Iron Overload Accelerates Development of Leukaemia: Evidence From a Mouse Model

Transfusion-related iron overload is common in MDS. Iron catalyzes, via the Fenton reaction, excess production of reactive oxygen species (ROS), which are known to cause cell senescence and death, promote DNA damage and accelerate carcinogenesis. In addition to the well characterized effects of iron overload on the heart, liver, and endocrine organs, clinical data have suggested iron also causes haematopoietic toxicity in MDS, since iron chelation can lead to dramatic reduction in transfusion requirements and in large registry studies leukaemia-free survival is inversely related to serum ferritin. These observations have led to controversy since they are supported by no mechanistic or animal model data.

We have demonstrated that iron overload increases intracellular ROS (iROS) in early haematopoietic cells in MDS. We hypothesize that iron, via increased iROS, promotes accumulation of DNA damage in MDS HSCs and thus, in the context of the genomic instability of the MDS clone, accelerates progression of MDS to AML. Here we report the results of experiments that establish the biological and mechanistic plausibility of this hypothesis.

To establish that the B6D2F1 mouse model, which has been used in studies of cardiac and hepatic iron overload, is also a suitable model of bone marrow iron overload, mice (n=5 per cohort) were given iron dextran (0-150 mg i.p.) and sacrificed 3 days later. Severe weight loss was noted in the iron-loaded animals. Iron deposition was confirmed by Prussian Blue staining in the bone marrow, liver, myocardium, and the red pulp of the spleen.

The cardiac effects of this degree of iron overload compromise survival, preventing assessment of longer-term effects of iron on haematopoiesis. We therefore evaluated the effects of lower doses of iron dextran (0, 5, 10, or 20 mg; n=5 per cohort) over 3 months. Increased iROS was seen in lineage negative (lin⁻) CD45⁺ bone marrow cells for animals that received 5 mg iron. However, iROS levels decreased progressively from the 10–20 mg treated animals, possibly representing an increase in apoptosis in early haematopoietic cells exposed to the greatest oxidative stress. Consistent with this, we observed increased apoptosis in early erythroid progenitors for the 20 mg iron treated animals (p<0.05).

We adapted the chronic iron overload mouse model to evaluate the effect of iron overload on leukaemogenesis. B6D2F1 mice were sublethally irradiated (300 cGy) followed by s.c. injection of 0.5 mg dexamethasone, a protocol which induces a pre-leukaemic state leading, in SJL mice, to AML in 50–75% with a 12 month latency. These mice were then loaded with 0 or 5 mg I.P. iron dextran, n=6 per cohort). Three mice from each cohort were sacrificed and analyzed 3 months after iron loading. Expansion of the splenic white pulp was observed in iron loaded mice and flow cytometric analysis of the bone marrow cells revealed expansion of the lin⁻ CD45⁺ early haematopoietic population. Furthermore, in one iron loaded mouse we observed a lin⁻ CD45^lo population with size and complexity similar to that of haematopoietic progenitors, suggesting blast accumulation. The remaining mice (n=3 per cohort) continued to be observed. One mouse in the iron loaded cohort died eight months after iron loading. Post-mortem examination revealed severe hepatomegaly and splenomegaly, massive splenic and hepatic infiltration by leukaemic blasts, and extensive bone marrow necrosis, fibrosis, and substantial blast accumulation.

To establish a plausible mechanism for the promotion of leukaemia development by iron, we tested the ability of iron to cause DNA damage in a haematopoietic cell line. HL60 cells line were treated with ferric ammonium sulfate (10 or 100 μ M) and DNA damage was assessed by flow cytometry for γH2AX, an indicator of DNA double-strand breaks. Elevated γH2AX was observed in HL60 cells 2 hours after iron loading, and sustained DNA damage was noted till the end of the experiment at day 4.

Our observations demonstrate that iron is mutagenic in haematopoietic cells and can promote progression of a pre-leukaemic state to frank AML. We postulate that iron is not itself leukaemogenic, but, by causing DNA damage, promotes clonal evolution in MDS. Further evaluation in animal models and in clinical trials is necessary to elucidate the clinical implications of these observations, especially in regard to the deployment of iron chelation therapy.

No relevant conflicts of interest to declare.