Hypoxic up-regulation of erythroid 5-aminolevulinate synthase

The first and regulatory step of heme biosynthesis in animals is catalyzed by 5-aminolevulinate synthase (ALAS).1 In addition to the ubiquitously expressed housekeeping isozyme (ALAS1), the erythroid-specific ALAS isozyme (X-chromosome–encoded ALAS2) provides the large quantity of heme required by erythroid cells for hemoglobin production.2,3ALAS2 is expressed exclusively in erythroid cells, and ALAS2 gene defects result in impaired heme biosynthesis, as observed in X-linked sideroblastic anemia.4,5 Enhanced synthesis of hemoglobin during erythropoiesis requires a coordinated production of heme molecules, globin chains, iron delivery, and iron uptake. Expression of proteins responsible for hemoglobin synthesis is therefore interconnected by several regulatory mechanisms, including erythroid-specific transcription factors, cellular heme concentration, and iron status through iron-responsive elements (IREs) located in untranslated mRNA regions.6,7 

Hypoxia up-regulates the expression of proteins involved in iron metabolism, including transferrin,8 transferrin receptor,9 and ceruloplasmin.10 The 5′ regulatory regions of these genes contain functional hypoxia-response elements (HREs), mediating transcriptional activation by the hypoxia-inducible factor–1 (HIF-1). HIF-1 is a heterodimeric transcription factor containing an oxygen-labile α-subunit.11 Because hypoxia induces several HIF-1 target genes involved in erythropoiesis, and considering that ALAS2 regulates heme biosynthesis, it is conceivable that ALAS2 expression might be regulated by the same oxygen-dependent mechanisms. Indeed, a single putative HIF-1 DNA-binding site (HBS) has been identified previously within the promoter region of the murine ALAS2 gene.12 Here, we present the first functional analysis of oxygen-regulated ALAS2 expression.

Murine erythroleukemia (MEL) and HeLa cells were cultured as described.13 Differentiation of MEL cells was induced with 1.8% dimethyl sulfoxide (DMSO) 24 hours before hypoxic (1% O₂) or normoxic (20% O₂) exposure. Where indicated, 16 μM ciclopirox olamine (CPX)14 or 100 μM desferrioxamine (DFX)15 were added to the medium.

Preparation of nuclear extracts, immunoblot analyses, and measurements of heme pseudoperoxidase activity were performed as described before.13,16,17 

Preparation of nuclear extracts from HeLa cells and electrophoretic mobility shift assays (EMSAs) were carried out as described.16 A double-stranded oligonucleotide (TCTCCACGTGGACTTGCC) containing the predicted ALAS2 HBS was used as a probe. Oligonucleotides containing a mutated HBS (TCTCCAAAAGGACTTGCC) or an erythropoietin 3′ enhancer–derived HBS16 served as negative and positive controls, respectively.

Total RNA isolation, probe preparation, and Northern blotting procedures were performed as described.8 Hybridization probes were based on the murine ALAS2 sequence.18 

A pGL2-derived firefly luciferase reporter construct encompassing the −718 to +1 region of the mouse ALAS2 promoter12 was used as a template for site-directed mutagenesis of the HBS at position −323/−318. The polymerase chain reaction (PCR) procedure was performed as described,12 with the use of the following primers: ALAS-KpnI (5′-ATATGGTACCCTAGAATCCAATCCATTAC-3′); ALAS-mutant reverse (5′-GCAAGTCCAAAAGGAGAGGACC-3′); ALAS-mutant forward (5′-GGTCCTCTCCTTTTGGACTTGC-3′); and reverse ALAS-HindIII (5′-ATATAAGCTTAGTGCTGTGGGCTGGGCTG-3′). The PCR product was inserted into KpnI-HindIII–digested pGL2, resulting in pTH15. As a control, subcloning of a PCR product amplified with the primers ALAS-KpnI and RALAS-HindIII, which contained the wild-type HBS, resulted in pTH10. Mutations were confirmed by DNA sequencing.

