<<

Proc. Nati. Acad. Sci. USA Vol. 73, No. 9, pp. 3160-3164, September 1976

Affinity purification of synthetic peptides [solid-phase peptide synthesis/Cys-Met-peptides/organomercurial-agarose/ribonuclease A(111-124)tetradecapeptide/ histone H441-37)heptatriacontapeptidej DAVID E. KRIEGER, BRUCE W. ERICKSON, AND R. B. MERRIFIELD The Rockefeller University, New York, N.Y. 10021 Contributed by R. B. Merrifield, July 8,1976

ABSTRACT A general strategy and a specific tactic for af- tion strategy (Fig. 1) uses this chemical distinction to separate finity purification of polypeptides synthesized on solid.supports the growing peptides from nongrowing peptides. An affinity are described and demonstrated. The desired peptide chains reagent A-B, bearing an affinity group A and a bindinggroup were distinguished from terminated peptide chains before re- moval from the support by attachment of an affinity reagent B, is selectively attached to the free amino groups of the (cysteinyl-) bearing an affinity group (thiol) and a growing chains. After cleavage from the solid support, the af- binding group (carboxylic acid). After cleavage from the syn- finity-labeled peptides are separated from terminated peptides thetic support, the affinity-labeled peptides (Cys-Met-peptides) and other peptides lacking the affinity reagent by selective were bound to an affinity receptor (organomercurial-agarose) binding to an affinity receptor R. Once the affinity-labeled and thus separated from terminated peptides and all other peptides are separated from the affinity receptor, the affinity peptides lacking the affinity group. The desired synthetic pep- tide was obtained by separation of the affinity-labeled peptides reagent is removed to obtain the desired peptide. from the affinity receptor (displacement by cysteine or other This strategy is independent of the length or sequence of the thiol) followed by removal of the affinity reagent (loss of Cys- desired peptide. It permits separation of the desired peptide Met by cleavage). from all terminated peptides and other chains lacking the af- This general affinity purification strategy is independent of finity reagent, but modified or addition peptides will not be the length or sequence of the desired peptide. After removed. Terminated peptides include chains inadvertently assembly of ribonuclease-( 11-124)-tetradecapeptide, using ra- as diolabeled acetic anhydride for termination of uncoupled in- blocked during the synthesis by such processes acetylation termediates, essentially all (>98.5%) of the acetylated deletion (2), trifluoroacetylation (3), or pyroglutamyl formation (4,5). peptides were removed by employing the organomercurial In addition, chains that have failed to couple can be inten- Cys-Met tactic. Similarly, the purity of crude synthetic histone tionally blocked by a terminating agent, which prevents the H4(1-37)heptatriacontapeptide was increased six-fold by using formation of internal deletion peptides. Finally, some non- this tactic to remove terminated peptides. A related dimeric growing chains might be chemically reactive but physically Cys-Met tactic is outlined for affinity purification of peptides inaccessible due to their location in a crowded region of the solid containing internal cysteine and methionine residues. support (6). The affinity purification method should efficiently Purification of synthetic peptides is important if proper con- remove all such nongrowing chains. clusions are to be drawn from subsequent biological experi- ments. Small peptides can often be separated from the major The organomercurial Cys-Met tactic byproducts by recrystallization or by standard biochemical One suitable affinity reagent is the