Introduction
Sry type HMG box (Sox6) is a member of the Sox transcription factor family characterized by the conserved high mobility group (HMG) domain, consisting of 79 amino acids involved in DNA recognition and binding [1]. Sox transcription factors bind to the minor groove of DNA and cause a 70°–85° bend of the DNA that leads to local conformational changes [2,3], while most other transcription factors target the major groove of DNA [4]. Therefore, Sox proteins may perform part of their function as architectural proteins by organizing local chromatin structure and assembling other DNA-bound transcription factors into biologically active, sterically defined multiprotein complexes. Sox6 has been reported to be able to act as either an activator or a repressor, depending on its interactors and its target promoter context [5,6]. Intriguingly, Sox6 has also been shown to act as a general splicing factor that participates in pre-mRNA splicing [7]. Depletion of Sox6 in HeLa cell extracts blocked splicing of multiple substrates, and expression of the HMG domain of either Sox6, Sox9, or Sry in the extracts restored splicing, indicating functional overlap of these proteins [7]. Regardless of how Sox6 functions in regulating gene expression, previous studies have demonstrated that Sox6 is an important regulatory molecule that plays a role in the development of the central nervous system [8–11], cartilage [6,12,13], and muscle [14,15]. A Sox6-null mutant mouse (p100H) has previously been identified in our laboratory [14]. Mice homozygous for p100H show delayed growth, develop myopathy and arterioventricular heart block, and die within 2 wk after birth [14]. The p100H mutant allele is associated with a Chromosome 7 inversion that disrupts both the p gene and the Sox6 gene (and no other gene within 50,000 nucleotides of the chromosomal breakpoints) [14]. Because the p gene functions solely in pigmentation [16], the Sox6 transcription factor is implicated in all other phenotypes.
Among the HMG box proteins distantly related to Sry (the first member identified of the Sox transcription factor family) that similarly bind to the minor groove and bend DNA, but without sequence specificity, are the ubiquitously expressed HMG1 and HMG2 proteins [17]. Modulation of DNA structure by these and other HMG proteins can mediate long-range enhancer function on both DNA and chromatin-assembled genes by bringing together distant regions of DNA and associated factors. Specifically, HMG proteins have been shown to modulate β-globin genes [18–21].
The mouse β-globin genes {ɛy, βh1, β-major, and β-minor} are clustered on Chromosome 7 and they are highly homologous to their human counterparts in organizational structure and function [22]. High-level expression of these genes requires a regulatory element, the locus control region that is characterized by a set of nuclease hypersensitive sites spread over 25 kb located 5′ of the ɛy gene [23]. The β-globin genes are expressed in a tissue- and development-specific fashion. In mice, erythropoiesis originates in the embryonic yolk sac where primitive erythroid cells express ɛy and βh-1 globins [22]. At 11.5 d post coitus (dpc), erythropoiesis shifts to the fetal liver where definitive erythroid cells express adult β globins (β major and minor) [22]. The ɛy gene is silenced in definitive erythroid cells. The mechanism of silencing of its human counterpart, ɛ globin, has been studied extensively. In definitive erythropoiesis, ɛ is activated and silenced autonomously [24,25], although in primitive erythropoiesis ɛ also appears to be regulated competitively [26]. The γ-globin to adult β-globin switch is controlled by promoter competition for the LCR [24,25].
All the elements responsible for silencing the ɛ globin gene are within the ɛ gene or in adjacent sequences [27], suggesting that silencing is primarily gene autonomous. Using promoter deletion analyses in transgenic mouse models and cell transfection assays, multiple DNA elements important to the silencing process have been previously identified in both the proximal and the distal ɛ gene promoter [27]. Their corresponding transcription factors, such as GATA-1, YY-1, COUP-TF, and DRED have been identified and shown to directly bind to these DNA elements (as part of protein complexes) to regulate ɛ silencing [27]. Thus, it appears that the silencing of the ɛ gene involves a complicated network of multiple cis elements and transacting proteins.
In addition to playing an important role in the development of the central nervous system [8–11], cartilage [6,12,13], and muscle [14,15], it was shown that Sox6 is upregulated in long-term hematopoiesis stem cells (LT-HSC) compared with multipotent progenitors of adult mouse bone marrow lineage [28]. In this study, we describe that Sox6 also exerts pleiotropic effects on erythropoiesis. These effects include delayed maturation of erythrocytes (that normally enucleate prior to entering the bloodstream [27]) and higher expression of embryonic globin genes. The most extreme effect is the persistence of high expression of the embryonic ɛy globin gene. Here we describe and characterize the effects of Sox6 on the ɛy globin gene. We show that Sox6 binds to the proximal promoter of ɛy globin and represses its transcription. In wild-type (WT) mice, Sox6 is not expressed in yolk sac blood islands, but is expressed in fetal liver, the opposite expression pattern of ɛy globin. In the absence of Sox6, ɛy globin is ectopically expressed in the fetal liver, demonstrating that Sox6 functions in definitive erythropoiesis.