3-(4-Hydroxymethylphenylsulfanyl)propanoic acid (HMPPA) as a new safety catch linker in solid phase peptide synthesis

Article abstract of DOI:10.1016/j.tetlet.2006.04.138

A new safety catch linker, 3-(4-hydroxymethylphenylsulfanyl)propanoic acid (HMPPA), is described for use in solid phase peptide synthesis. The linker is readily synthesized from commercially available chemicals in a more cost efficient way compared to similar reported linkers. The HMPPA linker is easily attached to an amino derivatized solid support followed by on-resin oxidation of the thioether to sulfoxide, thereby making the linker very stable towards strong acid treatment. Final resin cleavage is performed by reductive acidolysis.

Full text of DOI:10.1016/j.tetlet.2006.04.138

Tetrahedron Letters 47 (2006) 5829–5832
3-(4-Hydroxymethylphenylsulfanyl)propanoic acid (HMPPA) as
a new safety catch linker in solid phase peptide synthesis
*
´
Mikael Erlandsson and Anders Unden
Department of Neurochemistry, Stockholm University, 10691 Stockholm, Sweden
Received 28 February 2006; revised 14 April 2006; accepted 26 April 2006
Available online 27 June 2006
Abstract—A new safety catch linker, 3-(4-hydroxymethylphenylsulfanyl)propanoic acid (HMPPA), is described for use in solid
phase peptide synthesis. The linker is readily synthesized from commercially available chemicals in a more cost efficient way com-
pared to similar reported linkers. The HMPPA linker is easily attached to an amino derivatized solid support followed by on-resin
oxidation of the thioether to sulfoxide, thereby making the linker very stable towards strong acid treatment. Final resin cleavage is
performed by reductive acidolysis.
ꢀ 2006 Elsevier Ltd. All rights reserved.
The synthesis of complex peptides often requires protec-
tive groups that can be selectively removed. For a num-
ber of reasons, final resin detachment and side-chain
deprotection should be performed using anhydrous
acid.1 However, this requirement excludes many acid-
cleavable protective groups, as the resin linkage could
be jeopardized. The use of a safety catch linker is an
attractive method to circumvent these problems, as most
safety catch linkers are stable towards the reaction con-
ditions commonly used in solid phase peptide synthesis.
A safety catch linker is cleaved in a two-step process
where the first step activates the linker and the second
step involves the actual cleavage. Safety catch linkers
based on TFA-stable arylalkyl sulfoxides have previ-
ously been reported.2
size cyclic peptides on solid support.4 A conceptually
similar approach was later presented by Getman, who
anchored 4-mercaptobenzylalcohol to Merrifield resin.5
The resulting p-alkylthiobenzyl alcohol resin can be re-
garded as a sulfur analogue of the original Wang resin.6
When oxidizing the thioether to the sulfoxide, the lin-
ker’s stability towards acids is dramatically increased.
The linker was finally cleaved by the ‘low-TFMSA’
method.7 Another arylalkyl sulfoxide based linker for
the synthesis of peptide acids is the DSB-linker de-
scribed by Kiso et al.8 Other linkers, such as the SCAL-
and DSA-linkers, were developed for the synthesis of
peptide amides.9
In projects related to the synthesis of various cyclic pep-
tides, we came to the conclusion that the arylalkyl sulfox-
ide safety catch concept had many advantages but the
currently described linkers had some specific disadvan-
tages. The phenolic linkage of the Marshall linker cannot
be expected to be stable during the repetitive nucleophile
treatments needed for Fmoc protecting group removal.
The linker described by Getman offers a simple route
to introduce the linker on a solid support, but the start-
ing material is extremely expensive. Anchoring the first
residue to the DSB linker requires acylation of a second-
ary alcohol, which increases the risk of racemization.
Marshall and Leiner coupled 4-mercaptophenol to
Merrifield resin, with subsequent anchoring of the first
amino acid to the phenolic oxygen and repetitive solid
phase synthesis.3 Oxidation of the thioether to the sulfone
dramatically increased the linker’s sensitivity towards
nucleophiles. Treating the resin with the sodium salt of
an amino acid resulted in both linker cleavage and C-
terminal extension. This resin was later used to synthe-
Abbreviations: AcN, acetonitrile; EDT, 1,2-ethanedithiol; HMPPA(O),
3-(4-hydroxymethylbenzenesulfanyl)propanoic acid; p-ether, petro-
leum ether; TFAA, trifluoroacetic anhydride; TFE, trifluoroethanol;
TMSBr, bromotrimethylsilane.
