Biomaterials

Zinc-doped bioactive glass/Polycaprolactone hybrid scaffolds manufactured by direct

and indirect 3D printing methods for bone regeneration
--Manuscript Draft--

Manuscript Number: jbmt64303

Article Type: FLA Original Research

Keywords: Sol-gel; Bioactive glass; Organic-inorganic hybrid; Additive manufacturing; Human

mesenchymal stem cells; Bone Tissue Engineering

Corresponding Author: Jonathan Lao, Ph.D, Habil.

University of Clermont Auvergne
Aubiere, FRANCE

First Author: Nafise Elahpour

Order of Authors: Nafise Elahpour

Isabella Caroline Chiara Cichon

Cédric Bossard

Nora Abdellaoui

Valérie Montouillout

Franck Fayon

Christine Taviot-Guého

Tina Frankenbach

Alexander Crispin

Pardis Khosravani

Boris Michael Holzapfel

Edouard Jallot

Susanne Mayer

Jonathan Lao, Ph.D, Habil.

Abstract: A novel organic-inorganic hybrid, based on SiO2-CaO-ZnO bioactive glass (BG) and
polycaprolactone (PCL) associating the highly bioactive and versatile bioactive glass
with clinically established PCL was examined. The BG-PCL hybrid is obtained by acid-
catalyzed silica sol-gel process inside PCL solution either by direct or indirect printing.
Apatite-formation tests in simulated body fluid (SBF) confirm the ion release along with
the hybrid’s bone-like apatite forming. Kinetics differ significantly between directly and
indirectly printed scaffolds, the former requiring longer periods to degrade, while the
latter demonstrates faster calcium phosphate (CaP) formation. Remarkably, Zn
diffusion and accumulation are observed at the surface within the newly-formed active
CaP layer. Zn release is found to be dependent on printing method and immersion
medium. Investigation of BG at the atomic scale reveals the ambivalent role of Zn,
capable of acting both as a network modifier and as a network former linking the BG
silicate network. In addition, hMSCs viability assay evidences no cytotoxicity of the Zn-
hybrid. LIVE/DEAD staining of scaffolds demonstrates excellent biocompatibility, cell
attachment and proliferation for over seven weeks. Overall, the hybrid material either
non-doped or doped with metal trace element is a promising candidate to be translated
to clinical applications for bone regeneration .

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
Author Declaration Click here to access/download;Author Declaration;AUTHOR
DECLARATION.docx

AUTHOR DECLARATION

We wish to confirm that there are no known conflicts of interest associated with this publication and
there has been no significant financial support for this work that could have influenced its outcome.
We confirm that the manuscript has been read and approved by all named authors and that there
are no other persons who satisfied the criteria for authorship but are not listed.

We further confirm that the order of authors listed in the manuscript has been approved by all of us.
We confirm that we have given due consideration to the protection of intellectual property
associated with this work and that there are no impediments to publication, including the timing of
publication, with respect to intellectual property. In so doing we confirm that we have followed the
regulations of our institutions concerning intellectual property.

We further confirm that any aspect of the work covered in this manuscript that has involved either
experimental animals or human patients has been conducted with the ethical approval of all relevant
bodies and that such approvals are acknowledged within the manuscript.

We understand that the Corresponding Author is the sole contact for the Editorial process (including
Editorial Manager and direct communications with the office). He/she is responsible for
communicating with the other authors about progress, submissions of revisions and final approval of
proofs. We confirm that we have provided a current, correct email address which is accessible by the
Corresponding Author.

Sincerely,

France, 5th May 2023, Prof. Jonathan Lao on behalf of all authors
Cover Letter Click here to access/download;Cover Letter;Cover_letter-
1.docx

Muskuloskelettales Universitätszentrum
München - MUM

Muskuloskelettales Universitätszentrum München

Klinik für Orthopädie und Unfallchirurgie Prof. Dr. med. Susanne Mayer
LMU Klinikum Großhadern, Marchioninistraße 15, 81377 München Gesamtlaborleitung

Kontakt

Tel. +49 89 4400-74860

Mobil +49 152 54889941

E-Mail: [email protected]
muenchen.de

www.mum-lmu.de

Dear Prof. Leong,

dear Dr. Quek,

please find enclosed our manuscript for a full-length article entitled “Zinc-doped
bioactive glass/Polycaprolactone hybrid scaffolds manufactured by direct
and indirect 3D printing methods for bone regeneration”, which we would
like to submit for publication in the Journal “Biomaterials”.

Biomaterials and Bone Tissue Engineering (BTE) provide new treatment options for critical-
Vorstand
sized bone defects. The purpose of this study was to assess the impact of incorporating Zn,
Ärztlicher Direktor:
a known antibacterial and bone regeneration promoting agent, inside additively-
Prof. Dr. Markus M. Lerch
manufactured hybrids combining the highly bioactive and versatile bioactive glass with (Vorsitz)
clinically established polycaprolactone. Kaufmännischer Direktor:
Markus Zendler
Investigation of BG at the atomic scale reveals the ambivalent role of Zn, capable of acting Pflegedirektor (komm.):
both as a network modifier and as a network former linking the BG silicate network. Zn Alfred Holderied

diffusion and accumulation are observed at the surface within the newly-formed active CaP Vertreter der Medizinischen
Fakultät:
layer, with adequate Zinc release depending on immersion medium and printing technique. Prof. Dr. med. Thomas Gudermann
Furthermore, the hMSCs viability assay shows no cytotoxicity of the Zn-hybrid with (Dekan)
LIVE/DEAD stainings of 3D-culture proving excellent biocompatibility, cell ingrowth into Institutionskennzeichen:
pores, and proliferation for more than seven weeks, demonstrating the ability to perform as 260 914 050

a mid-term host for tissue formation Umsatzsteuer-ID:

DE813536017
As potential reviewers we suggest Prof. Paul Dalton, University of Oregon
Das Klinikum der Universität
([email protected], [email protected]) or Prof. Georgi Wassilew, University of München ist eine Anstalt des
Greifswald ([email protected]). Öffentlichen Rechts
We hope that after the reviewing process you will be able to judge our manuscript suitable
for publication in “Biomaterials”. Thanking you for your consideration, we are looking very
much forward to hearing from you.

Sincerely,

Prof. Lao & Prof. Mayer-Wagner

Manuscript Click here to
access/download;Manuscript;Biomaterials_manuscript_05May.
Click here to view linked References