dipeptide cysteinyl-me- techniques, but purification of larger synthetic peptides or thionine. The carboxyl group of methionine functions as the proteins is more difficult to achieve because the differences in binding group and the thiol group of cysteine as the affinity physical or chemical properties are relatively smaller. These group. Cys-Met can be selectively and efficiently cleaved by difficulties are magnified with the stepwise solid-phase method cyanogen bromide (7, 8) to restore the free terminal amino (1), where extensive purification is possible only at the end of group. Organomercurial-agarose (9) can serve as the affinity the synthesis. In this case, highly selective affinity purification receptor to bind the thiol group selectively and reversibly procedures can be extremely useful. through a covalent mercury-sulfur bond. A column of orga- Synthetic peptide byproducts are conveniently classified by nomercurial-agarose has often been used to bind proteins or the nature of the structural defect (Table 1) as modified pep- peptides bearing thiol groups (10-15). tides, which contain one or more chemically altered residues; A specific protocol using these components is shown in Fig. addition peptides, which contain one or more extra residues; 2. Although Cys-Met could be added as a protected dipeptide, deletion peptides, which lack one or more residues; and ter- it is more easily added by using the conventional procedures minated peptides, which have permanently stopped growing of solid-phase synthesis to couple first methionine as Boc-Met during the synthesis for either chemical or physical reasons. This and then cysteine as Boc-Cys(4-CHsOBzl). After cleavage from paper describes an affinity strategy for the separation of the the solid support with liquid hydrogen fluoride, which liberates desired product of a solid-phase synthesis from the general class the protected thiol group of cysteine, the crude peptide is gel of terminated peptides. filtered to remove the by-products of low molecular weight and treated with dithiothreitol to reduce the thiol group. After re- The affinity purification strategy moval of the reducing agent by gel filtration, the peptide is After solid-phase assembly of a peptide is complete and before applied to a column of organomercurial-agarose, a beaded the peptide is cleaved from the solid support, the desired pep- carbohydrate matrix bearing covalently bound 4-(chloromer- tide and other growing peptides each bear a reactive free amino curi)benzoyl sites. Each Cys-Met-peptide molecule is selectively group, which terminated peptides lack. The affinity purifica- bound through its thiol affinity group to an organomercurial site on the affinity receptor. The column is eluted with aqueous Abbreviation: Acm, acetamidomethyl. buffer until all peptides lacking a thiol group are removed. 3160 Downloaded by guest on September 26, 2021 Biochemistry: Krieger et al. Proc. Natl. Acad. Sci. USA 73 (1976) 3161

Table 1. Examples of peptide by-products H2N- Class Example /~NH2 Desired peptide H-Gly-Ala-Thr-Val-OH 1. Attachment of + Boc-Met Coupling H-Gly-Ala-O Affinity Reagent * Boc-Cys(R) Coupling Modified peptide r H-Thr-Val-OH RS MetNH Addition peptide H-Gly-Gly-Ala-Thr-Val-OH Deletion peptide H-Gly-Thr-Val-OH Terminated peptide CH3CO-Thr-Val-OH 2. Cleavage from HF Cleavage Synthetic Support P-2 Gel Filtration Then the Cys-Met-peptides are displaced from the column by elution with an organic thiol such as cysteine. After removal of HS - HMtNH-00C0 this thiol by gel filtration, the Cys-Met affinity reagent is se- A-NH-Oo03o lectively and efficiently removed from the desired peptide by treatment with cyanogen bromide, which converts the