'3'.II-;.=.....-...~.:
`
`
`
`
`
`
`
`;
`:
`
`1 ANNUAI.
`PEVIEW 1
`OF
`a{‘HE:MIsT1«:1'
`
`,_o,u|-H-,q..-u-s.sb-r*hH.II|I-q.InI~1*-r'l‘-s-I'1-I‘'
`
`
`
`_‘._‘._.,‘._._,....-':'-.-.....I.:.'|‘---'.!n.I-‘.''"-
`
`
`
`SNELL
`430153
`MEISTER
`W’
`
`g.
`
`-1-
`
`.
`
`'1
`
`I
`
`u\
`
`-l
`
`I.
`
`.1
`.5
`1..
`
`11
`
`
`
`-y..~a-I-.q--‘I-11>!-||'p'A-ll-||""""
`
`.1n-.I--
`
`Page 1 of 28
`
`ILMN EXHIBIT 1019
`
`
`
`Page 1 of 28
`
`ILMN EXHIBIT 1019
`
`

`
`Page 2 of 28
`
`
`
`Page 2 of 28
`
`

`
`IE!
`
`ANNUAL REVIEW OF
`
`BIOCHEMISTRY
`
`ESMOND E. SNELL, Editor
`University of Texas at Austin
`
`PAUL D. BOYER, Associate Editor
`University of California, Los Angeles
`
`ALTON MEISTER, Associate Editor
`Cornell University Medical College
`
`CHARLES C. RICHARDSON, Associate Editor
`Harvard Medical School
`
`I2 - I
`
`2;: 0432
`‘L5A901‘
`
`VOLUME 50
`1981
`
`ANNUAL REVIEWS INC.
`
`4139 EL CAMINO WAY PALO ALTO. CALIFORNIA 94306 USA
`
`Page 3 of 28
`
`
`
`Page 3 of 28
`
`

`
`\
`
`ANNUAL REVIEWS INC.
`
`Palo Alto, California, USA
`
`corrvruarrr © 1981 av ANNUAL REVIEWS INC., PALO ALTO. cauronwtn. USA. ALL
`
`rum-rrs RESERVED. The appearance of the code at the bottom of the first page of
`an article in this serial indicates the copyright owner's consent that copies of the
`article may be made for personal or internal use, or for the personal or internal use
`of specific clients. This consent is given on the condition, however, that the copier
`pay the stated per-copy fee of $1.00 per article through the Copyright Clearance
`Center, Inc. (21 Congress Street, Salem, MA 01970) for copying beyond that permit-
`ted by Sections 107 or 108 of the US Copyright Law. The per-copy fee of $1.00 per
`article also applies to the copying, under the stated conditions, of articles published
`in any Annual Review serial before January 1, 1978. Individual readers. and non-
`profit libraries acting for them. are permitted to make a single copy of an article
`without charge for use in research or teaching. This consent does not extend to other
`kinds of copying. such as copying for general distribution, for advertising or promo-
`tional purposes, for creating new collective works, or for resale. For such uses,
`written permission is required. Write to Permissions Dept, Annual Reviews Inc.,
`4139 El Camino Way, Palo Alto, CA 94306 USA.
`
`The conspicuous number aligned in the margin with the title of each
`n.EPiuN'rs
`article in this volume is a key for use in ordering reprints. Available reprints are
`priced at the uniform rate of $2.00 each postpaid. The minimum acceptable reprint
`order is 5 reprints and/or $10.00 prepaid. A quantity discount is available.
`
`International Standard Serial Number: 0066-4154
`International Standard Book Number: 0-8243-0850-6
`
`Library of Congress‘ Catalog Card Number: 32-25093
`
`Annual Reviews Inc. and the Editors of its publications assume no responsibility
`for the statements expressed by the contributors to this Review.
`
`PRINTED AND BOUND [N ‘HIE UNITED STATES OF AMERICA
`
`Page 4 of 28
`
`
`
`Page 4 of 28
`
`

