, no. 1. 1990
`
`'
`
`ISSN 0022-1759‘
`
`JIMMBG 134(1)1—152(199o)
`
`
`
`ICAL ROOM L
`
`2 -4: 1990
`
`.mrr.:,r-re : =r-~-- *"
`'r- r-J‘:
`
`Uiiiml
`
`“nmm3m@@n
`ME@EE(D@S
`
`EDITORS:
`
`
`
`V. NUSSENZWEIG
`
`M.W. TURNER
`
`Page 1 of 14
`
`ILMN EXHIBIT 1034
`
`
`
`Page 1 of 14
`
`ILMN EXHIBIT 1034
`
`

`
`ir@Imim GJF
`g@§E
`mE@na1@@s
`
`itors: V. NUSSENZWEIG
`
`Department of Paihoiogy, New York University Me-dicai Center. Schooi of Medicine,
`New York, NY 10016. U.S.A.
`
`M.W. TURNER
`
`Department of irnmunoiogy, institute of Chiid Health, University of London. 30 Guiiford
`Street, London WC1N IEH, U.K.
`
`Volume 134, 1990
`
`
`
`= LSEVIER SCIENCE PUBLISHERS B.V. — AMSTERDAM
`
`Page 2 of14
`
`
`
`Page 2 of 14
`
`

`
`1'9 1990 Elsevier Science Publishers B.V. (Biomedical Division)
`All rights reserved. No part ol this publication may be reproduced. stored in a retrieval system. or transmitted in any form or by any
`means. electronic. mechanical. photocopying, recording or otherwise. without the prior written permission of the publisher. Elsevicr
`Science Publishers B.V. (Biomedical Division). P.O. Box 15.27. 1000 BM Amsterdam. The Netherlands.
`
`No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability.
`negligence or otherwise. or from any use or operation ol any methods. products, instructions or ideas contained in the material
`herein. Because ol the rapid advances in the medical sciences. the publisher recommends that independent verification oi‘ diagnoses
`and drug dosages should be made.
`
`inclusion in this publication does not
`is expected to conform to ethical (medical) standards.
`Although all advertising material
`constitute a guarantee or endorsement of the quality or value of such product or of the claims made of it by its manufacturer.
`
`Special regulations for aurhc.-rs. Upon acceptance of an article by the journal. the author(s) will he asked to transfer copyright of the
`article to the publisher. This transfer will ensure the widest possible dissemination of information.
`
`Submission of a paper to this journal entails the author‘s irrevocable and exclusive authorization of the publisher to collect any sums
`or considerations for copying or reproduction payable by third parties (as mentioned in article 17, paragraph 2 of the Dutch
`Copyright Act of 1912. and in the Royal Decree of June 20. 1974 (S. 351] pursuant to article 16b of the Dutch Copyright Act of l9l2)
`and/or to act in or out of Court in connection therewith.
`
`Special regulations for readers in the U.S.A. This journal has been registered with the Copyright Clearance Center. Inc. Consent is
`given for copying of articles for personal or internal use. or for the personal use of specific clients. This consent is given on the
`condition that the copier pays through the Center the per—copy fee stated in the code on the first page of each article lor copying
`beyond that perrnitted by sections 10'! or 108 of the U5. Copyright Law. The appropriate fee should be forwarded with a copy of the
`first page of the article to the Copyright Clearance Center. Inc.. 2'."-' Congress Street. Salem. MA 01970. U.S.A. If no code appears in
`an article. the author has not given broad consent to copy and permission to copy must be obtained directly from the author. All
`articles published prior to 1980 may be copied for a per-copy tee of US. $2.25. also payable through the Center. This consent does
`not extend to other kinds of copying. such as for general distribution. resale. advertising and promotion purposes. or for creating new
`collective works. Special written pennission must be obtained from the publisher for such copying.
`
`This journal is printed on acid-free paper.
`
`Printed in The Netherlands
`
`Page 3 of14
`
`
`
`Page 3 of 14
`
`

`
`23
`
`
`
`al of Immunological Methods, 134 (1990) 23-33
`trier
`
`Multi-pin peptide synthesis strategy for T cell
`determinant analysis
`
`N. Joe Maeji, Andrew M. Bray and H. Mario Geysen
`Coselco Mimaropes Pty. Lid. (0 whofifv owned subsidiary of the Commimweairh Serum Laboratories). Crrr Duerdtrt at Martin Streets.
`P. 0. Box 40, Clayton, Victoria. Australia
`
`(Received 13 April 1990. revised received 28 June 1990, accepted 3 July 1990}
`
`Techniques to synthesize many peptides simultaneously exist, however their individual cleavage and
`.-subsequent purification constitutes a bottleneck to total throughput. Biological screening of peptides is
`Di-generally carried out at physiological pH in aqueous solutions. However. peptides. unless individually
`-'-‘purified are usually contaminated by residual compounds used in their preparation such as trifluoroacetic
`-"acid, organic solvents. scavengers etc. In testing with cellular systems. such as T cell determinant analysis.
