Clinical Chemistry
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 47: 2012-2022, 2001;
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mills, K.
Right arrow Articles by Winchester, B. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mills, K.
Right arrow Articles by Winchester, B. G.
Related Collections
Right arrow Molecular Diagnostics and Genetics
Right arrow Proteomics and Protein Markers
Right arrow Automation and Analytical Techniques
(Clinical Chemistry. 2001;47:2012-2022.)
© 2001 American Association for Clinical Chemistry, Inc.

Articles

Identification of {alpha}1-Antitrypsin Variants in Plasma with the Use of Proteomic Technology

Kevin Mills1, Philippa B. Mills1, Peter T. Clayton1, Andrew W. Johnson1, David B. Whitehouse2 and Bryan G. Winchester1a

1 Biochemistry Endocrinology and Metabolism Unit, Institute of Child Health at Great Ormond Street Hospital, University College London, 30 Guilford St., London WC1 N 1EH, United Kingdom.

2 Medical Research Council Human Biochemical Genetics Unit, Galton Laboratory, University College London, London NW1 2HE, United Kingdom.

aAuthor for correspondence. Fax 44-0207-404-6191; e-mail B.Winchester{at}ich.ucl.ac.uk.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Background: Proteomic technology permits the investigation of genetic metabolic diseases at the level of protein expression. Changes in the expression, polypeptide structure, and posttranslational modification of individual proteins can be detected in complex mixtures of proteins.

Methods: We used high-resolution two-dimensional polyacrylamide gel electrophoresis to separate isoforms of plasma proteins and detect abnormalities of mass and/or charge. We confirmed the identity of the separated proteins by in-gel digestion with proteases and N-glycanases and then analyzed the released peptides and glycans by matrix-assisted laser-desorption ionization–time-of-flight mass spectrometry.

Results: Complete characterization of the polypeptide sequences and glycosylation of {alpha}1-antitrypsin isoforms was achieved in plasma from controls and from patients with three different known {alpha}1-antitrypsin deficiencies and congenital disorder of glycosylation type Ia.

Conclusions: This study shows that proteomic techniques are a powerful and sensitive means of detecting changes in the amino acid sequence and abnormal posttranslational modifications of specific proteins in a complex biologic matrix.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
The rapid expansion in proteomic research over the last 5 years has been brought about by the need to bridge the gap between the sequence of a gene and the functional activity of its protein product. The genome of a cell allows the prediction of the amino acid sequence of a protein, but it does not indicate whether that protein is actually expressed or the extent of its expression (1). Some potential secondary and tertiary structural features and posttranslational motifs in a protein can be recognized in a gene sequence. However, it is not possible to predict whether these features and motifs will be utilized in the mature functional protein unless the physiologic environment of its site of action is known, i.e., the pattern of expression of proteins in its tissue of origin (2). Proteomics not only has the ability to provide fundamental insights into the molecular basis of the physiologic state of a cell, it also has the ability to reveal how cascades of protein expression change as a result of specific disorders or drug treatment (3).

In the investigation of a genetic metabolic disorder caused by mutations in the gene encoding a protein (which may be an enzyme or transporter or structural protein), proteomic analysis should be capable of the following: (a) detection of the altered amino acid sequence in the mutated protein; (b) detection of altered posttranslational modification of the mutated protein and other proteins; (c) measurement of the degree of expression of the mutated protein and of other proteins involved in the pathogenesis of tissue damage; and (d) detection of polymorphisms in the mutated protein (which may modulate the mutation) and in other proteins (which may exacerbate or diminish the effect of the mutation by affecting the overall activity of the metabolic pathway involved). Given the number of proteins in tissue, these are no small tasks.

This report examines two applications of proteomic technology: (a) the detection of mutant and polymorphic amino acid sequences and (b) the detection of posttranslational modifications. We have chosen plasma {alpha}1-antitrypsin as a target protein because plasma is readily available and is an important substrate for clinical investigation and because there are known diseases that result from changes in the amino acid sequence and the posttranslational modifications of {alpha}1-antitrypsin [the {alpha}1-antitrypsin deficiencies (4)]. A deficiency in {alpha}1-antitrypsin can lead to emphysema in the third and fourth decades of life, and some defects in {alpha}1-antitrypsin [notably the protease inhibitor Z (PIZ) 1 variant] can cause liver disease presenting in infancy or childhood (5)(6). The PIZ {alpha}1-antitrypsin deficiency is the most common metabolic disease for which individuals currently undergo orthotopic liver transplantation (7). The primary function of {alpha}1-antitrypsin is the inhibition of circulatory serine proteases or serpins, with the main target enzyme being neutrophil elastase, which is found primarily in the lung (8). The pathophysiology of the lung disease has been explained by the deficiency of plasma {alpha}1-antitrypsin, permitting an elastolytic attack on the lung by neutrophil elastase (9). The etiology of the liver disease has been more contentious (8)(9)(10). Most investigators believe that the amino acid substitution leads to the misfolding of the protein, leading to polymerization and a subsequent build up of {alpha}1-antitrypsin in the hepatocytes of the liver (9)(11)(12). However, little is known about the way in which this build up leads to inflammation and cirrhosis.

Mature {alpha}1-antitrypsin has a molecular weight of Mr 52 000, consists of a single polypeptide chain of 394 amino acids, and is ~12% carbohydrate by weight (4). The {alpha}1-antitrypsin molecule is synthesized in the liver before secretion into the circulation where it has a plasma concentration of ~1.3 g/L (4) and a half-life of 4–5 days (5). The {alpha}1-antitrypsin molecule carries a high negative charge because of sialic acid residues on the three complex glycans attached to asparagine residues 46, 83, and 247 (13).

