The genotoxic effect of gold nanoparticles in cultured human cells

The genotoxic effect of gold

nanoparticles in cultured human cells

DW Mulder

26718944

Dissertation submitted in partial fulfilment of the requirements

for the degree

Magister Scientiae

in

Biochemistry

at the

Potchefstroom Campus of the North-West University

Supervisor:

Prof PJ Pretorius

Co-supervisor:

Dr CJF Taute

Abstract

Gold nanoparticles are ultrafine particles which are being used diversely in the medical field. Despite their versatility, the toxicity research which has been conducted on these particles is incredibly limited as well as contradictory.

The aim of this study was to determine the genotoxicity of gold nanoparticles on human hepatocellular carcinoma (HepG2) cells. This was done by first synthesizing spherical gold particles; functionalizing, characterizing and then testing their stability in conditions similar to cellular environments. Once this was done the gold particles were then added to the cultured HepG2 cells and cytotoxicity was determined by the WST-1 assay and xCELLigence technology. The ApoPercentage assay was used in order to determine the cell death mechanism as a result of the gold particles and inductively coupled plasma mass spectrometry (ICP-MS) was used to determine their internalization. The genotoxicity caused as a result of the gold particles was analyzed by the use of an alkaline single cell gel electrophoresis.

The results obtained indicated that 18nm spherical gold nanoparticles were successfully synthesized and functionalized. The particles were stable in cellular environments except for thiol interactions and phosphate buffered saline. Cytotoxicity was evident however the cells did recover. Some of the gold nanoparticles did induce genotoxicity however the cells were able to repair the damage with the exception of the cells exposed to gold nanoparticles functionalized with mercaptoundecanoic acid (MUA).

i

Keywords

Single cell gel electrophoresis Stability

xCELLigence ApoPercentage WST-1

ii

Declaration

I, Danielle Wingrove Mulder, hereby declare that the work: “Genotoxicity of gold nanoparticles” is my own work, that it has not been submitted for any degree or examination in any other University and that all the sources I have used or quoted have been indicated and acknowledged by complete references.

_______________________________ Danielle Wingrove Mulder

iii

Acknowledgements

I would really like to take this opportunity to thank some people who played a crucial role in this study and without them would have made it difficult to finish this project. Firstly I would like to thank my supervisors Professor Piet Pretorius; for your constant encouragement and insight regarding the cell biology aspect of the project and who also taught me the importance of scrutinizing the data as well as the importance of careful data interpretation, for Dr Francois Taute for your constant encouragement and patience in repeatedly explaining the chemistry of the gold nanoparticles and the importance there of as well as the constant push to think outside the box pertaining to what the results could mean. I really appreciate your guidance, kindness, patience and open door policy that both of you constantly showed me.

To my husband Garryth-lee Mulder, thank you for your support and help during this project which gave me the time to write up and do what I needed to do in the lab. I really appreciate you and love you. To Letitchia Trollip and Sharon Rhodes for looking after my children in the afternoon when I was running late in the lab, you ladies really took a large portion of stress off my shoulders. To my mom, Mandy Tedder for the proofreading of my thesis and my dad Terry Tedder for pretending he understands.

I would also like to convey gratitude to Mari van Reenen for doing the box plots for the single cell gel electrophoresis for me, to Dr Anine Jordaan for the stunning TEM images and Dr Johan Hendriks for the ICP-MS results, your information added to the story of the project.

Last but not least, I would like to thank Alnari Matthyser, working with you on this project was amazing. Thank you for your encouragement, work ethic, praying with me when things got tough and the constant supply of cookies (Proverbs 16:3).

iv

List of abbreviations

BSA - Bovine serum albumin

Citrate - Trisodium citrate

DLS - Dynamic light scattering

g - grams

GNPs - Gold nanoparticles

GSH - Glutathione

HepG2 - human hepatocellular carcinoma

hr - hours

HR-TEM - high resolution transmission electron microscopy

ICP-MS - Inductively coupled plasma mass spectrometry

mL - milliliters

MUA - Mercaptoundecanoic acid

NaCl - Sodum Chloride

nM - nanoMolar

PEG - Polyethyleneglycol

pM - picoMolar

PSSNA - Poly(sodium 4-styrenesulfonate)

PVP - PolyVinylpyrrolidone

v

List of figures

Figure 2-1: A) The Lycurgus cup appearing green with direct light (reflected). B) The Lycurgus cup appearing red as a result of transmitted light. (Image taken from (Freestone et al., 2007))………..2

Figure 2-2: Damascus sabre sword

(Photograph was taken by Tina Fineberg and published in the New York Times 28 November 2006.)………..3 Figure 2-3: Photographs of gold nanoparticles and rod aqueous solutions with their corresponding transmission electron micrographs. Scale bars are 100nm. The gold spheres are images a-e and rods are f-k (Mody et al., 2010)………5 Figure 2-4: Various morphologies of gold nanoparticles. a) spheres; b) rods; c) hollow Silicon/gold nanoshells; d) gold nanobowls with gold seeds ; e) silver nanocubes and gold nanocages ; f) stars; g) bipyramids; h) octahedrals (Khlebtsov and Dykman, 2011)………...6 Figure 2-5: Applications of gold nanoparticles (Cabuzu et al., 2015)……….8 Figure 2-6: An indication of literature on nanoparticle toxicology. CA) comet assay; MN) micronucleus assay; CHA) chromosomal aberration test; Ames) bacterial reverse mutation assay (Magdolenova et al., 2014)………...11 Figure 2-7: Indicates the determinants of how nanoparticles will interact (including uptake) with cells such as shape (rod, sphere, triangle, star, cubes), size, cell type, surface chemistry, ligand and aggregation status (Kettler et al., 2014)……….12 Figure 2-8: Schematic representation of induced genotoxic and affected intracellular signaling pathways by nanoparticles. These mechanisms include; (a) genotoxic effects as a result of reactive oxygen species overload. (b) Altered protein or gene expression resulting in enlarged lysosomes which hinder function of transcription and translation machinery. (c) Free metal ions binding to molecules altering the levels of protein or gene expression. (d) Interference of factors such as cell-surface receptors thereby altering the activation status of proteins. (e) Nanoparticles induce added cellular stress which then alters the gene expression levels (Soenen et al., 2011)………17 Figure 2-9: Flow diagram based on (Ajnai et al., 2014)………..18 Figure 2-10: Schematic representation of some of the stressors of reactive oxygen species and the damage it may cause which may affect cellular processes resulting in cell death (Scandalios, 2005)………..19 Figure 2-11: Reaction scheme taken from (Valavanidis et al., 2009) showing the reaction of 2-deoxyguanosine with hydroxyl radicals. The C8-OH-adduct radical is followed by reduction to 7-hydro-8-hydroxy-2-deoxyguanosine which is in turn oxidised

