Further Studies of the Frequency and Significance of the Ty-Chain of Human Fetal

W. A. SCHROEDER, T. H. J. HUISMAN, G. D. EFREMOV, J. ROGER SHELTON, JOAN B. SHELTON, ROBERT PHILLIPS, ALICE REESE, MARSHA GRAVELY, J. MICHAEL HARRISON, and HENRY LAM, Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125, Laboratory of Chemistry and the Comprehensive Sickle Cell Center, Department of Cell and Molecular Biology, Medical College of Georgia, Augusta, Georgia 30901, and Veterans Administration Hospital, Augusta, Georgia 30904

ABSTRACT A further study of the Ty_chain in a position (the Ayy-chain). The recent discovery of the variety of conditions has revealed its presence in the Ty-chain by Ricco et al. (3) has added yet another facet cord bloods of ethnic groups previously unstudied. to the many ways in which the study of Hb-F has pro- Heterozygous newborn average 17-19% Ty_chain vided data about the genetics of this fetal protein. In the while the mean value in four presumed homozygotes Ty-chain, the isoleucyl residue at position 75 of the was 31%. y-chain is replaced by a threonyl residue. Although The Ty-chain is readily detectable in f3-thalassemia Ricco et al. (3) particularly examined the frequency of of various ethnic groups (although infrequent in Blacks) the Ty-chain in f8-thalassemia, they also detected its as well as in 8f3-thalassemia. Studies of a few families presence in cord bloods. Our studies of the T y-chain have provided an opportunity to determine whether or (4) substantiated their data in f8-thalassemia of in- not certain individuals are heterozygous or homo- dividuals of Mediterranean origin, but detected no zygous for the Ty_gene. The Ty_chain has not been T-y-chains in the ,8-thalassemia of Blacks. Moreover, our detected in the human that is syn- data from cord blood suggested that the frequency of thesized in increased amounts in persons with the the Tyy-chain might be related to ethnic source or hereditary persistence of fetal hemoglobin. geographical locality, and quantitatively was present in Although the Ty_chain is detectable in sickle cell greater percentage than reported by Ricco et al. Be- anemia, its frequency appears to be lower than in cause of the limited data, conclusions could only be normal individuals. tentative, and statistical evaluation had limited mean- By focussing upon the relationship of the percentage ing. Consequently, we have extended these studies in of Ty-chain to the sources of human fetal globulin from several areas and examined other previously unstudied determinants in cis and in trans, the conclusion has abnormalities. been reached that the Tyy-chain is the product of a mutant Ay-locus which should be named the TAy_chain. METHODS Many samples of Hb-F were available from previous in- INTRODUCTION vestigations of the Gy_ to Ay-ratio in Hb-F (1, 2). In most of these, not only y- and a-chains, but considerable percentages The y-chains of human fetal hemoglobin (Hb-F)l are of (3-chains were also present. For most samples, the Gy- to produced by nonallelic (1, 2). Evidence for this Ay-ratio was known as well as the percentages of Hb-F and is derived from the observation that Hb-F contains two Hb-A2; procedures for these determinations have been one with in position described in detail (1, 2). types of y-chains; namely, glycine The methods of tryptic hydrolysis, column chromatographic 136 (the Gy_chain) and a second with alanine in this isolation of the yT-9 peptides (residues 67-76, inclusive), their amino-acid analysis, and the calculation of results have Receivedfor publication 7 August 1978 and in revisedform been described in detail (4) and were used without major 25 September 1978. modification. To assess the reproducibility of results within 1Abbreviations used in this paper: AS, sickle cell trait; and between laboratories in Pasadena, Calif. and Augusta, Hb, hemoglobin; Hb-F, human fetal Hb; HPFH, hereditary Ga., a series of samples was repeatedly analyzed (Table I). persistence of fetal Hb; SC, Hb-S,Hb-C disease; SS, sickle Major discrepancy of data occurred only in the results from cell anemia. Cases M.D., M.H., and O.J. The original data in Augusta sug-