Isoelectric focusing of plasma {alpha}1-antitrypsin leads to the detection of eight bands, which are numbered M1 to M8 (anodal-low pH to cathodal-high pH) (14). The bands M4 and M6 are the most abundant of the isoforms, making up 40% and 34% of the total plasma {alpha}1-antitrypsin, respectively, whereas M3 and M5 are present in only trace amounts (4). The multiple forms of {alpha}1-antitrypsin are predominantly attributable to the presence of different numbers of sialic acid residues present on the glycans. Isoforms containing more triantennary complex glycans, and thus more sialic acid residues, have lower pI values. Bands M4 and M6 have one triantennary plus two biantennary glycans and three biantennary glycans, respectively (13). The two minor cathodal isoforms, bands M7 and M8, have the same carbohydrate structure as the major bands, M4 and M6, but M7 and M8 lack the five N-terminal amino acids (Glu-Asp-Pro-Gln-Gly) (15). The loss of these five amino acids by posttranslational cleavage causes an additional cathodal shift of the isoforms because of the loss of the negatively charged amino acids, glutamic acid (Glu 1) and aspartic acid (Asp 2) (15).

The combination of a well-characterized protein at a high concentration in plasma and a well-characterized disease resulting from the deficiency of that protein makes {alpha}1-antitrypsin an ideal candidate for this preliminary study in the application of proteomic techniques to genetic metabolic disease. This report describes the proteomic characterization of {alpha}1-antitrypsin from controls and the analysis of four {alpha}1-antitrypsin variants resulting from changes in a single amino acid in the sequence and/or posttranslational modifications.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
All chemical reagents used were of research grade and obtained from Sigma-Aldrich, unless stated otherwise. Ultrapure electrophoretic-grade acrylamide (30% by weight) was obtained from National Diagnostics Ltd. Anti-{alpha}1-antitrypsin antibodies were coupled to Hi-Trap NHS-affinity chromatography columns (Amersham Pharmacia Biotech) according to the manufacturer’s instructions. Immobiline DryStrip isoelectric focusing strips (18 cm; pH 4.5–5.5) were obtained from Amersham Pharmacia Biotech. Piperazine diacrylamide cross-linker was obtained from Bio-Rad. All of the proteases used were of sequence grade and were obtained from Promega Ltd. N-glycanase and Glyco H graphite columns were obtained from Glyko.

molecular variants of {alpha}1-antitrypsin
Plasma was obtained from four patients who had been diagnosed previously by biochemical and molecular genetic techniques with specific alterations in the structure of {alpha}1-antitrypsin (Fig. 1 ).



View larger version (44K):
[in this window]
[in a new window]
  		Figure 1. Amino acid sequence of {alpha}1-antitrypsin showing mutants analyzed in this study.

———, N-terminal amino acids (five); {circ}, N-glycosylation sites 46, 83, and 247; , PIZBristol (threonine85 to methionine); {square}, MIA213 (valine213 to alanine); {diamond}, PIZZ (glutamate342 to lysine).

Patient 1 was homozygous for a valine-to-alanine substitution at amino acid 213 in {alpha}1-antitrypsin, leading to a neutral amino acid substitution (MIA213).

Patient 2 was homozygous for a glutamate-to-lysine substitution at amino acid 342 in {alpha}1-antitrypsin, leading to a charged amino acid substitution (PIZZ).

Patient 3 was homozygous for a deficiency in phosphomannomutase [congenital disorder of glycosylation type Ia (CDG-Ia)] (16), which causes underglycosylation of {alpha}1-antitrypsin, leading to a change in molecular weight and charge.

Patient 4 was heterozygous for a C->T transition at codon 85, which changes threonine 85 (ACG) to methionine (ATG) and abolishes the N-glycosylation sequon (NLT) at amino acids 83–85, preventing glycosylation of asparagine 83 (PIZBristol) (17).

gel electrophoresis
All two-dimensional polyacrylamide (2D-PAGE) analyses were performed according to Swiss-Prot protocols (18)(19)(20)(21)(22) with minor modifications.

one-dimension isoelectric focusing
We focused plasma samples for both analytical (2.5 µL) and preparative gels (10 µL for peptide mass fingerprinting and 30 µL for glycan analysis) on Immobiline DryStrips (18 cm; pH 4.5–5.5) using a LKB-Multiphor II focusing unit (Amersham Pharmacia Biotech). Isoelectric focusing was carried out for a minimum of 75 kV h and a maximum of 100 kV·h. Focused strips were snap-frozen in liquid nitrogen and stored at -80 °C until later use.

two-dimension sodium dodecyl sulfate–page
Analytical gels consisted of 10% (by weight) acrylamide with 1.6% cross-linker piperazine diacrylamide. We prepared preparative gels for in-gel proteolytic digestions using 10% (by weight) acrylamide with 0.1% or 1.3% bis-acrylamide for chymotrypsin and trypsin, respectively. We performed resolubilization of the focused proteins and carboamidomethylation of cysteine residues according to Diettrich et al. (23). Sodium dodecyl sulfate–PAGE was carried out in a Protean II Multicell electrophoresis unit (Bio-Rad).

detection of separated proteins
Analytical gels were silver-stained with an automatic gel stainer (Amersham Pharmacia Biotech) as described by Hochstrasser et al. (22), and the preparative gels were silver-stained according to Shevchenko et al. (24).

in-gel proteolytic digestion and extraction of peptides and glycans
In-gel tryptic digestions were performed as described by Shevchenko et al. (24). In-gel chymotryptic digestions were performed in a similar manner but with the following minor modifications. The gel slices were incubated overnight with an excess of chymotrypsin (60 µL of 25 g/L), and the ammonium bicarbonate buffer was replaced with 100 mmol/L Tris-HCl, pH 7.8. The released peptides were extracted from the gel by shaking the gel slices with 300 µL of 50% acetonitrile containing 0.1% trifluoroacetic acid on a rotary mixer for 60 min. The acetonitrile was removed from the gel slice and taken down to dryness by centrifugal evaporation. Any residual peptides remaining in the gel slice were extracted by shaking the gel slice with 300 µL of 4 mol/L urea on a rotary mixer for 30 min. After the second extraction, the urea solution was removed from the gel slice and added to the dried peptides extracted with acetonitrile. This solution was vortex-mixed thoroughly for 60 s to resolubilize all peptides before desalting on a C-18 stationary-phase microcolumn as described by Mills et al. (25). The desalted peptide solution was dried by centrifugal evaporation and reconstituted in 10 µL of 1 mL/L trifluoroacetic acid for mass spectrometry.

