies in KU70- and KU80-deficient mice. Collectively, our work has uncovered important roles for TOX in regulating NHEJ by elevating genomic instability during leukemia initiation and sustaining leukemic cell proliferation following transformation.Significance: TOX is an HMG box-containing protein that has important roles in T-ALL initiation and maintenance. TOX inhibits the recruitment of KU70/KU80 to DNA breaks, thereby inhibiting NHEJ repair. Thus, TOX is likely a dominant oncogenic driver in a large fraction of human T-ALL and enhances genomic instability. Cancer Discov; 7(11); 1336-53. ©2017 AACR.This article is highlighted in the In This Issue feature, p. 1201.

Conflict of interest statement

The authors have no disclosures or conflicts of interest.


Figure 1
Figure 1. TOX collaborates with MYC to accelerate T-ALL onset
(a) Schematic of transgenic screen. (b) Kaplan-Meier analysis (p<0.05, Log-Rank Statistic). Leukemic fish have >50% of their body overtaken by T-ALL cells. Number of animals analyzed per genotype is shown in parenthesis. (c) Images of T-ALL transgenic fish at 21, 28 and 35 days post-fertilization (dpf). Asterisks denote auto-fluorescence. Arrowheads show leukemias initiating in the thymus. Scale bar equals 2mm. (d) EdU proliferation analysis of zebrafish T-ALL. Not significant (NS). +/− STD noted. (e) Annexin-V apoptosis staining as assessed by flow cytometry. Asterisk denotes p<0.05, Two-tailed Student’s t-test. +/− STD noted. ≥9 primary T-ALLs were analyzed per genotype in d,e. (f) Cytospin showing lymphoblast morphology (n>5 leukemias/genotype analyzed). Scale bar is 20μm. (g) RT-qPCR gene expression comparing MYC and MYC+TOX expressing T-ALL (n=5 per genotype, run in triplicate) with sorted thymocytes isolated from rag2:GFP transgenic fish (purity >95%, viability >95%). *: p<0.05; ***: p<0.001, Two-tailed Student’s t-test. Not significant (NS). Error bars denote standard error of the mean. (h) Number of TCRβ clones per primary leukemia (MYC+TOX: n=9 animals; MYC: n=24 animals). Asterisk indicates p<0.05, Two-tailed Student’s t-test. Error bars denote standard error of the mean.
Figure 2
Figure 2. TOX promotes genomic instability
(a) Flow cytometry analysis showing DNA content of zebrafish T-ALL (blue) compared to normal blood-derived DNA (red). Representative leukemias are shown with changes in DNA index noted (DI). 2n has a DNA index of 1. (b) Quantization of DNA content. Asterisk denotes p=0.013, Fisher Exact Test. (c) Genomic DNA alterations identified by whole genome sequencing. The log-log scatter plot represents read counts within non-overlapping window size of 10 kb across the genome comparing leukemia and control tissue from the same, representative animal. Amplifications found in the MYC+TOX leukemia are denoted by the green circle. (d) Manhattan plot representing the copy number variation across the genome of a representative MYC (top) and MYC+TOX (bottom) expressing T-ALL. Regions of significant gain, amplification (AMPL), high level amplification (HLAMPL), and neutral 2N copy number (NEUT) noted. (e-f) Analysis of metaphase spreads from MEF cells infected with control (MOCK), full-length Wild-type TOX (TOX WT), or TOX that lacks the HMG box domain (TOX dHMG). Arrows denote chromosome abnormalities. (f) Quantification of cells with genomic abnormalities. >50 nuclei were counted per condition and replicated three independent times. Error bars denote standard deviation. *, p<0.05 and **, p<0.005, Student’s t-test. Not significant (NS).
Figure 3
Figure 3. TOX is highly expressed in human T-ALL and transcriptionally regulated in a subset of leukemias by the TAL1/MYB complex
(a) TOX microarray gene expression from the Cancer Cell Line Encyclopedia. Box-and-whisker plots denote median TOX expression (black line), the inter-quartile range (box) and 1.5× the inter-quartile range (bars). (b) Volcano plot comparing gene expression between human T-ALL patient samples and bone marrow. (c) Western blot analysis of primary human leukemia (top) and T-ALL cell lines (bottom). The same thymus sample was run in lane 1, top and lane 2, bottom. (d) Flow cytometry analysis showing TOX expression in primary human T-ALL compared with B-ALL. Control is T-ALL cells stained with conjugated IgG control antibody (ctr). (e) Quantification showing the percentage of TOX+ cells found in each patient sample (right panel, **, p=0.006, Two-tailed Student’s t-test). Mean and standard error of the mean are denoted. (f-h) TOX is associated with two distinct, H3K27 acetylated super-enhancers (SE) in human T-ALL. (f) ChIP sequencing of human T cells, CD34+ progenitor cells from the marrow (black peaks), and T-ALL cell lines (red peaks). Super-enhancer 1 (SE 1) is found in MOLT-3 and Jurkat, while SE 2 is found in all three human T-ALL cell lines. (g) Magnified view of SE 2 in Jurkat T-ALL cells showing super-enhancer occupancy by the TAL1/MYB complex. (h) TOX gene expression in Jurkat cells following knockdown of T-ALL transcription factors found within the H3K27 acetylated super-enhancer (***, p<0.001; **, p<0.01; *, p<0.05, data set from GSE29179; +/− standard error of the mean denoted).\
