Incrementally increasing the length of a peptide backbone: Effect on macrocyclisation efficiency

Article abstract of DOI:10.1039/c4ob00492b

Three novel analogues of the cyclic pentapeptide sansalvamide A have been synthesised in high yield. A leucine residue in the lead compound is replaced with either a glycine, β-alanine or GABA residue, and the corresponding linear precursor peptides are found to cyclise with dramatically improved efficiency. This correlates with an increase in the effective molarity (EM) of the cyclisation reactions. This journal is

Full text of DOI:10.1039/c4ob00492b

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
COMMUNICATION
Incrementally increasing the length of a peptide
backbone: effect on macrocyclisation efficiency†
Cite this: Org. Biomol. Chem., 2014,
12, 4598
Md. Iqbal Ahmed, Jason B. Harper and Luke Hunter*
Received 5th March 2014,
Accepted 27th May 2014
DOI: 10.1039/c4ob00492b
www.rsc.org/obc
Three novel analogues of the cyclic pentapeptide sansalvamide A access small cyclic peptides,7 and harnessing the templating
have been synthesised in high yield. A leucine residue in the lead effect of
suitably-sized metal ion during cyclisation.8
a
compound is replaced with either a glycine, β-alanine or GABA However, despite these considerable advances no individual
residue, and the corresponding linear precursor peptides are found method is suitable for every peptide, and so there is an
to cyclise with dramatically improved efficiency. This correlates ongoing need to develop new strategies for improving the
with an increase in the effective molarity (EM) of the cyclisation efficiency of peptide cyclisation.
reactions.
A potential new strategy is inspired by the recently reported
total synthesis of unguisin A (1, Fig. 1).9 This natural product
is a cyclic heptapeptide derived from the marine fungus
Emericella unguis,10 and it uniquely contains the nonproteino-
genic amino acid γ-aminobutyric acid (GABA) within the ring.
During the total synthesis of the peptide 1,9 cyclisation of the
corresponding linear precursor peptide was found to be excep-
tionally rapid and efficient (<1 min reaction time, 81% isolated
yield after preparative HPLC), and the ease of cyclisation was
attributed, in part, to the flexibility imparted by the GABA
residue that was positioned in the centre of the linear pre-
cursor peptide. It was hypothesised that this result could form
the basis of a more general strategy for assisting the cyclisation of
Peptide based drug design is an important branch of contem-
porary medicinal chemistry research.1 Of particular interest
are small peptides that can modulate protein–protein inter-
actions, fundamental process that are involved in a myriad of
diseases including cancer, metabolic disorders, and diseases
associated with hormone dysfunction.2 Cyclic peptides often
exhibit superior biological activity relative to their linear
counterparts, due to their conformational rigidity and resist-
ance to proteolytic degradation.3 However, a limiting factor in
the development of cyclic peptide drugs is that their synthesis
is often very inefficient. During the macrocyclisation process,
there are energetic and entropic costs associated with pre-
organising the linear precursor peptide into a conformation
that is amenable to cyclisation. This can lead to the appear-
ance of sideproducts arising from polymerisation, cyclo-
oligomerisation and C-terminal epimerisation processes, which
erode the yield of the desired cyclic peptide.3 The difficulty is
greatest in the case of peptides containing seven or fewer
amino acids, as these have the least conformational
mobility.3,4
Several strategies have been developed to improve the
efficiency of peptide macrocyclisation,3 for example employing
highly dilute reaction conditions,5 incorporating turn-inducing
components such as pseudoproline residues into the linear
precursor,6 exploiting a two-step ring contraction strategy to
School of Chemistry, UNSW Australia, Sydney, NSW 2052, Australia.
E-mail: l.hunter@unsw.edu.au
†Electronic supplementary information (ESI) available: Synthetic procedures,
characterisation data, spectra of novel compounds, bioassay data, and details of
EM calculations. See DOI: 10.1039/c4ob00492b
Fig. 1 Structures of unguisin A (1),9,10 sansalvamide A amide (2),11,12 and
the new targets of this work (3–5).
4598 | Org. Biomol. Chem., 2014, 12, 4598–4601
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Communication
analogues of “difficult” peptides: the concept would be to these conditions none of the lead compound (2) was isolated
incrementally increase the backbone length of one constituent (Scheme 1).14 Gratifyingly however, a dramatic improvement
amino acid, and investigate whether the resulting increase in was observed for each of the new analogues 3–5, with these
flexibility could improve the cyclisation efficiency.
targets being isolated in up to 43%, 84% and 64% yields
For the present work, the cyclic pentapeptide sansalvamide respectively after HPLC purification (Scheme 1).