268 J. Clin. Invest. © The American Society for Clinical Investigation, Inc., 0021-973817910210268/08 $1.00 Volume 63 February 1979 268-275 TABLE I TABLE II Results of Replicate Determinations within and The Ty-Chain in Newborn Babies of Different Populations between Laboratories Total No. of Percentage of Ty-chain Population Condition No. positive Designation Pasadena Augusta Georgia Blacks AA; AS; AC; AR; Fx* 81 21 Georgia Blacks SS; SCt 13 2 F.N. 22, 20 Los Angeles Blacks AA; AS5 10 2 L.H. 11, 16 Ghana Blacks AA; AS; AC" 21 1 P.W. 16, 18 Kenya Blacks AA 17 3 C.K. 42, 42 L.P. 16, 22 Total 142 29 India 6,233 18, 22 P.B. 23, 22*, 214 25, 21*, 224 Los Angeles Whites AA 12 5 Chinese 1 18, 17*, 154 13, 14, 17*, 134 Holland Whites AA 11 4 B.V.V. 19, 18, 16 16 Malta Whites AA; FM heterozygotesl 17 3 N.W. 31, 31 32 Others ** 12 4 M.D. 16, 12 22 M.H. 16, 13 28 Total 52 16 O.J. 17 33 * 10 babies with AS, AC, or A-Richmond; 5 babies with F-Port In these experiments, the peptides TyT-9 and 'yT-9 were Royal (Fx); AA, normal. isolated in Pasadena, Calif., but portions were analyzed both t Two babies with SC; both negative. in Pasadena, Calif. and Augusta, Ga. § One baby with AS; negative. * Data from Pasadena, Calif. Three babies with AS or AC; all negative. 4 Data from Ga. Seven babies with Hb F-Malta I; one positive. Augusta, ** From India (five babies), and one each of Chinese, Jap- anese, Eskimo, Egyptian, Turkish, American Indian, and gested that technical problems existed in addition to the use of Australian aborigine origin. the one-column system (4). The remaining data reveal that duplicate determinations may be expected to agree within +3%. This degree of concordance between duplicate deter- The 31% T y-chain corresponds rather well with the minations is somewhat unexpected, because neither of the average value of 29% Ay_chain in the newborn babies yT-9 peptides is pure as isolated, and the interpretation of the (5). It had previously been suggested that one sample amino-acid analyses is subjective to a degree. with 35% Ty-chain (this is N.W. of Table II [4] which now averages 33% in three determinations) repre- RESULTS sented a Ty-homozygote. The detection of three others The T-chain in cord bloods. Table II summarizes in this sample range strengthens this conclusion. One the results from the examination for Ty-chains in cord bloods from a variety of ethnic and geographic sources. 35r MEAN: These data include the information previously reported ° 31% (4). Initially, Hb-F from 98 newborn was studied; to 30 these data may now be added the results from 96 samples. Two other ethnic groups, Kenya Blacks and 25 I Asiatic Indians, have been examined in these addi- 0 tional samples; positive samples were detected in both. 201 8 19%; Of the original 98 samples, 70 were negative. Con- o- sequently, the frequency of the Tyy- in this group 5 was 0.14. There are 149 negative results in the present 194 so that the T y-frequency (0.12) is about the same. It 101 is noteworthy that Blacks as well as Caucasians from Q =AA * =AS orA-Richmond several geographic areas had positive samples in fre- 5 quencies that probably were not greatly different from - * =SS each other. The distribution of positive data in terms of the per- Black Non-Bock centage of Ty-chain is presented in Fig. 1. Two groups FIGURE 1 Distribution ofthe percentage ofTy-chains in posi- were observed: One group of four averaged 31% Ty- tive cord bloods. Data from two definitely positive samples chain, and the remaining 39 averaged 18% (17% for from Blacks are omitted because of technical problems in the 23 Black babies and 19% for 16 Caucasian newbom). quantitation. AA, normal.

Frequency and Significance of the Ty_Chain 269 of these putative homozygotes is from Kenya whereas TABLE IV the remaining three are from a Georgia community. The TyChain in SS A family study was possible for one assumed homo- zygous newborn who had 31% Tyy-chain. Both parents Case Age Ty Gly in yCB-3 had Hb-F levels of <2% and were positive for the Ty-chain. D.C. Adult 29*, 23* 0.55 The Trchain in sickle cell anemia. The Hb-F of 73 cases of sickle cell anemia (11 Black newborn, 51 Black M.P. Adult 27*, 36* 0.77 adults, 10 Veddoid Indians, and 1 Israeli citizen) has L.P. Adult 28 0.42 been examined; 7 from Blacks and 0 from Veddoid T.A.t1 28 mo 22 0.50 Indians were positive for the Ty-chain (Tables II and 31 mo 18 III). All information on the seven positive cases is 34 mo 25 detailed in Table IV; initial detection was done in cord Tw.A.t 11 mo 24 0.63 blood samples of F.N. and C.N. and several samples 14 mo 22 0.61 from the former (and from two additional sickle cell 19 mo 22 anemia [SS] infants) could be examined. Although the F.N. Cord 19 0.75 data are somewhat erratic, no definite trend of in- 6 wk 16 0.71 creasing percentage of Tychain from birth to adulthood 14 mo 29 was observed. 