American Journal of Hematology 6 : 17-25 (1979)
Heterogeneity of Hemoglobin Gamma Chains in Normal Newborns and in Cases of Alpha and Beta Thalassemia Luan Eng Lie-lnjo, 2.1. Randhawa, J.P. Kane, J. Ganesan, and R. George G. W. Hooper Foundation and international Center for Medical Research (ICMR), University of California School of Medicine, San Francisco (L. E. L.-I., Z. 1. R.), Department of Medicine and the Cardiovascular Research Institute, University of California School of Medicine, San Francisco (J.P. K.), Institute for Medical Research, Kuala Lumpur, Malaysia (J. G.), and Pediatric Unit, General Hospital, Kuala Lumpur, Malaysia (R.G.)
The ratio of Gy t o Ay chains was studied in 1 3 normal healthy newborns and in eight neonates with hydrops fetalis due t o homozygous a-thalassemia. The findings in the normal healthy newborns agreed with those of earlier reports. In homozygous a-thalassemia the Gy and Ay ratio appeared generally lower than in healthy control newborns, but o n e of the hydrops fetalis cases had a very high Gy value. In all 1 3 normal healthy newborns, in 8 patients with homozygous athalassernia, and in 4 patients with homozygous 0-thalassemia, T y chains with threonine a t position 75 were detected in addition t o Iy chains with isoleucine at position 7 5 . In homozygous a-thalassemia, the Ty-to-Iy ratio seemed lower and in homozygous 0-thalassemia higher than in normal newborns. Key words: heterogeneity of gamma chains, newborn, thalassemia
INTRODUCTION
Schroeder e t a1 [ l ] reported two types o f gamma chains in fetal hemoglobin depending on whether glycine or alanine occupied position 136 of the gamma chain. The synthesis of these chains, designated Gy and Ay, is controlled b y nonallelic loci [ 1-31 . Although b o t h G y and Ay chains show identical electrophoretic and chromatographic behavior, b y studying the C-terminal cyanogen bromide peptides (Gy CB-3 and Ay CB-3) they can be differentiated and quantitated.
Received for publication July 15, 1978; accepted November 25, 1978.
Address reprint requests to Dr. L.E. Lie-Injo, G.W. Hooper Foundation, University of California, San Francisco, CA 94143. 0361-8609/79/0601-0017$02.00@ 1979 Alan R. Liss, Inc.
18
Lie-Injo et a1
to In newborns from many different parts of the world, the ratio of synthesized Ay chains is approximately 3: 1 [ 1-31 . By the time the baby is 6 months old, the ratio has become 2 Gy t o 3 Ay, the same ratio as seen in adults [4]. The problem of heterogeneity of fetal hemoglobin became more complicated when it was shown that on the gamma chains, the amino acid occupying position 75 can differ [5-61; it can be either isoleucine (Iy) or threonine (Ty). The reported findings of Ricco et a1 [6] suggest that the synthesis of the Ty chain is controlled by a gene nonallelic with those coding for Gy and Ay chains. The presence of Ty chains in normal newborns and different disease conditions was confirmed by Huisman et a1 [ 7 ] and in the present study [ 8 ] . The possible specific function of the different gamma chains and the factors influencing control of their synthesis are still obscure. Study of the relative proportions synthesized under differing abnormal conditions might lead t o a better understanding of their relative significance. We report the heterogeneity of gamma chains of fetal hemoglobin in homozygous a thalassemia [9] and in /3 thalassemia, as well as in normal healthy newborns. METHODS
Cord blood samples from 13 normal healthy newborns were obtained from the delivery room of Moffitt Hospital, San Francisco, and placed in acid-citrate dextrose (ACD) solution. Blood from eight neonates (seven Chinese and one Malay) with hydrops fetalis due to homozygous CY thalassemia and from patients with homozygous /3 thalassemia (one Chinese and three Malay) were obtained from the General Hospital, Kuala Lumpur, Malaysia. The diagnoses of these conditions were supported by hematologic and biochemical studies in the relatives, but no globin chain synthesis studies could be carried out. The washed, packed red blood cells were flown in dry ice to San Francisco. Hemolysates were prepared by the standard method using deionized distilled water and toluene. Hb Bart's from the hemolysates of the babies with hydrops fetalis was isolated and purified according to the inverted-bottle method of Lie-Injo [ 101 . Globin was prepared by acid-acetone precipitation. The study of cyanogen bromide (CB) digests was carried out according to Schroeder et a1 [ 11 , as modified by Nute et a1 [l 11 ; however, four columns of Sephadex G-50 (fine) in series, each column being 1.6 X 160 cm, were used. For each study, the cyanogen bromide digest of 80-150 mg globin was placed on the column. The CB fragments were hydrolyzed under vacuum in 6 N HCl for 48 hours at 108"C, and the hydrolyzed CB fragments were analyzed with a two-column technique on a Beckman Model 121 M amino acid analyzer. For the study of residue 75 of the gamma chain, trypsin digests of the globin of fetal hemoglobin or purified gamma chains were fingerprinted and the tryptic peptide 9 was characterized by amino acid analysis [12]. When Ty chains were present, peptide 9b was observed in addition t o the usual peptide 9a (Fig. 2). The peptide 9b was identified by amino acid analysis with a Beckman 121 M automatic amino acid analyzer, wliichis capable of determining amino acids in the nanomole range. The relative proportion of Iy and Ty was evaluated by staining the fingerprint with cadmium acetate, eluting peptides 9a and 9b, and estimating their relative optical density at 500 nrn [13]. (This staining method was employed at the advice of Dr. L.F. Bernini.)
Hemoglobin Gamma Chains
19
RESULTS
Because Ay CB-3 peptide has three residues of alanine and none of glycine, and Gy CB-3 peptide has only one glycyl and two alanyl residues, the proportion of Gy chains in a mixture containing Gy and Ay chains can be readily calculated from the number of gIycine residues, while the proportion of Ay chains is obtained by quantitating any alanine residues in excess of two. We obtained quite satisfactory separation of cyanogen bromide peptides because the relevant y CB-3 is entirely free of other fragments. The chromatogram of the CB peptides obtained from globin for one of the normal newborns is shown in Figure la. All chromatograms of the CB peptides prepared from the HB Bart's globin (which consists solely of y chains) of the babies with hydrops fetalis resembled one another. One is shown in Figure Ib. All chromatograms showed a complete separation of the relevant y CB-3 fragment. The different peaks, identified by amino acid analysis of the hydrolysates, are indicated in the figures. The amino acid composition of the first two peaks did not agree with the expected composition of any of the CB fragments of fetal y chains but they most resembled y CB-2.
a
mi e f f l u e n t
E 0 00 N
+ m
b Fig. 1. Chromatogram of cyanogen bromide peptides of globin prepared from hemolysate. Four 1.6 X 160 cm columns of Sephadex G-50 (fine) in series were developed with 1% formic acid. a) From a healthy newborn; b) from a case of homozygous CY thalassemia.
