I n [ . J . Peptide Protein Res. 31, 1988, 8 6 9 7

Applications of BOP reagent in solid phase synthesis Advantages of BOP reagent for difficult couplings exemplified by a synthesis of

[Ala 15]-GRF(1-29)-NHz

ALAIN FOURNIER, CHING-TSO WANG and ARTHUR M. FELIX

Peptide Research Department, Roche Research Center, Hoffmann-La Roche Inc., Nutley, New Jersey, USA

Received 19 May, accepted for publication 30 July 1987

The BOP reagent [benzotriazol-l-yl-oxy-tris-(dimethylamino)phosphonium hexa- fluorophosphate] introduced by Castro et al. [Tetrahedron Lett. (1 975) 14, 121 9-1 2221 is ideally suited for solid phase peptide synthesis. The rate of coupling using BOP compared favorably to DCC and other methods of activation including the symmetri- cal anhydride and DCC/HOBt procedures. BOP couplings using the solid phase procedure proceeded more rapidly and to a greater degree of completion for peptide bond formations that were previously determined to be very slow using the conven- tional DCC method. Stepwise solid phase peptide synthesis using BOP was success- fully utilized for the preparation of the (22-29) and (13-29) fragments of [AlaI5]- GRF(I-29)-NH2. Single couplings with 3 equiv. BOP and Boc-amino acids and 5.3 equiv. of diisopropylethylamine in D M F were used for each cycle. The yields of the fragments were superior and the purities comparable using the BOP procedure (single couplings) to those observed using multiple couplings via the DCC coupling method. A total synthesis of [Ala15]-GRF(1-29)-NH, was also carried out using the BOP procedure (single couplings and 3 equiv. BOP and Boc-amino acids and 5.3 equiv. diisopropylethylamine in D M F for each cycle). Multiple couplings were only required for Boc-Asn-OH due to the proposed formation of Boc-aminosuccinimide during activation. The resultant GRF( 1-29) analog was comparable to a control prepared with multiple DCC couplings under optimized conditions. In a parallel study, un- protected Boc-(hydroxy)-amino acids were successfully coupled with the BOP reagent. However, the number of coupling cycles after the introduction of unprotec- ted hydroxy-amino acid must be minimal (< 10). The use of the BOP reagent with unprotected Tyr in solid phase peptide synthesis was also clearly established.

Key words: BOP [benzotriazol-l-yl-oxy-tris-(dimethylamino)phosphonium hexafluoropho- sphate]; Growth Hormone Releasing Factor; solid phase peptide synthesis

Abbreviations follow the recommendations of the hexafluorophosphate; DIEA, diisopropylethylamine; IUpAC-IUB Commission on Biochemical N ~ ~ ~ ~ - TFA, trifluoroacetic acid; MeOH, methanol, DMSO, clature [Bioehem. 1. 126, 773-780 (1972)l. Additional dimethylsulfoxide; Et, N, triethylamine; DTE, dithio- abbreviations: DCC, dicyclohexylcarbodiimide; HOBt, ethane; BHA, benzhYdrYlamine; CHlCI,. methYlene hydroxybenzotriazole; DMF, dimethylformamide; BOP, chloride, CH, CN, acetonitrile; DMS, dimethylsulfide. benzotriazol-l-yl-oxy-tris-(dimethylamino)phosphonium

86

BOP reagent in solid phase synthesis

reagent in solid phase with a ininiinal side- chain protection scheme [unprotected Boc(hydroxy)-amino acids including Ser and Tyr] was also investigatcd.

The use of dicyclohexylcarbodiimide proposed in 1955 by Sheehan & Hess ( I ) has since been widely adopted for peptide bond formation. However, this reagent is not without any shortcomings (2) and side-reac- tions such as racemization or intramolecular rearrangement of the 0-acylisourea deriva- tive have been reported during the activation step. Nevertheless, DCC has been widely used for peptide bond formation during solid phase peptide synthesis (3). Repetitive cou- plings arc often required for the complete in- troduction of a rcsidue in the peptide chain during solid phase synthesis, resulting in the consumption of protected amino acids and time consuming cycles of synthcsis.

Among the several coupling reagents sug- gested for replacement of DCC, the "BOP reagent" [benzotriazol-l-yl-oxy-tris- (dimethy1amino)phosphonium hexafluoro- phosphate] proposed by Castro YI ul. (4) re- presents an attractive alternative. This is a non-hygroscopic salt, very stable and soluble in the usual organic solvents used in peptide synthesis. The activation of the COOH- function is reported to proceed via the forma- tion of an acyloxyphosphonium derivative, subsequently transformed into the HOBt active ester, which undergoes rapid coupling with the growing peptide chain ( 5 ) . The HOB( needed for this activation step is reported to bc produced rapidly in situ by the BOP reagent itself (5 ) .

