An unexpected entry into the globin real estate market

Comment on Basu et al, page 2566

Once EKLF (KLF1) was found to be instrumental for adult β-globin expression, related activators for embryonic and fetal β-like genes have been sought. One, first identified 10 years ago, has unexpectedly reappeared.

The regulation of β-like globin gene expression has historically been a fertile ground for examining cellular and developmental control mechanisms. The β-like cluster exhibits “switches” in expression in all blood-forming vertebrates. In humans, these switches give rise to changes in β-like globin polypeptide chain expression, which yield hemoglobin proteins uniquely expressed in the embryonic yolk sac, fetal liver, and adult bone marrow. Conserved, recurring sequences within the promoter and enhancing elements of these genes suggest that GATA, CACCC, and NFE2 motifs are important. Among the tissue-restricted transcription factors that bind to these sites is a zinc finger protein named erythroid Krüppel-like factor (EKLF).¹  Structural modeling of the EKLF zinc fingers led to the a priori prediction that it would bind the extended adult β-globin CACCC element (CCACACCCT), a postulate subsequently verified by a range of molecular analyses. The highly specific nature of this interaction was supported by the results of genetic ablation of EKLF in the mouse, which led to lethality due to the virtual absence of the developmental switch to adult β-globin transcription; however, the prior globin developmental program was not affected. This led to a conundrum: given that the other β-like globin genes (in both mice and men) also have important CACCC elements in their promoters, which protein binds there?

Based on the idea that if it looks and smells like a CACCC element, there must be a KLF protein nearby, a number of groups began homologous cloning strategies in search of just such an EKLF-related factor, contributing to the discovery of kindred proteins that now total 16. One early discovery was lung KLF (LKLF; KLF2).²  Although expressed in hematopoietic cells, it was highly expressed in the lung, and although genetic ablation analyses in the mouse led to embryonic lethality by embryonic day (E) 12.5, KLF2 was found to play critical roles in blood vessel wall integrity, T-cell viability, and lung development. These studies did not reveal any obvious red blood cell phenotype.

This was the status of KLF2 for the ensuing 6 years until the present study. By quantitative analyses of globin RNAs expressed in blood cells from staged murine embryos genetically altered in KLF2 expression, Basu and colleagues have now shown a KLF2 dosage-dependent expression of the murine embryonic, but not adult, β-like globin genes. By crossing these mice to a transgenic line that contains the complete human β-like globin locus, the authors find in addition that the human embryonic β-like globin gene is more highly affected by the absence of KLF2 than the fetal β-like globin gene; again, the adult gene is not affected. These studies not only show a gene-specific role for KLF2, but also help explain why the earlier studies, focused on the definitive red-cell population, did not detect an effect. Interestingly, KLF2-deficient primitive erythroid cells are morphologically altered compared to their wild-type counterparts in yolk sac tissue sections. Such defects may result from or contribute to the significantly higher percentage of apoptotic cells that are seen in the absence of KLF2.

As with many unexpected results, a number of questions arise from the new data. Might apoptotic pathways be involved in the gradual disappearance of yolk-sac erythroid cells during early development, possibly by the downregulation of KLF2 during primitive erythroid maturation? Does KLF2's role in endothelial and T-cell genetic regulation (with respect to cell cycle, apoptosis, and responses to external stimuli) suggest any additional functions in the erythroid compartment? Does KLF2 play a role in the transcriptional synthesis of membrane proteins, such that its absence leads to increased cellular irregularities? As KLF2 ablation leaves a residual level of embryonic globin expression, is there another KLF protein also involved? Given the same extended recognition sequences at their promoters, how is the divergent effect of KLF2 absence at the murine embryonic and human fetal globin genes explained? Does any KLF family member positively regulate the human fetal globin gene? Future studies will certainly focus on such intriguing issues raised by this late entry into the globin market. ▪