Replacement of the Disulfide Bridge in a KLK3-Stimulating Peptide Using Orthogonally Protected Building Blocks

Abstract

Peptide “B-2”, which is one of the most potent kallikrein-related peptidase 3 (KLK3)-stimulating compounds, consists of 12 amino acids and is cyclized by a disulfide bridge between the N- and C-terminal cysteines. Orthogonally protected building blocks were used in the peptide synthesis to introduce a disulfide bridge mimetic consisting of four carbon atoms. The resulting pseudopeptides with alkane and E-alkene linkers doubled the proteolytic activity of KLK3 at a concentration of 14 μM. They were almost as potent as the parent “B-2” peptide, which gives a 3.6-fold increase in the proteolytic activity of KLK3 at the same concentration.

Kallikrein-related peptidase 3 [KLK3, also known as prostate specific antigen (PSA)] is currently the most widely used serum marker for prostate cancer.¹⁻³ KLK3 has been shown to inhibit angiogenesis.⁴⁻⁶ This anti-angiogenic activity is associated with the proteolytic activity of KLK3 and can be enhanced by peptides that stimulate KLK3.^5,6 Furthermore, a high level of KLK3 expression in tumors is associated with a low microvessel density, and the level of expression of KLK3 decreases in poorly differentiated tumors, suggesting a tumor suppressing role for KLK3.⁷⁻⁹

Peptide 1a (B-2) and its equipotent amide analogue, 1b, stimulate the proteolytic activity of KLK3.¹⁰ Peptide 1a along with two other peptides with similar biological activity was identified using phage-display technology.^10,11 The two most potent peptides are 1a/1b and C-4 (structure not shown).¹⁰ Although previously published results indicate that peptide C-4 has a stronger stimulating effect on KLK3 with small peptide substrates, 1a has proven to be more potent with larger protein substrates and to have a higher affinity for KLK3 (J. M. Mattson and H. Koistinen, unpublished results). These peptides represent the most potent KLK3-stimulating compounds, and the search for low-molecular weight compounds with the same biological activity has so far been unsuccessful, although one low-molecular weight compound possessing both a very weak stimulating effect on its own and a weakly enhancing effect on the stimulation by 1a has been identified.¹²

In an earlier study of 1a,¹³ it was shown that most of the amino acid side chains are essential for the KLK3 stimulating activity. To improve the in vivo stability of the KLK3-stimulating peptides, alternatives to the disulfide bridge for cyclizing the peptide have been studied.¹⁴ These alternatives are based on cyclizing the peptide in a head-to-tail fashion using an amide bond. The only potent replacement for the terminal disulfide bridge consisted of γ-aminobutyric acid at the N-terminus and aspartic acid at the C-terminus connected via an amide bond (peptide 2). The analogue had improved stability in plasma compared to that of 1a.¹⁴ The same study also showed that the terminal NH₂ group could be removed from the peptide without a significant loss of biological activity.

Figure 1.

Structures of synthesized pseudopeptides and the linkers used.

We were interested in replacing the disulfide bridge in peptide 1a with some of the orthogonally protected disulfide bridge mimetics based on hydrocarbon linkers that we had developed previously,¹⁵ to afford pseudopeptides 3a and 3b. As the terminal NH₂ group does not seem to be critical for biological activity, analogues 3c and 3d were also included in the series. Replacing the disulfide bridge with a hydrocarbon linker has been reported to considerably increase the in vivo stability and bioavailability of peptides.^16,17 A similar alkene linker has already been shown to be a successful disulfide bridge replacement in the other KLK3-stimulating peptide, C-4.¹⁸ However, in the C-4 peptide, the alkene linker was introduced after the peptide synthesis using a ring-closing metathesis reaction that gave an inseparable mixture of the E- and Z-alkenes.

The synthetic route to pseudopeptides 3a and 3b started with the synthesis of the mimetics containing orthogonal protecting groups compatible with the Fmoc protocol for solid phase peptide synthesis (SPPS). The synthesis of building blocks 8a and 9a needed for synthesis of pseudopeptides 3a and 3b, respectively, is presented in Scheme 1 and is based on our previously published procedure.¹⁵

Scheme 1. Synthesis of Building Blocks 8a, 8b, 9a, and 9b.

Mimetics 8b and 9b lacking the terminal NH₂ group were synthesized in a manner analogous to that of 8a and 9a (Scheme 1). However, the obtained cross metathesis product (6b) was in this case an inseparable 3:1 mixture of the E- and Z-isomers (as determined by ¹H and ¹³C NMR spectroscopy, assignment of peaks using gCOSY, gHSQC, and gHMBC spectra). This was further confirmed by the reduction of the double bond in 6b that yielded only one product (7b).

Interestingly, the hydrolysis of the methyl ester in 6b was found to enrich the E-form of building block 8b, yielding the E-isomer in pure form. The cause for this difference in reactivity between the two isomers is unclear, but the observation is confirmed by NMR spectroscopy of both the isolated product and unreacted starting material.

