Investigating Zeta Globin Gene Expression to Develop a Potential Therapy for Alpha Thalassemia Major

Introduction: Alpha globin mutations are very common worldwide, and the severity of resulting anemia depends on the number and type of mutated alleles. While the 4 gene mutation (alpha thalassemia major, ATM) was previously deemed fatal except in rare cases, emerging evidence indicates that survival to birth and good postnatal outcomes are possible with in utero transfusions. We hypothesized that the embryonic zeta globin gene, which is expressed early in gestation prior to alpha globin, may compensate for the lack of alpha globin and that induction of zeta globin after it has naturally been silenced may become a new therapy for patients with ATM. Methods: We evaluated mutations in the UCSF international registry of patients with ATM to understand factors related to patient survival with and without in utero transfusions. We then engineered Human Umbilical Cord Derived Erythroid Progenitor Cells (HUDEP-2 cells) carrying the common SEA alpha globin deletion, in which zeta globin expression is preserved (H-SEA), as well as those on which the zeta globin genes were deleted (HBZ-/-) using CRISPR-Cas9. We evaluated the expression of alpha and zeta globins using qPCR, Western blot, and flow cytometry. We generated lentiviral vectors expressing zeta globin under the control of beta-globin promoters to examine changes in both zeta and alpha globin in a dynamic way. Results: None of the registry patients who survived to birth spontaneously (n=11) had a mutation that involves a concomitant deletion in zeta globin (such as the -FIL, -THAI, or -MEDII mutation), while alpha globin mutations extending into the zeta globin gene were found in 14 of 37 (38%) patients who were diagnosed prenatally, suggesting that the presence of zeta globin may play a role in the ability to survive to birth without fetal therapy. Interestingly, we found that H-SEA clones express higher levels of zeta globin than WT cells, as illustrated by quantitative real-time PCR (Fig 1A), Western blot (Fig 1B) and flow cytometry (Fig 1C). These cells also developed beta globin dimers due to excess unpaired beta-globin chains, as demonstrated by Western blotting with and without reducing agents, indicating that they are an appropriate cell model for ATM. We next generated HUDEP-2 clones lacking zeta globin (HBZ KO) and transduced these clones with lentiviral vectors expressing high levels of zeta globin (lenti-zeta) (Fig 1D). Western blotting revealed that increasing the levels of zeta globin in these cells resulted in decreased expression of alpha globin, suggesting reciprocal control between these genes (Fig 1E). Most importantly, we saw a reduction in toxic beta-globin dimers in H-SEA cells expressing high levels of zeta-globin after transduction with lenti-zeta, suggesting that zeta globin could functionally replace the missing alpha-globin (Fig 1 F,G). To understand transcriptomic differences in H-SEA cells that may result in increased zeta globin expression, we performed bulk RNA sequencing of wild type and H-SEA clones. We confirmed that zeta expression is significantly upregulated in H-SEA compared to wild type (log2 fold change of 4.25, p=2.24E-38). Pathway analysis of differentially expressed genes is ongoing. Conclusions: Our international patient registry suggests that expression of zeta globin may play a role in the spontaneous survival to birth in a subset of patients. Zeta globin expression is increased in the setting of H-SEA cells in vitro, and restoration of zeta expression by lentivirus results in a reduction of toxic beta globin dimers in these ATM cells. Furthermore, expressing zeta globin at high levels in H-WT cells decreased alpha globin expression, suggesting a reciprocal regulation of these two genes. This concept is similar to the relationship between fetal gamma and adult beta globins which has been exploited for therapeutic editing approaches in patients with beta-thalassemia. At this point, the natural repressor of zeta globin is not yet known, but our data suggests that a strategy of upregulating zeta globin could potentially be developed to mimic the ongoing trials of using the BCL11A repressor to induce gamma globin in patients with beta thalassemia and sickle cell disease. Disclosures Wienert: Integral Medicines: Current Employment. Kohn:Allogene Therapeutics: Consultancy, Membership on an entity's Board of Directors or advisory committees; Orchard Therapeutics: Consultancy, Patents & Royalties, Research Funding. MacKenzie:Acrigen: Membership on an entity's Board of Directors or advisory committees; Ultragenyx: Research Funding.