Preparative gels for glycan analysis were stained with Coomassie blue, and glycans were released by in-gel digestion with PNGase F according to Kuster et al. (26). The released glycans were desalted with graphite Glyco H columns (Glyko) before analysis by mass spectrometry.

identification of proteins separated by 2d-page
Proteins separated by 2D-PAGE were identified by mass-mapping studies (see below). The pI and molecular weight coordinates for our 2D-PAGE system were then determined by comparison of the identified proteins with the Swiss-Prot online plasma 2D-PAGE database (Geneva, Switzerland; http://www.expasy.ch/cgi bin/map2/def?PLASMA_HUMAN).

mass spectrometry
Mass spectrometry was carried out on a matrix-assisted laser desorption ionization–time-of-flight mass spectrometry (MALDI-TOF-MS) instrument fitted with a reflectron and a 337-nm ultraviolet laser (TOF Spec E; MicroMass). Peptide analyses were performed in positive-ion mode with the following voltages: source, 20 kV; extraction, 19.95 kV; focus, 16.5 kV; and reflectron, 25 kV. Spectra were acquired by calculating the mean value of five scans with the highest signal. Data were acquired in reflectron mode, operating over a mass range of m/z 6000 with matrix suppression set at 650 Da. Peptides were analyzed with an {alpha}-cyano-4-hydroxycinnamic acid–fucose comatrix as described by Mills et al. (25).

Glycan analyses were performed in negative-ion mode with voltages of 20 kV for the source, 19.95 kV for the extraction, 16.5 kV for the focus, and a pulse voltage of 900 V. Spectra were acquired by calculating the mean value of five scans with the highest signal. Data were acquired in linear mode, operating over a mass range of m/z 4000 with matrix suppression set at 800 Da. Glycans were analyzed with trihydroxyacetophenone–20 mmol/L ammonium citrate comatrix according to Papac et al. (27).

Data analysis was carried out with MassLynx data analysis software (Protein Prospector database software, the University of San Francisco; http://falcon.ludwig.ucl.ac.uk/mshome3.2.htm) and PAWS proteomic analysis software.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
2D-PAGE analysis of control plasma indicated the presence of six putative isoforms of {alpha}1-antitrypsin on the basis of their pI value and size (Fig. 2A ). Tryptic peptide mass mapping confirmed that all of these components were {alpha}1-antitrypsin and identified several other proteins. Variation in the electrophoretograms for healthy subjects was extensively studied. No changes in the expression or the ratio of pI/Mr were observed in any of the 2D-PAGE analyses of the many control plasmas analyzed. The expression of the {alpha}1-antitrypsin was remarkably consistent among the control individuals, and the electrophoretogram included is representative of a typical 2D-PAGE pattern for {alpha}1-antitrypsin in control plasma samples. Several control samples were analyzed at the same time as every {alpha}1-antitrypsin variant to avoid batch-to-batch differences during electrophoresis and staining. The precise identity of each isoform was deduced from glycan analysis and the intensity and peptide mass spectra of the isoform. Peptide mass mapping of the two most abundant isoforms, the putative M4 and M6 isoforms, with trypsin and chymotrypsin, respectively, showed the presence of peptides of masses m/z 2819.2 and m/z 3704.7, respectively (Fig. 3, A and B ). These masses correspond to peptides containing the amino acids 1–25 and 1–33, respectively, in the sequence of {alpha}1-antitrypsin and indicate that these isoforms do contain the first five amino acids, as predicted by Jeppsson et al. (15). Both of these peptides contain missed cleavage sites at lysine 10 and phenylalanine 23, respectively. No ions corresponding to peptides containing the five N-terminal amino acids were observed in the analysis of the predicted M7 (Fig. 3C ) and M8 isoforms (Fig. 3D ), supporting their identification. However, a tryptic peptide mass (m/z 1779.8) corresponding to residues 11–26 was detected. We postulate that tryptic cleavage after lysine 10 occurs more readily in the truncated protein than in the intact protein, preventing the appearance of a peptide corresponding to residues 6–25 (m/z 2293.0). The use of the two proteases independently increased the coverage of the amino acid sequence to 86% from 65.2% and 52.3% for trypsin and chymotrypsin, respectively.



View larger version (89K):
[in this window]
[in a new window]
  		Figure 2. Analytical 2D-PAGE of control and MIA213 plasma samples and a sample from a heterozygote for PIMZBristol.

Enlarged 2D-PAGE sections of plasma {alpha}1-antitrypsin (A), MIA213 plasma {alpha}1-antitrypsin (B), and plasma {alpha}1-antitrypsin from a heterozygote for PIMZBristol (C) within a narrow pH range of 4.5–5.5 for focusing in the first dimension. Ordinate values are Mr of known proteins. (C), spots 1–4 show additional {alpha}1-antitrypsin isoforms.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 3. MALDI-TOF-MS of tryptic and chymotryptic digests of {alpha}1-antitrypsin isoforms in control plasma.

(A), tryptic digest of the M4 isoform showing a peptide of m/z 2819.2, corresponding to amino acids 1–25. (B), chymotryptic digest of the M6 isoform showing a peptide of m/z 3704.7, corresponding to amino acids 1–33. (C), tryptic digest of the M7 isoform showing no extra peptides corresponding to amino acids 1–25 in the sequence. (D), chymotryptic digest of the M8 isoform showing no extra peptides corresponding to amino acids 1–33 in the sequence.

Further evidence for the identity of the major isoforms M4 and M6 was obtained by analysis of glycans released by in-gel digestion with PNGase F. Fig. 4A shows the glycan analysis of the M4 isoform. Two major glycans were detected, disialo-biantennary (m/z 2223.4) and trisialo-triantennary (m/z 2880.1), in a ratio of ~2:1. The next most abundant peak, m/z 1932.2, is probably attributable to in-source loss of sialic acid from the major biantennary glycan. Masses corresponding to fucosylated bi- (m/z 2369.6) and triantennary (m/z 3026.3) structures were also observed and constituted 5% and 7% of the total glycan signal, respectively. The glycan analysis of the predicted M6 isoform showed the presence of only one major species, the disialo-biantennary complex glycan structure (Fig. 4B ). As with the analysis of the M4 spot, a mass corresponding to a fucosylated biantennary structure was also observed, which constituted ~8% of the total glycan signal.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 4. MALDI-TOF-MS of glycans released from {alpha}1-antitrypsin isoforms.