methi- 3. Binding to * DTT Reduction Affinity Receptor Gel Filtration onine residue into homoserine and liberates the amino group P-2 of the desired peptide. A final gel filtration to remove excess reagent and the cysteinyl-homoserine dipeptide furnishes the any from chains desired peptide plus other peptides derived _$J Mt tH-000000 still growing at the end of solid-phase assembly. Peptides that EKjS still bear the Cys-Met affinity reagent are removed during a 4. Separation from Cys Displacement second pass through the organomercurial-agarose column. Affinity Receptor + P-2 Gel Filtration This organomercurial Cys-Met tactic has been used suc- cessfully to purify two biologically active peptides, bovine HSHMetNH-000000 pancreatic ribonuclease A-(111-124)-tetradecapeptide and H4- [12-N6-[14C]acetyllysine,16-N6-[3H]acetyllysine]histone 5. Removal of + CNBr Cleavage (1-37)-heptatriacontapeptide. Affinity Reagent * P-2 Gel Filtration MATERIALS AND METHODS H2N--OCOC0 Organomercurial-Agarose. This affinity support (9) was FIG. 2. The organomercurial Cys-Met tactic for affinity purifi- prepared as described (10). Sepharose 4B was activated with cation of peptides lacking cysteine and methionine. DTT, di- cyanogen bromide and treated with ethylenediamine hydro- thiothreitol. chloride to form 2-aminoethylisoureido-Sepharose, which was mercurial-agarose (Affinose 501, Bio-Rad) was used with coupled with sodium 4-(hydroxymercuri)benzoate using N- comparable results. When packed in a column, the affinity ethyl-N'-(3-dimethylaninoethyl)carbodiimide. In several later supports bound 0.9-1.2 gmol of thiol per ml of settled support. experiments, a related commercial preparation of organo- The binding capacity was measured (11) by reaction of the column with a solution of 5,5'-dithio-bis(2-nitrobenzoic acid) (Ellman's reagent). If the capacity test is omitted, thiol-con- taining proteins are reported (11, 12) to bind in an irreversible manner. The column was regenerated after use by elution with 0.05 M sodium acetate (pH 5) containing 2 mM mercuric $1. Attachment of Affinity Reagent chloride followed by the starting buffer. Ribonuclease-(111-124-tetradecapeptide. The desired peptide-resin was prepared by the solid-phase method (16, 17) using a Beckman model 990 automated peptide synthesizer and 2. Cleavage from Synthetic Support reagent-grade or purified reagents and solvents (18). Boc- Val-OCH2-bromoresin (2.5 g, 0.50 mmol of Val) was converted cZLEcooooo +A into the tetradecapeptide-resin (3.0 g) by successive addition 0g $f 3. Binding to Affinity Receptor of 13 Boc-amino acids. The side-chain protecting groups (1) were O-benzyl for , , and glutamic acid; 0-2,6-dichlorobenzyl for tyrosine; N4-4,4'-dimethoxy- 4. Separation from Affinity Receptor benzhydryl for asparagine, and N '-2,4-dinitrophenyl for histidine. Each residue was added through the following seven-step cycle: (1) deprotection with 50% (vol/vol) $ 5. Removal of Affinity Reagent CF3CO2H/CH2Cl2 for 1 min and 30 min, (2) neutralization three times with 5% (vol/vol) diisopropylethylamine/CH2Cl2 000C00 for 2 min each, (3) coupling with Boc-amino acid and N,N'- dicyclohexylcarbodiimide (each 1.5 mmol, 3 eq) for 30 min, K flAffinity Group E Affinity Receptor (4) repetition of neutralization step (2), (5) repetition of the | Binding Group A Terminating Group| coupling step (3), (6) monitoring (19), and (7) acetylation with FIG. 1. A general strategy for affinity purification of solid-phase [3H]acetic anhydride (0.365 mmol, 2.00 X 109 cpm) and im- peptides. 