Grafting of Molecules and Nanoparticles onto Bioactive Glass for Added Functionaliy

(1)

GRAFTING OF MOLECULES AND NANOPARTICLES ONTO BIOACTIVE GLASS FOR ADDED FUNCTIONALITY

Faculty of Medicine and Health Technology Master of Science Thesis May 2020

(2)

ABSTRACT

Hanna Hassinen: Grafting of Molecules and Nanoparticles onto Bioactive Glass for Added Func- tionality

Master of Science Thesis Tampere University

Degree Programme of Bioengineering May 2020

The aim of this work was to graft bioactive glass with antibacterial molecules and superparamagnetic iron oxide nanoparticles (SPIONs) for added functionality. Bioactive glasses are promis- ing material not only for bone, but also soft tissue repair. They are bioactive, biocompatible and biomimicking, which makes them an attractive option for current bone regeneration treatments.

Attaching of the antibacterial molecules on the surface as a uniform layer would enhance bioactive glass properties and prevent infection at implantation site. The antibacterial agents should prevent bacterial adhesion and subsequent biofilm formation on the implant surface. An- tibacterial molecule grafting was done in co-operation between the Faculty of Engineering and Natural Sciences and the Faculty of Medicine and Health Technology. The group of Synhetic Chemistry developed powerful antibacterial molecules and modified five molecules for grafting of bioactive glass.

Additionally, bioactive glass could be grafted with SPIONs on the glass surface, which could be utilized as a cancer treatment. The idea is based on SPION properties to generate heat in changing magnetc field. SPIONs need biocompatible coating, and ligands that direct them into cancer cells that could not be removed by surgery. These cancer cells are killed by hyperthermia, whereas bioactive glass act as a construct for surgically removed tissue.

The grafting of antibacterial molecules and SPIONs are two different approaches and their grafting was performed separately. The work is divided into two parts: in the literature review we shall discuss the bone biology and have a short review on infections and bone cancer. Bioactive glasses and their properties are discussed as well. The experimental part includes synthesis of SPIONs. The grafting was attempted first with antibacterial molecules and similar procedures were adapted to SPION grafting. Grafting was attempted with few variations of one method.

Two different bioactive glasses were used: 1393 and S53P4. Both glasses are silicate-based, yet they show different activities under immersion. The glass surface treatment was varied through the experiments to see if it has effect on grafting. Furthermore, grafting environment was varied:

we increased the grafting solution concentration and volume, we attempted grafting at room temperature and at 37 °C, as well as under reduced pressure (vacuum). Treatment time was also varied, however, after repeated tests with and without bioactive glass discs, based on UV-VIS measurements we noticed that molecule is degrading by its own, and therefore grafting was most likely unsuccessful. Moreover, the lack of reliable study methods hinders drawing definite conclu- sions. The glass surface was studied with FTIR and Raman.

SPIONs were successfully synthesized and coated, yet transmission electron microscopy (TEM) revealed too thick silica-coating, which prevents magnetic separation. Hence silica-coated nanoparticles were not taken to grafting. Nonetheless, citric-acid coated SPIONs showed promis- ing results. After several days of drying they formed sticky wax layer on glass discs, and FTIR measurements performed before and after washing indicated rather firm attachment on the glass surface.

Keywords: bioactive glass, nanoparticles, magnetic, biomaterials, induced hyperthermia, antibacterial molecule, S53P4, 1393, added functionality, superparamagnetic, cancer, treatment, grafting, biomedical

The originality of this thesis has been checked using the Turnitin OriginalityCheck service.

(3)

TIIVISTELMÄ

Hanna Hassinen: Antibakteeriaalisten molekyylien ja nanopartikkelien liittäminen bioaktiiviseen lasiin

Diplomityö

Tampereen yliopisto

Biotekniikan koulutusohjelma Toukokuu 2020

Työn tavoitteena oli parantaa bioaktiivisen lasin ominaisuuksia lisäämällä lasin pinnalle bakteereille myrkyllisiä molekyylejä ja superparamagneettisia rautaoksidinanopartikkeleita (engl. superparamagnetic iron oxide nanoparticle, SPION). Bioaktiivisia laseja käytetään luukuddostek- nologisissa implanteissa. Bioaktiiviset lasit ovat kiinnostavia materiaaleja kudosteknologian so- velluksiin, sillä ne eivät aiheuta negatiivista reaktiota implantoinnin jälkeen, vaan ne jopa edistä- vät luukudoksen muodostumista sekä implantin ja kudoksen yhteenkasvamista. Ne muistuttavat myös rakenteeltaan luukudosta.

Bakteereille myrkyllisten aineiden liittäminen lasin pintaan voisi estää leikkauksen jälkeisiä bakteeri-infektioita implantin ympäristössä. Bakteeri-infektio voi johtaa biolfilmin muodostumiseen implantin ympärille, joka myöhemmin estää bioaktiivisen lasin vuorovaikutuksen isäntäkudoksen kanssa. Implantin pinnan käsittelyllä voidaan estää bakteerien adheesio ja siten myös biofilmin muodostuminen. Molekyylien lisäys lasin pinnalle toteutettiin yhteistyössä Synteettisen kemian tutkimusryhmän kanssa. Tuttkimusprojektissaan he kehittivät bakteereille myrkyllisiä molekyylejä, ja muokkasivat näistä viisi tähän työhön sopiviksi.

Molekyylien lisäksi tässä työssä tutkittiin SPIONien lisäämistä lasiin. SPIONeita voidaan hyö- dyntää syöpähoidoissa: bioaktiivisesta lasista tehty implantti toimisi leikkauksessa poistetun luukudoksen korvaajana, ja pinnalle liitetyt SPIONit kohdennettaisiin jäljelle jääneisiin syöpäsoluihin syöpäsoluille spesifeillä ligandeilla. SPIONit tuottavat lämpöenergiaa muuttuvassa magneettiken- tässä, ja täten tappava annos lämpöä voitaisiin kohdentaa syöpäsoluihin.

Työ on jaettu kahteen osaan. Kirjallisuuskatsaus esittelee luukudoksen ja implantteihin liittyvät infektiot, sekä luusyövän. Viimeisenä kappaleena kirjallisuuskatsauksessa esitellään bioaktiiviset lasit ja niiden ominaisuuksia. Kokeellisessa osassa liittämisreaktioita yritetään ensin molekyyleillä käyttäen muutamaa erilaista metodia sekä syntetisoidaan SPIONeita sitruunahappo- ja silikaatti- kuorella. SPIONien liittämistä pintaan kokeiltiin samanlaisilla metodeilla kuin molekyylien.

Työssä käytettiin kahta silikaattipohjaista lasia, jotka tunnetaan lyhenteillä S53P4 ja 1393. Mo- lekyylien lisäys tutkimuksen metodeilla ei näyttänyt tuottavan haluttua tasaista molekyylikerrosta.

Molekyylien tulisi liittyä kovalenttisesti lasin pintaan, ja liittymistä yritettiin parantaa nostamalla reaktion lämpötilaa, käytetyn molekyyliliuoksen konsentraatiota, sekä testaamalla reaktiota alen- netussa paineessa. Tämän lisäksi reaktioaikaa pidennettiin, mutta UV-VIS mittausten perusteella päättelimme, että molekyylit ovat herkkiä hajoamaan liuoksessa, eikä liittyminen tämän vuoksi to- dennäköisesti onnistunut. Tämän lisäksi tutkimusmenetelmät olivat vajavaisia, joten luotettavien johtopäätösten tekeminen on vaikeaa. Lasien pintaa tutkittiin FTIR-spektrometrillä ja Ramanilla.