`
`Ann. Rev. Biochem. 1981. J'0.'26!-84
`Copyright © 193! by Annual Review: Inc. All rig}!-‘.1 reserved
`
`ADVANCES IN PROTEIN
`
`o12o79
`
`SEQUENCING
`
`Kenneth A. Walsh, Lowell H. Ericsson, David C Parmelee,
`and Koiti Tfrom‘
`a
`
`Department of Biochemistry, University of Washington,
`Seattle, Washington 98195
`
`CONTENTS
`
`264
`264
`266
`266
`
`270
`270
`272
`273
`274
`2'75
`276
`277
`
`280
`
`SUMMARY ........................................................................
`_
`‘_
`
`....................................... ..
`
`PERSPECTIVES
`
`Our present understanding of the detailed relationship of the structure and
`function of proteins has grown from discoveries in the nineteen fifties that
`these macromolecules have unique amino acid sequences that lit unique
`three-dimensional structures. Since that time, technological advances have
`facilitated the determination of the details of the amino acid sequences of
`a wide variety of proteins, and established: (a) that their complex biological
`functions can be described in terms of simple chemical concepts; (b) that
`
`0066-4154/81/0701-026l$0l.00
`
`261
`
`Page 5 of 28
`
`TACTICAL PROGRESS ..............................................................................................
`Cleavage Techniques
`.............................................................................................. ..
`Sepamtiorl Techniques ..............................................................................................
`General method:
`.................................................................................................. ..
`SDS-gel .wparu!ian.;
`.....
`............
`.......................................................
`High
`rforrnance '
`in‘ c rvnurtogra
`......................................................
`"E.'dnwnpeDeg1udarioriq:"‘ ......................
`....................................................................
`Chemistry ofdie degradation ...... ..
`Modification of the instrumentation
`Product identification
`Hiauuuuendng ..................
`Other Degradation .5'clIemes ............ ..
`Problems of Covalently Modified Resirlues
`STRATEGIES ........................................................................... ..
`
`
`
`
`
`..
`
`
`
`Page 5 of 28
`
`

`
`262
`
`WALSH ET AL
`
`A many proteins comprise domains and subunits that specify discrete sub-
`functions; (c) that proteins can be grouped into homologous, evolutionarily
`related families; and (d) that the structure of a protein can be directly
`related to its encoding gene.
`In 1976, Edman (1) estimated that the data base of established sequences
`consisted of 80,000 amino acid residues. Three years later, this number had
`doubled, and about l60,(IJO residues had been placed in about 1100 com-
`pleted protein sequences [W. C. Barker, personal communication from an
`analysis of the current data base of the Atlas of Protein Sequence and
`Structure (2)]. Ongoing research is providing a continuous stream of data, ‘
`including sequences of proteins of larger molecular weight and analyses of
`increasing sensitivity and reliability.
`Despite a recent view that "the decline and fall of protein chemistry” is
`at hand (3), it appears that detailed analyses of protein sequences will
`continue to provide a chemical base for understanding proteins and their
`role in biology. While it is true that recent advances in DNA sequencing
`technology provide more rapid methods, even analysis of cDNA (which
`the problem of untranslated information) fails to identify the
`sites of disulfide bonds, of attachment of prosthetic groups, or of removal
`of peptide segments, any of which may influence expression of function.
`Procedures to obtain pure nucleic acid starting material are rapidly improv-
`ing, and one can confidently predict that rean1s of new sequence data will
`be translated from these analyses. In fact, DNA sequence technology will
`probably provide the simplest and fastest source of raw polypeptide back-
`bone structure. Nevertheless, it is vital that sequences inferred in this way
`be correlated with those of the mature protein products, to avoid frame shift
`errors, to correct for intron excisions, and to identify the myriad of process-
`ing events (4) that modulate both structure and function. Ideally, one would
`like to describe the complete process of maturation of the protein from the
`control of gene expression through the control of its ultimate function, and
`this requires a cooperative blend of studies by molecular biologists, protein
`chemists, and X—ray crystallographers. Such cooperative descriptions are
`beginning to erge, but there are veryifew examples to date where a
`combination of nucleic acid analysis, amino acid sequence analysis, and
`X-ray crystallography have focused on a single protein.
`The increasing rate of accumulation of amino acid sequence data has
`resulted less from a proportional increase in the number of analysts, than
`from a striking increase in efiiciency (Figure 1). Whereas in the fifties and
`sixties a small research group might place, on the average, only a single
`amino acid residue per week, by analyzing many small overlapping peptides
`isolated from a large protein, the availability of automated sequential de-
`gradative technology in the seventies has changed the strategy (5-9) and
`
`Page 6 of 28
`
`
`
`Page 6 of 28
`
`

