Complete nucleotide and derived amino acid sequences of sex-limited protein (Slp), nonfunctional isotype of the fourth component of mouse complement (C4)

The nucleotide sequence coding for sex-limited protein (Slp), the testosterone-regulated isotype of the fourth component of mouse complement (C4), has been determined from cloned genomic DNA and cDNA fragments. The complete deduced amino acid sequence for the single chain precursor protein of Slp (pro-Slp) consists of 1716 residues. The mature beta, alpha, and gamma subunits contain 654, 763, and 291 amino acids, respectively. One potential carbohydrate attachment site is predicted from the beta-chain, five for the alpha-chain, and none for the gamma-chain. From the comparison with the mouse C4 sequences, an extensive overall sequence homology, 96.0% in nucleotides and 94.2% in amino acids, is observed. Only one deletion/insertion event is recognized between C4 and Slp sequences: three residues near the Cls cleavage site are deleted from Slp. The distribution of cysteine residues is completely conserved between pro-Slp and pro-C4.     

Biosynthesis and post-synthetic modification of a precursor (pro-C5) of the fifth component of mouse complement (C5).

Mouse peritoneal macrophages synthesized and secreted a precursor (pro-C5) of the fifth component of serum complement (C5) in short-term tissue culture. Approximately 0.2% of the newly synthesized intracellular protein and 0.8% of secreted protein were precipitable with antiserum to mouse C5. The precursor is similar in size to the native serum protein (210,000 daltons), but consists of a single polypeptide chain. In contrast, serum C5 consists of two polypeptide chains (m.w. 125,000 and 83,000) linked by disulfide bridges. An electrophoretic variant of pro-C5 distinct from intracellular C5 was detected in medium from macrophage cultures.     

Isolation and sequence analysis of a cDNA clone encoding the fifth complement component

We have used available protein sequence data for the anaphylatoxin (C5a) portion of the fifth component of human complement (residues 19-25) to synthesize a mixed-sequence oligonucleotide probe. The labeled oligonucleotide was then used to screen a human liver cDNA library, and a single candidate cDNA clone of 1.85 kilobase pairs was isolated. Hybridization of the mixed-sequence probe to the complementary strand of the plasmid insert and sequence analysis by the dideoxy method predicted the expected protein sequence of C5a (positions 1-12), amino-terminal to the anticipated priming site. The sequence obtained further predicted an arginine-rich sequence (RPRR) immediately upstream of the N-terminal threonine of C5a, indicating that the promolecule form of C5 is synthesized with a beta alpha-chain orientation as previously shown for pro-C3 and pro-C4. The C5 cDNA clone was sheared randomly by sonication, subcloned into M13 mp8, and sequenced at random by the dideoxy technique, thereby generating a contiguous sequence of 1703 base pairs. This clone contained coding sequence for the C-terminal 262 amino acid residues of the beta-chain, the entire C5a fragment, and the N-terminal 98 residues of the alpha'-chain. The 3' end of the clone had a polyadenylated tail preceded by a polyadenylation recognition site, a 3'-untranslated region, and base pairs homologous to the human Alu concensus sequence. Comparison of the derived partial human C5 protein sequence with that previously determined for murine C3 and human alpha 2-macroglobulin has indicated regions of pronounced sequence similarity. Examination of cytoplasmic RNA prepared from human liver and the human hepatoma cell line Hep G2 by Northern transfer has indicated a C5 mRNA species of about 5.2 kilobase pairs.     

Functional studies on the secreted form of human C4 (C4s), two incompletely processed two-subunit C4 molecules (beta - alpha + gamma and beta + alpha - gamma), and pro-C4