BOP reagent has been used in solution peptide synthesis (4, 6, 7) and there are reports of its application in solid phase peptide synthesis for a fragment coupling (8) and for a stepwise synthesis (9). In this man- uscript we report details on the use of BOP reagent in the solid phase procedure and on the advantages for the stcpwise synthesis of peptides using BOP in the solid phase meth- odology. As a first step. we explored the coupling properties of BOP by evaluating its capacity to couple residues found to be par- ticularly difficult when attempted by the con- ventional DCC method. The eRicacy of BOP as a condensing reagent in solid phase peptide synthcsis was demonstrated by the successful stcpwise preparation of [Ah"]-GRF( I .29)- NH2 (10). The feasibility of using the BOP

RESULTS AND DISCUSSION

Dificult coupling rencvions Our prior experience with the solid phase synthesis of GRF analogs (10) using DCC, symmetrical anhydride or DCClHOBt ac- tivation allowed thc identification of residues particularly difficult to introduce into the peptide chain. Investigation of numerous coupling reagents ( I I ) revealed that the BOP reagent introduced by Castro el ul. (4) is ideally suited for peptide bond formation. Therefore, we compared the rate of coupling obtained with DCC or other methods of ac- tivation to that observed with BOP when used for the difficult couplings previously identified in both the GRF( 1-44) and GRF (1-29) series (Table 1). As shown in this Table, the extent of coupling of Boc-Leu''- OH or Roc-Val"-OH proceeds to a greater extent of completion after a first coupling using the BOP procedure (92% vs. 81% and 86% vs. 64%, BOP vs. DCC, respectively). When a second coupling step is carried out using the same excess of reagents, approxi- mately equivalent yields are attained for both activation reagents. These estimates are based on the quantitative ninhydrin reaction ( 12) and were confirmed by analytical hplc of the crude products after HF cleavage reaction.

The BOP coupling method also compared favorably to the DCC/HOBt and symmetri- cal anhydride procedures. Whcn Boc-Am"- OH is activated with BOP, the percentage of coupling was shown to be better than that observed using DCCiHOBt activation (Table I) . A second coupling under the same con- ditions increased the amount of Boc-Asn-OH incorporation by both procedures. Since yields only reached values of approximately 80%, this suggests that this residue is par- ticularly difficult to introduce into the peptidc chain. The faster and more complete coupling using the BOP procedure is unexpected in view of the proposed mechanism bj Castro (5) which suggests that both the BOP and

87

A. Fournier et al.

TABLE 1 Comparison of BOP vs other methods: dijicult couplings

Activation Conditions Time % Coupling Double Coupled Hplc Quant. Nin. YO Coupling

Quant. Nin.

I. Boc-Leut4-OH + H-[Ala15]-GRF( 15-29)-BHA- @ + Boc-[A~~’~]-GRF( I&29)-BHA- @

BOP [2.1 equiv.] DMF 2h 92%

2. Boc-Val”-OH + H-[AlatS]-GRF(1429)-BHA- @ + Boc-[Ala”]-GRF( 13-29)-BHA- @

DCC [2.1 equiv.] CH,Cl, 2h 64 %

DCC [2.1 equiv.] CH, C1, 2h 8 1 Yo

DCC [2.1 equiv.] DMF 2h 0 Yo BOP [2.1 equiv.] DMF 2h 86%

3. BOC-AS~~~-OH + H-GRF(36-44)-BHA- @ + Boc-GRF(3544)-BHA- @I

DCCiHOBt [2.7equiv.] DMF 2 h 55% BOP [2.7 equiv.] DMF 2 h 67%

4. Boc-Gln3’-OH + H-GRF(31-44)BHA- @ + Boc-GRF(3W)-BHA- @ Symm. Anhydride [3.2 equiv.] DMF/CH,Cl, 2h19 75%

BOP [3.2equiv.] DMF 2h 97%

99 yo 95%

97% 94% 0% n.d.“

93% 94%

79% 83%

1988% 97%

”Not determined by hplc but confirmed by amino acid analysis after 6~ HCI-propionic acid hydrolysis.

DCC/HOBt procedures have a common ac- tivated HOBt ester. Although no attempt has been made to reexamine the mechanism of this reaction, it appears that the kinetics for the formation of the HOBt ester may be favored by BOP activation.

As shown in Table 1, the coupling of Boc- Gln3’-OH is also more favored by the BOP method than with the symmetrical anhydride (SA) procedure (97% vs. 75% coupling). Even double coupling using the SA method is not sufficient to obtain a yield comparable to that observed using BOP after a single cou- pling step.

Kinetics of BOP coupling reaction In order to evaluate the suitability of the BOP reagent for other types of difficult peptide bond formation, we evaluated the yield of coupling observed after the incorporation of the hindered residue, Boc-Ile, which often gives a poor coupling rate (13). Table 2 confirms that in two systems Boc-Ile was in- troduced into a model peptide more readily with BOP than by the DCC procedure.

A kinetic study was carried out for the 88

coupling of Boc- ValI3-OH (Fig. 1). The use of 1.5 or 3 equiv. BOP reagent resulted in a rate of coupling much faster than DCC. After 1 h, the percentage yield of coupling with 1.5 equiv. BOP is almost 100% while it takes nearly 24h with 3 equiv. DCC to reach a similar level of Boc-Val incorporation. This result suggests that the BOP reagent can be

TABLE 2 BOP vs other coupling methods: difficult couplings-Ile-Leu

and Ile-Ile

Activation Conditions Time YO Coupling Quant. Nin.