Työn toinen osio keskittyi SPIONien valmistamiseen ja liittämiseen lasin pinnalle. SPIONien rautaoksidi-ydin valmistettiin onnistuneesti, mutta lopulta piioksidi-päällystys muodostui liian pak- suksi, mikä esti SPION-liuoksen jatkokäsittelyn. Piioksidi-SPIONeja ei siten lopulta käytetty työs- sä, mutta liittämistä kokeiltiin sitruunahappo-SPIONeilla. Tämä menetelmä osoitti rohkaisevia tu- loksia, sillä sitruunahappo-SPIONit eivät huuhtoutuneet lasin pinnalta yhden pesukerran jälkeen FTIR mittauksien perusteella.

Avainsanat: bioaktiivinen lasi, nanopartikkeli, biomateriaali, antibakteriaalinen, molekyyli, S53P4, 1393, superparamagneettinen, syöpä, hoito, biofilmi, tulehdus, infektio, kudosteknologia,

Tämän julkaisun alkuperäisyys on tarkastettu Turnitin OriginalityCheck -ohjelmalla.

(4)

PREFACE

This Master of Science thesis study was performed in Bioceramics, -glasses and -composites at Tampere University. The antibacterial molecule study was done in collaboration with Synthetic Chemistry group of Tampere University.

First, I’d like to thank Associate Professor Jonathan Massera for guidance and valuable comments. I really appreciate that you listened my interests in the beginning and gave me such a interesting topic to work with. I also want to thank Nuno Candeias, who always had time for my questions and explained things with great patience.

Special thank you to Tatu Rimpiläinen, who synthesized the molecules for this work, and helped me. I also want to thank Turkka Salminen for excellent advice regarding analysis methods and keeping my mind focused on research questions. The Wire is waiting for post-thesis time.

Finally, I’d like to thank my friends, who offered their help when needed and were there for me. To my family, thank you for supporting, let’s see what kind of adventures are ahead of me!

Frankfurt am Main, 15th May 2020 Hanna Hassinen

(5)

LIST OF FIGURES

2.1 Bone structure in different levels (Gamagedara & Ziana 2018). . . 4

2.2 Bone fracture healing phases and temporal activity of different cell types. (Shrivats et al. 2014) . . . 5

2.3 Biofilm formation and biofilm prevention. 1) Bacterial colonization and at biomaterial surface and biofilm formation, which protects them from host immune reaction and antibiotics (ATB). 2) If host cells establish irreversible attachment on the biomaterial surface first, the bacterial cells cannot attach. 3) Use of antimicrobial coating/modification can prevent microbial attachment. Adapted from the work by Gallo et al. (2014) . . . 8

3.1 Structures of molecules synthesized for bioactive glass grafting . . . 14

4.1 Magnetic hyperthermia treatment. (Lemine 2019) . . . 15

4.2 Size of nanoparticles compared to other entities (Roy 2016) . . . 16

5.1 Glass-ceramic compositional diagram for bone-bonding (Hench 2006). . . . 22

5.2 Silica network: the role of bridging and non-bridging oxygens and network modifiers and formers. (Stanic 2017). . . 24

5.3 An illustration of reactions on BAG surface leading to HCA formation. 1) Ion exchange; 2) Si-O-Si bond hydrolysis into Si-OH groups; 3) Silica re- polymerization and formation of amorphous calcium phosphate (ACP); 5) ACP crystallizes into HCA (Gunawidjaja et al. 2012) . . . 26

5.4 Chemical structures of coupling agents a) GPTMS and b) APTES. (Karakoy et al. 2014 . . . 31

5.5 APTES bonding with glass surface. (Kyaw et al. 2015) . . . 31

5.6 FTIR measurement on ATR mode. (FT-IR Spectroscopy—Attenuated Total Reflectance (ATR2005) . . . 33

6.1 Surface preparation of glass discs. Polished glass discs were washed with acetone and then 3 times with water. Some glass discs were only washed with acetone and water, whilst majority of them were washed further with buffer. Depending on test, half of glass disc batch remained washed with acetone and/or buffer, while the other half was also silanized. In buffer washing glass discs were divided into three groups and buffer wash was performed with one buffer three times for each group. . . 37

6.2 Drop grafting method, which was used first for grafting, but later also to glass silanization. . . 38

(8)

6.3 Reaction between antibacterial molecule and OH-groups exposed on glass

surface. . . 39

6.4 The immersion technique of molecule grafting: the glass discs were im- mersed completely on the solution, which volume was measured before and after the treatment. The glass discs were washed after the incubation by immersing them into fresh DMSO and sonicating for 5 minutes. . . 42

6.5 The two step process of superparamagnetic nanoparticle synthesis. . . 46

6.6 Dried SPIONs . . . 46

6.7 Aqueous SPION solutions for grafting . . . 48

7.1 FTIR spectra of antibacterial molecules TJRD17 and TJRF21, as well as pure DMSO. . . 50

7.2 1393 glass disc treated with TJRD17 at room temperature for 2h. . . 51

7.3 Untreated glass surface . . . 52

7.4 Microscope images of glass discs treated with 35mM TJRF21 solution for 4 hours at 37°C . . . 53

7.5 Raman spectra of 1393 glass treated with TJRF21 for 6 hours and microscope images of measured spots. . . 54

7.6 Raman spectra of S53P4 APTES silanized glass disc treated with 35mM TJRF21 solution at 37 °C for 6 hours and microscope images of measured spots. . . 55

7.7 Microscope images of buffer washed glass discs treated with TJRF21 at 37 °C for 6 hours. . . 56

7.8 Citrate buffer washed S53P4 glass samples treated with TJRF21 molecule. The spectrum is measured from a colourful ring seen in Figure 7.7a. . . 57

7.9 Microscope images of glass discs treated with pure DMSO, as well as TJRD17 and TJRF21 at 37 ° C for 6 hours. . . 58

7.10 Microscope images of glass discs treated with TJRF45, TJRF33 and TJRF81 treated at 37 ° C for 6 hours . . . 59

7.11 Raman spectra measured from APTES silanized S53P4 glass disc treated with TJRF33 molecule for 6 hours at 37° C. . . 60

7.12 TJRF81 treated APTES silanized S53P4 glass disc measured by Raman. . 61

7.13 Absorbances measured from buffer washed, TJRF21 treated sample. Graft- ing refers to grafting solution collected from a well and washing to washing solution. . . 62

7.14 Amount of TJRF21 molecule before and after 6h treatment in vacuum or 37°C. based on UV-VIS measurements. . . 63

7.15 Microscope image of silanized glass surface. . . 63

7.16 Microscope images of glass discs treated with TJRF21. . . 64

7.17 Molecule consumption of silanized glass discs treated at 37 °C for 3-16h and 6 hours in vacuum based on UV-VIS measurements. . . 65

7.18 Raman spectra of glass disc treated with TJRF21 in vacuum for 6 h. . . 66

7.19 Microscope images of glass discs treated with TJRF81 for 6 hours. . . 67

(9)

7.20 Molecule consumption of silanized glass discs treated at 37 °C for 3-16h

and at room temperature for 6h based on UV-VIS measurements. . . 68

7.21 FTIR spectra of glass disc treated with TJRF81 at 37° for 6h. . . 69

7.22 XRD pattern of uncoated nanoparticle powder. . . 69

7.23 TEM images of synthetized SPIONs. . . 70

7.24 FTIR spectra of SPION grafted S53P4 glass discs . . . 71

(10)

LIST OF TABLES

2.1 Variables related to bacterial adhesion to biomaterials (Campoccia et al.

2013) . . . 9 2.2 Three most common types of primary bone cancer (Picci 2007 B. Widhe &

T. Widhe 2000 Krol et al. 2008) . . . 10 3.1 Systematic names of antimicrobial molecules for bioactive glass project. . . 13 4.1 Comparison between coprecipitation, thermal composition and sol-gel meth-

ods for SPION synthesis. Adapted from the work by Dadfar et al. (2019) . . 19 5.1 Surface reactions of BAG (Hulsen et al. 2017, Hupa 2018) . . . 26 5.2 Bioactive glass compositions presented as weight and mole percentages.