The functional properties of the secreted form of C4 (C4s), which has a Mr approximately 5000 greater than the predominant C4 molecule found in plasma (C4p), two incompletely processed two-chain C4 molecules (beta - alpha + gamma and beta + alpha - gamma), and the extracellular C4 precursor (designated pro-C4(E)] were evaluated. All four molecules are secreted in parallel by a human hepatoma-derived cell line (Hep G2). Secretion of hemolytically active C4 is linear up to approximately 12 hr, peaks at 24 hr, and then progressively decreases over the next 48 hr. This loss of C4s functional activity parallels the proteolytic conversion of C4s to C4bs. To compare the hemolytic efficiencies of C4s and C4p, a solid-phase competitive radioimmunoassay was developed to permit measurement of the small quantities of C4 antigen in these cultures. The hemolytic efficiencies of C4s and C4p were similar. These results indicate that extracellular processing of C4s to C4p does not modulate the hemolytic activity of the molecule. Consistent with their ability to bind methylamine, both the alpha s-chain and the alpha - gamma subunit undergo denaturation-induced autolysis. The extracellular and intracellular pro-C4 molecules are also sensitive to autolytic cleavage. Interestingly, the beta - alpha subunit is resistant to autolysis. In experiments in which C4s and C4p were cleaved by C1-s to C4bs, C4(beta - alpha + gamma), C4(beta + alpha - gamma), and pro-C4(E) were resistant to C1-s cleavage and thus hemolytically inactive relative to C4s. These data indicate that processing of C4 to a three-chain structure is required to provide the proper conformation for efficient activation by C1.     

Complete nucleotide and derived amino acid sequences of the fourth component of mouse complement (C4). Evolutionary aspects

The nucleotide sequence coding for the fourth component of mouse complement (C4) has been determined from a cloned genomic DNA fragment and a cloned cDNA fragment. The amino acid sequence of the protein was deduced. The single chain precursor protein (pro-C4) consists of 1719 amino acid residues. The mature beta, alpha, and gamma subunits contain 654, 766, and 291 amino acids, respectively. One potential carbohydrate attachment site is predicted for the beta chain, three for the alpha chain, and none for the gamma chain. From a comparison with human C4 cDNA sequence an extensive overall sequence homology, 79% in nucleotides and 76% in amino acids, is observed. There is conservation in both the position and number of cysteine residues in human and mouse C4. We compared the mouse C4 amino acid sequences with those of mouse C3 and human alpha 2-macroglobulin and the evolutionary relationship among these three proteins is discussed.     

Purification and characterization of subcomponent C1q of the first component of bovine complement

Bovine C1q, a subcomponent of the first component of complement, was purified in high yield by a combination of euglobulin precipitation, and ion-exchange and molecularsieve chromatography on CM-cellulose and Ultrogel AcA 34. Approx. 12-16mg can be isolated from 1 litre of serum, representing a yield of 13-18%. The molecular weight of undissociated subcomponent C1q, as determined by equilibrium sedimentation, is 430000. On sodium dodecyl sulphate/polyacrylamide gels under non-reducing conditions, subcomponent C1q was shown to consist of two subunits of mol.wts. 69000 and 62000 in a molar ratio of 2:1. On reduction, the 69000-mol.wt. subunit gave chains of mol.wts. 30000 and 25000 in equimolar ratio, and the 62000-mol.wt. subunit decreased to 25000. The amino acid composition, with a high value for glycine, and the presence of hydroxyproline and hydroxylysine, suggests that there is a region of collagen-like sequence in the molecule. This is supported by the loss of haemolytic activity and the degradation of the polypeptide chains of subcomponent C1q when digested by collagenase. All of these molecular characteristics support the structure of six subunits, each containing three different polypeptide chains, with globular heads connected by collagen triple helices as proposed by Reid & Porter (1976) (Biochem. J.155, 19-23) for human subcomponent C1q. Subcomponent C1q contains approx. 9% carbohydrate; analysis of the degree of substitution of the hydroxylysine residues revealed that 91% are modified by the addition of the disaccharide unit Gal-Glc. Bovine subcomponent C1q generates full C1 haemolytic activity when assayed with human subcomponents C1r and C1s.     

Proton-magnetic-resonance studies on the interaction of rabbit skeletal-muscle troponin I with troponin C and actin.

1. The p.m.r. spectra of the larger CNBr-cleavage peptides of troponin I from rabbit fast-twitch skeletal muscle corresponded largely to those of fairly flexible solution structures. 2. On addition of troponin C to each of the CNBr-cleavage peptides in turn, perturbations of side chains were noted only for peptides CN5 (residues 1-21) and CN4 (residues 96-116). 3. In the presence of Ca2+, troponin C induced perturbations of the side chains of threonine-11, alanine, isoleucine and arginine residues of peptide CN5. 4. In the presence of Ca2+, troponin C induced perturbations of the side chains of phenylalanine, lysine and leucine residues of peptide CN4. 5. Irrespective of the presence or absence of Ca2+, specific interaction with actin was observed only with peptide CN4. In this case the side chains of arginine residues were perturbed. 6. It is concluded that actin interacts with the C-terminal region of peptide CN4, whereas troponin C interacts with the N-terminal region of peptide CN4 and with peptide CN5.     

