Tet2 is a tumor suppressor in the preleukemic phase of T-cell acute lymphoblastic leukemia

TO THE EDITOR:

T-cell acute lymphoblastic leukemia (T-ALL) is a rare but aggressive leukemia affecting the T-cell lineage. During the malignant transformation of immature T cells, a clonal expansion of progenitor cells is selected via the gradual accumulation of advantageous genetic mutations and epigenetic changes.¹ Alterations in the DNA methylome have been suggested as 1 of the potential initiating mechanisms that allow this gradual oncogenic transformation to occur.² During DNA methylation, a methyl group is covalently attached to a cytosine in the CpG-rich dinucleotide sequences by DNA methyltransferases.³ It is a plastic and reversible process in which active demethylation is catalyzed by the ten-eleven translocation oxidase enzymes.^4,5,Tet2 knockouts are not only predisposed to myeloid malignancies but are also prone to develop lymphoid malignancies, including T-ALL.^6,7 Although TET2 mutations in T-ALL are rare (0.7%), TET2 is transcriptionally repressed or silenced in 71% and 17% of T-ALL, respectively. This silencing is often associated with hypermethylation of the TET2 promoter.^8,9 Based on these observations, restoration of TET2 tumor suppressor function was proposed as a potential effective therapeutic strategy.

Long-lived preleukemic stem cells (pre-LSCs) have been uncovered as an initiating event in various blood-born cancers.¹⁰ Pre-LSCs are phenotypically normal but molecularly poised to transform. Their self-renewing capacity allows clonal expansion and subsequent acquisition of extra oncogenic driver mutations, eventually leading to the onset of full-blown leukemia. Of interest, these pre-LSCs have a competitive advantage compared with untransformed progenitors and are more chemo/radio-resistant.^11,12 As a result, they are considered to reflect a population that potentially can give rise to relapse.^13-15 In T-ALL, the existence of a pre-LSC population has first been described and identified in a spontaneous T-ALL mouse model overexpressing LMO2, CD2-Lmo2^tg/+ mice.^16-18 Unlike the thymus of normal mice, which is continuously replenished by progenitors from the bone marrow, the preleukemic thymus of CD2-Lmo2^tg/+ mice is self-sustaining from a young age and many months before leukemia development. Similar thymocyte transplantation experiments, to demonstrate the presence of a pre-LSC self-renewing thymocyte population, have not been reported for other spontaneous T-ALL mouse models. To investigate the presence or absence of pre-LSCs in an unrelated second T-ALL model, thymocytes from Lck-cre^tg/+Pten^fl/fl mice¹⁹ or their Cre-negative wild-type (WT) littermate controls were transplanted into sublethal irradiated syngeneic Ly5.1 (CD45.1⁺) recipients. CD2-Lmo2^tg/+ thymocytes and WT bone marrow cells were used as positive control. Using flow cytometric analysis, we analyzed how many donor vs recipient cells contributed to the repopulation of the thymus 6 weeks after transplantation. In contrast to CD45.2+ CD2-Lmo2^tg/+ thymocytes, no CD45.2⁺Lck-cre^tg/+Pten^fl/fl thymocytes were detected in recipient mice, demonstrating the absence of a preleukemic phase in the Lck-cre^tg/+Pten^fl/fl model (Figure 1A).

Currently, detailed molecular insights are lacking in how thymocytes can gain this aberrant self-renewal and whether altered DNA methylation plays a pivotal role in this initiation event. Of interest, in our previous study,²⁰ we already demonstrated that CD2-Lmo2^tg/+ and Lck-Cre^tg/+Pten^fl/fl murine T-ALLs have significantly different DNA methylation profiles. To investigate the potential role of Tet2 in pre-LSCs, we started with expression analysis of Tet2 in thymocytes isolated from preleukemic CD2-Lmo2^tg/+ and WT littermate mice at the age of 8 and 24 weeks. Tet2 messenger RNA (mRNA) levels were significantly downregulated after 8 weeks and were almost completely absent after 24 weeks (Figure 1B). A similar Tet2 mRNA expression analysis was performed in preleukemic Lck-Cre^tg/+Pten^fl/fl thymocytes. Because these mice develop T-ALL faster (supplemental Figure 1), we could only collect thymocytes from nonleukemic Lck-Cre^tg/+Pten^fl/fl mice at the age of 8 weeks. In contrast to CD2-Lmo2^tg/+ mice, no difference in Tet2 levels was observed between WT and Lck-Cre^tg/+Pten^fl/fl mice (Figure 1C). Based on these findings, we hypothesized that a gradual loss of Tet2 during preleukemic expansion of self-renewing CD2-Lmo2^tg/+ pre-LSCs may contribute to the progression of overt leukemia in this model.