Mass spectra of the glycans obtained from the in-gel N-glycanase digestion of the M4 (A) and M6 (B) isoforms of {alpha}1-antitrypsin.

These results are in accord with previously published data for the microheterogeneity of glycosylation of the major isoforms of {alpha}1-antitrypsin in human plasma (13)(14)(15). The M4 isoform of {alpha}1-antitrypsin has been reported previously to contain two biantennary and one triantennary complex glycan, which is in agreement with our observation of bi- to triantennary glycans in a ratio of 2:1 (4). Similarly, our detection of only biantennary glycans on M6 agrees with the data published by Jeppsson et al. (12). Naitoh et al. (28) reported that the M4 and M6 {alpha}1-antitrypsin isoforms contain small amounts of fucosylated glycans, which again is in agreement with our observations.

The glycosylation analyses for the putative M7 and M8 isoforms were the same as those for the M4 and M6 isoforms, respectively, confirming that their different pI values were attributable to the loss of the five N-terminal amino acids and not to a difference in glycosylation.

Unfortunately, the mass spectral signals for the analyses of the glycans on the M1 and M2 isoforms of {alpha}1-antitrypsin were extremely weak, and it was not possible to identify the microheterogeneity of the glycosylation of these isoforms unequivocally. Peptide analyses from these spots indicated that the first five amino acids in the {alpha}1-antitrypsin sequence were present in both of these isoforms. The presence of the first five amino acids indicates that the different pI values of these isoforms must be attributable to glycan heterogeneity and not to posttranslational proteolysis. We therefore postulate that the isoforms M1 and M2 contain three triantennary glycans and two tri- plus one biantennary glycan, respectively, for the isoforms to migrate to the pI values observed with 2D-PAGE.

The use of high-resolution isoelectric focusing over the pH range 4.5–5.5 permitted complete separation of the isoforms of {alpha}1-antitrypsin, which was not possible with the use of wider pH ranges of ampholines. The ability to "focus" on particular pH ranges allows greater resolution of individual proteins and, more significantly, increases the probability of resolving the protein of interest from other proteins of similar pI and mass for analysis by mass spectrometry. The combination of glycan analysis to determine the microheterogeneity of glycosylation and peptide mass fingerprinting permitted the accurate identification of the abundant isoforms, M4, M6, M7, and M8, and a reasonable prediction of the structure of the M1 and M2 isoforms of {alpha}1-antitrypsin separated by 2D-PAGE (see Fig. 2AUp ).

analysis of molecular variants of {alpha}1-antitrypsin
Neutral amino acid substitution (V213A).
No marked changes in the protein intensity, pI, or molecular weight of the {alpha}1-antitrypsin isoforms were observed when the 2D-PAGE analyses of control plasma samples were compared with patient (V213A) plasma samples (Fig. 2, A and BUp , respectively). This is consistent with a neutral amino acid substitution, which would not be expected to alter the physical characteristics of {alpha}1-antitrypsin.

The MALDI-TOF-MS analyses of the tryptic digests of the control and patient {alpha}1-antitrypsin isoforms were similar, except for the fragment corresponding to the tryptic peptide containing amino acids 192–217. Masses of m/z 3148.5 and m/z 3120.5 were observed from the each {alpha}1-antitrypsin equivalent isoform from the control and V231A variant isoforms, respectively (Fig. 5, A and B , respectively). These masses differ by 28 m/z and are consistent with a peptide containing a valine-to-alanine substitution at residue 213 in the variant. The mass of m/z 3148.5 derived from the control {alpha}1-antitrypsin was not observed in the {alpha}1-antitrypsin from the V213A patient and vice versa.



View larger version (28K):
[in this window]
[in a new window]
  		Figure 5. MALDI-TOF-MS of the tryptic in-gel digestion of control, MIA213, and PIZZ (GLU342) {alpha}1-antitrypsin.

(A), tryptic in-gel digest of the control M4 isoform showing a peptide of mass m/z 3148.5, corresponding to amino acids 199–217 containing valine at position 213. (B), tryptic in-gel digest of the MIA213 M4 isoform showing a peptide of mass m/z 3120.5, corresponding to amino acids 199–217 containing alanine at position 213. (C), enlarged 2D-PAGE section of control plasma {alpha}1-antitrypsin within a narrow pH range of 4.5–5.5. (D), enlarged 2D-PAGE section of PIZZ (GLU342) plasma {alpha}1-antitrypsin within a narrow range pH 4.5–5.5.

Charged amino acid substitution (E342K): PIZ variant.
The 2D-PAGE analysis of PIZZ patient plasma showed changes in the intensity and pI of all the {alpha}1-antitrypsin isoforms relative to the control plasma sample (Fig. 6, A and B ). The intensities of the main M4 and M6 isoforms in the PIZZ plasma were ~20% of the intensities of the equivalent spots in the control plasma. The concentration of {alpha}1-antitrypsin in PIZZ plasma has been reported by Sveger et al. (29) to be 17% ± 3% of that in nondiseased plasma. The six {alpha}1-antitrypsin isoforms in the PIZZ plasma showed a cathodal shift relative to their counterparts in control plasma. The mean cathodal shift in pI between each of the PIZZ isoforms and its equivalent control isoform was 0.1 pH units. This is approximately double the pI difference observed between adjacent isoforms in the same plasma sample (0.05 pH units), which differ by one sialic acid residue or one negative charge. Thus the increase in pI of each PIZZ isoform relative to its control counterpart can be explained by the substitution of a positively charged lysine residue for a negatively charged glutamate residue with a net decrease in negative charge of 2, which is equivalent to an increase in pI of ~0.1 pH units.



