- Email: [email protected]
Albumin Church Bay: 560 Lys→Glu a new mutation detected by electrospray ionisation mass spectrometry
Biochimica et Biophysica Acta 1382 Ž1998. 305–310
Albumin Church Bay: 560 Lys ™ Glu a new mutation detected by electrospray ionisation mass spectrometry Evan K.M. Chua ) , Stephen O. Brennan, Peter M. George Molecular Pathology Laboratory, Department of Pathology, Christchurch School of Medicine, P.O. Box 4345, Christchurch, New Zealand Received 18 August 1997; accepted 27 August 1997
Abstract Albumin Church Bay is a fast migrating genetic variant of human serum albumin which, in a heterozygous subject, formed about 50% of the circulating albumin. Reversed phase peptide mapping and electrospray ionisation mass spectrometry ŽESI-MS. indicated that the C-terminal CNBr peptide had decreased polarity associated with a 1 Da increase in mass. Subdigestion of this peptide with trypsin and chymotrypsin revealed that the increased mass was associated with the chymotrypsin fragment VEKCCKADDKETCF Ž555–568. which had a mass of 1791.1 compared to 1790.2 for its normal counterpart. Sequence analysis of PCR-amplified DNA indicated an A ™ G mutation at position 98 of exon 13, which causes a point mutation of 560 Lys ™ Glu and results in a 1 Da mass increase. q 1998 Elsevier Science B.V. Keywords: Serum albumin; Amino acid substitution; Mass spectrometry; Point mutation; Formylation; Albumin Church Bay
1. Introduction Human serum albumin is the most abundant plasma protein Ž35–50 grl. with a circulating half life of 19 days in humans. It functions primarily as a transport protein for ligands such as long-chain fatty acids, drugs, bilirubin and thyroid hormones. There is a single copy of the albumin gene on chromosome four and this is made up of nearly 17 000 bp organised into 15 exons and 14 introns with exon 15 untranslated. The translated albumin is a single-chain protein consisting of 585 amino acids organised into three internally homologous domains w1x. There are nine Žactually eight and one-half. double loops formed by disulphide bonds involving adjacent half-cystine
)
Corresponding author. Fax: 64-3 364 0545.
residue. Albumin is synthesised in the liver as preproalbumin with an 18-residue prepeptide and a 6 residue propeptide Arg–Gly–Val–Phe–Arg–Arg. The prepeptide is cleaved off by signal peptidase when the precursor protein enters the lumen of the endoplasmic reticulum w2x. When the resulting proalbumin reaches the Golgi vesicles, its propeptide is removed by a membrane-bound enzyme Ž proalbumin convertase. to form mature albumin with aspartic acid as the amino-terminal residue w3x. The processed product is then constitutively secreted into the circulation w4x. More than thirty genetic variants of albumin have been identified so far w5x. However, unlike haemoglobin and coagulation factors variants, no adverse clinical effect on molecular function has been attributed to albumin variants, except for a condition known as FDH Ž familial dysalbuminaemic hyperthy-
0167-4838r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 7 - 4 8 3 8 Ž 9 7 . 0 0 1 4 7 - 7
306
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
roxinaemia. where a mutation of 218 His ™ Arg results in increased T4 binding w6x. None the less, the characterisation of genetic variants of albumin has provided unique insights into the specificity of the proteases involved in its processing w7,8x and facilitated the identification of the hepatic propeptide convertase w7,8x. Additional new variants will undoubtedly provide important clues about the molecular stability and help define ligand binding sites in the molecule w9x. We report here the molecular defect of a new fast albumin variant ŽAlbumin Church Bay. , which was found during routine clinical protein electrophoresis screening. ESI-MS played a key role in the identification of the new 560 Lys ™ Glu mutation. 2. Material and methods 2.1. Protein gel electrophoresis Agarose gel electrophoresis, performed in 1% agarose using Tris–barbitone buffer ŽpH 8.6., and Ni 63 autoradiography were performed as previously described w10x. 2.2. Albumin purification EDTA anticoagulated plasma was dialysed against 16 mM sodium acetate buffer ŽpH 5.2. and applied to a DEAE Sephadex A-50 anion exchange column
Ž1.6 = 30 cm., equilibrated in the same buffer. Bound albumin was eluted using a pH gradient from pH 5.2 to 4.5 w10x. Albumin peaks were pooled separately, concentrated and lyophilised. 2.3. Limited tryptic cleaÕage Native albumin, 5 mgrml in 10 ml 10 mM Tris– HCl, pH 8.0, was incubated at 208C for 120 min with 3 mg of trypsin and aliquots were analysed directly by agarose gel electrophoresis and by SDS-PAGE w11x. 2.4. Protein fragmentation Proteolytic digestions with trypsin and S. aureus V8 proteases were carried out on S-carboxamidomethylated albumin which was reduced and derivatised in 8 M urea as described previously w12x. 85 nmoles of S-carboxamidomethyl albumin was incubated with 0.1 mg TPCK trypsin Ž2% wt.rwt.. in 500 ml 0.5 M NH 4 HCO 3 , for 3 h at 378C. S. aureus V8 protease digestion was performed under the identical conditions except the digestion was for 24 h at room temperature. For CNBr cleavage, 4.5 mg of S-carboxamidomethyl albumin was incubated with ; 400 fold excess CNBr in 150 ml 70% HCOOH, for 24 h at room temperature in the dark. After dilution with 10
Fig. 1. Agarose gel electrophoresis ŽpH 8.6.. Lane 1, Normal human plasma; lane 2, patient plasma; lane 3, purified normal albumin component; lane 4, purified Albumin Church Bay; lane 5 and 6, partial tryptic digestion of native normal albumin Žtime 0 min and time 120 min respectively.; lane 7 and 8, partial tryptic digestion of native Albumin Church Bay Žtime 0 min and time 120 min respectively..