`such contaminations must be rigorously exc1uded._ We have extended the pin synthesis technique of
`synthesizing and screening large number of peptides (Geysen et al., 1984)
`to the analysis of T cell
`sieterrnjnants. Peptides can be synthesized on polyethylene pins, the side chain protective groups removed
`Fund the peptides washed free of contaminants. A linker system stable under these conditions can then be
`,-triggered to cleave the peptides from the pins in an aqueous solution at neutral pH. This strategy enables
`the rapid mapping of T cell determinants. It is also applicable to other systems where large numbers of
`solution phase peptides are required, for example, in the study of hormone analogues.
`
`Key words: Simultaneous multiple peptide synthesis; T cell delenninant analysis
`
`Qlntroduction
`
`The development of a number of different
`methods for the simultaneous synthesis of many
`-peptides (Geysen et a1., 1984; Houghten, 1985;
`Frank and During, 1988; Hudson, 1988) empha-
`fsises the need for such technologies in the fields of
`
`Correspondence to: H.M. Geysen, Coselco Mimotopes Pty.
`1.Id.. R0. Box 40. Clayton. Victoria 3168. Australia.
`t—butoxy-
`Abbreviations." Ac, acetyl;
`,8.
`,E'—alanine; Boc,
`enrbonyl; DCC. dicyclohexylcarbodiimide: DMAP. 4-(dime1h-
`Eytaminoipyndine; DMF, dimethylformamide; DN B, 2,4-di-
`liitxophenyl: Fmoc, 9-fluorenylmethoxycarbonyl; HOBT. 1-hy-
`flroxybenzotriazole; Me0H, methanol; PITC. phenylisothio-
`tyartate; r-Bu, t-butyl; TEA. triethylamine: TFA. trilluoracetic
`-acid.
`
`immunology and pharmacology.
`biochemistry.
`Each of these methods has its own advantages and
`disadvantages. However.
`in terms of the sheer
`number and overall simplicity with which peptides
`can be simultaneously synthesized and subse-
`quently tested the method proposed by Geysen
`(1984) is superior. Reservations about this method
`of peptide synthesis stem from the inability to
`assess directly the purity of the peptides cova-
`lently bound to polyethylene pins, despite the
`consistency of the results in numerous immuno-
`logical studies (Geysen et al._. 1987a).
`For the systematic study of T cell determinants
`in cell-proliferation systems.
`large numbers of
`solution-phase peptides are required.
`Individual
`purification of peptides is impractical if the poten-
`tial of the pin method of peptide synthesis is to be
`
`,¢.;lJ022-1759/90/$03.50 '1‘) 1990 Elsevier Science Publishers B.V. (Biomedical Division)
`{Ir._,'_
`
`Page 4 of14
`
`
`
`Page 4 of 14
`
`

`
`Peptide- NH
`
`Peptide-NH
`
`§_(O
`
`N
`
`Boc-NH
`
`t1l
`
`0
`
`y
`°'P'“
`
`0
`
`N
`
`HN
`
`0 ”
`HO-Pin
`1.3
`
`W
`
`Peptide-NH
`
`Peptide-“NH
`
`0
`
`Base or
`
`+
`H3l'.|
`
`Buffer
`N Tb HZN
`
`0
`
`N
`
`:2:
`
`0
`
`_
`0"'Pll"'l
`
`:31
`
`C’
`
`I
`O-Pll
`
`Fig. 1. Peptide cleavage by diketopiperazine formation.
`
`,_
`
`been recognized as a problem in solid-phi
`synthesis of dipeptides. especially when the 9.
`terminal residue is an imino acid such as :-.-;
`(Rothe and Mazanek. 1974). This reaction can a-'
`used as the basis of a cleavable linker as shown '
`.-
`
`Fig. 1. The peptide is synthesized on the 5-anti
`group of the protected dipeptide ester (I ). M ='
`Boc deprotection and subsequent deprotonati
`of (2),
`the Lys-Pro moiety of (3) cyclizes =
`cleaves the peptide from the solid support.
`-
`‘
`though the cleaved peptide (4) has a C ter
`diketopiperazine group,
`this ending is uncharg
`the assembly of the cleavable linker is simple .
`_
`diketopiperazine formation proceeds under -n
`mild conditions.
`
`
`
`Materials and methods
`
`Symhesfs of linker and peptides
`Synthesis of peptides was on solid polyethyl
`pins that were radiation grafted with acrylic . ‘
`N-r-butoxycarbonyl-1.6-diaminohexane was
`pled to this grafted polyacrylic acid matrix .-
`dicyclohexylcarbodiimide mediated coupling.