A amide (2, Fig. 1) was selected as a test case to investigate this
It was of interest to rank these new peptide cyclisations
hypothesis. Peptide 2 is a heat shock protein 90 inhibitor, and (8 → 3, 9 → 4 and 10 → 5) in terms of their synthetic
has previously been shown to exhibit potent in vitro activity efficiency. However, simply comparing the isolated yields of
against multiple cancer cell lines including colon, pancreatic, 3–5 was an unsatisfactory measure since these results were
breast and prostate.11 However, the medicinal development of highly variable across different experimental repetitions
this lead molecule is somewhat hampered by difficulties (Scheme 1), presumably due to mechanical losses of material
associated with peptide cyclisation: although a large number during HPLC purifications. Therefore, the rates of reaction
of analogues of 2 have been synthesised they are often were investigated as an alternative way to rank the cyclisation
obtained in very low (<1%) yield.11 Previous structure–activity efficiencies. Time-course LCMS analyses were attempted, but it
relationship studies have shown that alteration of one leucine was found that all three cyclisations (8 → 3, 9 → 4 and 10 → 5)
residue of 2 is possible without eroding the biological were surprisingly rapid; all appeared to reach completion
activity.12 Therefore, the new targets of the present work are within 1 min, and so it was not possible to rank the cyclisa-
peptides 3–5 (Fig. 1), in which a leucine residue of the parent tions by reaction rate either. Finally, the effective molarity
compound 2 is replaced with glycine, β-alanine and GABA (EM) of each linear peptide was measured as a novel way of
respectively. If the new analogues turn out to be more syntheti- comparing the peptide cyclisation efficiencies.
cally accessible this could benefit the medicinal development
Effective molarity is defined as the ratio between the rate
of the lead compound 2, but more importantly this could also constants of an intramolecular reaction and the corresponding
represent a general approach for synthesising analogues of intermolecular process.15 The EM concept is commonly associ-
other “difficult” cyclic peptides.
ated with descriptions of enzyme-catalysed reactions,16 where
The requisite linear precursor peptides (7–10, Scheme 1) high intramolecular reaction rates involving enzyme–substrate
were readily assembled by solid phase peptide synthesis, conjugates correspond to high EM values. However, to our
employing Wang resin as the solid support and HTBU/DIPEA knowledge the EM concept has not previously been employed
as the coupling reagents.13 With the linear peptides 7–10 in to measure and/or explain peptide macrocyclisation efficiency.
hand, attention was next turned to their macrocyclisation. For To determine EM values of the linear peptides 7–10, a series of
these reactions, identical conditions were employed to those inter-/intramolecular competition experiments were performed
reported for the total synthesis of unguisin A (1),9 i.e. the coup- (Table 1). These competition experiments involved repeating
ling reagents DMTMM·BF4/DIPEA were used,13 and the reac- the peptide cyclisation reactions in the presence of excess
tions were performed at 5 mM peptide concentration. Under phenylalanine methyl ester. In this way, two competing pro-
ducts are formed: the cyclic pentapeptides 2–5 resulting from
the intramolecular process; and the linear hexapeptides 11–14
resulting from an intermolecular process. By quantifying the
Table 1 Competition experiments were performed to calculate the
effective molarity (EM) of linear peptides 7–10
Ratio of cyclic
Starting
material
pentapeptide :
linear hexapeptidea
EMb (mM)
7
8
9
10
12 : 88
21 : 79
27 : 73
26 : 74
7
13
18
18
1
3
2
3
a Determined by analytical HPLC. b Uncertainties are reported as half-
the-range from at least three experiments.
Scheme 1 Synthesis of cyclic peptides 2–5.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4598–4601 | 4599

View Article Online
Communication
Organic & Biomolecular Chemistry
relative amounts of these intra- and intermolecular reaction be to incorporate conformationally-biased GABA analogues19
products, the effective molarity of the cyclisation reactions of in order to optimise the biological activity while retaining
species 7–10 can be calculated.17
synthetic efficiency.
