blood the journal of - the journal of the american socieq of hematology vol 78, no 9 november 1,...

Download BLOOD The Journal of - The Journal of The American SocieQ of Hematology VOL 78, NO 9 NOVEMBER 1, 1991

Post on 19-Sep-2019




0 download

Embed Size (px)


  • BLOOD The Journal of The American SocieQ of Hematology VOL 78, NO 9 NOVEMBER 1, 1991


    Hemoglobin A,: Origin, Evolution, and Aftermath

    By Martin H. Steinberg and Jun ius G. Adams 111

    H E MELANGE of hemoglobins present in the erythro- T cytes of humans includes hemoglobin A, (HbA,),’., a tetramer of a- and &globin chains (a26,). Its unique characteristics reside in the n~n-a-chain.~ HbA, is physio- logically unimportant because it is normally less than 3% of the total Hb. The study of HbAz has provided insights into the evolution and phylogeny of globin genes and enhanced our understanding of gene expression and globin synthesis. HbA, has substantial clinical relevance. Its concentration fluctuates in the thalassemia syndromes and some acquired diseases, so that its measurement provides a useful diagnos- tic aid. This review will focus on the structure, function, and synthesis of the &-globin chain and HbA,, as well as the features of this Hb that give it clinical utility.


    Linkage relationships and chromosomal location of the &globin gene. It has long been known that the p- and &globin genes were closely linked. These initial linkage data were derived from the study of families in which both p- and &globin variants were ~egregating.~.“’ The location of the p-globin gene family on the short arm of chromosome 11 involved a variety of molecular techniques.”.” The linear arrangement of the @-globin gene cluster was determined from data derived primarily from gene mapping and is shown in Fig 1.”’-*‘ These genes have also been completely sequenced.”-” The general form of the &gene is akin to all other globin genes with three coding regions (exons) and two intervening sequences (introns).

    Because of the large body of sequence data that is available from the globin genes as well as the proteins that they produce, this gene family has provided an excellent opportunity to examine its molecular evolution. The globin genes are all thought to have arisen from a common globinlike heme protein (Fig 2). The earliest duplication of this ancestral gene led to the diver- gence of myoglobin and the globins that comprise Hb and most likely occurred approximately 700 million years before present (Mybp).32 The next duplication gave rise to the divergence of the a-globin genes and occurred approxi- mately 450 Mybp.32 The P-globin gene is thought to have duplicated about 180 to 200 Mybp into an ancestral gene for E- and y-globin and an ancestral gene for 6- and P-gl~bin.’~ Approximately 110 to 130 Mybp, the E/? parent gene diverged to establish the E and y lineages.

    Evolution of the &globin gene.

    The evolution of the &globin gene was initially very confusing, because HbA, was present in humans, apes, and New World monkeys, but not in Old World monkeys.34” This finding was seemingly at variance with the evolutionary data which indicated that humans and Old World monkeys diverged after the divergence from New World Monkeys. This conundrum was ultimately solved with the finding that &globin genes are indeed present in Old World monkeys, but have been inactivated by mutation.38 42

    After examination of the gene sequence of a number of primate species, the origin of the &globin gene was first thought to have occurred relatively recently (about 40 nlillion years However, studies of the globin genes of mice, rabbits, and other primates make it more likely that this divergence occurred before the mammalian radiation approximately 85 to 100 Mybp, at about the same time as the E- and y-globin genes diverged.M4n It is clear from the comparisons of &globin genes among mammals that the &globin locus has not evolved as an independent lineage, but has evolved in concert with the p-globin gene. In each of the mammalian orders examined to date, the &globin locus has acquired characteristics of the P-globin locus through gene conversion (a nonreciprocal exchange of genetic material between the two linked homologous

    These gene conversions have most often occurred in the coding regions, rendering these regions useless in the quest for the primordial %globin gene. Thus, the evolutionary origin of the &globin gene has been performed using flanking and intervening sequence data (especially IVSII ) .~~

    The nonallelic gene conversion events postulated to occur during the evolution of the &globin gene are thought to have been rare when compared with the gene conversion between the two y-globin loci.44 However, Petes” has made an interesting observation concerning the structural variants of HbA,.

    genes).” 44 4‘47 cu

    Gene conversion and the &globin gene.