`|"_
`tailed methodology for the preparation of pins u
`been described elsewhere (Geysen et al., 1987'
`
`24
`
`realized. Hence. the ‘crude’ peptide product must
`be of good purity as well as being free of toxic
`chemicals commonly used in peptide synthesis
`such as trifluoroacetic acid, organic solvents and
`scavengers. This can be achieved by incorporating
`an appropriate linker between the pin and the
`synthesized peptide which is stable to relatively
`harsh acidolysis. but
`is cleavable in biologically
`compatible solutions. The linker design is depen~
`dent on the method of peptide synthesis and the
`Fmoc/t-butyl protection strategy (Atherton et al.,
`1978; Chang and Meienhofer. 1978) rather than
`the Boc/benzyl system of Merrifield was selected
`because of its milder side chain deprotection con-
`ditions.
`
`Cleavable linkers requiring hydrogenolysis and
`photolysis are not practical with the pin method.
`Cleavage with cyanogen bromide on peptides con-
`taining methionine (Hancock and Marshall. 1975),
`or aqueous formic acid cleavage on Asp-Pro bonds
`(Landen, 1977) yield peptides which require fur-
`ther processing before they can be used. These
`methods also obviously restrict the peptides which
`can be made to sequences which do not contain
`the labile bonds. Landen's approach was recently
`taken to characterize T cell determinants in con-
`
`junction with the pin synthesis method (Van der
`Zee et al., 1989). Following side chain deprotec-
`tion and washing to remove non—covalently bound
`impurities. peptide/pin cleavage was effected by
`using 70% formic acid for 20 h at 37°C. The
`solutions had to be lyophilized to remove excess
`formic acid before the peptides were redissolved in
`phosphate
`buffer. Extensive post-cleavage
`processing is an inherent weakness of this strategy.
`which is also limited by its incompatibility with
`peptides containing acid-labile peptide bonds.
`Other potential linker groups include those based
`on p-(hydroxymethyhbenzoic acid (Atherton et
`al.. 198]) and hydroxyacetic acid (Baleux et al..
`1986). Both of these linker groups are stable to
`low pH and will cleave smoothly under basic
`conditions. However,
`the resulting peptide solu-
`tions are basic and require subsequent neutraliza-
`tion. Since an uncharged C terminal of the peptides
`was preferred to avoid potential complications
`due to electrostatic effects, a well documented
`‘side reaction‘ was used as the basis of a cleavable
`
`linker group. Diketopiperazine formation has long
`
`Page 5 of14
`
`
`
`Page 5 of 14
`
`

`
`25
`
`from Novabiochem (Switzerland). All had the at-
`amino group protected by the Fmoc group and
`had their side chains protected by:
`I-butyl
`for
`threonine. serinc. aspartic acid, glutarnjc acid and
`tyrosine: I-butoxycarbonyl for lysine: 4-rnethoxy-
`2.3.6-trimethylbenzencsulfonyl for arginine; Fmoc
`for histidine: and. trityl for cysteine. Subsequent
`synthesis of the peptides on to the 1-:—amino group
`of the lysine was completed by repetitive cycles of
`Fmoc deprotection and amino acid couplings (Ta-
`ble I) at the rate of one or two residues per pin per
`day.
`After removal of the final Fmoc protecting
`group,
`the amino terminus was routinely capped
`using the
`following conditions: DMF : acetic
`
`H2N
`
`research purposes are available from
`for
`Kits
`Cambridge Research Biochemicals. Cheshire, U.K.
`After Boc deprotections, the cieavable linker was
`“assembled as shown in Fig. 2. The amide bonds
`i were formed by DCC/HOBT-mediated coupling
`as previously described (Geysen et al.,
`l987a).
`Ester formation was effected by carbodiimide con-
`l_ densation catalyzed by 4—(dimethy|amino)pyridine
`(DMAP) as described by Steglich and Hofle
`{" (1969). DMAP being a weak base can cause par-
`
`-
`
`__ lia] cleavage of the Fmoc group from the -amino
`‘" group of proline (Atherton et al., 1978), so Boc-Pro
`I was used instead of Fmoc-Pro in esterification.
`Boc-Lys(Fmoc)-OH was purchased from Auspep
`(Australia). All other amino acids were purchased
`
`t'BU— 0
`
`HN
`
`FI"l'1OC—Nj\( ‘pin
`
`H
`
`O
`
`HN
`
`_
`
`l
`
`0
`
`Fig. 2. Assembly of the diketopiperazine-forming cleavable linker.
`
`
`
`Page 6 of 14
`
`

`
`using a Titertek Multiskan MCC/340 Mkll. Each
`data point presented is the average of four pins.
`
`HPLC evaluation of cleaved peptides
`HPLC analysis was carried out with a Waters
`Associates liquid chromatography system consist-
`ing of two 510 pumps, WISP 710B autosampler,
`model 440 UV detector (254 nm) with an ex-
`tended wavelength module (214 rim) and DEC
`Professional 350 work station. Analytical runs were
`carried out on a 5 pm Merck Lichrosphere 100
`RP-13 (250 X 4 mm i.d.) column. The buffer sys-
`tem comprised: A. 0.1% TFA in water. and B.
`0.1% TFA in waterzacetonitrile (40:60 v/v).
`Gradient elution from A to B was carried out in
`5-20 min.