Lie-Injo et a1
20
The relative proportion of 0.75 glycyl and 2.27 alanyl residues obtained for our normal controls (Table I) agrees well with the finding of Schroeder and co-workers [2] who found the proportion of glycyl and alanyl residues in the peptides prepared from more than 100 cord blood samples obtained from different areas of the world to be 0.71 and 2.32 respectively. Our results also agree with those of Nute and colleagues [ l l ] , who found the ratio 0.74 glycine to 2.30 alanine in nine normal newborns. In our cases of hydrops fetalis, the estimation of the residues in the y CB-3 peptide was carried out in duplicate, usually on two or three independently isolated y CB-3 peptides. Results are listed in Table I. The ratio of Gy to Ay chains in the cases of hydrops fetalis appears t o be generally lower than the mean found in normal newborns except for Case 5. If Case 5 was not included, statistical analysis using a standard t test showed a highly significant (P < 0.001) difference between the group of hydrops fetalis and the group of normal newborns. However, if Case 5 was included, the difference was not significant (P > 0.05). The ratio of Gy t o Ay chains found in the isolated Hb Bart’s does not differ significantly from the ratio obtained from the globin of whole hemolysate. In all subjects, the sum of the glycyl and analyl residues per y CB-3 peptide does not deviate more than 0.09 from the expected sum of 3.00. The relative proportions of Ty and Iy chains in normal newborns, and in cases of homozygous LY thalassemia and /3 thalassemia, are listed in Table 11. Figure 2a,b shows the fingerprint of trypsin digests of Hb F from a normal newborn and a patient with p thalassemia and Figure 3a,b shows fingerprints of pure y chains from cases with homozygous LY thalassemia. In all our normal newborns studied, the Ty chain was present. It was noticed that in a few cases with a low level of Ty chains, the 9b peptide was not demonstrable or barely evident at the first examination, but appeared in a second study if we increased the amount of tryptic digest applied for fingerprinting. In one case with Ty chain levels below 15%, the peptide 9b was examined by amino acid analysis t o make sure that the peptide represented peptide 9 of the Ty chain. Of those samples in which peptide 9b TABLE 1. Heterogeneity of Fetal Hemoglobin: y136 (cyanogen bromide digest) Normal healthy newborns No.
Glycine
Alanine
1 2 3 4 5 6 7 8 9 10 11 12 13
0.80 0.79 0.73 0.74 0.75 0.72 0.76 0.77 0.72 0.78 0.69 0.72 0.77
2.21 2.20 2.35 2.26 2.26 2.29 2.23 2.25 2.30 2.24 2.33 2.37 2.23
Mean
0.75
2.27
G, = 75%
Ay = 27%
Homozygous o! thalassemia (Hydrops fetalis)
No.
Mean
Glycine
Alanine
0.65 0.66 0.64 0.70 0.92 0.70 0.64 0.63
2.39 2.35 2.45 2.28 2.16 2.30 2.30 2.33
0.71
2.32
C, = 69%
ky = 32%
21
Hemoglobin Gamma Chains
TABLE 11. Heterogeneity of Fetal Hemoglobin: 7” (tryptic digest) Normal healthy newborns
Homozygous a thalassemia (hydrops fetalis)
Homozygous p thalassemia
‘7%
11 12 13
90 82 85 80 93 78 79 94 80 88 80 85 77
10 18 15 20 7 22 21 6 20 12 20 15 23
Mean
83.9
16.1
1 2 3 4 5 6 7 8 9 10
95 88 79 79 83 84 92 94 Mean
81
5
1
12 21 21
2 3 4
72 15 I0 I7
28 25 30 23
13.5
26.5
17 16 8 6
Mean
13
Fig. 2. Fingerprints of tryptic digest of Hb F (a + y) chains. a) From a normal healthy newborn; b) from a case of homozygous p thalassemia.
spot was conspicuous, many, but not all, were analyzed for the 9b peptide amino acid composition. In cases of homozygous a thalassemia, the level of Ty chains is generally lower than in normal healthy newborns; in the small number of cases of homozygous 0 thalassemia examined in this study, the level appeared t o be higher. DISCUSSION
The synthesis of Gy and Ay chains is under the control of nonallelic loci [ 1-31 . IR attempts to better understand the specific function and possible purpose for production of different types of y chains in the fetus, as well as the regulation of the synthesis of the different types of y chains, research workers have studied abnormal conditions in which fetal hemoglobin is produced. The ratio of Cy t o Ay chains has been studied in physiologic
22
Lie-Injo et a1
Fig. 3. Tryptic digest fingerprints of y chains from two cases of hydrops fetalis homozygous a! thalassemia. a) With a very low amount of Ty chains. (Amino acid analysis of peptide 9b showed that, in comparison to peptide 9a, it had an additional threonine residue and no isoleucine.) b) With a higher amount of Ty chains.
