A Biphasic Calcium Sulphate/Hydroxyapatite Carrier Containing Bone Morphogenic Protein-2 and Zoledronic Acid Generates Bone

Abstract

In orthopedic surgery, large amount of diseased or injured bone routinely needs to be replaced. Autografts are mainly used but their availability is limited. Commercially available bone substitutes allow bone ingrowth but lack the capacity to induce bone formation. Thus, off-the-shelf osteoinductive bone substitutes that can replace bone grafts are required. We tested the carrier properties of a biphasic, calcium sulphate and hydroxyapatite ceramic material, containing a combination of recombinant human bone morphogenic protein-2 (rhBMP-2) to induce bone, and zoledronic acid (ZA) to delay early resorption. In-vitro, the biphasic material released 90% of rhBMP-2 and 10% of ZA in the first week. No major changes were found in the surface structure using scanning electron microscopy (SEM) or in the mechanical properties after adding rhBMP-2 or ZA. In-vivo bone formation was studied in an abdominal muscle pouch model in rats (n = 6/group). The mineralized volume was significantly higher when the biphasic material was combined with both rhBMP-2 and ZA (21.4 ± 5.5 mm3) as compared to rhBMP-2 alone (10.9 ± 2.1 mm3) when analyzed using micro computed tomography (μ-CT) (p < 0.01). In the clinical setting, the biphasic material combined with both rhBMP-2 and ZA can potentially regenerate large volumes of bone.

Full text

1
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
www.nature.com/scientificreports
A Biphasic Calcium Sulphate/
Hydroxyapatite Carrier Containing
Bone Morphogenic Protein-2 and
Zoledronic Acid Generates Bone
Deepak Bushan Raina1,2, Hanna Isaksson1,3, Werner Hettwer4, Ashok Kumar2, Lars Lidgren1 &
Magnus Tägil1
In orthopedic surgery, large amount of diseased or injured bone routinely needs to be replaced.
Autografts are mainly used but their availability is limited. Commercially available bone substitutes
allow bone ingrowth but lack the capacity to induce bone formation. Thus, o-the-shelf osteoinductive
bone substitutes that can replace bone grafts are required. We tested the carrier properties of a
biphasic, calcium sulphate and hydroxyapatite ceramic material, containing a combination of
recombinant human bone morphogenic protein-2 (rhBMP-2) to induce bone, and zoledronic acid (ZA) to
delay early resorption. In-vitro, the biphasic material released 90% of rhBMP-2 and 10% of ZA in the rst
week. No major changes were found in the surface structure using scanning electron microscopy (SEM)
or in the mechanical properties after adding rhBMP-2 or ZA. In-vivo bone formation was studied in an
abdominal muscle pouch model in rats (n = 6/group). The mineralized volume was signicantly higher
when the biphasic material was combined with both rhBMP-2 and ZA (21.4 ± 5.5 mm3) as compared to
rhBMP-2 alone (10.9 ± 2.1 mm3) when analyzed using micro computed tomography (μ-CT) (p < 0.01).
In the clinical setting, the biphasic material combined with both rhBMP-2 and ZA can potentially
regenerate large volumes of bone.
Bone regeneration is one of the most commonly explored areas for tissue engineering1. Partly due to a demo-
graphic shi towards an older population, there is an increased demand for bone gras in non-unions, large bone
defects in infections, aseptic prosthetic loosening with osteolysis, aer tumor surgery and in fragility fractures2.
In these situations, bone autogras have been used for decades to regenerate bone but the amount of autogra
is limited and the harvest of large quantities of autogra is associated with substantial morbidity3–5. e other
alternative to autogras is the use of allogras. However, their ecacy depends on the donor age as well as tissue
banking sources6–8, have the risk of disease transmission9 and are less ecacious compared to autogras. e
increasing demand and the absence of a viable solution for replacing large volumes of bone, poses a scientic
challenge that requires new innovative bone gra solutions.
Bone is a combination of both organic and inorganic components. Ceramic, polymer or composite materials
have been used to mimic the natural bone, all with the aim to restore bone and improve bone regeneration10,11.
Many osteoconductive scaolds allow some formation and ingrowth of bone from surrounding tissue, but in
large defects it remains a challenge to recruit and dierentiate the inducible cells to remodel the bone defect12,13.
Biomaterials designed for bone regeneration therefore are required to induce bone formation at the desired loca-
tions. Desired scaold pre-requisites include optimal physical properties such as sucient internal space for
new bone to grow in with space for the exchange of nutrients and gases, sucient mechanical stability, the right
surface properties and bioresorbability11,14,15. Also biological properties are important, and in particular signa-
ling molecules are required to recruit mesenchymal progenitor cells. ese molecules by preference should be
included already at the time of the setting of a bone substitute without additional steps.
1Department of Orthopedics, Clinical Sciences, Lund, Lund University, Lund, 221 85, Sweden. 2Department of
Biological Sciences and Bioengineering, Indian Institute of Technology Kanpur, Kanpur, 208016, India. 3Department
of Biomedical Engineering, Lund University, Lund, 221 85, Sweden. 4Department of Orthopedic Surgery,
Rigshospitalet, University of Copenhagen, Copenhagen, 2100, Denmark. Correspondence and requests for materials
should be addressed to L.L. (email: lars.lidgren@med.lu.se)
Received: 05 November 2015
Accepted: 26 April 2016
Published: 18 May 2016
OPEN