0, Protecting group; 0, amino acid residues; 0 resin sup- idazole (1.0 mmol) in CH2G12 (10 ml) for 60 min. port. Attachment ofAffinity Reagent. -Boc-Methionine and then Downloaded by guest on September 26, 2021 3162 Biochemistry: Krieger et al. Proc. Natl. Acad. Sci. USA 73 (1976) Boc-S-(4-methoxybenzyl)cysteine were coupled to the ribo- respectively. Uncoupled amino groups were terminated with nuclease-(111-124)-resin using the solid-phase procedure out- nonradiolabeled N-acetylimidazole (19). The neutralization lined above. Nonradioactive acetic anhydride and imidazole and coupling steps were repeated one to four times as judged were used to terminate any unreacted amino groups after each necessary by automated picrate monitoring (20). double coupling. The Nim-dinitrophenyl group was removed The peptide was cleaved in HF/anisole containing 20 eq of by thiolysis for 30 min with 2% thiophenol in dimethylform- methionine. Part of the crude peptide (5.8 X 107cpm) in a buf- amide. fer (pH 8.9; 4 ml) 6 M in guanidine-HCI, 0.5 M in Tris, and 2 Cleavage from Synthetic Support. The peptide-resin (0.50 mM in ethylenediaminetetraacetic acid was flushed with g) was treated with 9:1 (vol/vol) anhydrous HF/anisole for 60 oxygen-free nitrogen and stirred with dithiothreitol (200 mg, min at 00. 150 eq) for 10 hr. After more dithiothreitol (50 mg) was added, Binding to Affinity Receptor. The crude peptide (4.38 X the solution was stirred at 400 for 6 hr and gel filtered on a 1.5 106 cpm) was dissolved in 8 M urea buffered to pH 9.0 with 0.1 X 85-cm column of Bio-Gel P-2 with 1% acetic acid to remove M Tris and stirred with dithiothreitol (150 eq) for 18 hr under the dithiothreitol. Exposure to oxygen was minimized to pre- nitrogen, which was freed of oxygen by bubbling through 10% vent reoxidation of the thiol groups. The radioactive eluate was pyrogallol in 1 M NaOH. The reduced mixture was gel filtered lyophilized and dissolved in a nitrogen-flushed buffer (pH 5; on a 1.5 X 30-cm column of Bio-Gel P-2 and eluted with 1% 4.5 ml) 0.05 M in sodium acetate and 0.1 M in potassium chlo- aqueous acetic acid at 20 ml/hr. The eluate absorbance was ride. This solution was applied to a 1.5 X 23-cm column of or- monitored continuously at both 206 and 280 nm with an LKB ganomercurial-agarose (40 ml; 36 ,umol of Hg) equilibrated Uvicord III UV monitor and the radioactivity was measured with the sodium acetate buffer. Elution with this buffer for 15 with a Beckman model 355 liquid scintillation counter. Frac- hr at 6.0 ml/hr removed the nonthiol peptides. Subsequent tions comprising the first peak were pooled and lyophilized to elution of the organomercurial-agarose column with 0.5 M furnish a mixture of the larger peptides (23.5 mg; 2.27 X 106 cysteine for 6 hr at 10.4 ml/hr displaced the Cys-Met-peptides cpm). Part of this peptide mixture (17.1 mg; 1.6 X 106 cpm) was (1.8 X 107 cpm). dissolved in 0.05 M sodium phosphate buffer (pH 6.0; 2.0 ml) After lyophilization and gel filtration, the Cys-Met-peptide without stirring, flushed with oxygen-free nitrogen, and applied (1.5 X 107 cpm) was stirred with cyanogen bromide (160 mg) to a 1.5 X 23-cm column of organomercurial-agarose (38 ml; in aqueous 70% formic acid (6.5 ml) for 11 hr. More cyanogen 35 Mmol of Hg) equilibrated with 0.01 M sodium phosphate bromide (105 mg) in 70% formic acid (2 ml) was added, and buffer (pH 6.0; 250 ml). The column was eluted with this buffer the mixture was stirred for 10 hr and filtered on a 1.5 X 85-cm for 16 hr at 6.0 ml/hr and then with 0.01 M sodium acetate (pH column of Bio-Gel P-2 with 1% acetic acid to give a single ra- 5.0) for 2.8 hr at 30 ml/hr to minimize nonspecific binding. dioactive peak at the void volume. Part of this peptide (4.6 X Separation from Affinity Receptor. The Cys-Met-tetrade- 106 cpm) was lyophilized, dissolved in the reducing buffer, capeptide was