conditions and abnormal gentic as well as acquired conditions associated with the presence of fetal hemoglobin. During postnatal development of the newborn, the ratio of Gy t o Ay was observed to change from about 3: 1 at birth to about 2:3 toward the end of the first year [4], which led to the conclusion that the mechanism of y-gene deactivation involves an unequal repression of the Gy and Ay structural genes. It was further observed that an abnormal fetal hemoglobin variant with either glycine or alanine at position 136 differs in the level present in the blood [ 141. This led Huisman et a1 [14] t o postulate the existence of four y loci - two Gy loci, one producing more than the other; and two Ay loci, one producing more than the other - which contribute t o the total ratio of 4:2:2: 1 at birth. This hypothesis was strengthened by the observation that two Gy types of Hb F variant decline at different rates after birth in the heterozygote [15]. During the initial months after birth the activity of the major Gy gene declines, allowing the Ay chain t o dominate to the ratio of 2 Gy:3 Ay in the older child or adults. Direct estimate of the number of y-globin genes indicates that there are at least two y-chain genes per haploid human genome, but that the presence of four y-chain genes is unlikely [16]. In 1977 Huisman et a1 [17], in an attempt t o come t o a compromise, suggested that there might be three y-chain genes, two Gy-chain genes, one major and one minor, and one Ay-chain gene (Gy, gy, Ay). Up to the time of birth the ratio of product or the activity of these genes is assumed to be G.g.A, contributing to the total in a ratio 4:2:3. The postnatal decline from 0.7 t o 0.4 residues of glycine is explained by a nonsynchronized suppression of the postnatal activity of the major Gy gene while the minor Gy and the Ay genes remain active. In different types of 0 thalassemia, in hereditary persistence of fetal hemoglobin (HPFH), in patients with various hemoglobinopathies, and in acquired conditions associated with the presence of increased amounts of fetal hemoglobin, the Gy-to-Ay ratios vary greatly. Attempts t o correlate the ratio with certain conditions and thus shed more light on why different types of fetal hemoglobins are produced have not been very successful. In certain cases of hereditary persistence of fetal hemoglobin thought t o be caused by a deletion of the 0-and 6-globin genes, the fetal hemoglobin produced can be of both types, C q and Ay, or can be Gy only or Ay only [18-201. If either the Gy or Ay gene were deleted along with the 80 gene complex, only Gy or Ay would be produced. If the SO gene
Hemoglobin Gamma Chains
23
complex were deleted, but the y genes continued functioning, Gy as well as Ay chains would be produced. The ratios of Gy to Ay in patients with a thalassemia ill which the QI gene is deleted [21,22] have not been examined before, although the ratio in Hb Bart’s (74) in a black newborn carrier has been reported [ 11 . In our present study, eight cases of homozygous a thalassemia were examined. In one case (Case 5),the y chain appeared t o be almost entirely of the Gy type, while in the others the ratio of Gy t o Ay was generally lower than in our normal healthy newborn controls. Further study is needed to ascertain the significance of these findings. However, the findings cannot be explained, as is done for the HPFH, by assuming that the Gy or Ay chain gene is or is not deleted along with the a-chain gene, because the a-chain gene is located on chromosome 16 [23] and y and p genes are located on chromosome 11 [24] . In our cases, the production of the different types of y chain must have been under the control of genes other than the structural ones. Extensive recent studies of Huisman et a1 [17,25] add t o the complexity of the problem. They found in a black study population that fetal hemoglobin in newborns and adults showed a bimodal distribution of glycine values, that different patterns of decline of e - t o - A y ratio existed postnatally, and that, rather than a distinct “adult glycine value,” there was a variable ratio in adults. The existence in some individuals of Ty chains with threonine at postion 75 in addition to Iy chains with isoleucine [5, 61 poses another intriguing problem. Ricco et al [6] found +40% of normal newborns and premature infants t o have Hb F with Ty chains in addition to Hb F with Iy chains. Hb F with Ty chains was also found in almost all their patients with homozygous /3 thalassemia, in a 14-week-old fetus, and in one of three patients suffering from aplastic anemia with raised levels of Hb F. Their findings suggest that the synthesis of Ty chains is controlled by a gene nonallelic t o the genes coding for Ay and Gy chains. The presence of Ty chains in healthy normal newborns and several disease conditions has been confirmed by findings of Huisman et a1 [7] and by the findings presented in this paper. However, there are discrepancies between the findings of Ricco et al [6], those of Huisman et a1 [7] and our findings. Only a proportion of the newborns and patients were found to have Ty chains by Ricco et al [6] and by Huisman et a1 [7]. Our study showed the presence of Ty chains in all 13 normal, healthy newborns, 8 cases of homozygous a thalassemia (7 Chinese and 1 Malay) and 4 cases of homozygous /3 thalassemia (3 Chinese, and 1 Malay), if enough tryptic digest was applied t o the fingerprint. However, the number of newborns we examined was small and study of more samples may turn up negative cases. Nevertheless, the difference in findings are quite obvious and we have no explanation for this discrepancy. Further, while Ricco et a1 [6] found levels of lower than 10% in the positive cases, most of the positive cases examined by Huisman et a1 [7] in a larger group had levels around 20%. This last group also found that different ethnic groups showed widely different percentages of positive cases. Most of our newborns had Ty chain levels around 20%, a level which agrees with those of Huisman et a1 171, but we also found cases with levels below lo%, similar t o those in the group of Ricco et a1 [ 6 ] . Again, the reason for these discrepancies is not clear. Huisman et a1 [7] suggested that those who do not show the presence of Ty chains in laboratory tests do not carry the gene for Ty chains. Those who do show the presence of Ty chains in laboratory tests do carry Ty chain gene. If so, it is difficult to explain the findings of Ricco et a1 [ 6 ] , which showed that only 38.9% of their normal newborn babies had Ty chains (Ty gene) while almost all (90.6%) of their cases of homozygous /3 thalassemia had Ty chains (Ty gene). It may be that all or almost all individuals d o carry the gene for Ty chains, but that the presence of the Ty chains in the blood is not always detectable by laboratory methods and that the
24
Lie-Injo et a1
expression of this gene is increased in certain types of homozygous fl thalassemia but not in other types. This would explain the findings of Ricco et a1 [6] that only 38.9% of their normal newborns had the Ty chains while in almost all their cases of homozygous p thalassemia the Ty chains were easily detectable. Most of the cases of 0 thalassemia studied so far by all groups had Ty chains, and the levels of the Ty chains tend to be higher than those in normal healthy newborns. In our cases of homozygous p thalassemia the values appear generally higher, and in homozygous a! thalassemia they appear lower, than in normal healthy newborns. The difference between the mean level of Ty chains in the homozygous thalassemia group and in the homozygous a! thalassemia group is highly significant (P < 0.01). Interestingly, Huisman et a1 [7] did not find the Ty chains in seven black 0 thalassemia homozygotes, while they did find it in homozyogous 0thalassemia of Italian-Greek origin. In all the cases of homozygous fl thalassemia listed in the papers of Ricco et a1 [ 6 ] , of Huisman et a1 [7] , and in our own group, the type of homozygous thalassemia was not determined. In contrast to cr thalassemia, there are many variants of homozygous p thalassemia known, even in the same ethnic groups. The different findings in different ethnic groups may be due to the different types of homozygous 0 thalassemia, as was suggested by Huisman et a1 [7] . From the data obtained so far, it can still not be concluded whether the Ty chain is the product of an allele of the Gy or A-y chain structural genes or of an additional nonallelic gene. Further, the available data do not lend themselves t o an easy interpretation of why different types and levels of y chains are produced, nor d o they explain the mode of control of their synthesis, but it is clear that many factors other than the structural genes, still unknown to us, play a role in the control of y chain synthesis. ACKNOWLEDGMENTS
We would like to thank the Department of Microbiology, University of California (Dr. J.W. Goodman), San Francisco, for the use of its amino acid analyzer Beckman Model 1 1 8 in the early part of our study; Ms Ksenia Tomitch of the Cardiovascular Research Institute, San Francisco, for performing the amino acid analysis on the Beckman No. 121 M amino acid analyzer, and Mr. Alex Herrera for valuable technical help; the staff of the delivery room of the Moffitt Hospital, San Francisco, for collecting cord blood samples; and the staff of the delivery room, General Hospital, Kuala Lumpur, Malaysia, for referring cases of hydrops fetalis to us. This work was supported by research grants AM 25157, HL 10486, UC ICMR A1 10051 and HL 14237 from the National Institutes of Health, Public Health Service, and by the Committees on Research of the University of California and the School of Medicine, University of California, San Francisco. J.P. Kane was an Established Investigator of the American Heart Association during these studies. REFERENCES 1. Schroeder WA, Huisman THJ, Shelton JR, Shelton JB, Klqihauer EF, Dozy AM, Robberson B: Evidence for multiple structural genes for the 7 chain of human fetal hemoglobin. Proc Natl Acad Sci 60537-544, 1968. 2 . Schroeder WA, Shelton JR, Shelton JB, Ape11 G, Huisman THJ, Bouver NG: World-wide occurrence of nonallelic genes for the 7-chain of human foetal haemoglobin in newborns. Nature (New Biol) 240:213-274, 1972.
