A nu mouse model of diffuse large B-cell lymphoma in constitutional Atm loss

In this issue, Davies et al. present a new mouse model of B-cell lymphomagenesis in constitutional Atm loss.¹ ATM is a serine/threonine kinase that responds to DNA double stranded breaks by phosphorylating p53 and other targets to promote cell cycle arrest and/or apoptosis.² Biallelic loss of AT M in humans causes Ataxia Telangiectasia (A-T), a congenital disorder associated with high rates of cancer, most often diffuse large B-cell lymphoma (DLBCL).² Somatic mutations of ATM are also found in human B-cell lymphomas, especially mantle cell lymphoma, but occur at a lower frequency in sporadic DLBCL.^3,4 In mouse models, Atm^-/- alone leads to the early development of thymomas, while the complete abrogation of T cells in an ATMKO.CD3eKO mouse leads to the development of DLBCL-like lymphomas.^5,6 By crossing Atm^-/- mice with nude (nu^-/-) mice, which have incomplete loss of T cells and retain the ability to form germinal centers, Davies et al. were able to study B-cell lymphomas in a model system that more closely resembles the pathology of A-T in humans.

In order to understand the biology of DLBCL tumors arising in Atm^-/- nu^-/- mice, the authors first used RNA sequencing to show that most Atm^-/- nu^-/- DLBCL cluster with the activated B-cell-like (ABC) cell-of-origin (COO) subtype of sporadic human DLBCL, similar to the ATMKO.CD3EKO mouse model.⁵ Cytogenetic analysis identified several rearrangements in each tumor, consistent with genomic instability expected in Atm loss. Differential expression and gene set enrichment analysis comparing DLBCL tumors to splenic B cells from Atm^-/- nu^-/- mice identified several key pathways upregulated in the lymphomas, including cell cycle regulation, MYC targets, and DNA repair. Although the Myc transcript itself was not differentially expressed, MYC protein was significantly elevated relative to control non-tumor B cells, suggesting a non-transcriptional mechanism of MYC overexpression leading to MYC pathway activation in these lymphomas.

Next, the authors performed a genome-wide CRISPR screen in two Atm^-/- nu^-/- lymphoma cell lines with the highest MYC expression. In total, 197 genes were essential in both cell lines, and these were enriched for pathways involved in nucleotide metabolism and DNA replication and repair. Having demonstrated that several such pathways are both upregulated and essential in Atm^-/- nu^-/- lymphomas, the authors went on to explore the effects of four drugs targeting these pathways: a MYC inhibitor (MYCi361); an inhibitor of de novo purine synthesis (MMF); a WEE1 inhibitor (AZD1775), which inhibits dNTP synthesis by promoting the degradation of the ribonucleotide reductase subunit RRM2; and a NEDD8 inhibitor (pevonedistat) that stabilizes RRM2. MMF and AZD1775 displayed dose-dependent synergistic cytotoxicity caused by replication fork stalling and apoptosis in Atm^-/- nu^-/- lymphomas (Figure 1), and this synergy was partially rescued by either MYCi, dNTP supplementation, or NEDD8i. This finding replicated in several human DLBCL cell lines, with the most potent cytotoxicity observed in the cell lines with the highest MYC expression levels.

MYC is a transcription factor that drives the expression of genes in several cell proliferation and growth pathways, many of which overlap with upregulated pathways identified in Atm^-/- nu^-/- lymphomas.⁷ In normal germinal centers (GC), MYC is only transiently expressed at the initiation of GC formation and when GC B cells are transitioning from the GC light zone, where antigen selection takes place, to the dark zone, where surviving cells undergo rounds of proliferation and B-cell receptor hypermutation.⁷ Enhancer hijacking rearrangements that place MYC adjacent to an immunoglobulin locus or other highly expressed genes occur in 100% of Burkitt lymphomas (BL) and 15-20% of sporadic DLBCL, where it is associated with more aggressive disease.⁷ Although Davies et al. observed several rearrangements near the Myc gene in Atm^-/- nu^-/- lymphomas, none appeared to directly increase expression of the Myc transcript, alluding to a novel and still-unresolved mechanism underlying MYC dysregulation in these tumors.

Paradoxically, MYC also drives pro-apoptotic gene expression programs; therefore MYC rearrangements require cooperative survival-promoting alterations such as TCF3/ID3/CCND3 mutations in BL or BCL2 rearrangement in high-grade B-cell lymphoma double hit with MYC and BCL2 rearrangements (HGBCL-DH-BCL2).⁷ Notably, Atm loss-of-function is another established anti-apoptotic event that is permissive for MYC overexpression,⁸ therefore congenital Atm loss appears to enable excessive MYC activity and lymphomagenesis. Future studies could explore whether somatic AT M loss might similarly contribute to MYC activity in sporadic DLBCL, or whether Atm^-/- nu^-/- lymphomas remain dependent on Atm loss after transformation. In addition, genomic studies in DLBCL point to the existence of different founder mutations that lead to the formation of lymphomas representing distinct genetic subgroups.^9,10 Studying the mutational profile and evolutionary trajectory of Atm^-/- nu^-/- lymphomas through whole genome sequencing may provide new insights into the process of lymphomagenesis with Atm loss as a unique founder mutation.

Figure 1.Effects of MYC overexpression in Atm loss. Left panel: Atm loss is permissive for high MYC expression and activity, which drives high rates of proliferation dependent on continued dNTP synthesis. Without ATM, lymphoma cells do not undergo apoptosis in response to the replicative stress induced by high MYC activity. Right panel: when dNTP synthesis is inhibited, cells can no longer compensate for replicative stress and undergo apoptosis. DLBCL: diffuse large B-cell lymphoma.

A promising aspect of this study is the finding that nucleotide biosynthesis may be an underexplored vulnerability in A-T and MYC overexpressing lymphomas. Patients with A-T are uniquely sensitive to chemo- and radiotherapy due to the inherent genomic instability associated with ATM loss. In sporadic lymphoma, while patients with HGBCL-DH-BCL2 have inferior outcomes to R-CHOP chemoimmunotherapy, no biologically targeted therapies for front line treatment have yet been developed. Targeted therapies (including AZD1775/ adavosertib, which is under active clinical development) that exploit the synthetic lethality created by dNTP depletion with high MYC activity may be valuable tools for the treatment of the most aggressive and difficult to treat lymphomas.

Footnotes

• Received July 12, 2024
• Accepted July 30, 2024

Correspondence

L.K. Hilton lhilton@bccrc.ca

Disclosures

No conflicts of interest to disclose.