Comparative study of the asparagine-linked sugar chains of natural human interferon-beta 1 and recombinant human interferon-beta 1 produced by three different mammalian cells

The asparagine-linked sugar chains of natural interferon-beta 1 secreted from human foreskin fibroblasts by poly I:poly C induction and of three recombinant human interferon-beta 1 produced by Chinese hamster ovary cells, mouse epithelial cells (C127), and human lung adenocarcinoma cells (PC8) were released quantitatively as oligosaccharides by hydrazinolysis followed by N-acetylation. After being reduced with either NaB3H4 or NaB2H4, their structures were comparatively analyzed. More than 80% of the sugar chains of natural interferon-beta 1 occur as biantennary complex-type sugar chains, approximately 10% of which contain N-acetyllactosamine repeating structure in their outer chain moieties. The remainders are 2,4- and 2,6-branched triantennary complex-type sugar chains. The sugar chains of the recombinant interferon-beta 1 derived from Chinese hamster ovary cells were very similar to those of its natural counterpart. In contrast, two other recombinant proteins contain quite different sugar chains. The protein derived from C127 cells contains complex-type sugar chains with the Gal alpha 1----3Gal beta 1----4GlcNAc group in their outer chain moieties. Their sialic acid residues occur solely as the Sia alpha 2----6Gal group, where Sia is sialic acid. In contrast, the sialic acid residues of other interferon-beta 1 occur as the Sia alpha 2----3Gal group only. A part of the sugar chains of the protein derived from PC8 cells contains bisecting N-acetylglucosamine residue in addition to the Gal alpha 1----3Gal beta 1----4GlcNAc group.     

High affinity HERG K(+) channel blockade by the antiarrhythmic agent dronedarone: resistance to mutations of the S6 residues Y652 and F656

Pharmacological inhibition of human-ether-a-go-go-related gene (HERG) K(+) channels by structurally and therapeutically diverse drugs is associated with the 'acquired' form of long QT syndrome and with potentially lethal cardiac arrhythmias. Two aromatic amino-acid residues (Y652 and F656) on the inner (S6) helices are considered to be key constituents of a high affinity drug binding site within the HERG channel pore cavity. Using wild-type (WT) and mutant HERG channels expressed in mammalian cell lines, we have investigated HERG channel current (I(HERG)) blockade at 37+/-1 degrees C by dronedarone (DRONED), a non-iodinated analogue of the Class III antiarrhythmic agent amiodarone (AMIOD). Under our conditions WT I(HERG) tails, measured at -40 mV following activating pulses to +30 mV, were blocked with IC(50) values of approximately 59 and 70 nM for DRONED and AMIOD, respectively. I(HERG) inhibition by DRONED was contingent upon channel gating, with block developing rapidly on membrane depolarization, but with no preference for activated over inactivated channels. High external [K(+)] (94 mM) reduced the potency of I(HERG) inhibition by both DRONED and AMIOD. Strikingly, mutagenesis to alanine of the S6 residue F656 (F656A) failed to eliminate blockade by both DRONED and AMIOD, whilst Y652A had comparatively little effect on DRONED but some effect on AMIOD. These findings demonstrate that high affinity drug blockade of I(HERG) can occur without a strong dependence on the Y652 and F656 aromatic amino-acid residues.     

Sequence studies of the Fd section of the heavy chain of rabbit immunoglobulin G

A partial amino acid sequence was given by Cebra, Steiner & Porter (1968b) of the N-terminal half of the heavy chain of rabbit immunoglobulin G. This was extended and in part corrected to give a continuous sequence of 136 residues, which together with other work accounts for three-quarters of the total sequence. Evidence is given suggesting that there is a limited region of 10–15 residues that are exceptionally variable in the heavy chains from pooled rabbit immunoglobulin G.     

