Bioactive Glass Derived Scaffolds with Therapeutic Ion Releasing Capability for Bone Tissue Engineering Dreidimensionale bioaktive Glasgerüste mit therapeutischer Ionenfreisetzung zur Gewebezüchtung von Knochen Der Technischen Fakultät der Friedrich-Alexander-Universität Erlangen-Nürnberg zur Erlangung des Grades Doktor-Ingenieur vorgelegt von Herrn Dipl.-Ing. Alexander Hoppe aus Kustanaj
195
Embed
Bioactive Glass Derived Scaffolds with Therapeutic Ion Releasing ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Bioactive Glass Derived Scaffolds with Therapeutic Ion
Releasing Capability for Bone Tissue Engineering
Dreidimensionale bioaktive Glasgerüste mit therapeutischer
Ionenfreisetzung zur Gewebezüchtung von Knochen
Der Technischen Fakultät
der Friedrich-Alexander-Universität Erlangen-Nürnberg
zur Erlangung des Grades
Doktor-Ingenieur
vorgelegt von
Herrn Dipl.-Ing. Alexander Hoppe
aus Kustanaj
Als Dissertation genehmigt
von der Technischen Fakultät
der Friedrich-Alexander-Universität Erlangen-Nürnberg
Tag der mündlichen Prüfung: 14.07.2014
Vorsitzende des Promotionsorgans: Prof. Dr.-Ing. habil. Marion Merklein
Gutachter: Prof. Dr. Aldo R. Boccaccini
Prof. Dr. Uwe Gbureck
Acknowledgment
Acknowledgment
My greatest gratitude and appreciation goes to my supervisor Prof. Boccaccini who
gave me the opportunity to start my PhD research at the newly formed Institute of
Biomaterials. I want to thank him for his support and advices during all stages of
my PhD and also for giving me the freedom to develop my own research ideas and
framing my entire PhD project. Exceptionally, I want to thank him for supporting
and enabling numerous research fellowships and workshops abroad including stays
in Turin, Montreal, São Carlos (Brazil) or Buenos Aires which have had huge
impact not only my scientific career but also on my private life.
I want to thank Dr. Julia Will who has been supporting me from the very beginning
of my scientific career by supervising my diploma thesis whose great outcome
encouraged me to stay in academia and to pursue a PhD. She has always succeeded
to motivate me in all my research efforts and was helping with any issue during my
PhD studies.
My gratitude extends to Dr. Rainer Detsch whom I thank for introducing me to the
“world of cell biology”, for all the fruitful discussions and proof reading of parts of
my thesis. I also thank him for persistently motivating me to finish my PhD and his
unlimited trust in my skills.
I am thankful to Alina Grünewald for teaching me practical skills for the work in
the cell lab and for always being helpful no matter what issue would come up. I also
thank her for creating an extremely nice working atmosphere with many enjoyable
hours in the cell lab.
I want to thank Tobias Zehnder, Bapi Sarker and Raquel Silva and the complete
“Henkestrasse-crew” for the great time sharing the office, the labs and the lunch
breaks while exploring the infinite number of culinary options the nearby Aldi
offered us.
I would like to thank all my fellow PhD students for the grate time in and outside of
the institute.
I thank Dr. Gerhard Frank for his tireless patience and his help with the
management of the institute’s projects and administration issues.
Acknowledgment
I am grateful to Heinz Mahler for his help with technical and computer related
problems.
I would like to express my gratitude to all bachelor and master student I was lucky
to supervise who all contributed to my PhD project: Alexander Kent, Vincent
Bürger, Lukas Weidenbacher, Harald Unterweger, Stefan Grimm, Tobias Reichel,
Florian Ruther and Katharina Rzepka.
All collaborators who contributed to the completion of my PhD work are gratefully
acknowledged, in particular: Eva Springer for countless SEM images, Robert
Meszaros and Prof. Wondraczek for the help with glass melting, Chris Stähli and
Prof. Nazhat for helping with glass structural and degradation studies, Stefan
Romeis and Jochen Schmidt for ICP measurements and fruitful discussion, Jonathan
Lao and Prof. Jallot for micro-PIXE-RBS measurements, Juan Catallini and Prof. V.
Mourino for capillary electrophoreses measurements, Prof. Janackovic and Bojan
Jokic for the collaboration on bioactive glass.
I want to express my deepest gratitude to my parents and my brother for their
support during the last years who never gave up their faith that I would finally
graduate.
Abstract
Abstract
Regarding the need for highly vascularised engineered bone constructs to regenerate
bone defects, loading biomaterials with angiogenic therapeutics, like metallic ions,
has emerged as promising approach to develop novel biomaterials for regenerative
medicine.
In the present work, bioactive silicate glasses based on 45S5 and 1393 compositions
containing the well-known angiogenic ions Cu and Co were fabricated and used for
templating 3D scaffolds by foam replica technique.
It was shown that bioactive glasses can be used as carrier for therapeutic metal ions
with controlled release kinetics tuneable via the glass composition. The degradation
and bioactivity studies revealed that incorporation of metallic ions does not impair
the bioactivity and that traces of metallic ions are incorporated in the calcium
phosphate layer formed on the scaffolds surface in contact with simulated body
fluid. Cu and Co ions in therapeutic range were released and cell culture assays
confirmed high compatibility of 45S5-Cu and 1393-Co derived particulate glasses
and corresponding scaffolds with MG-63 osteoblast-like cells, human bone marrow
derived stem cells (hBMSCs) as well as human dermal micro vascular endothelial
cells (hDMECs). Cellular response, however, was shown to be dependent on the
concentration of the metallic dopant. Furthermore, Cu ions were shown to stimulate
angiogenesis in vitro by enhanced VEGF (vascular-endothelial growth factor)
expression in hBMSCs likely due to the activation of the HIF-1 transcriptional
factor. In a co-culture study of hDMECs and hMSCs 1 wt% Cu containing 45S5
scaffolds also enhanced the expression of endothelial cell specific markers vWF and
VEGFR and stimulated the hDMECs towards formation of prevascular tube-like
structure indicating the overall angiogenic potential of Cu. Altogether, Cu or Co
doped BG scaffolds represent a new family of highly promising materials for
applications in bone regeneration.
Kurzzusammenfassung
Kurzzusammenfassung
Die Vaskularisierung von gezüchtetem Knochengewebe ist von entscheidender
Bedeutung für die klinische Applikation dieser Gewebekonstrukte und stellt hohe
Anforderungen an das als Trägerstruktur (Scaffold) eingesetzte Biomaterial.
Ein vielversprechender Ansatz ist dabei das Dotieren von anorganischen
Biomaterialien mit Spurenelementen mit therapeutischer, angiogener Wirkung um
die Vaskularisierung von Scaffolds zu stimulieren.
In der vorliegenden Arbeit wurden ausgehend von zwei bekannten
Zusammensetzung, 45S5 und 1393, Cu and Co dotierte bioaktive Gläser (BG)
synthetisiert und anschließend zur Herstellung von 3D Poröser Scaffolds verwendet.
Es konnte gezeigt werden, dass bioaktive Glas-Scaffolds als Trägerstrukturen zur
kontrollierten Freisetzung von Metallionen geeignet sind und die
Freisetzungskinetik über die Glaszusammensetzung gesteuert werden kann.
Bioaktivitäts- und Degradationsuntersuchungen in simulierter Körperflüssigkeit
ergaben, dass die dotierten Elemente keinen negativen Einfluss auf die Bioaktivität
haben und Cu bzw. Co in der Hydroxylapatitschicht auf der BG Oberfläche
substituiert sind. Außerdem konnten Cu und Co in einem als therapeutisch wirksam
geltenden Bereich freigesetzt werden, wobei Zelluntersuchungen hohe
Biokompatibilität der Cu und Co dotierten Scaffolds mit osteoblastähnlichen Zellen,
mesenchymalen Stammzellen (MSCs) sowie Endothelzellen (ECs) bestätigten.
Darüber hinaus, wurde deutlich, dass Cu2+
Ionen die Expression des angiogenen
Signalmoleküls VEGF in humanen Stammzellen signifikant erhöhen und in einer
Co-Kultur aus MSCs und ECs die Ausbildung tubulärer prevaskulärer Strukturen
durch ECs stimuliert. Dies bestätigt das angiogene Potential von Cu dotiertem BG.
Insgesamt stellen Cu and Co dotierte bioaktive Gläser aufgrund ihrer hohen
Bioaktivität und stimulierenden Wirkung auf relevante Zelltypen vielversprechende
Materialen für den Einsatz in der regenerativen Medizin dar.
Cell viability and cell proliferation .................................................. 57 Analysis of osteogenic and angiogenic markers / Gene
expression and protein release ......................................................... 58 Cell morphology and cell adhesion ................................................. 59
Evaluation techniques of 2D experiments with EC ......................... 60 Statistical analysis ............................................................................ 62
3.7.4 Evaluation of the scaffolds ................................................................ 63
Indirect study (2D) ........................................................................... 63 Direct study (3D) ............................................................................. 63 Co-culture of hBMSCs and ECs ...................................................... 63
3.8 In vivo study .............................................................................................. 64
4 Results and Discussion.................................................................................... 66
4.1 Cu containing 45S5 bioactive glasses ....................................................... 66
Mechanical properties ...................................................................... 76 Acellular bioactivity in SBF ............................................................ 77 Degradation of 45S5 derived scaffolds and Cu release ................... 90
4.1.4 In vitro cell response .......................................................................... 94
Powder cytotoxicity ......................................................................... 94 Cell attachment on 2D pellets .......................................................... 97 In vitro cell studies with 3D scaffolds ........................................... 100
4.1.5 In vivo evaluation ............................................................................. 112
4.2 Cobalt containing 13-93 based glasses .................................................... 115
Sigma-Aldrich, Germany) and basic Cu carbonate (CuCO3*Cu(OH)2, Sigma
Aldrich, Germany). The raw materials were well homogenized and melted in a
platinum crucible at 1450 °C for 45 min (Furnace LHT 3 KW, C295 Control Unit,
Nabertherm, Germany). The nominal compositions of the glasses investigated in
this study are given in Table 7.
Table 7: Nominal 45S5 Bioglass® derived glass compositions with different Cu contents.
Glass Composition in wt%
SiO2 Na2O P2O5 CaO CuO
45S5 45 24.5 6 24.5 -
45S5-0.1Cu 45 24.5 6 24.4 0.1
45S5-1Cu 45 24.5 6 23.5 1
45S5-2.5Cu 45 24.5 6 22 2.5
3.1.2 Co containing 13-932
Co-containing bioactive glasses based on the 1393 composition (1393-Co) were
fabricated using melt-derived technique. Silicic-acid (pure anhydrated, Riedel de
Haën, Germany), Sodium-carbonate (Riedel de Haën, Germany), Potassium-
carbonate (Riedel de Haën, Germany), Sodium-dihydrogenphosphate (Riedel de
1 Melting of the 45S5-Cu glass series was carried out at the Institute of Glass and Ceramics (Glass group, Prof. Wondraczek) under assistance of R. Meszaros 2 Melting of the 1393-Co glasses was carried out by B, Jokic and Prof. Jankovic at Universtiy of Belgrade
Scientific iCAP 6500). Values were calibrated against certified standards serially
diluted with SBF to 100, 10, 1 and 0.1 and ppm. (4% nitric acid was added to all
samples and standards).