`
`PROTEIN SEQUENCING
`
`263
`
`increased the efliciency by perhaps two orders of magnitude. Nevertheless,
`more than 95% of completed sequence analyses describe proteins smaller
`than 40,000 daltons (Figure 2), and the efficiency of their analysis falls off
`markedly with increasing chain length (Figure 1). Larger proteins require
`more sophisticated strategies than smaller ones, and call for the develop»
`ment of yet more efficient and more sensitive tactics.
`In this review we focus on recent advances in tactics that have led to
`improvements in strategy requiring smaller amounts of proteins and being
`applicable to proteins of larger molecular weight. General reviews of meth-
`odology (10-13) and specific reviews of technology (14-16) have been pub-
`lished during the last five years. Shorter discussions of strategy have also
`
`00
`
`5°
`
`40
`
`so
`
`20
`
`to
`
`"
`
`
`in
`
`0
`
`Automated
`-<12!» roslduoa
`c125-280
`D:-230 residues
`o
`
`lI1IflLlII=I
`
`I
`
`.
`
`.0
`
`° 0
`
`.
`
`O
`
`I
`
`0
`
`to:
`
`o 6
`.9-
`
`-
`
`I‘I
`
`afi
`
`.__————3-an--u"""'l'_7"-1-I
` +:
`-\°
`‘9
`..¢"*.é=” 4*’ er“ .9
`is
`.9‘
`Your Publlnhod
`
`Progress in the efliciency of amino acid sequence analysis. The efliciency is calcu-
`Figure 1
`lated as the number of residues finally placed in sequence divided by the number of fragments
`isolated to prove that sequence. Of the 1100 proteins sequenced by 1978 (Figure 2), only 45
`were randomly chosen for this illustration. Those analyzed by manual degradation procedures
`(1) are distinguished from the three symbols designating automated Edman degradative proce-
`dures on proteins of three arbitrarily dilferent size classes. Sequences completed by automated
`techniques were lirst reported in 1971. The diagram demonstrates both the increased efliciency
`of automated analyses and a generally inverse relationship between chain length and eificiency.
`
`Page 7 of 28
`
`
`
`Page 7 of 28
`
`

`
`264
`
`WALSH ET AL
`
`_80(}
`
`600
`
`200
`
`Number
`or
`Proteins
`sequenced
`1950-1978
`
`9
`
`:-
`
`'V:°‘fl-FuP‘§&‘3§§.p‘_’§$
`
`Mal. an at 10 3
`
`Figure 2 Molecular weight distribution of proteins that have been sequenced.
`
`appeared (17-18). Periodical discussion of recent methods appear bienially
`in the proceedings of a symposium series (e. g. 19) and more frequently in
`a manufacturer's news service (20).
`
`TACTICADPROGRESS
`
`From a technological point of view, a range of procedures are now available
`for cleaving proteins into appropriate fragments, for separating these frag-
`ments from each other, and for degrading the pure fragments to pheny1thi—
`ohydantoins, which are identified by increasingly sensitive techniques.
`Recent improvements in each of these areas are continuing to provide
`broader strategies, more miniaturized separation systems, and more sensi-
`tive degradative analyses.
`
`Cleavage Techniques
`
`Table 1 is a partial list of specific chemical and enzymatic cleavage tech-
`niques that have been eifective in sequence analysis. The list does not
`
`Page 8 of 28
`
`
`
`Page 8 of 28
`
`