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
volumes of distilled water, the fragments were lyophilised and analysed by RP-HPLC. 2.5. HPLC Reverse-phase peptide maps were run on total tryptic and S. aureus V8 digests using the 50 mM phosphate buffer ŽpH 2.9.racetonitrile solvent system as previously described w13x. The column, however was a Waters 4 mm C 18 Novapak reverse-phase column. Peptide maps of CNBr digests were run on a Phenomenex W-Porex 5 C4 reverse-phase column. The initial solvent was 0.05% TFA and the final solvent was 0.05% TFA in 60% acetonitrile. The flow rate was 1 mlrmin. Because the aberrant peptide co-eluted with another CNBr fragment, it was further purified by re-injection and elution with a 25 mM ammonium acetate–25 mM ammonium acetater50% acetonitrile solvent system. The pure aberrant peptide was collected and lyophilised. All CNBr peptides were analysed and identified by ESIMS Želectrospray ionisation mass spectrometry..
307
nomic DNA encoding exon 13 of the albumin gene w15x Ž residues 528–571. was amplified by PCR w16x using primers A25CŽCATGCAGATGAGAATATTGAGAC . and A26B Ž CCTAAGCCCTAGCCTAACCAAAC.. Each 50 ml amplification reaction contained 1 unit Taq DNA polymerase Ž Boehringer Mannheim., 50 mM KCl, 15 mM MgCl 2 , 200 mM each of dATP, dTTP, dCTP and dGTP, 0.2 mM of each primer and 100 ng of template DNA. Forty cycles of denaturation Ž948C, 1 min. , annealing Ž 508C, 2 min. , and extension Ž 728C, 3 min. with a final extension phase of 5 min was performed in a PETC-1 thermal cycler Ž Perkin Elmer.. The PCR product was purified using Hi-Pure PCR purification cartridges ŽBoehringer Mannheim. and
2.6. Subdigestion of CNBr 7 The normal and aberrant peptide CNBr 7 were subdigested with trypsin and chymotrypsin. For chymotryptic digests, 17 nmoles of peptide was incubated with 50 mg chymotrypsin in 25 ml 0.5 M NH 4 HCO 3 , and incubated at 378C for 3 h. Similar conditions were used for tryptic subdigestion. 2.7. Screening of the albumin digests by mass spectrometry ESI-MS was performed on 10 ml of either total digest or purified peptide at a concentration of 10– 100 pmolrml in 50% acetonitriler0.1% formic acid. The instrument, a VG Platform II Ž Micromass. was operated in the positive ion mode and samples were delivered to the source at 5 mlrmin.The mrz range, 200–1400 was scanned every 3 s and data was acquired over 2 min. Mass spectrometer control and data processing used MassLynxe software supplied with the instrument. 2.8. Polymerase chain reaction (PCR) and DNA analysis Genomic DNA was extracted from whole blood by a standard method w14x. A 381 bp fragment of ge-
Fig. 2. Reverse-phase CNBr peptide maps of albumin A Žlower. and Albumin Church Bay Župper.. CNBr fragments were identified from their masses and are numbered from the N-terminal. Note that 7A which eluted at 14 min in albumin A elutes at 15.5 min Ž7CB. in the variant. Modified forms of these peptides, designated 7Af and 7CB f are also indicated.
308
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
sequenced directly by the dideoxy chain termination method w17x using 33P-labelled ddNTP ŽAmersham. and ThermoSequenaseTM ŽAmersham..