21 mo 22, 20§ 0.56 The data of Table III indicate that positive results were obtained in one of five adults with sickle cell trait C.N. Cord 19 0.75 and in one of seven adults with Hb-S,Hb-C (SC) * Data from two independently isolated samples of Hb-F disease. Assuming that one SS patient is homozygous for each individual. for the Ty-chain (M.P., to be discussed later) and all t Siblings. others (SS, sickle cell trait [AS], and SC) hetero- 5 Duplicate determinations on same sample. zygous for the Ty-chain, the frequency of the Ty-gene in sickle-cell disease approximates 0.06, which suggests that the occurrence of the Ty-gene in combination with number of samples. The data from a newborn infant the p5-gene is less than in combination with the p_ with the Gy-HPFH heterozygosity (N. 0. C. [6]) sug- gene. gest that the Ty-chain can indeed be present in trans to The Ty.chain in hereditary persistence offetal hemo- the HPFH determinant (Table VI). At birth, 17% Ty globin 8 chain and -14% A y-chain were detected. Both values (HPFH). homozygotes and 34 heterozygotes decreased with time, of various 6 which suggests that cessation of types, including individuals with Hb- Ay-chain synthesis by a locus in trans to the Gy-HPFH Kenya who have the Gy-HPFH phenotype, all were determinant parallels the decrease of the Tyy-chain negative for the Ty-chain (Table V). These results do not exclude the presence of a Ty-gene in trans to the HPFH production. determinant in the heterozygote, but do suggest that it TABLE V is absent in cis within the limits of this relatively small The Ty-Chain in Adults with HPFH

TABLE III Total No. of The Ty Chain in Persons* with Abnormal Population Condition No. positives

Total No. of Black GyAy-heterozygote* 19 0 Population Condition No. positivesI Black GyAy-homozygote 7 0 India GyA-y-heterozygote 1 0 Black SS 51 5 India Gy-heterozygote t 2 0 Black AS 4 1 India Gy-homozygote 1 0 Black AC 1 0 Black Gy-heterozygote§ 3 0 Black CC 1 0 Caucasian Ay-heterozygotell 3 0 Black SC 7 1 Black Gy-heterozygotel with India AS 1 0 Hb-Kenya 6 0 India SS 10 0 Israel SS 1 0 * Includes all types; two subjects were also heterozygous for Hb-S. * 11 months and older. I Was also heterozygous for Hb-E. I The Ty-chain values of the five positive SS patients are § See (6). listed in Table IV; percent Ty in one AS person was 41%, and Ii See (7, 8). 17% in one SC individual. ¶ See (9); two subjects were also heterozygous for Hb-S.

270 Schroeder et al. TABLE VI Black individuals with heterozygous or homozygous Postnatal Change of the Gy,Ay, and Ty-Chains in a ,8-thalassemia were positive for the Ty_chain. GY-HPFH Heterozygote Fig. 2 shows the wide distribution from 19 to 70% Ty-chain in the Hb-F of nine non-Black individuals with Residues in position 136 ,8-thalassemia trait who had Hb-F of the order of 2%; although one had only 0.3% and two had 5.5 and 10%. Age of individual Gly Ala* Ty-chains The meaning of this wide range is not apparent. 16 8,8-thalassemia heterozygotes and 7 8,3-thalas- semia homozygotes have been studied. However, the Birth 0.86 0.14 17 is not 56 d 0.90 0.13 15 large number of positive cases ( 11 heterozygotes) 112 d 0.95 0.08 8 indicative for a high frequency of the Ty_chain in this condition because 7 belong to the same family. More- * The residues of alanine in yCB-3 minus 2 because alanine over, none of the seven homozygotes had the Ty_chain. is always present at positions 138 and 140. The percent Ty_chain in the 11 8,3-thalassemia hetero- zygotes averaged m61% with a range of 51-71% (Fig. 2). The T-chain in f3thalassemia and 8,3thalassemia. Family studies. Data on members of four families The data of Table VII show that Hb-F has been exam- with /8-thalassemia or 8,8-thalassemia are presented in ined for the presence of the Ty_chain in many in- Figs. 3-5. dividuals with f3-thalassemia from a wide range of Family C. Hematological data about this family ethnic groups and