Erythroid-Specific Variant of the Nuclear Exportin Xpo7 Conserved Only in Mammals May Explain Functional Differences Between Mammalian Definitive and Lower Vertebrate (or Primitive) Erythropoiesis

Exportin 7 (Xpo7) was previously found to be crucial for late murine erythroid nuclear maturation when its knockdown significantly inhibited chromatin condensation and ultimate enucleation in a primary mouse erythroblast culture [1]. Additionally, Xpo7 was found to have an erythroid-specific isoform, transcribed from an erythroid-specific transcription start site (TSS) in the first intron of Xpo7 (named Xpo7-1^B in contrast to the ubiquitously expressed variant -1^A) [1]. Xpo7-1^B differs from Xpo7-1^B in 8 amino acid residues in the first exon and 9 residues in the 5th exon, and these differences are not conserved either in lower vertebrates or in primitive mouse erythropoiesis. It is intriguing that both lower vertebrate and primitive mammalian erythroblasts do not enucleate.

In order to understand the in vivo function of Xpo7 and determine if these differences in the two isoforms confer differential functions to the proteins or are simply regulatory, we examined the definitive fetal liver erythropoiesis (considered adult "stress erythropoiesis") and adult steady state erythropoiesis of transgenic mouse models lacking either Xpo7-1^A, Xpo7-1^B, or both. We had previously knocked down Xpo7 in zebrafish using morpholinos (zebrafish only express the Xpo7-1^A homologue) and found no effect on adult erythroid differentiation or proliferation. Adult Xpo7-1^A knockout (KO) mice are born in normal Mendelian ratios, are not anemic, and express normal Xpo7-1^B only in the blood. Closer examination reveals that fetal liver erythroid proliferation is actually higher in Xpo7-1^A KO mice than either heterozygous (HET) or wild type (WT) while differentiation is normal. The double KO (Xpo7-1^AB KO) appears to be embryonic lethal as only 4 live knockout pups were born out of 71 live births and even these 4 pups appear to be partial chimeras. We characterized the Xpo7-1^AB HET mice in more detail and fetal liver erythroblasts of HET mice showed normal proliferation and differentiation compared to WT, and adults showed no differences in red blood cell (RBC) counts, hematocrit or hemoglobin. No complete live Xpo7-1^AB KO has been born so we are now evaluating embryos at E3, E9.5, and E14.5. Given the possibility of embryonic lethality, in order to evaluate the erythroid-specific isoform Xpo7-1^B, we have crossed a mouse floxed for the Xpo7-1^B TSS/1st exon with a tamoxifen-inducible Cre-recombinase mouse. These mice have been given tamoxifen after reaching adulthood and we are currently evaluating their hematopoietic parameters. We have also set up HET-HET timed matings and administered tamoxifen to pregnant females to evaluate fetal liver erythropoiesis in Xpo7-1^B HET, WT, and KO embryos.

Not only will this work further our understanding of the process of global chromatin condensation during erythropoiesis, data from this in vivo study may reveal differences in erythropoiesis between mammals and lower vertebrates, which may also reflect intrinsic differences in mammalian primitive versus definitive erythropoiesis. Based on our in vitro knockdown studies, we anticipate that the phenotype of the Xpo7-1^B KO mice may be similar to congenital dyserythropoietic anemia (CDA), a human disorder involving abnormalities in erythroid progenitor chromatin condensation and nuclear membrane substructure, so examination of these mice may also uncover a novel mechanism for this form of ineffective erythropoiesis.

[1] Hattangadi, SM, et al. Blood (2014) 124, 1931-40.