`
`Amino acid unafysis of peptides
`Pre-column deriuatizarion with phen_vIi's0rhio-
`cyrmate (PITC). Crude peptide solutions were
`dried and then hydrolysed in 1 ml of 6 N HCl
`under nitrogen at 112°C for 24 h. 200 til aliquots
`of hydrolysate were dried in a SpeedVac con-
`centrator (Savant Instruments, Farmingdale. NY)
`under vacuum and redried twice after reconstitu-
`
`tion in 40 iii water: MeOH :TEA (2 : 2 : 1).
`Derivatization was carried out with 20 pl
`MeOH:H2O:TEA:PITC (7:1 :1 : 1) for 20 min
`at room temperature. The derivatised samples were
`dried and reconstituted in. 400 it] eluent A (Cohen _
`et al.. 1988) whose pH had been adjusted to 5.9.
`Separation of the phenylthiocarbarnyl derivatives
`was carried out on a 5 pm glass-lined octadecylsi-
`lane (C18) column (250>< 4 mm i.d.) (Scientific
`Glass Engineering. Melbourne. Australia) at 50°C
`with UV detection at 254 nm. All HPLC
`
`equipment was from Waters Associates (Milford,
`MA).
`
`T-cell proliferation assay
`Duplicate assays were performed in 96 Well.
`round—bottorned microtitre plates (Nunc) with a
`final volume of 200 p.l/well. 5 X 103 T cells were
`added with an equal number of v-irradiated (SON
`rads) Epstein-Barr virusdransformed autologous B.
`cells and antigen to each well. Peptide antigens-
`were used at an estimated concentration of 0.5-l
`
`'
`
`
`
`and 5 pg/ml while the tetanus toxoid positive:
`control was used at 10 Lf/tnl. Cultures were in-
`
`26
`
`TABLE I
`
`Fmoc SYNTHESIS CYCLE
`
`Reagent
`
`Number of
`treatments
`
`Treatment
`time
`
`DMF
`MeOH
`Air dry for at least 10 min
`DMF
`
`20% piperidine in DMF
`DMF
`MCOH
`
`Air dry for at least 10 min
`DMF
`Ami no acid coupling
`
`1
`1
`
`1
`
`1
`1
`3
`
`1
`
`2 min
`10 min
`
`2 min
`
`30 min
`5 min
`2 min
`
`5 min
`5 h or
`overnight
`
`anhydride:TEA (50: 5 : 1 v/v/v) for 90 min. Side
`chain deprotection was carried out with triflu0ro-
`acetic
`acid : phenol : ethandithiol
`(95 : 2.5 : 2.5
`v/w/v) for 4 h at room temperature. Side chain
`deprotection also removed the or-Boc group from
`lysine to set up the conditions for subsequent
`diketopiperazine formation and cleavage of the
`peptides from the pins. Non—covalent|y bound
`contaminants were removed by the following
`washing steps: vacuum desiccation forl h; sonica-
`tion in 1% acetic acid in MeOH/water (1:1) for
`15 min; and an overnight wash in 0.1 M phos-
`phate buffer at pH 5. The synthesized peptides
`were then cleaved from the pins. or, dried in vacuo
`and stored in air tight plastic bags containing
`desiccant for later cleavage.
`
`Cleavage test
`To determine the efficiency of cleavage, a test
`system was developed in which 2,4-dinitrot'luoro-
`benzene (BDI-I cat. no. 44032) was reacted for 4 h
`with Boc-Lys-Pro-O-pin to form the clinitrophenyl
`derivative. Aqueous
`sodium hydroxide (0.03%,
`0.06%, 0.08%. 0.12% and 0.16%) and phosphate
`buffer (pH 6.0. 7.0 and 8.0, and, 0.05 M, 0.1 M
`and 0-2 M) were appraised as peptide cleavage
`media. Tests were performed in 96-well microtitre
`plates using 150 pl of cleavage solution/well. The
`rate of diketopiperazine formation, and therefore
`peptide/rod cleavage, was assessed by the in-
`crease in absorbance of the cleavage solution at
`340 rim due to the dinitrophenyl chromophore
`
`Page 7 of14
`
`
`
`Page 7 of 14
`
`

`
`27
`
`ABSOREANCEaL340nm F
`
`0
`
`30
`
`90
`30
`TIME (min)
`
`120
`
`150
`
`Fig. 4. Cleavage of H-Lys(DNB)-Pro-Q-pin with 0.2 M phos-
`phate buffer: pH values 6. 7 and 8.
`
`sorbance was obtained by cleavage with 1% sodium
`hydroxide. At lower base concentrations the rate
`of cleavage slows considerably after 2 h even
`though peptide cleavage is incomplete. However,
`cleavage under
`these conditions
`is probably
`dominated by saponification of the ester linkage
`rather than intramolecular aminolysis to form di-
`ketopiperazine. Boc-Lys-Pro-O-pin ester (1) was
`shown to cleave in 1% sodium hydroxide even
`though diketopiperazine formation is not possible
`(data not given). Examples of cleavage using phos-
`phate buffer are shown in Figs. 4 and 5. Increas-
`ing pH increases the rate of diketopiperazine for-
`mation. as expected.