vi

to 8-hydroxy-2-deoxyguanosine (8-OHdG) or its tautomer

8-oxo-7-hydro-2-deoxyguanosine (8-oxodG)………..20 Figure 2-12: The different pathways and enzymes used for short patch-base excision repair. The middle pathway with OGG1 is the pathway of interest in terms of 8-oxodG removal……….22 Figure 3-1: Diagram illustrating how the measurements were taken on the gel in order to determine particle size. The total distance is a theoretical distance chosen which is equidistant on either side of the well………..31 Figure 4-1: UV-Vis spectrum for citrate capped gold particles………..39 Figure 4-2: HR-TEM micrograph of the citrate capped particles………...42 Figure 4-3: Particle size distribution as determined from the TEM micrographs. The average size of the particles was 18nm out of the 198 particles counted……….43 Figure 4-4: Hydrodynamic diameter for citrate covered particles………..44 Figure 4-5: UV-Vis spectra of the functionalised GNPs post sample clean-up………..45 Figure 4-6: (A) Gel electrophoresis done in TBE pH8. (B) Gel electrophoresis done in TAE pH8………..…………48 Figure 4-7: Protonation of PVP at pH 6.5……….50 Figure 4-8: UV-Vis spectra for PEG coated GNPs in a range of NaCl at the 6hr time point……….53 Figure 4-9: UV-Vis spectra for PEG coated particles in 1mM NaCl over a 48hr time period………..54 Figure 4-10: UV-Vis spectra obtained for PSSNA in β-mercaptoethanol over a 48hr time period………..56 Figure 4-11: UV-Vis spectra obtained for PSSNA in supplemented media over a 48hr time period……….58 Figure 5-1: Cytotoxicity of the functionalised GNPs at a range of dosages on HepG2 cells (See Table 3-4 for dosage concentrations)………63 Figure 5-2: Illustration of how GSH couples to copper pumped into the cells………...64 Figure 5-3: Reaction mechanism between GSH and GSSG resulting in the production of NADP……….65 Figure 5-4: Apoptosis in HepG2 cells exposed to MUA………69

vii

Figure 5-5: Micrographs taken of the apoptotic determination of HepG2 cells exposed to hydrogen peroxide over time. Micrograph A=0hrs, B=3hrs, C=6hrs, D=12hrs and E=24hrs………70 Figure 5-6: Normalised real time growth curve of the HepG2 cells after 48hrs GNP exposure (The original graph can be found in the Appendix D, Figure A5-9). This graph time point is taken from when the GNPs were added to the cells….……….72 Figure 5-7: Normalised calculated cytotoxicity using xCELLigence data of the GNPs to the HepG2 cells over 48hrs (The original graph can be seen in the Appendix D, Figure A5-10)……… 73 Figure 5-8: Comet assay micrographs taken of the HepG2 cells after 0hrs of exposure to the GNPs……….76 Figure 5-9: Comet assay micrographs taken of the HepG2 cells after 3hrs of exposure to the GNPs……….77 Figure 5-10: Comet assay micrographs taken of the HepG2 cells after 24hrs of

exposure to the GNPs……….. 78 Figure 5-11: Cellular uptake of nanoparticles as well as their interactions with some organelles. (Magdolenova et al., 2014)……… 82

viii

List of tables

Table 2-1: Different gold nanoparticle shapes and their uses………..7

Table 2-2: Gold nanoparticle based drugs which are either approved or under investigation by the FDA for cancer therapeutics………9

Table 2-3: List of genes which are down-regulated after exposure to gold nanoparticles…………...………...14

Table 2-4: Toxic effect of a variety of nanoparticles on various cell types………...15

Table 3-1: Chemical characteristics of ligands used to functionalise GNPs………27

Table 3-2: Compound category and parameters used for testing the functionalised GNP stability………30

Table 3-3: Cell culturing parameters used………33

Table 3-4: Dosing limits used in the WST-1 assay………..34

Table 4-1: UV-Vis results obtained for the citrate capped particles………..40

Table 4-2: Concentration of functionalised GNPs………46

Table 4-3: Rf values calculated from figure 4-6………...………….49

Table 4-4: Physical properties of GNPs obtained from three of the analysis methods………...51

Table 4-5: Stability of the functionalised GNPs in a varied salt concentration environment………55

Table 4-6: Stability of the functionalised GNPs in a thiol rich environment……….57

Table 4-7: Stability of the functionalised GNPs in media over time………...59

Table 4-8: Stability of the particles in environments similar to those which will be carried out during the cell biology section of the experiment………60

Table 5-1: The concentration at which the IC30 was determined by the WST-1 assay………65

Table 5-2: Analysis of particle distribution……….67

ix

Table 5-4: Comparison of ICP-MS internalisation percentage and xCELLigence

data………..74 Table 5-5: Overall DNA damage induced by the GNPs………..79 Table 5-6: Comparison of GNP physical parameters and their effect on

x Keywords……….. ….i Declaration………ii Acknowledgements……… iii List of abbreviations………....iv List of figures………v List of tables………..vii

CHAPTER 1:

INTRODUCTION………..1

CHAPTER 2:

LITERATURE REVIEW……….. …2

2.1 A brief history of nanoparticles in the ancient world………...2

2.2 Gold nanoparticles………..3

2.3 Physiochemical techniques for gold nanoparticle analysis…………...10

2.4 Toxicity studies………...11

2.5 Reactive oxygen species……….18

2.5.1 Reactive oxygen species, DNA damage and DNA repair…………...20

2.5.1.1 DNA base excision repair………...21

2.6 Problem statement……….22

2.7 Aims and objectives ……….23

CHAPTER 3:

MATERIALS AND METHODS………..24

3.1 Gold nanoparticle synthesis and ligand exchange………..24

3.1.1 Gold nanoparticle synthesis using the Turkevich method……… 24

3.1.2 Particle determination and concentration using UV-vis...24

3.1.3 Particle shape was determined visually by HR-TEM………..25

3.1.4 Dynamic light scattering………..26

3.2 Surface functionalisation of GNPs………26

3.2.1 Ligand exchange and pH optimisation……….28

3.2.2 Flocculation assay………...28

3.2.3 Aqueous stability evaluation of the surface functionalized GNPs: The role of pH, ionic strength and complex solution matrixes………..29

3.2.4 Gel electrophoresis………..31

3.3 General cell culturing of the human hepatocellular carcinoma (HepG2) cell line……….32

xi

3.3.1 Trypsinization, cell counting and cell seeding………32

3.3.2 Cryopreservation………..33

3.3.3 Cytotoxicity………....33

3.3.4 Inductively coupled plasma mass spectrometry (ICP-MS)………34

3.3.5 ApoPercentage apoptosis assay………...35

3.3.6 xCELLigence cytotoxicity………35

3.3.6.1 Resistor plate read………..35

3.3.6.2 Programme set-up………..35

3.3.6.3 Plate set-up………..36

3.3.6.4 Cytotoxicity………...36

3.4 DNA damage determination using the single-cell gel electrophoresis (Alkaline Comet Assay)………36

3.4.1 Sample preparation………..36

3.4.2 Slide preparation………...37

3.4.3 Comet assay procedure………...37

CHAPTER 4:

NANOCHEMISTRY RESULTS AND DISCUSSION..38

4.1 Nanochemistry……….38

4.1.1 Synthesis and physiochemical characterisation of GNPs prior to ligand exchange………...38

4.1.2 Ligand exchange of GNPs: Biofunctionalisation, characterisation and stability evaluation………...45

4.2 Functionalised GNPs stability in different buffer, salt, ionic strength and pH conditions………52

CHAPTER 5:

MAMMALIAN CELL BIOLOGY RESULTS

AND DISCUSSION………62

5.1 Mammalian cell biology………...62

5.1.1 Cytotoxicity assay………..62

5.1.2 Particle distribution (ICP-MS)………..…66

5.1.3 ApoPercentage………..68

5.1.4 xCELLigence………..72

5.1.5 Alkaline single cell gel electrophoresis……….75

CHAPTER 6:

RESULT SUMMARY AND CONCLUSION………….85

6.1 Results summary………...85

6.2 Conclusion………..85

xii

Appendix A: References………...87

Appendix B: Materials, suppliers and stock solutions………94

Appendix C: Nanochemistry supplementary data………..100

1

CHAPTER 1: INTRODUCTION

Gold nanoparticles are ultrafine particles which are being used diversely in the medical field. Despite their versatility, the toxicity research which has been conducted on these particles is incredibly limited as well as contradictory.

