Developmental effect of the XmnI site on Ggamma-globin gene expression among newborn Hb F-Malta-I [Ggamma117(G19)His-->Arg, CAT-->CGT] heterozygotes and adult beta+ -Thalassemia homozygotes.

Hb F-Malta-I [Ggamma117(19)His-->Arg, CAT-->CGT] is a stable and benign variant of Hb F found in 1.8% of Maltese newborn. We studied 120 Hb F-Malta-I heterozygotes and four Hb F-Malta-I homozygotes. The mean proportion of Ggamma-F-Malta-I in Hb F was 0.26 +/- 0.03 for the Hb F-Malta-I heterozygotes and 0.58 +/- 0.06 for the Hb F-Malta-I homozygotes. The Hb F-Malta-I allele was shown to occur on a background of the common Mediterranean haplotype Va [+ + - - - - - + + -]. Furthermore, the common Mediterranean haplotypes Va, IIIb [- + + + - + + + + -], I [+ + - - - - - + + +] and II [- + - + + - + + + +] accounted for most (66.2%) of the wild-type alleles among the tested Hb F-Malta-I heterozygotes. Different genotypes at the 5' epsilon HincII, Ggamma and Agamma HindIII, and 3'psibeta HincII sites (but not at the 5' Ggamma XmnI site) were found to be linked to significant variations in the proportion of Ggamma-F-Malta-I and Ggamma-globins in the Hb F of newborn Hb F-Malta-I heterozygotes. Moreover, the 5' Ggamma XmnI site was found to be associated with variations in Hb F and Ggamma-globin levels in a population of adult Maltese beta-thalassemia (thal) homozygotes. This implies that a determinant linked to the XmnI site which effects Ggamma-globin gene expression is active in anemic adults but not in normal infants.

A review of cis-trans interplay between DNA sequences 5' to the (G)gamma- and beta-globin genes among Hb F-Malta-I heterozygotes/homozygotes and beta-thalassemia homozygotes/compound heterozygotes, and the effects of hydroxyurea on the Hb F/F-erythrocyte; the need for large multicenter trials.

The biosynthesis of Hb F in place of the deficient Hb A could be a suitable treatment for beta hemoglobinopathies. Among newborn Hb F-Malta-I heterozygotes, it could be shown that the XmnI sequence alone had little, if any effect on gamma-globin gene expression, but interplay with the (AT)(X)T(Y) sites in cis and in trans may occur. In contrast, while the XmnI sequence is clearly correlated with gamma-globin levels in anemic adult beta-thalassemia (thal) homozygotes, the effect on F-erythrocyte numbers and Hb F/F-erythrocyte appears independent of the (AT)(X)T(Y) sites. Even at levels of hydroxyurea (HU) as low as 1.65 mg/kg/day (vs. 10 mg/kg/day on the high dose regime) it can be shown that although even a small increase of Hb F could be obtained, the effect was rarely translated into an increase in circulating hemoglobin (Hb). In most cases, the elevated Hb F level was dependent on the XmnI sequence and was due to increased numbers of F-erythrocytes or Hb F/F-erythrocyte or both. It seems that the bone marrow of thalassemia homozygotes may be more sensitive to myelosuppression by HU possibly due to medullary inflammation. While the data are consistent with loop models of globin switching mechanisms, there is urgent need for large, hypothesis driven, multicenter trials of molecules that could maintain or re-induce high Hb F levels in beta-thal and subject to genetic and epigenetic constraints including inflammation.