Boc-Ile-OH + H-Leu-Ala-Gly-Val-PAM- @+ Boc-Ile-Leu-Ala-Gly-Val-PAM- R

DCC [1.5 equiv.] CH,CI, 2h 92% BOP [1.5 equiv.] DMF 2h 97 %

Boc-Ile-OH + H-Ile-Leu-Ala-Gly-Val-PAM- @ + Boc-Ile-Ile-Leu- Ala-Gly-Val-PAM- @

DCC [I .5 equiv.] CH,CI, 2 h 83% BOP [1.5 equiv.] DMF 2h 99%

Boc -Vol”-tOOH + HzN-[Alo‘s] -CRF ( 14-29)-BHA- @

I BOP

1 1 1 1 1 1 1 1 1 A 1 “0 3 0 60 90 I20 I50 100 210 240“ 23h

TIME ( m m l

FIGURE I Kinetic coupling study or Boc-Val”-COOH to H2N- [Ala1’l-GRF-(I4-29)-BHA resin using 1.5 (0. 0) or 3 equiv. ( + . x ) Boc-amino acid and coupling reagents, BOP or DCC. Reactions were carried out with BOP in DMF containing 5.3 equiv. DIEA and with DCC in CHZCII.

used for rapid couplings even when the amount of acylating reagent is kept minimal.

Riic*enii,-iction study A model study was designed to evaluate the extent of racemization during stepwise solid phase coupling reactions using the BOP reagent. The racemization-susceptible residue, Boc-Phe-OH (3 equiv.), was coupled to a model peptide-resin, H-Leu-Ala-Gly- Val-PAM under our standard conditions using 3 equiv. BOP reagent in DMF contain- ing 5.3 equiv. DIEA (1.5% v/v). An analyti- cal hplc system was developed* which re- solved L-Phe-Leu-Ala-Gly-Val-OH from D- Phe-Leu-Ala-Gly-Val-OH. From this study it was estimated that < 0.2% racemization occurred during the coupling of either BOC-L- Phe-OH or Boc-D-Phe-OH to the model pep- tide-resin. In a parallel experiment using DCC for the coupling of either Boc-L-Phe- OH or Boc-D-Phe-OH comparable results wcrc obtained and similar trace amounts of racemization could be detected.

* Waters PBondapak C, , (0.39 x 2Ocm) using an eluant

with a gradient of 20-30% (B) in 5min; flow rate I mL/ min; detection, 215 nm. Retention times: L-Phe-Leu-Ala- Gly-Val-OH (8 min); o-Phe-Leu-Ala-Gly-Val-OH ( I 2 niin).

of(A) H,O(O.I2S0/o TFA)-(B)CH,CN(0.125% TFA)

BOP reagent in solid phase synthesis

Synthesis of’fAla‘-’J-GRN1-2Yf-NH, (BOP vs. DCC procedures with Ser protection) Star t ing w i t h Arg(Tos)-BHA-resin (0.313 mmol/g), parallel syntheses of [Ala”]- GRF( I-29)-NH2 were carried out under various coupling conditions. BOP coupling conditions included the use of only single couplings (3 equiv./cycle) and 5.3 equiv. DIEA in DMF and DCC/HOBt couplings were also carried out with 3 equiv./cycle (single couplings). The DCC coupling con- ditions consisted of two or more couplings (2.5 equiv./cycle) until a negative ninhydrin test was observed. Two analytical check points were chosen to evaluate the relative progress of the syntheses; aliquots were removed corresponding to GRF(22-29)- BHA-resin and GRF( 13-29)-BHA-resin, cleaved with HF and the crude products com- pared by analytical hplc. Syntheses werc carried out with or without Ser side-chain protection. Fig. 2 shows the chromatograms of the crude GRF(22-29)-NH2 obtained (using Ser side-chain protection) from the multiple coupling DCC procedure (Fig. 2b). the single coupling DCC/HOBt procedure (Fig. 2d). and the single coupling BOP method (Fig. 20. We observed that the single coupling BOP procedure gave crude product of superior purity to that obtained by the DCC/HOBt procedure and similar purity to that obtained from the multiple DCC coupl- ing procedure.

As a result of the poor results obtained from the DCC/HOBt procedure i t was dis- continued. Parallel syntheses were carried out to the next check point. GRF(13-29)-NHz. using the multiple coupling DCC procedure and the single coupling BOP method. Fig. 3 shows the chromatograms of the crude GRF( 1 3-29)-NH2 obtained (Ser’’, Ser” side- chain protection) from the multiple DCC coupling procedure (Fig. 3b) and the single BOP coupling method (Fig. 3d). The BOP synthesis gave crude product of excellent quality and similar to that obtained from the multiple DCC coupling procedure. The yield of peptide from the BOP synthesis was superior to that from the DCC method (Table 3).

The parallel syntheses were completed and

A. Fournier et al.

s

DCC 5 e q /couplin(

BOP 3 0 eq/cycle

SerZ8 unorotectc

D C C / H O B t 3 0 e q /cycle r"8unprotectc

DCC/HOBt 3 .0eq/cycle !r protected

28

d) NJpa

BOP 3.0 eq /cycle er protected

2e Purified Standard couplings/cy

b, 5 0 9 a f) 4 5 9 9

i E

W 0 N

W u z

c

a m

m a

n 0 v)

J * 10

" A

10 2 0 I

0 30 0 10 20 30 0 10 20 30 0 10 20 30 0 10 20

FIGURE 2

TIME (minutes)

Analytical hplc of crude GRF(22-29)-NH2 obtained under various conditions of coupling with Ser'* side-chain protected and unprotected. Merck Lichrosorb RP-8 column, 5pm (0.46 x 25cm). Eluant: (A) 0.1 M NaCIO,, pH 2.54B) CH,CN. Linear gradient, 2045% (B) in 30min. Flow rate: 1 mL/min, detection 206nm (0.2AUFS).