Hemoglobin Gamma Chains
25
3. Schroeder WA, Huisman THJ: Investigations of molecular variation in human fetal hemoglobin in the infant and in certain hematological conditions in the adult. In “Protides of the Biological Fluids, Proceedings of the 17th Colloquium, Bruges, 1969.” Oxford: Pergamon, 1970, pp 249-255. 4. Schroeder WA, Huisman THJ, Brown AK, Uy R, Bouver NG, Lerch PO, Shelton JR, Shelton JB, Apell G: Postnatal changes in the chemical heterogeneity of human fetal hemoglobin. Pediatr Res 5:493-499, 1971. 5. Grifoni V, Kamuzora H, Lehman H, Charlesworth D: A new Hb variant: HbF Sardiniay7 s (El 9) Isoleucine-+Threonine found in a family with Hb G Philadelphia, p-chain deficiency and a Lepore-like haemoglobin indistinguishable from Hb A,. Acta Haematol53:347-355, 1975. 6. Ricco G, Mazza U, Turi RM, Pich PH, Camaschella C, Saglio G, Bernini LF: Significance of a new type of human fetal hemoglobin carrying a replacement isoleucine threonine at position 75 (El 9) of t h e y chain. Hum Genet 32:305-313, 1976. 7. Huisman THJ, Schroeder WA, Reese A, Wilson JB, Lam H, Shelton JR, Shelton JB, Baker S: The T r chain of human fetal hemoglobin at birth and in several abnormal hematologic conditions, Pediatr Res 11:1102-1105, 1977. 8. Lie-Injo LE: The fetal hemoglobins. Proc. 15th Internatl Congress of Pediatrics, India, 1977, pp. 137-140. 9. Lie-Injo LE: Alpha-chain thalassemia and hydrops fetalis in Malaya: Report of five cases. Blood 20:581-590, 1962. 10. Lie-Injo LE: Simple method for the isolation of hemoglobin. J Chromatogr 117:53-58, 1976. 11. Nute PE, Pataryas HA, Stamatoyannopoulos G: The Gy and Ay hemoglobin chains during human fetal development. Am J Hum Genet 25:271-276, 1973. 12. Lehmann H, Hunstman RG: “Man’s Haemoglobins.” Philadelphia: JB Lippincott, 1974. 13. Baily JL: “Techniques in Protein Chemistry.” New York: Elsevier, 1967. 14. Huisman THJ, Schroeder WA, Bannister WH, Grech JL: Evidence for four nonallelic structural genes for t h e y chain of human fetal hemoglobin. Biochem Genet 7:131-139, 1972. 15. Schroeder WA, Bannister WH, Grech JL, Brown AK, Wrightstone RN, Huisman THJ: Non-synchronized suppression of postnatal activity in non-allelic genes which synthesize the Gy chain in human foetal haemoglobin. Nature (New Biol) 244: 89-90, 1973. 16. Old J, Clegg JB, Weatherall DJ, Ottolenghi S, Comi P, Giglioni B, Michell J, Tolstoshev P, Williamson R: A direct estimate of the number of human y globin genes. Cell 8: 13-18, 1976. 17. Huisman THJ, Harris H, Gravely M, Schroeder WA, Shelton JR, Shelton JB, Evans L: The chemical heterogenity of the fetal hemoglobin in normal newborn infants and in adults. Mol Cell Biochem 17:45-55, 1977. 18. Huisman THJ, Schroeder WA, Dozy AM, Shelton JR, Shelton JB, Boyd EN, Apell G: Evidence for multiple structural genes for the gamma-chain of human fetal hemoglobin in hereditary persistence of fetal hemoglobin. Ann NY Acad Sci 165:320-331, 1969. 19. Sukumaran PK, Huisman THJ, Schroeder WA, McCurdy PR, Freehafer JT, Bouver N, Shelton JB, Apell G: A homozygote for the HBGy type of foetal haemoglobin in India: A study of two Indian and four Negro families. Br J Haemat 23:403-417, 1972. 20. Huisman THJ, Schroeder WA, Stamatoyannopoulos G, Bouver N, Shelton JR, Shelton JB, Apell G: Nature of fetal hemoglobin in the Greek type of hereditary persistence of fetal hemoglobin with and without concurrent p-thalassemia. J Clin invest 49: 1035-1040, 1970. 21. Taylor JM, Dozy A, Kan YW, Varmus HE, Lie-Injo LE Ganesan J, Todd D: Genetic lesion in homozygous or thalassemia (hydrops fetalis). Nature (London) 251 :392-393, 1974. 22. Ottolenghi S, Lanyon WG, Paul J, Williamson R, Weatherall DJ, Clegg JB, Pritchard J, Pootrakul S, Wong HB: The severe form of or thalassemia is caused by a haemoglobin gene deletion. Nature (London) 25 1:398-392,1974. 23. Deisseroth A, Nienhuis A, Turner P, Velez R, Anderson WF, Lawrence J, Creagan R, Kucherlapati R: Localization of the human or-globin structural gene to chromosome 16 in somatic cell hybrids by molecular hybridization assay. Cell 12:205-218, 1977. 24. Deisseroth A, Nienhuis A, Ruddle J, Turner P: Chromosomal localization of the human p globin gene to human chromosome 11. Blood (Suppl) 50:105, 1977. (Abstracts of the Twentieth Annual Meeting of the American Society of Hematology. San Diego, December 3-6, 1977.) 25. Huisman THJ, Schroeder WA, Felice A, Powars D, Ringelhann B: An anomaly in the y chain heterogeneity of the newborn. Nature (London) 265:63, 1977.