1 Zinc-doped bioactive glass/Polycaprolactone hybrid

2
3 scaffolds manufactured by direct and indirect 3D
4
5 printing methods for bone regeneration
6
7
8 Nafise Elahpour1†, Isabella Cichon2†, Cédric Bossard1, Nora Abdellaoui1, Valérie Montouillout3,
9
10 Franck Fayon3, Christine Taviot-Guého4, Tina Frankenbach2, Alexander Crispin6, Pardis
11 Khosravani5, Boris Michael Holzapfel2, Edouard Jallot1, Susanne Mayer2*† and Jonathan Lao1*†
12
13 1 Université Clermont Auvergne, CNRS/IN2P3, LPC, F-63000 Clermont-Ferrand, France ; [email protected]
14 2 Department of Orthopaedics and Trauma Surgery, Musculoskeletal University Center Munich (MUM), University Hospital,
15 LMU Munich, Munich, Germany; [email protected]
3 Conditions Extrêmes et Matériaux : Haute Température et Irradiation, CEMHTI, CNRS-UPR3079, Univ. Orléans, F-45071
16
17 Orléans, France ; [email protected]
4 Université Clermont Auvergne, Institut de Chimie de Clermont-Ferrand, CNRS/UMR 6296, F-63000 Clermont-Ferrand, France.
18
19 ; [email protected]
5 Flow Cytometry Core Facility, Biomedical Center, Ludwig-Maximilians-Universität München, 82152 Planegg-Martinsried,
20
Germany;
21 6 Institut für Medizinische Informationsverarbeitung Biometrie und Epidemiologie (IBE), Ludwig-Maximilians-Universität
22
München, 81377 München, Germany;
23
* Correspondence: [email protected] , [email protected]
24
†These authors contributed equally to this work
25
26
27
28 Abstract: A novel organic-inorganic hybrid, based on SiO2-CaO-ZnO bioactive glass (BG) and polycaprolactone (PCL)
29 associating the highly bioactive and versatile bioactive glass with clinically established PCL was examined. The BG-
30 PCL hybrid is obtained by acid-catalyzed silica sol-gel process inside PCL solution either by direct or indirect printing.
31
Apatite-formation tests in simulated body fluid (SBF) confirm the ion release along with the hybrid’s bone-like apatite
32
33 forming. Kinetics differ significantly between directly and indirectly printed scaffolds, the former requiring longer
34 periods to degrade, while the latter demonstrates faster calcium phosphate (CaP) formation. Remarkably, Zn diffusion
35 and accumulation are observed at the surface within the newly-formed active CaP layer. Zn release is found to be
36 dependent on printing method and immersion medium. Investigation of BG at the atomic scale reveals the ambivalent
37 role of Zn, capable of acting both as a network modifier and as a network former linking the BG silicate network. In
38 addition, hMSCs viability assay evidences no cytotoxicity of the Zn-hybrid. LIVE/DEAD staining of scaffolds
39
demonstrates excellent biocompatibility, cell attachment and proliferation for over seven weeks. Overall, the hybrid
40
41 material either non-doped or doped with metal trace element is a promising candidate to be translated to clinical
42 applications for bone regeneration.
43
44 Keywords: Sol-gel 1; Bioactive glass 2; Organic-inorganic hybrid 3; Additive manufacturing 4; Human
45 mesenchymal stem cells 5; Bone Tissue Engineering 6
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Graphical abstract. Created with BioRender.com.
21
22
23 1. Introduction
24 Critical size bone defects cannot heal without additional surgical treatment. They pose a major
25 orthopaedic challenge in healthcare, with high costs and patient morbidity [1]. With over two
26
million bone grafting surgeries performed each year [2], bone grafting remains the second most
27
28 common tissue transplantation after blood transfusion. Autografts are still considered the "gold
29 standard" for treatment; however, they have several inherent limitations, including limited
30 availability and morbidity due to the need for a surgical donor site. Biobanked allografts as an
31 alternative are costly due to tissue harvesting and storage procedures carry risk of infection and
32 have limited osteoinductive and mechanical properties. Therefore, synthetic substitutes are
33 increasingly in the focus of research groups [3, 4]. Numerous requirements are placed on ideal
34
35 bone substitute materials in bone tissue engineering (BTE) – in addition to good
36 biocompatibility and adequate mechanical properties, the scaffolds should have high
37 osteoinductivity, osteoconductivity and biodegradability in order to be colonised with cells and
38 function as suitable tissue templates for bone formation [5].
39 In line with these requirements, bioactive glasses (BG) are of the highest interest among
40 synthetic bone substitutes. Once implanted, they trigger physicochemical reactions with body
41
42
fluids and subsequent delivery of osteostimulating degradation products (e.g. Si, Ca, P ions)
43 capable of regulating and even amplifying in vivo osteogenesis: for instance, among BG-
44 leaching products, Si(OH)4 orthosilicic acid species are known to stimulate osteoblast
45 proliferation and differentiation [6]. This interaction between the implant and host finally
46 results in a robust interfacial bioactive layer that will be arduous to be explanted even when
47 applying force, due to the strong BG bonding with bone.
48
Beyond bone/cartilage regeneration, recent research discloses promising results in wound
49
50 healing and soft tissue repair [7]. Indeed BG consists of a rather easily tuneable glassy matrix,
51 especially when synthesized through sol-gel route, which allows the incorporation of a wide
52 variety of “therapeutic” ions that can trigger targeted biological assets once delivered in the
53 body. Those desirable assets include: osteogenesis (e.g. Sr2+ [8-10], Li+ [11]), angiogenesis
54 (Co2+[12], Cu2+ [13, 14] , Zn2+ [15]), immune-modulatory (BO33—, Cu2+, Zn2+ [16]), anti-
55 inflammatory (Zn2+ [17]) and anti-bacterial activity (Ag+ [18], Cu2+ [16], Zn2+[19]). As can be
56
57
seen, Zn ions are especially attractive due to their simultaneous actions on multiple levels. In
58 fact zinc as the sixth most abundant trace element in human body plays a crucial role in bone
59 homeostasis, turnover, mineralization and ECM (Extracellular matrix) synthesis [20]. Bone
60 tissue is the major reservoir of Zn in human body [21, 22]. Studies suggest that Zn
61
62
63
64
65
 supplementation upregulates osteoprotegrin (OPG) expression. Despite scarce knowledge
1 about cellular/molecular pathways through which zinc promotes bone growth, it is known that
2 it can favourably affect osteoblast and chondrocyte functions while prohibiting osteoclastic
3 bone resorption [21, 23]. Worthy of note that all Zn amount in blood is bound with albumin,
4
α2-macroglobulin which increases the apparent binding of Zn2+ and diminishes free Zn2+
5
6 concentration in the serum to less than 1 nM [24]. From a materials point of view, a first
7 challenge is thus to be able to deliver significant amounts of Zn from the glass to the medium.
8 Zn in the glass is present in the form of zinc oxide (ZnO), which partially dissolves in aqueous
9 media to yield Zn2+ cations, which, either in a free or complex form that significantly contribute
10 to the biological effect, as demonstrated for the antibacterial activity in literature [25]. A second
11 challenge is related to the bone-like apatite forming ability of BG ̶ usually referred as in vitro
12
13
bioactivity, Zn being known to impede the transformation of amorphous calcium phosphate
14 (CaP) to crystalline hydroxycarbonate apatite (HCA) [26]. The higher the Zn2+ content, the
15 slower the HCA deposition; same with the rate of nucleation and later precipitation [27]. Also
16 pH adjustment, media circulation/refreshment, porosity of the material, incubation duration can
17 obviously influence the HCA precipitation rates [28, 29].
18 Here, our strategy to tackle these two challenges was to first act on the porosity of the BG
19
20
material with the aid of additive manufacturing. Considering the brittleness of raw BG
21 scaffolds, we associate the inorganic BG phase with a tough but “3D printing friendly”
22 polymeric phase, namely polycaprolactone (PCL), which is a polymer of choice in the
23 bioprinting field due to its adequate melting point, rheological and shear-thinning properties
24 [30]. We investigate the impact of Zn incorporation inside SiO2-CaO BG/PCL hybrid scaffolds
25 that previously showed remarkable performance supporting bone growth in a challenging
26
critical-mice calvaria model [31]. Hybrids consisting of 30 wt. % BG based on SiO2-CaO-ZnO
27
28 (75/15/10 Si/Ca/Zn relative atomic percent) and 70 wt% PCL were produced by conducting the
29 silica sol-gel process inside a solution of PCL [32]. SiO2-CaO-ZnO BG/PCL hybrid scaffolds
30 were additively manufactured by FDM-based (Fused deposition modelling) techniques
31 involving direct and indirect 3D printing. We hypothesized that the two completely different
32 porous structures obtained could be used to modulate the Zn delivery and thus inhibitory effect
33 on Hydroxyapatite (Hap) formation and subsequent biological effects. Similarly, the
34
35
degradation of the materials, ion delivery and Hap-formation ability were investigated in two
36 different media, a protein-free medium (standard saline SBF solution) and a protein-based
37 solution (Mueller-Hinton broth medium), hypothesizing that the ability of proteins to create
38 soluble complexes with metallic ions could impact the Zn release and in vitro bioactivity. Along
39 with a complete set of physicochemical characterization, the in vitro cell behavior is evaluated
40 through direct and indirect cell viability assays involving bone marrow derived human
41
42
mesenchymal cells (hMSCs) to evaluate the immense potential of hybrid scaffolds for BTE
43 applications.
44
45 2. Materials and Methods
46
47 2.1. Synthesis
48
The sol-gel synthesis of bioactive glasses proceeded according to our conventional protocol
49
50 detailed elsewhere [32]. Glasses with relative atomic percentage ratios of 75/25 Si/Zn, 75/15/10
51 Si/Ca/Zn (relative atom%) were synthesized. The binary system is being used as a control group
52 to evaluate the influence of Zn incorporation. Analytical grade reagents comprising of
53 Tetraethylorthosilicate (TEOS) (99 % purity, Sigma-Aldrich®), zinc methoxide (99.9 % purity,
54 Aldrich®) were used as the sol-gel precursors for Si and Zn. The calcium source is obtained by
55 an overnight calcination of a CaCO3 powder (Normapur, VWR®) at a heating rate of 1°/min
56
57 up to 1000°C plateaued for 1 hour. After one hour of cooling down under vacuum, obtained
58 CaO powder was freshly used.
59 Sol-gel process is started by hydrolyzing TEOS in absolute ethanol with the addition of 2 M
60 HCl (diluted from 37% fuming HCl, Sigma-Aldrich) for 30 min (molar ratio of ethanol: H2O:
61
62
63
64
65
 TEOS: HCl = 3.7: 2: 1: 0.07). This was pursued by calcium oxide and/or zinc methoxide
1 addition, and then same amount of ethanol as above was used to aid dissolution of the powder
2 while letting the mixture stirring (250 rpm) to form the glass network.
3 Just a while before the gelation of the bioactive glass solution, sol was divided into two parts.
4
Half of it was kept for 72 hours of aging in the same sealed flask and consecutive room
5
6 temperature (RT) drying for further analyses (XRD, NMR, PDF). The obtained xerogel were
7 ground into very fine powder by an agate mortar and pestle prior to analysis.
8 The other half of the BG sol was used to produce hybrids by mixing it with a solution of (18.2
9 w/v %) PCL in tetrahydrofuran (99.9 % purity, Sigma-Aldrich) with a 3:7 (wt/wt) BG sol/PCL
