Oct1 and Oca-B Regulate the Myeloid Leukemic Phenotype

Oct1, encoded by the Pou2f1 gene, is a widely expressed transcription factor that is associated with metabolic control, stress response, stem cell phenotypes and malignant transformation. Oct1 associates with a number of cofactors including the activating cofactor OCA-B encoded by Pou2af1. OCA-B is normally expressed specifically in B cells, but is misexpressed in about 30% of AML cases in myeloid cells and is a poor prognostic indicator. We observed a conserved Oct1 binding site in the promoter of Cdx2, a homeodomain transcription factor that is normally expressed solely in the gastrointestinal tract of adult humans, but has been shown to misexpressed in blood cells in almost all cases of AML. Our study looks at the interaction of Oct1 with Cdx2, and the effects of Oct1 and OCAB-B knockout in a mouse model of AML.

Using ChIP we determined that Oct1 is bound to the Cdx2 promoter in normal patient samples and AML patient samples. However, in normal patient samples a repressive cofactor (the NuRD complex) is also bound. In AML patient samples an activating cofactor (Jmjd1a) is bound. This suggests that in AML Oct1 is bound to Cdx2 and is associating with Jmjd1a, which can occur through ERK signaling or OCA-B expression, and activating expression of Cdx2.

We next wanted to determine the effect of Oct1 knockout in AML using a MLL-AF9 mouse model. Our lab has developed a conditional Oct1 allele, as Oct1 germline knockout is embryonic lethal. We harvested bone marrow from Oct1^+/fl;MX1-Cre, Oct1^fl/D;MX1-Cre, and Oct1^+/+;MX1-Cre mice and transduced the bone marrow with virus expressing MLL-AF9. Recipient mice were lethally irradiated and transplanted with the transduced bone marrow. Mice were injected with Poly IC at 5 weeks post-transplant to delete Oct1 foxed alleles. Mice that were wildtype for Oct1 developed leukemia around 7 weeks. Mice with one allele of Oct1 survived longer but still developed leukemia. Mice with no Oct1 survived longer still, but eventually suffered from bone marrow failure.

In order to assess whether the lethality in Oct1 deficient mice with MLL-AF9-induced leukemia was due to the presence of the MLL-AF9 oncoprotein or if Oct1 deletion is lethal in adult mice under normal conditions, we treated non-leukemic mice that were Oct1^+/fl;MX1-Cre or Oct1^fl/D;MX1-Cre with Poly IC to delete Oct1 at 5 weeks. Following Poly IC injection, mice with and without Oct1 appeared normal and healthy by examination and by blood counts. At 20 weeks, all mice were injected with 5-flurouracil (5-FU). We observed that mice that were deficient in Oct1, namely Oct1^fl/D;MX1-Cre mice injected with Poly IC, died within 2 weeks of 5-FU treatment. Put together, our data suggests that Oct1 is not necessary for normal hematopoiesis, as adult mice with no Oct1 appear normal under homeostatic conditions, but Oct1 is necessary for the hematopoietic proliferative stress response.

The protective effect of Oct1 deletion against leukemia in mice suggested an important role for Oct1 in regulating the leukemic phenotype. However, since Oct1 deletion is also lethal in the leukemic context, we looked at the effect of deletion of OCA-B in leukemia. OCA-B germline deficient mice are viable and the only detectable phenotype is the lack of germinal center formation. We harvested OCA-B germline deficient and wildtype control bone marrow and followed the same protocol as outlined above with the transduction of the bone marrow with MLL-AF9. Mice with no OCA-B did not develop leukemia and had 100% survival. This striking result suggests that OCA-B could be a therapeutic target in MLL-AF9 AML.