DCC 2 5eq/coupling

2 2couplings/cycle

BOP 3 Oeq/cycle

Ser" and SerZ8 unprotected

BOP 3Oeq/cycle

Serl'and Ser" protected

Purified Standard

*) 31251rq ') 40-q

k u 10 20 30 40

: 0 N

W u z

LL

v)

a m

a

f u

TIME (minutes)

FIGURE 3 Analytical hplc of crude [Ala'5]-GRF(13-29)-NH, obtained under various conditions of coupling with Ser" and SerZR side-chain protected and unprotected. Merck Lichrosorb RP-8 column, 5pm (0.46 x 25cm). Eluant: (A) 0.1 M NaCIO,, pH 2.5-(B) CH,CN. Linear gradient, 3&50% (B) in 30min. Flow rate: 1 mL/min, detection 206nm (0.2 AUFS).

90

13OP reagent in solid phase synthesis

Purhed Slondord

occ 2.5eq/coupt1ng

2 ? couplmgs /cycle

it I

I i I

I I I t

I

i l i '- .j

FIGURE 4 Analyticid hplc of crude [Alats]-CJRF( I 29)-NHL obtained undcr various conditions of coupling with I y r ' side-chiiin protected and unprotected. Merck Lichrosorb RP-8 column, SItm (0.46 x 25cm). Eluant: (A) 0.1 M NaCIO,. pl l 2.S-(B) Cf1,CN. Lienar gradient, 3S.-60% (B) in 40min. Flow rate: I mllmin. detection: 206nm (0.2 At!FS).

the peptides cleaved from the resin with HF and the crude peptides evaluated by hplc ( Fig. 4). [Ala"]-GRF( I-29)-NH2 from the single BOP coupling procedure** (Fig. 4d) was comparable in purity to that obtained from the multiple DCC coupling method (Fig. 4b). The overall yields as estimated from analytical hplc were also comparable for the two syntheses (Table 3). It should be em- phasized that when DCC was used for activa- tion, conditions were optimized [(2.5 equiv./ coupling), and couplings were repeated until the ninhydrin test was negative (2 2 cou- plings/cycle)J. Moreover, Boc-Gin was incor-

* * Although single BOP couplings were used for each cycle, Boc-Asnu-OH required four successive BOP coupl- ings to achieve the ninhydrin end point. Nevertheless thc coupling of Boc-Asn-OH proceeds better with the BOP reagent than by the DCC/HOBt procedure (Table I ) . The difficulty in coupling Boc-Asn-OH may be explained by i ts tendency to form Boc-aminosuccinimide asa major side product when activated (14).

porated by the single BOP coupling method. On the other hand for the DCC synthesis. couplings with Boc-Asn-OH and Boc-Gln- OH were carried out by the DCC/HOBt and symmetrical anhydride procedures, respec- tively, and 6 equiv. were used for each cou- pling step and cycles repeated t o the nin- hydrin end point.

Table 3 also gives the average percent of coupling per cycle observed during the syn- thesis of [Ala'S]-GRF( I-29)-NH2 and its fragments. It is clear from the Table that the BOP coupling procedure is more efficient than the DCC/HOBt method. This can also be seen in Fig. 5 which plots the Oio coupling for each residue. The coupling yields with DCC/HOBt (with Sef" protected) were ex- tremely variable while the BOP reagent gave nearly quantitative incorporation in to the peptide chain at least through the first 10 coupling reactions. The [Ala'.']-GRF( 1-29)- NH2 obtained from the BOP synthesis was purified by preparative hplc. fully charac-

1) I

TA

BL

E 3

Synt

hesis

of

[Ala

'5]-

GR

F(I

-29)

-NH

2 un

der

dige

rent

con

ditio

ns of p

rote

ctio

n an

d ac

tivat

ion.

Int

erm

edia

te:

GRF

122-

29)-

NH

2 m -

28

R. B

oc-S

r

OH

A

ctiv

atio

n C

ondi

tions

A

vera

ge

Am

ount

of

pept

ide-

C

rude

wei

ght

Wei

ght p

eptid

eb in

Th

eore

tical

YO of

%

cou

plin

g/cy

clea

re

sin

clea

ved

(mg)

(m

g)

crud

e pr

oduc

t (m

g)

yiel

d (m

g)

pept

ideb

7-

X

Bzl

OH

B

zl

OH

B

zl

DC

C19

>

5 eq

uiv.

(C

HzC

12)C

-

DC

C/H

OB

t 3 e

quiv

. (D

MF)

86

D

CC

iHO

Bt

3equ

iv. (

DM

F)

76

BO

P 3e

quiv

. (D

MF)

99

B

OP

3 eq

uiv.

