Strontium Incorporated Coralline Hydroxyapatite for Engineering Bone

1 Department of Orthopaedics & Traumatology, e University of Hong Kong, Pokfulam, Hong Kong Department of Orthopaedics, Shanghai Institute of Orthopaedics & Traumatology, Shanghai Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, 227 South Chongqing Road, Shanghai 20025, China 3 Center for Human Tissues and Organs Degeneration, Shenzhen Institute of Advanced Technology, Chinese Academy of Science, 1068 Xueyuan Avenue, Shenzhen University Town, Shenzhen 518055, China Department of Spine Surgery, Shenzhen People’s Hospital, Jinan University Second College of Medicine, 1017 Dong Min Bei Lu, Shenzhen 518020, China


Introduction
Reconstruction of massive bone defects caused by severe traumas and tumors remains a great challenge in orthopaedics.e preferred treatment is autologous bone gra, but the supply is usually limited due to the mobility and pain caused to the patient [1,2].Immunological response of our body and possibility of contamination have limited the use of allogra.Emergence of tissue engineering has been growing promisingly in orthopaedics as an alternative approach for bone regeneration.Before supplementing scaffolds with marrow stromal cells and growth factors to further enhance bone regeneration, the choices of scaffold with osteoconductivity or even osteoinductivity are important to start with.
Coral, which usually refers to the exoskeleton of natural coral, has been a material of interest in orthopaedic biomaterial due to its chemical composition and structural and mechanical properties.Other materials have to be further modi�ed to meet the essential structural, physical, or mechanical requirements of ideal bone substitutes; however, natural coral, primarily reef-building coral, has certain degrees of similarity to human bone without the need of modi�cation [3].Major component of coral is calcium carbonate (97-99%), in aragonite form, with 1-1.5% of organic substances and little trace elements (0.5-1% of magnesium, sodium, potassium, strontium, �uorine, phosphorous, etc.) [3][4][5].
Since Roy and Linnehan �rst published in 1974, hydrothermal reaction has been widely applied to convert natural coral from calcium carbonate to calcium hydroxyapatite, also known as coralline hydroxyapatite (CHA), by keeping the porous microstructure of coral [6].Both natural coral and CHA have their own advantages and drawbacks; while the decision is predominantly based on the different weighting on the mechanical property and biodegradability and compositional similarity to bone [3].More importantly, natural coral and CHA can supply plenty of calcium and possess micropores with interconnectivity, which are helpful in bone regeneration and essential for bone ingrowth in terms of tissue engineering scaffold [4,[7][8][9].
Strontium has been well published for its antiresorptive and anabolic effects on bone, so it is gaining attention on its applications in orthopaedic biomaterials [10].It is believed that the addition of Sr can modify or further enhance the bioactivity of materials, particularly for the materials to be applied in osteoporotic and osteopenic patients.In this study, Sr was incorporated to CHA converted from Goniopora as a new scaffolding material for bone tissue engineering; Goniopora has been one of the few genera of corals, in which it has been widely studied in the literature for biomedical applications.e mechanical properties, chemical composition, and morphological changes were characterized; the biological properties were investigated in vitro and in vivo with an OVX rat model.

Hydrothermal Conversion and Incorporation of Strontium.
Coralline hydroxyapatite (CHA) conversion was based on the calcium available in the coral reacted with diammonium hydrogen phosphate converting it into hydroxyapatite.Goniopora was obtained from the South China Sea near Sanya, China.Once received, it was immersed in 5% sodium hypochlorite solution and rinsed with distilled water several times until all the proteins were removed as widely reported [3].e molar ratio of CaCO 3 to (NH 4 ) 2 HPO 4 is 10 : 6 in order to prepare 1 mole of Ca 10 (PO 4 ) 6 (OH) 2 [6].Coral specimens were �lled with 2 M diammonium hydrogen phosphate (pH 8.0) solution in excess (5 times of the required amount) and put in a hydrothermal bomb [11].e bomb was set in a 180 ∘ C oven for 15 h and replaced with fresh (NH 4 ) 2 HPO 4 solution to being further treated for 15 h as a 2day treatment.Strontium was hydrothermally incorporated into CHA (i.e., Sr-CHA) by adding strontium nitrate (0.1 M or 0.5 M) solutions for 15 h at 180 ∘ C; the molar ratios of Sr to Ca for 0.1 M Sr(NO 3 ) 2 and 0.5 M Sr(NO 3 ) 2 solutions were Sr : Ca = 4 : 10 and 12 : 10, respectively.All chemicals used were purchased from Sigma-Aldrich (St. Louis, MO, USA) and �ako Pure Chemical Industries (Japan) unless speci�ed.