`
`PROTEIN SEQUENCING
`
`265
`
`include the classical enzymatic and partial acid cleavage techniques that
`were at one time the only known cleavage methods. Although these are still
`used as primary tools in some analyses, particularly in conjunction with
`dansyll -Edman procedures, the more restrictive cleavage techniques listed
`are more productive in concert with automated Edman degradations. In
`principle, it is more efficient to separate and analyze a small number of large
`fragments than a larger number of smaller ones. Three of the more par-
`simonious methods are finding wider usage, namely limited proteolysis of
`native proteins (17) and chemical cleavage of Asn-Gly or Asp-Pro bonds
`(12). The rarity of these dipeptides augments their value in subdividing a
`protein into large segments. Limited proteolysis of a native protein, al-
`though unpredictable in its site of attack, often separates domains at their
`“hinges” or nicks the protein on its “fringes" (31).
`Among the chemical cleavage agents, iodosobenzoate shows promise for
`cleaving tryptophanyl bonds, providing p-cresol is present to minimize a
`side reaction with tyrosine (27). New enzymes have expanded the cleavage
`repertoire by severing bonds at the amino side of lysyl residues (28, 29) and
`at the carboxyl side of glutamyl residues (23). Other reports describe sub»
`strate modifications that redirect tryptic cleavage to aspartyl residues (32,
`32a) and the Armiilaria protease to the amino side of cystcinyl residues
`(33).
`
`Table 1 Summary of useful specific cleavage techniques“
`
`Bond cleaved
`
`Cleavage agent
`
`Met-X
`Asn—G1y
`Asp-Pro
`Arg-X
`
`Lys-X
`
`Glu-X
`Various
`X-Cys
`
`Trp—X
`X-Lys
`
`Cyanogen bromide
`Hydroxylamine
`pH 2.5
`Trypsin (N-acylated substrate). Clostripain, or mouse
`submaxillary protease (21)
`Trypsin [block arginine with cyclohexanedione or
`with malonic dialdehyde (22)|
`Staphylococcal protease (23)
`Limited proteolysis of native substrate (17)
`2-nitro-5-thiocyanobenzoic acid (24, 24a) followed by
`Raney nickel (25)
`Skatole derivative (26) or iodosobenzoate (27)
`Myxobacter (28) or Annillarfa (29) proteases
`
`“General references (12. 30) describe techniques that are not specifically
`cited.
`-
`
`‘Dansyl, 5-dimethylamino l-naphthalenesulfonyl.
`
`Page 9 of 28
`
`
`
`Page 9 of 28
`
`

`
`266
`
`WALSH ET AL
`
`Separation Techniques
`
`GENERAL METHODS Isolation of pure fragments of suitable size has
`been a major operational problem in sequence analysis. Classical methods
`such as paper chromatography, paper electrophoresis, and Dowex column
`chromatography are still applied to mixtures of small peptides in many
`studies, with some modifications. The former two techniques have been
`replaced by thin-layer plate techniques in some cases (34, 343), and the
`latter by improved resins or ion exchange cellulose in microbore columns
`with automated apparati (35, 36). However, these methods are not generally
`applicable to mixtures of large, denatured fragments, which tend to aggre-
`gate and precipitate. By careful choice of denaturing solvents, such frag-
`ments have been separated by gel permeation chromatography and gel
`electrophoresis, and more recently by high performance liquid chromatog-
`raphy.
`Gel filtration has been applied to large fragments in denaturing solvents
`such as high contxntrations of formic, acetic, or propionic acid for basic and
`neutral fragments, and ammonium hydroxide or ammonium bicarbonate
`for acidic fragments. Certain chemical modifications, such as succinylation
`(37) or citraconylation (38) of lysine residues, and esterification (39) or
`amidation (40) of acidic residues, result in striking changes in the solubility
`of fragments and permit the separation of large fragments without using
`urea or detergents. Hydrophobic fragments have been separated on Se-
`phadex LI-l columns with organic solvents (41) or by partitioning between
`butanol and dilute acid (42). Urea, guanidine hydrochloride. or detergents
`provide alternative solvents (17), but urea can lead to carbamylation of
`amino groups at high pH, guanidine hydrochloride and ionic detergents
`preclude ion exchange chromatography, and other detergents may be so
`strongly bound to apolar fragments as to perturb further separation. Re-
`cently, severalmethods have been reported for removal of SDS from pro-
`teins or fragments (43, 44).
`Polystyrene-based ion exchange columns (e.g. Dowex columns) provide
`high resolution of small peptides but are not applicable to large fragments
`because of strong hydrophobic interactions and the small pore size of the
`resins. Various grades of hydrophilic cation or anion exchange columns are
`more useful for large fragments (35), provided that urea is present to
`solubilize the fragments. Examples of high resolution in urea solutions
`include the separation of ribosomal proteins on phosphocellulose or DEAF.-
`cellulose (45), of B-galactosidase fragments on CM-cellulose (42, 46), and
`of phosphorylase fragments on SP-Sephadex (47).
`Selective separation techniqus have been described for methionine-, cys-
`teine-, and tryptophan-containing peptides on special matrices (48-50).
`
`Page 10 of 28
`
`
`
`Page 10 of 28
`
`