Quasi-dynamic conditions4
Ion release (Cu, Si) under quasi-dynamic conditions from scaffolds soaked in SBF
was measured using the ICP-OES Optima 8300 (Perkin Elmer, USA). Five point
3 ICP measurements of ions released under static conditions were carried out by C. Stähli at McGill University, Montreal, Canada 4 ICP measurements under quasi-dynamic conditions were carried out by S. Romeis and J. Schmidt at the Institute of Particle Technology, University of Erlangen-Nuremberg
Materials and Methods 53
calibrations (25; 10; 5; 1; 0.1 ppm) were performed by diluting certified standards
(Carl Roth, Germany) with SBF. Given errors are estimated by linear regression
analysis. The samples were measured in triplicate and mean values with standard
deviation were derived. For scaffold digestions a of 2 ml hydrofluoric acid (48 %,
supra purity, Roth Chemicals, Germany), 2 ml nitric acid (69 %, supra purity, Roth
Chemicals, Germany), and 4 ml hydrochloric acid (35 %, supra purity, Roth
Chemicals, Germany) was used. After the microwave procedure 2 g of boric acid
(>99.8 %, Roth Chemicals, Germany) was added and the total volume was set to
250 ml.
3.3.5 Micro-Ion beam5
Analyses of the glass derived scaffolds/biological fluids interface were carried out
using nuclear microprobes at CENBG (Centre d’Études Nucléaires de Bordeaux-
Gradignan, France).264
PIXE-RBS analyses were performed on the nanobeam line
with a proton scanning micro-beam of 3 MeV energy and 60 pA in intensity. The
beam diameter was nearly 1 µm. An 80 mm2 Si(Li) detector was used for X-ray
detection, orientated at 135° with respect to the incident beam axis and equipped
with a beryllium window 12 µm thick. An aluminium funny filter of 100 µm in
thickness with a hole of 2 mm in diameter was added on the detector. The software
SUPAVISIO was used to define the different regions of interest with the use of
masks. These masks isolate the spectra corresponding to the region of interest in
order to calculate the elemental composition in that region. Quantification of PIXE
spectra was done using GUPIX software.265
The data was calibrated against NIST
standard reference glass materials. Relating to RBS, a silicon particle detector
placed at 135° from the incident beam axis provided the number of protons that
interacted with the sample. Data were treated with the SIMNRA code.29 The
thickness of reaction layers formed during immersion in SBF was determined using
ImageJ software. Three measurements at different spots were taken per chemical
map and the mean value and standard deviation were derived from 3 chemical
maps.
5 PIXE-RBS analysis was carried out by J. Lao and E. Jallot, fClermont Université, Université
Blaise Pascal
Materials and Methods 54
3.3.6 XRD
X-ray diffraction analysis was performed using a D8 ADVANCE X-ray
diffractometer (Bruker, Madison, US) in a 2Theta range from 20-80°. BG powders
were dispersed in ethanol. Then, the solution was dripped on off-axis cut, low
background silicon wafers (Bruker AXS, Germany). BG derived scaffolds were
powdered and prepared in the same way.
3.3.7 X-ray microtomography (µCT)6
µCT analysis was performed with a SkyScan 1172 (SkyScan, Kontich, Belgium).
Briefly, 45S5 glass-ceramic scaffolds were analyzed through a 360° flat-field
corrected scan at 40 kV and 250 µA, and then reconstructed (NRecon software,
SkyScan) with a beam hardening correction of 10, a ring artifact correction of 20
and an “auto” misalignment correction. The 2D analysis (software CTAn, SkyScan)
of reconstructed microCT transverse cross-sections of 45S5 glass-ceramic scaffolds
was carried out using a greyscale intensity range of 16 to 255 (8 bit images) in order
to remove background noise. 3D reconstruction and visualization of the scaffold
microstructure was achieved using CTVol software, Skyscan.
3.4 Scaffold fabrication
BG derived scaffolds were produced using the foam replica technique as
schematically shown in Fig. 9.68
Briefly, polyurethane foam (45 ppi, Recticel, UK)
was immersed in slurry containing 60 wt% BG-particles and 1.1 wt% PVA as
binder. The green bodies were dried at 60 °C for 24 h and sintered at 1050 °C for 2
h for glass series 45S5, 45S5-0.1Cu and 45S5-1Cu and 1000 °C for 45S5-2.5Cu
glass composition. The 1393-Co glass derived scaffolds were densified at 700 °C,
680 °C, and 670 °C for 1393. 1393-1Co and 1393-5Co, respectively.
Also, multiple coatings were applied by dip-coating the sintered foam and
centrifugation in order to remove the excess slurry.
6 µCT analysis was performed by B. Marelli at McGill Universtiy, Montreal, Canada
Materials and Methods 55
Fig. 9: Scheme of the foam replica technique used for scaffold fabrication.
3.5 Acellular bioactivity in simulated body fluid (SBF)
In order to assess the in vitro bioactivity, the scaffolds (5x5x5 mm3) were immersed
in 50ml simulated body fluid (SBF) for 1, 3 7, 14 and 21 days. SBF was produced
according to the protocol by Kokubo et al.99
After immersion, the samples were
gently washed with deionized water, dehydrated with acetone and dried at 60 °C for
12 h. Analysis of CHA formation was performed with SEM, FT-IR and micro-Ion
beam techniques.
3.6 Compressive strength
Compressive strength of the scaffolds (5x5x5 mm3) was measured using a tensile
testing machine (Z050, Zwick Roell, Germany). The testing speed was 10 mm min-
1. In order to assure homogenous loading of the scaffolds a polymeric rubber
interlayer was placed between the sample and the steel plates. 10 samples were
measured per scaffolds series and the standard deviation was derived.
3.7 Cell tests
In vitro cell investigations were carried out in order to assess the biocompatibility of
the developed Cu and Co containing bioactive glasses. Different cell types were
used in order to evaluate specific cell response analysing the cytotoxicity,
osteogenic differentiation and angiogenic potential. Hereby, 2D and 3D experiments
were conducted using glass powders, dense pellets and 3D scaffolds. Table 9 gives
Materials and Methods 56
an overview of cell types materials modifications used in the cell culture
experiments.
Table 9: Overview of the cell types and material modifications used for biocompatibility
assessment.
Cell type
Osteoblast-like
cells (MG-63)
Human bone marrow
derved stem cells
(hBMSC)
Human dermal
microvascular
endothelial cells
(hDMECs)
Co-culture of
hBMSC and
hDMECs
Aim Cytotoxicity Osteogenic
differentiation Angiogenic potential
Assessment Cell viability
Cell morphology
Cell vitality
Gene expression
Cell vitality, formation of tube-like
structures, VEGF release
Powder , - - -
2D pellets - - -
3D scaffolds
direct , - -
indirect - ,
...45S5-Cu, ...1393-Co
3.7.1 Cells and culture
Osteoblast-like cell line MG-63
MG-63 osteoblast-like cells Human osteosarcoma cell line (Sigma-Aldrich,
Germany) were used cultured at 37 °C in a humidified atmosphere of 95 % air and 5
% CO2, in DMEM (Dulbecco’s modified Eagle’s medium, Gibco, Germany)
with primers as shown in Appendix Fig. A 2 (CFX 96 real time systems, Bio-Rad,
Munich, Germany). The gene expressions were normalized to internal GAPDH
expression, and the relative fold change was expressed by comparing to that of
45S5 samples.
Cell morphology and cell adhesion
SEM analysis of the cells
After cell culture experiments the scaffolds were washed with PBS, fixed with a
solution containing 3 vol.% glutaraldehyde (Sigma, Germany) and 3 vol.%
paraformaldehyde (Sigma, Germany) in 0.2 M sodium cacodylate buffer (pH 7.4)
and finally rinsed three times with PBS For SEM analysis (ESEM, Quanta 200, FEI,
Netherlands). All samples were dehydrated in a graded ethanol series (30, 50, 75,
90, 95 and 99.8 vol.%). Samples were maintained at 99.8 vol.% ethanol and critical-
point dried (Leica, EM CPD 300). Prior to SEM examination the samples were
sputtered with gold.
Materials and Methods 60
Actin staining
In order to observe and to quantify the cell attachment on the BG samples labelling
of the cell skeleton (Actin labelling) was performed and the spreading behaviour of
the cells was analysed using fluorescence microscopy (Axio Scope.A1, Leica,
Germany). Actin staining was performed using a commercially available dye kit
(Alexa Fluor® 488 Phalloidin, Life Technologies GmbH, Darmstadt, Germany) at 5
units/ml concentration to see the cell spreading on the bio-active glass scaffold. The
nuclei of the cells are counterstained by 300 nM DAPI (SelectFX®, Life
Technologies GmbH, Darmstadt, Germany).
Vybrant™ cell-labeling
To analyse the adherent growth and distribution of cells on the BG scaffold
samples, commercially available Vybrant™ cell-labeling solution (Molecular
Probes, The Netherlands) was used. After 48 hours of incubation, cell culture
medium was removed and staining solution (5 μl dye labelling solution to 1 ml of
growth medium) was added and incubated for 15 min. Afterwards the solution was
removed, the samples were washed with PBS (phosphate buffered saline, Gibco)
and cells on the BG samples were fixed by 3.7 vol.% paraformaldehyde. The
samples were prepared for confocal scanning laser microscopy (CSLM, Leica TCS
SP5 II, Germany) to analyse cell morphology and distribution.
Evaluation techniques of 2D experiments with EC
Flow Cytometry analysis (FACS)
The human endothelial cell surface markers (Cluster of differentiation, CD31; von
Willebrand factor, vWF, Vascular endothelial growth factor 2, VEGFR2) were
stained at 5x104 cells for each antigen. CD31 is a known platelet adhesion protein
which is expressed at cell connecting junctions and is found in early and mature
endothelial cells and is known to be involved in activation of integrins. CD31 was
stained by mouse anti-human CD31-Biotin followed by a second staining step with
Streptavidin PerCP-eFluor® 710 (both from eBioscience, San Diego, CA, USA).
Materials and Methods 61
vWF is an important platelet adhesion factor which is also used as a marker for any
pathological dysfunctions of endothelial cells as increased release of vWF from
endothelial cells is related to endothelia cells damage. vWF was stained by sheep
anti-human vWF-FITC (Abcam, Cambridge, UK). VEGFR2, an important
signalling molecule for endothelial cell mitogenesis and migration, was stained by
mouse anti-human CD309 (VEGFR2)-Alexa Fluor® 647 (Biolegend, San Diego,
CA, USA). All staining steps were performed on ice for 30 minutes in dark. Stained
cells were analyzed by FACS-Calibur (BD Biosciences, San Diego, CA, USA) and
Cell Quest software (Beckton Dickinson, Heidelberg, Germany). Data
quantifications were perfomred by FlowJo software (Tree Star, Inc., Ashland, OR,
USA).
Matrigel™ sprouting assay
For analysis of capillary tube formation, 75 µl of Matrigel™ (Becton Dickinson,
Heidelberg, Germany), an extracellular mouse sarcoma matrix known to produce
pro-angiogenetic stimulus both in vitro and in vivo, was pipetted into each well of a
96-well plate (Falcon, Heidelberg, Germany) and incubated at 37°C for 60 minutes.
HDMECs were harvested at week 1 or week 2 and suspended at 50,000 cells per
150 µl of EGM MV2 media. 150 µl of this media were added to the Matrigel coated
96-well plates and incubated for 24 h at 37°C. Capillary tube formation on Matrigel
was observed with a light microscope (DMIL, Leica, Germany) and images were
processed with Leica application suite software (Leica GmbH, Wetzlar, Germany).
LDL (Low density lipoprotein)-staining
At end time points, the media were removed and the cells were washed once with
PBS to remove non-adherent cells. The wells were incubated with 2.5 µg/ml ac-
LDL-Alexa Fluor®
488 (Life Technologies GmbH, Darmstadt, Germany) for 4 h at
37°C. Afterwards cells were washed twice with PBS and further analysed by a
fluorescence microscope.