www.nature.com/scientificreports/
2
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
A number of inorganic bone substitutes have been used clinically11. Ceramic materials mimic the inorganic
components of bone but their ability to induce bone is limited if they are not used in a supportive local envi-
ronment or supplied with growth factors that serve as signaling molecules16–18. Autogras act as reservoirs of
important signaling molecules like the pro-osteogenic proteins from the transforming growth factor - β (TGF- β )
family19 at the bone defect but locally administered BMP treatment has also been explored11,20,21. In the few
randomized clinical non-union and spine fusion series, BMP’s have never been proven to induce bone heal-
ing better than autogra22,23. A possible explanation for this has been a rising insight into their function, and
identifying BMP’s as an inducer of not only bone formation but also bone resorption24 due to a RANKL-RANK
(osteoblast-preosteoclast) interaction leading to increased osteoclastogenesis25–27. We have previously shown
that it is possible to pharmacologically modulate the excessive bone resorption caused by the use of BMP,
without decreasing the increased bone formation, by adding osteoclast-inhibiting28,29 bisphosphonates16,20.
Bisphosphonates bind to the mineral phase of the bone with strong anity and when resorbed induce apoptosis
of osteoclasts28,30. Bisphosphonates today are administered systemically16,20, but there are unwanted side eects of
systemic treatment, like reduced bone remodeling31, osteonecrosis of the jaw32, gastric problems, u-like symp-
toms and a low risk of acute renal failure33,34 and local delivery of these drugs at the site of action is preferable.
e dosage, stability, delivery and release of BMP’s have always been a concern and dierent carriers and
methods have been suggested35–38. One of the most common methods has been soaking the carrier material in
a solution containing the protein, which leads to physical absorption of the protein to the material surface11,39.
is method has some limitations. e soaking time is not standardized, which may inuence the clinical eect40.
Moreover, the release kinetics depends on the type of carrier system being used and an optimal carrier system
has not been fully developed yet. A few biphasic preset carrier materials have been published recently includ-
ing porous polymer-inorganic composites41. However, these materials have the risk of antigenicity, uncontrolled
degradation, and insucient mechanical strength. Moreover, the non-injectability of these materials requires
invasive surgical procedures.
To overcome these drawbacks, we hypothesized that local co-delivery of soluble, carrier free rhBMP-2 and
ZA by means of physical entrapment or chemical binding in a ceramic, injectable biphasic carrier (Cerament™
Bone Void Filler)42,43 can improve the results. e soluble calcium sulphate phase will resorb over time and thus
release the osteoinductive protein initiating osteogenic dierentiation of mesenchymal progenitors. ZA bound to
the material will then protect the newly formed bone from early resorption caused by the addition of rhBMP-2.
e material mimics natural bone matrix and have several advantages such as high degree of protein encapsu-
lation, sustained release behavior and improved surgical handling. e biocompatibility and bioresorbability of
the biphasic microporous carrier that sets in situ makes it suitable as a carrier material with a controlled release
of encapsulated or chemically bound additives. e biphasic material is used clinically as bone void ller and
consists of hydroxyapatite (HA) and α - hemihydrate and dihydrate calcium sulphate mixed with a radiopaque
compound, Iohexol42. Aer injection, the components set in situ into a microporous, osteoconductive matrix.
e possibility of adding BMP’s or bisphosphonates to the biphasic material has been documented before44. ZA
is chemically bound to the HA while the rhBMP-2 is physically entrapped within the resorbing calcium sulphate
phase of the ceramic carrier. In this rst in-vitro and in-vivo trial, we incorporated both soluble rhBMP-2 and ZA
in the ceramic powder and analyzed the release and biological response with an aim to design an osteoinductive
tool for large bone reconstructions. We hypothesized that a gradual release of BMP’s from the biphasic carrier
could induce osteogenesis while chemically bound ZA will provide protection against premature bone resorption.
Results
In vitro rhBMP-2 release. A gradual but constant release of the protein was detected in the supernatants. At
day-3, approximately 50% of the protein was detected while nearly 90% of rhBMP-2 was found in the supernatant
on 7th day. Figure1A represents the in-vitro release kinetics of rhBMP-2 from the biphasic material.
In vitro ZA release. A dose dependent apoptosis of A549 cells aer treatment with ZA was observed in our
experiment (Fig.1B) with higher doses of free ZA causing apoptosis of A549 cells. Aer plotting a dose vs. cell
viability curve, a sustained and minimal release of ZA from the biphasic material was seen. Nearly 6% of ZA
originally loaded was released on day 1 that increased to a value of nearly 10% on day 7 (Fig.1C). Microscopically,
no signicant dierences were seen in the experimental group with cells showing healthy proliferation and mor-
phology when compared to controls (Fig.1B). ough a very little fraction of ZA was released by the material
over 7-days, the cytotoxicity (A549 cells) caused by the released fraction had a decreasing trend with time as seen
in Fig.1D.
Eect of bound zoledronic acid to the biphasic material. e proliferation of A549 cells on the bipha-
sic material with and without ZA was signicantly higher than A549 cells treated with free ZA (p < 0.01) on both
day 1 and 3 as shown in Fig.1E.
In vitro SEM analysis. e surface as well as the pore distribution across the three groups was similar on
day 0 (Fig.2). High magnication image showed a closed architecture of pores at the time of casting (d = 0). Aer
incubation in saline for 28-days, the samples appeared to be rougher at the surface and an open pore structure
was observed. However, the addition of rhBMP-2 or ZA appeared to have no eect on the surface structure of
the biphasic material and no dierences in the pore structure or surface morphology were observed between the
three groups even on day 28 (Fig.2A).
In vitro mechanical analysis. e stiness was higher in biphasic material + rhBMP-2 group when com-
pared to the group with the biphasic material alone (p < 0.05). ere were no signicant dierences in stiness

www.nature.com/scientificreports/
3
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
between biphasic material + rhBMP-2 + ZA group compared with biphasic material + rhBMP-2 group or bipha-
sic material alone. All groups showed similar values of absorbed energy (Fig.2B).
Assessment of ectopic bone formation. All animals were sacriced aer 4-weeks as described before.
One animal in rhBMP-2 and rhBMP-2 + ZA group had to be sacriced a few days aer implantation due to rup-
ture of the skin suture. A total of 10-samples per groups were harvested for rhBMP-2 and rhBMP-2 + ZA groups
while 12 samples were obtained in the group containing the biphasic material only.
Physical assessment of scaolds at harvest. e samples containing the biphasic material alone were
morbid and smaller than the implanted dimension. e samples in the group containing the biphasic mate-
rial + rhBMP-2 had increased dimensions. On applying a gentle force on the scaold, blood like uid oozed
out of the scaolds. e scaolds belonging to the biphasic material + rhBMP-2 + ZA group were the biggest in
dimensions by gross analysis. Moreover, they appeared to be hard and could not be compressed.
Radiography. Radiographically, the material alone had deformed with respect to its original shape at the
time of implantation (Fig.3). Biphasic material with rhBMP-2 + ZA was most dense radiographically (Fig.3).
Qualitatively, the dimensions of both rhBMP-2 and rhBMP-2 + ZA treated groups based on the radiopaque area
appeared larger than the material alone (Fig.3).
Micro computed tomography (μ-CT). The biphasic material loaded with a combination of
rhBMP-2 + ZA showed both quantitatively and qualitatively a higher amount of mineralized volume (remaining
biphasic material + newly formed bone) (21.4 ± 5.5 mm3) than the other groups (p < 0.01) (Fig.4). 3-D and 2-D
rendering of the slides shows that the mineralized areas are retained in this group (Fig.4). e biphasic mate-
rial + rhBMP-2 showed a higher amount of mineralized volume (10.9 ± 2.1 mm3) than the biphasic material alone
(Fig.4). However, the 2-D rendering in the middle panel shows a hollow core in the material treated with only
rhBMP-2 (Fig.4). e mineralized volume was the least in the biphasic material implanted alone (4.9 ± 0.9 mm3).
Histology and histomorphometry. e samples containing only the biphasic material did not show
any bone formation. It can be seen as brown to black crystals of HA covered by a layer of muscle and some
Figure 1. In-vitro rhBMP-2 and ZA release kinetics. (A) Indicates rhBMP-2 release (%) from the biphasic
material over a period of 7-days detected using ELISA. (B) Shows the microscopic eect of free ZA (increasing
concentrations using tissue culture plates (2D TCP)) and ZA released from the biphasic material (dierent day
fractions) on A549 lung cancer cells (Yellow arrow indicates healthy, epithelial morphology of cells, red arrow
points at round, dead cells while green arrow shows oating apoptotic bodies). (C,D) Fraction of ZA released
(%) from the biphasic material over a period of 7-days and the cytotoxicity induced in A549 cells by the released
fraction using the MTT assay, respectively. (E) Eect of bound and free ZA on A549 cells aer seeding the
cells directly on the biphasic material alone, in combination with ZA and plastic control treated with free ZA
using the MTT assay. **Indicates p < 0.01, #indicates non-signicant. Data is expressed as mean ± SD. Scale bar
represents 30 μ m.