Developmental silencing of the embryonic zeta-globin gene: concerted action of the promoter and the 3'-flanking region combined with stage-specific silencing by the transcribed segment.

Globin gene switching is a well-described model of eucaryotic developmental control. In the case of the human alpha-globin gene cluster, migration of erythropoietic activity from the embryonic yolk sac to the fetal liver is parallaled by the zeta-globin gene silencing and enhanced expression of the alpha-globin genes. To map critical cis determinants of this switch, the human zeta-globin gene, the alpha-globin gene, and chimeric recombinants were introduced into the mouse genome. Consistent with previous studies, expression of the individual alpha- and zeta-globin transgenes was found to be developmentally appropriate. Contrary to current models, however, the alpha- and zeta-globin gene promoters were not sufficient to establish this control. Instead, full silencing of the zeta-globin gene required the combined activities of this promoter, transcribed region, and 3'-flanking sequences. Individually, the silencing activities of the zeta-globin gene promoter and 3'-flanking region were minimal but increased markedly when both regions were present. The zeta-globin transcribed region appeared to contribute to gene silencing by a mechanism specifically activated in definitive erythroblasts in the fetal liver. These data demonstrate that a complex set of controls, requiring at least three determinants and involving at least two independent mechanisms, is necessary for full developmental silencing of the human zeta-globin gene.

Human embryonic zeta-globin gene expression in mouse-human hybrid erythroid cell lines

The human alpha-globin-like embryonic zeta-globin chains are present in abundance during the first 5 to 6 weeks of gestation. Subsequently, zeta-globin chains are present in fetal blood at a very low level, which is supplanted by the expression of alpha-globin chains. Adult individuals who are carriers of the (--SEA/) alpha-thalassemia deletion, in contrast to normal adults, have low levels of embryonic zeta-globin chains in their circulating erythrocytes. In this investigation, we constructed stable mouse-human hybrid cells with murine erythroleukemia cells bearing human chromosome 16, with either the normal alpha-globin gene cluster (alpha alpha/) or the (--SEA/) type of alpha-thalassemia deletion. The results on the human zeta- globin gene expression in these hybrid cells indicate that murine adult erythroid transcription factors can induce the expression of human embryonic zeta-globin gene is cis to the (--SEA/) deletion, in parallel with the endogenous mouse alpha-globin gene expression. These data also show the importance of the DNA sequences within the (--SEA) deletion in regulating the expression of zeta-globin gene in cis during normal human hemoglobin ontogeny.

Human embryonic zeta-globin gene expression in mouse-human hybrid erythroid cell lines

The human alpha-globin-like embryonic zeta-globin chains are present in abundance during the first 5 to 6 weeks of gestation. Subsequently, zeta-globin chains are present in fetal blood at a very low level, which is supplanted by the expression of alpha-globin chains. Adult individuals who are carriers of the (--SEA/) alpha-thalassemia deletion, in contrast to normal adults, have low levels of embryonic zeta-globin chains in their circulating erythrocytes. In this investigation, we constructed stable mouse-human hybrid cells with murine erythroleukemia cells bearing human chromosome 16, with either the normal alpha-globin gene cluster (alpha alpha/) or the (--SEA/) type of alpha-thalassemia deletion. The results on the human zeta- globin gene expression in these hybrid cells indicate that murine adult erythroid transcription factors can induce the expression of human embryonic zeta-globin gene is cis to the (--SEA/) deletion, in parallel with the endogenous mouse alpha-globin gene expression. These data also show the importance of the DNA sequences within the (--SEA) deletion in regulating the expression of zeta-globin gene in cis during normal human hemoglobin ontogeny.