Materials and Methods 62
VEGF quantification ELISA
At each time point, the cultured media were collected from 3 different samples,
pooled together, labeled and frozen at -20°C. At the end of all experiments, the
frozen media were thawed overnight at 4°C and the VEGF content of the media was
quantified by using an ELISA kit (Thermo Fisher Scientific GmbH, Schwerte,
Germany).
Statistical analysis
The statistical significant of the results was evaluated by one-way analysis of
variance (ANOVA). The level of the statistical significance was defined at p < 0.05
(Origin 8.1G, OriginLab Corporations, USA).
3.7.3 Powder cytotoxicity
Initial cytotoxicity tests of the glasses fabricated were assessed by culturing BG
powders in direct contact with MG-63 cells. The BG powder was dispersed in
DMEM and ultra-sounded for 5 min to break agglomerates. MG-63 cells were
cultured in direct contact with BG particles at concentrations from 0.1 to 1000 µg
ml-1
) in 600 µl DMEM for 48 hours. As positive control cells cultured without BG
particles, as negative control ZnO was used.
Cell distribution and morphology were evaluated using phase contrast light
microscopy (LM, Nikon Eclipse TE 2000-U, Japan). Cells viability was assessed
through WST-8 test (mitochondrial activity) and LDH assay (cell number) after 48h
of incubation. From these the LC50 value (lethal concentration of the used particles
where the activity of the cells is reduced to 50%) was derived.
Materials and Methods 63
3.7.4 Evaluation of the scaffolds7
Indirect study (2D)
In the indirect setup the cells were seeded on the bottom of the cell culture well
exposed to the ionic dissolution products released from the scaffold which was
placed in a permeable insert as shown in Fig. 10. Indirect studies were performed
with hBMSCs. Cell morphology was detected by light microscopy, ALP activity,
VEGF and Runx2 expression was analysed with PCR technique.
Fig. 10: Indirect setup for cell culture studies with scaffolds.
Direct study (3D)
Additionally, in order to test the interaction of cells and the scaffold surface the cells
were directly seeded on the BG derived scaffolds. MG-63 cells and hBMSCs were
used. Before cell seeding the scaffolds were preconditioned in DMEM for 5d. The
cells attachment and growth was evaluated with SEM, Actin staining, Live staining
(Calcein/Vybrant). Further, ALP activity was detected.
Fig. 11: Direct setup for cell culture studies with scaffolds.
Co-culture of hBMSCs and ECs
A co-culture model of hDMECs and hBMSCs was applied to 45S5-Cu derived
scaffolds as shown in Fig. 11. hBMSCc were seeded directly on the scaffolds and
7 Cell studies with hMBSCs were carried out in collaboration with D. Hiller and S. Rath, U. Kneser and R. Horch, Plastic and Handsurgery Department, University Clinic Erlangen
Materials and Methods 64
the constructs where placed into permeable membranes. hDMECS were seeded in a
concentration of 100.000 cells/ml and on the well bottom being exposed to ionic
dissolution products released from the scaffods as well as effects of the hBMSCs.
Cell morphology of the hDMECS was observed with light microscopy. Further,
HDMSCs specifically the LDL uptake and FACS analysis of endothelial markers
vWF, VEGFR and CD31 was performed.
Fig. 12: Set-up for the co-culture experiment with hBMSc and hDMECSs using permeable
insert containing the scaffold seeded with hBMSCs.8
3.8 In vivo study
In vivo evaluation of the angiogenic potential of Cu-doped glass scaffolds was
carried out at the Department for Hand plastic surgery, University of Erlangen9.
Six male Lewis rats were used and the experiments were carried out in compliance
with the animal care committee of the FAU and the government of Mittelfranken.
45S5-1Cu BG derived scaffolds were chosen as most promising candidates for in
vivo study, to prove their angiogenic potential in relevant model in addition to in
vitro results. Fabricated scaffold were crushed into small pieces with a diameter of ~
2-3 mm in order to create granulate like units. A Teflon chamber with an inner
diameter of 10 mm and a height 10 mm was used. The half of the chamber was first
filled scaffold granules and fibrin gel with a fibrinogen concentration of 10mg ml-1
and thrombin concentration of I.U ml-1
. Afterwards the AV loop was placed on the
BG/fibrin gel matrix and the chamber was completely filled with BG granules/fibrin
gel. The construct was, then, implanted subcutaneously and sutured to the
musculature of the rats. The arrangement of the AV-loop and the BG granules in the
8 Cell studies with hDMECs were carried out in collaboration with A. Brandl and O. Bleizifffer, U. Kneser and R. Horch, Plastic and Handsurgery Department, University Clinic Erlangen 9 In vivo studies were carried out by Ulrike Rothensteiner from the group by Dr. A. Arkudas, Dr. U. Kneser and Prof. Dr. Horch
Materials and Methods 65
teflon chamber are shown in Fig. 13. After 3 weeks of implantation the constructs
were explanted and were investigated by means of microCT, histological and
morphometrical analysis. 3D reconstruction of the vessel sprouting, vascular
density as wells as vessel number and vessel density were derived were derived.
Fig. 13: In vivo setup for assessing the angiogenesis of 45S5 and 45S5-1Cu BG scaffolds. Firstly,
the teflon chamber is filled half with BG granules and the AV-loop is placed (a). Afterwards the
chamber is fully filled with BG granules and implanted subcutaneously (b).
Results and Discussion 66
4 Results and Discussion
In this chapter the properties of different bioactive glass compositions and the
respective glass-derived scaffolds are described. The influence of metal ion doping
on the glass structure and glass properties is discussed. Further the structural and
mechanical properties of the glass derived scaffolds are shown. The apatite forming
ability as indicator for acellular bioactivity is assed in simulated body fluid and the
physic-chemical reactions are monitored in detail by SEM, FT-IR and RBS-PIXE
techniques. Moreover, the ion release kinetics from the BG scaffolds is observed
and the suitability of glass derived scaffolds as “carrier” of therapeutic metal ions is
discussed. The cellular biocompatibility is analysed in relevant cell culture studies
and also an in vivo model is applied.
4.1 Cu containing 45S5 bioactive glasses
4.1.1 Glass properties
In this section the glass structure of the synthesised Cu containing glasses is
discussed on the basis of FT-IR, Raman and NMR-analyses and is related to glass
thermal properties derived from DSC measurements.
Glass structure
Fig. 14 shows the FT-IR spectra of as-fabricated glass powder. The bands at ~ 500
cm-1
and ~1080 cm-1
can be assigned to Si-O-Si stretching and Si-O-Si bending
modes, respectively.267
The peak at 1460 cm-1
is related to carbonate species
adsorbed on the BG surface.268
No major differences for Cu-containing glasses can
be observed compared to undoped reference 45S5 glass indicating that the main
structure of the glass network remains unchanged. However, the intensity of the
non-bridging oxygen, Si-ONBO peak (~930 cm-1
), is decreasing with increasing Cu-
content in the 45S5 glass.
Similar observations can be made from Raman analysis, Fig. 15. The bands at ~620
cm-1
and 1060 cm-1
can be assigned to the rocking and stretching mode of Si-O-Si
Results and Discussion 67
band, respectively. While no major differences are observed for Cu-containing
glasses the band at 860 cm-1
which is assigned to non-bridging oxygen-silica band
Si-O2NBO is reduced in intensity for the 45S5-2.5Cu. Also the Si-O band at 1060 cm-
1 increases in intensity for the glasses with higher Cu content (>1%).
Fig. 14: FT-IR spectra of as-fabricated Cu-doped 45S5 glasses. A. Hoppe et al. J. Mater. Chem.
B 1 (2013), p. 5659. - Reproduced by permission of The Royal Society of Chemistry.
Fig. 15: Raman spectra of as-fabricated Cu-doped 45S5 glasses. A. Hoppe et al. J. Mater.
Chem. B 1 (2013), p. 5659. - Reproduced by permission of The Royal Society of Chemistry.
Results and Discussion 68
Fig. 16 shows the UV-VIS spectra of the glasses. For reference 45S5 glass a
continuous transmission in the range from 300 to 2000 nm is observed. Cu
containing glasses 45S5-0.1Cu, 45S5-1Cu and 45S5-2.5Cu, however, show an
absorption band at 800 nm which can be assigned to the presence of Cu2+
in the
glass network.269, 270
Fig. 16: UV-VIS spectra of 45S5-Cu BGs showing an absorption band at ~800 nm which is
assigned to Cu2+
. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. - Reproduced by
permission of The Royal Society of Chemistry.
For 45S5-0.1Cu no NMR measurements were performed since the Cu content is too
low and its influence on the 45S5 glass network structure cannot be resolved by
NMR. Fig. 17 shows the 29
Si MAS NMR spectra of the 45S5and 45S5-2.5Cu
glasses. The effect of Cu on 45S5 silicate glass structure is discussed on the basis of
45S5-2.5Cu data only as the NMR spectra for 45S5-1Cu (data not shown) and
45S5-2.5Cu show almost exactly the same characteristics and are completely
overlapped. An asymmetrical peak is observed for both glasses with the overall
peak position of the Si-related signal slightly shifted to negative values for 45S5-
2.5Cu glasses by 0.75 ppm. The NMR results show a broadening of the 29
Si peak.
Compared to reference 45S5 sample the FWHM (full width at half maximum) of
the Si signal for 45S5-2.5 Cu increased to 13.8 ppm (+ 1ppm).
Results and Discussion 69
Table 10: Distribution of Qn units calculated from the peak area of the deconvoluted
29Si NMR
signal and network connectivity NC for the 45S5 and 45S5-2.5Cu samples. A. Hoppe et al. J.
Mater. Chem. B 1 (2013), p. 5659. - Reproduced by permission of The Royal Society of
Chemistry.
45S5 45S5-2.5Cu
Q1 26.9% 12.7%
Q2 39.5% 47.6%
Q3 33.1% 38.9%
NC* 1.99 2.17
Commonly the structure of silicate glasses is described by quantifying the Qn
distribution (a Q species is a Si atom with n bridging oxygen atoms). Hence, for
detailed analysis of the NMR results for the 29
Si NMR signal of 45S5 and 45S5-2.5
Cu the curves were fitted and peak deconvolution applying Gauss profiles was
performed, Fig. 17. The glass structure of both glasses is dominated by Q2 with
smaller fractions of Q1 and Q
3 species as depicted in Fig. 17. For 45S5-2.5Cu a
slightly larger fractions of Q2 and Q
3 species is observed at the expense of Q
1. The
fractions of Qn and resulting network connectivity (NC) can be calculated from the
peak area. The exact values for the Qn distribution obtained for 45S5 and 45S5-
2.5Cu and corresponding values for the network connectivity (NC) are given in
Table 10. NC was calculated according to:
𝑁𝐶 = 1𝑄1 + 2𝑄2 + 3𝑄3 Eq. 4
Results and Discussion 70
Fig. 17: MAS NMR 29
Si analysis for 45S5 and 45S5-2.5Cu glass samples. Gauss fit and peak
deconvolution show the distribution of the Qn species in the glass network. A. Hoppe et al. J.
Mater. Chem. B 1 (2013), p. 5659. - Reproduced by permission of The Royal Society of
Chemistry.