3. Results Routine clinical electrophoresis in agarose gels detected an individual with bisalbuminaemia resulting from a genetic variant with increased negative charge ŽFig. 1, lane 2.. The variant represented some 50% of the total albumin and the increased anodal migration suggested a Lys ™ Glu substitution. Anion exchange chromatography of plasma on DEAE-Sephadex gave two major albumin peaks. Electrophoretic analysis indicated that the earlier eluting peak was normal albumin Žlane 3. and the later peak the variant Ž lane 4. . Again, peak areas indicated that the variant accounted for 50% of the total albumin. Partial tryptic digestion was used to assess which domain was abnormal. Cleavage of normal native albumin Ž lane 6. at Arg-197 w18x produces a 43 kDa C-terminal fragment which is detectable by both
agarose gel electrophoresis and SDS-PAGE Žnot shown.. The N-terminal domain is further degraded and is not detectable by electrophoresis. Partial tryptic digestion of the variant Žlane 8. showed that the increased negative charge was in the large C-terminal fragment. These results established that the abnormality was in the second or third domain, beyond residue 197. To establish the precise location of the mutation, the albumin components were carboxamidomethylated and digested with trypsin and S. aureus V8 protease. However, no abnormalities were detected by either 2-dimensional peptide mapping w19,20x or RP-HPLC mapping w13x. Further tryptic and S. aureus V8 digests were analysed by ESI-MS, but again no abnormalities were detected. However, the observed mrz signals only accounted for ; 90% of the known albumin sequence; unaccounted for regions included residues 318–337 and 558–574. Carboxamidomethylated albumin fractions were subjected to CNBr cleavage and the fragments mapped by RP-HPLC Ž Fig. 2. . 10 ml of each peak was analysed directly by ESI-MS and the measured mass was used as the basis of the peak identification
Fig. 3. ESI-MS spectra of CNBr 7 fragments, showing normal control Ž7A. and variant Ž7CB. which differ by 1 Da. Also shown are modified forms of these peptides, 7Af and 7CB f , which have an increased mass of 28 Da compared to their respective counterparts.
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
in Fig. 2. Maps of the normal and variant albumin differed in one major respect. The C-terminal fragment Ž7A. was missing and replaced by a new more hydrophobic peptide Ž7CB. . Minor modified forms of these two peptides, designated 7Af and 7CB f , were also detected eluting later in the chromatogram. Raw ESI-MS spectra of these four peptides are shown in Fig. 3. All displayed major wM q 3Hx and wM q 4Hx ions, and all peptides were homogeneous except for 7CB which was contaminated with CNBr 4. This was removed by rechromatography before subsequent subdigestion Žsee below.. Transformation of the raw spectra gave a measured mass of 4093 Da for peptide 7A and 4094 Da for 7CB. This same 1 Da increase was reiterated when peptides 7Af and 7CB f were compared; these had masses of 4121 and 4122 Da respectively. Of all the allowable point mutations, only four, Lys ™ Glu, Ile ™ Asn, Asn ™ Asp, and Gln ™ Glu, result in 1 Da increases in mass. Of these possibilities, Ile ™ Asn can be disregarded since it does not involve a change in charge, and Asn ™ Asp is not possible because there is no Asn in CNBr 7. This leaves either a Lys or Gln ™ Glu mutation as the most probable cause of the genetic variation. In order to define the precise location of the mutation, CNBr peptides 7A and 7CB were subdigested with trypsin and chymotrypsin and then analysed by ESI-MS. Scans of the tryptic peptides indicated that T1 Ž 549–557. and T6 Ž575–585. were present in both digests. These peptides had identical masses between the two digests indicating that 557, 574 Lys and 580 Gln were present and intact ŽFig. 4. . Scans of the chymotryptic subdigestion of CNBr 7 showed that Y4 Ž569– 585. was present and had the
Fig. 4. Amino acid sequence of CNBr 7, showing the location of tryptic ŽT. and chymotryptic ŽY. subpeptides together with their observed masses. At 1791 Da, peptide ŽY3X . was 1 Da more than its normal partner ŽY3.. C sCarboxamidomethyl Cysteine.
309
Fig. 5. Sequencing gel of PCR-amplified DNA from a normal control and Albumin Church Bay, showing the A ™G substitution in exon 13.
same mass Ž1654.7 Da. in both digests. However, peptide Y3 Ž555–568. showed a mass increase of 1 Da when the abnormal Ž1791.1 Da. and normal Ž1790.2 Da. forms were compared. Taken together, these data indicated a mutation of either 560 or 564 Lys to Glu. In order to establish which of these two lysine residues had mutated, genomic DNA was amplified using the PCR procedure and sequenced directly. The primers A25C and A26B were used to amplify a 381 -bp fragment encoding exon 13 of the albumin gene; residues Ala-528 through to Glu-571. Dideoxy chain termination sequencing analysis using A25C and A26B gave the expected sequence w9x up to base 15310 Žamino acid 559.. However both A and G were found at position 15311 ŽFig. 5.. This establishes that the proposita is heterozygous for a mutation of 560 Lys ŽAAG. ™ Glu ŽGAG.. This new variant has been named Albumin Church Bay.