This led us to develop a new arylalkyl sulfoxide safety
catch linker that could be synthesized in high yields
using inexpensive reagents in a few simple manipu-
lations. A sulfur analogue of the frequently used
*
Corresponding author. Tel.: +46 8 164117; fax: +46 8 161371; e-mail:
andersu@neurochem.su.se
0040-4039/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.04.138

5830
M. Erlandsson, A. Unde´n / Tetrahedron Letters 47 (2006) 5829–5832
3-(4-hydroxymethylphenoxy)propionic acid linker is
particularly attractive as this, and similar linkers, can
be introduced onto a solid support by simple acylation
of amino derivatized resins.10
RP-HPLC revealed that only 60% of the product was
the sulfone and the remaining 40% was the sulfoxide.
Attachment of the first residue to the linker can be
achieved using standard protocols for derivatizing
hydroxymethyl resins.6
Starting from the inexpensive and commercially available
4-(methylthio)benzaldehyde 1, the linker 3-(4-hydroxy-
methylphenylsulfanyl)propanoic acid (HMPPA, 5) was
synthesized in three simple and high yielding steps,
as outlined in Scheme 1.11 It is clear that a more direct
route to 5 would involve direct alkylation of commer-
cially available 4-mercaptobenzaldehyde or 4-mercapto-
benzyl alcohol. However, the cost of these reagents forced
us to start from 1.
A valuable application of the HMPPA safety catch
linker is in the synthesis of side-chain-to-side-chain
cyclic peptides. This was demonstrated by synthesizing
the model peptide Fmoc-K(Boc)F-D-APE(OtBu)G-
HMPPA(O) on aminomethyl resin. This model peptide
has previously been used in on-resin lactamization
studies.18 The peptide was treated for on-resin side-
chain lactamization after acidolytic cleavage of the
side-chain protective groups on Lys and Glu. Subse-
quent resin cleavage and analysis by RP-HPLC and
MALDI-TOF revealed that both the linear and the cyc-
lic peptides had been synthesized in very good purity.
The only by-product detected after on-resin lactamiza-
tion was the cyclic dimer.
An important intermediate in the synthesis of the
HMPPA linker is the a-trifluoroacetoxy sulfide moiety
3. The intermediate can be obtained through a Pum-
merer rearrangement by treating 2 with trifluoroacetic
anhydride, yielding 3, which in turn is readily hydro-
lyzed to the corresponding 4-mercaptobenzaldehyde.12
However, under these conditions, the aldehyde moiety
in 3 is simultaneously transformed into a geminal bis-
trifluoroacetate. This intermediate is quite unstable,
and may undergo polymerization through intermolecu-
lar thioacetalization.13 Polymerization can be avoided
by performing the Pummerer rearrangement under mild
reaction conditions and by adding the electrophile,
3-iodopropanoic acid, during hydrolysis.14 This also
provides for a simple work-up, as 4 can be separated
by simple extraction. The HMPPA linker is finally
obtained by reducing 4 with NaBH4 in EtOH, which,
after work-up, crystallizes readily from EtOAc/petro-
leum ether.
Peptides synthesized on the oxidized HMPPA linker can
be resin cleaved in a one-pot reductive acidolysis reac-
tion.19 By adding 0.2 M EDT/0.1 M TMSBr in TFA
to the resin bound peptide, the linker sulfoxide is re-
duced to the corresponding thioether, thereby making
the ester linkage susceptible for acidolysis.20 The reduc-
ing agent is the bromine anion and EDT functions as a
bromine scavenger.