geographical localities. Frequencies have been reported by Friedman et al. (10). Fig. 3 in individual groups appear meaningless because of presents the relevant part ofthe pedigree with quantita- limited data, but the results do suggest that the Ty_ tive data of several types. The propositus, 111-7, is chain is widespread in f3-thalassemia as shown also by homozygous for 8-thalassemia and is at least hetero- Ricco et al. (3). This is particularly true in ethnic groups zygous for the 8+-type. In addition, he has both Hb- surrounding the Mediterranean area (Italian-Greek- A2 and Hb-A'. In the mother, her sister, and two sib- Yugoslavian) where, from the data of Table VII, the lings of the propositus, Hb-A2 is in cis to the f8-thalas- frequency approaches 0.16. semia determinant, whereas the father and one sibling As previously noted (4) the Ty-chain is not very have Hb-A2 in cis to the f3-thalassemia determinant. evident in Blacks with /8-thalassemia, and only 6 of 50 The data show that Ty-chain is not produced in cis to ,B- thalassemia and Hb-A' on the maternal side. On the TABLE VII other hand, the propositus, his father (II-15), and a sib- The T-Chain in /& and 8,-Thalassemia ling with Hb-A2 (III-6) do produce Ty_chains. The per- centage of Ty-chains in the Hb-F of the propositus and Total No. of Population Condition No. positives 80r Blacks /8-Thal. Heterozygotes* 29 5t Blacks /8-Thal. Homozygotes 21 1 70[ 0 0 Non-Blacks fl-Thal. Heterozygotes§ 35 9 CD Non-Blacks fl-Thal. Homozygotes" 65 23 60p Non-Blacks 8,/-Thal. Heterozygotesi 16 11 0 c) O Non-Blacks 8/3-Thal. Homozygotes 7 0 50- o O Blacks Heterozygotes 8#f-Thal. 0 (G_ytype) 3 0 40p Yugoslavia Lepore Heterozygotes 2 0 0 Yugoslavia Lepore Homozygotes 2 0 301 Yugoslavia Lepore /3-Thal. 4 1 o 8 201 )- * Includes two persons with C-,8 Thalassemia and three with S-,p Thalassemia. 10 t Three are members of the same family. § Includes four persons with E-,8 Thalassemia. "18 from Italy or Greece (12 pos.); 13 from Yugoslavia (5 pos.); 4 4 4 4 A-,8Th. A- $Th. A-8,8Th 11 from Israel (1 pos.); 10 from Turkey and Iran (2 pos.); Block Non-Block Non-Block 5 from India (O pos.); one from Malta (pos.); one from Holland PTh- Th. ,?Th -RTh. (neg.); one Chinese (pos.); 4 from the Far East (1 pos.); one Block Non-tlock unknown (neg.). FIGuRE 2 Distribution of the percentage of Ty-chain in posi- I Many belong to one Yugoslavian family; see Fig. 5. tive samples from adults with different types of thalassemia.

Frequency and Significance of the Ty-Chain 271 FAMILY C FAMILY W FAMILY B 2 Tr 15 n~ LO O A2= 5.3% A2 = 4.9% A2 = 6.0% A2= 2.5% A2 = n.d. T F =8.8% F = 1.3% A21 = 0% A21=2.4% A21= 3.5% Gly= 0.40 Gly = 0.55 F = 1.2% F =0.6% F = 1.4% Ty= neg Ty =55% Gly = 0.64 Gly=0.41 Gly =0.37 1 2 T- = 43% Ty = neg. Ty = neg A2=4.6% A2 = 2.6% A2 = 2.5% Cord Blood F =21% F = 74% F =Tronsfused 8X- 7 ]IE6E Gly = 0.34 GI y= 0.54 Gly= 0.60 Gly = 0.73 = 70% Ty = 34% = 46% Ty =36% A2= 6.9% A2 = 2.4% A2= 2.4 % A2= 2.4% Ty Ty % A21=0 A2'=2.9% A21 = 2.5% A2 3.5% a * /3- Thalossemia without Ty-gene incis F =0.3% F =71% F =0.6% F =0.5% q ,-3 Tholossemia with Ty-gene incis Gly =043 Gly =0.56 Gly =0.54 Gly =0.57 FIGURE 4 = 55%(t) Ty = 23% Ty= neg. = neg. Partial pedigrees of two families with /3-thalas- Ty TyI semia. a'U I3- Tholassemia without Ty-gene incis E3~ /3 -Tholussemia with Ty-gene incis Hb A2I (a2 82 16 Gly _ Arg ) trait. On the paternal side, the percentage ofthe Tyy-chain has I E Hb C (a2,82 6 Glu. Lys) trait. been determined in the Hb-F of the father (II-7) and three of his siblings as well as in the grandfather and FIGuRE 3 Partial pedigree of a family with f3-thalassemia, two cousins of the proposita. The father ofthe proposita Hb-A', and Hb-C. and the other affected relatives have the Ty-gene in cis to the 8/3-thalassemia determinant and are hetero- his father could be accurately determined; however, zygous for both. The 35% T.y-chains in the Hb-F of the that in the sibling III-6 was somewhat high because proposita, in contrast to the average of 62% in the seven of technical problems. These data indicate that the 8,3-thalassemia heterozygotes, again reflects the dilu- propositus who must be heterozygous for the Ty_ tion effect from the non_T y-containing Hb-F of the (8- determinant had about one-half the percentage of thalassemia determinant. Ty-chain