Observations on the change from foetal to adult erythropoiesis.

The replacement of foetal haemoglobin (Hb F) by the adult form (Hb A) in the circulating blood of an infant is a striking event, but its mechanism is incompletely understood. The dramatic cessation of haemopoiesis in the liver at about the time of birth coincides roughly with the time when Hb F begins to disappear most rapidly from the peripheral blood; and it is not surprising that some writers, most recently Langley (1951) and Brody (1959), have taken the view that the two events are closely connected, and assume that Hb F production is limited to the liver, and Hb A production to the marrow. This view would require that the change-over was the result of the emergence and ultimate domination of erythrocyte precursors in the marrow which, because of their genetic equipment or their environment, were from the start capable only of the adult type of haemoglobin synthesis. Such an interpretation is not acceptable to all, however, and the view is gaining ground that the two types of haemoglobin molecule can be synthesized by normoblasts in each of the main anatomical sites of erythropoiesis. One requirement of this view-that a single normoblast can manufacture two types of molecule-is met by the undoubted facts that whenever Hb S occurs with another Hb type, both haemoglobins are present in the same cell, since all the cells can be made to undergo sickling; and that in adults who are heterozygous for the normal and for the 'high-F' gene, Hb F is demonstrable in every erythrocyte, and therefore each must contain both Hb F and Hb A (Thompson, Mitchener and Huisman, 1961). On this second view the change-over would be explained by the progressive maturation of the intracellular enzyme systems concerned in synthesis, taking place at about the time of birth in all erythropoietic cells whatever their site, so that successive generations of normoblasts would manufacture more Hb A and less Hb F. The main difficulty in the past has been that it has only been possible to study the Hb F and Hb A content of the blood in bulk, and to relate the proportions of the two types of haemoglobin in the whole peripheral blood to the actual mass of erythropoietic tissue in the viscera and marrow respectively. This gives only a very general correlation; the amount of Hb F in the blood is reduced from about the time that the mass of blood-forming tissue in the marrow exceeds that in the viscera (Custer, 1949). Consequently other methods have been sought. Attempts have been made to determine whether different proportions of Hb F exist in blood withdrawn from the liver and marrow respectively in the perinatal period (Jonxis, 1949; Schulman, 1959). These have not yielded any definite result, as might indeed be expected, since aspiration at either site is liable to sample not only the blood formed locally, but also much mixed peripheral blood. The critical fact, if it could be determined, is what type of haemoglobin is present in the local normoblasts that have not yet entered the circulation. A more promising way of investigating the synthetic activity of haemopoietic cells is to study them in culture in vitro. Thomas, Lochte, Greenhough and Wales (1960) incorporated glycine containing A4C in liver, spleen and marrow cells from a 1 7-week foetus by incubating them for 24 hours in vitro. They then separated haemoglobins A and F from a haemolysate of the cells by the device of adding a proportion of 'carrier' Hb A. By determining the specific activity of the two fractions obtained by electrophoresis, they were able to estimate the relative capacities for forming the two haemoglobin types for cells from each of the three sites. Relatively more Hb F was produced by cells from the liver and spleen than by those from the marrow; nevertheless, the bone marrow produced 1-8 times as much Hb F as Hb A, and the viscera of this foetus, and of a 9-week foetus, produced Hb A in amounts varying from 1/4 8 to 1/2 3 of the Hb F manufactured. Perhaps these results, especially the last-