Resin cleavage studies showed that cleavage efficiency is
about 90% as measured by quantitative ninhydrin test.21
Furthermore, the overall yield in the synthesis of the lin-
ear peptide KF-D-APGE-OH was 52% as calculated
from resin loading.22
The linker is easily attached to an amino derivatized
solid support in the form of an active ester.15 The thioether
is oxidized to the corresponding sulfoxide by treating
the solid supported linker with 10 equiv of H2O2 in
TFE/DCM (2:1) for 4 h.16 TFE prevents further oxida-
tion of the sulfoxide by forming strong hydrogen bonds,
thereby diminishing the nucleophilicity of the sulfoxide
for further oxidation by H2O2.17 This was also con-
firmed experimentally by treating the model peptide
Boc-K(Boc)FAFA-HMPPA(O)-AFK(Fmoc)LG-Merri-
field resin, with 100 equiv of H2O2 in TFE/DCM (2:1)
for 7 days. After cleaving the model peptide from the
solid support by a 24 h treatment with neat TFA,
During reductive acidolysis, benzyl-type protective
groups commonly used in solid phase peptide synthesis
are also cleaved.23 This is especially interesting, as it
would open the possibility of using the Boc/Bzl synthetic
strategy when synthesizing peptides on a solid support,
yet avoiding the use of hazardous super acids for final
resin cleavage. However, more acid resistant side-chain
protective groups, in particular Arg(Tos) and Cy-
s(MeBzl), are not cleaved by reductive acidolysis.24 We
therefore speculated that these protecting groups could
be removed via a high-acid SN1 mechanism on-resin,
HO
CHO
CHO
CHO
CHO
a
b
c
S
O
S
S
S
S
O
1
2
O
O
O
CF3
OH
OH
3
4
5
Scheme 1. Synthesis of the HMPPA linker. Reagents and conditions: (a) H2O2 (2 equiv) in TFE/DCM (2:1), 4 h; (b) 3 equiv TFAA in AcN/ 2 equiv
2,6-lutidine, 1 h, 0 ꢁC ! rt, then 1.0 equiv 3-iodopropanoic acid in TEA/MeOH (1:1); (c) NaBH4 in EtOH.

M. Erlandsson, A. Unde´n / Tetrahedron Letters 47 (2006) 5829–5832
5831
without cleaving the product from the HMPPA(O) lin-
Acknowledgements
ker, as the linker should be stable towards even very
strong acid treatment.25 Studies on model peptides con-
taining the oxidized HMPPA linker in homogeneous
solution suggest that the linker is very stable towards
strong anhydrous acids and could potentially be used
when cleaving side-chain protective groups through an
SN1 mechanism in Boc chemistry. Relatively low levels
of linker cleavage (14%) were observed after a 3 h treat-
ment with TFMSA/TFA (1:9) at 0 ꢁC. However, com-
mon scavengers had a negative effect on the linker
stability and integrity. When treating the model peptides
with TFMSA/TFA and p-cresol or anisole (1:9:1), a sin-
gle by-product was formed. Although the structure of
this side-product was not determined, it is likely that it
represents S-arylation of the HMPPA linker by the
scavenger as previously reported.26 S-arylation pre-
vents subsequent resin cleavage by reductive acidolysis.
The scavengers thioanisole and thiophenol, on the other
hand, resulted in reduction of the HMPPA sulfoxide
and subsequent cleavage of the linker.27 These initial re-
sults were somewhat disappointing but suggested that if
a scavenger can be identified that does not react with the
linker under strong acidic conditions, it should be possi-
ble to selectively cleave most side-chain protecting
groups commonly used in the Boc/Bzl protective groups
strategy, while the peptide is still attached to solid sup-
port. Final resin cleavage is performed by a relatively
mild method; hence, this approach would circumvent
many of the problems associated with the use of super
acids for final resin cleavage in Boc chemistry.
The authors thank Kristina Romare at the Department
of Organic Chemistry, Stockholm University, for
providing the NMR spectra.
References and notes
1. Barany, G.; Merrifield, R. B. In The peptides; Erhart,
G., Meienhofer, J., Eds.; Solid-phase Peptide Synthesis;
Academic Press: New York, 1980; Vol. 2, pp 1–
284.
2. Guillier, F.; Orain, D.; Bradley, M. Chem. Rev. 2000, 100,
2091–2157.
3. Marshall, D. L.; Liener, I. E. J. Org. Chem. 1970, 35, 867–
868.
4. Flanigan, E.; Marshall, G. R. Tetrahedron Lett. 1970, 27,
2403–2406.
5. Getman, D. P.; Heintz, R. M. Chem. Abstr. 110:213351,
Eur. Pat. Appl. 1988.
6. Wang, S.-S. J. Am. Chem. Soc. 1973, 95, 1328–1333.
7. Tam, J. P.; Heath, W. F.; Merrifield, R. B. J. Am. Chem.
Soc. 1986, 108, 5242–5251.
8. Kiso, Y.; Fukui, T.; Tanaka, S.; Kimura, T.; Akaji, K.
Tetrahedron Lett. 1994, 35, 3571–3574.
´
9. Patek, M.; Lebl, M. Tetrahedron Lett. 1991, 32, 3891–
3894; Kimura, T.; Fukui, T.; Tanake, S.; Akaji, K.; Kiso,
Y. Chem. Pharm. Bull. 1997, 45, 18–26.
10. Albericio, F.; Barany, G. Int. J. Pept. Protein Res. 1985,
26, 92–97.