Treatment of rhamnogalacturonan I with lithium in ethylenediamine

Rhamnogalacturonan I is a pectic polysaccharide that is solubilized from the walls of suspension-cultured sycamore cells (Acer pseudoplatanus) by the action of a highly purified endo-1,4-α-polygalacturonanase. Rhamnogalacturonan I has a linear backbone consisting of the diglycosyl repeating unit, →4)-α-d-GalpA-(1→2)-α-l-Rhap-(1→. Approximately half of the α-l-rhamnosyl residues of the backbone are branched at O-4. Selective cleavage at the galactosyluronic acid residues of the backbone by treatment of rhamnogalacturonan I wit lithium in ethylenediamine resulted in the release of the neutral glycosyl-residue sidechains that had been attached to the backbone. Various analytical techniques, including combined liquid chromatography-mass spectrometry, combined gas-liquid chromatography-mass spectrometry, and ¹H-nuclear magnetic resonance spectroscopy, were used to determine the structure of the side chains. The majority of the sidechains were isolated as oligoglycosylalditols, with rhamnitol at the “reducing” end. Terminal 2-, 4-, or 6-linked galactosyl residues were found attached to O-4 of the rhamnitol residues The 2-, 4-, and 6-linked galactosyl residues had terminal or 2-linked arabinosyl, or additional galactosyl, residues attached to them. Based on the results of fast-atom-bombardment mass spectrometry, the side chains were found to range in size from one to fourteen glycosyl residues. The side-chain structures suggest that there are four or more distinct families of side chains attached to the backbone of rhamnogalacturonan I.     

Structural characterization of the N-terminal oligomerization domain of the bacterial chromatin-structuring protein, H-NS

The H-NS protein plays a key role in condensing DNA and modulating gene expression in bacterial nucleoids. The mechanism by which this is achieved is dependent, at least in part, on the oligomerization of the protein. H-NS consists of two distinct domains; the N-terminal domain responsible for protein oligomerization, and the C-terminal DNA binding domain, which are separated by a flexible linker region. We present a multidimensional NMR study of the amino-terminal 64 residues of H-NS (denoted H-NS1-64) from Salmonella typhimurium, which constitute the oligomerization domain. This domain exists as a homotrimer, which is predicted to be self-associated through a coiled-coil configuration. NMR spectra show an equivalent magnetic environment for each monomer indicating that the polypeptide chains are arranged in parallel with complete 3-fold symmetry. Despite the limited resonance dispersion, an almost complete backbone assignment for 1H(N), 1H(alpha), 15N, 13CO and 13C(alpha) NMR resonances was obtained using a suite of triple resonance experiments applied to uniformly 15N-, 13C/15N- and 2H/13C/15N-labelled H-NS1-64 samples. The secondary structure of H-NS1-64 has been identified on the basis of the analysis of 1H(alpha), 13C(alpha), 13Cbeta and 13CO chemical shifts, NH/solvent exchange rates, intra-chain H(N)-H(N) and medium-range nuclear Overhauser enhancements (NOEs). Within the context of the homotrimer, each H-NS1-64 protomer consists of three alpha-helices spanning residues 2-8, 12-20 and 22-53, respectively. A topological model is presented for the symmetric H-NS1-64 trimer based upon the combined analysis of the helical elements and the pattern of backbone amide group 15N nuclear relaxation rates within the context of axially asymmetric diffusion tensor. In this model, the longest of the three helices (helix 3, residues 22-53) forms a coiled-coil interface with the other chains in the homotrimer. The two shorter N-terminal helices fold back onto the outer surface of the coiled-coil core and potentially act to stabilise this configuration.     

Comparative studies on the extractability of collagen from aortas of stroke-prone spontaneously hypertensive and normotensive rats.

The molecular states of collagen in the aortas of age-matched stroke-prone spontaneously hypertensive (SHRSP) and normotensive Wistar Kyoto rats (WKY) were studied by analyzing its extractability under defined conditions. The monomeric and oligomeric collagen extractable with 0.5 M acetic acid/6 M urea from aortic homogenates of 9-month-old SHRSP and WKY comprised approx. 0.6 and 2.0%, respectively, of the total collagen. On incubation of the acetic acid/urea-extracted residues with pepsin at 4 degrees C, the levels of the collagen alpha 1(I) and alpha 2(I) chains solubilized from the SHRSP residues were both less than 50% of those from the WKY residues. When the residues were incubated with pepsin at 15 or 25 degrees C, the differences became smaller. When the acetic acid/urea residues were hydrolyzed with cyanogen bromide, nearly identical peptide maps were obtained for SHRSP and WKY. The aortas from 2-month-old SHRSP and WKY contained much larger proportions of acid/urea-extractable collagen than those of the older rats (8.2 and 13% of the respective total collagen). The levels of the alpha 1(I) and alpha 2(I) chains solubilizable from the respective residues by pepsin at 4 degrees C were similar to each other. These results indicate that aortic collagen fibrils in SHRSP are stiffened more prominently than those in WKY.     