`Increasing ionic strength
`caused a decrease in the rate of diketopiperazine
`formation within the range (0.05-0.2 M) studied.
`The differences in cleavage rate observed with
`variation in pH and ionic strength were within
`lim.its which allows flexibility in adjusting the
`
`AHSORBANCEat340nm
`
`'
`
`0
`
`so
`
`so
`so
`TIME (min)
`
`120
`
`150
`
`Fig. 5. Cleavage of H- Lys( DNB)-Pro-O—pin with pH 6.0 phos-
`phate buffer: 0.05 M. 0.1 M and 0.2 M.
`
`
`
`, bated for 72 h after which they were pulsed for
`;fi h with 1 ,uCi of tritiated thymidine. Cells were
`lhen harvested on to glass fibre filter mats, pre-
`pared for scintillation counting and the incorpo-
`iated tritiated thymidine measured using a flat-bed
`scintillation counter (LKB—Betaplate). Assay re-
`sults are expressed as the mean counts/min of
`replicate cultures (Ho et at., 1989).
`
`Results
`
`Diketopiperozine Formation
`As shown in Fig. 1, as soon as the at-Boc
`I
`it protecting group of the lysine is removed,
`the
`'tlH-Lys-Pro-O-pin moiety (3) can be triggered to
`-"-cyclize to form the diketopiperazine derivative (4).
`!'The stability of the protonated assembly (2) to
`iivarious washing solvents was investigated. Acetic
`| acid, methanol and water in various ratios as well
`
`H as low pH aqueous phosphate buffer were as-
`sessed as potential washing solvents. Most wash-
`ing solvents lead to no appreciable diketopipera-
`..zi.ue formation. Under the chosen conditions, pins
`iremained bright yellow indicating that. at worst.
`H|-only a small proportion of the chromophore was
`_lost from the pins. Peptide cleavage can be ef-
`lpfected by dilute basic solutions or neutral buffers.
`" Fig. 3 presents the cleavage rate at various con-
`centrations of sodium hydroxide. Maximum ab-
`
`naoenm
`Iflltllllflfll
`
`0.0
`
`0.5
`
`1.0
`
`1.5
`
`2.5
`2.0
`Tlll (hr)
`
`3.0
`
`5.5
`
`4.0
`
`4.5
`
`Fig. 3. Cleavage of H-Lys(DNB)-Pro-O-pin with aqueous
`medium hydroxide. The horizontal line indicates the maximum
`qbsorbance obtained with 1% sodium hydroxide after 3.5 h
`EXDOSLI 1'8.
`
`Page 8 of14
`
`
`
`Page 8 of 14
`
`

`
`28
`
`TABLE II
`
`CLEAVAGE DATA FOR H-Lys(DNB)-Pro-O-pin
`
`Cleavage
`solution
`
`0.06%NaOH
`0.03% NaOH
`0.12% N30!-I
`0.] M phosphate bufler (pH 7}
`
`Proline ‘
`
`25 nmol 1 3
`34 nmoli 3
`50 nmol 1 7
`42 nmol 1 6
`
`" Average prolinc levels for four pins after a cleavage time of
`3.5 h. The proline level on pins prior to cleavage was about 80
`nmol.
`
`cleavage solution to comply with the conditions
`required for subsequent biological assays. The
`amount of peptide cleaved from the pins was
`confirmed by amino acid analysis. Table 11 lists
`
`the proline levels of some cleavage solutions after
`3.5 h.
`
`HPLC assessmem of cleaved peptides
`A set of analogues of the well-characterized
`antibody epitope, NDFLEKI“, of myohemeryth-
`rin (Getzoff et al., 1987. Geysen et al., 1987b) was
`synthesized on two sets of 96 pins. These com-
`prised the peptides "FLEKl and DFLEK* where
`indicates positions where the amino acid was
`varied to include each member of the 20 geneti-
`cally coded amino acids. Each peptide was synthe-
`sized
`in quadruplicate
`as were
`two control
`peptides, PLRQGG and GLAQGG. Additional
`copies of
`the parent sequence DFLEKI were
`synthesized on the remaining 16 pins. In another
`synthesis, 21 well-characterized antibody epitopes
`(Geysen et at.. 1988) were synthesized. also in
`
`$585
`
`mil‘
`
`1E3
`
`158%
`
`[IN
`
`LBBE
`
`5'39
`
`.,l
`GFLEKI
`
`IFLEKI
`
`28
`
`Tl ME
`
`(min)
`
`Fig. 6. HPLC traces of Ac—DFLEKl—B—eyda (KP) and analogues. The D residue has been replaced by A. C. E. F. G. H, I. K and L
`The peptides were cleaved into 0.08% sodium hydroxide. Detection at 214 nm. " indicates phenol.