www.nature.com/scientificreports/
4
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
brous tissue, although it was well inltrated with granular cells (Fig.5). On the other hand, samples containing
rhBMP-2 exhibit bone like morphology in many places (Fig.5). However, it appears to have undergone oste-
oclastic resorption characterized by the presence of fatty marrow cells visible in dierent areas of the scaold.
e group treated with a combination of rhBMP-2 and ZA has developed a cortical shell around the biphasic
material with islands of trabecular bone abundantly present all across the scaold and even in the middle (Fig.5).
Histomorphometrically, signicantly higher area of bone formation was seen in the biphasic material combined
with rhBMP-2 and ZA when compared to the biphasic material + rhBMP-2 group (p < 0.01).
Scanning electron microscopy (SEM). e implanted samples loaded with rhBMP-2 have developed a
hollow core in the middle with bone formation on the sides (Fig.6). Biphasic material treated with rhBMP-2 and
ZA appears to be intact and bony ossicles can be seen spread all across the material showing bone progression
towards the center of the scaold (Fig.6). Typical trabecular morphology was observed even in the middle of
the bone/material composite (Fig.6, right lower panel). Most of the bone/material composite is retained. Only
Figure 2. In-vitro SEM and mechanical analysis. (A) Surface architecture and pore morphology of the
biphasic material alone and with rhBMP-2 and rhBMP-2 + ZA was compared aer casting (Day 0) and aer
28-days of incubation in saline. SEM images on the le panels at Day 0 and day 28 have been captured at 500X
while images in the right panel depict high magnication images (8000X). e lower le panel (B) shows the
stiness of the biphasic material alone or aer addition of rhBMP-2 and ZA. e lower right panel (B) indicates
the absorbed energy by the samples in dierent groups. *Indicates p < 0.05, #indicates non-signicant. Data is
expressed as mean ± SD. n = 5.

www.nature.com/scientificreports/
5
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
biphasic material appears to have undergone no bone formation and porous material and HA particles can be
seen in Fig.6.
Mechanical compression. e stiness of the biphasic material alone was signicantly higher than the
group where the biphasic material was treated with rhBMP-2 (p < 0.001) (Fig.7). No signicant dierences were
found between biphasic material alone and biphasic material + rhBMP-2 + ZA groups. Also the biphasic mate-
rial + rhBMP-2 + ZA group was signicantly stier than biphasic material + rhBMP-2 group (p < 0.05). In terms
of absorbed energy, the biphasic material combined with rhBMP-2 and rhBMP-2 + ZA absorbed signicantly
more energy as compared to the material alone (p < 0.05). No signicant dierences between the rhBMP-2 and
rhBMP-2 + ZA groups were seen in terms of absorbed energy.
Discussion
e osteoblast eect and osteoinductive potential of BMP’s is well established45, but the simultaneous osteoclast
induction, causing premature resorption of the newly formed callus is less known20,24,27. Although bisphospho-
nates have primarily been given to increase or maintain bone mineral content in patients with low bone density,
in animal models, systemic bisphosphonates can also be used to block the BMP mediated bone resorption and
pharmacologically balance the anabolic and anti-catabolic eects of the two drugs16,21. Further, there is room
for improvements regarding the delivery of the protein and BMP’s have been delivered using carboxymethyl
cellulose (CMC) or bovine collagen particles16 as carriers , or by impregnation of collagen sponges12,39,41,46 and
other polymers47. ese carriers have limitations, like the impregnation time that can lead to varying clinical
outcomes. Further, the carrier collagen may cause an adverse tissue reaction, mainly due to a local inammatory
reaction. In the present study, we addressed both the premature resorption and the burst release and investigated
a one-stage method of delivering the combination of rhBMP-2 and ZA, locally at the site where bone formation
is needed. rhBMP-2 and ZA were encapsulated within a ceramic matrix consisting of hydroxyapatite and calcium
sulphate. We were able to show a constant but yet high rhBMP-2 release (Fig.1A), still at the end of 7-days in the
current in-vitro study. e biphasic nature of the material led to release of the soluble calcium dihydrate sulphate
containing the embedded protein. e SEM analysis showed that the samples have more pronounced pores and
a rough surface aer incubation in saline for 28 days (Fig.2A). Rough and porous materials have been reported
to support osteogenesis10,11. However, the material may have dierent release kinetics in-vivo enabling a gradual
and more extended release of the protein (due to lesser volume of body uids, hematoma and cellular inltration
around the material). e results from the in-vitro release experiment suggested that the protein did not interact
with the material by other means than physical entrapment, which can ensure its availability in recruiting and
dierentiating osteoprogenitor cells. Also, more than 90% of the embedded protein was recovered aer 7-days,
which also emphasizes high degree of protein encapsulation.
Additionally, the release of ZA from the biphasic material was measured in-vitro. Bisphosphonates, in contrast
to BMP are most oen delivered systemically16,20, and there are reports that local bisphosphonate treatment may
interfere with osteoblastic bone formation48, which seems to be possibly tackled by combining it with BMP’s21. A
few reports state successful co-delivery of BMP’s and ZA by a local scaold. ZA is known to induce apoptosis in
A549 cells in a dose dependent manner49 and we used a bioassay for ZA release kinetics in-vitro. It is known that
bisphosphonates in general, and third generation such as ZA in particular, have high anity for hydroxyapatite30
and we speculate that the concentration of HA in our material led to a strong binding of ZA leaving only a limited
Figure 3. X-ray radiography of implanted samples in the abdominal muscle pouch. X-ray radiographs
of biphasic material alone and combined with rhBMP-2 (10 μ g) and rhBMP-2 (10 μ g) + ZA (10 μ g) in the
abdominal muscle pouch aer 28-days of in-vivo implantation. Scale bar represents 1 cm.