Purification and characterization of subcomponent C1q of the first component of mouse complement.

1. Mouse C1q, a subcomponent of the first component of complement, has been purified in a highly haemolytically active form by a combination of precipitation with EGTA, ion-exchange chromatography and gel filtration. Yields ranged from 3 to 5 mg/200 ml of serum, and the activity of final preparations was in the range of 2 X 10(13)-4 X 10(13) C1q effective molecules/mg. 2. The molecular weight of mouse C1q was 439 500 +/- 1586, as determined by polyacrylamide-gel electrophoresis in the presence of sodium dodecyl sulphate. 3. Mouse C1q was shown to be composed of non-covalently linked subunits, all being in the molecular-weight range 45 000-46 000, and three covalently linked chains each having a molecular weight of approx. 23 000 as determined on polyacrylamide-gel electrophoresis in the presence of sodium dodecyl sulphate by using non-covalently and covalently linked subunits of human C1q as markers with known molecular weights calculated theoretically previously [Porter & Reid (1978) Nature (London) 275, 699-704]. 4. Mouse C1q contained hydroxyproline, hydroxylysine, a high percentage of glycine and approx. 9% carbohydrate. The absorption coefficient and nitrogen content of C1q were also determined.     

Amino acid sequence of the N-terminal 139 residues of light chain derived from a homogeneous rabbit antibody

The amino acid sequence of the N-terminal 139 residues of the L (light) chain derived from a homogeneous rabbit antibody (designated BS-1) to type III pneumococci was determined. A combination of methods involving tryptic cleavage restricted to the 2 arginine residues of the molecule and mild acid hydrolysis of a labile peptide bond between the V (variable) and C (constant) regions of the L chain (Fraser et al., 1972) allowed the isolation of two large peptides comprising the entire V region (residues 1-109); these peptides were suitable for automated Edman degradation. The complete sequence analysis of the V region was carried out with only 4mumol of L chain. This material was homogeneous, although minor variant sequences, if present at the 10% value, would not have been detected. The L chain contains 3 intrachain disulphide bridges, whose pairing was established by diagonal electrophoresis: there is one V-region bridge between positions 23 and 88 and one C-region bridge between positions 134 and 194; the third one connects V and C domains between positions 80 and 171. When compared with the basic sequence of human kappa chains, rabbit L chain BS-1 appears to be more similar to the V(KI) prototype sequence than to V(KII) or V(KIII) sequences, where V(KI), V(KII) and V(KIII) represent subgroups I, II and III respectively of V regions of kappa light chains. The V regions of rabbit heavy and light chains are homologous to each other. The presence of two clusters of 3 glycine residues in positions 94-96 and 99-101 respectively is remarkable. Residues 94-96 may be related to antibody complementarity whereas residues 99-101 function probably as a pivot permitting the combining region of the L chain to make optimal contact with the antigenic determinant (Wu & Kabat, 1970).     

pH dependence of the enzymatic processing of collagen I by MMP-1 (fibroblast collagenase), MMP-2 (gelatinase A), and MMP-14 ectodomain

The proteolytic processing of collagen I by three matrix metalloproteinases (MMPs), a collagenase (MMP-1), a gelatinase (MMP-2), and the ectodomain of a membrane-type metalloproteinase (MMP-14), has been investigated at 37 °C between pH 6.0 and 9.2, a pH range reflecting conditions found in different body compartments under various physiopathological processes. In the proteolytic degradation the native collagen triple helix must be partially unwound to allow the binding of α chains to the protease’s active-site cleft. We have found that MMP-1 interacts with the two types of collagen I α chains in a similar fashion, whereas both MMP-2 and MMP-14 bind the two α chains in a different way. The overall enzymatic activity is higher on the α-2 chain for both MMP-1 and MMP-2, whereas the MMP-14 ectodomain preferentially cleaves the α-1 chain. In MMP-2 a marked difference for substrate affinity (higher for the α-1 chain) is overwhelmed by an even more marked propensity to cleave the α-2 chain. As a whole, the three classes of MMPs investigated appear to process collagen I in a significantly different fashion, so various MMPs play different roles in the collagen homeostasis in various compartments (such as bloodstream, synovial fluid, normal and tumoral tissues), where different pH values are observed.     