The observed EM values of the cyclisations of 7–10 are illu-
In summary, three novel analogues of the medicinally rele-
minating. The cyclisation of 7 has the lowest EM value vant cyclic pentapeptide 2 have been synthesised in high yield.
(Table 1, entry 1), which was to be expected since this reaction Replacing a leucine residue in the lead compound with glycine
was already known to be low-yielding (Scheme 1).18 The corres- leads to a dramatic improvement in macrocyclisation efficiency
ponding EM value for linear peptide 8 is higher (Table 1, entry 2), (analogue 3), and this is attributed to the increased flexibility of
and this correlates with the higher yield in the cyclisation the linear precursor peptide 8. Progressing to analogues con-
of 8 relative to 7 (Scheme 1). Upon proceeding to peptide 9, taining the backbone-homologated residues β-alanine (4) and
which contains an additional methylene group relative to GABA (5) gives a further increase in flexibility and cyclisation
peptide 8, a further increase in EM is observed (Table 1, entry 3). efficiency. These results should facilitate the medicinal develop-
Finally, the cyclisation of linear peptide 10 was found to have ment of 2 towards anticancer therapeutic applications, but more
an EM value that was essentially identical to that of 9 (Table 1, broadly this work may also represent a novel strategy for assist-
entry 4), demonstrating that the incorporation of yet another ing the synthesis of analogues of “difficult” cyclic peptides.
methylene group has no further effect on the cyclisation
efficiency.
Overall, the EM values of 7–10 (Table 1) confirm that the
incorporation of a glycine or backbone-homologated residue
Acknowledgements
can improve peptide cyclisation efficiency. Also, comparing the This work was funded by the Australian Research Council
EM values constitutes a more precise method of ranking these (Discovery Project grant awarded to LH). The authors thank
peptides’ propensity for cyclisation than other measurements S. McAlpine for helpful discussions, A. Islam for assistance
attempted in this work (i.e., reaction rate and product yield). It with bioassays, and S. Videnovic for LCMS technical support.
is notable that the changes in EM are quite small in magni-
tude; this implies that while adding extra rotatable bonds does
reduce the enthalpic penalty of cyclisation, this is somewhat
offset by an increased entropic penalty. The maximum benefit
Notes and references
is achieved with β-alanine (i.e., 9 → 4).
1 (a) S. H. Joo, Biomol. Ther., 2012, 20, 19–26; (b) D. Raucher,
S. Moktan, I. Massodi and G. L. Bidwell III, Expert Opin.
Drug Delivery, 2009, 6, 1049–1064; (c) P. Vlieghe,
V. Lisowski, J. Martinez and M. Khrestchatisky, Drug Dis-
covery Today, 2010, 15, 40–56.
Having identified three synthetically accessible analogues
of the lead compound 2, it became of interest to investigate
whether these new compounds maintained useful levels of
biological activity. Accordingly, the cytotoxicities of 3–5 were
measured against the HCT-116 human colon cancer cell line
using the CCK assay method (Table 2). Analogue 3 was found
to suffer a reduction in activity relative to parent 2 (only 10%
inhibition by 3 at 100 μM, cf. 35% inhibition by 2 at 50 μM),
which was disappointing in light of previous structure–activity
data suggesting that the variable amino acid was not critical
for activity.12 However, it was gratifying to observe that some
activity was recovered in analogues 4 and 5 (Table 2). Com-
pound 5 now appears to be an interesting candidate for
further development; one possible avenue of future work may
2 B. Groner, Peptides as Drugs: Discovery and Development,
Wiley-VCH, Weinheim, 2009.
3 C. J. White and A. K. Yudin, Nat. Chem., 2011, 3, 509–524.
4 (a) A. Ehrlich, H.-U. Heyne, R. Winter, M. Beyermann,
H. Haber, L. A. Carpino and M. Bienert, J. Org. Chem.,
1996, 61, 8831–8838; (b) R. Hili, V. Rai and A. K. Yudin,
J. Am. Chem. Soc., 2010, 132, 2889–2891.
5 M. Malesevic, U. Strijowski, D. Bächle and N. Sewald,
J. Biotechnol., 2004, 112, 73–77.
6 D. Skropeta, K. A. Jolliffe and P. Turner, J. Org. Chem.,
2004, 69, 8804–8809.