`
`Page 9 of14
`
`
`
`Page 9 of 14
`
`

`
`29
`
`quadruplicate. In all a total of 62 different peptides
`were synthesized on 276 separate pins. In the case
`of
`the analogues of DFLEKI, after side chain
`_ deprotection with TFA, the peptides were washed
`only with MeOH (2 X 2 min) and cleaved using
`' 0.03% NaOH (150 pl/well). The HPLC traces of
`the first half of
`the set *FLEK.i analogues is
`shown in Fig. 6. The quality of the peptides was
`promising considering that
`these data were the
`very first obtained without optimization of the
`chemistry. The peaks at 16.3 min and 16.3 min
`appear in every trace and are due to phenol and
`I-butylated phenol
`respectively,
`indicating a re-
`quirement for a more stringent final washing pro-
`
`cedure. No other non-peptide impurity was ob-
`served at either 214 or 254 nm. Fig. 7 shows a
`selection of HPLC traces from the 21 peptides
`made in the latter synthesis. These peptides under-
`went the washing protocol given in the methods
`section. Peptide/rod cleavage was effected with
`0.1 M phosphate buffer at pH 7. The more vigor-
`ous washing procedure reduced the phenol con-
`taminants to trace levels. Table III presents the
`amino acid analyses of the peptides shown in Fig.
`7. Analyses were done on both cleaved peptide
`solutions and peptide remaining on the pins after
`cleavage. Table IV gives the absolute quantities of
`peptides cleaved and still remaining on pins. In all
`
`1538
`
`IV
`
`1883
`
`589
`
`8
`
`15%
`
`IV
`
`1393
`
`538
`
`.
`
`.
`I
`
`I
`
`’
`
`i
`
`16
`
`15
`Minute;
`
`EB
`
`25
`
`18
`
`15
`Minutes
`
`EB
`
`25
`
`15
`
`15
`Minutes
`
`20
`
`25
`
`AKDGKEID
`
`GPSDTPIL
`
`APVVHNPA
`
`r‘ L
`
`‘
`
`I
`
`,
`
`A
`
`I
`
`28
`15
`(rninl
`TIME.
`Fig. 3'. HPLC traces of a selection of peptides cleaved with 0.1 M. pH 7 phosphate buffer. All peptides are N terminal acetylaled and
`carry a C terminal ,8-c_vc.*o(KP) moiety. Detection at 214 nm_
`
`IE
`
`15
`
`23
`
`$_':"L-
`
`Page10of14
`
`6 -
`
`E5
`
`16
`
`85
`
`19
`
`15
`
`29
`
`25
`
`
`
`Page 10 of 14
`
`

`
`30
`
`TABLE I11
`
`AMINO ACID ANALYSIS OF SOME PEPTIDES CLEAVED WITH 0.1 M PHOSPHATE BUFFER pH 7 3"’
`
`Amino acid
`
`Ratios ‘
`GTDFKYKG
`
`NSAPN LAT
`
`TEYYLN HG
`
`AKDGKEID
`
`GPSDTPIL
`
`APVVHNPA
`
`0.55 (0.88)
`
`1.79 (1.71)
`
`0.63 L‘
`
`1.06 (1.35)
`1.24(1.18)
`2.00 (2.06)
`2.52 (1.85)
`
`1.00 (1.08)
`1.l0(1.02)
`
`0.67 (1.52)
`I.I6 (0.83)
`0.96 (1.50)
`l.0S(1.22)
`
`1.30 (1.13)
`2.09 (-1
`
`2.01 (2.00)
`1.06(l.15)
`
`0.81 (1.35)
`
`1.16 (1.27)
`
`1.06 (1.06)
`1.49 (0.90)
`
`1.23(1.02)
`
`0.87 (0.66)
`
`1.02 (0.81)
`
`0.77 “
`0.78 (1.44)
`
`0.90 (1.21)
`1_14(1.13)
`
`3.60 (2.63)
`
`1.10(1.04)
`1.07(l.04)
`
`1.05 (0.81)
`1.00 (1.44)
`
`1.98 (2.30)
`3.66 (2.92)
`
`1.43 (1.75)
`
`D
`E
`S
`G
`H
`B
`T
`A
`P
`Y
`v
`I
`L
`F
`K
`
`1.75 (2.66)
`
`1.00 (1.3?)
`1.19(0.98)
`
`1.37 (0.82)
`1.09 (0.94)
`
`1.33 (0.86)
`2.6? (2.15)
`
`1,140.03)
`
`1_09(1.08)
`
`0.63 (0.86)
`
`0.64 (1.08)
`
`2.09 (2.10)
`
`0.70 (0.81)
`
`0.86 (0.92)
`
`a All peptides are N terminal acetylated and carry a C terminal ,8--:,vcl'o(KP) moiety.
`" Amino acid analyses were performed on sample of high salt content and low peptide content (10-10 nmol). These conditions are
`not ideal for the PITC derivatization procedure.