View larger version (76K):
[in this window]
[in a new window]
  		Figure 6. Analytical 2D-PAGE of control and PIZ plasma.

Enlarged 2D-PAGE section of control plasma {alpha}1-antitrypsin (A) and PIZZ (GLU342) plasma {alpha}1-antitrypsin (B) within a narrow pH range of 4.5–5.5.

As the concentration of {alpha}1-antitrypsin in PIZZ plasma is quite low, it was initially purified by immunoaffinity chromatography followed by one-dimensional PAGE to optimize the digestion conditions for analysis by mass spectrometry. A small amount of contaminating albumin (confirmed by mass mapping) was detected in the one-dimensional PAGE, but it was clearly resolved from the {alpha}1-antitrypsin. The separated {alpha}1-antitrypsin was digested in-gel with trypsin and the resulting peptide mixture analyzed by MALDI-TOF-MS. Only peptides containing a lysine residue at position 342, as opposed to a glutamate, were detected in the analysis.

Having defined the digestion and mass spectrometry conditions for detection of the amino acid change, we applied these conditions to the {alpha}1-antitrypsin isoforms separated by 2D-PAGE. The spectra observed from the in-gel digestion of isoform M4 in the control and PIZZ plasma samples over the mass range m/z 2020–2640 are shown in Fig. 5, C and DUp , respectively. Three extra masses, m/z 2387.2, m/z 2403.2, and m/z 2419.2, were observed in the PIZZ {alpha}1-antitrypsin analysis, which were not observed in the control analysis. These masses correspond to a peptide covering amino acids 343–365 in the sequence of {alpha}1-antitrypsin after cleavage between the two lysine residues at positions 342 and 343. The observation of three masses for a single peptide is explained by the presence of the same peptide in the three forms, i.e., native (m/z 2387.2) and with one (m/z 2403.2) or two (m/z 2419.2) oxidized methionine residues. This confirmed that the PIZZ patient was homozygous for the E342K mutation. No peptide covering amino acids 343–365 in the sequence was detected in the analysis of the control plasma. An ion corresponding to residues Ala-336–Lys- 342 (m/z 887.6) could not be detected because it would fall into the matrix noise signal observed on the mass spectrometer. A possible explanation for the detection of a peptide covering the mutation region in the PIZZ sample, but not in the control, could be the nature of the amino acid substitution. The substitution of a lysine for a glutamate residue adjacent to another lysine residue at amino acid 343 creates a highly positively charged dilysine motif in the variant. This would be expected to enhance tryptic cleavage between the two lysines. In contrast, the negatively charged glutamic acid at position 342 in the normal sequence would be expected to inhibit cleavage after lysine 343. In fact, this area of the sequence is intransigent to in-gel tryptic digestion even after prior deglycosylation (25). The three extra masses of native m/z 2387.2, m/z 2403.2, and m/z 2419.2 were found in all the {alpha}1-antitrypsin isoforms from the PIZ variant but not in the control.

underglycosylation
CDG type I results from a deficiency of phosphomannomutase, which leads to a decreased capacity for asparagine-linked glycosylation of proteins. An important diagnostic manifestation of CDG-Ia is the appearance of more cathodal glycoforms of plasma/serum glycoproteins on electrophoresis attributable to their underglycosylation and consequent loss of sialic acid residues. Seven extra {alpha}1-antitrypsin isoforms (spots 1a–7a) were observed in the 2D-PAGE analysis of plasma from a patient with CDG-Ia (Fig. 7A ). These were analyzed by peptide mass fingerprinting to establish the molecular basis of their different pI values and masses. Spots 1a–5a all had Mr ~2000 less than the corresponding M series of {alpha}1-antitrypsin isoforms in control plasma, which were also present in the CDG-Ia plasma sample (Fig. 7A ). This decrease in mass is consistent with the loss of one complex N-linked glycan. Spots 6a and 7a had a Mr ~4000 less than the control M series, consistent with the loss of two complex glycans. These mass differences suggest that isoforms 1a-7a are underglycosylated compared with the control M isoforms.



View larger version (56K):
[in this window]
[in a new window]
  		Figure 7. 2D-PAGE and mass spectra of plasma {alpha}1-antitrypsin from CDG, control, and PIMZBristol plasma.

(A), 2D-PAGE of CDG plasma {alpha}1-antitrypsin and mass spectra of the tryptic in-gel digestion of spot 3a, showing a peptide of mass m/z 3691.8, corresponding to amino acids 70–101 with an asparagine at position 83. (B), 2D-PAGE of control plasma {alpha}1-antitrypsin and mass spectra of the tryptic in-gel digestion of the M6 isoform after prior enzymatic removal of the glycans. A new peptide of mass m/z 3692.8 corresponding to amino acids 70–101 with an aspartic acid at position 83 is observed. (C), 2D-PAGE of control plasma {alpha}1-antitrypsin and mass spectra of the tryptic in-gel digestion of the M6 isoform. (D), 2D-PAGE of PIMZBristol plasma {alpha}1-antitrypsin and mass spectra of the tryptic in-gel digestion of spot 3, showing a peptide of m/z 3737.8, corresponding to amino acids 70–101 with an asparagine at position 83 and a methionine at position 85.

The state of glycosylation of spots 1a–7a was investigated by mass mapping studies. It has been shown previously (25) that glycopeptides containing complex glycans do not produce detectable ions under the conditions used here for mass spectrometry, whereas the same peptides in which the asparagine is not glycosylated or in which the asparagine has been converted to aspartic acid by enzymatic deglycosylation can be detected (Fig. 7BUp ). These observations provide a means for assessing the occupancy of a glycosylation site.