Luciferase constructs described above, as well as positive and negative control constructs,16 were transiently transfected into HeLa cells and analyzed for oxygen-dependent regulation as described before.13 

Hypoxic exposure as well as treatment with the iron chelators CPX14 and DFX15 led to stabilization of HIF-1α in MEL cells differentiated by adding DMSO 24 hours before hypoxic exposure (Figure 1A). To analyze whether ALAS2 mRNA steady-state levels were inducible by hypoxia like other genes involved in erythropoiesis, MEL cells were cultured under hypoxic (1% O₂) or normoxic conditions for 24 and 48 hours, respectively. As shown in Figure 1B, DMSO treatment led to a strong up-regulation of ALAS2 mRNA expression. Interestingly, hypoxic exposure of differentiated cells induced the mRNA levels of ALAS2 and glucose transporter-1 (Glut-1), a known HIF-1 target gene,19 up to 3-fold, indicating that expression of both genes is influenced by oxygen-regulated pathways. Since ALAS2 protein catalyzes the rate-limiting step in heme synthesis, relative heme content was indirectly estimated following normoxic or hypoxic exposure by measuring its pseudoperoxidase activity. As shown in Figure1C, the heme content was significantly (P < .05; n = 3; Student t test) increased after 48 hours of hypoxic exposure, demonstrating that hypoxic up-regulation of ALAS2 mRNA results in physiologic heme production during erythroid differentiation.

Previous analysis of the 5′-flanking region (−1440 to +1) of the mouse ALAS2 gene revealed the presence of several binding sites for erythroid-specific transcription factors as well as a putative HBS (CCACGTGG) at position −318 to −323.12 The predicted HBS shares a high similarity with HBSs found in the promoters of genes encoding insulinlike growth factor–binding protein and lactate dehydrogenase A. To test whether oxygen regulation of ALAS2 expression is mediated by HIF-1, fragments ranging from −716 to +1 and containing either a wild-type (pTH10) or a mutated (pTH15) HBS were inserted upstream of a promoterless luciferase reporter gene vector. Because differentiated MEL cells could not be transfected (data not shown), HeLa cells were used instead. Previous reports showed that the ALAS2 promoter activity is comparable in MEL and HeLa cells.12,20 Hypoxic exposure led to a 2.7-fold induction of luciferase expression (Figure 1D), suggesting that hypoxic induction of ALAS2 is regulated at the transcriptional level. The similarity of the hypoxic induction factors in ALAS2 promoter activity and ALAS2 mRNA levels do not suggest (but also do not formally exclude) that increased mRNA stability contributes to the ALAS2 mRNA induction. However, hypoxic induction of the ALAS2 promoter activity persisted when a construct containing a mutation in the predicted HBS was used (Figure 1D).

Iron chelators are known to functionally stabilize HIF-1α.14,15 However, in contrast to HeLa cells transfected with a control plasmid containing erythropoietin-derived HBSs, no significant increase in ALAS2 promoter activity was observed following DFX and CPX treatment (Figure2A). In line with this result, Northern blot analysis revealed that DFX leads to an up-regulation of the HIF-1 target gene Glut-1, but not of ALAS2 (Figure 2B).

In EMSAs, an erythropoietin-derived HBS16 resulted in an HIF-1 DNA-binding complex that could be supershifted with the use of a specific anti–HIF-1α antibody (Figure 2C). In contrast, no hypoxia-inducible signal could be observed with the putative HBS derived from the ALAS2 promoter. We therefore conclude that despite the high similarity of the putative ALAS2 HBS core sequence to other HIF target genes, mismatches in the flanking region inhibit an effective DNA-protein interaction. Nonfunctional putative HBSs were also found in the mouse gene encoding HIF-1α itself.21 

Taken together, our data demonstrate that ALAS2 transcription is hypoxically induced in an HIF-1–independent manner, leading to an increase in heme content. HIF-1–independent hypoxia-inducible pathways have previously been reported for other genes, including IAP-2 and tumor necrosis factor (TNF) receptor type 2.22,23 As iron depletion is known to inhibit ALAS2 translation,24 hypoxic induction of ALAS2 transcription has probably evolved to be HIF-1 independent, as the HIF-1 pathway is activated by iron depletion.

We are grateful to C. Bauer for helpful discussions.