`
`PROTEIN SEQUENCING
`
`267
`
`Novel de facto separations can be achieved by selective attachment of
`fragmts through C-terminal homoserine or lysine residues in preparation
`for solid phase sequencing (S1).
`The ninhydrin method has been routinely applied to detect small pep-
`tides, whereas absorbance measurements at 200-280 nm are suitable for
`
`large peptides. Fluorescamine- or o-phthalaldehyde—based methods have
`been developed for more sensitive detection of peptides (52, 53).
`
`SDS-GEL SEPARATIONS Polyacrylamide gel electrophoresis (PAGE) is
`a powerful analytical tool for assessing the purity of proteins and large
`peptides before subjecting them to sequence analysis. The usefulness of this
`technique is due largely to its high resolving capability. For example. a
`two-dimensional electrophoresis system has resolved 1100 different compo-
`nents from Escherichia call‘ and has the potential of separating 5000 pro-
`teins (54). This procedure utilizes isoelectric focusing in the first dimension
`and SDS electrophoresis in the second. The results of isoelectric focusing
`may be evaluated prior to the second electrophoresis by fixing and staining
`with Coomassie brilliant blue (55). This procedure does not alter the protein
`pattern, but does cause sharper spots. The order of iso-electric focusing and
`SDS-electrophoresis may be reversed (S6) to yield material free from SDS.
`Such two-dimensional systems have been used analytically and prepara-
`tively on complex mixtures of proteins or peptides (57).
`Diflicult separations have been achieved by varying the experimental
`conditions. For example, the use of SDS-gradient gels selectively improves
`resolution and provides reliable molecular weight determinations (58); sub-
`stitution of SDS by a cationic detergent, cetyltrimethyl ammonium bro-
`mide, yields better data with certain proteins (59).
`As microsequencing techniques become more extensively used, sensitive
`analytical gels are needed to avoid waste. Recently, a cupric-silver stain has
`been reported (60) to be 100 times more sensitive than the conventional
`Coomassie blue stain (e.g. it detects 0.33 ng/mm’ of serum albumin). These
`results were comparable to an autoradiogram of "C-methylated proteins
`'following a five-day exposure. Another rapid staining procedure, ferrous
`bathophenanthroline sulfonate (61), has a detection limit of l p.g for bovine
`serum albumin. The sensitivity of fluorescent labeling techniques allows
`continuous monitoring during electrophoresis. For example, a combination
`of microgel electrophoresis and fluorescent labeling with dansyl chloride
`produced distinct bands with 50 pmol of material (62). The fluorogenic
`reagent, 2-methoxy-2,4-diphenyl-3(2H)-furanone detected as little as 10 ng
`of peptide in a CNBr digest (63), whereas iluorescein-labeled lectin probes
`detected less than 100 ng of hexose in carbohydrate-containing proteins
`(64)-
`
`Page 11 of 28
`
`
`
`Page 11 of 28
`
`