An extra mass at m/z 3691.8, which was not detected in the control M series of {alpha}1-antitrypsin, was detected in all of the tryptic peptide maps of spots 1a–5a (Fig. 7AUp ). The mass of m/z 3691.8 corresponds to the peptide containing amino acids 70–101 with an asparagine residue at amino acid 83, which indicates that this glycosylation site is not occupied (Fig. 7AUp ). Masses corresponding to peptides containing asparagine at the two other glycosylation sites, asparagines 46 and 247, were not detected, implying that these glycosylation sites were occupied. Thus the decrease in mass and increase in pI of spots 1a–5a are probably attributable to nonglycosylation of asparagine 83. This unglycosylated peptide (70–101) was also present in the tryptic peptide map of spots 6a and 7a, together with mass of m/z 1755.9, which is also absent from the control isoforms. This mass corresponds to a peptide covering the amino acids 244–259 with asparagine at position 247. These data indicate that the isoforms 6a and 7a are glycosylated only at asparagine 46 and that this accounts for their positions on 2D-PAGE.

altered amino acid sequence and posttranslational modification
The 2D-PAGE analysis of the plasma from the PIMZBristol heterozygote (Fig. 2CUp ) revealed the presence of four additional putative {alpha}1-antitrypsin isoforms that were not present in the control plasma (Fig. 2AUp ). All four of these additional spots were identified as {alpha}1-antitrypsin by mass mapping studies. Spots 1, 2, 3, and 4 had pI values of 5.07, 5.10, 5.16, and 5.27, respectively, and Mr ~2000 less than the control M series of {alpha}1-antitrypsin isoforms. These changes in pI and mass are consistent with the loss of a complex biantennary glycan, as was seen in some of the {alpha}1-antitrypsin isoforms in CDG-Ia. The substitution of a methionine for threonine at position 85 in PIMZBristol {alpha}1-antitrypsin leads to the loss of a sequon for N-linked glycosylation, i.e., NLT85E->NLM85E. The loss of this sequon would be expected to lead to isoforms of {alpha}1-antitrypsin that are not glycosylated at asparagine 83 and have only two of the three glycosylation sites occupied. The observed changes in pI and molecular weight suggest that the four novel isoforms contain the PIZBristol mutation. The presence of two sets of isoforms in the 2D-PAGE shows that both the wild-type (M) and variant alleles of {alpha}1-antitrypsin are expressed in the plasma of PIMZBristol heterozygotes and can be detected by this protocol.

To demonstrate that the novel isoforms 1–4 were attributable to the PIMZBristol allele, we subjected them to in-gel digestion with trypsin followed by mass spectrometry of the resulting peptide mixture. An additional mass of m/z 3737.8 was detected in each of the isoforms (Fig. 7DUp ), which was not present in the M series of {alpha}1-antitrypsin in both the PIMZBristol heterozygote and the control (Fig. 7CUp ). A peptide of m/z 3737.8 can be accounted for by a peptide containing amino acids 70–101 with methionine at position 85 and asparagine at position 83, i.e., it is derived from the PIMZBristol allele with a nonglycosylated asparagine 83. The lack of glycosylation at asparagine 83 permits in-gel tryptic digestion of this region, as was seen in the underglycosylated isoforms from CDG-Ia. A glycopeptide covering sequon 83 was not observed in any of the M series of {alpha}1-antitrypsin isoforms from control or mutant samples. This is attributable to the difficulty of analyzing complex sialic acid-containing glycopeptides with MALDI-TOF-MS. We obtained further evidence that a peptide containing amino acids 70–101 can be produced by tryptic digestion and detected by MALDI-TOF-MS if it is not glycosylated by N-glycanase to remove glycans before mass mapping studies. A peptide of mass m/z 3692.8 was detected in the mass spectrum after in-gel tryptic digestion of the main M4 isoform after prior enzymatic removal of the glycans with N-glycanase F (Fig. 7BUp ). This mass corresponds to a peptide covering amino acids 70–101 in the sequence, but with an aspartic acid residue at position 83. The mass of this peptide (m/z 3692.8) is one mass unit greater than the mass of the unglycosylated peptide (m/z 3691.8) that contains amino acids 70–101 found in {alpha}1-antitrypsin from CDG-Ia because of the conversion of an asparagine 83 to aspartic acid during the release of the glycan by the N-glycanase.

Further characterization of the novel isoforms 1, 2, 3, and 4 was obtained by analysis of their N-terminal peptides and glycans. The major mutant isoforms, spots 2 and 3 (Fig. 7DUp ), both had the same glycan profile as the control M4 and M6 isoforms (Fig. 2AUp ). The peptide analysis of spots 2 and 3 also contained masses corresponding to amino acids 1–20 in the amino acid sequence. Therefore, spots 2 and 3 are the equivalents of the major control M4 and M6 isoforms but lack glycosylation at asparagine 83. The combined glycan and peptide analysis of spot 4 showed that it was the unglycosylated equivalent of the control M8 isoform of {alpha}1-antitrypsin. It was not possible to carry out glycan analysis of spot 1, but its pI and mass suggest that it is the equivalent of the normal M2 isoform. Thus the mutant spots 1, 2, 3, and 4 observed in the PIZBristol plasma are the underglycosylated (at asparagine 83) equivalents of the M2, M4, M6, and M8 isoforms of {alpha}1-antitrypsin. The reason for the absence of an equivalent to the M7 {alpha}1-antitrypsin isoform in the PIMZBristol plasma is not clear. It could be that it is not secreted from the liver as in the PIZZ phenotype where decreased amounts of mutant {alpha}1-antitrypsin are observed in the plasma. However, only the mutant M7 isoform was absent from the plasma samples of the PIZBristol heterozygotes, whereas all isoforms were observed, albeit in decreased amounts, in the plasma of PIZZ patients.


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
We believe this work is a good example of the usefulness of proteomics as a tool in the detection and further understanding of genetic metabolic diseases. Proteomics is a natural development of the pioneering work on the 2D-PAGE separation of multiple forms of {alpha}1-antitrypsin (30)(31).

However, it is important to recognize the limitations of proteomic methodology. As reported previously (25), the use of two or more proteases may be required to obtain peptides that cover the complete amino acid sequence of a protein. Analysis of {alpha}1-antitrypsin by in-gel tryptic digestion allows only ~60% of the amino acid sequence to be covered. However, when we used chymotrypsin as well, it was possible for us to cover 86% of the sequence. All the mutations analyzed in this study were chosen deliberately to be in peptides produced by in-gel tryptic digestion.