4. Discussion Albumin Church Bay is a new genetic variant with a substitution of 560 Lys ™ Glu resulting from a
310
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
mutation of AAG ™ GAG at nucleotide 98 of the 13th exon. The increased anodal mobility on agarose gel electrophoresis is explained by the charge change associated with the Lys ™ Glu mutation. The codominant expression suggests that there is little structural perturbation associated with the substitution and this can be understood in the context of its location in the 3-dimensional structure of the molecule w9x. Residue 560 is situated in subdomain IIIB and immediately follows the Cys–Cys sequence at the end of helix III-9. The S–S bridges fix it rigidly in space with the charged side chain on the surface of the molecule. The substitution may however, have an impact on free fatty acid binding since the highest affinity long chain binding sites are located in this domain. Indeed the neighbouring mutation, 563 Asp ™ Asn in Albumin Paris 2, results in a doubling of its affinity for in vivo binding of total fatty acids and a specific increase in binding of the longer polyunsaturated essential fatty acids w21x. ESI-MS played an pivotal role in the identification of the 560 Lys ™ Glu substitution since earlier approaches using 2-dimensional and RP-HPLC peptide mapping failed to identify an aberrant peptide. ESI-MS analysis of the total tryptic digests however, allowed the immediate exclusion of ; 90% of the primary sequence and by default focused attention on the C-terminal region 558–574. This region was excised as a CNBr fragment Ž549–585. and found to have an altered reversed phased retention and a 1 Da increase in mass. Subdigestion with chymotrypsin confirmed a mass increase of 1 Da, from 1790.2 to 1791.1, in the segment 555–568 and DNA sequencing confirmed the mutation of 560 Lys ™ Glu. Minor modified forms of CNBr peptides 7A and 7CB were identified in the reversed-phase maps Ž Fig. 2.; at 4121 and 4122 Da respectively. These more hydrophobic components Ž 7Af and 7CB f . both differed from their more abundant counterparts by 28 Da and probably represent formylated forms of the peptides. Over time after purification, these masses reverted to normal Ž4093 and 4094 respectively. suggesting that the modification was a reversible one associated with the conditions of CNBr cleavage rather than an endogeneous protein modification.
Acknowledgements The authors wish to thank Andrew Fellowes and Howard Potter for expert technical assistance. References w1x T. Peter Jr., in: C.B. Anfinsen, J.T. Edsall, P.M. Richards ŽEds.., Advances in Protein Chemistry, vol. 37, Academic Press, New York, 1985, pp. 161–245. w2x T. Peters Jr., Clin. Chem. 33 Ž1987. 1317–1325. w3x S.O. Brennan, T. Myles, R.J. Peach, D. Donaldson, P.M. George, Proc. Natl. Acad. Sci. U.S.A. 87 Ž1990. 26–30. w4x W.C. Zan, W.F. Xu, C.W. Chi, Int. J. Pept. Protein Res. 41 Ž1993. 441–446. w5x J. Madison, M. Galliano, S. Watkins, L. Minchiotti, F. Porta, A. Rossi, F. Putnam, Proc. Natl. Acad. Sci. U.S.A. 91 Ž1994. 6476–6480. w6x C.E. Petersen, C.E. Ha, D.M. Jameson, N.V. Bhagavan, J. Biol. Chem. 271 Ž32. Ž1996. 19110–19117. w7x S.O. Brennan, Mol. Biol. Med. 6 Ž1989. 87–92. w8x S.O. Brennan, R.J. Peach, J. Biol. Chem. 266 Ž32. Ž1991. 21504–21508. w9x T. Peters Jr., ed., All About Albumin—Biochemistry, Genetics, and Medical Applications, Academic Press, New York, 1996. w10x S.O. Brennan, P.M. George, R.J. Peach, Clin. Chim. Acta 176 Ž1988. 179–184. w11x O.M.J. Bos, J.E. Fisher, J. Wilting, L.H.M. Jansen, Biochim. Biophys. Acta 953 Ž1988. 37–47. w12x C. Nelson, M. Noelkan, C. Buckley, C. Tanford, R. Hill, Biochemistry 4 Ž1965. 1418–1426. w13x S.O. Brennan, Biochim. Biophys. Acta 830 Ž1985. 320–324. w14x L.M. Kunkel, K.D. Smith, S.H. Boyer, D.S. Borgaonkar, S.S. Wachtel, O.J. Miller, W.R. Berg, H.W. Jones Jr., J.M. Rary, Proc. Natl. Acad. Sci. U.S.A. 74 Ž1977. 1245–1249. w15x P.P. Mingetti, D.E. Ruffner, W.J. Kuang, O.E. Dennison, J.W. Hawkins, W.G. Beattie, A. Dugaiczyk, J. Biol. Chem. 261 Ž1986. 6747–6757. w16x R.K. Saiki, D.H. Gelfand, S. Stoffel, S.J. Scharf, R. Higuchi, G.T. Horn, K.B. Mullis, H.A. Erlich, Science 239 Ž1988. 489–491. w17x C. Wong, C.E. Dowling, R.K. Saiki, R.G. Higuchi, H.A. Erlich, H.H. Kazazian Jr., Nature 330 Ž1987. 384–386. w18x O.M.J. Bos, J.E. Fischer, J. Wilting, L.H.M. Jansen, Biochim. Biophys. Acta 953 Ž1988. 37–47. w19x S.O. Brennan, Hemoglobin 1 Ž1977. 571–576. w20x R. Archer, C. Crocker, Biochim. Biophys. Acta 9 Ž1952. 704–708. w21x H. Nielsen, U. Kragh-Hansen, L. Minchiotti, M. Galliano, S.O. Brennan, A.L. Tarnoky, M. Franco, F.M. Salzano, O. Sugita, Biochim. Biophys. Acta 1342 Ž1997. 191–204.