`
`268
`
`WALSH ET AL
`
`SDS-polyacrylamide gel electrophoresis probably will be used more rou-
`tinely to purify suflicient quantities of material for microsequence analysis.
`A small portion of the sample _can be dansylated as a marker of the un-
`derivatized material (65), but with caution, because the mobility of low-
`molecular-weight peptides may be altered by the label. Alternatively,
`phosphorescence (66), alkylbenzyldimethyl-ammonium chloride (67), or
`precipitation with sodium acetate (68) may be employed to locate the
`components of interest. After the materials have been separated they can
`be electrophoretically eluted, with recoveries of more than 90% (69). If
`necessary, residual SDS can be quantitatively removed by a micro method
`utilizing ion-pair extraction with triethylamrnoniutn cations (44). The re-
`sulting small amounts of proteins or peptides can then be subjected to amino
`acid microanalysis using o-phthalaldehyde (70). At
`low protein-to-gel
`ratios it may be necessary to apply a correction for possible background
`contamination (71). The ultimate goal of the various separation techniques
`has been achieved by subjecting the eluted materials (100 pmol) to microse-
`quence analysis (e.g. T2, 73).
`
`HIGH PERFORMANCE LIQUID CHROMATOGRAPHY (I-IPLC) This
`technique is potentially the most powerful procedure currently available for
`both analytical and preparative isolation of peptides and proteins. The most
`eflicient column materials are silica-based microporous particles (5-10 pm)
`that have coatings covalently linked to their surface silanol moieties (74-
`76). Several commercial I-IPLC pacl-tings are currently available that have
`been utilized for molecular exclusion (77) or high-capacity ion exchange
`chromatography (78). These materials have only recently been introduced
`and subsequent improvements will, it is hoped, result in products superior
`to the classical alternatives. The microporous particles with bonded nonpo-
`la: moieties (74-76) have been responsible for the growing interest in HPLC
`to fractionate peptides and proteins. These column packings are used in
`“reversed-phase" chromatography in which the eluent is more polar than
`the stationary phase. This is reversed from the conventionally more polar
`stationary packing material.
`Reversed-phase HPLC has advantages of speed, sensitivity (79-82), and
`high resolving power (83, 84) when compared with other separation tech-
`niques. These features are only beginning to be utilized in primary struc- '
`tural studies. Although this application is far from routine, and experiments
`must be closely monitored to ensure that all components applied to the
`various ‘columns are actually eluted,
`the practicality of reversed-phase
`HPLC for separation and recovery of biologically active peptides and pro-
`teins has been confirmed by many investigators. For example, one species
`T of human leukocyte interferon (mol wt 18,000) has been purified to
`
`Page 12 of 28
`
`
`
`Page 12 of 28
`
`

`
`PROTEIN SEQUENCING
`
`269
`
`homogeneity (85). Thirty-two hormonal polypeptides and six proteins,
`ranging from live to 584 amino acid residues in length, have been success-
`fully chrornatographed (86). However, lactalburnin, elastase, and ovalbu—
`min were apparently irreversibly bound to octadecyl silica (86). In many
`cases components fail to elute because of the use of inappropriate buffers
`and organic eluents.
`Several bulfer systems are routinely used. These include 0.1% phosphoric
`acid (87), 0.1% trifluoroacetic acid (88), pyridine (1 M) -— acetic acid (0.5
`M) (89), and 0.25 N phosphoric acid titrated to pH 3-3.5 with triethylamine
`(90). Other ion-pairing reagents, e.g. alkylsulfonates (80) or alkylam-
`monium salts (91), have also been added to these buflers to improve the
`resolution of various peptides (92). Water-miscible organic solvents, such
`as methanol, rt -propanol, or acetonitrile are common eluent components for
`peptides and proteins (83, 84, 89, 90). The most frequently used reversed-
`phase supports are cyanopropyl-, octadecyl-, octyl-, and phenyl-silica (74.
`83, 84, 90). Recommendations have been made by several investigators (41,
`33, 84, 90) for optimizing the various experimental conditions, such as flow
`rate, ionic strength, pH, temperature, capacity, gradient elution, and sample
`preparation.
`Many have utilized reversed-phase HPLC to analyze small peptides pro-
`duced by enzymatic digests of various proteins (41, 81, 82, 89). This is
`essentially an analytical peptide mapping technique that provides the means
`to characterize subnanomol quantities, to observe the extent of digestion,
`and to prepare sufiicient material for further structural studies. For exam-
`ple, the total amino acid composition of bovine intestinal calcium-binding
`protein (75 residues) was accounted for on the basis of two tryptic maps that
`used a total of 4 ptg of protein (81). Differences in the HPLC peptide maps
`of the tryptic digests of normal human hemoglobins and 25 variants were
`used to purify the specific abnormal peptides. Amino acid analysis of those
`peptides established the nature of the alterations (93). Many of these variant
`peptides showed diflerences in elution position that correlated well with that
`expected from the changed amino acid compositions. Methods are now
`available for the prediction of the relative retention times of small peptides
`(< 20 residues) solely on the basis of their amino acid compositions (94).
`However, the absolute elation times vary with the chromatographic condi-
`tions.
`The ultimate goal in using HPLC for primary structural studies is to
`rapidly purify relatively large, denatured peptides to homogeneity for se-
`quence analysis or subdigestion and refraciionation. For example, two
`I chymotryptic fragments (19,000 and 6,900 daltons) of bacteriorhodopsin
`have been separated (41). These were further cleaved with CNBr, and two
`of the resulting peptides were sequenced after reversed-phase HPLC purifi-
`cation. Similarly, all three CNBr fragments (32, 44, and 66 residues) from
`
`Page 13 of 28
`
`
`
`Page 13 of 28
`
`