The aim of this study was to determine the genotoxicity of gold nanoparticles on human hepatocellular carcinoma (HepG2) cells. This was done by first synthesising spherical gold particles; functionalising, characterising and then testing their stability in conditions similar to cellular environments. Once this was done the gold particles were then added to the cultured HepG2 cells and cytotoxicity was determined by the WST-1 assay and xCELLigence technology. The ApoPercentage assay was used in order to determine the cell death mechanism as a result of the gold particles and their internalisation was determined by inductively coupled plasma mass spectrometry (ICP-MS). The genotoxicity caused as a result of the gold particles was analysed by the use of an alkaline single cell gel electrophoresis.

The results obtained indicated that 18nm spherical gold nanoparticles were successfully synthesised and functionalised. The particles were stable in cellular environments except for thiol interactions and phosphate buffered saline. Cytotoxicity was evident however the cells did recover. Some of the gold nanoparticles did induce genotoxicity however the cells were able to repair the damage with the exception of the cells exposed to gold nanoparticles functionalised with mercaptoundecanoic acid (MUA).

2

CHAPTER 2: LITERATURE REVIEW

Nanotechnology is the application of nanoparticles with sizes on the nanoscale usually

cited between 1nm – 100nm (Ju-Nam and Lead, 2008, Maynard et al., 2006). These

particles can be classified into a variety of different categories such as; organic (i.e. fullerenes and carbon nanotubes), inorganic (i.e. Zinc oxide, iron oxide and titanium oxide), metals (gold, silver and iron) and quantum dots (i.e cadmium sulfide) (Ju-Nam and Lead, 2008). Due to their chemical and physiochemical properties gold nanoparticles have drawn particular interest.

Without nanotechnology being defined, it was being used as far back as the 4th century.

The famous Lycurgus cup was made by the Romans in the 4th century which comprised

of colloidal gold and silver. The fascinating allure of the cup is the light scattering properties of the particles which make the cup appear to change colours. If light shines directly (reflected light) on the cup then the cup appears green in colour however if the light is transmitted light (e.g. light shines from behind the cup towards the viewer’s eyes) then the glass appears ruby red in colour (Freestone et al., 2007). Figure 2-1 is photographs taken of the Lycurgus cup.

Figure 2-1: A) The Lycurgus cup appearing green with direct light (reflected). B) The Lycurgus cup

appearing red as a result of transmitted light. (Image taken from (Freestone et al., 2007))

3

Between the 13th and 18th century the Damascus swords were crafted by the use of an

unknown method lost in the 18th century. These blades were extremely strong and

sharp but remained flexible. In 2006 a high-resolution transmission electron micrograph was taken of a sword sample revealing that the swords comprised of nanocarbon rods mixed with cementite nanowires (Reibold et al., 2006). Figure 2-2 is a photograph of a typical Damascus sabre sword.

Figure 2-2: Damascus sabre sword

(Photograph was taken by Tina Fineberg and published in the New York Times 28 November 2006.)

During modern times, different nanoparticles have been used in a variety of commercialised products such as paints, targeted drug delivery, robotics and molecular devices (Roco, 2011).

2.2 Gold nanoparticles

Gold is a noble metal with an atomic number of 79 and atomic weight of 197 at ground state (Merchant, 1998, Thakor et al., 2011). Apart from its ground state (valency 0), two common oxidation states of gold include +1 (Au[1] – known as aurous compounds) and

+3 (Au[3] – known as auric compounds) (Jain et al., 2012). Metallic gold (Au[0]) does

not oxidise or burn easily in air and has also been shown to be inert to strong acidic and alkaline reagents, as a result, gold is considered one of the least alkali metals (Thakor et al., 2011). Although ground state gold (Au[0]) is considered stable, Au[3] and Au[1] are unstable (Thakor et al., 2011). Au[3] is a strong oxidising agent and can be easily

4

reduced to Au[1] by biological molecules such as those containing thiols (Thakor et al., 2011). On the basis of gold’s chemistry, Au[1] is less reactive than Au[3], it is water soluble and can be easily stabilised in a complex by added ligands and is therefore used as a therapeutic agent (Thakor et al., 2011).

Gold nanoparticles (GNPs) can be synthesised into different sizes and geometries via a facile synthesis. The different sizes and their shape will have a unique solution colour; an example of this phenomenon is illustrated in Figure 2-3. For spherical particles the smaller particles are redder in colour and the red shade varies with the change in size with the larger particles being more purple, in comparison the rod shaped particles differ drastically in colour between the different sizes as a result of the localised surface electrons (Mody et al., 2010).

5

Figure 2-3: Photographs of gold nanoparticles and rod aqueous solutions with their corresponding

transmission electron micrographs. Scale bars are 100nm. The gold spheres are images a-e and rods are f-k. With permission from (Mody et al., 2010).

6

Not only can the gold particles be synthesised into rods and spheres of different sizes, they are also easily synthesised into different morphologies of which examples are visible in Figure 2-4. These anisotropies have led to the gold nanoparticles being used for a variety of functions of which a few have been listed in Table 2-1 (Khan et al., 2014).

Figure 2-4: Various morphologies of gold nanoparticles. a) spheres; b) rods; c) hollow Silicon/gold

nanoshells; d) gold nanobowls with gold seeds ; e) silver nanocubes and gold nanocages ; f) stars; g) bipyramids; h) octahedrals (Khlebtsov and Dykman, 2011).

7

Table 2-1: Different gold nanoparticle shapes and their uses

(Khan et al., 2014).

As a result of their shape and size nanoparticles behave very differently from their bulk state i.e. bulk gold will reflect light at 520nm, whereas 5nm GNPs will strongly absorb the light (Klien and Godnic-Cvar, 2012). The difference in their reactivity is that the reactivity of bulk gold is governed by classical mechanics, whereas GNPs are governed by quantum mechanics (El Naschie, 2006). As a result of their unique chemical and physiochemical properties such as; surface plasmon resonance in the visible range, colorimetric, high electric conductivity, non-linear optical properties, high surface area, biocompatibility and their ability to bind to a variety of ligands, GNPs have gathered particular interest in a variety of fields (Daniel and Astruc, 2004, Jain et al., 2012, Rosarin and Mirunalini, 2011, Patra et al., 2010) and as a result have been used in a variety of applications such as those in Figure 2-5.

8

Figure 2-5: Applications of gold nanoparticles. Modified from (Cabuzu et al., 2015).

Gold nanoparticles are revolutionizing the discipline of nanomedicine by introducing new ways in which diagnosis, treatment and preventative measures are carried out (Panyala et al., 2009). These particles also have the potential to overcome some of the drug delivery challenges by improving the drug bioavailability and aqueous solubility, preventing drug degradation which results in a prolonged drug release as well as a reduction in toxic side effects and provides a platform for targeted delivery which allows for easy administration (Parveen et al., 2012). Table 2-2 shows current cases with the Food and Drug Administration (FDA) which are either undergoing investigating or approved cancer therapeutics which are gold nanoparticle based (Ajnai et al., 2014).