www.nature.com/scientificreports/
6
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
amount of unbound ZA behind (Fig.1B–D). In our model, the strong binding of ZA to the hydroxyapatite is
advantageous in retaining high bone turnover. Our results also indicate that bound ZA does not seem to have
a cytotoxic eect on A549 cells when seeded directly on the biphasic material mixed with ZA (Fig.1E). is
rearms that only a limited amount of ZA is released since no negative eects were observed regarding cell pro-
liferation. A similar release pattern of ZA from a porous collagen-HA scaold has previously been shown to have
a protective eect on pre-osteoblastic cells by the bound ZA50. e mechanism is unclear but we speculate that
A549 cells do not interact with HA unlike the osteoclasts and thus the negative eects of ZA bound to HA were
not seen. In case of free ZA, available for the cells in the culture medium, the drug may enter the cell inducing
apoptosis.
No signicant delay was observed regarding the setting of the ceramic material which has been reported to be
a concern51. In-vitro mechanical compression (Fig.2B) showed no reduction in the stiness and absorbed energy
in the groups treated with rhBMP-2 and rhBMP-2 + ZA when compared to biphasic material only, contrary to
what has been reported earlier51. e in-vitro mechanical and SEM analysis of the samples alone or in combina-
tion with rhBMP-2 and ZA does not show signicant dierences between the groups. is strengthens the feasi-
bility of using this biphasic material clinically since the additives do not compromise the surface or mechanical
properties of the biphasic material.
e in-vivo ectopic bone formation model has been widely used to assess the carrier properties or osteoin-
ductivity of various scaolds. e doses for rhBMP-2 and ZA in our study were based on previous studies50,52.
In surgery, a one step mixing and delivery of a soluble rhBMP-2 makes this biphasic material an encouraging
carrier. No apparent loss in the bioactivity of protein has been observed due to physical entrapment in the set
calcium dihydrates and importantly, signicant bone formation was seen also with the low doses of rhBMP-2. In
a previous study using the ectopic bone formation model, we were able to lower the dose and still observe bone
formation at a 2-μ g rhBMP-2 added to the biphasic material (Raina DB, 2014, unpublished data). Delivery of low
Figure 4. Micro-computed tomography results 28-days post implantation. Images in the top panel represent
full 3-D rendering of the samples in the three groups (biphasic material, biphasic material + rhBMP-2 (10 μ g)
and biphasic material + rhBMP-2 (10 μ g) + ZA (10 μ g)) while images in the middle panel show sliced 2-D
images in the middle of the samples in order to emphasize on the internal content of the samples across dierent
groups. e bottom panel shows the mineralized volume in each group. **Indicates p < 0.01, #indicates non-
signicant. Data is expressed as mean ± SD. n = 5 for biphasic material + rhBMP-2 and rhBMP-2 + ZA groups
and n = 6 for biphasic material alone. Scale bar represents 1 mm.

www.nature.com/scientificreports/
7
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
doses of BMP remains an important goal by itself, given the recent report of negative eects with supraphysiolog-
ical doses presently used53.
The in-vitro ZA release kinetics is in accordance with the in-vivo results. While both rhBMP-2 and
rhBMP-2 + ZA treated groups were able to induce bone formation, signicant dierences were observed. e
ectopic bone found using radiography, μ -CT, histological, histomorphometry and SEM analysis are well in
accordance and clearly emphasize that the co-delivery of BMP-2 and ZA using the biphasic ceramic material
is successful in the current muscle model compared to previous studies20,21. ere are several studies that show
systemic bisphosphonate treatment also has similar protection on premature bone resorption21,54 and it is spec-
ulated, based on these studies, that systemic administration of ZA in this study would also have led to similar
results as with local delivery. However, the paradigm is shiing towards local ZA treatment keeping in view the
side eects of long term systemic bisphosphonate treatment50. Signicantly higher mineralized volume was found
in the biphasic material + rhBMP-2 + ZA group as seen from μ -CT (p < 0.01) (Fig.4). Addition of rhBMP-2 to
the biphasic material doubled the mineralized volume compared to the biphasic material alone while the combi-
nation of rhBMP-2 and ZA increased the mineralized volume approximately 4-times compared to the biphasic
Figure 5. Histological representation and histomorphometric analysis of the samples implanted in the
abdominal muscle pouch. Images in top panels provide an overview of the whole sample (12.5X) in the three
groups aer 28-days of implantation. Images in the middle panel emphasizes on the periphery of the material/
bone composite (100X) while images in the bottom panels indicate the innermost tissue/material construct
(100X). A indicates apatite, B shows bone, F represents brous tissue and “BM” shows bone marrow. Data in
(J) shows Histomorphometric quantication of bone area across the two groups with bone formation. Data is
expressed as mean ± SD. n = 5.

www.nature.com/scientificreports/
8
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
material alone and 2-times more than the biphasic material combined with rhBMP-2. Histomorphometry results
also support the μ -CT data with signicantly higher bone formation in biphasic material + rhBMP-2 + ZA group
compared to the biphasic material + rhBMP-2 group. (p < 0.01) (Fig.5). Bone formation was noticed even in the
middle of the scaold (Figs5 and 6) when rhBMP-2 was combined with ZA. is implies that the scaold gets
porous over time, acting synergistically, enhancing osteoconductivity by dynamic structural remodeling of the
material. Without the addition of an anti-resorptive the central part of the scaold is empty and lled with brous
tissue or marrow fat. e in-vivo results also corroborate with the in-vitro rhBMP-2 release, indicating that the
carried protein leads to bone formation in-vivo. It could be argued that a limitation in the current animal study
is the absence of a biphasic material + ZA group. However, from our earlier experiments20, the addition of only
ZA does not enhance bone formation when compared to the combination of BMP + ZA. Moreover, we are not
aware of any reports on ectopic bone formation caused by ZA alone and thus we complied with the 3R’s principle
by reducing a group that is known to have little or no impact.
Figure 6. SEM analysis of implanted samples aer 28-days of in-vivo implantation. Top panel provides a
low magnication (50X) overview of samples in the three groups while the images in the lower panels have been
taken at comparatively higher magnications to emphasize on the appearance and surface structure of the bone/
material composite. e arrows in lower le and middle panel indicate apatite particles while the arrows in the
lower right panel show typical trabecular bone formation on the biphasic material loaded with rhBMP-2 and ZA.
Figure 7. Mechanical testing of implanted samples aer 28-days of in-vivo implantation. Le panel shows
the stiness of the bone/material composite in the three groups while right panel indicates the absorbed energy
across the three groups. *Indicates p < 0.05, ***p < 0.001, #indicates non-signicant. Data is expressed as
mean ± SD. n = 5.