Heterogeneity of Hemoglobin Gamma Chains in Normal Newborns and in Cases of Alpha and Beta Thalassemia Luan Eng Lie-lnjo, 2.1. Randhawa, J.P. Kane, J. Ganesan, and R. George G. W. Hooper Foundation and international Center for Medical Research (ICMR), University of California School of Medicine, San Francisco (L. E. L.-I., Z. 1. R.), Department of Medicine and the Cardiovascular Research Institute, University of California School of Medicine, San Francisco (J.P. K.), Institute for Medical Research, Kuala Lumpur, Malaysia (J. G.), and Pediatric Unit, General Hospital, Kuala Lumpur, Malaysia (R.G.)
The ratio of Gy t o Ay chains was studied in 1 3 normal healthy newborns and in eight neonates with hydrops fetalis due t o homozygous a-thalassemia. The findings in the normal healthy newborns agreed with those of earlier reports. In homozygous a-thalassemia the Gy and Ay ratio appeared generally lower than in healthy control newborns, but o n e of the hydrops fetalis cases had a very high Gy value. In all 1 3 normal healthy newborns, in 8 patients with homozygous athalassernia, and in 4 patients with homozygous 0-thalassemia, T y chains with threonine a t position 75 were detected in addition t o Iy chains with isoleucine at position 7 5 . In homozygous a-thalassemia, the Ty-to-Iy ratio seemed lower and in homozygous 0-thalassemia higher than in normal newborns. Key words: heterogeneity of gamma chains, newborn, thalassemia
INTRODUCTION
Schroeder e t a1 [ l ] reported two types o f gamma chains in fetal hemoglobin depending on whether glycine or alanine occupied position 136 of the gamma chain. The synthesis of these chains, designated Gy and Ay, is controlled b y nonallelic loci [ 1-31 . Although b o t h G y and Ay chains show identical electrophoretic and chromatographic behavior, b y studying the C-terminal cyanogen bromide peptides (Gy CB-3 and Ay CB-3) they can be differentiated and quantitated.
Received for publication July 15, 1978; accepted November 25, 1978.
Address reprint requests to Dr. L.E. Lie-Injo, G.W. Hooper Foundation, University of California, San Francisco, CA 94143. 0361-8609/79/0601-0017$02.00@ 1979 Alan R. Liss, Inc.
18
Lie-Injo et a1
to In newborns from many different parts of the world, the ratio of synthesized Ay chains is approximately 3: 1 [ 1-31 . By the time the baby is 6 months old, the ratio has become 2 Gy t o 3 Ay, the same ratio as seen in adults [4]. The problem of heterogeneity of fetal hemoglobin became more complicated when it was shown that on the gamma chains, the amino acid occupying position 75 can differ [5-61; it can be either isoleucine (Iy) or threonine (Ty). The reported findings of Ricco et a1 [6] suggest that the synthesis of the Ty chain is controlled by a gene nonallelic with those coding for Gy and Ay chains. The presence of Ty chains in normal newborns and different disease conditions was confirmed by Huisman et a1 [ 7 ] and in the present study [ 8 ] . The possible specific function of the different gamma chains and the factors influencing control of their synthesis are still obscure. Study of the relative proportions synthesized under differing abnormal conditions might lead t o a better understanding of their relative significance. We report the heterogeneity of gamma chains of fetal hemoglobin in homozygous a thalassemia [9] and in /3 thalassemia, as well as in normal healthy newborns. METHODS
Cord blood samples from 13 normal healthy newborns were obtained from the delivery room of Moffitt Hospital, San Francisco, and placed in acid-citrate dextrose (ACD) solution. Blood from eight neonates (seven Chinese and one Malay) with hydrops fetalis due to homozygous CY thalassemia and from patients with homozygous /3 thalassemia (one Chinese and three Malay) were obtained from the General Hospital, Kuala Lumpur, Malaysia. The diagnoses of these conditions were supported by hematologic and biochemical studies in the relatives, but no globin chain synthesis studies could be carried out. The washed, packed red blood cells were flown in dry ice to San Francisco. Hemolysates were prepared by the standard method using deionized distilled water and toluene. Hb Bart's from the hemolysates of the babies with hydrops fetalis was isolated and purified according to the inverted-bottle method of Lie-Injo [ 101 . Globin was prepared by acid-acetone precipitation. The study of cyanogen bromide (CB) digests was carried out according to Schroeder et a1 [ 11 , as modified by Nute et a1 [l 11 ; however, four columns of Sephadex G-50 (fine) in series, each column being 1.6 X 160 cm, were used. For each study, the cyanogen bromide digest of 80-150 mg globin was placed on the column. The CB fragments were hydrolyzed under vacuum in 6 N HCl for 48 hours at 108"C, and the hydrolyzed CB fragments were analyzed with a two-column technique on a Beckman Model 121 M amino acid analyzer. For the study of residue 75 of the gamma chain, trypsin digests of the globin of fetal hemoglobin or purified gamma chains were fingerprinted and the tryptic peptide 9 was characterized by amino acid analysis [12]. When Ty chains were present, peptide 9b was observed in addition t o the usual peptide 9a (Fig. 2). The peptide 9b was identified by amino acid analysis with a Beckman 121 M automatic amino acid analyzer, wliichis capable of determining amino acids in the nanomole range. The relative proportion of Iy and Ty was evaluated by staining the fingerprint with cadmium acetate, eluting peptides 9a and 9b, and estimating their relative optical density at 500 nrn [13]. (This staining method was employed at the advice of Dr. L.F. Bernini.)
Hemoglobin Gamma Chains
19
RESULTS
Because Ay CB-3 peptide has three residues of alanine and none of glycine, and Gy CB-3 peptide has only one glycyl and two alanyl residues, the proportion of Gy chains in a mixture containing Gy and Ay chains can be readily calculated from the number of gIycine residues, while the proportion of Ay chains is obtained by quantitating any alanine residues in excess of two. We obtained quite satisfactory separation of cyanogen bromide peptides because the relevant y CB-3 is entirely free of other fragments. The chromatogram of the CB peptides obtained from globin for one of the normal newborns is shown in Figure la. All chromatograms of the CB peptides prepared from the HB Bart's globin (which consists solely of y chains) of the babies with hydrops fetalis resembled one another. One is shown in Figure Ib. All chromatograms showed a complete separation of the relevant y CB-3 fragment. The different peaks, identified by amino acid analysis of the hydrolysates, are indicated in the figures. The amino acid composition of the first two peaks did not agree with the expected composition of any of the CB fragments of fetal y chains but they most resembled y CB-2.
a
mi e f f l u e n t
E 0 00 N
+ m
b Fig. 1. Chromatogram of cyanogen bromide peptides of globin prepared from hemolysate. Four 1.6 X 160 cm columns of Sephadex G-50 (fine) in series were developed with 1% formic acid. a) From a healthy newborn; b) from a case of homozygous CY thalassemia.