Structural Evaluation and Compressive
Properties. e microstructure and compressive properties of specimens were studied by micro-CT and compression test.Raw Goniopora was scanned using SkyScan 1076 (Kontich, Belgium) micro-CT scanner with the pixel size set to 11.53 m.reedimensional model and porosity measurement were performed using SkyScan CTAn soware.Compression test was performed on the specimens cut in the size of 6 × 6 × 8 mm 3 .Testing was undergone on the 6 × 6 mm surface along the height of 8 mm with a recording material testing machine (MTS 858 Bionix machine, MTS System, Minneapolis, MN, USA) with the compression rate of 2 mm/min.Five specimens each of the untreated coral, CHA, and Sr-CHA were taken for the compression test.

2.3.
Characterization.Degree of conversion, percentage of Sr substitution, and surface morphology of CHA and Sr-CHA scaffolds were studied with SEM, EDX, and XRD.Specimens for SEM were sputter coated with a 100 Å layer of gold palladium and imaged using a scanning electron microscope (Hitachi S-4800, Hitachi, Japan).For investigating the depth of CHA conversion, CHA specimen was embedded in epoxy (Epo�xTM, Electron Microscopy Sciences, PA, USA) and cut into 300 m cross-sectional slides.Line EDX was done on these cross-sections of embedded CHA, and general EDX was done on untreated coral, CHA and Sr-CHA to identify different elements on the surfaces.e compositions of all specimens were further characterized by XRD (Model D/max 2550V, Rigaku, Japan) using CuKa radiation (  1406 Å) in step-scan mode (2  002 ∘ per step).

2.4.
In Vitro Study.MC3T3-E1 (ATCC, VA, USA), a mouse osteoblastic cell line, was used to study the cell proliferation on these scaffolds.e cells were cultured in Minimum Essential Alpha (-MEM) Medium (without ascorbic acid) supplemented with 10% fetal bovine serum (Biosera, UK), Fungizone (1.0 g/mL), and 1% penicillin/streptomycin in a 37 ∘ C incubator with 95% CO 2 .Goniopora was cut into  ×  ×  mm 3 scaffold and converted to CHA and Sr-CHA.Cells were seeded by putting the scaffolds in a polypropylene container or centrifuge tube with 1 mL (2 × 10  cells/mL) cell suspension for each scaffold.e container was set on a shaker at 300 rpm for 2 h at room temperature.Scaffolds with cell suspension were then transferred to a 24 multiwell plate, and the plate was put in the incubator with the medium changed every other day.Cell proliferation on the scaffolds was studied using the common cell proliferation MTT (3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay by measuring the metabolic activity of the cells.Viable cells of CHA and Sr-CHA were studied by MTT assay on days 3 and 5.

Surgical
Procedure.An ovariectomized (OVX) model using Sprague-Dawley rat was adopted to preliminarily investigate the in vivo properties of the converted corals.Female Sprague-Dawley rats aged 3 months and weighed about 250 g were included.All rats were operated under general anesthesia with 10% ketamine (67 mg/kg) and 2% xylazine (6 mg/kg) (Alfasan, Holland).Bilateral ovariectomy was performed on these rats.ree months aer ovariectomy, each of these OVX rats was randomly arranged to receive untreated coral, CHA, or Sr-CHA implants in a disc shape of 3 mm diameter and 2 mm height.A 3 mm diameter bone defect was created on the tibial sha at about 5 mm below tibia plateau and randomly inserted with implants on both legs.e wound was closed and wrapped with dressing aer implantation.

Micro-CT Evaluation.
Under general anesthesia, these OVX rats were taken for micro-CT scanning on week 0, 1, 2, and 4 aer the surgery to evaluate the repairing process of the bone defects and degradation of implants.e images were taken through 180 ∘ rotation at a rotation step of 0.60 ∘ with the source voltage 88 kV, source current 100 A, and exposure time of 560 ms.e scaffold volume was processed by SkyScan CTAn and CTvox sowares.