found in the Hb-F of his father. Assuming that The Ty-chain in miscellaneous conditions. Table his Hb-F comes about equally from both maternal and VIII lists a variety of miscellaneous conditions whose paternal ,8-thalassemia determinants, Hb-F without Ty Hb-F has been examined for the presence of T y-chains. chain from the maternal determinant has diluted the The frequency of the Ty_gene in this very hetero- Hb-F with the Ty-chain from the paternal determinant. geneous collection of conditions is 0.09 and does not In the Chinese Family W., the daughter II-1, who is differ significantly from the frequency in all cord bloods heterozygous for 13-thalassemia, inherited the Ty-deter- (vide supra). Consequently, it seems probable that the minant from her mother and not from her father (Fig. 4). The sibling II-2, who is homozygous for 8-thalas- semia, is also TABLE VIII heterozygous for the Ty-determinant. The The Ty-Chain in Persons with Some Acquired Hematological 70% Ty-chains in the Hb-F of II-1 and one-half that per- Conditions and in Normal Individuals centage in II-2 suggest that, as in Family C., there is a dilution of Ty-chains in this 83-thalassemia homozygote. Total No. of The expression of the Ty-chain in Family B. is also Condition No. positives illustrated in Fig. 4. Because homozygote 11-2 had no JCML* 7 3 Hb-A at birth, nor did she produce any before her first Erythroleukemia 3 0 transfusion at 112 d (11), ,o-thalassemia must be Leukemiat 8 2 present. The percentages of Ty_chain in the father, the Fanconi's Anemia 14 1 mother, and one older sister, who also had a thalassemia Blackfan-Diamond Anemia 3 1 homozygozity, are rather similar; thus, one might con- Aplastic Anemia 7 3 clude that there is no dilution effect in these homo- D-trisomy 9 2 zygotes who are also homozygous for the Ty_chain. Normals 8 3 The relationships in Family D., which have been Miscellaneous§ 35 2 described in other respects (12), are given in Fig. 5. The * Juvenile chronic myelogenous leukemia. proposita (III-4), who has both 8/3-thalassemia and 18- t Various forms. thalassemia, received the /8-thalassemia determinant § Undiagnosed cases of which many had normal hematology, from her mother (II-8) who is negative for the TVy-chain. but all had elevated levels of Hb-F to some degree.

272 Schroeder et al. presence of the Ty_chain in the Hb-F of these in- FAMILY D dividuals simply reflects the appropriate component in their genetic makeup and is unrelated to the specific A2 2.8 % condition. F =72% Gly =0.45

T y = 62 % DISCUSSION 2 34 4db 5 6 7 8 A2= 2.1% A =2.6% A2=2.6% A2 31% A22=2.3% A2=53% The incidence of the Ty_chain. The Ty-chain occurs F= 11% F =10% F =11% F =14% F rz7% F - 1.9% widely among the human races and various ethnic Gly =0.61 Gly=042 Gly =0.43 Gly z 0.49 Gly=0.38 Gly =0.64 groups. Although the available data are still limited, it Ty=64% Ty = n.d. TY-58% Ty= 62% Ty-67% Ty= neg appears that its incidence is the lowest in the Black ] 26 3 4d race. Moreover, no Ty_chain was found in the Hb-F of A2 =2 5% A2 = 2.5% A2 = 2.0% A2 =2.5% persons with the HPFH condition and in a few F = 11% F = 9% F z12% F =84% only Gly=0.53 Gly= 0.48 Gly'0.40 Gly = 0.61 Black patients with sickle cell anemia and homo- Ty=n.d. Ty =56% Ty=67% Ty z 35% zygous f3-thalassemia. The absence of the Ty_chain in ,B- Thasserria without Trgene in cis the Hb-F of HPFH heterozygotes does not exclude its Q a -Tholssmmo with Ty gene incis presence in trans to the HPFH determinant. The ob- servations made in the newborn Gy-HPFH hetero- FIGURE 5 Pedigree of Family D. with 88-thalassemia and zygote (Table VI) show that the Ty-chain gene may be ,B-thalassemia. present in trans, but that with the decrease in the ac- tivity of the Gy. and Ay_genes in trans, the relative pro- of two additional alanyl residues in the yCB-3 pep- duction of the Ty-chain decreases accordingly to a level tide.) Assessment of these data relies heavily upon the at which it will not be detectable any longer. The fre- accuracy and precision of the quantitative results and quency of Ty-chain in Caucasian f8-thalassemia homo- the agreement that may be expected from determina- zygotes (23 of 65 cases studied, which corresponds to a tions in two laboratories (Table I). The data of Fig. 6 frequency of 0.18) is less than that reported by Ricco