This study shows that it is possible to detect known alterations in the amino acid sequence of a protein with the use of MALDI-TOF-MS to detect a peptide that is not present in the digestion of the normal protein. Reliable detection of unknown amino acid substitutions will require sequencing of all the peptides generated by proteolytic digestion of the protein. Although MALDI-TOF-MS has the capability of sequencing peptides by postsource decay analysis, sequencing can be carried out more efficiently and sensitively with a quadrupole time-of-flight mass spectrometer, such as the Q-TOF (MicroMass) or ion-trap mass spectrometer.

The data generated in this study were essentially qualitative. To derive the maximum information from proteomic analysis it will be important to be able to quantify the degree of expression of all the important proteins. Methods need to be developed to ensure that all peptides are sequenced and measured quantitatively for the reliable detection of mutations and measurement of the degree of expression of proteins in a metabolic pathway. Prior immunoprecipitation of less abundant proteins could be used to increase the sensitivity of this strategy.


   Acknowledgments
 
We would like to thank Drs. Kevin Rogers and Richard Tyldesley of MicroMass UK for help with the mass spectrometry and Dr. Ole Diettrich for advice and refinements to the 2D-PAGE protocol. We would also like to thank Dr. Ted Tarelli for help and advice in the analysis of glycans by MALDI-TOF-MS. The financial support of the Wellcome Trust, the Sir Jules Thorn Charitable Trust and the European Union (Euroglycan) are gratefully acknowledged. Research at the Institute of Child Health and Great Ormond Street Hospital for Sick Children NHS Trust benefits from R & D funding received from the NHS Executive.


   Footnotes
 
1 Nonstandard abbreviations: PI, protease inhibitor; CDG-Ia, congenital disorder of glycosylation type Ia; 2D-PAGE, two-dimensional polyacrylamide gel electrophoresis; and MALDI-TOF-MS, matrix-assisted laser desorption–ionization time-of-flight mass spectrometry.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Humphery-Smith I, Blackstock W. Proteome analysis: genomics via the output rather than the input code. J Protein Chem 1997;16:537-544.[Web of Science][Medline] [Order article via Infotrieve]
  2. Varki A. Biological roles of oligosaccharides: all of the theories are correct?. Glycobiology 1993;3:97-130.[Abstract/Free Full Text]
  3. Parekh R. Proteomics and molecular medicine. Nat Biotechnol 1999;17:19-20.[Web of Science][Medline] [Order article via Infotrieve]
  4. Cox DW. {alpha}1-Antitrypsin deficiency. 7th ed. Scriver CR Beaudet AC Sly WS Valle D eds. The metabolic and molecular bases of inherited diseases 1995;Vol. III:4125-4158 McGraw-Hill New York. .
  5. Brantly M, Nukiwa T, Crystal RG. Molecular basis of {alpha}1-antitrypsin. Am J Med 1988;84:13-31.[Web of Science][Medline] [Order article via Infotrieve]
  6. Eriksson S. {alpha}1-Antitrypsin deficiencies. Acta Med Scand 1964;175:197-205.[Web of Science][Medline] [Order article via Infotrieve]
  7. Sherlock S, Dooley J. Hepatic transplantation. Diseases of the liver and biliary system 9th ed. 1992:613-620 Blackwell Scientific Publications Oxford, United Kingdom. .
  8. Travis J, Salvesen GS. Human plasma proteinase inhibitors. Ann Rev Biochem 1983;52:655-709.[Web of Science][Medline] [Order article via Infotrieve]
  9. Kim D, Yu MH. Folding pathway of human {alpha}1-antitrypsin. Biochem Biophys Res Commun 1996;226:378-384.[Web of Science][Medline] [Order article via Infotrieve]
  10. Bathurst IC, Travis J, George PM, Carrell RW. Structural and functional characterisation of the abnormal Z {alpha}1-antitrypsin isolated from human liver. FEBS Lett 1984;177:179-183.[Web of Science][Medline] [Order article via Infotrieve]
  11. Foreman RC. Disruption of the Lys290-Glu342 salt bridge in human {alpha}1-antitrypsin isolated from human liver does not prevent its synthesis or secretion. FEBS Lett 1987;216:79-82.[Web of Science][Medline] [Order article via Infotrieve]
  12. Bottomley SP, Chang WS. The effect of the reactive centre loop length upon serpin polymerisation. Biochem Biophys Res Commun 1997;241:264-269.[Web of Science][Medline] [Order article via Infotrieve]
  13. Vaughan L, Lorier M, Carrell RW. {alpha}1-Antitrypsin microheterogeneity: isolation and the physiological significance of isoforms. Biochim Biophys Acta 1982;701:339-345.[Medline] [Order article via Infotrieve]
  14. Fagerhol MK, Laurell CB. The polymorphism of ‘prealbumins’, {alpha}1-antitrypsin in human sera. Clin Chim Acta 1967;16:199-203.[Web of Science][Medline] [Order article via Infotrieve]
  15. Jeppsson JO, Lilja H, Johansson M. Isolation and characterization of 2 minor fractions of {alpha}1-antitrypsin by high performance liquid chromatography chromatofocusing. J Chromatogr 1985;327:173-177.[Web of Science][Medline] [Order article via Infotrieve]
  16. Keir G, Winchester BG, Clayton PT. Carbohydrate-deficient glycoprotein syndromes: inborn errors of protein glycosylation. Ann Clin Biochem 1999;36:20-36.
  17. Lovegrove J, Jeremiah D, Gillet GT, Temple IK, Povey S, Whitehouse DB. A new {alpha}1-antitrypsin mutation, THR-MET85 (PIZBristol), associated with novel electrophoretic properties. Ann Hum Genet 1997;61:385-391.[Web of Science][Medline] [Order article via Infotrieve]
  18. Hughes GJ, Frutiger S, Paquet N, Ravier F, Pasquali C, Sanchez JC, et al. Plasma protein map–an update by microsequencing. Electrophoresis 1992;13:707-714.[Web of Science][Medline] [Order article via Infotrieve]
  19. Bjellqvist B, Pasquali C, Ravier F, Sanchez J-C, Hochstrasser DF. A nonlinear wide range immobilized pH gradient for 2D-PAGE and its definition in a relevant pH scale. Electrophoresis 1993;14:1357-1365.[Web of Science][Medline] [Order article via Infotrieve]
  20. Hochstrasser DF, Patchornik A, Merril CR. Development of polyacrylamide gels that improve the separation of proteins and their detection by silver staining. Anal Biochem 1988;173:412-423.[Web of Science][Medline] [Order article via Infotrieve]
  21. Hochstrasser DF, Merril CR. Catalysts’ for polyacrylamide gel polymerization and detection of proteins by silver staining. Appl Theor Electrophor 1988;1:35-40.[Medline] [Order article via Infotrieve]
  22. Hochstrasser DF, Harrington MG, Hochstrasser AC, Miller MJ, Merril CR. Methods for increasing the resolution of two-dimensional protein electrophoresis. Anal Biochem 1988;173:424-435.[Web of Science][Medline] [Order article via Infotrieve]
  23. Diettrich O, Mills K, Johnson AW, Hasilik A, Winchester BG. Application of magnetic chromatography to the isolation of lysosomes from fibroblasts of patients with lysosomal storage disorders. FEBS Lett 1998;441:369-372.[Web of Science][Medline] [Order article via Infotrieve]
  24. Shevchenko A, Wilm M, Vorm O, Mann M. Mass spectrometric sequencing of proteins from silver stained polyacrylamide gels. Anal Chem 1996;68:850-858.[Medline] [Order article via Infotrieve]
  25. Mills K, Johnson AW, Clayton PT, Diettrich OGP, Winchester BG. A strategy for the identification of site specific glycosylation in glycoproteins using MALDI TOF MS. Tetrahedron Asymmetry 2000;11:75-93.
  26. Kuster B, Hunter AP, Wheeler SF, Dwek RA, Harvey DJ. Structural determination of N-linked carbohydrates by matrix-assisted laser desorption/ionization-mass spectrometry following enzymatic release within sodium dodecyl sulphate-polyacrylamide electrophoresis gels: application to species-specific glycosylation of {alpha}1-acid glycoprotein. Electrophoresis 1998;19:1950-1959.[Web of Science][Medline] [Order article via Infotrieve]
  27. Papac DI, Wong A, Jones AJ. Analysis of acidic oligosaccharides and glycopeptides by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Anal Chem 1996;15:3215-3223.
  28. Naitoh A, Aoyagi Y, Asakura H. Highly enhanced fucosylation of serum glycoproteins in patients with hepatocellular carcinoma. J Gastroenterol Hepatol 1999;14:436-445.[Web of Science][Medline] [Order article via Infotrieve]
  29. Sveger T. Plasma protease inhibitors in {alpha}1-antitrypsin deficient children. Pediatr Res 1985;19:834-835.[Web of Science][Medline] [Order article via Infotrieve]
  30. O’Farrell PH. High resolution two-dimensional electrophoresis of proteins. J Biol Chem 1974;250:4007-4021.[Abstract/Free Full Text]
  31. Anderson L, Anderson NG. High resolution two-dimensional electrophoresis of human plasma proteins. Proc Natl Acad Sci U S A 1977;74:5421-5425.[Abstract/Free Full Text]