`
`270
`
`WALSH ET AL
`
`the S-aminoethylated at-chain of human hemoglobin were recovered in
`yieids of 59, 70, and 102%. respectively (82). Two of these were sequenced
`using 4-N,N—dimethylaminoazobenzene 4'-isothiocyanate (DABITC) with
`solid phase and liquid phase instruments, respectively (95, 96). Another
`investigation demonstrated that all but two of the CNBr peptides from
`human hemoglobin e.- , 8- , and 7-chains (13-91 residues in length) could
`be purified and recovered in yields of 71-95% (88). These various peptides
`were not sequenced, but their amino acid compositions agreed well with
`expected values.
`These reports indicate that reversed-phase HPLC is an extremely power-
`ful analytical and preparative alternative to more conventional techniques.
`The excellent resolution and high capacity may provide single-step purifica-
`tion of many peptides needed in primary structural studies. Unresolved
`fragments may be collected and rechromatographed on other reversed-
`phase columns or with dilferent eluents. At present, the most useful I-IPLC
`separations utilize reversed-phase column packings. As improved alterna-
`tive supports become available, analogous puiifications may well be made
`in conjunction with molecular exclusion HPLC or ion-exchange I-IPLC.
`
`Edman Degradation
`
`The principal chemical method to determine the amino acid sequences of
`fragments of a protein is the stepwise degradation initiated with phenyliso-
`thiocyanate and completed with“ acid (1), as devised by Edman three
`decades ago. This “Edman degradation" has been used in a “subtractive"
`mode whereby the composition of the shortened peptide establishes the
`missing amino terminus (97), in the dansyl-Edman mode whereby the new
`amino-terminal _residue is identified by end-group analysis (98), and in
`direct modes whereby the phenylthiohydantoin (PTH) of the released
`amino acid is identified directly by various chromatographic techniques or
`as the free amino acid after hydrolysis (99). A direct manual method was
`well described in 197'! (100). Other direct methods, as automated in spin-
`ning cups in 1967 (5) and on solid matrices in 1971 (9) oifer controlled
`reaction conditions and reproducibility that permit extended degradation of
`fragments. Improvements in the spinning cup methodology (16) have fo-
`cused recently on the chemistry, the instrumentation (101-193), and tech-
`niques of product identification. Together, these advances have led in 13
`years to increases by 3 to 6 orders of magnitude in sensitivity, which has
`now approached 10 pmol in one laboratory (103) and 0.1 pmol in special
`cases (see section on microsequencing).
`
`CHEMISTRY OF 'I'HE. DEGRADATION To achieve degradation yields of
`94-98%, side reactions must be minimized by paying close attention to the
`
`Page 14 of 28
`
`
`
`Page 14 of 28
`
`