eluted with freshly prepared 0.4 M mercapto- reduced with dithiothreitol (75 mg; then 15 mg), gel filtered, ethanol at 6 ml/hr. The absorbance at 280 nm and the relative and lyophilized. Part of the lyophilized peptide (2.9 X 106 cpm) radioactivity were measured. After elution of the Cys-Met- was dissolved in a buffer (pH 5) 0.05 M in sodium acetate and peptide, the baseline was higher due to the absorbance of 0.1 M in potassium chloride and reapplied to the organomer- mercaptoethanol. The peptide fractions were pooled, lyophi- curial-agarose column. Elution with this buffer for 11.5 hr at lized, and desalted by gel filtration to furnish the Cys-Met- 9.9 ml/hr gave histone H4-(1-37) (2.5 X 106 cpm; 87% yield). tetradecapeptide (6.0 mg; 2.28 X 104cpm). Subsequent elution with 0.5 M cysteine afforded uncleaved Removal of Affinity Reagent. The peptide was dissolved in Cys-Met-histone H4-(1-37) (0.4 X 106 cpm; 13%). The histone 70% formic acid (1 ml) and a solution of cyanogen bromide (26 H4-(1-37) gave the expected analysis after acid hydrolysis: Lys mg; 100 eq) in 70% formic acid (0.6 ml) was added. The mix- 6.1 (6), His 0.8 (1), Arg 5.9 (6), Asp 2.0 (2), Thr 1.1 (1), Ser 1.0 ture was stirred for 22 hr, diluted 30-fold with distilled water, (1), Glu 1.2 (1), Pro 1.0 (1), Gly 8.5 (9), Ala 2.1 (2), Val 0.9 (1), lyophilized, and desalted by gel filtration as described above Cys 0.0 (0), Met 0.0 (0), Ile 3.2 (3), Leu 3.1 (3). to afford the purified tetradecapeptide (4.7 mg; 1.53 X 104 cpm). RESULTS AND DISCUSSION Characterization. Amino acid analysis was performed by Ribonuclease-(111-124). At the end of each synthetic cycle hydrolysis in 6 M HCl at 110° for 18 hr followed by ion-ex- during assembly of this carboxyl-terminal tetradecapeptide (17, change separation with a Beckman model 121 analyzer. The 21), the amino groups of any peptide chains that had failed to affinity-purified tetradecapeptide gave the expected analysis couple were permanently blocked by acetylation with [3H] (theoretical yield in parentheses): His 0.8 (1), Asp, 1.8 (2), Ser acetic anhydride. Thus potential deletion peptides were con- 1.1 (1), Glu 1.0 (1), Pro 1.9 (2), Gly 1.0 (1), Ala 1.1 (1), Val 3.1 verted into radiolabeled terminated peptides. After assembly (3) Tyr 1.0(1), Phe 1.1 (1). Enzymatic assay was performed by of the desired peptide was complete, Boc-Met and then Boc- combining the purified peptide with natural ribonuclease- Cys(4-CH30Bzl) were added to the growing chains and the (1-118) and measuring the ribonuclease activity of the resulting protocol of Fig. 2 was carried out. Most of the terminated binary complex towards cyclic cytidine 2':3'-phosphate (17). peptides were eluted near the void volume of the organomer- A parallel assay using ribonuclease-(111-124) highly purified curial-agarose colume (Fig. 3), but some bound nonspecifically by ion-exchange chromatography (17) served as the positive to the affinity column and were eluted with a different buffer. control. Elution with 2-mercaptoethanol displaced the Cys-Met-pep- Histone H441-37)heptatriacontapeptide. The protected tides. The small amount of [3H]acetylated peptide observed in Cys-Met-(1-37)-peptide was assembled on an unbrominated this peak (0.13 mol %) may be explained by slow acid cleavage polystyrene support as described for Cys-Met-ribonuclease- (22) of the benzyl group from serine-123 followed by acetyla- (111-124). Arginine was protected by the Ng-4-toluenesulfonyl tion of its free hydroxyl group. Thus, the organomercurial group and lysine was masked with the N6-2,4-dichloroben- Cys-Met tactic removed essentially all of the terminated pep- zyloxycarbonyl group (18) except for lysines 12 and 16, which tides, which amounted to about 5 mol % of the peptide chains. were blocked by N6-[14C]acetyl and N6-[3H]acetyl groups, After removal ofCys-Met, the tetradecapeptide was combined Downloaded by guest on September 26, 2021 Biochemistry: Krieger et al. Proc. Natl. Acad. Sci. USA 73 (1976) 3163