Albumin Church Bay: 560 Lys ™ Glu a new mutation detected by electrospray ionisation mass spectrometry Evan K.M. Chua ) , Stephen O. Brennan, Peter M. George Molecular Pathology Laboratory, Department of Pathology, Christchurch School of Medicine, P.O. Box 4345, Christchurch, New Zealand Received 18 August 1997; accepted 27 August 1997
Abstract Albumin Church Bay is a fast migrating genetic variant of human serum albumin which, in a heterozygous subject, formed about 50% of the circulating albumin. Reversed phase peptide mapping and electrospray ionisation mass spectrometry ŽESI-MS. indicated that the C-terminal CNBr peptide had decreased polarity associated with a 1 Da increase in mass. Subdigestion of this peptide with trypsin and chymotrypsin revealed that the increased mass was associated with the chymotrypsin fragment VEKCCKADDKETCF Ž555–568. which had a mass of 1791.1 compared to 1790.2 for its normal counterpart. Sequence analysis of PCR-amplified DNA indicated an A ™ G mutation at position 98 of exon 13, which causes a point mutation of 560 Lys ™ Glu and results in a 1 Da mass increase. q 1998 Elsevier Science B.V. Keywords: Serum albumin; Amino acid substitution; Mass spectrometry; Point mutation; Formylation; Albumin Church Bay
1. Introduction Human serum albumin is the most abundant plasma protein Ž35–50 grl. with a circulating half life of 19 days in humans. It functions primarily as a transport protein for ligands such as long-chain fatty acids, drugs, bilirubin and thyroid hormones. There is a single copy of the albumin gene on chromosome four and this is made up of nearly 17 000 bp organised into 15 exons and 14 introns with exon 15 untranslated. The translated albumin is a single-chain protein consisting of 585 amino acids organised into three internally homologous domains w1x. There are nine Žactually eight and one-half. double loops formed by disulphide bonds involving adjacent half-cystine
)
Corresponding author. Fax: 64-3 364 0545.
residue. Albumin is synthesised in the liver as preproalbumin with an 18-residue prepeptide and a 6 residue propeptide Arg–Gly–Val–Phe–Arg–Arg. The prepeptide is cleaved off by signal peptidase when the precursor protein enters the lumen of the endoplasmic reticulum w2x. When the resulting proalbumin reaches the Golgi vesicles, its propeptide is removed by a membrane-bound enzyme Ž proalbumin convertase. to form mature albumin with aspartic acid as the amino-terminal residue w3x. The processed product is then constitutively secreted into the circulation w4x. More than thirty genetic variants of albumin have been identified so far w5x. However, unlike haemoglobin and coagulation factors variants, no adverse clinical effect on molecular function has been attributed to albumin variants, except for a condition known as FDH Ž familial dysalbuminaemic hyperthy-
0167-4838r98r$19.00 q 1998 Elsevier Science B.V. All rights reserved. PII S 0 1 6 7 - 4 8 3 8 Ž 9 7 . 0 0 1 4 7 - 7
306
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
roxinaemia. where a mutation of 218 His ™ Arg results in increased T4 binding w6x. None the less, the characterisation of genetic variants of albumin has provided unique insights into the specificity of the proteases involved in its processing w7,8x and facilitated the identification of the hepatic propeptide convertase w7,8x. Additional new variants will undoubtedly provide important clues about the molecular stability and help define ligand binding sites in the molecule w9x. We report here the molecular defect of a new fast albumin variant ŽAlbumin Church Bay. , which was found during routine clinical protein electrophoresis screening. ESI-MS played a key role in the identification of the new 560 Lys ™ Glu mutation. 2. Material and methods 2.1. Protein gel electrophoresis Agarose gel electrophoresis, performed in 1% agarose using Tris–barbitone buffer ŽpH 8.6., and Ni 63 autoradiography were performed as previously described w10x. 2.2. Albumin purification EDTA anticoagulated plasma was dialysed against 16 mM sodium acetate buffer ŽpH 5.2. and applied to a DEAE Sephadex A-50 anion exchange column
Ž1.6 = 30 cm., equilibrated in the same buffer. Bound albumin was eluted using a pH gradient from pH 5.2 to 4.5 w10x. Albumin peaks were pooled separately, concentrated and lyophilised. 2.3. Limited tryptic cleaÕage Native albumin, 5 mgrml in 10 ml 10 mM Tris– HCl, pH 8.0, was incubated at 208C for 120 min with 3 mg of trypsin and aliquots were analysed directly by agarose gel electrophoresis and by SDS-PAGE w11x. 2.4. Protein fragmentation Proteolytic digestions with trypsin and S. aureus V8 proteases were carried out on S-carboxamidomethylated albumin which was reduced and derivatised in 8 M urea as described previously w12x. 85 nmoles of S-carboxamidomethyl albumin was incubated with 0.1 mg TPCK trypsin Ž2% wt.rwt.. in 500 ml 0.5 M NH 4 HCO 3 , for 3 h at 378C. S. aureus V8 protease digestion was performed under the identical conditions except the digestion was for 24 h at room temperature. For CNBr cleavage, 4.5 mg of S-carboxamidomethyl albumin was incubated with ; 400 fold excess CNBr in 150 ml 70% HCOOH, for 24 h at room temperature in the dark. After dilution with 10
Fig. 1. Agarose gel electrophoresis ŽpH 8.6.. Lane 1, Normal human plasma; lane 2, patient plasma; lane 3, purified normal albumin component; lane 4, purified Albumin Church Bay; lane 5 and 6, partial tryptic digestion of native normal albumin Žtime 0 min and time 120 min respectively.; lane 7 and 8, partial tryptic digestion of native Albumin Church Bay Žtime 0 min and time 120 min respectively..