Thermal properties
Fig. 18 shows the DSC curves for the 45S5-Cu glasses as fabricated. The
characteristic temperatures of glass transition Tg, crystallization onset and maximum
Tc and Tp and melting temperature Tm are given in Table 11. Both glasses show two
crystallisation events described by Tc1, Tc2 and Tp1, Tp2, respectively. Compared to
the 45S5 reference, for 45S5-2.5Cu a decrease of Tg, Tc1, Tp1 and Tm is observed
while Tc2 increases and Tp2 remains constant. For the lower Cu concentrations,
however, no significant differences in the DSC diagram for the characteristic
Results and Discussion 71
temperatures were observed. The origin of the crystalline phases occurring during
the sintering of 45S5 bioactive glass is discussed in 4.1.2.
Table 11: Characteristic features (temperatures in °C) of the Cu-containing glasses observed
with DSC.
Glass Tg Tc1 Tc1-Tg Tp1 Tc2 Tp2 Tm
45S5 522 592 70 682 756 908 1105
45S5-0.1Cu 526 611 85 696 - - 1110
45S5-1Cu 519 610 91 698 783 832 1112
45S5-2.5Cu 496 610 114 645 806 908 1044
Fig. 18: DSC analysis of 45S5-Cu glasses showing the glass transition point and two
crystallisation events at ~650-700 °C and 850 °C, respectively. A. Hoppe et al. J. Mater. Chem.
B 1 (2013), p. 5659. - Reproduced by permission of The Royal Society of Chemistry.
Role of CuO in 45S5 glass structure
The 45S5-1Cu and 45S5-2.5Cu glasses show almost the same characteristics.
Hence, 45S5-2.5Cu glass is used for deeper discussion of the effect of Cu doping on
the structure of 45S5 silicate glass.
FT-IR analysis, Raman spectroscopy, NMR and UV-VIS spectroscopy were applied
in order to investigate the glass structure and the effect of Cu-doping. It was
confirmed through UV-VIS spectroscopy that Cu is present as Cu2+
in the glass
Results and Discussion 72
structure which has been also proposed for soda silicate glasses containing CuO
showing Cu2+
valence state of Cu.271
Cu2+
(as well as Cu+) predominantly work as
network modifiers and are incorporated in the glass work matrix in octahedral
coordination.272
According to Abdrakhmanov et al.273
Cu2+
is surrounded by two
non-bridging oxygens (NBO) in order to achieve electro neutrality. Since no major
differences were observed in FT-IR and Raman spectroscopy the Q2 SiO2 network
of the 45S5 glass seems to be dominant and is not significantly affected by the Cu-
doping. Basically, according to FT-IR and Raman analyses a decreased intensity of
the Si-ONBO bond with increasing Cu content and at the same time increased
intensity of asymmetric Si-O vibration mode for glass with high Cu contents in the
glass was observed.
Si-NMR is a well-established experimental technique for characterising the
structure of silicate glasses which allows detailed insight about the distribution of
Qn species. As described in 2.3.2 the best descriptive model describing the structure
of 45S5 BG is achieved by assuming a trinodal distribution of Q1, Q
2 and Q
3 silica
species. According to this for the experimental NMR data the best fit of the 29
Si
peak (R2=0.99) was achieved by applying a ternary peak deconvolution, Fig. 17.
Concerning the 29
Si MAS-NMR analysis a shift of the peak maximum was observed
for 45S5-2.5Cu compared to 45S5 control. According to Elgayar et al.274
the
negative shift of the maximum to lower frequencies of 0.75 ppm between the 45S5
and 45S5-2.5Cu is likely caused by changing the cation in the glass structure which
is accomplished with an increase of the Q2 species in the glass network neutralized
with Cu2+
ions. Furthermore, the increased line width of the 29
Si peak indicate more
structurally disordered Q2 and Q
3 species neutralized by Cu
2+.274
This might be due
to the larger ion radius of Cu2+
replacing Ca2+
.
The higher fractions of Q2 and Q
3 species and hence the resulting higher network
connectivity in 45S5-2.5Cu samples suggests a repolymerisation effect caused by
introducing CuO into the glass network. Cu are also bigger ion, so the SiO2 network
is disrupted, a higher amount of silica chains Q2 are present and also there is
indication that the chains are repolymerised which results in higher Q3 fractions and
less terminating Q1
species. This is likely due to the more covalent character of the
Cu-O bond compared to Ca-O bond allowing repolymerisation of Si-NBOs.
Results and Discussion 73
Accordingly, FT-IR and Raman spectra revealed decreased intensity of the NBO
peak which is consistent with higher Q2 and Q
3 fraction and thus lower amounts of
free NBO in the glass network.
Fig. 19: Model of the Cu-containing 45S5 BG structure. Cu2+
(blue) replace Ca2+
ions (yellow)
as network modifiers.
Even though the main glass structure remains similar consisting mainly of Q2 and
Q3 units, the NMR results show higher Q
2 and Q
3 fractions for the 45S5-2.5Cu glass
which indicates higher silicate network distortion. The weakening of the glass
matrix is confirmed by the DSC measurement which revealed a decrease in glass
transition temperature for Cu containing 45S5 glasses. Incorporation of Cu into the
45S5 glass matrix was found to reduce the Tg of 45S5 glass and to stabilize the
amorphous phase during sintering.
The ionicity, iG, of the Me-O bonds within the glass structure can be used as
explanation of the decrease of Tg. With increasing Cu content in the glass the Tg
decreases likely due to the more covalent character of the Cu-O (iG=0.617) bond
compared to Ca-O (iG=0.707) bond resulting in higher degree of structure relaxation
and hence lower glass transition temperature.275
During heat treatment the 45S5 glass tends to crystallise and two crystalline phases,
combeite and silicorhenanite, form which have been described in literature for this
glass.276, 277
Results and Discussion 74
4.1.2 Scaffold properties
Macro-Structure
Scaffold macro and microstructure was analysed with SEM and µCT. The porosity
was derived by the Archimedes method.
Fig. 20a-d show the macrostructure of 45S5-Cu derived scaffolds. High pore
interconnectivity as well as pore sizes of ~ 200-300 µm were observed for all glass
compositions. Porosity values of 92 %, 91 %, 92 % and 93% and were observed for
45S5, 45S5-0.1Cu, 45S5-1Cu and 45S5-2.5Cu respectively. Fig. 20e) shows
additionally high magnifications of a 45S5 derived strut showing a hollow strut as it
is typical for polyurethane foam derived scaffolds. Furthermore, Fig. 20f shows the
3D reconstruction of a 45S5 derived scaffold from µCT data confirming the
interconnected pore system. The total porosity was calculated to be 93.5 ±2.0 % and
the strut thickness of ~67 µm was derived. The average pore size was 314 ± 87 m.
99.1±0.4 % of the total porosity was open indicating a completely interconnected
pore system. High porosity values and interconnected pore system of the scaffold
should enable vascularisation and tissue ingrowth when applied as engineered bone
construct. Vascularisation has been shown to be enhanced in scaffolds with pores >
250 um. 278
. However, high interconnectivity is considered even more important for
blood vessel and tissue ingrowth.43, 279
Hence, 45S5 derived scaffolds fulfil these
requirements.
Results and Discussion 75
Fig. 20: SEM images of 45S5-Cu derived scaffolds: a) 45S5, b) 45S%-0.1Cu, c) 45S5-1Cu and
d) 45S5-2.5Cu. e) typical hollow strut of 45S5 derived scaffolds and f) microCT reconstruction
of a 45S5 derived scaffold. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. - Adapted by
permission of The Royal Society of Chemistry.
Fig. 21 shows the XRD analysis of the 45S5-Cu derived scaffolds after sintering.
Two crystalline phases were observed after the heat treatment of the glass: a sodium
calcium silicate phase (combeite, Na2Ca2Si3O9) and silicorhenanite
(Na2Ca4(PO4)2SiO4) which have been shown to occur during high temperature
treatment of Bioglass®.67, 276, 280
Basically, no qualitative differences among the Cu-
containing and 45S5 reference were observed. All glasses show high crystallinity
after sintering. However, Rietveld analysis was performed in order to quantify the
crystalline phases. Since the sensitivity of the XRD technique is limited only 45S5
with the highest copper content (45S5-2.5Cu) was investigated in order to
qualitatively evaluate the role of copper during the crystallisation of 45S5 glass. The
Results and Discussion 76
amount of the amorphous phase was higher for 45S5-2.5 Cu compared to 45S5
reference. The content for the amorphous phase was calculated to 3.5 ±2.2 and 10.5
±2.8 % for 45S5 and 45S5-2.5Cu, respectively.
Fig. 21: XRD analysis of 45S5-Cu derived scaffolds showing two crystalline phases forming
during heat treatment. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. - Reproduced by
permission of The Royal Society of Chemistry.
Mechanical properties
For the 45S5-Cu glass series no differences in the compressive strength was
observed between the different glass compositions (data not shown) likely due to
the low amount of CuO presented in the glass. Fig. 22 summarises the mechanical
properties of the 45S5 derived scaffolds for multiple coatings regimes. Fig. 22a and
Fig. 22b shows show the mean values and typical stress-strain curves, respectively,
for 45S5 after multiple coatings with BG slurry were applied to the polyurethane
template. Maximal values of 0.16 ±0.07 MPa were observed for 45S5 scaffolds
after application of 3rd
coating with BG slurry.
Results and Discussion 77
Fig. 22: Mechanical properties of the 45S5 BG derived scaffolds: a) mean values calculated
from 10 measurements per scaffold; b) selected typical stress-displacement curves.
The enhancement of the compressive strength through multiple coating of the foam
is not significant and the overall mechanical strength of the scaffolds is rather poor.
The values observed here are in agreement with literature reports where
compressive strength values of 0.3-0.4 MPa (but with lower porosity compared to
scaffolds prepared in this work) were observed for 45S5 derived scaffolds by foam
replica technique. For clinical application, however, the mechanical properties of
the scaffolds should also be considered. The mechanical properties of the 45S5
derived scaffolds are discussed in the context of bone tissue engineering in 0.
Acellular bioactivity in SBF
Despite recent critical discussion in the literature regarding the suitability of SBF
studies for predicting the bioactivity,101, 102
this well-known and established
technique was employed in this work since the objective was to establish a close
comparison with non-doped 45S5 Bioglass® which have been largely investigated
by the SBF test as well as in vitro cell culture studies in the last decades.
The reaction stages of 45S5 bioactive glass derived scaffolds, as documented by
SEM, are summarised in Fig. 23. Fig. 23a shows the typical hollow macrostructure
of the initial 45S5 BG scaffolds strut. After 1d of immersion in SBF, Fig. 23b, the
surface is homogenously covered with calcium phosphate (CaP) precipitates as
indicated through their typical morphology. A higher magnification images of the
cross-section of a scaffold strut, Fig. 23c-d reveal a more detailed view on the
Results and Discussion 78
reaction stages. Fig. 23c shows a porous structure of roughly 8-10 µm which is
covered by a thinner dense layer (roughly estimated to be ~1-2 µm). These two
layer s are likely to be identified as a silica rich layer and CaP layer, respectively.
However, at higher magnification according to Fig. 23d no clear discrimination of
the silica and the CaP layer can be made. The CaP enrichment on the BG scaffold
surface is further indicated through small round-shaped particles marked as ACP
precipitates in Fig. 23 c). After 3 d in SBF, the phosphate layer continues to grow
reaching 3-5 µm after 3d (Fig. 23d). After 7 days only minor further growth of the
CaP layer was observed which remained roughly at 5 µm (Fig. 23e). Furthermore,
the inner region of the scaffold (inner part of the struts) shows signs of dissolution
turning into a porous silica structure with the modifier ions Ca and Na leached out.
After 7 days the entire scaffold strut seems to have transformed into pure silica
phase.