9

Table 2-2: Gold nanoparticle based drugs which are either approved or under investigation by the FDA for cancer therapeutics

(Ajnai et al., 2014)

As mentioned previously, the synthesis methods for the particles are simplistic. To prepare GNPs with a core diameter size of 10-150nm the Turkevich method is the

common method used which is the reduction of chloroauric acid (HAuCl4) with sodium

citrate in water whereby the citrate caps the particles stabilising them (Kimling et al., 2006). The size of GNPs synthesised can also be controlled by the amount of citrate added (Kumar et al., 2007). For a core diameter size of 1.5-5nm, two methods are

applied. Either the biphasic reduction of HAuCl4 using sodium borohydride in the

presence of dodecanethiol (Brust et al., 1994) whereby the alkanethiol acts as the

stabilising agent (Brust-Schiffren method) (Ghosh et al., 2008) or

chloro(triphenylphosphine)gold(1) (AuCl(PPh3)) is reduced with either diborane or

sodium borohydride. In this case, phosphine acts as the capping agent (Ghosh et al., 2008).

10

2.3 Physiochemical techniques for gold nanoparticle analysis

The metallic compositions of GNPs, as well as the organic or biological surface functionalization needed for biomedical applications, require various physicochemical techniques to fully characterize. Typically, the following techniques are frequently used (Oksel et al., 2015).

i) Ultraviolet-visible spectrophotometry (UV-Vis) is a technique in which the

electric field of ultraviolet light (380nm-800nm) interacts with the molecule’s electrons. The intensity and position of the spectrum obtained is a result of the energies of the external molecular electrons and their dynamic characteristics compared to the core of the molecule, this information is used for both qualitative and quantitative analysis (Ohannesian and Streeter, 2005).

ii) Transmission electron microscopy (TEM) works on a similar principle to the

light microscope, however uses an electron beam instead of light. Due to the electron beam being much smaller than the light beam, the resolution for the TEM images are of a higher quality and magnitude which allow the visualisation of internal structures such as individual atoms (Williams and Carter, 2009).

iii) Inductively coupled plasma mass spectrometry (ICP-MS) is an element

specific detection technique. This technique uses a noble gas, such as Argon, under high temperature to fragment every sample into its individual atomic constituents which can then be quantified (Profrock and Prange, 2012).

iv) Dynamic light scattering (DLS) is a sensitive method used to measure the

hydrodynamic size, size distribution and in some cases shape of the nanoparticles by measuring their Brownian motion and then equating it to a hydrodynamic diameter (Brar and Verma, 2011, Pecora, 2000).

11

2.4 Toxicity studies of gold nanoparticles

There is still a large need for toxicity studies of GNPs as the number of publications regarding this aspect is minimal (Panyala et al., 2009, Alkilany and Murphy, 2010, Ajnai et al., 2014). A Pubmed search on toxicity and nanoparticles as a whole was conducted and the results which were obtained are given in Figure 2-6 (Magdolenova et al., 2014). It is observed from the Figure 2-6 that the number of genotoxicology studies that have been published on nanoparticles is largely limited.

Figure 2-6: An indication of literature on nanoparticle toxicology available in the year 2013. CA) comet assay; MN) micronucleus assay; CHA) chromosomal aberration test; Ames) bacterial reverse mutation assay (Magdolenova et al., 2014).

Challenges of advanced drug delivery systems for nanomedicine which are relevant to nanoparticles were compiled by (Crommelin and Florence, 2013). These matters include inconsistent and confusing nomenclature of the nanoparticles, as well as them being prematurely claimed as potential delivery systems (Etheridge et al., 2013). When investigating the toxicity of gold nanoparticles a variety of factors such as; particle size,

12

surface chemistry, ligand, shape, coordination geometry of the complex should be considered, as these affect the biodistribution, pharmacokinetics and blood stability of the particles (Gerber et al., 2013, Magdolenova et al., 2014, Milacic et al., 2008, Ajnai et al., 2014). These factors also have an effect on the cellular uptake of the particles which will determine whether the desired ligand (drug) will make it to the designated target (Jenkins et al., 2015, Shang et al., 2014, Kettler et al., 2014). Figure 2-7 shows the various factors which need to be considered when assessing particle toxicity. Another aspect for investigation in which there is lack of understanding, is how these particles interact and affect living cells (Almeida et al., 2014)

Figure 2-7: Indicates the determinants of how nanoparticles will interact (including uptake) with cells such

as shape (rod, sphere, triangle, star, cubes), size, cell type, surface chemistry, ligand and aggregation status. With permission from (Kettler et al., 2014).

13

A study by (Ajnai et al., 2014) showed that 10nm particles could be found circulating in the animal model’s blood for up to 24 hours, however larger sizes were excreted at a faster rate. Apart from the stability of the GNPs in blood, they were also strongly connected to the aggregation and cytotoxicity in the blood cells. Contradictory findings in various GNP toxicity studies showed that the size of the particles had an effect on the toxicity of the particles. Hwang and colleagues investigated the hepatotoxic effect of polyethyleneglycol (PEG) functionalized 15nm GNPs on malnourished rodents. Pathology done post-administration of the GNPs found hepatic inflammation, severe cell damage and significantly higher reactive oxygen species levels (Hwang et al., 2012). A study done by Kim and co-workers showed that gold nanoparticles of 1.3nm which were coated with N,N,N-trimethylammoniumethanethiole, induced systemic cytotoxicity to zebrafish embryos resulting in neuronal damage as well as malpigmented small eyes (Kim et al., 2013). Trisphenylphosphine coated GNPs (diameter of 1.4nm) induced cell necrosis on a range of cells such as fibroblasts, macrophages, melanoma and epithelial cells compared to that of bigger particles (Pan et al., 2007). Contradictory results found by Gosens and colleagues: GNPs between the sizes of 50nm and 200nm showed no cytotoxicity to animals (Gosens et al., 2010). Supporting these findings, in 2012 Wistar rats were exposed to 18-mg gold particles of 2nm, 20nm and 200nm (which the rodents inhaled), the results obtained indicated that the GNPs had no systemic, local or genotoxic effect in the lungs (Schulz et al., 2012). In contradiction to the negative genotoxic effect of the Wistar rat study Singh and colleagues showed GNPs having a genotoxic effect by the down-regulation of DNA repair genes such as Hus1 ,

ATV1/ATV2, BRCA1 and ATLD/HNGS1 (Singh et al., 2009). The study done by

Schaeublin and co-workers complement this study as they also showed a change in the gene expression of a number of genes after exposure to gold nanoparticles (Schaeublin et al., 2011). The results obtained from this study can be found in Table 2-3. In Table 2-3 the gene symbol is the name of the gene. The functional group is the area the gene is involved with regards to cell functions and the function is the specific location the matching protein is involved with e.g. APEX1 is the gene involved with DNA repair, its’ translated protein APEX1 is responsible for initiating base excision repair.

14

Table 2-3: List of genes which are down-regulated after exposure to gold nanoparticles.

Portion of the table taken from (Schaeublin et al., 2011).

A study done by Paino and colleagues showed that toxicity was cell type dependent, the immune cell lines were less affected by the particles compared with a cancer cell line, which had both cytotoxic and genotoxic side effects (Paino et al., 2012).

An exhaustive search of the literature pertaining to the epigenetic effects of the particles, was done and it was noted that this area of research is very limited. The first paper which investigated the epigenetic effect of gold nanoparticles, observed the effect of 20nm citrate capped gold nanoparticles on MRC5 human fetal fibroblasts. The outcome of the investigation was an up-regulation of miR-155, which had an epigenetic down-regulation effect of the PROS1 gene (Ng et al., 2011). This finding is important as miR-155 has a function in multiple pathways, which include LPS and MAPK signaling pathways, as well as pathways which contribute to conditions such as cardiovascular, inflammation and carcinogenesis (Ng et al., 2011).