et show a wide scattering. However, one group of data al. (3) (39 of 42 cases studied, corresponding to a fre- suggests, with a considerable degree of confidence, an quency of 0.46) and is not significantly different from inverse correlation between the percent Ty_chains and that of newborn (Table II). It appears desirable to the percent Gy-chain (represented by the solid line); analyze the Hb-F from newborn of different racial and these results are from all ,88-thalassemia heterozygotes, ethnic backgrounds to evaluate the world-wide occur- rence of the Ty-chain. The Ty-chain in relation to the Gy_ and Ay-chains. o /3 Tholassemia trait Because all samples of Hb-F have either Gy_ or Ay_ 801- *, Thalasserria homoz. chains or both, that is, either glycine or alanine or both o 8S Thalassemia trait at position 136, it follows that the Ty_chain must have 70 0 a Sickle Cell Anemia one or both. Ricco et al. (3) concluded from indirect v SC disease evidence that the Ty-chain "certainly has glycine in 60 0\ o position 136" and might also have alanine. In their in- 0 vestigation of Hb-F-Sardinia (which is another example 50p- 0 of the Ty-chain in a thalassemia individual), Grifoni et 0 al. (13) came to the same conclusion. However, when H1 40 p 0 0 in previous studies the fraction of Tyy-chain was I0" plotted against the fraction of y-chains that are Gy_ 30 0. * r=-092 chains, a rather ill-defined inverse correlation appeared , '> *;S* f( (n= 24) to exist (4); conversely, this would indicate a direct cor- 0 20 I'* relation with the fraction of Ay-chains in Hb-F. Such a d °v* " plot, however, is too simple and ignores completely the 10 r= -0.55 source of Hb-F and the possibility that the Ty_chain (n=23) gene is in cis or in trans (or both) of the primary genetic determinant of the condition involved. The data given 0.4 0.5 0.6 0.7 0.8 0.9 in Fig. 6 again compare the fraction of Gy-chains with Glycine in yCB-3 that of the Ty-chain. (A comparison of the percent Ty-chain with the percent Ay_chain is avoided because FIGURE 6 Relationship between the percentage of Ty-chain and the fraction of Gy-chain in the Hb-F from persons with the accuracy ofthe percent Ay_chain might be somewhat different forms of thalassemia, and with SS and SC diseases. less than that of the Gy-chain as a result of the presence See text for details.

Frequency and Significance of the Ty_Chain 273 from several (but not all) 18-thalassemia heterozygotes, from slightly <50% of the Ty-chain positive /8-thalas- semia homozygotes, and from one of the five Ty-chain 50 - AS positive SS patients. The others form a group in which relation between the percent Ty_chain and the percent 40 - Gy-chain is much less clear (represented by the broken line). F- 30 ~ Thus, the first set of data suggests an inverse correla- 20 tion between the percent Ty-chains and the fraction of Gy-chains (which is similar to a direct correlation with the fraction of Ay-chains). Individuals belonging to this group may be either homozygous for the Ty_gene and 10 O ° produce Hb-F equally in cis and in trans (examples are several homozygous B-thalassemia patients and the one SS patient), or are heterozygous for the Ty_gene 0.3 0.4 0.5 0.6 0.7 0.8 0.9 and produce Hb-F only in cis (examples are several Glycine in yCB-3 ,3-thalassemia traits and all 38-thalassemia traits). How- FIGuRE 7 Relationship between the percentage of Ty_chain ever, if an individual is heterozygous for the Ty-gene and the fraction of Gy-chain in the Hb-F from normal persons and produces Hb-F equally in cis and in trans (exam- and from individuals with acquired hematological disorders. ples are most,8-thalassemia homozygotes, the four SS patients, and perhaps a few /3-thalassemia traits) a dilu- orangutans have shown that each of these samples have tion effect is evident, and the percent Ty_chain is a Vy-chain, i.e., a chain with a valyl residue in position reduced considerably. 