`C Figures in brackets are the post cleavage ratios for peptides remaining on the pins. As serine was used as part of the cleavable linker
`assembly, the levels found on pins are not given.
`
`cases. the phosphate buffer solutions (150 til/pin)
`contained more than 10 nmol of peptide after a
`cleavage time of 3.5 h.
`
`T (‘eff determinant studies
`
`A well-characterized system was selected to test
`the suitability of using cleaved peptides in T cell
`proliferation assays. 1-lo et al.
`(1989)
`identified
`several peptides
`(15-mers)
`from tetanus toxin
`which encompassed the determinant for a particu-
`
`TABLE IV
`
`QUANTITY OF PEPTIDE IN SOLUTION AND ON PINS
`AFTER PHOSPHATE BUFFER CLEAVAGE (3.5 h) 3
`
`Peptide "
`
`GTDFKYKG
`NSAPN LAT
`TEYYLNHG
`AKDGKEID
`GPSDTPIL
`APVVHNPA
`
`Pin
`(nmol)
`25
`17
`20
`20
`24
`18
`
`Well
`(nmol)
`47
`30
`16
`14
`14
`16
`
`Total
`( nmol)
`‘T2
`47
`36
`34
`33
`34
`
`lar T cell clone isolated from a human donor.
`
`A“.
`
`These peptides were synthesized by the method of
`I-[oughten (1985) spanning the appropriate region
`of the sequence of tetanus toxin. Only peptides
`containing the sequence YSYFPSV stimulated the
`T cells. and it was therefore inferred that
`the
`determinant for this c1one"was mYSYFPSV5"°.
`To validate our method of peptide preparation,
`overlapping peptides of four lengths (7-mer, 8-mer,
`9-mer and 10-mer) spanning the region of the '
`sequence of tetanus toxin containing the putative _
`determinant, YSYFPSV. were synthesized and ii
`tested for their ability to activate the T cell clone.
`‘_
`Fig. 8 shows that cell proliferation was observed
`for the octapeptide corresponding to, and most
`longer peptides
`encompassing the
`sequence i
`593YSYFPSV16°°. No toxic effects due to residual
`chemicals from either the synthesis. side chain '
`deprotection or cleavage steps remaining in the
`final solution, were observed.
`
`Discussion
`
`" The results are an average of two wells or two pins.
`h All peptides are N terminal acetylated and carry a C tenni-
`nal B-eyclo(KP) moiety.
`
`a well-cha.r-
`formation is
`Diketopiperazine
`acterized side reaction in peptide synthesis. It is
`
`Page11of14
`
`
`
`Page 11 of 14
`
`

`
`31
`
`generally accepted that near-quantitative loss of
`peptide from resin is possible and both acid (Gisin
`and Merrifield, 1972) and base (Khosla et at..
`1972) catalyzed mechanisms have been proposed.
`The use of this ‘side reaction’ as a method of
`
`for cleavage of
`cleaving peptides has potential
`rnany pepfides sunuhaneoufly to ymhd pepfide
`solutions which are compatible with biological
`testing. The main disadvantage is that each peptide
`remains linked at
`its carboxy-terminus to the e-
`amino group of cyc-!o(Lys-Pro). This is not neces-
`sarily a problem where the method is being used
`as a screening process where every overlapping
`peptide of various lengths (say. of lengths 6-rner,
`7-mer, 8-mer and 9-mer) from the sequence of a
`protein is synthesized. Peptides of interest could
`be resynthefized twing ahernadve chenustfies to
`confirm or reject the initial screening result.
`The cleavable linker group is easily assembled
`using commercially available reagents. The linker
`group exists as its stable TFA salt after side chain
`deprotection and can be washed with a range of
`solvents under acid or near neutral conditions (pH
`5.0) without appreciable cleavage. The cleavage of
`peptides from pins occurs in aqueous buffer within
`the physiological pH range. To our knowledge
`there is no other handle/linker group which is
`cleavable under such conditions. Initially, cleavage
`for T cell determinant studies was done with 0.08%
`NaOI-I. This cannot be construed as harsh condi-
`
`tions because the pins themselves are slightly acidic
`because of their polyacrylic acid matrix. Under
`the couphng condifions used. hfibbutoxycarbon-
`yl-1,6-diaminohexane couples only to the freely
`available. surface layer of polyacrylic acid leaving
`the less accessible groups buried inside the poly-
`mer matrix. However. these carboxylic acid groups
`can be titrated and they probably account for the
`drop in pH which was observed in the cleavage
`solution over 4 h (data not presented). They also
`probably assist in the stabilization of the Lys-Pro
`emer(2)dufingthewaflungcydesaherfidechmn
`deprotection has been effected.