To functionally investigate the Tet2 tumor suppressor role in T-ALL initiation, we developed a conditional Tet2 gain-of-function mouse model using our previously reported transgenic pipeline (Figure 2A).²¹ We crossed this newly developed R26-Tet2^tg with the CD2-iCre transgenic line,²² expressing Cre recombinase from the common lymphoid progenitor stage onward. In the resulting R26-Tet2^tg/tg;CD2-iCre mice, Cre-mediated deletion of the floxed transcriptional stop cassette enabled expression of a bicistronic mRNA encoding for TET2 and an enhanced green fluorescent protein (eGFP)-luciferase fusion reporter in lymphoid progenitors and their progeny (Figure 2A). Next, we crossed R26-Tet2^tg with the CD2-Lmo2^tg/+ T-ALL model¹⁸ and generated an aging cohort of CD2-Lmo2^tg/+ animals without (Cre^–,CD2-Lmo2^tg/+R26-Tet2^tg/tg, hereafter named Lmo2) and with Tet2 overexpression (CD2-iCre^tg/+CD2-Lmo2^tg/+R26-Tet2^tg/tg, hereafter named Lmo2-Tet2), to evaluate the effects of TET2 gain on T-ALL initiation. Mice were euthanized when signs of hematological malignancy were observed. Lmo2-Tet2 mice showed a significant delay in T-ALL formation (median survival of 266 days) compared with Lmo2 mice (median survival of 201 days), which confirmed the role of TET2 as a tumor suppressor in T-ALL (Figure 2B). Tet2 overexpression in Lmo2-Tet2 leukemia samples was confirmed by quantitative polymerase chain reaction (Figure 2C) and was in line with the gain of GFP (supplemental Figure 2A) and loss of the transcriptional stop cassette (supplemental Figure 2B). Luciferase activity (supplemental Figure 2C-D) confirmed conditional expression of the transgene transcript in Lmo2-Tet2 mice. Endogenous Tet2 mRNA levels did not differ between the groups (supplemental Figure 2E), indicating that the observed increase in T-ALL latency was exclusively caused by Rosa26-driven Tet2 overexpression.

Next, we wondered whether TET2 could act as a tumor suppressor in a mouse model that lacks pre-LSCs. For this, we crossed the R26-Tet2^tg/tg mice with the Lck-cre^tg/+Pten^fl/fl T-ALL model and generated an aging cohort of mice with (Lck-Cre^tg/+Pten^fl/lfR26-Tet2^tg/tg, hereafter named Pten-Tet2) or without Tet2 overexpression (Lck-Cre^tg/+Pten^fl/flR26-Tet2^wt/wt, hereafter named Pten). No significant difference in overall survival was observed (Figure 2D), although gain of Tet2 mRNA expression was confirmed (Figure 2E) in obtained leukemia samples. Similar to the LMO2 aging cohort, the increased Tet2 mRNA levels corresponded with the gain of the GFP-Luciferase reporter gene and loss of the STOP codon (supplemental Figure 3A-D) in the Pten-Tet2 mice. Endogenous Tet2 levels did not differ between the groups (supplemental Figure 3E).

Currently, the tumor suppressor role of TET2 in T-ALL has only been studied in knockout models and rescue experiments using vitamin C or 5-azacitidine.⁹ Here, for the first time, we investigated the effects of gain of TET2 in T-ALL using genetically overexpressing mouse models. We demonstrated a tumor suppressor role for TET2 in the preleukemic phase of T-ALL, whereas no effects were seen on T-ALL maintenance. The difference in the effect of Tet2 overexpression between the Pten and Lmo2 T-ALL mouse models can be correlated with our previous observations and reported differences in the DNA methylation profile of these 2 models.²⁰ The observed decrease in Tet2 mRNA levels in preleukemic Lmo2 mice, is associated with an increase in methylation at CpG islands, whereas this is not seen in the Pten model.²⁰ Based on these findings, we hypothesized that a critical TET2 threshold is necessary to protect preleukemic cells from obtaining secondary mutations that in the end will lead to the formation of T-ALL.

All in vivo experiments were approved by the ethical committee for animal experimentation of the Faculty of Medicine and Health Sciences of Ghent University.

Acknowledgments: The authors thank Utpal Davé for providing CD2-Lmo2^tg mice.

S.D.C. and J.R. were supported by a starting fellowship from Kom op tegen Kanker (Stand up to Cancer; the Flemish Cancer Society). The Goossens and Van Vlierberghe laboratories are supported by the Research Foundation Flanders (FWO-G049920N, SBO-S002322N), Ghent University start-up funds (STA/201909/007), and a Flanders interuniversity consortium grant (BOF23/IBF/042).

Contribution: S.D.C., J.R., D.J.C., S.G., and P.V.V. contributed to conceptualization; S.D.C., B.L., and S.T. contributed to the methodology; S.D.C. and J.R. performed the formal analysis; S.D.C., B.L., and S.T. contributed to the investigation; T.T., D.J.C., T.P., S.G., and P.V.V. contributed the resources; S.D.C. curated and visualized the data; S.D.C., J.R., T.P., S.G., and P.V.V. edited the original draft of the manuscript; S.D.C., J.R., B.L., S.T., T.T., D.J.C., T.P., and S.G. reviewed and edited the data; S.G. and P.V.V. contributed to supervision and funding acquisition; and S.D.C., S.G., and P.V.V. contributed to project administration.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Steven Goossens, Department Diagnostic Sciences, Ghent University, Corneel Heymanslaan 10, Medical Research Building II (entrance 38), Ghent B-9000, Belgium; email: steven.goossens@ugent.be.