Figure 4
Figure 4. TOX is required for continued human T-ALL growth and maintenance
(a) Western blot analysis following stable shRNA knockdown in human T-ALL cells. Percent knockdown noted. (b) Cell viability following knockdown as assessed by Cell TiterGlo. (c) EdU proliferation analysis. (d) Annexin-V staining. Asterisks in b-d denote p<0.05, Student’s t-test. Xenograft studies performed with human HPB-ALL (e-g) and CCRF-CEM (h-j). (e,h) Flow cytometry analysis showing efficiency of TOX knockdown. (f,i) Luciferase bioluminescent imaging of representative animals engrafted at 0 days compared with 21 days. Scramble shRNA control (left flank) or shRNA-TOX #2 (right flank). (g,j) Quantification of xenograft growth using two independent shRNAs. Not significant (NS), p<0.05 (*), p<0.01 (**), and p<0.001 (***), ANOVA test.
Figure 5
Figure 5. TOX binds directly to the KU70/KU80 and inhibits Non-Homologous End Joining by suppressing recruitment of KU70/KU80 to sites of DNA damage
(a) TOX immunoprecipitation followed by Western blot analysis in HPB-ALL cells in the presence or absence of DNAseI treatment. The right panel shows the reciprocal immunoprecipitation using anti-KU70. (b) TOX immunoprecipitation followed by Western blot analysis for members of the NHEJ pathway. (c) Western blot analysis following cell fractionation in non-irradiated and 3Gy-irradiated HPB-ALL cells. TUBULIN, SP1 and Histone H4 (H4) are controls for assessing cytoplasmic, nucleoplasmic and chromatin fractions, respectively. (d) In vitro binding of TOX (WT) or HMG-box (ΔHMG) deletion mutant with KU70/KU80 followed by Western blot analysis. Negative control was KU70/KU80 added to beads and then purified in the absence of TOX (Neg. Ctr). Loading control contains only purified KU70/KU80 proteins (Load Ctr). (e) Western Blot analysis of 3T3 cells following transfection with full-length TOX (+), ΔHMG (ΔH), or ΔNLS (ΔN). Below is the quantification of the NHEJ assay (NS: Not significant; ***, p<0.0001, Two-tailed Student’s t-test comparing experimental samples with control, +/− STD noted). (f) Western Blot analysis of 3T3 cells following transfection with full-length TOX, KU70 and/or KU80. Below is shown the quantification of the NHEJ assay (**, p<0.001, ***, p<0.0001, Two-tailed Student’s t-test, +/− STD noted). (g) Confocal imaging of 3T3 cells transfected with mRuby2-fused constructs. Dashed lines denote nucleus as assessed by Hoechst stain (n>100 cells/construct analyzed). The lower right panel shows the fluorescence-based NHEJ assay completed in 3T3 (*, p=0.03, Two-tailed Student’s t-test). +/− STD noted. (h,i) Quantitative assessment of recruitment of KU80-GFP (h) or XRCC4-GFP (i) to sites of UV laser-induced DNA damage in 3T3 cells. Asterisks denote significant differences (p<0.05, Two-tailed Student’s t-test). +/− STD noted. Representative fluorescent images of cells following laser-induced damage are shown to the right.
Figure 6
Figure 6. TOX loss-of-function increases DNA repair and accelerates the time to break resolution in human T-ALL cells
(a–f) Transient repair assays in HPB-ALL (a–c) and CCRF-CEM (d–f) cells that have shRNA knockdown. (a,d) Western blot analysis with percent knockdown noted. (b,e) Flow cytometry analysis of knockdown and control cells following transient fluorescence-based NHEJ assay. (c,f) Graphical summary of data. *, p<0.05; ***, p<0.001 when compared with control shRNA treated cells, Two-tailed Student’s t-test. (g) Western blot analysis of TOX expression in CCRF-CEM cells with stable TOX shRNA #2 or scramble control shRNA knockdown used in the Traffic Light Reporter assay. (h) Flow cytometry analysis of CCRF-CEM cells stably integrated with the Traffic Light Reporter. Cells were nucleofected with Tamoxifen-inducible I-SceI meganuclease and induced as noted in the figure. Percent of cells that undergo NHEJ repair and express mCherry are noted (+/−1STD). ***, p=0.0001, Two-tailed Student’s t-test. (i,j) Kinetics of the DNA repair in CCRF-CEM control and stable knockdown down T-ALL cells (TOX shRNA #2). The number of 53BP1 (i) and γH2A.X (j) foci per nucleus following 3Gγ irradiation are denoted. Each point represents data from a single cell and the black bars denote the median foci per cell (***, p<0.0001, Two-tailed Student’s t-test, >150 cells counted/condition). Box-and-whisker plots denote expression with the median 75% of samples and bars 90%.

Similar articles

See all similar articles

Cited by 11 articles

See all "Cited by" articles

MeSH terms


LinkOut - more resources