www.nature.com/scientificreports/
9
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
e results from the mechanical analysis aer harvest of the in-vivo samples provide an insight into the
mechanics of the material/bone composite (Fig.7). Due to the absence of bone in the center of the ceramic
samples treated with rhBMP-2 alone (Figs4–6), the stiffness of the material was found to be less than the
rhBMP-2 + ZA group (Fig.7). However, the material alone was stier even though it appeared to have no bone
formation, which maybe due to inherent material properties and the small but dense remaining volume of the
biphasic material. is phenomenon was supported by the calculations of the absorbed energy, which demon-
strated higher energy absorption for the groups where mineralized tissue had been observed.
e material by itself, without the addition of any signaling molecule did not induce bone formation histo-
logically in this muscle model (Fig.5). We observed an increase in the mineralized volume (Fig.4), which may
be due to precipitation of hydroxyapatite on the surface of the material due to its composition and irregular sur-
face10. is is an expected result in a rodent model10,11,55. As our main aim was to study the carrier properties of
the material for rhBMP-2 and ZA by inducing bone at an ectopic location, using the rodent muscle model is well
apted for such studies.
Our results show an easy method with pronounced eect of the co-delivery of ZA and rhBMP-2 and a possible
mechanism for the delivery of these bioactive components to induce bone formation in a non-union or in bone
regeneration (Fig.8).
e carrier encapsulates and releases enough rhBMP-2 to induce bone formation in a non-osseous environ-
ment by interaction with the BMP-receptor containing surrounding muscle cells56. e molecular mechanisms
behind the interaction of BMPs with mesenchymal cells have been explained in literature. BMP binds to the
receptors of inducible cells causing Smad activation. is leads to transcription of RunX-2 and osteogenic dif-
ferentiation of the progenitor cells and thus osteoinduction57. e combination of the biphasic material with
rhBMP-2 provides an osteoinductive tool and we speculate that the molecular mechanisms of BMP mediated
osteoinduction remain the same as described above. However, the material just acts as a carrier for the delivery
of the osteoinductive molecule and provides a template for bone regeneration. us, at this point we do not claim
that the biphasic material itself has any osteoinductive properties. Moreover, the bound ZA also enables higher
new bone retention by inducing osteoclastic apoptosis28, delaying the remodeling of the new-formed bone.
In future studies, we would like to evaluate the osteoinductive potential of the material in large animal models
(non-human primates). Moreover, we are evaluating the eect of released BMP and ZA on various progenitor
cells to conrm that the molecular mechanism of osteoinduction using the biphasic material and rhBMP-2 fol-
lows the conventional molecular pathways that have been described earlier.
Figure 8. Schematic of rhBMP-2 and ZA delivery from the biphasic material and possible bone formation
process. (1) indicates a set disc of the biphasic material with ZA bound to HA while rhBMP-2 is encapsulated
between the two phases. Aer in-vivo implantation, the material releases sulphate, rhBMP-2 and little ZA as
shown in (2). Muscle stem cells interact with rhBMP-2 via BMP receptors56 and a change in their phenotype
occurs leading to their osteogenic dierentiation. Subsequently the bone formation approaches inwards into
the scaold. Due to sulphate resorbing over time, the scaold gets more porous and the bone formation is
substantiated by rhBMP-2 as shown in (3). Aer a signicant amount of bone is formed, RANKL-RANK
(Osteoblast- Osteoclast progenitor) interaction causes osteoclastogenesis as shown in (4)25. However, due to the
presence of ZA, osteoclastic apoptosis occurs28 leading to a preserved bone turnover (5).

www.nature.com/scientificreports/
10
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
us in conclusion, the selected combination of HA particles and calcium sulphate ensures sucient chem-
ical binding of ZA to the HA, while an in situ setting calcium sulphate phase, physically incorporating BMP,
leads to release of the bioactive protein to the inducible cells. e addition of rhBMP-2 and ZA does not alter
the structural properties of the material. e high degree of non-resorbing HA attracting ZA with high anity
is advantageous for local administration of ZA. e results overall suggest that the material can act as a carrier
for the co-delivery of rhBMP-2 and ZA with a synergistic eect and it can potentially be used for regeneration of
large bone defects.
Methods
Study Plan. In-vitro. e release of rhBMP-2 and ZA from the biphasic material in-vitro was analyzed.
Mechanical analysis and SEM imaging was performed to ensure that addition of rhBMP-2 and ZA did not cause
dramatic changes in the surface or mechanical properties of the material.
In-vivo. In the in-vivo study, discs of the biphasic material alone, or containing rhBMP-2 or the combination of
rhBMP-2 and ZA were implanted in an abdominal muscle pouch. Aer harvest, the samples were analyzed using
radiography, μ -CT, histology, SEM and mechanical compression test.
Materials. rhBMP-2 was purchased from Medtronic (Medtronic, Infuse® Bone Gra) and zoledronic acid
(Novartis) was purchased from the local pharmacy. e biphasic material, Cerament™ was supplied by Bone
Support AB, Lund, Sweden. Alpha modied eagles medium (α -MEM) was purchased from ermo scientic,
U.S.A. Heat inactivated fetal bovine serum (FBS) and MTT reagent was purchased from Sigma Aldrich, Germany.
A549 lung cancer cells were kindly provided by Aab Nadeem (Biomedical center, Lund University). rhBMP-2
ELISA kit was purchased from Abcam, Cambridge, U.K. All other reagents were of high purity.
Material preparation. In v itro rhBMP-2 release. e biphasic material was mixed with rhBMP-2 solution (dis-
solved in Iohexol and saline) and casted in the shape of cylinders (5 mm diameter, 1.8 ± 0.2 mm height containing
40 μ l of ceramic paste) in a sterile polypropylene mold. Each disc contained 2.5 μ g of rhBMP-2.
Material preparation for in vitro ZA release and bound ZA experiment. e biphasic material was mixed with
ZA (56.25 μ g ZA in 500 mg of ceramic powder), as per the concentrations we have used in a few clinical cases
(Hettwer et al., unpublished data) and allowed to set in 24-well polystyrene plates.
In vivo muscle pouch model. e animals were divided into three groups for the animal experiments, animals
receiving biphasic material alone, biphasic material in combination with rhBMP-2 and biphasic material with
rhBMP-2 and ZA. e discs were casted in specially designed polypropylene moulds as described earlier. Each
cylindrical disc contained 83,33 mg ceramic powder (Calcium Sulphate-60%, Hydroxyapatite-40% by weight),
13,5 μ l saline and 22,3 μ l of iodine based contrasting agent (Iohexol). e discs in the rhBMP-2 group additionally
contained 10 μ g of rhBMP-2/disc dissolved in saline, while the discs in rhBMP-2 + ZA group contained 10 μ g of
rhBMP-2 and 10 ug of ZA/disc dissolved in saline. Aer all additives were added, the paste was rigorously mixed to
homogenize the contents and casted as cylinders in the mould described above in aseptic conditions. Aer a period
of 15 min, the discs were monitored for hardness to record any delays in setting time and the casted discs were stored
for further use. Discs from all three groups were also used as unimplanted controls for all evaluation methods.
In-vitro rhBMP-2 release. In order to achieve a release kinetics curve, discs of the biphasic material containing
rhBMP-2 (described in 2.1.1) were immersed in saline in a low-protein binding tube at physiological pH and
incubated at 37 °C for 7 days. At each time-point (Day 1, 3, 5 & 7) the supernatants were collected and stored at
− 20 °C until assay. Aer 7-days, the collected supernatants were analyzed using ELISA to determine the concen-
tration of protein in the supernatant.
In-vitro ZA release. e analysis of ZA release from the biphasic material was investigated by an indirect method.
ZA is known to induce apoptosis of lung cancer cell line A549 in a dose dependent manner49. A549 cells were thus
used to quantify the release of ZA from the biphasic material in-vitro. Various doses of ZA (ranging from 0 μ M to
320 μ M) were given to 1 × 104, A549 cells/well seeded on 96-well polystyrene plates containing 300 μ l of culture
medium α -MEM/well (containing 10% FBS by volume). e cells were incubated with ZA for 48 h and the cell
viability was assessed using the MTT assay and microscopic visualization. Pre-set discs (described in 2.1.2) were
incubated with 0.5 ml of α -MEM without FBS for dierent time points (1, 3, 5 & 7 days) and supernatants were
collected at each time point and stored at 4 °C until the assay at the 7th day. e collected supernatants were nally
mixed with 10% (v/v) FBS and added to A549 cells (104/well). e cells were incubated with the supernatants for
48 h and the cell viability was calculated using the MTT assay and cell morphology was analyzed using microscopy.
Recorded absorbance was then used to calculate the released drug fraction from the standard curve plotted above.
Eect of bound ZA to the biphasic material. ZA was mixed (same concentration as used in 2.3) with the biphasic
material and seeded with 104, A549 cells on the material. Material without ZA served as negative control while
direct addition of free ZA to A549 cells served as positive controls. e cells were incubated for 24 and 72 h and
cell viability was calculated using the MTT assay. e experiment was performed to assess the eect of bound ZA
on the proliferation of A549 cells.