However, the analysis of the reaction layers on 45S5 bioactive glass scaffold and
identifying surface reaction layer with SEM is rather speculative. Hence, FT-IR
analysis and micro-PIXE-RBS analysis are presented in the following. Fig. 24
shows the FT-IR spectra of a 45S5 BG derived scaffold after immersion in SBF for
1d, 3d, and 7d. As-fabricated 45S5 scaffold shows bands at 629 cm-1
, 575 cm-1
and
530 cm-1
which can be attributed to apatite-like crystalline phase as the P-O bending
modes are located at 580 and 620 cm-1
.281
This crystalline phase corresponds to the
XRD results showing silicoarhenite (Na2Ca4(PO4)2SiO4) phase which is
isostructural to apatite.277
The bands at 450 and 930 cm-1
can be assigned to
symmetric Si-O-Si vibrations and to non-bridging oxygen Si-ONBO, respectively.267,
282 In contrast to the amorphous glass powder (Fig. 14) the broad silicate band at
~1020 cm-1
is split into two bands which are related to the vibrations of isolated Si
tetrahedral.67
Results and Discussion 79
Fig. 23: Reaction stages of HAp formation on crystallized 45S5-0.1Cu derived scaffold: a)
initial strut; b) scaffold after 1 day in SBF; c) and d) scaffold after 1 day in SBF at higher
magnification showing the formation of CaP-SixOy layer and amorphous calcium phosphate
(ACP) precipitates; e) CHA formation and f) further growth of carbonated hydroxyapatite
(CHA) on BG scaffold surface. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. -
Reproduced by permission of The Royal Society of Chemistry.
During treatment in SBF, the following changes in the FT-IR spectra occur:
- (after 1d): the P-O bands resulting from the crystalline, rhenanite-like phase
in the BG scaffold transform to a broad peak at 600 cm-1
which is typical for
amorphous calcium phosphate phase (ACP).283
The sharp peak of the Si-O-
Si band at 450 cm-1
is reduced in intensity and becomes broader indicating
the formation of a phosphate phase as O-P-O (bending mode) band absorbs
in this region.284
The shoulder at 960 cm-1
can be attributed to HPO42-
units
incorporated in the calcium phosphate phase.284
Furthermore, a broad peak
appears at 800 cm-1
which corresponds to silica.91
Results and Discussion 80
- (after 3d): with increasing immersion time in SBFa a sharp double band was
detected at 560 cm-1
and 600 cm-1
which is assigned to the vibration modes
of the PO43-
groups284
typically observed in crystalline HAp
- (after 7d): the HAp bands increase in intensity and, additionally, CO32-
band
at 870 cm-1
appears which are observed for natural carbonated
hydroxyapatite (CHA)284
but, however, could also be attributed to surface
bonded carbonate compounds.91
Also, double peaks at 1430 and 1500 cm-1
are observed which can be assigned to v3
(CO32-
) of carbonate ions
incorporated in CHA.284
Fig. 24: Evolution of FT-IR spectra of 45S5 derived scaffold during immersion in SBF for 1d,
3d and 7d. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. - Reproduced by permission of
The Royal Society of Chemistry.
FTIR-analysis did not reveal any effect of Cu on the CHA formation. Fig. 25 shows
the FTIR-spectra for 45S5-Cu derived scaffolds after 3d. Clear formation of CHA
was detected after 3 days of immersion in SBF independently of Cu content in the
glasses.
Considering the time point of appearance of the CHA layer as marker for bioactivity
one can conclude that the bioactivity of the 45S5 derived scaffolds is not impeded
Results and Discussion 81
by the incorporation of Cu. In order to gain more detailed information about the
nature of the CHA layer formed on Cu.
Ion beam method allows a chemical analysis with an excellent sensitivity of several
ppm due to very good signal to background ratio. Compared to other techniques like
SEM/EDS, micro-PIXE-RBS method allows an improvement of the sensitivity up
to 3 orders of magnitude. This is a great advantage to study the distribution and the
role of relevant bone trace elements. In addition to other conventional methods
PIXE-RBS analysis enables to detect the local chemical composition of the reaction
surface layer.
Fig. 25: FT-IR spectra of Cu-containing scaffolds after 3d of immersion in SBF compared to
reference 45S5 glass. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. - Reproduced by
permission of The Royal Society of Chemistry.
Chemical maps were acquired for the distribution of Si, Ca, P, (Na is not shown for
simplicity) and Cu in the fabricated Cu-containing scaffolds. 45S5-0.1Cu serves as
representative example for the element distributions in the inner part of the scaffold
and the periphery layer and for the elemental evolution during immersion in SBF.
The glass composition with the lowest Cu amount was also chosen in order to
confirm high sensitivity of the micro ion beam technique enabling monitoring the
evolution of low concentrations of trace elements during the reaction of a bioactive
Results and Discussion 82
glass scaffold. Fig. 26 shows typical multichemical elemental maps for Si, Ca, P,
and Cu for an as-fabricated 45S5-0.1Cu cross section (Na is not shown for
simplicity). All elements are homogenously distributed.
Fig. 26: Element distribution of a 45S5-0.1Cu scaffold cross-section in the as-fabricated stage
derived from micro-PIXE-RBS measurements. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p.
5659. - Reproduced by permission of The Royal Society of Chemistry.
Fig. 27-Fig. 29 are representative elemental maps of the scaffolds after immersion
in SBF for different time periods. In correlation to SEM/FT-IR results, during
immersion in SBF the following different regions can be distinguished from the
chemical maps:
i) primary bioactive glass network
ii) silica rich layer on the surface of the scaffolds (clearly distinguishable after 1d)
iii) Ca-P rich layer in the scaffold periphery
For the areas i) and iii) additionally the elemental concentrations were quantified in
these two distinct regions of interests. The elemental evolution for all 45S5-Cu
glass derived scaffolds as function of immersion time in SBF is shown in Fig. 30
and Fig. 31 for the inner region (i) and the periphery region (iii), respectively.
Results and Discussion 83
Fig. 27: Element distribution of a 45S5-0.1Cu scaffold after 1 day in SBF monitoring the
formation of a calcium phosphate layer. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. -
Reproduced by permission of The Royal Society of Chemistry.
Fig. 28: Element distribution of a 45S5-0.1Cu scaffolds after 3 days in SBF monitoring the
formation of a calcium phosphate layer. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. -
Reproduced by permission of The Royal Society of Chemistry.
Results and Discussion 84
After 1 day, a silica rich layer is detected at the scaffold surface and Ca and P
enrichment is observed as shown in Fig. 27. The silicon content increased from 21
wt% in the inner part of the scaffold to 33 wt% in the silicon enriched layer (12.0 ±
0.2 µm). The Ca/P ratio for the (iii) periphery layer was determined as 1.3 ± 0.1 µm.
The P layer is rather distinct whereas Ca traces can be also found in the Si-rich
layer. This corresponds to the SEM observation where no clearly distinguishable
CaP but rather a mixed Ca-P-SiO2 layer was detected.
After 3d, Fig. 28, a CaP layer was observed on the pore surface with a Ca/P ratio of
1.88 ± 0.06, Fig. 31. The layer thickness was determined to 10.5 ±0.9 µm. The inner
parts of the BG scaffolds showed further dissolution and depletion in Ca and P. At
the same time the relative amount of silicon present within the glass matrix
increases since Ca, P and Na are leached out. After further immersion in SBF for 7d
no significant changes occured: the Ca-P rich layer thickness remains nearly
constant (10.9 ±1.6 µm), Fig. 29, while the Ca/P ratio slightly increases to 1.9 ±0.1.
Fig. 29: Element distribution of a 45S5-0.1 scaffold after 7 days in SBF showing further growth
the calcium phosphate layer. A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. - Reproduced
by permission of The Royal Society of Chemistry.
Results and Discussion 85
Elemental concentrations for all 45S5-Cu derived scaffolds were calculated in two
areas, the inner area (i) of a strut and the outer layer (surface periphery (iii)).
Hereby, the outer layer was defined as the area of interest once the Ca and P rich
peripheral layer has formed. The evolution of the element concentration during
immersion in SBF is shown in Fig. 30 and Fig. 31 for the inner part and the
peripheral layer, respectively. Si release from the periphery was slowed down for
the 45S5 and 45S5-0.1Cu scaffolds as compared to 45S5-1Cu and 45S5-2.5Cu
indicating slower surface degradation kinetics of these glasses. Accordingly, Ca and
P enrichment is also slower for 45S5 and 45S5-0.1Cu. With higher Cu contents in
the glass (≥ 1wt.-%) Ca and P seem to diffuse faster through the silica layer and are
faster accumulated in the scaffold periphery.
Fig. 30: Evolution if elements in the inner region of the 45S5-Cu derived scaffolds as function
of immersion time in SBF. Lines are for eyes guidance only. A. Hoppe et al. J. Mater. Chem. B
1 (2013), p. 5659. - Reproduced by permission of The Royal Society of Chemistry.
Results and Discussion 86
However, the evolution of Ca and P in the inner region of the scaffolds, Fig. 30, is
not significantly different among the glasses indicating that the global dissolution of
the glass is independent of the Cu content in the glass.
The kinetics for Na release from the scaffolds was found to be similar for all 45S5-
Cu derived scaffolds. However, for higher Cu contents (45S5-1Cu and 45S5-2.5Cu)
small amounts of Na can be detected even after 7d of immersion in SBF (in the
inner part of the scaffold), whereby no Na is detectable for 45S5 reference and
45S5-0.1Cu.
Fig. 31: Evolution if elements in the periphery layer of the 45S5-Cu derived scaffolds as
function of immersion time in SBF. Lines are for eyes guidance only. A. Hoppe et al. J. Mater.
Chem. B 1 (2013), p. 5659. - Reproduced by permission of The Royal Society of Chemistry.
Results and Discussion 87
Altogether the mechanism for apatite formation on 45S5-Cu derived scaffolds can
be described as follows:
Basically, the present observations confirm the reaction mechanisms of
(amorphous) bioactive 45S5 glass as originally proposed by Hench et al.9 and
Kokubo.285
For Cu containing 45S5 glasses, the relevant physico-chemical reactions
were shown exemplary for the 45S5-0.1Cu composition. However, since no
significant differences in the reaction within the Cu-glass series were detected by
FT-IR (Fig. 25) it is possible to assume that the proposed model is valid also for the
other Cu-containing glass composition.
(1d-3 d): Initial reaction of the BG surface and release of Na+ and Ca
2+ from the
scaffold periphery in an exchange reacting with H+ from the solution according to
Eq. 5.13, 286
. Free Si-NBO are protonated forming silanols groups.
𝑆𝑖 − 𝑂−𝑁𝑎+ + 𝐻2𝑂 → 𝑆𝑖 − 𝑂𝐻 + 𝑂𝐻− + 𝑁𝑎+ Eq. 5
Soluble Si species are released upon breakage of Si-O bonds by hydrolysis in the
surface region of the glass:13, 287
𝑆𝑖 − 𝑂 − 𝑆𝑖 + 𝐻2𝑂 → 𝑆𝑖 − 𝑂𝐻 + 𝐻𝑂 − 𝑆𝑖 Eq. 6
The formation of silanols is triggered by both reactions Eq.5 and Eq.6.288
A SiO2 layer is formed upon condensation of SiOH groups on the BG surface:
2𝑆𝑖𝑂𝐻𝑐𝑜𝑛𝑑.→ 𝑆𝑖 − 𝑂 − 𝑆𝑖 + 𝐻2𝑂 Eq. 7
From SEM images and from micro-PIXE-RBS analysis the thickness of the SiO2
layer was estimated to approx. 8-10 µm. The combination of the two methods
enables a reasonable estimation. The formation of silica-rich region (ii) was indicted
by FT-IR measurements, Fig. 24, and also confirmed in the elemental maps which
revealed increase of Si content. At the same time a thin ACP layer (iii) is formed (2-
3 µm) through migration of Ca and P ions to the BG surface.13
Indeed, the inner
side of BG scaffold (i) is depleted in P, as shown by ion beam measurements.