Table 2-4 is a summary of some of the nanoparticles and the toxic effect they had on a variety of cell types. From the table, it can be concluded that cytotoxic effects have been investigated, however the genotoxic investigations remain limited.

15

16 Table taken from (Fraga et al., 2013)

In light of the studies mentioned, it is clear that there has been no consensus reached pertaining to the toxicity of the particles. Duncan and co-workers mention that the particles have low toxicity and in another article, mention that depending on the charge of the particles, they can interact electrostatically with proteins of the opposite charge and that more research needs to be conducted regarding their toxicity (Duncan et al., 2010, Xu et al., 2013)

In terms of cellular interactions, inorganic nanoparticles can disrupt cellular homeostasis by several mechanisms as illustrated in Figure 2-8.

17

Figure 2-8: Schematic representation of induced genotoxic and affected intracellular signaling pathways

by nanoparticles. These mechanisms include; (a) genotoxic effects as a result of reactive oxygen specie overload. (b) Altered protein or gene expression resulting in enlarged lysosomes which hinder function of transcription and translation machinery. (c) Free metal ions binding to molecules altering the levels of protein or gene expression. (d) Interference of factors such as cell-surface receptors thereby altering the activation status of proteins. (e) Nanoparticles induce added cellular stress which then alters the gene expression levels. (f) Nanoparticles can interact directly with the DNA. Modified from (Soenen et al., 2011).

18

The general cytotoxic mechanism that gold nanoparticles follow is as follows;

The internalization of the particles

Release of toxic ions

Ion toxicity together with the inactivation of associated enzymes

Mitochondria membrane depolarization

Disruption of redox balance

Lysosomal dysfunction in cell

The disruption of the cellular homeostasis results in increased reactive oxygen species levels

Apoptosis

Figure 2-9: Flow diagram based on (Ajnai et al., 2014).

2.5 Reactive oxygen species

Based on Figure 2-9, the disruption of the redox balance is an interesting aspect of the cytoxicity pathway to examine, as it is the point in the pathway that reactive oxygen species are present, due to the gold nanoparticles inducing oxidative stress on the cells. Reactive oxygen species (ROS) are a by-product of cellular aerobic metabolism. The

main reactive oxygen species are hydrogen peroxide (H2O2), the hydroxyl radical (OH•)

and the superoxide anion (O2•-) (Thorpe et al., 2004). These molecules are involved in

19

and protein-DNA interactions for signal transduction pathways as well as defense against microbes (Scandalios, 2005). ROS are found in equilibrium with antioxidants in the cell but when these levels are elevated above the normal threshold or the antioxidants are depleted or both scenarios occur, then the cell enters into a state of oxidative stress (Scandalios, 2005). When oxidative stress occurs, the cell will respond by either activating or silencing the genes responsible for transcription factors, structural proteins or defensive enzymes in order to re-establish the ROS-antioxidant equilibrium (Scandalios, 2005). Oxidative stress may occur as a result of a variety of factors, which in turn may result in damage to proteins, lipids, carbohydrates and DNA. This affects cellular processes which may ultimately lead to cell death (Scandalios, 2005, Thorpe et al., 2004). This process is illustrated in the schematic representation in Figure 2-10.

Figure 2-10: Schematic representation of some of the stressors of reactive oxygen species and the

damage it may cause which may affect cellular processes resulting in cell death. Obtained with permission from (Scandalios, 2005).

20

2.5.1 Reactive oxygen species, DNA damage and DNA repair

Due to the increased ROS levels caused by gold nanoparticles, the DNA is damaged and therefore needs to be repaired. The hydroxyl radical is the ROS which triggers DNA damage, despite its short life span in the cell (Henkler et al., 2010). The hydroxyl radical interacts with the guanine nucleotide/nucleoside forming 3 different radical adducts. These radical adducts are then oxidised forming 8-hydroxy-2-deoxyguanosine (8-OHdG) or its tautomer 8-oxo-7-hydro-2-deoxyguanosine (8-oxodG) (Valavanidis et al., 2009). The reaction scheme is shown in Figure 2-11.

Figure 2-11: Reaction scheme taken from (Valavanidis et al., 2009), with permission, showing the

reaction of 2-deoxyguanosine with hydroxyl radicals. The C8-OH-adduct radical is followed by reduction to 7-hydro-8-hydroxy-2-deoxyguanosine which is in turn oxidised to 8-hydroxy-2-deoxyguanosine (8-OHdG) or its tautomer 8-oxo-7-hydro-2-deoxyguanosine (8-oxodG).

21

2.5.1.1 DNA base excision repair

DNA base excision repair (BER) is the process which cells use in order to repair endogenous DNA damages (Wallace et al., 2012). Approximately 30 000 damages occur per cell per day including a plethora of oxidative damages, deaminations, depurinations, aminations and alkylations (Wallace et al., 2012). There are two routes in which the BER process can be accomplished and these are known as short-patch BER (or single nucleotide base excision repair) and long-patch BER (Wallace et al., 2012). Since the guanine nucleotide is oxidised during oxidative stress and only a single nucleotide is being repaired, the guanine base is repaired via the short-patch BER system. 8-OHdG/8-oxodG is recognized by 8-oxoguanine DNA glycosylase (OGG1). When it recognises the oxidised guanine, OGG1 flips it out of the helix into its active groove detaching the base from its N-glycosylic-bond (Kazak et al., 2012). AP endonuclease 1 (APE1) binds to the 3’PUA (3’-phospho-α,β-unsaturated aldehyde) and creates the hydroxyl group required by the DNA polymerase for DNA synthesis to begin (Kazak et al., 2012). Polymerase β then removes the sugar and fills the gap with the correct guanine nucleotide (Kazak et al., 2012). Finally the ligation of the DNA strands are carried out by ligase 3 (Lig III)/XRCC1 complex which link the 5’ and 3’ ends of the two DNA strands together forming the original continuous strand (Kazak et al., 2012). This process is shown in figure 2-12

22

Figure 2-12: Pathway and enzymes used for short patch-base excision repair for removal of 8-oxodG.

Modified from (Wallace et al., 2012).

2.6 Problem statement

Gold nanoparticles are ultrafine particles which are being used diversely in the medical field (Gerber et al., 2013) such as platforms for drug and genetic material delivery, as well as diagnostic agents. This has proven valuable in disease diagnosis, treatment and prevention (Panyala et al., 2009). Despite all the beneficial uses for GNPs, toxicity

23

studies of how these particles may impact human health, are largely limited (Gulumian et al., 2012, Magdolenova et al., 2014). This has led to the global drive in developing standardised testing procedures for risk assessment, which is being led by the Organization for Economic Cooperation and Development (OECD) (Gulumian et al., 2012). As a result of GNPs being considered biologically inert, they do not get digested in the body, which results in them accumulating in the liver (the smaller particles exit the body via the urine) (Papasani et al., 2012). It has been said that GNPs also contribute to oxidative stress and cytotoxicity; however, there are conflicting results with regards to the studies which have been done (Tedesco et al., 2010). GNPs have also been considered to induce toxicity by binding to the DNA grooves (Soenen et al., 2011). This information led to the inspiration of looking at the genotoxic effect that GNPs have in cultured human cells.

