Amino Acid Substitutions in Albumin Variants Found in Brazil
Total Page：16
File Type：pdf, Size：1020Kb
Proc. Natl. Acad. Sci. USA Vol. 86, pp. 1821-1825, March 1989 Biochemistry Amino acid substitutions in albumin variants found in Brazil (alloalbumins/genetic polymorphism/population markers/point mutation) KUNIO ARAItt, KAREN Husst§, JEANNE MADISONt, FRANK W. PUTNAMt¶, FRANCISCO M. SALZANO11, MARIA H. L. P. FRANCO'1, S. E. B. SANTOStt, AND MARIA J. M. FREITAStt tDepartment of Biology, Indiana University, Bloomington, IN 47405; IIDepartamento de Gendtica, Instituto de Biocidncias, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS, Brazil; ttDepartamento de Gendtica, Centro de Cidncias Biol6gicas, Universidade Federal do Pard, Beldm, PA, Brazil; and #Laboratori6 de Gendtica, Instituto de Ciencias Biol6gicas, Universidade Federal do Amazonas, Manaus, AM, Brazil Contributed by Frank W. Putnam, December 15, 1988 ABSTRACT Conventional horizontal starch-gel electro- capital of Brazil's southernmost state (30010' S, 5105' W), phoresis in four buffer systems and structural studies were while Coari (40'5 S. 6308' W) and Oriximini (1045' S, 5508' W) performed on four albumin variants, and the findings were are located, respectively, in the northern states of Amazonas compared with similar previous data. Albumins Coari I and and Para', in the Amazonian region. All of the variants were Porto Alegre I have a previously unreported amino acid from heterozygous individuals. Familial occurrence was substitution (glutamic acid replaced by lysine at position 358, established for Porto Alegre I and Oriximina' I but not for denoted 358 Glu -* Lys). The alteration in albumin Porto Coari I or Porto Alegre II. Plasma samples from the probands Alegre II (501 Glu -* Lys) is the same as that found for three were first restudied in Porto Alegre by using starch gel alloalbumins of Asiatic origin, designated Vancouver, Bir- electrophoresis in four different buffer systems as described mingham, and Adana. Albumin Oriximina I has the same by Franco and Salzano (5): (i) sodium acetate/EDTA, pH exchange as albumin Maku (541 Lys -> Glu). Some of these 5.3; (ii) Tris/EDTA/borate, pH 6.9; (iii) Tris/EDTA/borate, findings can be explained only by the occurrence of indepen- pH 6.1; and (iv) Tris/lithium succinate/citrate, pH 6.0. In dent mutations at the same site in the albumin gene. They also these tests the four variants were compared with Manaus I, point to a third cluster of mutations in that gene. indicating Mura I, and 10 other reference albumin types, including some hypermutability in some of its segments. specimens for which the structural change had earlier been determined in Bloomington. The structural studies were done in Bloomington and Systematic electrophoretic screening ofplasma for a series of involved the same general strategy, with some modification genetic variants has been extensively conducted in Brazil (1- as required in individual cases. The strategy consists of six 5). In studies including the albumin locus the Porto Alegre steps, the principles and details ofwhich have been described group conducted population surveys of a total of 14,650 (6-11). The steps are: (i) cellulose acetate (Microzone) persons. Of these, 9615 individuals surveyed at 13 locations electrophoresis at pH 8.6; (ii) purification ofthe total albumin were classified phenotypically as of mixed ancestry (White/ (normal albumin A plus variant) by HPLC, reduction and Black in Porto Alegre and in the interior ofRio Grande do Sul, carboxymethylation, and cleavage with CNBr (7, 10, 11); (iii) and White/Black/Indian elsewhere), and 2110 individuals at analytical isoelectric focusing of the CNBr digest to identify 4 locations as White. In addition, 2925 Indians of 17 different the CNBr fragment in which the substitution was localized (7, tribes were included in the surveys. Most of these studies 10); (iv) HPLC peptide mapping on a preparative scale to have been published by the Porto Alegre group, including a purify the variant CNBr fragment (10, 11); (v) HPLC peptide series of articles pertinent to the present investigation (1-5). mapping of a tryptic or Staphylococcus aureus V8 protease As a result of these studies 12 instances of albumin variants digest of the purified CNBr fragment (10, 11); and (vi) amino have been detected. Franco and Salzano (5) gave details acid analysis of the variant peptide(s) with the Beckman about the probands and used horizontal starch gel electro- model 121M amino acid analyzer and automated sequence phoresis in four buffer systems to compare 10 of these determination with the Beckman model 890C sequencer (6). alloalbumins with each other and also with 6 reference Tryptic peptides are designated T and V8 protease peptides albumin variants. Some of the variants could not be distin- are designated S, and