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
volumes of distilled water, the fragments were lyophilised and analysed by RP-HPLC. 2.5. HPLC Reverse-phase peptide maps were run on total tryptic and S. aureus V8 digests using the 50 mM phosphate buffer ŽpH 2.9.racetonitrile solvent system as previously described w13x. The column, however was a Waters 4 mm C 18 Novapak reverse-phase column. Peptide maps of CNBr digests were run on a Phenomenex W-Porex 5 C4 reverse-phase column. The initial solvent was 0.05% TFA and the final solvent was 0.05% TFA in 60% acetonitrile. The flow rate was 1 mlrmin. Because the aberrant peptide co-eluted with another CNBr fragment, it was further purified by re-injection and elution with a 25 mM ammonium acetate–25 mM ammonium acetater50% acetonitrile solvent system. The pure aberrant peptide was collected and lyophilised. All CNBr peptides were analysed and identified by ESIMS Želectrospray ionisation mass spectrometry..
307
nomic DNA encoding exon 13 of the albumin gene w15x Ž residues 528–571. was amplified by PCR w16x using primers A25CŽCATGCAGATGAGAATATTGAGAC . and A26B Ž CCTAAGCCCTAGCCTAACCAAAC.. Each 50 ml amplification reaction contained 1 unit Taq DNA polymerase Ž Boehringer Mannheim., 50 mM KCl, 15 mM MgCl 2 , 200 mM each of dATP, dTTP, dCTP and dGTP, 0.2 mM of each primer and 100 ng of template DNA. Forty cycles of denaturation Ž948C, 1 min. , annealing Ž 508C, 2 min. , and extension Ž 728C, 3 min. with a final extension phase of 5 min was performed in a PETC-1 thermal cycler Ž Perkin Elmer.. The PCR product was purified using Hi-Pure PCR purification cartridges ŽBoehringer Mannheim. and
2.6. Subdigestion of CNBr 7 The normal and aberrant peptide CNBr 7 were subdigested with trypsin and chymotrypsin. For chymotryptic digests, 17 nmoles of peptide was incubated with 50 mg chymotrypsin in 25 ml 0.5 M NH 4 HCO 3 , and incubated at 378C for 3 h. Similar conditions were used for tryptic subdigestion. 2.7. Screening of the albumin digests by mass spectrometry ESI-MS was performed on 10 ml of either total digest or purified peptide at a concentration of 10– 100 pmolrml in 50% acetonitriler0.1% formic acid. The instrument, a VG Platform II Ž Micromass. was operated in the positive ion mode and samples were delivered to the source at 5 mlrmin.The mrz range, 200–1400 was scanned every 3 s and data was acquired over 2 min. Mass spectrometer control and data processing used MassLynxe software supplied with the instrument. 2.8. Polymerase chain reaction (PCR) and DNA analysis Genomic DNA was extracted from whole blood by a standard method w14x. A 381 bp fragment of ge-
Fig. 2. Reverse-phase CNBr peptide maps of albumin A Žlower. and Albumin Church Bay Župper.. CNBr fragments were identified from their masses and are numbered from the N-terminal. Note that 7A which eluted at 14 min in albumin A elutes at 15.5 min Ž7CB. in the variant. Modified forms of these peptides, designated 7Af and 7CB f are also indicated.
308
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
sequenced directly by the dideoxy chain termination method w17x using 33P-labelled ddNTP ŽAmersham. and ThermoSequenaseTM ŽAmersham..