Results and Discussion 88
Interestingly, the inner part of the scaffold is not depleted in Ca. The scaffold
peripheral layer (ii), however, shows depletion in Ca content with a slight increase
in Ca concentration in the region (iii) on top of the Si-enrichment, Fig. 27. From the
ion beam measurements a Ca/P ratio of 1.28 was calculated for this specific region
of interest (iii) which corresponds to typical values of ACP.54
However, as shown in
Fig. 23c-d the silica layer and the ACP layer cannot be clearly distinguished from
the SEM figures. Also ion beam analysis shown in Fig. 27 reveals that Ca is also
found in the (ii) region which is described previously as silica-rich layer. Hence, it
can be concluded, that instead of a distinct ACP layer on top of the silica gel as
proposed by Hench et al.13
more likely a mixed layer of CaP+SixOy is formed on the
scaffolds surface after immersion for 1d in biological fluid as described by Aguiar et
al.289
for bioactive glasses.
(3d-7d) of immersion in SBF, a CHA layer was formed as confirmed by FT-IR
measurements, Fig. 25. Also from micro-PIXE-RBS analysis a Ca/P ratio of 1.85
(after 3d) and 1.77 after 7d was determined which are closer to stoichiometric
apatite (Ca/P~ 1.67) confirming the crystallisation of ACP to CHA. Furthermore,
traces of Cu are incorporated in the CHA layer.
Altogether, the combination of FT-IR, SEM and micro-PIXE-RBS results give an
comprehensive picture of physico-chemical reaction on a 45S5-Cu derived scaffolds
during immersion in SBF. The analysis performed confirmed enhanced apatite
forming ability of the 45S5 derived scaffolds after 3d of immersion in SBF. Despite
some reports in literature on the potential negative impact of crystallinity on the in
vitro bioactivity of BG derived (sintered) scaffolds,147
in this work high bioactivity
was observed which is comparable to those of amorphous bioactive glass products
widely reported in literature.30, 35
Similar rapid CHA formation was also found on
micron sized 45S5 bioactive glass particles after 3 d in SBF has been shown.94
Despite some critical considerations related to the “bioactivity test” by SBF
immersion102
this method can still be considered a relevant technique for assessing
the reactivity of BGs in physiological environment. Particularly, the “SBF test”
serves as an internal control for new developed Cu-containing bioactive glasses in
order to compare the reactivity to the undoped 45S5 bioactive glass as reference.
Results and Discussion 89
Moreover, in this work for the first time the reaction stages of a highly crystalline
45S5 derived (glass-ceramic) scaffold in simulated body fluid were observed in
high detail using the micro-PIXE-RBS technique. On top of FT-IR measurements
and SEM observations revealing the reaction on the BG scaffold surface the micro
ion beam method enabled detailed evaluation of elements distributed and evolution
in the 45S5 scaffold. Moreover, the chemical composition of the CHA layer was
derived. Detailed insight into the reaction scheme of 45S5 derived scaffolds was
gained. On top of the conventional methods thickness layer and details of the
composition of the reaction layers formed on the BG scaffold surface was obtained.
In particular for metal ion containing bioactive glass derived scaffolds the
distribution of the dopant element in the BG matrix is of great interest which cannot
be assed with conventional spectroscopic methods.
Additionally, it was confirmed that Cu inclusion in the 45S5 BG does not have any
negative impact on the CHA forming ability of the 45S5 bioactive glasses was
observed. Formation of CHA was observed after 3 days of immersion in SBF for all
glasses investigated,Fig. 25. It was shown that a homogenous CHA layer was
formed covering the BG scaffolds surface completely after 3d in SBF.
These observations are important for better understanding of the physico-chemical
reactions occurring on 45S5 Bioglass® derived scaffolds in physiological fluids.
Since chemistry of the materials surface strongly affects cell adhesion and cell
changes of the surface chemistry when testing BG scaffolds in vitro. CaPs for
instance are known to influence osteoblast cell attachment and differentiation.
Hence, the transformation of the silicate bioactive glass to a CaP phase is an
important parameter dictating the biological performance of BG derived scaffolds.
Interestingly, the chemical maps derived from micro-PIXE-RBS analyses revealed
that traces of Cu are also incorporated into the HAp layer formed on the BG
scaffolds surface. Thus, this effect should also be considered as parameter affecting
cell response beside the ionic dissolution products from the BG scaffold. Also, the
kinetics of the transformation of BG to CaP should determine the bone ingrowth
kinetics of BG scaffolds in vivo, which has been shown for ingrowth of 13-93 based
scaffolds.71
Results and Discussion 90
Degradation of 45S5 derived scaffolds and Cu release
Two different conditions were used in order to assess the degradation behaviour of
the BG derived scaffolds as it has been shown that degradation behaviour of
bioactive glasses and glass derived scaffolds depend on immersion condition
whether static or dynamic.291
Fig. 32 shows the Cu and Si release from the BG
derived scaffolds when immersed in SBF for a period up to 21d under static
condition. The error bars indicate the standard deviation derived from 3 samples
measured. Depending on the glass composition Cu ions in the range from 0.04 up to
3.4 ppm were released. For all glasses, a rapid increase in Cu release was observed
in the initial stage of degradation reaching a plateau after 7d with ~0.04 ppm, ~0.4
ppm and ~3.4 ppm Cu released from 45S5-0.1Cu, 45S5-1Cu and 45S5-2.5Cu
scaffolds, respectively.
Similar trend was observed for Si released from the 45S5-Cu derived scaffolds
under static conditions. Burst release of Si was detected in the first 7d of reaction in
SBF reaching a plateau at Si levels of ~30-35 ppm.
Depending on the glass composition Cu concentrations in the range from 0.03 up to
3.5 ppm were released in SBF under static conditions. These values were reached
after 7d immersion in SBF and remained constant during further reaction in SBF.
Also, the Si values reach a plateau after 7d immersion in SBF which indicates a
decrease in degradation of the scaffolds due to saturation of the solution. Being a
silicate based glass the release of silicon can be used as marker for degradation of
45S5 derived scaffolds. Since no further increase in Si concentration in SBF was
observed after 7d it can be concluded that the degradation of the scaffolds is
inhibited due to saturation of the solution. This behaviour corresponds to literature
reports which have shown that amorphous BG scaffolds (mol%: 70SiO2-30CaO)
reach a degradation stop after 3d in SBF under static conditions due to saturation of
the solution in Si.291
Variations in Si release from different glasses are detected for
1d and 3d of immersion in SBF, while no significant difference was observed for an
immersion time after longer than 7d. However, the effect of Cu is inconsistent for
this data: 45S5-2.5Cu and reference 45S5 derived scaffolds show highest Si values
for initial stage of immersion (1-3d). Also, micro-PIXE-RBS derived data in Fig. 31
show that 45S5-1Cu and 45S5-2.5Cu exhibit faster Si release from the periphery of
Results and Discussion 91
the scaffold indicating enhanced surface reactivity for BG with higher Cu contents
(≥ 1wt%). Altogether it seems that Cu-addition to the 45S5 matrix enhances the
initial surface reactivity of the scaffold while after longer reaction time no effect of
Cu is visible.
Fig. 32: Cu and Si release from 45S5-Cu derived scaffolds in SBF under static conditions.
Fig. 33 shows the Cu and Si release under quasi-dynamic conditions. For all
glasses, the overall released (cumulative) Cu concentrations were found to be higher
compared to the release rates under static conditions.
For 45S5-0.1Cu, 45S5-1Cu and 45S5-2.5Cu glasses, highest Cu levels of ~0.6 ppm,
~2.8 ppm and ~4.6 ppm, respectively, were detected. This was expected as under
quasi-dynamic conditions the concentration gradient and, thus, the ion diffusion is
enhanced. Cu concentrations reached a plateau after 14 days of immersions in SBF.
Accordingly, Si concentrations in the SBF were continuously increasing until 14
days of immersion reaching a plateau suggesting that the overall degradation of the
scaffolds is decelerated. In the same way, Si concentration increased during the first
days of immersion in SBF reaching a saturation level after 7d whereas Si values in
the range of 30-40 ppm depending on the glass composition were obtained.
Considering the standard deviation within the triplicate sample no major differences
in Si release is observed for different glass composition after 7 days of immersion in
Results and Discussion 92
SBF as the mean Si values is in the range between 30 ppm and 50 ppm. Solely, in
the initial state of reaction higher values of Si were observed for 45S5 and 45S5-
2.5Cu scaffolds after 1 and 3 days in SBF. Among the Cu-doped glasses the final Si
levels decreased released decreased with higher Cu-content. This might be a result
of higher network connectivity of the silica network in the glass caused by Cu
incorporation which inhibits the release of soluble silica.
Under quasi-dynamic conditions higher absolute Cu as well Si levels were detected
for all glasses indicating overall enhanced degradation of the scaffolds under quasi-
dynamic conditions compared to static setup. These observations should be
considered for in vitro experiment of bioactive glasses.
In literature reports silicate based glass derived scaffolds were shown to follow an
linear degradation profile when cultured in SBF up to 7 days under quasi-dynamic
conditions similar to the parameters used in this work.291
However, after 14 days the
degradation slowed down and Si as well as Cu values reach constant constant
levels. Furthermore, with increasing Cu-content in the glass lower Si values are
released from the scaffolds. Considering the NMR results this is likely due to higher
Q2 and Q
3 fractions in the glass network resulting in higher SiO2 network stability
and less amount of soluble silica being released.
Fig. 33: Cu and Si release from 45S5-Cu derived scaffolds in SBF under quasi-static conditions.
A. Hoppe et al. J. Mater. Chem. B 1 (2013), p. 5659. - Reproduced by permission of The Royal
Society of Chemistry.
Results and Discussion 93
After 21 days under both, static and dynamic conditions no significant mass gain
was observed for all 45S5-Cu derived scaffolds. Despite high release rates of
soluble silica the material loss is seemingly compensated with the mass gain
through the formation of hydroxyapatite on the 45S5 scaffolds surface.
Cu ions exhibit dose-dependent effects on human cells. For example, 50 ppm of
CuSO4 (313 µM) were shown to be optimal for stimulation of tube-like structures
formed by ECs when exposed to a Cu concentration range from 0 ppm to 100
ppm.197
Similarly, Wu et al. showed that Cu levels from 60.4 to 152.7 ppm favor
angiogenesis and osteogenesis via expression of osteogenic markers (ALP, OPN,
OCN) and secretion and expression of the angiogenic marker VEGF in human bone
marrow derived stem cells (hBMSCs).226
However, also lower Cu concentrations of
4 ppm released from phosphate glasses (similar to the Cu release ranges observed in
our study) were favorable for HUVEC cells via down regulating apoptosis.260
Furthermore, related values of 1.6 ppm to 8 ppm of CuSO4 were reported to
significantly increase the VEGF expression in keratinocytes.292
In this work Cu
levels from 0.3 ppm to 4.6 ppm were released in SBF from Cu-containing 45S5
bioactive glass derived scaffolds depending on the culturing conditions, hence it can
be concluded that 45S5-Cu bioactive glass might be a potential material for bone
tissue engineering applications where both osteogenic and angiogenic properties are
required.
