Synthesis and biological evaluation of the osteoblast proliferating cyclic peptides dianthins G and H

Article abstract of DOI:10.1016/j.tet.2014.04.035

The first synthesis and osteoblast proliferative activity of the naturally occurring cyclic peptides dianthins G and H is described. The greater potency of naturally occurring dianthin G over dianthin H at physiological concentrations mirrored the osteobl

Full text of DOI:10.1016/j.tet.2014.04.035

Tetrahedron xxx (2014) 1e7
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Synthesis and biological evaluation of the osteoblast proliferating
cyclic peptides dianthins G and H
,
,
,
c
,
,d,e
Harveen Kaur a b, Amanda M. Heapy b c, Renata Kowalczyk b d, Zaid Amso b
,
Maureen Watson e, Jillian Cornish b e, Margaret A. Brimble b
a School of Biological Sciences, The University of Auckland, 3A Symonds St, Auckland 1142, New Zealand
b Maurice Wilkins Centre for Molecular Biodiscovery, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand
c School of Chemical Sciences, The University of Auckland, 23 Symonds St, Auckland 1142, New Zealand
d Institute for Innovation in Biotechnology, The University of Auckland, 3A Symonds St, Auckland 1142, New Zealand
e Department of Medicine, The University of Auckland, Private Bag 92019, Auckland 1142, New Zealand
,
,c,d,
*
a r t i c l e i n f o
a b s t r a c t
Article history:
The first synthesis and osteoblast proliferative activity of the naturally occurring cyclic peptides dia-
nthins G and H is described. The greater potency of naturally occurring dianthin G over dianthin H at
physiological concentrations mirrored the osteoblast proliferative activity observed for synthetic dia-
nthins G and H. Six alanine-scan analogues of the more potent dianthin G were also synthesised and
osteoblast assays revealed that four of the six residues can be further modified for improved activity. We
also confirmed by variable temperature 1H NMR spectroscopic analysis that the sets of major and minor
signals observed for dianthins G and H in DMSO-d6 are in fact due to cisetrans rotational isomers of the
proline ring.
Received 3 March 2014
Received in revised form 6 March 2014
Accepted 10 April 2014
Available online xxx
Keywords:
Cyclic peptides
Dianthin G
Osteoblast proliferation
Macrolactamisation
Ó 2014 Elsevier Ltd. All rights reserved.
1. Introduction
include denosumab, calcium, vitamin D, hormone therapy, selec-
tive oestrogen-receptor modulators and calcitonin.2,4
Osteoporosis is a prevalent skeletal disease that affects 1 in 2
women and 1 in 4 men over the age of 50.1 It is characterised by loss
of bone mass and strength due to an imbalance in regulating bone
turnover, whereby the bone resorption activity of osteoclasts ex-
ceeds the bone formation activity of osteoblasts.1e3 This results in
increased bone fragility and susceptibility to fractures, which
commonly occur at the hip, spine or wrist.3
Due to its asymptomatic nature, it is usually diagnosed after the
first clinical fracture and patients are henceforth affected by chronic
pain, loss of mobility, a reduction in quality of life as well as an
increased risk of mortality.3 It is a significant medical and socio-
economic issue that is further exacerbated by a growing world-
wide elderly population and a high cumulative rate of fractures.4
Current therapies for the treatment of osteoporosis mainly uti-
lise antiresorptive agents, which slow down bone resorption by
reducing osteoclast activity. The largest class of antiresorptive
drugs are the widely used bisphosphonates, as they are in-
expensive, available orally or intravenously and can be used for
several osteoporosis types.3 Other antiresorptive treatments
In cases where there is significant loss of bone however, alter-
native therapies, such as the use of anabolic agents, which promote
bone formation are required. Currently the only approved anabolic
drug is recombinant human parathyroid hormone (rhPTH), which
is administered by injection over a maximum of 2 years.1 The ef-
fectiveness of rhPTH however is limited by its cost, a plateauing of
the net anabolic effect, decreased responsiveness and potential link
to osteosarcoma.1
Recently, two novel cyclic peptides, dianthins G (1) and H (2),
isolated from the aerial parts of Dianthus superbus, a traditional
Chinese medicinal plant from Shandong province, were found to
exhibit osteoblast proliferative activity (Fig. 1). Dianthins G (1) and
H (2) are sequentially similar cyclic hexapeptides that differ from
each other by two branched hydrophobic amino acids. The struc-
ture of dianthins G and H were elucidated as cyclo-(Pro-Leua-Thr-
Leub-Phe-Gly) and cyclo-(Pro-Val-Thr-Ile-Phe-Gly), respectively.5
Both dianthins G and H were tested for their rat osteoblast
MC3T3-E1 proliferative activity using an MTT assay.5 At 10ꢀ5 mM,
both dianthins G and H were found to significantly promote
MC3T3-E1 osteoblast proliferation by 38% and 34%, respectively.5
Interestingly, this activity was found to decrease at higher and
lower concentrations.5
* Corresponding author. Fax: þ64 (9)3737422; e-mail address: m.brimble@
auckland.ac.nz (M.A. Brimble).
0040-4020/$ e see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2014.04.035
Please cite this article in press as: Kaur, H.; et al., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.04.035

2
H. Kaur et al. / Tetrahedron xxx (2014) 1e7
Fig. 1. Structures of dianthins G (1) and H (2) and dianthin G alanine scan analogues 3e8 with the alanine substituted residues highlighted in red.
Intrigued by the osteoblast proliferative activity of these novel
threonine does not require protection as the more nucleophilic N-
terminus reacts faster to form the desired amide bond.10 The use of
unprotected linear precursors during solution phase macro-
cyclisations is favourable for our synthetic strategy as it overcomes
the poor solubility issues associated with fully protected peptide
segments.11 Moreover the achiral glycine residue was selected as
the C-terminal residue of the linear precursors as glycine minimises
steric hindrance at the cyclisation site and also prevents any op-
portunity for racemisation during macrolactamisation.7
The eight linear precursors to the cyclopeptides 1e8 were syn-
thesised by Fmoc-SPPS (see Supplementary data). Briefly, commer-
ciallyavailable hydroxymethylphenoxypropionic acid (HMPP) linker-
bound C-terminal glycine and alanine residues were attached to in-
house prepared aminomethyl polystyrene resin using DIC (Scheme
1). The remaining sequence to the eight linear precursors were
then synthesised by automated Fmoc-SPPS on a TributeÔ peptide
synthesiser. The Fmoc-protecting groups were removed with 20%
piperidine in DMF and each subsequent amino acid was sequentially
coupled to the growing resin-bound peptide using a mixture of HBTU
and NMM in DMF. Following the final Fmoc-deprotection, trifluoro-
acetic acid mediated cleavage of the resin-bound peptides released
the linear precursors from the acid-labile HMPP linker with con-
comitant removal of the side-chain protecting groups and in suffi-
cient purity for subsequent macrolactamisation.
cyclic peptides and the need to find new anabolic agents, we herein
report the first total synthesis and evaluation of the primary rat
osteoblast proliferative activity of synthetic samples of the naturally
occurring cyclic peptides dianthins G (1) and H (2), and six alanine-
scan analogues 3e8 of the more active dianthin G (Fig. 1). The six
alanine-scan analogues of dianthin G have undergone sequential
substitution of each residue with alanine. Alanine substitution al-
lows the contribution of each individual amino acid’s side-chain to
the peptide’s function and bioactivity to be evaluated without in-
troducing additional conformational freedom. As dianthin G consists
of only L-amino acids, L-alanine was used for the alanine scan.
2. Results and discussion
There are many methods available for the synthesis of head-to-
tail cyclic peptides, of which the final macrolactamisation reaction
is commonly accomplished by classical solution phase methods us-
ing protected linear precursors under high dilution conditions.6e8
Alternatively, when the isolation of protected linear precursors and
final cyclopeptides is challenging, the macrolactamisation reaction
can be performed by side-chain anchoring of a trifunctional amino
acid or the use of ‘safety-catch’ linkers on solid support with low
resin loading, which creates a pseudodilution effect.6e8
The key macrolactamisation reaction was then successfully ach-
ieved by adding a mixture of each linear precursor and HBTU in
DCM:DMF (4:1) dropwise at a rate of 0.5 mL/h to a stirring solution
of DIPEA in DCM with a final concentration of 0.5 mg/mL.12 LCeMS
analysis of the reaction mixture after complete addition of all re-
agents showed the corresponding cyclic peptide as the predominant
peak with the linear precursor undetected. The cyclic peptides were
then isolated by concentration of the reaction mixture under re-
duced pressure and purification by semipreparative RP-HPLC.
The spectroscopic data and measured optical rotations for both
synthetic dianthins G (1) and H (2) were in agreement with that
reported for the corresponding naturally occurring cyclo-
hexapeptides thus confirming the elucidated structures (Tables 1
and 2). For dianthin H (2), although the 13C NMR chemical shift
The success of head-to-tail cyclisations is predominantly influ-
enced by the ability of a linear precursor to adopt an intermediate
conformation similar to that of its corresponding cyclic peptide.6
The formation of this reactive intermediate conformation is pro-
moted by the ability of residues, such as glycine and proline to
induce torsion angles that are characteristic of
b-turns, thereby
positioning the reactive N- and C-termini in closer spatial prox-
imity.9 In addition, the cyclisation of hexapeptides is also promoted
by the potential to form a stabilised two b
-turn structure.7,9
Fortunately all eight cyclic peptides 1e8 herein contain these
favourable characteristics and consist of predominantly hydro-
phobic residues, hence the key macrolactamisation reaction was
accomplished by solution phase using unprotected linear pre-
cursors. The less reactive secondary hydroxyl functionality of
Please cite this article in press as: Kaur, H.; et al., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.04.035

H. Kaur et al. / Tetrahedron xxx (2014) 1e7
3
Scheme 1. Synthesis of dianthin G (1). Reagents and conditions: (i) Fmoc-Gly-O-CH2-phi-OCH2-CH2-COOH, DIC, DCM:DMF (1:1), 4 h; (ii) 20% piperidineeDMF (2ꢁ5 min); (iii)
Fmoc-amino acid, HBTU, NMM, DMF, 40 min; (iv) repeat (iii) and (ii) ꢁ5; (v) TFA-i-Pr3SiHeH2O-2,20-(ethylenedioxy)diethanethiol (94:1:2.5:2.5), 3 h; (vi) HBTU, DIPEA, DCM:DMF
(4:1), 36 h, 53%.
Table 1
Table 2
1H and 13C NMR assignment in DMSO-d6 for the major conformer of synthetic and
isolated dianthin G (1)
1H and 13C NMR assignment in DMSO-d6 for the major conformer of synthetic and
isolated dianthin H (2)a
Residue
Synthetic (1)
Lit.
Residue
Synthetic (2)
Lit.
dH
dC
dH
dC
dH
dC
dH
dC
Gly
CO
NH
aa
Gly
CO
NH
aa
d
166.7
d
d
N/Oc
3.75 (br d, 15.5)
4.10 (m)
166.7
d
d
167.0
d
d
167.0
d
7.43 (m)
3.74 (dd, 17.4, 1.9)
4.10 (m)
7.50 (dd, 5.7, 2.1)
3.87 (dd, 17.0, 1.9)
4.06 (dd, 17.0, 5.8)
7.49 (d, 3.5)
3.87 (br d, 15.5)
4.05 (dd, 17.0, 6.0)
42.0
42.0
42.6
42.6
ab
ab
Pro
CO
a
ba
bb
ga
gb
da
Pro
CO
a
ba
bb
ga
gb
da
d
171.5
61.2
29.0
d
171.4
61.2
29.0
d
171.5
61.2
29.6
d
171.5
61.2
29.6
4.14 (m)
1.74 (m)
2.15 (m)
1.91 (m)
1.96 (m)
3.52 (m)
3.60 (m)
4.14 (dd, 7.5, 4.5)
1.74 (m)
2.15 (m)
1.91 (m)
1.96 (m)
3.50 (m)
3.60 (m)
4.19 (dd, 8.5, 5.6)
1.79 (m)
2.21 (m)
1.93 (m)
1.96 (m)
3.50 (m)
3.74 (m)
4.19 (dd, 8.5, 5.5)
1.79 (m)
2.20 (m)
1.93 (m)
1.96 (m)
3.50 (m)
3.75 (m)
24.7
45.7
24.7
45.7
24.6
46.1
24.5
46.0
db
db
Leua
CO
NH
a
ba
bb
g
Val
CO
NH
a
b
ga
gb
d
171.2
d
d
171.2
d
d
170.4
d
d
170.4
d
8.12 (d, 8.8)
4.11 (m)
1.65 (m)
1.75 (m)
1.48 (m)
8.15 (d, 9.0)
4.11 (m)
1.65 (m)
1.74 (m)
1.48 (m)
7.45 (d, 9.9)
4.24 (dd, 9.9, 6.3)
2.26 (m)
0.80 (3H, d, 6.7)
0.85 (3H, d, 6.8)
7.53 (d, 10.0)
4.23 (dd, 9.5, 6.5)
2.26 (m)
0.79 (3H, d, 6.5)
0.86 (3H, d, 7.0)
51.7
40.0
51.7
40.0
57.9
30.2
18.1
19.7
57.9
30.7
18.1
19.7
24.7
20.9
23.4
24.7
20.9
23.3
da
db
0.84 (3H, d, 6.5)
0.89 (3H, d, 6.6)
0.84 (3H, d, 6.5)
0.88 (3H, d, 6.5)
Thr
CO
NH
a
b
g
d
170.4
d
d
170.4
d
Thr
CO
NH
a
b
g
7.32 (d, 7.6)
4.35 (dd, 7.2, 2.7)
4.30 (m)
7.39 (d, 7.0)
4.35 (dd, 7.5, 2.5)
4.30 (m)
d
169.6
d
d
169.6
d
56.8
68.3
19.2
56.8
68.2
19.1
7.43 (m)
4.38 (m)
4.32 (m)
1.02 (3H, d, 6.3)
7.45 (d, 7.5)
4.39 (dd, 8.0, 3.0)
4.32 (m)
56.6
67.6
18.6
56.6
67.6
18.5
1.10 (3H, d, 6.2)
1.10 (3H, d, 6.0)
Ile
1.02 (3H, d, 6.5)
CO
NH
a
d
170.3
d
d
170.3
d
Leub
CO
NH
a
ba
bb
g
8.02 (d, 4.7)
3.74 (m)
1.61 (m)
1.01 (m)
0.58 (3H, d, 6.8)
0.68 (3H, dd, 7.2, 7.4)
8.04 (d, 4.5)
3.74 (m)
1.61 (m)
0.98 (m)
0.58 (3H, d, 7.0)
0.67 (3H, d, 7.0)
d
171.5
d
d
171.5
d
59.8
34.8
23.9
15.1
11.3
59.8
34.8
23.9
15.0
11.2
8.05 (d, 5.5)
3.82 (dt, 9.5, 5.5)
1.14 (m)
1.32 (m)
1.42 (m)
8.07 (d, 5.5)
3.83 (dt, 10.0, 5.5)
1.14 (m)
1.32 (m)
1.42 (m)
b
53.3
39.0
53.2
39.0
gaCH2
gbCH3
d
24.5
21.5
22.6
24.5
21.4
22.6
Phe
CO
NH
a
da
db
0.72 (3H, d, 6.5)
0.79 (3H, d, 6.5)
0.72 (3H, d, 6.5)
0.79 (3H, d, 6.5)
d
170.2
d
d
170.2
d
7.87 (d, 8.3)
4.42 (m)
2.87 (m)
3.34 (dd, 14.2, 4.0)
d
7.85 (d, 7.0)
4.42 (m)
2.85 (m)
3.32 (m)
d
Phe
CO
NH
a
54.2
36.2
54.2
36.2
d
170.1
d
d
170.1
d
ba
bb
10
7.96 (d, 8.8)
4.37 (m)
2.84 (dd, 13.8, 10.8)
3.29 (dd, 14.0, 4.0)
d
7.97 (d, 9.0)
4.37 (m)
2.83 (dd, 11.5, 14.0)
3.29 (dd, 14.0, 4.0)
d
53.9
36.3
53.9
36.3
138.4
128.1
129.0
126.1
129.0
128.1
138.4
128.0
128.9
126.0
128.9
128.0
ba
bb
10
20
7.24 (m)
7.19 (m)
7.18 (m)
7.19 (m)
7.24 (m)
7.15e7.32 (m)
7.15e7.32 (m)
7.16 (d, 7.0)
7.15e7.32 (m)
7.15e7.32 (m)
30
138.4
128.0
129.1
126.0
129.1
128.0
138.4
128.1
129.0
126.0
129.0
128.1
40
20
7.25 (m)
7.16 (m)
7.19 (m)
7.16 (m)
7.10e7.32 (m)
7.10e7.32 (m)
7.10e7.32 (m)
7.10e7.32 (m)
7.10e7.32 (m)
50
30
60
40
a
The proton signals integrate to 1H, unless indicated.
50
60
7.25 (m)
a
The proton signals integrate to 1H, unless indicated.
A signal is not expected.
N/O an expected signal is not observed.
with the Val Cb 13C chemical shift of dC 30.2 recorded for synthetic
dianthin H (2). In addition, the trans-conformation of the proline
residues adopted by both of the naturally occurring dianthins G (1)
and H (2) were also retained in the synthetic samples of these
cyclohexapeptides.
b
c
for Val Cb was reported as dC 30.7, the 13C spectra recorded for
dianthin H exhibited a chemical shift of 30.1.5 This is in agreement
Please cite this article in press as: Kaur, H.; et al., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.04.035

4
H. Kaur et al. / Tetrahedron xxx (2014) 1e7
Importantly, the 1H NMR spectra provided in the literature for
In addition, to further confirm that the observed 1H NMR signals
are due to the presence of cisetrans conformers at the proline resi-
due, the 1H NMR spectrum of proline-deficient dianthin G analogue
3 was also recorded (Fig. 2c and Table 3). This spectrum showed only
one set of signals, indicating that proline-deficient analogue 3 exists
as a single conformer. Overall the spectroscopic analysis conducted
in this study confirms that the major and minor signals observed for
both isolated and synthetic dianthins G (1) and H (2) are due to
cisetrans isomerism of the proline ring in DMSO-d6.
We nextevaluated the osteoblast proliferative activity of synthetic
cyclohexapeptides 1e8 using primary rat osteoblasts isolated from
neonatal rat calvarial bones.16 This osteoblast culture system gives
more accurate and representative results than the osteoblast-like
MC3T3-E1 cells originally used to probe the osteoblast proliferative
activity of the natural dianthins G (1) and H (2) (Fig. 3).17e19 Osteo-
blast proliferative activity was measured over the same 10ꢀ7 to
10ꢀ9 M physiological concentration range asTonget al. and the extent
of proliferation was determined by [3H]thymidine incorporation into
cells exposed to the cyclohexapeptides relative to the control.20e22
At all three concentrations synthetic dianthin G (1), like naturally
occurring dianthin G, significantly promoted osteoblast proliferation.
both naturally occurring dianthins G (1) and H (2) exhibited two
sets of signals in DMSO-d6, however Tong et al. only assigned and
reported the major set of signals for the natural products.5 Likewise
our 1H NMR spectra for both synthetic dianthins G (1) and H (2)
closely resembled the literature spectra in DMSO-d6, with a 1.0:0.1
ratio of the two sets of signals being observed. Due to the achiral C-
terminal glycine at the cyclisation site, the two signals are not di-
astereomers as there is no opportunity for racemisation during
cyclisation.
As both cyclic hexapeptides contain a proline residue that can
adopt a cis or trans conformation, we performed variable temper-
ature 1H NMR experiments on synthetic dianthin G (1) from 300 to
350 K to probe this cisetrans isomerisation (Fig. 2a and b).13 The
amide region of the 1H NMR spectrum is depicted as both cis- and
trans-conformers exhibit well resolved signals in this region.13,14 At
the lowest temperature the proton signals for both the cis- and
trans-conformers are well resolved due to slow cisetrans isomer-
isation.15 As the temperature increases, the line broadening and
coalescence of some signals indicates that cisetrans isomerisation
between the two conformers is accelerated.15
Fig. 2. Amide region of the 1H NMR spectra of (a) dianthin G (1) at 300 K with a 1.0:0.1 ratio of major:minor signals, (b) dianthin G (1) at 340 K and (c) dianthin G analogue cyclo-
(ALTLFG) 3 at 300 K. The minor signals of (1) could not be assigned.
Please cite this article in press as: Kaur, H.; et al., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.04.035

H. Kaur et al. / Tetrahedron xxx (2014) 1e7
5
Table 3
Synthetic dianthin H however was not significantly active at any of
these concentrations, unlike naturally occurring dianthin H, and this
could be due to the different cells used. At lower concentrations, the
greater potency of synthetic dianthin G over dianthin H is in accor-
dance with the reported activities of the natural products.
1H and 13C NMR assignment in DMSO-d6 for cyclo-(ALaTLbFG) 3
Residue
dH
dC
Gly
CO
NH
aa
d
168.6
d
7.83 (m)
3.80 (dd, 16.7, 5.4)
3.61 (dd, 16.7, 5.0)
Similar to synthetic dianthin G (1), analogues 5e8 where Thr,
Leub, Phe and Gly, respectively were substituted with alanine, all
exhibited significant activity. This indicates that these residues are
not essential for activity and substitutions at these positions are
tolerated. As the substitution of glycine for alanine in 8 also reduces
conformational flexibility, these initial studies suggest that rigidi-
fication of the cyclic backbone or increased bulk at this position is
allowed. In contrast, the substitution of Pro and Leua for alanine in 3
and 4, respectively resulted in loss of osteoblast proliferative ac-
tivity, which indicates that these two residues were essential for
proliferative activity. Interestingly, as both 4 and dianthin H (2)
where Leua is substituted for alanine and valine, respectively were
inactive, suggests that variation at Leua might also explain the in-
significant activity observed for dianthin H (2) in our studies.
42.3
ab
Ala
CO
NH
a
d
171.8
d
8.46 (d, 7.0)
4.10 (p, 7.0)
1.26 (d, 7.2)
49.7
16.7
b
Leua
CO
NH
a
ba
bb
g
d
171.5
d
8.10 (d, 8.8)
4.21 (m)
1.75 (m)
1.65 (m)
1.55 (m)
0.84 (d, 6.4)
0.90 (d, 6.5)
52.0
37.7
24.5
21.2
23.1
da
db
3. Conclusion
Thr
CO
NH
a
b
g
d
170.51
d
7.54 (d, 8.0)
4.27 (m)
4.26 (m)
1.04 (d, 6.1)
In conclusion, we herein report the first synthesis of the natu-
rally occurring cyclic peptides dianthins G (1) and H (2) and dia-
nthin G analogues 3e8. Importantly, spectroscopic analysis of
synthetic dianthins G and H confirmed the structure of the natural
products elucidated by Tong et al.5 Moreover, variable temperature
NMR analysis of dianthin G (1) and proline-deficient analogue 3
confirmed that the major and minor set of signals for the natural
products are due to cisetrans isomerism of the proline ring in
DMSO-d6. Biological evaluation of the cyclohexapeptides 1e8
showed that significant osteoblast proliferative activity was ob-
served for dianthin G (1) and analogues 5e8, whereas dianthin H
(2) and analogues 3 and 4 were inactive. These results highlight the
importance of dianthin G (1) as a promising new candidate for the
treatment of osteoporosis, particularly for its osteoblast pro-
liferative activity and its potential as an alternative anabolic agent.
We are currently further investigating the synthesis and biological
activity of other dianthin G analogues.
67.5
57.5
19.4
Leub
CO
NH
a
ba
bb
g
d
171.6
d
7.88 (d, 5.4)
3.97 (ddd, 5.2, 5.4, 10.6)
1.32 (m)
1.27 (m)
1.50 (m)
52.3
37.9
23.8
21.5
23.0
da
db
0.77 (d, 6.6)
0.80 (d, 6.6)
Phe
CO
NH
a
d
170.55
d
8.04 (d, 6.9)
4.29 (m)
3.15 (dd, 14.0, 5.2)
2.82 (dd, 14.0, 10.0)
d
54.5
36.2
ba
bb
10
137.9
128.1
129.0
126.3
129.0
128.1
20
7.25 (m)
7.17 (m)
7.20 (m)
7.17 (m)
30
4. Experimental section
4.1. General information
40
50
60
7.25 (m)
a
The proton signals integrate to 1H, unless indicated.
A signal is not expected.
All reagents were purchased as reagent grade and used without
further purification. Solvents for RP-HPLC were purchased as HPLC
grade and used without further purification. Fmoc-Gly-O-CH2-phi-
b
OCH2-CH2-COOH and Fmoc-
purchased from Polypeptide Laboratories Group. Fmoc-
-Leu, Fmoc- -Val, Fmoc- -Ala and Fmoc-
-Thr(tBu), Fmoc-
purchased from GL Biochem. Fmoc- -Phe and Fmoc-Gly were pur-
L
-Ala-O-CH2-phi-OCH2-CH2-COOH were
-Pro, Fmoc-
-Ile was
L
L
L
L
L
L
L
chased from CEM. Infrared spectra were recorded on a Perkin Elmer
Spectrum 100 infrared spectrometer and reported in wavenumbers
(n
, cmꢀ1). Optical rotations were determined at the sodium D line
(589 nm) at 20 ꢂC on a Perkin Elmer 341 polarimeter and are given in
units of 10ꢀ1 deg cm2 gꢀ1. HRMS analysis was performed on a Bruker
micrOTOFQ mass spectrometer. Nuclear magnetic resonance spectra
were recorded on a Bruker AVANCE 400 spectrometer (1H 400 MHz;
13C 100 MHz). Assignments were made with the aid of HSQC, COSY
and HMBC experiments. Analytical RP-HPLC spectra were acquired
on a Dionex Ultimate 3000 System using an analytical column
ꢀ
(Phenomenex Gemini C18, 110 A, 250 mmꢁ4.6 mm, 5
mm) heated to
50 ꢂC at a flow rate of 1 mL minꢀ1. A linear gradient of 1% solvent B
(0.1% TFA/MeCN (v/v)) to 80% B over 40 min was used with detection
at 210 nm or 254 nm. Semi-preparative RP-HPLC was performed on
a Dionex Ultimate 3000 System using a semi-preparative column
Fig. 3. Osteoblast proliferative activity of dianthins G (1), H (2) and dianthin G alanine
scan analogues 5e8. The data is expressed as a ratio of treatment to control, mean-
ꢃSEM; *significantly different from control (p<0.05).
Please cite this article in press as: Kaur, H.; et al., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.04.035

6
H. Kaur et al. / Tetrahedron xxx (2014) 1e7
ꢀ
(Phenomenex Gemini C18, 110 A, 250 mmꢁ10 mm, 5
m
m) at a flow
was concentrated under reduced pressure, diluted with 0.1% TFA/H2O
(v/v) (12 mL) and purified by semi-preparative RP-HPLC to afford the
title compound (1) as a colourless solid (10.7 mg, 53%). tR: 29.9 min
(59.5% B). HRMS (EI): m/z [MþH]þ calculated for C32H49N6O7:
629.3663; observed: 629.3673. IR (cmꢀ1): 3316, 2954, 1669, 1639,
rate of 5 mL minꢀ1 using a linear gradient of 1% B to 80% B over a time
period adjusted according to the elution and peak profiles obtained
from the analytical RP-HPLC chromatograms with detection at
210 nm. LRMS analysis and LCMS spectra were acquired on an Agi-
lent Technologies 1120 Compact LC equipped with a Hewlett Packard
1529,1452, 701. [
a
]
20 ꢀ34.0 (c 0.10, MeOH) [lit. ꢀ44.0 (c 0.10, MeOH)].
D
1100 MSD mass spectrometer using an analytical column (Phe-
To a stirring solution of DIPEA (32 mL, 184 mmol) in DCM (24 mL)
ꢀ
nomenex Jupiter C18, 300 A,150 mmꢁ4.6 mm, 5
mm) at a flow rate of
was added a mixture of linear dianthin H (11) (24 mg, 38
mmol) and
0.3 mL minꢀ1 using a linear gradient of 5% B to 95% B over 40 min.
HBTU (43 mg, 134 mol) in DCM:DMF (4:1, 22 mL) dropwise at
m
a rate of 0.5 mL/h. After complete addition of the reagents, the
reaction mixture was concentrated under reduced pressure, diluted
with 0.1% TFA/H2O (v/v) (12 mL) and purified by semi-preparative
RP-HPLC to afford the title compound (2) as a colourless solid
(11.8 mg, 51%). tR: 28.6 min (56.8% B). HRMS (EI): m/z [MþNa]þ
calculated for C31H46N6NaO7: 637.3326; observed: 637.3341. IR
4.2. General procedure for peptide synthesis
Linear peptides were synthesised on a 0.1 mmol scale. A mixture
of the C-terminal residue, either Fmoc-Gly-O-CH2-phi-OCH2-CH2-
COOH (2 equiv) or Fmoc-L-Ala-O-CH2-phi-OCH2-CH2-COOH
(2 equiv), and DIC (2 equiv) in DCM:DMF (1:1, 4 mL) was added to
pre-swollen (DCM) in-house prepared aminomethyl polystyrene
resin. The reaction mixture was gently agitated for 4 h, after which
a Kaiser test confirmed the coupling. The remaining sequence was
assembled by Fmoc-SPPS on a TributeÔ peptide synthesizer (Pro-
tein Technologies, Inc.). The Fmoc group was deprotected using 20%
piperidine in DMF (3 mL, 2ꢁ5 min) followed DMF washes (2 mL,
5ꢁ0.5 min). Couplings were performed using a mixture of the
Fmoc-protected amino acid (5 equiv), HBTU (4.8 equiv) and NMM
(10 equiv) in DMF (3.5 mL, 40 min), followed by DMF washes (2 mL,
5ꢁ0.5 min). The completed linear peptides were cleaved from the
linker by gently agitating the resin-bound peptides with a mixture
of TFA:TIS:H2O:2,20-(ethylenedioxy)diethanethiol (94:1:2.5:2.5;
3 mL) for 3 h. The resin was filtered and washed with neat TFA
(2ꢁ1 mL). Cold Et2O (35 mL) was added to the combined TFA fil-
trates and the solution cooled to 0 ꢂC. The precipitated product was
isolated by centrifugation (4000 rpm, 10 min), after which the
liquid layer was decanted. Further cold Et2O (35 mL) was then
added and the pellet disturbed by vortex. The mixture was again
cooled to 0 ꢂC and the isolation procedure repeated twice.
(cmꢀ1): 3307, 2962, 1640, 1516, 1453, 1165, 700. [
a
]
20 ꢀ33.0 (c 0.10,
D
MeOH) [lit. ꢀ45.0 (c 0.10, MeOH)].
4.3.1. Cyclo-(ALaTLbFG) (3). tR: 29 min (58% B). LRMS (EI): m/z
[MþH]þ calculated for C30H47N6O7: 603.4; observed: 603.4.
4.3.2. Cyclo-(PATLbFG) (4). tR: 26 min (52% B). LRMS (EI): m/z
[MþH]þ calculated for C29H43N6O7: 587.3; observed: 587.4.
4.3.3. Cyclo-(PLaALbFG) (5). tR: 30.6 min (61% B). LRMS (EI): m/z
[MþH]þ calculated for C31H47N6O6: 599.4; observed: 599.4.
4.3.4. Cyclo-(PLaTAFG) (6). tR: 25.7 min (51% B). LRMS (EI): m/z
[MþH]þ calculated for C29H43N6O7: 587.3; observed: 587.4.
4.3.5. Cyclo-(PLaTLbAG) (7). tR: 25.1 min (50% B). LRMS (EI): m/z
[MþH]þ calculated for C26H45N6O7: 553.3; observed: 553.4.
4.3.6. Cyclo-(PLaTLbFA) (8). tR: 30.2 min (60% B). LRMS (EI): m/z
[MþH]þ calculated for C33H51N6O7: 643.4; observed: 643.2.
4.2.1. Linear dianthin G (10). tR: 23.0 min (45.5% B). HRMS (EI): m/z
[MþH]þ calculated for C32H51N6O8: 647.3768; observed: 647.3754.
4.4. Osteoblast proliferative assays
Osteoblasts were isolated from 20 day foetal rat calvariae. Briefly,
calvariae were excised and the frontal and parietal bones, free of
suture and periosteal tissue, were collected. The calvariae were se-
quentially digested using collagenase (Sigma) and the cells from
third and fourth digests were collected, pooled and washed. Cells
were grown in T75 flasks in 10% FBS/Dulbecco’s modified eagle
4.2.2. Linear dianthin H (11). tR: 21.3 min (42% B). HRMS (EI): m/z
[MþH]þ calculated for C31H49N6O8: 633.3612; observed: 633.3596.
4.2.3. Linear ALaTLbFG (12). tR: 22.2 min (44% B). LRMS (EI): m/z
[MþH]þ calculated for C30H49N6O8: 621.4; observed: 621.4.
medium (DMEM) (Invitrogen) and 5
phosphate (Sigma) for 2 days and then changed to 10% FBS/MEM
(Invitrogen)/5 g/mL -ascorbic acid 2-phosphate and grown to 90%
confluency. Cells were then seeded into 24 well plates in 5% FBS/
MEM (Invitrogen)/5 g/mL -ascorbic acid 2-phosphate for 24 h.
Cells were growth-arrested in 0.1% bovine serum albumin (BSA) (ICP,
Auckland, New Zealand)/5 g/mL -ascorbic acid 2-phosphate for
mg/mL L-ascorbic acid 2-
4.2.4. Linear PATLbFG (13). tR: 19.7 min (39% B). LRMS (EI): m/z
[MþH]þ calculated for C29H45N6O8: 605.3; observed: 605.4.
m
L
4.2.5. Linear PLaALbFG (14). tR: 22.6 min (44.6% B). LRMS (EI): m/z
[MþH]þ calculated for C31H49N6O7: 617.4; observed: 617.4.
m
L
m
L
4.2.6. Linear PLaTAFG (15). tR: 19.6 min (39% B). LRMS (EI): m/z
[MþH]þ calculated for C29H45N6O8: 605.3; observed: 605.4.
24 h. Cells were pulsed with [3H]thymidine 6 h before the end of the
experimental incubation. The experiments were then terminated
and thymidine incorporation assessed, as measurement of cell
growth. A Trilux counter (Wallac 1450 Microbeta counter) was used
for the data collection. Each of the analogues was screened at three
different concentrations. There were six wells in each group and
each experiment was repeated 3 or 4 times. All treatments were
compared to an untreated control. Data was analysed using analysis
of variance with post-hoc Dunnett’s tests for significant main effect.
A 5% significance level (two-tailed) was used throughout.
4.2.7. Linear PLaTLbAG (16). tR: 18.5 min (36.5% B). LRMS (EI): m/z
[MþH]þ calculated for C26H47N6O8: 571.4; observed: 571.4.
4.2.8. Linear PLaTLbFA (17). tR: 23.3 min (46.3% B). LRMS (EI): m/z
[MþH]þ calculated for C33H53N6O8: 661.4; observed: 661.4.
4.3. General procedure for macrolactamisation
To a stirring solution of DIPEA (28
was added a mixture of linear dianthin G (10) (21 mg, 32
HBTU (37 mg, 98 mol) in DCM:DMF (5:1,18 mL) dropwise at a rate of
0.5 mL/h. Aftercomplete additionof thereagents,thereactionmixture
m
L, 161
m
mol) in DCM (21 mL)
Acknowledgements
mmol) and
m
We thank the Maurice Wilkins Centre for Molecular Bio-
discovery for financial support.
Please cite this article in press as: Kaur, H.; et al., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.04.035

H. Kaur et al. / Tetrahedron xxx (2014) 1e7
10. Kaur, H.; Heapy, A. M.; Brimble, M. A. Synlett 2012, 275.
7
Supplementary data
11. Kent, S. B. H. Chem. Soc. Rev. 2009, 38, 338.
12. Kaur, H.; Heapy, A. M.; Brimble, M. A. Synlett 2012, 2284.
13. Kopple, K. D.; Wang, Y. S.; Cheng, A. G.; Bhandary, K. K. J. Am. Chem. Soc. 1988,
110, 4168.
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.tet.2014.04.035.
14. Singh, E. K.; Ramsey, D. M.; McAlpine, S. R. Org. Lett. 2012, 14, 1198.
15. Kern, D.; Schutkowski, M.; Drakenberg, T. J. Am. Chem. Soc. 1997, 119, 8403.
16. Cornish, J.; Callon, K. E.; Lin, C. Q. X.; Xiao, C. L.; Mulvey, T. B.; Cooper, G. J. S.;
Reid, I. R. Am. J. Physiol. Endocrinol. Metab. 1999, 277, E779.
17. Ito, N.; Yamazaki, H.; Nakazaki, M.; Miyahara, T.; Kozuka, H.; Sudo, H. Calcif.
Tissue Int. 1987, 40, 200.
18. Kartsogiannis, V.; Ng, K. W. Mol. Cell. Endocrinol. 2004, 228, 79.
19. Julien, M.; Khairoun, I.; LeGeros, R. Z.; Delplace, S.; Pilet, P.; Weiss, P.; Daculsi,
G.; Bouler, J. M.; Guicheux, J. Biomaterials 2007, 28, 956.
20. Sugimoto, E.; Yamaguchi, M. Biochem. Pharmacol. 2000, 59, 471.
21. Walsh, S.; Jordan, G. R.; Jefferiss, C.; Stewart, K.; Beresford, J. N. Rheumatology
2001, 40, 74.
References and notes
1. Baron, R.; Hesse, E. J. Clin. Endocrinol. Metab. 2012, 97, 311.
2. Kraenzlin, M. E.; Meier, C. Nat. Rev. Endocrinol. 2011, 7, 647.
3. Rachner, T. D.; Khosla, S.; Hofbauer, L. C. Lancet 2011, 377, 1276.
4. Sambrook, P.; Cooper, C. Lancet 2006, 367, 2010.
5. Tong, Y.; Luo, J. G.; Wang, R.; Wang, X. B.; Kong, L. Y. Bioorg. Med. Chem. Lett.
2012, 22, 1908.
6. White, C. J.; Yudin, A. K. Nat. Chem. 2011, 3, 509.
7. Davies, J. S. J. Pept. Sci. 2003, 9, 471.
8. Lambert, J. N.; Mitchell, J. P.; Roberts, K. D. J. Chem. Soc., Perkin Trans. 1 2001, 471.
22. Cornish, J. Biometals 2004, 17, 331.
9. Rizo, J.; Gierasch, L. M. Annu. Rev. Biochem. 1992, 61, 387.
Please cite this article in press as: Kaur, H.; et al., Tetrahedron (2014), http://dx.doi.org/10.1016/j.tet.2014.04.035