75, in addition to the 'y-chain which has an isoleucyl Some of the available data of Fig. 6 are difficult to residue in position 75 (14). This heterogeneity is in explain because it cannot be determined with certainty addition to the one in position 135 which is occupied into which category they belong. It is noteworthy, how- by either a threonyl residue (the Tyy-chain) or an alanyl ever, that the three ,B-thalassemia homozygotes of Fig. residue (the Ay-chain) (15). The present information 4 fall in the expected categories, namely II-2 of Family suggests that Ile-75 and Thr-135 are in one chain (the W. with a Ty-chain heterozygosity in the category of ITy-chain) and Val-75 and Ala-135 in the other chain (the decreased percent Ty_chain, and cases 11-1 and 11-2 with VAyy-chain). Thus, this type of heterogeneity is similar a Ty-chain-homozygosity in the category in which the to that of the human fetal hemoglobins, although the correlation between the percentage of Ty- and Ay-chain incidence of the human TAy-chain appears to be con- is clearly evident. siderably less than that of the VAy-chain of the orangutan. Fig. 7 gives similar data for individuals without Some genetic considerations. Is the human TAy_ hemoglobinopathies, and, although the data scatter gene that produces the TAy_chain a mutant of the considerably, an inverse correlation between the per- "normal" Ay-gene or genes, or is it a duplication? We centages of Ty- and Gy_chains appears to exist. A com- have presented evidence (16, 17) for pairs of nonallelic parison of these data with those of Fig. 6 suggests that Gy- and Ay_genes which are arranged in the order all persons are heterozygous for the Ty_chain. G:g:A:a and which produce y-chains in the ratio of Thus, the data demonstrate a distinct inverse rela- 4:2:2:1; for simplicity, they are abbreviated G, g, A, and tionship between the Ty- and the Gy_chains: as the a where upper case letters denote the more active Gyy-chain decreases in a sample, the Ty_chain increases. gene and lower case the less active gene. Hybridiza- We conclude, therefore, that the Ty-chain is an Ay-chain: tion experiments (18, 19) do not agree with these ideas; that is, has an alanyl residue in position 136 and should however, their accuracy may not distinguish between be designated TAy_chain. This conclusion is given sup- models with three or four genes per haploid chromo- port by the fact that none ofthe samples of Hb-F that has some. Although some lines of evidence point to a pair exclusively Gy_chains also has Ty_chains. It is also sup- of nonallelic Gy-genes (17, 20, 21), the evidence for two ported by the cord-blood data because the value of nonallelic Ay-genes is much less direct. The linear cor- 31% Ty-chain in the four presumed homozygotes cor- relation between the percentages of Ty- and Ay_chains relates well with the 29% A y-chain found in newborn that is illustrated in Figs. 6 and 7, and the nearly identi- babies (5). Furthermore, the data listed in Table VI cal Ty-chain and Ay_chain values in the four newborn show that in this baby, who is positive for the Ty-chain with a Ty-homozygosity, suggest that these chains are and heterozygous for the Gy-HPFH condition, the the product of a single Ay_gene which is also a Ty_gene; production of both the Ay-chain and the Ty-chain because of the relatively low general frequency of decreases simultaneously and to the same extent. Ty-genes, it seems unlikely that the Ty_mutation would Recent studies of the fetal hemoglobins of five be present both in A and a genes. All evidence, there-

274 Schroeder et al. fore, implies a single Ay_gene. Furthermore, the linear 9. Kendall, A. G., P. J. Ojwang, W. A. Schroeder, and T. H. J. correlation points to the TAy-gene as a mutant of the Huisman. 1973. Hemoglobin Kenya, the product of a y-,8 fusion gene: Studies of the family. Am. J. Hum. Genet. normal Ay-gene rather than as a duplication such that 25: 548-563. nonallelic Ay and TAy_genes are present on each haploid 10. Friedman, Sh., E. Schwartz, V. Ahern, and E. Ahern. 1974. chromosome. Globin synthesis in the Jamaican Negro with beta- thalassemia. Br. J. Haematol. 28: 505-513. 11. Huisman, T. H. J., W. A. Schroeder, A. K. Brown, C. B. ACKNOWLE DGMENTS Hyman, J. A. Ortega, and P. K. Sukumaran. 1975. Further studies of the postnatal change in chemical heterogeneity This work was supported in part by U. S. Public Health Service of human fetal hemoglobin in several abnormal condi- grants HL-02558 and HL-05168. This is Contribution 5847 tions. Pediatr. Res. 9: 1-6. from the Division of Chemistry and Chemical Engineering, 12. Efremov, G. D., N. Nikolov, H. Duma, W. A. Schroeder, California Institute of Technology, Pasadena, Calif. A. Miller, and T. H. J. Huisman. 