Statistical Analysis.
e results presented in this study are expressed as mean ± standard deviation (   unless speci�ed).Analysis of variance (ANOVA) was applied to calculate the signi�cance level of all data.Statistically signi�cant differences were considered as   ; Tukey's post-hoc test was used to study the correlation if statistical signi�cance was obtained.

Macrostructure, 3D Image, and Porosity Measurement.
e macrostructure of Goniopora was imaged with an optical microscope (Eclipse 80i, Nikon, Japan), SEM, and micro-CT scanning (as shown in Figure 1).Pore size and thickness of the skeleton can be clearly observed from the optical microscopy and SEM images.e pores are irregular in shape with the majority of the pores having pore sizes between 40-300 m.Other than the pore size, the thickness of the skeleton is also an important factor determining the porosity; the skeleton thicknesses of Goniopora could be estimated to be around 20-100 m.e micro-CT scanning is used to visualize the 3D macrostructure and measure the porosity of Goniopora; the porosity is found to be 683 ± 123%. .Moreover, the line EDX result also con�rmed the complete conversion of CHA from the surface to the core, which was about 1.5 mm from the surface as shown here, and full thickness of the skeleton with the detection of phosphorus coincided with the calcium detected as seen in Figure 2(d).

SEM Scanning, EDX and XRD
Surface morphology of the corals aer CHA and Sr-CHA treatments was observed with SEM.Goniopora before any treatment was imaged at various magni�cation to understand the microstructure of coral.e skeleton of coral has been reported to be formed by nano�bers and nanograins �19].As seen in Figure 3(a), the coral was formed by nanograins of diameters <100 nm and nano�bers of diameters 100-500 nm with a few mm in length.Aer CHA conversion, the surface was covered with �ake-like structures (as shown in Figure 3(b)), unlike the surface of raw coral covered with bundles of �bers.A similar structure can be seen in Figures 3(c) and 3(d) aer Sr incorporation at different Sr(NO 3 ) 2 concentrations, but the �akes were smaller in si�e and more spiky in shape for 500 mM Sr(NO 3 ) 2 than 100 mM Sr(NO 3 ) 2 .In addition, there were some hexagon rods on the specimens of Sr-CHA, while more and longer rods were found in the 500 mM Sr(NO 3 ) 2 group; these rods were mainly composed of Sr as con�rmed by point EDX as Figure 4. Degree of CHA conversion and percentage of Sr incorporation were carried out by EDX analysis.e elements detected are shown in Figure 5, while the ratio of Ca + Sr/P and percentage of Sr incorporation calculated using EDX data are shown in Table 1.Before hydrothermal conversion, it can be seen that P and Sr were barely detected in the untreated coral.However, the peak of P increases signi�cantly aer CHA conversion.e signal of Sr peak becomes distinguishable aer Sr incorporation treatment.e Ca + Sr/P calculated on all 3 groups of CHA and Sr-CHA are 1.55 and 1.60, which are lower than the stoichiometric HA of 1.67.Regarding the detection of Sr, 0.8-1.28% of Sr was present in untreated coral and CHA before Sr incorporation; while about 5.94% Sr was detected with less concentrated Sr solution (100 mM Sr(NO 3 ) 2 ), and 13.78% Sr was detected with more concentrated Sr solution (500 mM Sr(NO 3 ) 2 ).
e XRD of untreated coral, CHA, and Sr-CHA conversions are shown in Figure 6.e differences between these two treatments on coral can be clearly identi�ed with their

Cell Culture and Proliferation.
In vitro properties of CHA and Sr-CHA (100 mM and 500 mM Sr(NO 3 ) 2 ) were studied by MTT assay on days 3 and day 5 as shown in Figure 8.On day 3, an estimate of 100,000-140,000 cells was calculated with the greatest number found on CHA; statistical differences were found in CHA comparing to 100 mM and 500 mM Sr(NO 3 ) 2 Sr-CHA.e number of cells increased  quite signi�cantly on day 5 in the range of 500,000 to 800,000 cells; CHA was calculated with the smallest number of cells and found to be statistically different from 100 mM Sr(NO 3 ) 2 Sr-CHA.e greatest number of cells is observed with 100 mM Sr(NO 3 ) 2 Sr-CHA specimens.