The synthesized mimetics 8a, 8b, 9a, and 9b were introduced into the peptide following a previously published protocol.¹⁴ The peptide was attached to the resin (Rink Amide AM) from the side chain of Asn6, followed by coupling of amino acids and mimetics using Fmoc chemistry, and then finally cyclizing the pseudopeptide between residues Asn6 and Tyr7 (Scheme 2). Longer reaction times (4 h) were used for coupling the orthogonally protected building blocks to ensure a high yield in these steps. The yield-limiting step in the syntheses of the pseudopeptides was expected to be the head-to-tail cyclization. The main cause of this is the length of the peptide and the bulky tyrosine at the terminal site of the cyclization point that has previously resulted in low yields.¹⁴ One common impurity according to MS studies of pseudopeptides 3a and 3b was a linear peptide resulting from incomplete cleavage of the allylic ester used for protecting the main chain carboxylic acid of the aspartic acid, thus preventing the head-to-tail cyclization.

Scheme 2. Synthesis of Pseudopeptides 3a–3d.

The X in intermediate 10 refers to a bridge mimetic corresponding to building block 8a, 8b, 9a, or 9b.

To verify the stereochemistry of the double bond in the purified pseudopeptide 3c, we also analyzed it by NMR using a method we have utilized in a previous study.¹⁸ This showed that the peptide exclusively contained the E-isomer. After purification of the pseudopeptides to >90% purity, the yields of peptides were 0.6–5%.

The effect of the pseudopeptides on the proteolytic activity of KLK3 was determined in an assay using a chromogenic chymotrypsin substrate.^13,14,19 The ability of pseudopeptides 3a–3d to stimulate the activity of KLK3 is summarized in Table 1. All four pseudopeptides stimulated the activity of KLK3; however, the effects were weaker than that of 1b, and the activity of the best one (3b) was equal to that of the previously published analogue with a linker consisting of aspartic acid and γ-aminobutyric acid as the bridge mimetic (peptide 2, 250% stimulation compared to that with KLK3 alone).¹⁴ Pseudopeptides 3a and 3b containing the disulfide bridge mimetics increased the enzymatic activity to 160 and 250%, respectively, of that of KLK3 alone. Similarly, pseudopeptides 3c and 3d stimulated KLK3 activity to 160 and 170%, respectively.

Table 1. Stimulation of KLK3 Activity by the Synthesized Pseudopeptides at 14 μM (N = 2).

compound	activity ± standard error (%)
KLK3 alone	100
1b	360 ± 5.4
3a	160 ± 8.1
3b	250 ± 10
3c	160 ± 0.5
3d	170 ± 15

In this work, we have shown that it is possible to replace the disulfide bridge in KLK3-stimulating peptide 1a with a hydrocarbon linker without a significant loss of biological activity. Interestingly, the four-carbon linkers containing an alkane or E-alkene seem to be able to mimic the disulfide bridge well despite the carbon atoms in the bridge being smaller in size than the corresponding sulfur atoms in the natural disulfide bridge. Furthermore, we have also confirmed that the terminal NH₂ group next to the disulfide bridge is not critical for the biological activity of the pseudopeptides, although removing it slightly decreases the stimulating effect of the pseudopeptide. Even though the stimulating effect of the synthesized pseudopeptides is weaker than that of the parent peptide, they still significantly stimulate KLK3. In the pseudopeptides both with and without the terminal NH₂ group, the fully saturated linker seems to result in a KLK3 stimulator slightly more potent than the E-alkene linker. This can possibly be attributed to the higher flexibility of the saturated linker that allows the peptide to adopt a conformation closer to that of the biologically active conformation of the parent peptide.

In this work, we have presented the replacement of a terminal disulfide bridge using an orthogonally protected building block. Together with the pseudopeptides of the KLK3-stimulating peptide C-4, which we have published previously,¹⁸ we have now successfully introduced similar hydrocarbon-based disulfide bridge replacements in two structurally different KLK3-stimulating pseudopeptides. Notably, pseudopeptides 3a–3d have been synthesized employing a synthetic strategy different from that used for pseudopeptides based on C-4. We thereby conclude that the presented disulfide bridge mimetics and the synthetic strategies for introducing them could also be applied to other biologically relevant peptides.

Glossary

Abbreviations

KLK3

human kallikrein-related peptidase 3

DCE

1,2-dichloroethane

HOAT

1-hydroxy-7-azabenzotriazole

HATU

2-(1H-7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate

HBTU

N,N,N′,N′-tetramethyl-O-(1H-benzotriazol-1-yl)uronium hexafluorophosphate

TBTU

N,N,N′,N′-tetramethyl-O-(benzotriazol-1-yl)uronium tetrafluoroborate

DCM

dichloromethane

DIPEA

N,N-diisopropylethylamine

Fmoc

9-fluorenylmethoxycarbonyl

Boc

N-tert-butoxycarbonyl

tBu

tertiary butyl

SPPS

solid phase peptide synthesis

Supporting Information Available

Experimental procedures for the synthesis and purification of peptides 3a–3d and compounds 6a and 7–9 and information about the biological assay. This material is available free of charge via the Internet at http://pubs.acs.org.

Author Present Address

^∥ M.P.: Nesteoil Corp., P.O. Box 310. FIN-06101 Porvoo, Finland.

This work was supported by the Academy of Finland (Grants 126969, 135937, and 140929), the Graduate School of Organic Chemistry and Chemical Biology, Research Funds of the University of Helsinki, the Helsinki University Central Hospital, the Finnish Cancer Foundation, the Sigrid Juselius Foundation, Finska Läkaresällskapet, and the Magnus Ehrnrooth foundation.

The authors declare no competing financial interest.