www.nature.com/scientificreports/
11
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
Scanning electron microscopy (SEM). SEM was performed to analyze and compare the structure and surface
architecture of the biphasic material with or without the addition rhBMP-2 and ZA. e samples were analyzed
either aer casting or by incubating in saline for 28-days to allow dissolution of the calcium sulphate phase and
release of other additives. Later, the samples were dried at 37 °C in a vacuum desiccator for 1 day followed by gold
coating. e samples were analyzed using a FEI Quanta, SEM analyzer at an operating voltage of 12.5-15 kV.
Mechanical compression test. Compression tests were performed on the samples prepared by following the
method described in detail in section 2.1.3. e samples were prepared 24 h before the test. e samples were ana-
lyzed on an Instron mechanical analyzer (Instron 8511 load frame, MTS FlexTest 40 Controller, MTS TestSuite
Multipurpose Elite Soware). A pre load of 1 N was used, followed by loading with 0.1 mm/s until a maximum
load of 100 N was reached. Load and displacement data were used to calculate the stiness of the materials and
absorbed energy in all three groups.
Surgical Procedure for ectopic bone formation. All discs for in-vivo implantation were casted in molds as
described above in 4.1.3. A total of 12 Sprague-Dawley rats of 6-weeks age divided into two groups were used. e
animals were anaesthetized using a combination of pentobarbital sodium (15 mg/ml) and diazepam (2.5 mg/ml)
administered intraperitoneal (I.P). e animals were given intramuscular antibiotic prophylaxis (Streptocilin).
Each rat was placed in the dorsal supine position and an approximately 2 cm long mid-line skin incision was
made in the abdomen. A muscle pouch was created in the external oblique muscle and the material was placed
between two layers of the muscle. e muscle pouch was closed using a non-resorbable 5-0 Ethilon to identify the
implants post sacrice. e skin was sutured using a 5-0 Vicryl bioresorbable suture. All twelve animals received
two discs of the biphasic material alone on the le side of the abdominal midline at a minimum distance of 1.5 cm
apart. Six of the rats also received two discs of biphasic material + rhBMP-2 on the right side of the abdominal
midline, separated from each other by at least 1.5 cm. e remaining 6 animals received discs containing biphasic
material + rhBMP-2 + ZA on their right side of the abdominal midline. Animals had access to food and water ad
libitum and aer a period of 4-weeks the animals were sacriced using an I.P overdose of pentobarbital sodium.
Assessment of ectopic bone formation. In order to evaluate the total bone formation in the implanted scaolds,
various techniques were used. All assessments were done aer harvesting the samples.
Physical assessment of scaolds at harvest. At the time of sacrice, all samples were assessed manually by pal-
pating the implant site.
Radiography. Aer sacrice, all samples were immediately placed in sterile gauze drenched in saline and placed
in plastic tubes. All scaolds were imaged in a similar orientation using a GE Healthcare discovery X-ray machine
(CT, USA).
Micro computed tomography (micro-CT). All samples were scanned using an isotropic voxel size of 21 μ m
(nanoScan, Mediso Medical Imaging Systems, Budapest, Hungary) at an operating voltage of 65 kV, 123 μ A cur-
rent using 360 projections and a RAMLAK lter. Images were reconstructed post hoc to a voxel size of 10 μ
m (Nucline soware, VivoQuant 1.22, inviCRO, Boston, MA, USA). e bone mineral density was calibrated
by using two hydroxyapatite phantoms of known densities (0.25 g/cm3 and 0.75 g/cm3) separated by water and
a density of 0.46 g/cm3 and above was considered mineralized tissue. e entire sample volume was chosen
as the region of interest, and the total mineralized volume in the samples was quantied (including both the
hydroxyapatite from the remaining biphasic material and the newly formed bone).
Histology and histomorphometry. Five samples from each group were xed in 4% (v/v) formalin solution for 2
days. e samples were cut in two halves using a surgical scalpel and one half was used for histology while the other
half was used for SEM. Histology specimens were decalcied in 10% (w/v) ethylenediaminetetraacetic acid (EDTA)
for 10 days at room temperature by constant shaking (Grant-Bio, multi rotator) and the EDTA solution was changed
every second day. Post decalcication, samples were dehydrated in increasing ethanol gradient and xylene treatment
for 1 h each and embedded in paran. e samples were sectioned using a microtome (HM355S, ermo Fischer
Scientic, MA, USA) to a thickness of 5 μ m and stained using hematoxylin and eosin. Histomorphometry was per-
formed using HALO v1.92 (Leica Biosystems, histomorphometry soware) by selecting the areas of bone formation
(bone matrix and osteocytes). e area of bone formation in both groups was then plotted for comparison.
Scanning electron microscopy (SEM). Five undecalcied samples were cut into halves and processed for SEM
analysis (Carl Zeiss GmbH, EVO18 Special Edition, Germany). e samples were dehydrated using an increasing
ethanol gradient, and the excess muscle tissue around the sample was carefully scraped o under a stereomicro-
scope. e samples were placed in a vacuum desiccator until imaging and just before imaging the samples were
gold coated using a cressignton sputter coater. e samples were analyzed at an operating voltage of 15 kV.
Mechanical compression. e remaining 5 samples/group were subjected to mechanical compression using an
Instron mechanical analyzer (Instron 8511 load frame, MTS FlexTest 40 Controller, MTS TestSuite Multipurpose
Elite Soware) connected to a 250 N load cell. Each sample was compressed between two cylindrical metal rods
(1 cm diameter). A pre-load of 1 N was applied and the loading rate was 0.1 mm/s until failure. e load and