(D

MF)

10

0

750

500

500

500

500

229

I53

I45

I72

153

I30 40

42

70

102

250

52

144

28

144

29

144

49

144

71

INT

ER

ME

DIA

TE

[A

lat5

]-G

RF(

l MY

)-N

H2

18.2

8 B

oc-S

r

OH

A

ctiv

atio

n C

ondi

tions

A

vera

ge

Am

ount

of

pept

ide-

C

rude

wei

ght

Wei

ght p

eptid

eb in

Th

eore

tical

%

of

% c

oupl

ing/

cycl

ea

resi

n cl

eave

d (m

g)

(mg)

cr

ude

prod

uct (

mg)

yi

eld

(mg)

pe

ptid

eb

7-

X

Bzl

DC

C

> S

equi

v. (C

H2C

l2)'

-

500

240

100

240

42

OH

B

OP

3 equ

iv. (

DM

F)

96

610

515

27

263

10

Bzl

BO

P 3 e

qurv

. (D

MF)

98

62

0 47

4 21

7 26

7 81

PRO

DU

CT

: [A

la'5

]-G

RF(

I-29

)-N

H2

9.18

.28

I B

oc-S

r

OH

B

oc-T

r

OH

A

ctiv

atio

n C

ondi

tions

A

vera

ge

Am

ount

of

pept

ide-

C

rude

wei

ght

Wei

ght p

eptid

eb in

Th

eore

tical

%

of

% c

oupl

ingi

cycl

e' re

sin

clea

ved

(mg)

(m

g)

crud

e pr

oduc

t (m

g)

yiel

d (m

g)

pept

ideb

7-

Y

e-

X

Bzl

2,6-

C12

Bzl

D

CC

>

5eq

uiv

(CH

2CI,)

c -

Bzl

O

H

BO

P 3 e

quiv

. (D

MF)

9g

d Bz

l 2,

6-C

l2 Bz

l B

OP

3equ

iv (

DM

F)

Ygd

2000

96

6 73

0 33

8 73

0 33

7

153

1111

14

30

38

3 8

43

37s

12

a D

eter

min

ed f

rom

qua

ntita

tive

ninh

ydrin

rea

ctio

n bE

stim

ated

by

hplc

C

2.5e

quiv

. per

cou

plin

g; 2

cou

plin

gs (m

inim

um)

per

cycl

e un

til n

egat

ive

Kai

ser

ninh

ydrin

tes

t (1

2)

* Boc

-A!n

-OH

re

quire

d 4

coup

lings

(B

OP

proc

edur

e)

BOP reagent in solid phase synthesis

FIGURE 5 Percentage yield of coupling observed For each residue during the syntheses oF [Ala"I-GRF( I 29-NH: and its Fragments 122 29) and ( 13-29) using various con- ditions of coupling and sidc- chain protection. Yields evaluaicd by nicilns or ihc quantitative ninhydrin icsi (12). Asn' required four co nsecut i vc coupling st cps 10 rcach the poini indicated by the asierirk.

terized and shown to be chemically identical to a standard prepared by the DCC solid phase procedure. The biological activity of the two samples (rat pituitary bioassay) were also equivalent. Tryptic digestion gave a frag- mentation of peptides which was examined by hplc (Fig. 6). The tryptic digest from the [Ala'']-GRF( I-29)-NH2 prepared by the DCC procedure (Fig. 6b) and that from the BOP synthesis (Fig. 6d) were identical to a purified standard (Fig. 6a). It was concluded that the BOP coupling procedure offers im- portant advantages over the DCC method. Coupling reactions proceed more rapidly and less excess of protected amino acids are re- q ui red.

Synthesis of [Ala" ]-GRF(1-29)-NH, (BOP vs. DCC' procedures with Ser and Tjtr uriprotected) The incorporation of Ser or Thr in solid phase synthesis without side-chain protection has been reported to leave open the serious danger of 0-acylation and branching of peptide chains (15). This side reaction has even been observed with active ester coupling reactions (16). Although Tyr has been used in the unprotected form when it is near the NH?- terminal end of short peptide sequences (17- 19), the free phenolic hydroxyl has been re- ported to undergo 0-acylation reactions in solution (20, 21).

The feasibility of using the BOP reagent with unprotected Boc-( hydroxy)-amino acids was also investigated during the course of this

work. Therefore, during the synthesis of GRF(22-29bBHA-resin an aliquot from the DCC assemblage was coupled with unprotec- ted Boc-Ser"-OH via the DCC/HOBt procedure. In parallel, an aliquot from the BOP asemblage was also coupled with un- protected Boc-Ser"-OH using the BOP procedure. Fig. 2 shows the chromatograms of the crude GRF(22 - 29)-NH2 obtained (using SeiX unprotected) from the DCC/ HOBt procedure and the BOP method. The peptide synthesized by the BOP procedure with SerZX unprotected (Fig. 2e) was of higher purity than that obtained from the DCC/ HOBt with (Fig. 2d) or without (Fig. 2c) Ser" side-chain protection. This is confirmed from the estimation of the percent of peptide found in the crude preparation (Table 2). I t appears that the yield for the GRF(22- 29)-NH2 syn- thesis using unprotected Boc-Ser"-OH and BOP as a condensing agent (single coupling,! cycle) is comparable to that observed after a DCC synthesis using protected Boc-Scr" (BLI)-OH under optimal conditions ( 2 2 cou- plings/cycle). Table 3 also shows that the best yield was obtained using BOP reagent with protected Boc-Ser'x(Bzl)-OH. I t is particular- ly important to observe from Fig. 5 that the coupling yield with unprotected Boc-Ser''- OH was nearly quantitative by the BOP procedure but only 56% using DCC/HOBt.