11. 4-Methylthiobenzaldehyde (5 g, 33 mmol) was dissolved in
TFE/DCM (2:1, 75 ml), the solution was cooled on ice
and then 30% H2O2/H2O (7.45 ml, 66 mmol) added. The
reaction was left on ice for 30 min and then at ambient
temperature for 4 h. Subsequent TLC analysis (EtOAc/
p-ether 1:1) indicated complete oxidation. Excess H2O2was
decomposed by adding Pd/C. When the gas evolution had
stopped, the catalyst was filtered and the solvents removed
in vacuo. The resulting white solid (2, 5.5 g, 33 mmol) was
dissolved in AcN (100 ml) containing 2,6-lutidine (11.5 ml,
99 mmol), and then cooled on ice under a blanket of N2.
TFAA (9.14 ml, 66 mmol) in AcN (25 ml) was added
dropwise. Upon completion, the reaction was allowed to
reach rt and then left for 30 min. The yellowish solution
was added to a pre-cooled solution of TEA/MeOH (1:1,
100 ml), containing 3-iodopropanoic acid (6.6 g, 33 mmol)
and then left at ambient temperature for 4 h. The solvents
were removed in vacuo and the resulting oil was taken up
in a saturated Na2CO3 solution (100 ml) and extracted
with EtOAc (3 · 75 ml). The solution was acidified (concd
HCl) resulting in voluminous precipitation. The precipi-
tate was filtered, washed with cold 1 M HCl(aq) and dried
in vacuo, yielding 4 (5.73 g, 83%). The crude product was
dissolved in ethanol (50 ml) and cooled on ice. NaBH4
(3.1 g, 82 mmol) was added and the reaction left for 5 h.
Water (100 ml) was added, the ethanol removed in vacuo
and the pH adjusted to 3 by adding concd HCl. The
product was filtered and washed with 1 M HCl and dried,
yielding 5 (5.06 g, 88%). A sample of the linker was
crystallized from EtOAc/p-ether for NMR analysis. 1H
NMR 5: (400 MHz, DMSO-d6): d 7.28, 7.26 (2d, 4H,
J = 5.8, 8.6), 4.44 (s, 2H), 3.08 (t, 2H, J = 7.1), 2.48 (t, 2H,
J = 3.8). 13C NMR: (100 MHz, DMSO-d6): 28.3, 33.9,
39.1, 62.5, 127.4, 128.9, 133.4, 140.7 and 172.8.
Relatively few studies have been carried out on side-
reactions associated with the reductive acidolysis meth-
od presented here. It can, however, be expected that
the indole side-chain functional group in Trp residues
would be the most vulnerable residue for modifications
during reductive acidolysis, in particular to bromina-
tion. Therefore, the model peptide Fmoc-K(Boc)W-
(Boc)FAK(Boc)PVA-HMPPA(O)-aminomethyl resin
was synthesized using standard Fmoc/tBu protective
group strategy and subsequently resin cleaved by reduc-
tive acidolysis. Analysis by RP-HPLC and MALDI-
TOF revealed that the KWFAKPVA-OH peptide was
obtained as the major product (87%). Two minor by-
products were detected, that is, the 2-indolylindoline
dimerization product (7%) and EDT scavenger addition
products (6%).28 However, it was noted that higher
amounts of by-products were isolated if the removal of
TFA was slow during work-up. On the other hand, it
is noteworthy that no brominated by-products could
be observed. It can, therefore, be suggested that the
side-reactions observed are more a reflection of the use
of EDT and a relatively weak acid during final cleavage,
rather than the reductive acidolysis method.
In conclusion, although a large number of linkers are
available for the solid phase synthesis of peptides, rela-
tively few of these are safety-catch linkers. The HMPPA
linker might, therefore, prove to be a valuable new tool
in the synthesis of a number of complex peptides, which
require selective removal of the side-chain protecting
groups during synthesis.
12. Pummerer, R. Chem. Ber. 1910, 43, 1401–1412; Young, R.
M.; Gauthier, J. Y.; Coombs, W. Tetrahedron Lett. 1984,
25, 1753–1756.

5832
M. Erlandsson, A. Unde´n / Tetrahedron Letters 47 (2006) 5829–5832
13. Gryko, D. T.; Clausen, C.; Lindsey, J. S. J. Org. Chem.
1999, 64, 8635–8647.