Adapted from the work by Fagerlund et al. (2016) . . . 28 5.3 A summary of possible methods to prevent biofilm formation on BAG im-

plant surface. Adapted from the work by Romano et al. (2015) . . . 29 5.4 Functional groups and their FTIR bands . . . 32 6.1 Composition of bioactive glasses used in the experiments as mol-%. . . 35

(11)

LIST OF SYMBOLS AND ABBREVIATIONS

APTES (3-Aminopropyl)triethoxysilane BAG Bioactive glass

BMP bone morphogenetic protein BO Bridging oxygen

DMSO Dimethyl sulfoxide ECM Extracellular matrix

GPTMS 3-Glycidyloxypropyl)trimethoxysilane Hap Hydroxy apatite

HCA Hydroxyl carbonated apatite MNP Magnetic nanoparticle

MRI Magnetic resosnance imaging NBO Non-bridging oxygen

NP Nanoparticle PEG Poly-ethyl glycolide

SPION Superparamagnetic iron oxide nanoparticle TE Tissue engineering

TEOS Tetraethyl orthosilicate

the FDA The U.S. Food and Drug Administration

TJRD17 4-nitro-2-((5-nitroindolin-1-yl)(p-tolyl)methyl)phenol

TJRF21 4-nitro-2-((5-nitroindolin-1-yl)(4-(1-(3-(triethoxysilyl)propyl)-1H- 1,2,3-triazol-4-yl)phenyl)methyl)phenyl acetate

TJRF33 4-nitro-2-((5-nitroindolin-1-yl)(4-(1-(3-(triethoxysilyl)propyl)-1H- 1,2,3-triazol-4-yl)phenyl)methyl)phenol

TJRF45 4-nitro-2-((5-nitroindolin-1-yl)(4-((1-(3-(triethoxysilyl)propyl)-1H- 1,2,3-triazol-4-yl)methyl)phenyl)methyl)phenol

TJRF81 4-nitro-2-((5-((1-(3-(triethoxysilyl)propyl)-1H-1,2,3-triazol-4- yl)methoxy)indolin-1-yl)(4-(trifluoromethyl)phenyl)methyl)phenol USPION Ultrasmall superparamagntic iron oxide nanoparticle

(12)

1 INTRODUCTION

Critical-sized defects in bone – caused by primary tumor resection, trauma or selective surgery – have proven to be great challenge for current bone repair methods (Porter et al.

2009). Critical-sized bone defects are wounds of a particular bone and animal species that will not heal spontaneously during the lifetime of the animal (Hollinger & Kleinschmidt 1990), and therefore they need to be treated with bone regeneration supporting methods.

The number of bone defect cases is around 1 million per year, and bone defects are a major concern for both the United States and the European Union due to aging population.

(De Santis et al. 2019)

Current method to treat critical-sized bone defects is the use of autografts: host bone from another site is removed and taken to fill the defect site. The method is having high com- plication rate (30%) and may include donor site morbidity, pain paresthesia, prolonged hospitalization and rehabilitation. (Silber et al. 2003) There is also an increased risk of deep infection, hematoma and inflammation, and moreover, the method is associated with restricted availability (Arrington et al. 1996; Banwart et al. 1995).

Another treatment option is the use of allograft that is harvested from another human being (typically a cadaver). Usage of allografts is a FDA approved procedure that has been utilized for years. (Hou et al. 2005) However, allografts and autografts rely on physical and biological similarity between donor (site or patient) and recipient. Allografts carry a risk of infection, the rate is as high as 13%, as well as disease transmission and host immune responses. Moreover, they are scarcely available and require life-long medication to prevent the rejection reaction in the host. (Mankin et al. 2005; Nishida & Shimamura 2008) Beside autografts and allografts, one option is the use of xenografts harvested from a non-human species. Xenografts, like allografts, are associated with a high risk of infection, disease transmission and host immune response. Moreover, their use can be against religious belief. (Yang & Sykes 2007)

To overcome these difficulties faced with current methods, bioactive glasses (BAGs) have raised great interest (Hench & Polak 2002; Hollister 2005). The first version was invented already 50 years ago by Larry Hench. They are a synthetic material, that utilize the body’s natural biological response to tissue damage. They are described as biocompatible, biodegradable, and biomimic, which allows the native bone tissue to interact and integrate with BAG. Moreover, BAGs imitate the multidimensional hierarchical structure of native bone. BAGs are accepted as a treatment method, and they could compete with the actual golden standard. (Hench 2006; Hench, Roki, et al. 2014)

(13)

Ideal synthetic tissue engineered (TE) scaffold should be capable of mimicking physio- chemical environment of tissue, and at the same time be biodegradable as native tissue integrates into it. Moreover, scaffold should actively promote desirable physiological reactions while preventing undesirable ones. (Kretlow & Mikos 2007)

To meet these requirements, a synthetic bone scaffold must (Porter et al. 2009)

1. support mechanically the affected area for a short time period and degrade control- lably as load is tranferred to developing bone,

2. degrade a non-toxic manner,

3. enhance and provide a substrate for osteoid deposition and a surface bone cell migration

4. accomodate vascularization and bone in-growth by a porous structure

5. induce osteogenesis i.e. support and promote osteogenic differentiation de novo (osteoinduction)

6. enhance the integration of scaffold and host tissue by increasing cellular activity 7. not activete chronic inflammatory response and

8. sterilizabale without loss of bioactivity.

Ideally, a bone tissue scaffold should also be capable of eluting bioactive molecules, such as growth factors to ease tissue integration, or drugs, such as antibiotics to prevent negative biological reaction like sepsis, or cancer treatment agents for cancer recurrence (Porter et al. 2009). One bioactive glass type, S53P4, has the inherent property of in- hibiting bacterial growth, however, the antibacterial property is linked to steep pH change within small particle size, which cannot be achieved in TE scaffolds (Drago et al. 2018).

Moreover, most of the bioactive glasses do not have such antibacterial property, and therefore bioactive glass needs to be functionalized.

The aim in this thesis is to study if antibacterial molecules or superparamagnetic iron oxide nanoparticle (SPION) could be grafted onto BAG. The grafted antibacterial molecules could solve implant related infection issues, and SPIONs could overcome drawbacks faced in current cancer treatment methods. After successful grafting, there should be evenly distributed layer of antibacterial molecules / SPIONs on the surface. The grafting can be studied directly from the glass surface with Fourier-transform infrared spectroscopy (FTIR) and Raman spectroscopy, and undirectly from used molecule solutions with the ultraviolet-visible (UV-VIS) spectroscopy. We attempted similar surface treatment methods for BAGs as Zhang et al. (2013a) and (2013b) used in their studies for gallic acid and polyphenol grafting. Verne et al. (2009) studied bone morphogenic protein grafting, and we adapted one of their silanization method in this work. There has not been studies related to iron oxide nanoparticle grafting on BAG surface, so similar methods are attempted in SPION grafting.

The work also includes SPION synthesis. Several studies have described efficient methods to synthesize SPIONs without using toxic chemicals. (Campelj et al. 2008; Dadfar et

(14)

al. 2019; Sodipo & Aziz 2016) We adapted a coprecipitaion method described by Campelj et al (2008). Their synthesis method was used in a study by Kralj et al. (2010), in which they coated SPIONs with silica. Our intial aim was to study silica coated SPION grafting on BAG surface, and later we attempted also citric acid (CA) coated SPIONs for grafting as well.