As a result of the poor results obtained for the unprotected Ser" using DCC/HOBt couplings, this synthesis was abandoned but the parallel synthesis using BOP coupling of

93

A. Fournier et al.

Purified Standard

'a)

0 10 20 30 40

FIGURE 6

occ 2.5 eq /coupli L 2 couplings,

0 10 20 30 40

BOP 3.0 eq / cycle

Ser9,1e*2e protected Tyr' unprotected

( C )

0 10 20 30 40

TIME (minutes)

BOP 3.0 eq/ cycle

Ser9*1e*2e and Tyr' protected

r

0 10 20 30 40

Analytical hplc of [Alai5]-GRF(l-29)-NH, from various syntheses after enzymatic digestion with trypsin. Waters pBondapak C,, column, 10pm (0.39 x 30cm). Eluant: (A) H,O (0.025% TFA-(B) CH,CN (0.025% TFA). Linear gradient, &25% (B) in 30min. Flow rate: 2mL/min, detection: 206nm (0.5 AUFS).

unprotected Boc-Ser"-OH and Boc-Se?'- OH was continued to GRF( 13-29)-NH2. As noted earlier, the BOP synthesis using the full protection scheme gave product of excellent quality (Fig. 3d). The GRF( 1 3-29)-NH, ob- tained when unprotected Boc-Ser"-OH were introduced serially by the BOP procedure gave crude product that was extremely con- taminated (Fig. 3c). This can also be seen in Fig. 5 which shows a constant decrease in YO coupling. Fig. 5 also shows that the YO yield for the addition of the four residues following Ser" falls drastically suggesting that side- chain acylation becomes more serious and contributes to the detrimental effect created by the presence of free Ser''. Although the introduction of a second unprotected Ser is partly responsible for the low quality of the peptide, the repeated exposure of the un- protected Se?' side-chain to acylation is probably the major reason for the presence of

the large amount of side products. This suggests that unprotected Boc(hydroxy) amino acids can be used with the BOP con- densing agent but the number of coupling cycles after its incorporation must be minimal ( < 10) as observed for the successful synthesis

An aliquot of the [Ala" J-GRF(2-29)- BHA-resin from the BOP assemblage was also coupled with unprotected Boc-Tyrl-OH by the BOP procedure. Although the purity (Fig. 4c) and yield (Table 3) of [Alal']- GRF( 1-29)-NH, were slightly lower, the potential of using the BOP reagent with un- protected Tyr in solid phase peptide synthesis is also clearly established. Structure con- firmation was provided by a tryptic digest of the [Ala15]-GRF( 1-29)-NH2 prepared with Tyr unprotected. The fragmentation of pep- tides was examined by hplc (Fig. 6c) and found to be identical with the standard (Fig.

of GRF(22-29)-NH2.

94

BOP reagent in solid phase synthesis

Charac rerizat ion of' pep rides Crude and purified peptide preparations were analyzed by hplc on Lichrosorb RP-8 ( 5 pin) using a 0.1 M NaCIO, (pH 2.5): acctonitrilc eluant system (10). Purified peptides were also characterized by amino acid analysis after hydrolysis for 24 h at I 10' in 6 N HCI contain- ing 0.1% thioglycolic acid. Fitst atom bom- bardment mass spectrometry and optical rotations were performed with purified preparations.

Synrhesis of"Ala"J-GRF( I-ZY)-NH, usbig BOP method solid phase pepride synrltesis Peptide syntheses were carried out manually by the protocol described in Table 4 using wrist-action shakers. W-Boc protection wits used for all amino acids and sidc-chain protecting groups were: Arg(Tos); Ser( Bzl); Asp(0cHex); Lys(2-CI2); Tyr(2.6-C12 Bzl): Thr(Bz1). After proper washings of the resin following the TFA deprotection and the DIEA neutralization, Boc-amino acid cou- plings using DCC were performed in CH2C12 while DMF was preferred for couplings with BOP reagent (1S0/o v/v DIEA). Every cou- pling step was monitored using the quan-

6a). The general application and potential in solid phase peptide synthesis of using BOP in conjunction with a minimal protection scheme will be reported independently (22).

MATERIALS A N D METHODS

Rmgents and snltvnts Boc-protected amino acids were purchased from Bachem Inc. and their purities were carefully checked before use. The benz- hydrylamine resin (- 0.7 mequiv./g) was also obtained from Bachem Inc. ACS grade methylene chloride, dimethylformamide and methanol were purchased from Fisher Scien- titic or Burdick and Jackson and trifluoro- acetic acid (glass distilled) was obtained from Halocarbon and wcre used without further purification. Diisopropylethylamine was purchased from Pfaltz and Rauer Inc. and distilled from CaO and ninhydrin. DCC and BOP reagent were obtained from Cheniical Dynamics Corp.