`
`PROTEIN SEQUENCING
`
`271
`
`purity of the sample, reagents, solvents and gaseous atmosphere. and to the
`cleanliness of the surrounding mechanical structures.
`The nature and the purity of the subject peptide must be carefully consid-
`ered. Prior to sequce analysis, impurities and salts can be removed by gel
`filtration, dialysis, or electrophoresis (104), but contaminating SDS niay
`actually be advantageous during subsequent analysis (105). Blocking groups
`at the amino terminus are particularly troublesome. Enzymatic methods to
`remove amino-terminal pyrollidone carboxylic acid (106) have met with
`varying success (I07). Formyl groups have been removed from amino-
`terminal methionine (108). Although no method is known to remove the
`acetyl groups that block the amino termini of many proteins, a technique
`is described (109) to prevent its introduction during in vitro syntheses.
`Other covalently modified residues (4) present difficulties during both
`degradation and product identification. For example, glycosyl groups com-
`plicate estimations of molecular weight, isolation of peptides, and the Ed-
`man degradation. The PTH derivatives of glyeosylated amino acids have
`been extracted during solid phase degradation, but not from the spinning
`cup (110). However, a general method for partial deglycosylation with HF
`has been reported (111).
`General procedures for purifying reagents and solvents have been de-
`scribed (5, 112), but the strategy pf continuously scavenging contaminants
`in the rwgent bottles provides a simple alternative (1 13) that yields impres-
`sive results (114) and
`problems of reagent instability. The most
`widely used coupling bufer, Quadrol, is diflicult to purify, but Begg recom-
`mends its replacement by tetrahydroxyethylethylenediatnine (THEED),
`and claims very high degradative efficiency (115). Significant advantages in
`peptide retention are achieved by reducing the Quadrol concentration from
`1 M to 0. l-0.18 M (116, 117), although many laboratories have now
`adopted intermediate concentrations of 0.25-0.33 M, provided that protect-
`ing carriers are included. Alternatives to phenylisothiocyanate have been
`employed, usually in order to achieve greater sensitivity of product detec-
`tion. These are reviewed in the section on microsequencing.
`An important operational improvement is the inclusion of a polyionic
`carrier to improve retention of small peptides and small quantities of pep-
`tide or protein in the spinning cup. The most widely used is Polybrene, a
`polycationic carrier introduced in 1978 (118, 119), whereas a polyanionic
`alternative should be better with basic peptides containing arginine or histi-
`dine (120).
`_
`The chemistry of the degradation can be optimized by appropriate con-
`trol of the automation program. These programs are continually being
`modified in various laboratories, but minor changes tend not to be published
`until enough of them accumulate to merit a detailed description (eg. 102,
`
`'
`
`Page 15 of 28
`
`
`
`Page 15 of 28
`
`

`
`272
`
`WALSH ET AL
`
`116, 117. 121, 122). Improvements include schemes to avoid a postdelivery
`drip from the reagent delivery tubes during vacuum steps (117), to control
`film height by cup speed changes (102), to reduce partial acid cleavage by
`minimizing N to 0 shifts (122), to achieve azeotropic removal of acid (116,
`l2l), and to apply- gentler conditions for drying the film and extracting the
`product with minimal loss of peptide ([16, 117, 121). More complex pro-
`grams define shorter cleavage times, double coupling, and/or double cleav-
`age. These can provide residue-specific experimental conditions at
`preselected degradation cycles (I02, 103, 123). It should be noted that one
`impact of the widespread use of the Polybrene carrier has been to reduce
`the rate of evolution of programs designed to minimize peptide washout.
`
`MODIFICATIONS or TI-IE INSTRUMENTATION Etfective improvement
`of the sensitivity and efficiency of automated Edman degradation requires
`a combination of engineering and chemical skills. Several laboratories have
`introduced minor mechanical changes, (e.g. 124, 125), and two have de-
`scribed major redesigns of the spinning cup instrument (101, 103). Unfortu-
`nately, the major modifications are not yet commercially available, and
`other laboratories have only introduced a few changes consistent with their
`engineering capabilities.
`The vacuum system in the conventional instrument has many problems.
`The pneumatic valve in the fine vacuum system has a significant failure rate,
`but can be replaced by a more reliable alternative (125). A more complex
`vacuum control manifold has been designed (101, 126) and adopted in a few
`other laboratories. Edman’s original simple design (5) may yet be rea-
`dopted. The reliability of the vacuum sys