www.nature.com/scientificreports/
12
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
displacement data were used to calculate the stiness of the implanted materials and the energy absorbed by the
implanted scaolds.
Ethical Permission. All animal procedures were approved and performed in accordance with the directives of
Jordbruksverket (animal ethics and licensing committee), the Swedish regulatory authority for the use of animals
for experimental purposes. (Ethical approval number M124-14).
Statistical Analysis. Statistical analysis was performed on Prism 6 for Mac OS X (Version 6.0 d, GraphPad
Soware, Inc., CA, USA). Data is represented as mean ± SD. All analysis were performed using unpaired t-test
and p < 0.05 was considered to be statistically signicant. All results are expressed in triplicates unless otherwise
specied.
References
1. Amini, A. ., Laurencin, C. T. & Nuavarapu, S. P. B one tissue engineering: recent advances and challenges. Crit. ev. Biomed. Eng.
40, 363–408 (2012).
2. Lidgren, L., Gomez-Barrena, E. N., Duda, G., Puhl, W. & Carr, A. European musculoseletal health and mobility in Horizon 2020:
Setting priorities for musculoseletal research and innovation. Bone Joint es. 3, 48–50 (2014).
3. Arrington, E. D., Smith, W. J., Chambers, H. G., Bucnell, A. L. & Davino, N. A. Complications of iliac crest bone gra harvesting.
Clin. Orthop. el. es. 329, 300–309 (1996).
4. Giannoudis, P. V., Dinopoulos, H. & Tsiridis, E. Bone substitutes: an update. Injur y 36, S20–S27 (2005).
5. Boyan, B. D., anly, D. M. & Schwartz, Z. Use of Growth Factors to Modify Osteoinductivity of Demineralized Bone Allogras:
Lessons for Tissue Engineering of Bone. Dent. Clin. 50, 217–228 (2006).
6. Schwartz, Z. et al. Ability of commercial demineralized freeze-dried bone allogra to induce new bone formation. J. Periodontol. 67,
918–926 (1996).
7. Schwartz, Z. et al. Ability of commercial demineralized freeze-dried bone allogra to induce new bone formation is dependent on
donor age but not gender. J. Periodontol. 69, 470–478 (1998).
8. Wei, L., Miron, . J., Shi, B. & Zhang, Y. Osteoinductive and osteopromotive variability among dierent demineralized bone
allogras. Clin. Implant Dent. el. es. 17, 533–542 (2013).
9. Tomford, W. W. Transmission of disease through transplantation of musculoseletal allogras. J. Bone Joint Surg. 77, 1742–1754
(1995).
10. Habibovic, P. & de Groot, . Osteoinductive biomaterials—properties and relevance in bone repair. J. Tissue Eng. egen. Med. 1,
25–32 (2007).
11. Oryan, A., Alidadi, S., Moshiri, A. & Maulli, N. Bone regenerative medicine: classic options, novel strategies, and future directions.
J. Orthop. Surg. es. 9, 1749–1799 (2014).
12. Blacwood, . A., Boc, N., Dargaville, T. . & Ann Woodru, M. Scaolds for growth factor delivery as applied to bone tissue
engineering. Int. J. Polym. Sci. 2012, 25 Pages (2012).
13. Place, E. S., Evans, N. D. & Stevens, M. M. Complexity in biomaterials for tissue engineering. Nat. Mater. 8, 457–470 (2009).
14. Hutmacher, D. W. S caolds in tissue engineering bone and cartilage. Biomaterials 21, 2529–2543 (2000).
15. Mandal, B. B., Das, S., Choudhury, . & undu, S. C. Implication of sil lm GD availability and surface roughness on cytoseletal
organization and proliferation of primary rat bone marrow cells. Tissue Eng. A 16, 2391–2403 (2010).
16. Bosemar, P., Perdiouri, C., Pelonen, M., Isasson, H. & Tägil, M. e Masquelet Induced Membrane Technique with BMP and a
Synthetic Scaold Can Heal a at Femoral Critical Size Defect. J. Orthop. es. 33, 488–495 (2015).
17. Gao, T. J. et al. Enhanced healing of segmental tibial defects in sheep by a composite bone substitute composed of tricalcium
phosphate cylinder, bone morphogenetic protein, and type IV collagen. J. Biomed. Mater. es. 32, 505–512 (1996).
18. oempel, J. A., Patt, B. S., O’Grady, ., Wozney, J. & Toriumi, D. M. e eect of recombinant human bone morphogenetic protein-2
on the integration of porous hydroxyapatite implants with bone. J. Biomed. Mater. es. 41, 359–363 (1998).
19. Caballé-Serrano, J., Bosshardt, D. D., Buser, D. & Gruber, . Proteomic Analysis of Porcine Bone-Conditioned Medium. Int. J. Oral
Maxillofac. Implants 29, 1208–1215 (2013).
20. Mathavan, N., Bosemar, P., Isasson, H. & Tägil, M. Investigating the synergistic ecacy of BMP-7 and zoledronate on bone
allogras using an open rat osteotomy model. Bone 56, 440–448 (2013).
21. Belfrage, O., Flivi, G., Sundberg, M., esteris, U. & Tägil, M. Local treatment of cancellous bone gras with BMP-7 and zoledronate
increases both the bone formation rate and bone density: a bone chamber study in rats. Acta Orthop. 82, 228–233 (2011).
22. Johnsson, ., Strömqvist, B. & Aspenberg, P. andomized adiostereometric Study Comparing Osteogenic Protein-1 (BMP-7) and
Autogra Bone in Human Noninstrumented Posterolateral Lumbar Fusion: 2002 Volvo Award in Clinical Studies. Spine 27,
2654–2661 (2002).
23. Friedlaender, G. E. et al. Osteogenic protein-1 (bone morphogenetic protein-7) in the treatment of tibial nonunions. J. Bone Joint
Surg. 83, S151–S158 (2001).
24. aneo, H. et al. Direct stimulation of osteoclastic bone resorption by bone morphogenetic protein (BMP)-2 and expression of BMP
receptors in mature osteoclasts. Bone 27, 479–486 (2000).
25. Yasuda, H. et al. A novel molecular mechanism modulating osteoclast dierentiation and function. Bone 25, 109–113 (1999).
26. Gravallese, E. M., Galson, D. L., Goldring, S. . & Auron, P. E. e role of TNF-receptor family members and other TAF-dependent
receptors in bone resorption. Arthritis es. 3, 6–12 (2001).
27. Giannoudis, P., anaaris, N. & Einhorn, T. Interaction of bone morphogenetic proteins with cells of the osteoclast lineage: review
of the existing evidence. Osteoporos. Int. 18, 1565–1581 (2007).
28. Hughes, D. E. et al. Bisphosphonates promote apoptosis in murine osteoclasts in vitro and in vivo. J. Bone Miner. es. 10, 1478–1487
(1995).
29. ogers, M. J., Crocett, J. C., Coxon, F. P. & Mönönen, J. Biochemical and molecular mechanisms of action of bisphosphonates.
Bone 49, 34–41 (2011).
30. Nancollas, G. H. et al. Novel insights into actions of bisphosphonates on bone: Dierences in interactions with hydroxyapatite. Bone
38, 617–627 (2006).
31. Odvina, C. V. et al. Severely suppressed bone turnover: a potential complication of alendronate therapy. J. Clin. Endocrinol. Metabol.
90, 1294–1301 (2005).
32. uggiero, S. L., Mehrotra, B., osenberg, T. J. & Engro, S. L. Osteonecrosis of the jaws associated with the use of bisphosphonates:
a review of 63 cases. J. Oral Maxillofac. Surg. 62, 527–534 (2004).
33. Chang, J.-T., Green, J. & Beitz, J. enal Failure with the Use of Zoledronic Acid. New Eng. J. Med. 349, 1676–1679 (2003).
34. Diel, I. J., Bergner, . & Grotz, . Adverse eects of bisphosphonates: current issues. J. Support. Oncol. 5, 475–482 (2007).
35. Hollinger, J. O. & Leong, . Poly (α -hydroxy acids): carriers for bone morphogenetic proteins. Biomaterials 17, 187–194 (1996).
36. Minamide, A., awaami, M., Hashizume, H., Saata, . & Tamai, T. Evaluation of carriers of bone morphogenetic protein for
spinal fusion. Spine 26, 933–939 (2001).