Micro-CT Evaluation
. e 3D structures of the untreated coral, CHA, and Sr-CHA (  ) were taken for micro-CT scanning at 0, 1, 2, and 4 weeks aer implantation.e micro-CT images of each scaffold taken at the same leg at week 2 are shown in Figure 9, in which the scaffolds can be clearly distinguished from the surrounding cortical bone.All 3 types of scaffolds were undoubtedly degraded in week 4 by a decrease in sizes of the scaffolds and an increase of porosities comparing to week 0. e relative scaffold volumes calculated by the micro-CT results are presented in Figure 10 to give semiquantitative results of the degradation of scaffold or bone growth in vivo.ere are only about 80% of the volume remained for untreated coral and CHA specimens in the �rst week aer surgery; on the contrary, there are >25% increase in volume, possibility of apatite formation, found in Sr-CHA scaffold.At week 4, more than 50% of the untreated coral and CHA scaffolds degraded, while about 90% of Sr-CHA scaffold remained.

Discussion
In this study, Sr was successfully incorporated into coral by hydrothermal conversion.e highly porous structure with interconnectivity of coral has made this naturally occurring material a very attractive material in the area of bone tissue engineering.Over the years, many studies have been done on various genera of corals both in the forms of raw coral or CHA.e interest of Sr in the area of orthopaedic biomaterial �ourishes gradually in the last few decades.As a calcium-based material, raw coral, or CHA could be readily incorporated with Sr by simply replacing calcium to a certain extent; the incorporation of Sr should potentially enhance the osteoconductivity or even osteoinductivity of coral as a scaffold.
Like in any orthopaedic biomaterials, interconnected porosity is particularly important in the area of bone tissue engineering because the scaffold is meant to be fully in�ltrated with cells as the natural tissue.In addition, porosity is known to be critical in determining the degradation and osteoconductivity of a material [3,20,21].Some of the typical stony corals used as biomaterials are Acropora, Goniopora, and Porties because of the open porosity of their structures.From the porosity measurement measured by micro-CT scanning, Goniopora is found to have a porosity of around 68%. e porosity measurement of Goniopora (∼68%) is similar to the results in the literature, in which it is similar to the porosity of cancellous bone [3,8,16,22,23].Moreover, the pore sizes and skeleton thicknesses are also critical factors to facilitate the ingrowth of bone or �brovascular tissue and affect the mechanical properties [3,4].e pore sizes of Goniopora are reported in the range of 40-300 m, which favors the ingrowth of osteoblasts into the scaffold [4,9].
Hydrothermal conversion of calcium carbonate into CHA with diammonium hydrogen phosphate solution has been frequently performed; however, the degree of conversion was seldom studied since Roy and Linnehan �rst reported the method in 1974.e standard parameters of hydrothermal treatment were suggested and commonly adopted to be around 180-250 ∘ C for 24-48 h [6].In a previous study of hydrothermally converting conch shell into HA, 2 days and 5 days were proposed in order to convert 120 and 200 m of calcium carbonate into HA [24].e thickness of the Goniopora can be seen to be <200 m from the SEM images, but unlike conch shell, coral is a highly porous structure so that a 2-day and 15 h/day hydrothermal conversion protocol at 180 ∘ C was chosen.Line EDX characterization on the cross-sections of embedded CHA con�rmed the complete conversion of carbonate by phosphate of this modi�ed protocol.
Aer HA conversion, there is transformation on the surface morphology compared to untreated coral as imaged by SEM.Over the years, the microstructures of coral have been suggested to be composed of two main structures: center of calci�cation (COC) (or early mineralization zone) and �brous zone; structures of nanograins are observed in COC, while nano�bers formed by growth layers of nanograins are found in the surrounding �brous zone [19,[25][26][27].For the untreated Goniopora, nanogranular and nano�brous structures are clearly observed in the SEM images.Contrastingly, �ake-like structures seen in the SEM images of the HA converted corals are the typical surface morphology of CHA [28].Furthermore, Sr was successfully incorporated into CHA with the presence of hexagon rods aer Sr(NO 3 ) 2 treatment, in which the rods were con�rmed to be primarily Sr by EDX.However, the percentage of Sr incorporated into CHA was much less than the amount added in the reaction.It indicates the low Sr incorporation efficiency by this current method and the possibility of a limit of Ca replacement in a readily formed microstructure like coral.
A lot of studies applied the hydrothermal reaction to form CHA, but rarely did these studies characterize the Ca to P ratio aer conversion.In this study, Ca/P on the specimens aer hydrothermal conversion is found to be 1.55 to 1.60.ere are two possible explanations for CHA to be in a Ca/P value lower than the stoichiometric value 1.67; calcium-de�cient hydroxyapatite (Ca/P � 1.5-1.67)was formed instead of HA or the �nal product was a mixture of HA and -TCP.Calcium-de�cient HA can be regarded as HA with certain calcium ions replaced by other ions [29].Although other trace elements are present in corals, which Sr contributed a great percentage and was already taken into the calculation of Ca/P, calcium-de�cient HA should not be the form of calcium phosphate in CHA.Xu et al. have suggested that coral was converted to HA via two parallel pathways: (1) direct conversion from aragonite to HA, which starts early during the reaction; (2) from aragonite to calcite, then from calcite to -TCP, and �nally -TCP conversion to HA [30].A total of 17 days were required for complete conversion of coral into HA through both pathways mentioned previously.As a result, the calcium phosphate formed aer 2 days of hydrothermal reaction in this study should be a mixture of HA and -TCP but not purely HA.