2.7 Aims and objectives

The aim of the study is to determine the relation between phenotypic and genotypic toxicity of differentially surface functionalized 18nm gold nanoparticles in HepG2 cells. The objectives are:

 Synthesize and surface functionalize 18nm gold nanoparticles

 Physiochemical characterization of the above mentioned GNPs

 Evaluation of the aqueous stability of the above mentioned GNPs in various

chemical (e.g.: pH, ionic strength etc.) and biological conditions.

 Assess the cellular toxicity of the GNPs on HepG2 cells using the WST-1 assay,

Apopercentage assay as well as the xCELLigence technology.

 Assess the genotoxic effects of the GNPs using the alkaline single cell gel

24

CHAPTER 3: MATERIALS AND METHODS

3.1 Gold nanoparticle (GNP) synthesis and ligand exchangeA list of chemicals, suppliers and stock solutions for this chapter can be found in Appendix.

3.1.1 Gold nanoparticle synthesis using the Turkevich method

(Turkevich et al., 1951)

All glassware was washed with phosphate free soap and then stripped with aqua regia (Appendix B). Prior to use, the glassware was rinsed five times with deionised water

(milliQ). A rapidly stirring 100mL aqueous solution containing 0.25mM HAuCl4 was

heated to 90º C in a 500mL Erlen-Meyer flask on a stirrer-hotplate (Stuart). The temperature of the solution was continuously monitored with a thermometer. Once the rapidly stirring solution reached the desired temperature, 2300µl of a 1% aqueous trisodium citrate was added all at once. An immediate change in the colour of the solution took place (light yellow → colourless → black → purple → rose red). The reaction was left to proceed at 90ºC for 15min, after the rose red colour developed, where the colour was indicative of near-spherical GNPs in solution. The solution was then placed in an ice-bath to cool the solution down. The GNP solution was filter sterilized (0.2µm, GVS) and kept at 4°C until needed.

3.1.2 Particle size determination and concentration using ultra violet- visible spectroscopy (UV-Vis)

UV-Vis spectrum for the GNPs was obtained, using an HT Synergy (BioTEK) micro

plate reader with a UV-Vis spectrometer (200 – 1000nm) and Gen5.1, as the

corresponding software. A 50µL aliquot of the GNPs was transferred to a 96-well

micro-well plate. The spectral range chosen was 350nm – 800nm, in accordance with the

literature (Haiss et al., 2007), where a blank reading was obtained using ddH2O (18.2

MΩ). The maximum absorbance peak, also known as the surface plasmon resonance

peak (λSPR), as well as the absorbance at 450nm, was used to estimate the size and

25

and table S-3 found in the supplementary information of (Haiss et al., 2007). The calculations were as follows:

The diameter of the particle was then determined from table S-2. Once the diameter of

the particle was known, the molar decadic extinction coefficient (ε) at wavelength

450nm was determined from table S-3.

The concentrations of the particles were calculated by the OD obtained at 450nm, as divided by the molar decadic extinction coefficient. The concentration unit is calculated as mol/L.

Where C is concentration

Where A450 is the absorbance at 450nm

Where ε450 (Epsilon) is the molar decadic extinction coefficient for 450nm

An example calculation can be found in results 4.1.1

3.1.3 Particle shape was determined visually by high resolution transmission electron microscopy (HR-TEM).

The morphology, as well as GNP size distribution, was determined using HR-TEM. The citrate capped GNP (direct from Turkevich synthesis, 3.1.1) sample was prepared for HR-TEM, by centrifugation at 800xg for 2 hours, to sediment and subsequently concentrate the GNPs. HR-TEM was done with a Tecnai F20 high resolution field emission transmission electron microscope by Dr Anine Jordaan (Electron Microscopy Specialist, Electron Microscopy Unit, Potchefstroom Campus, NWU). ImageJ V1.46r. was used to analyse the micrographs for automated particle counting and size dispersion.

26

3.1.4 Dynamic light scattering (DLS)

The hydrodynamic diameter of all the prepared GNPs (3.2) was done to determine the changes in the hydrated diameter of the differentially surface functionalized GNPs i.e. particle size. DLS was done using the Zetasizer Nano (Malvern) module using

(Zietasizer version 6.20 software) and polystyrene cuvettes. To prevent dust

interference, polystyrene cuvettes were prepared by being rinsed three times with

deionized water and inverted overnight on a paper towel and stoppered once dry. A

dilution series of the GNPs (Turkevich synthesis, 3.1.1) was prepared to optimize the assay.

3.2 Surface functionalisation of GNPs

Ligands were chosen based on their toxicity and biological applications. The ligands represented altered surface charges (e.g.: cationic, anionic, zwitterionic), molecular weights as well as biocompatibility. The citrate coated particles were the gold controls used in all the experimental work.

27

Table 3-1: Chemical characteristics of ligands used to functionalise GNPs.

Ligand Structure

Net charge at physiological

pH

pKa GNP pH

stability Application Ref*

BSA Negative (Bueno and Petri, 2014) 5.4 pH 7 Transport vehicles for proteins, hormones and diagnostic agents in biomedical field Sigma (Nadi et al., 2014) Citrate negative 3.138 4.76 6.40 - Anti-coagulation (Tolwa ni et al., 2001) GSH Negative (Okumura et al., 2012) 2.12 3.53 8.66 9.12 pH 8 Antioxidant (Towns end et al., 2003) MUA Negative (Simonian et al., 2002) 6.5 pH 6 Drug delivery (Nation al et al., 2015)

PEG Neutral 16 pH 7 Tablet

coating Sigma (Gans and Chavki n, 1954) PSSNA Negative (Shovsky et al., 2012) 1-2 pH 6 Drug delivery (Venka tesan et al., 2013) PVP Positive 5 pH 5 Cosmetics, food and drug delivery (Nair, 1998)

28

3.2.1 Ligand exchange and pH optimisation

The GNP solution (which was synthesised in 3.1.1) was divided into equal volume aliquots for ligand exchange in 20mL Pyrex tubes. Each gold nanoparticle aliquot was

pH adjusted (NaOH, NaCO3 or HCl) to the corresponding ligands’ isoelectric point. The

ligands were added to their corresponding pH adjusted GNP solutions all at once and vortexed for 30s. After the 30s was over, all GNP-ligand solution agitation was stopped and the solution mixture was left to stand at room temperature overnight. The isoelectric points, pH modifier as well as concentrations for the ligands can be seen in Tables 3-1 and A3-1. The control experiment was the addition of milliQ water (18.2MΩ) to an aliquot of GNPs and treated exactly like the other samples.

For all samples, centrifugation was used to remove excess ligands. The samples were centrifuged at 1000g (Lasec Z 206 A, fixed angle rotor) for 90min. The supernatant was removed and the pellet was re-suspended in half the original sample volume to concentrate the samples. Sample preparation conditions for ligand exchange which demonstrated the highest colloidal stability (Table A3-1) for the GNPs post-ligand exchange (surface functionalized) were used for all further experiments.

3.2.2 Flocculation assay (Salt stress test)

The success of surface functionalisation after ligand exchange (3.2.1) on the GNPs was evaluated by a flocculation assay. This assay is based on the principles described in (Toma et al., 2010) and evaluates two variables: the strength of the GNP-ligand interaction (e.g.: electrostatic VS covalent bonds) as well as the salt tolerance of the surface functionalized GNP. The assay was adapted for this study as follows: 200µl of the ligand/gold sample was transferred to an Eppendorf tube, followed by the addition of 10µl of a 1.5M NaCl solution. The samples were vortexed for 10 seconds, transferred to a 96-well microplate to obtain the UV-Vis spectra of the sample. A blue-shift in the

UV-Vis spectrum of a GNP surface plasmon peak (λSPR) was indicative of etching (Lee

29

aggregation (Yang et al., 2007). Readings were obtained immediately upon addition of the saline to the GNPs, with subsequent readings at 1h, 2h, 6h, 12h and 24h.