Identification of the binding site for activated protein C on the light chain of factors V and VIII

Activated protein C has been observed to bind to the light chains of factor Va and factor VIII. Fragments of the factor VIII light chain were produced by recombinant DNA techniques and expressed in Escherichia coli. Three fragments of the light chain were studied; L4 (residues 1974-2332), L3.2 (residues 1560-1829 and 2046-2332), and L3.3 (residues 1560-2052). Two fragments, L4 and L3.3, which overlapped sequences between residues 1974-2052, inhibited the anticoagulant activity of activated protein C. Comparison of the sequences of factors V and VIII in this region revealed that residues 2005-2018 in the factor VIII sequence were homologous with residues 1861-1874 in the factor V sequence. The peptides Arg-Ala-Gly-Met-Gln-Thr-Phe-Leu-Ile (RAGMQTPFLI; residues 1865-1874) from the factor V sequence and His-Ala-Gly-Met-Ser-Thr-Leu-Phe-Ile-Val (HAGMSTLFIV; residues 2009-2018) from the factor VIII sequence were synthesized. Both peptides were observed to inhibit the anticoagulant activity of activated protein C and its inactivation of factors Va and VIII. Furthermore RAGMQTPFLI quenched the fluorescence of the dansyl-Glu-Gly-Arg-modified protease. Polyclonal antibodies against RAGMQTPFLI bound to factor Va and inhibited the anticoagulant activity of activated protein C and the inactivation of factor Va. These results indicate that a portion of the binding sites for activated protein C on the light chains of factors V and VIII are contained in the sequences RAGMQTPFLI or HAGMSTLFIV, respectively.     

Structure of Plant Cell Walls : XII. Identification of Seven Differently Linked Glycosyl Residues Attached to O-4 of the 2,4-Linked l-Rhamnosyl Residues of Rhamnogalacturonan I

Seven differently linked glycosyl residues have been found to be glycosidically linked to O-4 of the branched 2,4-linked l-rhamnosyl residues contained in the rhamnosyl and galacturonosyl backbone of the cell wall pectic polysaccharide rhamnogalacturonan I. These seven glycosyl residues are, therefore, the first residues of at least seven different side chains attached to the rhamnogalacturonan backbone. These first side chain glycosyl residues are 5-linked l-arabinofuranosyl and terminal 3-, 4-, 6-, 2,6-, and 3,6-linked d-galactopyranosyl residues. The existence of at least seven different side chains in rhamnogalacturonan I indicates that rhamnogalacturonan I is either an exceedingly complex polysaccharide or that rhamnogalacturonan I is a family of polysaccharides with similar or identical rhamnogalacturonan backbones substituted with different side chains.     

Amino acid sequence of the monomer subunit of the extracellular hemoglobin of Lumbricus terrestris

The giant extracellular hemoglobin (3,800 kDa) of the oligochaete Lumbricus terrestris consists of four subunits: a monomer (chain I), two subunits each of about 35 kDa (chains V and VI), and a disulfide-bonded trimer (50 kDa) of chains II, III, and IV. The complete amino acid sequence of chain I was determined: it consists of 142 amino acid residues and has a molecular weight of 16,750 including a heme group. Fifty-nine residues (42%) were found to be identical with those in the corresponding positions in Lumbricus chain II (Garlick, R. L., and Riggs, A. F. (1982) J. Biol. Chem. 257, 9005-9015); 45 (32%), 56 (40%), 44 (31%), and 45 (32%) residues were found to be in identical positions in the sequences of chains I, IIA, IIB, and IIC, respectively, of Tylorrhynchus heterochaetus hemoglobin (Suzuki, T., and Gotoh, T. (1986) J. Biol. Chem. 261, 9257-9267). When the sequences of all six annelid chains are compared, 18 invariant residues are found in the first 104 residues of the molecule; very little homology exists among the annelid chains in the carboxyl-terminal 38-residue region. Nine of the 18 invariant residues are also found in the human beta-globin chain.