Asp(0cHex); Lys(2-CIZ); Tyr(2.6-C12 Bzl); Thr(Bz1). After proper washings of the resin following the TFA deprotection and the

TABLE 4 Siundrrrcl pro~oc.rd /or LI sutihciir q ~ l e usrnX B O P rcqi,yair'

Step Reagent Tinir

I I % DMS!CH:C13 I x I niin

3 I?" DMSiCH?CI, I x I min 1 5ou/b IFA.CHICI, + 1 % DMS ( V V ) I x 20niin

6 10% DIEA/CH:CI, I x Jmin 7 C'H,CI, I x In i in

1 - 50% TFAKHZCI, + 196 DMS (v . 'v) I x I niin

5 C'H?C'I: 4 x I lllill

n 10% DIEA!CH,C12 I x 5 111111 9 CH,CI, I x I111111

10 MtQH 2 x I niin I I CH?CI, I x I 111111

I? DMF 2 x I 111111

13 3 Cqul\.. Boc-AA-COOH;DMF +

I4 DMF ? x I Illill 15 McOH 2 x I niin I6 CH,CI, 3 x In i in

3 cquiv. BOP reagent DMF + 5.3 cquiv. DlEA ( I S?U v $ 1 I x 120niin

"Solvent For all washings and couplings wcrc measured to volumes of 10- 20mL g rcsin.

A. Fournier et al.

titative ninhydrin test described by Sarin et al. (12).

After neutralization and washing (Steps 6-12, Table 4) the benzhydrylamine (BHA) resin (80g, - 0.7mequiv./g resin) was charged into a reaction vessel. A solution of Boc-Arg(Tos)-OH (48 g, 0.1 12 mol, 2 equiv.) in DMF (l00mL) was diluted with CHzC1, (600mL) and DCC (23.1 g, 0.112mo1, 2 equiv.) in CH2Clz (100 mL) were added to the reaction vessel. The reaction was carried out overnight and the resin was carefully washed with DMF, MeOH, and CH2C1,. The resin was dried in vacuo and a sample was used for quantitative amino acid analysis. The sub- stitution of Boc-Arg(Tos) was determined to be 0.3 13 mmol/g resin. Boc-Arg(Tos)-BHA resin (2g, 0.626mol) was deprotected and neutralized with DIEA according to the protocol described in Table 4. After the washing steps, a solution of Boc-Ser(Bz1)-OH (554mg, 1.88mmo1, 3 equiv.) in DMF (20mL) and a separate solution of BOP reagent (844 mg, 1.88 mmol, 3 equiv.) also in DMF (20 mL) were introduced into the reactor. To the resin suspension was added 0.6mL (3.32mmo1, 5.3 equiv.) of DIEA (final concentration of 1.5% DIEA) and the reac- tion carried out for 2 h at room temperature. After the coupling, the resin was washed and was ready for the next coupling cycle. All the following residues were added to the growing peptide chain by the same protocol, using single couplings, with the exception of Boc- Asn' which required four successive coupling steps using 3 equiv. of reagents each time.

Cleavage of the peptide from the support After the last amino acid was introduced, the TFA deprotection was executed. The TFA salt peptide-resin (0.73 g) was dried overnight in vacuo and placed in the reaction vessel of a liquid H F apparatus (Protein Research Foundation). HF (7-8 mL) was added to the vessel already containing 730 pL of dithio- ethane (DTE) as scavenger. The cleavage reaction proceeded for 2 h at 0" and H F was rapidly evaporated in vacuo. The resin was washed with ethyl acetate (EtOAc) and the peptide was extracted with TFA, evaporated in vacuo, and the remaining oily material pre-

cipitated with ether to give 337 mg of product. A portion of the crude material (200mg)

was purified by preparative hplc on 2 Syn- chropak RP-P reverse-phase columns (10 x 250 mm) connected in series using an eluant of (A) H,O (0.025% TFA)<B) CH,CN (0.025% TFA). The peptide was eluted with a linear gradient of 2 W 5 % (B) in 2 h at a flow rate of 2 mL/min and detection at 208 nm. Analytical hplc of the individual fractions was carried out using a Merck Lich- rosorb RP-8 column (0.46 x 25cm) and an eluant of (A) 0.1 M NaCIO, (pH 2.5HB) CH,CN in a linear gradient mode (30% (B) to 55% (B) in 30 min) at a flow rate of 1 mL/ min and detection at 206 nm (0.2 AUFS). The fractions corresponding to the purified peptide (retention time: - 55 min) was lyo- philized. Although no attempt was made to optimize the yields for the BOP-synthesis of [AlaI5]-GRF( I-29)-NH2 an overall yield of 12% was estimated from the analytical hplc of the crude product.

The [Ala'']-GRF( 1-29)-NHz obtained from the BOP synthesis was characterized by amino acid analysis after a 24 h hydrolysis at 110" with 6~ HCI containing 0.1% thiogly- colic acid: Arg(3):2.96, Ser(3):2.66, Met(1):0.87, Ile(2): 1.75, Asx(3):2.92, Glx(2):2.08, Leu(4):4.13, Lys(2): 1.94, Ala(4):3.94, Val(l):0.90, Tyr(2): 1.80, Thr(l):0.87, Phe(l):0.91. The peptide was also found to be homogeneous by analytical hplc and the structure was confirmed by (+) FAB [2000v] mass spectrometry [(M + H)+; calculated 3372.9; observed, 3373. I]. Specific rotation: [u]E - 54.02" (c 1.06; 0.2 N AcOH). Circular dichroism study in 50% TFE-phosphate buffer (pH 7.4) and 75% TFE-phosphate buffer (pH 7.4) showed that the molecular ellipticity curve of the BOP-synthesized [Ala"]-GRF( I-29)-NH, was superimposable to that measured with the peptide obtained from DCC synthesis.