`HPLC analysis shows that 46 of the 62 peptides
`have purities greater than 80% if the phenol peaks
`are ignored. implying a mean coupling efficiency
`of more than 96% for each step in the peptide
`synthesis. HPLC analyses of replicate syntheses of
`the same peptide showed no significant pin to pin
`
`‘O00
`
`3000
`
`2000
`
`1000
`
`4000
`
`mm
`
`2000
`
`1000
`
`4000
`
`3000
`
`2000
`
`1000
`
`O
`
`4000
`
`3000
`zoom
`
`1000
`
`Yfrne r
`peptides
`
`II
`xvugsaocr
`E05
`610
`
`00
`
`Bfmer
`pept1de.-5
`
`I II
`ISKVNQGAQG]_f
`00
`605
`610
`
`Qfme r
`peptides
`
`565
`
`séo
`
`I I I I I I I I I
`sgrssvisxvuocaoe;
`sea
`son
`605
`510
`
`r
`pep '1 es
`
`IIIIIIIII
`IIIIIIII
`nnanrnsrarvsgrpsvrsxvnocaoer
`555
`5'90
`595
`sno
`605
`610
`
`°
`
`Fig. 3. Peptides spanning the region of tetanus toxin encom-
`passing the putative determinant 5°’YSYFPSV5"° were pre-
`pared and tested for their ability to stimulate the correspond-
`ing T cell clone. The results are expressed as counts/rnin of
`incorporated tritiated thyrnidine.
`
`Page12of14
`
`
`
`Page 12 of 14
`
`

`
`32
`
`variation in purity. The retention times for the
`DFLEKI analogues were internally consistent,
`gjving longer times with hydrophobic substitution
`and correct 214 nm absorbance to 254 nm ab-
`sorbance ratios on substitution with aromatic re-
`
`sidues. Amino acid analyses were consistent with
`these results. These findings strongly indicate that
`the correct peptide has been synthesized with good
`purity. Even peptides synthesized with less than
`80% purity, are promising because nearly all HPLC
`traces showed a major peak flanked by one or two
`others.
`In cysteine- and methionine-containing
`peptides, extra peaks are expected because of
`oxidation. An excess of DCC was used in the
`
`coupling procedure and the shoulder observed in
`asparagine- and glutamine—c0ntaining peptides
`could be the result of nitrile formation. The HPLC
`
`peaks of peptides cleaved with NaOH were rela-
`tively broad compared with those cleaved with
`phosphate buffer. This is believed to be due to
`partial cleavage by saponification or possibly the
`susceptibility of diketopiperazines
`to undergo
`base-catalyzed epimerization (Kemp, 1979).
`We chose to ‘rediscover’ the T cell determinant
`
`recognized by a known T cell clone as a test of the
`suitability of these peptides in cell culture assays
`(Fig. 8). The minimum stimulatory peptide was
`shown to correspond to the residues 593-600 of
`the sequence of tetanus toxin. This is consistent
`with the original proposal by Ho et al. (1989),
`based on the ability of 15-mers to caused T cell
`pggoliferation. that the determinant corresponds to
`vsvFi=sv5"". T cell proliferation was not ob-
`served for this heptapeptide. This could be for one
`of several reasons: (a) a longer peptide than the
`essential core region maybe required to provide
`some structural stability, e.g., helix formation; (b)
`the presence of the diketopiperazine moiety which
`remains after cleavage may sterically interfere with
`the ability of the antigen-presenting cell or, subse-
`quently,
`the T cell
`to interact
`thus requiring an
`additional amino acid to space the moiety from
`the actual T cell determinant; and, (c) the mini-
`mum reactive sequence is actually an octapeptide,
`a possibility which cannot be excluded on the
`basis of the original results obtained by Ho et al.
`(1989).
`In a more extensive study of this and several
`other T cell determinants which includes all single
`
`amino acid substitutions of appropriate peptides
`as well as corresponding peptides with carboxylate
`and amide C terminal groups the effect of the
`diketopiperazine ending was not apparent (Rodda
`et al..
`in preparation).
`In this same study.
`the
`minimum T cell stimulatory sequence was found
`to require either I5” or 1"“) in addition to the core
`sequence 593YSYFPSV5"", suggesting that the in-
`ference of the minimum reactive sequence from
`longer, nested peptides. is not always valid.
`In one calendar month, 82 peptides were
`synthesized, cleaved and tested in a T cell prolifer-
`ation assay as part of this study. The ‘window’ set
`format (Geysen et al., 1987a) clearly identifies the
`minimum reactive sequence without the need to
`infer the site of the epitope or determinant from
`the data obtained with fewer, but longer peptides.
`Furthermore,
`it becomes possible to use tech-
`niques similar to those previously described for
`the resolution of B cell epitopes at
`the level of
`single amino acids, to T cell determinants (Geysen
`et at., 1987a). We have described a method by
`which many peptides can be synthesized simulta-
`neously, cleaved, and used directly without further
`purification in cell culture assays. This technique
`should quickly lead to an understanding of the
`mechanism by which peptides bind to MHC mole-
`cules and the resulting complex (MHC/peptide) is
`subsequently recognized by the T cell receptor.
`The evaluation of this -method can be extrapo-
`lated to peptides synthesized on pins using non-
`cleavable linkers. Both HPLC and amino acid
`
`a