    From the Department of Medicine, University of Mississippi School

    Submitted June 17, 1991; acceptedAugust 21, 1991. Supported by Research Funds of the Department of Veterans Affairs. Address reprint requests to Martin H. Steinbe%, MD, (151), VA

    This is a US government work. There are no restrictions on its use.

    of Medicine, and VA Medical Center, Jackson, MS.

    Medical Center, 1500 E Woodrow Wilson Dr, Jackson, MS 39216.

    0006-4971 191 17809-0049$0.00/0

    Blood, Vol78, No 9 (November 1). 1991: pp 2165-2177 2165


    Chromosome 11

    H 5kb

    Fig 1. The arrangement of the plike globin gene cluster.

    Two of the then-described 10 &-globin variants could have arisen by gene conversion. The &globin variants HbA, Flatbush (622 ala + glu) and HbA, Coburg (6116 arg - his) both contain a P-globin amino acid in 1 of the 10 amino acid residues where these two globins differ. The occurrence of these variants is much greater than would be expected for random mutation. Therefore, Petes suggested that these two variants could represent gene conversion events. An- other variant that was found subsequent to Petes’ hypothe- sis, Hb Parchman (622 ala + glu, 50 ser thr), could also be due to gene conversion rather than to a double crossover as proposed by the investigators.s2 This hypothesis is easily testable in the variant HbA, Coburg, because a gene conversion event would result in a codon 116 change of CGC to CAT, while a point mutation would result in a CGC to CAC change.’’

    Clearly gene conversion has maintained strong sequence homology between the human P- and &globin genes. This homology is especially striking in the coding regions where there are 10 amino acid differences and 31 nucleotide differences. However, the quantitative expression of these two genes is strikingly different. HbA makes up more than 95% of the adult hemolysate, while HbA, comprises only 2% to 3%. The molecular stability of these two molecules appears to be almost identical, making it highly unlikely that difference in posttranslational survival of the two molecules accounts for the low proportion of HbA,. In studies where reticulocytes were incubated with radioactive precursor amino acids, it became apparent that the &globin chain is synthesized at a reduced rate in the bone marrow and not at all in reticulo- cyte~.’~.’~ These experiments were extended by fractionating bone marrow erythroblasts into fractions of different levels

    Synthesis of the &globin chain.

    P I My b.p. A -190

    A A -130 p c* 85 (Mammalianradiation) e Y 6

    Fig 2. The evolutionary history of the &-globin gene.

    of maturity. These experiments showed that there was a progressive decrease in &-globin synthesis in relation to P-globin in increasingly mature cells. A relative instability of 8-globin mRNA was proposed as a mechanism for the premature decrease in &globin synthesis? Using highly selective probes for P- and &globin mRNA, it was found that the half-life of &globin mRNA was less than one third that of P-globin mRNA, supporting this hypothesis.”

    Despite the strong evidence that 6-globin synthesis de- creases during the maturation of erythroid precursors, this decrease does not account for the great discrepancy be- tween P- and &globin synthesis. In normal bone marrow cells, the synthesis of &globin chains is less than 2% of total non-a-globin synthesis. In addition, even in the youngest fraction of bone marrow erythroid cells examined, the 6:P-globin synthesis ratio did not exceed the usual HbA2: HbA ratio found in the peripheral blood. Furthermore, the translation rates of the two globins were the ~ a m e . ~ ‘

    These findings strongly suggested that the rate of tran- scription of S-globin gene must be less than that of the P-globin gene. When the in vitro transcription of non-a- globin genes was compared, it was found that P-, E-, and y-globin genes are transcribed with equal efficiencies, but that the transcription of the 6-globin gene is far less efficient.” Humphries et a P compared the expression of a-, P-, and &globin genes in monkey kidney cells. Under conditions that promoted optimal transcription of each globin gene in this system, &globin gene transcription was found to be 50 times less efficient than that of the P-globin gene. This transcriptional deficiency of the &globin gene approximates the synthesis of &globin in normal erythroid cells. Humphries et al also made hybrid constructs of the 6- and P-globin genes. When the 5‘ end of the &globin gene was replaced by the homologous portion of the P-globin gene, transcription of this hybrid gene was equal to that of the normal P-globin gene. However, when the 5’ portion of the p-globin gene was replaced by the ho


View more >