Inhibition of Splicing Factor SF3B1: Evaluating Effects on Iron Metabolism in K562 and CD34+ Cells

Somatic mutations in the RNA splicing factor SF3B1 have been found in hematopoietic cells from patients with myelodysplastic syndrome. The association is particularly strong for patients with anemia with ringed sideroblasts, where 65 - 75 % show alterations in SF3B1. Ringed sideroblasts are erythroid precursors with insoluble iron aggregates in mitochondria and deficient hemoglobin synthesis. Thus these cells demonstrate a defect in iron metabolism. SF3B1 is a core component of the U2 snRNP spliceosome, responsible for the majority of mRNA maturation, raising the question of how alteration of SF3B1 might perturb cellular iron metabolism. Meayamycin B, a synthetic analog of the natural product FR901464, is an inhibitor of SF3B1, and here we describe the effects of this compound on cellular iron metabolism in the K562 leukemic cell line and in CD34+ cells isolated from cord blood.

Inhibition of SF3B1 by meayamycin B was confirmed by analysis of spliced forms of Mcl-1 mRNA, a known target of SF3B1. In untreated cells, a long form (Mcl-l_L) arises from inclusion of exon 2 in the final mRNA. In treated cells, inhibition of SF3B1 splicing generates a short form (Mcl-l_S) that lacks exon 2. In the presence of 3 to 10 nM meayamycin B the short splice-form (Mcl-1_S) became the dominant species. To examine the effect of inhibition of SF3B1 on iron metabolism, K562 cells were exposed to meayamycin B for 24 hours, with controls of 50 µM hemin and 50 µM deferoxamine (DFO). Steady state mRNA levels of transferrin receptor (TFR-1), ABCB7 (mutations in which are associated with inherited sideroblastic anemia and ataxia), and NDRG1 (n-myc downstream regulated gene, previously shown to be induced by DFO) were significantly down regulated by meayamycin B, as determined by qRT-PCR. NDUFS8 mRNA that encodes a subunit of mitochondrial complex 1 and has previously been shown to be iron-responsive, did not change in response to meayamycin B, hemin, or DFO. However, MRLP19 that encodes a mitochondrial ribosomal component, was strongly down-regulated by meayamycin B. In contrast, cytosolic ribosomal protein RPLP0 was not affected by meayamycin B, hemin or DFO and was used as the endogenous control.

Analysis of protein levels by western blotting indicated that meayamycin B increased levels of ferritin (heavy chain) and decreased TFR-1. In concordance with the decreased TFR-1 protein levels, meayamycin B treated K562 cells exhibited significant downregulation of iron uptake from ⁵⁵Fe-transferrin, similar to that observed for hemin treated cells. DFO treatment significantly increased transferrin iron uptake, compared to untreated controls.

A similar set of experiments was performed with CD34+ cells, before or after differentiation with erythropoietin. Meayamycin was efficacious in these cells, as shown by disappearance of the long splice-form Mcl-l_L and appearance of the short splice form Mcl-1_S in cells exposed to the compound for 16 h. TFR-1 mRNA exhibited a dose dependent decrease in response to meayamycin (both pre and post epo treatment). Neither hemin nor DFO altered TFR-1 mRNA levels, or protein levels. In non-epo treated CD34+ cells, exposure to meayamycin caused a small decrease in ferritin protein levels, but ferritin protein increased in epo-treated cells exposed to meayamycin. Hemin increased ferritin protein in both non-epo and epo treated cells. Mitochondrial ribosomal protein MRPL19 RNA was significantly decreased by meayamycin in all cases, but cytoplasmic ribosomal protein RPLP0 was unaffected by meayamycin. Ribosomal protein levels were not responsive to iron manipulations with hemin or DFO.

To summarize, treatment of cells with meayamycin B inhibits SF3B1 function in K562 cells or CD34+ cells. The effects are quicker and cleaner than siRNA, making possible closer functional correlations. Meayamycin treated K562 cells exhibited downregulation of TFR-1 and TFR-1 iron uptake, as if they were sensing high cellular iron levels. Note that this is the opposite of the expected result, because iron loading in the setting of sideroblastic anemia would be expected to be associated with increased cellular iron uptake. In CD34+ cells, a similar downregulation of TFR-1 levels was observed, although iron uptake has not yet been directly measured. The strong decrease in mitochondrial ribosomal protein expression indicates a specific mitochondrial effect of SF3B1 inhibition, which may be an important mediator of iron metabolism.