1975. 8,8-Thalassaemia in two Yugoslavian families. Scand. J. Haematol. 14: 226-232. REFERENCES 13. Grifoni, V., H. Kamuzora, H. Lehmann, and D. Charles- 1. Schroeder, W. A., T. H. J. Huisman, J. R. Shelton, J. B. worth. 1975. A new Hb variant: Hb F Sardinia v75 (E 19) Shelton, E. F. Kleihauer, A. M. Dozy, and B. Robberson. Isoleucine-Threonine found in a family with Hb G 1968. Evidence for multiple structural genes for the y Philadelphia, 13-chain deficiency and a Lepore-like chain of human fetal hemoglobin. Proc. Natl. Acad. Sci. haemoglobin indistinguishable from Hb A2. Acta U. S. A. 60: 537-544. Haematol. (Basel). 53: 347-355. 2. Huisman, T. H. J., W. A. Schroeder, G. D. Efremov, H. 14. Schroeder, W. A., J. R. Shelton, J. B. Shelton, and T. H. J. Duma, B. Mladenovski, C. B. Hyman, E. A. Rachmilewitz, Huisman. 1978. The Vy chain of fetal hemoglobin of the N. Bouver, A. Miller, A. Brodie, J. R. Shelton, J. B. orangutan. Biochem. Genet. 16: 1203-1205. Shelton, and G. Apell. 1974. The present status of the 15. Huisman, T. H. J., W. A. Schroeder, M. E. Keeling, N. heterogeneity of fetal hemoglobin in f8-thalassemia: An Gengozian, A. Miller, A. R. Brodie, J. R. Shelton, J. B. attempt to unify some observations in thalassemia and Shelton, and G. Apell. 1973. Search for nonallelic related conditions. Ann. N. Y. Acad. Sci. 232: 107-124. structural genes for y-chains of fetal hemoglobin in some 3. Ricco, G., U. Mazza, R. M. Turi, P. G. Pich, C. Cama- primates. Biochem. Genet. 10: 309-316. schella, G. Saglio, and L. F. Bernini. 1976. Significance 16. Huisman, T. H. J., W. A. Schroeder, A. M. Dozy, J. R. of a new type of human fetal hemoglobin carrying a Shelton, J. B. Shelton, E. M. Boyd, and G. Apell. 1969. replacement isoleucine -- threonine at position 75 (E19) Evidence for multiple structural genes for the gamma- of the y chain. Hum. Genet. 32: 305-313. chain of human fetal hemoglobin in hereditary per- 4. Huisman, T. H. J., W. A. Schroeder, A. Reese, J. B. Wilson, sistence of fetal hemoglobin. Ann. N. Y. Acad. Sci. 165: H. Lam, J. R. Shelton, J. B. Shelton, and S. Baker. 1977. 320-331. The Ty chain of human fetal hemoglobin at birth and in 17. Huisman, T. H. J., W. A. Schroeder, W. H. Bannister, and several abnormal hematological conditions. Pediatr. Res. J. L. Grech. 1972. Evidence for four nonallelic structural 11: 1102-1105. genes for the y chain of human fetal hemoglobin. Bio- 5. Huisman, T. H. J., W. A. Schroeder, A. Felice, D. Powars, chem. Genet. 7: 131-139. and B. Ringelhann. 1977. An anomaly in the y chain 18. Old, J., J. B. Clegg, D. J. Weatherall, S. Ottolenghi, P. heterogeneity of the newborn. Nature (Lond.). 265: Comi, B. Giglioni, J. Mitchell, P. Tolstochev, and R. 63-65. Williamson. 1976. A direct estimate of the number of 6. Sukumaran, P. K., T. H. J. Huisman, W. A. Schroeder, P. R. human y-globin genes. Cell. 8: 13-18. McCurdy, J. T. Freehafer, N. Bouver, J. R. Shelton, J. B. 19. Tolstochev, P., R. Williamson, J. Eskdale, G. Verdier, J. Shelton, and G. Apell. 1972. A homozygote for the HbGy Godet, V. Nigon, G. Trabuchet, and M. Benabadji. 1977. type of foetal haemoglobin in India: A study of two Indian Demonstration of two a-globin genes per human haploid and four Negro families. Br. J. Haemotol. 23: 403-417. genome for normals and Hb J Mexico. Eur. J. Biochem. 7. Huisman, T. H. J., W. A. Schroeder, H. R. Adams, J. R. 78: 161-165. Shelton, J. B. Shelton, and G. Apell. 1970. A possible 20. Schroeder, W. A., W. H. Bannister, J. L. Grech, A. K. subclass of the hereditary persistence of fetal hemoglobin. Brown, R. N. Wrightstone, and T. H. J. Huisman. 1973. Blood. 36: 1-9. Non-synchronized suppression of postnatal activity in 8. Huisman, T. H. J., W. A. Schroeder, G. Stamatoyan- non-allelic genes which synthesize the Gy chain in human nopoulos, N. Bouver, J. R. Shelton, J. B. Shelton, and G. foetal haemoglobin. Nat. New Biol. 244: 89-90. Apell. 1970. Nature of fetal hemoglobin in the Greek type 21. Brennan, S. D., M. B. Smith, and R. W. Carrell. 1977. of hereditary persistence of fetal hemoglobin with and Haemoglobin F Melbourne Gy16 Gly -+ Arg and Haemo- without concurrent f8-thalassemia. J. Clin. Invest. 49: globin F Carlton Gy121 Glu -* Lys. Biochim. Biophys. 1035-1040. Acta. 490: 452-455.

Frequency and Significance of the Ty_Chain 275