Type I interferons: leukemia's old foe in the limelight again

In this issue of Haematologica, Smeets et al.¹ shed new light on the regulation of the classical immunomodulatory cytokine family, type I interferons (IFN-I), in the ETV6-RUNX1 subgroup of acute lymphoblastic leukemia (ALL). Discovered in 1957 by Isaacs and Lindenmann, IFN-I gained prominence as critical regulators of antiviral innate immune responses. IFN-I production in response to viral infection or DNA damage induces the expression of IFN-stimulated genes (ISG). These ISG mediate the anti-proliferative, pro-apoptotic, and pro-inflammatory functions of IFN-I, a process that ultimately results in the elimination of the infected and/or damaged cells either directly or by host immune cells.² Immune responses in cancers mimic those seen during viral infections.² Consistent with this, Dunn and colleagues made the landmark discovery that IFN-I restrict solid tumorigenesis by activating both innate and adaptive arms of anticancer host immune defenses.³ IFN-I are thus widely regarded as ‘anticancer’ cytokines. IFN-α2, a member of the IFN-I family, was the first immunotherapeutic agent to be approved by the US Food and Drug Administration in 1986 for the treatment of hairy cell leukemia.⁴ The remarkable response rates observed in patients with hairy cell leukemia led to the expanded use of this cytokine in the treatment of other hematopoietic malignancies, including ALL.⁵ In ALL, IFN-I were shown to increase relapse-free survival in patients who received an allogeneic bone marrow hematopoietic stem cell transplant.⁵ Despite the remarkable improvements in clinical outcomes of patients treated with IFN-I, IFN-I gradually lost their charm as ‘wonder drugs’ due to off-target toxicity associated with their administration.² Nevertheless, their strong anticancer function could never be refuted. IFN-I were unfortunately used in clinics at a time when their mode of therapeutic action was not completely understood. However, recently identified mechanisms of action of IFN-I in human ALL by Smeets et al.¹ and us⁶ rekindle the interest of the scientific community in harnessing the therapeutic potential of this age-old cytokine family in treating ALL. In 2005, Einav et al.⁷ discovered that patients with the most common and treatable form of childhood B-cell precursor (BCP) ALL, the ETV6-RUNX1⁺ subtype, were approximately three times more likely to exhibit an enhanced expression of ISG in comparison to children with high-risk BCP ALL subtypes such as those with MLL, E2A-PBX1, or BCR-ABL1 rearrangements and hypodiploidy. However, they did not delve into why ETV6-RUNX1 BCP ALL were associated with interferonopathy, which cells in the leukemia microenvironment cause such interferonopathy, and which class(es) of IFN were induced.

In the current issue of Haematologica, Smeets et al.¹ answer some questions arising from the publication by Einav and colleagues.⁷ They found that in pediatric patients with BCP ALL, bone marrow mesenchymal stromal cells (MSC) are a critical source of interferons, specifically the IFN-I α and b. Among childhood BCP ALL subtypes, they found that ETV6-RUNX1⁺ ALL most profoundly induces ISG in co-cultured bone marrow MSC derived from healthy donors and from children with ETV6-RUNX1⁺ and other ALL (hyperdiploid and those with DUX4, CRLF2, and EPOR translocations). They show that induction of ISG in MSC co-cultured with ETV6-RUNX1⁺ ALL cells occurs partly by direct contact between leukemic cells and MSC via tunneling nanotubes (Figure 1). Overall, the findings of Smeets and colleagues suggest that ETV6-RUNX1 could be a direct inducer of the IFN-I pathway in normal and ALL patient-derived MSC. Mechanistically how expression of ETV6-RUNX1 in leukemic blasts triggers the IFN-I signature in surrounding non-malignant MSC remains to be delineated. Together, the studies by Einav et al. and Smeets et al. raise interest in studying these biological mechanism(s).

Smeets and colleagues also observed that IFN-I from MSC co-cultured with ETV6-RUNX1⁺ BCP ALL cells do not directly impact the viability of the leukemic cells or their sensitivity to chemotherapeutic agents.¹ This observation is consistent with Dunn et al.’s seminal finding that only cells of the hematopoietic lineage mediate the anticancer effects of IFN-I. Dunn et al. made their discovery in a solid tumor model of fibrosarcoma that is very different from leukemia in which malignancy arises in the hematopoietic cells themselves.³ Studying the role of IFN-I in BCP ALL, we showed that IFN-I mediate their anti-leukemic effects indirectly by activating host immune defenses. We found that IFN-I enhance the production and maturation of the non-malignant, innate immune cytotoxic natural killer cells in the ALL microenvironment by stimulating the production of interleukin-15, the IFN-I-induced cytokine critical for natural killer-cell homeostasis.⁶ We found significantly higher expression of interleukin-15 in patients with ETV6-RUNX1⁺ BCP ALL than in higher risk ALL subgroups, suggesting more intact IFN-I-induced immune responses in the former.⁶ The work by Smeets et al. thus complements ours by demonstrating that: (i) MSC-derived IFN-I mediate their effects on ALL indirectly; and (ii) ETV6-RUNX1⁺ BCP ALL have a distinctly higher IFN-I-driven immune response signature in comparison to other ALL subtypes (Figure 1).

Figure 1.Type I interferons in B-cell precursor acute lymphoblastic leukemia. Smeets et al. found that acute lymphoblastic leukemia (ALL) cells, primarily the ETV6-RUNX1⁺ subtype, induce the paracrine expression of type I interferons (IFN-I) and IFN-I-stimulated genes (ISG) in co-cultured bone marrow mesenchymal stromal cells (MSC). IFN-I pathway induction in MSC is partially mediated via direct contact between leukemia cells and MSC. Induction of ISG and IFN-I production from MSC has an indirect effect on the ALL. This indirect effect of MSC-derived IFN-I may be potentially mediated by non-malignant host immune cells.

The publication by Smeets and colleagues is topical and opens additional avenues for research. The ETV6-RUNX1 subgroup of BCP ALL is unique in terms of its clinical outcome and etiology. Children with this form of ALL have one of the most favorable clinical outcomes with event-free survival for these patients after standard therapies exceeding 90%.⁸ The increased IFN-I pathway signature in ETV6-RUNX1-driven BCP ALL as compared to other ALL subtypes and the heightened ability of ETV6-RUNX1⁺ ALL cells to induce ISG in surrounding MSC¹ could explain the favorable clinical outcomes of patients with this form of ALL. Another interesting feature of ETV6-RUNX1⁺ BCP ALL is that not all individuals who acquire this translocation in utero go on to develop overt leukemia.⁹ The extent of IFN-I-mediated antileukemic responses by MSC may determine the risk of development of overt leukemia in individuals born with the ETV6::RUNX1 rearrangement.¹⁰ The above theories require experimental testing. Lessons learnt from the role of IFN-I in ETV6-RUNX1⁺ ALL will also inform the development of safer therapeutic alternatives to direct IFN-I administration for treating higher risk BCP ALL subgroups.

Footnotes

• Received March 1, 2024
• Accepted March 14, 2024

Correspondence

S. Swaminathan sswaminathan@coh.org

Disclosures

No conflicts of interest to disclose.

Contributions

Both authors contributed equally.