Results and Discussion 94
4.1.4 In vitro cell response
Theoretically the included Cu into the 45S5 Bioglass could affect cell behaviour in
two ways: Firstly, Cu ions released from the 45S5 Cu derived scaffold might
enhance vascularisation as Cu2+
ions are well-known angiogenic agents. As
indicated above, the Cu levels released into physiological environment are within
the therapeutic range according to literature reports for Cu2+
ions.293
Since ionic
dissolution products have been shown to stimulate cells on molecular level towards
osteogenic differentiation18
the Cu doping is expected to impart additional
angiogenic functionalities and enhance the overall biological activity of the 45S5
BG derived scaffolds. Secondly, Cu was incorporated in the CHA layer formed on
the scaffold surface after immersion in SBF, which might influence the attachment
and growth of relevant cell types used in regenerative medicine.290
Metal ion doped
hydroxyapatite, for instance, has been shown to influence osteoblast adhesion and
differentiation.294
The assessment of such biological effects of the incorporation of
Cu in the basic 45S5-Bioglass® material is described in this section.
Powder cytotoxicity
Cu in high dosage might be toxic to human cells and organism and therefore
cytotoxicity of Cu-containing glasses was investigated. Commercial use of
bioactive glass products also includes applications of particulate glass, such as
powder or granula (BonAlive® or NOVABONE
®). Hence in the first place the
cytotoxicity of the particulate 45S5-Cu powder was assessed. Fig. 34 shows light
microscopy evaluation of MG-63 cell morphology after 48 h of incubation with
45S5 bioactive glass particles at different concentrations. The cells show a slightly
elongated triangle morphology which is typically observed for healthy MG-63 cells.
Even at BG particle concentrations of 1 mg ml-1
no negative effects on cell growth
were observed. It seems that the particles sediment during the culturing time and the
cell grow on the top of the particles.
For comparison of the Cu containing glasses the cell morphology of MG-63 cells
cultured with 45S5-Cu containing samples at 100 and 1000 µg ml-1
is shown in Fig.
35. Similarly, no toxic effects are seen for the Cu samples for particle
Results and Discussion 95
concentrations up to 100 µg ml-1
. For 1000 µg ml-1
, however, less cells were visible
indicating possible negative effect of the BG particles at 1000 µg ml-1
. Particularly,
for the 45S5-2.5Cu lees cells are visible compared to 45S5-Cu glasses with lower
Cu contents.
.
Fig. 34: Light microscope of MG-63 after 48h of incubation in drect ccontact with 45S5 BG
particles at different concentrations.
Fig. 35: Light microscope images of MG-63 after 48h of incubation in direct contact with 45S5-
Cu BG particles at 100 µg/ml and 1000 µg/ml.
In order to gain more detailed analysis of the cell viability the mitochondrial
activity and the cell number were derived. Fig. 36 shows the mitochondrial activity
and the cell number indicated by LDH activity of osteoblast-like cells after
Results and Discussion 96
cultivation time of 48 h as function of BG concentration in the cell culture medium.
All Cu containing 45S5 BGs show high cell mitochondrial activity of > 50% of the
reference control for the entire concentration range investigated. Also the cell
number remains nearly constant for all particle concentrations tested. This indicated
that 45S5-Cu particles do not show any toxic effects on the cells during the first 48
h of incubation. Moreover, statistically significant enhancement of the
mitochondrial activity of cells in contact with 45S5-0.1Cu and 45S5-1Cu for
particles concentration between 0.1 and 100 µg ml-1
were observed compared to
plain 45S5 reference. At 1000 µg ml-1
for the Cu containing glasses a reduction of
the mitochondrial activity to 70 % (for 45S5-0.1Cu and 45S5-1Cu) and ~50 % (for
45S5-2.5Cu) was observed while remaining at ~100 % for the reference 45S5.
Hence, even though the cells remain at high viability higher than 50 % this indicates
possible cytotoxic effects of the 45S5-Cu particles at too high concentrations.
However, this effect could be also assigned to the alkaline pH shift due BG
dissolution when applied at a critical concentration ≥1000 µg ml-1
.
These observations confirm good biocompatibility of Cu-containing 45S5 glass
powders and beyond that stimulating effect on MG-63 cells when applied at
concentration range of 0.1-100 µg ml-1
. In particular, 45S5-0.1Cu and 45S5-1Cu
glass particles enhanced the mitochondrial activity of MG-63 cells compared to the
undoped 45S5 reference.
Fig. 36: Mitochondrial activity and cell number (LDH activity) as function of particle
concentration for 45S5-Cu glass series. *p<0.5 and **p<0.001 compared to 0 µg/ml reference.
Results and Discussion 97
It has been shown in literature that Cu can stimulate the cell activity and
proliferation of osteoblastic cells.258
Accordingly, 45S5-0.1Cu and 45S51Cu glass
samples showed enhanced MG-63 cell activity compared to 45S5 reference.
In literature it has reported that hat BG particles of ~100 µm did not have any
significant effect on osteoblastic cell proliferation and metabolic activity.295
On the
other hand nano-sized bioactive glass particles can be cytotoxic as shown by means
of reduced cell activity of mesenchymal stem cells when cultured with BG particles
of ~70 nm at a concentration of 0-200 µg/ml (similar range as tested in this
work).296
The BG glass particles used in this study are ~ 6 µm and were shown to
stimulate the activity of osteoblast like cells indicating good biocompatibility of
micron-sized BG particles.
These findings are important for designing of new in vitro studies incorporating BG
particles. Further, these fundamental studies confirm the biocompatibility of the
45S5-Cu glasses which is a first step in the evaluation of biomaterials for use in
clinical applications. Indeed, commercial use bioactive glasses also involve the
application of particulate bioactive glass (Novabone, BoneAlive).
Cell attachment on 2D pellets
Fig. 37 shows the attachment and growth of MG-63 cells on 45S5-Cu derived
pellets after culturing for 48 hours. A dense cell layer was observed on the 45S5,
45S5-0.1Cu and 45S5-1Cu samples. The cells are widely spread indicating high
compatibility of the samples surface towards MG-63 cells. On 45S5-0.1Cu and
45S5-1Cu even multilayer growth of the cells was observed. However, on the 45S5-
2.5Cu samples no cells were detected indicating possible cytotoxicity of the high
Cu concentration in the 45S5-2.5Cu samples. These results are in agreement with
the cytotoxicity tests carried out on powdered 45S5-Cu glasses which indicated
possible cytotoxic effects of 45S5-2.5Cu glass.
Results and Discussion 98
Fig. 37: Fluorescence microscopy of cytoskeleton (red) and cell nucleus (green) staining of
osteoblast-like cells seeded on 45S5-Cu derived pellets for 48h. No cells were observed on 45S5-
2.5Cu.
One well accepted explanation for copper-induced cytotoxicity is related to the
formation of reactive oxygen species (ROS) by Cu ions the via Fenton reaction.297
with the consequence of peroxidative damage of membrane lipids.298
In order to test
this hypothesis a western dot-blot analysis was performed with HOS cells seeded on
45S5-Cu pellets and the formation of 4-hydroxynonenal (HNE), one of the best
known and well-studied products of lipid peroxidation.10
A detailed description of
the experiment is given elsewhere.299
Fig. 38 shows the HNE formation for 2D pellets of Cu-containing 45S5 glasses. For
pure 45S5 (reference material) the cell growth was associated with low HNE
formation which slightly increased from after 7d and 14d of cell culture. With
10
The immuno-blot analysis of the HNE formation was carried out in collaboration with L.
Milkovic and Prof. N. Zarkovic, Laboratory for Oxidative Stress, Rudjer Boskovic Institute, Bijenicka 54, 10000 Zagreb, Croatia. These results are also part of the PhD thesis of L. Milkovic entitled “Beneficial effects of lipid peroxidation in the bone cell growth on bioactive glass - New perspectives in tissue engineering and regenerative medicine”
Results and Discussion 99
addition of Cu higher values of HNE were detected whereby the strongest
enhancement was observed for 45S5-1Cu and 45S5-2.5Cu samples.
HNE is an important signaling molecule involved in various cellular processes
including cell proliferation and differentiation.300, 301
However, HNE acts in
concentration-depended manner and hence high HNE levels can also correspond to
cell death.302
This dese-dependent role is in accordance to the results shown in this
study: while for 45S5, 45S5-0.1Cu and 45S5-1Cu the enhanced HNE formation and
(hence peroxidation) are correlated to cell growth, for 45S5-2.5Cu the enhanced
HNE formation is related to cytotoxicity.
From these results it can be concluded lipid peroxidation is involved in the
interaction between 45S5 BG and osteoblast-like cells and that the effect of copper
is dose-depending and is correlated to the formation of lipid peroxidation products.
These correlations should be further explored in future studies in order to get more
information on the mechanism of the interaction of BG and human cells.
Fig. 38: Immuno-blot analyses of the HNE-protein adducts formation in osteoblast-like cells
(HOS) cells after 3, 7 and 14 days. The results are expressed as nmol of HNE-protein
adducts/mg of protein. Enhancement of the lipid peroxidation was in particular pronounced
after 1d (*** p<0.0001) and 14d (** p<0.01) for the cells grown on 45S5-2.5Cu samples.*p<0.05
Results and Discussion 100
In vitro cell studies with 3D scaffolds
Response of osteoblast-like cells (MG-63)
From the powder cytotoxicity and the 2D cell attachment studies it was concluded
that 45S5-2.5Cu BG seems to be cytotoxic and, hence, this composition was not
further considered for biological investigations. For the 3D studies with MG-63
cells only 45S5, 45S5-0.1Cu and 45S5-1Cu were tested.
Good cell attachment and growth of osteoblast-like cells was observed with direct
seeding of the cell on the 3D scaffolds. Fig. 39 shows the attachment and growth of
osteoblast-like cells on 45S5-Cu derived scaffolds. Similar to 2D experiments no
signs of cytotoxicity of Cu was observed. The cell can attach and proliferate on the
scaffolds surface. MG-63 are an established cell culture model for assessing
osteoblast-like behaviour of cell in the context of bone tissue engineering.303
In
particular the attachment of human osteoblast cells can be well monitored by using
the MG-63 cell line as they show similar integrin profile as human osteoblasts. Also
it is a widely used cell model to test therapeutic agents and cytocompatibility testing
of materials.303
Hence, it can be concluded that 45S5-Cu derived scaffolds are
suitable for attachment and growth of osteoblast-like cells. Combined with the
results from the studies with powders and dense pellets in general a good
biocompatibility of 45S5-Cu scaffolds was observed. Hence, Cu levels up to 4 ppm
as shown in the degradation studies are not toxic to osteoblast-like cells up to a
period of 21 days. Also independently of the BG morphology applied as particulate,
dense pellets or porous scaffolds 45S5-Cu BG is biocompatible in vitro.