Analysis of the human zeta-globin gene promoter in transgenic mice

zeta-Globin is the embryonic form of the alpha chain of hemoglobin. Transgenic mice generated with zeta-globin constructs containing the zeta-globin gene, 557 bp of 5′ flanking sequence, and 2-kb of 3′ flanking sequence linked to the beta-globin locus control region hypersensitive site 2 (HS2) expressed human zeta-globin only in embryonic yolk sac erythroid tissue, and not in definitive erythroid tissue in the fetal liver or in adult peripheral blood. To determine what sequences in the 5′ flanking region of the zeta-globin gene might be important for developmental specificity, a series of 5′ deletion constructs of the zeta-globin gene were made and used to generate transgenic mice. The 5′ ends of these constructs were located 417, 207, and 128 bp 5′ to the zeta-globin transcriptional start site, and HS2 was included to increase the level of erythroid-specific expression. In all lines of mice tested, human zeta-globin was expressed only in embryonic tissue, and not in fetal livers or in adult peripheral blood. Expression was independent of copy number and appeared to be dependent on the site of transgene insertion. These data suggest that the proximal 128 bp of the zeta-globin promoter is sufficient to properly regulate zeta-globin expression during development.

Analysis of the human zeta-globin gene promoter in transgenic mice

zeta-Globin is the embryonic form of the alpha chain of hemoglobin. Transgenic mice generated with zeta-globin constructs containing the zeta-globin gene, 557 bp of 5′ flanking sequence, and 2-kb of 3′ flanking sequence linked to the beta-globin locus control region hypersensitive site 2 (HS2) expressed human zeta-globin only in embryonic yolk sac erythroid tissue, and not in definitive erythroid tissue in the fetal liver or in adult peripheral blood. To determine what sequences in the 5′ flanking region of the zeta-globin gene might be important for developmental specificity, a series of 5′ deletion constructs of the zeta-globin gene were made and used to generate transgenic mice. The 5′ ends of these constructs were located 417, 207, and 128 bp 5′ to the zeta-globin transcriptional start site, and HS2 was included to increase the level of erythroid-specific expression. In all lines of mice tested, human zeta-globin was expressed only in embryonic tissue, and not in fetal livers or in adult peripheral blood. Expression was independent of copy number and appeared to be dependent on the site of transgene insertion. These data suggest that the proximal 128 bp of the zeta-globin promoter is sufficient to properly regulate zeta-globin expression during development.

Detection of the hemoglobin E mutation using the color complementation assay: application to complex genotyping

The color complementation assay (CCA) is a method of allele-specific DNA amplification by which competitive priming and extension of fluorescently labeled oligonucleotide primers determine the color of DNA amplification product. This diagnostic method precludes the need for radioisotopes, electrophoresis, and multiple high-stringency reaction conditions. The multiplicity of mutant globin genes present in Southeast Asians complicates clinical diagnosis and underscores the importance of DNA-based diagnostic methods. We have applied CCA to distinguish beta A and beta E alleles. Competing 15mer primers were a fluorescein-labeled complement to beta A and a rhodamine-labeled complement to beta E, identical except for their central nucleotides. A common unlabeled primer was used to amplify DNA product, the color of which was determined by the perfectly complementary primer. Color photography and spectrofluorometry, as well as a method of black-white photography that we developed to distinguish fluorescein- and rhodamine- labeled DNA, were used to record results. We applied CCA to define the complex genotype of a Thai woman with thalassemia intermedia, 96% HbE, and 4% HbF whose possible genotypes included several permutations of alpha-thalassemia, beta-thalassemia, and beta E genes. zeta-Globin gene mapping of DNA doubly digested with Bg/II and Asp 718 showed the -alpha 3.7/--SEA genotype, and CCA confirmed homozygous beta E/beta E. The CCA is useful for diagnosing the compound hemoglobin genotypes of Southeast Asians and could be applied also to prenatal diagnosis in this population.