7 W. D. F. Meutermans, S. W. Golding, G. T. Bourne,
L. P. Miranda, M. J. Dooley, P. F. Alewood and M. L. Smythe,
J. Am. Chem. Soc., 1999, 121, 9790–9796.
Table 2 Cytotoxicity of 2–5 towards human colon cancer cells. The
positive and negative controls were 17-AAG and DMSO respectively
8 M. Liu, Y.-C. Tang, K.-Q. Fan, X. Jiang, L.-H. Lai and
Y. H. Ye, J. Pept. Res., 2005, 65, 55–64.
9 L. Hunter and J. H. Chung, J. Org. Chem., 2011, 76,
5502–5505.
10 J. Malmstrøm, J. Nat. Prod., 1999, 62, 787–789.
11 R. P. Sellers, L. D. Alexander, V. A. Johnson, C.-C. Lin,
J. Savage, R. Corral, J. Moss, T. S. Slugocki, E. K. Singh,
M. R. Davis, S. Ravula, J. E. Spicer, J. L. Oelrich,
A. Thornquist, C.-M. Pan and S. R. McAlpine, Bioorg. Med.
Chem., 2010, 18, 6822–6856.
% inhibition
of HCT-116
cell growth
Concentration
(μM)
Compound
2
3
4
5
50a
100b
100b
100b
35a
10 1c
15 1c
19 1c
a Data taken from ref. 11. b Higher concentrations were also investigated
in an attempt to match the level of inhibition shown by 2, but this led
to solubility problems. c Uncertainty is reported as the standard error of
at least four experiments.
4600 | Org. Biomol. Chem., 2014, 12, 4598–4601
This journal is © The Royal Society of Chemistry 2014

View Article Online
Organic & Biomolecular Chemistry
Communication
12 M. R. Davis, E. K. Singh, H. Wahyudi, L. D. Alexander, 15 (a) A. J. Kirby, Adv. Phys. Org. Chem., 1980, 17, 183–
J. B. Kunicki, L. A. Nazarova, K. A. Fairweather, A. M. Giltrap,
K. A. Jolliffe and S. R. McAlpine, Tetrahedron, 2012, 68,
1029–1051.
278; (b) M. A. Winnik, Chem. Rev., 1981, 81, 491–524;
(c) R. Cacciapaglia, S. Di Stefano and L. Mandolini, Acc.
Chem. Res., 2004, 37, 113–122.
13 W. C. Chan and P. D. White, Fmoc Solid Phase Peptide 16 V. M. Krishnamurthy, V. Semetey, P. J. Bracher, N. Shen
Synthesis: A Practical Approach, Oxford University Press
Inc., New York, 2000. HBTU = N,N,N′,N′-tetramethyl-O-(1H-
and G. M. Whitesides, J. Am. Chem. Soc., 2007, 129,
1312–1320.
benzotriazol-1-yl)uronium hexafluorophosphate; DIPEA = 17 See ESI.†
N,N-diisopropylethylamine; DMTMM·BF4 = 4-(4,6-dimethoxy- 18 It is noteworthy that the EM of the cyclisation of peptide 7
1,3,5-triazin-2-yl)-4-methyl-morpholinium tetrafluoroborate.
14 It is possible to obtain 2 in moderate yield under different
reaction conditions, see: (a) W. Gu, S. Liu and R. B. Silverman,
Org. Lett., 2002, 4, 4171–4174; (b) A. I. Fernández-Llamazares,
J. García, V. Soto-Cerrato, R. Pérez-Tomás, J. Spengler
(i.e., 7 mM) is close to the concentration at which the
reaction was actually performed (i.e., 5 mM), which may
explain the low yield of 2. We therefore repeated the
reaction at a lower peptide concentration of 0.5 mM, but
unfortunately still did not obtain an isolable quantity of 2.
and F. Albericio, Chem. Commun., 2013, 49, 6430–6432. 19 (a) L. Hunter, K. A. Jolliffe, M. J. T. Jordan, P. Jensen and
However, we were primarily interested in comparing 2 with
the other targets (1, 3–5) which is why the quoted con-
ditions were employed.
R. B. Macquart, Chem. – Eur. J., 2011, 17, 2340–2343;
(b) L. Hunter, S. Butler and S. B. Ludbrook, Org. Biomol.
Chem., 2012, 10, 8911–8918.
This journal is © The Royal Society of Chemistry 2014
Org. Biomol. Chem., 2014, 12, 4598–4601 | 4601