3. Results Routine clinical electrophoresis in agarose gels detected an individual with bisalbuminaemia resulting from a genetic variant with increased negative charge ŽFig. 1, lane 2.. The variant represented some 50% of the total albumin and the increased anodal migration suggested a Lys ™ Glu substitution. Anion exchange chromatography of plasma on DEAE-Sephadex gave two major albumin peaks. Electrophoretic analysis indicated that the earlier eluting peak was normal albumin Žlane 3. and the later peak the variant Ž lane 4. . Again, peak areas indicated that the variant accounted for 50% of the total albumin. Partial tryptic digestion was used to assess which domain was abnormal. Cleavage of normal native albumin Ž lane 6. at Arg-197 w18x produces a 43 kDa C-terminal fragment which is detectable by both
agarose gel electrophoresis and SDS-PAGE Žnot shown.. The N-terminal domain is further degraded and is not detectable by electrophoresis. Partial tryptic digestion of the variant Žlane 8. showed that the increased negative charge was in the large C-terminal fragment. These results established that the abnormality was in the second or third domain, beyond residue 197. To establish the precise location of the mutation, the albumin components were carboxamidomethylated and digested with trypsin and S. aureus V8 protease. However, no abnormalities were detected by either 2-dimensional peptide mapping w19,20x or RP-HPLC mapping w13x. Further tryptic and S. aureus V8 digests were analysed by ESI-MS, but again no abnormalities were detected. However, the observed mrz signals only accounted for ; 90% of the known albumin sequence; unaccounted for regions included residues 318–337 and 558–574. Carboxamidomethylated albumin fractions were subjected to CNBr cleavage and the fragments mapped by RP-HPLC Ž Fig. 2. . 10 ml of each peak was analysed directly by ESI-MS and the measured mass was used as the basis of the peak identification
Fig. 3. ESI-MS spectra of CNBr 7 fragments, showing normal control Ž7A. and variant Ž7CB. which differ by 1 Da. Also shown are modified forms of these peptides, 7Af and 7CB f , which have an increased mass of 28 Da compared to their respective counterparts.
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
in Fig. 2. Maps of the normal and variant albumin differed in one major respect. The C-terminal fragment Ž7A. was missing and replaced by a new more hydrophobic peptide Ž7CB. . Minor modified forms of these two peptides, designated 7Af and 7CB f , were also detected eluting later in the chromatogram. Raw ESI-MS spectra of these four peptides are shown in Fig. 3. All displayed major wM q 3Hx and wM q 4Hx ions, and all peptides were homogeneous except for 7CB which was contaminated with CNBr 4. This was removed by rechromatography before subsequent subdigestion Žsee below.. Transformation of the raw spectra gave a measured mass of 4093 Da for peptide 7A and 4094 Da for 7CB. This same 1 Da increase was reiterated when peptides 7Af and 7CB f were compared; these had masses of 4121 and 4122 Da respectively. Of all the allowable point mutations, only four, Lys ™ Glu, Ile ™ Asn, Asn ™ Asp, and Gln ™ Glu, result in 1 Da increases in mass. Of these possibilities, Ile ™ Asn can be disregarded since it does not involve a change in charge, and Asn ™ Asp is not possible because there is no Asn in CNBr 7. This leaves either a Lys or Gln ™ Glu mutation as the most probable cause of the genetic variation. In order to define the precise location of the mutation, CNBr peptides 7A and 7CB were subdigested with trypsin and chymotrypsin and then analysed by ESI-MS. Scans of the tryptic peptides indicated that T1 Ž 549–557. and T6 Ž575–585. were present in both digests. These peptides had identical masses between the two digests indicating that 557, 574 Lys and 580 Gln were present and intact ŽFig. 4. . Scans of the chymotryptic subdigestion of CNBr 7 showed that Y4 Ž569– 585. was present and had the
Fig. 4. Amino acid sequence of CNBr 7, showing the location of tryptic ŽT. and chymotryptic ŽY. subpeptides together with their observed masses. At 1791 Da, peptide ŽY3X . was 1 Da more than its normal partner ŽY3.. C sCarboxamidomethyl Cysteine.
309
Fig. 5. Sequencing gel of PCR-amplified DNA from a normal control and Albumin Church Bay, showing the A ™G substitution in exon 13.
same mass Ž1654.7 Da. in both digests. However, peptide Y3 Ž555–568. showed a mass increase of 1 Da when the abnormal Ž1791.1 Da. and normal Ž1790.2 Da. forms were compared. Taken together, these data indicated a mutation of either 560 or 564 Lys to Glu. In order to establish which of these two lysine residues had mutated, genomic DNA was amplified using the PCR procedure and sequenced directly. The primers A25C and A26B were used to amplify a 381 -bp fragment encoding exon 13 of the albumin gene; residues Ala-528 through to Glu-571. Dideoxy chain termination sequencing analysis using A25C and A26B gave the expected sequence w9x up to base 15310 Žamino acid 559.. However both A and G were found at position 15311 ŽFig. 5.. This establishes that the proposita is heterozygous for a mutation of 560 Lys ŽAAG. ™ Glu ŽGAG.. This new variant has been named Albumin Church Bay.