Results and Discussion 101
Fig. 39: SEM images of osteoblast-like cells cultured for 21d on 3D scaffolds at different
Citizenship: German Date/place of birth: 1st July 1984 in Kustanaj/Kasachstan
Education
November 2009
(on-going)
PhD research (Institute of Biomaterials, University of Erlangen-Nuremberg) Topic: Bioactive glass based scaffolds with therapeutic ion release for bone tissue engineering Supervisor: Prof. Aldo R. Boccaccini
2004-2009
Study of Materials Science and Engineering (University of Erlangen-Nuremberg) completed with diploma degree (grade 1.3=”very good”) Thesis: Metal Ion doped biomimetic hydroxyapatite coatings Supervisor: Prof. P. Greil
2001-2004 Secondary education leading to Abitur (A levels); grade: 1.2 (=very good)
Main Research Activities / Expertise
Fabrication and characterisation of biomaterials
Synthesis and structural characterisation of melt-derived bioactive glasses and fabrication of 3D glass derived scaffolds
Compositional design of (bioactive) glasses
In vitro bioactivity and biomineralisation studies in biological fluids Materials surface characterisation (roughness, topography, wettability)
Biomimetic hydroxyapatite coatings; surface functionalisation Development of bioactive glass/polymer composites Characterisation of nano-scaled bioactive glass particles Degradation and ion release assessment of degradable (bio)materials
vessel-like tube formation, expression of angiogenic marker
Characterisation techniques
Materials: X-ray diffraction (XRD), Fourier-Transform Infrared and Raman Spectroscopy (FT-IR), Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES), Scanning electron microscopy (SEM)+Energy Dispersive Spectroscopy (EDS), X-ray Photoelectron spectroscopy (XPS), Nuclear magnetic resonance spectroscopy (NMR) Cell biology: Gene expression (RT-PCR), Laser Confocal Fluorescence Microscopy (LCFM), SEM, Protein assay (Bradford), Viability (WST-8, AlamarBlue), ELISA assay (VEGF), DNA quantification (PicoGreen, BrdU assay)
International Experience / Fellowships / Grants
Mar 2014 Short Term Scientific Mission (STSM) within a COST (European Cooperation in Science and Technology) action (Project NAMABIO) at 3B's Research Group, University of Minho, Braga, Portugal Supervisor: Prof. Rui Reis
Nov-Dec 2012 and Nov-Dec 2013
Research stay at the University of Buenos Aires, Argentina in the framework of the project: Design and development of novel matrices for tissue engineering and/or drug delivery for pathologies of high social and economic impact Supervisor: Prof. Viviana Murino
2012-2014 (on-going)
Project Assistant for the bilateral research exchange programme with Rudjer Boskovic Institute, Zagreb Coratia, Prof. Neven Zarkovic Research topic: Effect of novel Cu-doped 45S5 bioactive glass and lipid peroxidation products on bone regeneration
Mar 2012 Travel grant to the First São Carlos School of Advanced Studies in Materials Science and Engineering (SanCAS–MSE), São Carlos, Brazil
Aug-Nov 2011
Research stay at McGill University, Montreal, Canada in the framework of the Quebec-Bavarian project Mesenchymal stem cell seeded nano-composite constructs for bone tissue engineering Supervisor: Prof. Showan N. Nazhat
Oct-Nov 2010 KMM-VIN Research Fellowship at Polytechnic University of Turin Topic: Glass-ceramic scaffolds with antibacterial capability for bone tissue engineering Supervisor: Dr. Enrica Verne
Tutorial lecture on Materials for Biomedical Engineering within the Masters programme Life Science Engineering
Teaching assistant in undergraduate laboratory courses on physical optics
Supervising activities (selected)
Florian Ruther, Bachelor thesis, 2013: Metal ion containing bioactive glass derived 3D scaffolds with enhanced mechanical strength
Stefan Grimm, Master thesis 2011: Fabrication and characterization of 3D highly porous scaffolds with enhanced biocompatibility based on ion doped glasses
Vincent Bürger, Master thesis 2010: Effects of residual surface stresses on the bioactivity of bioactive silicate glasses (Winner of the 2011 Oldfield Award)
Alexander Kent, Bachelor thesis 2010: Fabrication and characterisation of highly porous three-dimensional bioactive glass-based scaffolds for bone tissue engineering
Language and other skills
Language German (native speaker), Russian (native speaker), English (fluent), Spanish (intermediate)
Memberships Member of the American Ceramic Society (ACerS), Member of the German Biomaterials Society (DGBM)
Reviewer activity Reviewer in international peer to peer journals: Materials Letters, Acta Biomaterialia, Surface and Coating Technology, Journal of American Ceramic Society, Colloids and Surfaces B: Biointerfaces, NANO
Other interests I am interested in playing guitar. I also enjoy various kinds of sports including squash, basketball, running, racing bicycle and trekking.
Publications
Publications in peer-reviewed journals Hoppe A, Boccaccini A. R. On the degradation of bioactive glass scaffolds. 2014 (in preparation) Hoppe A, Brandl A, Bleiziffer O, Jokic B, Janackovic D, Boccaccini A R. In vitro cell response to Co-containing 1393 bioactive glass. 2014 (submitted). Hoppe A, Jokic B, Janackovic D, Fey T, Greil P, Romeis S, et al. Cobalt-Releasing 1393 Bioactive Glass-Derived Scaffolds for Bone Tissue Engineering Applications. ACS Appl Mater Interfaces. 2014;6(4):2865-77. Hoppe A, Sarker B, Detsch R, Hild N, Mohn D, Stark WJ, et al. In vitro reactivity of Sr-containing bioactive glass (type 1393) nanoparticles. J Non-Cryst Solids. 2014;387: 41-6. Hoppe A, Will J, Detsch R, Boccaccini AR, Greil P. Formation and in vitro biocompatibility of biomimetic hydroxyapatite coatings on chemically treated carbon substrates. J Biomed Mater Res, Part A. 2014;102: 193-203. Hoppe A, Meszaros R, Stähli C, Romeis S, Schmidt J, Peukert W, et al. In vitro reactivity of Cu doped 45S5 Bioglass® derived scaffolds for bone tissue engineering. J Mater Chem B. 2013;1: 5659-74. Hoppe A, Mourino V, Boccaccini AR. Therapeutic inorganic ions in bioactive glasses to enhance bone formation and beyond. Biomater. Sci. 2013;1:254.
Hoppe A, Mačković M, Detsch R, Mohn D, Stark WJ, Spiecker E, et al. Bioactive glass (type 45S5) nanoparticles: in vitro reactivity on nanoscale and biocompatibility. J Nanopart Res. 2012;14: 1-22. Hoppe A, Güldal NS, Boccaccini AR. A review of the biological response to ionic dissolution products from bioactive glasses and glass-ceramics. Biomaterials. 2011;32: 2757-74. Rath S N, Brandl A, Hiller D, Hoppe A, Gbureck U, Horch R E, Boccaccini A R, Kneser U; Copper doped 45S5 bioactive glass scaffolds stimulate endothelial cells in a co-culture with mesenchymal stem cells. Biomaterials. Acta Biomater 2014 (submitted). Papageorgiou GZ, Papageorgiou DG, Chrissafis K, Bikiaris D, Will J, Hoppe A, et al. Crystallization and melting behavior of poly(butylene succinate) nanocomposites containing silica-nanotubes and strontium hydroxyapatite nanorods. Ind Eng Chem Res. 2014;53(2):678-92. Grigoriadou I, Nianias N, Hoppe A, Terzopoulou Z, Bikiaris D, Will J, et al. Evaluation of silica-nanotubes and strontium hydroxyapatite nanorods as appropriate nanoadditives for poly(butylene succinate) biodegradable polyester for biomedical applications. Composites Part B: Engineering. 2014;60: 49-59. Milkovic L, Hoppe A, Detsch R, Boccaccini AR, Zarkovic N. Effects of Cu-doped 45S5 bioactive glass on the lipid peroxidation-associated growth of human osteoblast-like cells in vitro. J Biomed Mater Res, Part A. 2013. Romeis S, Hoppe A, Eisermann C, Schneider N, Boccaccini AR, Schmidt J, et al. Enhancing In Vitro Bioactivity of Melt-Derived 45S5 Bioglass® by Comminution in a Stirred Media Mill. J Am Ceram Soc. 2014;97: 150-6. Arkudas A, Balzer A, Buehrer G, Arnold I, Hoppe A, Detsch R, et al. Evaluation of angiogenesis of bioactive glass in the arteriovenous loop model. Tissue Eng, Part C. 2013;19: 479-86. Strobel LA, Hild N, Mohn D, Stark WJ, Hoppe A, Gbureck U, et al. Novel strontium-doped bioactive glass nanoparticles enhance proliferation and osteogenic differentiation of human bone marrow stromal cells. J Nanopart Res. 2013;15: 1-9. Strobel LA, Rath SN, Hoppe A, Beier JP, Arkudas A, Boccaccini AR, et al. Influence of leptin on osteogenic differentiation of human marrow stromal cells (hMSC) and modulation of BMP-2-mediated osteoinduction. J Tissue Eng Regen Med. 2012;6: 271-. Cabal B, Malpartida F, Torrecillas R, Hoppe A, Boccaccini AR, Moya JS. The Development of Bioactive Glass-Ceramic Substrates with Biocide Activity. Adv Eng Mater. 2011;13(12):B462-B6. Will J, Hoppe A, Müller FA, Raya CT, Fernández JM, Greil P. Bioactivation of biomorphous silicon carbide bone implants. Acta Biomater. 2010;6: 4488-94 Book chapters A. Hoppe A, AR. Boccaccini. 7 - Bioactive glass foams for tissue engineering applications. In: Netti PA, editor. Biomedical Foams for Tissue Engineering Applications: Woodhead Publishing; 2014. p. 191-212. A. Hoppe, A. R. Boccaccini. Biological response to ionic dissolution products from bioactive glasses and glass ceramics. In S. Deb, editor. Karger Frontiers in Biology: Biomaterials in Regenerative Medicine and Dentistry. Basel: Karger. (submitted) Presentations in international conferences (selected) A. Hoppe, A. Brandl, O. Bleiziffer, D. Janackovic and A. R. Boccaccini. Cobalt releasing bioactive glass (type 13-93) derived scaffolds for bone tissue engineering applications. Oral presentation. Annual meeting of the German Biomaterials Society Sep 2013, Erlangen, Germany. A. Hoppe. 3D scaffolds for bone tissue engineering based on Cu-releasing 45S5 Bioglass®. Invited oral presentation at the Symposium on “Bioceramics as Carrier” Jun 2013, Medical Center, University of Freiburg, Germany. A. Hoppe, J. Will, R. Detsch, A.R. Boccaccini und P. Greil. Formation and biocompatibility of biomimetic hydroxyapatite on chemically treated carbon substrates. Poster presentation. Annual meeting of the German Biomaterials Society Nov 2012, Hamburg Germany.
A. Hoppe, D. Hiller, S. Narayan Rath, A. Arkudas, U. Kneser, A. R. Boccaccini. In vitro and in vivo studies of Cu-doped 45S5 bioactive glass derived scaffolds. Oral presentation. 3rd International Conference "Strategies in Tissue Engineering" May 2012, Würzburg Germany. A. Hoppe, T. Reichel, R. Detsch, A. Lenhart, A. R. Boccaccini. Ion doped bioactive glass derived scaffolds for bone tissue engineering. Poster presentation. 3rd International Conference "Strategies in Tissue Engineering" May 2012, Würzburg Germany A. R. Boccaccini, A. Hoppe. Cellular response to ionic dissolution products from bioactive glasses and glass ceramics. Invited oral presentation. Meeting of the American Ceramic Society Jan 2012, Daytona Beach (FL), USA. A. Hoppe, D. Hiller, U. Kneser, A. R. Boccaccini Novel scaffolds made from metallic ion doped bioactive glasses: fabrication and characterization. Oral presentation in the young researcher forum. Annual meeting of the American Ceramic Society Jan 2012, Daytona Beach (FL) USA A. Hoppe, D. Hiller, S. Narayan Rath, U. Kneser, A. R. Boccaccini Novel Cu-doped bioactive glass (45S5) derived scaffolds for bone tissue engineering. Oral presentation, Annual meeting of the American Ceramic Society Jan 2012, Daytona Beach (FL) USA A. Hoppe, M. Miola, E. Verné, A.R. Boccaccini. Novel Zn-doped bioactive glasses developed by ion-exchange. Poster presentation, EUROMAT Sep 2011, Montpellier, France A. Boccaccini, A. Hoppe. Oral presentation, The biological effect of ionic dissolution products from bioactive glasses, EUROMAT Sep 2011, Montpellier, France