14. Sugihara, H.; Tanikaga, R.; Kaji, A. Synthesis 1978, 12,
881.
15. Rich, D. H.; Singh, J. In The Peptides; Gross, E.,
Meinenhofer, J., Eds.; The Carbodiimide Method; Aca-
demic Press: New York, 1979; Vol. 1, pp 241–261.
16. Ravikumar, K. S.; Kesavan, V.; Crousse, B.; Bonnet-
Delpon, D.; Begue, J.-P. Org. Synth. 2003, 80, 184–189.
17. Curci, R.; DiPrete, R. A.; Edwards, J. O.; Modena, G. J.
Org. Chem. 1970, 35, 740–745.
18. Cavallaro, V.; Thompson, P.; Hearn, M. J. Pept. Sci.
1998, 4, 335–343.
19. Beck, W.; Jung, G. Lett. Pept. Sci. 1994, 1, 31–37.
20. Landini, D.; Modena, G.; Montanari, F.; Scorrano, G. J.
Am. Chem. Soc. 1970, 92, 7168–7174.
DCM (1:1). Next, N-termini Fmoc protected residues
were coupled by OBt ester activation. The Fmoc groups
were removed by treatment for 20 min with 20% piperi-
dine in DMF. 50 mg of the resin bound model peptide was
treated for side-chain-to-side-chain lactamization by add-
ing 2 equiv of BOP and HOBt, respectively, and 4.2 equiv
of DIPEA in NMP to the side-chain deprotected peptide.
The peptide was washed after 4 h and the procedure
repeated once. The N-terminal Fmoc group was removed
and the cyclic peptide, and 50 mg of the starting linear
peptide, was resin cleaved by reductive acidolysis (0.2 M
EDT and 0.1 M TMSBr in 3 ml TFA) for 2 h. The
solution was filtered and washed with TFA (·3) and the
solvents then removed in vacuo. The peptides were taken
up in water and extracted with diethyl ether (·4). Both
peptides were analyzed by RP-HPLC and MALDI-TOF.
To estimate overall yields, the linear peptide was partially
purified by semi-preparative RP-HPLC, lyophilized and
analyzed gravimetrically, giving a total, overall yield of
52% as measured from initial resin loading.
21. Sarin, V. K.; Kent, S. B. H.; Tam, J. P.; Merrifield, R. B.
Anal. Biochem. 1981, 117, 147–157.
22. The model peptide was synthesized using the standard
Fmoc/tBu protecting group strategy on the HMPPA
linker, attached on aminomethyl polystyrene. In brief: The
HMPPA linker was coupled to 1 g of aminomethyl
polystyrene (0.62 mmol/g), by HOBt ester activation
(preformed by mixing 2.3 equiv HMPPA linker, HOBt
and DIC, respectively, in DCM 30 min at 0 ꢁC) for 4 h.
After washing and drying, gravimetrical analysis gave a
resin loading of 0.553 mmol/g resin. The thioether was
oxidized on-resin, by adding 10 equiv of H2O2 (0.63 ml) in
TFE/DCM (2:1, 18 ml) for 4 h. 200 mg of the resin was
transferred to another vessel and the first residue was
attached by OAt ester activation (formed by premixing
10 equiv of Boc-Gly, HOAt and DIC, respectively, and
0.05 equiv DMAP in DCM for 30 min at 0 ꢁC). The Boc
group was removed by treatment for 20 min with TFA/
23. Yajima, H.; Fujii, N.; Funakoshi, S.; Watanabe, T.;
Muryama, E.; Otaka, A. Tetrahedron 1988, 44, 805–819.
24. Hughes, J. L.; Leopold, E. J. Tetrahedron Lett. 1993, 34,
7713–7716.
25. Sakakibara, S.; Shimonishi, Y.; Kishida, Y.; Okada, M.;
Sugihara, H. Bull. Chem. Soc. 1967, 40, 2164–2167.
26. Funakoshi, S.; Fujii, N.; Akaji, K.; Irie, H.; Yajima, H.
Chem. Pharm. Bull. 1979, 27, 2151–2156.
27. Yajima, H.; Fujii, N. In The Peptides; Gross, E.,
Meienhofer, J., Eds.; Acidolytic Deprotection Procedures
in Peptide Synthesis; Academic Press: New York, 1983;
Vol. 5, pp 65–109.
28. Fontana, A.; Toniolo, C. Fort. Chem. Org. Nat. 1976, 33,
309–449.