e XRD patterns of the untreated and HA-converted specimens clearly indicate the change of composition on the surfaces of Goniopora.e untreated Goniopora in this study has the typical peaks of aragonite and looks very much like the XRD pattern of Goniopora reported in the literature [22,31].Although 2 small peaks are usually found around 28-29 ∘ C on the XRD patterns of HA, there is also a signi�cant peak 29 ∘ C reported on calcite [22,32,33].Calcite is a more thermodynamically stable form of calcium carbonate comparing to aragonite [30].e transformation usually occurs at about 300 ∘ C [34,35]; however, transformation to calcite is proposed to occur as an intermediate process during the hydrothermal conversion of coral aragonite into CHA at temperature <300 ∘ C [30].As a result, it is possible to detect calcite XRD peak in CHA or Sr-CHA.In general, CHA and Sr-CHA have similar XRD patterns, and all 3 groups have the representable peaks (26 ∘ , 32-34 ∘ , 40 ∘ , and 46-54 ∘ ) of CHA, which are also the peaks of well-crystallized HA [24,30,33,36,37].With the increase of Sr incorporation on synthetic HA particles, the XRD signal at 33-34 ∘ should diminish, but it is not the case in this study, while the patterns of CHA and Sr-CHA look almost the same [32].Only with very high Sr-incorporated HA (40-100%), there should be some shiing of peaks to the le at 28-29 ∘ to 27-28 ∘ (100% Sr), 32 ∘ to 31 ∘ , and 33 ∘ to 32 ∘ [10].Since only 6-14% of Sr was detected in CHA and Sr-CHA, no apparent signal shi was observed.
Incorporation of Sr altered the chemical composition and surface morphology of CHA, but the compressive strength was neither compromised nor enhanced.e compressive strengths of untreated coral, CHA, and Sr-CHA are in the range of 3.8 to 4.6 MPa, which were greater than the published value (2.62 MPa) for untreated Goniopora or similar to CHA (2.6-4.2MPa) made with Porites [8,22,38,39].e porosity and pore size of a material are crucial in determining the mechanical properties, while coral is such an interesting natural material that every single piece is unique; so it is difficult to have exactly the same compressive strength even for the same species of coral.Out of all mechanical properties of orthopaedic biomaterials, compressive strength is a critical factor to look into since our bone is usually under compression.Compressive strength of cancellous bone varied by locations, but in general it is in the range of 2-9 MPa [17,18,40].Based on the compressive strengths of this study, CHA and Sr-CHA converted from Goniopora are advisable for nonload bearing applications.
In order for Sr to have an effect on MC3T3-E1 proliferation, the incorporated Sr should �rst be able to be released from the scaffold.Sr is undoubtedly added to CHA, but the mechanism of incorporation is still unknown.As it can be seen on the SEM images of Sr-CHA, the Sr in the form of hexagon rod is strongly bounded with the �akelike crystal structures of CHA.e perfect structures of Sr and strong bonding with surface CHA might hinder the release of Sr.Initially, CHA had the greatest number of cells, but the result was reversed at d5, while Sr-CHA had more cells grown.HA conversion and Sr incorporation de�nitely made changes on the surface chemistry and morphology of coral giving the stimulatory effects of Sr-CHA on MC3T3-E1 cells.e increase of cell number on Sr-CHA indicates the potential of this Sr incorporation approach to enhance osteoblast proliferation on coral, in particular of relatively lower Sr content.
One specimen each of the micro-CT images was analyzed in this current study providing preliminary data of the in vivo properties of untreated coral, CHA, and Sr-CHA.Strontium ranelate is a registered medication for treating osteoporosis in many countries, and Sr is also gaining attention to be applied in orthopaedic biomaterials for the last one to two decades.An OVX rat model was chosen in this study to better examine the in vivo properties of CHA and Sr-CHA in poor quality bone such as osteoporotic bone in particular.During the period of study, a net volume loss was reported on untreated coral and CHA; more than 50% of the volumes were reduced at week 4 aer the surgery.e in vivo degradation of a scaffold should be site speci�c and could vary to a great extent among animals.Natural coral tends to be less favourable than CHA because of the rapid degradation with the scaffold almost completely degraded as early as 4 weeks [41][42][43][44].e loss of volumes of untreated coral and CHA showing the rate of degradation of scaffolds is agreed with the studies in the literature; while HA conversion has no obvious effect on delaying the degradation of coral or promotion of bone formation.On the other hand, there was an increase of the scaffold volume on Sr-CHA 1 week aer implantation.is may suggest a net gain of volume resulting from initial apatite formation and minimal scaffold degradation.While untreated coral and CHA degraded pretty quickly, Sr-CHA appeared to induce apatite formation and decelerate the degradation rate with 90% scaffold volume remained at week 4. Formation of apatite layer on the biomaterials has been suggested to induce bone formation [45]; thus, the increase of volume on Sr-CHA due to apatite formation suggests that it favours new bone formation.Unfortunately, it is technically challenging to differentiate apatite or new bone from a radiopaque scaffold primarily based on the micro-CT images.Nonetheless, it is clear that the degradation of Sr-CHA was slower than untreated coral, so this Sr incorporating treatment may be further evaluated to enhance osteoconductivity of coral.