10 solution. The PCL average molecular number varied dependent of the printing method (Mn =
11 80 kDa or 45 kDa, Sigma-Aldrich). The highly viscous BG/PCL sols were rigorously blended
12
13
manually and then sonicated for 15 minutes (10kW, 40 kHz), stirred in sealed flask for 1 h for
14 homogenization and further condensation. The homogeneous obtained hybrid sol is then
15 processed into scaffolds but in two different manners based on direct and indirect 3D printing
16 as explained in the following sections.
17
18 2.2. Additive manufacturing of hybrids
19
20 2.2.1. 3D direct printing
21
22 Synthesized hybrid mentioned in previous section was left under the laminar flow hood to be
23 dried to form a xerogel. The latter was crushed with a knife mill (IKA A11 basic analytical
24 mill) followed by rinsing in absolute ethanol and drying at RT. The obtained powder was loaded
25 in a metal syringe equipped with a 500-µm nozzle and heated for approximately 1 hour
26
(temperatures mentioned below). The syringe is mounted on a FDM bioprinter (Bioscaffolder
27
28 GeSim mbH) which allows the extrusion of highly viscous materials thanks to a piston-based
29 extrusion conducted at 82°C. The cartridge and nozzle temperatures were respectively chosen
30 to be 75°C and 82°C, feed rate and printing speed were equal to 1 and 0.8 mm/s. The material
31 flow experiencing a change in the section from a few centimeters diameter (syringe cylinder)
32 to a few mm (the conical attached tip) and finally through the micrometric nozzle made it
33 essential to increase the temperature of the nozzle a few degrees above the temperature of the
34
35 syringe. Multiple 3-layered scaffolds (radius = 10 mm, infill distance = 600 µm) were prepared.
36
37 2.2.1. 3D indirect printing
38 Based on a template leaching method, paraffin template molds were printed from paraffin
39 granules loaded in a metal syringe equipped with a 300-µm nozzle. The feed rate, speed,
40
41
cartridge and nozzle temperature were set to 1 mm/s, 1.8 mm/s, 46°C and 47°C, respectively.
42 The paraffin sacrificing templates were geometrically defined as circular cylinders (radius= 10
43 mm, height= 5 or 10 mm, infill distance = 0.8 mm). Once printed the paraffin molds were fitted
44 into plastic flat-bottom microtubes, then infiltrated with hybrid sol obtained at section 2.1, then
45 centrifuged for 1 min to ensure penetration of the viscous hybrid sol into the void spaces of the
46 paraffin template. Infiltrated molds were dried (for 3-5 days). Three consecutive overnight
47 cyclohexane bathes were needed to dissolve the paraffin templates and one overnight ethanol
48
49 bath to assure complete removal of cyclohexane. The obtained porous BG/PCL hybrid scaffolds
50 were finally dried for 1 day. Figure 1 shows an illustration of the synthesis and additive
51 manufacturing process.
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26 Figure 1. Schematic representation of synthesis and additive manufacturing of hybrid scaffolds, created with
27 BioRender.com.
28
29
2.3. Bioactivity Evaluation in SBF Solution
30
31 The SBF (Simulated body fluid) solution has ionic concentrations similar to human blood
32 plasma, and was produced as formulated by Bohner and Lemaitre [33]. Hybrid scaffolds
33 previously prepared and dried were carefully cut into equal 1 mm-thick sections. Each section
34
averagely weighing 20-25 mg was soaked in SBF (1 mg/mL). Samples were incubated at 37°C
35
36 on an orbital shaker (120 rpm) to avoid them settling. After 1 hour, 6 hours, 1 day, 3, 7 and 14
37 days (21 days also in first round of experiments), they were removed, then immersed in absolute
38 ethanol to inhibit any further interaction/CaP formation. These interacted sections were kept for
39 microscopy. Extracted SBF was filtered with 0.22 µm syringe filters at each time point.
40
41 2.4. ICP-OES (Inductively Coupled Plasma-Optical Emission Spectroscopy)
42
43 The concentrations of Ca, Zn, Si and P ions released into SBF medium were measured by ICP-
44 OES. The mean values combined with standard deviations from triplicates for each dissolved
45 ion were analysed and recorded. Calibration solutions were prepared to obtain a linear
46 correlation between ions’ intensity and concentration.
47
48 2.5. PIXE analysis
49
50 Elemental composition of the non-soaked and SBF-soaked hybrid scaffolds sections were also
51 obtained by Particle-Induced X-ray Emission (PIXE) nuclear microprobe. This is an elemental
52 non-destructive method, similar to Energy-Dispersive X-ray Spectroscopy (EDS) or X-ray
53 Fluorescence (XRF), with increased sensitivity due to low Bremsstrahlung background and is
54 capable of chemically mapping structures down to a submicronic resolution [34]. Samples were
55
56 first embedded in epoxy resin (Agar 100 Resin, agar scientific), then cut into slices of about
57 150-µm thickness with a low-speed diamond saw. PIXE quantitative chemical imaging was
58 then carried out at the AIFIRA platform (LP2I, UMR5797, Gradignan, France) using a 3 MeV
59 incident proton beam (beam diameter of 1 µm).
60
61
62
63
64
65
 An 80 mm2 lithium-doped silicon (Si(Li)) detector, commonly used in EDXS and low-energy
1 gamma rays detection, equipped with a 12 μm-thick beryllium window and a 100 µm thick
2 aluminum “funny filter” (central hole diameter = 2 mm). Prior to investigation of samples, soda-
3 lime flat glass composed of ten oxides (NIST-620, USA) was used as standard reference for
4
calibration purposes. Conversion of X-ray peaks intensities to ion concentrations was carried
5
6 out with Gupixwin software (version 2.2.4, University of Guelph, Canada) to identify the local
7 composition of desired ions.
8
9 2.6. Pair Distribution Function (PDF) analysis
10
Local atomic structure of Si/Ca 75/25, Si/Zn 75/25 and Si/Ca/Zn 75/15/10 sol-gel derived
11
12 grounded BG powders have been measured with X-ray PDF analysis. All samples studied here
13 were previously proved as amorphous according to their corresponding XRD patterns as
14 perquisite prior to PDF analysis. The atomic PDF were obtained from X-ray total scattering
15 data collected on a PANalytical Empyrean diffractometer equipped with a solid state
16 GaliPIX3D detector, a focusing X-ray multilayer mirror, and an Ag anticathode (Kα1 =
17 0.5594214 Å, Kα2 = 0.5638120 Å). Powder samples were placed in glass capillaries of 0.7 mm
18
19
diameter. An empty capillary of the same type was measured in the same way for background
20 subtraction. Data were recorded over the 1-145° 2θ range, which corresponds to an accessible
21 maximum value for the scattering vector Q max of 21.4 Å−1. Data merging, background
22 subtraction, and Kα2 stripping were done using HighScore Plus software provided by
23 PANalytical Corporation. It was also used to generate a corrected and normalized total
24 scattering structure functions S(Q) and considering the above-mentioned bulk chemical
25
26
compositions. Finally, the PDF or G(r) were calculated from the Fourier transforms of
27 S(Q) truncated at 20 Å−1. Since the silicon content is the same in all samples, the number of
28 the Si-O pairs (direct interatomic distance) is expected to be the same in all samples; the PDF
29 curves were thus scaled to get a similar intensity for Si-O PDF peak at 1.62 Å. The simulated
30 PDF were calculated using PDFgui software [35].
31
32 2.7. 29Si magic angle spinning nuclear magnetic resonance (MAS-NMR) spectroscopy
33
34 Qn silicon species distribution inside the binary and ternary BG compositions (75/25 Si/Zn and
35 75/15/10 Si/Ca/Zn) were determined using a Bruker Avance I spectrometer operating at a
36 magnetic field of 9.4 T (1H and 29Si Larmor frequencies of 400.2 and 79.5 MHz) using a 4
37 mm double resonance MAS probe head. The Si quantitative MAS spectra was recorded at a
38 rotor spinning frequency of 10 kHz after a RF pulse of 2.4 µs (corresponding to a flip angle of
39
40
30°) and a recycle delay of 10 s. 29Si chemical shifts were referenced relative to
41 tetramethylsilane. The spectra were simulated using DMfit software (CEMHTI, CNRS,
42 Orléans, France). Following equation was used to calculate the Degree of Condensation (DC)
43 of BG:
44 2 × Q2 + 3 × Q3 + 4 × Q4
45 DC = 100 × (1)
4
46
47
48 2.8. Scanning Electron Microscopy (SEM)
49
50 Samples were carbon coated prior to SEM observation and then the micro/macro porous
51 structure, topography and morphology of the scaffolds were observed with a field-emission gun
52 scanning electron microscope Regulus 8230 (Hitachi, Japan) operating at 1kV, images were
53 recorded with a secondary electron detector. Alongside X-ray microanalysis was performed in
54 areas of interest with EDS Ultim Max 170mm² detector (Oxford, UK) at an electron
55
56
accelerating voltage increased to 10kV.
57
58 2.9. X-ray Diffraction (XRD)
59 Diffraction patterns were collected on a Philips X-Pert Pro equipped with a X’celerator 1D
60 detector (2.122° active length) using CuKα1/Kα2 radiations (1.5406/1.5444 Å) in Bragg
61
62
63
64
65
 Brentano θ-θ geometry from 2 to 90° (2θ) with a scan step of 0.066°. Data analysis was
1 performed with HighScore Plus software provided by PANalytical Corporation and
2 Crystallography Open Database (COD) to identify the crystalline peaks.
3
4 2.10. Fourier-Transform Infrared Spectroscopy (FTIR)
5
6 FTIR spectroscopy was used to assess the chemical functional groups of scaffolds before and
7 after immersion in SBF. The acquisition of spectra was done with a Nicolet 380 FT-IR (Thermo
8 Fisher Scientific) equipped with a diamond Attenuated Total Reflectance (ATR) accessory. 32
9 spectral scans at resolution of 4 cm−1 were repeated over the wavenumber range 4000–500 cm-
10 1 (mid-IR region).
11
12
2.11. Cellular assays
13
14
15 2.11.1. Cell culture
16 Cells are cultured with α-Minimal Essential Medium (PAN-Biotech GmbH, Aidenbach,
17 Germany) supplemented with 10% fetal bovine serum (Anprotec, South Africa), 60 IU/mL
18
19
penicillin, 60 µg/mL streptomycin and 2.5 µg/mL amphotericin B as culture medium. Cells are
20 maintained at 37 °C, high relative humidity and 5% CO2. Cells from passages 3 to 5 are used
21 in the experiments. All supplements are purchased from Sigma-Aldrich Co., St. Louis, MO,
22 USA, if not mentioned otherwise. Falcons and Flasks are purchased from Sarstedt, Nümbrecht,
23 Germany.
24
25 2.11.2. Isolation of bone marrow derived hMSCs
26
27 The hMSCs are isolated from bone marrow of femoral heads of 4 patients (3 male, 1 female;
28 mean age 66,8 years; range 59-76 years) undergoing hip replacement surgery (Klinikum
29 Großhadern, Munich, Germany). The study is approved by the LMU medical ethics committee
30 (ID: 22-0379, 8-8-2022) with informed patient consent being required. After being transported
31 in saline solution, bone marrow is scraped out and washed with 30 ml phosphate buffered saline
32
33
(PBS, Biochrom, Berlin, Germany) before being poured through a cell strainer (pores size 70
34 µm, BD Bioscience, San Jose, CA, USA) into a 50 mL falcon tube. The remaining bone marrow
35 pieces are digested three times for 10 min each on a 3D-shaker in an incubator (37 °C) with 10
36 ml Collagenase II solution (1 g/1 ml in cell culture media). The 10 ml solution is then poured
37 into another falcon through a cell strainer (pore size 70 µm, BD Bioscience, San Jose, CA,
38 USA). Both falcons are centrifuged (500 g, 5 min, RT), media is discarded, and each cell pellet
39
40
is suspended in 10 ml cell culture media and transferred to culturing flasks. Following that, all
41 flasks are kept in the incubator (Binder, Tuttlingen, Germany), media was changed twice a
42 week. Cells are passaged at 90% confluence.
43 Multilineage differentiation is used to assess stem cell traits such as adipogenic, osteogenic,
44 and chondrogenic differentiation potential, following the recently published protocol [36].
45 After several weeks, successful differentiation is demonstrated using Bodipy stainings, Alizarin
46 Red staining and Safranin O staining. Images were captured using a light microscope
47
48 (Axioobserver, Zeiss, Oberkochen, Germany). Additionally, cells are characterized with flow
49 cytometry sample analysis for the expression of the stem cell markers CD73, CD90 and CD105.
50 After trypsinization and washing, 1*10⁶ cells are incubated in 100 μl staining volume with
51 antibodies against CD73, CD90, CD105, which are stem cell markers, and CD34 and CD45,
52 which serve as a negative control [37, 38]. For detailed information on staining reagents and
53 antibody panel, a list is attached in appendix (see Table S1). After incubation on ice for 30
54
55
minutes and washing, flow cytometry measurements are performed with a BD LSRFortessa™
56 Cell Analyzer (BD Bioscience, San Jose, USA). FlowJo™ v10.8.1 software are used to analyze
57 data.
58
59 2.11.3. Indirect cell viability assay with cell proliferation reagent WST-1
60
61
62
63
64
65
 Scaffolds (n = 2 per group) from each group [Hyb80 and Zn-Hyb80] (respectively, undoped
1 and Zn-doped hybrid with PCL 80k) are incubated in cell culture media for up to 1, 5 or 7 days
2 to determine the potential cytotoxicity of the scaffold material. The conditioned media (CM) is
3 collected and stored at -80°C. Cells are cultured in 96-well plates (Cellstar®, Greiner Bio-One
4
GmbH, Frickenhausen, Germany), at a density of 10⁴ cells/well and cultured under addition
5
6 of CM. As positive control, cells are treated with normal culture media. After 24 hours, cells
7 are washed with PBS to remove CM and cell proliferation reagent WST-1 (Roche Diagnostics
8 GmbH, Mannheim, Germany) is added (1:10 dilution) and incubated for 3.5 hours. All
9 quantifying experiments (indirect cell viability assay) are confirmed with four biological donors
10 and with at least three independent experiments performed for each donor. A scanning
11 microplate reader is used to measure absorbance at 450 nm (Multiskan FC, Thermofisher
12
13
Scientific, Schwerte, Germany). The wavelength 620 nm is chosen as a reference. An
14 illustration of the experimental procedure is schematically shown (Fig. S1).
15
16 2.11.4. Seeding and cultivation of the scaffolds
17 The hMSCs are expanded up to passage 5 and then drop seeded at a density of 2*10 5
18
19
cells/scaffold. Five drops of 5 ml each were pipetted onto the scaffolds, and the scaffolds are
20 incubated for 5h with an hourly moisture check. 5-20 ml culture media is added as needed. Cell
21 culture media is added after 5 h of incubation. The media is changed three times a week until
22 the cells are confluent, then media is changed to twice a week osteogenic media (α-MEM
23 supplemented with 10 % FBS, 60 IU/ml penicillin, 60 g/ml streptomycin, 100 nM
24 dexamethasone, 10 mM glycerophosphate, and 0.17 mM ascorbate-2-phosphate). 3D-
25 constructs are cultured for 7 weeks in total.
26
27
28 2.11.5. LIVE/DEAD Staining
29 Cytocompatibility of the scaffolds is further assessed by a LIVE/DEAD staining. Scaffolds are
30 stained with 10 µg/ml fluorescein diacetate (FDA) in PBS and 1 mg/ml Propidiumiodide in
31 PBS at 3, 5, and 7 weeks. Based on intracellular esterase activity, viable cells convert FDA to
32
33
green fluorescein, which stains viable cells green. As a nucleic acid dye, propidiumiodide can
34 only enter compromised cell membranes of dying or dead cells, dyeing them red. Pictures are
35 taken with a Leica SP8 WLL confocal microscope (Leica Microsystems GmbH, Wetzlar,
36 Germany). Non-quantifying experiments are confirmed with two biological donors.
37
38 2.11.6. Statistical evaluation
39
40 Descriptive analyses were performed with GraphPad Prism 9.5 (GraphPad Software, La Jolla,
41 CA, USA). Statistical inferences regarding For the indirect cell viability assay were based on
42 random intercept models using the GLIMMIX proxedure of the Statistical Analysis System
43 SAS release 9.04.01M6P11072018 for Linux (SAS Institute, Cary, NC, USA). Fixed effects
44 were material itself, time of incubation with the material and interaction between material and
45 length of incubation. Post-hoc tests were adjusted for multiple testing by multiplying the
46
47
uncorrected p value by the number of tests (Bonferroni correction). Adjusted p values <0.05
48 were considered statistically significant.
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 3. Results
1
2 3.1. NMR and PDF analysis of Si/Zn and Si/Ca/Zn BGs
3 The Qn species distribution is extracted from the 29Si MAS-NMR spectra visible in Fig 2. Their
4
5 respective proportion along with the degree of condensations of Si/Ca/Zn and Si/Zn are brought
6 in table 1. Qn silica species refer to Si atoms bonded to n bridging oxygen atoms in a silicate
7 network, a bridging oxygen (BO) being shared between two silica tetrahedras (Si ─ O ─ Si)
8 (fig. 2a). Therefore, the more Q4 species, the more polymerized and condensed the network, as
9 expected from a pure silicate network. Non-bridging oxygen (NBO) atoms can either arise from
10 the silanol groups (Si ─ OH) or from network modifier cations (alkaline or alkaline earth cations
11
12
as Ca²⁺ or Zn²⁺ in our case, as shown in Fig. 2a) which both disrupts the connections between
13 silica tetrahedras. The more Q2 and Q3 species, the more depolymerized is the silicate network.
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36 Figure 2. a) up: schematic representation of different Qn species in a silicate network (BO = Bridging Oxygen), down:
37 disordering the network by modifier cations e.g. Ca, M²⁺ is representing bivalent metal cation introduced to
38 bioactive glass network b) 29Si MAS-NMR spectra of Si/Zn (top) and Si/Ca/Zn (bottom) glasses and their fits.
39 Individual Qn contributions are shown in table 2.
40
41
42
43
In our case, the presence of Q2 and Q3 species is in favour of a successful incorporation of
44 network modifiers. This investigation of cation incorporation is of primary importance when
45 soft chemistry routes are used and thermal treatments are banished like in the present work, as
46 e.g. the temperature at which Ca is incorporated into a silicate network was found to vary
47 between RT and several hundred degrees depending on the calcium source [39]. From fig. 2b
48 and the deconvolution of 29Si MAS-NMR spectra, it is visible that both Si/Zn and Si/Ca/Zn
49 BG networks are poorly polymerized, with an abundant proportion of Q2 and Q3 species. For
50
51 the ternary Si/Ca/Zn BG, it is possible to identify a characteristic signature of Ca incorporation
52 with the presence of Q3(Ca) species. However, no specific signature of Zn incorporation can
53 be identified in the spectra, although it could account for the observed proportion of Q2 and Q3
54 species (if Zn acts as a network modifier). Nearly identical DCs (table 1) for the Si/Zn and
55 Si/Ca/Zn BGs are calculated, indicating the same degree of disruption in silicate glass networks
56 disrespectful of doping percentage of the glass with zinc. The DC are also close to the one
57
58
reported for the binary Si/Ca system in a previous work [40].
59
60
61
62
63
64
65
 1
Q species δiso FWHM % Q species δiso FWHM %
2 Si/Zn Si/Ca/Z
3 (ppm) (ppm) (ppm) (ppm)
4 Q4 -109.1 10.0 22 Q4 -108.0 10.0 28
5 Q3 -100.7 6.1 36 Q3 -100.6 6.0 17
6 Q 2
-91.5 10.0 42
3
Q (Ca) -97.0 6.0 17
7 Q2 -91.2 11.6 38
8
9
1 DC (Si/Zn) = 70, DC (Si/Ca/Zn) = 72.
10
11
12 Table 1. Relative intensities of Qn resonances obtained from fits and degrees of condensations (DC).
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28 Figure 3. Left: the pair distribution function (PDF) curve obtained from total X-ray scattering data from different
29 bioactive glass samples (respective compositions for blue: Si/Zn 75/25, red: Si/Ca 75/25 and black: Si/Ca/Zn 75/15/10
30 (all relative atom%)). Right: Zn–O coordination structures tetrahedral (C4), b) trigonal Bipyramidal (C5), c) Square
31 Pyramidal (C5´), d) Octahedral (C6). e) Structural model for Zn cations within the silicate network as tetrahedral
32 ZnO4 unites and charge balanced by Ca ions as proposed by Shahrabi et al. [41].
33
34 Extraction of the data from XRD analysis depends mostly on Bragg peaks and we like them to
35 be narrow, fully separated and well representative, however any sort of deficiencies or lack of
36 any long-range order makes it a challenge. As shown by research groups including Proffen et
37
38
al. [42, 43], the PDF analysis is an interesting method to investigate poorly crystalline and even
39 amorphous systems. The pair distribution functions G(r) for r-values up to 5 Å are given in
40 figure 3 (left). The low crystallinity of the present samples yields atomic PDF that rapidly decay
41 to zero at distances of 5-10 Å but still allow studying local atomic arrangement in materials.
42 Indeed, the peaks on the PDF observed below 5 Å correspond to interatomic distance between
43 first and second neighboring atoms while the intensity is proportional to the number of pairs at
44 a given distance or the coordination number in the different shells and to the product of the
45
46 scattering powers of the atoms forming the pair. The width of the peaks can give information
47 about the distribution of the distances; the broader the peak, the larger the distribution of
48 distances, which it can come from disorders. As a function of the nominal chemical bioactive
49 glass, there are changes in all atom-atom correlations shown in Figure 3 (left). Since all the
50 contributions of all pairs of atoms are considered in the PDF, it can be useful to refer to
51 structural models/ reference samples to determine to which pairs these correlations belong. In
52
53
the present case, we referred to SiO2 (quartz-type; COD 1526860) and ZnO (zincite-type; COD
54 1011258) structural models (Fig. S4) and also data reported for calcium silicate hydrate (C-S-
55 H) gels [44]. Based on this model, we were able to assign the peaks of PDF curve as follows:
56 r = 1.62, 2.05, 2.39, 2.66, 3.08, 3.14, 3.54 Å to Si-O, Zn-O, Ca-O, O-O, Si-Si/Si-Zn and Ca-Si
57 contributions respectively. The entire atom-atom correlations are figured out and mentioned in
58 figure 3 (left). Despite major differences in synthesis methods, it is interesting to note that the
59
60
short-range order of C-S-H gels mostly resembles to our bioactive glasses, particularly Si─O
61
62
63
64
65
 and Ca─O atom-atom distances are found nearly at the same positions [44]. Ca incorporation
1 into Si/Ca and Si/Ca/Zn BG silicate network is evidenced by the presence of Ca─O correlations
2 at 2.39 Å and Ca─Si distance observed at 3.54 Å for Si/Ca BG. Both distances obviously
3 disappear in Si/Zn, which does not contain Ca. On the other hand, in ternary Si/Ca/Zn BG, the
4
Ca-O distance is maintained and the Ca-Si distance disappears, suggesting that Ca is more
5
6 distant/ is farther from Si when Zn is introduced in the glass structure [44].
7
8 3.2. Primary morphological observations
9 A schematic presentation of printed paraffin template, obtained indirectly printed scaffold and
10 directly printed scaffold besides geometrical and morphological features as observed in SEM
11
12
is shown in figure 4. The mean strut diameter for the direct printed scaffolds is 0.547±0.038
13 mm and the maximum width equal to about 0.607 mm is usually observed at the contact point
14 of the strand with its beneath strand. Widening is observed in all these node-like places whereas
15 in-between to the next node we can see the thinning of the strand. Inter-strand distance is 0.650
16 mm. Considering indirect printing method, the images of negative paraffin mold indicate the
17 mean strut value of 0.371± 0.021 mm and an inter-strut space of 0.290± 0.006 mm.
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32 Figure 4. Imaging and SEM observations of the a) direct printing method of the hybrids. b) Paraffin template and
33 c) indirect printing of the hybrids
34
35
36
37 3.3. Directly printed vs indirectly printed scaffolds reactivity of Zn-BG/PCL of 45k hybrids
38
39 Over a three weeks’ time interval the ionic release behavior of direct and indirectly printed
40 scaffolds was investigated as reported in fig 5. Looking at the Si release in SBF, indirectly
41 printed scaffolds degrade quicker in comparison with directly printed scaffolds. A similar
42 behavior is found for Zn release, with surprising low Zn concentrations being released in SBF
43 : a maximum of 0.14 ppm is reached for indirectly printed scaffolds, while no more than 0.02
44
45
ppm Zn are detected in the fluids for directly printed scaffolds. Regarding Ca, the trends remain
46 unclear since Ca can be both released and its concentration increased in the medium as a result
47 of the material degradation, or decreased as a result of CaP precipitation at the surface of
48 hybrids. Therefore, the evolution of P in SBF is a good indication of the apatite forming ability
49 of the material. For the directly printed scaffolds, a slight decrease corresponding to a few ppm
50 loss in P, while for the indirect printed scaffolds the concentration of P is divided by two after
51
three days of interaction, followed by a constant plateau up to the whole 21 days, suggesting no
52
53 significant evolution of CaP formation or apatite formation being hindered in the long term.
54 Obviously, the thinner strands of the indirectly printed scaffold increase the surface reaction
55 and speed up the ionic leaching and exchanges and degradation of the material.
56 Chemical distributions of elements inside the scaffolds in the starting materials and after 14
57 days of interaction with SBF are depicted in fig.6. The ability to form CaP is evidenced for both
58 type of scaffolds. For directly printed scaffolds, CaP remain deposited within a thin layer at the
59
60 surface of the strands, while for indirectly printed, P seems quite evenly distributed within the
61
62
63
64
65
 strands, showing an extended CaP deposition through the whole material. Zn and P maps
1 overlap in mineralized areas, indicating the colocalization of CaP depositions with Zn in the
2 samples.
3 Considered simultaneously with the ICP-OES measurements, the chemical distribution in the
4
scaffolds show that the indirectly printed scaffolds are superior in terms of bioactivity and Hap
5
6 formation. However, the indirectly printed scaffolds were extremely fragile and subjected to
7 premature degradation in SBF solution. Observing severe brittleness for the indirectly printed
8 scaffolds and poor bioactivity potential of directly printed scaffolds led us to opt for the
9 indirectly printing technique but using a higher molecular weight (average Mn 80000; PCL80k
10 instead of lately 45k). Afterwards, these will be mentioned as respectively Zn-Hyb80 and
11 Hyb80 in the following sections.
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38 Figure 5. Evolution of Si, Ca, Zn and P concentrations in Simulated Body Fluid during the immersion of Zn-BG/PCL
39 45k hybrid scaffolds (DP: Directly printed, IDP: Indirectly printed) (1mg/1 mL SBF, 37°C) as determined by ICP-
40 OES, n = 3.
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27 Figure 6. Cross-sectional PIXE chemical mapping of Si, Ca, Zn and P in Zn-BG/PCL 45k hybrid scaffolds printed
28 differently via direct and indirect routes each before and after immersion in SBF for 14 days (scale bars are 200 µm
29 in two first lines and 100 µm in below ones).
30
31
32
33 3.4. ATR and XRD results on Zn-Hyb80 and Hyb80 scaffolds
34
35 Pre-immersion spectra of both Zn-Hyb80 and Hyb80 scaffolds exhibit features of symmetric
36 and asymmetric CH2 bands corresponding to peaks positioned at around respectively 2863 cm-
37 1 and 2942 cm-1 (3000-2800) cm-1 (fig. 7a, 7b) [45]. The most intense band at 1720 cm-1 is
38 correlated to carbonyl stretching ν(C=O). This carbon-oxygen esteric stretching band is
39
40
observed in all the other graphs and they are normalized based on this peak. Next band at
41 1293cm-1 is indicator of C-O and C-C stretching. Band observed at 1240 cm-1 correlates to
42 asymmetric C-O-C stretching and the one at 1175 cm-1 corresponding to ester C-O stretching.
43 The band shown in a rectangular red zone highlighted in both diagrams of Zn-Hyb80 and Hyb80
44 can be ascribed to C-O stretching. All these peaks are in perfect agreement with the
45 characteristic peaks of the polymer PCL reported elsewhere [46].
46 The first ATR pattern positioned at the bottom of each stack corresponds to the raw BG. In case
47
48 of Hyb80 the only band peak positioned at 1004 cm-1 is ascribed to Si–O–Si asymmetric
49 stretching. In search of signs confirming the apatite formation, the characteristic band peak
50 positioned at 962 ± 2 cm-1, despite its insignificant intensity in IR spectroscopy, is an
51 interesting proof (fig. 7c). This band representing the ν1(PO4)3- is reported at same position
52 for apatite phases or OCP but in case of amorphous calcium phosphate, monetite or brushite is
53 shifted to respectively 950, 995 and 985 cm-1 [47].
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19 Figure 7. Evaluation of the apatite-forming ability of BG-PCL hybrid scaffolds. ATR spectra of (a) Hyb80 scaffolds
20 and (b) Zn-Hyb80 scaffolds before and after immersion in SBF for 6 h, 1, 3, 7 and 14 days (1 mL of SBF per mg,
21 37°C). (c) Comparison between the FTIR spectra of Hyb80 and Zn-Hyb80 after 7 days of immersion in SBF
22 indicating nanocrystalline non-stoichiometric apatite formation highlighted by▼ in dashed zoomed zone.
23
24
25 The XRD results of 3D indirectly printed scaffolds of Hyb80 and Zn-Hyb80 are presented in
26 figure 8. Primarily there are two very sharp at about 21.24° and 23.68° (2θ) attributed to PCL
27 respectively corresponding to the (110) and (200) planes of the orthorhombic crystal structure
28
29
[48]. An additional peak positioned at 29.7° (2θ) is observed, which is related to alignment of
30 the polymer chains because of tetrahydrofuran evaporation and printing process (preferred
31 orientation). Due to the amorphous nature of the bioactive glass component, all the other non-
32 PCL characteristic peaks are related to newly-formed phases after immersion in SBF. The
33 patterns of the phases formed after the immersion disclosed broad peaks that can be related to
34 their low crystallinity and/or small grain sizes [49]. A shoulder feature positioned at 22.06° is
35
observed in all Zn-Hyb80 and Hyb80 samples revealing the formation of aragonite phase and
36
37 corresponding to (002) plane of its orthorhombic lattice.
38 In Hyb80 scaffold patterns, the formation of HCA (generic formula Ca10–x(PO4)6–x(CO3,
39 HPO4)x(OH)2–x), calcite and aragonite (CaCO3) phases was evident during SBF immersion.
40 The observation of characteristic double peak at 25.96°/26.88° (2θ) along with the peaks
41 positioned at 30.56°, 35.62°, 36.09°, 37.99°, 40.99°, 42.31° and 45.47°(aragonite JCPD,
42 reference code: 96-901-5562) support the presumption of aragonite phase. On the other hand,
43
44
Hap/HCA characteristic triplex peaks are detected at 31.69°/32.16°/32.81° (2θ) respectively
45 corresponding to d-spacing planes of (121), (112) and (030).
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36 Figure 8. a: XRD patterns of Hyb80 scaffolds, b: XRD patterns of Zn-Hyb80 scaffolds after determined time intervals
37 of immersion in SBF (▼: HCA/Hap feature peaks, ■: Calcite, ●: Aragonite)
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57 Figure 9. Microstructure of Hyb80 and Zn-Hyb80 scaffolds after 7 days of incubation in SBF (More dispersion of
58 post-immersion precipitation deposits on Hyb80 compared to Zn-Hyb80).
59
60
61
62
63
64
65
 3.5. SEM/EDS of Zn-Hyb80 and Hyb80 scaffolds
1 Polygonal crystals are fully covered on non-doped Hyb-80 scaffold after 7 days of immersion
2 in SBF. Less precipitates were observed at the surface of Zn-Hyb80. It is noticeable that the
3
4
sample doped with Zn exhibited less porosity on the surface and internal pore walls (fig.9).
5 SEM observation in Fig.9 and EDS elemental analysis (Fig. S3) show the presence of calcite
6 polymorphs, coexisting with CaP precipitates, as calcium and phosphorous are detected in most
7 of the nucleation sites. Such proofs are not observed for Zn-Hyb80, whose bioactivity seems
8 strongly diminished.
9
10 3.6. Ion release in Mueller Hinton (MH) Media compared to SBF
11
12 Given the very low release of Zn ion in SBF, we wanted to investigate the Zn-Hyb80 behavior
13 in a protein containing solution. The interaction between metals and proteins, including the
14 formation of metallocomplexes, is of interest for its potential to increase the Zn bioavailability.
15 The Mueller Hinton (MH) media is a yellowish transparent liquid commonly used in
16 microbiological assays. The media is composed of beef extract, casein hydrolysate, starch, agar
17
18 and distilled water. The Zn-Hyb80 scaffolds were immersed in MH for 1h, 6h, 1, 3, 7 and 14
19 days and the ionic release was determined by ICP-OES. For comparison, the same was done in
20 SBF for Hyb80 and Zn-Hyb80. In Fig. 10, the decrease of P concentration in SBF confirms the
21 apatite forming of the materials, but lower when hybrid is doped with Zn. As it can be seen in
22 figure 10, in MH we see a high Si, Ca release and a remarkable increase in Zn release, with Zn
23 concentration about 130 times higher than that measured in SBF (always below 0.5 ppm). This
24
25
could be attributed to proteins ability to enhance the dissolution of ZnO.
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45 Figure 10. Up: ICP-OES measured evolution of the Si, Ca and P of the Hyb80 and Zn-Hyb80 immersed in SBF up
46 to 14 days (Zn always below 0.5 ppm is not included). Down: Evolution of Si, Ca and Zn concentrations in Mueller
47 Hinton Medium during the immersion of Zn-Hyb80 scaffolds (1 mL of MH per mg, 37°C) as determined by ICP-
48 OES, n = 3.
49
50 3.7. Isolated cells prove stemness and multilineage differentiation
51
52
All donors included in the study underwent multilineage differentiation successfully
53 comparable to published data [36]. Flow cytometry characterization shows high expression for
54 all stem cell markers (<99%) for the stem cells, with only one exception for Donor D, where
55 CD105 expression was 62.5 % (see Table S2 with dot plots for visualization). It has been
56 reported that in bone marrow derived stem cells, CD105 negative subpopulations can be found,
57 which is associated with even higher osteogenic differentiation capacity [50]. Minor subsets of
58
59
cells are also CD34 and CD45 positive, which are markers for hematopoietic precursor cells.
60 We assumed that direct isolation of bone marrow derived stem from human bone marrow cells
61
62
63
64
65
 yielded different smaller subpopulations from the bone marrow including hematopoietic
1 precursor cells.
2
3 3.8. Hybrid materials show promising cytocompatibility with hMSCs under normal experimental
4 conditions
5
6 The indirect cell viability assay is used to assess the cytocompatibility of the hybrid materials.
7 WST-1 is a tetrazolium salt that is converted to formazan dye by cellular enzymes. The amount
8 of formazan product correlates with the number of metabolically active cells, making it a useful
9 reagent for spectrophotometric cell number quantification. WST-1 is cleaved by metabolically
10 active cells into formazan dye, which is measured spectrophotometrically with an ELISA
11
12
reader. Any increase or decrease in measured absorbance directly correlates with the number
13 of metabolically active cells, provided the seeding density is between 0.1 and 5 × 104 cells/well.
14 Incubating CM with scaffolds for 1 day, 5 days, or 7 days, results in an increasing proportion
15 of released scaffold products. The indirect cell viability assay (see fig. 11) reveals no significant
16 difference in absorbance between Hyb80 and control groups for 1, 5 and 7 days for all biological
17 donors. This suggests excellent cytocompatibility for the undoped hybrid group for all time
18
points since there is no significant change in the number of metabolically active cells. Similarly,
19
20 CM incubated for 1 d with Zn-Hyb80 scaffolds leads to no decrease in the number of
21 metabolically active cells, which implies that Zn-Hyb80 CM 1d had no cytotoxic effects.
22 The data for CM from the Zn-Hyb80 group incubated for 5 and 7 days with scaffolds was donor-
23 dependent. For donor B and D, there is no significant difference in absorbance compared to the
24 control group, i.e. not showing a cytotoxic effect for CM. The situation is different with donor
25 A and C, where there is little to no absorption at CM of the Zn-Hyb80 group of 5d and 7 d
26
27 (Donor A) or 7d (Donor C). This suggests a lack of metabolically active cells and thus either a
28 decrease in cell number or an inhibitory influence on cell metabolism. The CM of the Zn-Hyb80
29 material with the higher concentrations of soluble products seems to have a donor-dependent
30 effect on cell viability.
31
32 3.9. Scaffolds allow mid-term cell culture for up to 7 weeks with excellent cell attachment and
33 proliferation
34
35 In the LIVE/DEAD staining, viable cells can be detected on both scaffolds at all of the time
36 points for both donors (see fig. 11 and multimedia file 1 in Supplementary Information, showing
37 a 3D-image of cell ingrowth into the scaffold’s pores), indicating that both scaffold materials
38 allow cell attachment and mid-term cell culture for at least 7 weeks, indicating no change in
39 cytocompatibility compared to control scaffolds. Over time, there is a visible increase in cell
40
41 count and cell ingrowth into pores. LIVE/DEAD staining is not considered suitable for
42 quantification, since only small parts of the scaffolds can be imaged and drop seeding of
43 scaffolds does not guarantee a homogenous cell layer.
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
 1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29 Figure 11. Left, LIVE/DEAD Staining of 3D constructs at 3, 5 and 7 weeks. Scale bars are 500 µm. right: WST-1
30 indirect cell viability assay, all data is normalized against the control group (treated with normal cell culture media)
31 and expressed as arithmetic mean ± standard error; *statistically significant difference between d7-ZnHyb and all
32 other groups (p<0.001 since p values were adjusted for multiple testing).
33
34 4. Discussion
35
36 This study aimed at evaluating the impact of incorporating Zn inside hybrids consisting of BG
37 and PCL. The materials were synthesized using an acid-base catalyzed sol-gel route, the BG
38 phase being doped with Zn. Hybrids were additively manufactured through the direct printing
39 of hybrid granules or indirect printing of the sol involving a 3D sacrificial template. Insights
40 into the atomic structure of the BG were gained through solid-state NMR and X-ray PDF
41 analyses, the former evidencing a limited impact of Zn incorporation on the silicate network
42
43 itself and calcium incorporation, while the latter suggests an ambivalent role of Zn due to the
44 tetrahedral coordination of ZnO that could act as a bridge between silicate tetrahedras. Taken
45 together, the NMR and PDF analyses suggest doping of the SiO2/CaO bioactive glass system
46 with Zn leads to a structural change, which is not a mere substitution of Ca by Zn. From their
47 ‘role in BG network’ point of view, Ca is known as a network modifying cation in the glass
48 structure that produces non-bridging oxygen atoms, while ZnO is known to be capable of
49
50
serving both as a network modifier and as an intermediate oxide depending on its proportion,
51 with some supports for the former being gained by the 29Si MAS-NMR analysis, and supports
52 for the latter being gained from the PDF analysis as discussed in section 3.1. Therefore at least
53 a fraction of Zn is present in the form of tetrahedral species (ZnO42-) that play an ambivalent
54 role, capable of connecting silica tetrahedras [51] but these species need Ca ions to stabilize
55 and balance the charge [41]. This could explain the reduction observed in the peak
56
corresponding to “short” Si─Ca distances in the PDF of Si/Ca/Zn BG, along with the decrease
57
58 in the rate of dissolution of the silicate network in SBF as observed with ICP-OES
59 measurements. Accordingly, bioactivity is simultaneously decreased and delayed with Zn
60
61
62
63
64
65
 substitution for Ca, because of both the impact of Zn on the silicate structure and of the lower
1 amount of Ca in the glass for Si/Ca/Zn BG.
2 XRD and ATR studies, collectively point out a tedious decreased ion exchange with the
3 immersion media for Zn-Hyb80. Same observations of reduction in the leaching activity of Zn-
4
doped glasses are reported elsewhere [41, 52].
5
6
7 In vitro apatite-forming ability tests in SBF confirm the ion release along with the hybrid’s
8 bone-like apatite forming ability. The presence of Zn ions in the bioactive glass structure is
9 supposed to potentially lead into Zn-doped calcium phosphates. Unlike Hyb-80, traces of HCA
10 formation are scarcely found in Zn-Hyb80 scaffold even after longer periods of SBF immersion.
11 Our results are consistent with previous studies about zinc inhibiting role on crystal formation
12
13
/growth of CaPs. In fact, Legeros et al. found that Zn negatively affects de novo apatite
14 formation and even remarkably diminishes the crystal size. At concentrations higher than
15 2mM/L (130.76 ppm), Zn in solution even promotes the formation of more soluble CaP phases
16 and mostly ACP instead of all the other multiple possible phases [53]. Other studies report a
17 strong inhibition of Zn ions against apatite formation in vitro even up to 31 days of SBF
18 immersion. The formation of CaPs was evidenced in section 3.3 and Fig 6, but here the XRD
19
20
analysis suggest that it remains in an amorphous state for Zn-Hyb80. The lower bioactivity we
21 can conclude for the Zn-Hyb80 can be ascribed to the lower bioactivity of Zn-BG compared to
22 binary BG. Zn is reported recurrently as an ion inhibiting the ion exchange with the surrounding
23 media since it rivals the Ca ions and exceeds them in formation of a Zinc calcium phosphate
24 precipitation, thus mostly retarding any HCA precipitation. Moreover, the lower bioactivity
25 could also be related to the Zn insertion in the silicate network. The Infra-red peak positioned
26
at about 960 cm-1 as a Raman shift in Raman spectroscopy clearly reveals the phosphate
27
28 vibrational bands with major intensities [54]. This could be suggested for further investigations.
29 However, here in IR spectra of our hybrid the overlap of PCL and post-immersion hybrids in
30 this same position, makes it obscure to surely relate this band to CaP functional groups.
31 Magnification of ν4(PO4)3- band positioned at 550 and 600 cm-1 is shown in fig.7c and its
32 evolution during the immersion time in SBF solution. These bands can be indicator of biological
33 Hap formation and are observed in soaked Hyb80 IR after 7 days of immersion in SBF while
34
35
they are not observable after 7 days of immersion for Zn-Hyb80. A feature of HCA formation
36 is the asymmetric stretching vibration positioned at wavenumber of about 1030 cm-1, however
37 in both Zn-Hyb80 and Hyb80 scaffold samples IR spectra the PCL peak at about 1045 cm-1 is
38 overlapping at all time points, hindering a deduction on having P-O asymmetric stretching
39 proof. P–O bending vibration at estimate wavenumber of 600 cm-1 in post-immersion patterns
40 of Hyb80 is also a sign of Hap formation and hence bioactivity (fig.7c).
41
42
43 Regarding Zn insertion in the silicate network, a classification of different possible Zn sites in
44 the primary coordination sphere is shown in fig.3 (right), forming 4, 5 or 6 coordination
45 complexes. Zn-O distance observed in 2.05 Å here is typical of a tetrahedral environment (4
46 coordination complex), which is in agreement with the most prevalent option mentioned in the
47 literature namely tetrahedral [55]. Interestingly, the gradual shift of the peak maximum
48 attributed to Si-Si/Si-Zn correlations to longer distances when increasing zinc content (3.08 Å
49
50 for Si/Ca 75/25, 3.10 Å Si/Ca/Zn 75/15/10 and 3.14 Å for Si/Zn 75/25), also simultaneous
51 increase in the intensity of this peak and its broadening strongly suggest that at least part of
52 these [ZnO4] tetrahedra are connected to the silicate network, as shown in Fig 3-e. It may be
53 recalled that in the case of zincite, the distance between neighbouring is 3.20 Å (COD 1011258).
54
55 The release kinetics of Zinc differ significantly between directly and indirectly 3D printed
56
57
scaffolds. The structure obtained with 3D direct printing is denser because it results from the
58 coarse (several hundred µm) printing resolution and hot extrusion associated with FDM, and
59 requires longer soaking periods in SBF to degrade, while the 3D indirect template method led
60
61
62
63
64
65
 to finer pore sizes, thinner struts with internal porosity generated as a result of solvent
1 evaporation, therefore enhancing apatite nucleation.
2
3 Remarkably, diffusion and accumulation of zinc are observed at the surface of both kind of
4
hybrids, within the newly-formed active CaP layer. Zn release was found to be dependent on
5
6 the scaffold printing method but also on the medium, with Zn2+ concentrations released being
7 2 orders of magnitude higher in Muller-Hinton Broth bacterial culture medium compared to
8 SBF, reinforcing observations from previous studies about the ability of proteins to enhance the
9 dissolution of ZnO by creating highly soluble complexes. Regarding protein’s ability to
10 enhance the dissolution of ZnO, casein phosphopeptides are well known to improve zinc
11 absorption, caseien hydrolysate is responsible for maintaining the Zn in solution during Ca and
12
13
P precipitation [56]. Zn ions interact with certain amino acids and phosphate groups on the
14 protein's surface as determined by spectroscopic techniques and electron microscopy imaging