-20 0.25 a T b 0.01 M No H2 P04 0.05M NoOAc RSH 0 pH 5.0 pH 6.0 x E E 0.15 E Q. 0cq CD 0 OD 0.5- -lo-g IT 1~~~~~~~~~~~~~~~~~~~~~~~ + 0.05 04 I t'

0 0 50 00 50 200 250

Volume (ml) 0 -i-

0 50 100 [3H]Ac 960% 2.5% 1.5% 0 50 100 Volume (ml) FIG. 3. Affinity purification of Cys-Met-ribonuclease-(111-124) Volume (ml) by means of organomercurial-agarose. RSH, 2-mercaptoethanol. FIG. 5. Gel filtration of Cys-Met-histone H4-(1-37) on Bio-Gel P-2 before (a) and after (b) affinity purification on organomercu- with natural ribonuclease-(1-118) to reconstitute 85% of the rial-agarose. ribonuclease activity observed with highly purified synthetic tetradecapeptide (17,.21). nine, but only chains still growing at the end of chain assembly Histone H4-1-37). The-amino-terminal 37-residue peptide will contain serine. By amino acid analysis (Table 2) the Thr:Ser from histone H4 (23) was synthesized by the stepwise solid- ratio was 6.0:1 before the affinity column but was 1.0:1 after. phase method. The radiolabels in this synthesis were in the Thus, although less than 20% of the chains were still growing N6-acetyl groups of lysine 12 and 16, not in the terminating at the end of assembly, essentially all terminated peptides and agent. Since known synthetic difficulties were encountered at other nongrowing chains were removed by the organomercu- several points during this particular synthesis, the organomer- rial-agarose column. curial Cys-Met procedure was considered to be a stringent test The Cys-Met affinity reagent was removed from the desired of the affinity purification strategy as well as an important peptide by cleavage of the Met-Ser peptide bond with cyanogen initial step in purification of the synthetic peptide. About 70% bromide. Gel filtration of the product gave a single radioactive of the radiolabeled peptides lacked Cys-Met and passed directly peak showing constant 3H:14C ratio across the peak. Thus by though the affinity column (Fig. 4). The Cys-Met-peptide was the double-label criterion (24) no terminated peptides longer eluted with aqueous cysteine. The purification achieved is than 22 but shorter than 26 residues were present. The Cys- shown by the gel filtration profiles of the synthetic peptide Met-peptide may not be cleaved if the methionine side chain before and after the affinity column (Fig. 5). Since the first is sterically inaccessible, if the sulfide group suffered air oxi- radiolabeled residue was N6-[3H]acetyllysine-16, most radio- dation to the sulfoxide (25), or if the iminolactone intermediate labeled peptides should be longer than 21 residues. Before the is intercepted before hydrolysis by the hydroxyl group of an affinity column three UV-absorbing peaks were seen in front adjacent serine (or ) residue (26). Any uncleaved of the salt peak. Most of the radiolabeled peptides were present chains were separated from histone H4-(1-37) by a second in the void-volume peak and thus were longer than about 25 passage through the organomercurial-agarose column. About residues. The Cys-Met-peptides retained by the affinity column, All however, eluted as a single peak at the void volume. peptides Table 2. Amino acid composition of synthetic shorter than 25 residues and even some larger peptides were Cys-Met-histone H4-(1-37) during affinity removed. purification with organomercurial-agarose The histone peptide contains a single serine residue at the amino terminus and a single