Conclusions
A comprehensive study was done to modify and study the mechanical, in vitro and in vivo properties of Goniopora aer HA conversion with Sr incorporation.By characterizing the materials using EDX, SEM, and XRD, the surface morphology and chemical composition were detected with the presence of Sr.It is the �rst report on incorporating foreign element, Sr in this case, into CHA; it maintains the porous coral structure with the addition of bone stimulating factor Sr. CHA and Sr-CHA were treated to convert the carbonate of Goniopora into phosphate, while the in vitro results were encouraging with the possibility of low Sr release.Micro-CT analysis of Sr-CHA has suggested that this Sr-incorporated CHA could enhance bone formation and slow down the degradation of coral.To sum up, this study explored the idea of incorporating Sr into the natural porous structure of coral as potential scaffolding materials.

Con�ict o� �nte�ests
e authors declare no con�ict �nancial of interests in relation to the work described.
Analyses.Surface morphology, depth of HA conversion, and composition of the corals were studied by SEM, EDX, and XRD.Line EDX was performed on the epoxy embedded cross-sections of untreated coral and CHA to study the depth of conversion as shown in Figures2(a)-2(d).e dark grey and light grey areas in the SEM images (Figures2(a) and 2(c)) represent epoxy and coral, respectively.It can be observed that phosphorus was barely detected in raw coral while it reached about 1.3-4 k in CHA specimen (Figures 2(b) and 2(d))

F 3 :F 4 :
SEM images of: (a) untreated coral, (b) CHA, corals with Sr incorporation in (c) 100 mM, and (d) 500 mM Sr(NO 3 ) 2 solutions.Point EDX was done on a structure formed by hexagonal rods aer CHA conversion with Sr incorporation.T 1: Ratio of Ca + Sr/P and percentage of Sr incorporated aer CHA conversion calculated by EDX analysis.

F 5 :
EDX microanalysis spectra of untreated coral, CHA, and Sr-CHA.T 2: Measurements of compressive strength and porosity of untreated Goniopora comparing to allogenic cancellous bone, HA, or other bioactive glasses in the literature.

F 6 :
XRD patterns of untreated coral, CHA and Sr-CHA.

F 8 :F 9 :
e number of MC3T3-E1 cells on CHA and Sr-CHA on day 3 and 5 with initial seeding of 100,000 cells/scaffold measured by MTT assay ( * indicates statistically signi�cant difference for   0.05).Micro-CT images of the front and cross-sectional views of the untreated coral, CHA, and Sr-CHA at week 2 aer implantation.

F 10 :
Relative scaffold volumes of different corals at 0, 1, 2, and 4 weeks aer implantation calculated by micro-CT results.