The peptide was also analyzed by hplc after digestion of the [Ala15]-GRF(I-29)-NH, with trypsin (Fig. 6). Identical chromatograms were obtained with the purified standard (Fig. 6a), and the peptides obtained from the DCC synthesis (Fig. 6b) and the BOP- synthesis with Tyrl unprotected (Fig. 6c) or

96

with Tyrl protected (Fig. 6d). Finally the biological activity of the BOP-synthesized [Ala”]-GRF( I-29)-NH2 was evaluated. Both the DCC-synthesized and the BOP- synthesized peptides had identical potency for the secretion of growth hormone in the in v i m superfusion of rat pituitary cell bioassay (Brazeau, P., personal communication).

ACKNOWLEDGMENT

’l’he authors thank Dr. F. Scheidl and his staff for the optical rotations and amino acid analyses. We also thank Dr. V. Toome and his staff for the circular dichroism spectra. Dr. W. Benz for carrying out mass spectroscopy and Drs. T. Mowles and P. Brazccau for the bioassay.

REFERENCES

I

-I L

3

3

5.

6.

7.

n.

Y.

Sheehan. J.C. & Hcss, G.P. (1955) J. Ant. Chent. Soc. 77, 1067 Bodanszky, M. (19x4) Principlev o/ Pupticlo S.w- fhrsis. p. 37. Springer-Vcrlag. New York Biirany, G . & Merrifield. R.B. (1980) in 17tr Pep- riclc:~ (Gross. E. & Meienhofer, J.. eds.), vol. 2. pp. I 2x3. Academic Press, Ncw York Castro, B., Dormoy. J.R.. Evin. G. & Selve. C . (1975) T(wuhrdron Lett.. 1219 1222 Castro. B.. Dormoy. J.R.. Evin, G. & Selve. C . ( 1977) J. C/tent. Rev. (S). I82 Castro, B.. Dormoy. J.R.. Dourtoglou. R.. Evin. G.. Selve. C.& Zregler, J.C. ( 1976) .Smh~*.vi.v. 751- 752 Npuyen, D.L.. Seyer. R., Heitz. A. & Castro. B. (1985) J. Chetn. Soc. Perkin 7‘rcrtr.v. 1. 1025-1031 Rivaille. P., Gautron, J.P.. Castro, B. & Milhaud, G . ( 1980) Tiwulteclron 36, 34 13- 34 19 Audousset-Pucch, M.P.. Dufour. M.. Kevran. A,. Jarrousse. C. , Castro. B.. Bataille, D. & Martinez, I . (1986) PEBS Lefr. 200, 181-185

BOP reagent in solid phase synthesis

10. Felix. A.M., Heimer. E.P., Mowles, T.F.. Elsenbeis. H.. Leung. P.. Lamhros. T.J. Ahmad. M.. Wiing. C-T. & Brazeau. P. ( 1987) frcic.. I Y f h E i i r r p ~ m fepridc Swipo.\iunt (Theodoropoulos. D.. ed.), pp. 4x1 484. W. de Gruytcr & Co.. Berlin Felix. A.M.. Wang. C.-T.. Heirner, E.P.. Fournicr. A.. Bolin, D.R.. Ahmad, M.. I,iiinhros. T.J.. Mowles. T.F. & Miller. L.F. (1987) PIw. /Odt . 4 ~ rriwn fc.pfiilu Smposiunt (in press)

12. Sarin. V.K.. Kent, S.B.H.. Tarn. J.P. & Mcrrilicld. R.B. (1981) And. Bioclient. 177, 147 I57

13. Gross. E. & Meienhofer. J . (1979) in Tlto P~yrid(~.v (Gross. E. & Meienhorer. J.. eds.). vol. I. pp. I -64. Academic Press, New York

14. Castro. B.. Evin. G . . Selve. C . & Seycr. R. (1977) Sl.trthesrs. 4 I 3

IS. Stewart. J.M.. Mizoguchi. T. & Wt~ilcy. D W. (1967) Abstr.. IS3rd Nail. Meet.. Am Cliem Soc. No. 0 206

16. Garner. R.. Schafer. D.J. & Young.G.T. (1971) 111

North Holliind Publishers, Amsterdam 17. I-lruby. V.J.. llpson. D.A. & Agarwal. N.S. (1077)

.I. Org. C‘hmi. 42, 3552- 3566 18. Meienhofer. J.. Trrcciak. A.. Havran. R.T. &

Walter, R. (1970) J. .4ni. C h m . .So(,. 92, 719Y 7202 19. Livc. D.H., .\gosta. W.C. &Cowburn. D. (1977)J.

Org. ( * h c ~ ~ . 42, 3556-3561 20. Paul, R. (1963) J . Org. Chwi . 28, 2.16 ,737 21. Ramachandran. J. & Li. C.-H. (1963)J Org. C/teni.

28, 173 177 22. Fournier. A.. DiInho. W. iyc Felix. A.M. (in

preparation)

I I .

Pcpricie.c 1969 (Scoffone, E.. cd.), pp. I02 lox.

Address:

Dr. .4rr/iur hi. Frli.1

Peptide Research Department Roche Research Center Hoffrnann-la Roche Inc Nutley. NJ 071 10 USA

97