15 [57]. For example, Zn2+ might bind to the phosphate groups of serine, the histidine imidazole
16 ring (mainly nitrogen) or to the deprotonated carboxyl groups of glutamic and aspartic acids.
17 Zn2+ may participate in the cross-linking of two or more neighbouring charged amino acids
18 groups [57]. Zinc is involved in many biological processes and has unique coordination
19
20
chemistry, allowing it to form stable aqua complexes in acidic aqueous solutions that can
21 exchange water molecules when binding to other ligands. Its affinity for functional groups such
22 as carboxyl and amino groups of proteins and subsequent formation of metal ligands is probably
23 the reason why the Zn release is observed to be more than 100-fold in a protein medium such
24 as MH unlike a protein free one such as SBF. This is a good finding promising that significant
25 amount of Zn2+ ions can be delivered in vivo [58].
26
27
28 As we aimed for secondary biological assets with fractional substitution of Ca source with Zn
29 source, it was concluded that in first place the Zn-doped hybrid biomaterial and following
30 scaffolds are of no toxicity to mesenchymal human cells under normal experimental conditions
31 and can perform as good as simple binary Si/Ca bioactive glass derived hybrids. The
32 maintenance of cytocompatibility for the hybrid material was expected since both undoped and
33 doped bioactive glass, PCL and their hybrid materials are reported to possess good
34
35
biocompatibility in literature [32, 59-62] and Zn-doped and undoped bioactive glass had shown
36 a beneficial effect on bone formation [62, 63]. Similar results are obtained for the LIVE/DEAD
37 staining – on all scaffolds, the vast majority of cells is alive for up to 7 weeks. Over time, there
38 is a visible increase in cell count and cell ingrowth into pores, indicating that the scaffolds
39 possess excellent potential as tissue templates for 3D constructs and bone substitutes. Undoped
40 and doped hybrid scaffolds showed the ability to function as template for cells to attach, grow
41
42
and proliferate desirably leading to bone remodeling.
43 In order to quantify any potential cytotoxic effects, WST-1 cell proliferation assays are
44 conducted with CM. Undoped hybrid showed no cytotoxic effect for all time points for all
45 biological donors, since there are no statistically significant differences in the absorbance
46 measurements compared to the other groups. This underlines the impressions from the
47 LIVE/DEAD staining. Similarly, there are no statistically significant differences for Zn-Hyb80
48 CM for 1 d, suggesting that there are no cytotoxic effects for the CM from 1d of doped hybrid
49
50 scaffolds.
51 Regarding the CM from 5d and 7d there are donor-dependent differences in the number of
52 metabolically active cells. In vivo, Zn is considered a relatively non-toxic metal element that is
53 only toxic to humans and animals at really high doses [64], but free Zn ions outside a narrow
54 concentration range have been shown to be cytotoxic to various cell lines and primary cells
55 [65], demonstrating the oligodynamic effect of the metal Zn [66]. It may have good
56
57
biocompatibility and antibacterial properties at low concentrations, but it also has cytotoxic
58 effects at high concentrations [65]. Researchers demonstrated that biodegradability and
59 extensive mass loss of the polymer matrix of composite-based scaffolds consisting of a
60 bioceramic phase and a polymer phase at some time points can lead to high concentrations of
61
62
63
64
65
 release products, which may cause cell death [59, 67]. Regarding the donor-specific data, we
1 assume that there is inter-donor variability with Zn tolerance. Random intercept models were
2 calculated to exclude donor-dependent Zn tolerance and only evaluate material-specific
3 influences on cell viability. Only for 7d, the Zn-Hyb80 CM with the highest concentration of
4
released products had led to a significant decrease in the number of metabolically active cells
5
6 in the random intercept model. This implied that only Zn-Hyb80 CM with rather high
7 concentration exhibited cytotoxic effects.
8 As cell culture on Zn-Hyb80 scaffolds included three media changes per week, incubating CM
9 for 7 days with Zn-Hyb80 material would appear to be an experimental exaggeration of the real
10 conditions. Nonetheless, potential cytotoxic effects at higher concentrations within the
11 scaffold’s material cannot be excluded. Cytocompatibility is the most important trait for novel
12
13
biomaterials in order to facilitate bone regeneration and minimize tissue damage [5]. Overall,
14 our findings show that Hyb80 and Zn-Hyb80 materials have excellent biocompatibility with
15 cell attachment, growth, and proliferation on scaffolds for at least 7 weeks and have no negative
16 effect on cell viability under normal experimental conditions, making them candidates for BTE.
17
18
19 5. Conclusions
20
21 This study was conducted for the design of novel Zn-doped hybrid materials consisting of BG
22 and PCL. Performance of these scaffolds was evaluated through physicochemical
23 characterization and in vitro 3D cell culture, showing promising release kinetics and excellent
24 biocompatibility. Direct and indirect cellular assays involving hMSCs were conducted. Both
25 groups of hybrid scaffolds show the ability to perform as a mid-term host for cells to attach,
26
27 grow and proliferate desirably leading to bone remodeling. Indirect cell viability assay
28 evidences no cytotoxic effects of the Zn-hybrid under normal experimental conditions, a major
29 concern to be addressed given the relative sensitivity of eukaryote cells exposed to Zn and
30 possible Zn2+ homeostasis deregulation. LIVE/DEAD staining of hMSCs 3D cultured in the
31 scaffolds for prolonged periods demonstrate excellent biocompatibility with cell attachment,
32 growth and proliferation for at least 7 weeks.
33
34
Overall, the hybrid material either non-doped or doped with metal trace element is a promising
35 candidate to be translated to clinical applications for bone regeneration. Considering the
36 antibacterial potential, the improved Zn ion release in protein-containing media can open new
37 avenues in biomedical applications attached to different conditions.
38
39
40 Author Contributions:
41
42 Nafise Elahpour: Writing - original draft preparation, Methodology, Visualization, Formal analysis.
43 Isabella Cichon: Writing - original draft preparation, Methodology, Visualization, Formal analysis.
44 Cédric Bossard: validation, Methodology. Nora Abdellaoui: validation, Methodology. Valérie
45 Montouillout: software, data curation, validation. Franck Fayon: software, data curation, validation.
46 Christine Taviot-Guého: software, data curation, Conceptualization, Supervision, Formal analysis,
47 Writing - review & editing. Tina Frankenbach: Conceptualization, Supervision, Writing - review &
48
editing. Boris Holzapfel: Writing - review & editing. Pardis Khosravani: Resources, Methodology.
49
50 Alexander Crispin: Software, Validation. Edouard Jallot: Conceptualization, Supervision, Formal
51 analysis, Project administration. Susanne Mayer: Conceptualization, Supervision, Formal analysis,
52 Writing - review & editing, funding acquisition. Jonathan Lao: Conceptualization, Supervision, Formal
53 analysis, Writing - review & editing, funding acquisition.
54
55 Funding: The project was supported by the EU under the European Regional Development Fund
56 program (“Fonds Européen de Développement Régional” FEDER) and Région Auvergne-Rhône-Alpes
57 under the “pack Ambition Recherche” program (BIOSTEON project, grant agreement No. AV0016426).
58
59 Institutional Review Board Statement: The study was conducted in accordance with the Declaration
60 of Helsinki, and approved by the Ethics Committee of LMU Munich (ID: 22-0379, 8-8-2022).
61
62
63
64
65
 Informed Consent Statement: Informed consent was obtained from all subjects involved in the study.
1 Acknowledgments: We acknowledge the Core Facility Flow Cytometry at the Biomedical Center, LMU
2 Munich and Dr. Benjamin Tast and Pardis Khosravani, for providing equipment and their expertise,
3 services and assistance with data acquisition and analysis. The authors also want to acknowledge
4 Christelle Blavignac and Claire Szczepaniak at the Centre Imagerie Cellulaire Santé (Clermont-Ferrand
5
Faculty of Medicine) and Emmy Voyer at the Laboratoire Magma et Volcans (CNRS and Université
6
7 Clermont Auvergne mixed unit UMR 6524) for SEM-EDXS analysis, and Claire Fonquernie at the
8 Laboratoire Magma et Volcans for ICP-OES measurements.
9 Conflicts of Interest: The authors declare no competing conflict of interest. The funders had no role in
10
the design of the study; in the collection, analysis, or interpretation of data; in the writing of the
11
12 manuscript; or in the decision to publish the results”.
13
14 Data availability: Upon request, all data can be obtained from the corresponding authors.
15
16 Supplementary Information: available as a separate file
17
18
19
20
21 References:
22
23
1. Mayfield, C.K., et al., 3D-Printing for Critical Sized Bone Defects: Current Concepts and Future
24
25 Directions. Bioengineering, 2022. 9(11): p. 680.
26 2. Van der Stok, J., et al., Bone substitutes in the Netherlands–a systematic literature review. Acta
27 biomaterialia, 2011. 7(2): p. 739-750.
28 3. Campana, V., et al., Bone substitutes in orthopaedic surgery: from basic science to clinical practice.
29 Journal of Materials Science: Materials in Medicine, 2014. 25(10): p. 2445-2461.
30
31 4. Archunan, M.W. and S. Petronis, Bone Grafts in Trauma and Orthopaedics. Cureus, 2021. 13(9): p.
32 e17705.
33 5. Scheinpflug, J., et al., Journey into bone models: a review. Genes, 2018. 9(5): p. 247.
34 6. Zhou, X., et al., Orthosilicic acid, Si (OH) 4, stimulates osteoblast differentiation in vitro by
35 upregulating miR-146a to antagonize NF-κB activation. Acta biomaterialia, 2016. 39: p. 192-202.
36
37 7. Baino, F., et al., Bioactive glasses: Special applications outside the skeletal system. Journal of Non-
38 Crystalline Solids, 2016. 432: p. 15-30.
39 8. Zreiqat, H., et al., The incorporation of strontium and zinc into a calcium–silicon ceramic for bone tissue
40 engineering. Biomaterials, 2010. 31(12): p. 3175-3184.
41 9. Isaac, J., et al., Effects of strontium-doped bioactive glass on the differentiation of cultured osteogenic
42
cells. Eur Cell Mater, 2011. 21(130): p. e43.
43
44 10. Wu, Q., et al., Strontium-incorporated bioceramic scaffolds for enhanced osteoporosis bone regeneration.
45 Bone Research, 2022. 10(1): p. 55.
46 11. Farmani, A.R., et al., Preparation and In Vitro Osteogenic Evaluation of Biomimetic Hybrid
47 Nanocomposite Scaffolds Based on Gelatin/Plasma Rich in Growth Factors (PRGF) and Lithium-Doped
48
45s5 Bioactive Glass Nanoparticles. Journal of Polymers and the Environment, 2022: p. 1-16.
49
50 12. Wu, C., et al., Hypoxia-mimicking mesoporous bioactive glass scaffolds with controllable cobalt ion
51 release for bone tissue engineering. Biomaterials, 2012. 33(7): p. 2076-2085.
52 13. Bührer, G., et al., Evaluation of in vivo angiogenetic effects of copper doped bioactive glass scaffolds in
53 the AV loop model. Biomedical glasses, 2016. 2(1).
54 14. Gorustovich, A.A., J.A. Roether, and A.R. Boccaccini, Effect of bioactive glasses on angiogenesis: a
55
56 review of in vitro and in vivo evidences. Tissue Engineering Part B: Reviews, 2010. 16(2): p. 199-207.