both types are numbered consecutively guished from each other; for example, Coari I, Porto Alegre in their predicted order in the amino acid sequence (6, 11). I, and several others were indistinguishable. Likewise, Orix- imina' I from a triracial individual appeared to be the same as the Amerindian variant Maku (3). RESULTS AND DISCUSSION Because of mutual interests arising out of previous inves- Starch Gel Electrophoresis. Examples of starch gel elec- tigation of alloalbumins occurring in North and South Amer- trophoresis at pH 6.1 and pH 6.9 are given in Fig. 1. At pH ican Indians, Japanese, and Caucasians (6-11), we undertook 6.1 and 6.9 Maku, Oriximina I, and Mura I had a fast mobility a collaborative structural study offour albumin variants from (-2 net charge relative to normal albumin), and they ap- Brazil that were available for analysis. The results obtained peared identical in all four buffer systems. Coari I and Porto are presented herein. Alegre I had a slow mobility (about +2) and could not be differentiated from each other under these conditions, or at MATERIALS AND METHODS pH 5.3 or pH 6.0. In all four buffer systems Porto Alegre II, which had a slow mobility (2 +2), could readily be differ- The four variants studied structurally were Coari I, Porto entiated from Coari I and Porto Alegre I (see Fig. 1 for pH 6.1 Alegre I, Porto Alegre II, and Oriximina I, named after the and pH 6.9). However, despite some slight differences in towns where they have been discovered. Porto Alegre is the tPresent address: Josai University, Department of Chemistry, Fac- The publication costs of this article were defrayed in part by page charge ulty of Science, Keyakidai, Sakado, Saitama 350-02, Japan. payment. This article must therefore be hereby marked "advertisement" §Present address: Pitman-Moore, Terre Haute, IN 47802. in accordance with 18 U.S.C. §1734 solely to indicate this fact. ITo whom reprint requests should be addressed. 1821 Downloaded by guest on September 28, 2021 1822 Biochemistry: Arad et al. Proc. Natl. Acad. Sci. USA 86 (1989) I& pH 6X1 pH 6.9 S S 2 0.5I 3 _ 4 E .g_. ._ S. " ~~~~~~A..L 5 co 'at cm 6 I-. R _E,, U = iEAM_ 7 U 8 To U 0: 9 11 1 2 1 3 1 4 TI TI ME I MIN) 1 5 FIG. 2. HPLC elution profile on a Vydac C18 column of the ~ + + tryptic peptides of CB5 from albumin Coari I. The lyophilized tryptic FIG. 1. Comparative mobility of human albumin variants in two digest was dissolved in 0.1% trifluoroacetic acid (buffer A) and eluted Tris/EDTA/borate systems, pH 6.1 and pH 6.9. The anode is at the at a flow rate of 1 ml/min over 100 min with a linear gradient from right. From top to bottom: A (normal serum) (lane 1), Coari I (lane 0 to 50% buffer B (acetonitrile/0.1% trifluoroacetic acid). T48 and 2), Porto Alegre I (lane 3), Gainesville (Christchurch type of T49 are peptides of albumin A. T48A* denotes a variant peptide with proalbumin, serum 3433) (9) (lane 4), Pollibauer (Lille type of an amino acid substitution. proalbumin, serum 6535) (9) (lane 5), Porto Alegre II (lane 6), artefact T48A* had the same amino-terminal hexapeptide sequence as (denatured B serum 3006) (lane 7), Vancouver (11) (lane 8), Manaus I (lane 9), Cooperstown (Reading type) (10) (lane 10), Mexico (serum the normal T48, but it differed because the substitution 358 276682) (8) (lane 11), Maku (lane 12), Oriximind I (lane 13), Mura I Glu -+ Lys provided a new site for tryptic cleavage, causing (lane 14), and A (normal serum) (lane 15). All the sera with variants the loss of the carbpxyl-terminal Lys-359 (Fig. 3). This are from individuals heterozygous at the albumin locus. References substitution had not previously been reported for an alloal- cited for non-Brazilian sera give the source of the serum and the bumin, and it required confirmation because of the lack of an results of structural study of the variant. overlap into the following sequence. To confirm the structural change in Coari I a V8 protease apparent mobility, it was difficult to differentiate Porto digest was made of a second preparation of the CB5 variant Alegre II unambiguously from Vancouver and Manaus I. Fig. fragment. Only the Vydac C18 HPLC step was used for 1 shows that all ofthe variants mentioned above could readily preparation. The variant CB5 was separated from normal be distinguished at pH 6.1 and 6.9 from albumin Mexico CB5, but it was contaminated with CB3. The V8 protease (slow, + 1, aspartic acid replaced by glycine at position 550, peptides were separated by HPLC on an Ultrasphere ODS denoted 550 Asps- Gly) (8), albumin Cooperstown (fast, -1, column (Anspec). Fraction 31 (F31 in Fig. 4) contained the 313 Lys -+ Asn) (10), and the two types of proalbumins, variant peptide (designated S37*-38); this was purified by Gainesville (Christchurch type, -1 Arg -- Gln) (9) and rechromatography on a Vydac C18 column (see Fig.