Lie-Injo et a1
20
The relative proportion of 0.75 glycyl and 2.27 alanyl residues obtained for our normal controls (Table I) agrees well with the finding of Schroeder and co-workers [2] who found the proportion of glycyl and alanyl residues in the peptides prepared from more than 100 cord blood samples obtained from different areas of the world to be 0.71 and 2.32 respectively. Our results also agree with those of Nute and colleagues [ l l ] , who found the ratio 0.74 glycine to 2.30 alanine in nine normal newborns. In our cases of hydrops fetalis, the estimation of the residues in the y CB-3 peptide was carried out in duplicate, usually on two or three independently isolated y CB-3 peptides. Results are listed in Table I. The ratio of Gy to Ay chains in the cases of hydrops fetalis appears t o be generally lower than the mean found in normal newborns except for Case 5. If Case 5 was not included, statistical analysis using a standard t test showed a highly significant (P < 0.001) difference between the group of hydrops fetalis and the group of normal newborns. However, if Case 5 was included, the difference was not significant (P > 0.05). The ratio of Gy t o Ay chains found in the isolated Hb Bart’s does not differ significantly from the ratio obtained from the globin of whole hemolysate. In all subjects, the sum of the glycyl and analyl residues per y CB-3 peptide does not deviate more than 0.09 from the expected sum of 3.00. The relative proportions of Ty and Iy chains in normal newborns, and in cases of homozygous LY thalassemia and /3 thalassemia, are listed in Table 11. Figure 2a,b shows the fingerprint of trypsin digests of Hb F from a normal newborn and a patient with p thalassemia and Figure 3a,b shows fingerprints of pure y chains from cases with homozygous LY thalassemia. In all our normal newborns studied, the Ty chain was present. It was noticed that in a few cases with a low level of Ty chains, the 9b peptide was not demonstrable or barely evident at the first examination, but appeared in a second study if we increased the amount of tryptic digest applied for fingerprinting. In one case with Ty chain levels below 15%, the peptide 9b was examined by amino acid analysis t o make sure that the peptide represented peptide 9 of the Ty chain. Of those samples in which peptide 9b TABLE 1. Heterogeneity of Fetal Hemoglobin: y136 (cyanogen bromide digest) Normal healthy newborns No.
Glycine
Alanine
1 2 3 4 5 6 7 8 9 10 11 12 13
0.80 0.79 0.73 0.74 0.75 0.72 0.76 0.77 0.72 0.78 0.69 0.72 0.77
2.21 2.20 2.35 2.26 2.26 2.29 2.23 2.25 2.30 2.24 2.33 2.37 2.23
Mean
0.75
2.27
G, = 75%
Ay = 27%
Homozygous o! thalassemia (Hydrops fetalis)
No.
Mean
Glycine
Alanine
0.65 0.66 0.64 0.70 0.92 0.70 0.64 0.63
2.39 2.35 2.45 2.28 2.16 2.30 2.30 2.33
0.71
2.32
C, = 69%
ky = 32%
21
Hemoglobin Gamma Chains
TABLE 11. Heterogeneity of Fetal Hemoglobin: 7” (tryptic digest) Normal healthy newborns
Homozygous a thalassemia (hydrops fetalis)
Homozygous p thalassemia
‘7%
11 12 13
90 82 85 80 93 78 79 94 80 88 80 85 77
10 18 15 20 7 22 21 6 20 12 20 15 23
Mean
83.9
16.1
1 2 3 4 5 6 7 8 9 10
95 88 79 79 83 84 92 94 Mean
81
5
1
12 21 21
2 3 4
72 15 I0 I7
28 25 30 23
13.5
26.5
17 16 8 6
Mean
13
Fig. 2. Fingerprints of tryptic digest of Hb F (a + y) chains. a) From a normal healthy newborn; b) from a case of homozygous p thalassemia.
spot was conspicuous, many, but not all, were analyzed for the 9b peptide amino acid composition. In cases of homozygous a thalassemia, the level of Ty chains is generally lower than in normal healthy newborns; in the small number of cases of homozygous 0 thalassemia examined in this study, the level appeared t o be higher. DISCUSSION
The synthesis of Gy and Ay chains is under the control of nonallelic loci [ 1-31 . IR attempts to better understand the specific function and possible purpose for production of different types of y chains in the fetus, as well as the regulation of the synthesis of the different types of y chains, research workers have studied abnormal conditions in which fetal hemoglobin is produced. The ratio of Cy t o Ay chains has been studied in physiologic
22
Lie-Injo et a1
Fig. 3. Tryptic digest fingerprints of y chains from two cases of hydrops fetalis homozygous a! thalassemia. a) With a very low amount of Ty chains. (Amino acid analysis of peptide 9b showed that, in comparison to peptide 9a, it had an additional threonine residue and no isoleucine.) b) With a higher amount of Ty chains.