In the next chapters, we shall discuss basics of bone biology and implant related infections, as well as bone cancer. After that, there are short introductions to antibacterial molecules and magnetic nanoparticles, and in the last chapter of literature review BAGs are discussed. The experimental part is described in materials and methods, followed by the results and the discussion. Finally, we shall have conclusion chapter focusing on the limitations and sugggestions how the work could be improved.

(15)

2 BONE TISSUE ENGINEERING

2.1 Bone tissue

Bone is a complex organ, which provides many functions for human body. It enables movements, offers protection and support to more sensitive and critical organs while serving as a mineral storage and housing multiple progenitor cells, just to name few.

(Athanasiou et al. 2000)

Bone tissue can be divided into two categories based on their physical appearance:

dense cortical bone and spongy/cancellous bone. Cortical bone porosity varies between 5-10% whereas cancellous bone porosity can be up to 90%. The difference is explained by their structure: cortical bone consists closely packed osteons and canals for bone cells, blood vessels, nerves and lymphatic vessels, whereas spongy bone is built of trabecular network of single trabeculae. (Porter et al. 2009; Gamagedara & Ziana 2018) These two types, however, have similar microstructure, mineralized collagen fibril (Fig.

2.1) and bone cells. Bones are around 70% inorganic, and most of it is found in the form of hydroxyapatite (HA) crystals [Ca₁₀(PO₄)₆(OH)₂] (Figure 2.1). There are also organic components present, of which the majority is type I collagen (90% of organic content).

(Athanasiou et al. 2000) Moreover, other organic components of bone tissue include other collagen types, proteoglycans, glycoproteins, bone specific vitamin K-dependent proteins and growth factors and cytokines (Porter et al. 2009).

Figure 2.1. Bone structure in different levels (Gamagedara & Ziana 2018).

(16)

There are several bone cell types. Osteoprogenitors are pre-osteoblast cells committed to differentiate into bone cells. They are derived from mesenchymal stem cells. There are also osteoblasts, osteocytes and osteoclasts. Osteoblasts are bone builders, which secrete bone matrix turning into osteocytes once they are surrounded by the secreted matrix. Alongside bone forming cells, there are bone resorbing cells called osteoclasts.

They become active when bone tissue is remodeled or damaged. (Porter et al. 2009) Beside physical appearance, cortical and cancellous bone have different functions: cortical bone forms the exterior part of bones providing strength, whereas porous cancellous bone is found interior of the bones. This cancellous bone forms continuous trabecular network, which holds bone marrow and blood vessels. Bone marrow is divided into red and yellow blood marrow, of which the first one is the development site for blood cells and prominent in young people, and the second one containing mostly fat cells and being abundant in adults. (Athanasiou et al. 2000)

2.2 Bone regeneration and osteointegration

Process of fracture healing can be divided into 4 phases: inflammatory (0-3 days), proliferation (3-7 days), soft callus formation (1-4 weeks) and hard callus formation (1-6 months) (Shrivats et al. 2014). These phases are presented in Figure 2.2.

In the inflammation phase, a hematoma or a blood clot forms at the fracture site. The acute inflammation activates the non-specific immune system including platelets, leuko- cytes, macrophages, growth factors and cytokines delivered by blood and post-capillary venues. (Shrivats et al. 2014; Mavrogenis et al. 2009) This inflammation phase is often

Figure 2.2. Bone fracture healing phases and temporal activity of different cell types.

(Shrivats et al. 2014)

(17)

referred as a destructive phase, which also characterized by local hypoxia (Mavrogenis et al. 2009).

The implant is in reached by blood proteins and interstitial fluids, which are adsorbed on the surface. (Mavrogenis et al. 2009) The protein layer guides inflammation process, and subsequent healing, which makes the implant surface chemistry fundamentally important in terms of implant successfulness (Jimbo et al. 2010). These first interfacial reactions of BAG and bone are discussed in depth in section 5.2 explaining the formation of silica- rich layer and subsequent formation of hydroxyl carbonated apatite (HCA) layer. These are the bioactive layers playing important role on protein adsorption. (Hench, Roki, et al.

2014)

Hypoxia induces angiogenesis i.e. the formation of new vasculature. This leads to the second phase of the fracture healing process, as mesenchymal stem cells are recruited to the site. In the bone regeneration process biological cues, such as bone morphogenetic proteins (BMPs), guide the stem cells to differentiate into chondrocytes and osteoblasts.

First, chondrocytes build cartilage matrix (soft callus), which is substituted by osteoblasts bone formation (hard callus). Finally, chondrocytes will enter apoptosis (programmed cell death), whereas osteoblasts and osteoclasts remodel bone to a physiological status, which is indistinguishable from the bone tissue before fracture. (Shrivats et al. 2014) These steps are seen when the implantation and time after it has been ideal. Next, we are focusing on what happens if there is bacterial infection before any interaction of blood cells and biomaterial has taken place.

2.3 Implant infections

As the insertion of implants and medical devices has emerged as a common and often life-saving procedure, also the number of infections related to implantation procedures has been growing. It has been estimated that the total rate of infection after joint replace- ment is around 1-3% being highest within patients suffering osteoarthritis. (Jarvis 1996;

Arciola et al. 2015) Even though the rate is rather low, there are approximately one million hip replacements and 250,000 knee replacements in the world in a year. With such volumes, infections at implant interfaces become remarkable problem, not to mention the pain and hazard which it causes to the patients. The implant infection can result in chronic infection or tissue necrosis. (Schierholz & Beuth 2001; Song et al. 2013)

The infections can be classified based on the route of the infection (Ribeiro et al. 2012):

• perioperative: bacterial emergence through the surgical side or immediately there- after,

• hematogenous: through blood or the lymphatic system from a distant focus of infection and

• contiguous: spread from an adjacent focus of infection (e.g. pre-existing infection, skin lesions)

(18)

Or according to onset of symptoms after implantation:

• early: within 3 months after surgery,

• delayed: 3-24 months after surgery, and

• late: more than 24 months from surgery,

the highest incidence of infections occurs during the first two years from the surgery.

(Zimmerli et al. 2004)

Studies have shown that there are multiple bacterial species causing prosthesis related infections, such as S. aureus, including methicillin-resistant strain (MRSA), coagulase- negative Staphylococci (CNS) (e.g. S. epidermidis, S. haemolyticus, S. hominis, S.

warneri), Propionibacterium acnes, P. aeruginosa, Haemophilus influenzae, Providencia, Enterococci, Streptococcus viridans, Escherichia coli, Citrobacter, Lactobacillus, Acine- tobacter, Serratia marcescens, Klebsiella pneumoniae, and Corynebacterium. (Song et al. 2013)

The most common bacteria responsible for prosthesis-related infections are S. aureus and coagulase- negativeStaphylococci of above-mentioned species. Approximately half of the infections are related to them (Arciola et al. 2015). There is also linkage between bacterial species and the on-set of symptoms: early infection are in most of the cases caused by virulent microorganisms like S. aureus, while delayed infections are usually caused by low virulent microorganisms such as coagulase-negativeStaphylococci. (Song et al. 2013)

The infection starts with few bacterial cells inhabiting the biomaterial surface. They ad- here, proliferate and secrete extracellular matrix containing polysaccharides. This accu- mulation of micro-organisms in matrix is called a biofilm, and they account for over 80% of microbial infections in the body including prostheses and internal fixation devices. (Song et al. 2013; Arciola et al. 2015) Inside the biofilm bacteria activate a cell-density dependent mechanism (quorum sensing, QS), which affect bacteria’s adhesion mechanisms and expression of virulent factors (Arciola et al. 2015). The biofilm formation is presented in Figure 2.3 (1). Since the bacterial cells are now under protection of their self-made matrix and changed physiology, the immune system struggles to eliminate the pathogens.

The same applies to antibiotics, which leads the biofilm infection turning chronic. Stan- dard therapy for biofilm infected implant is removal of implant. (Schierholz & Beuth 2001;

Song et al. 2013))

(19)

Figure 2.3. Biofilm formation and biofilm prevention. 1) Bacterial colonization and at biomaterial surface and biofilm formation, which protects them from host immune reaction and antibiotics (ATB). 2) If host cells establish irreversible attachment on the biomaterial surface first, the bacterial cells cannot attach. 3) Use of antimicrobial coating/modification can prevent microbial attachment. Adapted from the work by Gallo et al. (2014)

Even though infection caused by the bacterial biofilm elicit less immune response, it still hampers the interactions between an implant and a host tissue. After maturation, some bacteria are able to leave the biofilm in planktonic state, which might cause infection in other parts of the body as well. (Arciola et al. 2015) However, the bacterial cells are not able to inhabit the surface after the host’s cells have attached into it. This short time scale is called window of opportunity in Figure 2.3 (1) and the prevention of bacterial adhesion below it.

Biomaterial surface could be treated so that bacterial attachment is not possible (Fig.

2.3 (3)), and it can be achieved by coating the implant with antibacterial molecules or other surface treatment methods, which are discussed in the section 5.3 Many of these methods rely on affecting bacterial attachment, and variables affecting the adhesion are collected on Table 2.1. Antibacterial coating could be released in a sustained way into the local micro-environment of implants. With such method systemic side-effects could be prevented, and moreover, exceed usual systemic concentrations by several order of magnitude. (Schierholz & Beuth 2001)

(20)

Table 2.1. Variables related to bacterial adhesion to biomaterials (Campoccia et al. 2013)

Surface morphometry Macroporosity Microporosity Micro-roughness Nano-roughness

Physico-chemical properties Surface energy Hydrophilicity Hydrophobicity

Hydrophobic functional groups Polar functional groups

Charged functional groups

Functional groups with specific activities Degree of hydration

Environmental conditions Electrolytes pH

Temperature

Host proteins/host adhesins Shear rate/fluid viscosity Fluid flow rate

Pathogens Gram-positive/-negative

Genus/ Species Bacterial shape Surface energy Strain type

Specific set of expressed adhesins

(21)

2.4 Cancerous bone tissue

Another life-threatening condition of bone disease is bone cancer. Estimates for cancer of the bones and joints for 2020 are around 3600 new cases with 78-82% survival rate in the USA. (Siegel et al. 2020)

Cancer begins with mutations in DNA. Accumulated mutations in specific DNA sequences result in the continual unregulated proliferation of cancer cells. Cancer cells do not re- spond appropriately to the signals that are used to control cell behavior resulting in a tumor of poorly formed cells, that can invade and destroy healthy tissue. Cancer cells can eventually spread (metastasize) throughout the body as well. There are benign and malignant tumors, of which latter is aggressive and have higher risk of growing and spreading, therefore needing immediate medical attention. (Cooper 2000)

Bone cancer can be divided into primary and secondary bone cancer, of which the first has begun from the bone tissue, whereas the second has developed in another part of the body and metastasized in bone tissue. Treatment of secondary bone cancer aims controlling the cancer rather than curing it. (Waldt 2016) Therefore the condition is out of the scope of this thesis.

There are three main types of malignant primary bone cancer: osteosarcoma, Ewing sarcoma, and chondrosarcoma. As seen in Table 2.2, the patients suffering these conditions usually receive solely or a combination of surgery, chemotherapy, or radiation therapy.

In surgical procedure, a section of cancerous tissue is removed. Critical sized excision should be reconstructed. (Krol et al. 2008; Picci 2007)

Chemotherapy relies on powerful chemicals that affect fast growing cells. Even though chemotherapy is an effective way to treat cancer, it also has systemic effect and side effects such hair loss, damage to other tissues and immune system depending on the used drug. (Schirrmacher 2019) Radiation therapy utilizes beams of intense energy to Table 2.2. Three most common types of primary bone cancer (Picci 2007 B. Widhe &

T. Widhe 2000 Krol et al. 2008)

Cancer type Tissues / cells affected Treatment

(sole/ combination)

Osteosarcoma

direct formation of immature bone or osteoid tissue

soft tissues near bone (rarely)

chemotherapy surgery Ewing sarcoma the pelvis, long bones

several soft tissue types

chemotherapy radiation therapy surgery

Chondrosarcoma

excessive growth of cartilage pelvis, hip, shoulder

in rare cases the base of the skull soft tissue near bones

surgery

(22)

kill cancer cells. Most often these are X-rays, but also other types of energy can be used.

The beam is aimed at a precise point of the body (tumor), destroying as little as possible healthy cells. (Yagawa et al. 2017) However, there is always a risk of damaging healthy cells. These treatment methods have also long-term side effects, confining patients to long term control. (Cooper 2000)

Solely used radical surgical treatments fail in about 85-90% of patients due to formation of micro-metastases in high grade osteosarcoma. This rate can be improved by utilizing combination therapy to fight micro-metastases. One option is to utilize traditional chemotherapy, yet as mentioned above, it bears side-effects. (Picci 2007) Therefore there is a need for new kind of treatment methods, such like magnetically induced hyperthermia, which is discussed in Chapter 4.

(23)

3 ANTIBACTERIAL MOLECULES

Antibacterial molecules are an attractive option to enhance biomaterial properties. In- fections of implants are caused by several types of bacteria, which can form a capsule around the tissue implant and hinder interaction between the implant and the host tissue, as described in the section 2.3. This could be prevented by antibacterial molecules, which inhibit bacterial growth on BAG surface. (Gristina et al. 1988)

Polyphenols have gathered great interest due to their antibacterial, antioxidant, anti- cancer and bone stimulation properties. (Quideau et al. 2011; Bravo 1998) However, their limitations include relatively low stability and poor bioavailability when administrated through systemic routes. They are sensible to light, temperature and pH changes. How- ever, they are stable in acidic environment, which is the case for example on the ini- tial stage of bone healing process. (Friedman & Jürgens 2000; Cazzola et al. 2016;

Hazehara-Kunitomo et al. 2019)

Moreover, currently used antibiotics are powerless against multidrug-resistant bacteria.

These multidrug-resistant bacteria are causing most of nosocomial infections and are great threat to health care, including TE implants. Therefore polyphenols, which have not been used in their whole potential, could be utilized in preventing infections. (Zaman et al.

2017; O’Neill et al. 2016) Their limitations of bioavailability could be solved by delivering them locally to the site of need. As they are needed primarily on the BAG surface after implantation, they should be immobilized on the surface as a uniform coating. (Hasan et al. 2013) These methods are discussed in section 5.3 in detail.

Polyphenols can be obtained from low-cost sources, such as grapes and wine by-products (El Gharras 2009). They can also be synthesized, as in a cooperation project between two research groups of Synthetic Chemistry (Tampere, Finland and Lisboa, Portugal).

They produced powerful amino-alkylphenols through Petasis borono-Mannich multicom- ponent reaction and the synthesized molecules were found to inhibit the visible growth of a microorganism after overnight incubation even with low concentrations. (Rimpiläinen, Andrade, et al. 2018)

These amino-alkylphenols were studied with gram-positive bacteria, such as Staphylo- coccus aureus and Enterococcus faecalis, which are main pathogens responsible for infections. The most active amino-alkylphenol, 2-((4-Chlorophenyl)(indolin-1-yl)methyl)- 4-nitrophenol, was selectively active against these strains, yet only moderately active against S. pneumoniae and S. agalactiae. Further studies led to the refinement of

(24)

Table 3.1. Systematic names of antimicrobial molecules for bioactive glass project.

Name Systematic name

TJRD17 4-nitro-2-((5-nitroindolin-1-yl)(p-tolyl)methyl)phenol

TJRF21 4-nitro-2-((5-nitroindolin-1-yl)(4-(1-(3-(triethoxysilyl)propyl)-1H-1,2,3-triazol-4- yl)phenyl)methyl)phenyl acetate

TJRF33 4-nitro-2-((5-nitroindolin-1-yl)(4-(1-(3-(triethoxysilyl)propyl)-1H-1,2,3-triazol-4- yl)phenyl)methyl)phenol

TJRF45 4-nitro-2-((5-nitroindolin-1-yl)(4-((1-(3-(triethoxysilyl)propyl)-1H-1,2,3-triazol-4- yl)methyl)phenyl)methyl)phenol

TJRF81 4-nitro-2-((5-((1-(3-(triethoxysilyl)propyl)-1H-1,2,3-triazol-4-yl)methoxy)indolin- 1-yl)(4-(trifluoromethyl)phenyl)methyl)phenol

the molecular structure, resulting in a more active aminoalkylphenols, namely TJRD17.

(Rimpiläinen, Nunes, et al. 2020)

The phenol moiety was found to be essential to achieve the antibacterial properties, but they were not found to be related to antioxidant properties of these compounds. More- over, quinone methides were found to be highly reactive, yet the antibacterial mode of these compounds remains uncertain. (Rimpiläinen, Andrade, et al. 2018) The other properties (anti-tumoral, antioxidant etc.) related to polyphenols were not reported of these molecules. The systematic names of antibacterial molecules are presented in Table 3.1 and the structure of the molecules in Figure 3.1.

TJRD17 is a starting point in our study, as it is a antibacterial molecule without triethoxy group in its structure. The four other molecules, related to TJRD17, were derivatized to introduce a triethoxysilane moiety that serves as functionality for glass modification. The presence of the triethoxysilane would enhance the molecule reactivity with OH-groups on the glass surface, allowing the use of grafting procedures as described for APTES - a commonly used coupling agent in bioactive glass grafting (Verne et al. 2009; Ferraris

& Verne 2016), that contains the same triethoxysilane and three-carbon chain moiety (further discussed in section 5.3). The APTES-like unit is attached to the core of the molecule via a triazole linker, which is likely to have an additional role of serving as a cleaving site for the release of the antibacterial molecule.

(25)

Figure 3.1.Structures of molecules synthesized for bioactive glass grafting

(26)

4 MAGNETIC NANOPARTICLES

The current cancer treatments face systemic effect of drug administration, lack of drug specificity, large drug dose, non-specific toxicity and other adverse events. (Schirrmacher 2019; Yagawa et al. 2017; Cooper 2000))

However, magnetic nanoparticles (MNPs) could offer a novel treatment option without these problems, and MNPs are already in clinical use: they are accepted as contrast agents for magnetic resonance imaging (MRI) (Chaughule et al. 2012; Dadfar et al. 2019)) and they are used as a cancer treatment method as well. The treatment is based on local administration: SPIONs are injected inside tumor, and the patient is taken into MRI, where alternating magnetic field induces particles to rotate and to generate heat delivering toxic amount of thermal energy to cancer cells. The treatment idea is presented in Figure 4.1.

It is currently used for brain tumors (Thiesen & Jordan 2008), yet clinical trials for prostate cancer are being conducted as well (Winter et al. 2018).

This thesis focuses on a bit different treatment idea: SPIONs would be grafted onto bioactive glass and the glass-delivery system is implanted to the tumor site. As described in section 2.4 the current cancer treatment options include surgeries, and one of their drawbacks is the need for a replacing construct for lost tissue. (Cooper 2000) So, this

Figure 4.1. Magnetic hyperthermia treatment. (Lemine 2019)

(27)

treatment would provide a scaffold in a place of lost tissue and magnetic hyperthermia would be directed to cancer cells, which could not be removed in the surgery. This could be achieved by careful selection of functional groups/ligands on SPION surface. Even though functionalization of SPIONS – as vital as it is for targeting SPIONs to specific cells– is out of the scope of this thesis. Functionalization offer precise and specific elimi- nation (or illumination) of cancer cells when specific ligands are used to guide SPIONs to tumor site. (Dadfar et al. 2019; A. Gupta & M. Gupta 2005))

Especially iron oxide nanoparticles (IONPs), which are ultra-fine units with diameter of 1- 100 nm, are studied extensively for medical applications due to their versatile properties such as superparamagnetism, high magnetic moment, magnetothermal effect and ideal biocompatibility. Superparamagnetic nanoparticles are made of ferromagnetic crystals and due to their nano-sized nature, they have zero magnetism in the absence of external magnetic field (that is, superparamagnetism). (Mahmoudi et al. 2011; Xie et al. 2018)).

One reason for their attractiveness for application is their easy synthesis, magnetically governable properties, and controllable sizes. Their size compared to biological systems and visible objects is presented in Figure 4.2. They are smaller or equal to biological systems, such as cell (10-100 um), virus (20-450 nm), protein (5-50 nm) and gene (2nm wide, 10-100 nm long). Due to their submicroscopic size, nanoparticles possess unique capabilities and negligible side effects. (Mahmoudi et al. 2011)

Ferrous and ferric iron oxides present seven crystalline phases, the most common are

• a-Fe₂O₃ (hematite)

• g-Fe2O3 (maghemite)

• Fe₃O₄ (magnetite)

• Fe₁O_x (wustite)

and the less commonly found are theb- and e-Fe2O3 phases and the low-temperature rhombohedral structure of magnetite. Of these crystalline phases, the magnetite (and

Figure 4.2.Size of nanoparticles compared to other entities (Roy 2016)

(28)

less maghemite) have gathered great interest due to their superparamagnetic properties.

(A. Gupta & M. Gupta 2005; Xie et al. 2018; Veiseh et al. 2010)) Magnetite has been shown great potential, and its biocompatibility has already been proven (Schwertmann

& Cornell 2008) Moreover, SPIONs are favored for their reactive surface, which can be easily modified with biocompatible coatings and well as functionalized for targeted drug delivery. (Dadfar et al. 2019) Coating of SPIONs is discussed in section 4.3.

4.1 Biocompatibility of SPIONs

SPIONs have proven to be advantageous in a variety of applications due to their good biocompatibility and safety (Veiseh et al. 2010). Currently, magnetic nanoparticles (MNPs) are used in in vitro and in vivo applications: magnetic nanoparticles can be utilized in bioseparations and enzyme immobilization, as well as in diagnostic applications as contrast agents. Nevertheless, they could be used in therapies also, and their use in targeted drug delivery, tissue repairs and hyperthermia has been studied in great detail. (Laurent et al. 2008; Xie et al. 2018; Veiseh et al. 2010) These biomedical and bioengineering applications require small particle size with high magnetization values and narrow size distribution. (Xie et al. 2018) Since we are interested in cancer treatment, the emphasis is on characteristics and properties that are advantageous for magnetic hyperthermia.

Biocompatibility of SPIONs is affected by shape, size, surface properties and structure.

(Xie et al. 2018; Veiseh et al. 2010) Nanotoxicology is an emerging area of research and not all aspects of it are fully understood. (Elsaesser & Howard 2012)

4.1.1 Size of SPIONs

The IONPs can be divided into two groups according to their hydrodynamic size (including coating):

• SPION greater than 50 nm

• SPION smaller than 50 nm, which are called ultra-small superparamagnetic iron oxide nanoparticles (USPION)

In this thesis the focus is on USPIONs, since they are advantageous for magnetic hyperthermia. (Di Marco et al. 2007, Sodipo & Aziz 2016) Small size and narrow size distribution are required for magnetic hyperthermia to have particles with uniform physical and chemical properties. However, these properties also lead to tendency to agglomeration of SPIONs. Therefore, a biocompatible coating is needed, which is preferably functionalized to targeted delivery to cancer cells. (A. Gupta & M. Gupta 2005)

Superparamagnetism is defined as magnetism observed under magnetic field, whereas no magnetism remains after removal of magnetic field. (Xie et al. 2018). There are several estimations for size range of IONPs to have superparamagnetic properties. Some sources claim the IONPs smaller than 30nm as a maximum of core size (Kandasamy &

Maity 2015). Nevertheless, some sources claim that the maghemite and magnetite single

(29)

domains are about 5-20nm in diameter (A. Gupta & M. Gupta 2005), and particles below a critical particle size (15nm) consist a single magnetic domain, which in other words, leads to uniform magnetization and high saturation magnetization values and hence superparamagnetism (Chatterjee et al. 2003).

On the other hand, the hydrodynamic size have effect on NPs clearance from the circulation. Small NPs (< 20 nm) are excreted renally, whereas medium sized NPs (30–150 nm) accumulate in the bone marrow, heart, kidney and stomach, whereas large NPs gather in the liver and spleen. (Elsaesser & Howard 2012)

4.1.2 Shape and structure

SPIONs are not typically excreted from the body as a construct, and therefore the use of different components should be carefully considered. SPION toxicity is should be considered by its structure, shape, components, and end-products inside the body. (Veiseh et al. 2010)

SPIONs can be manufactured to multiple shapes such as nanooctahedrons, nanorods, nanocubes, nanohexagons, nanowires, nanotubes, nanoplates, and nanocapsules. How- ever, only a limited number of comparative studies have been performed to evaluate biodistribution of spherical, non-spherical and rod-shaped nanoparticles. Geng et al (2007) found in their study, that there is a relationship between an increase in the length- to-width aspect ratio of a nanostructure to increased blood circulation time in vivo. En- hanced blood circulation time seems to hold true with the high aspect ratio compared to spherical nanoparticles. (Park, Maltzahn, et al. 2008; Park, von Maltzahn, et al. 2009)) At least nanooctahedrons, nanorods and nanocubes with different morphologies were found to be quite safe to use in the concentration range of 10-100 mg/ml in an in vitro study. (Zhou et al. 2012) Moreover, concentration of SPIONs is a key factor for cell viability: Ankamwar et al. (2010) studied cytotoxicity as a function of concentration from low (0.1mg/ml) to high concentration (100mg/ml) using human glial, breast cancer and normal cell lines. Concentrations below 100mg/ml showed great biocompatibility whereas higher concentration could result in noticeable cytotoxicity, yet this was also dependable from the cell line. (Ankamwar et al. 2010)

The structure represents the shape, size and spatial distribution of mineral aggregations.

Several studies have been conducted to study IONPs structures including aggregates, mesopores, magnetic micro needles, clusters and hollow microspheres. (Xie et al. 2018) In general, IONPS aggregates greater than 400nm seem to arise necrotic mechanism in tested cell lines (Sadhukha et al. 2014), whereas with other studies found that magnetic particles with different structures have low cytotoxicity under controlled concentration (Kavaldzhiev et al. 2017; J. Yang et al. 2017).

(30)

4.2 Synthesis of SPIONs

SPIONs can be obtained either by biological, physical or chemical synthesis. The chemical route is the most feasible: they have low production cost and high yield (Dadfar et al. 2019; Xie et al. 2018. With chemical routes, it is possible to obtain SPIONs with uniform composition and size (Willard et al. 2004) Physical approaches include powder ball milling, electron beam lithography, and their drawback is a lack of nanoparticle size control. Biological methods employ reduction-oxidation reactions involving microbial en- zymes or plant phytochemicals. Such method is considered eco-friendly and products show good biocompatibility, yet the method suffer from low yield and broad size distribution. (Dadfar et al. 2019)

Several chemical synthesis methods are listed in Table 4.1. In these methods, uniform sized SPIONs are obtained via homogenous precipitation reactions. This is achieved by conditions that allow a short nucleation reaction followed by a slower growth phase.

(Schwertmann & Cornell 2008) Nucleation requires supersaturated solution, which will lead to nucleation burst. Nucleation phase is over and concentration drops, and the formed nuclei will grow as solutes diffuse onto the nuclear surface from the solution.

These two steps need to be separated in order to have uniform size and structure. (Lodhia et al. 2010)

While each method has its advantages, the most common preparation method is coprecipitation of Fe²⁺/Fe³⁺ salt solutions by adding a base. If the focus is on production of magnetite SPIONs, they should be synthesized under an inert atmosphere, otherwise formation of iron hydroxides might hamper end-product properties (A. Gupta & M. Gupta 2005). Properties of the synthesized SPIONs depend on the type of employed salts (e.g.

sulfates, nitrates etc.), ratio of Fe²⁺/Fe³⁺, temperature, pH and ionic strength. (Dadfar et al. 2019; Lodhia et al. 2010)

Table 4.1. Comparison between coprecipitation, thermal composition and sol-gel methods for SPION synthesis. Adapted from the work by Dadfar et al. (2019)

Method Co-precipitation Thermal decom- position

Sol-gel

Synthesis procedure Very simple Very complicated Relatively simple

Process temperature Low High Low

Process time Minutes Hours-days Hours

Size distribution Broad Vary narrow Relatively narrow

Shape control Bad Very good Good

Morphology Irregular sphere Cube- sphere Porous or non- porous sphere

Degree of crystallinity Low High Low

Yield High High Medium

Scalable Yes Yes Yes

(31)

However, standard co-precipitation methods struggle to yield consistent SPION size, shape and polydispersity. Co-precipitation may also result in impurities and surface defects in the particles, violating their magnetic properties. (A. Gupta & M. Gupta 2005) Nevertheless, as co-precipitation is easy to implement and do not require high temperature nor pressure, it has been considered an efficient method to produce SPIONs. Fur- thermore, less hazardous materials are needed in coprecipitation than in the other techniques, and therefore it is usually favored over others. (Veiseh et al. 2010; A. Gupta &

M. Gupta 2005)

To achieve narrow size distribution, one should select thermal decomposition as a synthesis method. High control over size and shape, as well as high crystallinity can be achieved with thermal decomposition. The method relies on organoiron precursors in high-boiling point organic solvent within stabilizing surfactants. Use of surfactants allow the size control and obtaining well dispersed, crystalline samples. The method is considered unfriendly to environment, since it requires toxic chemicals such as chloroform, hex- ane, and iron pentacarbonyl. Moreover, for biomedical applications the SPIONs produced via thermal decomposition require additional modification step to obtain water-dispersible and biocompatible SPIONs. The surface of SPIONS produced by thermal composition have a hydrophobic coating, which should be replaced by hydrophilic one. (Dadfar et al.

2019)

One commonly employed method to produce silica-coated nanoparticles is sol-gel technique. The method is based on colloidal solution (sol) acting as a precursor for a network of discrete paricles (gel). Metal precursor hydrolyses rapidly and a metal hydroxide solution is formed. This is followed by a condensation and formation of gel, which is later dried. The gel need to be crushed and solvent to be removed. As a result, monodis- persed and relatively large SPIONs are obtained. The method suffers from by-product contaminations and require post-treatment purification. Moreover, further heat treatment is required to achieve crystalline structures. (Dadfar et al. 2019)

4.3 Surface properties and coating of SPIONs

The uncoated IONPs are relatively inert, which hinders interactions with biomolecules and/or other materials. The interactions are needed, however, since IONPs should stay unrecognized from immune system. Moreover, they should be functionalized in order to target to cancer cells and their surface properties affect their grafting onto biomaterials.

Lastly, uncoated SPIONs tend to agglomerate in colloidal suspension. Therefore, coating of SPIONs is a key factor for successful SPION utilization. (Veiseh et al. 2010)

By altering surface charge and hydrophobicity of SPIONs, the biodistribution can be affected. Interactions with the adaptive immune system, plasma proteins, extracellular ma- trices and non-targeted cells can be either enhanced or limited by altering these properties. (Davis 2002). Hydrophobic and charged nanoparticles tend to have short circulation times due to opsonization i.e. adsorption of plasma proteins. Positively charged nanopar-

(32)

ticles may bind with non-targeted cells leading to non-specific internalization. (Xie et al.

2018, Chouly et al. 1996)

Several polymers have been tested for coating to enhance biocompatibility and stability in colloidal suspensions. The most interesting organic coating materials are polyethylene glycol (PEG), dextran and chitosan. (Cole et al. 2011, Remya et al. 2016, Mohammadi- Samani et al. 2013) There have been also studies related to inorganic coating with materials such as silica. This material could provide not only biocompatibility, but also glass compatibility. Silica coating can be acquired by Stöber sol-gel process (Deng et al. 2005) or via microemulsions, which lead to polycondensation of TEOS. (Santra et al. 2001) The coating is usually aimed to be continuous, homogenous and as thin as possible, since the coating is non-magnetic shell, it will affect SPION magnetic properties (Kandasamy

& Maity 2015). Gomez-Lopera et al. (Gomez-Lopera et al. 2001) found that a polymer coating of poly(lactide-co-glycolide) affected magnetization values of particles to about one-half compared to pure magnetite. Kralj et al. (2010) studied coating of SPIONs with silica, and they reported that even with a monolayer of silica on USPION surface reduced the magnetization by 11.5% and with 2nm thick coating it was more than 50%.

Furthermore, they concluded that 2nm thick coating is needed, and thinner coating would not be protective enough for SPIONs. (Kralj et al. 2010)

To assess physicochemical properties of SPIONs, there are several different analytical techniques: iron concentration can be determined by inductively coupled plasma mass spectroscopy or via the 1,10-phenantroline assay. With transmission electron microscopy (TEM) and with atomic force microscopy, the core size, core size distribution and morphology can be studied. Furthermore, X-ray diffraction (XRD) can be used to determine crystalline structure and to calculate crystalline size of SPION by using Scherrer’s equa- tion. (Dadfar et al. 2019)

(33)

5 BIOACTIVE GLASS

New era of biomaterials started in the late 60’s, when Larry Hench developed a new-kind material, the first composition for bone-bonding glass-ceramic. Previously developed, the first-generation biomaterials were selected as bioinert as possible in order to prevent fi- brous capsule and scar tissue formation. The first-generation material were metals or synthetic polymers, which were susceptible to be rejected by the body. Hench was chal- lenged to find a solution to overcome rejected implants by an army doctor. He developed an inorganic material inspired by Na₂O-CaO-SiO₂ diagram, shown in Figure 5.1. The diagram presents how glass properties change if SiO₂, CaO and Na₂O percentages are modified, whereas P₂O₅ is kept constant 6%. (Hench 2006)

The developed glass was implanted into rats, and Hench et al. reported, that instead of being inert, the implants made of that new glass were bonding with bone tissue, whereas more traditional inert implants would easily slide off bone. The material was later com- mercialized and sold as Bioglass ®. (Hench 2006)

Nowadays there are several types of bioactive glasses (BAG) including phosphate-based and borate glasses. Silicate-based glasses are all variants from Bioglass ®, known also as 45S5. Most of the bioactive glass application are for clinical use: they are used in

Figure 5.1. Glass-ceramic compositional diagram for bone-bonding (Hench 2006).

(34)

craniomaxillofacial and otolaryngologic surgery, spine and trauma surgery, bone cavity treatment as well as bone infection treatment. (Hulsen et al. 2017, Baino et al. 2018) However, there are several hinderances in usage of bioactive glasses as a TE implants.

First, glasses have brittle structure, which is problematic in load-bearing applications.

Secondly, the major problem in orthopedic and trauma-related surgery is risk of infection and unwanted inflammation reaction. (Zimmerli et al. 2004) If that inflammation could be prevented, there would be no need for second operations nor suffering (Kunutsor et al.

2015).

BAG implants could be utilized as drug delivery devices. Issues related to infection could be solved by making BAG surface antibacterial either by surface modification methods or by grafting antibacterial molecules at it. These methods are described in detail below in section 5.3. BAGs could also be utilized in bone cancer treatment, as described in Chapter 4. In this work, the focus is on bioactive silicate glasses, which by definition must contain SiO₂ less than 60%, otherwise the glass will not bond with bone and is therefore bioinert (Hench 2006). Additionally, there are other ions in the structure as well (Na₂O, CaO) and high CaO/P₂O₅ ratio (Kiran et al. 2017; Hench 2006).

5.1 Structure

Glasses are inorganic materials that are described to be brittle and transparent. Glasses are widely used in packaging, photonics, construction and decoration, however, nowadays their application reach to clinical side as well. This is due to their excellent biocompatibility, which stems from their structure and composition. (Hench 2006)

Soda-lime glass, that is used in windows and other applications requiring transparent material, is with high-content of silica (SiO₂), which results in excellent chemical resistance.

Within silica rich glasses, the glass network is strong due to silicon’s ability to form four covalent bonds. SiO₂exhibits sp³hybridization giving the molecule tetrahedron form, which is the base of silica network. Each silicon will form bond with 4 oxygen atoms, which are bonded to other silicon atoms, presented in Figure 5.2 Each oxygen is shared by two adjacent tetrahedral blocks, which together build up covalent network of glass. The oxygens connecting silicon atoms are called bridging oxygens (BO), and SiO₂ as network formers.

In other glass types, the network former is different: within phosphate-based glasses that would be -PO₄ group, and with borate glass B₃O₆. (Le Bourhis 2014)

Nevertheless, strong covalently bound structure makes glass inert. Therefore, in bioactive glasses alkaline oxides and alkaline earth oxides are used to disrupt that strong network. They are called modifiers and they reside the interstitial spaces between said SiO₂ tetrahedrons. Modifiers enable the formation of non-bridging oxygen (NBO) and ionic bonding, which will replace strong and stable covalent bonds in glass structure.

Shift from BO to NBO happens when the oxygen loses connection to a neighboring tetrahedron. It will compensate the loss of electron by forming an ionic bond with a modifier.

(Fernandes et al. 2018) The modifier are as well presented in Figure 5.2. The introduction

(35)

Figure 5.2. Silica network: the role of bridging and non-bridging oxygens and network modifiers and formers. (Stanic 2017).

of modifiers weakens the silica network, which lowers chemical resistance of the glass and the glass formation temperature. (Le Bourhis 2014)

These SiO2tetrahedrons could arrange themselves into repeating units giving ultimately crystalline solid, such as quartz crystals. The long-range order will take place during cooling from molten state, however, the time needed for such arrangement takes long cooling time, which makes the approach impractical. In quenching method, the rapid cooling rate forces the viscosity of the molten to increase gradually, which will eventually hinder ion mobility. The result is a frozen state of the random network of a liquid, which is having short range order in solid state, which resembles much of a frozen structure of a liquid, namely super-cooled liquid. (Le Bourhis 2014)

This gives glass its intrinsic amorphous nature. Crystalline solids, after reaching the melting point, have an abrupt change in physical appearance, whereas glasses have a gradual transition from solid to liquid. This is due to glass network structure, and its ability to become gradually looser. Above this temperature span, glasses are viscous liquid while being an amorphous solid below it. (Le Bourhis 2014)