threonine residue at position 30 Amino By- near the carboxyl terminus. The ratio of these diagnostic resi- acid Before products After Theory dues provides an independent criterion of homogeneity. All peptide chains longer than seven residues will contain threo- Lys 2.4 2.2 5.4 6 His 0.5 0.4 0.8 1 0.005 M NaOAc Arg 4.2 4.0 5.3 6 20 - 0.Io M KCI Asp 1.9 1.6 2.0 2 pH 5.0 0.5M Cys Thr* 1.0 0.9 1.0 1 Ser* 0.2 0.0 1.0 1 Glu 1.0 1.0 1.1 1 Pro 1.1 1.1 1.0 1 E Gly 2.7 1.9 8.0 9 10 Ala 1.3 1.2 1.9 2 Val 0.7 0.7 0.9 1 Cys 0.2 0.0 0.9 1 Met 0.2 0.0 0.9 1 Ile 2.9 2.8 3.1 3 Leu 2.1 2.0 2.9 3 Thr:Ser* 6.0 46 1.0 1 0 100 200 Volume (ml) * Threonine values were increased 5% and serine values 8% to com- FIG. 4. Affinity purification of Cys-Met-[Lys([14C]Ac)12, pensate for destruction during acid hydrolysis in 6 M HCl at 1100 Lys([3HJAc)16]histone H4-(1-37) by means of organomercurial-aga- for 18 hr. Thr:Ser ratios were calculated with amino acid data rose. containing one more significant figure. Downloaded by guest on September 26, 2021 3164 Biochemistry: Krieger et al. Proc. Nati. Acad. Sci. USA 73 (1976) 87% of the CNBr-treated peptide passed directly through the Mojsov and Miss Anita Bach for experimental assistance with this column in acetate buffer and thus had undergone removal of peptide. This research was supported in part by Grant AM 01260 from the Cys-Met affinity reagent. the U.S. Public Health Service, by funds from the Hoffmann-La Roche Foundation, by Grant-in-Aid 74-844 from the American Heart Asso- The dimeric Cys-Met tactic ciation, and with funds contributed by the New York Heart Associa- Application of the organomercurial Cys-Met tactic to poly- tion. peptides containing cysteine or methionine residues introduces major complications. If the thiol group of each internal cysteine 1. Erickson, B. W. & Merrifield, R. B. (1976) in The Proteins, eds. is not masked during affinity purification, peptides containing Neurath, H. & Hill, R. H. (Academic Press, New York), 3rd ed., cysteine the amino-terminal Cys-Met moiety will Vol. 2, pp. 255-527. but lacking 2. Brunfeldt, K., Christensen, T. & Roepstorff, P. (1972) FEBS Lett. also bind to organomercurial-agarose. If the sulfide group of 25, 184-188. each internal methionine residue is not blocked, cyanogen 3. Yamashiro, D. & Li, C. H. (1974) Proc. Nati. Acad. Sci. USA 71, bromide treatment will fragment the desired peptide after each 4945-4949. methionine. Thus the cysteine and methionine residues should 4. Takashima, H., du Vigneaud, V. & Merrifield, R B. (1968) J. Am. be protected by groups that survive not only assembly of the Chem. Soc. 90,1323-1325. desired peptide but also each step of the affinity purification 5. Beyerman, H. C., Lie, T. S. & van Veldhuizen, C. J. (1973) in procedure. Peptides 1971, ed. Nesvadba, H. (North-Holland, Amsterdam), None of the presently available protecting groups is entirely pp. 162-164. P. R. & R. satisfactory. For example, even the HF-stable S-acetamido- 6. Hancock, W. S., Prescott, D. J., Vagelos, Marshall, G. (1973) J. Org. Chem. 38,774-781. methyl (Acm) group (27) for cysteine and S-oxide group (28) 7. Gross, E. & Witkop, B. (1962) J. Biol. Chem. 237,1856-1860. for methionine are not compatible with the organomercurial 8. Spande, T. F., Witkop, B., Degani, Y. & Patchornik, A. (1970) Cys-Met tactic. The Cys(Acm) residues are stable during Adv. Protein Chem. 24,97-260. solid-phase synthesis, HF cleavage, thiol reduction, and CNBr 9. Cuatrecasas, P. (1970) J. Biol. Chem. 245,3059-3065. treatment; the Met(O) residues are stable during synthesis, HF 10. Ruiz-Carrillo, A. (1974) in Methods in Enzymology, eds. Jakoby, cleavage, binding to organomercurial-agarose and CNBr W. B. & Wilchek, M. (Academic Press, New York), Vol. 37, pp. cleavage. But the organomercurial Cys-Met tactic will generally 547-552. A. & fail for polypeptides containing Cys(Acm) or Met(O) residues 11. Sluyterman, L. A. Wijdenes, J. (1970) Biochim. Biophys. Acta 200,593-595. because the Acm group is removed by mercuric reagents and 12. Der Terrossian, E., Pradel, L., Kassab, R. & Desvages, G. (1974) the sulfoxide is reduced to the sulfide by the thiols needed to Eur. J. Biochem. 