ADH5/ALDH2 Dehydrogenases and DNA Polymerase Theta Protect Hematopoietic Cells from Formaldehyde Challenge: Therapeutic Implications

Background: Formaldehyde is a highly reactive, carcinogenic aldehyde that can accumulate in the hematopoietic stem/progenitor cells (HSPCs) from both endogenous and exogenous sources. Therefore, HSPCs are exposed to steady-levels and occasionally to higher concentrations of formaldehyde. Evidence suggests that exposure to formaldehyde can lead to myeloid leukemia and bone marrow failure.

Formaldehyde can generate interstrand crosslink (ICL) and DNA-protein crosslinks (DPCs); if not repaired properly, they result in DNA double stranded breaks (DSBs) which can be lethal. Alcohol dehydrogenase 5 (ADH5) and aldehyde dehydrogenase 2 (ALDH2) are two detoxifying enzymes in HSPCs that are essential for breakdown of formaldehyde into the less toxic metabolite formate and for preventing the accumulation of DNA damage. Fanconi anemia D2 (FANCD2) plays a critical role in repairing of ICLs. Thus, it appears that cellular genome is protected from the toxicity of intracellularly generated formaldehyde by a two-tier mechanism: AHD5 and ALDH2 (tier-1) and FANCD2 (tier-2). However, it remained unknown if there is another mechanism responsible for protecting cells from DNA damage resulting from challenges of high levels of formaldehyde.

DNA polymerase theta (Polθ) is a promising new target in homologous recombination (HR)-deficient tumor cells. We reported that Polθ-mediated end-joining repair of DSBs might play a critical role in the repair of formaldehyde-mediated DSBs. Thus, we hypothesized that Polθ works in concert with ADH5 and/or ALDH2 to protect HSPCs from formaldehyde-mediated toxic DSBs.

Results: To determine whether ADH5 and/or ALDH2 work together with Polθ to protect HSPCs from continuous physiological concentrations of endogenous formaldehyde, we crossed Polq-/-, Adh5-/-, and Aldh2-/- mice to generate Adh5-/-;Polq-/- and Aldh2-/-;Polq-/- mice. Overall, simultaneous knockouts of Polq and Adh5 or Aldh2 did not cause any major deficiency in hematological parameters of young (3-4 months old) and aged (18-24 months old) mice. Therefore, it seems unlikely that Polθ and ADH5 or ALDH2 play a major role in protecting HSPC from physiological levels of endogenous formaldehyde.

We next wanted to test if Polθ and ADH5 or ALDH2 also protect hematopoietic cells from exogenous formaldehyde challenge. Remarkably, Adh5-/-;Polq-/- cells displayed much higher sensitivity to 5-20 μM formaldehyde when compared to Polq-/- and Adh5-/- counterparts. Next, mice were injected with methanol to increase the levels of formaldehyde. Methanol-induced formaldehyde did not change the basic peripheral blood parameters in Polq-/-, Adh5-/-, Aldh2-/-, Adh5-/-;Polq-/- and Aldh2-/-;Polq-/- mice. Remarkably, clonogenic capacity of HSPCs obtained from methanol-treated Adh5-/-;Polq-/- and Aldh2-/-;Polq-/- mice was reduced when compared to that of the single knockouts.This effect was associated with accumulation of DSBs. Altogether, these results clearly indicate that Polθ and ADH5/ALDH2 provide two-tiered protection against formaldehyde challenge in HSPCs.

We reported that oncogenic tyrosine kinases such as FLT3(ITD) and JAK2(V617F) stimulated one carbon metabolism to generate formaldehyde. To test if Polθ collaborate with ADH5 or ALDH2 in protecting leukemia cells from formaldehyde, FLT3(ITD) and JAK2(V617F) were expressed in HSPCs wild-type, Polq-/-, Adh5-/-, Aldh2-/-, Adh5-/-;Polq-/- and Aldh2-/-;Polq-/- mice. Clonogenic and proliferation potential of FLT3(ITD) and JAK2(V617F)-positive Adh5-/-;Polq-/- and Aldh2-/-;Polq-/- cells was abrogated when compared to Polq-/-, Adh5-/-, Aldh2-/- counterparts. Moreover, proliferation of FLT3(ITD);Adh5-/- and FLT3(ITD);Aldh2-/- HSPCs was inhibited by suboptimal concentrations of Polθ inhibitor which were not toxic for FLT3(ITD) cells.

In addition, HR-deficient human leukemia cells were highly sensitive to the combination of Polθ and ADH5 or ALDH2 inhibitors. This effect was associated with accumulation of DSBs and stimulation of apoptosis. Altogether, we postulate that ADH5 and ALDH inhibitors enhance synthetic lethal effect of Polθ inhibitor against HR-deficient leukemia.

Conclusion: Here we show that Polθ provides a tier-3 protection against extracellular and oncogene-induced formaldehyde challenges. In addition, simultaneous targeting of Polθ and ADH5 or ALDH2 exerted therapeutic effect against HR-deficient leukemias.