57 15. Zhao, X., Bioactive glass based fibre mats for wound healing. 2022.
58 16. Schuhladen, K., et al., Cu, Zn doped borate bioactive glasses: antibacterial efficacy and dose-dependent
59 in vitro modulation of murine dendritic cells. Biomaterials science, 2020. 8(8): p. 2143-2155.
60
61
62
63
64
65
 17. Sun, H., et al., Incorporation of zinc into binary SiO2-CaO mesoporous bioactive glass nanoparticles
1 enhances anti-inflammatory and osteogenic activities. Pharmaceutics, 2021. 13(12): p. 2124.
2 18. Kargozar, S., et al., Mesoporous bioactive glasses: Promising platforms for antibacterial strategies. Acta
3 Biomaterialia, 2018. 81: p. 1-19.
4
5 19. Sergi, R., et al., Zinc containing bioactive glasses with ultra-high crystallization temperature, good
6 biological performance and antibacterial effects. Materials Science and Engineering: C, 2019. 104: p.
7 109910.
8 20. O’Connor, J.P., et al., Zinc as a therapeutic agent in bone regeneration. Materials, 2020. 13(10): p.
9 2211.
10
11 21. Amin, N., et al., Zinc supplements and bone health: The role of the RANKL-RANK axis as a therapeutic
12 target. Journal of Trace Elements in Medicine and Biology, 2020. 57: p. 126417.
13 22. Huang, T., G. Yan, and M. Guan, Zinc homeostasis in bone: zinc transporters and bone diseases.
14 International journal of molecular sciences, 2020. 21(4): p. 1236.
15 23. Liang, D., et al., Zinc upregulates the expression of osteoprotegerin in mouse osteoblasts MC3T3-E1
16
through PKC/MAPK pathways. Biological trace element research, 2012. 146: p. 340-348.
17
18 24. Foote, J. and H. Delves, Albumin bound and alpha 2-macroglobulin bound zinc concentrations in the
19 sera of healthy adults. Journal of clinical pathology, 1984. 37(9): p. 1050-1054.
20 25. Pasquet, J., et al., The contribution of zinc ions to the antimicrobial activity of zinc oxide. Colloids and
21 Surfaces A: Physicochemical and Engineering Aspects, 2014. 457: p. 263-274.
22
26. Jaroch, D. and D. Clupper, Modulation of zinc release from bioactive sol–gel derived SiO2‐CaO‐
23
24 ZnO glasses and ceramics. Journal of Biomedical Materials Research Part A: An Official Journal
25 of The Society for Biomaterials, The Japanese Society for Biomaterials, and The Australian
26 Society for Biomaterials and the Korean Society for Biomaterials, 2007. 82(3): p. 575-588.
27 27. Balasubramanian, P., et al., Zinc-containing bioactive glasses for bone regeneration, dental and
28 orthopedic applications. Biomedical glasses, 2015. 1(1).
29
30 28. Lu, X., et al., Effect of solution condition on hydroxyapatite formation in evaluating bioactivity of B2O3
31 containing 45S5 bioactive glasses. Bioactive Materials, 2019. 4: p. 207-214.
32 29. Tavangarian, F. and R. Emadi, Nanostructure effects on the bioactivity of forsterite bioceramic.
33 Materials Letters, 2011. 65(4): p. 740-743.
34 30. Fathi, A., et al., Three-dimensionally printed polycaprolactone/multicomponent bioactive glass scaffolds
35
36 for potential application in bone tissue engineering. Biomedical glasses, 2020. 6(1): p. 57-69.
37 31. Granel, H., et al., Bioactive glass/polycaprolactone hybrid with a dual cortical/trabecular structure for
38 bone regeneration. ACS Applied Bio Materials, 2019. 2(8): p. 3473-3483.
39 32. Bossard, C., et al., Polycaprolactone/bioactive glass hybrid scaffolds for bone regeneration. Biomedical
40 glasses, 2018. 4(1): p. 108-122.
41
33. Bohner, M. and J. Lemaitre, Can bioactivity be tested in vitro with SBF solution? Biomaterials, 2009.
42
43 30(12): p. 2175-2179.
44 34. Wyroba, E., et al., Biomedical and agricultural applications of energy dispersive X-ray spectroscopy in
45 electron microscopy. Cellular and Molecular Biology Letters, 2015. 20(3): p. 488-509.
46 35. Farrow, C., et al., PDFfit2 and PDFgui: computer programs for studying nanostructure in crystals.
47
Journal of Physics: Condensed Matter, 2007. 19(33): p. 335219.
48
49 36. Mayer-Wagner, S., et al., Membrane-based cultures generate scaffold-free neocartilage in vitro:
50 influence of growth factors. Tissue Engineering Part A, 2010. 16(2): p. 513-521.
51 37. Lv, F.-J., et al., Concise review: the surface markers and identity of human mesenchymal stem cells. Stem
52 cells, 2014. 32(6): p. 1408-1419.
53
38. Consentius, C., et al., In situ detection of CD73+ CD90+ CD105+ lineage: Mesenchymal stromal cells
54
55 in human placenta and bone marrow specimens by chipcytometry. Cytometry Part A, 2018. 93(9): p.
56 889-893.
57 39. Yu, B., et al., Effect of calcium source on structure and properties of sol–gel derived bioactive glasses.
58 Langmuir, 2012. 28(50): p. 17465-17476.
59
60
61
62
63
64
65
 40. Bossard, C., et al., Mechanism of calcium incorporation inside sol–gel silicate bioactive glass and the
1 advantage of using Ca (OH) 2 over other calcium sources. ACS Biomaterials Science & Engineering,
2 2019. 5(11): p. 5906-5915.
3 41. Shahrabi, S., et al., Structural discrepancies and in vitro nanoapatite formation ability of sol–gel derived
4
5 glasses doped with different bone stimulator ions. Ceramics international, 2011. 37(7): p. 2737-2746.
6 42. Proffen, T., et al., Atomic pair distribution function analysis of materials containing crystalline and
7 amorphous phases. Zeitschrift für Kristallographie-Crystalline Materials, 2005. 220(12): p. 1002-
8 1008.
9 43. Bordet, P. Local structure studies using the pair distribution function. in EPJ Web of Conferences. 2015.
10
11 EDP Sciences.
12 44. Morandeau, A.E. and C.E. White, In situ X-ray pair distribution function analysis of accelerated
13 carbonation of a synthetic calcium–silicate–hydrate gel. Journal of Materials Chemistry A, 2015.
14 3(16): p. 8597-8605.
15 45. Tallia, F., et al., Bioactive, Degradable and Tough Hybrids Through Calcium and Phosphate
16
Incorporation. Frontiers in Materials, 2022: p. 430.
17
18 46. Sahoo, S., et al., Synthesis of chitosan–polycaprolactone blend for control delivery of ofloxacin drug.
19 Carbohydrate Polymers, 2010. 79(1): p. 106-113.
20 47. Drouet, C., Apatite formation: why it may not work as planned, and how to conclusively identify apatite
21 compounds. BioMed research international, 2013. 2013.
22
48. Balu, R., et al., Electrospun Polycaprolactone/Poly (1, 4-butylene adipate-co-polycaprolactam) blends:
23
24 Potential biodegradable scaffold for bone tissue regeneration. Journal of Biomaterials and Tissue
25 Engineering, 2011. 1(1): p. 30-39.
26 49. Ofudje, E.A., et al., Synthesis and characterization of Zn-Doped hydroxyapatite: scaffold application,
27 antibacterial and bioactivity studies. Heliyon, 2019. 5(5): p. e01716.
28 50. Anderson, P., et al., CD105 (endoglin)-negative murine mesenchymal stromal cells define a new
29
30 multipotent subpopulation with distinct differentiation and immunomodulatory capacities. PloS one,
31 2013. 8(10): p. e76979.
32 51. Rabiee, S.M., et al., Effect of ion substitution on properties of bioactive glasses: A review. Ceramics
33 International, 2015. 41(6): p. 7241-7251.
34 52. Aina, V., et al., Zinc-containing bioactive glasses: surface reactivity and behaviour towards endothelial
35
36 cells. Acta Biomaterialia, 2009. 5(4): p. 1211-1222.
37 53. LeGeros, R., et al., Zinc effect on the in vitro formation of calcium phosphates: relevance to clinical
38 inhibition of calculus formation. American journal of dentistry, 1999. 12(2): p. 65-71.
39 54. Nosenko, V.V., et al., Nature of some features in Raman spectra of hydroxyapatite‐containing
40 materials. Journal of Raman Spectroscopy, 2016. 47(6): p. 726-730.
41
55. Patel, K., A. Kumar, and S. Durani, Analysis of the structural consensus of the zinc coordination
42
43 centers of metalloprotein structures. Biochimica et Biophysica Acta (BBA)-Proteins and
44 Proteomics, 2007. 1774(10): p. 1247-1253.
45 56. Feng, Y., et al., Prevention of zinc precipitation with calcium phosphate by casein hydrolysate improves
46 zinc absorption in mouse small intestine ex vivo via a nanoparticle-mediated mechanism. Journal of
47
agricultural and food chemistry, 2019. 68(2): p. 652-659.
48
49 57. Rodzik, A., et al., Study on zinc ions binding to the individual casein fractions: αS1-, β-and κ-casein.
50 Journal of Molecular Structure, 2023. 1272: p. 134251.
51 58. Kambe, T., et al., The physiological, biochemical, and molecular roles of zinc transporters in zinc
52 homeostasis and metabolism. Physiological reviews, 2015.
53
59. Hutmacher, D.W., et al., State of the art and future directions of scaffold‐based bone engineering from
54
55 a biomaterials perspective. Journal of tissue engineering and regenerative medicine, 2007. 1(4): p.
56 245-260.
57 60. Manoochehri, H., et al., Strontium doped bioglass incorporated hydrogel-based scaffold for amplified
58 bone tissue regeneration. Scientific Reports, 2022. 12(1): p. 1-18.
59
60
61
62
63
64
65
 61. Filip, D.G., et al., Current Development in Biomaterials—Hydroxyapatite and Bioglass for Applications
1 in Biomedical Field: A Review. Journal of Functional Biomaterials, 2022. 13(4): p. 248.
2 62. Balamurugan, A., et al., Development and in vitro characterization of sol–gel derived CaO–P2O5–
3 SiO2–ZnO bioglass. Acta biomaterialia, 2007. 3(2): p. 255-262.
4
5 63. Paramita, P., et al., Sol–gel based synthesis and biological properties of zinc integrated nano bioglass
6 ceramics for bone tissue regeneration. Journal of Materials Science: Materials in Medicine, 2021.
7 32(1): p. 1-11.
8 64. Kaneko, J.J., J.W. Harvey, and M.L. Bruss, Clinical biochemistry of domestic animals. 2008:
9 Academic press.
10
11 65. Bozym, R.A., et al., Free zinc ions outside a narrow concentration range are toxic to a variety of cells in
12 vitro. Experimental biology and medicine, 2010. 235(6): p. 741-750.
13 66. Müller, H., Oligodynamic action of 17 different metals on Bacillus subtilis, Enterobacteriaceae,
14 Legionellaceae, Micrococcaceae and Pseudomonas aeruginosa. Zentralblatt fur Bakteriologie,
15 Mikrobiologie und Hygiene. Serie B, Umwelthygiene, Krankenhaushygiene, Arbeitshygiene,
16
praventive Medizin, 1985. 182(1): p. 95-101.
17
18 67. Vert, M., S. Li, and H. Garreau, New insights on the degradation of bioresorbable polymeric devices
19 based on lactic and glycolic acids. Clinical materials, 1992. 10(1-2): p. 3-8.
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
Figure Click here to access/download;Figure;Figure1.jpg
Figure Click here to access/download;Figure;Figure2.jpg
Figure Click here to access/download;Figure;Figure3.jpg
Figure Click here to access/download;Figure;Figure4.jpg
Figure Click here to access/download;Figure;Figure5.jpg
Figure Click here to access/download;Figure;Figure6.jpg
Figure Click here to access/download;Figure;Figure7.jpg
Figure Click here to access/download;Figure;Figure8.jpg
Figure Click here to access/download;Figure;Figure9.jpg
Figure Click here to access/download;Figure;Figure10.jpg
Figure Click here to access/download;Figure;Figure11.jpg
Figure Click here to access/download;Figure;FigureS1.jpg
Figure Click here to access/download;Figure;FigureS2.jpg
Figure Click here to access/download;Figure;FigureS3.jpg
Figure Click here to access/download;Figure;FigureS4.jpg
Graphical Abstract Click here to access/download;Graphical Abstract;Graphical abstract.jpg
Movie/Animation

Click here to access/download

Movie/Animation
MultimediafileM1_video.mp4
Supplementary Files

Click here to access/download

Supplementary Files
Supplementary Information.docx
Supplementary Files

Click here to access/download

Supplementary Files
Figure size.docx
Movie/Animation

Click here to access/download

Video Still
MultimediafileM1_still.jpg