The following articles in journals at HighWire Press have cited this article:


Home page
 J. Biol. Chem.Home page
B. B. Cheung, J. Bell, A. Raif, A. Bohlken, J. Yan, B. Roediger, A. Poljak, S. Smith, M. Lee, W. D. Thomas, et al.
The Estrogen-responsive B Box Protein Is a Novel Regulator of the Retinoid Signal
J. Biol. Chem., June 30, 2006; 281(26): 18246 - 18256.
[Abstract] [Full Text] [PDF]

Home page
 GlycobiologyHome page
Y. Wang, S.-l. Wu, and W. S. Hancock
Approaches to the study of N-linked glycoproteins in human plasma using lectin affinity chromatography and nano-HPLC coupled to electrospray linear ion trap--Fourier transform mass spectrometry
Glycobiology, June 1, 2006; 16(6): 514 - 523.
[Abstract] [Full Text] [PDF]

Home page
 Genome ResHome page
B. Thomsen, P. Horn, F. Panitz, E. Bendixen, A. H. Petersen, L.-E. Holm, V. H. Nielsen, J. S. Agerholm, J. Arnbjerg, and C. Bendixen
A missense mutation in the bovine SLC35A3 gene, encoding a UDP-N-acetylglucosamine transporter, causes complex vertebral malformation
Genome Res., January 1, 2006; 16(1): 97 - 105.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Chem.Home page
L. Valmu, A. Paju, M. Lempinen, E. Kemppainen, and U.-H. Stenman
Application of Proteomic Technology in Identifying Pancreatic Secretory Trypsin Inhibitor Variants in Urine of Patients with Pancreatitis
Clin. Chem., January 1, 2006; 52(1): 73 - 81.
[Abstract] [Full Text] [PDF]

Home page
 Clin. Chem.Home page
E. Ballot, A. Bruneel, V. Labas, and C. Johanet
Identification of Rat Targets of Anti-Soluble Liver Antigen Autoantibodies by Serologic Proteome Analysis
Clin. Chem., April 1, 2003; 49(4): 634 - 643.
[Abstract] [Full Text] [PDF]

Home page
 GlycobiologyHome page
K. Mills, P. B. Mills, P. T. Clayton, N. Mian, A. W. Johnson, and B. G. Winchester
The underglycosylation of plasma {alpha}1-antitrypsin in congenital disorders of glycosylation type I is not random
Glycobiology, February 1, 2003; 13(2): 73 - 85.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (21)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Mills, K.
Right arrow Articles by Winchester, B. G.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Mills, K.
Right arrow Articles by Winchester, B. G.
Related Collections
Right arrow Molecular Diagnostics and Genetics
Right arrow Proteomics and Protein Markers
Right arrow Automation and Analytical Techniques

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS