Development of three-component conjugates: To get nano-globes with porous surfaces, high in vivo anti-osteoporosis activity and minimal side effects

Article abstract of DOI:10.1039/c2jm34370c

Three novel three-component conjugates of a succinyl spacer, a pharmacophore of 17β-amino-11α-hydroxyl-androst-1,4-diene-3-one, and a targeting sequence of RGD-tetrapeptide (peptide Arg-Gly-Asp-AA) were designed, synthesized, and evaluated. This paper presents evidence that inserting a succinyl functional group into the pharmacophore and the targeting sequence improves the oral anti-osteoporosis activities in mouse glucocorticoid model of secondary osteoporosis, which were reflected by dry weight, ash weight, Ca ²⁺ and bone mineral content of the femur of examined mice. These novel conjugates have no side effects on estrogen, and form nano-globes with a porous surface. This paper emphasizes that the hydrogen bond between the carbonyl group of the succinyl group and the carboxylic group of the C-terminal amino acid residue of RGD-tetrapeptides is the key to forming pores on the surface of the nano-globes of these novel conjugates.

Full text of DOI:10.1039/c2jm34370c

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links <
C
Journal of
Materials Chemistry
Cite this: J. Mater. Chem., 2012, 22, 21740
www.rsc.org/materials
PAPER
Development of three-component conjugates: to get nano-globes with porous
surfaces, high in vivo anti-osteoporosis activity and minimal side effects†
*
Guifeng Kang,‡ Yuji Wang,‡ Jiawang Liu, Jianhui Wu, Ming Zhao, Guochun Li, Ning Li, Li Peng,
*
Xiaoyi Zhang, Li Li, Nathan Mair and Shiqi Peng
Received 5th July 2012, Accepted 31st August 2012
DOI: 10.1039/c2jm34370c
Three novel three-component conjugates of a succinyl spacer, a pharmacophore of 17b-amino-11a-
hydroxyl-androst-1,4-diene-3-one, and a targeting sequence of RGD-tetrapeptide (peptide Arg-Gly-
Asp-AA) were designed, synthesized, and evaluated. This paper presents evidence that inserting a
succinyl functional group into the pharmacophore and the targeting sequence improves the oral anti-
osteoporosis activities in mouse glucocorticoid model of secondary osteoporosis, which were reflected
by dry weight, ash weight, Ca2+ and bone mineral content of the femur of examined mice. These novel
conjugates have no side effects on estrogen, and form nano-globes with a porous surface. This paper
emphasizes that the hydrogen bond between the carbonyl group of the succinyl group and the
carboxylic group of the C-terminal amino acid residue of RGD-tetrapeptides is the key to forming
pores on the surface of the nano-globes of these novel conjugates.
for the onset and progression of osteoporosis in postmenopausal
1. Introduction
women. Therefore, estrogens or hormone replacement therapy
(HRT) have been used clinically because they increase the
activity of osteoblasts and decrease the activity of osteoclasts.20,21
However, the potential increase in the risk of thrombosis and
cancers limit the clinical use of estrogens. RGD-tetrapeptides
(peptide Arg-Gly-Asp-AA) block the adhesiveness of osteoclasts
onto bone surface, which inhibits bone resorption.22 To improve
estrogen therapy and reduce the side effect profile (endometrial
hyperplasia and thrombosis, etc.) estrogen was conjugated with
RGD-tetrapeptides in our previous studies.23,24
Nano-scale assembly is important for forming various ordered
structures in solution, at bulk state, and on a solid surface which
includes spherical, lamellar, and honeycomb structures.1 The
development of nano-species pushes self-assembly into a new
field.2 At the nano-scale, numerous self-assembling molecules
can be converted into nano-particles. These molecules include
highly fluorinated amphiphilic molecules,3 amphiphilic triblock
copolymers with polyrotaxane as a central block,4 amphiphilic
dodecyl ester derivatives from aromatic amino acids,5 dendritic
molecules,6 the shape anisotropy of non-spherical colloidal
building blocks,7 alkylated polycyclic aromatic hydrocarbons,8
porphyrins, graphenes and fullerenes.9 Of these molecules,
peptides have particular importance,10–17 due to their ability to
form nano-structures and having excellent bioactivity.18,19
It is well known that bone formation and bone resorption are
mediated by osteoblasts and osteoclasts, respectively. Osteopo-
rosis occurs when the body fails to form enough new bone or too
much old bone is reabsorbed by the body. All postmenopausal
women confront osteoporosis to varying extents (primary oste-
oporosis or postmenopausal osteoporosis), with over 50%
suffering from bone fractures and secondary complications. The
lack of estrogen resulting from ovarium atrophy is responsible
Primary osteoporosis is common in premenopausal women,
while secondary osteoporosis (senile osteoporosis) occurs after
the age of 75 and is seen in both females and males. The
secondary causes of osteoporosis in older men account for 50–
80% of the cases of bone loss leading to fracture.25 Many prostate
cancer patients develop secondary osteoporosis when taking
glucocorticoids during androgen deprivation therapy (ADT).26,27
Glucocorticoids are ubiquitously prescribed in the fields of
rheumatology, respirology, neurology, hematology, derma-
tology, gastroenterology, and transplant medicine.28 Chronic
exposure to pharmacological doses of glucocorticoids causes
multiple deleterious effects on osteopenia, osteoporosis, and
bone fracture, in particular.29 Carcinoma of prostate is one of the
most common diseases in older men. After the surgery or radi-
ation therapy the male patients with localized or metastatic
prostate cancer generally suffer from ADT.30 Though male
patients on ADT usually have good prognosis, osteoporosis is a
very common consequence of this therapy. Up to 20% of these
patients will have a bone fracture within 5 years.31 The preven-
tion of osteoporotic fracture in the female patients treated with
College of Pharmaceutical Sciences, Capital Medical University, Beijing
100069, P.R.China. E-mail: sqpeng@mail.bjmu.edu.cn; mingzhao@mail.
bjmu.edu.cn; Fax: +86-10-8391-1528; +86-10-8280-2482; Tel: +86-10-
8391-1528; +86-10-8280-2482
† Electronic supplementary information (ESI) available. See DOI:
10.1039/c2jm34370c
‡ These authors contributed equally to this work.
21740 | J. Mater. Chem., 2012, 22, 21740–21748
This journal is ª The Royal Society of Chemistry 2012

View Article Online
glucocorticoids and the male patients receiving ADT needs novel
effective agents.32–34
spacer, a pharmacophore of 17b-amino-11a-hydroxyl-androst-
1,4-diene-3-one and a targeting sequence of RGD-tetrapeptide.
The anti-osteoporosis activity and the side effects were evaluated
in vivo, the relationship between the anti-osteoporotic efficacy
and the nano-structure was discussed.
Recently, the two-component conjugates of 17b-amino-11a-
hydro-xyandrost-1,4-diene-3-one, an anti-osteoporosis andro-
gene without the side effect of endometrial hyperplasia, and
RGD-tetrapeptides were reported.35 We demonstrated that these
conjugates effectively inhibited the development of osteoporosis
but did not induce endometrial hyperplasia and thrombosis.
Furthermore, in ultrapure water the conjugates of 17b-amino-
11a-hydro-xyandrost-1,4-diene-3-one with RGDS (Arg-Gly-
Asp-Ser, Fig. 1, left), RGDV (Arg-Gly-Asp-Val, Fig. 1, middle)
and RGDF (Arg-Gly-Asp-Phe, Fig. 1, right) formed nano-
globes of 55–200 nm in diameter, 24–182 nm in diameter, and 48–
188 nm in diameter, respectively, and most of which had smooth
surfaces.
2. Experimental
2.1. General
The protected amino acids with L-configuration were purchased
from Sigma Chemical Co. The coupling and deprotective reac-
tion was carried out in anhydrous conditions. Qingdao silica gel
H was used for chromatography. The purities of the intermedi-
ates were identified on TLC (Merck silica gel plates of type 60
F
254, 0.25 mm layer thickness) and were greater than 95%. The
To improve the self-assembly, which could be achieved with
the assistance of the intra-molecular hydrogen bond within the
RGD-tetrapeptide moiety, the structure of the two-component
conjugates mentioned above was modified by using a succinyl
group as a spacer connecting 17b-amino-11a-hydroxyandrost-
1,4-diene-3-one and RGD-tetrapeptide to form three novel
three-component conjugates. To predict if such intramolecular
hydrogen bonds between the succinyl oxygen and the carboxyl
group at the C-terminus of RGD-tetrapeptide could form and
improve the self-assembly, the computer-based hydrogen bond
investigation was performed by using the hydrogen bond
monitor method of Discovery Studio 2.1 (Accelrys, San Diego,
CA). The 3D structure of three-component conjugate consisted
of 17b-amino-11a-hydroxy-androst-1,4-diene-3-one, succinyl
group and RGDS or RGDV or RGDF was built with
3D-sketcher module. According to the requisition, if the distance
between the succinyl oxygen and the carboxyl group at the
C-terminus of RGD-tetrapeptide does not exceed the threshold
set by the distance criterion an intramolecular hydrogen bond
could be observed. The default value for the distance threshold is
2.5 angstroms. The computer-based intramolecular hydrogen
bond investigation gave positive prediction with desirable ster-
eoview (see ESI†). Due to this hydrogen bond the peptide chain
will form the hydrophilic ring. In water the hydrophobic 17b-
amido-11a-hydroxyl-androst-1,4-diene-3-one and the hydro-
philic ring could form the pores on the surface of the nano-globes
(Fig. 2). Since the pores on the surface of the nano-globes could
help the nano-globes adhering to the bone surface and thus this
probably benefits the in vivo anti-osteoporosis. In this context the
present paper provided three-component conjugates of a succinyl
purities of the products were identified on HPLC (Waters, C18
column 4.6 ꢀ 150 mm) and were greater than 97%. All reagents
with commercial quality were used as received. Solvents were
dried and purified using standard techniques when it was
necessary. Reactions were monitored by thin layer chromatog-
raphy on glass plates coated with silica gel with a fluorescent
indicator. Melting points were measured on a XT5 hot stage
microscope (Beijing keyi electrooptic factory) and were uncor-
rected. Infrared spectra were recorded on a spectrophotometer
(Perkin-Elmer 983 instrument), and the solid compounds were
compressed into KBr pellets. 1H (500 M Hz) and 13C (125 M Hz)
NMR spectra were recorded on a Bruker Avance II 500 MHz
spectrometer by using DMSO-d6 or CDCl3 as the solvent and
tetramethylsilane as an internal standard. ESI/MS was tested on
ZQ2000 (Waters, US). The content of calcium was measured on
a Varian SpectrAA 220 instrument. Total vBMD and images of
pQCT scanning and trabecular vBMD were measured on a
STRATEC XCT Research SA+ instrument. The statistical
analysis of all the biological date was carried out by use of
ANOVA test with p < 0.05 as significant cut-off.
2.2. Synthesis of 4a–c
As indicated in Scheme 1, via a four-step-procedure 4a–c were
prepared. In brief, the reduction of 11a-hydroxyandrost-1,4-dien-
3,17-dione resulted in 17b-amino-11a-hydroxyandrost-1,4-dien-
3,17-dione (1, 90% yield). The amidation of 1 and succinic anhy-
dride produced 17b-(30-carboxylpropionylamono)-androst-1,4-
diene-3-one(2, 90%yield). Threeprotectedpeptides, i.e. Arg(Tos)-
Gly-Asp(OBzl)-Phe-OBzl, Arg(Tos)-Gly-Asp-(OBzl)-Val-OBzl,
Fig. 1 TEM-images of the two-component conjugates of 17b-amino-11a-hydro-xyandrost-1,4-diene-3-one with RGDS (left), RGDV (middle) and
RGDF (right). R of the formula represents the side chain of S, V or F.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 21740–21748 | 21741

View Article Online
Fig. 2 Graphical representation of the assembly of 17b-(aminoethylcarbonyl-Arg-Gly-Asp-AA)-androst-1,4-diene-3-one in water to form a porous
nano-globe, R of the formula represents the side chain of S, V or F.
and Arg(Tos)-Gly-Asp(OBzl)-Ser-OBzl, were prepared by using a
solution-phase method and followed a strategy from C-terminus
to N-terminus. The coupling of 2 and the protected peptides
formed 17b-[aminoethyl carbonyl-Arg-(Tos)-Gly-Asp-(OBzl)-
AA-OBzl]-androst-1,4-diene-3-ones (3a–c, AA ¼ Phe, Val or Ser,
in 86–94% yield). In the presence of CF3CO2H/CF3SO3H all the
protective groups of 3a–c were removed and provided 4a–c with a
76–82% yield.
phosphorus) and running water. Each mouse in the prednisone
control group was injected with 6.3 mg kgꢂ1 of prednisone
intramuscularly twice a week and orally administered 0.2 mL of
distilled water once a day. Each mouse in the treated group was
injected with 6.3 mg kgꢂ1 of prednisone intramuscularly twice a
week and orally administered 110 nmol kgꢂ1 of 4a–c in 0.2 mL of
distilled water once a day. All mice were treated according to our
protocol for 4 weeks. On the day after the last administration the
mice were weighed, the blood was drawn from the eye orbit
and then the mice were sacrificed with pentobarbital sodium
(40.0 mg kgꢂ1, i.p.) anesthesia. The mice were then immediately
dissected to obtain the femur and organ.
2.3. In vivo anti-osteoporosis and side effect assay
2.3.1. Animals and administration. The described assessments
herein were performed according to a protocol reviewed and
approved by the ethics committee of Capital Medical University.
The committee assured that the welfare of the animals was in
conformity with the requirements of the animal welfare act and
according to the guide for care and use of laboratory animals.
ICR mice (weighing 21.1 ꢁ 1.8 g, purchased from Animal Center
of Peking University) were housed under temperature and
humidity controlled conditions at a 12-hour light and 12-hour
dark (6 p.m. to 6 a.m.) cycle for one day before being used. Each
mouse in drug receiving groups was administered special food
(containing 0.1% of calcium and 0.4% of phosphorus) and
distilled water, while each mouse in blank control was adminis-
tered normal food (containing 1.76% of calcium and 1.01% of
2.3.2. Total vBMD and trabecula vBMD. Patients receiving
glucocorticoid therapy and androgen deprivation therapy (ADT)
usually suffered from secondary osteoporosis.33,34,36,37 Chronic
exposure to treatment doses of prednisone caused multiple
deleterious effects on osteopenia, osteoporosis and bone fracture
in particular.38–40 The activity of agents was measured with the
total volumetric bone mineral density (vBMD) and trabecula
vBMD.8–10 Computed tomography (CT) reflects the 3D bone
geometry and the size-independent vBMD. Peripheral quanti-
tative CT (pQCT) gives quantitative 3D bone geometry and the
size-independent vBMD. With pQCT tests we used NS (normal
saline) alone as the background control and NS plus prednisone
Scheme 1 Synthetic route of 17b-(aminoethylcarbonyl-Arg-Gly-Asp-AA)-androst-1,4-diene-3-ones. AA ¼ Ser, Val or Phe.
21742 | J. Mater. Chem., 2012, 22, 21740–21748 This journal is ª The Royal Society of Chemistry 2012

View Article Online
as the blank control to identify the total vBMD and trabecular
anhydrous ethanol was added to assist with the removal of water.
At first the grid was dried thoroughly in air and then at 35 ꢃC for
24 h to get the sample for photography under the TEM. The
shape and size of the TEM image was obtained from counting
over 100 species in a randomly selected region. All determina-
tions were carried out in triplicate grids. The TEM was operated
at 80 kV, electron beam accelerating voltage. The nano-images
were recorded on an imaging plate (Gatan Bioscan Camera
Model 1792) with 20 eV energy windows at 6000–400 000ꢀ. The
images were digitally enlarged.
vBMD in the femur of the mice receiving 4a–c plus prednisone.
2.3.3. Bone weight and mineral content. After the mice were
sacrificed the left femur of the mice was collected immediately.
After the complete removal of the muscle, the length of the femur
was measured and then immersed in a mixed solution of chlo-
roform and methanol (2 : 1) twice (each 3 h). To remove the fat
ꢃ
the femur was heated at 120 C for 6 h, and then cooled. After
recording dry weight the femur was calcined in a furnace at
800 ꢃC for 8 h and then cooled to weigh the ash weight. The
weight ratio of the ash to dry femur (the mineral content of the
femur) was calculated. The femur ash was dissolved in 0.5 mL of
hydrochloric acid (6 N) and diluted to 5.00 mL with ultrapure
water, from which 0.05 mL of the solution was drawn and diluted
to 1.00 mL with ultrapure water before use. The calcium content
of the aqueous solution was measured with the method of
o-methyl-phenolphthalein complexing ketone.41,42 The phos-
phorous content of the aqueous solution was measured with the
use of molybdenum blue.43,44
2.4.3. The size of the nano-particle of 4a–c in normal saline
(NS). The size of the nano-particle of 4a–c in NS was measured
on a Malvern’s Zeta Sizer (Nano-ZS90) with DTS (Nano)
Program. The concentration of the solution of 4a–c in NS was
1.1 mM. The testing temperature was 25 ꢃC. The time interval was
30 s. The mean size and half-peak width of ten times were
recorded.
2.4.4. The zeta-potential of 4a–c in NS. The surface zeta-
potential of the nano-particle of 4a–c in NS was measured on a
ZetaPlus Potential Analyzer (ZetaPlus S/N 21394, Brookhaven
Instruments Coorpo-ration) with BIC Zeta Potential Analyzer.
The concentration of the solution of 4a–c in NS was 1.1 mM and
2.3.4. Bleeding ceased time. After the 4-week treatment of
estradiol, the mice from the groups 4a–c were weighed to record
the body weight. Thirty and ninety minutes after the last
administration, the mice were given an in vivo tail bleeding ceased
time assay following a standard procedure. In brief, the mouse
was placed in a tube holder with its tail protruding, and a 2 mm
cut was made on the tail. Flowing blood was gently wiped away
with a filter paper every 30 s until it stopped to yield the bleeding
ceased time.45,46
ꢃ
the testing temperature was 25 C. The zeta potential measure-
ment was repeated for 6 runs per sample, and the data was
calculated automatically using the software from the electro-
phoretic mobility based on Smoluchowski’s formula.
3. Results and discussion
2.3.5. Uterine weight. After the 4-week treatment of estradiol
the mice in groups 4a–c were sacrificed with pentobarbital
sodium (40.0 mg kgꢂ1, i.p.) anesthesia and dissected immediately
to obtain the uterine weight.
3.1. Total vBMD and trabecula vBMD of the femur of 4a–c
treated mice
The anti-osteoporosis efficacy of the mice in groups of 4a–c was
quantitatively measured by CT and pQCT to determine 3D bone
geometry and the size-independent vBMD. This is shown in
Fig. 3 and Fig. 4. In Fig. 3 we found that the total vBMD in the
femur of the mouse receiving NS plus prednisone was signifi-
cantly lower than that of the femur of the mouse receiving NS
alone. We can infer that prednisone effectively induced the
2.4. Characterizing the nano-image
2.4.1. Scanning electron microscope (SEM) image of 4a–c in
solid state. The characterization of the nano-image of the powder
from a 1.1 mM solution of 4a–c in ultrapure water was identified
on SEM (JEM-1230, JEOL, Tokyo, Japan) at 50 kV. The solu-
tion was attached on to a copper plate and lyophilized via
double-side tape (Euromedex, France). The specimens were
coated with 20 nm of gold–palladium using JEOL JFC-1600
AUTO FINE COATER. The coater was operated at 15 kV,
30 mA, 200 mTorr (argon) for 60 seconds. The shape and size
distribution of the nano-species were measured by counting more
than 100 particles in a randomly selected region on the SEM
alloy. All the measurements in triplicate grids were performed.
The images were recorded on an imaging plate (Gatan Bioscan
Camera Model 1792) with 20 eV energy windows at 100–
10 000ꢀ, and were digitally enlarged.
Fig. 3 Total vBMD and images of pQCT scanning at a distance from
the proximal femur. Growth plate corresponding to <6% of the total
length of the femur of the treated mice. (a) Total vBMDs is represented
with mean ꢁ SD mg cmꢂ3, n ¼ 12, PDN ¼ prednisone, dose of 4a–c ¼
110 nmol kgꢂ1. (b) Compared to NS + PDN, 4b + PDN and 4c + PDN
p < 0.01; (c) compared to NS + PDN and 4c + PDN p < 0.01; (d)
compared to NS + PDN p < 0.01.
2.4.2. Transmission electron microscopy (TEM) images of 4a–
c in water. The characterization of the nano-image of 4a–c in
aqueous solution was identified on TEM (JSM-6360 LV, JEOL,
Tokyo, Japan). A 1.1 mM solution of 4a–c in ultrapure water was
dripped onto a former-coated copper grid and then a drop of
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 21740–21748 | 21743

View Article Online
respectively. It was found (Fig. 5) that the dry weight of the
femur of the mice receiving NS plus prednisone was significantly
lower than that of the femur of the mice receiving NS alone,
suggesting that prednisone effectively induced the mice to lose
bone mass. The dry weight of the femur of the mice receiving 4a–
c plus prednisone was significantly higher than that of the femur
of the mice receiving NS plus prednisone, suggesting that 4a–c
effectively inhibited prednisone treated mice to lose bone mass.
Similarly, the femur ash weight, femur Ca2+ and femur BMC of
the treated mice also suggested that 4a–c effectively inhibited
prednisone treated mice to lose bone mass. We also found the
activity order of 4a–c inhibiting prednisone treated mice to lose
femur’s dry weight, femur’s ash weight, femur’s Ca2+ and femur’s
BMC was 4a > 4b > 4c.
Fig. 4 Trabecular vBMD in the femurs of treated mice at a distance
from the proximal femur. Growth plate corresponding to <6% of the
total length of the femurs of the treated mice. Trabecular vBMD is rep-
resented with mean ꢁ SD mg cmꢂ3, n ¼ 12, PDN ¼ prednisone, dose of
4a–c ¼ 110 nmol kgꢂ1. (a) Compared to NS + PDN, 4b + PDN and 4c +
PDN p < 0.01; (b) compared to NS + PDN and 4c + PDN p < 0.01; (c)
compared to NS + PDN p < 0.01.
3.3. Side effect of 4a–c in vivo
Two side effects of estrogen are endometrial hyperplasia and
thrombosis. The risk of endometrial hyperplasia of 4a–c was
represented with the uterine weight of the treated mice and the
data are shown in Fig. 6. The thrombosis risk of 4a–c was rep-
resented with the tail bleeding time of the treated mice and the
data are shown in Fig. 7.
treated mice to have a lower vBMD. The total vBMDs of the
femur of the mice receiving 4a–c plus prednisone were signifi-
cantly higher than that of the femur of the mice receiving NS plus
prednisone. We can also infer that 4a–c effectively inhibited
prednisone treated mice to lose total vBMDs. In further analysis,
we compared the femur vBMD of the mice receiving 4a–c plus
prednisone. The total vBMDs of the femur of the mice receiving
4a plus prednisone was significantly higher than that of the femur
of the mice receiving 4b plus prednisone, and the total vBMDs of
the femur of the mice receiving 4b plus prednisone was signifi-
cantly higher than that of the femur of the mice receiving 4c plus
prednisone. It can be inferred that the activity order of inhibiting
vBMD loss in prednisone treated mice was 4a > 4b > 4c.
It was found (Fig. 6) that the uterine weight of the mice
receiving estradiol plus prednisone was significantly higher than
that of the mice receiving NS alone, suggesting that estradiol
induced the mice to develop endometrial hyperplasia. In contrast
to estradiol, the uterine weight of the mice receiving 4a–c plus
It was found (Fig. 4) that the trabecular vBMD of the femur of
the mice receiving NS plus prednisone was significantly lower than
that of the femur of the mice receiving NS alone, suggesting that
prednisone effectively induced the mice to lose trabecular vBMD.
The trabecular vBMD of the femur of the mice receiving 4a–c plus
prednisone was significantly higher than that of the femur of the
mice receiving NS plus prednisone, suggesting that 4a–c effec-
tively inhibited prednisone treated mice to lose trabecular vBMD.
In further analysis, we compared the femur trabecular vBMD of
the mice receiving 4a–c plus prednisone. The trabecular vBMD of
the femur of the mice receiving 4a plus prednisone was signifi-
cantly higher than that of the femur of the mice receiving 4b plus
prednisone, and the trabecular vBMD of the femur of the mice
receiving 4b plus prednisone was significantly higher than that of
the femur of the mice receiving 4c plus prednisone. It can be
inferred that the activity order of inhibiting trabecular vBMD loss
in prednisone treated mice was 4a > 4b > 4c.
Fig. 5 Dry weight, ash weight, Ca2+ and BMC of the femur of 4a–c
treated mice. Weight of dry femur and femur ash are represented with
mean ꢁ SD mg, femur Ca2+ and BMC are represented with mean ꢁ SD
%, n ¼ 12, PDN ¼ prednisone, dose of 4a–c ¼ 110 nmol kgꢂ1. For dry
femur: (a) compared to NS + PND, 4b + PND and 4c + PND p < 0.01;
(b) compared to NS + PND and to 4c + PND p < 0.01; (c) compared to
NS + PND p < 0.01. For femur ash: (a) compared to NS + PND, 4b +
PND and 4c + PND p < 0.01; (b) compared to NS + PND and to 4c +
PND p < 0.01; (c) compared to NS + PND p < 0.01; (c) compared to NS
+ PND p < 0.01. For femur BMC: (a) compared to NS + PND, 4b +
PND, and 4c + PND p < 0.01; (b) compared to NS + PND and to 4c +
PND p < 0.01; (c) compared to NS + PND p < 0.01. For femur Ca2+: (a)
compared to NS + PND, 4b + PND and 4c + PND p < 0.01; (b)
compared to NS + PND p < 0.01, to 4c + PND p > 0.05; (c) compared to
NS + PND p < 0.01.
3.2. Dry weight, ash weight, Ca2+ and bone mineral content
(BMC) of the femur of 4a–c treated mice
Physical and chemical tests were widely used in the in vivo anti-
osteoporosis assay. To explore the dry weight, ash weight, Ca2+
and BMC of the femur of 4a–c treated mice, we used NS alone
and NS plus prednisone as the background and blank control
21744 | J. Mater. Chem., 2012, 22, 21740–21748
This journal is ª The Royal Society of Chemistry 2012

View Article Online
significantly higher than that of 5a–c treated mice. The data
suggest that, in respect to inhibiting the prednisone treated mice
to lose trabecular vBMD, the efficacy of 4a–c is significantly
higher than that of 5a–c. On the other hand, with respect to
inhibiting the prednisone treated mice to lose total vBMD, the
efficacy of 4a and 4b is significantly higher than that of 5a and 5b.
This increased efficacy of 4a–c could be attributed to the inser-
tion of a succinyl group and to the formation of three-component
conjugate.
Fig. 9 indicated that at 110 nmol kgꢂ1 of dose the dry weight of
the femur of the mice receiving 4a–c plus prednisone was
significantly higher than that of the femur of the mice receiving
5a–c plus prednisone. These data suggest that the efficacy of 4a–c
in inhibiting prednisone treated mice to lose bone mass is
significantly higher than that of 5a–c. Similarly, the femur ash
weight, femur Ca2+ and femur BMC of 4a–c plus prednisone
treated mice is significantly higher than that of the femur of 5a–c
plus prednisone treated mice. These data also suggest that the
efficacy of the 4a–c in inhibiting prednisone treated mice to lose
bone mass is significantly higher than that of 5a–c. Again, this
increased efficacy of 4a–c could be attributed to the insertion of a
succinyl group and to the formation of three-component
conjugate.
Fig. 6 Uterine weights of 4a–c treated mice. Uterine weight is repre-
sented by mean ꢁ SD mg, n ¼ 12, PDN ¼ prednisone, E2 ¼ estradiol.
Dose of E2 and 4a–c: 110 nmol kgꢂ1. (a) Compared to NS alone p < 0.01;
(b) compared to NS alone p > 0.05.
prednisone was equal to that of the mice receiving NS alone,
suggesting that 4a–c had no endometrial hyper-plasia risk.
When measuring tail bleeding time we found (Fig. 7) that
30 min and 90 min after the last oral administration the tail
bleeding time of the mice receiving estradiol plus prednisone was
significantly shorter than that of the mice receiving NS plus
prednisone. This suggests that estradiol had thrombosis risk.
However, 30 min and 90 min after the last oral administration the
tail bleeding time of the mice receiving 4a–c plus prednisone was
equal to that of the mice receiving NS plus prednisone, sug-
gesting that 4a–c had no thrombosis risk.
3.5. SEM and TEM nano-image of 4a–c
SEM and TEM have been widely used for the investigation of the
three-dimensional structure of hydrogels, cells entrapped in
nano-fiber matrices, or the individual nano-structure of the
peptides. To visualize the size and morphology of 4a–c in solid
state and in solution their SEM-image and TEM-image were
measured and are shown in Fig. 10–12 and Fig. 13–15,
respectively.
3.4. Comparing the in vivo activity and side effects of 4a–c with
two-component conjugates 5a–c
The in vivo anti-osteoporosis activity of 4a–c was further
compared with that of 17b-(RGD-AA-amido)-11a-hydroxyan-
drost-1,4-diene-3-one (5a, 5b and 5c, herein AA represents Phe,
Val and Ser, respectively),35 and the data are shown in Fig. 8 and 9.
Fig. 8 indicated that at 110 nmol kgꢂ1 of dose the total vBMD
and trabecular vBMD of the femur of 4a–c treated mice was
Fig. 8 Total vBMD and trabecular vBMD of the femur of 4a–c and 5a–c
treated mice. Total and trabecular vBMDs are represented with mean ꢁ
SD mg cmꢂ3, n ¼ 12, PDN ¼ prednisone, dose of 4a–c and 5a–c ¼
110 nmol kgꢂ1. (a) Compared to NS + PDN and 5a + PDN p < 0.05; (b)
compared to NS + PDN and 5b + PDN p < 0.05; (c) compared to NS +
PDN and 5c + PDN p < 0.05.
Fig. 7 Effect of 4a–c on the tail bleeding time of the treated mice. Tail
bleeding time is represented by mean ꢁ SD s, n ¼ 12, PDN ¼ prednisone,
E2 ¼ estradiol. (a) Compared to NS + PDN p < 0.01; (b) compared to
NS + PDN p > 0.05.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 21740–21748 | 21745

View Article Online
Fig. 11 SEM image of 4b in solid state. Globe of 2.9 mm in diameter
(left), globes of 5.0–6.4 mm in diameter and eggs of 5.7–12.9 mm in length
(middle), as well as globes of 2.8–3.5 mm in diameter (right).
correlating the molecular constitutions with the nano-structures
of 4a–c we got an impression that 17b-ethylcarbonylamino-
androst-1,4-diene-3-one was responsible for forming a globe-like
body, and RGD-tetrapeptide was responsible for characterizing
globe-like body.
The TEM images (Fig. 12–14) demonstrate that in water 4a–c
consistently forms nano-globes with porous surfaces. When
comparing the nano-globes of 17b-(Arg-Gly-Asp-AA-amido)-
11a-hydroxyandrost-1,4-diene-3-ones,35 which had no porous
surface, with the nano-globes of 4a–c, which had a porous
surface, we got an impression that the insertion of succinyl group
was key to forming the nano-globes with porous surface,
17b-ethyl-carbonylaminoandrost-1,4-diene-3-one was respon-
sible for forming nano-globe structures, and RGD-tetrapeptide
was responsible for characterizing the surface feature and size of
the nano-globes, in particular. For instance, RGDS causes 4a to
form dispersing nano-globes of 8–150 nm in diameter and having
porous surfaces, RGDV causes 4b to form dispersing nano-
globes of 29–150 nm in diameter and having porous surfaces, and
RGDF causes 4c to form dispersing nano-globes of 76–343 nm in
diameter and having porous surfaces.
Fig. 9 Dry weight, ash weight, Ca2+ and BMC of the femur of 4a–c and
5a–c treated mice. Weight of dry femur and femur ash are represented
with mean ꢁ SD mg, femur Ca2+ and BMC are represented with mean ꢁ
SD %, n ¼ 12, PDN ¼ prednisone, dose of 4a–c and 5a–c ¼ 110 nmol
kgꢂ1. (a) Compared to NS + PDN and 5a + PDN p < 0.01; (b) compared
to NS + PDN and 5b + PDN p < 0.01; (c) compared to NS + PDN and 5c
+ PDN p < 0.01.
3.6. Sizes and zeta-potential of nano-particles of 4a–c
As seen in Fig. 16, during the monitored 10 days the mean
diameter of the particles of 4a–c in NS gradually decreased from
250 nm to 119 nm. The mean diameter less than 250 nm could be
attributed to the hydrophobicity of 17b-ethylcarbonylamino-
androst-1,4-diene-3-one that could cause a stronger hydrophobic
interaction and form aggregators associated with closer
packing density. Moreover, the order of the mean diameter was
4c > 4b > 4a.
The SEM image (Fig. 10–12) demonstrates that in solid state
4a–c exist as nano-globes of 15 nm–6.4 mm in diameter, nano-
pine seeds of 286 nm–2.7 mm in length, nano-eggs of 1.3–12.9 mm
in length, nano-pinecones of 5.0–5.6 mm in length, nano-gear of
10 mm in diameter, nano-calabash of 4 mm in length, and
uncompleted nano-calabash of 11.3 mm in length.
The zeta-potential of the nano-particles of 4a–c in NS was in a
range of 110–152 mV during the monitored 10 days, and the
order of the zeta-potential was 4a > 4b > 4c (Fig. 17). This order
matches the order of the mean diameters, i.e., 4a < 4b < 4c. The
lower diameter and the higher zeta-potential imply that the
formed nano-particle is a stable nano-structure and could benefit
the in vivo transportation.
The coexistence of nano-globes having nano-egg, nano-pine
seed and nano-pinecone structures, and uncompleted nano-
calabash having nano-calabash structure, could imply that the
nano-egg, nano-pinecone and nano-calabash are built by nano-
globes, nano-pine seeds and uncompleted nano-calabash. When
Fig. 10 SEM image of 4a in solid state. Nano-globes of 15 nm–2 mm in
diameter and nano-calabash of 4 mm in length (left), nano-globes of
600 nm–1.3 mm in diameter and nano-egg of 1.3 mm in length (middle), as
well as nano-globes of 3.0–5.0 mm in diameter and uncompleted nano-
calabash of 11.3 mm in length (right).
Fig. 12 SEM image of 4c in solid state. Nano-pine seeds of 286–625 nm
in length and nano-pinecones of 5.0–5.6 mm in length (left and middle), as
well as nano-pine seeds of 0.7–2.7 mm in length and nano-gear of 10 mm in
diameter (right).
21746 | J. Mater. Chem., 2012, 22, 21740–21748
This journal is ª The Royal Society of Chemistry 2012

View Article Online
Fig. 13 TEM images of 1.1 mM of 4a in ultrapure water. The dispersing
globes of 8–150 nm in diameter (4a-A), the dispersing globes of 17–94 nm
in diameter (4a-B), as well as the dispersing globes of 27–82 nm in
diameters (4a-C).
Fig. 17 Zeta-potentials of the particles of 4a–c within 10 days.
4a–c have porous surfaces. It could be hypothesized that the
nano-globes with porous surfaces benefited 4a–c in increasing the
oral anti-osteoporosis efficacy and in decreasing the oral side
effect profile. The desirable diameter (<100 nm) of the most
nano-globes may help 4a–c to escape phagocytosis from
macrophages.47
Fig. 14 TEM images of 1.1 mM of 4b in ultrapure water. The dispersing
globes of 29–69 nm in diameter (4b-A), the dispersing globes of 70–
120 nm in diameter (4b-B) and the dispersing globes of 67–150 nm in
diameter (4b-C).
Acknowledgements
This work was finished in Beijing Area Major Laboratory of
Peptide and Small Molecular Drugs, supported by PHR (IHLB,
KZ200910025-004), Innovation Platform Project of Education
Committee of Beijing, Special Project (2011ZX09302-007-01),
and Natural Scientific Foundation of China (81072522).
References
1 H. Li, B. Song, L. Qin, Q. Liu, L. Wu and J. Shen, J. Colloid Interface
Sci., 2005, 290, 557.
2 S. C. Glotzer, M. A. Horsch, C. R. Iacovella, Z. Zhang, E. R. Chan
and X. Zhang, Curr. Opin. Colloid Interface Sci., 2005, 10, 287.
3 R. T. M. Jakobs, J. van Herrikhuyzen, J. C. Gielen,
P. C. M. Christianen, S. C. J. Meskers and A. P. H. J. Schenning,
J. Mater. Chem., 2008, 18, 3438.
Fig. 15 TEM images of 1.1 mM of 4c in ultrapure water. The dispersing
globes of 320–343 nm in diameter (4c-A), the dispersing globes of 76–
139 nm in diameter (4c-B) and the dispersing globes of 120–171 nm in
diameter (4c-C).
4 X. Zhang, X. Zhu, F. Ke, L. Ye, E. Chen, A. Zhang and Z. Feng,
Polymer, 2009, 50, 4343.
5 R. Vijay, S. Angayarkanny and G. Baskar, Colloids Surf., A, 2008,
317, 643.
6 E. Mehdipoor, M. Adeli, M. Bavadi, P. Sasanpour and B. Rashidian,
J. Mater. Chem., 2011, 21, 15456.
7 S. Sacanna and D. J. Pine, Curr. Opin. Colloid Interface Sci., 2011, 16,
96.
ꢀ
8 V. Palermo, M. Palma, Z. Tomovic, M. D. Watson, K. Mullen and
ꢁ
P. Samorı, Synth. Met., 2004, 147, 117.
€
ꢂ
9 S. Gadipelli, I. Calizo, J. Ford, G. Cheng, A. R. Hight Walker and
T. Yildirim, J. Mater. Chem., 2011, 21, 16057.
10 L. S. Witus, J. R. Rocha, V. M. Yuwono, S. E. Paramonov,
R. B. Weisman and J. D. Hartgerink, J. Mater. Chem., 2007, 17, 1909.
11 G. Palui, S. Ray and A. Banerjee, J. Mater. Chem., 2009, 19, 3457.
12 V. Castelletto and I. W. Hamley, Biophys. Chem., 2009, 141, 169.
€
13 H. G. Borner, Prog. Polym. Sci., 2009, 34, 811.
14 A. Carlsen and S. Lecommandoux, Curr. Opin. Colloid Interface Sci.,
2009, 14, 329.
15 E. K. Johnson, D. J. Adams and P. J. Cameron, J. Mater. Chem.,
2011, 21, 2024.
Fig. 16 Particle size of 4a–c in NS (1.1 mM) within 10 days.
16 N. Wiradharma, Y. W. Tong and Y. Yang, Biomaterials, 2009, 30,
3100.
4. Conclusions
17 V. Gribova, T. Crouzier and C. Picart, J. Mater. Chem., 2011, 21,
14354.
By using a succinyl group as a spacer of RGD-tetrapeptides and
17b-amino-11a-hydroxyandrost-1,4-diene-3-one, three novel
conjugates (4a–c) were constructed. In contrast to the nano-
globe of RGD-tetrapeptides and 17b-amino-11a-hydroxyan-
drost-1,4-diene-3-one without this spacer, the nano-globes of
18 N. Li, G. Kang, L. Gui, M. Zhao, W. Wang, J. Zhang, Y. Yue and
S. Peng, Nanomed.: Nanotechnol., 2011, 7, 403.
19 Y. Lim and M. Lee, J. Mater. Chem., 2008, 18, 723.
20 K. J. Majeska, J. T. Ryaby and T. A. Einhorn, J. Bone Jt. Surg., Am.
Vol., 1994, 76, 713.
This journal is ª The Royal Society of Chemistry 2012
J. Mater. Chem., 2012, 22, 21740–21748 | 21747

View Article Online
21 X. Y. Zang, Y. B. Tan, Z. L. Pang, W. Z. Zhang and J. Zhao, Chin.
Med. J., 1994, 107, 600.
22 M. A. Horton, M. L. Taylor, T. R. Arnett and M. H. Helfrich, Exp.
Cell Res., 1991, 195, 368.
23 Y. Xiong, M. Zhao, C. Wang, H. Chang and S. Peng, J. Med. Chem.,
2007, 50, 3340.
35 Y. Wang, J. Wu, G. Kang, M. Zhao, L. Gui, N. Li, L. Peng,
X. Zhang, L. Li and S. Peng, J. Mater. Chem., 2012, 22, 4652–
4659.
36 R. W. Ross and E. J. Small, J. Urol., 2002, 167, 1952–1956.
37 D. Cohen and J. D. Adachi, J. Steroid Biochem. Mol. Biol., 2004, 88,
337–349.
24 C. Li, S. Peng, M. Zhao, N. Li and F. Wu, Application no:
CN200710090291.8.
38 F. Manelli and A. Giustina, Trends Endocrinol. Metab., 2000, 11, 79–
85.
25 R. W. Ross and E. J. Small, J. Urol., 2002, 167, 1952–1956.
26 D. Cohen and J. D. Adachi, J. Steroid Biochem. Mol. Biol., 2004, 88,
337–349.
27 F. Manelli and A. Giustina, Trends Endocrinol. Metab., 2000, 11, 79–85.
28 S. M. H. Alibhai, S. Gogov and Z. Allibhai, Crit. Rev. Oncol.
Hematol., 2006, 60, 201–215.
39 S. M. H. Alibhai, S. Gogov and Z. Allibhai, Crit. Rev. Oncol.
Hematol., 2006, 60, 201–215.
40 R. A. Adlera, Maturitas, 2011, 68, 143–147.
41 N. Wang, C. Wu, Y. Wu and T. Xu, J. Membr. Sci., 2010, 363, 128–
139.
ꢀ
ꢀ
42 J. Siska, Hydrometallurgy, 2005, 76, 155–172.
29 R. A. Adlera, Maturitas, 2011, 68, 143–147.
€
43 K. Grudpan, P. Ampan, Y. Udnan, S. Jayasvati,
S. Lapanantnoppakhun, J. Jakmunee, G. D. Christian and
J. Ruzicka, Talanta, 2002, 58, 1319–1326.
30 B. Straub, M. Muller, M. Schrader, R. Heicappell and K. Miller, J.
Urol., 2001, 165, 1783–1789.
31 V. B. Shaninian, Y. F. Kuo, J. L. Freeman and J. S. Goodwin, N.
Engl. J. Med., 2005, 352, 154–164.
32 B. Frediani, P. Falsetti, F. Baldi, C. Acciai, G. Filippou and
R. Marcolongo, Bone, 2003, 33, 575–581.
33 A. Kelman, Best Pract. Res. Clin. Rheumatol., 2005, 19, 1021–1037.
34 H. Liu, N. M. Paige, C. L. Goldzweig, E. Wong, A. Zhou,
M. J. Suttorp, B. Munjas, E. Orwoll and P. Shekelle, Ann. Intern.
Med., 2008, 148, 685–701.
44 J. Huo and Q. Li, Spectrochim. Acta, Part A, 2012, 87, 293–297.
45 J. C. Martinichen-Herrero, E. R. Carbonero, P. A. J. Gorin and
M. Iacomini, Carbohydr. Polym., 2005, 60, 7–13.
46 J. C. Martinichen-Herrero, E. R. Carbonero, G. L. Sassaki,
P. A. J. Gorin and M. Iacomini, Int. J. Biol. Macromol., 2005, 35,
97–102.
47 Y. Fujita, M. Mie and E. Kobatake, Biomaterials, 2009, 30, 3450–
3457.
21748 | J. Mater. Chem., 2012, 22, 21740–21748
This journal is ª The Royal Society of Chemistry 2012