45,243-251. reduce the Cys-Met-peptides and to elute them from orga- 13. Anderson, C. D. & Hall, P. L. (1974) Anal. Biochem. 60,417- nomercurial-agarose. 423. This dilemma could be resolved by replacing the organo- 14. Murayama, A., Raffin, J. P., Remy, P. & Ebel, J. P. (1975) FEBS mercurial column by another affinity receptor. One Cys- Lett. 53, 23-25. Met-peptide molecule can effectively act as an affinity receptor 15. Boegman, R. J. (1975) FEBS Lett. 53,99-101. for another such molecule by dimerization through formation 16. Merrifield, R. B. (1963) J. Am. Chem. Soc. 85,,2149-2154. of a disulfide bond between their amino-terminal cysteine 17. Hodges, R. S. & Merrifield, R. B. (1975) J. Biol. Chem. 250, 1231-1241. tactic uses acetamidomethyl residues. This dimeric Cys-Met 18. Erickson, B. W. & Merrifield, R. B. (1973) J. Am. Chem. Soc. 95, protection for internal cysteine residues and sulfoxide protection 3757-3763. for internal methionine residues during assembly and purifi- 19. Markley, L. D. & Dorman, L. C. (1970) Tetrahedron Lett., cation. After cleavage from the solid support, the free thiol 1787-1790. group of the amino-terminal cysteine is air oxidized to the di- 20. Hodges, R. S. & Merrifield, R. B. (1975) Anal. Biochem. 65, sulfide. Then the Cys-Met-peptide dimer is separated by gel 241-272. filtration from the monomer and any smaller by-products 21. Gutte, B., Lin, M., Caldi, D. G. & Merrifield, R. B. (1972) J. Biol. lacking Cys-Met. After cyanogen bromide cleavage of the Chem. 247, 4763-4767. dimer, the Met(O) residues are reduced with thiol and the 22. Erickson, B. W. & Merrifield, R. B. (1973) J. Am. Chem. Soc. 95, Cys(Acm) residues are deprotected with mercuric acetate to 3750-3756. have 23. Krieger, D. E., Levine, R., Merrifield, R. B., Vidali, G. & Allfrey, provide the desired polypeptide. Preliminary experiments V. G. (1974) J. Biol. Chem. 249,332-334. shown that this dimeric Cys-Met tactic is feasible. 24. Erickson, B. W. & Krieger, D. E. (1976) in Antibodies in Human Several other promising groups can be envisaged in the role Diagnosis and Therapy, eds. Haber, E. & Krause, R. M. (Raven of affinity receptor, affinity group, or binding group. For ex- Press, New York), in press. ample, avidin-agarose (29) could be used as the affinity re- 25. Gross, E. (1967) in Methods in Enzymology, ed. Hirs, C. H. W. ceptor, a biotin derivative such as lipoic acid (30) as the affinity (Academic Press, New York), Vol. 11, pp. 238-255. group, and a photolabile group as the binding group. The two 26. Schroeder, W. A., Shelton, J. R., Shelton, J. B., Robberson, B. & Cys-Met tactics described above, however, have the distinct Appel, G. (1969) Arch. Blochem. Biophys. 130,551-556. advantage of using commercially available components and 27. Veber, D. F., Milkowski, J. D., Denkewalter, R. G. & Hirschmann, well-established chemical reactions. R. (1968) Tetrahedron Lett., 3057-3058. 28. Iselin, B. (1961) Helv. Chim. Acta 44,61-78. We thank Dr. Robert S. Hodges for valuable discussions and for 29. Green, N. M. & Toms, E. J. (1973) Biochem. J. 133,687-700. synthesis and bioassay of the ribonuclease peptide, and Miss Svetlana 30. Green, N. M. (1963) Biochem. J. 89,599-609. Downloaded by guest on September 26, 2021