3.2.3 Aqueous stability evaluation of the surface functionalized GNPs: The role of pH, ionic strength and complex solution matrixes

Assessment of the aqueous stability of the GNPs (citrate capped as well as other surface functionalization, (3.1.1 and 3.2.1) was necessary to determine the most feasible range of conditions for which the nanoparticles could be applied. The GNPs

were all diluted to the same optical density at 450nm (OD450) in 96-well micro plates for

UV-Vis spectrometry. The aqueous parameters investigated were various typically used biological buffers at different ionic strengths, pH ranges as well as mammalian tissue culture medium (with and without foetal bovine serum).The experimental layout is given in Table A3-2. Readings were obtained immediately upon addition of the saline to the GNPs, with subsequent readings at 1h, 2h, 6h, 12h, 24h.

30

Table 3-2: Compound category and parameters used for testing the functionalised GNP stability

Compound

type Compound Parameters

GNPs 1.6nM Salt NaCl 1mM 10mM 50mM 100mM Thiols Β-Mercaptoethanol 20% Medium Supplemented Medium 70% Non-supplemented medium 70% Buffers 10mM Mops pH 9 pH 10 100mM Hepes pH 7 pH 8 EDTA 0.5X 1X 2.5X 5X PBS 0.25X 0.5X 1X 5X pH 50% MilliQ water (acidic) pH 3 pH 4 pH 5 pH 6 50% MilliQ water (basic) pH 7 pH 8 pH 9 pH 10 Organic molecules 20mMGlycine pH 2 pH 3 pH 9 pH 10 5.04mMcitrate pH 4 pH5 pH 6

31

3.2.4 Gel electrophoresis

Size and charge, of the GNPs, were determined by running a TBE gel electrophoresis. The 0.25% agarose was placed in the microwave until the agarose was completely melted. The gel was then cast into the casting tray and was allowed to set. The GNPs were loaded in a proportion of 1/5 80% glycerol and 4/5 GNPs. The samples were run at 50volts, 033mA for 30minutes (Baygene, BG-power 300, Vacutec).

The size of the particles was determined by using the theoretical calculation:

Where Ad is the distance the band of sample A travelled from the well Where Td is the total distance

Figure 3-1: Diagram illustrating how the measurements were taken on the gel in order to determine particle size. The total distance is a theoretical distance chosen which is equidistant on either side of the well.

Well Band Sample I.D Distance A travelled A B C Total distance

32

Cell biology

3.3 General cell culturing of the human hepatocellular carcinoma ([HepG2] ATCC HB-8065) cell line

The flow cabinet (Clearflow, 6FT laminar flow cabinet, Vivid Air) was switched on at least 30 minutes prior to use and sterilized with 70% ethanol. Sterile techniques were used at all times when the cells were handled. Cells were retrieved from cryostorage

(-80°C freezer), slowly thawed on ice and were then decanted into a T25 (25cm2) vented

cap flask containing 10ml RMPI1640 medium supplemented with 10% FBS and 1% Pen-strep. The cells were cultured in a humidified incubator (HeraCELL, Heraeus) at

37˚C and 5% CO2. Cells which failed to attach to the flask after a 16-24h period were

discarded. The cells were then rinsed with 1xPBS to remove the DMSO and 10mL fresh medium was added.

3.3.1 Trypsinization, cell counting and cell seeding

The cells (3.3) were trypsinized upon reaching 80-90% confluency. The cells were rinsed twice with 1xPBS and 1mL 1xtrypsin was added, followed by incubation at 37ºC for 5min. The cells were dislodged from the T25 flask by gently tapping the flask with the palm of a hand and 2mL medium was added. The contents of the T25 flask was transferred to a 15mL Falcon centrifuge tube to sediment the cells and remove the trypsin-medium at 1000xg for 5min. The cell pellet was suspended in supplemented medium (3.3) to a final volume of 10mL.

Cell counting was done using a hand held cell counter (Merck). In an Eppendorf tube, 900µl 1xPBS and 100µL cell suspension (3.3.1) was mixed together. The cells were then counted. The reading obtained by the machine is a 10x dilution and therefore this value must be corrected to obtain the cell number in the cell suspension.

33

Table 3-3: Cell culturing parameters used Culturing vessel PBS wash Trypsin concentration Supplemented medium volume Seeding density (cells/cm2) T75 5mL 3mL 1x 15mL - T25 1mL 1mL 1x 10mL - 96-well plate 100µL - 100µL 7 500 24-well plate 500µL 150µl 0.33x 500µL 37 00 3.3.2 Cryopresevation

The cells were trypsinized (3.3.1) and the equivalent volume of medium was added to neutralise the trypsin. The supernatant was removed and 1mL medium containing 1% DMSO was added to the cell pellet. The cells were then re-suspended and transferred into a cryovial. The cells were then stored immediately at -80°C.

3.3.3 Cytotoxicity

The WST-1 assay was used to determine whether the GNPs (3.1.1 and 3.2.1) were cytotoxic to the HepG2 cells in a 24h period. The cells were seeded (3.3.1) in a 96-well plate at a cell density of 7500 cells per well (as per Table 3-3) in a 100µL medium and allowed to attach overnight in culturing conditions as described in 3.4.1. The following day the spent medium was removed from the cells, rinsed twice with sterile 1x PBS and fresh 100µL supplemented medium was added to the cells. The GNPs, controls and blanks, respectively, were added to the cells to a final volume of 200µL per well followed by incubation (3.3.1) for 24h. The volumes, concentrations and controls used in this experiment are given in Table 3-4.

34

Table 3-4: Dosing limits used in the WST-1 assay.

Dosage Final concentration (nM) Dose 1 1.71 Dose 2 0.86 Dose 3 0.43 Dose 4 0.22 Dose 5 0.10

Positive control milliQ water

Negative control 5mM H2O2

Ligand only control (i.e no gold particles) 2.2

After 24h the medium mixtures (Table 3-4) were removed from the cells, washed twice with sterile PBS (x1) and 100µL fresh medium was added to each well, followed by the addition of 10µL WST-1 reagent per well. The 96-well plate was then transferred to the BioTek Synergy HT plate reader. The plate reader allowed automated shaking of the plate for 10 seconds and the plate was incubated at 37º C for 30 minutes. When the desired time-point was reached optical density readings were obtained at 460nm and a reference wavelength of 660nm. The readings obtained from the reference wavelength were subtracted from the readings obtained at 460nm to correct for the presence of phenol red in the media. The Gen 5.1 software allowed for an automated blank subtraction. The % cytotoxicity / viability were calculated as follows:

% Cytotoxicity: – Where the negative control = cells treated with medium and water only. The positive control = WST1 + medium

The citrate capped particles were the gold control used for comparison right throughout the research.

3.3.4 Inductively coupled plasma mass spectrometry (ICP-MS)

The cells used in the WST1 assay (Roche) were then taken for MS analysis. ICP-MS was used to determine how many gold particles had internalised into the cells. The

35

test was done by Dr Johan Hendriks from the Division of Microbiology. The cells were digested with neat nitric acid overnight and then diluted to 1% nitric acid concentration after which the samples were then analysed. The apparatus used was the Agilent 7500CE, RF power of 1550W, sample depth of 8 mm, carrier gas was 0.92 L/min, makeup gas was 0.25 L/min, nebulizer contained MicroMist, the spray chamber temp was 2ºC, with an integration time of 100ms.