4. Discussion Albumin Church Bay is a new genetic variant with a substitution of 560 Lys ™ Glu resulting from a
310
E.K.M. Chua et al.r Biochimica et Biophysica Acta 1382 (1998) 305–310
mutation of AAG ™ GAG at nucleotide 98 of the 13th exon. The increased anodal mobility on agarose gel electrophoresis is explained by the charge change associated with the Lys ™ Glu mutation. The codominant expression suggests that there is little structural perturbation associated with the substitution and this can be understood in the context of its location in the 3-dimensional structure of the molecule w9x. Residue 560 is situated in subdomain IIIB and immediately follows the Cys–Cys sequence at the end of helix III-9. The S–S bridges fix it rigidly in space with the charged side chain on the surface of the molecule. The substitution may however, have an impact on free fatty acid binding since the highest affinity long chain binding sites are located in this domain. Indeed the neighbouring mutation, 563 Asp ™ Asn in Albumin Paris 2, results in a doubling of its affinity for in vivo binding of total fatty acids and a specific increase in binding of the longer polyunsaturated essential fatty acids w21x. ESI-MS played an pivotal role in the identification of the 560 Lys ™ Glu substitution since earlier approaches using 2-dimensional and RP-HPLC peptide mapping failed to identify an aberrant peptide. ESI-MS analysis of the total tryptic digests however, allowed the immediate exclusion of ; 90% of the primary sequence and by default focused attention on the C-terminal region 558–574. This region was excised as a CNBr fragment Ž549–585. and found to have an altered reversed phased retention and a 1 Da increase in mass. Subdigestion with chymotrypsin confirmed a mass increase of 1 Da, from 1790.2 to 1791.1, in the segment 555–568 and DNA sequencing confirmed the mutation of 560 Lys ™ Glu. Minor modified forms of CNBr peptides 7A and 7CB were identified in the reversed-phase maps Ž Fig. 2.; at 4121 and 4122 Da respectively. These more hydrophobic components Ž 7Af and 7CB f . both differed from their more abundant counterparts by 28 Da and probably represent formylated forms of the peptides. Over time after purification, these masses reverted to normal Ž4093 and 4094 respectively. suggesting that the modification was a reversible one associated with the conditions of CNBr cleavage rather than an endogeneous protein modification.
Acknowledgements The authors wish to thank Andrew Fellowes and Howard Potter for expert technical assistance. References w1x T. Peter Jr., in: C.B. Anfinsen, J.T. Edsall, P.M. Richards ŽEds.., Advances in Protein Chemistry, vol. 37, Academic Press, New York, 1985, pp. 161–245. w2x T. Peters Jr., Clin. Chem. 33 Ž1987. 1317–1325. w3x S.O. Brennan, T. Myles, R.J. Peach, D. Donaldson, P.M. George, Proc. Natl. Acad. Sci. U.S.A. 87 Ž1990. 26–30. w4x W.C. Zan, W.F. Xu, C.W. Chi, Int. J. Pept. Protein Res. 41 Ž1993. 441–446. w5x J. Madison, M. Galliano, S. Watkins, L. Minchiotti, F. Porta, A. Rossi, F. Putnam, Proc. Natl. Acad. Sci. U.S.A. 91 Ž1994. 6476–6480. w6x C.E. Petersen, C.E. Ha, D.M. Jameson, N.V. Bhagavan, J. Biol. Chem. 271 Ž32. Ž1996. 19110–19117. w7x S.O. Brennan, Mol. Biol. Med. 6 Ž1989. 87–92. w8x S.O. Brennan, R.J. Peach, J. Biol. Chem. 266 Ž32. Ž1991. 21504–21508. w9x T. Peters Jr., ed., All About Albumin—Biochemistry, Genetics, and Medical Applications, Academic Press, New York, 1996. w10x S.O. Brennan, P.M. George, R.J. Peach, Clin. Chim. Acta 176 Ž1988. 179–184. w11x O.M.J. Bos, J.E. Fisher, J. Wilting, L.H.M. Jansen, Biochim. Biophys. Acta 953 Ž1988. 37–47. w12x C. Nelson, M. Noelkan, C. Buckley, C. Tanford, R. Hill, Biochemistry 4 Ž1965. 1418–1426. w13x S.O. Brennan, Biochim. Biophys. Acta 830 Ž1985. 320–324. w14x L.M. Kunkel, K.D. Smith, S.H. Boyer, D.S. Borgaonkar, S.S. Wachtel, O.J. Miller, W.R. Berg, H.W. Jones Jr., J.M. Rary, Proc. Natl. Acad. Sci. U.S.A. 74 Ž1977. 1245–1249. w15x P.P. Mingetti, D.E. Ruffner, W.J. Kuang, O.E. Dennison, J.W. Hawkins, W.G. Beattie, A. Dugaiczyk, J. Biol. Chem. 261 Ž1986. 6747–6757. w16x R.K. Saiki, D.H. Gelfand, S. Stoffel, S.J. Scharf, R. Higuchi, G.T. Horn, K.B. Mullis, H.A. Erlich, Science 239 Ž1988. 489–491. w17x C. Wong, C.E. Dowling, R.K. Saiki, R.G. Higuchi, H.A. Erlich, H.H. Kazazian Jr., Nature 330 Ž1987. 384–386. w18x O.M.J. Bos, J.E. Fischer, J. Wilting, L.H.M. Jansen, Biochim. Biophys. Acta 953 Ž1988. 37–47. w19x S.O. Brennan, Hemoglobin 1 Ž1977. 571–576. w20x R. Archer, C. Crocker, Biochim. Biophys. Acta 9 Ž1952. 704–708. w21x H. Nielsen, U. Kragh-Hansen, L. Minchiotti, M. Galliano, S.O. Brennan, A.L. Tarnoky, M. Franco, F.M. Salzano, O. Sugita, Biochim. Biophys. Acta 1342 Ž1997. 191–204.