www.nature.com/scientificreports/
13
Scientific RepoRts | 6:26033 | DOI: 10.1038/srep26033
37. Winn, S. ., Uludag, H. & Hollinger, J. O. Carrier Systems for Bone Morphogenetic Proteins. Clin. Orthop. el. es. 367, S95–S106
(1999).
38. isiel, M. et al. Critical assessment of rhBMP-2 mediated bone induction: An in vitro and in vivo evaluation. J. Control. el. 162,
646–653 (2012).
39. Friess, W. Collagen – biomaterial for drug delivery1. Eur. J. Pharm. Biopharm. 45, 113–136 (1998).
40. Stenfelt, S. et al. Pre-incubation of chemically crosslined hyaluronan-based hydrogels, loaded with BMP-2 and hydroxyapatite, and
its eect on ectopic bone formation. J. Mater. Sci: Mater. Med. 25, 1013–1023 (2014).
41. Quinlan, E., ompson, E. M., Matsio, A., O’Brien, F. J. & López-Noriega, A. Long-term controlled delivery of rhBMP-2 from
collagen–hydroxyapatite scaolds for superior bone tissue regeneration. J. Control. el. 207, 112–119 (2015).
42. Wang, J.-S. et al. In vitro and in v ivo biological responses to a novel radiopacifying agent for bone cement. J. oyal Soc. Interface 2,
71–78 (2005).
43. Nilsson, M., Wielane, L., Wang, J. S., Tanner, . E. & Lidgren, L. Factors inuencing the compressive strength of an injectable
calcium sulfate–hydroxyapatite cement. J. Mater. Sci: Mater. Med. 14, 399–404 (2003).
44. arlsson, V., Nilsson, M. & Lidén, E. Inventors; Bone Support AB, assignee. Preparation of bone cement compositions. European
Patent EP2353619 A1. 2011 Aug 10.
45. Urist, M. . Bone: Formation by Autoinduction. Science 150, 893–899 (1965).
46. Geiger, M., Li, . H. & Friess, W. C ollagen sponges for bone regeneration with rhBMP-2. Adv. Drug Del. ev. 55, 1613–1629 (2003).
47. Yu, N. et al. Spatial control of bone formation using a porous polymer scaold co-delivering anabolic rhBMP-2 and anti-resorptive
agents. Eur. Cell. Mater. 27, 98–109 (2014).
48. Belfrage, O., Isasson, H. & Tägil, M. Local treatment of a bone gra by soaing in zoledronic acid inhibits bone resorption and
bone formation. A bone chamber study in rats. BMC Musculoselet. Disord. 13, 240 (2012).
49. Oztur, O. H. et al. Cisplatin cytotoxicity is enhanced with zoledronic acid in A549 lung cancer cell line: Preliminary results of an in
vitro study. Cell Biol. Int. 31, 1069–1071 (2007).
50. Murphy, C. M. et al. A collagen–hydroxyapatite scaold allows for binding and co-delivery of recombinant bone morphogenetic
proteins and bisphosphonates. Acta Biomater. 10, 2250–2258 (2014).
51. Jindong, Z. et al. Evaluation of a novel osteoporotic drug delivery system in vitro: alendronate-loaded calcium phosphate cement.
Orthopedics 33, 561 (2010).
52. Schindeler, A. et al. Modeling bone morphogenetic protein and bisphosphonate combination therapy in wildtype and Nf1
haploinsucient mice. J. Orthop. es. 26, 65–74 (2008).
53. Carragee, E. J. et al. Cancer ris aer use of recombinant bone morphogenetic protein-2 for spinal arthrodesis. J. Bone Joint Surg. 95,
1537–1545 (2013).
54. Little, D. G., McDonald, M., Bransford, ., Godfrey, C. B. & Amanat, N. Manipulation of the Anabolic and Catabolic esponses
With OP1 and Zoledronic Acid in a at Critical Defect Model. J.Bone Miner. es. 20, 2044–2052 (2005).
55. Ohgushi, H., Goldberg, V. M. & Caplan, A. I. Heterotopic osteogenesis in porous ceramics induced by marrow cells. J. Orthop. es.
7, 568–578 (1989).
56. Naamura, Y. et al. Temporal and Spatial Expression Proles of BMP eceptors and Noggin During BMP2Induced Ectopic Bone
Formation. J. Bone Miner. es. 18, 1854–1862 (2003).
57. Chen, G., Deng, C. & Li, Y.-P. TGF-beta and BMP signaling in osteoblast dierentiation and bone formation. Int. J. Biol. Sci. 8,
272–288 (2012).
Acknowledgements
Authors would like to acknowledge the Foundation for disabled people in Skåne and the Medical Faculty, Lund
University for funding the project. AK acknowledges DBT-TATA innovation fellowship. We would also like to
thank Mea Pelkonen (Department of Orthopedics, Lund University) for her advice with histological procedures,
Christina Perdikouri (Department of Biomedical Engineering, Lund University) and Gustav Grafström (Lund
Bio Imaging Center (LBIC), Lund University) for their assistance with micro-CT analysis.
Author Contributions
D.B.R., M.T. and L.L. designed the study. M.T. and D.B.R. performed the in-vivo experiments. D.B.R. performed
the in-vitro experiments. H.I. and D.B.R. performed mechanical testing and micro-CT and H.I. analyzed the data
from both experiments. A.K. performed the SEM analysis and provided assistance with in-vitro experimentation.
W.H. assisted with addition of ZA to the material and ZA dosage. D.B.R. wrote the rst dra of this paper and all
authors revised the manuscript. e nal manuscript has been read and approved by all authors for publication.
Additional Information
Competing nancial interests: Prof. Lars Lidgren is a board member of Bonesupport AB, Lund, Sweden and
Orthocell, Australia. None of the other authors declare any conict of interest.
How to cite this article: Raina, D. B. et al. A Biphasic Calcium Sulphate/Hydroxyapatite Carrier Containing
Bone Morphogenic Protein-2 and Zoledronic Acid Generates Bone. Sci. Rep. 6, 26033; doi: 10.1038/srep26033
(2016).
is work is licensed under a Creative Commons Attribution 4.0 International License. e images
or other third party material in this article are included in the article’s Creative Commons license,
unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license,
users will need to obtain permission from the license holder to reproduce the material. To view a copy of this
license, visit http://creativecommons.org/licenses/by/4.0/