3.3.5 ApoPercentage apoptosis assay

The ApoPercentage assay (BioColor, UK) was used to determine the induction time-point of apoptosis, using the LD70/IC30 as determined by the WST-1 assay (3.3.3) Cells were seeded into a 96-well plate (Table 3-3) and allowed to attach as described in 3.3.3. A time-dose experiment was done for the time-points 0h, 3h, 6h, 12h and 24h. The assay was conducted as described in the assay kit manual but was adapted to a 96-well micro plate format. At the zero hour time point, the GNPs were added to the medium and then immediately removed from the cells. To each well, 95µl medium mixed with 5µl dye was added and the plate was placed back into the incubator for 30min, after which the cells were rinsed with 200µl 1 x PBS. In order to visualise the cells under the microscope, 50µl 1 x PBS was added to each well and micrographs were taken of the cells using a standard camera, by viewing through the eye piece of the microscope. In order to release the dye, 50µl dye release solution (ApoPercentage kit) was then added to the wells. The plate was shaken for 10minutes in the plate reader machine (Biorad) and then read at 550nm. The same procedure was followed 3hrs, 6hrs, 12hrs and 24hrs later.

3.3.6 xCELLigence cytotoxicity

3.3.6.1 Resistor plate read

The resistor plate was set-up in the RTCA SP machine (Roche) as per manufacturer’s (RTCA software manual version 1.2) instructions

36

Under the layout tab – all the wells that were being utilised were highlighted and “Apply” was selected. The sample identification was entered under “compound name”.

Under the schedule tab – Step_1 was the medium background check. The set-up for

this was 5 sweeps and 1 minute intervals.

Step_2 was the addition of the cells. The set-up for this was 20 sweeps and 3 minutes intervals. The sub-step was 999 sweeps and 10 minutes intervals.

Step_3 was the addition of the gold nanoparticles. The same set-up was used as in Step_2. The machine was left to run for 24hrs and then the programme was terminated.

3.3.6.3 Plate set-up

To the wells being used, 100µl medium was added and step_1 of the programme was executed. The 96-well xCELLigence plate was seeded as per 3.3.1 after which Step_2 was executed. The cells were given approximately 16hrs to attach before the cells were treated with the different GNPs, at a 70% cell viability dosage, as determined via the WST-1 assay, after which step_3 was executed.

3.3.6.4 Cytotoxicity

Cytotoxicity of the GNPs on the cells was calculated with the same formula mentioned in (3.3.3) after the GNP reference was subtracted. The GNP reference was the capped particles loaded into the wells containing no cells for all seven particle types in order to account for the visible particle interference with the machine readings.

3.4 DNA damage determination using the single-cell gel electrophoresis (Alkaline Comet Assay)

This assay was slightly modified from Singh (Singh et al., 1988). 3.4.1 Sample preparation

Cells were grown and subcultured into a 24-well plate as previously (3.3 – 3.3.1)

37

3-4) at the 0hrs, 3hrs and 24hr time points in preparation for comet assay analysis. The

negative control contained milliQ water and 60µM H2O2 was added to the cells 30min

prior to trypsinization, for the positive control. After the trypsinization, the samples were neutralised with 450µl supplemented medium (3.3), transferred into autoclaved Eppendorf tubes and placed back in the humidified incubator for an hour recovery time. The preparative steps as well as the procedure will be discussed in the following paragraphs.

3.4.2 Slide preparation

The slides were coated with 300µl 1% high melting point agarose and then left to set. Once the slides had set they were then inserted into the comet assay manifolds ready for the addition of the cells. Approximately 20 cells per well were added into the wells of the manifolds. To 100µl 0.5% low melting point agarose, 50µl cell mixture was added, the gel and cells (3.4.1) were mixed gently together and then of the cell/gel mixture 20µl was added to a manifold well. The station was placed on ice allowing the gel time to set. Once the gel had set, the slides were carefully removed from the manifolds ensuring that no gels were detached from the slide.

3.4.3 Comet assay procedure

The slides were then placed into lysis buffer for 16-20 hours at 4°C after which they were then transferred into the electrophoresis buffer and incubated for 30minutes at 4°C. The electrophoresis power pack (Biorad) was switched on and left to run at 40Volts for 30min at 4°C. Once this was done, the slides were then transferred into Tris-HCl (pH 7.4) buffer for 15min at at 4°C after which they were then transferred into ethidium bromide staining solution and stained for 30min at 4°C. The excess stain was then rinsed off by incubating the slides in milliQ water for 5min at room temperature. The comets were then viewed at 20 x magnification on the Olympus lx70 light microscope fitted with a green filter. Approximately 50 comets were analysed per slide. The software used to score the comets was Comet Assay IV (Perceptive Instruments).

38

CHAPTER 4: NANOCHEMISTRY RESULTS AND DISCUSSION

4.1 Nanochemistry Gold nanoparticles (GNPs) were synthesised using the Turkevich method (3.1.1) and were also coated surface functionalised (3.2). The GNPs were characterised by ultra-violet visible spectrometry (UV-Vis) (3.1.2), dynamic light scattering (DLS) (3.1.4), high resolution-transmission electron microscopy (HR-TEM) (3.1.3) and gel electrophoresis (3.2.4). The aqueous stability of all prepared GNPs was assessed in different salts as well as buffer molecules using parameters such as ionic strength, pH and time.

4.1.1 Synthesis and physiochemical characterisation of GNPs prior to ligand exchange

The Turkevich method was chosen for the synthesis of the GNPs due to the fact that this method is a popular, cost-effective and rapid method. The surface charge of the GNPs (2.4) plays a significant role in the cytotoxicity, therefore a range of surface functionalizations were done to assess the effect of surface charge, hydrophilicity, hydrophobicity and molecular weight.

There are various physiochemical techniques available to determine synthesis and ligand exchange success (Oksel et al., 2015), where the following techniques were chosen:

(i) UV-Vis spectrometry (2.3 and 3.1.2) – Was used to determine the average

diameter, size distribution and concentration of the GNPs.

(ii) HR-TEM (2.3 and 3.1.3) - Was used to determine the physical size and

morphology of the citrate capped GNPs in a dry state.

(iii) DLS (2.3 and 3.1.4) – Was used to determine the hydrodynamic diameter of

the GNPs

(iv) Agarose gel electrophoresis (3.2.4) – Was used to estimate the size and

39

A typical UV-Vis spectrum (3.1.2) of the GNPs is illustrated in Fig 4-1 with a corresponding HR-TEM (3.1.3) micrograph in Figure 4-2.

Figure 4-1: UV-Vis spectrum for citrate capped gold particles

The expected maximum optical density at 520nm for 10-20nm spherical GNPs (Huang and El-Sayed, 2010). As can be seen by the sharp peak and narrow distribution, the particles are mostly monodispersed and are stable as they have a half width of 20nm which is close to the size of 18nm determined by the following Haiss calculations (Khlebtsov and Khlebtsov, 2011). A comparison graph of a polydispersed and monodispersed sample can be seen in (Appendix C, Figure A4-1) where the polydispersed sample has a half width of 30nm which is much larger than the calculated 18nm gold particle size.

OD λ450 = 0.166 OD λmax = 520nm 0 0.05 0.1 0.15 0.2 0.25 0.3 400 450 500 550 600 650 700 Optical De nsity (OD ) Wavelength (nm) Half width = 20nm