conditions and abnormal gentic as well as acquired conditions associated with the presence of fetal hemoglobin. During postnatal development of the newborn, the ratio of Gy t o Ay was observed to change from about 3: 1 at birth to about 2:3 toward the end of the first year [4], which led to the conclusion that the mechanism of y-gene deactivation involves an unequal repression of the Gy and Ay structural genes. It was further observed that an abnormal fetal hemoglobin variant with either glycine or alanine at position 136 differs in the level present in the blood [ 141. This led Huisman et a1 [14] t o postulate the existence of four y loci - two Gy loci, one producing more than the other; and two Ay loci, one producing more than the other - which contribute t o the total ratio of 4:2:2: 1 at birth. This hypothesis was strengthened by the observation that two Gy types of Hb F variant decline at different rates after birth in the heterozygote [15]. During the initial months after birth the activity of the major Gy gene declines, allowing the Ay chain t o dominate to the ratio of 2 Gy:3 Ay in the older child or adults. Direct estimate of the number of y-globin genes indicates that there are at least two y-chain genes per haploid human genome, but that the presence of four y-chain genes is unlikely [16]. In 1977 Huisman et a1 [17], in an attempt t o come t o a compromise, suggested that there might be three y-chain genes, two Gy-chain genes, one major and one minor, and one Ay-chain gene (Gy, gy, Ay). Up to the time of birth the ratio of product or the activity of these genes is assumed to be G.g.A, contributing to the total in a ratio 4:2:3. The postnatal decline from 0.7 t o 0.4 residues of glycine is explained by a nonsynchronized suppression of the postnatal activity of the major Gy gene while the minor Gy and the Ay genes remain active. In different types of 0 thalassemia, in hereditary persistence of fetal hemoglobin (HPFH), in patients with various hemoglobinopathies, and in acquired conditions associated with the presence of increased amounts of fetal hemoglobin, the Gy-to-Ay ratios vary greatly. Attempts t o correlate the ratio with certain conditions and thus shed more light on why different types of fetal hemoglobins are produced have not been very successful. In certain cases of hereditary persistence of fetal hemoglobin thought t o be caused by a deletion of the 0-and 6-globin genes, the fetal hemoglobin produced can be of both types, C q and Ay, or can be Gy only or Ay only [18-201. If either the Gy or Ay gene were deleted along with the 80 gene complex, only Gy or Ay would be produced. If the SO gene
Hemoglobin Gamma Chains
23
complex were deleted, but the y genes continued functioning, Gy as well as Ay chains would be produced. The ratios of Gy to Ay in patients with a thalassemia ill which the QI gene is deleted [21,22] have not been examined before, although the ratio in Hb Bart’s (74) in a black newborn carrier has been reported [ 11 . In our present study, eight cases of homozygous a thalassemia were examined. In one case (Case 5),the y chain appeared t o be almost entirely of the Gy type, while in the others the ratio of Gy t o Ay was generally lower than in our normal healthy newborn controls. Further study is needed to ascertain the significance of these findings. However, the findings cannot be explained, as is done for the HPFH, by assuming that the Gy or Ay chain gene is or is not deleted along with the a-chain gene, because the a-chain gene is located on chromosome 16 [23] and y and p genes are located on chromosome 11 [24] . In our cases, the production of the different types of y chain must have been under the control of genes other than the structural ones. Extensive recent studies of Huisman et a1 [17,25] add t o the complexity of the problem. They found in a black study population that fetal hemoglobin in newborns and adults showed a bimodal distribution of glycine values, that different patterns of decline of e - t o - A y ratio existed postnatally, and that, rather than a distinct “adult glycine value,” there was a variable ratio in adults. The existence in some individuals of Ty chains with threonine at postion 75 in addition to Iy chains with isoleucine [5, 61 poses another intriguing problem. Ricco et al [6] found +40% of normal newborns and premature infants t o have Hb F with Ty chains in addition to Hb F with Iy chains. Hb F with Ty chains was also found in almost all their patients with homozygous /3 thalassemia, in a 14-week-old fetus, and in one of three patients suffering from aplastic anemia with raised levels of Hb F. Their findings suggest that the synthesis of Ty chains is controlled by a gene nonallelic t o the genes coding for Ay and Gy chains. The presence of Ty chains in healthy normal newborns and several disease conditions has been confirmed by findings of Huisman et a1 [7] and by the findings presented in this paper. However, there are discrepancies between the findings of Ricco et al [6], those of Huisman et a1 [7] and our findings. Only a proportion of the newborns and patients were found to have Ty chains by Ricco et al [6] and by Huisman et a1 [7]. Our study showed the presence of Ty chains in all 13 normal, healthy newborns, 8 cases of homozygous a thalassemia (7 Chinese and 1 Malay) and 4 cases of homozygous /3 thalassemia (3 Chinese, and 1 Malay), if enough tryptic digest was applied t o the fingerprint. However, the number of newborns we examined was small and study of more samples may turn up negative cases. Nevertheless, the difference in findings are quite obvious and we have no explanation for this discrepancy. Further, while Ricco et a1 [6] found levels of lower than 10% in the positive cases, most of the positive cases examined by Huisman et a1 [7] in a larger group had levels around 20%. This last group also found that different ethnic groups showed widely different percentages of positive cases. Most of our newborns had Ty chain levels around 20%, a level which agrees with those of Huisman et a1 171, but we also found cases with levels below lo%, similar t o those in the group of Ricco et a1 [ 6 ] . Again, the reason for these discrepancies is not clear. Huisman et a1 [7] suggested that those who do not show the presence of Ty chains in laboratory tests do not carry the gene for Ty chains. Those who do show the presence of Ty chains in laboratory tests do carry Ty chain gene. If so, it is difficult to explain the findings of Ricco et a1 [ 6 ] , which showed that only 38.9% of their normal newborn babies had Ty chains (Ty gene) while almost all (90.6%) of their cases of homozygous /3 thalassemia had Ty chains (Ty gene). It may be that all or almost all individuals d o carry the gene for Ty chains, but that the presence of the Ty chains in the blood is not always detectable by laboratory methods and that the
24
Lie-Injo et a1
expression of this gene is increased in certain types of homozygous fl thalassemia but not in other types. This would explain the findings of Ricco et a1 [6] that only 38.9% of their normal newborns had the Ty chains while in almost all their cases of homozygous p thalassemia the Ty chains were easily detectable. Most of the cases of 0 thalassemia studied so far by all groups had Ty chains, and the levels of the Ty chains tend to be higher than those in normal healthy newborns. In our cases of homozygous p thalassemia the values appear generally higher, and in homozygous a! thalassemia they appear lower, than in normal healthy newborns. The difference between the mean level of Ty chains in the homozygous thalassemia group and in the homozygous a! thalassemia group is highly significant (P < 0.01). Interestingly, Huisman et a1 [7] did not find the Ty chains in seven black 0 thalassemia homozygotes, while they did find it in homozyogous 0thalassemia of Italian-Greek origin. In all the cases of homozygous fl thalassemia listed in the papers of Ricco et a1 [ 6 ] , of Huisman et a1 [7] , and in our own group, the type of homozygous thalassemia was not determined. In contrast to cr thalassemia, there are many variants of homozygous p thalassemia known, even in the same ethnic groups. The different findings in different ethnic groups may be due to the different types of homozygous 0 thalassemia, as was suggested by Huisman et a1 [7] . From the data obtained so far, it can still not be concluded whether the Ty chain is the product of an allele of the Gy or A-y chain structural genes or of an additional nonallelic gene. Further, the available data do not lend themselves t o an easy interpretation of why different types and levels of y chains are produced, nor d o they explain the mode of control of their synthesis, but it is clear that many factors other than the structural genes, still unknown to us, play a role in the control of y chain synthesis. ACKNOWLEDGMENTS
We would like to thank the Department of Microbiology, University of California (Dr. J.W. Goodman), San Francisco, for the use of its amino acid analyzer Beckman Model 1 1 8 in the early part of our study; Ms Ksenia Tomitch of the Cardiovascular Research Institute, San Francisco, for performing the amino acid analysis on the Beckman No. 121 M amino acid analyzer, and Mr. Alex Herrera for valuable technical help; the staff of the delivery room of the Moffitt Hospital, San Francisco, for collecting cord blood samples; and the staff of the delivery room, General Hospital, Kuala Lumpur, Malaysia, for referring cases of hydrops fetalis to us. This work was supported by research grants AM 25157, HL 10486, UC ICMR A1 10051 and HL 14237 from the National Institutes of Health, Public Health Service, and by the Committees on Research of the University of California and the School of Medicine, University of California, San Francisco. J.P. Kane was an Established Investigator of the American Heart Association during these studies. REFERENCES 1. Schroeder WA, Huisman THJ, Shelton JR, Shelton JB, Klqihauer EF, Dozy AM, Robberson B: Evidence for multiple structural genes for the 7 chain of human fetal hemoglobin. Proc Natl Acad Sci 60537-544, 1968. 2 . Schroeder WA, Shelton JR, Shelton JB, Ape11 G, Huisman THJ, Bouver NG: World-wide occurrence of nonallelic genes for the 7-chain of human foetal haemoglobin in newborns. Nature (New Biol) 240:213-274, 1972.
Hemoglobin Gamma Chains
25
3. Schroeder WA, Huisman THJ: Investigations of molecular variation in human fetal hemoglobin in the infant and in certain hematological conditions in the adult. In “Protides of the Biological Fluids, Proceedings of the 17th Colloquium, Bruges, 1969.” Oxford: Pergamon, 1970, pp 249-255. 4. Schroeder WA, Huisman THJ, Brown AK, Uy R, Bouver NG, Lerch PO, Shelton JR, Shelton JB, Apell G: Postnatal changes in the chemical heterogeneity of human fetal hemoglobin. Pediatr Res 5:493-499, 1971. 5. Grifoni V, Kamuzora H, Lehman H, Charlesworth D: A new Hb variant: HbF Sardiniay7 s (El 9) Isoleucine-+Threonine found in a family with Hb G Philadelphia, p-chain deficiency and a Lepore-like haemoglobin indistinguishable from Hb A,. Acta Haematol53:347-355, 1975. 6. Ricco G, Mazza U, Turi RM, Pich PH, Camaschella C, Saglio G, Bernini LF: Significance of a new type of human fetal hemoglobin carrying a replacement isoleucine threonine at position 75 (El 9) of t h e y chain. Hum Genet 32:305-313, 1976. 7. Huisman THJ, Schroeder WA, Reese A, Wilson JB, Lam H, Shelton JR, Shelton JB, Baker S: The T r chain of human fetal hemoglobin at birth and in several abnormal hematologic conditions, Pediatr Res 11:1102-1105, 1977. 8. Lie-Injo LE: The fetal hemoglobins. Proc. 15th Internatl Congress of Pediatrics, India, 1977, pp. 137-140. 9. Lie-Injo LE: Alpha-chain thalassemia and hydrops fetalis in Malaya: Report of five cases. Blood 20:581-590, 1962. 10. Lie-Injo LE: Simple method for the isolation of hemoglobin. J Chromatogr 117:53-58, 1976. 11. Nute PE, Pataryas HA, Stamatoyannopoulos G: The Gy and Ay hemoglobin chains during human fetal development. Am J Hum Genet 25:271-276, 1973. 12. Lehmann H, Hunstman RG: “Man’s Haemoglobins.” Philadelphia: JB Lippincott, 1974. 13. Baily JL: “Techniques in Protein Chemistry.” New York: Elsevier, 1967. 14. Huisman THJ, Schroeder WA, Bannister WH, Grech JL: Evidence for four nonallelic structural genes for t h e y chain of human fetal hemoglobin. Biochem Genet 7:131-139, 1972. 15. Schroeder WA, Bannister WH, Grech JL, Brown AK, Wrightstone RN, Huisman THJ: Non-synchronized suppression of postnatal activity in non-allelic genes which synthesize the Gy chain in human foetal haemoglobin. Nature (New Biol) 244: 89-90, 1973. 16. Old J, Clegg JB, Weatherall DJ, Ottolenghi S, Comi P, Giglioni B, Michell J, Tolstoshev P, Williamson R: A direct estimate of the number of human y globin genes. Cell 8: 13-18, 1976. 17. Huisman THJ, Harris H, Gravely M, Schroeder WA, Shelton JR, Shelton JB, Evans L: The chemical heterogenity of the fetal hemoglobin in normal newborn infants and in adults. Mol Cell Biochem 17:45-55, 1977. 18. Huisman THJ, Schroeder WA, Dozy AM, Shelton JR, Shelton JB, Boyd EN, Apell G: Evidence for multiple structural genes for the gamma-chain of human fetal hemoglobin in hereditary persistence of fetal hemoglobin. Ann NY Acad Sci 165:320-331, 1969. 19. Sukumaran PK, Huisman THJ, Schroeder WA, McCurdy PR, Freehafer JT, Bouver N, Shelton JB, Apell G: A homozygote for the HBGy type of foetal haemoglobin in India: A study of two Indian and four Negro families. Br J Haemat 23:403-417, 1972. 20. Huisman THJ, Schroeder WA, Stamatoyannopoulos G, Bouver N, Shelton JR, Shelton JB, Apell G: Nature of fetal hemoglobin in the Greek type of hereditary persistence of fetal hemoglobin with and without concurrent p-thalassemia. J Clin invest 49: 1035-1040, 1970. 21. Taylor JM, Dozy A, Kan YW, Varmus HE, Lie-Injo LE Ganesan J, Todd D: Genetic lesion in homozygous or thalassemia (hydrops fetalis). Nature (London) 251 :392-393, 1974. 22. Ottolenghi S, Lanyon WG, Paul J, Williamson R, Weatherall DJ, Clegg JB, Pritchard J, Pootrakul S, Wong HB: The severe form of or thalassemia is caused by a haemoglobin gene deletion. Nature (London) 25 1:398-392,1974. 23. Deisseroth A, Nienhuis A, Turner P, Velez R, Anderson WF, Lawrence J, Creagan R, Kucherlapati R: Localization of the human or-globin structural gene to chromosome 16 in somatic cell hybrids by molecular hybridization assay. Cell 12:205-218, 1977. 24. Deisseroth A, Nienhuis A, Ruddle J, Turner P: Chromosomal localization of the human p globin gene to human chromosome 11. Blood (Suppl) 50:105, 1977. (Abstracts of the Twentieth Annual Meeting of the American Society of Hematology. San Diego, December 3-6, 1977.) 25. Huisman THJ, Schroeder WA, Felice A, Powars D, Ringelhann B: An anomaly in the y chain heterogeneity of the newborn. Nature (London) 265:63, 1977.
